CULTURE AND HUMAN NATURE

CULTURE AND HUMAN NATURE

Melford E. Spiro

Edited by
Benjamin Kilborne and L.L. Langness

With a new introduction by the author

Originally published in 1987 by the University of Chicago Press.
Published 1994 by Transaction Publishers

Published 2019 by Routledge

2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
52 Vanderbilt Avenue, New York, NY 10017

Routledge is an imprint of the Taylor & Francis Group, an informa business

New material this edition © 1994 Taylor & Francis

All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by
any electronic, mechanical, or other means, now known or hereafter invented, including
photocopying and recording, or in any information storage or retrieval system, without permission in
writing from the publishers.

Notice:
Product or corporate names may be trademarks or registered trademarks, and are used only for
identification and explanation without intent to infringe.

Library of Congress Catalog Number: 93-30304

Library of Congress Cataloging-in-Publication Data

Spiro, Melford E.
Culture and human nature / Melford Spiro ; edited by Benjamin Kilborne
and L. L. Langness.
p. cm.
Originally published: Chicago : University of Chicago Press, 1987.
With new introd.
Includes bibliographical references and index.
ISBN 1-56000-702-8
1. Culture. 2. Anthropology—Philosophy. 3. Ethnopsychology.
I. Kilborne, Benjamin. II. Langness, L. L. (Lewis L.), 1929- III. Title.
GN357.S68 1994 93-30304
306—dc20 CIP

ISBN 13: 978-1-56000-702-9 (pbk)

ISBN 13: 978-1-138-52182-7 (hbk)
Contents

Introduction to the Transaction Edition

Editors’ Introduction

I. CULTURE AND HUMAN NATURE

1. Culture and Human Nature

2. Some Reflections on Cultural Determinism and Relativism with
Special Attention to Emotion and Reason
3. Preculture and Gender
4. Is the Oedipus Complex Universal?

II. FUNCTIONAL ANALYSIS

5. Social Systems, Personality, and Functional Analysis

6. Religious Systems as Culturally Constituted Defense Mechanisms
7. Collective Representations and Mental Representations in Religious
Symbol Systems

III. RELIGION AND MYTH

8. Religion: Problems of Definition and Explanation

9. Virgin Birth, Parthenogenesis, and Physiological Paternity: An Essay
in Cultural Interpretation
10. Whatever Happened to the Id?
11. Some Reflections on Family and Religion in East Asia
12. Symbolism and Functionalism in the Anthropological Study of
Religion
Index
Introduction to the Transaction Edition

THE ARTICLES collected in this volume, first brought together and re-
published in 1987 by Benjamin Kilborne and L.L. Langness, explore some
of the relationships that, in my view, obtain among culture, society, and the
human mind. Within this framework this collection addresses a range of
theoretical issues including the limitations of cultural determinism and
cultural relativism, the problem of explanation in the social sciences, and
the importance of a comparative approach for the study of social and
cultural systems.
One of the critical emphases of this collection is the importance not only
of cultural variables, but also of psychological and social variables, for the
understanding of human behavior and human belief systems. Seeing as
most of the articles are based on research conducted in the Micronesian
atoll of Ifaluk, an Israeli kibbutz, and a village in Upper Burma, they hardly
minimize the importance of culture for such understanding. Nevertheless
they argue that social actors possess motivational dispositions and cognitive
orientations—both parochial and universal, and both conscious and
unconscious—that, while derived in part from their cultural heritage, are
also derived from their social experience (most especially early experience
in the family), as well as from the evolutionary status of our species.
Spanning the years between 1961 and 1984, some of the views expressed
in these articles have undergone change, not only in the period subsequent
to that time interval, but also within it. Moreover, even when they have not
changed, in retrospect I wish that these views had been expressed with
greater clarity and felicity. Nevertheless, except for typographical errors, I
have decided (as did Kilborne and Langness) to republish these articles in
their original form.
An implicit theme running throughout this collection is that
contemporary anthropological orthodoxy has replaced an older biological
and psychological reductionism with its own cultural reductionism. That is,
in the case of many contemporary anthropologists, their interpretations and
theories of human behavior, belief systems, emotions, and motivation rest
on a wholesale form of cultural determinism and a radical type of cultural
relativism which leaves little (if any) theoretical space for the concept of a
common human nature. Drawing, however, on my own research in Ifaluk,
Israel, and Burma, as well as on published reports on other societies, the
following articles argue that despite the wide diversity in human cultural
systems, important panhuman characteristics nevertheless are to be found,
for example, in sex-role differentiation, aggression, world view, and the
Oedipus complex, among others.
Although my reasons for rejecting wholesale cultural determinism and
radical cultural relativism are suggested in various of the following
chapters, and explicated in detail elsewhere (Spiro 1992: chs.1–2),
nevertheless because of the ambiguities surrounding “culture,” “cultural
determinism,” and “cultural relativism,” not to mention “human nature,” in
order to preclude misunderstanding it is perhaps useful to indicate at the
outset how I am using these terms here. Let us begin with “culture.”
Although for the past quarter century most anthropologists have tended
to view culture as a symbolic or ideational system—some focusing on the
symbols, some on their meanings—nevertheless they have not consistently
employed the term “culture” to designate either the one or the other. Rather,
they also employ that term, just as it had formerly been employed, to
designate two other things, as well: (a) a social group (“one of the critical
characteristics of Ifaluk culture is the virtual absence of interpersonal
aggression”) and (b) the totality of a group’s customs, beliefs, artifacts, and
technology (“Ifaluk culture is characterized by matrilineal descent,
hereditary chieftanship, a belief in ghosts, horticulture, outrigger canoes,”
etc.).
Since such inconsistency in the use of a single term is not only confusing,
but also (as we shall see below) theoretically misleading, in this volume
“culture” designates an ideational system exclusively, that is, a system of
ideas, values, and norms. More specifically, it designates the subset of
ideas, values, and norms that (a) are encoded in public symbols and other
public signs, (b) are acquired by cultural novices in virtue of their social
transmission, and hence (c) are in varying degrees socially shared. In short,
“culture,” as that term is used here, designates traditional ideas, values, and
norms. Correspondingly, “a culture” designates the traditional ideas, values,
and norms that are found in some particular social group, and consequently
constitute the “cultural heritage” of that group.
Often these elements of their cultural heritage are transmitted to cultural
novices intentionally, and by means of verbal instruction, in which case
they are conveyed in the form of propositions, descriptive (“the Buddha
discovered the Noble Eightfold Path”), prescriptive (“it is wrong to steal”),
and directive (“a successful rice crop requires that the paddy fields be
irrigated”). Often, however, traditional values, ideas, and norms are
acquired by the novices indirectly, as a consequence of their observation of,
and participation in, social relations, religious rituals, and the like. Although
in the latter case these cultural elements are not acquired in the form of
propositions, still that is the form, or so I would claim, in which the novices
typically represent them in their minds.
In either event, from the perspective of cultural novices the ideas, values,
and norms that comprise their cultural heritage are wholly “outside” them.
That is, they are “located” in public symbols and other public signs that
“represent” them, in the behavior of social actors that “express” them, and
in the minds of cultural initiates that hold them. As a consequence,
however, of their enculturation, the erstwhile novices themselves come to
represent these cultural elements in symbols and other signs (both public
and private), and to express them in their own behavior, because now these
elements are also located in their minds. In short, following the novice’s
enculturation, one or another element of the cultural heritage come to be
located not only “outside,” but also “inside,” him or her.
However, for cultural novices to become enculturated—that is, for their
psyches and actions to be importantly influenced by the ideas, values, and
norms that comprise their cultural heritage—it is not sufficient that the
novices acquire these cultural elements, but they must also (as I argue in
chapter 2) internalize them. That is, they must acquire them as cognitively
salient personal beliefs.
In the previous paragraph I used the neutral expression “importantly
influenced” by the cultural heritage, rather than “determined” by it, because
the expression “cultural determinism,” confusingly, has four very different
meanings—culturally constituted, culturally caused, culturally shaped, and
socially caused—which must now be distinguished. (Here, I follow some of
the distinctions made by D’Andrade [1992:13] in his analysis of the
expression “culturally constituted.”)
 First, “cultural determinism” is employed in connection with a category
of things which are, as D’Andrade (ibid.) says, created, or brought into
existence, by culture. In contemporary usage this category not only
(properly) includes such symbolic things as nationality, money, baseball,
and unicorns, but it also (improperly) includes such nonsymbolic things as
emotions, gender, the sense of self, and sociality. Let us begin with the
latter.
Although the members of this set are no doubt effected by culture, that
they are, however, created or constructed (whatever that might mean) by
culture is to say the least problematic. Thus, there are substantial grounds
for believing that most emotions (though not their behavioral expression)
are biologically grounded, that many (but not all) gender differences follow
a phylogenetic template, that the sense of self is acquired as a function of
early social interaction, and that human sociality is rooted in our primate (if
not mammalian) needs. In short, to say (as many cultural determinists say)
that the members of this set are “culturally constructed” is perversely to
reject the relevant findings of the neurosciences, ethology, biological
evolution, primate studies, clinical and developmental psychology, and (I
might add) comparative ethnography.
To say, however, that the members of the first set (nationality, money,
baseball, unicorns, and the like) are culturally created or constructed is not
problematic at all. For they, unlike the members of the second set, are as
D’Andrade (ibid.) says, “purely symbolic”; that is, they exist “solely
because [as a function of their cultural heritage] people hold that [these
things] exist and that various actual events count as these things or are
expressions of them.” Since, then, these symbolic entities are constructed
from, and consist of, cultural ideas, rules, norms, and the like, they may be
said to be “culturally determined” in the sense of their being constituted. by
culture.
“Cultural determinism” is employed, in the second place, in connection
with things which, though not themselves cultural, are caused by culture. I
refer here to a class of actions and psychological characteristics (especially
desires and beliefs) of social actors which may be said to be culturally
“caused” in two senses. First, the existence of culture—any culture—is a
necessary condition for the existence of this class of actions, beliefs, and
desires. Second, seeing as the members of this class vary from one social
group to another, the internalization of a particular culture is a sufficient
condition for their existence.
Thus, all cultures prescribe some type of action for greeting one’s
fellows, but whether actors greet each other by shaking each other’s hands,
or instead by placing their own palms together and holding them against
their own forehead, is a consequence of having internalized one particular
culture, rather than another. Again, all cultures contain rules for conflict
resolution, but whether actors desire to resolve their conflicts by litigation,
or instead by conciliation, is also a consequence of having internalized a
particular culture. Similarly, all cultures contain principles for
distinguishing kin from non-kin, but whether actors believe that mothers
(but not fathers) are kin, or instead that fathers (but not mothers) are, or
again that both are, is once again a consequence of having internalized a
particular culture.
Now since the members of this class of actions, beliefs, and desires do
not in themselves consist of culture, it is surely improper to refer to them as
“culturally determined” in the sense of their being constituted by culture.
Since, however, the existence of this class depends on the existence of
culture, and since the particular members of the class vary as a function of
the variability of particular cultures, then both the class and its members
may be said to be “culturally determined” in the sense of being caused by
culture.
“Cultural determinism” is employed, in the third place, in connection
with a class of social actions which, while not caused by culture, are (to use
D’Andrade’s term [op. cit.]) shaped by it. This class consists of actions that
are instigated by biologically or experientially acquired emotions, needs, or
drives. Since, now, it is in the nature of emotions to press for expression, of
needs to press for fulfillment, and of drives to press for satisfaction, it can
be presumed that whatever their culture, or even in the absence of culture,
human actors (like nonhuman actors) would express these emotions, fulfill
these needs, and satisfy these drives in some form of action.
Since, however, in diverse social groups these actions take one form
rather than another, it can also be presumed that often their form is a
function of the values, rules, and norms of each group’s particular culture.
Thus, as a function of the emotional display rules of their respective
cultures, in one social group males express their panhuman emotion of
anger by verbal aggression against its elicitor, whereas in another group
they displace their anger and express it in headhunting. Similarly, as a
function of the gender norms of their respective cultures, in one social
group females satisfy their panhuman maternal drive by bearing many
children, whereas in another group they do so by the intensive nurturance of
a few children. Again, as a function of the achievement values of their
respective cultures, in one social group actors fulfill their panhuman need
for self-esteem by acts of religious piety, whereas in another group they do
so by the accumulation of wealth.
In sum, since all human beings possess biologically and experientially
acquired emotions, needs, and drives which, regardless of culture, press for
expression, fulfillment, and satisfaction respectively, it is entirely
misleading to claim that the actions by which these purposes are achieved
are “culturally determined” in the sense of their being caused by culture.
Seeing, however, that the forms they take are a function of the values and
norms of particular cultures, these actions may properly be said to be
“culturally determined” in the sense of their being shaped by culture.
Finally, “cultural determinism” is employed in connection with a large
class of drives, desires, beliefs, and actions which, although they may be
shared by social actors, are in no sense “culturally” determined, as we may
now see.
From the moment of birth infants, and later children, participate in an
ever expanding circle of social relations, and they undergo an ever
increasing variety of social (and other) experiences, all of which occur
many months and sometimes many years prior to their acquisition of one or
another of the ideas, values, and norms that comprise the cultural heritage
of their group. This is so for at least three rather obvious reasons
First, seeing as they are transmitted in the medium of language, it is
impossible for children to acquire any of these cultural elements prior to
their development of the requisite linguistic competence; until then, they
perforce remain cultural novices. Second, even after they attain the requisite
linguistic competence, many of these cultural elements may nevertheless
not be transmitted to children on the grounds, actual or presumed, that they
have not achieved the requisite maturity, either biological or intellectual.
Third, even after their attaining the requisite biological and intellectual
maturity, others of these cultural elements may not be transmitted to
children on the grounds that they have not yet achieved the culturally
stipulated social age for their acquisition.
 Does this interval between birth and the acquisition of culture (which, of
course, varies from one social group to another) entail that the precultural
child is a cognitive, emotional, and motivational tabula rasal Although
implicit in the regnant theory of wholesale cultural determinism (according
to which virtually all nontrivial psychological characteristics of social
actors are either constituted or caused by culture), nevertheless such an
entailment is surely not warranted, as any observer of children (whether
child psychologist or parent) can testify. For apart from their biologically
determined drives and emotions, precultural children also possess any
number of socially determined beliefs and desires, that is, beliefs and
desires acquired from their social (but precultural) experience.
Thus, as a consequence of their socially variable interactions with parents
and other socializers, precultural children may come to believe, for
example, that their social world (to put it more starkly than is actually the
case) is nurturant or punitive or both; that they must compete, or need not
compete, with others in order to receive care and attention; that adults fulfill
or frustrate their needs, or both; that certain of their actions elicit approval
and love, whereas others elicit disapproval and punishment; and so on. All
of these beliefs, and many more, are constructed by the children from their
precultural social experience.
Similarly, as a consequence of their social interactions, and/or the beliefs
to which they give rise, precultural children also acquire any number of
desires: desires, for example, to eat certain things and avoid others; to
perform certain acts, but inhibit others; to achieve certain goals, but eschew
others; to comply with some demands, but violate others; to imitate certain
persons, but refrain from imitating others; to display some feelings, but
suppress others; and so on. All of these desires, and many more, are
socially, but not culturally, acquired.
It is clearly misleading to refer to any of the previously mentioned beliefs
and desires as culturally determined when, in fact, they are not determined
by culture at all. Instead, they are socially determined, in the sense of being
either caused or shaped by children’s social (but precultural) experience. It
might be argued, of course, that insofar as the behavior of parents and
others is governed by cultural norms and values, and insofar as they express
these norms and values in their interaction with their precultural children,
then although the children’s social experience may be the direct and
proximal cause of their beliefs and desires, nevertheless culture is the
indirect and distal cause. Such an argument, however, is valid to only a
limited degree.
Consider, for example, the socialization of aggression. Thus, even on the
assumption, for example, that parents and other socializers respond to the
aggressive action of children with physical punishment as a means to
inculcating cultural values regarding aggression, nevertheless it is unlikely
that from this punitive experience the children internalize such values at all.
For while it is likely that as a consequence of this experience they inhibit
their subsequent aggressive desires, it is also likely that it arouses anger and
fear in the children who, consequently, construct the belief that their
socializers are oppressive, and acquire the desire to aggress against them
(which, from fear, they probably repress).
As a consequence of their social experience, children may acquire a set
of cognitions and motivations which, not so paradoxically, are the very
opposite of the cultural values and norms that their socializers intended to
inculcate. According to personality studies, moreover, these cognitions and
motivations persist, whether consciously or unconsciously, as more or less
durable components of their character structure. Thus, in our example, as
the children become adults they may displace onto other authority figures
their belief that parents and other socializers are oppressive, and they may
express their feelings of anger and hostility to them by similarly displacing
them onto these transference objects. A similar process operates mutatis
mutantis in the case of experientially acquired prosocial cognitions and
motivations.
Let us now summarize this discussion of the terms “culture” and
“cultural determinism.” Formerly, when it was employed globally, “culture”
designated at one and the same time a social group, the customary behavior
patterns of the group, and its traditional ideas, values, and norms. Since this
polysemic usage was too heavy a semantic load for one and the same term
to carry, latterly most anthropologists have come to employ “culture” to
designate the third referent only.
This restricted meaning of the term “culture” entails that the expression
“cultural determinism” should correspondingly be restricted to the shared
behavioral and psychological characteristics of social actors that they
acquire as a consequence of the transmission of the traditional values,
norms, and ideas of their social group. Given this restricted, and logically
consistent, meaning of “cultural determinism,” the persistent
anthropological claim that virtually all of a group’s nontrivial behavioral
and psychological characteristics are culturally determined is clearly
invalid, for a large share of these characteristics are not culturally, but
biologically or socially determined.
Even, however, in this restricted meaning, “cultural determinism” is a
confusing expression because it is used not only in reference to those
psychological and behavioral characteristics that are caused by culture, but
also in reference to those that are shaped by it, as well as to any number of
purely symbolic entities that are constructed by culture, and are therefore
constituted by it.
By viewing all nontrivial characteristics of social actors as determined by
culture, thereby precluding any biological and social determinants, and by
conflating the separate concepts of constitute, cause, and shape, many
anthropologists (and other social scientists) claim that virtually everything
in the human world is culturally determined. I shall return to this concept of
wholesale cultural determinism below, but first it is necessary to examine
the concept of cultural relativism.
Like “cultural determinism,” the expression “cultural relativism” is used
in a number of different, and confusing, ways. Thus, although it is
frequently used to denote the fact that the cultural ideas, values, and norms
of any one group may importantly differ from that of any other group, this
usage confuses cultural relativism with cultural diversity. To say, for
example, that religious ideas, family values, and sexual norms are culturally
diverse is one thing, but to say that they are culturally relative is quite
another, for whereas “cultural diversity” denotes a (well-documented) fact,
“cultural relativism” denotes a (controversial) judgment regarding that fact.
Moreover, even when “cultural relativism” is used in the latter (and only
proper) sense, this expression is still confusing, because in current usage it
denotes not one, but three judgments, which might be termed “descriptive,”
“normative,” and “epistemological,” respectively. Since elsewhere (Spiro
1986, 1992: chapter 1) I have explicated the meanings of these judgments in
some detail, here I shall describe them only briefly.
If, as the Oxford English Dictionary puts it, “relative” means “arising
from, depending on, or determined by relation to something else,” then
descriptive cultural relativism might refer to two very different judgments:
(a) the judgment that the diversity in cultural variables (traditional ideas,
values, and norms) across social groups is a function of the diversity in
noncultural variables (ecology, biology, subsistence economy, and the like);
(b) the judgment that the diversity in social and psychological variables
(social behavior, motivation, cognition, emotions, and the like) across social
groups is a function of the diversity in cultural variables.
Because most contemporary culture theorists, with the exception of
cultural materialists, reject the first judgment (wrongly in my view) on the
grounds that it is materialist or reductionist, it is only the second to which I
shall refer to here as “descriptive” cultural relativism. Accordingly, as a
judgment regarding the fact of cultural diversity, its claim is not that culture
is relative, but that the diverse social and psychological characteristics of
different social groups are culturally relative, in short, that their diversity is
dependent upon, and exists in relation to, the diverse cultures of these
different groups.
Since any judgment, whether explicitly or implictly, is based of course on
some theory, it is now perhaps obvious that descriptive cultural relativism is
based on the theory of cultural determinism (in one or all of the senses
distinguished above). I shall return to this point below after examining the
other two types of cultural relativism.
Normative cultural relativism (like descriptive relativism) also consists of
two judgments regarding cultural diversity which may be stated as follows.
First, because evaluative norms are culturally diverse, and hence any
evaluation of different cultures on a scale of merit or worth is ethnocentric,
the only valid judgment that can be made regarding such variable cultural
systems as religion, science, ethics, kinship, and the like, is that none is
better or worse than any other. They are only different. Second, because any
evaluation of the mental and emotional characteristics of different social
groups is perforce similarly ethnocentric, the only valid judgment that can
be made regarding their logical ability, emotional maturity, and the like, is
that no group is superior or inferior, either intellectually or emotionally, to
any other. They are only different.
Since these claims of normative relativism are critically evaluated in
chapter 2 (though for a more recent, and far-searching critique, see
Edgerton 1992) it would be redundant to repeat my criticisms here. Instead,
I shall proceed directly to the third type of cultural relativism,
epistemological relativism.
A recent arrival, epistemological relativism takes its point of departure
from (what might be termed) the “strong” form of descriptive relativism, a
form which combines a radical view of cultural diversity (the view that the
latter is virtually limitless, so that every culture is unique in the nontrivial
sense of being incommensurate with any other) with wholesale cultural
determinism. From these premises, epistemological relativism draws two
far-reaching epistemological conclusions: (a) panhuman generalizations
regarding culture, human behavior, and the human psyche are likely to be
vacuous, if not false and (b) panhuman theories of culture, human behavior,
and the human psyche are likely to be trivial, if not invalid.
Although these conclusions are logically valid, nevertheless in my view
they are false because, as I have argued elsewhere (Spiro 1992: chapters 1
and 2), their premises (wholesale cultural determinism and radical cultural
diversity) are false. I shall return to this substantive argument below, but
here I wish to make a methodological observation.
Rather than constituting an unbridgable obstacle to the discovery of
nonvacuous generalization and nontrivial theories, as epistemological
relativism claims, cultural and psychological diversity constitute, instead,
an unparalleled opportunity for their discovery. Thus, by the application of
Mill’s canons of similarity and difference to comparative ethnographic data,
anthropology has been able to forge a powerful method for formulating, as
well as testing, generalizations and theories, a method by which some of
them have been confirmed, and others disconfirmed. For an inventory and a
summary discussion of these findings, see Levinson and Malone (1980), as
well as some of the other references cited below.
Based on the foregoing discussion of cultural determinism and cultural
relativism, we may turn, finally, to the second term comprising the title of
this book, “human nature.” Although for many years I had been interested
in the philosophical relevance of the problem of human nature, I had not
developed an interest in its anthropological relevance until 1952, when I
was invited by A.I. Hallowell to present a paper (Spiro 1954) to a
symposium on that question to the annual meeting of the American
Anthropological Association.
Although from its very beginning, American anthropology had a special
interest in the parameters of human nature (Boas 1911), Hallowell’s
decision to organize this symposium was motivated by Robert Redfield’s
response to a paper by Hallowell (1953) on culture and personality, in the
course of which which he, Redfield, had remarked, “In this territory we find
ourselves characterizing human nature as a recognized and subject matter of
our science….” (quoted in Hallowell:1976: 224).
Now if someone, like Redfield, who was hardly known for an interest in
human psychology, had come to recognize the anthropological importance
of the concept of human nature, then—so Hallowell observed in his
introduction to the symposium—the time had come to publically reassess
the anthropological relevance of this concept. And to better promote such a
reassessment, he invited representatives from various subfields of
anthropology—Walter Goldschmidt, G.P. Murdock, S. L. Washburn, and
myself—to participate in the symposium.
Unlike these participants who, like many other members of their
anthropological generation, viewed the concept of human nature—or, as
many of them preferred, “the psychic unity of mankind” (e.g., Kroeber
1948, Wallace 1961)—as indispensable for understanding the mutually
reciprocal relationships among human biology, psychology, society, and
culture, many contemporary anthropologists reject this concept. Their
rejection is based on two arguments.
First, since nature and culture are binary opposites, with the emergence
of culture the latter has virtually become the sole determinant of the human
psyche which, in turn, has come to transcend nature. That being the case,
the concept of a human nature is both “essentialist” and “reduction-ist.”
Second, even though culture is virtually the sole determinant of the human
psyche, nevertheless given the wide diversity in culture, rather than a
common human psyche, there are only culturally parochial and diverse
psyches.
Consider, for example, Geertz’s comment on the psychological
differences that (presumably as a function of their subcultural differences)
distinguish the various categories of actors—scholars, artists, scientists,
professionals, and administrators—comprising the American academic
world. These actors, Geertz (1983:160) writes, “are radically different not
just in their opinions, or even in their passions, but in the very foundations
of their experience.”
If because of their subcultural differences, the actors comprising one
category of only one subgroup of a single society are radically different in
their psychological characteristics from the actors comprising the other
categories of that subgroup, then afortiori the psychological differences
across these subgroups must be even more radical, while the psychological
differences across societies (given the magnitude of their cultural
differences) must be more radical still. Indeed, for the proponents of radical
cultural diversity, the alleged incommensurability of cultures entails a
corresponding incommensurability in the human psyche.
Let us evaluate these two arguments for rejecting the concept of a
panhuman nature, beginning with the first. That nature and culture are
mutually exclusive concepts is a notion that I would have thought had been
laid to rest many years ago, not only by philosophers such as John Dewey
(1929), but also by generations of anthropologists (both cultural and
biological) who have argued that because we are a generalized, plastic, and
instinctless species, culture is our species-specific mode of adaptation. In
short, culture is to a human mode of existence what the biological
specializations of other species are to infrahuman modes of existence.
If culture is an adaptive requirement of our species, then culture and its
psychological (cognitive, motivational, and emotional) products are no less
a part of human nature than the human biological organism and its
psychological products.
But even granting this point, still that would not entail, according to
many contemporary anthropologists, that there is a panhuman nature.
Although culture, as they view it, is virtually the sole determinant of the
human psyche, nevertheless since there is not one culture, but many diverse
cultures, so also there are many diverse human psyches. As Shweder
(1990:2) has put it, the differences among cultures result “less in psychic
unity for humankind than in ethnic divergences in mind, self, and
emotions.”
This, however, is a straw man argument because the concept of psychic
unity never meant, as Shweder presumably takes it to mean, that all human
groups possess identical psychological characteristics, but rather that they
possess a common (and critical) subset of them (Kroeber 1948:00, Wallace
1961). As Hallowell put it,
man is characterized by a unique psychic structure the generic form of which we have only begun
to discern in the common features that underlie the range and variation of personality structure
that have been empirically investigated in recent years. (Hallowell 1955:5, my italics)

To understand the full import of this passage it is perhaps useful to point

out, lest it has been forgotten, that it was not interpretive and postmodern
anthropologists who first called our attention to the wide diversity in human
cultures, but rather Hallowell, Kroeber, and Wallace—together with many
colleagues and predecessors—who did so. Moreover, they demonstrated
this diversity by their own intensive field work. In short, these proponents
of the concept of psychic unity can hardly be considered to have been cross-
cultural naifs.
It is also useful to point out, lest perhaps this too has been forgotten, that
it was not interpretive and postmodern anthropologists who first called our
attention to the wide diversity in the human psyche, but rather that same
previous generation of anthropologists who did so. It was, after all, the
members of the (now much maligned) “culture and personality” school,
which included Hallowell and Wallace, who (in their studies of modal
personality, national character, basic personality structure, and the like) first
established the existence of “ethnic differences in mind, self, and
emotions.” In short, these proponents of psychic unity can hardly be
considered to have been psychological naifs.
It can be presumed that this earlier generation of anthropologists insisted
upon the psychic unity of mankind not because they were a bunch of
country bumpkins, innocent of cultural and psychological diversity, but
because they believed (as I do) that there are sound empirical grounds—
biological, social, and cultural—for doing so. Let us now briefly consider
these grounds starting with the biological. Once again I begin with
Hallowell.
When we have more knowledge of the range and variation in the human personality structure in
relation to major provincial [social and cultural] determinants we shall be able to state with more
precision what is common to man everywhere. By that time we may be able to construct a better
picture of the psychobiological structure of man as an evolving primate. (Hallowell 1955:13)

And again,
The distinctive psychological structuralization of men, the novel qualitative features that
characterize the psychodynamic of a human level of adjustment, combined with the generic
capabilities of man that are rooted in the phylogenetic status of our species, lie at the very core of
our human nature. (Hallowell 1976:227).

As these passages indicate, for Hallowell (and his colleagues) the concept
of a universal human nature is, in the first instance, an inevitable
consequence of the findings of biological evolution, according to which all
human beings, having evolved from a common primate ancestor, comprise
a single biological species. On the assumption now that our biology is not
totally irrelevant for our psychological functioning, then it can be presumed
that as members of a single species all human beings share a set of
biologically determined psychological characteristics which comprise one,
but only one, component of a common human nature.
Granted that we evolved from an early African primate (and that even
today we share approximately 98 percent of our molecular DNA with
contemporary African chimpanzees and gorillas), still it is not our simian,
but our hominid phylogeny that accounts for the specifically human (as
opposed to the simian) component of human nature. In this regard I would
emphasize, in addition to the frequently mentioned expansion of the
hominid brain (which accounts, in the largest part, for the critically
important cognitive dimension of a uniquely human nature), prolonged
infantile dependency and the suppression of estrus.
I single out these two biological characteristics in particular because,
jointly, they probably account for some of the most important social
characteristics of our species, including the relatively permanent biparental
family, the prolonged and bisexual socialization of children, and intrafamily
incest avoidance. These characteristics of human social systems, in turn, are
critical for the emergence of human nature because, arguably, they are the
most important social determinants of the human psyche (particularly its
emotional and motivational dimensions) which, as some anthropologists
have argued, are no less important than its biological determinants. Once
again Hallowell:
The distinctive psychological organization of the human being, whether described as mind or
personality structure, is just as much a function of his membership in a [homind type] social
group as it is a function of his inherited organic equipment. (Hallowell 1955:5)

Having now briefly discussed the likely biological and social

determinants of a putatively panhuman nature, we may finally turn to
culture, the third (and last) of these likely determinants. Seeing, however,
that many contemporary anthropologists (mostly wholesale cultural
determinists) argue that because of the wide range of cultural diversity, it is
culture, above all else, that precludes the very possibility of a panhuman
nature, on what grounds then do other anthropologists (both past and
present) argue that the opposite is the case?
 The grounds are straightforward and simple. Since cultural diversity is
constrained by the adaptive requirements (ecological, biological, economic,
and social structural) of any human society, the range of diversity, though
dramatic, is not unlimited. Consequently, diverse human groups not only
display (nontrivial) cultural differences, but they also display (nontrivial)
cultural universals, and because of the universals culture constitutes one,
but only one, determinant of a putatively panhuman nature.
Cultural universals, it may now be emphasized, are neither esoteric nor
underreported. Not only can many of them be readily, if randomly,
described by anyone with more than a cursory knowledge of comparative
ethnography, but more important, a large share of them, and of the known
social and psychological universals, as well, have been abundantly
documented and extensively commented upon in numerous systematic
studies. Drawing on a small sample of both older and recent studies, these
include Arnheim (1988), Brown (1991), Davenport (1987), Edgerton
(1992), Hallowell (1955: chs. 1 and 4, 1976: chs. 5 and 6), Hockett (1973),
Kluckhohn (1953, 1959), LaBarre (1954), Levinson and Malone (1980),
Lloyd and Gay (1981), Murdock (1945), Staal (1988), Tiger and Fox
(1971).
That in their concern with cultural diversity, interpretive and postmodern
anthropologists have overlooked, when they have not denied, the existence
of the many cultural universals summarized in these, and other studies, is
surely paradoxical. For if culture is, as they claim, the sole determinant of
what is truly “human” about human beings, then either (a) all human groups
comprise a common humanity, in which case they must share a set of
nontrivial cultural universals, or (b) they do not share such universals, in
which case they do not comprise a common humanity.
Since I myself believe that the existence of cultural universals is as
firmly established as cultural differences, and no less important for the
development of the human psyche, I have no doubt that all human groups,
no matter how diverse their cultures, comprise a common humanity. Since,
however, I do not subscribe to wholesale cultural determinism, I believe this
on the grounds not only of cultural universals, but also (as I discussed
above) of biological and social structural universals.
Believing in the existence of cultural universals, and their importance for
the development of the human psyche, I also believe (as I argue in chapter
2) that the process of enculturation is at one and the same time a process of
humanization. That is, by internalizing their respective cultures, cultural
novices not only become Ifalukians, Burmans, and Israelis (each different
from one another, and from us), but they also, and simultaneously, become
human beings (each similar to each other, and to us). That is so because our
common humanity is constituted by universal culturally determined
characteristics, as well as by universal biologically and socially determined
characteristics.
What some, at least, of these characteristics—the characteristics that
make up our “psychic unity,” or our “human nature”—might be, and what
some of their biological, social, and cultural determinants might be, are the
subjects for various of the following chapters.

References
Arnheim, Rudolf. 1988. Universals in the arts. Journal of Social and
Biological Structures 11:60–65.
Boas, Franz. 1911. The mind of primitive man. New York: Macmillan.
Brown, Donald E. 1991. Human Universals. New York, McGraw-Hill.
D’Andrade, Roy G. 1992. Toward a Theory of Emotion and Culture.
Ms.
Davenport, William H. 1987. The cultural anthropology of sex:
Description (ethnography) and comparison (ethnology). In
Theories of human sexuality, James H. Geer and William T.
O’Donohue, eds., 197–236. New York: Plenum.
Edgerton, Robert B. 1992. Sick societies: Challenging the myth of
primitive harmony. New York: The Free Press.
Geertz, Clifford. 1983. Local knowledge. New York: Basic Books.
Hallowell, A. Irving. 1955. Culture and experience. Philadelphia:
University of Pennsylvania Press.
---------. 1976. Contributions to anthropology.. Chicago: University of
Chicago Press.
Hockett, C.F., 1973. Man’s place in nature. New York: McGraw-Hill.
Kluckhohn, Clyde. 1953. Universal categories of culture. In
Anthropology today: An encyclopedic inventory, 507–23. Chicago:
University of Chicago Press.
______. 1959. Common humanity and diverse cultures. In The human
meaning of the social sciences, Daniel Lemer, ed., 245–284. New
York: Meridian.
Kroeber, A.L. 1948. Anthropology. New York: Harcourt Brace and
Company.
La Barre, Weston. 1954. The human animal. Chicago: University of
Chicago Press.
Levinson, David and Martin J. Malone. 1980. Toward explaining
human culture. New Haven: HRAF Press.
Lloyd, Barbara, and John Gay, eds. 1981. Universals of human
thought: Some african evidence. Cambridge: Cambridge
University Press.
Murdock, George Peter. 1945. The common denominator of cultures.
In The science of man in the world crisis, Ralph Linton, ed., 123—
42. New York: Columbia University Press.
Shweder, Richard A. 1990. Cultural psychology—What is it? In
Cultural psychology: Essays on comparative human development,
James E. Stigler, Richard A. Shweder, and Gilbert Herdt, eds., 1–
46. Cambridge: Cambridge University Press.
Spiro, Melford E. 1954. Human nature in its psychological
dimensions. American Anthropologist 56:19–30.
---------. 1986. Cultural relativism and the future of anthropology.
Cultural Anthropology 1:259–86, 1986.
---------. 1992. Anthropological other or Burmese brother: Studies in
cultural analysis. New Brunswick: Transaction Publishers.
Stall, Frits. 1988. Universals: Studies in Indian Logic and Linguistics.
Chicago: University of Chicago Press.
Tiger, Lionel, and Robin Fox. 1971. The Imperial animal. New York:
Holt, Rinehart & Winston.
Wallace, Anthony F.C. 1961. Culture and personality. New York:
Random House.
Editors’ Introduction

IN THIS volume are assembled for the first time twelve of the finest
theoretical papers of Melford E. Spiro, one of the most distinguished figures
in anthropology today. Professor Spiro founded the department of
anthropology at the University of California, San Diego, where he is
currently Presidential professor of anthropology. A member of the National
Academy of Sciences, he combines long-standing interest in the
philosophical problems of human existence with extensive anthropological
fieldwork. In short, he is perhaps the most prominent figure in the country
in psychological anthropology and culture and personality, and clearly one
of the outstanding scholars in the areas of social and cultural theory.
This collection of papers, spanning the years between 1961 and 1984,
appropriately demonstrates Spiro’s contributions to the exploration of
human nature and culture. It also addresses the fundamental issues of
cultural relativism, the problem of explanation in the study of culture, and
the anthropological study of religion and belief systems. Grouped according
to theme rather than chronology, these papers have been selected and
arranged so as to give a comprehensive view of Professor Spiro’s
theoretical works.
During the course of his career Spiro has developed a coherent and
systematic body of theory modified by fieldwork, experience, and changes
of mind. This body of theory has over the years provided him with a
framework within which to develop his ideas and pursue various research
projects in the field. Thus Spiro stands out as a scholar of unusual integrity,
at once capable of changing his mind, of modifying his theories in the light
of evidence, and of far-reaching efforts to understand the human condition.
With an eye for well-constructed arguments, he incorporates ideas from a
wide variety of sources.
One of the most fundamental emphases in Spiro’s work has been on the
need to consider both psychological and cultural forces and structures as
part of any explanation of human behavior. In fact, a strong case can be
made for the ineluctable necessity of doing this in the explanation of all
behavior, although some anthropologists and sociologists would either deny
or ignore it. Anthropologists inevitably employ psychological theories and
premises whether or not they admit to doing so. As S. F. Nadel once put it,
they “sneak them in through the back door.” Psychological factors are,
Spiro holds, an explicit and necessary part of any adequate theory of culture
and human action. Indeed, it is difficult to see how a genuine and useful
theory of culture could persist in the absence of an attempt to integrate
within it a theory of the human mind. Spiro stands out as a foremost
American anthropologist who has not wavered in his efforts to address the
problems of human nature, and who has systematically linked theories of
psychological motivation with social structure and belief systems.
The reader will note that Spiro consistently uses elements of
psychoanalytic theory in his explanations and in his own theoretical
constructions. While essentially a proponent of the ethnographic method,
Spiro nonetheless holds that there are deep motivational structures
underlying all human behavior, rational and irrational alike. His emphasis
on psychoanalytic theory, which entails a theory of panhuman nature,
constitutes one of the most fundamental thrusts of this book and, indeed, of
Spiro’s work in general. Thus, although psychoanalytic theory traditionally
has been criticized and often dismissed by anthropologists, who assume that
it is culturally relative and therefore of no use in ethnographic fieldwork
and interpretation, Spiro makes it one of the keystones of his theorizing and
methodology. He argues that psychoanalysis provides the best available
coherent body of theory linking the (panhuman) structure of the family to
concepts of individual motivation, on the one hand, and to concepts of
culture and social organization, on the other. For example, in his Oedipus in
the Trobriands Spiro demonstrates that the conclusions reached by
Malinowski were far too hastily drawn. Spiro’s reanalysis of Malinowski’s
material shows that there is an Oedipal constellation in the Trobriands,
whereas Malinowski’s work, Sex and Repression in Savage Societies, is
commonly assumed to have proven there is not. Thus Spiro contends that
the Oedipal complex can be used to explain Malinowski’s Trobriand data
with distinctly more parsimony and comprehensiveness than Malinowski
employs in his theoretical revision of Freud.
 In the first section, on human nature and culture, Spiro’s debt to Freudian
theory stands out clearly. While disagreeing with sociobiologists, he
emphasizes the need for a theory of human nature implying panhuman
psychological characteristics rooted in early childhood experience. Such
psychological characteristics are, Spiro demonstrates, indispensable in
understanding both anthropological questions and anthropological data.
In his interest in theories of human nature, Spiro resembles his friend and
mentor, A. Irving Hallowell. The affinity between the two writers is
palpable also in the second section of this book, where Spiro deals with
motivations for the performance of roles defined as culturally important.
Such sober doggedness in pursuing questions of rather fundamental
significance characterizes both Hallowell and Spiro, and distinguishes these
two writers from more fickle and sometimes also more popular
contemporaries.
Notions of human nature and culture do not stay put in any bed of
Procrustean theory, whether of functionalism, structuralism,
environmentalism, or of the superorganic. Because of this, these notions are
better addressed directly and honestly than indirectly by way of whatever
theories happen to be popular or appealing at the moment. Spiro has built
his career largely on direct and forthright attempts to deal with the questions
of human nature and culture, personality, and cultural values and beliefs,
and he has sought to grasp these questions in all their complexity.

The first section of this book, Culture and Human Nature, includes
papers representing Spiro’s emphasis on a theory of human nature grounded
in individual needs. The opening chapter, as well as those which follow,
speak to the tendency in American anthropology toward both materialism
and economic determinism. Spiro energetically criticizes the radical cultural
relativism and determinism which he sees as characterizing much American
anthropology. He has argued consistently over the past thirty years for the
importance of a theory of human nature rooted in individual human needs,
yet seen in terms of the perceptual configurations and meanings of symbols
and symbol systems in specific societies at particular times.
In “Some Reflections on Cultural Determinism and Relativism…
(chapter 2), for instance, Spiro disputes the claims of cultural relativists by
linking personality to culture as well as by refusing to concede that
personality can ever be reducible to culture. Insisting on the psychological
basis for the persistence of religious beliefs and cultural institutions, he also
challenges the claims of the cultural materialists. Moreover, he strongly
emphasizes both culture and human nature as manifested by the dynamics
of the Oedipus complex and the nature of the human family. For Spiro,
human nature (panhuman psychological characteristics) results from the
interaction of phylogenetically determined biological (inherited)
characteristics, and the functional requirements of any social system. As all
human social systems depend upon the biparental family, it is in such
families that the child’s basic cognitive and motivational orientations are
molded. Therefore, it is on these orientations that all social systems depend.
In this sense, Spiro demonstrates that in fundamental respects the problem
of the locus of culture is a specious one.
“Preculture and Gender” examines concepts of gender identity in kibbutz
members. Whereas in works written earlier Spiro held that gender identity
could be learned, in this chapter he explicitly holds that it cannot be
understood merely in terms of learned roles. Spiro’s change in emphasis is
the direct result of fieldwork and his growing conviction that learning
theory, while valuable, is not enough to account for what he saw in the
kibbutz. A theory of human nature, of the human family, and of the ways in
which families respond to human needs was required.
This theme (of the necessity of developing a theory of human nature and
culture) is further elaborated in chapter 4, “Is the Oedipus Complex
Universal?” In this paper Spiro argues that the Oedipus complex is
structurally invariable (i.e., an expression of invariant features of the human
family on the one hand, and of early stages of emotional cognitive
development on the other). However, he points out, the Oedipus complex is
variable when viewed functionally or culturally. The same material
Malinowski relied upon has been reinterpreted convincingly by Spiro to
demonstrate the presence of an Oedipus complex in the Trobriands as well
as elsewhere.

The second section, Functional Analysis, presents an overview of

Professor Spiro’s theory of functionalism, and illustrates how he
conceptualizes the functions of psychological drives and cultural systems
and institutions.
The first paper of this section, “Social Systems, Personality, and
Functional Analysis,” attempts to explain the interrelationship between
personality and culture. This chapter focuses, of course, on mutually
dependent questions which have occupied social thought for centuries: How
do human societies persuade their members to behave in conformity with
cultural norms, and why do individuals feel motivated to conform to social
values? Or, to put the questions differently: How are cultural values
inculcated in individuals such that these individuals are motivated to
conform to cultural norms? Where do social laws come from and why are
they obeyed? In exploring these questions, Spiro analyzes the way in which
human personality and motivational configurations affect the functioning of
social systems. Indeed, he investigates how they make the functioning of
social systems possible. When compared with scholars concerned only with
putatively nonpsychological kinds of analysis (e.g., Leslie White, Marvin
Harris), Spiro’s position constitutes an important contribution to the entire
field of social theory.
Spiro develops a “functional” view of personality and culture, which are
interdependent. This position is illustrated in chapters 6 (“Religious
Systems …”) and 7, where he applies functional definitions of culture and
personality to the analysis of religious systems. Once again he stresses the
importance of early childhood experiences in linking family structure with
religious taboo and guilt. Consequently, he can speak of “culturally
constituted defense mechanisms,” for religious systems (and other belief
systems) serve defensive functions. These chapters, then, explore how the
psychodynamic notions of “drive” and “defense” can be used in the
functional analyses of cultural belief systems. Broadly speaking, chapters
5–7 illustrate the functional analysis of psychodynamic processes and
cultural belief systems, as represented both in individual projective
identification and projective mechanisms, and in culturally shared (and
reinforced) means available to individuals for defensive reactions.
Chapter 7 (“Collective Representations and Mental Representations in
Religious Symbol Systems”) constitutes an attempt to show the relationship
between cultural symbols and personal symbolism, between conscious and
unconscious meanings both in the ways in which symbols are held and
believed in, and in the ways in which they are explained. The chapter also
discusses with particular cogency the distinctions between primary and
secondary process thinking as defined and used in psychoanalysis. Primary
process thinking reflects unconscious drives and fantasies; secondary
process thinking is more rational, purposeful, acceptable, and explicable.
 The final section of this book, Religion and Myth, illustrates in different
ways Spiro’s use of early family experience as an organizing principle.
Religion, he shows, is necessarily rooted in the family. Furthermore, family
dynamics are internalized in all human beings, who, as it were,
“familiarize” themselves with the outside world on the basis of their own
family experiences. Children internalize not only their perceived,
experienced, and remembered relationships with their parents, but also their
assessment of the judgments and evaluations which their parents have made
of their own behavior, thoughts, and feelings. This process continues
throughout life as adults read into adult relationships the wishes, fantasies,
and fears of their own childhoods.
Thus Spiro’s emphasis on the nature of the human family enables him to
continue to pursue the questions—so alive in the nineteenth century but
frequently believed to be superceded today—of taboo and the family. In this
way he allows for the irrational, linking the human family to religious belief
systems. Professor Spiro does not subordinate religion to a superordinate
notion of social function, any more than he subordinates individual
psychodynamics to such functions—and for similar reasons. However, the
anthropological emphasis on context, together with the fieldworker’s need
to understand, for example, the Burmese beliefs from the inside, do not lead
Spiro to the conclusion that human behavior is infinitely plastic (infinitely
relative). What varies cross-culturally is surface structure, not deep
motivational structure.
In “Religion: Problems of Definition and Explanation.” Spiro reviews
major definitions of religion, developing his own view. For him religion is
“an institution consisting of culturally patterned interaction with culturally
postulated superhuman beings.” He then elaborates still further the
arguments concerning the “truth” of religious beliefs, and analyzes their
various manifest and latent functions, both psychological and sociological.
Finally, he assesses “causal” as distinct from “functional” explanations for
religious belief.
In “Virgin Birth: Parthenogenesis and Physiological Paternity,” Spiro
explores rationally and irrationally held beliefs. He points out that there are
many kinds of connections between evaluations of how “true” or “false,”
how “rational” or “irrational” beliefs may be judged to be, thereby
challenging structuralist assumptions that it is irrational to hold a false
belief and rational to hold a true one. Once again Spiro stresses the
pertinence of the functional/psychocultural view of explanation in
anthropological studies of religion. Anthropologists are responsible for
determining how beliefs are held (which entails psychodynamic analysis),
as well as for understanding how they are labeled “true” and “false” by
members of the society in question. When anthropologists proceed this way,
Spiro shows, what beliefs “do” for their believers may be assessed with
respect to empirical evidence.
“Whatever Happened to the Id?” succinctly lays out Spiro’s propositions
about human nature and culture. He shows that the often fanciful claims of
structural analysis are frequently just that: fanciful. Reinterpreting the
Bororo myth which Lévi-Strauss uses as evidence for the transition from
the stage of nature to that of culture in The Raw and the Cooked (1969), he
argues for a simpler, more obvious (less arcane) interpretation derived from
psychoanalytic theory.
Chapter 11, “Some Reflections on Family and Religion in East Asia.”
applies Spiro’s ideas about the importance of the family (and the Oedipus
complex) toward understanding religion, emphasizing the panhuman
character of the human family system and speculating on potential links
between family tensions in East Asia and certain aspects of East Asian
religions.
Finally, “Symbolism and Functionalism in the Anthropological Study of
Religion” deals with the relationship of symbolic to functional analyses,
and particularly with the relationship of cultural symbols to social
experience. Spiro points to fundamental similarities between the symbolism
of Judaism and Buddhism, linking certain religious beliefs to (culturally-
defined) family roles. Religious symbols, he argues, are created not only, as
Lévi-Strauss contends, to “think by,” but also to “live by.” Indeed, one of
the strongest points of Spiro’s theoretical orientation is his consistent
attempt to ground theory in both behavioral and psychic reality.

This volume finally makes available a comprehensive collection of

Professor Spiro’s theoretical writings. Elaborations and variations on the
major theoretical positions laid out in the papers here assembled may also
be found in Spiro’s numerous and important books. Ifaluk: An Atoll Culture
(with E. G. Burrows) and Kibbutz, Venture in Utopia are early, primarily
ethnographic works which lay the foundations for Spiro’s subsequent
interpretations of these cultures. Children of the Kibbutz explores the role of
learning theories in making sense of observation and experience in the
kibbutz. In this, work Spiro analyzes the psychological dimensions of
socialization and the acquisition of utopian values and aspirations.
In Gender and Culture (1979) Spiro returns to Israel, to the kibbutz
materials and to his kibbutz experiences. He examines the relationship
between human nature, sexual identity and social structure, social roles and
family constellations in a radical attempt to engineer a utopian religious
community. The entire venture of the kibbutz, together with its vicissitudes
can, Spiro holds, be understood not only in terms of successful or
unsuccessful social programs, but also in terms of the limits all social
programs encounter in bending human nature to what might be construed
by the actors as social purposes. What can be modified is surface structure;
deep structure, rooted as it is in early childhood experience, is panhuman,
and therefore beyond the reach of even the most high-minded and ambitious
social reformers. The concluding chapter from Gender and Culture appears
in this book (as chapter 3), demonstrating that human nature will show itself
beneath the clothing of any social system.
Spiro’s contributions to studies of religion and to articulations between
personality systems, belief systems, and social systems are perhaps most
apparent in his work on Burma and particularly on Burmese Buddhism.
Spiro’s Burmese Supernaturalism represents his first book-length treatment
of a religious system; in this work he brings his concepts of cultural symbol
systems and personality drive-defense theories to bear on the subject of
Burmese folk religion.
Buddhism and Society (1970) develops the themes of religious needs
(drive/defense) and social organization, of personality and culture.
Specifically, he examines the meaning for the actors of discrepancies
between their actual beliefs and the normative (canonical) doctrines of
Buddhism.
Subsequently Spiro published yet another monograph, Kinship and
Marriage in Burma (1977), pursuing his investigations of the cultural and
psychodynamic dimensions of the Burmese world and applying functional,
historical, and psychodynamic explanations to Burmese customs relating to
kinship and marriage.
The papers that make up this volume, we hope, faithfully represent
Spiro’s most impressive corpus of scholarly anthropological inquiry and
achievement. The portrait which emerges depicts his exceptional dedication
to the problem of human suffering, to religious belief and scientific
understanding. Such dedication and such contributions are, we believe, of
particular value at a time when contemporary anthropology often appears to
be rudderless, drifting aimlessly into literary criticism, idiosyncratic self-
revelation, institutionalized routine and what Freud termed “the narcissism
of small differences.”
In sum, Spiro’s theoretical contributions in this volume must be
considered standard fare for all students of human nature, personality, and
culture. This body of work cogently calls into question basic—and too often
unquestioned—assumptions made by social scientists about the nature of
belief and of scientific evidence, about the relation between society,
personality, and culture. If anthropology fails to address problems of human
nature, of human motivations and their relationship to cultural institutions,
and of collective representations and individual action, the discipline of
anthropology runs the risk of becoming trivial and, consequently, obsolete.
Dealing with precisely these questions, the papers of Professor Spiro in this
volume set exemplary standards for all students of culture and the human
mind.
I
CULTURE AND HUMAN NATURE
1
Culture and Human Nature

ALTHOUGH I have been asked to describe the development of my research

findings and ideas, I cannot describe what I have done, or how I did it,
without explaining why I did it.1 Much of this paper, therefore, will be
concerned with intellectual motivations and research strategies rather than
with research operations or detailed research findings. Moreover, since I
believe that any valid explanation is ultimately historical (genetic,
evolutionary, developmental, etc.), a reliable account of why I did what I did
cannot begin with my anthropological research, but must rather be rooted in
my intellectual history. Hence, I shall first describe the intellectual (and
other) interests that brought me to the study of culture and personality, and I
shall then discuss one of the themes that, until relatively recently, has run
through a great deal of my research. Since I am concerned with explaining
the over-determined motivational structures that lie behind research choices
and the complex decision structures on which research strategies are based,
I do not have space to discuss other themes or to describe my more recent
interests.
1. I am grateful to Theodore Schwartz for his extremely helpful criticisms of an earlier draft of this
paper.

Beginnings
In an important sense, my intellectual interests have always been more
philosophical than scientific, and just as in the Middle Ages philosophy was
the handmaiden of theology, so for me anthropology has been the
handmaiden of philosophy, a tool for the empirical investigation of some
central issues concerning the nature of man. Although I have worked in four
different societies—the Ojibwa, Ifaluk, an Israeli kibbutz, and Burma—I
have never been interested in ethnographic description, per se; and although
I have published on a variety of institutions—family, kinship, politics,
socialization, and religion—I have had little interest in institutional
analysis, as such. Ultimately—so, at least, I have believed—these
enterprises are useful to the degree that they can illuminate some aspect of
the nature of man. Since, however, anthropology is primarily interested in
society and culture, and since until recently it has been much more
concerned with social and cultural differences than with universals, my
choice of anthropology—rather than, for example, psychology—might
seem rather strange. In the light, however, of the intellectual and political
Zeitgeist of the intellectually formative years of my life, this choice was not
so strange after all.
For liberal intellectuals, like myself, coming to maturity in the late
thirties and early forties, politics was an overriding concern. Existentialists
without knowing it, we had to come to grips with the twin traumata of our
time—the great depression at home, and the rise of Fascism abroad. Having
escaped the seductions of Soviet Communism, while yet deploring the
“poverty amidst plenty” which seemed to be characteristic of capitalism, we
perceived in democratic socialism the only viable alternative to the horrors
of both Fascist and Communist totalitarianism. As Marxists—and, in some
sense, we were all Marxists in those days—we believed that men were the
creatures of their social systems. If American society was characterized (as
we thought) by competitiveness, exploitation, and injustice, these
characteristics were not expressions of human nature, but of a particular
social system. Hence, to abolish these evils it was only necessary to change
the social system that produced them. In a social system, such as
democratic socialism, whose institutions were based on equality and justice,
they would disappear.
If Marxism was the ideological inspiration for these convictions, their
intellectual underpinnings, at least for me, were derived from my
philosophical studies. As a philosophy major, I had been persuaded—to be
sure, I was prepared to be persuaded—by Locke, Hume, Rousseau, and the
philosophers of the Enlightenment that man comes into the world as a
tabula rasa; that anything that is eventually inscribed on the blank slate is
put there by experience; and (though this was not shared by all these
thinkers) that the most important types of experience are those derived from
encounters with social institutions. Hence, the notion of a society in which
men are motivated by cooperation, altruism, and mutual aid was not viewed
as a utopian quest—nor was it viewed as a secular derivative of the
religious visions of Amos or Isaiah, those Hebrew prophets who had earlier
influenced me—but as a logical deduction from the social theories of the
eighteenth century philosophers of Reason. For if, indeed, man has no
“nature,” then his characteristics must be a product of history, as the latter is
distilled by and concretized in the social institutions of his society.
But the capstone—and “proof’!—of the thesis of the malleability of man,
and the omnipotence of social institutions, came from the writings of
Durkheim, for Durkheim (as I viewed him) was not just another speculative
philosopher; he was, rather, an empirical scientist. And if, as he had shown
in The Elementary Forms of the Religious Life, the very forms of thought
(space, time, causality, etc.) are ultimately derived from the structure of
society, what else could be said on the subject? Durkheim, though a
sociologist, had used anthropological data to sustain his arguments. This
convinced me that it was from anthropology that I could best derive future
intellectual nourishment, and when some years later I decided to pursue
graduate studies in the social sciences, I turned to anthropology.
Although I earlier discovered Durkheim was not in good favor in
anthropology—mostly because of Goldenweiser’s critique of his theory of
totemism—the regnant anthropological notions of that time were entirely
consistent with (indeed, they might just as well have been derived from)
Durkheim’s view of society and culture. Still, these notions—cultural
determinism and cultural relativism—provided a different conceptual, as
well as rhetorical, basis for my views concerning the relationship between
man and society. Since cultures vary across space and time, and since
behavior is culturally variable, the inference that man’s nature is similarly
culturally variable seemed irrefutable. The concept of culture not only
seemed to provide the definitive refutation of the notion of a universal
human nature, but it appeared to be a refined tool for understanding group
differences in behavior. Unlike the vague and rather metaphysical concept
of social or historical “forces,” the concept of culture could not only be
broken down into observable units of empirical investigation, but it
generated theories which seemed to account for group differences. (The
culture theories of the time were inadequate to account for cultural
invention, change, or deviance, but I did not perceive this as impugning
their validity.)
Having discovered anthropology, I lost whatever interest I might
otherwise have had in sociology or psychology. Focused on only one
cultural variant—that of Western man—the findings of sociology, so it
seemed to me, could shed little light on Man. Similarly, psychology, with its
concern with subcultural psychological processes (learning, perception,
cognition, and the like) as they could be studied in the laboratory, also
seemed unlikely to shed important light on Man. Its findings were based
either on the study of lower animals who, since they had no culture, seemed
inappropriate models for the study of humans; or, they were based on the
study of Western subjects, and since, as I believed, psychological processes
are culturally shaped, these findings seemed culture-bound. Anthropology,
on the other hand, seemed admirably suited to my interests. If man is the
creature of culture, then the proper study of man, so 1 believed, is the study
of culture; and since culture is variable, then the proper study of Man is the
study of culture in all of its variability. If, moreover, each culture (as Ruth
Benedict had argued in her seductive metaphor) has carved out a different
arc from the total circle of cultural variability, and if (as she also contended)
each primitive culture constitutes a natural experiment in cultural
variability, then it further followed that the proper study of cultural
variability entailed the study of primitive cultures. For me, then,
anthropology was clearly the science of choice. This being so, it is
important to sketch in greater detail the dominant anthropological notions
concerning culture and human nature of that time, notions to which I and
most of my contemporaries became, if we were not already, committed.
Since human behavior is culturally determined, and since cultures vary
enormously, the only valid generalization that can be made about human
nature is that it is enormously malleable (“plastic”). The existence of
cultural universals does not require any qualification of this conclusion
since culture, being “superorganic,” does not reflect (and, hence, cannot be
“reduced” to) noncultural, panhuman, biological or psychological attributes.
Although, so far as biology is concerned, such anatomical attributes as hair
form, eye color, and so on, are biologically determined (subject of course to
cultural selection), biology has little, if any, determination on social
behavior. Since the latter can be explained, without residue, by cultural
determinants, the organism, like the psyche, is conceived as either an
“empty” or a “black box.” Since, according to one “empty box” model,
culture affects behavior by a kind of Newtonian action-at-a-distance, the
box remains perpetually empty. According to a second model, however, the
empty box does not long remain empty for, as a consequence of
“enculturation,” the external culture is somehow incorporated by the social
actors. Nevertheless, since the resultant “inside” (psychological)
determinants of behavior are merely the “outside” (cultural) determinants
that have become “internalized,” and since, therefore, they vary as culture
varies, these psychological determinants can hardly constitute the basis for
a species-specific human nature. The “black box” differs from these two
“empty box” models in that it does not deny the existence of “inside”
variables; it merely denies their influence on social behavior. Since the
latter, so the argument goes, is culturally variable, how can it be explained
by biological or psychological determinants which are constant? Hence,
even if these “inside” variables may be said to constitute or to determine a
species-specific human nature, the latter is irrelevant to the understanding
of man’s social behavior.
Whatever the differences among these three models, it will be noted that
all agree (if only by implication) there can be no conflict between the
individual and culture. If there is nothing inside the individual (the first
model), or if what is inside is either orthogonal to, or represents the
internalization of, culture (the third and second models respectively), then
the individual and culture are in a state of harmony. That is, what the
individual wants to do is identical with what his culture requires him to do.
It should also be noted that although all three models explain behavior by
reference to culture, culture itself is left unexplained. To explain culture by
reference to some set of biological and psychological determinants
(construed as the core, or nucleus, of human nature) is not only inconsistent
with the empty and black box models, but it illustrates the fallacy of
“reductionism.” (For unexplained reasons, reductionism was taken to be a
self-evident fallacy.) Hence, like Aristotle’s First Cause, culture remains an
Unmoved Mover. To be sure, since culture is in large part a symbolic
system, man must everywhere have the capacity to invent, transmit, and
acquire cultural symbols, and to that extent symbolization is to be included
with plasticity as a second attribute of human nature. (Indeed, functionally
or adaptively viewed, symbolization is to human plasticity what instinct is
to animal specificity.) But since symbolization is culturally variable, the
universality of symbolic processes does not imply the universality of
symbolic meanings. Hence, so far as content is concerned, there is no
universal human nature; there are only culturally specific—and therefore
culturally variable—character structures.
Lest the above conceptions be used as a handle for racist arguments, the
concept of cultural relativism—according to which all cultures, and
therefore all culturally variable character structures, are equally valuable—
was enlisted to do battle against racist notions in general, and the notion of
primitive mentality, in particular. It should be noted, however, that cultural
relativism was also used, at least by some anthropologists, to perpetuate a
kind of inverted racism. That is, it was used as a powerful tool of cultural
criticism, with the consequent derogation of Western culture and of the
mentality which it produced. Espousing the philosophy of primitivism—
akin to, but not identical with, Rousseau’s notion of the Noble Savage—the
image of primitive man was used by some few anthropologists as a vehicle
for the pursuit of personal utopian quests, and/or as a fulcrum to express
personal discontent with Western man and Western society. The strategies
adopted took various forms, of which the following are fairly
representative: (1) attempts to abolish private property, or inequality, or
aggression in Western societies have a reasonably realistic chance of
success since such states of affairs may be found in many primitive
societies; (2) compared to at least some primitives, Western man is uniquely
competitive, warlike, intolerant of deviance, sexist, and so on; (3) paranoia
is not necessarily an illness, because paranoid thinking is institutionalized in
certain primitive societies; homosexuality is not deviant because
homosexuals are the cultural cynosures of some primitive societies;
monogamy is not viable because polygamy is the most frequent form of
marriage in primitive societies. Needless to say, the anthropologists of that
period were neither the first, nor—as recent politically motivated
resolutions advocated at meetings of the American Anthropological
Association indicate—the last to use anthropological “findings” as
“scientific” support for personal and political weltanschauungen.

Qualifications
Although many years were to pass before I finally discarded as untenable
many of the ideas described in the previous section, the corrosive process
began when I encountered the theories of A. I. Hallowell, my mentor at
Northwestern University, and (through him) psychoanalytic and learning
theory. From Hallowell’s writings and teachings I came to realize that
culture does not impinge directly on behavior, but is mediated through
personality processes relating to individuals. The contours of these
processes, I came to believe, were best delineated by Freud. It was the work
of Kardiner, however, that persuaded me of the importance of the family
and of socialization in the formation of these processes. From Kardiner, too,
I became convinced of the importance of “projective systems” for the
understanding of those aspects of culture that are not “reality” based. There
remained, however, a missing ingredient. If social actors monitor their own
behavior in accordance with cultural norms and rules, it was necessary to
explain the acquisition of culture in each generation of social actors. Here,
the social learning theorists—and especially Miller and Dollard—provided
the key.
Hence, when I completed graduate work, my theoretical framework
comprised an inchoate synthesis of cultural determinism, and cultural
relativism, neo-Freudian personality theory, and social learning theory. In
this loosely integrated synthesis, culture was viewed as a kind of master
plan for group adaptation; it was transmitted primarily in the family, and in
the early years of life, by traditional methods of socialization and
enculturation; it was acquired as part of the personality by techniques of
social reinforcement; its acquisition (“internalization”), however, often
conflicted with and frustrated other personality “needs”; the conflict was
resolved by the disguised satisfaction of these needs through culturally
mediated symbolic systems (“projective systems”).
It will be noticed that this synthesis, by including explicit attention to
“personality,” represented an important qualification of my earlier thinking.
But this shift from a cultural to a culture-personality framework did not
alter my earlier notions concerning human nature; on the contrary, it
strengthened them. For, although influenced by psychoanalytic thought, the
main thrust of the culture-personality school was precisely the reverse of
psychoanalysis. While the latter postulated invariant stages and processes
(and even invariant symbol formations and symbolic meanings) in the
formation, structure, and functioning of personality, culture-personality
(with some important exceptions) was primarily concerned to demonstrate
their cultural variability. Since, according to this school, personality was
determined by and constituted the internalization of culture, the range of
personality variability across groups could hardly be smaller than the range
of cultural variability. Indeed, since personality characteristics and
personality configurations, respectively, were viewed as isomorphic with
cultural characteristics and cultural patterns, the notion of a pancultural
human nature was viewed as highly unlikely. On the contrary, it was
assumed as almost self-evident that Zunis, Germans, Hawaiians, and Thais
are much more different, not only in culture but also in personality, than
they are similar.
These, then, were the views which I took with me on my first field trip—
to the atoll of Ifaluk in the Central Carolines.

First Field Trip—Ifaluk

In those far off days—it was 1947–48—very few anthropologists thought of
going into the field with a “problem,” let alone with a hypothesis for
“testing.” Rather, the typical fieldworker immersed himself in the local
society and culture and, with luck, he then came up with a problem that
might provide the basis for a doctoral dissertation. This, too, was my
approach, and the problem I came up with was aggression. The choice of
this problem was hardly accidental. In the first place, after some few
months on the atoll, I was struck by the fact that I had not observed a single
act of overt aggression. Reflecting on this remarkable fact, and reflecting on
my own adjustments to atoll living, it was hard to escape the conclusion
that the control of aggression was the central—or at least a central—
problem confronting any group of 250 people attempting to live together on
a land mass sixtenths of a mile square.
But, of course, the decision to study aggression was not that simple.
What about the kinship system? The Ifaluk had a matrilineal descent
system, based on matrilineages—a subject concerning which we knew very
little at that time. What about the political structure? Ifaluk, though a small
face-to-face society, had a complex system of hereditary chieftainship—a
combination which would have evoked the attention of scores of
anthropologists with backgrounds different from mine. What about the
complicated redistribution system? The products of the fields and the
oceans were periodically redistributed in Ifaluk by a mechanism which any
Oceanist worth his salt would have investigated in minute detail. Although
these, and other equally fascinating aspects of Ifaluk social structure
commanded and deserved as much attention as aggression, there must have
been other reasons for choosing the latter for study.
First, given my interest in human nature, it is understandable I would
have been much more interested in social processes than in social structure.
Second, among these processes, aggression had been a salient concern even
prior to my trip to Micronesia, beginning with my political interest in the
viability of socialism. This interest was supported by the culture-personality
literature of that period, which had persistently held up the Pueblos as
examples of nonaggressive, non-competitive peoples, who constituted
refutations of the Western notion that aggression and competition were
universal social characteristics. Indeed, in this literature, the Hopi and Zuni
were constantly contrasted with the Kwakiutl and Dobuans in support of the
cultural relativist thesis concerning aggression and cooperation, and the
Ifaluk seemed to offer yet additional support for this thesis. More
immediately, as a student of Hallowell, my interest in aggression had been
stimulated by three brilliant papers he had published on the psychocultural
determinants of, and solutions to, the problem of aggression among the
Berens River Saulteaux. These had left a deep impression on me. Finally,
under Hallowell’s guidance, and together with a group of graduate students,
I had conducted field work on the Lac Du Flambeau (Wisconsin) Ojibwa
reservation the summer prior to leaving for Ifaluk, and I had been struck
with (what I thought to be) a high incidence of aggression on that
reservation. It was found at the social level (in interpersonal and intergroup
behavior), at the cultural level (spontaneously told jokes and folk tales
almost invariably displayed an aggressive theme) and at the personality
level (as revealed in the Rorschach test). That the Ojibwa and Saulteaux are
the same people, and that, though separated in space and living under
different ecological and economic conditions, they yet exhibited similar
aggressive tendencies—which, in turn, were similar (as Hallowell had
shown in an ethnohistorical analysis) to those of their historical forbears—
was an exciting finding, which reinforced my interest in aggression.
 It is little wonder, then, that when observations in Ifaluk revealed almost
no manifest aggression, I was struck by the contrast with what I had
observed among the Ojibwa only six months earlier. The near absence of
aggression in Ifaluk seemed clearly to support my views concerning the
cultural relativism of human nature, and the cultural determinism of
personality. Since the Ifaluk had a cooperative social system, since their
ethos stressed the value of nonaggression, since there were few cultural
pressures for competition for scarce resources, and since in this
sociocultural context there was no observable aggression, it seemed to
follow—as the Pueblo data had already suggested—that aggression is a
function of historically specific cultural determinants, rather than an
attribute of “human nature.” This conclusion was also consistent with, and
lent support to, the leading psychological theory of aggression—the
frustration-aggression theory. For, when formulated in psychological terms,
the Ifaluk observations seemed to indicate that in the absence of
sociocultural frustration, there is no aggression.
After some additional work in Ifaluk, other observations, however, not
only began to intrude on my attention, but they eventually compelled a
change in my views about Ifaluk aggression. First, I had begun collecting
personality data—Rorschach and TAT protocols and dreams—which clearly
revealed the existence of hostile impulses at the personality level. In short,
from the absence of aggression—a behavioral variable—in Ifaluk, 1 had
wrongly inferred the absence of hostility—a psychological variable. (Here,
and in what follows, “hostility” refers to such motivational and affective
variables as anger, rage, hatred, and so on, while “aggression” refers to
observable social action—verbal and behavioral—whose intent is harm,
injury, damage, and so on.) Second, I had begun collecting data on cultural
systems—folk tales, myths, and religious beliefs—which revealed
important aggressive themes; and on the assumption that these systems
constituted projective expressions of personality dispositions, these cultural
data, too, revealed the existence of hostility in Ifaluk.
These findings not only challenged my conception of Ifaluk personality,
but they posed two immediate problems. First, in the absence of a
competitive culture, what explanation, other than recourse to the outmoded
notion of “instinct,” could be offered for Ifaluk hostility? Second, given the
existence of hostility, why was it not expressed in aggressive social
behavior?
 Actually, I had been aware of what later appeared to be a partial solution
to the first problem even prior to the empirical challenges to my original
observations. Shortly after arriving in Ifaluk I was (rather annoyingly)
awakened every morning by the cries of babies who, at dawn, are brought
by their mothers to be bathed in the chilly waters of the lagoon. Invariably,
the infants would react to this experience with cries of rage, and although I
could not say the experience was traumatic, there was no doubt that it was
painful for the infant. When queried, mothers said they knew that it was a
painful experience, but custom nevertheless required that infants be bathed
in that way and at that time. When, some months later, I was perplexed by
the presence of hostility in people living in a cooperative and nonfrustrating
culture, it occurred to me that it might be related, at least in part, to this
bathing experience. Since Ifaluk adults did not suffer important frustrations,
perhaps, so I reasoned, their hostility was not situational, but
characterological, their hostile impulses representing motivational
dispositions produced by those conditions that are formative of personality
dispositions in general, viz., infant and childhood experiences. Although it
seemed to me doubtful that one type of frustrating and painful experience—
even one that was repeated daily for a long period—could produce a
permanent character trait, it seemed reasonable that perhaps other types of
early frustrating experiences might also be discovered. Hence, changing the
focus of my study from adults to children, I began a three-pronged program
of investigation, including observations of parent-child interaction,
interviews of parents and of children concerning processes of socialization,
and interviews of children concerning their reactions to these socialization
processes and their consequent feelings toward parents and siblings.
From these investigations it was discovered that in addition to their early
bathing experience, Ifaluk children do indeed have other frustrating
(perhaps traumatic) experiences, and I became increasingly convinced that
adult hostility in Ifaluk was traceable to this configuration of early
childhood frustrations. Although, as is the case with all naturalistic studies,
the absence of controls did not permit this conviction to be converted into a
conclusion, its status as a highly likely hypothesis seemed warranted on
theoretical-deductive grounds. That is, on the assumption that hostility is
not “instinctive,” and on the further assumption that it is produced by
frustration, then, since there appeared to be few situational determinants of
hostility in Ifaluk adulthood, it seemed to follow that it was produced by the
frustrations of childhood.
Having discovered a tentatively satisfactory answer to the first question
raised by Ifaluk hostility, I still had to find an answer to the second: given
that the Ifaluk do indeed have hostile impulses, what accounts for their
nonaggressive behavior? Why isn’t hostility expressed in observable social
aggression? Two answers suggested themselves—one based on the Freud-
Kardiner theory of religion, the other on my observations of chieftainship.
Since Ifaluk religion postulates the existence of a class of spirits who are
purely evil—their sole aim is to cause human suffering—I reasoned that
much of the hostility of the Ifaluk is expressed (both displaced and
projected) in their hostile feelings to the alus, as these spirits are called.
Moreover, since rituals are periodically performed to drive away these evil
spirits, I inferred that much of their hostility is discharged in the
performance of these “aggressive” rituals. Since the characteristics of these
spirits are, on a number of dimensions, isomorphic with those of the parents
of their childhood, the expression of hostility in this form could be seen as a
symbolic expression of the hostility which, though repressed, was originally
aroused by the frustrating parents. In short, viewed as a projective system,
Ifaluk religion seemed to afford one avenue for the expression of hostility.2
2. Since that time, I have become convinced that there are sources of frustration—especially
related to esteem—in the adult social system, and that there are other expressions of hostility in
addition to projective expressions. Moreover, Theodore Schwartz (personal communication) has
called my attention to the aggression in the Ifaluk treatment of infants and children.
Although, ex hypothesi, religious beliefs and rituals channeled the
expression of hostility, permitting the discharge of hostile impulses through
projection and displacement, this hypothesis did not explain why the Ifaluk
complied with this culturally approved form of aggression, rather than
expressing their hostility in aggression against their fellows. Why, in short,
did they follow their cultural ethos, which prohibits any social aggression?
The ethos itself, so I assumed, is a highly adaptive cultural trait for a group
living in a tiny land mass (in which physical avoidance of others is
impossible), but, like any other functional explanation of culture derived
from a biological evolutionary model, it does not explain the conditions for
the persistence of the ethos, or for compliance with its dictates. The latter
condition must be explained not by some functional requirement—a “final”
cause—but by some condition in the immediate social field of the actors—
an “efficient” cause. This condition, it seemed to me, consisted in the
institution of hereditary chieftainship, especially the prescribed behavior
and persona of the chiefs, qua chiefs.
In Ifaluk, the chiefs are moral mentors. At periodic assemblies they
exhort the people to do “good,” and much of this exhortation is concerned
with admonitions to behave in accordance with the ethos of nonaggression.
In addition, they periodically monitor the behavior of their subjects by
regular inspections of their districts—there is one chief for each district—to
assure that this ethos is complied with. The chiefs, to use their own
expression, are the “fathers” of their people. Moreover, from observations
of the Ifaluk in interaction with the chiefs, and from interview and test
protocols, it seemed as if, reciprocally, the chiefs were, in the people’s eyes,
benevolent parental figures, whose approval was of vital importance for
their self-esteem and positive self-image. Desire for the approval of chiefs,
and fear of their disapproval, seemed to be the most important social
determinant of the Ifaluk adherence to the ethos of nonaggression.
With this, I felt that I had tentatively, and in large part, solved the
problem of aggression in Ifaluk, and in a manner entirely consistent with
my views of culture and human nature. However disappointing, especially
after my original impressions, to discover hostility in the Ifaluk, this
discovery neither supported the notion of a pancultural human nature, nor
did it challenge my own notion of historically specific cultural determinism.
Although the social system did not engender hostility, the socialization
system did, and this clearly supported my view that hostility is culturally
relative. For, if the Ifaluk socialization system engenders hostility, then, so
it seemed to me, other socialization systems could surely be found which do
not. Moreover, despite the presence of hostility at the personality level, the
Ifaluk case demonstrated that its expression, like its instigation, is culturally
relative. Instead of permitting hostility to be expressed in social relations,
Ifaluk culture directed its expression into other, less disruptive, channels.
Although for me, at least, the foregoing interpretation of the instigations
to, and vicissitudes of, aggression in Ifaluk seemed convincing, it was
obvious, as I have already indicated, that its various hypotheses remained
unproven. Without controls there can be no proof, and in naturalistic studies
of society and culture, there are really only two types of controls. One can
study different groups under the same conditions (this, in effect,
characterizes the comparative method in all of its variants), or the same
group under different conditions. The latter method (which, for reasons I
cannot develop here is the much more satisfactory) exploits historical
change to test interpretations previously offered for the status quo ante.
Thus, if my interpretations of Ifaluk aggression are valid, it follows that
were its religious beliefs and rituals to change, and were chieftanship to be
abolished (or otherwise lose its meaning), then—if there were no
functionally equivalent structural alternatives for these institutions—
hostility should be expressed in overt social aggression, or (if it is inhibited
by external sanctions) in predictable clinical symptoms. This prediction not
only provides a clear empirical test of the hypotheses discussed above, but
it is now possible to perform this test. In the quarter century that has elapsed
since my study of Ifaluk, there have been important changes both in its
religious and its political system. Christianity, introduced by the
missionaries, has replaced the traditional religion, and elective government,
introduced by the United States, has replaced traditional chieftainship. I
hope to return to Ifaluk to assess the psychological consequences of these
changes and, thereby, to test the hypothesis described here.

Second Field Trip—Israel

Although the Ifaluk findings were consistent with traditional culture-
personality views, they nevertheless left unanswered the basic question with
which I had been concerned—can culture (in the holistic sense) form the
psychological structure of human beings into any mold it chooses? Still, I
was sufficiently wedded to the traditional anthropological paradigm to
remain committed to the conventional culturalist answer to this question. In
1950, having recently completed my Ph.D. thesis on Ifaluk aggression, I
decided to explore this question in an Israeli kibbutz.
Although I still believed, on the grounds adduced by Benedict, that
anthropology ought to be concerned with primitive societies, the rationale
for studying this, a modern group, was precisely that which Benedict had
offered for the study of primitives—it represented yet another, and an
unexplored, arc of the total circle of cultural variation. Within this particular
arc, I was concerned with two problems especially. First, since children are
reared outside of the domestic family, to what extent do they exhibit the
same stages of psychosexual development which psychoanalysis had
postulated as universal, and more especially, to what extent do they develop
an Oedipus complex? Second, referring back to the Ifaluk study, to what
extent had the kibbutz succeeded in producing children without hostility? It
seemed particularly desirable to explore the second question in a kibbutz
since the latter is not only one of the few modern examples of a cooperative
group, but it is one that practiced a form of democratic socialism which had
been of such great interest to me in my early life. Here, I shall treat the
latter question only.
On the basis of the meager available literature, it was evident that the
kibbutz movement had ushered in a new type of human society. Rejecting
the traditional social structures of the West, this movement had initiated a
radical experiment in voluntary, comprehensive, cooperative living. Not
only were the means of production collectively owned, and not only was the
system of distribution based on the socialist principle of “from each
according to his abilities, to each according to his needs,” but even the
children were raised by (and for) the community, rather than in individual
families. Living in communal children’s houses, rather than with their
parents, they were reared by professional nurses and teachers, whose
primary goal was the transmission of the cooperative values of kibbutz
culture. If, then, there is no universal human nature, if the psychological
characteristics of social actors represent the internalization of the
historically specific cultural values of their group—if, in a word,
personality is the culture writ small—then, if the kibbutz values of sharing
and cooperation were in fact internalized by the children, it would be
expected that they would exhibit little if any competitive or hostile
characteristics. This at least was the premise which guided my kibbutz field
work.
The findings of the kibbutz study had a strong—and thus far a lasting—
impact on my image of man and on my conception of anthropology.
Although the kibbutz children were raised in a totally communal and
cooperative system; although their socialization had as its primary aim the
inculcation of a cooperative, noncompetitive ethic; although the techniques
of socialization were mild, loving, and permissive; although the target
responses were properly reinforced; although, in a word, almost all of the
culture conditions were designed to exclusively promote cooperation and
sharing, the data clearly indicated that kibbutz children, like other children,
do not wish to share scarce and valued goods—they want them for
themselves and they resist the attempts of adults to get them to share them.
They view as rivals those with whom they are obliged to share, and they
aggress against those who frustrate their desires to monopolize (or at least
to maximize) these scarce goods. Although they learn to cooperate, their
cooperative motives do not lead to the extinction of their learned
competitive and rivalrous motives. In short, although they learn to view
aggression as wrong, when they are frustrated they become angry, and their
anger—when not controlled—leads to overt aggressive behavior.
This does not mean, I hasten to point out, that the kibbutz has been
unsuccessful in transmitting its ethic of sharing, equality, mutual aid, and
cooperation to its children. It has, on the contrary, been surprisingly
successful, if “success” is defined as the perpetuation, by successive
generations of adults, of the social and cultural systems for which they were
socialized. But if “succss” means not only the internalization of the above
values, but the absence of any competing and conflicting tendencies—i.e.,
if “success” means the development of a “new man,” (as the kibbutz puts it)
—one without competitive, hostile, or acquisitive motives—then, of course,
the kibbutz has not been successful. But only a utopian or—what is the
same thing—a radical cultural relativist would have ever thought that the
creation of such a “new man” was possible. Indeed, from a nonutopian
point of view the real mark of kibbutz success is that although its children
have developed competitive and acquisitive, as well as cooperative and
sharing dispositions, when, as adults, they experience conflict between
them, they usually resolve the conflict in favor of the latter dispositions. In
short, the kibbutz values have penetrated to that part of the personality
which is the true measure of the internalization of cultural values—the
superego.
On the basis of the kibbutz study (and on the basis of everything I have
studied, read, and reflected upon since) I slowly and painfully came to the
conclusion that the belief that competition, rivalry, hostility, and so on, are
culturally relative rather than generically human is a misguided and false
notion which, invented by Rousseau, continues to be perpetrated by the
latter day believers—true believers—in the noble savage. This is so, I came
to believe, because although the intensity of competitive and hostile
motives is culturally (and individually) variable, and although culture can
tame and domesticate these motives, it is culture itself (interacting with
characteristics of man’s mammalian biology) which also, and universally,
creates them. This conclusion is based on a number of assumptions that will
be examined in the last section of the paper.

Third Field Trip—Burma

From my earlier exposure to Durkheim, I had acquired a persistent interest
in the study of religion, but it was the influence of Max Weber, whose
religious sociology I became acquainted with only after returning from
Israel, that led me to turn my research attention to religion. The Protestant
Ethic opened my mind to a point of view and method of analysis which
were revolutionary and enormously exciting. Moreover, Weber’s essays on
the sociology of religion, and especially his work on the religions of India,
stimulated an interest in Asia and Asian religions which has not yet run its
course. Again, however, this interest was very much related to my concern
with human nature. Although man’s basic motives, I had already decided,
are culturally universal, Weber’s work strongly suggested that world views
are culturally relative, and the religions of India, Hinduism and Buddhism
—but especially the latter—seemed to constitute convincing proof for this
thesis. If the adherents of one of the world’s great religions believe that the
self is an illusion, that there are no gods, that life is suffering, that suffering
can only be avoided by rejecting the world and all worldly desires, that
(therefore) the quest for immortality is a quest for eternal suffering, that
salvation consists in the cessation of life (i.e., of rebirth)—if, that is,
Buddhists believe in these, and in many other, principles that are directly
opposed to accepted Western principles, it seemed to follow that world
views—those fundamental cognitive orientations by which men order their
lives—are historically conditioned and therefore culturally relative.
In order to study both the determinants and the consequences of the
Buddhist world view “on the ground,” I determined that my next field trip
would take place in Asia, and from 1961 to 1962 I was fortunate to have the
opportunity to conduct such a study in Upper Burma. Subsequently, in the
summers of 1969 to 1972 I continued these studies among the Burmese
expatriates in Thailand. As an anthropologist, especially one committed to
functionalism, I did not, of course, restrict my studies to Buddhism. But I
shall avoid any discussion here of the findings concerning kinship and
politics, of folk religion and folk medicine, not to mention aggression,
though the latter is directly related to the previous sections of this paper.
Instead, this discussion will be confined to elements of the Burmese
Buddhist world view in their relationship to human nature.
As was the case in the earlier field trips, the expectations I took to Burma
proved to be chimerical, for I discovered that the religious beliefs of
Burmese Buddhists—and subsequent studies of other scholars revealed this
was also the case in Ceylon and Thailand—were in many respects rather
discrepant from those described in Western works on Buddhism, not
excluding that of Weber. The discrepancies were based on two errors. One
error was mine, in expecting congruence between canonical texts and
beliefs held by religious actors, for, of course, it is never the case that the
belief systems of religious actors accurately mirror their canonical texts.
The other error was that of Western interpreters of Buddhism, who, by their
often inaccurate renditions of canonical beliefs, had presented a distorted
conception of Buddhism. One example of such a distortion is the alleged
atheism of this religion. Although in the metaphysical or theological sense
of denying the existence of a Creator or Redeemer, canonical Buddhism is
indeed atheistic, it is nevertheless far from being the kind of Ethical Culture
movement that it is often portrayed to be. On the contrary, Buddhism
explicitly affirms the existence of a host of superhuman beings—gods,
godlings, and spirits—and it prescribes various types of ritual for enlisting
the help of the benevolent, and appeasing the wrath of the malevolent, ones.
It is metaphysically important that these superhuman beings are not eternal,
that they, like humans, are subject to the law of karma and, hence, to the
wheel of rebirth, but it is irrelevant to the religious concerns of the Buddhist
actor who, like religious actors everywhere, turns to these beings for help in
time of need, and invokes their activity to explain his suffering. For him,
life is inconceivable without the help of the benevolent superhuman beings,
and most of life’s vicissitudes are inexplicable without his belief in the
malevolent.
To be sure, the superhuman helpers of Buddhism are of no assistance—as
they are in some other religions—in the Buddhist’s quest for salvation, for
their help is confined to this world. Even the Buddha, Himself, though He
has shown the Path to salvation, is not—as is Christ, for example—a Savior.
To be saved every man must walk alone on that Path, without external
guidance or intercession. Man, as it were, must save himself. To this extent,
Western interpretations of Buddhism are correct. What they usually fail to
observe, however, is that this Path is not one on which many Buddhists
desire to walk because its soteriological goal, nirvana, is not one which they
usually aspire to attain. The reason is simple. Nirvana, a knotty concept at
best, is not a state of being—not, at least, in the ordinary sense of “being.”
Indeed, in the latter sense, nirvana is best characterized as a state of
nonbeing, signalling the end of the wheel of rebirth in all of the planes of
existence postulated by Buddhist cosmology. But Buddhists, for the most
part, are no more interested in the extinction of their future rebirths than are
Westerners in the extinction of their present birth. Although paying lip
service to the notion of nirvana, most Buddhists, in fact, conceive of
salvation in highly concrete and material terms, much as it is conceived in
most other religious systems. That is, salvation for them is a state of being
in which wants and desires are fulfilled, and in which suffering is
extinguished. Such a state of affairs they project into future rebirths, in
which they are reborn either as wealthy human beings, or—if they dare to
aspire to it—as gods in one of the Buddhist material heavens. (They explain
that they do not aspire to the normative goal of nirvana by the fact that their
present spiritual attainments are still underdeveloped. They hope, however,
that after many future rebirths they will eventually develop higher spiritual
qualities by means of which they will be able to aspire to this superior
vision of salvation. In this regard they are like the young St. Augustine,
who prayed he might achieve celibacy, “but not yet.”)
But in addition to these findings, there were still others that also
disabused me of my earlier view that basic human cognitive orientations are
entirely culturally relative and wholly historically conditioned. Thus, for
example, the all too human hope which the Burmese evince for a
continuous existence is expressed not only in their desire for continuous
rebirths, but also in the prevalence of alchemic beliefs and practices whose
aim—like the alchemic aims of China—is not the transmutation of base into
precious metals, but the attainment of “immortality.” To be sure, this goal is
not defined as immortality because, as Buddhists, the Burmese profess
belief in the normative doctrine of the impermanence of sentient existence.
Hence, this goal is defined as—and the alchemic beliefs hold up the
possibility of—a continuous existence of only sixty thousand years
duration. One could, of course, debate the issue (though probably without
much profit) of whether, for the average human mind, sixty thousand years
is or is not tantamount to immortality.
Now in pointing, by implication, to the similarities between certain
dimensions of the Burmese world view and those of traditional Western
culture, I am not suggesting the two are similar on all dimensions. Since,
even within the same social group, there are individual differences in
cognitive orientations, it is a fortiori the case that there are group
differences in cognitive orientations, as well, and the latter are obviously a
function of cultural differences. I am suggesting, however, there are some
basic cognitive orientations that are not culturally variable (though they
may be culturally determined), and among these must be counted those
mentioned here in this brief description of some of the cognitive
orientations found in Buddhist societies, viz.: the desirability of life, and
hence the desire for its prolonged duration; the desirability of material and
physical pleasures, and hence the desire to maximize these pleasures; belief
in superhuman beings, both of good and evil, and the corollary belief that
human action (religious ritual, etc.) can influence their activity by invoking
the assistance of the former and repelling the harm of the latter. It is
especially instructive that Buddhist actors share these panhuman
orientations because in their case these orientations (except for the third) are
in direct opposition to the normative teachings of the religion which they
genuinely—indeed passionately—revere.
In sum, although the differences in the normative values and official
codes of different cultures may, in some instances, lead to the conclusion
that culturally constituted cognitions are so different as to constitute (as
many contemporary ethnoscientists believe) incommensurable world views,
on the basis of the Burmese study I have come to believe this conclusion is
entirely unwarranted. I now believe, on the contrary, that when the values
and codes of social actors are studied directly, rather than inferred from
normative cultural concepts, many cultural differences in basic cognitive
orientations are seen to be more apparent than real; to a large extent these
differences are merely the manifest expressions of the same underlying
cognitive orientations which, being historically conditioned, take different
cultural forms.
In short, just as the kibbutz and Ifaluk studies convinced me that many
motivational dispositions are culturally invariant, the Burmese study
convinced me that many cognitive orientations are also invariant. These
invariant dispositions and orientations stem, I believe, from panhuman
biological and cultural constants, and they comprise that universal human
nature which, together with received anthropological opinion, I had
formerly rejected as yet another ethnocentric bias.

Conclusions
Having described how I gradually and painfully came to discard much of
the received anthropological wisdom, I now wish to explore (what I
consider to have been) the fallacies in traditional anthropological thinking
that led to the adoption of cultural determinism and, hence, to the cultural
relativistic view of human nature.
I have already noted that in the traditional conception, “culture” referred
to all aspects of a group’s environment, except the physical, and to all
aspects of man, except the biological. Given, then, that man constitutes a
single biological species and that, nevertheless, social behavior reveals a
wide range of cross-cultural variability, it seemed to follow that the
organism is an empty or a black box, and that social behavior, therefore, is
determined by culture. Given, too, that personality characteristics were
believed to exist in a one-to-one relationship with behavioral
characteristics, the former being thought to be directly deducible from the
latter, it seemed to follow that personality is also culturally determined. This
meant that no psychological characteristic—no affect, no need, no wish, no
belief—could be part of human personality unless culture put it there.
Given, moreover, that culture is variable, personality (it was further argued)
must similarly be variable; historically specific cultures produce culturally
variable personality characteristics. Given, finally, that depending on their
culture, human beings may acquire any empirical subset of the total
conceivable set of human psychological characteristics, any member of the
latter set (it was concluded) is culturally relative; none is an invariant
characteristic of a universal human nature. Any psychological characteristic
—the feeling of love or hate, the wish for mortality or immortality, the
belief in the talion principle or its reverse—might or might not be found in
any social group as a function of its cultural program.
 Skipping over some logical and empirical problems in its global3 and
holistic4 conception of culture, this conceptual structure falters, I believe,
on two related, but separable, theses, both of which are (I believe)
untenable. These are: (1) culture is the exclusive determinant of personality;
and (2) personality consists exclusively of the internalization of culture. Let
us begin with the first.
3. The logical problem raised by the global conception of culture is easily stated. If culture
includes (among other things) behavioral patterns and personality traits, to say that culture
determines behavior and personality is obviously circular. Although this difficulty is easily resolved
—by excluding behavior and personality from the definition of “culture” and thereby rendering the
culture concept less global—yet another problem with respect to personality remains. For if
personality is not part of culture, and, moreover, if personality (as we shall see below) is not
isomorphic with behavior—if, on the contrary, personality is viewed as a system of cognitive,
perceptual, motivational, and affective dispositions “underlying” behavior—it cannot be so blithely
assumed that cultural (and, therefore, behavioral) variability is associated with personality variability.
This latter thesis now becomes an empirical hypothesis, to be tested by direct examination of
personality, rather than an a priori assumption whose proof consists in pointing to the cross-cultural
variability in behavior. In short, if personality is not, merely by definition, a part of culture, and if the
isomorphism of personality and behavior is not accepted as an unchallenged assumption, then the
question of pancultural psychological characteristics remains open.
4. The empirical problem raised by the holistic conception of culture is also easily stated for, given
this conception, it is difficult to know which of the various elements comprising a culture are the
determinants of personality: Art styles? Religious beliefs? Descent systems? Modes of production?
Child rearing? Some of these? All of these? In short, so long as culture is taken as an undifferentiated
whole, a seamless web, the thesis of cultural determination of personality is not very illuminating,
since (even in the now restricted conception of culture) the only things that are excluded as possible
determinants are characteristics of the organism and of the physical environment. Even Margaret
Mead, much of whose work has concentrated on (and has illuminated so much of) the relationship
between child training and personality, has disclaimed any notion that the former is an especially
important cultural determinant; rather, she claims, it is merely a convenient way of entering into the
total configuration of cultural determinants.

Culture is the exclusive determinant of personality. Culture determinist

theories of behavior and of personality were developed in the first instance
as alternatives to and refutations of biological determinism. Given the
demonstrable cross-cultural variability in behavior and personality,
anthropology (validly) argued that social behavior cannot represent an
expression of instincts, and personality formation cannot represent an
unfolding of genetically programmed psychological traits. Rather, both
must be (in a large part, at least) a result of learning, which is to say (since
anthropology is concerned with group, rather than individual, variability)
that they are the products (in large part, at least) of social and cultural
determinants. So far, so good. To have concluded, however, that cultural
determinism implies the absence of any invariant pancultural psychological
characteristics—in short, that all such characteristics are culturally relative
—is an invalid conclusion which is based, I believe, on three interrelated
fallacies.
The fallacy of assuming that the cross-cultural variability in social
behavior and personality implies that the organism is an empty or black
box. Despite the variability in culture, it does not follow that man has no
invariant psychological characteristics, because some, at least, are
biologically determined. Far from being an empty or black box, some of the
properties of the organism are important determinants of social behavior as
well as personality. Although space does not permit a detailed, let alone an
exhaustive catalogue of these properties, I might mention a small subset,
viz., those relevant to the problem of hostility and aggression discussed in
earlier sections of this paper.
Because of prolonged helplessness, requiring dependency on others for
the satisfaction of their survival needs, children are everywhere raised in
family or family-like groups whose members, to a greater or lesser extent,
provide them with nurturance, gratification, and protection. As a result,
children everywhere have the following characteristics: the need to receive
love from, and the motivation to express love for, the loving and loved
objects; feelings of rivalry toward those who seek love from the same
(scarce) love objects; hostility toward those who would deprive them of
these objects; and so on. In short, everywhere (due to Oedipal struggles and
conflicts with siblings) children’s need for love is necessarily thwarted, as
well as gratified, to some extent. Moreover, because everywhere man lives
in social groups—yet another biological requirement of the human
organism—and since the requirements of a human social order demand that
children learn to behave in compliance with cultural norms, everywhere
their needs, desires, and wishes are necessarily frustrated, as well as
gratified, to some extent. Everywhere, therefore, if for only these reasons,
hostile, rivalrous, and competitive feelings, as well as those of love and
mutual aid, will be found to some extent.
The expression “to some extent” underscores the fact that the intensity of
these various types of frustrations, and hence the intensity of these feelings,
are highly culturally variable. Similarly, and for the same reason, the targets
of these feelings, the degree to which they are permitted expression in
aggressive behavior, the social domains in which they are permitted
expression (whether in kinship, religion, politics. and so on), the manner in
which they are expressed (whether directly, through displacement,
projection, and so on)—all these are culturally determined and culturally
variable. Nevertheless, although different cultures channel these feelings in
a bewildering variety of social and cultural forms—some more, some less
adaptive—all cultures produce them to some extent. They are, in short,
among man’s invariant psychological characteristics.
The fallacy of confusing “a culture” tvith “culture” in the expression
“cultural determinism.” The theory of cultural determinism was based on
the following (valid) argument. Given that human beings are born without
instincts, the gratification of human “needs” depends on learning. Given,
moreover, that they are bom entirely helpless, they are wholly dependent on
adults for the acquisition of the means for their gratification. Given, finally,
that they live in social groups, these means must be shared and, therefore,
prescribed, i.e., they must be cultural. Thus, the properties of the organism,
interacting with those of the social environment, require that a human
existence be a culturally constituted existence. In short, if other animals
adapt by means of species-specific biological specializations, human
adaptation is achieved by means of a species-specific nonbiological
specialization, viz., culture. If this is the case, then, for a human primate to
be classified as man—i.e., to be characterized as more than a bipedal, big-
brained, primate—it is not enough that he have those biological
characteristics a zoologist would designate as the distinguishing features of
Homo sapiens. It is also necessary that he have those characteristics—
socially shared and transmitted symbols, values, rules, and so on—that an
anthropologist would designate as the distinguishing features of a cultural
mode of adaptation. But if culture is as important an adaptive human
requirement as food or water, then culture and the psychological products of
culture—the drives, needs, cognitions, and the like, that are produced by
culture—are as much a part of man’s nature as his biological characteristics
and their psychological products. Indeed, since many of man’s biological
and biologically derived psychological characteristics are found in the
entire class of mammals, it may be said that since his cultural and culturally
derived psychological characteristics are man’s species-specific
characteristics, they are the uniquely human part of his nature.
These, then, were the insights that were originally captured by the notion
of cultural determinism. They were valid then and, in my opinion, they are
valid now. When, however, anthropology became increasingly impressed
with cultural differences, this original notion of cultural determinism
underwent (as we have seen) two changes. First, instead of being taken as a
determinant, culture was taken as the determinant, of human nature, for
when anthropologists began to view culture as internalized by social actors,
they also began to argue that culture was the exclusive (or the exclusively
relevant) content of the black or empty organism. Second, since culture is
found in a bewildering variety of local manifestations, anthropologists came
to view each culture as a more or less historically unique creation, each
producing a culturally unique human nature. Consequently, it came to be
believed that there is little basis for postulating any set of invariant
psychological characteristics that might comprise a universal human nature.
Almost without recognition, therefore, the meaning of “cultural
determinism” has undergone a gradual shift. Beginning as a statement about
man’s nature, it has come to be a statement about his history. From a theory
of the generic cultural determination of single pan-cultural human nature, it
has become a theory of the historically specific cultural determination of
many culturally relative human natures.5 The fallacy in this shift is obvious.
Since cultural variability does not mean that culture is distributed across
space and over time in a series of discrete configurations, each
incommensurable with every other; since, on the contrary, culture is known
to be distributed in a series of overlapping configurations; since, in short,
despite the variability of culture, there is clearly discernible (what Wissler
termed long ago) a “universal culture pattern,” then, if personality is
determined by culture, there must also be a universal personality pattern—a
set of invariant psychological characteristics produced by all cultures. In
sum, even an exclusively cultural determinist theory of personality does not
entail a cultural relativistic theory of human nature. For if certain cultural
characteristics are pancultural, then, ex hypothesi, certain personality
characteristics are also pancultural, and they, at least, comprise man’s
pancultural human nature.
5. Certain trends in current anthropology are even more relativistic. Thus, for example, the
movement alternatively referred to as the “new ethnography,” ethnosemantics, or ethnoscience,
rejects the very notion of objective categories of ethnographic description. Arguing that cultures are
incommensurable, the members of this movement insist that “etic” (i.e., cross-cultural) categories
distort, when they do not falsify, the meanings of intracultural concepts, and that “emic” categories
alone can convey their meanings. In my view this is not only the reductio ad absurdum of cultural
relativism, but it leads to the demise of anthropology as a science.
 The fallacy of not distinguishing the phenotypic from the genotypic, or
(using a more fashionable metaphor) surface structure from deep structure,
in culture. Culture, as man’s most important adaptive specialization,
mediates the interaction between the properties of his psychobiological
organism and those of his social and physical environments. But in its role
as mediator, culture is necessarily variable across space and over time
because, as a product of man’s symbolic capacities, culture can (and must)
vary with a host of variable historical conditions—ecological settings,
diffusionary opportunities, politically powerful or charismatic leadership,
unpredictable physical and social events (war, drought, invasions), and so
on. If, then, culture is the means by which men and groups adapt to the
functional requirements of individual and group existence, it is not
surprising to find a wide range of differences in the form and content of
culture as a function of an equally wide range of differences in man’s
historical experience.
Variability in the form and content of culture does not, however, imply
variability in the substance of culture, nor, a fortiori, does it entail
variability in personality. Indeed, identical psychological structures can be
associated with (and eventuate in) manifestly dissimilar cultural structures,
for though phenotypically different, the latter may be genotypically
identical, i.e., they may be functionally equivalent structural alternatives for
coping with the identical functional requirements of individual and group
existence. Different religions, for example, exhibit wide variability in their
belief and ritual systems, and yet these different systems may all satisfy
(among other things) a common human wish for dependency on powers
greater than man. Thus, the gods of different cultures may differ in form
and content (cultural phenotype), but, as powers greater than man (cultural
genotype), they all reflect a pancultural human psychological characteristic
(dependency need).
If, then, differences in cultural phenotypes do not imply differences in
genotype, culture is less variable than it seems. Hence, phenotypic cultural
differences do not in themselves justify the inference of personality
differences, and phenotypic cultural variability does not in itself imply the
cultural relativity of human nature. Unless these phenotypic differences can
be assumed to be more than historically conditioned variable expressions of
the same cultural genotype, such differences are entirely compatible with a
pancultural human nature.
 The foregoing critique of cultural determinism, and its corollary, cultural
relativism, may be summarized as follows: the nature/history dichotomy is
a false dichotomy; although personality, to some extent, is culturally
relative, man has a nature as well as a history; this is so because even a
radical cultural determinism does not imply a radical cultural relativism;
however much societies may differ, they all must cope with man’s common
biological features, especially his prolonged infantile dependency; the
adaptively viable means for coping with the latter condition exhibit
common social and cultural features across a narrow range of social and
cultural variability; these common biological, social, and cultural features
are a set of constants which, in their interaction, produce a universal human
nature.
2. Personality consists exclusively of the internalization of culture.
Although the thesis that culture is the exclusive determinant of personality
accounts, in part, for the traditional denial of a pancultural human nature,
the thesis that personality is the internalization of culture has been even
more decisive in this regard. For since culture is variable, and since, ex
hypothesi, personality is merely culture as it is internalized in social actors,
it follows necessarily that there can be no pan-cultural human nature; there
can only be culturally relative human nature. But this undifferentiated
model of personality (personality as constituted exclusively of culture) is
even less tenable than the global model of culture held by its champions. If,
instead, we were to adopt a more sophisticated model of personality—one
in which personality, conceived as a system, consists of differentiated
structures, each with distinctive functions—cultural variability need not
imply personality variability (and, for that matter, cultural similarities need
not imply similarities in personality).
Consider, for example, the Freudian model according to which
personality, as a system, consists of three differentiated, but interrelated
structures. Over-simply put, these structures consist of an impulse system,
or id; a cognitive-perceptual system, or ego; and a normative-prescriptive
system, or superego. This model is not only much more complex than the
internalization-of-culture model, but one of its postulated structures—the id
—comprises wishes and desires, many of which are in frequent, if not
persistent, conflict with culture. Moreover, since the id is only one structure
of the personality, and since another of its structures—the superego—
comprises (among other things) internalized cultural values, many wishes
of social actors are not only in conflict with the cultural requirements of
their group (external conflict), but one part of their personality is frequently
in conflict with another (internal conflict). In the latter case, the ego
experiences conflict between impulses which seek gratification and
internalized cultural values which proscribe their gratification. To
complicate the picture even more, this conflict may be unconscious, as well
as conscious.
According to this model of personality, social behavior is often neither a
direct expression of an undifferentiated personality, nor a simple result of
the influence of external cultural norms. It is more likely to be the end
product of a chain of interacting psychological events, including impulse
(id), cultural and personal values (superego), conflict between them, and
defense against conflict (ego), which only then eventuates in behavior. In
most cases, to be sure, behavior—the end product of this chain—conforms
to cultural norms, for the actor usually complies with his superego, and
resolves the conflict by controlling the forbidden impulses—either by the
conscious mechanism of suppression, or by a variety of unconscious
mechanisms of defense (including the defense of repression). In short,
according to this model of personality, social actors are not merely the
creatures of their culture, formed in its mold, and reflecting its values.
Although typically conforming with them, a comprehensive description of
their values is only a partial description of their personality. Indeed, their
values, which comprise one part of their motivational system, are often in
conflict with equally strong motives which oppose them. The latter may be
kept under control, but since control does not mean extinction, culture, as
Freud argued, necessarily produces “discontents”—a condition in which the
social actor is frequently at odds with himself and with his culture.
Since both models of personality, the Freudian structural model and the
internalization-of-culture model, make the same behavioral predictions—
both predict behavior will conform with cultural norms—what difference
does it make regarding which of the two one chooses? (Indeed, the principle
of parsimony would suggest the adoption of the latter, or cultural
determinist model.) Although this choice makes little difference for the
prediction of social behavior, it makes a great difference for the
understanding of personality (not to mention the understanding of the
psychic costs of culture, mental illness, social deviance, and so on).
According to the internalization-of-culture model, behavior is the direct
(unconflicted) expression of cultural norms as they are internalized in the
personality. Hence, for example, if one observes little aggressive behavior
in some social group, it could be inferred that, having internalized a set of
cultural values opposing aggression, the social actors are nonhostile—for
hostility, according to this model, is determined by cultural values which
favor aggression. According to the Freudian model, however, this inference
may be entirely unwarranted, for these nonaggressive actors may (as a
result of any number of punitive and frustrating conditions) have strong
hostile impulses which, however, because of strong superego disapproval of
aggression, are kept under rigid control.
Just as the internalization-of-culture model may invalidly deduce
personality from behavior, it may similarly invalidly deduce culture from
behavior. Thus, if one were to observe a great deal of aggression in a social
group, then, in accordance with this model, it could be inferred that the
actors have internalized a set of cultural values which favor aggression.
According to the Freudian model, however, this may not be the case at all.
Rather than expressing cultural values, aggressive behavior may in fact be
in direct violation of them. The aggression (stemming from socially induced
hostility) may instead be the result of an immature ego (with little impulse
control), or a weak superego (which has insufficiently internalized the
cultural values opposing aggression), or of a powerful id (with strong
hostile impulses).
In sum, when personality is viewed as a system with differentiated
structures and functions, the simple isomorphism between culture, behavior,
and personality postulated by the internalization-of-culture model is
frequently found wanting. As we have seen, a group which exhibits very
little aggressive behavior may, nevertheless, be characterized by strong
(probably unconscious) hostility; and a group which exhibits a great deal of
aggressive behavior may nevertheless be characterized by cultural values
which oppose aggression. (In the latter case, of course, the aggression will
probably be expressed—displaced, projected, and the like—in socially
acceptable forms). To take a classic case, the Hopi may be no less hostile
than the Sioux, despite the fact that the latter exhibit much more social
aggression, and that their cultural values concerning aggression are much
different. Rather than reflecting differences in hostility, resulting from
differences in culture, the differences in their aggressive behavior may
instead reflect differences either in the social canalization of aggression, or
in the relationships among the id, ego, and superego components of their
respective personalities. Indeed, their strict avoidance of interpersonal
aggression might suggest the hypothesis that Hopi hostility may be even
stronger than that of the Sioux, but that, because of a stricter superego, they
inhibit their dangerous hostile impulses.
This example suggests that although group differences in behavior say a
great deal about differences in culture, in themselves they may say little
about differences in personality, and even less about the cultural relativity
of human nature. It also suggests that different models of personality do
indeed make a difference for our understanding of human nature. The
internalization-of-culture model, which assumes that personality differences
are isomorphic with behavioral and cultural differences, leads to the cultural
relativis tic view of human nature. The model of a structurally and
functionally differentiated personality, which makes no assumptions about
social-eulture-personality isomorphisms—and which therefore requires that
personality be investigated independently of both behavior and culture—
supports the view of a pancultural human nature.

Note
Reprinted from The Making of Psychological Anthropology, edited by George D. Spindler
(Berkeley: University of California Press, 1978), pp. 331–60.

References

The following entries pertain exclusively to the author’s three research

projects discussed in this chapter.

Ifaluk
1950. A psychotic personality in the South Seas. Psychiatry
13:189–204.
1951. Some Ifaluk myths and folk tales. Journal of American
Folklore 64:280–303.
 1952. Ghosts, Ifaluk and teleological functionalism. American
Anthropologist 54: 497–503.
1953. An atoll culture, with E. G. Burrows. New Haven: Human
Relations Area Files.
1953. Ghosts: An anthropological inquiry into learning and
perception. Journal of Abnormal and Social Psychology 48:376–
82.
1959. Cultural heritage, personal tensions, and mental illness in a
South Sea culture. In Culture and mental health, ed. M. K. Opler.
New York: Macmillan.
1961. Sorcery, evil spirits, and functional analysis. American
Anthropologist 63:820–24.
Kibbutz
1954. Is the family universal? American Anthropologist 56:839–
46.
1955. Education in a communal village in Israel. American
Journal of Orthopsychiatry 25:283–92.
1956. Kibbutz: Venture in Utopia. Cambridge: Harvard University
Press. (New augmented edition, 1971, with a new chapter.)
1957. The Sabras and Zionism: A study in personality and
ideology. Social Problems 5:100–110.
1958. Children of the kibbutz. Cambridge: Harvard University
Press. (New augmented edition, 1975, with a new chapter.)
Burma
1965. Religious systems as culturally constituted defense
mechanisms. In Context and meaning in cultural anthropology,
ed. Melford Spiro. Glen-coe: Free Press.
1966. Buddhism and economic saving in Burma. American
Anthropologist 68: 1163–73.
1967. Burmese supernaturalism: A study in the explanation and
resolution of suffering. Englewood Cliffs, N.J.: Prentice-Hall.
1968. Religion, personality, and behavior in Burma. American
Anthropologist 70:359–63.
1968. Politics and factionalism in Upper Burma. In Local level
politics, ed. Marc Swartz. Chicago: Aldine.
1969. Religious symbols and social behavior. Proceedings of the
American Philosophical Society 113:341–50.
1969. The psychological functions of witchcraft: The Burmese
case. In Mental health in Asia and the Pacific, ed. Caudill and
Lin. Honolulu: East-West Center Press.
1971. Buddhism and society: A great tradition and its Burmese
vicissitudes. New York: Harper and Row.
1972. Violence in Burmese history: A psychocultural
interpretation. In Collective violence, ed. Short and Wolfgang.
Chicago: Aldine.
1973. Social change and functional analysis: A study in Burmese
psycho-cultural history. Ethos 1:263–97.
1974. The Oedipus complex in Burma. Journal of Nervous and
Mental Disease 157:389–95.
1975. Some psychodynamic determinants of household
composition in village Burma. Contributions to Asian Studies
8:126–38.
1977. Kinship and marriage in Burma: A cultural and
psychodynamic analysis. Berkeley: University of California
Press.
1979. Symbolism and functionalism in the anthropological study
of religion. Proceedings of the Study Conference of the
International Association for the History of Religion. In Science
of religious studies in methodology, ed. Lauri Honko. The Hague:
Mouton.
2
Some Reflections on Cultural
Determinism and Relativism with
Special Reference to Emotion and
Reason

A Conception of Culture
AS I SEE IT, “culture” designates a cognitive system, that is, a set of
“propositions,” both descriptive (e.g., “the planet earth sits on the back of a
turtle”) and normative (e.g., “it is wrong to kill”), about nature, man, and
society that are embedded in interlocking higher-order networks and
configurations. Cultural and noncultural propositions differ in two
important dimensions. First, cultural propositions are traditional, that is,
they are developed in the historical experience of social groups, and as a
social heritage, they are acquired by social actors through various processes
of social transmission (enculturation) rather than constructed by them from
their private experience. Second, cultural propositions are encoded in
collective, rather than private, signs (indices and icons, to employ Peirce’s
distinctions, as well as symbols). Hence, they exist and (in the first
instance) are discoverable by anthropologists in the collective
representations of social groups without their having to probe for them in
the private representations of social actors (though, as I shall soon argue,
many of them are also found there, albeit in a different form). This is not to
say that cultural statements, rules, values, norms, and the like are always
stated in propositional form, for clearly they are not, but that they are
susceptible of statement in that form.
This restricted conception of culture has important implications for what
“culture” does not designate. First, although by this conception, culture is
obviously an important—though only one—determinant of behavior,
culture as such does not consist of behavior. Moreover, although culture—
to broaden this implication—includes propositions referring to social
structure, social organization, social behavior, and the like, culture as such
does not consist of them. It might be added that because the latter have
noncultural (situational, ecological, economic, political, biological,
emotional, etc.) as well as cultural (ideational) determinants, the cultural
propositions comprising the emic models of social structure, social
organization, and the like cannot in themselves provide a reliable basis for
predicting their content or their shape. For the same reason, cultural models
cannot be reliably deduced from the observation of social behavior alone.
Second, although by this conception “culture” designates a cognitive
system, it is not the only—though it is clearly the most important—source
of the cognitions and schemata held by social actors. The other source, of
course, consists of their own experience. Thus, on the basis of their social
experience, a group of actors may come to construct a conception of their
social universe as hostile and threatening even though such a conception is
not transmitted to them by intentional enculturative processes and may even
be inconsistent with cultural propositions that convey the very opposite
message. Although conceptions of this type—beliefs or cognitive
orientations constructed by social actors as an unintended consequence of
social experience—have important effects on the social behavior, social
structure, and world view of social groups, they are not, by our definition,
“cultural.” Usually, but not always, unconscious, they are not encoded in
collective signs, and they are (therefore) not transmitted by means of
intentional enculturative processes. (They are discovered by the
anthropologist as an inference from clinically oriented interviews, dreams,
projective tests, and culturally constituted projective systems such as myths,
folklore, and the like.)
Third, although many cultural propositions have emotional antecedents,
and although others have emotional consequences—they arouse emotional
responses in social actors—and although some even prescribe the proper
conditions for the expression of emotions, culture as such—a cognitive
system encoded in collective representations—does not consist of emotions.
The exclusion of emotion from this conception of culture is not based on
the view—which, it might be thought, I hold either from ignorance or
foolishness—that emotions occur without thought or thought without
emotions. Even if I held such a misguided view (which I do not), it would
hardly be germane to the present issue because—although I have defined
“culture” as a cognitive system—culture does not consist of thought
(thinking) any more than it consists of emotion (feeling). As thinking and
feeling are properties of persons, and as culture—neither by this definition
nor by any other that I am aware of—does not consist of persons—though
society does—it is hard to see how either could be part of culture.
Although not a part of culture, thinking and feeling are often determined
by culture. That is, we most often think by means of the concepts
comprising cultural propositions, and our emotions are often aroused by
them; in short, many of our thoughts and emotions are (what might be
termed) “culturally constituted.” But just because thoughts and emotions
are culturally constituted, it is logically impermissible to conclude that
culture, as such, consists of emotion or thought any more than it can be
concluded that culture consists of behavior or social structure, although they
too are culturally constituted (cultural propositions serve to motivate
behavior and to provide a model for special structure). The conflation of
culture and culturally constituted phenomena is based on a confusion of
logical types, of cause with effect, structure with function, producer with
product. (It is that very confusion that invalidates my earlier attempt ([Spiro
1950] to understand the relationship between mind and culture by arguing
—as I now see it—fallaciously that “culture” and “personality” are
synonymous.)
If, then, the exclusion of emotion from this conception of culture is not
based on the misguided view that feeling is isolated from thinking, neither
is it based on the view that while emotions (because they are “located” in
the mind) are private events, thoughts (because culture is “located” in
collective representations) are public. In fact, I reject that view, not because
I believe that emotions are public—although their expression is—but
because I believe that thoughts are also private. Hence, as culture is a public
system, thoughts no less than emotions are, by definition, excluded from
“culture.”
 Culture is public because it consists of propositions that are encoded in
collective representations, or public signs, usually, but not always, symbols.
It is precisely for that reason that anyone—and not only a native—can learn
the culture of any social group. Since, then, the “meanings” of those signs
consist of the concepts they designate and represent, cultural propositions
are public because—as Saussure (1966) has shown—their meanings are
“located,” as it were, in the signs themselves: the signs, to employ his
terms, function simultaneously as both signified (concept) and signifier
(sign vehicle). When, however, cultural propositions are learned by social
actors, they become personal thoughts that, like emotions, are private; they
are now “located” in the mind.
However, because these private thoughts are derived from, though they
may be less than isomorphic with, cultural propositions, they are (to employ
a term introduced earlier) culturally constituted. Nevertheless, inasmuch as
they are “located” in the (private) mental representations of social actors,
rather than in the (public) collective representations of their group, to
conclude that culture consists of these culturally constituted thoughts is to
commit the logical fallacies mentioned earlier.
Finally, this conception of culture does not imply that the meanings of the
symbols and other signs—both “discursive” and “presentational”—in which
cultural propositions are represented are only conscious, nor (in the latter
case) only denotative. This conception of culture recognizes that cultural
symbols have unconscious and connotative meanings, as well—the
meanings that are expressed in metaphor, metonymy, and other tropes.
Moreover, as it is in these meanings that most of the emotional action of
symbols is found, it is in respect to them that the interaction between
culture and emotion is most pronounced. As these meanings, however, do
not constitute the (conventional) meanings of symbols but are rather the
meanings that social actors, consciously or unconsciously, intend for them
to have, they are “located” not in the symbols themselves but in the minds
of social actors. Hence, if we distinguish between the meanings of cultural
symbols and the meanings that they have for social actors, the range of
meanings that a culture has for social actors is much broader than the range
of meanings of (conventional) cultural symbols.
Because of this difference in range, although anyone, not only the
natives, can learn a culture, the meanings that it has for the natives may be
very different from those that it has for non-natives. In learning a foreign
culture a non-native may acquire as firm a grasp of the meanings of the
culture as a native, but not having been socialized in the group, he or she
has not had those social experiences that, alone, serve to invest the culture
with those surplus meanings, that it has for the native.
That is why, too, we must distinguish between learning a culture and
becoming enculturated. To learn a culture is to acquire its propositions; to
become enculturated is, in addition, to “internalize” them as personal
beliefs, that is, as propositions that are thought to be true, proper, or right.
This is especially the case for those propositions that Shweder (1984) calls
“cultural frames,” propositions that can be neither proved nor disproved.
Thus, for example, a non-Buddhist scholar of Buddhism may have studied
its textual doctrines and may know their meanings in much greater detail
than a Buddhist. Nevertheless, the scholar is rarely converted to Buddhism
—its textual doctrines do not become personal beliefs—because, not having
been socialized in a Buddhist society, its textual doctrines do not have for
him the connotative meanings, nor do they arouse in him the emotional
responses, that alone serve to transform cultural frames into culturally
constituted beliefs. I shall return to this thesis later.
That is also why, even for the natives themselves, many cultural frames
are, or become, what might be called “cultural cliches”—propositions to
which actors may give nominal assent but which are not “internalized” by
them as personal beliefs. Although this may occur for many reasons, the
most important, in my view, is that these particular propositions do not
have, or have lost, emotional importance for them.
Finally, that is why the anthropologist who is interested in the meanings
that a culture has for the social actors must investigate the personality of the
actors with the same diligence that he investigates their cultural symbol
systems. For, in addition to the fact that, as D’Andrade has observed (1984),
there is considerable slippage in the transformation of cultural propositions
into (culturally constituted) beliefs, the surplus meanings of cultural
symbols are “located” in the minds of the actors.

The Aims of This Chapter

Given the cognitive and public conception of culture sketched in the
preceding section, the primary questions I wish to address in this chapter
are the questions of cultural determinism and relativism. The reason for
addressing these traditional questions in particular is that they form the core
issues of the essays I was asked to discuss at the SSRC conference—the
essays by Shweder and Rosaldo—and the excellence of those essays,
together with the ensuing discussion of them, served to focus in sharp relief
those dimensions of cultural determinism and relativism from which I
dissent. In my view the essays and discussion alike attribute much more
social and psychological power to particular cultures than I believe is
justified; and from the variability among cultures, they draw relativistic
conclusions that are more radical than I believe is warranted.
Almost all anthropologists, I take it, would agree that descriptive cultural
propositions related to scientific (or ethnoscientific) domains can be judged
to be true or false by normal canons of scientific evidence. Hence, what is
at issue here is that subset of cultural propositions, both descriptive and
normative, that Shweder calls “cultural frames.” Cultural frames, of course,
constitute a large—perhaps the largest—proportion of all the propositions
comprising any culture.
That cultural systems display a wide range of variability in their cultural
frames is of course an empirical generalization that anthropology has
documented both richly and abundantly. However, that no comparative
judgments can be made, as some relativists claim, concerning these frames
is, in my view, a non sequitur. For if, as I believe, the processes that
characterize the working of the human mind are the same everywhere—
even though human cultures are different—then there are certain
psychological criteria by which such judgments can be made. For example,
judgments concerning the rationality of cultural frames can be made
according to the criterion of “reality testing”: that is by the degree to which
they rest on a failure to distinguish fantasy from reality. Again, judgments
concerning their utility can be made according to the criterion of functional
consequences, that is by the degree to which they lead to an emotional
disturbance and to disruptive social relationships rather than the reverse,
and so on. These and other criteria will be examined in detail.
The validity of these criteria, of course, rests on the assumption that, as
mentioned earlier, the human mind works (or has the capacity to work) the
same everywhere so that, in respect to my examples, their validity presumes
that all humans have the capacity to distinguish fantasy from reality, to
prefer pleasurable to painful feelings, to welcome nonconflictual over
conflictual relationships, and so forth.
Having mentioned two psychological criteria for comparing the relative
merits of different cultures, I am not suggesting that culture can be
“reduced” to psychology. Far from it. For to say (as I have) that the
propositions comprising culture are “traditional” propositions is to say that
culture is quintessentially a historical product. Hence, any adequate account
of culture is necessarily a historical account. Nevertheless, once a cultural
proposition is acquired as a personal belief by social actors, its acquisition
is a psychological event, and that event requires a “psychological”
explanation. This can perhaps be best understood by reference to the
concept of the cognitive salience of cultural propositions.
Somewhat schematically, and in ascending order of importance, those
propositions that, following Shweder, we are calling “cultural frames” may
be said to comprise a hierarchy of cognitive salience consisting of the
following five levels.

1. As a result of normal enculturative processes, social actors learn about

the propositions; they acquire an “acquaintance” with them, as
Bertrand Russell would say.
2. In addition to learning about the propositions, the actors also
understand their traditional meanings as they are interpreted, for
example, in authoritative texts or by recognized specialists.
3. Understanding their traditional meanings, the actors “internalize” the
propositions—they hold them to be true, correct, or right. It is only
then that they are acquired as personal beliefs. The transformation of a
cultural proposition into a culturally constituted belief does not in
itself, however, indicate that it importantly affects the manner in which
the actors conduct their lives, which leads to the fourth level.
4. As culturally constituted beliefs, cultural propositions inform the
behavioral environment of social actors, serving to structure their
perceptual worlds and hence, to guide their actions.
5. At this level, culturally constituted beliefs serve not only to guide but
to instigate action, that is, they possess emotional and motivational, as
well as cognitive, salience. Thus, for example, one who acquires the
religious doctrine of infant damnation as a personal belief at this level
 of cognitive salience not only incorporates it as part of his
(theological) belief system but also internalizes it as part of his
motivational system: It arouses strong affect (anxiety), which, in turn,
motivates him to action (the baptism of his children) whose purpose is
to save them from damnation. (For a detailed explication of the
consequences of this hierarchy, see Spiro 1982.)

A cultural frame that is acquired as a personal belief at the highest three

levels of cognitive salience has a legitimacy and a moral and emotional
urgency that are absent from other cultural frames. Thus, as was noted
earlier, although the existence of such a frame as part of the cultural
heritage of the group requires a historical explanation, its internalization as
a personal belief requires a “psychological” explanation. This can perhaps
be clarified by the example of the Augustinian doctrine of infant damnation.
The doctrine of infant damnation, a cultural proposition that is part of the
theological system of certain Christian denominations, must be explained,
in the first instance, by the history of Christian theology. The moment,
however, that some Christian actor says, “By God, it’s true!” it is
internalized as a personal belief, and its internalization requires a
psychological explanation. For why this actor (or any other) should believe
that a just God would condemn little children to damnation must now be
explained by reference not to the history of Christian doctrine (which the
believer is probably ignorant of anyway) but to the cognitive orientations,
perceptual sets, and motivational dispositions that led him to assent to its
truth.
Again, in saying that the transformation of a cultural frame into a
personal belief entails an explanatory shift from a historical to a
psychological perspective, I am not saying that this represents a shift from a
group to an individual level of explanation. For the psychological char-,
acteristics of social actors must also be explained by reference to a group
phenomenon—not, however, by reference to culture but, rather, at least in
my view, to society, a distinction that (for me at any rate) is crucial. If
“culture” refers to traditional propositions about nature, man, and society,
then “society,” as I am using that term, refers to traditional forms of social
relations, in which “social” refers to a range extending from a dyad to a
nation-state.
 Cultural acquisition begins in childhood, and children acquire culture
from persons who are their “significant others,” that is, persons—usually
parents or parent surrogates—with whom they have a powerful emotional
involvement, both positive and negative. Hence, their mental
representations of these parenting figures not only persist; they often
constitute as well a template for their perception of other superordinate
beings, both human and divine. It is the mental representation that a child
forms of a particular type of parenting figure that explains (in part) how an
otherwise benevolent Christian might come to believe that a just God can
send unbaptized infants to hell. For in their modes of socialization and
enculturation, not only do parents intentionally transmit the meanings or
messages of cultural propositions to their children, but they also
unintentionally transmit another set of messages to them: messages about
the kinds of persons they (the parents) are, the conditions under which they
offer and withhold love, and punishment, and the like. It is from these latter
messages, acquired by (as Bateson calls it) deutero-learning, that children
form their parental representations. If the particular concept of God that is
implicit in the doctrine of infant damnation is consistent with the child’s
parental representation, then, when he later learns that doctrine, it makes
“psychological” sense to him and that is one reason he might internalize it
as a personal belief.
That cultural frames make emic, or intracultural, sense does not imply,
however, that they are immune from valid transcultural, or etic, judgment
(as Shweder argues), let alone that they are the primary, if not the exclusive,
determinants of what social actors think and feel or how they behave (as
Rosaldo argues). Before examining these arguments, however, I wish to
emphasize that where I disagree with them, my disagreement is with the
two dominant themes of con temporary culture theory that they represent
and not with Rosaldo’s and Shweder’s able formulations of these themes.
Indeed, because their formulations are clearer and more cogent than any I
previously encountered, I was able to more clearly discern my own
disagreement with these two themes—the themes (as I shall call them) of
particularistic cultural determinism and normative cultural relativism.

Particularistic Cultural Determinism

“Culture patterns,” Rosaldo wrote in an earlier draft of her essay, “provide a
template for all human action, growth, and understanding” (my italics), and
again, “culture does not dictate simply what we think but how we feel about
and live our lives” (italics in original).1 Such comprehensive claims, no
doubt, gratify our anthropological egos, for as the main keepers of the flame
of culture, we thereby assure ourselves that we travel the royal road to
human understanding. Gratifying as it is, we must nevertheless ask to what
extent such claims are true. Is there nothing in biology or in social relations,
for example, that might also affect our action, growth, and understanding?
And is there nothing in imagination and fantasy that might also affect the
way we feel about and live our lives? In short, to advert to the first question,
are our “action, growth, and understanding” never influenced by our trans-
cultural biologically acquired drives and socially acquired selves or are they
only influenced by culture patterns? Moreover, to advert to the second
question, do our egos never transcend the constraints of all three
determinants—the biological, the social, and the cultural alike—and
thereby achieve some degree of autonomy, if only in fantasy and
imagination?
1. In its published version, this quotation has been modified to read, “Culture makes a difference
that concerns not simply what we think but how we feel about and live our lives.” Since this
modified version is one that every anthropologist, not only a cultural determinist, would
unqualifiedly assent to, I have used the earlier version as representing the view of particularistic
cultural determinism, even though Rosaldo herself apparently no longer subscribed to it.
Now because, as Rosaldo claims, culture dictates what we think and feel,
we (following the dictates of our culture) “think of hidden or forgotten
affects as disturbing energies repressed [and] see in violent actions the
expression of a history of frustrations buried in a fertile but unconscious
mind,” but the Ilongots (following the dictates of their culture) think and
see nothing of the kind.
Although I am not convinced that, except for some few Freudian
theorists, most of us do not think and see things as the Ilongot do, the
relevant issue is not whether we, but not they, entertain the former theory
about the relationship between affects, frustration, and violence, but
whether that theory—let us call it the “Western” theory—can account both
for their behavior and ours. For if it can, then much of their behavior and
ours, alike, must be explained not by the “dictates” of our respective (and
different) culture patterns but rather, as the “Western” theory claims, by the
transcultural characteristics of the human mind. Let us briefly examine that
claim by examining the paradigmatic case that Rosaldo cites from her
fieldwork among the Ilongot.
In that case, a man who was ostensibly offended by the carelessness of
his “brother” in making some plans got drunk and fought with him. That his
aggression was not caused (as the Western theory would claim) by some
“repressed” resentment, which finally surfaced when he was drunk, is
proved, Rosaldo says, by the absence of “symptoms” of resentment in his
subsequent sober behavior. Hence, she argues, these “brothers” (and the
Ilongots in general) were and are successful in “keeping ‘anger’ from
disrupting bonds of kin suggests that in important ways their feelings and
the ways their feelings work must differ from our own” (1984).
Although that is one possible conclusion, it is not the only one. In the
first place, so far as the working of “our own” feelings is concerned, the
Ilongot case can surely be duplicated in spades in our own society. The
documentary evidence—from history, biography, sociology,
psychoanalysis, and, of course, fiction, let alone the evidence from our
personal observations of kin relations (if not of others, then of our own)—is
too abundant to require comment. Indeed, it is the very abundance of that
evidence that led, in the first instance, to the development of the theory of
repression and to the study of the various defense mechanisms by which,
according to the “Western” theory, repressed emotions and motives find
disguised expression.
In the second place, to return to the Ilongot case, it is not at all clear from
the available data that the brothers in fact did not “repress” their anger for
each other or that the frustration-induced anger in their relationship is in
fact not expressed in violent action—not, to be sure, against each other, but
in culturally approved violence against enemies. In the absence of the
relevant psychological data for testing this hypothesis, I am not claiming
that this in fact is what does occur, but in the absence of the relevant data it
cannot be concluded that it does not occur. At the same time there are hints
in Rosaldo’s description that suggest that it might occur.
The Ilongots, Rosaldo informs us, “think of ‘anger’ as a thing that, if
expressed, will necessarily destroy social relations …. [they] respond to
conflict with immediate fear of violent death; they say they must forget
things lest expression make men kill; [therefore] disputing persons either
separate or fight—and the expression of violent feelings is seen as always
dangerous” (ibid.).
 Now I would submit that a psychocultural context in which it is believed
that the expression of anger destroys social relations and that any conflict
arouses fear of homicide is a context, par excellence, in which the
“repression” of anger and the “denial” of conflict, most especially in kin
relations, are most expected, as the Ilongots themselves explicitly
recognize. For what else can the injunction to “forget things lest expression
make men kill” possibly mean if it does not mean, “Since anger (my own
and alter s) is so powerful that once aroused it leads to murder, in order to
avoid such a consequence (either for me or for alter), it is crucial that I (and
he) ‘deny’ those frustrating events that arouse our anger.”
Now what there is about Ilongot social relations that arouses anger of
such intensity or what there is about the lack of controls in Ilongot
personality that transforms anger into homicidal impulses, we do not know.
However, that the one or the other (or both) should lead to “‘forgetting
anger in those contexts [kinship] where a show of violence has no place”
(ibid.) is now not only understandable but also predictable by the “Western”
theory. Is there any other way that a kin-based, face-to-face society, whose
members have those personality characteristics, might possibly survive? In
short, rather than confuting the view of the “Western” theory that the human
mind, including its affective dimension, works the same everywhere, the
Ilongot case in fact supports it. More particularly, it supports its claim that
profoundly disturbing affects and events are dealt with, inter alia, by
repression and denial regardless of cultural differences and the “dictates” of
culture patterns.
But what about the other claim of the “Western” theory, the claim that
“violent actions [are] the expression of a history of frustrations buried in a
fertile but unconscious mind.” That claim is especially confuted, Rosaldo
argues, by the fact that, in the case of the two brawling Ilongot “brothers,”
no “symptoms” of any antagonism were evident in the antagonist once he
became sober. But given the psychocultural context that has just been
described, the “Western” theory would not expect such “symptoms” to be
evident in the relationship between the “brothers.” Rather, it would predict
precisely what Rosaldo describes, namely, that the Ilongot would be “quite
capable of ‘forgetting anger’ in those contexts where a show of violence has
no place,” the fraternal relationship being, par excellence, one of those
contexts.
 That theory would also predict, however, that their “forgetting anger” in
their relationship with each other does not mean that the anger of the
“brothers” had disappeared. Rather, it would predict that having “repressed”
the anger that is aroused in the “history of frustrations” with each other,
they would express it in symbolic disguise in other “contexts where a show
of violence has [its] place.”
Given that prediction, however, I do not know enough about the Ilongots
to describe or even to guess at the various symbolic forms by which, or the
various contexts in which, repressed anger between siblings (or anyone
else) might be expressed. However, because the Ilongots have a long
tradition of headhunting and because few actions are as violent as hacking
off a man’s head (which is what Ilongot head-hunters do), I would suggest
that the headhunting expedition is perhaps the most important symbolic
form by which—just as “enemy” territory is a most important context in
which—repressed anger toward their fellows (including siblings) is both
displaced and gratified. This suggestion, of course, can only be verified by a
symbolic investigation of Ilongot headhunting to determine whether the
headhunter’s victim is an unconscious symbolic representation of his
frustrating fellows. Although the data for that kind of analysis are not
available, this suggestion is consistent with the following characteristics of
Ilongot headhunting, described in R. Rosaldo’s Ilongot Headhunting.
Item 1. Although headhunting expeditions are usually instigated by
insults or wrongs, some of them (at least from our point of view) are trivial
in the extreme. Moreover, they are sometimes undertaken without any
recognizable motive at all as, for example, when the prospective victim is in
a “vulnerable” state and happens to fall victim to “youths with characteristic
relentless zeal to take heads as they grow into manhood” (Rosaldo
1980:63).
Item 2. Even when ostensibly instigated by some insult, retaliation is
seldom directed to the offender or even his family; rather, it will
“encompass a wider target of population, including men and women and
adults and children, all equal in their shared liability as victims” (ibid.).
Item 3. One of the characteristics of the relentlessly zealous head-hunting
youths is their emotional volatility, “their moods and passions [being]
subject to dramatic ups and downs; and many youths describe themselves in
song and story alike as weeping in their fierce, as yet frustrated, desire to
‘arrive’ and take a head” (ibid., 139).
 Item 4. The source of their fierce desire “is above all envy of their peers
and elders, those men who … have taken a head and thus won the coveted
right to wear [the insignia of a headhunter]” (ibid., 140).
Item 5. “The point in Ilongot headhunting … [is] not to capture a trophy,
but to ‘throw away’ a body part, which by a principle of sympathetic magic
represents the cathartic throwing away of certain burdens of life—the
grudge an insult has created, or the grief over the death in the family, or the
increasing ‘weight’ of remaining a novice when one’s peers have left that
status” (ibid.).
Item 6. “Taking a head is a symbolic process designed less to acquire
anything … than to remove something. What is ritually removed, Ilongots
say, is the weight that grows on one’s life like vines on a tree. Once
cleansed [by taking a head] the men are said to become light’ in weight,
‘quick’ of step, and ‘red’ in complexion…. In other words, the raiders
regress through this ritual process to a culturally idealized phase of life”
(ibid.).
Item 7. After hacking off the victim’s head, the men (in one reported
case, presumably not atypical, that was taped by Rosaldo) “vented their
pent-up anger on the cadaver and chopped it up until it had no body and you
couldn’t see its bones,’ until ‘it was like ashes”’ (ibid., 162).
Item 8. Following the decapitation, the men fled, pausing “now and again
to shout and to sing the song of celebration” (ibid., 162).
Item 9. Before stopping for the night, they collected some fern leaves and
tucked them into their armbands, “in order to modify and preserve the smell
of their victim” (ibid., 163).
Item 10. Upon their return to the village, both men and women joined in
the song of celebration and dancing and singing throughout the night while
their “hearts lengthened with joy” (ibid., 163).
Although these characteristics of Ilongot headhunting speak for
themselves as implicit support of the prediction of the “Western” theory,
still it might be well perhaps to indicate explicitly why I believe they
support my suggestion that the Ilongot handling of anger is entirely
consistent with the “Western” theory of emotions. Because the victims of
headhunting are clearly not the instigators of the hunters’ anger (Items 1
and 2), the violence that the Ilongot display in hacking off the head of an
‘enemy’ and in their treatment of the cadaver (Item 7) expresses anger that
is clearly displaced from somewhere else. That that anger, whatever its
instigation, is repressed and consequently (in an almost classical hydraulic
model of the emotions) presses for discharge is equally clear (Items 1 and
3). That their anger is instigated in the first instance by frustrations within
the group, frustrations related (among other things) to envy and status
rivalry, is also clear (Item 4), and I would guess that their envy and,
therefore, their repressed anger have a long history, beginning with sibling
rivalry in early childhood. That this might be so is suggested by the Ilongot
metaphor according to which men carry a “weight” in them, the removal of
which is the motive for, and is achieved after, a successful headhunting raid
(Item 6). The latter has the cathartic effect of removing that burden (Item
5), so that the headhunter “regresses” to that “idealized phase of life” (pre-
sibling rivalry?) prior to the imposition of that weight (Item 6). That that
“weight” represents their “pent-up” (repressed?) anger is shown both by the
savoring of their victim’s memory even after their hacking him to bits (Item
9) and especially by the exaltation (a manic state?) that they both feel and
express following the violence displayed in the raid (Items 8 and 10).
To sum up this discussion, I would suggest that it is not the case, insofar
as anger is concerned at least, that “in important ways [Ilongot] feelings and
the ways their feelings work must differ from our own” (Rosaldo 1984). To
be sure, their anger seems to be much more intense than ours, and its
expression is much more violent, but, these quantitative dimensions aside,
their anger and ours seem to work in similar ways. They, like we, get angry
when frustrated, and they, like we, usually repress their anger in culturally
inappropriate contexts only to express it symbolically in culturally
appropriate ones. This indicates, I would suggest, that human feelings and
the ways in which they work are determined not so much by the
characteristics of particularistic culture patterns but by the transcultural
characteristics of a generic human mind.
That those characteristics are in turn determined in large measure,
however, by the transcultural characteristics of a set of universal culture
patterns is of course—for me at least—axiomatic. But that kind of generic
cultural determinism—one that rests on certain assumptions about the
universality of human biological and social characteristics (which we
cannot go into here) and their symbolic expressions and cultural
transformations—is very different from the particularistic cultural
determinism proposed by cultural determinists of the relativistic school.
 Briefly put, particularistic cultural determinism views the process of
enculturation as a process by which a neonate learns to become a
completely enculturated Iatmul, Ifaluk, Ifugao, and the like—a process
according to that view, by which each of the three learns to become
radically different from the others. Although in one sense that thesis is also
held by generic cultural determinism, the latter views enculturation as first
and foremost a process of humanization, a process, that is, in which the
neonate, by becoming an Iatmul, Ifaluk, or Ifugao, is also transformed from
a mammal into a human being.
This is not to deny, on the one hand, that having become encultu-rated
the human being is no longer a mammal—much of this chapter argues that
the contrary is the case. Nor is it to deny, on the other hand, that an Iatmul
human being is different in important respects from an Ifaluk human being
—human beings do come in different cultural shapes and forms. Rather, it
is to affirm (to advert to the second point) that insofar as it is culture that
makes us human and insofar as an enculturated Iatmul, Ifaluk, and Ifugao
are equally human, the deep structural similarities in their cultures (and in
all other cultures) comprise a set of universal culture patterns, which, in
interaction with a common biological heritage and common features of
social interaction, create a generic human mind. It is to affirm, in short, that
despite the surface structure differences in their cultures—and perhaps in
their deep structures too—the minds of the Iatmul, the Ifaluk, and the
Ifugao (and everyone else) work in accordance with the same principles
(“the psychic unity of mankind’). Starting from different premises and
concerned with different intellectual problems, Chomsky and Lévi-Strauss,
of course, among others, have most notably propounded the psychic unity
thesis in our time.
So far as the emotions are concerned, the perspective and strategy of
generic cultural determinism is very different from that of particularistic
determinism. Whereas the former type leads to a search for a set of
universal principles that, insofar as they are derived from the transcultural
characteristics of the human mind, underlie the manifest cultural differences
in the display of emotions, the latter type views such a search as futile.
Perhaps so. But if the preceding analysis of Ilongot anger is valid, it may
not be so futile after all.
Normative Cultural Relativism
Because, according to the theory of particularistic cultural determinism,
emotions and the self (and almost everything else) are determined by
culture, and because particular cultures are markedly different, it follows,
according to what might be termed “descriptive cultural relativism,” that
emotions and the self (and almost everything else) vary as a function of the
variability in cultures. To what extent those cultural variations are
themselves susceptible of judgments concerning their relative worth is a
question that descriptive cultural relativism does not address, and Rosaldo
does not address it.
Shweder, on the other hand, addresses that very question, for his essay
(1984) is a defense of what might be termed “normative cultural
relativism,” the second strand in the classical theory of cultural relativism.
Because, according to this strand, there are no transcultural standards by
which the variable propositions of particular cultures can be validly
evaluated, there is no way by which their relative worth can be judged. (A
similar distinction between these two kinds of relativism is made by Swartz
and Jordan 1976:701–2.)
Shweder, however, both broadens and narrows this claim. He narrows it
by restricting it to a subset of cultural propositions, which he calls “cultural
frames.” He broadens it to include explanation as well as evaluation. That
is, he also claims (for reasons to be elucidated later) that there are no
transcultural theories by which cultural frames can be validly explained.
“Cultural frames.” as I have already indicated, consist of that subset of
cultural propositions that are susceptible of neither confirmation nor
disconfirmation. They can be neither confirmed nor disconfirmed according
to Shweder, because they are (1) arbitrary, being grounded in neither logic
nor experience, and (2) being arbitrary, they are neither rational nor
irrational but nonrational. These characteristics of cultural frames have two
implications so far as their explanation and evaluation are concerned. As for
the former, cultural frames “fall beyond the sweep of logical and scientific
evaluation,” so that the most the anthropologist can hope to do is
“document” the differences among the cultural frames of different cultures
and explicate the “internal rules of coherence” of those comprising the same
culture. As for evaluation, “there are no standards worthy of universal
respect dictating what to think or how to act,” so that alternative frames are
neither “better” nor “worse” but only “different.” In short, the “whole thrust
[of this position] is to defend the co-equality of fundamentally different
‘frames’ of understanding” (ibid.).
Finally, the cultural frames of different cultures are not only
“fundamentally different,” but they are “irreconcilable” (ibid.); and because
there are no “deep structures” underlying their surface content, “the more
we attend to surface content, the less common is the culture of man” (ibid.).
Since, therefore, cultural frames do not reflect but rather construct reality,
there is no single reality—none at least that can ever be known; there are
only “diverse realities.”
Although there is much in this argument with which I agree, I believe
that it is both over- and understated. Specifically, I believe, on the one hand,
that cultural propositions are less arbitrary and, on the other, that they are
more irrational than normative cultural relativism claims. That being the
case, I believe that cultures are susceptible of a greater degree of
explanation and comparative evaluation than a relativistic view allows. I
shall begin with the degree of arbitrariness of cultural propositions.
In my view cultural relativism overstates the arbitrariness of cultural
propositions because, despite their undoubted variability, their range is not
all that variable. For any cultural domain, and for any sample of cultures,
the diversity of cultural propositions is much smaller than the number of
cultures comprising the sample; and the diversity becomes smaller yet when
we observe that even at the level of surface content—let alone at the level
of deep structure, which Shweder rejects—the diversity can be radically
reduced to, and yet adequately described by, a small number of types.
Although cultural propositions (including cultural systems of
classification) may be arbitrary, they are not all that arbitrary. Shweder is
correct, of course, when he says that “logically, any classification is
possible.” But he himself adduces an important constraint when he observes
that any actual classification is a function of “some special purpose of
man.” Given this relationship between “purpose” and classification (and
even omitting unconscious purposes), the fact that in our kinship system
parents’ brothers and parents’ sisters’ husbands are classified as “uncle”
does not, as Shweder claims, make such a classification of these kin types
“arbitrary.” Indeed, in discovering their classificatory principles, Kroeber
(1909) and others have been able to elucidate the logic not only of the
putatively nonrational and arbitrary classification of these kin types but also
of the entire classificatory systems in which they are embedded. Moreover,
by demonstrating systematic relationships between kinship systems and a
small set of ecological, political, economic, and other determinants,
Murdock (1949) and others have been able to causally explain the
distribution of these types.
To be sure, were we to discover a system in which not only parents’
sisters’ husbands, but parents’ brothers’ wives, or brothers’ sons’ wives, as
well, were also classified with parents’ brothers, that would certainly count
as “arbitrary.” But there is no need to reach for such extreme counterfactual
examples to make my simpleminded point that, inasmuch as the range of
cultural variability is not unconstrained, many cultural classifications are
only apparently nonrational.
That still leaves, of course, a large number of cultural propositions that
are unquestionably nonrational: They are grounded in neither logic nor
evidence. That they are, therefore, arbitrary (thereby eluding the net of
causal explanation) is perhaps true of some of them. To claim, however, that
it is true of all of them is to overlook another possible determinant of
cultural propositions.
Although, beginning with its title and persisting as a pervasive theme of
the chapter, Shweder argues that “there’s a lot more in the mind than reason
and evidence,” including “culture, the arbitrary, the symbolic, the
expressive, the semiotic,” he does not mention the emotional part of that
“more,” nor does he explicitly attend to emotions in his discussion. If,
however, we attend to that emotional part, then, although many cultural
propositions may not be determined by logic or evidence, it may
nevertheless not be the case that they are arbitrary, for they may be
determined by emotion. In that case, they do not—in principle at least—
elude the net of scientific explanation, if by “explanation” is meant the
discovery of systematic relationships or regularities between cultural
propositions and noncultural conditions that might account for cross-
cultural variation.
Even if relativists were to concede that some cultural propositions might
be grounded in emotional needs, they might counter that they are
nevertheless arbitrary precisely because there seems to be no systematic or
predictable relationship between such propositions and their putative
emotional determinants. Food taboos and witchcraft propositions, to take
but two examples, are often claimed to have emotional determinants, and
yet there is no obvious reason why this should be the case, nor are there any
predictable emotions to which they are attached.
This riposte is well founded if we attend exclusively to the public and
conscious meanings of cultural propositions. If, however, we also attend to
the cognitive meanings that they have for social actors—especially their
unconscious meanings—it is often the case that their emotional
determinants can be discovered, for (like the emotional determinants of
dreams) they are most frequently related to the unconscious meanings that
they have for social actors.
That conclusion rests on two assumptions. First, in some cases, at least,
the conscious meanings of cultural propositions (like the manifest content
of dreams) are “surface structure” meanings, and their unconscious
meanings (like the latent content of dreams) are “deep structure” meanings.
Second, the latter meanings are unconscious because conscious awareness
of them (like the conscious awareness of the latent content of dreams)
arouses anxiety, shame, guilt, and the like. On these assumptions the
emotional determinants of cultural propositions can only be discovered by
investigating their possible deep structure meanings. Until such
investigations are undertaken in regard, say, to food taboos, witchcraft
propositions, and others, we cannot assume a priori that they are arbitrary
even when they appear to be in respect to their conscious meanings.
This thesis also applies to seemingly arbitrary and inexplicable
classifications. Classes whose members appear to share no distinctive
features may nevertheless be shown to be logically classified when their
unconsciously perceived similarities are discovered. In that case it is usually
found that class membership is based on the same logical principles by
which tropes are constituted, particularly the type of trope referred to as
synechdoche. For example, one of the reasons that, in his “maternal
transference.” a psychoanalytic patient classified his therapist with his
mother is that (as he saw it) the vital “nourishment’’ that the therapist gave
him in the form of words was an unconscious symbolic representation of
the nourishment that, as a child, he had received from his mother in the
form of milk. The use of unconscious distinctive features as the basis for
seemingly bizarre classifications applies not only to private but to cultural
classification as well.
 By attending to the relationship between emotion and culture, we may
discover not only that many cultural propositions (because they are not
arbitrary) are susceptible of scientific explanation but also that alternative
cultural frames are susceptible of being judged as “better” or “worse.” That
is because cultural propositions may have not only emotional antecedents,
as in the case of the previous example, but emotional consequences as well.
If those consequences are ignored, then, viewed “in terms of comparative
adequacy,” it might indeed be the case, to take Shweder’s example, that the
only judgment that can be made as between those who empathize with
starving Armenians and those who take the heads of their neighbors is that
they are “obviously different.” Many of the headhunters that I have read
about—and the one that I knew—do not, however, share that view, not
because they necessarily share the Western cultural frame that to give life is
better than to take it but because in headhunting societies the hunter is also
the hunted, a prospect which leaves him in a state—actual or potential—of
terror.
Now some relativists might retort that the emotional consequences of
nonrational ideas and actions cannot constitute a measure of “comparative
adequacy,” because inasmuch as the assessment of emotions, like the
assessment of reality, is culturally relative, headhunters, unlike Americans,
may prefer to live in terror. Such a retort, however, would constitute an
empirical disagreement, which, presumably, could be resolved by collection
of the relevant data.
By not attending more explicity to the relationship between emotion and
culture, Shweder, in my view, not only overstates the arbitrariness of
culture but he also understates its irrationality. For Shweder, the “irrational”
is confined to errors in thinking or thought (the verb). But inasmuch as
cultural propositions are traditional ideas or thoughts (the noun), they are
acquired by social actors from the cultural heritage of their group, not as a
product of their own thought (the verb). Since they therefore elude
Shweder’s criterion for assessing “irrationality,” by an unintended sleight of
hand many cultural propositions become “nonrational”—neither rational
nor irrational. To the degree, however, that the internalization of cultural
propositions is motivated by the wish to satisfy emotional needs, many of
those that are internalized at the highest three levels of cognitive salience
mentioned previously are arguably “irrational” rather than “nonrational.” I
am referring to the subset of emotionally motivated propositions that are
emotionally driven and are therefore obsessional and “magical.” To better
understand this contention, let us briefly describe the characteristics of
“obsessional” and “magical” thoughts, beginning with the former.
An obsessional idea or thought is an all-consuming thought, that,
typically, is driven by a powerful emotional need to prevent an unconscious
thought, which it has replaced, from entering consciousness because the
latter arouses an uncontrollable, forbidden, or painful emotion, such as lust,
hatred, or guilt, respectively. In short, an obsessional thought is a defense
against powerful and potentially overwhelming anxiety. The reason for
calling thoughts of this type, whether privately or culturally constituted,
“irrational” is that they are impervious to reason or evidence. Even when
confronted with logical inconsistencies in holding them, or with
disconfirmatory evidence in respect to them, those who hold such thoughts
resist any change—lest the anxiety they are defending against erupt into
consciousness. Held in thrall by these thoughts, they cannot let them go.
Although clinical psychiatry is a storehouse of thoughts (including cultural
propositions) of this type, let us illustrate them by an example taken from
the storehouse of ethnography.
Consider, for example, the widely prevalent cultural proposition that
various forms of misfortune are caused by witchcraft. As a proposition—
but not necessarily as an internalized belief—this appears to be a good
example of a cultural frame. Indeed, when found in prescientific societies,
one might go so far as to call it a “rational” proposition even though, from a
modern scientific perspective, it is false. In the former societies, such a
proposition is validly deduced from a culturally constituted world view; it is
inductively supported by a great deal of empirical evidence; it is compatible
with whatever reliable scientific knowledge is available in those societies;
and it is consistent with all the other cultural propositions of those societies.
In short, it is a proposition that in prescientific societies does not rest on
fallacious inductive or deductive reasoning, does not violate the law of
contradiction, and—given their scientific knowledge—is not empirically
absurd.
Whether rational or nonrational, however, if the social actors, driven by
imperious emotional needs, develop an obsessional concern with witchcraft
—as happened, for example, in sixteenth- and seventeenth-century Europe
and seventeenth-century New England—their witchcraft propositions can
be said to be neither rational nor nonrational, but irrational. Having become
obsessive, they are no longer just another of those many cultural
propositions that social actors are taught and to which they give assent.
Rather, being emotionally driven, the actors are not free—regardless of
evidence, logic, or reason—to let them go.
The same may be true not only of relatively dramatic cultural
propositions like those concerning witchcraft but also of seemingly trivial
ones like those concerning food preferences. One need only observe, for
example, the obsessional concern of some few orthodox Muslims, Jews, or
Hindus with their respective food taboos to realize that some food
preferences, at least, can hardly be characterized as merely nonrational.
That they are rather irrational is evidenced by the acute anxiety the actors
suffer even by the prospect, let alone the act, of violating them.
That in neither of these cases, clearly, is there anything in the (conscious)
meanings of these propositions that would explain—or permit us to predict
—why they should be emotionally driven, and hence obsessive, suggests
that the explanation lies in the (unconscious) meanings that they have for
the social actors. To uncover those meanings, of course, requires a separate
investigation of each of these frames, case by case, in their ethnographic
context.
Emotionally driven cultural frames are sometimes not only obsessional
but also magical, and all the more irrational. A “magical” thought, as I am
using it here, is based on an impairment in “reality testing,” that is, the
ability to distinguish mental events that occur in the inner world (the mind)
from physical events that occur in the external world. By this definition, if a
cultural proposition asserts that heaven is populated by seven gods, each
possessing seven heads, it is not based on any impairment in reality testing.
Nor could it properly be said to rest on such an impairment even if, driven
by powerful emotional needs, the social actors have a fantasy of such a
heaven. If, however, they were to construe such a fantasy—an image in
their minds—as a perception of such a heaven, then inasmuch as the fantasy
had become a hallucination, the proposition is now based on impaired
reality testing, and is “irrational.”
In short, to say that a “magical” thought is based on an impairment of
reality testing is to say that it is a thought in which inner reality is mistaken
for outer reality, impulses stemming from the internal world are mistaken
for stimuli coming from the external world, a mental representation of an
object or event is mistaken for the object or event that it represents, and so
on. If, then, because they are emotionally driven, cultural propositions are
based on an impairment in reality testing, they are clearly irrational as they
are based on the most primitive form of empirical error. (For a smiliar view
of the relationship between culture and the irrational, see Devereux 1980:
chap. 1.)
For an actual, rather than a hypothetical, example of the relationship
between cultural frames and magical thinking, let us turn once again to
witchcraft. From the cultural frame that misfortune is caused by witches,
not only do social actors acquire a belief in the existence of such a class of
malevolent human beings, but they also form composite mental
representations of them, which (more or less) correspond to the collective
representation of witches. Their mental representations of witches can be
called a fantasy—a culturally constituted fantasy—and so long as they are
just that, they reflect no impairment of reality testing. If, however, driven by
emotional conflicts, an actor comes to believe that some specific person, X,
is a witch and that X is bewitching him, his belief is characterized by
magical thinking because the fantasy is now mistaken for reality.
Take, for example, the ailing Burmese villager, Mr. G, who told me that
he was being attacked by a witch, whose spirit or soul has possessed him,
and that the witch was in fact his wife. This belief, in my view, was based
on a complex set of obsessional and magical thoughts. Enraged by his
wife’s repeated and flagrant adulteries, but unable to cope with his rage, Mr.
G projected it onto his wife, thereby constructing a mental representation of
her as malevolent and punitive. Identifying this mental representation of his
wife with his composite mental representation of witches, he thereby
perceived his wife to be a witch. That is, he modified his mental
representation of his wife to correspond to that of a witch, although she is
not one.
But if that is not irrational enough, consider then that his belief that his
wife, now a witch, possessed him was based on an even more serious
impairment in reality testing. For his sensation of his body having been
invaded by her could only be based on first, the reification of his mental
representation of the witch, and second, mistaking it for the witch herself.
As a reification, her mental representation was no longer experienced by
him as an image that he “located” in his mind; rather, he experienced it as
his wife herself (more accurately, as her spirit or soul), which he now
“located” in his body.
 Although witchcraft beliefs, as such, may be nonrational, if—because
emotionally driven—they are based on, or eventuate in, an impairment of
reality testing, they are clearly irrational. Relativism rejects this view on the
grounds, it will be recalled, that inasmuch as culture constructs reality, there
is no other reality that can be used to judge them. Hence, there is no
criterion “worthy of universal respect” by which such beliefs and other
beliefs that are logically consistent with them can be judged on a scale of
rationality. And that is where the issue is joined. Although Mr. G’s belief
that he was possessed by a witch is entirely consistent with the cultural
frame of Burmese witchcraft, and although his fellow villagers, therefore,
fully shared his belief that he was possessed, in my view his belief was no
less irrational than if it had been inconsistent with it. For in both cases such
a belief is based on an hallucination—on a confusion of fantasy with reality
—as those terms have been defined here.
In sum, in ignoring—either from principle or from neglect—the
emotional motivation for the internalization of cultural propositions,
normative cultural relativism underestimates the degree of irrationality in
cultural propositions by construing many of them as nonrational—as
cultural frames—when (as I have attempted to demonstrate) they are more
properly construed as irrational. Hence, by the two criteria that I suggested
for the assessment of irrationality—there may be others—we may conclude
this section by drawing the following—antirelativistic—conclusions. (1)
The alternative cultural propositions comprising different cultural systems
can be compared on a scale of rationality. (2) On such a scale, many cultural
propositions are cultural frames—nonrational—and are therefore merely
“different” from each other. (3) Many cultural propositions that are
seemingly nonrational, and are so viewed by cultural relativists, can be
assessed by this scale as irrational, and they may therefore be judged as
either “better” or “worse” than their cross-cultural alternatives. (4) If total
cultural systems differ in the extent to which they comprise irrational
propositions, by such a scale cultural systems can be similarly judged to be
“better” or “worse.”

Concluding Remarks
I should like to conclude with some personal observations. Thus far I have
examined particularistic cultural determinism and normative cultural
relativism in theoretical terms alone. But theories also have consequences,
and these theories have some (largely unintended) consequences, which—
given my values—I consider unfortunate. Because values, whether
personally constructed or culturally constituted, may be frames, it would be
a conceit to claim that the values on which this judgment is based are
grounded in logic or evidence or that they are free of emotional
determinants. Because the contrary is very likely the case, I refer to these
remarks as “personal observations.”
Shweder, entirely correctly in my view, places (normative) cultural
relativism in the “romantic’’ movement, a movement that tends to place a
high value on the nonrational. Hence, (normative) cultural relativism not
only defends the “coequality of fundamentally different ‘frames’ of
understanding” but, consistent with its intellectual heritage, also celebrates
them precisely because they are nonrational. Shweder puts this view
forthrightly: “Don’t knock the mystical, the transcendental, or the
arbitrary,” he exhorts us even though the “arbitrary” in his discussion
ranges, on the one hand, from differences between one kin type
classification and another to the difference between taking heads and
feeding starving Armenians, on the other.
Now, by this time it should be abundantly clear that I fully agree with
Shweder’s twin contentions that “there’s something more to thinking than
reason and evidence” and that reason is in short supply. But it is precisely
because reason is in short supply, and because (in my view) its alternatives
are not only the nonrational and the logically fallacious irrational but also
the emotionally driven irrational, that I believe that it is reason and those
aspects of culture that are based on it that should be celebrated. To be
concrete, it is not only because the emotionally driven irrational leads to
witchcraft accusations, food taboos, and head hunting, but because it also
leads to the destruction of scholarly papers and scientific laboratories—and
I am referring not only to the Arabian armies that sacked Alexandria in 640
but also to the university students who sacked Columbia, Cornell, and
Wisconsin in the 1960s—that I am dubious about celebrating the mystical,
the transcendental, and the arbitrary. In short, it is because reason is in short
supply that I prefer—as Freud once put it in expressing the goal of
psychoanalytic therapy—“where id was, there ego shall be.”
 By the same token, it is because the emotionally driven irrational has no
limits or—to be more cautious—because its limiting case is Auschwitz, that
I believe that there are standards “worthy of universal respect” by which
cultural frames can be evaluated. That is the case, of course, only if it can
be agreed that those standards—to advert to Freud’s metaphors for the
different parts of the mind—are the standards not of the “id” but of the self-
reflective “ego.” Assuming agreement on that, then, inasmuch as the
preference for headhunting over the feeding of starving Armenians is based
on the standards of the former, while the contrary preference is based on
those of the latter, the latter preference can then be judged to be superior to
the former.
To be sure, not all social actors have self-reflective egos. But on the
assumption that self-reflection is a characteristic at least of philosophers
and saints, and that all societies, including—as Radin has observed—
primitive societies, have their philosophers (and probable saints too), the
preference for feeding starving Armenians over the taking of heads is based
on a standard that, I would argue, is universal in that it is found in at least
the reflective members of all societies.
If, however, it turned out that in some headhunting societies—the
Ilongot, for example—even the philosophers judged headhunting to be
preferable to feeding, say, the starving Ifugao, I would then argue that the
contrary judgment of the Buddha and Christ, of Isaiah and Laotzu, of
Socrates and Gandhi is worthy of greater respect. For, I would contend, the
Ilongot philosophers are unable because of emotional constraints to apply
the standards of their self-reflective ego to this question.
Given my values, particularistic cultural determinism and normative
cultural relativism have a second consequence that I consider to be
unfortunate. As these theories have come increasingly to dominate
anthropological thought, the vision of anthropology as the “study of man”
has been gradually eroded and replaced by the vision of anthropology as the
“study of men.”
The latter vision, which represents a departure from an attempt to
discover cross-cultural regularities and to formulate theories that might
explain them, challenges the status of anthropology as a “science” (as I
characterized that term earlier) because it views each culture (so far at least
as its cultural frames are concerned) as a kind of Leibnitzian cultural
monad. The diversity of cultures, from the perspective of this new vision,
represents a very Babel of voices, each expressing a set of arbitrary ideas
encoded in a set of arbitrary symbols. The noncommensurability of cultural
frames renders them incomprehensible to each other and recalcitrant to
scientific explanation, and their power—like that of God’s voice at Creation
—is sufficient to construct reality itself.
The result of this vision is well known. Although anthropology has made
much progress in the meticulous recording and reporting of the ideas and
institutions of particular cultures, it has made little progress in relating its
findings to the development of a theory of culture. The latter enterprise, we
are now increasingly told, is precluded in principle by the proclaimed
incommensurability of cultures. Thus, we can’t have a cross-cultural theory
of incest taboos, for example, because “incest” has different meanings in
different cultures; we can’t have a cross-cultural theory of religion, because
“religion” means very different things in different cultures and so on.
That view, in my opinion, is unfortunate, and has the potentiality for
becoming even more unfortunate. For as the premises of this new vision
gain currency, it might then be contended—as indeed some have already
contended—that one or more of the following conclusions may be derived
from those premises. (1) Anthropologists cannot adequately describe, let
alone explain, any culture different from their own. (2) For any culture to be
adequately described and understood, it must be investigated by an
anthropologist who himself has been enculturated in it. (3) For the latter to
adequately convey the ideas and institutions of that culture, they must be
reported in the native language, for there is no adequate way of rendering
the conceptual system of one culture by the concepts of another—not even
those of (anthropological) science, which is just another culture-bound
(Western) conceptual system. All science is ethnoscience.
Such a scenario may be the reductio ad absurdum of the new vision of
culture rather than its logical entailment. Nevertheless, we have seen it
enacted (sometimes with the encouragement of anthropologists) in another
form in some American universities. Thus, there are some few programs in
ethnic and women’s studies that claim that only ethnics are qualified to
teach the former and women the latter. Given the new vision of culture, why
should the same logic not apply to the teaching of Confucian philosophy,
the Old Testament, or the French Revolution?
However, even if the former scenario is only the reductio of the new
vision of culture, insofar as it is a potential consequence of that vision, I
would hope that the present ascendancy of particularistic cultural
determinism and normative cultural relativism might be subjected to
continuous and searching scrutiny.

Notes
Reprinted from Culture Theory: Essays on Mind, Self, and Emotion, edited by Richard A. Shweder
and Robert A. LeVine (Cambridge: Cambridge University Press, 1984), pp. 323–46. © 1984 by
Cambridge University Press. Reprinted by permission of the publisher.
Although we continue to disagree on some key issues, this chapter has importantly benefited from
the suggestions and comments of Roy D’Andrade, Richard Shweder, Marc Swartz, and Donald
Tuzin.

References

D’Andrade, Roy G. 1984. Cultural meaning systems. In Culture

theory: Essays on mind, self, and emotion, ed. Richard A. Shweder
and Robert LeVine, 88–119. Cambridge: Cambridge University
Press.
Devereux, George. 1980. Basic problems of ethnopsychiatry. Chicago:
University of Chicago Press.
Kroeber, A. L. 1909. Classificatory systems of relationship. Journal of
the Royal Anthropological Institute 39:77–85.
Murdock, George Peter. 1949. Social structure. New York: Macmillan.
Rosaldo, Michelle Z. 1984. Toward an anthropology of self and
feeling. In Culture theory: Essays on mind, self and emotion, ed.
Richard A. Shweder and Robert LeVine, 137–57. Cambridge:
Cambridge University Press.
Rosaldo, Renato. 1980. Ilongot headhunting. Stanford, Calif.: Stanford
University Press.
Saussure, Ferdinand de. 1966. Course in general linguistics. New
York: McGraw-Hill.
Shweder, Richard A. 1984. Anthropology’s romantic rebellion against
the enlightenment, or there’s more to thinking than reason and
 evidence. In Culture theory: Essays on mind, self, and emotion, ed.
Richard A. Shweder and Robert LeVine, 27–66. Cambridge:
Cambridge University Press.
Spiro, Melford E. 1950. Culture and personality: The natural history of
a false dichotomy. Psychiatry 13:9–204.
_____ 1970. Buddhism and society: A great tradition and its Burmese
vicissitudes. New York: Harper & Row.
_____ 1982. Collective representations and mental representations in
religious symbol systems. In On symbols in cultural anthropology:
Essays in honor of Harry Hoijer, ed. Jacques Maquet. Malibu,
Calif.: Udena.
Swartz, Marc J., and David K. Jordan. 1976. Anthropology:
Perspective on humanity. New York: Wiley.
3
Preculture and Gender

A Precultural Interpretation
ANY ATTEMPT to assess the possible determinants of the counter-
revolutionary changes that have occurred in the kibbutz movement in such
institutions as marriage, the family, and sex-role differentiation is beset with
formidable difficulties. The problem is too complex, the data are too
limited, and our methods of investigation were too primitive to permit an
unequivocal interpretation. The weight of the evidence nevertheless
suggests that although several possible cultural determinants may have
contributed to these changes, on balance they do not appear to have been
decisive. Evidence from sabra childhood behavior suggests, instead, that
these counterrevolutionary changes were more probably brought about not
primarily as a response to external cultural conditions nor by culturally
acquired motives, but by precultural motivational dispositions.
Since, however, it is a basic axiom of the social sciences that human
behavior and motives are primarily, if not exclusively, culturally
programmed, I wish to observe, lest this conclusion be rejected on
axiomatic grounds, that the counterrevolutionary changes in the above
domains were not the only (nor even the most dramatic) changes brought
about by the sabras. A perhaps even more dramatic change occurred in the
sexual domain. Hence I would like to examine briefly this latter change
which constitutes a rather unequivocal exception to our social science
axiom.
According to the ideology of the kibbutz pioneers, attitudes and
orientations to sexual behavior and sexual anatomy are cultural artifacts.
Hence, so they believed, if children were raised in a sexually permissive
and enlightened environment, in which boys and girls, living together, were
acquainted with each other’s bodies and were taught to view nudity as
natural, so that notions of shame were not attached to the exposure of sex
organs—in such an environment, differences in sexual anatomy would
assume little more importance than any other kind of anatomical
differences. This belief was important for the pioneers not only because of
their commitment to healthy sexual attitudes, but also because of their
conviction that sexual equality (in its “identity” meaning) required an
attitude of indifference to sexual di-morphism. If, as they believed, the only
“natural” difference between the sexes consists in differences in sexual
anatomy, if children were raised to view this difference as inconsequential,
the road to sexual equality (as they conceived it) would then have been
paved.
Acting upon their beliefs, the pioneers established an entirely
“enlightened” sexual regime in the children’s houses. Boys and girls used
the same toilets, dressed and undressed in each other’s presence, walked
about their dormitory rooms (if they chose) in the nude, showered together
in one shower room, and so on. This system worked (and works) as the
pioneers expected until the first intimations of puberty in the girls—in
general, girls enter puberty a year or two before the boys—at which time
the very girls who had been raised in a sex-blind environment developed
intense feelings of shame at being seen in the nude by the boys. Sometime
before our 1951 study, the girls in Kiryat Yedidim, for example, initiated an
active rebellion against the mixed showers: they began to shower separately
from the boys, refusing to admit them into the shower room at the same
time. Consistent with this attitude, some of the girls would return early to
their children’s house at night to undress and be in their pajamas before the
boys arrived.
Despite the girls’ active opposition, the educational authorities refused to
change the system of mixed showers. Moreover, when high schools were
built in the kibbutzim, mixed showers and bedrooms were instituted in the
high school dormitories as well. By 1951, however, the mixed showers in
most kibbutz high schools had been unofficially abandoned. As one teacher
in Kiryat Yedidim put it, the mixed showers had become “a form of torture”
for the girls, their shame at exposing their nude bodies in front of the boys
being intensified by the latter’s teasing. Hence, though the high school
authorities did not officially sanction it, arrangements were made for boys
and girls to shower at different times. In a survey I conducted in Kiryat
Yedidim in 1951, only three students in the entire student body favored a
return to the mixed showers. Today, the sexes not only shower separately,
but in almost all kibbutz high schools there are now separate shower rooms
for boys and girls.
The same process has taken place with respect to the dormitory rooms. I
have already noted that even in the grade school the older girls felt
considerable discomfort about undressing in the presence of the boys. Their
discomfort was exacerbated in the high school. In 1951, for example,
although boys and girls in the high school in Kiryat Yedidim shared the
same rooms (usually three boys and three girls to a room), they were careful
to undress in the dark with their backs to each other. Moreover, so that their
bodies would not be exposed, the girls wore pajamas (regardless of the
heat) even though they slept under sheets. Despite these precautions,
succeeding generations of students have been persistently unhappy with
these living arrangements until, seven years ago (and after many
generations of female protest), the high school authorities capitulated to the
girls’ demands, and instituted unisexual bedrooms. Similar changes have
been introduced in most other kibbutzim as well.
In sum, the original kibbutz belief, that in the proper learning
environment children would be sex-blind, was proven to be false even in
the sexually enlightened conditions in which these children were raised.
Even if it were the case that the only natural difference between males and
females is one of sexual anatomy, this one difference apparently is not as
trivial as had been assumed. In this instance, at least, it had important social
and psychological consequences which could hardly have been culturally
determined, for these children (as we have seen) developed a sense of
sexual shame not as a result of, but in opposition to, the cultural values of
their learning environment.1 Apparently, nudity on an impersonal and
anonymous bathing beach is one thing; but in an intimate and potentially
sexually charged small group, it is quite another. When, then, the social
institutions that embodied these cultural values became too painful for the
children, they pressed for their abolition in violation of the attitudes in
which they had been imbued and over the opposition of the adults.2
1. In 1951, under the influence of kibbutz ideology, the high school personnel attributed the sabras’
reactions to the influence of students from the city who had imbued them with feelings of sexual
shame. Being a cultural determinist at that time, I too found this to be a persuasive explanation
although, in retrospect, its flaws are obvious. First, these shameful feelings were almost always
aroused during (or shortly before) pubescence when most girls, still in the grammar school, were not
yet exposed to city students. Second, even for those whose puberty was delayed till high school, the
assumption that the cultural values of a tiny minority of outside students could prevail over those of
the majority, especially when the latter were natives (supported by the entire weight of their native
and much more prestigeful environment) makes little sense. Moreover, if the absence of sexual
shame is natural and its acquisition cultural, this explanation makes even less sense, for one would
then have expected the cultural to give way to the natural. If the kibbutz students were indeed
influenced by the city students, it is more reasonable to believe that they were ready to be influenced
because this influence was syntonic with their natural dispositions.
2. It is pertinent to observe here that these children, whose behavior refuted the assumption that the
shame aroused by sexual dimorphism is cultural, are the same children who, upon becoming adults,
reversed the attempts of the pioneers to minimize the importance of dimorphism by eschewing
feminine clothing, jewelry, and cosmetics. Today, as we have seen, female sabras attempt to enhance
their feminine appearance by these cultural means, and male sabras obviously approve of these
attempts.

Is this not the same process that describes the counterrevolutionary

changes in the family and sex-role differentiation which were instituted by
the sabras upon becoming adults? In the case of these children, reared in a
learning environment that was predicated on the assumption that sex
differences in behavior and psychology are cultural artifacts, that boys and
girls differ only by virtue of their sexual anatomy, and that this difference
becomes socially important only so far as culture makes it so—in the case
of these children the sex differences in behavior that they exhibited very
early in their lives were exhibited in spite of and in opposition to their
learning environment. This being so, it seems most likely that these sex
differences (like their sense of sexual shame) were brought about not by
culture, but by the triumph of human nature over culture, that is, by
motivational dispositions based on sex differences in precultural, rather than
culturally constituted, needs. If, then, the counterrevolution of the female
sabras was motivated by precultural needs, these needs cannot be unique to
them; rather, all things being equal, it is probable that they are shared by
females in any society. Hence, having thus far avoided any discussion of the
types of precultural needs that might explain these sex differences in
motivation, we must finally address this issue directly.

Precultural Needs and the Sabra Counterrevolution

In a typology of possible determinants of the sabra counterrevolution three
types of precultural needs can be distinguished. One type (“biological
needs’) consists of genetically inherited drives. The other two
(“psychosocial” and “psychobiological” needs) consist of experientially
acquired wishes and desires. Here, then, we have three types of
determinants of sex differences in motivational dispositions which are
present prior to (or, as in some cases, independent of) the acquisition of
culturally constituted motives. Although the latter two types are
experientially acquired, they are no less panhuman than those genetically
inherited because the experiences by which they are acquired are dependent
either on certain invariant characteristics of the human organism or on those
characteristics of human society that are invariant. Since the invariant
characteristics of human society (biparental families, group living,
socialization systems, and the like) are institutional solutions to adaptive
requirements of human beings (the satisfaction of early dependency needs,
for example) which they share by virtue of their constituting a common
biological species, these needs too are indirectly “psychobiological.” From
this perspective, then, those pre-cultural needs that are experientially
acquired are no less a part of “human nature” than those that are genetically
inherited. In the present stage, at least, of human biological and social
evolution, both are invariant characteristics of human personality and both
constitute pan-human bases for human behavior. (For the most important
anthropological statement of this thesis, see La Barre 1954.)
Although precultural needs, then, may be either genetically or
experientially acquired, the research strategy employed in this study does
not permit us to decide whether the motivational determinants of the sabra
counterrevolution—and, therefore, of precultural sex differences in
motivation anywhere—are the one or the other. That the counterrevolution
was motivated by precultural needs is an interpretation, it will be
remembered, that was adopted only after the cultural hypotheses
comprising our explanatory paradigm were finally rejected as incompatible
with the data. The precultural interpretation was then adopted not only
because it was the one remaining hypothesis in the explanatory paradigm
but because it was the one interpretation that was compatible with the entire
array of data. On the basis of these data, however, there is no way of
deciding whether the sex differences in precultural needs that are reflected
in the counterrevolution are genetically or experientially acquired. Hence,
the only thing we can do is delineate the shape of these competing types of
precultural interpretations by offering examples of the more prominent
theories which exemplify each type.
To simplify our task, I shall concentrate on only one need for which,
according to our analysis of the behavior of sabra children, there are
precultural sex differences. For this purpose I have chosen the parenting
need because, in one sense, it is the cornerstone of all the changes that
comprise the counterrevolution. The aim of the feminist revolution of the
pioneers, it will be recalled, was to minimize the woman’s involvement in
family, and especially in mothering roles because (it was believed) this
would maximize her involvement in extrafamilial roles. This, in turn, was
expected to lead to the dissolution of sex-role differentiation, and thereby to
sexual equality (in its “identity” meaning). Hence, the feminine
counterrevolution, as I have often emphasized, is essentially a phenomenon
of sabra females, for while the males have persisted in economically
traditional male roles, the females have rejected the more “masculine” of
the traditional male roles in favor of other kinds. Moreover, to a much
larger extent than the males, the females have also reemphasized the very
family—and especially parenting—roles which the pioneers had attempted
to deemphasize. Since, then, the parenting need (like most other precultural
needs) is shared by both sexes, it is with respect to its greater strength in
females that examples of alternative types of precultural interpretations will
be examined. In the following discussion, then, “the female parenting need”
is used as an ellipsis for “the greater strength of the parenting need in
females.”
According to one prominent example of a biological interpretation, the
female parenting need is an instance of those precultural needs which are
genetically determined. Phylogenetically inherited, this need is interpreted
in the same manner as any other biological characteristic that is the product
of biological evolution, namely, by natural selection. Such an interpretation
would hold that in the conditions obtaining in the early history of our
species, a strong mother-child bond was an adaptive requirement, so that a
strong parenting need in women had a selective advantage. Tracing this
advantage to the adaptive requirements of the hunting stage of human
evolution, this is precisely the interpretation offered by Tiger and Shepher
(1975:274–77) for the female parenting need and, therefore, for the
counterrevolution in the orientation of the female sabras to the family.3
According to this interpretation, then, the female parenting need is
conceptualized as a “biological” need which, phylogenetically inherited,
serves as an internal stimulus to behavior.
3. There are, of course, alternative biological interpretations of the female parenting need which,
departing from classical Darwinian theory, one might mention—parental investment theory, for
example (Trivers 1972). I do not discuss this theory here because although our research design does
not permit us to decide whether, as a precultural need, the female parenting need is biologically or
experientially acquired, this does not hold for this particular biological theory which (for reasons
which would require an extensive discussion) does not seem to adequately explain the kibbutz data.
We now can turn to examples of those theories of the female parenting
need according to which this need, though precultural, is experientially
acquired. According to one example, the female parenting need is a
“psychobiological” need which is acquired as a result of psychological
experiences derived from a biological characteristic of the female organism.
Specifically, this need is explained as the motivational consequence of the
girl’s cognitive and emotional reactions attendant upon her psychic
awareness of the structure of her reproductive organs. This theory, as most
prominently formulated by Erikson (1963:91), anchors many sex-linked
needs in what he calls the “ground plan” of the body. Given that boys and
girls have a different “ground plan,” each is characterized by “a unique
quality of [inner] experience.” The experience of girls is different from that
of boys as a function of (among other things) the “inner space” that
characterizes the female reproductive organs. This experience, which is
“founded on the preformed functions” of the “future childbearer,” provides
girls with a motivational disposition for childbearing and hence for
parenting. This thesis has been more extensively developed by Bardwick
(1971:15). It is because of the girls’ creative inner space, so her thesis goes,
that “an anticipatory pleasure and rehearsal of future maternity … looms
large in the girl.”
Perhaps the most influential (and controversial) examples of a
“psychosocial” interpretation of the female parenting need are those
formulated by Freud (1964, ch. 33). For Freud, like Erikson, this need is
acquired as a result of experiences related to female sexual anatomy, but
since for Freud these experiences are social (consisting in the girl’s
interaction with significant others), it seems more accurate to say that for
him the female parenting need, though precultural, is more a “psychosocial”
than a “psychobiological” need.
According to the first of Freud’s hypotheses, if the young girl feels loved
by her mother, then, given the dependency need of children, she develops a
libidinal attachment to and identifies with her. Since, for the growing girl,
the mothers parenting role is her most important characteristic, the girl’s
identification with her mother is the basis for her desire to emulate that role
particularly. To be sure, the earliest identification of the boy is also, and for
the same reason, with the mother; and by this explanation of the girls’
acquisition of the parenting need, one would expect that boys would acquire
a parenting need no less strong. This is exactly what psychoanalytic
theorists like Bettelheim (1954) and others have suggested, a suggestion
which receives support from the fantasy play of the sabra children in which,
it will be recalled, the second most frequent identification of the boys was
with parenting women. Nevertheless, the boy’s identification with the
mother does not persist because, according to these latter theorists, with his
discovery of the anatomical differences between the sexes, he realizes that
he cannot become like her. For Freud, however, it is not because he cannot
become like her, but because his fear of castration leads him to give up his
desire to become like her, that is the crucial factor in the boy’s
disidentification with his mother. On either interpretation, although boys
may subsequently come to envy women for their childbearing function,
they give up their identification with the mother.
For the girl, on the other hand, the discovery of the anatomical
differences between the sexes has a rather different consequence, which
leads to Freud’s second hypothesis. When the girl makes this discovery,
disappointment supersedes her attachment to the mother as the basis for her
parenting need. Viewing herself as having been deprived of a penis, the girl
develops a strong wish to acquire one. Eventually, however, she must
accept the fact that she cannot gratify this wish (just as the boy must accept
the fact that he cannot gratify his wish to bear a child). When, then, the girl
gives up her wish for a penis, she puts in its place a wish for a child, and the
latter wish acquires all the intensity of the former.
The above four theories are among the most prominent examples of
precultural interpretations of the female parenting need. Since, in our
present state of knowledge, there is no way of assessing their relative merit,
we can only say that all of them can account (in principle) for the
precultural existence of this need. But even if all four examples were to be
discontinued, this would not invalidate the conclusion of this study that the
female parenting need is precultural. For if the findings reported here are
reliable, the disconfirmation of the above examples of precultural
interpretations of this need would merely oblige us to search for alternative
interpretations.
On the assumption that the female parenting need is a precultural need,
we can not only explain the counterrevolutionary attitudes of the female
sabras to the family, but we can also explain the vicissitudes of the
revolutionary attitudes of their mothers and grandmothers. For on this
assumption, when the kibbutz pioneers rejected (and physically abandoned)
their biological family of origin, it is entirely understandable (and in
hindsight, at least predictable) that they would have created a sociological
family to take its place. Thus it is that the kibbutz, as we have seen, became
for them a surrogate family, one, however, in which culture took the place
of biological kinship as its basis. Moreover, their repressed parenting need
—the women’s exaggerated expressions of affection for their grandchildren
is evidence for its repression and for the subsequent “return of the
repressed”—was initially satisfied by the maternal attitudes they displayed
to all kibbutz children. In short, although in the early years of the kibbutz
few women performed the role of genitrix, any could (and many did)
perform the role of mater.
But a surrogate family can take the emotional place of the biological
family only until one’s own family of procreation becomes psychologically
important; and on the assumption that the female parenting need is
precultural, this must inevitably happen unless the initial motive for the
repression of this need is transmitted from one generation to the next. In the
kibbutz case, the motive for its repression (whatever it may have been) was
obviously not transmitted to the second generation, for the sabras have not
only established larger biological families than their mothers, but they have
also transferred a significant measure of their familial emotions from the
sociological family (the kibbutz), which had been the focus of the familial
emotions of the kibbutz founders, to the biological family which each has
created herself. This is the process, or so at least it seems to me, by which
the kibbutz has been transformed from one, undifferentiated child-oriented
community to a structurally differentiated community consisting of separate
(though integrated) child-oriented families.
But this is not all. Insofar as the female sabras value parenting as a phase-
specific role in the life cycle, the gratifications they derive from this
“feminine” role obviate the need to strive for status in “masculine” roles.
Confident in and valuing their status in the family domain, their desire,
however, for sexual equality in extrafamilial domains has in no way
diminished, although it has taken a different form from that desired by
women who disvalue the maternal role. Instead of seeking “status identity”
with men in a system of sex-role uniformity, the sabras seek “status
equivalence” in a system of sex-role differentiation. It is all the more
significant, therefore, that although many of them have been frustrated in
this attempt by the narrow range of occupational opportunities available to
women, they have neither abandoned their familistic orientation, nor have
they attempted to reinstate the pioneers’ “identity” meaning of sexual
equality.
These kibbutz findings, if I may be permitted a personal note, forced
upon me a kind of Copernican revolution in my own thinking. When I
returned to Kiryat Yedidim in 1975, I realized that my understanding of
what I thought I had been doing in the kibbutz in 1951 was very different
from what I found myself doing in 1975. As a cultural determinist, my aim
in studying personality development in Kiryat Yedidim in 1951 was to
observe the influence of culture on human nature or, more accurately, to
discover how a new culture produces a new human nature. In 1975 I found
(against my own intentions) that I was observing the influence of human
nature on culture; alternatively, I was observing the resurgence of the old
culture (in modern garb) as a function of those elements in human nature
that the new culture was unable to change. If this is so, then what is really
problematic about the data I have presented is not the feminine
counterrevolution of the sabras, but the feminist revolution of their parents
and grandparents. For if, as these data suggest, many of the motivational
differences between the sexes are precultural, and if, moreover, these
differences are more or less accurately reflected in the system of sex-role
differentiation presently found in the kibbutz (and in almost every other
human society), then the challenge for scientific inquiry presented by the
kibbutz experience is not why the sabras, in their system of sex-role
differentiation, conform to “human nature,” but why the kibbutz pioneers
had attempted to undo it. Since, however, a nonspeculative answer to this
question requires historical data which I do not command, and since in any
event the question is best answered by a study of contemporary movements
in the West that are making the same attempt today, there would be little
gain in offering a speculative answer. Instead, I wish to turn to some of the
broader issues implicit in the kibbutz experience.
 Unlike cultural theories, which attribute sex differences to sexually
appropriate role modeling, our analysis of the kibbutz data has suggested
that the obverse is closer to the truth; that is, sexually appropriate role
modeling is a function of precultural differences between the sexes. Implicit
in this difference between cultural and precultural interpretations of the
motivational bases for role modeling is an even more important difference
with respect to the origin and persistence of systems of sex-role
differentiation. Since, according to cultural interpretations, there are no
precultural differences between the sexes, it follows that sex-role
differentiation is itself culturally determined. Hence, it is just as feasible for
social systems to be constructed on (or to evolve into) a “plan” of sex-role
uniformity as of sex-role differentiation. According to precultural
interpretations, however, the former alternative is not feasible, for the
precultural motivational differences between the sexes render it highly
probable that these differences will inevitably be institutionalized in some
type of sex-role differentiation.
Of course, the content of any system of sex-role differentiation is
culturally constituted, so that such systems can—and many do—become
ossified and exploitative. If, then, as a reaction to such a situation, a
particular system were to be abolished, it is highly likely, as the kibbutz
experience suggests, that another, albeit nonexploitative system, would take
its place. For if many sex differences in motivation are precultural, then
systems of sex-role differentiation not only create sex differences in
motivational dispositions, but they also constitute important
institutionalized means for the expression and gratification of these
precultural dispositions. Lest I be misunderstood, I should like to make
explicit some of the implications of this conclusion.

1. To say that sex-role differentiation is a consequence of sex differences

in precultural needs does not imply that all differences in sex roles are
a result of these differences; this inference is both theoretically
untenable and empirically false. Moreover, to say that the sexes differ
in precultural needs is not to say that they differ in all precultural
needs, nor is it to say that they differ only in precultural needs, for both
statements, again, are theoretically untenable and empirically false.
2. To say that sex-role differentiation, as such, has its origin in sex
differences in precultural needs is not to say that sex roles are
 themselves precultural in origin. Any system of sex-role differentiation
is a culturally constituted system; that is, it consists of a set of rules
and norms which, viewed as cognitive messages, inform social actors
of the appropriate behavioral means by which their needs may be
gratified. This being so, although the motivation for performing certain
sex roles may stem from a desire to gratify needs, their performance is
governed by cultural rules and norms.
3. To say that the performance of some sex-roles gratifies precultural
needs (among others) does not imply that sex differences in these
needs are differences in kind; rather (as the evidence from sabra
children demonstrates) they are typically differences in degree. This is
especially true of those needs whose expression and gratification are
institutionalized in sex-role systems. Hence, the fact that such systems
tend to classify social roles categorically as either male or female does
not mean that sex differences in precultural needs are categorically
different. On the contrary, so far as these needs are concerned, human
beings are most probably bisexual. The behavior of sabra children
indicates that both sexes share the same needs, the differences between
them consisting of differences in the strength of these needs.
Nevertheless, although the differences are in degree, rather than in
kind, if the sex-role system does not recognize these differences, then,
as the kibbutz data suggest, the social actors will eventually change it.
There is, however, another side to this coin. Whether they are
genetically or experientially acquired, it often happens that a reversal
occurs in the relative strength of precultural needs. Some males, for
example, may exhibit an especially strong parenting need, while some
females may exhibit a relatively weak one. This being the case, we
may expect that in any society there will be a certain percentage of
social actors for whom the culturally appropriate sex roles are
psychologically inappropriate. If, then, inflexible boundary rules deny
these actors access to the complementary set of sex roles found in their
society, or if they are not provided with alternative roles, we may also
expect that such actors will exhibit psychological dislocations which,
in the absence of relevant structural changes, will lead to sociological
dislocations.
4. From the last point it follows that, as a principle of social policy, no
social role should be barred to any person on the grounds that his or
 her recruitment is inconsistent with the current system of sex-role
differentiation. In short, no individual or group of individuals should
be prohibited from achieving sexual equality in the “identity” meaning
of equality. If, however, our findings are reliable, attempts to correct
the inequities in any particular system of sex-role differentiation
should most effectively be addressed to the achievement of sexual
equality in its “equivalence” meaning, for it is the latter meaning of
equality that is important for most people to achieve. Hence, for any
group of individuals to attempt to impose their particular reversal of a
panhuman distribution in sex differences upon others, is an insult to
their basic human dignity. If, moreover, the political or media
influence of such a group assures their attempts a measure of success,
the ensuing social and psychological dislocations for the larger society
can be expected to be as serious as those attendant upon the reverse
kind of straightjacketing (except that in the latter case the
consequences are felt only by a minority). For if systems of sex-role
differentiation, as such, are in large part a function of sex differences
in motivational disposition, attempts to convince women that sexual
equality, for example, is worthwhile only in the “identity” meaning of
equality, and that “feminine” careers—even if they achieve equality in
its “equivalence” meaning—are unseemly pursuits imposed on them
by a sexist society, may (if successful) deprive them of important
sources of human gratification. Moreover, to the extent that some
women are persuaded by this ideology, but continue to be motivated
by powerful countervailing needs, the resulting inner conflict may
lead, as one psychiatric study has shown (Moulton 1977), to painful
feelings of guilt and depression.

Single cases prove little; they are primarily useful insofar as they
challenge received opinion. The kibbutz case does not prove the existence
of precultural sex differences. Rather, it challenges the current intellectual
and political pieties which deny the existence of such differences (just as
they deny the existence of other group differences) on the grounds that to be
different is ipso facto to be unequal. That individuals and groups must be
identical in order to be equal is surely one of the more pernicious dogmas of
our time, and the fact that, ironically enough, it has become a liberal dogma
does not make it any the less so. Until or unless the kibbutz data are
interpreted differently, the kibbutz case constitutes a challenge to this
dogma so far as sex differences are concerned. Of course, the strength of
this challenge cannot be determined without much more extensive research
—especially longitudinal research—in a variety of cultural settings. Until
then, prudence suggests that scientific formulations and public policies
related to sex differences proceed with caution.

Note
Reprinted from Melford E. Spiro, Gender and Culture (Durham, N.C.: Duke University Press,
1979), pp. 97–110. © 1979 by Melford E. Spiro.

References

Bardwick, Judith M. 1971. Psychology of women. New York: Harper

and Row.
Bettelheim, Bruno. 1954. Symbolic wounds. Glencoe, 111.: The Free
Press.
Erikson, Erik H. 1963. Childhood and society. 2d. ed. New York: W.
W. Norton.
Freud, Sigmund. 1964. New introductory lectures on psychoanalysis
[1933]. In The standard edition of the complete psychological
works of Sigmund Freud, vol. 22. London: Hogarth Press.
La Barre, Weston. 1954. The human animal. Chicago: University of
Chicago Press.
Moulton, Ruth. 1977. Some effects of the new feminism. American
Journal of Psychiatry 134:1–6.
Tiger, Lionel, and Joseph Shepher. 1975. Women in the kibbutz. New
York: Harcourt Brace Jovanovich.
Trivers, Robert L. 1972. Parental investment and sexual selection. In
Sexual selection and the descent of man 1871–1971, ed. Bernard
Campbell. Chicago: Aldine.
4
Is the Oedipus Complex Universal?

IN PREVIOUS work I attempted to achieve two separate, but related, goals.

First, I attempted to demonstrate that its empirical foundation is too weak to
support Malinowski’s argument that a matrilineal complex exists in the
Trobriands. Second, I attempted to show that the evidence indicates to the
contrary that an unusually strong Oedipus complex exists in the Trobriands.
Here I wish to examine the implications of these Trobriand findings for
human societies in general.
Malinowski and those who follow his lead have argued that the male
Oedipus complex is culturally relative because it is produced not by the
boy’s experience in the family in general, but by his experience in the
“patriarchal” family uniquely. Since, so they argue, it is the authoritarian
father—the nineteenth-century European father being paradigmatic of the
class—who arouses filial hostility, matrilineal societies whose family
structure does not conform to the “patriarchal” type—the Trobriands being
paradigmatic of the class—would not be expected to produce an Oedipus
complex.
In a sophisticated and perceptive explication of that argument, Campbell
and Naroll (1972:437) quite properly observe that in principle the boy’s
hostility to the father may be motivated either by his rivalry with him for
the love of the mother (as Freud claimed) or by the father’s punitive
authority (as Malinowski claimed). Freud, they argue, confounded these
two motives in his construction of the Oedipus complex because, in his
European patient population, the father was both the authority figure and
the mother’s lover. Since in the Trobriands, however, these roles are
performed by different persons rather than by one and the same person, and
since the boy’s hostility, according to Malinowski, is directed to the
authority figure (mothers brother) rather than the mother’s lover (father),
the hostility dimension of the Western Oedipus complex is brought about
not by the son’s sexual rivalry with the father, they argue, but by his
resentment of the latter’s authority.
That argument can now be faulted on three grounds. First, there are no a
priori grounds for assuming that hostility to the father cannot be motivated
by both sexual rivalry and resentment against his authority. Indeed, that is
the view I have adopted in distinguishing between Oedipal and non-Oedipal
grounds for filial hostility. Second, there are no evidential grounds for
claiming that in the Trobriands the jural authority of the mother’s brother is
exercised either frequently or punitively. Hence, even if the boy’s conscious
hostility is directed toward the mother’s brother, it cannot be a function of
the latter’s punitive authority. Third, the findings I have presented
demonstrate rather conclusively that in fact it is the father who is the prime
target of the boy’s hostility in the Trobriands, and moreover his hostility is
Oedipal in motivation, i.e., it is motivated by his rivalry with him for the
love of the wife-mother.
Even, however, without that empirical demonstration, the claim that the
male Oedipus complex would not be expected in matrilineal societies
because the father is not an authority figure can be faulted on theoretical
grounds alone. In order to sustain that claim, it would have to be
demonstrated that in matrilineal societies it is the case not only that the
father is not an authority figure for the son (however important that might
be for the non-Oedipal dimensions of the father-son relationship), but that
the mother is not a love-object for him. For if it is the case that the son has a
libidinal attachment to the mother in matrilineal as well as in “patriarchal’’
societies, the contention that he would nevertheless not be hostile to the
father—or anyone else that he perceived to be the rival for her love—would
be warranted only on the assumption that hostility to a rival is a
phenomenon which is restricted to patriarchal societies. That assumption,
however, is easily refuted, if only by the abundant evidence to the contrary
reported by Malinowski for the Trobriands. That being the case, unless it
were demonstrated that a libidinal attachment to the mother is restricted to
boys in patriarchal societies, then, all things being equal, it would be no less
likely for the Oedipus complex to be found in matrilineal societies than in
patriarchal ones.
Malinowski, who waffled on this issue of the boy’s relationship to the
mother—on the one hand he claimed that in the Trobriands the young boy
has a “passionate” attachment to his mother, on the other that this
attachment disappears spontaneously prior to the normally expected onset
of the Oedipus complex, a point to which we shall return—nowhere
adduced any structural feature(s) peculiar to matriliny that might account
for the absence of a libidinal attachment to the mother in matrilineal
societies. Unless such a feature can be identified, it seems judicious to go
along with the large body of evidence which suggests that a motivational
disposition to nuclear family incest in general, and to mother-son incest in
particular, is a panhuman characteristic. Since Lindzey (1967) has brought
together much of the evidence pertaining to incest in general, our discussion
of that topic will take its departure from his excellent paper, following
which we shall turn to mother-son incest, which is our primary concern
here.
The evidence marshaled by Lindzey can be classified into two categories,
indirect and direct. Beginning with the former, we might mention in the first
place the abundant social-psychological findings which suggest that
“personal attractiveness and interpersonal choice are mediated, or
determined, by similarity in attitudes, values, needs, and background
factors,” and that “positive social choice is strongly facilitated by physical
or geographic proximity” (Lindzey 1967:1056). When these social-
psychological findings are combined with the sociological findings
concerning homogeneity of mate selection and the demographic findings
concerning assortative mating (which indicate that the grounds for
preferential choice in nonsexual domains apply to the sexual domain as
well), their cumulative force suggests that in the absence of countervailing
factors (most notably, incest taboos) sexual choice within—but not
restricted to—the nuclear family would be a likely outcome.
That suggestion is supported by the direct evidence regarding the
motivational disposition to nuclear family incest. Consider, first, the
psychiatric findings concerning the existence of incestuous wishes at least
in clinical populations. Consider, again, the psychological findings
concerning the frequency of incestuous wishes, both overt and covert, in the
dreams of normal populations. Consider, moreover, the anthropological
findings concerning the near-universal incidence of the incest motif in
myths, legends, and folktales, as well as the universality of nuclear family
incest taboos. Consider, in addition, the sociological findings concerning
the prevalence of incestuous behavior in the United States. For example,
one out of every twenty persons questioned in 1970, according to an
authoritative estimate, had had an incestuous experience (Justice and Justice
1979:17). Consider, finally, the ethological findings (Bischof 1975) which
indicate a near-universal motivation to incest in infrahuman mammalian
societies.1 If, then, we consider the cumulative force of these various kinds
of evidence, it is hard to avoid the conclusion that the existence of a
panhuman motivational disposition to incest is a highly probable
hypothesis. This hypothesis, I hasten to add, does not entail the conclusion
that nuclear family members are the strongest objects of sexual desire—
though that too may sometimes occur—but only that they, among others,
are members of that class.
1. Despite their motivational disposition, incestuous behavior is infrequent in mammals because of
a variety of structural constraints—most notably the formation of all-male adolescent groupings,
adolescent extrusion, and dominance structures—which are summarized by Bischof (1975) in his
admirable survey of the mammalian evidence. Although Bischof, a zoologist, shows that in the
absence of these structural arrangements incest would be prevalent in mammalian societies, and
although he argues that these arrangements are the evolutionary result of a selection process to
constrain its occur-rence, he then, strangely, concludes that the infrequency of mammalian incest
indicates the absence of motivation to incest. It is hard to account for this non sequitur in an
otherwise exemplary study.
Although almost all anthropologists, representing the entire range of
anthropological thought, accept this conclusion, it has always had its
opponents. If, so their counterargument goes, incestuous behavior is rare in
human societies, it is not because of social or cultural pressures, most
notably the implementation of incest taboos, which lead to its inhibition (as
the proponents of the incest hypothesis argue), but because (to quote a
recent statement of a distinguished critic) “human beings are naturally’ non-
incestuous” (Fox 1980:14). (Fox’s view is much more complex than this
quotation suggests, as will become apparent when we return to it below.)
Hence, before turning to the motivational disposition to incest in the
mother-son dyad (our concern here) it is necessary to examine this
counterargument, especially since prominent findings from China and Israel
have recently been adduced on its behalf.
Various of its proponents base the counterargument to the incest
hypothesis on different theoretical grounds, but they all agree with
Westermarck (who first proposed it) that sexual indifference or aversion,
rather than attraction, develops between any persons (including family
members) who live together from an early age. It is this naturally
developing aversion that then accounts for the relative infrequency of
incestuous behavior (Westermarck 1906–8:vol. 2, p. 368). In order to fully
grasp this counterargument it must be emphasized that none of the
proponents of the incest hypothesis argue that—except for young children
—members of the nuclear family have a conscious wish to take one another
as sexual partners. They, no less than Westermarck, are fully aware that,
consciously, most individuals are either indifferent or aversive to having
sexual relations with family members. For them, however, sexual
indifference or aversion is a consequence not of childhood propinquity but,
rather, of the internalization of incest taboos and the consequent repression
or extinction of the incestuous wishes.
Although Westermarck’s theory has been persistently rejected by social
scientists, most particularly on the grounds—first adduced by Frazer
([1887] 1910:vol. 4, p. 97 ff.)—that the universality of nuclear family incest
taboos implies that the motivational disposition to incest is also universal, it
has recently been defended on the basis of two sets of findings, one from
China, the other from Israel. Since these findings, both confined to the
question of sibling incest, have been widely heralded as constituting
conclusive support (cf. Bischof 1975; Demarest 1977; Fox 1980: chapter 2;
Money 1980; Wilson 1978) for West-ermarck’s theory, it is important to
examine them rather carefully.
The Chinese case, which has been extensively studied by Wolf (1966,
1968, 1970) and by Wolf and Huang (1980), concerns a type of marriage
known as simpua marriage. In this marriage type, a boy’s parents choose as
his future bride a young girl, often orphaned or from a poor and socially
inferior family, who is adopted as a daughter, who lives in the family
household throughout childhood, and whom he subsequently marries at the
appropriate age. In short, prior to their marriage the structural relationship
between the boy and his chosen bride is little different from that of
biological siblings raised in the same family. According to Wolf’s findings,
the adultery and divorce rates found in simpua marriages are higher and the
fertility rate is lower than those found in regular Chinese marriages. Taking
these rates as a measure of sexual dissatisfaction, Wolf argues that the
greater sexual dissatisfaction in simpua marriages is a consequence of the
sexual aversion that develops between the boy and girl in childhood, which
supports West-ermarck’s theory that propinquity leads to sexual aversion.
Convincing as it might seem, Wolf’s argument, I believe, is invalid on
both empirical and theoretical grounds. Empirically, Wolfs findings
regarding simpua marriage do not sustain the conclusion that children who
live together typically develop a sexual aversion for each other, for although
these marriages are characterized by higher rates of divorce and adultery
and a lower rate of fertility than regular marriages, it is nevertheless the
case that the great majority of them do not display these characteristics. If it
is then claimed that these findings indicate that there is a greater likelihood
for childhood propinquity to lead to sexual aversion, even this lesser claim
can be sustained only on the dubious assumption that adultery, divorce, and
low fertility are necessarily (as Wolf takes them to be) measures of sexual
dissatisfaction. Surely, other factors also contribute to these forms of
behavior in China—in regular and simpua marriages alike—as we know to
be the case elsewhere.
But even accepting the assumption that sexual dissatisfaction is the sole
determinant of divorce, adultery, and low fertility, Wolfs ethnographic data
cast considerable doubt on his contention that the postulated sexual
dissatisfaction found in simpua marriages is primarily determined by the
sexual aversion that, putatively, the couple develop as a result of their living
together in childhood.
Consider in the first place that the boy, according to Wolf, feels cheated
and frustrated by a simpua marriage because it deprives him not only of
honor and prestige—these marriages are viewed as “vulgar and inferior”
and are therefore “socially despised”—but of a dowry, affinal alliances, and
other advantages of a regular marriage, as well. Considering all of these
disadvantages, it is hard to credit Wolfs contention that although the boys
“resent their having their best interests sacrificed by their parents,” their
resentment is nevertheless “not likely to disrupt permanently their
relationship as husband and wife” (Wolf 1970:506). I would assume, on the
contrary, that their resentment would have that effect precisely.
Consider, again, that a girl adopted for a simpua marriage is an object of
“abuse” by her adoptive family; indeed, such girls are treated so badly that
they are the very “symbol of the life of misery.” Consider, too, that
(presumably as a result of this treatment) the girl is hostile to the members
of her adoptive family, including her “brother” (and future husband), of
whom she is jealous and toward whom she displays “sibling” rivalry. It is
again hard to credit Wolf’s contention that these factors do not importantly
affect the girl’s subsequent relationship with the boy when she becomes his
wife. It is also difficult to believe that the girl’s abusive treatment by his
parents does not affect the boy’s perception of her as an inferior person, one
who is unworthy of esteem and affection. Indeed, since the main reason for
a simpua marriage is the wish of the boy’s mother to have a subordinate
daughter-in-law, one who will not be a rival for her son’s affection (as is the
case in regular marriages), it would seem not unlikely that she goes out of
her way to prevent her son from establishing an affectionate relationship
with his “sister.”
In short, given all of these considerations, it is hard to credit Wolf’s claim
that the boy’s “resentment” and the girl’s “misery” have no influence on
their feelings for each other, and that these feelings, in turn, have no effect
on their subsequent relationship as husband and wife, most especially their
sexual relationship. Indeed, everything that we know about the influence of
emotional attitudes on sexual desire and performance supports the contrary
assumption. Hence, even if it were the case that the differences in the rates
of fertility, divorce, and adultery in regular and simpua marriages are
exclusively determined by a higher incidence of sexual dissatisfaction in the
latter marriages, it is hard to believe that all of these social and cultural
factors have no influence on that dissatisfaction. Indeed, given those factors
it is a wonder that these marriages do not display a much greater degree of
dissatisfaction.
When, then, to these social and cultural impediments to a satisfactory
sexual relationship that are contained in Wolf’s data we add the observation
that in simpua marriage the boy and girl must marry each other even if,
being sexually unattracted, they would not have married had they been
raised separately; and when to that we add the additional observation that
almost everywhere divorce in early marriages is higher than in later ones, so
that the higher divorce rate of simpua marriages (in which twice as many
couples marry before seventeen than in regular marriages [44 as against 22
percent]) would be expected as a function of the couple’s age at marriage—
when these observations are also taken into account, I would then submit
that Wolf’s contention that the case of simpua marriage proves
Westermarck’s theory that childhood propinquity leads to sexual aversion
rests on a very shaky foundation.
Let us then turn to the second ethnographic case that allegedly proves
Westermarck’s theory, the Israeli kibbutz movement. In 1958 I reported that
in the kibbutz I had studied none of the children who had been reared
together from birth had married each other, and that to the best of my
knowledge none had had sexual intercourse with each other (Spiro
1958:347–48). These findings were replicated in a later study by Talmon
(1964) in three kibbutzim. Still later, based on a study of the marriage
records of kibbutz children in all kibbutzim—2,769 marriages in all—
Shepher (1971) reported that in not one case had there been a marriage
between children reared together from birth through the age of six years.
Shepher also reported that so far as he was able to ascertain, this finding
also applied to love affairs, those at least that were publicly known.
Now what makes these three reports of considerable interest is that no
marriages (and probably no love affairs) have taken place among members
of the same peer groups despite the fact that such marriages are not
prohibited. It is understandable, therefore, that these findings have aroused
a great deal of attention.
I interpreted my own findings as indicating that members of the kibbutz
peer group repress their sexual feelings for each other. This interpretation
was based on two sets of data. First, with one exception they themselves
attributed the absence of marriage within the peer group to the sexual
indifference which, they explained, arose from their perception of each
other as “siblings.” Second, the data regarding sexual behavior and
socialization in these peer groups, to which I shall advert below, suggested
rather strongly that their sexual indifference did not develop spontaneously.
Some (but hardly all) incest theorists, especially those acquainted with
Shepher’s subsequent large-scale report, interpreted the kibbutz findings, as
did Shepher himself, as supporting Westermarck’s theory that children
reared together in childhood (whether or not they are siblings) develop a
sexual aversion for each other. Since the latter interpretation, however, was
offered without giving any consideration to (what I believe to be) the
crucial data on sexual socialization, in order to evaluate these competing
interpretations it is necessary to summarize those data for the particular
historical period—because much has changed since then—for which the
marriage findings were reported (for details see Spiro 1958: chapters 9, 11,
13, 14).
In that period, consistent with the kibbutz ideology of sexual freedom,
young children were almost entirely free to engage in sexual play without
interference or punishment by their caretakers. Since, therefore, boys and
girls of the same age not only lived together in one dormitory, but also slept
and showered together and had frequent other opportunities to see each
other in the nude, it is not surprising that they also engaged in (childlike)
sexual behavior and that they displayed little sexual shame (Spiro
1958:219–28). An important change in their behavior occurred, however,
around eleven or twelve, when girls, who were beginning to show the first
signs of puberty, refused to shower together with the boys, and in general
began to display overt signs of sexual shame. At the same time, sexual
behavior no longer occurred, and was replaced by a great deal of bickering
and hostility between the sexes. Although the bickering and hostility
gradually disappeared, sexual shame, including the avoidance of mixed
showers, persisted throughout high school, and there was no return to the
sexual play that characterized early childhood, nor was there an assumption
of sexual behavior of a more mature form.
With this brief behavioral description, we may now examine the cultural
and structural factors associated with these behavioral changes. First,
despite their differential physical maturation at puberty, boys and girls of
the same peer groups continued to live and sleep together. This meant that
physically immature boys continued to be the roommates of girls who may
have begun to menstruate, and whose secondary sexual characteristics—the
development of breasts, the growth of pubic hair, and the like—were
becoming prominent. Hence, although the girls were beginning to
experience the sexual tensions of puberty and a concern with the physical
changes in their bodies, the boys remained physically immature. It is not
surprising, then, that the girls displayed no interest in the boys, and that in
some cases the latter were put off—even frightened—by the girls, while
others (outsiders) were sexually aroused by them.
Even at a later age, however, when the boys’ maturation caught up with
the girls’, two other factors intervened. First, in the case of the older
children—beginning around prepuberty and continuing until the end of high
school—the permissive ideology of the kibbutz regarding the sexual play of
young children was replaced by a strong prohibition on sexual behavior.
The introduction of such prohibitions at this age was not based—not, at
least, officially—on sexual puritanism, but on the assumptions that sexual
behavior disrupts the learning process, that intellectual development is
enhanced when the sex drive is sublimated in intellectual pursuits, and that
pair bonding, which often results from sexual behavior, interferes with the
intensive group interaction and group identification that the kibbutz viewed
as paramount values. Hence, at this age strong pressure was exerted on the
children to defer all sexual encounters until graduation from high school,
and especially to avoid the formation of permanent liaisons. The latter in
particular were the object of strong social sanctions.
Second, despite these sexual prohibitions and their attendant sanctions,
kibbutz educators believed that to encourage children to develop a
wholesome attitude toward sex and a “natural” attitude toward the body and
its functions, it was important that boys and girls live together not only in
early childhood but throughout their educational careers until graduation
from high school. Beginning, however, in the seventh grade the group of
approximately sixteen age peers, who had previously shared one barrack-
like dormitory room, was divided into groups of four—two males and two
females—each group sharing one bedroom.
Here, then, I would submit, is a classic example of incompatible
demands. On the one hand, we have a group of teenagers, at a physiological
developmental stage of maximum sexual tension, who are exposed to
persistent sexual stimulation induced by living in close quarters with
members of the opposite sex, who dress and undress in one another’s
presence—though it was expected that they avert their eyes during this
process—and who sleep in adjacent beds. At the same time, these same
teenagers are expected to comply with a cultural norm which prohibits
sexual behavior between them on pain of serious social sanctions. Such a
contradiction, I would submit, can only result in intolerable conflict and
unbearable sexual frustration. Edith Buxbaum, a child therapist, highlights
this contradiction in her discussion of a fifteen-year-old kibbutz boy who
could not sleep because of his urge to touch a girl, and who was advised to
seek psychiatric help for his “problem.” It seems paradoxical. Buxbaum
(1970:286) writes
that people should consider it abnormal for a fifteen-year-old boy to want to touch a girl with
whom he sleeps in the same room. Indeed, it would be extraordinary if he did not want to. Yet,
this is what kibbutz educators expect and the children expect of themselves. They are not
supposed to have these feelings, or if they have them, they are not supposed to act on them.

That, however, only a small percentage of these teenagers required

professional help to cope with the tensions induced by this contradiction
suggests that the majority managed to erect strong psychological barriers
against them. These barriers, as I suggested above—and as Buxbaum
(1970:285–90) and Bettelheim (1969:235–40) have suggested elsewhere—
could only have consisted in the repression of their sexual wishes (which
would explain how they might have consciously become sexually
indifferent to each other) and, if that were not sufficient, in the formation of
a reaction against them (which would explain how they might have
consciously developed a sexual aversion for each other).2
2. If correct, this analysis confutes Fox’s attempt to salvage Westermarck’s theory. The latter
theory, according to Fox (1980:chapter 2), only applies to children who not only live together but
also engage in intense physical interaction. Since, he argues, their physical contacts lead to sexual
arousal, and since immature children have no means of physiological discharge, their painful
frustration leads to a consequent sexual aversion for the children (later adults) who are responsible
for their pain. If, however, children who live together have no physical interaction and, therefore, no
history of painful sexual frustration, Westermarck’s theory does not apply. The kibbutz case,
according to Fox, supports this argument because it involves physical interaction in childhood
followed by aversion at adolescence.
According to my analysis, however, it is not the physical interaction in childhood, but the sexual
prohibition at adolescence in the face of powerful sexual stimulation that leads to sexual aversion in
the kibbutz. By my argument, the sexual stimulation leads to sexual arousal, which is frustrated by
sexual prohibitions, and the resulting painful tensions are defended against by repression and reaction
formation which, if effective, lead to sexual aversion.
There are three reasons, I would submit, for preferring this interpretation
of these kibbutz findings to that of Westermarck’s followers. First, the early
sexual play of kibbutz children for the period under discussion was replaced
by sexual abstinence only after the sexual permissiveness of childhood was
replaced by the sexual prohibitions of adolescence. This suggests, pace
Westermarck, that their sexual aversion for each other—if that is what it
was—did not develop endogenously. Second, the sexual abstinence of the
adolescents applied not only in regard to the members of their peer group—
the group with whom they lived as children—but to other groups as well—
those with whom they did not live as children. Third, and most important,
this interpretation is supported by the recent findings of Kaffman (1977), a
psychiatrist employed by the kibbutz movement, regarding sexual behavior
in the kibbutz.
Although Kaffman’s findings regarding peer-group marriage do not
disagree with the early findings reported above, those regarding love affairs
are at variance with them. “There is hardly a kibbutz,” Kaffman writes,
‘without its report of heterosexual relationships between adolescents
brought up together from infancy” (Kaffman 1977:216). Such relationships,
he claims, “may not be typical, but [they are] not all that rare either” (ibid.).
Kaffman also reports that whereas in an earlier study of seventeen-year-old
kibbutz children, 66 percent of the sample were opposed to sexual relations,
in 1973 only 7 percent of the males and 11 percent of the females voiced
opposition.
How, then, are the discrepancies between Kaffman’s report and the
earlier reports to be explained? And what light do these discrepancies shed
on the conflicting interpretations of the earlier reports? Kaffman’s findings
are based on studies conducted some few years after two major changes had
begun to take place in the kibbutz movement. In the first place, many
kibbutzim changed the living arrangements in the high school dormitories
from mixed to unisexual bedrooms. In many of them, too, the prohibition
on teenage sexual behavior has been informally if not formally abolished.
According, then, to my interpretation of the earlier findings, Kaffman’s
findings are exactly what one would expect from such changes: no longer
suffering the severe sexual tensions aroused by a contradiction between
their living arrangements and sexual norms, adolescents in those kibbutzim
which had undergone these changes had no need to create psychological
barriers (repression and/or reaction formation) to their sexual feelings. As
more and more kibbutzim make these structural and cultural changes, we
would then expect the sexual changes reported by Kaffman to become more
widespread. It is difficult, however, to accommodate Kaffman’s findings to
Westermarck’s theory. Since, according to that theory, sexual aversion
results from joint living in childhood, and since no changes in the children’s
living arrangements were associated with the sexual changes reported by
Kaffman, his data are clearly inexplicable by that theory.
If, then, my interpretation of the earlier reports is correct, why is it that
the structural and cultural changes pointed to above have not had an even
greater effect on the incidence of intragroup love affairs and marriage?
There are at least two answers to that question. In the first place, the fact
that individuals who live together in childhood (whether they are siblings or
nonsiblings) acquire sexual wishes for each other does not mean, as I
observed above, that these are their strongest sexual wishes. When we
consider, then, that kibbutz peer groups, as Kaffman observes, are small and
the range of sexual choice limited, and when we consider too the different
maturation rates of boys and girls, we would not expect a surge of sexual
behavior within the per group despite those changes. In the second place,
since sexual attractiveness is hardly the only basis for marriage, we would
not expect all couples who are sexually attracted to each other to wish to
enter into a marriage.3
 3. Those incest theorists—the so-called alliance theorists—who see the unchecked motivational
disposition to incest as a barrier to the formation of interfamily alliances have primitive and
prehistoric families as their model. Since the latter (hunting and gathering) families are frequently
isolated from other families in their bands, mating with a family member is much more likely—
unless it is checked by incest taboos—than in families in which a wide range of sexual partners, in
addition to family members, are available. Physical isolation of the family, as we have previously
observed, is an important condition for incest in modern families as well.
I would conclude, then, that the kibbutz case no more than the Chinese
constitutes proof for the Westermarck theory. Indeed, if my analysis is
correct, the kibbutz case provides strong support for the contention of the
vast majority of incest theorists that individuals who live together (whether
they are family or nonfamily members) do indeed develop and retain sexual
feelings for each other unless they are inhibited by countervailing social and
cultural pressures.
This analysis of the Israeli case is supported by yet another case, one
which was reported only some few years following the claim that the
kibbutz and Chinese cases support the Westermarck theory. This case,
which has already been mentioned in a previous chapter, consists of
brother-sister marriage in Roman Egypt. For two centuries, brother-sister
marriage, according to Hopkins (1980:310), was a “frequent practice” in
Roman Egypt. According to second century census returns, for example, 23
of 113 recorded marriages were between brother and sister. More important
for the terms of this discussion is that these marriages, which were
considered to be entirely “normal”—hence, not prohibited by an incest
taboo—were based not on economic or other practical considerations, but
on “love and sexual passion.” Hence, when such marriages came to an end
in the third century, it was not because love and passion between brothers
and sisters came to an end, but because Egypt came under Roman law,
which prohibited incest.
If, then, the Chinese case does not refute the generalization that siblings
who are reared together develop incestuous feelings for each other, and if
the kibbutz and Egyptian cases offer strong support for that generalization,
we may now turn to the question of incestuous feelings in the dyad that is
relevant to the Oedipal problem, the mother-son dyad.
There is nothing in the evidence summarized above regarding the
existence of a motivational disposition to nuclear family incest which
suggests that the boy’s incestuous feelings for his mother are less strong
than those for his sister. Indeed, given the nature of the mother-son bond in
humans, it is reasonable to assume that, typically, his incestuous feelings for
the mother are stronger by far than those for his sister, and that this
generalization holds even when the mother is not seductive, as she is
(according to my argument) in the Trobriands. Moreover, given the nature
of the mother-child bond in humans, the incestuous feelings for the mother
in our species, so I shall argue below, are much stronger than those found in
any other species. In short, contrary to most incest theorists, I shall argue
that if the (mother-son) incest taboo represents the transition from the state
of nature to that of culture, it is because the intensity of the incestuous
attachment to the mother is much stronger in human (culture) than in
animal (nature) societies. A comparison of human with infrahuman primate
societies—for convenience’s sake this discussion will be restricted to
primates, our closest animal relatives—can reveal why this is the case.
For obvious biological reasons it is the mother with whom any primate
infant, human or infrahuman, sustains its most intimate emotional and
physical relationship. It is she primarily who suckles, trains, and plays with
him. It is her voice, her body, and her face that the infant knows best, and it
is she whom above all others he seeks out for pleasure, protection, and
comfort. As a consequence, it is the mother who is the primary object of the
infant’s emotional attachment. If, then, even the infant is a sexual creature
and capable of sexual arousal, it is not unreasonable to assume that the early
attachment to the mother has important libidinal overtones, and that the
mother, therefore, is the object of the earliest libidinal attachment. Despite
these similarities between humans and other primates, the incestuous
attachment to the mother is much stronger and especially problematic in
humans because of two important biological characteristics that are unique
to humans: the suppression of estrus and prolonged infantile dependency.
Among infrahuman primates there is an incompatibility between
sexuality and mothering because at the height of estrus, when the females
are in a state of sexual “mania” or “frenzy,” as the primatologists call it,
they have little interest in mothering or, for that matter, in any other activity
except sex. Hence, mothering can only occur during anestrus—the period in
which the female is not sexually receptive—which, apart from the normal
sexual cycle, begins immediately following pregnancy and continues until
the end of lactation. When, at the end of lactation, estrus returns, the
juvenile is already weaned, and since typically infantile helplessness does
not extend beyond weaning, the mother’s indifference to her offspring
during estrus does not endanger their welfare since they are no longer
dependent upon her for care and nurturance. Indeed, at the conclusion of
weaning the juvenile’s relationship with the mother is typically and abruptly
severed. Thus, in those primate societies consisting of mother-child
families, the juvenile is usually driven off by the mother when, at the
conclusion of weaning, she enters estrus, while in those with biparental
families, he is usually driven off by her consort. In either case, he usually
joins a group of juveniles who, with some few exceptions, are permanently
“peripheralized” (as the primatologists say).
Since, then, mother and son are separated following weaning, and since
their separation is socially enforced at least until the son achieves maturity
—when, in a few cases, he may force himself upon the dominant adult
male(s)—there is no opportunity for his early attachment to the mother to
become intensified by a continuing relationship with her. (For a more
extensive treatment, see Bischof 1975; Chance and Jolly 1970; Fox 1980;
Lancaster 1979; Rowell 1972.)
Among humans, the mother-child relationship is very different. Since
infantile helplessness is prolonged, and cultural acquisition complex, it is
necessary that the child remain dependent upon the mother (or mother
surrogate) long after the completion of weaning. And since, with the
suppression of estrus, the human female is not characterized by sexual
mania, her interest in sex, though continuous, does not interfere with her
motivation or ability to care for her young children. Since, then, there is no
incompatibility between mating and mothering among humans, human
offspring live with and remain dependent upon their mothers for many
years, not excluding the “phallic” period (when the early libidinal
attachment to the mother receives strong reinforcement) and puberty (when
sexuality may erupt explosively).
In short, since the son’s relationship with the mother may persist
throughout childhood and even into puberty, thereby intensifying his early
emotional attachment to her, and since that attachment is based on both
libidinal and dependency feelings, it is reasonable to expect the incestuous
attachment of the son to the mother to be much stronger in humans than in
primates. For the same reason, one would expect the attachment to the
mother to be much stronger than the incestuous attachment to the sister.
Although brother and sister live together, they never have—indeed, they are
prohibited from having—the intimate physical relationship that the son has
with his mother.
 Strangely enough, Malinowski (unlike most of his followers) fully
recognized the strength of the boy’s incestuous attachment to the mother, as
well as the critical challenge that it posed for his contention that the
Oedipus complex is confined to “patriarchal” families of the Western type.
Indeed, in a long rhapsodic passage he expatiates in concrete detail on the
erotic feelings of the young boy for his mother (SR:212–14). At the same
time he also observes that since lovers employ the same organ zones and
behavioral modes in their physical interaction with each other that mother
and child employ in their interaction, the male’s later induction into the
“erotic life” may arouse in him “disturbing memories” of the relationship
with his mother, because she “remains in the foreground of [his] emotional
interests throughout his life.” That being the case, Malinowski continues,
before the boy reaches sexual maturity it is necessary that “all sensuality
felt toward the mother become repressed.” That is achieved by inculcating
in the boy emotions of “reverence, dependence [and] respect” toward the
mother so that if a “subconscious temptation of incest” is aroused when he
is mature, it is muted by its “blending” with these nonsexual emotions.
The above analysis represents Malinowski’s conception of what happens
to the early incestuous attachment to the mother in the West (and some
other societies), but it is not what happens (according to him) in matrilineal
societies, in which the “passionate” attachment of the infant to the mother
spontaneously disappears by the time the boy reaches the “phallic” stage of
development, the very stage when genital love for her is normally expected
to appear. Unfortunately, however, his argument on behalf of this claim is
less than convincing. In a nutshell, Malinowski nowhere indicates how or
why the spontaneous disappearance of the boy’s passionate attachment to
the mother occurs. That it occurs, however, he is quite sure on the grounds
(it will be recalled) that, when he asked the Trobriand men whether they
wished to have intercourse with their mothers, they ridiculed the very
suggestion. Such a response, he argued, proves that a “repressed Oedipus
complex” cannot possibly exist in the Trobriands.
Although I have already dealt with this “proof,” it is nevertheless worth
repeating because most other critics of the notion that incestuous motives
might be repressed adduce exactly the same argument. Briefly, the
argument is invalid on three counts. First, although it now seems unlikely
that the boy’s incestuous wishes in regard to the mother are ever—to use
Freud’s metaphor—“smashed” in childhood (as a result of castration
anxiety), it does seem to be the case that in some individuals and in some
societies those wishes do undergo a gradual process of extinction,
coordinate with the cessation of childhood sexuality (sexual “latency”).
Second, even in those individuals and societies in which incestuous wishes
regarding the mother persist into adulthood, it is rare that the adult mother
is the object of these wishes. Rather, it is the mother of childhood—or,
rather, his mental representation of her—on whom the son’s incestuous
wishes remain fixated. In the latter event—to come to the third reason—the
son’s fixation cannot be discovered by the mere asking because, typically, it
is not in conscious awareness, which is precisely what is meant—to use
Malinowski’s expression—by a “repressed Oedipus complex.” To that
extent Westmarck and his followers are correct when they argue that incest
taboos do not have the function of prohibiting incest (with mother or
anyone else), not because incestuous wishes (as they contend) do not exist,
but because typically the implementation of the incest taboos in childhood
has achieved its intended function. In short, from this perspective, the
function of incest taboos is not to signal to adults that their incestuous
wishes must be inhibited, but rather to banish such wishes from children
before they become adults.
If, then, the implementation of the mother-son incest taboo (a subject to
which we shall return) typically results in the disappearance of the
incestuous attachment to the childhood mother—as a result either of
extinction, on the one hand, or of repression and reaction formation, on the
other—to challenge the universality of such an attachment on the grounds
that in some society (or societies) adults do not consciously experience any
sexual feelings for the mother, is clearly misguided. By that criterion, such
an attachment does not exist in any society, whether “patriarchal” or
matrilineal, not at any rate in its “normal” members. In short the evidence
for the universality of a motivational disposition to mother-son incest
remains unaffected by this challenge.
Since, then, that evidence is very strong, and since (as we argued above)
the existence of the incestuous dimension of the Oedipus complex renders
its aggressive dimension all but axiomatic, the only appropriate response to
the question, “Is the Oedipus complex universal?” is “How could it possibly
not be?” Even in those primate societies in which females have no
permanent consorts—the multimale societies of trooping monkeys—and in
which, therefore, there can be no rivalry between son and father, powerful
rivalry nevertheless exists between the peripheralized young males as a
group and the dominant males who monopolize the females (Chance and
Jolly 1970). In single-male primate societies, on the other hand, whether
those of the monogamous gibbon (Carpenter 1940) or the polygynous
hamadryas baboon (Kummer 1968), the rivalry is directly and explicitly
between father and son.
In sum, there is only one obvious retort to the above riposte. If there were
a human society in which mothers did not have male consorts—so that the
son had no adult rival for the love of the mother—in such a society the
Oedipus complex (by definition) would not exist. So far as we know,
however, no human society of that type exists, or has ever existed. The
“matrifocal” households—widely prevalent in the Caribbean (Smith 1956)
and among lower class American blacks (Rain-water and Yancey 1967)—
do not constitute an exception to that generalization because typically a
husband-father is intermittently present or else the mother brings a series of
temporary lovers into the household. That being the case, although there is
no evidence one way or the other, I would also expect these New World-
matrifocal households to produce an Oedipus complex. This expectation is
supported by the findings of Gough (1953) that an Oedipus complex is
found among the Nayar, an Indian caste of the Malabar Coast, despite the
fact that Nayar women typically take a series of lovers rather than living
with a permanent consort.
That the Oedipus complex, according to this analysis, would be expected
to be universal does not imply, I hasten to add, that it would also be
expected to be cross-culturally uniform. On the contrary, since we know
(from Western data) that the Oedipus complex is variable within societies,
we would then expect that it would exhibit a wide range of variability
across societies. That is the thesis I want to explore in the following section.

Cross-cultural Variability in the Oedipus Complex

Since the Oedipus complex may be said to have three important dimensions
—structure, intensity, and outcome—in principle at least it could be
expected to display cross-cultural variability in all three. Let us begin with
its structure. By the “structure” of the Oedipus complex I mean the
members—in addition to the boy himself—who make up the Oedipal
triangle and who are the objects, therefore, of his sexual and aggressive
wishes.
If in the classical Oedipus complex it is the boy’s biological parents who
are the objects of those wishes, that is not to be explained, surely, by the
closeness of their genetic relationship, nor again by some instinctual
vectorial dimension of the sexual and aggressive drives. Rather, that
particular Oedipal triangle is best accounted for by the sociological fact that
the boy and his biological parents, typically constituting a social group and
inhabiting a common household, sustain certain modes of social
relationships with each other. It is these relationships that account for the
biological parents, specifically, becoming the objects of his sexual and
aggressive wishes. Hence, variability in these relationships could be
expected to result in corresponding and predictable variability in the
structure of the Oedipal triangle. Thus, for example, if there were a society
in which, rather than belonging to a common residential household, parents
and children were distributed in different households, it might then be
expected that the child’s mother-surrogate (rather than his biological
mother) would be the focus of his libidinal wishes, and that her consort
(rather than his biological father) would be the focus of his aggressive
wishes. At the moment, however, such a society is not known.
If, then, there is no theoretical reason why the adult members of the
Oedipal triangle must consist of the boy’s biological mother and father, it
might well be the case that in some society this triangle consists of the boy,
his sister, and his mother’s brother. Malinowski’s claim that this is the case
in the Trobriands was rejected not on theoretical, but on empirical grounds:
neither the composition of the Trobriand household nor the social
relationships that obtain within the nuclear family display the characteristics
that might expectably produce that particular structural variant of the
classical Oedipus triangle. For exactly the same reason neither this nor any
other structural variant of the classical Oedipus triangle has been reported at
a total societal level for any other society, which does not mean that such a
variant or variants may not occur in individual cases or in certain subgroups
in some societies, or that some variant may not be reported in the future for
some (as yet unknown) total society.
Although the structure of the Oedipus complex, while variable in
principle, seems to be universal in fact, this is not the case in regard to its
two other attributes—its intensity and outcome—in which cross-cultural
variability is not only a theoretical expectation but an ethno-graphic fact.
Since variability in the intensity of the Oedipus complex has been discussed
elsewhere (Spiro 1982), we shall focus here on variability in its outcome,
especially since the latter attribute has important social and cultural
consequences. Moreover, since the implementation of the incest taboo is a
major determinant of its outcome, we shall limit our discussion to that
determinant alone.
Since libidinal desires for the mother may be present, as we have seen, in
the nursing infant, the implementation of the incest taboo may be said to
begin with weaning, which is also the time when the child is usually banned
from the mother’s bed and when, in general, he is discouraged from
continuing those more intimate forms of physical contact with the mother
that she had previously permitted, if not actively encouraged. The
diminution, if not cessation, of these intimate forms of bodily contact with
the mother during this—the pre-Oedipal—period does not, of course, lead
to the extinction of the boy’s libidinal desires for her, especially since those
desires are intensified when he enters the “phallic” stage of psychosexual
development. At the latter stage, therefore, the mother-son incest taboo is
implemented in all societies by still other means which (in varying degrees)
result in its internalization, i.e., in the acquisition of a motivational
disposition on the part of the boy to comply with the prohibition. These
means and the process of internalization are conceived of differently by
different theorists.
For cognitive theorists, the implementation of the mother-son incest
taboo, like the implementation of any other cultural prohibition, is achieved
by the usual processes of enculturation, i.e., by verbal instruction, including
instruction in the punitive consequences attendant upon its violation. For
social learning theorists that kind of instruction is not sufficient to achieve
the internalization of the taboo unless it is accompanied by various
socialization techniques—techniques of positive and negative
reinforcement—which provide the boy with important incentives for
complying with the taboo. Psychoanalytic theorists, who also stress the
importance of socialization, emphasize the signal importance of castration
anxiety as the motivational basis for compliance with the taboo. As they see
it, the threats (negative reinforcement) which are used to sever his
incestuous attachment to the mother lead to the son’s fantasized expectation
of castration as the punishment for incest with the mother.
Whether it is achieved by the one means or the other, if the
implementation of the incest taboo leads to its internalization, compliance
with the taboo is achieved either by the extinction of the boy’s incestuous
desire for his mother or by its repression, the latter often being accompanied
by a reaction formation against the desire, i.e., by an emotional aversion to
sexual contact with the mother. In sum, if the taboo is internalized, the
boy’s incestuous attachment to the mother either disappears entirely
(extinction) or, although persisting unconsciously (repression), disappears
from conscious awareness. Why the internalization of the taboo results in
the one consequence rather than the other is a question concerning which
there are many, but few satisfactory, answers.
Extinction and repression, however, are not the only possible outcomes
of the incestuous attachment to the mother. If, for example, the
implementation of the incest taboo by any of the processes mentioned
above is only partially successful in promoting its internalization, the son’s
libidinal attachment to the mother is not extinguished, and although it may
be repressed, it undergoes only weak or incomplete repression. In Bengal,
for example, mother and son—so Roy (1975:125) observes—remain
“highly cathected libidinal objects [for] … a lifetime.” (For other parts of
India, see also Carstairs 1956 and Kakar 1978.) Similarly, the typical male
in the Mexican town of San Juan, according to Hunt (1971:129), “has never
been able to transfer his libidinal energies from his mother to an outsider,”
thereby manifesting a “typical Mediterranean pattern.” (For other parts of
Mexico, see also Bushnell 1958.) That such an outcome is a “Mediterranean
pattern” is readily discerned, for example, from Parson’s (1969) description
of the mother-son relationship in Italy (Naples). The same pattern is also
found, however, in East Asia. In Japan, for example, as Tanaka (1981:16)
observes, the mother-son relationship is characterized by the “continuous
presence of unresolved libidinality.” In many societies in which the
incestuous attachment to the mother is incompletely repressed, other means
than those mentioned above are used for the implementation of the incest
taboo, as we shall see below.
Let us first, however, turn from the incestuous attachment to the mother
to comment briefly on the possible outcomes of the boy’s hostility to the
father, the other dimension of the Oedipus complex. Since, as we have seen,
the boy’s hostility to the father sustains a correlative relationship with his
love for the mother, the outcome of his Oedipal hostility is systematically
related to the outcome of his Oedipal love. In short, it is the entire Oedipus
complex whose outcome may variously take the form of extinction,
repression, or incomplete repression.
While all three outcomes may be found in a single society, as we know
from clinical evidence in the West, it is usually the case (as I have been
suggesting) that one of them is dominant. Since, however, the dominant
outcome (as I have also been suggesting) is not the same in all societies, we
may now say that this—the second dimension—of the Oedipus complex is
cross-culturally variable not only in principle but in fact.
Although of some interest in itself, the cross-cultural variability in the
outcome of the Oedipus complex is anthropologically important because its
outcome has social and cultural consequences. Since, then, the differences
in the psychological characteristics of these three outcomes are nontrivial,
the differences in their variable social and cultural consequences are
likewise nontrivial. The latter differences are especially marked when we
compare societies in which extinction and repression are the dominant
outcomes with those in which incomplete repression is dominant.
Operationally defined, a “weak” or “incomplete” repression of the
Oedipus complex is one in which repression is insufficiently powerful to
preclude the conscious arousal of the boy’s incestuous wishes for the
mother (and hence his hostile wishes toward the father) under conditions of
incestuous temptation. Hence, those societies in which incomplete
repression is the dominant outcome of the Oedipus complex are societies in
which the implementation of the taboos on mother-son incest and father-son
aggression by the enculturation and socialization techniques described
above is not entirely successful in achieving their internalization. This being
the case, rather than relying on the boy’s own psychological resources—
extinction, repression, and reaction formation—to ensure compliance with
those taboos, many of those societies achieve compliance by means of
social and cultural resources, as well.
In short, so far as their social and cultural consequences are concerned,
the first notable difference between socities in which extinction and
repression are the dominant outcome of the Oedipus complex and those in
which incomplete repression is its dominant outcome, is that the latter
societies (much more often than the former) practice child extrusion and
painful initiation rites. These customs ensure compliance with the twin
Oedipal prohibitions by reducing the opportunities for incestuous and
aggressive temptation or by strengthening the incomplete repression of the
boy’s sexual and aggressive Oedipal wishes. Let us briefly examine each of
these customs, beginning with child extrusion.
When boys—and in some societies, girls too—are extruded from the
parental household, their subsequent residence, as the cross-cultural record
indicates, is highly variable. It may be an age-graded dormitory, a men’s
house, a children’s village, some other household, a boarding school, or the
like. Whatever the conscious motives for these practices may be—
depending on the society, they include the reduction of the son’s rivalry
with the father (Wilson 1949; Spiro 1958), the prevention of his witnessing
the primal scene (Elwin 1968), economic apprenticeship (Aries 1962),
educational advancement (Gathome-Hardy 1977), the economic advantage
of adoption (Powell 1957), and others—they have the consequence (among
others) of separating the son from his parents, thereby reducing the
opportunities for the arousal of his sexual and aggressive Oedipal wishes.
When, at some later age, more frequent interaction with the parents is once
again resumed, his libidinal and aggressive impulses have typically been re-
channeled. Son extrusion in human societies has its analogue, it will be
recalled, in the peripheralization of male juveniles in primate societies. By
the latter process, the juveniles are deprived of the opportunity of acting
upon their sexual and aggressive impulses toward their mothers and adult
males, respectively.
The second consequence of the incomplete repression of the Oedipus
complex—which, like the first, constitutes a cultural resource for enhancing
the compliance with the taboos on mother-son incest and father-son
aggression—is the practice of painful initiation rites. Although the
conscious, culturally constituted explanations for these rites only
infrequently relate them to the Oedipal issue we have been addressing here,
the ethnographic descriptions of these practices and of the initiates’
psychological reactions to them provide strong evidence for the thesis that
these rites, like child extrusion, constitute an important cultural resource for
ensuring compliance with the taboos on incest with the mother and
aggression toward the father.4
4. For four superb descriptions of these rites and of the emotional reactions of the initiates in New
Guinea, I especially recomment Herdt (1980, 1982), Poole (1982), Read (1965), and Tuzin (1980).
 If, however, child extrusion achieves this end by removing the son from
the locus of Oedipal sexual and aggressive temptation, initiation rites
achieve the same end (according to some theorists at least) by removing
these Oedipal wishes from the son. That is, by hazing, isolation, physical
torture, ordeals, and phallic mutilation (circumcision, subincision, and
superincision), these rites arouse in the boys intense fear and anxiety—
often, in my view, castration anxiety—regarding the father and or the
father-figure initiators (Fox 1980:159), thereby serving to break the boys’
incestuous attachment to their mothers (Hiatt 1971:81) and inhibit their
aggression to their fathers. In sum, these rites, I am suggesting, serve to
strengthen the (incomplete) repression of the boys’ Oedipal (sexual and
aggressive) wishes, and in some cases they might perhaps lead to their
extinction. That these painful and often brutal rites also provide a culturally
sanctioned (and ritually limited) opportunity for men to express their
complementary Oedipal hostility to boys—rationalized, of course, by the
ideology of helping them achieve social and cultural maturity—seems
equally obvious, as Reik ([1919] 1946) suggested some years ago.
I must hasten to say, however, that an Oedipal interpretation of these rites
not only is not obvious to other commentators, but has typically been
rejected by anthropologists (see Langness 1974) and psychoanalysts (see
Lidz and Lidz 1977) alike. Since I believe that rituals, like most other
cultural activities, have multiple meanings, I have little difficulty in
accepting most of the sociological and psychological meanings that other
commentators have attributed to these rites. But I have great difficulty in
ignoring their Oedipal and complementary Oedipal meanings, which, in the
light of recent descriptions (see footnote 4), are just too blatant to overlook.
Indeed, from these descriptions I would argue that societies which practice
initiation rites of the ferocity found in—and perhaps confined to—New
Guinea and Australia are societies in which the incomplete repression of the
childhood Oedipus complex is most pronounced.
As an illustration of their “blatant” Oedipal and complementary
meanings, consider Herdts (1982) description of the boy’s initiation rite
among the Sambia of Papua New Guinea. The first stage of the rite, which
consists of forcible nose-bleeding, administered by the men on boys of
seven to ten years, is a “violent assault whose effects are probably close to
producing authentic trauma,” the boys themselves referring to their fright
by such expressions as “‘I feared they were going to kill me.” If the boys
resist the men and cry, as they often do, the men “have little pity” for them,
and they are “severely dealt with by prolonging the action and thereby
brutalizing it.” Herdt is hardly exaggerating in calling it “an act of raw
aggression.” Similarly, in the third stage, performed for boys thirteen to
sixteen, a line of warriors, appearing as ghosts and enemies, surround them
and, plucking bows and arrows, hooting, and shouting, they again forcibly
nose-bleed the boys, who are now in a state of “terror.”
That this aggression both gratifies the complementary Oedipal complex
of the men and is intended to contain the Oedipal wishes of the boys is
indicated by Herdt’s comments. Nose-bleeding is described by the Sambia
as “punishment” for the boys for their insubordination to their fathers and
elders. That this insubordination includes Oedipal insubordination is
reflected both in the warning given the boys that they may never again so
much as touch, hold, talk with, eat with, or look at their mothers and in the
conscious and unconscious symbolic equation of nose and penis in Sambia
thought and culture. Because of their early and exclusive attachment to
them, Herdt comments, the boys “must be traumatically detached from their
mothers and kept away from them at all costs.” In later stages of the rite,
those following puberty, references to the mother drop out, and, instead, the
boys are admonished by the initiators to avoid other women, especially
married women; should they disobey, they are warned, they will be killed.
Indeed, beginning with the separation of the boys from their mothers at the
first stage of the initiation rite until their marriage following the last stage,
males are never again alone with a woman, and throughout that period ritual
aggression is used by the older males to “instill [in them] fear and
obedience” so that their avoidance of females will be maintained.
Although the Sambia begin their first stage of initiation in early
childhood, it is important to note that there are important cross-cultural
differences in the age at which these rites, as well as child extrusion, occur.
In some societies they take place in childhood, in others not until puberty.
Insofar, then, as these customs are (among other things) cultural resources
for the implementation of the taboos on mother-son incest and father-son
aggression, it may perhaps be assumed that those societies in which they do
not take place until puberty are societies in which the boy’s repression of
his incestuous and aggressive wishes, though incomplete, is sufficiently
strong to prevent them from passing through the repressive barrier
throughout childhood, and that it is only at puberty, with the eruption of his
sexual urges, that his own psychological resources are inadequate to that
task. If that assumption is correct, it then follows that, conversely, those
societies in which these cultural resources are already brought into play in
childhood are societies in which the boy’s psychological resources for
coping with his Oedipal wishes are inadequate from the very beginning.
Whether the one or the other, however, if the Oedipus complex is
incompletely repressed in childhood, there is a special urgency for
containing the boy’s Oedipal wishes either before or at the time he reaches
puberty, for it is then that he is physiologically capable of acting upon his
sexual wishes and physically capable of acting upon his aggressive wishes.
That is the time, in short, by which it is especially urgent that cultural
resources be brought into play either to strengthen the incomplete
repression of his Oedipal wishes (painful initiation rites) or to reduce his
opportunities for acting upon them (extrusion).
If, then, son extrusion and painful initiation are most likely to be
practiced in societies in which the Oedipus complex is incompletely
repressed, and least likely in those which exhibit the two other outcomes,
we can now see that differences in the outcome of the Oedipus complex
have other important consequences as well. Since, for example, societies in
which the Oedipus complex is extinguished or fully repressed are unlikely
to practice child extrusion, the family is more likely to remain an intact
residential group than in those in which, as a consequence of the incomplete
repression of the Oedipus complex, the son is extruded from the household.
But not all intact family households are the same; among other things they
vary importantly in their emotional texture. Thus, the emotional texture of
the family in which the Oedipus complex is extinguished and in which,
therefore, children can live together with parents in emotional comfort,
without suffering intrapsychic conflicts regarding their sexual and
aggressive Oedipal wishes, is very different from the emotional texture of
the family in which the Oedipus complex is repressed and in which,
therefore, the integrity of the family household is purchased at the cost of
persistent unconscious struggles with Oedipal wishes. These differences, of
course, are much more pronounced during puberty than during childhood.
Differences of another type distinguish societies which, because of
incomplete repression of the Oedipus complex, practice painful initiation
rites from those which, because of its extinction or full repression, have no
need to practice them. Thus, in many tribal societies, initiation rites can be
of unimaginable complexity, extending over a period of many years, and
consuming a large proportion of their social and economic resources.
Indeed, in many of these societies—most notably those of New Guinea—
these rites may be said to dominate the lives of the members of the group,
constituting the main focus of their interests and action (see Herdt 1980;
Read 1965; Tuzin 1980). A funnel for so many social and economic
resources, and for so much emotional energy, a cultural focus of this type
has still other consequences. For example, it significantly limits the options
of these societies for choosing alternative (and perhaps more productive)
cultural means for the investment of emotional energy and the allocation of
social and economic resources. Moreover, inasmuch as a cultural focus of
this type is both a highly elaborated magical response to unconscious
wishes and fears, as well as a stimulus for the arousal of still others, it may
serve to reinforce the skewed ratio of magical to realistic thinking found in
many of these tribal societies. If so, this would account not only for the high
proportion of magical (alternatively, primary-process, prelogical, animistic)
thinking that, in my view, is one of their singular psychological
characteristics, but also for their seeming inability to evolve an alternative
cultural focus (or foci) based on realistic (alternatively, secondary-process,
logical, nonanimistic) thinking.
There are, of course, many other social and cultural consequences
attendant upon the variability in the outcome of the Oedipus complex
which, since their explication would require a separate book, can only be
mentioned here. Thus, in societies in which unconscious Oedipal conflicts
require persistent repression for their containment, the Oedipus complex
may undergo structural transformations as a result of defensively motivated
projections and displacements which importantly affect other social
relationships and institutions. That, indeed, is Jones’s (1925) contention
regarding the boy’s hostility toward the mother’s brother in the Trobriands,
which, so he argues, represents the displacement of his Oedipal hostility for
his father. Gough (1953) interprets the nephew’s hostility toward the
mother’s brother among the Nayar in the same manner.5
5. Structural transformations in the initial formation of the Oedipus complex are also found in its
symbolic expression in myth, as Paul (1980) has shown in his perceptive analysis of the hostility
dimension of the Oedipus complex in Greek, Judaic, and Christian Oedipal myths.
Kin relationships, however, are not the only social relationships in which
repressed Oedipal conflicts are projected and displaced. Political (Lasswell
1960), religious (Erikson 1958), and economic (Brown 1959) institutions,
to mention only a few, constitute some of the larger social arenas for the
symbolic expression of a repressed Oedipus complex. Thus, to advert to its
hostility dimension, rebellious attitudes toward authority figures often have
their psychodynamic source in repressed Oedipal hostility, the former being
a vehicle for the displacement or sublimation of the latter (see Erikson
1963; Feuer 1969; Rothman and Lichter 1982). The same process occurs in
regard to the sexual dimension of the Oedipus complex. Thus, for example,
the strong male involvement in the Marian cult in southern Italy (Parsons
1969) and Mexico (Bushnell 1958) is often interpreted as a sublimation of
the repressed Oedipal attachment to the mother.
Since religion (because, perhaps, of its frequent use of family idioms)
like politics (for the same reason) is an especially important cultural domain
for the expression of repressed Oedipal conflicts, it might be added that
other differences between societies in which the Oedipus complex is
repressed in contrast to those in which it is extinguished may also be seen in
such diverse religious phenomena as ritual circumcision and clitoridectomy,
ascetic abstinences and self-torture, sexualized goddesses and witches,
celibate priests and priestesses, mystical and trance states, and many others.
Although further examples would require a monograph, this brief
discussion has perhaps been sufficient to suggest that the continuing debates
over the Oedipus complex are debates not merely about a passing episode
in the psychological development of the child. Rather, they are debates
about a psychological constellation which, as this discussion has attempted
to show, has pervasive cultural, social, and psychological consequences.

The Making of a Scientific Myth

In the previous two sections I argued that although the universality of the
Oedipus complex is rendered highly likely by the child’s potentiality for
sexual and aggressive arousal, inasmuch as the intensity and distribution of
sexual and aggressive wishes are each a product of the social relationships
that the child sustains with those adults who constitute his “significant
others,” in principle we would expect to find cross-cultural variability in the
Oedipus complex as a function of the variability in those relationships. That
this expectation is most probably actualized, however, in regard only to the
intensity and outcome, but not the structure, of the Oedipus complex is not
surprising. Since the classical Oedipal triangle—consisting of mother,
father, and son—is determined by the twin facts that the child’s early
mothering figure is the first and most important object of his sexual wishes,
and that his perceived rival for her love is the consequent object of his
hostility, cross-cultural variability in this structure would be expected if, but
only if, there were some range of variability in the social recruitment to the
mothering role. Since at the societal level, however, it is the biological
mother who is the child’s central (if not exclusive) early mothering figure,
and the biological father his most salient rival for her love in all known
societies, it is understandable that the structure of the Oedipus complex,
although cross-culturally variable in principle, is most probably invariant in
fact.
This, of course, brings us back to our point of departure, the controversy
over the Trobriand Oedipus complex. For the very nub of Malinowski’s
contention that the structure of the nuclear complex (as he preferred to put
it)6 is culturally variable not only in principle, but also in fact, consists of
his claim that although the biological mother is the primary mothering
figure in the Trobriands, it is the sister who is the primary object of the
boy’s libidinal desires. Since that paradoxical claim has already been shown
to be refuted by Malinowski’s own evidence, there is no need to repeat that
demonstration. Rather, in this final section, I wish to address the remarkable
fact that this paradoxical claim has been accepted with almost no skepticism
or critical inquiry for fifty years.
6. Since, for Malinowski, the Oedipus complex is a cultural product, he used “nuclear complex” as
a cover term for all possible cultural variants, reserving “Oedipus complex” for that particular variant
in which mother and father are the objects, respectively, of the child’s sexual and aggressive
impulses.
Given the powerful evidence for the universality of the motivational
disposition to mother-son incest, if it were reported that in some society the
mother is not an incestuous object for the son, such a report would, of
course, be rather surprising. But since the history of science is a history of
persistent refutations of well-established generalizations, there would be no
reason to greet such a report with special skepticism. In claiming, therefore,
that it is a “remarkable fact” that Malinowski’s report of such a finding in
the Trobriands has not been greeted with sufficient skepticism, it is not
because that report challenges an established generalization, but because in
the context of Trobriand culture the absence of an incestuous attachment to
the mother is twice anomalous.
The first anomaly consists in the fact that of the three types of incestuous
desire that might, in principle, be found in any male, the desire of the son
for the mother is, alone, reported to be absent in the Trobriands, while that
of the father for the daughter is reported to be as strong as it is purported to
be in the West, and that of the brother for the sister stronger by far. This
gestalt is all the more anomalous since typically the boy’s incestuous
attachment to the mother is, if anything, much stronger than the other two;
and it becomes still more anomalous when it is recalled that the incestuous
attachment of the Trobriand daughter to the father, unlike that of the son to
the mother, is reported to be of normally expected Oedipal intensity.
In claiming that because it is anomalous Malinowski’s report should have
been greeted with skepticism, I am by no means suggesting that its truth or
accuracy should therefore have been impugned. If “anomaly” is glossed as
“puzzle,” and if we follow Kuhn (1962: chapter 4) in viewing science as a
puzzle-solving enterprise, then, rather than impugning the accuracy of an
anomalous finding, scientific skepticism merely signals the existence of yet
another puzzle to be solved. In the case of this Trobriand puzzle, one simple
solution immediately suggests itself. Since, as has already been observed, it
is their physical and emotional intimacy that accounts for the son’s
incestuous attachment to the mother, part of the Trobriand puzzle would be
solved if it were the case that the most intimate relationship of the young
boy is with some woman (or women) other than the biological mother. If,
moreover, it were also the case that this structural arrangement did not
affect the relationship of the male to either his sister or his daughter, the
other parts of the puzzle would be solved, as well, and the skepticism could
be laid to rest.
Unfortunately, however, the ethnographic facts are the very opposite of
this proposed solution. As we have seen, all the conditions that highlight the
mother-son relationship as the primary arena for the boy’s most intense
incestuous struggles in the “normally expectable” family are present in the
Trobriands as well—and then some! In short, there is nothing in
Malinowski’s account that might solve the puzzle of why it is that in a
family and socialization system of the Trobriand type the sister and
daughter are the objects of the male’s incestuous desires, whereas the
mother is not. That, indeed, is why I characterized Malinowski’s claim as
anomalous.
Perhaps, however, the solution to this anomaly might be found by taking
another tack. For just as the problem of inflation can be tackled from either
the supply or the demand side of the market, the solution to an
ethnographically anomalous finding can sometimes be found not in an
antecedent, but in a consequent condition. Specifically, if the reported
absence of a normally expectable incestuous attachment to the mother were
accompanied by the absence of a normally expectable mother-son incest
taboo, the former report would not then constitute an anomaly (although the
absence of such an attachment would still have to be explained by some as
yet unknown, and theoretically unexpected, antecedent condition).
Unfortunately, the ethnographic reality is once again the reverse of the
proposed solution: the mother-son incest taboo is, in fact, present in the
Trobriands. Moreover, the presence of this taboo not only leaves the
anomaly unresolved, but creates yet another: if the boy has no incestuous
desire for the mother, why should there be a taboo prohibiting incest with
her? This second anomaly, of course, could be resolved by simply rejecting
the assumption on which it is based, viz., that taboos exist to prevent the
practice of the tabooed actions. But that requires at least a brief discussion
of the historical debates regarding incest taboos.
It was Frazer ([1887] 1910:vol. 4, p. 97 ff.) who first converted the above
assumption about the function of taboos into the generally accepted theory
of incest taboos. Since, following from that assumption, the function of
incest taboos is the prevention of incest with their stipulated targets, the
existence of these taboos, so Frazer argued, implies the existence of
incestuous desire for those very targeted persons.7 This theory, however,
has had its opponents from its very inception. Most of them, following
Westermarck’s ([1906–8] 1924:vol. 2, p. 368) view that sexual aversion
develops between family members who live in proximity to each other from
an early age, argue that incest taboos are merely the institutionalized
expression of sexual aversions.
7. Why it is that all societies have wanted to prevent incest at least within the nuclear family is a
question that has been variously answered, but all the proposed answers point to one or more of the
following maladaptive consequences of the widespread practice of nuclear family incest which are
precluded by compliance with incest taboos: (1) biological impairment attendant upon inbreeding
(see Lindzey 1967), (2) collapse of the family authority structure, because of the sexual rivalry
between father and son, with the attendant difficulty of cultural transmission (see Malinowski [1927]
1955:216), (3) structural breakdown of the family attendant upon role confusion (see Parsons 1954),
(4) breakdown of social alliances attendant upon family endogamy (see Lévi-Strauss [1949]
1969:chapters 4 and 29).
Although the acceptance of this theory would resolve the anomaly of the
reported co-occurrence in the Trobriands of a mother-son incest taboo and
the absence of an incestuous desire for the mother, there are powerful
grounds for rejecting it as invalid. One could point, for example, to the
evidence in support of the motivational disposition to nuclear family incest
which was summarized earlier in this chapter. Again, one could point to
Lindzey’s reformulation of Frazers hypothesis in terms of the adaptive
framework of evolutionary biology, a reformulation which, in my view, is
unimpeachable. “It seems unlikely that there would have been universal
selection in favor of such a taboo if there were not widespread impulses
toward expression of the prohibited act” (Lindzey 1967:1055). Rather,
however, than evoking these general empirical and theoretical grounds in
refutation of Westermarck’s theory, it is enough—and in this context much
more relevant—to refute it on specific Trobriand grounds.
The latter grounds consist of evidence that proclaims loudly and clearly
that at least some Trobrianders who live in close propinquity develop strong
sexual feelings for each other, rather than sexual aversion. That is how
Malinowski, at any rate, describes the relationship between brothers and
sisters in the Trobriands, and (to a lesser degree) between fathers and
daughters. At the same time there are taboos in the Trobriands which both
prohibit—and (as Malinowski makes quite clear) are intended to prevent—
incest between the members of these dyads. For the Trobriands, then, there
can be no doubt about the functional relationship between incestuous
desires and incest taboos, so far at least as these two dyads are concerned.
Hence, for the Trobriands the co-occurrence of the mother-son incest taboo
and the absence of incestuous desires for her is clearly an anomaly, not
because it violates a (disputed) theoretical assumption, but because in the
Trobriand cultural configuration their co-occurence is incongruous.
If, however, it were still contended that this finding is not anomalous—
because the anomaly now derives from merely another theoretical
assumption, that of pattern consistency—this contention would convert an
unresolved anomaly into an unresolved dilemma. For if it were held, on the
one hand, that the taboo on incest with the mother does not imply the
existence of incestuous wishes for her, what possible explanation might
then be offered for the taboos on incest with the sister and daughter which,
in fact, do correspond to known incestuous wishes for them? And if it were
held, on the other hand, that the taboos on sister and daughter incest are
explained by known incestuous wishes for them, how could it then be
contended that it is invalid to infer the existence of incestuous wishes for
the mother from the taboo prohibiting incest with her?
Malinowski, himself a leading proponent of Frazer’s theory of incest
taboos, was keenly aware of this dilemma, and, it will be recalled, he
attempted to resolve it by holding to both of its horns. That is, while he held
that incest taboos are valid measures of incestuous wishes, he also held that
in the Trobriands the taboo on mother-son incest is “weak.” Hence, he
argued, just as the strong incestuous desire for the sister in the Trobriands is
reflected in a “strong” brother-sister taboo, the weak desire for the mother is
reflected in a “weak” mother-son taboo. Unfortunately, however, the facts
(see above) are otherwise: there is simply no evidence for the putative
weakness—whether absolute or relative—of the mother-son taboo. In short,
despite Malinowski’s attempt to resolve it, the second anomaly in his
reported absence of an incestuous attachment to the mother in the
Trobriands remains—like the first—unresolved.
The fact, then, that neither of these anomalies can be resolved—or, to put
it more cautiously, that neither has been resolved thus far—must surely
constitute sufficient reason to be skeptical of the report that the Trobriand
son has no incestuous attachment to the mother. That this report, then, has
not received the skeptical reception that normally greets an anomalous
scientific report is not only, as I said above, a remarkable fact, but one
which itself constitutes an intriguing intellectual problem. Since its solution
could shed important light on the influence of scientific paradigms on the
acceptance and persistence of scientific ideas, it is to be hoped that an
investigation of the problem might some day be undertaken by an
intellectual historian or a historian of science.
Although the causes for the unskeptical reaction to this anomalous report
are still to be discovered, its consequence—with some notable exceptions—
has been an uncritical acceptance of the putative Trobriand matrilineal
complex. I would suggest, then, that if this report had been subjected to the
probing scrutiny to which anomalous scientific findings are usually subject,
the matrilineal complex would have been rejected as empirically
unsupported rather than achieving the status of an incontrovertible finding
of anthropological science. Nevertheless, it is not its weak empirical
foundation that led me to characterize the Trobriand matrilineal complex a
“scientific myth.” This characterization stems, rather, from the uncritical
acceptance of the reported finding—the absence of an incestuous
attachment to the mother—on which its plausibility hangs, despite the fact
that this finding is not only once, but twice, anomalous.
A myth, Malinowski taught us, enjoys uncritical acceptance because it
serves important functions for those who believe it to be true. It would be
well, therefore, if our hoped-for historian were to address yet another
question in the course of his investigations: what functions might have been
served by the acceptance of this myth? Indeed, since the role of the “will to
believe” in the acceptance of scientific ideas is as prominent as the role that
William James attributed to it in the acceptance of religious doctrines, it is
entirely possible that the answer to this second question might
simultaneously provide the answer to the first.

Note
Reprinted from Melford E. Spiro, Oedipus in the Trobriands (Chicago: University of Chicago
Press, 1982), pp. 144–80.

References

Aries, Philippe. 1962. Centuries of childhood. New York: Alfred

Knopf.
Bettelheim, Bruno. 1969. The children of the dream. London:
Macmillan.
Bischof, Norbert. 1975. Comparative ethology of incest avoidance. In
Biosocial anthropology, ed. Robin Fox. London: Malaby Press.
Brown, Norman O. 1959. Life against death. Middletown: Wesleyan
University Press.
Bushnell, John. 1958. La Virgen de Guadalupe as surrogate mother in
San Juan Atzingo. American Anthropologist 60:261–65.
Buxbaum, Edith. 1970. Troubled children in a troubled world. New
York: International Universities Press.
Campbell, Donald T., and Raoul Naroll. 1972. The mutual
methodological relevance of anthropology and psychology. In
Psychological anthropology, ed. Francis L. K. Hsu. Cambridge,
Mass.: Schenkman Publishing Company.
Carpenter, C. R. 1940. A field study in Siam of the behavior and social
relations of the Gibbon (Hylobatis lar). Comparative Psychology
Monographs 16: 1–212.
Carstairs, G. Morris. 1956. The twice born. Bloomington: Indiana
University Press.
Chance, Michael R. H., and Clifford J. Jolly. 1970. Social groups of
monkeys, apes, and men. New York: Dutton.
Demarest, William J. 1977. Incest avoidance in human and nonhuman
primates. In Primate bio-social development, ed. Suzanne
Chevalier-Skolnikoff and Frank E. Poirier. New York: Garland
Publishing.
Elwin, Verrier. 1968. The kingdom of the young. Oxford: Oxford
University Press.
Erikson, Erik H. 1958. Young man Luther. New York: W. W. Norton.
_______ 1963. Childhood and society. New York: W. W. Norton.
Feuer, Lewis S. 1969. The conflict of generations. New York: Basic
Books.
Fox, Robin. 1980. The red lamp of incest. New York: Dutton.
Frazer, James George. 1910. Totemism and exogamy [1887]. London:
Macmillan.
Gathome-Hardy, Jonathan. 1977. The public school phenomenon, 597–
1977. London: Hodder and Stoughton.
Gough, E. Kathleen. 1953. Female initiation rites on the Malabar
coast. Journal of the Royal Anthropological Society 85:45–80.
Herdt, Gilbert H. 1980. Guardians of the flutes: Idioms of masculinity.
New York: McGraw-Hill.
____ 1982. Fetish and fantasy in Sambia initiation. In Rituals of
manhood, ed. Gilbert H. Herdt. Berkeley: University of California
Press.
Hiatt, L. R. 1971. Secret pseudo-procreation rites among the
Australian aborigines. In Anthropology in Oceania, ed. L. R. Hiatt
and C. Jayawardena. San Francisco: Chandler.
Hopkins, Keith. 1980. Brother-sister marriage in Roman Egypt.
Comparative Studies in Society and History 22:303–54.
Hunt, Robert C. 1971. Component of relationships in the family: A
Mexican village. In Kinship and culture, ed. Francis L. K. Hsu.
Chicago: Aldine.
Jones, Ernest. 1925. Mother-right and the sexual ignorance of savages.
International Journal of Psychoanalysis 6:109–30.
Justice, Blair, and Rita Justice. 1979. The broken taboo. New York:
Human Sciences Press.
Kaffman, Mordecai. 1977. Sexual standards and behavior of the
kibbutz adolescent. American Journal of Orthopsychiatry 47:207–
17.
Kakar, Sudhir. 1978. The inner world. Delhi: Oxford University Press.
Kuhn, Thomas S. 1962. The structure of scientific revolutions.
Chicago: University of Chicago Press.
Kummer, Hans. 1968. Social Organization of Hamadryas baboons.
Chicago: University of Chicago Press.
Lancaster, Jane B. 1979. Sex and gender in evolutionary perspective.
In Human sexuality: A comparative and developmental
perspective, ed. Herant A. Katchadorurian. Berkeley: University of
California Press.
Langness, L. L. 1974. Ritual power and male domination in the New
Guinea highlands. Ethos 2:189–212.
Lasswell, Harold D. 1960. Psychopathology and politics. New York:
Viking.
Lévi-Strauss, Claude. 1969. The elementary structures of kinship
[1949]. London: Eyre and Spottiswoode.
Lidz, Ruth W., and Theodore Lidz. 1977. Male menstruation: A ritual
alternative to the Oedipal transition. International Journal of
Psychoanalysis 58:17–31.
Lindzey, Gardner. 1967. Some remarks concerning incest, the incest
taboo, and psychoanalytic theory. American Psychologist 22:1051–
59.
Malinowski, Bronislaw. 1955. Sex and repression in savage society
[1927]. New York: Meridian Books (Sfi).
Money, John. 1980. Love and love sickness. Baltimore: Johns Hopkins
University Press.
Parsons, Anne. 1969. Belief magic, and anomie. New York: Free Press.
Parsons, Talcott. 1954. The incest taboo in relation to social structure
and the socialization of the child. British Journal of Sociology
5:101–17.
Paul, Robert A. 1980. Symbolic interpretations in psychoanalysis and
anthropology. Ethos 8:286–94.
Poole, Fitz John Porter. 1982. The ritual forging of identity: Aspects of
person and self in Bimin-Kuskusmin male initiation. In Rituals of
manhood, ed. Gilbert H. Herdt. Berkeley: University of California
Press.
Powell, H. A. 1957. Analysis of present-day social structure in the
Trobriands. Ph.D. Dissertation, University of London.
Rainwater, Lee, and William L. Yancey. 1967. The Moynihan report
and the politics of controversy. Cambridge: MIT Press.
Read, Kenneth E. 1965. The high valley. New York: Scribners.
Reik, Theodore. 1946. Ritual: Four psychoanalytic studies [1919].
New York. Grove Press.
Rothman, Stanley, and S. Lichter. 1982. The radical impulse. New
York: Oxford.
Rowell, Thelma E. 1972. Female reproduction cycles and social
behavior in females. Advances in the Study of Behavior 4:69–105.
Roy, Manisha. 1975. Bengali women. Chicago: University of Chicago
Press.
Shepher, Joseph. 1971. Mate selection among second generation
kibbutz adolescents and adults: Incest avoidance and negative
imprinting. Archives of Sexual Behavior 1:293–307.
Smith, Raymond T. 1956. The negro family in British Guiana: Family
structure and social status in the villages. London: Routledge and
Kegan Paul.
Spiro, Melford E. 1958. Children of the kibbutz. Cambridge: Harvard
University Press.
_______ 1982. Oedipus in the Trobriands. Chicago: University of
Chicago Press.
Talmon, Yonina. 1964. Mate selection on collective settlements.
American Sociological Review 29:491–508.
Tanaka, Masako. 1981. “Maternal” authority in the Japanese family.
Paper presented for the International Symposium on Religion and
 the Family in East Asia, National Museum of Ethnology, Osaka,
August 30-September 7, 1981.
Tuzin, Donald F. 1980. The voice of the Tambaran. Berkeley:
University of California Press.
Westermarck, Edward. 1924–26. The origin and development of the
moral ideas [1906–8]. London: Macmillan.
Wilson, Edward O. 1978. On human nature. Cambridge: Harvard
University Press.
Wilson, Monica. 1949. Nyakyusa age villages. Journal of the Royal
Anthropological Institute 79:21–25.
Wolf, Arthur P. 1966. Childhood association, sexual attraction, and the
incest taboo: A Chinese case. American Anthropologist 68:893–98.
_______ 1968. Adopt a daughter-in-law, marry a sister: A Chinese
solution to the incest taboo. American Anthropologist 70:864–94.
_______ 1970. Childhood association and sexual attraction: A further
test of the Westermarck hypothesis. American Anthropologist
72:503–15.
Wolf, Arthur P., and Chich-Shan Huang. 1980. Marriage and adoption
in China, 1845–1945. Stanford: Stanford University Press.
II
FUNCTIONAL ANALYSIS
5
Social Systems, Personality, and
Functional Analysis

Introduction
WHEN ANTHROPOLOGY was primarily interested in culture history, the
question of how societies get their members to behave in conformity with
cultural norms was of small concern. But when anthropology became
interested in the problem of how societies operate, this question became—
and has remained—salient, not only for culture-and-personality theorists but
for other anthropologists as well. “Our great problem as anthropologists,”
says Firth, is “… to translate the acts of individuals into the regularities of
social process” (1954: 11).
Since social systems are attributes of society and personality systems are
attributes of individuals, it was formerly assumed, both by anthropologists
and by psychologists, that there was little relationship between “the acts of
individuals” and the “regularities of social process.” Before the development
of culture-and-personality studies, this assumption seemed reasonable. First,
although there is but one social system for a society, there are as many
personalities as there are members of society. Secondly, since social systems
are normative, their constituent activities are prescribed; but since
personality systems are conative, their activities are motivated. Finally,
social systems serve social functions, while personalities serve individual
functions. In short, although the functions of social systems are served by the
activities of individuals, these activities were not seen as serving personal
functions. Hence, older theories of cultural conformity1 and social control
ignored personality as an irrelevant variable.
 1. The concept, “cultural conformity,” is here taken to mean, behavior which is in conformity with
cultural norms. Hence, “cultural conformity,” as used in this chapter, is to be distinguished from
“social conformity,” which refers to behavior which is in conformity with the behavior of others. In a
fully integrated and relatively unchanging society it would be difficult to distinguish between these
two types of conformity: the behavior of others would be more or less identical with the requirements
of the cultural heritage. In a somewhat less integrated and rapidly changing society (such as our own)
the distinction between these two types of conformity is clearer; Riesman’s (1950) other-directed
individuals, for example, represent social conformity rather than (or more than) cultural conformity. In
either case, though it might be difficult to distinguish between these types of conformity in overt
behavioral terms, it is not at all difficult to distinguish between them in motivational terms. Social
conformity is motivated by the desire to conform to the behavior of others; cultural conformity, by the
desire to conform to cultural norms. Cultural conformity, as we shall attempt to show, is a requisite for
the functioning of human social systems, whereas social conformity is not.
Classical cultural determinism, for example, attributed efficient causation
to the cultural heritage—people perform this or that activity of the social
system “because it’s part of their culture.” Although this theory represents an
advance over still older biologistic theories, it begs the very question which
is to be answered. As Nadel has put it: “… little is gained [in the study of
social control] by adducing the force of custom and tradition, that is, the
sheer inertia of habitual behavior and inherited practice” (1953:266). The
mere existence of a cultural heritage does not imply that it will be inherited;
or, if inherited, that behavior will be in conformity with its requirements.
The notion that cultural behavior is inherited automatically from the cultural
heritage is probably based on a confusion ultimately derived from Tylor’s
omnibus definition of culture (1874:1). For it would seem that the model
upon which the inheritance of cultural behavior is based is the inheritance of,
for example, tools, paintings, and houses—all of which are, of course,
inherited automatically, without either effort or motivation. Culture behavior,
too, is transmitted from a previous generation; but it is inherited by learning,
and not merely by being handed down.
Another answer to the problem of cultural conformity is provided by the
social sanctions theory. According to this theory, compliance with cultural
norms is achieved through positive and negative sanctions—rewards and
punishments—which function as techniques of social control. Although the
use of sanctions is probably universal, the thesis that cultural conformity is
achieved primarily or exclusively through the use of social sanctions rests, at
least implicitly, on two demonstrably false assumptions. These are the
Rousseauist assumption that culture is necessarily frustrating, and the super-
organistic assumption that cultural norms “exist” in the cultural heritage, but
are not internalized by the members of society.
 Agreeing with the first, but disagreeing with the second of the above two
assumptions, a third theory of cultural conformity views compliance with
cultural norms as a function, primarily, of their internalization within
personality. Although cultural norms are, indeed, internalized, and although
conscience does play an important part in achieving cultural conformity, this
theory too is but a partial theory for, as we shall attempt to show, social
control is frequently achieved without the necessity for norm internalization.
Culture-and-personality studies suggest that though there is a large
measure of truth in these theories, cultural conformity is most frequently
achieved because social systems satisfy personality needs. This chapter,
then, will attempt to show that there is an intimate relationship between
social systems and personality: social systems operate by means of
personality, and personality functions by means of social systems. Many of
the social functions of social systems can be served only when this intimate
relationship obtains.

Human Social Systems: The Problem

Unlike other social animals, the social system of any particular human
society cannot be predicted from a knowledge of the species (Homo sapiens)
of which the society is a member. Nevertheless, since human social systems
are rooted in man’s biological nature, any discussion of the generic attributes
of these systems must take its departure from certain biological dimensions
of human existence. From a comparative biological perspective a human
social system may be viewed as a functional requirement of human life.
Ultimately it stems from the psychobiological needs of what the biologist
terms a generalized, fetalized (Bolk 1929), and highly plastic (Montagu
1951:368–75) primate. Here we can only point to the consequences of these
biological attributes for human social systems. (But cf. La Barre 1954;
Roheim 1943.)
The combination of man’s mammalian drives (hunger, sex, etc.) and his
plastic hominoid constitution (paucity of instincts) requires that means of
drive-reduction be learned. Again, the combination of man’s organic needs
(protection against weather, predatory beasts, etc.) and his hominoid
constitution (generalized and fetalized) requires learned methods of
protection and adaptation. Moreover, man’s prolonged primate dependency
and his primate sexual behavior (lack of a breeding season) combine to
produce the relatively permanent biparental family, and—by extension—
larger collectivities (societies) consisting of two or more families. In the
absence, however, of an instinctual base—shopworn comparisons of human
with insect societies (Wheeler 1928) are still much to the point—human
social life demands that forms of social interaction, methods of social
cooperation, techniques of conflict resolution, and the like be learned. But
this is not enough. Social existence is necessarily an orderly and regulated
existence. Unless the members of a group are able to predict with some
probability far greater than chance the behavior of other members of the
group with whom they interact, social action, let alone interaction, would be
all but precluded. Hence, man must not only learn the various kinds of
behavior patterns mentioned above, but these learned behavior patterns must
be prescribed by society and shared with others. The configuration of these
socially prescribed, learned, shared and transmitted behavior patterns which
mediate and facilitate social relationships constitutes the social system of a
human society. We are here only concerned with those characteristics which
make social systems necessary for human survival. We are not concerned
with those characteristics—a complex brain and central nervous system and
the symbolic behavior to which they give rise (White 1940; Mead 1934;
Langer 1942; Cassirer 1944; Hallowell 1950)—which make their invention
and transmission possible.
To conclude: since man is a generalized, fetalized, and plastic animal and
since everywhere he is necessarily social, a typically human existence
depends on the existence of socially shared behavior patterns which satisfy
his (1) biological needs, (2) those group needs that are an invariant
concomitant of social life (Aberle 1950) and (3) those emotional needs that
develop in the interaction between biology and society. In this evolutionary
perspective a social system may be viewed as an “instrumental apparatus”
(Malinowski 1944) for the satisfaction of these needs. Social systems, then,
have three types of functions. They promote the physical survival of society
and of its constituent members (adaptive functions); they contribute to the
persistence of the social structure of a society and, hence, to orderly social
interaction (adjustive functions); they promote social solidarity by the
reduction of inter- and intra-personal tension (integrative functions).
 This is not to say, of course, that all aspects of every social system are
functional, or that all social systems are equally functional, or that any social
system is functional to the same degree for all the members of, or groups
within, a society. The collapse of some social systems, the oppressive means
used by powerful groups within a society to preserve others, the repeated
history of successful and of unsuccessful rebellions against still others—all
these testify to the powerful dysfunctional forces operative in some and
potentially in all social systems. But these observations serve to confirm,
rather than to confute, the major thesis. Social systems have vital functions;
that these functions be served is their raison d’être. If they are not served, to
a greater or lesser degree, the social system will, in the long-run, be
modified, or the society will not survive.
Before proceeding with this discussion, it is necessary to emphasize an
obvious characteristic of human social systems that is frequently obscured
by the ambiguity of the word “learned.” an ambiguity that sometimes leads
to hasty generalizations from small-group experiments to social behavior in
society. When it is observed that human social systems, as well as all other
aspects of culture, are learned, the word “learned” has one meaning in a
phylogenetic, and another in an ontogenetic context. “Learned” in a
phylogenetic context means invented or discovered; “learned” in an
ontogenetic context means acquired. Thus, the hypothetical Ur-mensch of
the Paleolithic was, culturally viewed, a tabula rasa. The adaptive, adjustive,
and integrative requirements of his society had to be satisfied by behavior
patterns of his own invention and discovery. Succeeding generations of
human societies have also, to be sure, invented and discovered behavior
patterns. Their incorporation into the configuration of existing behavior
patterns (which comprise their social systems) produces one of the unique
dimensions of culture—its cumulativeness. For the most part, however, all
generations subsequent to the hypothetical Ur-generation of a society have
acquired their social systems from a previous generation, rather than
inventing or discovering them themselves. In short, the social system of any
generation represents, in part, the cultural heritage of the succeeding
generation; the social system of the latter, is acquired from the social system
of the former.
Social systems, like any other large configuration, can be—and for certain
purposes must be—broken down into smaller components. These units,
proceeding from the largest to the smallest, are generally termed sub-
systems, institutions, roles. Thus, every social system includes an economic
system—an organized means for the production, consumption, and
distribution of goods and services; a kinship system—an organization of
behavior within the family and among kinsmen; a political system—a
sanctioned means for the acquisition and use of legitimate power, and so
forth. The universality of these subsystems is sometimes referred to as “the
universal culture pattern” (Wissler 1923: chap. 5).
Each of these broad categories can usually be classified, in turn, into
smaller units. It is rare for any one type of social group within society to
perform all the activities which comprise any of these broad subsystems.
Thus the kinship system may embrace the activities of nuclear families,
lineages and clans; or the economic system may include the activities of
trade unions, banks, factories, and accounting firms. In short, since any
society is differentiated and, therefore, consists of many types of social
groups, and since each type serves different functions, either for its own
members or for those of other social groups, each type of group performs
different activities. The configuration of activities which characterizes these
different types of groups may be termed an “institution.” Thus the activities
of the members of the family, qua family members, may be termed the
“family institution’; the activities of the members of the lineage, qua lineage
members, may be termed the “lineage institution.’’ Since, collectively, these
institutions comprise the kinship system of a society, each may be termed a
kinship institution.
Although each type of social group within a society is characterized by a
different institution, its constituent members do not, qua members, perform
the same activities. Each type of group, like the entire society of which it is a
part, is structurally differentiated so that various members of the group
occupy different positions within the group. Within the family, for example,
different members may occupy such positions as father, mother, son, or
daughter. Since each position (“status”) within the group is associated with
one (Linton 1936) or more (Merton 1957) sets of activities (“roles”), each
institution may be broken down into its constituent roles. Thus the set of
activities which comprises the role of father varies from that which
comprises the role of mother. Each is a family role; collectively they
comprise the family institution. The role, then, is the smallest unit of the
social system; the operation of the social system, ultimately and most
directly, depends on the proper performance of roles.
 To sum up: the survival of a society depends on the operation of its social
system; a social system is comprised of subsystems which, in turn, are
comprised of institutions; the functions of these institutions are served only
if their constituent roles are performed. In turn, this requires the recruitment
of individuals for the various statuses which comprise the social structure. If
these propositions are valid, we are brought back to the central issue of this
chapter—the problem of cultural conformity. How does society induce its
members to perform roles—those that are instrumental to the attainment of a
status, as well as those that are entailed by the occupancy of a status? (Nadel
1957: chap. 2, has suggested the terms “recruitment roles” and “achievement
roles” to refer to these different types of roles.) This problem is best
understood against the background of infrahuman societies.
Among lower social animals there is a remarkably high correlation
between species and social systems. If the environment is held constant, the
description of the social system of one society within a species is more or
less descriptive of all other societies within the species. Thus, if one knows
the species to which a particular subhuman organism belongs, one can
predict with high accuracy and with great detail the social system (assuming
that it is social) in which it participates (Hine and Tinbergen 1958;
Thompson 1958; Mayr 1958). It is quite meaningful, therefore, to speak of
species-specific social systems among lower animals.
But though it is meaningful to speak of the red deer social system (Darling
1937) or the howling monkey social system (Carpenter 1934) it is not at all
meaningful, except on the highest level of generality, to speak of the human
social system. Man differs dramatically from all other social mammals in the
great variety of his intraspecies social system differences. Indeed, the
magnitude of social system differences within the human species may be as
great as the magnitude of difference among animal species. Thus, for
example, while the mating pattern of an entire mammalian species may be
characterized, and thus distinguished from other species, by monogamy
(e.g., gibbons—Carpenter 1940) or polygyny (e.g., baboons—Zuckerman
1932) or group marriage (e.g., howling monkeys—Carpenter 1934), such
generalizations apply in the case of humans only to societies within the
species and not to the species as a whole. Hence, the fact that one human
society practices monogamy, or has patrilineal descent, or is governed by
hereditary chiefs, or is stratified by caste, does not enable us to predict that
other societies within the species will have the same marriage, descent,
political or stratification systems. In short, if one knows that a particular
organism belongs to the human species, one cannot predict in any detail the
social system in which he participates even if the physical environment is
specified. Thus, though paired groups such as California Indians and modern
California Americans, precon-tact Hawaiians and the contemporary
inhabitants of Hawaii, Alaskan Eskimos and contemporary modern Alaskans
have occupied the same physical environment, their respective social
systems are radically different.
It is a reasonable inference, then, that though much of the social behavior
of animals is not instinctive (Beach 1955; Lehrman 1953)—as was formerly
believed to be the case—so that each generation of social animals may learn
a large percentage of its behavior patterns and social roles from a preceding
generation, the range of species plasticity is so narrow that any animal has
little alternative, if he is to learn at all, but to learn the behavior patterns
which he is taught. What he must learn in order to participate in his society’s
social system and what he can learn are for the most part identical. Since
humans, on the other hand, are highly plastic, what an individual must learn
in order to participate in the social system of his society is not at all identical
with what he can learn; for what he is taught represents, as the cross-cultural
record clearly reveals, but one alternative among a large number of behavior
patterns and roles which he is potentially capable of learning or, at least, of
thinking of learning.
Since humans are so enormously plastic it is not enough, if human social
systems are to function properly, that social roles and the behavior patterns
of which they are comprised be socially learned, shared, and transmitted; it
is also necessary that these roles be prescribed (Newcomb 1950: chap. 3).
For, since what a person must do in order to participate in a given social
system is not identical with what he can do, it may be inconsistent with what
he would like to do. Hence in the process of socialization children are not
only taught how to behave, but they are taught that the ways in which they
are taught to behave are the ways in which they ought to behave. In short,
“every human social order,” as Hallowell has put it, “operates as a moral
order” (1950:169). This normative, or cultural dimension (Spiro 1951:31–
36) of the human social system is for humans the functional equivalent of
restricted plasticity for lower animals. It is the basis for relatively uniform
and, therefore, predictable role behavior. (The psychological basis for the
emergence of a moral dimension in experience—the self—is discussed in
Hallowell 1954.)
But this analogy cannot be pressed too far, and it is precisely at the point
where it breaks down that human societies are uniquely different from
animal societies. Since there is always a potential conflict between duty and
desire, between cultural heritage and personality, this potentiality—which
gives special poignancy to the human situation—sets the problem of our
present inquiry: how do human societies get their members to behave in
conformity with cultural norms? Or, alternatively, how do they induce their
members to perform culturally prescribed roles?
It is at this juncture in the analysis that the concept of personality becomes
salient for the understanding of human social systems, for it is in the concept
of role that personality and social systems intersect. If personality is viewed
as an organized system of motivational tendencies, then it may be said to
consist, among other things, of needs and drives. Since modes of drive-
reduction and need-satisfaction in man must be learned, one of the functions
of personality is the promotion of physical survival, interpersonal
adjustment, and intrapersonal integration by organizing behavior for the
reduction of its drives and the satisfaction of its needs. If some of these
needs can then be satisfied by means of culturally prescribed behavior—if,
that is, social roles are capable of satisfying personality needs—these needs
may serve to motivate the performance of the roles. But if social systems can
function only if their constituent roles are performed, then, in motivating the
performance of roles, personality not only serves its own functions but it
becomes a crucial variable in the functioning of social systems as well. This
is the thesis which will be explored in this chapter.

Extrinsic Cultural Motivation

Since role behavior is a subclass of learned behavior we may begin our
discussion by asking under what conditions any learned behavior pattern
arises. Many—but not all—behavioral scientists2 seem to agree that
behavior occurs when the contemplated action is believed by the actor to be
rewarding. An act is performed when a person wants something and when he
has reason to expect that the performance of the act will supply his want. A
simple ontogenetic model can illustrate how this expectation is established.
2. This discussion is based primarily on psychoanalytic (Rapaport 1951) and behavior theory
(Miller and Dollard 1941; Tolman 1951). Despite the differences among contemporary psychological
theorists, almost all agree that reward—different terms are used to refer to the same concept—is a
crucial motivational variable (Nebraska Symposia on Motivation 1953:v. 1, ff.). They differ primarily
in their analysis of its referents and its properties. It is with respect to performance, not to learning,
that the notion of reward is here held to be crucial.
The ontogenetic model begins with a “drive”—that is, with some felt
tension or discomfort. “Drive” is used here in a psychological, not a
physiological, sense, for even the biological drives are significant for
behavior only if they function as psychological stimuli; hence, “felt tension
or discomfort.” Behavior in an infant or in a naïve experimental animal is
instigated by the desire to “reduce” the drive. But since the drive is still
uncanalized (Murphy 1947: chap. 8)—it has no goal, no cathected object—
behavior approaches randomness. By trial-and-error some object or event,
which has the property of reducing the drive, is chanced upon; the drive is
gratified; homeostasis is restored. If this sequence is repeated a sufficient
number of times, the drive-reducing object or event becomes a “goal” and
the act which is instrumental to the attainment of the goal becomes a
behavior pattern. An expectation of gratifying a drive by means of the goal
attained by the behavior pattern has been established.
Using this psychological model two simple questions concerning behavior
may be answered. Why does a naïve organism behave at all? Because it has
a drive. Why, after experience or training, does it behave in this, rather than
in some other, way? Because it has learned that this way attains goals which
are rewarding, i.e., drive-reducing. It should be emphasized, of course, that
“drive” refers to both innate and acquired drives, and that rewards need not
be “physical” nor need they be administered by others. The rewards for
exploratory and cognitive activity are frequently—even in the case of lower
primates (Harlow 1953)—inherent in the very act of exploration or intrinsic
to the solution of a problem. It should also be emphasized that no
assumption is made concerning fixed homeostatic states such that the
achievement of drive-reduction leads to relative quiescence until the drive is
reactivated. It is assumed, on the contrary, that there is always some
discrepancy between achievement and aspiration levels (Lewin et al., 1944)
so that present goal achievement may become but a temporary way-station
for contemplated further and different goal achievement. It is assumed,
however, that drives are motivational variables and, although not every act is
instigated by the anticipation of drive-reduction, that every drive must
eventually be reduced, either directly or indirectly.
Can this psychological model help us to understand cultural behavior, in
general, or the performance of the constituent roles of a social system, in
particular? At first, the answer might appear to be negative. For social roles,
it will be remembered, are not discovered at random by each individual, and
social systems are not invented de novo by each generation. On the contrary,
since the social system of any generation is in the main acquired from its
cultural heritage, from a previous generation, the goals which are attained by
the performance of roles are either sanctioned or prescribed.3 The roles
which are instrumental for the attainment of these goals are prescribed. In
short, since social systems are normative systems, social roles—unlike other
learned behavior patterns—very likely are performed not because they are
rewarding but because they are mandatory.
3. “Sanctioned” goals are goals which are culturally approved; “prescribed” goals are goals which
are culturally mandatory. Thus, though all prescribed goals are sanctioned, not all sanctioned goals are
prescribed. The goal of achieving the status of physician, for example, is a sanctioned, not a
prescribed, goal in our culture. That is, we approve of those who aspire to achieve the goal, but we do
not expect everyone to aspire to it. On the other hand the goal of curing patients is not only a
sanctioned, but a prescribed goal for physicians. From now on the expression, “culturally stipulated”
will be used to embrace both “sanctioned” and “prescribed.”
It is this imperative dimension of human social systems that has led many
social theorists to interpret cultural conformity as a function, primarily, of
special techniques of social control. Since social systems have vital social
functions, which are served only if their constituent roles are performed,
their operation requires that individuals behave in culturally desirable, rather
than in personally desired, ways. The proponents of this theory see little
relationship between personal motivation and cultural behavior. Conformity
to cultural norms, they believe, is not a matter of personality drives
primarily, but of social sanctions.
All societies, of course, employ social sanctions as a means of achieving
social control, though the specific techniques and agents of control may
differ from society to society. Thus, the sanctions may consist in quite
different kinds of rewards or punishments. Similarly the agents who
administer these sanctions (agents of control) may be one’s peers who
exercise control through the ubiquitious (informal) techniques of public
opinion—shame and praise. This may be termed “alter-ego” control.
Alternatively, the agents may be one’s superiors, who exercise control
through numerous (formal) techniques of public recognition and punitive
sanctions available to constituted authority. This may be termed “super-
alter” control. These superiors, it might be added, may be natural or
supernatural authority.
It should be obvious, however, that even a social sanctions theory of
social control, despite the antipsychological bias of many of its proponents
(Radcliffe-Brown 1957:45–52), is essentially a motivational theory. No
social sanction can compel a person to conform; it can only motivate him to
do so. As Radcliffe-Brown himself observes (1933: 531), “The sanctions
existing in a community constitute motives in the individual for the
regulation of his conduct in conformity with usage.” Thus the positive
sanction of material reward does not compel a person to perform an
economic role; rather, it motivates him to perform it because the material
reward serves as a goal to reduce some drive such as hunger or prestige.
Similarly malicious gossip or a jail sentence can induce a person not to steal
only if either of these negative sanctions are painful to him; if incarceration
or gossip were not painful, they could not compel him to conform to the
injunction. Unless the members of society have certain personality drives
which can be reduced by acquiring positive, and avoiding negative,
sanctions, it is unlikely that these sanctions would serve as techniques of
social control. In short, social sanctions serve as techniques of social control
because they function as motivational variables.
If social sanctions become incentives for action because of their cathexes
as personal (positive or negative) goals, their efficacy may be explained in
terms of our psychological model. They function as anticipated rewards or
punishments. Since these rewards and punishments are extrinsic to the
performance of a role, and since they are administered by persons other than
the actor, this type of cultural motivation4 may be termed “extrinsic cultural
motivation,” and this type of social control may be termed “extrinsic social
control.”
4. The motivation for the performance of social roles is termed “cultural motivation” because these
roles are culturally sanctioned and prescribed.
To the extent that all societies employ extrinsic social control in some
degree as a means of achieving cultural conformity, personality motivation
enables the social system to serve its vital social functions. Though the
performance of roles may be motivated by the fear of punishment or the
desire for rewards, their social functions are served regardless of the
personal motives for their performance. Even if social sanctions were the
primary technique of social control, analysis of cultural conformity could not
avoid the concept of personality.
Although extrinsic social control is universal, it does not follow that its
importance in achieving cultural conformity is paramount. Social sanctions
may be necessary in order to achieve the conformity of some individuals in
some societies almost all of the time, and of most individuals in any society
some of the time. Moreover, they are necessary to resolve those conflicts
that frequently arise between two persons or groups, both of whom are
behaving in conformity with the cultural norms. It is probably safe to
assume, however, that this type of control is only rarely the primary type in
any society; it is most prevalent in those historical periods of a society which
are characterized by anomie. Thus, it is typically found as a primary type of
control in transitional periods in which changes, either in tension-producing
or tension-reducing social institutions seriously restrict the possibility of
satisfying personality needs by culturally stipulated techniques (Sapir 1924,
“Spurious culture”). In the long run, however, further changes in the social
system will restore its tension production-reduction balance (Henry 1953:
154), so that extrinsic control is no longer primary; or the psycho- and
sociopathology that result from this cultural pathology will become so
extensive that social life is no longer viable.

Intrinsic Cultural Motivation

Personal Motives and Manifest Social Functions

If social sanctions are not the primary means of achieving cultural

conformity, it is because social roles, though prescribed, satisfy personality
needs. Fromm is undoubtedly correct when he writes:
In order that any society may function well, its members must acquire the kind of character
which makes them want to act in the way they have to act as members of the society … They have
to desire what objectively is necessary for them to do. (1944:381)

In order to understand this transformation of duty into desire we must first

understand how the normative dimension of human social systems serves to
qualify our psychological model. In this model, it will be recalled, behavior
is instigated initially by the desire to reduce a drive, and any object or event
which serves this end will do. Subsequently those objects or events which
gratify the drive may become cathected so that they function as goals. When
this happens behavior is motivated by the desire not merely to gratify a
drive, but to gratify it by attaining a particular goal. Canalization, as Murphy
(1947: chap. 8) has termed this process of drive-goal connection, is
characteristic of much motivation. But cultural motivation is unique in that
these canalizations are ordained by the cultural heritage prior to individual
experience instead of arising in the context of individual experience. By
stipulating that only a limited, out of a potentially large, number of objects
or events may serve as goals for drives, and by prohibiting all others, the
cultural heritage insists that if a drive is to be gratified at all, it must be
gratified by means of these stipulated prescribed or sanctioned goals. Thus,
though a New Guinea headhunter must bring home a head if he is to gratify
his prestige drive, an Ifaluk must not; and though an American is permitted
to gratify his hunger drive by eating roast beef, a Hindu is not.
If the goal of a behavior pattern is distinguished from its drive, much of
the dramatic diversity found in the cross-cultural record reflects the
diversity, not of man’s “nature,” but of his history and of his cognitive
ingenuity. Since man is enormously plastic, a large variety of goals may,
potentially, reduce the same drive. In the absence of biologically rigid drive-
goal connections, different societies, as a function of their unique histories
and ecologies, have “chosen” different goals for the same drives as well as
different roles for the attainment of these goals. The resultant diversity in
goals has led some anthropologists to insist that each culture is not only sui
generis but that cross-cultural generalizations are impossible to achieve.
Such a position is, functionally viewed, wide off the mark. Although cultural
goals are parochial, most human drives—because of their rootedness in a
common biology and in common conditions of social life—are probably
universal. Hence, it is generally not too difficult to demonstrate (on a fairly
high level of generality, of course) that the quite diverse goals of different
societies, as well as the roles which are instrumental for their attainment, are
functionally equivalent; they serve to gratify the same drives (Murphy
1954:628–31).
But the fact that goals are prescribed does not imply—as some theorists
take it to imply—that they do not or cannot gratify drives. On the contrary,
culturally prescribed goals may be as rewarding as non-prescribed goals. By
prescribing goals the cultural heritage does not frustrate drives, it merely
limits the number of ways in which they may be gratified. To be sure, since
man has few instincts, he must learn to perceive the prescribed goals as
rewarding. But this is the function of child-training. In the process of
socialization, children acquire not only drives, but they acquire goals as
well; they learn which objects or events—the culturally prescribed goals—
are drive-reducing. In short, socialization systems—by techniques which
cannot be described here—are institutionalized means for transforming
culturally-stipulated goals into personally-cathected goals (Erikson 1950;
Whiting and Child 1953). If the personal cathexis of a stipulated goal is
termed a “need,” it is apparent that a need-satisfaction model is more
appropriate than a drive-reduction model as a description of cultural
motivation. For in general once a prescribed goal is sufficiently cathected,
that goal which is culturally viewed as the only desirable, if not the only
possible, goal for the gratification of a drive, becomes personally viewed as
the most desired goal. Indeed, in some instances it, and no other, is perceived
as drive-reducing, so that drive-frustration may be preferred to drive-
reduction by noncathected goals. An Orthodox Hindu, for example, refuses
to eat beef, not only because it is prohibited, but because it is not desired; the
very notion of eating beef may be disgusting to him. Hence the paradox:
although evolution has produced a species characterized by the absence of
drive-goal invariants, culture produces personalities who behave as if there
were. For after cultural goals are cathected, human beings sometimes behave
as if their drive-goal connections were the only ones possible.
To conclude: social roles, like other types of learned behavior, are
performed if they are rewarding. But if the culturally stipulated goals, which
are attained by their performance, are cathected, behavior is motivated by
the expectation of attaining a goal, not by the desire to reduce a drive. To be
sure, the goal is desired because its attainment produces drive-reduction; if it
did not, it would not, in the long run, continue to be desired. But this is
precisely the point of the “need” concept: it looks, so to speak, in two
directions. On the one hand it affirms that drive-reduction is rewarding, so
that acts that do not reduce drives are not performed. On the other hand, it
denies that the desire to reduce a drive is a sufficient explanation of cultural
motivation; for when culturally stipulated goals are sufficiently cathected,
action is motivated (with the possible exception of extreme deprivation) by
the expectation of attaining these cathected goals.5
5. This notion of “need” is almost identical with the notion of “need-disposition” in Parsons’ action
theory (Parsons and Shils 1951:114–20). There are other points of convergence, as well, between the
limited formulations of this chapter and those of Parsons (Parsons and Shils 1951: parts 1 and 2;
Parsons 1951: chaps. 1–3, 6–7). The serious student of the relationship between social system and
personality is urged to read these two important volumes.
It is now perhaps clear how cultural imperatives can become personal
desires—how, in short, people can want to perform social roles. If the
performance of social roles does in fact attain those culturally stipulated
goals for whose attainment they are intended by the cultural heritage, and if
these goals have been cathected by the members of society, these roles are
performed because of the desire to attain these goals. In short, although
social roles are prescribed by the cultural heritage, their performance is
motivated by the expectation of satisfying personality needs. Though these
roles must be performed so that their functions for society can be served,
individuals desire to perform them because personal functions are thereby
served.6 Thus, for example, the American army serves an important adaptive
function for American society by defending its people against foreign
enemies. Should an individual American become identified (in the
psychoanalytic sense) with his society, he will internalize this function as a
personal drive and, therefore, he might cathect this stipulated goal as a
personal goal. Should this happen he may be motivated to become a soldier
and to perform its prescribed role because its performance satisfies a
personality need. In short, if the social function of a role is internalized as a
personal drive, its performance, which is intended to serve a social function,
serves a personal function—albeit unintentionally—as well.
Diagrammatically, this can be represented thus:

6. For a preliminary typology of functionalism, see Spiro 1953. For a detailed analysis of
functionalism as “functional consequence,” see Merton 1949. For illuminating discussions of
functionalism, based on Merton’s analysis, see Nagel 1957: chap. 10, and Hempel 1959. For a general
review of recent functionalist theory and research, see Firth 1955.
But if cultural goals are cathected as personal goals only when social
functions are internalized as personal drives, the number of roles whose
performance can serve personal functions would be small indeed. Indeed, it
is precisely because the social functions of roles only rarely become personal
drives that some theorists have stressed the importance of social sanctions as
a means, par excellence, of assuring cultural conformity. And surely the
argument is plausible. For if a role has a social function, and if the serving of
its function is not a personality need, how can its performance be motivated
by the expectation of satisfying a personality need?
This argument, however plausible, neglects to consider still another
possibility. Stipulated goals may be cathected, and therefore social roles
performed, although social functions are not internalized; and social
functions may be served, although they are unintended. As Kroeber,
generalizing from his analysis of religious change among the Kota, has put
it: “!n manipulating their culture for their personal ends, the participants
often produce a cultural effect that may be enduring, as well as attaining
their individual goal or tension reliefs” (1948:507).
This can happen in two ways: when personal and social functions are
members of the same functional class, and when they are members of
different functional classes. Both of these ways, beginning with the first, can
be illustrated by returning to the soldier role and the motives for its
performance. Since individuals exist qua members of society as well as qua
individuals, their welfare is frequently dependent upon the welfare of
society. Thus, though an individual may not internalize the adaptive (social)
function of the soldier role as a personal drive, he may nonetheless, if he
believes that his personal survival depends on the survival of his society,
cathect its goal of national defense; and he may then be motivated to
perform the role in order to satisfy this personality need. But since the social
(adaptive) function of the soldier role consists in the summation of its
personal (adaptive) functions for individuals, the performance of the role not
only serves a personal function, but it serves its social function as well—
albeit unintentionally.
There is still another way in which culturally stipulated goals can become
personality needs. Not everyone who becomes a professional soldier, for
example, is motivated to achieve or to perform this role in order to defend
either himself or his society from enemy attack. Since the promotion of
national defense is, at least in our society, one means for the attainment of
prestige and power, this culturally stipulated goal may become the cathected
goal for the reduction of these drives. The performance of the soldier’s role
may then be motivated by the expectation of satisfying power and prestige
needs. Nevertheless though the personal functions (integrative and adjustive)
and social function (adaptive) of the role are members of different functional
classes, and though its performance is intended to serve personal functions,
its social function—although unintended—is served as well.
Diagrammatically, these last two cases can be represented thus:

To sum up, any act may be viewed from at least two perspectives: motive
and function. The motive of an act is the consequence, either for the actor or
for society, which is intended by its performance; its function is the actual
consequence of its performance, either for the actor or for society. Functions
may be positive or negative; that is, the consequence of an act may
contribute to the welfare of the actor or of society or it may detract from
their welfare. (We are here concerned with positive functions only, and the
generic term, “function,” refers to positive function exclusively.) Finally acts
have intended and unintended functions. That is, the consequence of an act
may be the consequence which was intended by its performance, or it may
be one which was not intended by its performance. Since social roles have
social functions, and since acts are performed only if they have personal
functions, it has often been assumed that there is little intrinsic relationship
between personality needs and the performance of roles except in those few
instances in which an act is intended to serve both personal and social
functions. A soldier, for example, might be motivated to play his role
because he intended to serve both himself and his country.
But if acts can have unintended as well as intended consequences, it is
possible for personal and social functions to be served in the performance of
the same acts or roles. And this can happen in two ways: when their personal
functions are intended and their social functions are unintended, and when
their personal functions are unintended and their social functions intended.
In either event since the performance of the roles serves personal as well as
social functions, their performance is motivated—without the operation of
social sanctions—because they satisfy personality needs. Since these needs
consist in the personal cathexes of culturally stipulated goals, the
performance of social roles is effected by, what may be termed, “intrinsic
cultural motivation.” Alternatively, the cultural conformity which results
from this type of motivation is achieved by “intrinsic social control,” for the
control function is, as it were, built into the very fabric of the role. By
satisfying personality needs, its performance is assured, and its social
functions performed, without the necessity for sanctions extrinsic to the role
or for agents external to the actor.

Personal Motives and Latent Social Functions

Thus far it has been contended that personality plays an important part in the
operation of social systems because, by motivating the performance of social
roles, it enables the social system to serve its social functions. The
discussion, however, has dealt with the manifest functions of roles
exclusively, that is, with those social functions which, whether intended or
unintended by the members of society, are recognized by them. But social
systems, have latent functions as well, and sometimes their latent functions
are more important for society than their manifest functions. It is here,
moreover, that personality is uniquely important for the functioning of
society.
If manifest functions are those consequences of role performance which
are recognized by the members of society, latent functions are those
consequences which—whether intended or unintended—are not recognized
by them.7 That the paradox of an intended but unrecognized function is
apparent rather than real, becomes clear when one considers that motives
may be unconscious, as well as conscious. In short, manifest (recognized)
functions are served by the performance of roles when at least one of the
motives for their performance is conscious; latent (unrecognized) functions
are served when at least one of the motives for their performance is
unconscious. Hence, before analyzing latent functions, it is necssary to
examine the concept of unconscious motive.
7. Merton (1949), whose now-classic analysis of functionalism remains the incisive treatment of
this subject, and who first introduced the terms “manifest” and “latent” into functional analysis,
ignored a potentially powerful mode of analysis by merging “intention” and “recognition.” As he
defines them, manifest functions are those which are both intended and recognized, while latent
functions are those which are neither intended nor recognized. Since manifest functions—as we have
seen—may be unintended, and since latent functions—as we shall see—may be intended, intention
and recognition may vary independently.
If the motive for behavior consists in an intention to satisfy a need by
performing a particular act (and if a need consists in a drive and a goal), a
motive may be unconscious, i.e., unrecognized, in any one or all of these
three dimensions. Thus a drive, its goal, and the desired means for the
attainment of the goal may all be unconscious. With the exception of
neurotic, i.e., idiosyncratic, repression, if any or all of these dimensions of a
motive are unconscious in a typical member of a society, the motive has
generally been rendered unconscious because of a cultural prohibition or
because of its systematic frustration. In the latter case, since need-frustration
as well as the memory of need-frustration are painful, repression of the
frustrating experience as well as of the need is one possible defense against
pain. We shall confine the discussion to the former basis for repression.
Thus, for example, the cultural heritage may prohibit any reduction of the
sex drive, as in sacerdotal celibacy; or it may prohibit a desired goal for its
reduction, such as intercourse with kinsmen who fall within the boundaries
of the incest taboo; or it may prohibit a desired means for the attainment of
the goal, such as some “perverted” technique of sexual relations. In short the
cultural heritage not only provides means of need-satisfaction for an animal
without instinctive means for drive-reduction, but—by prescribing these
means—it prohibits other means which this relatively plastic and
imaginative animal may come to prefer. Moreover it may completely
prohibit any conceivable (manifest) expression or reduction of certain drives.
But motives do not disappear simply because they are prohibited. Even if
the cultural prohibition is internalized as a personal norm, the culturally
prohibited canalization may continue to persist as a personally preferred
canalization, and the culturally prohibited drive may continue to seek
expression. The resultant incompatibility between internalized norm and
personal desire leads to inner conflict which must be “handled” in some way.
If these personally preferred, but culturally prohibited, canalizations are
stronger than the internalized cultural prohibition, they may be expressed
directly. If, then, the resultant behavior is categorically prohibited, it is
deemed criminal or psychologically aberrant (depending on the culture) by
the members of society. Alternatively, if the behavior is culturally aberrant,
but not clearly prohibited, it may be viewed as a cultural innovation—that is,
as a new, but culturally acceptable, behavior pattern.
On the other hand, should the internalized cultural prohibition be stronger
than the personal desire, the inner conflict may be resolved by repressing the
desire. The prohibited motive, in short, becomes unconscious. But
unconscious as well as conscious motives seek expression and satisfaction.
They may be expressed (and satisfied) in neurosis and psychosis; in private
fantasy (day-dreams and night dreams); in symbolic, but culturally creative,
ways (artistic and scientific work); or, and more germane to this chapter, in
the performance of culturally prescribed roles. Since unconscious motives
cannot be satisfied directly—if they could they would not be unconscious—
they may thus seek indirect satisfaction in the performance of culturally
sanctioned behavior. In short, in addition to its conscious motivation,
culturally sanctioned behavior, including role behavior, may be
unconsciously motivated as well. Since, in the latter case, the performance of
a role is motivated by an unconscious as well as by a conscious intention of
satisfying a need, the role may have unrecognized though intended personal
functions; and these, in turn, may produce unrecognized and unintended
social functions. This thesis may be illustrated by examples from two
societies: warfare among the Sioux Indians of the American Plains, and
religion among the Ifaluk of Micronesia.
Diagrammatically, the relationship between the motives for the
performance of the Sioux warrior role and its various personal and social
functions can be represented thus:
 Sioux warfare was motivated by two conscious needs: prestige and
protection from enemies. In satisfying these needs for prestige and
protection, warfare served manifest personal (integrative) and social
(adaptive) functions. It may be assumed, it served an unintended
(integrative), but manifest social function, as well—the promotion of social
solidarity by the creation of esprit de corps among the warriors.
 But the “choice” of warfare as a preeminent institutionalized means of
obtaining prestige leads us to suspect that the conscious motives for the
performance of the warrior role, though genuine, were not its only motives.
Warfare is an aggressive activity, why did the Sioux act aggressively when
the objective threat from the enemy was slight? Sioux war parties, it will be
recalled, attacked rather than defended; they preferred to attack when their
“enemies” were least prepared, that is, when they constituted no threat;
young bucks had to be restrained from going on the warpath, rather than
having to be encouraged to do so. And why did they seek prestige through
aggression when, as the cross-cultural record reveals, there are many
nonaggressive roles through which prestige can be obtained? Sioux warfare
apparently was motivated not only by the conscious needs of prestige and
protection, but by yet another, unconscious, motive—hostility against their
fellows.
As is the case in most societies, Sioux socialization, as well as the
conditions of adult Sioux social life, created in each new generation a motive
for aggression against their fellows (Erikson 1939, 1945). Like most
societies, moreover, the cultural heritage of the Sioux prohibited physical
aggression against the in-group. However, only one of the three dimensions
of this motive was prohibited. Neither the drive itself (hostility) nor the
means of its reduction (physical aggression), but only its object (the in-
group), was prohibited. It was assumed, then, that the specific dimension of
physical aggression against fellows was repressed, i.e., rendered
unconscious. But by displacing hostility from the in- to the out-groups, this
motive could now be expressed. This motive, one may suggest, sought
satisfaction in, and was therefore important in the motivation of, Sioux
warfare. In addition to their motives of prestige and protection, Sioux war
parties were also motivated by aggression. In satisfying this motive, the
warrior role served a latent personal function (integration), as well as its
manifest personal and social functions.
When the performance of roles is motivated by unconscious needs, it
serves unintended and unrecognized social functions as well as intended but
unrecognized personal functions. What possible unintended and
unrecognized function for society was served by the Sioux institution of
warfare? By displacing hostility, and its subsequent aggression, in warfare
against the outgroup, the warrior role protected Sioux society from the
aggression of its own members (adaptive function). Had the original hostility
not been displaced and subsequently gratified in socially sanctioned
aggression, it might have sought un-disguised and, therefore, socially
disruptive expression. It might have sought expression in other ways as well.
Indeed, Erikson interprets the sun dance, and its painful consequences for its
participants—staring into the sun and tearing of skewers from their flesh—as
the turning of aggression inward. It might be suggested, then, that in the
absence of war, even more aggression would have been turned against the
self. Hence, the performance of the warrior role served a latent, unintended
personal function, as well as its latent unintended social function.
But to return to the social functions of Sioux warfare: by deflecting
hostility from in- to out-group, the preponderance of positive over negative
sentiments concerning the members of the group was increased, thereby
promoting in-group solidarity (integrative function). Neither of these
unintended and unrecognized social functions would have been served had
the performance of the warrior role been motivated exclusively by the
motives of prestige and protection. Moreover, those anthropologists who are
unaware of, or uninterested in, the latent personal functions of roles—
because unaware of, or uninterested in, unconscious motives—would remain
ignorant of the important latent social functions which are served by this
role, and of the general functional significance of Sioux warfare within the
total social system.
The second example of the relationship between unconscious motives and
social functions not only illustrates the importance of unconscious
motivation in the functioning of social systems, but it also illustrates how a
society and its social system may be affected by the intrinsic motivation of
another cultural system—religion. Most public religious rituals in the
Micronesian atoll of Ifaluk are either therapeutic or prophylactic in nature;
they are designed to maintain or restore health by exorcising malevolent
ghosts (who cause illness by possessing their victims), or by preventing
these ghosts from executing their intentions in the first place. It is not within
the province of social science to decide whether one of the manifest,
intended, functions of these rituals—defeat of the ghosts—is served; the
Ifaluk, of course, believe that it is. But these rituals serve other manifest
functions to which the behavioral scientist can testify. By their performance
the twin fears of illness and of attack by ghosts are reduced (manifest
intended personal function), and by assembling and acting in concert for the
achievement of a common end, good fellowship is strengthened (manifest
unintended social function).
But the performance of these rituals requires another motive in addition to
its therapeutic motive. These are aggressive rituals in which malevolent
ghosts are attacked and, it is hoped, routed. It requires little insight to infer
that hostility, as well as fear, motivates the performance. Indeed, the Ifaluk
are quite consciously hostile toward the ghosts. But though consciously
hostile to ghosts, the Ifaluk, like all people, have occasion to be hostile to
their fellows, particularly to their close kinsmen. By displacing hostility
from fellows to ghosts, their hostility is acceptable, and their subsequent
aggressive motive can be gratified in a socially sanctioned manner in the
performance of these rituals (Spiro 1953a). A latent personal function
(integrative) of these rituals, then, consists in the opportunity which their
performance affords for the satisfaction of this aggressive need.
As in the Sioux example, however, the performance of these rituals also
serves a latent (unintended) social function, one which is vital for this
society. The Ifaluk social system, based on the strongly held values of
sharing, mutual aid, and kindliness, is highly cooperative. If the Ifaluk were
unable to express aggression symbolically in ritual, it is not improbable that
their hostility would eventually seek direct expression. If this were to
happen, the probability of physical survival on an atoll, six-tenths of a mile
square, would be effectively reduced. By serving to deflect aggression onto
malevolent ghosts, the performance of these rituals effectively increases the
chances for survival. Moreover, as in the case of the Sioux, the belief in
malevolent ghosts, which permits the displacement of hostility from the in-
group to the wicked out-group, assures the persistence of the warm
sentiments which the Ifaluk harbor towards each other. Hence the
psychological basis for their cooperative social system—probably the only
kind of system which is viable in this demographic-ecological balance—is
preserved. In short, by serving its latent personal function, this ritual is also
able to serve the latent social functions of promoting the group’s survival
and of preserving the viability of its social system (Spiro 1952).
Diagrammatically, the relationship between the motives for the
performance of Ifaluk rituals, which attack and exorcise malevolent ghosts,
and their personal and social functions can be represented in the following
way:
 This discussion of unconscious motives, and of the latent social functions
served by social roles (and other types of cultural behavior) whose
performance is motivated by them, has been somewhat extended because,
with the exception of culture-and-personality research, they are ignored in
most analyses of social systems. Unconscious motives are frequently
dismissed by social scientists as irrelevant to an understanding of society.
“Oh,” it is often said, “these unconscious motives may be important for
personality, but we re interested in the study of society.” This analysis has
attempted to demonstrate that anyone interested in society should also be
interested in unconscious motives. They are as important for the student of
society as for the student of personality, not only because they motivate the
performance of social roles but because the latent social functions which
they enable these roles to serve are often more important than those which
are served under conscious motivation. It should be emphasized, however,
that unconscious motives, however important, are not the only motives of
behavior, that conscious motives are not merely rationalizations. Though this
may sometimes be the case, the assumption that only unconscious motives
are genuine is as fallacious as the contrary assumption. If the conscious
motives of Sioux warfare and Ifaluk ritual are not sufficient explanations for
their performance, neither are the unconscious motives: both are necessary,
both are genuine, neither is sufficient. To assume that only unconscious
motives are genuine is to perpetuate that vulgar interpretation of
psychoanalytic theory in which schoolteaching, for example, is “nothing
but” the sublimation of an unconscious sexual motive, or surgery is “nothing
but” the displacement of unconscious aggression.
Theories of social systems that ignore unconscious motives are not only
truncated, but when social analyses which are based on such theories are
applied by administrators, they often lead to unfortunate results. If we were
to assume, for example, that Sioux warfare or Ifaluk religious rituals are
means merely for obtaining prestige or reducing anxiety concerning illness
respectively, and that by achieving these ends they also promote social
solidarity—the typical social anthropological functionalist analysis—then it
is a fair administrative conclusion that these “savage” and “superstitious”
practices can be abolished without harm to society as long as the “civilized”
practices with which they are replaced are their functional equivalents, as
long, that is, as the new practices are also means for obtaining prestige, for
reducing anxiety concerning illness, and for promoting social solidarity. But
despite these good intentions, the new practices are not the functional
equivalents of the old if they do not serve, as well, the latent personal
function of displacing unconscious hostility. Unless this function is
achieved, substitutes cannot serve the latent social function of deflecting
aggression from the in-group. Hence, this unconscious motive may seek
expression in numerous dysfunctional ways—dysfunctional both for
individuals and society. It may be expressed directly, leading to crime, or
indirectly, leading to drunkenness, etc.; it may be inverted, leading to anxiety
and depression (“race suicide”), and so forth. By ignoring the importance of
unconscious motivation in social behavior, the attempt of well-intentioned
administrators (acting upon the findings of psychologically uninformed
researchers) to substitute “unobjectionable” for “objectionable” native
practices has often been a history of grave disappointments to the
administrator and sordid results for the natives.
 It should be strongly emphasized that although personality needs are
satisfied in and therefore motivate the performance of social roles, a person’s
personality cannot necessarily be inferred from a knowledge of the roles he
performs. In the first place, although this chapter is concerned with the
relationship between personality and the social system, it is obvious that
only part of the personality is relevant to and is expressed through the social
system. The relationship between personality and other cultural systems
(religion, art, science, etc.), as well as those private aspects of personality
that are not caught up in the sociocultural net (Murphy 1958: part 3), are
deliberately ignored. In short, a description of a person’s various social roles
would not lead to an exhaustive description of his personality.
More important, however, for our purposes, a knowledge of a person’s
social roles would not even lead to an accurate prediction of those aspects of
his personality that are caught up in their performance. For, as this section
has attempted to show, since different goals may be cathected by the same
drive and since different roles may be instrumental for the attainment of the
same goal, “a high degree of role differentiation,” as Kaplan has put it, does
not necessarily require “a similar degree of differentiation at the personality
level” (1957: 100). At the same time, since the same goal may be cathected
by different drives, and since the same role may be instrumental for the
attainment of different goals, a high degree of personality differentiation
does not necessarily require a similar degree of differentiation at the social
system level. Thus, (1) different drives may be canalized by the same goal,
which is attained by the performance of the same role; (2) the same drive
may be canalized by different goals, which are attained by the performance
of different roles; and (3) different drives may be canalized by the same goal
which is attained by the performance of different roles. These alternatives
are shown in the following diagrams.
 But if this is true within a society, it is equally true across societies. Since
there are fewer drives in man than there are goals in all his societies, and
since there are fewer goals in all human societies than there are roles in their
social systems, it is reasonable to expect fewer modal personality systems
than social systems. On the other hand, since drives, goals, and roles may
vary independently of each other, it is possible for different modal
personality systems to be associated with similar social systems, and for
similar modal personality systems to be associated with different social
systems.
If this is so the student of social systems, who is interested in their
motivational well-springs, must at the same time be a student of personality;
and statements about the relationships between personality and social
systems must be based on personality investigations, and not inferred from a
description of social systems (Inkeles and Levinson 1954). Personality
investigation may entail the use of psychological instruments, such as
projective tests (Hallowell 1955), the analysis of dreams (Eggan 1952), the
collection of life histories (Kluckhohn 1945), and depth interviewing. It may
also be based, however, on the observation of behavior when viewed from
the perspective of, and interpreted in terms of, psychodynamic personality
theory. For if the same set of activities can serve both personal and social
functions, the same set of activities may be viewed from a personality
perspective (as a means for serving personal functions) or from a social
system perspective (as a means for serving social functions). If one’s focus is
on society and on those adaptive, adjustive, and integrative prerequisites of a
viable social life, a given set of activities is analyzed as a role within the
social system. If, on the other hand, one’s focus is on an individual and on
the adaptive, adjustive, and integrative prerequisites of a viable individual
life, the same set of activities is analyzed as a means for satisfying the needs
of the personality system. This last technique uses a powerful, but rare
instrument—a sensitive observer.

Internalized Cultural Motivation

Since intrinsic cultural motivation is based on the personal cathexis of
culturally stipulated goals, it obviously cannot serve as a technique of social
control when these goals are not cathected. There are various conditions
which reduce the probability of goal cathexis. The following are probably
most important: (a) the goals of many taboos and prohibitions, since they
lead to frustration, may increase rather than reduce the intensity of drives
(Freud 1930); (b) in societies undergoing rapid culture change, many new
goals will not reduce extant drives (Hal-lowell 1945); (c) a similar situation
will obtain in the case of subordinate groups, whose culture has largely been
imposed by a dominant group; (d) cathexis may be withdrawn from
previously cathected goals because of the realization that they cannot be
achieved (Merton 1938).
Even when goals are cathected it does not necessarily follow that the
culture patterns or roles that attain them will be performed. There are other
possibilities. Though the goal is desired, the role which is prescribed for its
attainment may be odious. Thus, though everyone may desire clean public
latrines, no one may desire to perform the role of latrine attendant. Again, a
nonsanctioned means for the attainment of a desired goal may be perceived
as more efficient or as less burdensome than the sanctioned role. Similarly,
the cultural goal may be scarce, so that not all who strive for its attainment
can be successful. Hence, competitive anxiety may motivate the
performance of proscribed, but more efficient, techniques. Finally, the social
structure, particularly in a stratified society, may effectively preclude certain
categories of persons from performing the roles which attain the goals
(Merton 1938).
 In all of these situations social conformity will be achieved only by some
technique of social control other than—or in addition to—intrinsic cultural
motivation. Extrinsic social control is one such technique; but it is not the
only one. For the importance of personality needs in the motivation of social
roles is not restricted to intrinsic cultural motivation. The latter type of
motivation is ultimately based on two kinds of personality needs—id and
ego needs, in psychoanalytic vocabulary. But personality has superego needs
as well; and many roles may be performed (though id and ego are not
satisfied, and may even be frustrated) because their performance satisfies
superego needs.
If the performance of roles is motivated by the expectation of satisfying
superego needs, social control is achieved by, what we may term,
“internalized cultural motivation.” For cultural conformity in this instance is
achieved, not through external sanctions (extrinsic control), nor by intrinsic
goals (intrinsic control), but by internalized norms. To put it in terms we
have been employing, if extrinsic control is achieved (in the case of positive
sanctions) by the cathexis of the social sanction, and if intrinsic control is
achieved by the cathexis of the cultural goal, internalized control is achieved
by the cathexis of the cultural norm.
There has been a great deal of discussion concerning the internalization of
norms. Some writers, following Ruth Benedict (1946: 222–27, 288–89),
have suggested that norm-internalization is a phenomenon restricted to
certain types of societies and absent from others. Cultures which give rise to
norm-internalization are termed “guilt-cultures,” for cultural conformity is
motivated by guilt. Those which do not produce norm-internalization are
termed “shame-cultures,” for the members of society conform to cultural
norms only when their fellows are present to shame them. Hence, in
societies with shame cultures extrinsic control is necessary to ensure cultural
conformity—assuming that the performance of roles is not intrinsically
motivated.
Although shame obviously operates as a control technique in any society,
the validity of this shame culture—guilt culture dichotomy is open to
question (Piers and Singer 1953). Since social systems are, to a great extent,
normative—many of their constituent roles and goals are prescribed by the
cultural heritage—it is improbable for the members of any society not to
have internalized these norms. If norms were not internalized, parents would
have none to transmit to their children because, ex hypothesi, they would not
have internalized any in the course of their own socialization. Further, if no
one has internalized the norms, who, in societies with shame cultures, would
do the shaming? The existence of agents of shame implies that at least some
members of society have internalized at least some norms.
In short, one may argue that although in any society there may
hypothetically be some individuals who have internalized very few norms
(the so-called psychopaths), and many individuals who have not internalized
some of the norms, in all societies most individuals not only (a) learn about
their cultural norms, but they also (b) accept them, (c) evaluate their own
acts in accordance with them, and (d) experience anxiety (“moral anxiety”)
should they desire to violate them. This anxiety serves as an important
deterrent to norm violation. Indeed, even in societies whose cultures
correspond most closely to the description of the ideal shame culture, “…
blame, ridicule, or holding up to shame are controls only if they express
commonly accepted values and correspond to the promptings of the
superego” (Nadel 1953:272).
How does this moral anxiety develop? And what does it represent? To
answer the second question first, this anxiety presumably represents the
largely unconscious expectation of punishment, as distinguished from the
rational, conscious fear of being punished. This distinction must be
explained.
The individual who has internalized a norm, and not merely learned about
it, perceives his anticipated violation of it as a transgression and hence as
deserving of punishment. This perception induces anxiety (the anticipation
of punishment). The mere intention of committing an act which he himself
labels as a “transgression,” or of not performing an act which he deems
compulsory, leads him to expect that his behavior (which in his eyes is
deserving of punishment) will in some way be punished. Where the
individual believes that punishment is his due, “expectation of punishment”
is but another term for “moral anxiety.”
On the other hand, the individual who has merely learned about the norm,
but has not internalized it, suffers no moral anxiety as a consequence of his
anticipated violation of it. Because he himself has not internalized the norm,
he does not (though others may) consider his anticipated violation to be
deserving of punishment, since he does not consider his act to be “wrong.”
In short, he experiences no moral anxiety. He may, of course, experience
considerable anxiety about the punishment which would be meted out to him
were he caught. In moral anxiety, however, it is not the fear that one might
be punished if caught, but the belief that one merits punishment that evokes
anxiety.
Moral anxiety, therefore, has both drive and cue properties. It informs the
individual that his anticipated behavior is wrong (worthy of punishment),
and that its performance will lead to punishment; and it motivates him to
reduce the anxiety by refraining from transgression. Hence, the anxiety
serves as a motive for conformity.
Since moral anxiety is not innate, how is it acquired? So far as our present
knowledge permits, we may suggest that it arises out of certain universal
features of human socialization systems. In all societies, agents of
socialization are not only trainers, but they are also nurturers, satisfying the
child’s most important need—the need for love. To the extent that these
agents employ rewards and punishments as part of their training methods,
and to the extent that such rewards and punishments are, for the child,
symbolic of their love, the child is motivated to comply with the demands of
these “significant others” (Mead 1934) in order to obtain their love or,
conversely, to preclude its withdrawal. Through their ability to give and
withhold love, the child not only learns what the agents of socialization
judge to be good and bad behavior, but he also learns to concur in their
judgment; in short, he models his behavior in accordance with their norms.
He learns to accept their judgment as his own because behavior which these
significant others judge to be bad is indeed “bad” for him—it leads to the
withdrawal of love (punishment) by those whose love he so strongly desires.
Since he agrees that certain acts are “bad,” and therefore deserving of
punishment, his mere intention to transgress leads to the anticipation of
punishment (moral anxiety). He has developed a superego, or a conscience.
But having questioned the validity of one distinction—that between
shame and guilt cultures—it is necessary to introduce another. The superego
has been implicitly defined operationally as the configuration of those
expectations of punishment, experienced as anxiety (either conscious or
unconscious) that are evoked by the anticipated violation of an internalized
cultural norm. But we have not yet specified the agent of punishment, the
“significant other” from whom punishment (withdrawal of love) was
originally expected. Two types of superego, based on the agent of
anticipated punishment, can be distinguished. This agent may be outside the
individual or “within” him. It is our hypothesis that societies in which the
child is trained by only a few agents of socialization, who themselves
administer punishments, produce individuals who not only internalize the
norms of the socializing agent but who “introject” the agent as well. This
introjected figure, then, is the significant other for such individuals and it is
withdrawal of its “love” that constitutes the anticipated punishment. Since
this punishment, when it comes—and it comes after the transgression is
committed—is experienced as guilt (“pangs of conscience”), this type of
superego may be termed “guilt-oriented.”
Other societies, we believe, in which the child is trained by a number of
socializing agents, or in which the trainers discipline the child by claiming
that other agents will punish him, do not produce individuals with “guilt-
oriented” superegos. For, though these individuals internalize the norms of
the socializing agents, they do not introject the agents themselves. Since the
significant others continue to remain external, it is withdrawal of the love of
others that constitutes the anticipated punishment. Because this punishment,
when it comes, is experienced as shame, this type of superego may be
termed “shame-oriented.” Of course, these two types of superego represent
the polar extremes, conceived as ideal types, of a superego continuum. Most
superegos would represent admixtures of the two, weighted toward one or
the other end of the continuum.
A shame-no less than a guilt-oriented superego constitutes a conscience.
By producing anxiety concerning anticipated punishment, both types inform
the individual that his anticipated behavior is wrong, and both motivate him
to refrain from transgressing a norm, whether others are present or not.
Nevertheless, they function differently after a transgression has occurred. A
person with a guilt-oriented superego suffers guilt when he transgresses,
even if no one perceives his transgression, because the agent of punishment
(the introjected figure) is always with him. However, a person with a shame-
oriented superego does not suffer shame when he transgresses unless others
witness his transgression, for no agent of punishment (the external others) is
present. Instead of experiencing actual punishment (shame), he continues to
anticipate punishment; he suffers from anxiety.8 This anxiety may be so
painful that it may lead some persons who live in societies with so-called
shame-cultures to commit suicide. Incidentally, this fact is sufficient to cast
doubt on the validity of the shame-culture guilt-culture dichotomy. The
Japanese, who allegedly have a shame-culture, may be driven to suicide
when they perceive themselves to have lost face, even in the absence of any
other perceiver. In the terms we have been employing, the Japanese would
be said to have shame-oriented superegos; they experience anxiety when
they anticipate performing a forbidden act or not performing a prescribed
act. After committing the transgression, they continue to anticipate
punishment, anxiety mounts, and suicide represents the last desperate
attempt to remove the anxiety.
8. For an empirical demonstration of this process, see Spiro 1958: chap. 15, from which part of this
discussion, with permission of the publisher, is taken.
To conclude, then, regardless of the type of superego that is preponderant
in the personalities of the members of any society, cultural conformity is
frequently achieved by means of internalized cultural motivation. Though
the goal attained by the performance of a role may not be cathected, and
though a means other than the prescribed role may be preferred, the role may
nevertheless be performed (and its social functions thereby served) without
the necessity for extrinsic control. If the members of society have cathected
and internalized their cultural norms, conformity with these norms serves to
reduce the drive of moral anxiety. In short, in internalized, as well as in
intrinsic, cultural motivation the members of society have acquired “the kind
of character which makes them want to act in the way they have to act …”

Conclusions
In the past, when the behavioral sciences were still reacting against
instinctivist theories of social behavior, the relationship between social
system and personality was viewed as primarily asymmetrical. Personality
was viewed as a relatively passive agent—affected by, but not affecting, the
social system. Faris (1937: ch. 3), for example, refers to personality as the
“subjective aspect of culture.” Recent work in culture-and-personality,
however, has tended to conceive of the social system—personality
relationship as more nearly symmetrical. These studies have suggested that
although personality is, indeed, affected by the social system, the social
system, in turn, is affected by personality.
This changing conception of the relationship between personality and
social system has had its influence on the study and analysis of social
systems by culture-and-personality theorists. Instead of merely asking how
the social system influences the development and structuring of personality,
we are now equally interested in how personality affects the functioning of
social systems. And, in general, it seems to be agreed that there is feedback
between social system and personality such that the social system creates
those personality needs which, in turn, are satisfied by and motivate the
operation of the social system (Kardiner 1939). Since society has but one
social system, while the component members of society have different
personalities, this feedback is effected because the component roles of the
social system can satisfy different needs, and its socialization system
produces common needs, or a modal personality (DuBois 1944).
This chapter has been exclusively concerned with the impact of
personality on the social system, and specifically on the importance of
personality for the motivation of role performance. Since the social system
can serve its functions for society only if its component roles are performed,
every society is confronted with the problem of social control—the problem
of getting people to behave in conformity with cultural norms. By supplying
the psychological basis for cultural motivation, personality is a vital
instrument in society’s attempt to achieve social control. It serves as such an
instrument in three ways.
In the first place, although society provides sanctions as a means for
achieving social control, these sanctions are effective only if the members of
society have drives which can be reduced by the attainment of these goals. If
this is the case these sanctions are cathected, and thereby become personality
needs which motivate role performance. Second, if the cultural norms, which
prescribe the performance of the role, are internalized by the members of
society, nonconformity induces anxiety. Since this anxiety can be reduced by
the performance of the role, conformity with these norms becomes a need
which motivates role performance. Finally, the prescribed goals which are
attained by role performance are, themselves, cathected and, hence, serve as
per sonality needs to motivate the performance of roles.
These three types of control have been termed, extrinsic, internalized, and
intrinsic, respectively. We may summarize their differences and similarities,
as follows: (a) In extrinsic control which is based on positive social
sanctions, and (b) in intrinsic control which is based on manifest personal
functions, the performance of roles is motivated by the desire to obtain a
rewarding goal—either the cathected social sanction or the cathected goal of
the role, (c) In extrinsic control which is based on negative social sanctions,
(d) in internalized control, and (e) in intrinsic control which is based on
latent personal functions, the performance of roles is motivated by the desire
to avoid pain—in the forms of physical or social punishment, moral anxiety,
or unrelieved needs, respectively.

Note
Reprinted with permission of Harper & Row, Publishers, Inc., from Studying Personality Cross-
Culturally, edited by Bert Kaplan, pp. 93–127. © 1961 by Harper & Row, Publishers, Inc.

References

Aberle, D. F., et al. 1950. The functional prerequisites of a society.

Ethics 9:100–111.
Beach, F. 1955. The descent of instinct. Psychological Review 62:401–
10.
Benedict, R. 1946. The chrysanthemum and the sword. Boston:
Houghton Mifflin.
Bolk, L. 1929. Origin of racial characteristics in man. American
Journal of Physical Anthropology 13:1–28.
Carpenter, C. 1934. A field study of the behavior and social relations of
the howling monkeys (Alouatta paliata). Comparative Psychology
Monographs, vol. 10.
Carpenter, C. 1940. A field study in Siam of the behavior and social
relations of the Gibbon (Hylobates lar). Comparative Psychology
Monographs, vol. 16.
Cassirer, E. 1944. An essay on man. New Haven: Yale University Press.
Darling, F. 1937. A herd of red deer. London: Oxford University Press.
DuBois, C. 1944. The people of Alor. Minneapolis: University of
Minnesota Press.
Eggan, D. 1952. The manifest content of dreams: A challenge to social
science. American Anthropologist 54:469–85.
Erikson, E. H. 1939. Observations on Sioux education. Journal of
Psychiatry 7:101–56.
_______ 1945. Childhood and tradition in two American Indian tribes.
Psychoanalytic Study of the Child 1:319–50.
_______ 1950. Childhood and society. New York: Norton.
Faris, E. 1937. The nature of human nature. New York: McGraw-Hill.
Firth, R. 1954. Social organization and social change. Journal of the
Royal Anthropological Institute 84:1–20.
_______ 1955. Function. Yearbook of Anthropology 1:237–58.
Freud, S. 1930. Civilization and its discontents. New York: J. Cape and
H. Smith.
Fromm, E. 1944. Individual and social origins of neurosis. American
Sociological Review 9:380–84.
Hallowell, A. 1945. Sociopsychological aspects of acculturation. In The
science of man in the world crisis, ed. R. Linton. New York:
Columbia University Press.
----------- 1950. Personality structure and the evolution of man.
American Anthropologist 52:159–73.
----------- 1954. The self and its behavioral environment. Explorations
2:106–65.
----------- 1955. The Rorschach test in culture and personality studies. In
Developments in the Rorschach technique, vol. 2, ed. B. Klopfer et
al. Yonkers-on-Hudson: World Book Company.
Harlow, H. 1953. Motivation as a factor in the acquisition of new
responses. Nebraska Symposia in Motivation 4:24–49.
Hempel, C. 1959. The logic of functional analysis. In Symposium on
sociological theory, ed. G. Gross. Evanston: Row, Peterson and Co.
Henry, J. 1953. Towards a system of socio-psychiatric invariants: A
work paper. Journal of Social Psychology 37:133–61.
Hilgard, E. 1956. Theories of learning. New York: Appleton-Century-
Crofts.
Hine, R., and N. Tinbergen. 1958. The comparative study of species-
specific behavior. In Behavior and evolution, ed. A. Roe and G.
Simpson. New Haven: Yale University Press.
Inkeles, A., and D. Levinson. 1954. National character: The study of
modal personality and sociocultural systems. In Handbook of social
psychology, ed. G. Lindzey. Cambridge: Addison-Wesley.
Kaplan, B. 1957. Personality and social structure. In Review of
sociology, analysis of a decade, ed. J. Grittier. New York: John
 Wiley and Sons.
Kardiner, A. 1939. The individual and his society. New York: Columbia
University Press.
Kluckhohn, C. 1945. The personal document in anthropological
science. In
Use of personal documents in history, anthropology, and sociology, ed.
L. Gottschalk, C. Kluckhohn, and R. Angel. Social Science
Research Council, Bull. no. 53.
______ _ and H. Murray, with the collaboration of D. Schneider. 1953.
Personality in nature, society, and culture. New York: Knopf.
Kroeber, A. L. 1948. Anthropology. New York: Harcourt, Brace.
La Barre, W. 1954. The human animal Chicago: University of Chicago
Press.
Langer, S. 1942. Philosophy in a new key. Cambridge: Harvard
University Press.
Lehrman, D. 1953. A critique of Konrad Lorenz’ theory of instinctive
behavior. Quarterly Review of Biology 28:337–63.
Lewin, K., T. Dembo, L. Festinger, and P. Sears. 1944. Level of
aspiration. In Personality and the behavior disorders, ed. J. Hunt.
New York: Ronald Press.
Linton, R. 1936. The study of man. New York: Appleton-Century.
Malinowski, B. 1944. A scientific theory of culture. Chapel Hill:
University of North Carolina Press.
Mayr, E. 1958. Behavior and systematics. In Behavior and evolution,
ed. A. Roe and C. Simpson, New Haven: Yale University Press.
Mead, G. 1934. Mind, self, and society. Chicago: University of Chicago
Press.
Merton, Robert K. 1949. Social theory and social structure. New York:
Free Press.
_______ 1957. The role set. British Journal of Sociology 8:106–20.
Miller, N. E., and J. Dollard. 1941. Social learning and imitation. New
Haven: Yale University Press.
Montagu, A. 1951. An introduction to physical anthropology.
Springfield, 111.: Thomas.
Murphy, G. 1947. Personality: A biosocial approach to origins and
structure. New York: Harper.
_______ 1954. Social motivation. In Handbook of social psychology,
ed. G. Lindzey. Cambridge: Addison-Wesley.
_______ 1958. Human potentialities. New York. Basic Books.
Nadel, S. F. 1953. Social control and self-regulation. Social Forces
31:265–73.
_______ 1957. The theory of social structure. London: Cohen and
West.
Nagel, E. 1957. Logic without metaphysics. Glencoe, 111: The Free
Press.
Nebraska symposia on motivation, ed. M. R. Jones. 1953. Lincoln:
University of Nebraska Press.
Newcomb, T. M. 1950. Social psychology. Dryden: New York.
Parsons, T. 1951. The social system. Glencoe, 111.: The Free Press.
______, and E. Shils. 1951. Toward a general theory of action.
Cambridge: Harvard University Press.
Piers, G. and M. B. Singer. 1953. Shame and guilt: A psychoanalytic
and a cultural study. Springfield: Charles C. Thomas.
Radcliffe-Brown, A. R. 1933. Sanctions, social. Encyclopedia of the
social sciences 13:531–34.
_______ 1957. A natural science of society. Glencoe, 111.: The Free
Press.
Rapaport, D. 1951. The conceptual model of psychoanalysis. Journal of
Personality 20: 56–81.
Riesman, D. 1950. The lonely crowd. New Haven: Yale University
Press.
Roheim, G. 1943. The origin and function of culture. New York:
Nervous and Mental Disorder Monograph Series, vol. 63.
Sapir, E. 1924. Culture, genuine and spurious. American Journal of
Sociology 29:401–29.
Spiro, M. E. 1951. Culture and personality: The natural history of a
false dichotomy. Psychiatry 14:19–46.
_______ 1952. Ghosts, Ifaluk, and teleological functionalism.
American Anthropologist 54:497–503.
_______ 1953a. Ghosts: An anthropological inquiry into learning and
perception. Journal of Abnormal Social Psychology 48:376–82.
_______ 1953b. A typology of functional analysis. Explorations 1:84–
95.
_______ 1958. Children of the kibbutz. Cambridge: Harvard University
Press.
Thompson, W. 1958. Social behavior. In Behavior and evolution, ed. A.
Roe and G. Simpson. New Haven: Yale University Press.
Tolman, E. C. 1951. Collected papers in psychology. Berkeley and Los
Angeles: University of California Press.
Tylor, E. B. 1874. Primitive culture. Chicago: Brentano.
Wheeler, W. M. 1928. The social insects—their origin and evolution.
New York: Harcourt, Brace.
White, L. 1940. The symbol: The origin and basis of human behavior.
Philosophy of Science 7:451–63.
Whiting, J. W. M., and I. Child. 1953. Child training and personality.
New Haven: Yale University Press.
Wissler, C. 1923. Man and culture. New York: Thomas Y. Crowell.
Zuckerman, S. 1932. The social life of monkeys and apes. New York:
Harcourt, Brace.
6
Religious Systems as Culturally
Constituted Defense Mechanisms

SINCE THE RANGE of beliefs, values, and rituals related to supernatural

beliefs and events is enormous, it is obvious, as Durkheim observed long
ago, that no belief, value, or ritual is intrinsically identifiable as “religious.”
Since the “religious,” on the contrary, is a quality capable of being attached
to almost any instance of these three dimensions of religious systems, the
latter, to use a modern idiom, are in large measure projective systems. It is
this characteristic of religion that poses a problem which has long
confronted its students, and which comprises the problem of this chapter: If
religious systems are indeed projective in character, how can we be sure
that religious behavior is not abnormal behavior, requiring psychiatric,
rather than sociocultural, analysis?
Anthropology, as is well known, has adopted a fairly uniform stance with
respect to the cross-cultural variability which characterizes notions of the
good, the true, and the beautiful. This stance, so far as the normal-abnormal
distinction is concerned, was given classic expression by Benedict (1934),
who maintained that judgments concerning abnormality are necessarily
relative to intracultural standards. What is judged to be abnormal in one
cultural setting may be properly characterized as normal in other cultural
settings. This relativistic approach continues to provoke somewhat heated
controversy, not only within the anthropological fraternity, but within the
social sciences in general.
In a series of seminal papers Professor Hallowell has offered us
important conceptual vehicles by which we can avoid the Scylla of nihilistic
relativism and the Charybdis of ethnocentric absolutism. His concept of a
“culturally constituted behavioral environment” (Hal-lowell 1955: chapter
8) allows us to view behavior relative to its cultural setting and, at the same
time, to assess its functional consequences in terms of pancultural scientific
criteria. I should like to examine the relationships between religion,
abnormality, and relativism in the light of these conceptual tools. Although
only these concepts are explicitly taken from Hallowell, this entire paper is
heavily indebted to his work. Indeed, it is difficult to say when his ideas end
and mine begin.
The persistent controversy over cultural relativism has been confounded
by implicit disagreements concerning its proper antithesis. Some scholars
conceive of (what I shall term) “universalism” to be the antithesis of
relativism, while others take (what may be termed) “absolutism” to be its
antithesis. The logic of relativism is quite different, depending upon which
of these two alternative conceptions is taken to be its antithesis; indeed
consensus concerning one or the other antithesis would mitigate, if not
resolve, the controversy. I should like to propose absolutism as the valid
antonym for relativism, and universalism as the antonym not for relativism,
but for what might be called “particularism.”

Universalism vs. Particularism

The question entailed by this distinction is whether a particular belief or
behavior pattern found in our society—and which a clinician would
characterize as pathological—occurs in all societies, in some societies, or
only in our own society. This is an empirical question, concerned
exclusively with the occurrence and relative incidence of a certain type of
behavior. Correspondingly, the answer to this question must also be
empirical, rather than judgmental. Whatever the answer to this question
may be, the research task—once the answer is provided—is to identify
those causal or antecedent conditions which produce the behavior or which
account for its differential occurrence. In the older arguments for relativism,
in which sociocultural variables were adduced as foils for, or as antidotes
to, biological variables (usually, but not always, associated with racism),
relativistic theorists turned to sociocultural determinants in order to explain
the differential occurrence of different types of behavior. Particularism, as I
shall label this variety of relativism, interprets the differential occurrence of
a certain type of behavior as relative to, because determined by, differential
sociocultural conditions.
Anthropological data—and, one might add, historical and sociological
data as well—are unambiguous so far as the universalisticparticularistic
dichotomy is concerned. The occurrence and relative incidence of certain
types of behavior which, when occurring in our society, are labeled as
schizophrenic, hysterical, paranoid, etc., vary not only from society to
society, but they vary as well within the same society at different periods in
its history, and among different social groups (class, ethnic, religious, etc.)
within the same historical period. This empirical finding is independent of
any judgment by which these types of behavior may be evaluated.

Absolutism vs. Relativism

This, rather than the former, dichotomy raises the genuine relativistic
question, viz., can certain types of behavior which are designated as
pathological when they occur in our society be properly designated by the
same label when they occur in other societies? The concern here is not with
the question of the differential occurrence of beliefs and practices—this is
taken for granted—but with their clinical assessment. Absolutism holds that
certain types of behavior are properly designated as pathological wherever
and whenever they occur. Relativism holds that judgments concerning
pathology are necessarily relative. There are different types of relativism,
however, differing according to what these judgments are conceived to be
relative to. One type, which we may label “social relativism,” holds that
judgments concerning pathology are relative to intrasocietal criteria,
because there are no pan-human criteria by which cross-cultural judgments
can be made. A second type, which may be termed “cultural relativism,”
holds that the criteria forjudging pathology are panhuman, but that
judgments based on these criteria are relative to the cultural context in
which the action occurs. I should like to examine these two types of
relativism, and their respective subtypes.
Social Relativism—Social relativism holds that judgments concerning
pathology can only be based on intrasocietal—and, as we shall see,
quantitative—criteria. One subtype, which might be termed “objective
social relativism,” insists that judgments concerning the pathology of some
type of behavior must be relative to its incidence in the society (or some
social group within a society) in which it occurs. Hence, behavior is
properly judged to be pathological if (but only if) it is statistically rare; if,
that is, it deviates from the statistical norms for that society. According to
this subtype, “schizophrenic” behavior is pathological in our society
because it characterizes a minority; if schizophrenics were to become a
majority, schizophrenia would be normal. For this subtype, then, the
abnormal represents social deviation.
“Subjective social relativism” holds that behavior may be judged to be
pathological if it is deemed to be pathological by the members of the
society in which it is found; it is not pathological if they deem it to be
normal, or, at least, if it is positively rather than negatively evaluated by
them. If, according to objective social relativism, judgments concerning
pathology are relative to the incidence of some type of behavior within a
society, according to subjective social relativism they are relative to its
evaluation by the members of society. For the former, pathology represents
deviation from statistical norms discovered by the anthropologist; for the
latter it represents deviation from evaluative norms held by the society.
Although their criteria differ, both subtypes of relativism agree that there
can be no panhuman criteria by which judgments concerning pathology can
be established. Hence, behavior which is properly labeled “paranoid” when
found in our society is “normal” if found among the Kwakiutl; hysteria is
normal in the case of St. Theresa, but abnormal in a contemporary middle-
class woman; the belief that most of mankind is doomed to eternal hell-fire
is normal for a Calvinist, but abnormal for a Buddhist, and so on.
In my opinion both subtypes of social relativism are untenable. In the
first place, since man (even homo religiosus) is a biological organism, there
are, I would suggest, panhuman biological criteria by which behavior may
be judged. If the majority in a given society have cancer, are we to say that
they, the majority, are biologically normal, while those who do not have
cancer, the minority, are biologically abnormal? This crude analogy
suggests the fallacy inherent in the first subtype of social relativism.
Judgments concerning biological normality or abnormality (health and
illness) are indeed relative; but they are relative to specified criteria for the
functioning of healthy organisms, criteria which are—or, at least, biological
science takes them to be—universally applicable. Only two of the criteria
which are relevant for the problem of behavioral pathology need be
mentioned here: adaptation and optimum functioning.
Any organic characteristic which, like cancer, produces death, and which,
if generalized, precludes the survival of the group—i.e., any maladaptive
characteristic—is biologically abnormal. Furthermore, any characteristic
which reduces efficient functioning—efficient relative to the potentiality of
either the organ or the organism—is biologically abnormal. Thus, impaired
vision or hearing, let alone blindness or deafness, are abnormal because
they reduce the optimum functioning of these organs and, ultimately, of the
organism. In short, maladaptive and inefficient functioning are, for the
physician and biologist, panhuman criteria by which the presence of
biological abnormality may be assessed; this assessment is independent of
the incidence of these characteristics in a society or of their evaluation by
its members. If we are prepared to accept these criteria it would seem to
follow that any condition which produces these effects—whether it be a
characteristic of the organism or a characteristic of behavior—is
pathological. Hence, if it is the case that a religious belief or a religious
ritual is maladaptive and/or reduces optimum biological functioning, such a
belief is, biologically viewed, abnormal.
But man is not only a biological, he is also a psychological and a social
animal. For psychological, as for biological processes, the home-ostatic
principle, tension→tension-reduction, holds: tension stimulates action
which is intended to reduce tension. Hence the persistence of tension is
abnormal. Most psychological tension stems from conflict, either between
one’s moral values and/or one’s self-image, or between those desires which
violate the one or the other (or both). Since the tension induced by such
conflict is painful, the typical actor—in accordance with the homeostatic
principle enunciated above—attempts to resolve the conflict and, hence, to
reduce tension. This is usually accomplished by a number of maneuvers,
termed “mechanisms of defense,” which defend the ego against the pain
induced by the tension. Although these defense mechanisms may reduce the
tension, thereby satisfying the homeostatic criterion of normality, they may
have other pathological consequences. Any resolution of conflict, however
successful in reducing tension, which results in the impairment of
psychological, social, or cultural functioning may be taken to be a universal
index of pathology.
 “Impairment of psychological functioning” refers to distortions of three
important modes of action: (1) Cognitive distortion: this refers to any
cognitive behavior in which a demonstrably false belief is held to be true;
(2) Perceptual distortion: this refers to any perceptual behavior in which
stimuli are perceived to be other than what they are; (3) Affective
distortion: this refers to any emotional behavior which, relative to the
stimulus condition, is characterized by hyper- or hypoaffectivity.
“Impairment of social functioning” refers to any condition of the actor
which precludes the performance of social roles. “Impairment of cultural
functioning” refers to any condition of the actor which precludes
compliance with cultural norms and rules.
The three psychological criteria of pathology are panhuman because, if
satisfied, the reality-testing function of the ego, the agent of man’s
adaptation (to nature) and adjustment (to society), is impaired, and the
consequence of such impairment may be either individual or social
disruption. The last two sociocultural criteria of pathology are panhuman
because, in the absence of adequate social and cultural functioning, the
survival of the individual—and, if generalized, the survival of the group—is
jeopardized. No society could persist if the behavior of its members was
characterized, for example, by genuine schizophrenia or genuine paranoia.
Hence any religious belief or religious ritual which is characterized by, or
which leads to, these three types of impairment is properly characterized as
abnormal, regardless of its incidence and regardless of its evaluation by the
members of society.
Cultural Relativism—Despite these animadversions on social relativism,
its rejection does not entail the acceptance of absolutism. According to
absolutism certain forms of behavior and certain kinds of beliefs are
pathological wherever or whenever they may be found. Thus, for example,
if a religious belief or ritual in a non-Western society is phenotypically
similar to a belief or behavior pattern characteristic of a Western psychotic,
the former, according to absolutism, is also abnormal. This position is as
untenable as social relativism. Although the criteria by which behavior is
judged may be universally applicable, it does not follow that all instances of
phenotypically identical behavior or belief, when evaluated by these
criteria, will lead to the same judgment. Although the criteria are
panhuman, the judgments based on these criteria are necessarily relative to
the sociocultural context within which behavior occurs. Distortion, for
example, implies the existence of some reality relative to which a cognition
is false, a perception is skewed, an affect is misplaced. But this reality, as
Professor Hallowell has so cogently shown, is not a universal “given.”
Different cultures structure reality in different ways. For any actor reality is
mediated through the world view and behavioral environment constructed
by his culture. Hence, judgment concerning the existence of distortion must
be based on the culturally constituted world view and the behavioral
environment of the actor whose behavior is being assessed. For a Western
man to believe that he is the reincarnation of some ancestor is to commit
severe cognitive distortion; a Buddhist who holds the same belief commits
no distortion. In short, the same criterion, applied to two identical acts,
yields different judgments because the judgment—but not the criterion—is
relative to the sociocultural context within which action occurs.
Social and cultural inadequacy must be assessed in the same manner.
Whether a particular condition does or does not permit normal social and
cultural functioning is necessarily relative to the repertory of social roles
and the set of cultural norms which are found in the actors, not the
observers, society.
If the preceding discussion is sound, it may then be concluded that beliefs
and rituals which characterize the behavior of religious actors in non-
Western societies, although phenotypically identical with beliefs and
behavior which may characterize abnormal individuals in our society, are
not necessarily (or even usually) abnormal when sanctioned or prescribed
by the religious systems of the former societies and taught to the actors as
part of their cultural heritage. There are a number of reasons for this
conclusion: (a) The religious actor acquires his beliefs and rituals, as he
acquires other aspects of his cultural heritage, through the usual techniques
of instruction and imitation. Hence these beliefs and rituals are expressions,
rather than distortions, of (his culturally constituted) reality. They are
consistent with, rather than obstacles to, social and cultural functioning.
Psychotic beliefs and behavior, on the contrary, are devised by the actor
himself, as an attempt to reduce the painful tension induced by inner
conflict. This attempt is necessarily based on his private distortion of
(culturally constituted) reality, resulting in serious impairment of his social
and/or cultural functioning; (b) Since religious beliefs and practices are
derived from tradition, they are frequently compulsory; but since they are
not created by the actor himself to defend against forbidden or shameful
impulses, they are not compulsive. Psychotic beliefs and practices, on the
contrary, are not compulsory—indeed they are usually prohibited—but as
attempts to reduce tension, they are compulsive; (c) Although not devised
by the actor to resolve conflict, religious beliefs and ritual may be used for
that end. When so used they may not only resolve conflict, but as
“culturally constituted defenses” (Spiro 1961:472–97) they are consistent
with, rather than distortions of, reality; they comprise culturally sanctioned,
rather than culturally prohibited, behavior; they protect the individual and
his society from the disruptive consequences both of his shameful and/or
forbidden needs and of his private defensive maneuvers. Conflict may also
be resolved by psychotic beliefs and practices; but these idiosyncratic
resolutions produce those psychological distortions and socio-cultural
impairments which, as I have argued, are properly characterized as
abnormal. I should like to examine these propositions in the context of one
empirical situation—Burmese Buddhist monasticism.

Burmese Monasticism 1
1. Materials in this section are based on field work carried out during 1961–62; I am grateful to the
National Science Foundation for a fellowship which made the research possible.

In Burma, one of the centers of Theravada Buddhism, the monastic

vocation is the most venerated of all patterns of life. Almost every village
contains at least one monastery with at least one resident monk. The monk,
in theory at least, lives an exclusively otherworldly existence. His
monastery is outside the village gates, and his interaction with the layman is
confined to occasional ritual situations. The monk is prohibited from
engaging in any form of physical labor, including any economic activity.
All of his wants are attended to by the laymen, who provide his daily meals,
his robes, and other necessities. Except for teaching young children, the
monk’s official responsibilities to the laymen are restricted to chanting of
“prayers’’ at funerals and to public recitation of the Buddhist precepts on
the “Sabbath’’ and other sacred days. His primary responsibility is to
himself and to his attempt to attain nirvana. The latter goal is achieved
through the study of Scripture and through various techniques of Buddhist
meditation. These activities are believed to be instrumental to the
attainment of Release from the round of rebirths because they lead to
ultimate comprehension of the true characteristics of existence, viz.,
impermanence, suffering, and the absence of an ego. This comprehension,
in turn, is believed to lead to the severance of all desire for, and cathexis of,
the world. With the destruction of desire or “clinging’’ (tanha), the basis for
rebirth is destroyed. Nirvana, whatever else it may be, is the cessation of
rebirth.
The true monk, then, is completely absorbed in his own “salvation.”
Although living in a state of absolute dependence on the laymen, he has
withdrawn both physically and psychologically from the physical and social
world, and even—in states of trance (dhyānas)—from his own self. This
extreme withdrawal from reality is similar to that withdrawal behavior
which, in our society, would be taken as symptomatic of severe pathology,
most certainly schizoid, if not schizophrenic. Is the Burmese monk to be
similarly characterized? Such a judgment, in my opinion, would be grossly
in error. Phenotypically, the behavior of the monk and that of the schizoid
or schizophrenic patient may be very similar. Genotypically and
functionally, however, they are importantly dissimilar. All of the criteria
suggested in the previous section for assessing pathology are applicable to
the schizophrenic; none is applicable to the monk.
In the case of the schizophrenic the actor resolves his inner conflict by
constructing private fantasy and action systems; in the case of the monk,
however, the actor uses culturally constituted fantasy and action systems
(Buddhism) to resolve his inner conflicts. This difference not only provides
the primary basis for a differential diagnosis of monk and schizophrenic,
but it also provides, parenthetically, an important insight into the nature of
religion. Culturally constituted religious behavior not only is not a symptom
of pathology but, on the contrary, it serves to preclude the outbreak of
pathology. The schizophrenic and the Burmese monk, alike, are
characterized initially by pathogenic conflict, and schizophrenia and
monasticism may each be interpreted as a means for resolving conflict. But
this is where the similarity ends. Although schizophrenia and monasticism
are both symptomatic of pathogenic conflict, the former represents a
pathological, whereas the latter represents a nonpathological, resolution of
the conflict. Let us examine these claims.
 An analysis of monastic personality, based on the Rorschach records of a
sample of Burmese Buddhist monks, and without reference to their
(monastic) behavioral environment, would surely lead to a diagnosis of
severe pathology. Dr. James Steele (1963), who has analyzed these records,
finds the following “pathological” features, among others: (1) a very high
degree of “defensiveness”; (2) “pathologically regressed” expression of
aggressive and oral drives; (3) cautious avoidance of “emotionally laden”
situations as a means of obviating the necessity of handling affect, for
which there are no adequate resources; (4) a “hypochondriacal self-
preoccupation” and “erotic self-cathexis,” instead of a cathexis of others;
(5) latent homosexuality; (6) above-average fear of female- or mother-
figures.
One of the significant characteristics of the Rorschach protocols of these
monks, according to Steele, is their similarity to the records of Burmese
laymen. It is not that the monastic records do not differ from those of the
laymen; but the difference is one of degree, not of kind. Monks differ from
laymen, not because they have different problems, but because they have
more of the same problems. The monk, in other words, is a Burman in
extremis. Burmese laymen (like Burmese monks) are constricted,
ruminative, defensive, anxious about females, distrustful of others, and,
perhaps, latently homosexual. The monks differ from the laymen only in
that, for all these characteristics, they are more so. Monks are more
constricted, more ruminative, more…, etc. For other characteristics,
however, monks are less so. Compared to laymen, monks are “… less
phallic, less self-confident, less striving, and less impulsive.” In summary,
Burmese monks not only appear to have “more of the basic problems”
which characterize Burmans in general, but they also seem to be
characterized by a “more constricted adjustment.” It is the latter feature,
still quoting Steele, which makes them “less accessible to social interaction
with the protection that this provides.”
This picture of the Burmese monks is surely a picture of pathology. Are
we to conclude then—holding in abeyance a specific psychiatric diagnosis,
and assuming that the Rorschach test is a reliable instrument—that these
monks are abnormal? If personality existed in a social and cultural vacuum,
the answer would be an unqualified “yes.” Acute psychological conflicts
and attendant intrapersonal tensions are marked. That these conflicts have
produced defensive distortions of various kinds—perceptual, cognitive,
affective—is clearly indicated. That social impairment is a most likely
consequence of these conflicts, tensions, and distortions can hardly be
doubted. In brief, if personality existed in vacuo one would probably
conclude that Burmese monks resolve their conflicts in a manner which
issues in severe pathology (perhaps paranoid schizophrenia).
But the proviso, “if personality existed in vacuo” is crucial. Although
Steele’s analysis of their Rorschach records is remarkably similar to my
clinical impressions of these monks, impressions derived from intensive
participant-observation in a score of monasteries, and from personal
interviews with more than twenty-five monks—thus providing a dramatic
test of the reliability of the instrument and of Steele’s skill in its use—I did
not ever feel that these monks, with but one exception, were pathological or,
specifically, schizophrenic. Nor is this a paradox. The psychological
analysis (based on Rorschachs and clinical impressions) provides a set of
statements concerning the emotional problems of the subjects; it also
provides, to a somewhat lesser extent, a picture of their idiosyncratic
defenses, i.e., of those defenses which the subjects have constructed for
themselves, in an attempt to resolve their problems. That these defenses are
hardly adequate to the task is obvious from the Rorschach analysis. Left
exclusively to their own inner resources many of these subjects would have
become, I believe, genuine psychotics.
But personality does not exist in vacuo, and Burmese males,
characterized by the problems described, are not confronted with the
necessity of solving these problems by means of their own resources. In
addition to their private resources, they are able to utilize a powerful
cultural resource for their solution, i.e., they can solve their problems by
recruitment to the monastic order. By utilizing the role-set prescribed for
this institution as a culturally constituted defense, Burmese monks can
resolve their conflicts with a minimum of distortion. Since, moreover, the
performance of the roles comprising this role-set satisfies their prohibited
and/or shameful needs and reduces their painful fears and anxieties, these
potentially disruptive psychological variables, rather than provoking
socially disruptive behavior, provide the motivational basis for the
persistence of the most highly valued institution—monasticism—in
Burmese society. As a culturally constituted defense, the monastic
institution resolves the inner conflicts of Burmese males, by allowing them
to gratify their drives and reduce their anxieties in a disguised—and
therefore socially acceptable—manner, one which precludes psychotic
distortion, on the one hand, and criminal “acting-out,” on the other. Hence,
the monk is protected from mental illness and/or social punishment; society
is protected from the disruptive consequences of antisocial behavior; and
the key institution of Burmese culture—Buddhist monasticism—is provided
with a most powerful motivational basis. Space permits only a brief
examination of these assertions.
The monastic rules which interdict all labor, and those Buddhist norms
which guarantee that the laity provide monks with all their wants, combine
to satisfy the monk’s “regressed oral drives.” The monastic life, moreover,
makes no demands, either social or psychological, which might render the
monk’s weak “phallic-orientation,” his low degree of “striving and
impulsivity,” and his lack of “self-confidence” nonviable modes of
adjustment. Quite the contrary, the physical isolation of the monastery, and
the monastic norms proscribing social participation, preclude the
stimulation of “disruptive affect.” At the same time, the monk’s “self-
preoccupation” and “erotic self-cathexis” is wonderfully expressed and
institutionalized in the prescribed techniques of Buddhist meditation. Latent
“homosexual” needs can be satisfied in the exclusively male setting in
which the monks live. Finally, the strong interdiction on interaction with
females provides little opportunity for encounters with them and for the
consequent fear attendant upon such encounters. Buddhist monasticism,
then, is a highly efficient means for coping with the psychological problems
of many Burmese men. The differences between a monastic and a psychotic
resolution of these problems are instructive.

1. In general the genesis of the psychotic’s conflict is idiosyncratic, while

the genesis of the monk’s problem is rooted in modal features of his
society. That is, the source of the monk’s conflict—which we cannot
discuss here—is to be found in culturally constituted experiences
which the monk shares with many other members of his social group.
The monk differs from other Burmans in one of three ways: his
potentially pathogenic experiences are more intense than those of other
males; other Burmans utilize alternative, non-Buddhist, institutions for
the resolution of equally intense problems; still others (a minority)
develop idiosyncratic methods of conflict-resolution (in extreme cases,
these take the form of mental illness or criminal behavior).
2. The psychotic resolves his problems by means of idiosyncratic,
private, defenses; the monk resolves his problems in an
institutionalized manner, by utilizing elements of his religious heritage
as a culturally constituted defense. The difference between these two
types of defense accounts for the following differences between the
psychotic patient, on the one hand, and the normal monk, on the other.
3. The behavior of the psychotic is incompatible with any normal social
role within his society, and inconsistent with important cultural norms
of the larger society. The psychotic is psychologically incapable of
performing social roles or of complying with those cultural norms
which he violates. The behavior of the monk, on the other hand, is
entirely appropriate to—indeed, it is the enactment of—a most
important and honorific social role. The monk may be psychologically
incapable of performing nonmonastic roles, but he is ideally suited for
performing the monastic role.
4. As a corollary of the above, the behavior of the psychotic is bizarre in
the eyes of, and disapproved by, his fellows. The behavior of the monk
is not only approved by the other members of his society, but it is most
highly valued.
5. Following from the last point, the behavior of the psychotic alienates
him psychologically from his fellows. The behavior of the monk, on
the contrary, though isolating him physically from his group, serves to
integrate him psychologically into the group; for in his behavior he
expresses the most cherished values of Burmese culture.
6. The world view constructed by the psychotic represents a dramatic
distortion of reality, as the latter is structured by the world view of his
culture; and the cognitions and perceptions that are derived from his
idiosyncratic world view are highly distorted, relative to the behavioral
environment in which expected social interaction of his society takes
place. The world view of the monk, on the other hand, rather than
being constructed from his private fantasies, is taught to him as an
integral part of the cultural heritage of his society. The private world
view of the monk corresponds to the public world view of his society;
his world view, in brief, is a culturally constituted world view. The
Buddhist world view, of course, may be false, a distortion of reality,
relative to the world view of modern science; but it is true, relative to
the knowledge available to Burmese society. True or false, however,
 the monk’s cognitions and perceptions are consistent with, rather than
distortions of, reality, as the latter is structured by the world view and
behavioral environment of his society. The perceptions and cognitions,
the fantasies and emotions experienced by the monk in the course of
Buddhist meditation and concentration may never be experienced by
other Burmans—because the latter do not meditate or concentrate—
but they are experiences consistent with the conception of reality
which all Burmans hold, and they are vouchsafed to any Burman who
is prepared to enter into these spiritual disciplines.
7. The psychotic sustains social relationships neither with the normal
members of his society, nor with other psychotics. Psychotics, in short,
do not participate in the society of which they are members, nor do
they comprise a social group, distinct from the larger society, but
nevertheless viable for its constituent members. Burmese monks, on
the other hand, although socially isolated from Burmese society, are
yet psychologically part of it. Moreover, although the monk may find
difficulty in participating in the larger society, and in forming social
relationships within it, he does enter into social relationships with
other members of the monastic order. Monks are members of
increasingly larger concentric groups, beginning with other members
of the local monastery and extending to the entire order of monks. In
short, while psychotics comprise a typological class, the monks
constitute a social group. The psychotic cannot live as a member of a
social group, even if it be but a subgroup of the larger society.
8. Finally, and as a corollary of the last point, the behavior of the
psychotic is anomic; it violates many of the rules of his society. The
monk, on the other hand, not only exemplifies the rules of Burmese
society, but he must, in addition, comply with the 227 rules of the
monastic order (as outlined in the Vinaya). Should he violate these
rules, he is expelled from the order as a charlatan, regardless of
whatever wondrous visions he is alleged to have had, or miraculous
powers he is supposed to possess.

In summary, then, a psychiatrically diagnosed psychotic is not only

incapable of participating in his own society, he is incapable of participating
in any society. An American psychotic would function no better in a
Buddhist monastery than in an American city. A Buddhist monk, to be sure,
while not capable of functioning in every cultural environment, functions
very well indeed within his cultural environment. This is hardly surprising,
since the latter environment is so structured that it satisfies his needs and
resolves his conflicts. That he cannot function in a radically different
environment does not render him “sick,” nor his adjustment precarious.
Typically, differential sets of human needs are differentially satisfied in
different types of cultural milieux. It is doubtful if the typical Burmese
peasant could adjust to an American urban environment, or a typical
American to a Buddhist monastery. For neither is the new environment
capable of satisfying the needs and resolving the conflicts which were
produced by the old.

Summary and Conclusions

Burmese monks, as Rorschach data and clinical observations agree, are
characterized by serious emotional conflicts. Their religious heritage,
however, provides them with institutionalized means for resolving conflicts,
and, moreover, for resolving them in a manner which satisfies none of the
five criteria suggested in the first section of this chapter as panhuman
indices of abnormality (psychological distortions and sociocultural
impairments). Employing the monastic system as a culturally constituted
defense obviates the necessity for Burmese males to erect private defenses
which, necessarily, would lead to one or more of these distortions and
impairments. Monasticism, in short, has important psychological functions
for individual actors. By the same token, however, the psychological
problems of these actors have the important cultural function of helping to
perpetuate the monastic system; that is, these conflicts are not only resolved
by means of the monastic system, but they provide the motivational basis
for recruitment to the monastery and, hence, for the persistence of the
system. The existence of the monastic system, moreover, not only permits
the resolution of emotional conflict (the latent psychological function of the
monastery), but by serving this function it reduces the probability of the
occurrence of other, nonsanctioned means by which these conflicts might be
expressed and resolved (latent social function of the monastery). If this
culturally constituted defense were not available for the resolution of
conflict, the consequent persistence of tension might lead to defenses of a
psychotic nature, psychosomatic disorders of various kinds, or many types
of antisocial, i.e., criminal behavior. I would suggest that a study of
Burmese crime, including dacoity and insurgency—both of which are, and
have been, endemic in Burma—would reveal that a large percentage of
dacoits and insurgents are recruited from the ranks of those for whom the
monastic life is not (for reasons still to be determined) a psychologically
viable means for resolving emotional conflict.2
2. This generalization excludes those—and I have met them—for whom insurgency and dacoity
are romantic, adventurous activities to be given up when the adventure palls. It also excludes those
whose emotional conflicts are idiosyncratic, rather than culturally patterned, and for which there are
no cultural institutions by which conflict can be resolved.
The monastic system, in short, not only serves the important personal
function of precluding psychotic breakdown, but it also serves the important
social function of allowing potentially disruptive, antisocial drives to be
channeled into culturally approved (institutional) behavior. Since the
monastery, moreover—for reasons beyond the scope of this paper—is the
most integrative institution in Burmese society, the social function of
psychological conflict, when resolved in this fashion, cannot be
overestimated.
If the foregoing analysis is correct, then—to return to the more general
problem of this chapter—abnormal behavior can be expected to appear
under one of three conditions: (1) when emotional conflict is idiosyncratic,
so that cultural means are not available as potential bases for culturally
constituted defense mechanisms; (2) when emotional conflict is modal, and
cultural means are available for conflict resolution, but these means have
been inadequately taught or inadequately learned; (3) when, under
conditions of rapid social change, culturally constituted defense
mechanisms are unavailable, either because older institutions have been
discarded or because the new situation creates a new set of conflicts. These
three conditions, however, are necessary, but not sufficient, conditions for
abnormal behavior. Although emotional conflict is potentially pathogenic, it
need not produce pathology. Emotional conflict issues in pathology only if
it is not resolved, or if it is resolved in a manner which is characterized by
psychological distortion and/or sociocultural impairment. The latter
resolutions are characteristic of neurotic and psychotic resolutions.
But neurosis and psychosis are not the only means for resolving conflict.
Other private defense mechanisms may be constructed which are
perceptually, cognitively, and affectively consistent with the behavioral
environment of the actors, and which, moreover, constitute no obstacle to
adequate sociocultural functioning. Finally, there is a third category of
defense mechanisms—culturally constituted defenses—which are not only
not disruptive of, but rather serve to perpetuate, the sociocultural system.
Conflict, in short, may indeed produce pathological defenses; but it may
also produce normal defenses, either private or culturally constituted.
In most traditional societies, where religious beliefs and practices
continue to carry conviction, religion is the cultural system par excellence
by means of which conflict-resolution is achieved. In such societies, in
which religious behavior is appropriate to, rather than disruptive of, the
behavioral environment of the actors, and in which a religious world view is
consistent with, rather than a distortion of, “reality,” religion serves as a
highly efficient culturally constituted defense mechanism.

Note
Reprinted with permission of The Free Press, a division of Macmillan, Inc., from Context and
Meaning in Cultural Anthropology, edited by Melford E. Spiro, pp. 100–113. © 1965 by The Free
Press.

References

Benedict, R. 1934. Patterns of culture. Boston: Houghton Mifflin.

Hallowell, A. I. 1955. Culture and experience. Philadelphia:
University of Pennsylvania Press.
Spiro, M. E. 1961. An overview and a suggested reorientation. In
Psychological anthropology, ed. F. L. K. Hsu. Homewood, 111.:
Dorsey Press.
Steele, J. 1963. A preliminary analysis of the Burmese Rorschachs.
Unpublished ms.
7
Collective Representations and Mental
Representations in Religious Symbol
Systems

IN THIS chapter I want to address the relationship between statements of the

following two types: “The omnipotence of God is a basic doctrine of
Christianity,” and, “As a Christian, John holds the doctrine of the
omnipotence of God to be true.” To put it more abstractly, I want to address
the relationship between religious doctrines as they are found in the
collective representations of a social group, on the one hand, and in the
mental representations of the religious actors, on the other. Divested of
jargon, I want to ask why it is that religious actors believe in the doctrines
that comprise the religious system of their culture. If this question seems
trivial, or if the answer seems obvious, it is only because we have for too
long—certainly since Durkheim—accepted the coercive power of cultural
symbols on the human mind to be a self-evident truth.
I have at the outset introduced the mind as one of the key terms of our
discussion because one of the unexamined assumptions of much of
anthropology is that any attempt to understand culture by reference to the
mind is at best to confuse levels of analysis, if not levels of ‘reality’. and at
worst, to perpetuate the intellectual sin of reductionism. Although I oppose
the confusion and abjure the sin, I will argue that inasmuch as cultural
doctrines, ideas, values, and the like exist in the minds of social actors—
where else could they exist?—to attempt to understand culture by ignoring
the human mind is like attempting to understand Hamlet by ignoring the
Prince of Denmark. To be sure, cultural doctrines and the like are encoded
in those public and visible signs (cultural symbols) which, following
Durkheim, we call ‘collective representations’; but since the latter neither
possess nor announce their meanings, they must be found in the minds of
social actors. If this is so, then to understand culture it is not sufficient to
attend to cultural symbol systems and how they work; it is also necessary to
attend to the mind and how it works. That, at any rate, is the thesis I wish to
explore here.
In referring to the mind, I am not only referring to those intellectual and
information-processing functions which are often exclusively associated
with that concept. Rather, I am referring to all of those psychological
processes—cognitive, affective, and motivational—which underlie any type
of complex behavior. Such a broad notion of ‘mind’ is especially important
if we are to deal adequately with the complex problem with which we are
concerned in this chapter. Its complexity is a function of the fact that
although religious systems are cognitive systems, they persist because of
powerful motivational dispositions and affective needs of the social actors
to which they are responsive. Moreover, although the culturally constituted
meanings of the symbols in which religious doctrines are encoded are
consciously held by the actors, the latter also invest them with private, often
unconscious meanings, whose cognitive salience is no less important for
their understanding. This being the case, in order to explain why social
actors believe in the religious doctrines of their culture we must attend to
the motivational and affective, as well as the cognitive properties of the
mind; and in attending to these properties, we must be concerned with their
unconscious, as well as their conscious dimensions.
In an important sense, this chapter may be viewed as a long, and
somewhat extended and delayed commentary on an observation I made
eleven years ago as a discussant at a symposium on symbolic anthropology.
“If symbolic behavior is even half as important as Freud, for example,
suggested, symbolic anthropology is the custodian of the richest of all the
mines which are worked by the science of man. To be sure, we have not yet
produced our Freud, but until we do, perhaps we would be wise to reread
The Interpretation of Dreams” (Spiro 1969:214).
Before embarking on this inquiry, I should perhaps define its basic terms.
By “belief’ I will mean any proposition concerning human beings, society,
and the world that is held to be true. By “religious belief’ I will mean any
belief that, directly or indirectly, relates to beings who possess greater
power than human beings and animals, with whom human beings sustain
asymmetrical relationships (interactions and transactions), and who affect
human lives for good or for ill. In short, “religious” beliefs comprise that
subset of beliefs which, directly or indirectly, are concerned with
‘superhuman’ beings. Not all beliefs, of course, are culturally-constituted,
and since the distinguishing feature of “culture,” as I shall use that term, is
tradition, a “culturally-constituted” belief—religious or any other—is a
traditional belief. That is, it is one which is acquired by learning a cultural
doctrine—a proposition about man, society, or nature—that originates and
develops in the history of a social group, and is transmitted from generation
to generation by means of those social processes that are denoted by such
terms as “education” and “enculturation.” Our definition of “symbol” will
be deferred until later.
With these definitions in mind, and considering what I have already said
about the need to study culture in its relationship to the social actors—the
culture-bearers, as they used to be called—the investigation of a culturally-
constituted belief involves—or at least it ought to involve—a six-fold task.
First, the cultural doctrine which is the basis tor the belief must be
described. Second, its traditional meanings must be ascertained. Third, its
relationship to the other doctrines comprising the system of which it is a
part must be delineated. Fourth, the structure of the system must be
explicated. Fifth, the grounds for the actors’ accepting the doctrine as their
own belief must be uncovered. Finally, the functions of holding this belief
—the consequences, for either the social actors or their society—must be
discovered.
In referring above to the “meanings” of cultural doctrines, I was, of
course, speaking elliptically; for their meanings, as I have already said, are
“located” not in the doctrines themselves, but in the minds of the actors
who hold and transmit them. Thus, when we ask, “what does the Christian
doctrine in the omnipotence of God mean?” we are really asking, “what
does this doctrine mean to Christians?” But since cultural doctrines are
acquired through social transmission (rather than constructed from personal
experience), the answer to questions of this type is far from obvious, as it
depends on the cognitive salience with which the doctrines are acquired as
personal beliefs.
Briefly, and in ascending order of importance, we may delineate a
hierarchy comprising five levels of cognitive salience: (a) The actors learn
about the doctrines; as Bertrand Russell would say, they acquire an
“acquaintance” with them; (b) The actors not only learn about the doctrines,
but they also understand their traditional meanings as they are interpreted in
authoritative texts, for example, or by recognized specialists; (c) The actors
not only understand the traditional meanings of the doctrines, but
understanding them, they believe that the doctrines so defined are true,
correct, or right. That actors hold a doctrine to be true does not in itself,
however, indicate that it importantly effects the manner in which they
conduct their lives. Hence (d) at the fourth level of cognitive salience,
cultural doctrines are not only held to be true, but they inform the
behavioral environment of social actors, serving to structure their
perceptual worlds and, consequently, to guide their actions. When cultural
doctrines are acquired at this level we may say that they are genuine beliefs,
rather than cultural cliches; (e) As genuine beliefs the doctrines not only
guide, but they also serve to instigate action; they possess motivational as
well as cognitive properties. Thus, one who has acquired, for example, the
doctrine of hell at this—the fifth—level of cognitive salience, not only
incorporates this doctrine as part of his cosmography, but he also
internalizes it as part of his motivational system; it arouses strong affect
(anxiety) which, in turn, motivates him to action whose purpose is the
avoidance of hell.
Although, in general, anthropologists have assumed that cultural
doctrines are acquired at the fourth, if not the fifth level of this hierarchy,
this assumption is all too often unwarranted, and it has led to erroneous
interpretations of the importance of particular cultural doctrines, as well as
to extravagant claims concerning the importance of culture in general in
human affairs. Thus, for example, many key features of the social structure,
political system, and economic organization of the Buddhist societies of
Southeast Asia have often been explained as a function of the putative
otherworldly’ orientation of Buddhist actors, an orientation which is
inferred from the Buddhist doctrine of nirvana. This explanation, however,
is typically invalid because, as recent anthropological studies of these
societies have shown, except for a few monastic virtuosi, this doctrine has
not been internalized by Buddhist actors as a culturally-constituted belief,
but is rather a cultural cliche.
My point, then, is that in order to explain a cultural doctrine—that is to
account for its existence—we must first interpret it; we must discover its
‘meaning’ for the actors. This requires that we ascertain at which of the five
levels of cognitive salience adumbrated above it has been acquired as a
belief. On the premise that cultural doctrines have been acquired as
personal beliefs at the fourth or fifth level of that hierarchy, anthropologists
have typically adopted two complementary intellectual modes in their
attempts to explain and interpret them. One mode is concerned with
culturally particular analysis and the other with trans-cultural analysis.
Conceived in the first mode, the analysis of cultural beliefs focuses on
the local ethnographic setting in all of its uniqueness. For this mode, the
question of why the Burmese, say, believe in cultural doctrines related to
gods, demons, and the Buddha evokes an unambiguous answer: as part of
the cultural heritage of their society, these doctrines have been acquired as
beliefs by each successive generation of Burmese from previous
generations. Hence, this mode is especially interested in discovering how
these doctrines are related to other aspects of Burmese culture and society
such that, together, they comprise an integrated “system.” To the extent that
this mode presses its analysis further, it turns in one of two directions. One
direction, culture history, is concerned with diachronic ‘explanation’. and
attempts to discover the socioeconomic and political conditions which
might have led to the invention or borrowing of these doctrines. The other,
symbolic anthropology, is concerned with symbolic interpretation’. and
attempts to discover the meanings of the symbols by means of which the
Burmese express and represent their doctrines. In both cases,
anthropologists who operate exclusively in this mode usually contend that
cultural beliefs can be understood only in the historical context and the
conceptual terms of the culture in which they are embedded. For, so they
claim, in as much as the history which produced them and the symbols
which represent them are unique, their meanings cannot be conveyed
nontrivially by transcultural concepts.
The analysis of cultural beliefs conceived in the second mode goes
beyond the first mode in that it seeks explanations not only of a culturally
parochial, but also of a transcultural provenance. This mode is necessary to
explain certain phenomena which are difficult, if not impossible to explain
in the first mode. Why, to take one example, do cultural doctrines
sometimes die out or sufficiently transform over time as to become
unrecognizable? Or, to take another example, why are some cultural beliefs
no more than cliches, while others are held with strong conviction and
emotional intensity? In the second mode, explanations for such questions
are usually not sought in parochial culture history, but in generic human
experience. Moreover, those analysts who employ the second, as well as the
first mode, usually disagree with the contention of those who employ the
first mode exclusively that the cross-cultural diversity in symbols and
symbol systems implies that the meanings of cultural beliefs are unique and
incommensurable. They contend, rather, that underlying the wide range of
variability in the ‘surface’ meanings of cultural symbols, there is a narrow
range of variability—if not universality—at some ‘deep’ level of meaning,
and that their interpretation must attend to the latter, as well as the former,
meanings.
Since it is this second intellectual mode whose application to the analysis
of religious symbol systems I wish to explore in this chapter, it is necessary
to examine the set of assumptions—at least as I see them—on which it
rests. First, despite the diversity of human cultures, the minds of culture-
bearers everywhere share panhuman characteristics (“psychic unity of
mankind”); second, these characteristics are the product of a shared
biological phylogeny, on the one hand, and a similar social ontogeny on the
other; third, the diversity of cultures represents variable attempts of the
human mind to cope with the existential and adaptive problems of
individual and social living, which vary as a function of diverse historical
experiences and ecological conditions; fourth, the diversity which is found
at the ‘surface’ meanings of cultural beliefs is associated with panhuman
regularities in their ‘deep’ meanings; fifth, these regularities are
transformed, by processes to be described below, into the historically
conditioned, and therefore culturally parochial, meanings of the symbols by
which these beliefs are represented and conveyed. It is primarily in their
surface meanings that the much heralded diversity and relativity of human
cultures, whose proclamation has become the hallmark of anthropology, is
to be found.
Just as the first mode in the study of cultural beliefs can be divided into
culture history and symbolic anthropology, the second mode can likewise
be divided into two types: structuralism and culture-and-personality.
Structuralism, whose founder and most eminent figure is Lévi-Strauss, is
wholly concerned with the cognitive, and more particularly, the intellectual
dimension of the psychic unity of mankind. Hence, it views the resolution
of intellectual puzzles and paradoxes as the crucial feature of the mind, so
far, at least, as its relationship to religious beliefs and mythic themes is
concerned. In their interpretation of myth, for example, structuralists not
only can (and do) ignore all other characteristics of the mind, but they also
exclude from their purview the social actors who acquire and transmit the
myths, focusing their attention on the myths themselves, as they are
recorded in texts.
Culture-and-personality theory, insofar at least, as it takes its departure
from the psychoanalytic conception of the mind, is concerned not only with
the intellectual dimension of cognition, but also with its other dimensions,
as well as with the affective characteristics of the social actors, as on the
structural characteristics of religion and myth, in order to arrive at an
adequate interpretation and explanation for them. It is this approach to
religion that I shall explicate in this chapter. First, however, we must
delineate those characteristics of the mind that are relevant for
understanding religious symbols.

That human beings attempt to maximize pleasure, and that pleasure

involves the gratification of needs, wishes, and desires—whether these be
biological, social, emotional, or intellectual—are two propositions
concerning the human mind which, I take it, would evoke fairly wide
assent. That, typically, wishes and desires can be gratified only by
transactions with the external environment—both physical and social—and
that these transactions are based on specifiable perceptual and mental
processes whose characteristics are universal are also (I expect) widely
accepted assumptions. Thus, with respect to perceptual processes, if there
ever was a society whose members could never distinguish fantasy from
reality, or hallucinatory from veridical perception, such a society, surely, is
part of the fossil record of human history. The same consideration holds
with respect to mental processes. Societies whose members were typically
unable to assess causal relationships, to grasp logical connections, to
construct valid inductive generalizations, or to make valid deductive
inferences—these are societies in which even the simplest of subsistence
activities would have little chance of success. In a word, such societies
would not have survived.
The perceptual and mental processes alluded to above—which, taken
jointly, I shall label as “cognitive”—comprise a type (in the Weberian
sense) in which mentation is governed by normally accepted rules of logic,
and in which ideas and thoughts are represented by verbal signs which are
combined and manipulated by conventional rules of grammar and syntax. It
is a type, moreover, in which the perception of internal stimuli is
distinguished from that of external stimuli. In the study of dreams, fantasy,
and related phenomena, however, we encounter yet another type of
cognition. So far as mentation is concerned, it is a type in which ideas and
thoughts are typically represented by visual signs (both iconic and
symbolic) and whose logic, as we shall see below, is analogous to that
exhibited in metaphor, metonymy, and other tropes. So far as perception is
concerned, it is a type in which stimuli originating in the inner world are
taken as objects and events in the outer world. Following Freud, this type
may be termed the “primary process” mode of cognition, in contrast to the
first type which may be termed the “secondary process” mode.
In this chapter, we shall be primarily concerned with the primary process
mode of cognition. For if, to turn from the individual to the sociocultural
level, economic and technological systems, for example, may be said to be
based more-or-less on the secondary process mode, then I would claim that
mythicoreligious and ritual systems are based to a much greater extent on
the primary process mode. In considering the above discussion of the latter
mode, this claim means, first, that religion and myth possess a logical
structure which differs importantly from that found in the technicoeconomic
domain, strictly conceived; second, it means that they are the cultural
domains, par excellence, in which fantasy is taken for reality. In short, they
are the domains in which the adaptive constraints on the satisfaction of
wishes and desires find an important exception. We shall begin this
discussion with the second claim, deferring the discussion of the first to the
following section.
To better understand the claim that in religion and myth fantasy is taken
for reality, it is instructive to compare these cultural phenomena with
dreams; those psychological phenomena in which this dimension of the
primary process is most obvious and best understood. The dream world is a
reified world. That is, although dreams consist of images of persons and
events, these images are believed by the dreamer to be, rather than to
represent, the persons and events they signify. In short, in the dream the
mental representation of a thing is taken for the thing itself. Similarly, based
on culturally acquired religious doctrines and rituals, as well as mythic
narratives, the religious believer constructs a mental representation—also in
the form of images—of a special and wondrous world of beings and events.
Unlike the dreamer, however, the religious believer does not confuse the
mythicoreligious world with his mental representation of it. Except in
trance and other altered states of consciousness, the images in his mind, as
well as the cultural symbols from which they are constructed, are not
believed to constitute the beings and events that comprise that world; rather,
they are believed to represent them. In short, unlike the world of the dream,
the mythicoreligious world is believed to exist independently of the mental
images and public symbols by which it is represented. In this sense, the
images and symbols of the mythicoreligious world are like the images of
the dream-as-recalled, rather than those of the dream-as-dreamt. For when
the dreamer awakens from his nocturnal slumber, he recognizes (in many
societies at least) that the persons and events comprising his dream-as-
dreamt had, in fact, consisted of his mental representations of them.
Nevertheless, the analogy between religion and the dream-as-dreamt still
holds. In the latter, the images in the mind are not only reified, they are
externalized, that is, these images of persons and events are not only taken
for actual persons and events, but they are located in the external world,
outside of the dreamer. In his waking life, however, the dreamer recognizes
not only that the dream consists of images in his representational world, but
that these images are representations offantasied events, constructions of his
mind, which occur not in the external, but in his internal world. In short, he
recognizes that the dream is a hallucination.1 The case of the religious
believer, however, is rather different. Although distinguishing between the
mythicoreligious world, on the one hand, and the private images and public
symbols by which it is represented, on the other, he nevertheless believes
(in accordance with the teachings of his religious tradition) that these
images and symbols signify real, rather than fantasied, beings and events
which, as in the dream-as-dreamt, he locates in the external world.
1. There are some primitive societies, of course, in which the dream-as-recalled is taken as a
memory not of a nocturnal hallucination, but of an actual happening. For them, in short, the line
between fantasy and reality (so far at least as dreams are concerned) is blurred. This does not mean
that they are “pre-logical” in a wholesale sense, for the boundary between fantasy and reality is
confined to a restricted domain. Nevertheless, it does mean that the primary process mode is a more
prominent feature in the mental functioning of these societies than in those that do recognize the
hallucinatory quality of the dream. It also means, moreover, that for such societies the reality of the
mythicoreligious world poses a problem of a much smaller magnitude than those in which the
boundary between fantasy and reality is drawn much more sharply.
 Since, then, except for those who enter into trance and similar
experiences, there is no experiential ground for believing in the external
reality of the mythicoreligious world, the first problem posed by the
analogy between religion and dreams is why the religious believer does not
(like the awakened dreamer) awaken from his religious slumber and
recognize that the mythicoreligious world exists not in some external
reality, but rather in the inner reality of the mind. The explanation, I think,
is twofold. First, there are obvious differences between the images and
symbols of the mythicoreligious world and those of the dream world.
Second, there are certain characteristics of the mind which predispose
human actors to believe in the external reality of the mythicoreligious (but
not of the dream) world.
Before examining the first explanation, something must be said about
that slippery word, ‘symbol’. which I have thus far been skirting. Following
Charles Peirce, I use ‘symbol’ as one type of sign, a ‘sign’ being any object
or event which represents and signifies some other object and event, or a
class of objects and events. A symbol, then, is that type of sign in which (to
shift from Peirce to Saussure) the relationship between the signifier and the
signified is arbitrary. It is this dimension of the symbol which distinguishes
it from Peirce’s other two types of signs: the ‘index,’ in which there is a
factual contiguity between signifier and signified, and the ‘icon,’ in which
there is a factual similarity between them. By these definitions, a relic of the
Buddha is (or is believed to be) an index of him; a sculpture of the Buddha
is (or is believed to be) an icon of him; and the word “Buddha,” is a symbol
for him. Icons and symbols (but not indices) may be internal (i.e., we may
have iconic or symbolic mental representations of objects or events) or they
may be external (i.e., objects and events may be represented by physical
icons and symbols in the external world). Cultural symbols and icons are, of
course, external signs whose meanings are public (or shared) and
conventional (handed down by tradition). With some few exceptions, of the
kind already mentioned, cultural signs are symbols.
On the basis of this brief definitional excursus—I shall have more to say
about icons and symbols below—we may now turn to the first difference
between the dream world and the mythicoreligious world which, as
suggested above, might account for the belief in the external reality of the
latter world, and its repudiation in the case of the dream world.
 In reference to the dream world, it is necessary to distinguish between the
dream-as-dreamt and the dream-as-recalled. (These two versions of the
dream—in both versions I am concerned with the manifest content only—
must also be distinguished, of course, from the dream-as-reported; but the
latter version, important as it is for other purposes, is tangential to the
purpose of this paper.) The dream-as-dreamt is a fantasy world which,
represented in private and internal images (iconic and symbolic signs), is
the dreamer’s own creation. In constructing this world, the dreamer of
course calls upon the memories of his own social experiences, the history of
his group, cultural performances and traditions, and the like; but whatever
its social and cultural inspirations, it is a representation of his own wishes
and fears, hopes and anxieties. The dream-as-recalled is a representation, in
memory (whether fragmentary or complete, distorted or veridical) of the
dream-as-dreamt. Unlike the latter, however, which is experienced as an
actual event, the dream-as-recalled is usually taken to be a memory of what
the dream really is—a fantasied event. That it might, rather, be a memory of
an actual event is usually contradicted by other events in the dreamer’s
waking life, as well as by the testimony of those of his fellows who may
have comprised the dramatis personae of the dream.
Although the mythicoreligious world is also a fantasy world, rather than
invented by himself, it is acquired by the religious believer from his
cultural traditions. These traditions, which proclaim the historicity and
factuality of that world, are transmitted by means of cultural signs—the
verbal symbols of religious doctrine and myth, and the visual symbols of
ritual and sacred drama. Hence, unlike the dream world, the
mythicoreligious world is represented not only by the private and internal
images of the religious believer, but also by the public and external symbols
of his culture. Indeed, it is from these collective representations of the
mythicoreligious world that the dreamer constructs his mental
representation of it.
In short, one possible explanation for religious believers holding to the
external reality of the mythicoreligious world, while denying such reality to
the dream world, is that the latter is constructed from private thoughts, the
former from cultural traditions. This difference has three important
consequences. First, the reality of the mythicoreligious world is not only
proclaimed by the full authority of tradition, but it is confirmed by the ever-
present (and psychologically compelling) cultural symbols from which the
believer’s representation of this world is constructed in the first place.
Second, the fantasy quality which characterizes any mental representation
consisting of images, is blunted in the case of the mythicoreligious world
because of its simultaneous representation in verbal symbols. Hence, unlike
the dream-as-dreamt whose reality, upon awakening, is challenged by its
chaotic, fragmentary, and often bizarre quality, the relatively systematic and
coherent character of religious belief systems and myth narratives presents
less of a challenge to the reality of the mythicoreligious world. Third, since
culturally-constituted symbols are public (and their meanings are therefore
widely shared), the actor’s belief in the correspondence between the
mythicoreligious world and his mental representation of it is confirmed by
the consensual validation of his fellow.
Important as they are, these formal differences between the private
images of the dream and the public symbols of religion are not, in my view,
a sufficient explanation for the belief in the reality of the mythicoreligious
world. They do not explain, for example, why religious doctrines persist
even in the face of competing, and often compelling, counter-claims of fact
or reason, nor why cognitive dissonance is resolved not by abandoning the
doctrines, but rather by resting their truth in faith. These facts (and others)
suggest, as William James pointed out long ago, that religious belief
ultimately rests on the actors’ “will to believe,’’ an intellectual posture
which perhaps finds its extreme expression in Tertullian’s precept credo
quia absurdum est, I believe because it is absurd.
This brings us to the second explanation for the belief in the reality of the
mythicoreligious world mentioned above, for to speak of the will to believe
is, of course, to shift our attention from the belief, and its representational
medium, to the believer and his mind. In introducing this shift, I should like
to distinguish “religion-as-a-doctrinal structure” from “religion-in-use,” a
distinction analogous to Saussure’s distinction between ‘langue’ and
‘parole’. By “religion-as-a-doctrinal structure” I refer to the organization of
religious doctrines taken as a cognitive system, that is, a system of
propositions together with their constituent meanings. In affirming these
doctrines, the religious actor, unlike the religious philosopher, is concerned
not only with their meanings, but also with what James termed their ‘cash
value’. That is, the religious actor is not so much interested in theory as in
praxis (to employ a much over-worked distinction), and it is the latter
dimension of religion to which I refer by the expression, “religion-in-use.”
As the expression indicates, this dimension refers to the purposes to which
the religious actor puts his beliefs.
To say that religious actors are primarily concerned with religion-inuse is
to say that although religious systems are cognitive systems, they persist not
only because of the cognitive basis for the belief in the reality of the
mythicoreligious world, nor even because its symbols are good to think, but
because the belief in its reality satisfies some powerful human needs. In
referring to the satisfaction of needs, I am of course alluding to the
functions of religion—not, however, to its social functions, but to those
which it serves for the religious actors themselves. Here I would follow
Max Weber’s contention that religion serves two universal functions, both
of which are related to the vexatious problem of suffering—illness, death,
drouth, loss, bereavement, madness, and so on. First, it provides answers to
the intellectual problem of the existence of suffering and its seemingly
unfair and inequitable distribution (the theodicy problem). Second, it
provides various means for overcoming suffering, both as a temporary
achievement and a permanent victory (salvation).
If, despite the remarkable cross-cultural diversity in its structure and
content, religion universally serves (at least) these two functions, then it
follows that the latter are related to two corresponding need-dispositions of
the human mind which preadapt social actors to believe in the reality of
their mythicoreligious worlds. These comprise the need to explain and find
‘meaning in that which is otherwise inexplicable and meaningless, and the
need to conquer the intolerable anxiety attendant upon painful and
frightening situations that are beyond human ability to effect or control. If
this proposition seems banal, it is nevertheless important to state as a
reminder that even the most radical cultural relativist can hardly begin to
understand the persistence of religion—or much else about human culture
—without postulating some set of need-dispositions as a universal
characteristic of the human mind.
Powerful as it might be, however, motivation alone is not a sufficient
explanation for the belief in the reality of the mythicoreligious world. Such
a belief persists, I would suggest, because social actors are preadapted
cognitively, as well as motivationally, to believe in its reality. Furthermore,
I would suggest, this cognitive preadaptation is derived from two biological
(hence universal) characteristics of childhood—prolonged helplessness and
extended dependency—as a result of which cultural systems, when viewed
ontogenetically, are not the first resource from which social actors construct
their representational world.
Beginning from birth—hence prior to the acquisition of language and the
culturally-constituted conceptions of the world which language makes
possible—children develop what might be called socially-constituted
conceptions as a consequence of (prelinguistic) transactions with parents
and other parenting figures. Hence, long before they are taught about the
powerful beings who inhabit the mythicoreligious world young children
have persistent and prolonged experiences, often accompanied by intense
affect, with these powerful beings who inhabit their family world. Entirely
helpless from birth, and absolutely dependent on these beings, young
children form highly distorted, exaggerated, and even bizarre
representations of these parenting figures. To be sure, as they grow older
most (but not all) children relinquish these representations—often, however,
after considerable struggle—in favor of more realistic conceptions of them.
At first, however, these bizarre and distorted images, the products of
primary process cognition, are unconstrained by the secondary process
cognition characteristic of mature ego-functioning; that type of cognition
which depends on the achievement of ‘object constancy,’ language
competence, and ‘reality-testing’. Let us examine each of these in turn.
Prior to the attainment of the developmental cognitive stage of object
constancy, the representations of different types of experience with one and
the same person are not yet integrated by the child so as to form an
organized, albeit differentiated, representation of him; rather, each type of
experience produces a separate representation. Hence, although the actual
parent is typically both good and bad, helpful and harmful, dependable and
undependable, the young child, by a process known as ‘splitting’ forms
separate images of a helping figure, a harming figure, a frustrating figure,
and so on.
But even with the achievement of object constancy, it is difficult to form
an integrated representation of one and the same person until the acquisition
of language. For when images, rather than words, constitute the
representational medium, the portrayal of different, but especially opposing,
attributes of the same person—nurturant and punitive, good and bad, etc.—
in a single image is difficult, if not impossible to achieve by means of such
a medium, as any dreamer or sculptor knows. Hence, typically, the
prelinguistic child forms different and opposing representations of the same
parent, rather than one, conceptually integrated representation of him or her.
In addition to this cognitive basis, however, there is an equally important
affective basis for the young child’s splitting the opposing characteristics of
his parents into separate representations. The integration of the loving and
loved parent and the frustrating and hated parent into a single representation
presupposes a degree of emotional maturity not yet attained by the young
child. The inner conflict resulting from hating the person he loves, and is
dependent upon, is beyond his emotional capacity to tolerate. Moreover,
given the child’s lack of reality testing, to hate someone is to destroy him,
and since he both needs and loves the parent, this potentially intolerable
conflict is obviated by splitting these opposing mental representations of the
parent.
Having mentioned the concept of ‘reality testing,’ we may now explore
its relevance for our thesis in greater detail. In order to do so, we shall once
again return to the dream. I have already noted certain ways in which, with
respect to their mental functioning, the cognitive stage of the prelinguistic
child is similar in some respects to the cognitive state of the dreamer. I now
wish to examine yet another similarity. The dream, I have already observed,
is a nocturnal hallucination in which the dreamer, whose reality-testing is
impaired, does not distinguish fantasy from reality, nor does he distinguish
the fantasies themselves from the images by which they are represented,
with the result that these images are reified.
Clinical data suggest that these same cognitive confusions may be found
in the mental functioning of the prelinguistic child, not because his reality-
testing is impaired, but because it is still undeveloped. Thus, for example,
the young child’s mental images of his parenting figures, just like dream
images, may be reified, and thereby experienced as autonomous agents.
Since, moreover, the boundary between inner and outer experience is
blurred at this age, these reified agents may be experienced as located
within himself (whence they are labeled, in the terminology of
psychoanalysis, ‘introjects’), or they may be externalized and located in the
outer world (in which case they are labeled ‘projections). Although as the
ego develops reifications are gradually given up, they are nevertheless not
relinquished easily, as is indicated by the projections which form the basis
for the imaginary playmates of children, and by the introjects which are the
basis for spirit possession. (Those few adults who never give up these
reifications suffer severe psychopathology; for example, psychotic
depression, in the case of persistent introjects, and paranoid delusions, in
the case of persistent projections.) Rather than being relinquished, however,
the externalized reifications of the early parental images may instead
undergo a transformation, and it is this vicissitude of these projections with
which we are concerned here.
In societies in which there is a high degree of integration between social
and cultural systems, the child’s early experiences with his parents may lead
him to construct mental representations of them which, structurally, at least,
are isomorphic with the mental representations of the superhuman beings of
the mythicoreligious world whose characteristics are only subsequently
conveyed through the verbal and visual symbols of his culture. If one
considers the typical mythicoreligious world—with its gods and demons,
saviors and satans, redeemers and destroyers—then it becomes apparent
that the socially-constituted images which young children form of the
powerful beings comprising their family world are highly similar to the
culturally-constituted images which, at a later age, they form of the
powerful beings comprising the mythicoreligious world. Since, then, the
former images, with all their bizarre distortions and exaggerations,
represent and signify actual beings whose reality they have personally
experienced, we may say that children are cognitively preadapted to believe
in the reality of the superhuman beings that are represented and signified in
the external collective representations of mythic narratives and religious
ritual, as well as in the mental images which children form of them.
But given the fact that the child’s early mental images of his parenting
figures are reified and externalized, I would claim even more. For, I would
suggest, when the child constructs his mental representations of the
superhuman figures of the religious world, they may be merged (identified)
with the corresponding representations he had previously constructed of the
parenting figures of his family world, thereby forming a single
representational world. When this occurs, the child’s projections of his
parental images may be retained without any psychopathological
entailments, for they are then assimilated to his images of the superhuman
beings whose existence is taught by religion and myth. At the same time,
this process assures the belief in the external reality of these superhuman
beings, for they are now merged with the reified and externalized images of
those powerful human beings whose external reality he has himself
experienced. (In rapidly changing societies, or in any other in which there is
only minimal integration between social and cultural systems, the self-
evident belief in the reality of the mythicoreligious world is maximally
jeopardized, with the result that the belief may be relinquished, or—as I
have already observed—proclaimed as an article of faith.)
Thus, to take an example near home, when God is referred to as “Our
Father who art in heaven,” the cultural symbol, “Father,” may be said to
have two simultaneous meanings for the religious believer, one at the
‘surface’ level, the other at a ‘deep’ level. Since it is God who is designated
as “Father,” and since He is not literally conceived by the religious believer
to be his father—whether genitor or pater—the surface meaning of “Father”
is obviously a metaphorical one. That is, with respect to certain of His
attributes—justice, mercy, love, etc.—God, who resides in heaven, is
conceived to be like his father (pater) or, at any rate, like the normative
conception of father, as that conception is informed by Western values
regarding fatherhood. If, however, his mental representations of
superhuman beings are merged with the religious believer’s projections of
his mental representation of the parents of childhood, then, in its deep
meaning, “Father” is taken literally. For although God is not conceived by
the believer to be his actual father (the one who is, or at least was, on earth),
He is conceived, according to this explanation, as one of the reified and
externalized representations of his childhood father—or a composite
representation of some of them—which, in accordance with religious
doctrine, the believer locates in heaven. Since in this sense, but in this sense
only, God is indeed his father who is in heaven, in its deep meaning,
“Father” is taken literally.

Let me summarize my argument thus far. My main point has been that
the belief in the reality of the mythicoreligious world, a belief in which
culturally-constituted fantasy is invested with the appearance of reality, may
be explained to a large extent as a function of the primary process mode of
cognition. The cultural conceptions of the superhuman beings who inhabit
that world are conveyed, of course, by the external cultural symbols by
which they are represented—words, icons (sculpture and painting), and
ritual—and from these collective representations the believer forms his
mental representations of them. That these beings are believed to exist
independently of the collective, as well as the mental, representations which
signify them is best explained by the correspondence that exists between
these representations and the mental representations that the young child
previously forms of those actual powerful beings whose reality he has
personally experienced—his parenting figures. These representations are
based on the primary process mode of cognition because, in the absence of
language, the representational medium consists of images; in the absence of
object constancy, these images are formed by the process of splitting; and in
the absence of reality testing, they are reified and externalized. It is the
merging of the believer’s mental representations of the parents of early
childhood with his or her mental representations of the superhuman beings
comprising the mythicoreligious world that constitutes the cognitive basis
for the belief that the mythicoreligious world exists independently of the
collective representations by which it is both represented and signified.
By this explanation for the belief in the reality of the mythicoreligious
world, religious symbols have both ‘surface’ and ‘deep’ meanings, and no
interpretation of any particular religion is complete unless its symbols are
interpreted at both levels.2 For this reason, the interpretation of religion
(and other cultural systems) is similar to the interpretation of a dream in that
the knowledge of its manifest content alone can be highly misleading
without knowledge of its latent content.
2. The current interest in cultural hermeneutics persistently distinguishes between interpretation
and explanation, interpretation being viewed as a humanistic endeavor, concerned with intentions’
purposes, goals, and the like, while explanation is viewed as a scientific or positivistic endeavor,
concerned with the search for causal and functional ‘laws.’ In my view, this is a false dichotomy. If
the former endeavor is concerned with producing valid, rather than just any kind of interpretations, it
must be no less concerned with ‘laws,’ than the former, because, of course, the cogency, if not
validity of the idiographic interpretation is dependent on the nomothetic theory from which it is
implicitly or explicitly deduced.
According, then, to this explanation, the external and public symbols of
religion—its collective representations—represent and signify at their
‘surface’ level the superhuman beings whose existence is affirmed by the
various culturally parochial, religious traditions—Jahweh, Allah, Siva, the
Madonna, Durga, and the like. These are their conscious and culturally
variable meanings. At their ‘deep’ level, however, these symbols represent
and signify the projections of the mental representations of the parents of
early childhood. These are their unconscious, and culturally universal
meanings. (Such an interpretation of the collective representations of
religion might be contrasted with that of Durkheim who, it will be recalled,
viewed them—in their ‘deep’ meanings—as signifying society.)
 In sum, I have argued thus far that underlying the cross-cultural diversity
in the surface meanings of culturally parochial religious symbols, there are
universal deep processes and meanings. If this is so, these cultural symbols
effect three important psychological transmutations in the religious actors:
transmutation of infantile into adult conceptions, of individual into public
meanings, and of unconscious into conscious concerns. The satisfaction of
these adult, public, and conscious concerns—especially those related to the
explanation and conquest of suffering—constitutes, so I have argued, the
most important manifest function of religion, providing a powerful
motivational basis for the belief in the reality of the mythicoreligious world.
However, if religious symbols also have deep meanings, then religion not
only has manifest functions related to the surface meanings of these
symbols, but it must also have latent functions related to their deep
meanings. Hence in this, the concluding section of this paper, I wish to
argue that religion attends not only to the conscious and public concerns of
the actor’s adult-like experience, but also to the unconscious and private
concerns of his child-like experience. For if religious symbols are
associated with unconscious infantile mental representations, it can only be
because in addition to their conscious, adult concerns, social actors retain
unconscious, infantile concerns, and it is their satisfaction that constitutes
the latent function of religion. The intention of satisfying these concerns
constitutes yet another—an unconscious—motivational basis for the belief
in the reality of the mythicoreligious world.
Since dreams constitute the most important private symbol system for the
gratification, in fantasy, of infantile needs, I shall turn once again to the
dream to help us understand the motivational aspects of unconscious
symbolic processing. Since in this context, however, we are interested not
in the hallucinatory, or ontological dimension of primary process cognition,
but in its “logical” dimension, we shall seek assistance from poetry as well.
(The argument of this section of the paper is similar to, but also differs to
some extent from one I have previously developed elsewhere. See Spiro
1977:xix-xxx.)
Should a poet wish to represent a conception of a friend—his bravery for
example—he may convey this conception in a simple prose sentence, “John
is brave”; in a figure of speech, such as the simile, “John is like a bull”; or
in a trope, such as the metaphor, “John is a bull.” In the metaphor, the
intended meaning of the verbal symbol, “bull,” is figurative rather than
literal, for it is intended to represent the poet’s conception of John, rather
than to signify the brave bull in Madrid’s corrida. Unlike the poet, the
dreamer has fewer degrees of freedom to express his thoughts because a
representational medium consisting of images cannot directly represent
qualities, such as bravery, which in language are represented by adjectives,
adverbs, and similar parts of speech. In such a medium, which only contains
the structural equivalents of nouns and verbs, the thought, “John is brave,”
cannot be represented in a form analogous to a simple sentence, let alone a
simile. Rather, given the constraints of his medium, the dreamer, just like
the painter or sculptor, can only represent such a thought in a form
analogous to a trope. Hence, to represent the thought that John is brave, he
may dream of a bull. Like the verbal trope of the poet, the visual trope of
the dreamer can be misleading to one who does not understand the code.
Thus, though the consciously intended meaning of the bull is figurative,
rather than literal, inasmuch as an image of a bull is conventionally taken to
be a representation of a certain species of bovine, it is a conventionally
inappropriate sign for a human being.
To put it in Peirce’s classification of signs, since the image of the bull is
intended as a symbol for John, its meaning will be misunderstood if it is
taken as an icon for a bull. And this potential confusion is precisely one of
the difficulties that is posed by the interpretation of dreams, as well as one
of the reasons for their seemingly bizarre qualities. For although in the
sleeping code by which he constructs his dream, the dreamer consciously
intends the bull to be a symbolic representation of John, in his waking code
by which he interprets the dream, it is taken by him to be an iconic
representation of a bovine. In short, in the dream-as-recalled, the image of
the bull is taken literally, though in the dream-as-dreamt it was intended
figuratively.
The poet, of course, does not have an analogous problem—though his
reader may—because he uses the same code in reading his poem that he had
used when composing it. On both occasions, the conventionally
inappropriate verbal symbol, “bull,” is consciously understood by him to be
a metaphor, a form which he chose in the first place in order to convey his
conception of John more effectively, forcefully, or artistically than might
have been achieved by a simple prose sentence. In short, both in the case of
the poem and of the dream-as-dreamt, the figurative meaning of the sign—
the word in the former, the image in the latter—is its consciously intended,
and only, meaning.
In addition, however, to the representational constraints of his medium,
the dreamer may set forth his thoughts in conventionally inappropriate
images for yet another reason: the wish to disguise them. All of us have
thoughts that are repugnant to our moral values, and since such thoughts are
painful—they arouse moral anxiety—we often repress them, i.e., eliminate
them from conscious awareness. Let us suppose, then, that in his waking
state a dreamer has a repressed thought concerning his friend, John—the
thought, for example, that he would like him to die. Let us further suppose
that this thought continues in his sleep. If the dreamer were to distort this
thought, by substituting a bull for John as the object of his wish, then he
might, with moral impunity, gratify this disguised wish in a dream in which
he kills a bull. In such a dream, the image of the bull has two meanings
simultaneously—one literal, the other figurative. Its literal meaning (bull) is
consciously intended, while its figurative meaning (John) is unconsciously
intended. Since in this dream, unlike the first, the image is an unconscious
symbol for John—consciously, of course, it is taken as an icon of a bull—
the substitution of a bull for John is an example not of a trope, but of a
defense mechanism; that is, it is a cognitive maneuver in which a forbidden
wish undergoes unconscious symbolic distortion in the service of a
disguise.
Let me now summarize very briefly the formal characteristics of defense
mechanisms, in contrast to tropes, in somewhat more technical terms: (a) In
a defense mechanism the symbolic distortion of the wish is overdetermined,
that is, it is based on multiple and simultaneous motives, including the
motives to gratify and—since it is forbidden—to disguise a wish; (b)
Disguise and gratification alike are achieved by displacement,, an
unconscious process in which a conventionally inappropriate sign is
substituted for an appropriate one. (Displacement is based on the same
criteria—similarity or contiguity between the original and substituted
objects signified by the two signs—that are employed in the symbolic
substitutions found in metaphor, metonymy, synecdoche, and other tropes.);
(c) Hence, the substitute sign is characterized by condensation, i.e., it has
two or more simultaneously intended meanings, at least one of which is
unconscious; (d) The conscious, or manifest meaning of the sign is its literal
meaning; its unconscious, or latent meaning is its figurative meaning.
 Let us now apply this analysis of the defensive use of the private symbols
comprising the dream to the cultural symbols comprising religion. As a
cultural system, religion attends in the first instance, as I have stressed more
than once, to the public and conscious concerns of the believers’ adult
experience, especially their concern with suffering in both its intellectual
and existential dimensions. That is, it attends to the needs to both explain
and overcome suffering. To achieve the latter end, the religious actor
engages in ritual transactions with the superhuman beings comprising the
mythicoreligious world. Some of these beings, kindly and benevolent, he
turns to for assistance and aid in his attempt to cope with suffering. Others
of them, aggressive and malevolent, are often viewed as the cause of
suffering, and these he attempts to drive out or drive off. The former type
arouse his wishes for and emotions of dependency and succorance; the
latter type arouse his aggression, fear, and hatred.
Although such postures of dependency and aggression—whether
expressed in the form of wishes, emotions, or actions—are culturally
appropriate for adults in the religious contexts in which they are aroused
and displayed, they are usually considered inappropriate for them in other
contexts. There is one context, in particular, in which they are especially
inappropriate; that context, of course, is the family. As the child’s most
significant others, his parents are at once his most important frustrating
figures (consequently, the targets of his most intense aggressive feelings
and wishes) and his most important nurturant figures (consequently the
objects of his most intense dependency feelings and wishes). Parents are
also, however, the very persons concerning whom, following an initial,
culturally variable period of indulgence, the cultural prohibitions against
dependency and aggression are most severe. The reasons are obvious.
Social survival requires that children eventually outgrow their dependency
on their family of origin, and that, having achieved independent status, they
establish their own families and become objects for the dependency of their
own children. Similarly, since aggression within the family is entirely
disruptive of its integration, if not survival—hence inimical to its vital
individual and social functions—it too must be prohibited.
This being so, every social actor and every society is confronted with an
acute existential dilemma. Although his parents are the objects of the
child’s most intense dependency and aggressive needs and wishes, they are
also the persons concerning whom their gratification is eventually most
strongly frustrated. For although as children grow older and become adults,
they learn to comply with the cultural norms prohibiting the overt display of
aggression toward and dependency upon parents, this does not mean that
these infantile needs are extinguished. That the contrary is the case is
indicated not only by an abundance of clinical evidence, but also by
commonplace observations of everyday life which indicate that these
emotions and wishes are capable of arousal—and not only in a displaced
form—in certain contexts, at least, and under certain provocations.
In sum, then, I am arguing that the intense dependency and aggressive
wishes of children concerning parents, though seemingly extinguished,
continue to exist in a repressed state in adults. Like all repressed wishes,
these too seek gratification, and like them they are typically gratified—if
they are gratified at all—in symbolic disguise. In addition to dreams,
repressed wishes may be represented and (partially) gratified in the many
privately constructed symbolic forms (including symptoms) which have
been described and classified by psychiatrists. Typically, however, such
wishes, particularly if they are widely shared, are represented and gratified
in culturally-constituted rather than privately-constructed symbolic forms.
Although many cultural systems—from games through politics—can be
and have been used for this purpose, I would argue that religion is the
system, par excellence, which is used for the disguised representation and
gratification of the repressed wishes with which we are concerned here—
dependency and aggressive wishes with regard to the parents of childhood.
This is certainly the case in traditional societies, and if newspaper reports
and television broadcasts can be taken as evidence, it is also the case, to a
larger extent than we usually credit, in certain strata of modern society as
well.
That religion should be a focal system for the gratification of these
wishes is hardly surprising if the explanation which I have offered for the
meaning of its symbols is valid. For if the cultural symbols which represent
the superhuman beings of the mythicoreligious world signify, in their ‘deep’
meaning, the reified and externalized mental representations of the parents
of childhood, what better way to express and gratify unconscious rage
toward and dependency longings for these parents than through the vehicle
of religious beliefs and rituals?
My thesis, then, is that when religious actors invoke the assistance of
benevolent superhuman beings, or exorcise malevolent superhuman beings,
they are not only consciously gratifying dependency and aggressive needs
in regard to beings who are their culturally appropriate objects and targets,
but they are also doing more than that. For if the actors’ mental
representations of these benevolent and malevolent superhuman beings are
merged with the reified and projected representations of their kindly and
hateful parents of childhood, then, they are simultaneously, but
unconsciously, gratifying their dependency and aggressive needs in regard
to their childhood parents, their culturally inappropriate objects and targets.
That religion-in-use serves this (latent) function explains at least one of the
unconscious motivational bases for the belief in the reality of the
mythicoreligious world. I might add that if this argument is valid, religion
also serves an equally important latent function for society. For if religion-
in-use is a means for the symbolic gratification of these powerful infantile
needs, society is thereby spared the highly disruptive consequences of their
direct gratification. But that is a topic for another paper.
We may now summarize the implications of this paper, with respect to
both its specifically religious argument and its more general cultural-
symbolic argument. The former argument has been concerned with only
one problem related to the explanation of religious systems, the problem of
why religious actors believe in the reality of the mythicoreligious world.
Whether or not the particular solution offered here is correct is less
important, however, than its underlying thesis that a comprehensive
explanation for such a belief must attend to at least three dimensions of the
problem: (a) the private, as well as public meanings of religious symbols;
(b) their ‘deep,’ or socially acquired meanings, as well as their ‘surface,’ or
culturally transmitted meanings; and (c) the latent, as well as the manifest
functions of the actor’s belief that these symbols signify a real, and not
merely a representational world. An explanation that ignores any of these
dimensions is, I have tried to show, incomplete.
To arrive at such a comprehensive explanation, I further attempted to
show, we must be as much concerned with the properties and processes of
the human mind as with the properties of cultural symbols and the doctrines
which they represent. Although Durkheim’s insistence that collective
representations constitute the focus of anthropological investigation marked
a giant leap forward in the study of sociocultural systems, he made a serious
error in ruling out the study of mental representations as irrelevant to their
study. For, as this paper has attempted to show, any cultural system is a vital
force in society so long as there is a correspondence between the symbols in
which cultural doctrines are represented and their representation as beliefs
in the minds of social actors. When such a correspondence does not obtain,
a cultural system may yet survive, but it survives as a fossil—as a set of
cliches—rather than as a living force. If this is so, then the study of mental
representations is no less important than that of collective representations
for the anthropological enterprise.
My argument, however, makes an even more radical claim, namely, that
in attending to the human mind it is as important to understand its
unconscious, as well as its conscious processes. Although a knowledge of
conscious cognitions and motives can help us to understand the ‘surface’
meanings and manifest functions of cultural symbols, knowledge of
unconscious cognitions and motives is required to understand their ‘deep’
meanings and latent functions. Lest I be misunderstood, I have not argued,
as an older generation of psychoanalytic theorists was sometimes prone to
do, that the latter meanings and functions are more important for the
understanding of symbols (whether cultural or noncultural) than the former.
I have argued, however, that they are no less important.

Notes
Reprinted from On Symbols in Anthropology: Essays in Honor of Harry Hoijer, edited by Jacques
Maquet (Malibu: Udena Publications, 1982), pp. 45–72. © 1982 by Melford E. Spiro.
I am indebted to Jacques Maquet and Walter Goldschmidt for inviting me to participate in the
second series of lectures in honor of Harry Hoijer, and to Roy G. D’Andrade, Fitz John Poole,
Theodore Schwartz, Marc J. Swartz, and Donald F. Tuzin for their valuable criticisms of an earlier
draft of this paper. I wish to acknowledge my gratitude to the National Institute of Mental Health for
its support of a research project on the comparative study of culturally-constituted defense
mechanisms, some of whose findings are incorporated in this paper.

References

Spiro, Melford E. 1969. Discussion. In Forms of symbolic action, ed.

Robert F. Spencer. Seattle: University of Washington Press (for the
American Ethnological Society), 208–14.
---------. 1977. Kinship and marriage in Burma. Berkeley: University
of California Press.
---------. 1978. Burmese supernaturalism. Expanded edition.
Philadelphia: Institute for the Study of Human Issues.
III
RELIGION AND MYTH
8
Religion: Problems of Definition and
Explanation

Introduction 1
1. Work on this paper is part of a cross-cultural study of religion supported by research grant M-
2255 from the National Institutes of Health, U.S. Public Health Service.

BEFORE EXAMINING various approaches to the explanation of religion, we

must first agree about what it is that we hope to be able to explain. In short,
we must agree on what we mean by ‘religion’. Anthropology, like other
immature sciences—and especially those whose basic vocabulary is derived
from natural languages—continues to be plagued by problems of definition.
Key terms in our lexicon—culture’, ‘social system’, ‘needs’, ‘marriage’,
‘function’, and the like—continue to evoke wide differences in meaning
and to instigate heated controversy among scholars. Frequently the
differences and controversies stem from differences in the types of
definition employed.
Logicians distinguish between two broad types of definition: nominal and
real definitions (Hempel 1952:2–14). Nominal definitions are those in
which a word, whose meaning is unknown or unclear, is defined in terms of
some expression whose meaning is already known. We all engage in such
an enterprise in the classroom when we attempt to define, i.e. to assign
meaning to, the new terms to which we expose our untutored
undergraduates. Our concern in this case is to communicate ideas efficiently
and unambiguously; and, in general, we encounter few difficulties from our
students, who have no ego-involvement in alternative definitions to our
own. We do have difficulties with our colleages, however, because they—
unlike us!—are ego-involved in their immortal prose and, intransigently,
prefer their nominal definition to ours. Despite their intransigence, however,
the problem of achieving consensus with respect to nominal definitions is,
at least in principle, easily resolved. We could, for example, delegate to an
international committee of anthropologists the authority to publish a
standard dictionary of anthropological concepts, whose definitions would
be mandatory for publication in anthropological journals.
The problem is more serious, and its resolution correspondingly more
difficult, in the case of real definitions. Unlike nominal definitions which
arbitrarily assign meaning to linguistic symbols, real definitions are
conceived to be true statements about entities or things. Here, three
difficulties are typically encountered in anthropology (and in the other
social sciences). The first difficulty arises when a hypothetical construct—
such as culture or social structure—is reified and then assigned a real
definition. Since that which is to be defined is not an empirically observable
entity, controversies in definition admit of no empirical resolution.
A second difficulty is encountered when real definitions are of the kind
that stipulate what the definer takes to be the ‘essential nature’ of some
entity. Since the notion of ‘essential nature’ is always vague and almost
always nonempirical, such definitions are scientifically useless. Kinship
studies represent a good case in point, with their—at least so it seems to a
nonspecialist—interminable controversies concerning the essential nature
of marriage, descent, corporality, and the like.
Sometimes, however, real definitions are concerned with analyzing a
complex concept—which has an unambiguous empirical referent—by
making explicit the constituent concepts which render its meaning. These
are known as analytic definitions. Thus, the expression ‘X is a husband’ can
be defined as ‘X is a male human, and X is married to some female human’.
But the possible objections which such a definition would evoke among
some anthropologists, at least, exemplifies the third definitional difficulty in
anthropology: what might be called our obsession with universality. Since
there are instances in parts of Africa of a phenomenon similar to what is
ordinarily termed ‘marriage’, but in which both partners are female, some
scholars would rule out this definition on the grounds that it is culturally
parochial. This insistence on universality in the interests of a comparative
social science is, in my opinion, an obstacle to the comparative method for
it leads to continuous changes in definition and, ultimately, to definitions
which, because of their vagueness or abstractness, are all but useless. (And
of course they commit the fallacy of assuming that certain institutions must,
in fact, be universal, rather than recognizing that universality is a creation
of definition. I am also at a loss to understand why certain institutions—
marriage, for example—must be universal, while others—such as the state
—need not be.)

The Problem of Definition in Religion

An examination of the endemic definitional controversies concerning
religion leads to the conclusion that they are not so much controversies over
the meaning either of the term ‘religion’ or of the concept which it
expresses, as they are jurisdictional disputes over the phenomenon or range
of phenomena which are considered to constitute legitimately the empirical
referent of the term. In short, definitional controversies in religion have
generally involved differences in what are technically termed ostensive
definitions. To define a word ostensively is to point to the object which that
word designates. In any language community, the fiery ball in the sky, for
example, evokes a univocal verbal response from all perceivers; and a
stranger arriving in an English-speaking community can easily learn the
ostensive definition of the word ‘sun’ by asking any native to point to the
object for which ‘sun’ is the name. Similarly the empirical referent of
‘table’ can be designated unequivocally, if not efficiently, by pointing to
examples of each subset of the set of objects to which the word applies.
The community of anthropologists, however, is not a natural language
community—more important, perhaps, it does not share a common culture
—and although there is little disagreement among anthropologists
concerning the class of objects to which such words as ‘sister’, ‘chief,
‘string figure’—and many others—properly do apply, there is considerable
disagreement concerning the phenomena to which the word ‘religion’ ought
to apply. Hence the interminable (and fruitless) controversies concerning
the religious status of coercive ritual or an ethical code or supernatural
beings, and so on. From the affect which characterizes many of these
discussions one cannot help but suspect that much of this controversy
stems, consciously or unconsciously, from extrascientific considerations—
such as the personal attitudes to religion which scholars bring to its study.
Since I am concerned with the logic of inquiry, I must resist a tempting
excursion into the social psychology of science.
The scientific grounds for disagreement are almost always based on
comparative considerations. Thus Durkheim rejects the belief in
supernatural beings as a legitimate referent of ‘religion’ on the grounds that
this would deny religion to primitive peoples who, allegedly, do not
distinguish between the natural and the supernatural. Similarly, he rejects
the belief in gods as a distinguishing characteristic of ‘religion’ because
Buddhism, as he interprets it, contains no such belief (1954:24–36). Such
objections raise two questions; one factual, the other methodological. I shall
return to the factual question in a later section, and confine my present
remarks to the methodological question. Even if it were the case that
Theravada Buddhism contained no belief in gods or supernatural beings,
from what methodological principle does it follow that religion—or, for that
matter, anything else—must be universal if it is to be studied
comparatively? The fact that hunting economies, uni-lateral descent groups,
or string figures do not have a universal distribution has not prevented us
from studying them comparatively. Does the study of religion become any
the less significant or fascinating—indeed, it would be even more
fascinating—if in terms of a consensual ostensive definition it were
discovered that one, or seven, or sixteen societies did not possess religion?
If it indeed be the case that Theravada Buddhism is atheistic and that, by a
theistic definition of religion, it is not therefore a religion, why can we not
face, rather than shrink, from this consequence? Having combatted the
notion that ‘we’ have religion (which is good) and ‘they’ have supersitition
(which is ‘bad), why should we be dismayed if it be discovered that society
x does not have ‘religion’, as we have defined that term? For the premise
‘no religion’ does not entail the conclusion ‘therefore superstition’—nor,
incidentally, does it entail the conclusion ‘therefore no social integration’,
unless of course religion is defined as anything which makes for
integration. It may rather entail the conclusion ‘therefore science’ or
‘therefore philosophy’. Or it may entail no conclusion and, instead,
stimulate some research. In short, once we free the word ‘religion’ from all
value judgments, there is reason neither for dismay nor for elation
concerning the empirical distribution of religion attendant upon our
definition. With respect to Theravada Buddhism, then, what loss to science
would have ensued if Durkheim had decided that, as he interpreted it, it was
atheistic, and therefore not a religion? I can see only gain. First, it would
have stimulated fieldwork in these apparently anomalous Buddhist societies
and, second, we would have been spared the confusion created by the
consequent real and functional definitions of religion which were
substituted for the earlier substantive or structural definitions.
Real definitions, which stipulate the ‘essential nature’ of some
phenomenon are, as I have already argued, necessarily vague and almost
always nonempirical. What, for example, does Durkheim’s ‘sacred’—which
he stipulates as the essential nature of religion—really mean? How useful is
it, not in religious or poetic, but in scientific discourse? It is much too vague
to be taken as a primitive term in a definitional chain, and it is useless to
define it by equally vague terms such as ‘holy’ or ‘set apart’. But if such
real definitions are unsatisfactory when the phenomenal referent of the
definiendum is universally acknowledged, they are virtually useless when,
as in this case, it is the phenomenal referent which is precisely at issue. If
there is no agreement about what it is that is being defined, how can we
agree on its essential nature? Durkheim, to be sure, circumvented this
problem by arguing that the sacred is whatever it is that a society deems to
be sacred. But even if it were to be granted that one obscurity can achieve
clarity by the substitution of another, real definitions of this type—like
functional definitions to which I now wish to turn—escape the trap of
overly narrow designata only to fall into the trap of overly broad ones.
Most functional definitions of religion are essentially a subclass of real
definitions in which functional variables (the promotion of solidarity, and
the like) are stipulated as the essential nature of religion. But whether the
essential nature consists of a qualitative variable (such as the sacred) or a
functional variable (such as social solidarity), it is virtually impossible to
set any substantive boundary to religion and, thus, to distinguish it from
other sociocultural phenomena. Social solidarity, anxiety reduction,
confidence in unpredictable situations, and the like, are functions which
may be served by any or all cultural phenomena—Communism and
Catholicism, monotheism and monogamy, images and imperialism—and
unless religion is defined substantively, it would be impossible to delineate
its boundaries. Indeed, even when its substantive boundaries are limited,
some functional definitions impute to religion some of the functions of a
total sociocultural system.
It is obvious, then, that while a definition cannot take the place of
inquiry, in the absence of definitions there can be no inquiry—for it is the
definition, either ostensive or nominal, which designates the phenomenon to
be investigated. Thus when Evans-Pritchard writes that ‘objectivity’ in
studies of religion requires that “we build up general conclusions from
particular ones” (1954:9), this caution is certainly desirable for discovering
empirical generalizations or for testing hypotheses. But when he tells us
that “one must not ask ‘what is religion?’ but what are the main features of,
let us say, the religion of one Melanesian people …” which, when
compared with findings among other Melanesian peoples, will lead to
generalizations about Melanesian religion, he is prescribing a strategy
which, beginning with the study of that one Melanesian people, cannot get
started. For unless he knows, ostensively, what religion is, how can our
anthropologist in his Melanesian society know which, among a possible n,
observations constitute observations of religious phenomena, rather than of
some other phenomenal class, kinship, for example, or politics?
Indeed, when the term ‘religion’ is given no explicit ostensive definition,
the observer, perforce, employs an implicit one. Thus, Durkheim warns that
in defining religion we must be careful not to proceed from our “prejudices,
passions, or habits” (1954:24). Rather, “… it is from the reality itself which
we are going to define” (ibid.). Since any scientist—or, for that matter, any
reasonable man—prefers ‘reality’ to ‘prejudice’, we happily follow his lead
and, together with him, “… set ourselves before this reality” (ibid.). But
since, Durkheim tells us, “religion cannot be defined except by the
characteristics which are found wherever religion itself is found,” we must
“… consider the various religions in their concrete reality, and attempt to
disengage that which they have in common” (ibid.). Now, the very
statement of this strategy raises an obvious question. Unless we already
know, by definition, what religion is, how can we know which ‘concrete
reality’ we are to ‘consider’? Only if religion has already been defined can
we perform either this initial operation or the subsequent one of
disengaging those elements which are shared by all religions.
In sum, any comparative study of religion requires, as an operation
antecedent to inquiry, an ostensive or substantive definition that stipulates
unambiguously those phenomenal variables which are designated by the
term. This ostensive definition will, at the same time, be a nominal
definition in that some of its designata will, to other scholars, appear to be
arbitrary. This, then, does not remove ‘religion’ from the arena of
definitional controversy; but it does remove it from the context of fruitless
controversy over what religion ‘really is’ to the context of the formulation
of empirically testable hypotheses which, in anthropology, means
hypotheses susceptible to cross-cultural testing.
But this criterion of cross-cultural applicability does not entail, as I have
argued above, universality. Since ‘religion’ is a term with historically rooted
meanings, a definition must satisfy not only the criterion of cross-cultural
applicability but also the criterion of intracultural intuitivity; at the least, it
should not be counter-intuitive. For me, therefore, any definition of
‘religion’ which does not include, as a key variable, the belief in
superhuman—I won’t muddy the metaphysical waters with
‘supernatural’—beings who have power to help or harm man, is counter-
intuitive. Indeed, if anthropological consensus were to exclude such beliefs
from the set of variables which is necessarily designated by ‘religion’, an
explanation for these beliefs would surely continue to elicit our research
energies.
Even if it were the case that Theravada Buddhism postulates no such
beings, I find it strange indeed, given their all-but-universal distribution at
every level of cultural development, that Durkheim—on the basis of this
one case—should have excluded such beliefs from a definition of religion,
and stranger still that others should have followed his lead. But this
anomaly aside, is it the case that Buddhism contains no belief in
superhuman beings? (Let us, for the sake of brevity, refer to these beings as
‘gods’.) It is true, of course, that Buddhism contains no belief in a creator
god; but creation is but one possible attribute of godhood, one which—I
suspect—looms not too large in the minds of believers. If gods are
important for their believers because—as I would insist is the case—they
possess power greater than man s, including the power to assist man in, or
prevent him from, attaining mundane and/or supermundane goals, even
Theravada Buddhism—Mahayana is clearly not at issue here—most
certainly contains such beliefs. With respect to supermundane goals, the
Buddha is certainly a superhuman being. Unlike ordinary humans, he
himself acquired the power to attain Enlightenment and, hence
Buddhahood. Moreover, he showed others the means for its attainment.
Without his teachings, natural man could not, unassisted, have discovered
the way to Enlightenment and to final Release.
The soteriological attributes of the Buddha are, to be sure, different from
those of the Judaeo-Christian-Islamic God. Whereas the latter is living, the
former is dead; whereas the latter is engaged in a continuous and active
process of salvation, the former had engaged in only one active ministry of
salvation. But—with the exception of Calvinism—the soteriological
consequences are the same. For the Buddhist and the Western religionist
alike the Way to salvation was revealed by a superhuman being, and
salvation can be attained only if one follows this revealed Way. The fact
that in one case compliance with the Way leads directly to the ultimate goal
because of the very nature of the world; and, in the other case, compliance
leads to the goal only after divine intercession, should not obscure the basic
similarity: in both cases man is dependent for his salvation upon the
revelation of a superhuman being. (Indeed, there is reason to believe—I am
now analyzing field data collected in a Burmese village which suggest that
this might be the case—that Buddhist worship is not merely an expression
of reverence and homage to the One who has revealed the Way, but is also a
petition for His saving intercession.)
But superhuman beings generally have the power to assist (or hinder)
man’s attempts to attain mundane as well as supermundane goals, and when
it is asserted that Buddhism postulates no such beings, we must ask to
which Buddhism this assertion has reference. Even the Buddhism of the
Pali canon does not deny the existence of a wide range of superhuman
beings who intervene, for good and for ill, in human affairs; it merely
denies that they can influence one’s salvation. More important, in
contemporary Theravada countries, the Buddha himself—or, according to
more sophisticated believers, his power—is believed to protect people from
harm. Thus Burmese peasants recite Buddhist spells and perform rites
before certain Buddha images which have the power to protect them from
harm, to cure snake bites, and the like. And Buddhist monks chant passages
from Scripture in the presence of the congregation which, it is believed, can
bring a wide variety of worldly benefits.
There are, to be sure, atheistic Buddhist philosophies—as there are
atheistic Hindu philosophies—but it is certainly a strange spectacle when
anthropologists, of all people, confuse the teachings of a philosophical
school with the beliefs and behavior of a religious community. And if—on
some strange methodological grounds—the teachings of the philosophical
schools, rather than the beliefs and behavior of the people, were to be
designated as the normative religion of a society, then the numerous gods
and demons to be found in the Pali canon—and in the world-view of most
Theravadists, including the monastic virtuosos—find more parallels in other
societies than the beliefs held by the numerically small philosophical
schools.
Finally—and what is perhaps even more important from an
anthropological point of view—the Pali canon is only one source for the
world-view of Buddhist societies. Indeed, I know of no society in which
Buddhism represents the exclusive belief system of a people. On the
contrary, it is always to be found together with another system with which it
maintains an important division of labor. Whereas Buddhism (restricting
this term, now, to Canonical Buddhism) is concerned with supermundane
goals—rebirth in a better human existence, in a celestial abode of gods, or
final Release—the other system is concerned with wordly goals: the
growing of crops, protection from illness, guarding of the village, etc.,
which are the domain of numerous superhuman beings. These are the nats
of Burma, the phi of Laos and Thailand, the neak ta of Cambodia, etc.
Although the Burmese, for example, distinguish sharply between Buddhism
and nat worship, and although it is undoubtedly true—as most scholars
argue—that these nonBuddhist belief systems represent the pre-Buddhist
religions of these Theravada societies, the important consideration for our
present discussion is that these beliefs, despite the long history of Buddhism
in these countries, persist with undiminished strength, continuing to inform
the world-view of these Buddhist societies and to stimulate important and
extensive ritual societies and to stimulate important and extensive ritual
activity. Hence, even if Theravada Buddhism were absolutely atheistic, it
cannot be denied that Theravada Buddhists adhere to another belief system
which is theistic to its core; and if it were to be argued that atheistic
Buddhism—by some other criteria—is a religion and that, therefore, the
belief in superhuman beings is not a necessary characteristic of ‘religion’, it
would still be the case that the belief in superhuman beings and in their
power to aid or harm man is a central feature in the belief systems of all
traditional societies.
But Theravada Asia provides only one example of the tenacity of such
beliefs. Confucianist China provides what is, perhaps, a better example. If
Theravada Buddhism is somewhat ambiguous concerning the existence and
behavior of superhuman beings, Confucianism is much less ambiguous.
Although the latter does not explicitly deny the existence of such beings, it
certainly ignores their role in human affairs. It is more than interesting to
note, therefore, that when Mahayana Buddhism was introduced into China,
it was precisely its gods (including the Boddhisatvas), demons, heavens,
and hells that, according to many scholars, accounted for its dramatic
conquest of China.
To summarize, I would argue that the belief in superhuman beings and in
their power to assist or to harm man approaches universal distribution, and
this belief—I would insist—is the core variable which ought to be
designated by any definition of religion. Recently Horton (1960) and Goody
(1962) have reached the same conclusion.
Although the belief in the existence of superhuman beings is the core
religious variable, it does not follow—as some scholars have argued—that
religious, in contrast to magical, behavior is necessarily other-worldly in
orientation, or that, if it is otherworldly, its orientation is ‘spiritual’. The
beliefs in superhuman beings, other-worldliness, and spiritual values vary
independently. Thus, ancient Judaism, despite its obsession with God’s will,
was essentially this-worldly in orientation. Catholicism, with all its other-
worldly orientation is, with certain kinds of Hinduism, the most
‘materialistic’ of the higher religions. Confucianism, intensely this-worldly,
is yet concerned almost exclusively with such ‘spiritual’ values as filial
piety, etc. In short, superhuman beings may be conceived as primarily
means or as ends. Where values are worldly, these beings may be viewed as
important agents for the attainment and/or frustration of worldly goals,
either ‘material’ or spiritual’. Where values are materialistic, superhuman
beings may be viewed as important agents for the attainment of material
goals, either in this or in an after life. Where values are otherworldly,
mystical union with superhuman beings may be viewed as an all-consuming
goal; and so on.
Although the differentiating characteristic of religion is the belief in
superhuman beings, it does not follow, moreover, that these beings are
necessarily objects of ultimate concern. Again, it depends on whether they
are viewed as means or as ends. For those individuals whom Weber has
termed ‘religiously musical’ (Gerth & Mills 1946:287), or whom Radin
(1957:9) has termed ‘the truly religious’, superhuman beings are of ultimate
concern. For the rest, however, superhuman beings are rarely of ultimate
concern, although the ends for which their assistance is sought may be.
Hence, though their benevolent ancestral spirits are not of great concern to
the Ifaluk, restoration of health—for which these spirits are instrumental—
most certainly is. Similarly, while the Buddha may not be of ultimate
concern to a typical Burmese peasant, the escape from suffering—for which
He is instrumental—can certainly be so designated.
Conversely, while religious beliefs are not always of ultimate concern,
nonreligious beliefs sometimes are. This raises a final unwar-ranted
conclusion, viz. that religion uniquely refers to the ‘sacred’, while secular
concerns are necessarily ‘profane’. Thus, if ‘sacred’ refers to objects and
beliefs of ultimate concern, and ‘profane’ to those of ordinary concern,
religious and secular beliefs alike may have reference either to sacred or to
profane phenomena. For the members of Kiryat Yedidim, an Israeli kibbutz,
the triumph of the proletariat, following social revolution, and the ultimate
classless society in which universal brotherhood, based on loving kindness,
will replace parochial otherhood, based on competitive hostility, constitutes
their sacred belief system. But, by definition, it is not a religious belief
system, since it has no reference to—indeed, it denies the existence of—
superhuman beings.
Similarly, if communism, baseball, or the stock market are of ultimate
concern to some society, or to one of its constituent social groups, they are,
by definition, sacred. But beliefs concerning communism, baseball, or the
stockmarket are not, by definition, religious beliefs, because they have no
reference to superhuman beings. They may, of course, serve many of the
functions served by religious beliefs; and they are, therefore, members of
the same functional class. Since, however, they are substantively dissimilar,
it would be as misleading to designate them by the same term as it would be
to designate music and sex by the same term because they both provide
sensual pleasure. (modern American society presents an excellent example
of the competition of sports, patriotism, sex, and God for the title, perhaps
not exclusively, of the sacred’. Indeed, if the dictum of Miss Jane Russell is
taken seriously—God, she informs us, is a “livin’ doll”—I would guess
that, whichever wins, God is bound to lose.)
A Definition of ‘Religion’
On the assumption that religion is a cultural institution, and on the further
assumption that all institutions—though not all of their features—are
instrumental means for the satisfaction of needs, I shall define ‘religion’ as
‘an institution consisting of culturally patterned interaction with culturally
postulated superhuman beings’. I should like to examine these variables
separately.

Institution. This term implies, of course, that whatever phenomena we

might wish to designate by ‘religion’, religion is an attribute of social
groups, comprising a component part of their cultural heritage; and that its
component features are acquired by means of the same enculturation
processes as the other variables of a cultural heritage are acquired. This
means that the variables constituting a religious system have the same
ontological status as those of other cultural systems: its beliefs are
normative, its rituals collective, its values prescriptive. This, I take it, is
what Durkheim (1954:44) had in mind in insisting that there can be no
religion without a church. (It means, too, as I shall observe in a later
section, that religion has the same methodological status as other cultural
systems; i.e. religious variables are to be explained by the same explanatory
schemata—historical, structural, functional, and causal—as those by which
other cultural variables are explained.)

Interaction. This term refers to two distinct, though related, types of

activity. First, it refers to activities which are believed to carry out, embody,
or to be consistent with the will or desire of superhuman beings or powers.
These activities reflect the putative value system of these superhuman
beings and, presumably, they constitute part—but only part—of the actors’
value system. These activities may be viewed as desirable in themselves
and/or as means for obtaining the assistance of superhuman beings or for
protection against their wrath. Second, it refers to activities which are
believed to influence superhuman beings to satisfy the needs of the actors.
These two types of activity may overlap, but their range is never
coterminous. Where they do overlap, the acfion in the overlapping sphere
is, in large measure, symbolic; that is, it consists in behavior whose
meaning, cross-culturally viewed, is obscure and/or arbitrary; and whose
efficacy, scientifically viewed, is not susceptible of ordinary scientific
‘proof. These symbolic, but definitely instrumental, activities constitute, of
course, a ritual, or symbolic action, system. Unlike private rituals, such as
those found in an obsessive-compulsive neurosis, religious rituals are
culturally patterned; i.e. both the activities and their meaning are shared by
the members of a social group by virtue of their acquisition from a shared
cultural heritage.

Superhuman beings. These refer to any beings believed to possess power

greater than man, who can work good and/or evil on man, and whose
relationships with man can, to some degree, be influenced by the two types
of activity described in the previous section. The belief of any religious
actor in the existence of these beings and his knowledge concerning their
attributes are derived from and sanctioned by the cultural heritage of his
social group. To that extent—and regardless of the objective existence of
these beings, or of personal experiences which are interpreted as encounters
with them—their existence is culturally postulated. Beliefs concerning the
existence and attributes of these beings, and of the efficacy of certain types
of behavior (ritual, for example) in influencing their relations with man,
constitute a belief system.

This brief explication of our definition of ‘religion’ indicates that, viewed

systemically, religion can be differentiated from other culturally constituted
institutions by virtue only of its reference to superhuman beings. All
institutions consist of belief systems, i.e. an enduring organization of
cognitions about one or more aspects of the universe; action systems, an
enduring organization of behavior patterns designed to attain ends for the
satisfaction of needs; and value systems, an enduring organization of
principles by which behavior can be judged on some scale of merit.
Religion differs from other institutions in that its three component systems
have reference to superhuman beings.
Having defined ‘religion’, our next task is to examine the types of
explanation that have been offered to account for its existence. First,
however, we must answer some elementary questions concerning the nature
of anthropological explanation.
Explanation in Social Anthropology
What do anthropologists attempt to explain? Of what do explanations
consist? How do these explanations differ from each other? Once we
penetrate beneath our jargon, it appears that always the phenomenon to be
explained is (a) the existence of some social or cultural variable, and (b) the
variability which it exhibits in a cross-cultural distribution. These
statements of course are really one, because the variable ‘exists’ in the
range of values which it can assume. If a theory purports to explain the
existence of religion, but its concepts are so general or so vague that it
cannot explain the variability exhibited by its empirical instances, it is
disqualified as a scientific, i.e. a testable, theory.
‘Existence’ is an ambiguous term. In asking for an explanation for the
existence of religion, we might be asking how it came to exist in the first
place—this is the question of religious origins—or how it is that it exists
(i.e. has persisted) in some ethnographic present. Since a testable, i.e.
scientific, theory of religious origins will probably always elude our
explanatory net, this chapter will be concerned with the persistence, not the
origin, of religion.
In all of our explanations, to answer the second question, we stipulate a
condition or a set of conditions in whose absence the variable to be
explained would not exist. Now, to say that any sociocultural variable—
religion, for example—‘exists’ is, in the last analysis, to say that in some
society—or, in one of its constituent social groups—a proposition is
affirmed, a norm complied with, a custom performed, a role practiced, a
spirit feared, etc. In short, the ‘existence’ of a sociocultural variable means
that in any sense of ‘behavior’—cognitive, affective, or motor—there
occurs some behavior in which, or by which, the variable in question is
instanced. Hence, a theory of the ‘existence’ of religion must ultimately be
capable of explaining religious ‘behavior’.
In general, theories of the existence—in the sense of persistence—of
sociocultural variables are cast in four explanatory modes: historical (in the
documentary, not the speculative, sense), structural, causal, and functional.
When analyzed, the first two can be reduced to either the third or the fourth.
Thus historical explanations are either no explanation at all, or they are
causal explanations. Surely, the mere listing of a series of events which are
antecedent to the appearance of the variable in question does not constitute
explanation—unless it can be shown that one or more of these events was a
condition, either necessary or sufficient, for its appearance. If this can be
demonstrated, the explanation is based on a causal theory, the fact of its
having originated in the past being incidental to the theoretical aim of
explaining a certain type of social or cultural innovation.
The key term here is ‘innovation’. Although a causal explanation of the
historical type may account for the existence of some sociocultural variable
during the period in which it made its appearance, it is not a sufficient
explanation for its persistence into a later period—no more than a genetic
explanation for the birth of an organism is a sufficient explanation for its
persistence. Thus while a historicogenetic explanation is necessary to
account for Burma’s adoption of Buddhism, it cannot account for its
persistence nine hundred years later. Alternatively, historical data explain
why it is that the Burmese, if they practice any religion at all, practice
Buddhism rather than, for example, Christianity; but they do not explain
why they practice any religion at all—and, therefore, they do not explain
why they practice Buddhism.
Structural explanations, too, can be reduced either to causal or to
functional modes. Those structural accounts which delineate the
configuration in, or relationships among, a set of sociocultural variables are
essentially descriptive rather than explanatory—unless of course some
theory, causal or functional, is offered to explain the configuration.
Structural explanations which purport to explain some variable by means of
a structural ‘principle’—such as the principle of the unity of the sibling
group—are either verbal labels which at best order a set of data according
to a heuristic scheme; or they are phenomenological principles’ of the
actors (cognitive maps), in which case they comprise a cognitive subset in a
set of causal variables. Similarly, structural explanations which stipulate
some variable as a “structural requirement’ of a system, on the one hand, or
as a ‘structural implication’, on the other, can be shown to be either causal
or functional respectively.
This brings us, then, to causal and functional explanations. Causal
explanations attempt to account for some sociocultural variable by
reference to some antecedent conditions—its ‘cause’. Functional
explanations account for the variable by reference to some consequent
condition—its ‘function’. (For a detailed analysis of the logic of causal and
functional explanation, cf. Spiro 1963).
How, then, is the existence of religion to be explained, causally or
functionally? The answer depends, I believe, on which aspect of a religion
is to be explained.
Typically, explanations for the existence of religion have been addressed
to one or both of two questions, (a) On what grounds are religious
propositions believed to be true? That is, what are the grounds for the belief
that superhuman beings with such-and-such characteristics exist, and that
ritual is efficacious in influencing their behavior? (b) What is the
explanation for the practice of religion? That is, what is the basis for belief
in superhuman beings, and for the performance of religious rituals? These
questions, though clearly related—religious practice presupposes religious
cognitions—are yet distinct; and they probably require different types of
explanation.

The ‘Truth’ of Religious Beliefs

Every religious system consists, in the first instance, of a cognitive system;
i.e. it consists of a set of explicit and implicit propositions concerning the
superhuman world and of man’s relationship to it, which it claims to be
true. These include beliefs in superhuman beings of various kinds, of rituals
of a wide variety, of existences—both prior and subsequent to the present
existence—and the like. To the extent that documentary evidence is
available, it is possible to discover a testable explanation for their origin and
their variability. Since, for most of man’s religions, however, such evidence
is lacking, explanations are necessarily speculative. I shall be concerned,
therefore, with explanations for their persistence.
This cognitive system, or parts of it, is of course acquired by the
members of a group, and, on the individual level, it becomes a ‘culturally
constituted belief system’. It is a ‘belief system because the propositions are
believed to be true; and it is ‘culturally constituted’ because the
propositions are acquired from this culturally provided religious system.
But the latter fact, surely, is not a sufficient basis for the belief that these
propositions are true. Children are taught about many things which, when
they grow up—often, before they grow up—they discard as so much
nonsense. The fact that my personal belief system is acquired from my
society explains why I might believe in the existence of the Lord Krishna
rather than that of the Virgin Mary; it does not explain the grounds on
which I believe in the existence of superhuman beings of any kind, whether
Krishna or the Virgin.
The notion of ‘need’ fares no better as an explanation. A need, in the
sense of desire, may provide the motivational basis for the acquisition of a
taught belief, but it cannot establish its truth. Similarly, a need, in the sense
of a functional requirement of society, may explain the necessity for some
kind of religious proposition(s), but—even if this need is recognized and its
satisfaction is intended—it does not explain why the proposition is believed
to be true.
Most theorists seem to agree that religious statements are believed to be
true because religious actors have had social experiences which,
corresponding to these beliefs, provide them with face validity. Thus
Durkheim and Freud, agreeing that the cognitive roots of religious belief
are to be found in social experience, disagree only about the structural
context of the experience. Durkheim (1954) argues that the two essential
attributes of the gods—they are beings more powerful than man, upon
whom man can depend—are the essential attributes of society; it is in
society that man experiences these attributes. These attributes are
personified in superhuman beings, or imputed to extra-social powers,
because of highly affective collective experiences in which the physical
symbols of the group are taken to be symbolic of the one or the other. Freud
(1928), emphasizing man’s helplessness in a terrifying world, stresses the
importance of personifying these terrifying forces so that, on a human
analogy, they can be controlled. These personifications—the gods—reflect
the child’s experience with an all-powerful human being, his father.
In both Freud and Durkheim, then, society is the cause for the fixation of
religious belief. It is ironical to observe, however, that it is Freud rather than
Durkheim who anchors this experience within a specified structural unit,
the family. In general, I believe that Freud has the better case. To be sure,
his ethnocentric conception of patriarchal fathers, and of gods reflecting this
conception of ‘father’, is entirely inadequate for comparative analysis: but
the general theory which can be generated from this ethnocentric model is
both adequate and, I believe, essentially correct. The theory, briefly, states
that it is in the context of the family that the child experiences powerful
beings, both benevolent and malevolent, who—by various means which are
learned in the socialization process—can sometimes be induced to accede
to his desires. These experiences provide the basic ingredients for his
personal projective system (Kardiner 1945) which, if it corresponds
(structurally, not substantively) to his taught beliefs, constitutes the
cognitive and perceptual set for the acceptance of these beliefs. Having had
personal experience with ‘superhuman beings’ and with the efficacy of
‘ritual’, the taught beliefs reenforce, and are reenforced by, his own
projective system (Spiro 1953).
This theory is superior to Durkheim’s, first, in its ability to explain these
latter two nuclear religious variables. It is difficult to see how they can be
deduced from Durkheim’s theory. More important, even if they were
deducible, it is difficult to see how Durkheim’s theory can explain their
cross-cultural variability. Since the antecedent condition—a power greater
than man upon which man can rely—is a constant, how can it explain the
consequent condition—religious belief—which is a variable? Hence,
although the general theory may be true, there is a serious question
concerning what empirical operations, if any, would permit us to decide
whether it is true or false. The Freudian-derived theory, on the other hand,
is capable of explaining cross-cultural differences and, therefore, it can be
tested empirically. If personal projective systems, which form the basis for
religious belief, are developed in early childhood experiences, it can be
deduced that differences in religious beliefs will vary systematically with
differences in family (including socialization) systems which structure these
experiences. A number of anthropological field studies have been able to
test—and have confirmed—this conclusion. And one need not go so far
astray as personality-and-culture studies to find the evidence. In his
illuminating analysis of Tallensi religion, Fortes (1959:78) concludes that,
“All the concepts and beliefs we have examined are religious extrapolations
of the experiences generated in the relationships between parents and
children.”
As important as they are, case studies do not constitute rigorous proof of
theories. The cross-cultural method in which large samples can be used for
statistical testing of hypotheses is more rigorous. A fairly large number of
hypotheses, predicting religious variables—the character of supernatural
beings, the means (performance of rituals or compliance with norms) which
are believed to influence them, the conception of ritual (coercive or
propitiatory) and the like—from child-training variables have been tested
by this method, and many of them have been confirmed. (These studies are
summarized in Whiting 1961.)
Despite the differences between Freud and Durkheim, both propose
causal explanations of the credibility of religious cognitions, in which
society, as cause, produces a religious (cognitive) effect by means of
psychological processes—in one case, in the feeling of dependency; in the
other, in the feeling of dependency combined with the personal projection
of nuclear experiences. For both theorists the independent, sociological,
variable may be said to ‘cause’ the dependent, religious, variable by means
of a set of intervening, psychological, variables.
Before passing to the next section it is necessary to counter one
assumption which is often linked to this type of explanation, but which is
not entailed by it. From the hypothesis that the antecedent condition for
belief in superhuman beings is to be found in specified sociological
variables, it does not follow that for the believer these beings are identical
with, have reference to, or are symbolic of these variables. To put it bluntly,
the fact that conceptions of God have their roots in society does not mean
that for the believer society is God, or that God is merely a symbol of
society, or that society is the true object of religious worship. Freud never
said this, nor—despite some claims to the contrary—did Durkheim. Hence,
it is no refutation of Durkheim’s sociological hypothesis—as Horton
(1960:204) believes it to be—that for the Kalahari—and, I would add, for
every other people—the statement “‘I believe in God’” does not imply, “‘I
subscribe to the system of structural symbolism of which this belief
statement is part.’”2
2. In any case, Durkheim’s sociology of religion is unconcerned with structural symbolism. It is
concerned with society as a collectivity, not as a configuration of structural units; with collective
representations, not with social structure.
Similarly it is not only among the Kalahari that a person who uses belief
or ritual “merely to make a statement about social relations or about his
own structural alignment” is viewed as one who “‘does not really believe’”
(ibid.:203). Indeed the contrary notion is so absurd that it is difficult to
believe that it could be proposed by anyone who has personally observed a
Micronesian exorcising a malevolent ghost, a Catholic penitent crawling on
hands and knees to worship at the shrine of Our Lady of Guadalupe, an
Orthodox Jew beating his breast on Yom Kippur to atone for his sins, a
Burmese spirit-medium dancing in a trance before a nat image. And yet,
Leach, writing of Kachin spirits, insists (1954:182) that “the various nats of
Kachin religious ideology are, in the last analysis nothing more than ways
of describing the formal relationships that exist between real persons and
real groups in ordinary Kachin society” (italics mine). For Leach, then, not
only are religious beliefs derived from social structure, but the only referent
of these beliefs, even for the believer, is social structure. Since Leach has, I
am sure, observed as many manifestations of religious belief as I have, what
are we to make of this extraordinary statement? I should like to suggest that
this assertion reflects a confusion between the practice of religion and its
manipulation; and although the former may be of as much interest to the
student of religion as the latter, not to recognize this distinction is to sow
confusion.
Leach, it will be recalled, interprets Kachin religion as almost exclusively
an instrument in the political struggle for power and prestige. That theology,
myth, or ritual may be manipulated for prestige and power, and that the
latter drives may provide motivational bases for their persistence are
documented facts of history and ethnology. But the manipulation of religion
for political ends tells us more about politics than about religion. In the
former, religion is used as a means, in the latter religious means are used,
for the attainment of certain ends. (Indeed, they may both be instrumental
for the same end.) This is an essential distinction. The differential
characteristic of religious, compared with other types of instrumental,
behavior consists in an attempt to enlist the assistance, or to execute the
will, of superhuman beings. Indeed, by what other criterion could religious
be distinguished from nonreligious behavior? Surely not in terms of ends:
with the exception of mysticism (confined to religious virtuosos) the range
of mundane ends for which religion, cross-culturally viewed, is conceived
to be instrumental is as broad as the range of all human ends. My argument,
then, is not that political power is disqualified as a religious end, but that
any attempt to achieve this end by means which do not entail a belief in the
existence of superhuman beings is disqualified as religious means. A
Kachin headman may attempt to manipulate myths and nats in order to
validate his, or his clan’s claim to power and authority; but this political
behavior is to be distinguished from his religious behavior, which consists
in his belief in the existence of the nats, and in his propitiation of them,
both at local shrines and during manaus. Indeed, only because Kachins do
believe that the verbal symbol ‘nat’ has reference to an existential being—
and is not merely a social structural symbol—is it possible to manipulate
this belief for political ends.

The Practice of Religious Belief

The religious actor not only believes in the truth of propositions about
superhuman beings, but he also believes in these beings—they are objects
of ‘concern’: he trusts in God, he fears and hates Satan. Similarly, he not
only believes in the efficacy of ritual, but he performs rituals. Explanations
for religion, then, are addressed not only to the truth of religious
propositions but also—and more frequently—to certain practices. In order
to explain the practice of religion, we must be able to explain the practice of
any sociocultural variable.3
3. The notion that religion necessarily eludes the net of naturalistic explanation, though implicit in
certain recent anthropological writing, is beginning to find explicit expression. Thus, Turner
(1962:92) claims that “one has to consider religious phenomena in terms of religious ideas and
doctrines, and not only, or principally, in terms of disciplines which have arisen in connection with
the study of secular institutions and processes…. Religion is not determined by anything other than
itself, though the religious find expression in sensory phenomena. … We must be prepared to accept
the fruits of simple wisdom with gratitude and not try to reduce them to their chemical constituents,
thereby destroying their essential quality as fruits, and their virtue as food.”
To say that religion is determined by religion is surely as meaningless as to say that tables are
determined by tables, or that social structure is determined by social structure. But even if this
statement is an ellipsis for a meaningful one, I would have thought that the determinants of religion
are to be established empirically rather than by verbal assertion. I would have thought, too, that to
‘reduce’ religion to its ‘chemical constituents’—which, I take it, refers to the discovery of its social
and psychological bases—is precisely the task of the student of culture. I would not have thought that
for the religionist—and the anthropologist, qua anthropologist, is of course neither a religionist nor
an anti-religionist—the ‘reduction’ of religion to its social and psychological ‘constituents’ destroys
its ‘essential quality’, any more than the ‘reduction’ of a Brandenburg concerto to its physical
‘constituents’ destroys its ‘essential quality’ for the lover of Bach.
All human behavior, except for reflexive behavior, is purposive; i.e. it is
instigated by the intention of satisfying some need. If a given response is in
fact instrumental for the satisfaction of the need, this ‘reinforcement’ of the
response ensures its persistence—it becomes an instance of a behavior
pattern. The motivational basis for the practice of a behavior pattern, then,
is not merely the intention of satisfying a need, but the expectation that its
performance will in fact achieve this end. Institutional behavior, including
religious behavior, consists in the practice of repeated instances of
culturally constituted behavior patterns—or customs. Like other behavior
patterns they persist as long as they are practiced; and they are practiced
because they satisfy, or are believed to satisfy, their instigating needs. If this
is so, an explanation for the practice of religion must be sought in the set of
needs whose expected satisfaction motivates religious belief and the
performance of religious ritual.4
4. There is, of course, no convincing evidence for the existence of a distinctively ‘religious need’.
That the belief in superhuman beings corresponds to and satisfies a ‘need’ for a belief in them is an
unfounded instinctivist assumption. Nor is there any evidence for the assumption that the
motivational basis for religious action, or the affect which it arouses, are unique. The meager
evidence from religious psychology suggests that any drive or affect connected with religious
behavior is also found in such activities as science, warfare, sex, art, politics, and others. The
‘religious thrill’ which Lowie (1924) and others have pointed to as a differentiating characteristic of
religion is still to be documented. A clinical psychologist, commenting on still another set of data
remarks: “Some of our tests seem able to tap fairly deep levels of personality functioning, and yet we
rarely encounter a clearly religious response to our Rorschach and Thematic Apperception tests….
For the psychology of religion this means that the clinical psychologist will not readily be able to
furnish new data” (Pruyser 1960:122).

Needs

As a concept in the social sciences, ‘need’ has been borrowed from two
sources: biology and psychology. Its ambiguity as a social science concept
stems from a confusion in its two possible meanings. In biology ‘need’
refers to what might be termed a ‘want’, i.e. to some requirement which
must be satisfied if an organism is to survive. In psychology, on the other
hand, it refers to what might be termed a ‘desire’, i.e. a wish to satisfy some
felt drive by the attainment of some goal. These two meanings may, of
course, overlap. Thus water satisfies an organic want and it may also be the
object of desire. Just as frequently, however, they do not overlap. The
circulation of the blood is a want, but—for most animals, at any rate—it is
not the object of desire. In sociological and, especially, in functionalist
discourse, much confusion has resulted from not distinguishing these two
meanings of ‘need’ when it is applied to society. In this paper I shall, when
clarity does not suffer, use the generic ‘need’ to refer both to sociological
wants and to psychological desires. Otherwise, I shall use ‘sociological
want to refer to any functional requirement of society; and I shall use
‘desire’ in its motivational sense. I hasten to add that desires are not
necessarily ‘selfish’, oriented to the welfare of the self. The goal, by whose
attainment a drive is satisfied, may be—and, obviously, it often is—the
welfare of an entire group, or of one or more of its constituent members.
We may now return to our question. If the practice of religion is
instigated by the expectation of satisfying needs, by which set of needs—
desires or wants—is its practice to be explained? It should be perfectly
obvious that although behavior can satisfy both wants and desires, it is
motivated by desires, not by wants. Wants in themselves have no causal
properties. The absorption of moisture, for example, is a functional
requirement of plants; but this requirement cannot cause the rains to fall.
Human behavior, to be sure, is different from the growth of plants. A social
group may recognize the existence of some, at least, of its functional
requirements, and these recognized wants may constitute a set of stimulus
conditions which evoke responses for their satisfaction. Notice, however,
that it is not the functional requirement—even when it is recognized—
which evokes the response, but rather the wish to satisfy it. A functional
requirement of society becomes a stimulus for a response, i.e. it acquires
motivational value if, and only if (a) it is recognized, and (b) its satisfaction
becomes an object of desire. If the functional requirement of social
solidarity, for example, is not recognized, or, if recognized, it is not an
object of desire, or, although both recognized and desired, it is not the
desire whose intended satisfaction motivates the practice of the variable to
be explained, it cannot be used to explain behavior, even though need-
satisfaction may be one of its consequences. If social solidarity is a
consequence—an unintended consequence—of the practice of religion,
social solidarity is properly explained by reference to the religious behavior
by which it is achieved; but religion, surely, is improperly explained by
reference to social solidarity. An unintended consequence of behavior—
however important it may be—can hardly be its cause. If religious behavior
is to be explained by reference to those functions which it serves—and,
indeed, it must be—the functions must be those that are intended, not those
that are unintended (and probably unrecognized). We must, therefore,
remind ourselves of some elementary distinctions among functions.

Functions
I should like, first, to distinguish between ‘psychological’ and ‘sociological’
functions. The psychological functions of behavior consist in the
satisfaction of desires,5 its sociological functions consist in the satisfaction
of functional requirements.
5. That a function ‘psychological’ does not mean that the object of desire is psychological. The
object of desire may be political, meteorological, economic, nutritional, sexual—and all other goals
known to man. These goals are cathected because they satisfy some drive, acquired or innate. The
attainment of the goal reduces the drive or, alternatively, satisfies the desire. The satisfaction of the
desire—or, more realistically, the set of desires—whose intended (conscious or unconscious)
satisfaction instigates behavior is its ‘psychological’ function.
Second, I should like to distinguish between ‘manifest’ and ‘latent’
functions. In his now-classic analysis of functional explanation, Merton
(1957) distinguished between intended and recognized functions—which he
termed ‘manifest’—and unintended and unrecognized functions—which he
termed ‘latent’. Merton’s dichotomous classification can be shown to yield
a four-class functional typology. In addition to intended-recognized and
unintended-unrecognized functions, we can also distinguish intended-
unrecognized and unintended-recognized functions (Spiro 1961a). The
latter is a simple concept to grasp; social solidarity may be a recognized
function of religious ritual, for example, although the intention of satisfying
this functional requirement may not motivate its performance. An intended-
unrecognized function, however, seems paradoxical. Assuming that
intentions may be conscious as well as unconscious, this paradox is more
apparent than real; if a behavior pattern is unconsciously motivated—or,
more realistically, if its motivational set includes both conscious and
unconscious intentions—one of its functions, although intended, is
unrecognized.
The final distinction I should like to make is between real and apparent
functions. ‘Real functions’ are those which, in principle at least, can be
discovered by the anthropologist, whether or not they are recognized by the
actors. ‘Apparent functions’ are those which the actors attribute to the
sociocultural variable in question, but which cannot be confirmed by
scientific investigation.
With these distinctions in mind, we may now attempt to answer the
question with which we began this section. If institutional behavior in
general is motivated by the expectation of satisfying desires, to what extent
can religious behavior, specifically, be explained within this framework?
That is, what desires are satisfied by religion? Since this question remains
one of the unfinished tasks of empirical research, I can only make some
tentative suggestions. As I interpret the record, I would suggest that there
are at least three sets of desires which are satisfied by religion and which—
for lack of better terms—I shall call cognitive, substantive, and expressive.
The corresponding functions of religion can be called adjustive, adaptive,
and integrative.

Cognitive

I believe that it can be shown that everywhere man has a desire to know, to
understand, to find meaning;6 and I would suggest—although this is a
terribly old-fashioned nineteenth-century idea—that religious beliefs are
held, and are of ‘concern’, to religious actors because, in the absence of
competitive explanations, they satisfy this desire. Religious belief systems
provide the members of society with meaning and explanation for otherwise
meaningless and inexplicable phenomena.
6. The most striking evidence, on the simplest perceptual level, of this ‘need’ for meaning is
provided by the cross-cultural use of the Rorschach test. As Hallowell has observed (1956:476–88),
the most dramatic finding of cross-cultural Rorschach investigations is that at every level of
technological and cultural development in which this test has been administered, subjects have
attempted to offer ‘meaningful’ responses to what are, objectively, ‘meaningless’ ink-blots. The
insistence, even on the part of ‘primitive’ peoples, on finding meaning in what is for them an exotic
task—something concerning which many anthropologists had been skeptical—is certainly consistent
with the assumption concerning a universal need for meaning.
‘Meaning’, of course, is often used in two senses. It may be used in an
exclusively cognitive sense, as when one asks for the meaning of a natural
phenomenon, of a historical event, of a sociological fact. In this sense, it has
the connotation of ‘explanation’, as that word is typically used. But
‘meaning’ is also used in a semantic-affective sense, as when one asks for
the meaning of unequal life-fates, frustration, or death. The phenomena for
which religion provides meaning, in this second sense of ‘meaning’, have
been classified by Weber under the general rubric of ‘suffering’ (Gerth &
Mills 1946: ch. 11). The main function of the higher religions, he argues, is
to provide meaning for suffering (and some means to escape from or to
transcend it).
Although the range of phenomena for which religious beliefs provide
meaning in the first, explanatory, sense of meaning’, occupies a broad
spectrum, some structuralists hold to the peculiar notion that man’s
curiosity is so limited that religious explanations, regardless of their
ostensible meaning, are concerned almost exclusively with phenomena of
social structure. Again I should like to use an example from Leach—
although the example concerns magical rather than religious belief—
because, with his usual verve, he adopts what I would think to be an
extreme position.
In his highly critical evaluation of Frazer, Leach (1961) tells us that
Frazer (and Roth, too) is naïve in interpreting Australian explanations for
conception as in fact referring to conception—and, therefore, as reflecting
ignorance of physiological paternity. According to the ‘modern
interpretation’, their notions of conception are to be seen, not as biological,
but as sociological, statements. Let us examine the ethnographic facts, as
Leach (p. 376) quotes them from Roth. Among the Tully River Blacks,
A woman begets children because (a) she has been sitting over the fire on
which she has roasted a particular species of black bream, which must have
been given to her by the prospective father, (b) she has purposely gone a -
hunting and caught a certain kind of bull-frog, (c) some man may have told
her to be in an interesting condition, or (d) she may dream of having the
child put inside her.
Both Frazer and Roth agree—and, I may add, I agree with them—that
these statements are addressed to the problem to which they appear to be
addressed—the problem of conception; and they agree that from these
statements it may validly be deduced that the aborigines are ignorant of
physiological paternity, believing rather that conception is the result of four
kinds of ‘magical’ causation. Leach will have none of this. For him (ibid.) it
is not
a legitimate inference to assert that these Australian aborigines were ignorant of the connection
between copulation and pregnancy. The modern interpretation of the rituals described would be
that in this society the relationship between the woman’s child and the clansmen of the woman’s
husband stems from public recognition of the bonds of marriage, rather than from the fact of
cohabitation, which is a very normal state of affairs.

The logic of this ‘modern interpretation’ is certainly not evident to me.

Ignoring the fact that only two of the four explanatory beliefs have
reference to a male—so that they, at least, are hardly susceptible of this
modern interpretation—by what evidence or from what inference can it be
concluded that the other two statements mean what Leach claims that they
mean? Is this the interpretation which the aborigines place on these beliefs?
There is certainly no evidence for this assumption. Perhaps, then, this is the
meaning which they intended to convey, even though they did not do so
explicitly? But even if we were to grant that, for some strange reason,
aborigines prefer to express structural relationships by means of biological
symbolism, how do we know that this was their intention? Perhaps, then,
the symbolism is unconscious, and the structural meaning which Leach
claims for these beliefs, although intentional, is latent? This interpretation is
certainly congenial to other ‘modern interpretations’. Again, however, we
are hung up on the problem of evidence. From what ethnographic data, or
from what psychological theory of the unconscious, can this meaning be
inferred? If, then, there is no way of demonstrating that either the manifest
or the latent content of these symbols has reference to the structural
relationship between a woman’s children and the clansmen of her husband,
I am compelled to discard this interpretation as not only implausible but
false. I shall insist, instead, that the aborigines are indeed ignorant of
physiological paternity, and that the four statements quoted in Roth are in
fact proffered explanations for conception.

Substantive

The most obvious basis for religious behavior is the one which any
religious actor tells us about when we ask him—and, unlike some
anthropologists, I believe him. He believes in superhuman beings and he
performs religious ritual in order that he may satisfy what I am calling
substantive desires: desires for rain, nirvana, crops, heaven, victory in war,
recovery from illness, and countless others. Everywhere man’s mammalian
desires (those which can be satisfied by naturalistic goals) must be satisfied,
and in the absence of competing technologies which confer reasonable
confidence, religious techniques are believed to satisfy these desires.
Almost everywhere, moreover, the human awareness of the cessation of
existence and/or of the unsatisfactory character of existence, produces
anxiety concerning the persistence of existence (in some cases, it is desired;
in others, it is not desired), and in the absence of competing goals for the
reduction of anxiety, belief that one is successfully pursuing these religious
goals (heaven-like or nirvana-like states) serves to reduce this anxiety.
 Most, if not all, of these substantive desires, then, can be classified as
attempts to overcome or transcend suffering. The religious actor wishes to
overcome specific suffering—economic, political, physical, and the like;
and he wishes to transcend more general suffering induced by some
conception of life and the world as being evil, frustrating, sinful, and so on.
Religion, as Weber (1930) points out, not only provides an explanation for,
but it also promises redemption from, suffering. Religious techniques—
performance of ritual, compliance with morality, faith, meditation, etc.—are
the means by which this promise is felt to be fulfilled.
For the religious actor, if we can believe him, the expectation of realizing
this promise is the most important motivational basis for religous behavior;
the realization of this promise is its function. For him, it is an intended and
recognized function. Believing in its reality, he clings tenaciously to his
religious beliefs and practices—however irrational they may seem, and
however dysfunctional with respect to other ends their consequences may
be. From the anthropologist’s point of view—and this is what presents such
a knotty problem to many classical functionalists—these functions are
apparent; they are not real. Ritual cannot effect rainfall, prayer cannot cure
organic diseases, nirvana is a figment of the imagination, etc. It is this
seeming irrationality of religion and, therefore, the apparent—rather than
real—nature of its intended functions, that has given rise, I believe, to
misplaced emphases on the importance of its sociological functions. Thus,
despite Merton’s incisive analysis of functional theory, it is highly
questionable if the persistence of Hopi rain ceremonies is to be explained
by the social integration to which he (Merton) thinks their performance is
conducive (their real, but latent, functions), rather than by the
meteorological events to which Hopi think they are conducive (their
manifest, but apparent, functions).
The Hopi belief in the efficacy of their rainmaking ritual is not irrational
—although it is certainly false—because the conclusion, rain ceremonies
cause the rains to fall, follows validly from a world-view whose major
premise states that gods exist, and whose minor premise states that the
behavior of the gods can be influenced by rituals. That the premises are
false does not render them irrational—until or unless they are discontinued
by evidence. But all available ‘evidence’ confirms their validity: whenever
the ceremonies are performed it does, indeed, rain. Hence, given their
‘behavioral environment’ (Hallowell 1955:75–110), Hopi beliefs are not
irrational; and given their ecological environment, the apparent function of
these ceremonies is surely a sufficient explanation for their persistence. (For
further argument, see Spiro 1964).
If it is not sufficient, however, no appeal to unintended sociological
functions will provide us with a better explanation—indeed, as we have
already seen, it can provide us with no explanation at all. For how can the
function of social solidarity explain the practice of these—or of any other—
rituals? Notice that the objection to such an explanation is not that social
solidarity may not be an object of desire—there is no reason why it cannot;
and it is not that social solidarity is not achieved by the practice of these
rituals—it often is. The objection, rather, is that the achievement of this end
is not the desire which the practice of these rituals is intended to satisfy.
Surely, not even the proponents of this type of explanation would suggest
that Hopi rain ceremonies, sacrifices to Kali, exorcism of demons,
celebration of the Mass, and the like are practiced with the conscious
intention of achieving social solidarity. Is it suggested, then, that this is their
unconscious intention? I would doubt that anyone would make this
suggestion, for this suggests that if the efficacy of these rituals for the
attainment of their designated ends were to be disbelieved, they would
nevertheless be performed so that their solidarious functions might be
served. This argument surely cannot be sustained. I can only conclude, then,
that the persistence of these rituals is explicable by reference to what, for
anthropologists, are their apparent, rather than their real, functions.7
7. One might, of course, wish to defend the weaker thesis, viz. that if these practices were
sociologically dysfunctional, they would eventually disappear. Even this thesis is somewhat doubtful,
however, when applied to multi-religious societies in which the practice of religion, however
solidarious it may be for each religious group, has important dysfunctional consequences for the total
social system. Still these religions persist, and with undiminished vigor, despite—one is tempted to
say, because of—these consequences.
Even if it were to be conceded that institutions must have real, rather than
apparent, functions, anthropologists must surely be aware of those real
functions of ritual which are recognized by religious actors and which may,
therefore, reenforce their practice. For although religious ritual may not, in
fact, be efficacious for the elimination of poverty, the restoration of health,
the bringing of rain, and the like, the belief that it does achieve these ends
serves the important psychological (real) function of reducing hopelessness
—and its attendant anxiety—concerning their otherwise impossible
attainment.
Expressive

A third set of desires which, I would suggest, constitutes a motivational

source of religious behavior consists of painful drives which seek reduction
and painful motives which seek satisfaction. By ‘painful drives’ I refer to
those fears and anxieties concerning which psychoanalysis has taught us so
much: fears of destruction and of one’s own destructiveness, castration
anxiety, cataclysmic fantasies, and a host of other infantile and primitive
fears which threaten to overpower the weak and defenseless ego of the
young child and which, if they become too overwhelming, result in
schizophrenic and paranoid breakdown. By ‘painful motives’ I refer to
those motives which, because culturally forbidden—prohibited forms of
aggression, dependency, (Oedipal) sexuality, and the like—arouse feelings
of shame, inadequacy, and moral anxiety. These drives and motives are
much too painful to remain in consciousness and they are generally
rendered unconscious. Although unconscious, they are not extinguished;
they continue to seek reduction and satisfaction.
In the absence of other, or of more efficient means, religion is the vehicle
—in some societies, perhaps, the most important vehicle—by which,
symbolically, they can be handled and expressed. Since religious belief and
ritual provide the content for culturally constituted projective, displacement,
and sublimative mechanisms by which unconscious fears and anxieties may
be reduced and repressed motives may be satisfied, these drives and
motives, in turn, constitute an important unconscious source of religious
behavior. Because the range of painful drives and motives which find
expression in religion remains to be discovered by empirical research, I
shall merely comment on two motives which I believe to be universally—
but not exclusively—satisfied by religion; and since I have already
attempted to deal with this problem elsewhere (Spiro 1961b), I shall be
brief.
Forbidden dependency needs inevitably seek satisfaction in religious
behavior, in that the religious actor depends on superhuman beings for the
gratification of his desires. Repressing his desire to remain in a state of
childlike dependency on powerful adult figures, he can still satisfy this
desire, symbolically, by his trust in and reliance upon superhuman beings.
Similarly, since all religions of which I am aware postulate the existence
of malevolent, as well as of benevolent, superhuman beings, repressed
hostility motives can be displaced and/or projected in beliefs in, and rituals
designed for protection against, these malevolent beings. Prevented from
expressing his hostility against his fellows, the religious actor can satisfy
this desire symbolically through religion (Spiro 1952).
These, then, are three sets of desires whose satisfaction partially explains
the persistence of religion. But though the persistence of religion is to be
found in motivation—a psychological variable—the sources of motivation
are to be sought, in part, in society. For just as the sociological causes
(social structural variables) of the truth of religious beliefs achieve their
effects through mediating psychological processes—feelings, projections,
perceptions, and the like—so too the psychological causes (desires) of
religious behavior may be explained by reference to those sociocultural
(and biological) variables by which they are produced. With few
exceptions, human drives are acquired rather than learned; and all human
goals, it is probably safe to assume, are acquired. Since the crucial context
for the learning of human drives and goals is social, and since most
religious drives—but not most goals—are acquired in the child’s early
experiences, it is the family that once again is the nuclear structural
variable.
Indeed, because these motivational variables are acquired within
specified structural contexts, and because these contexts exhibit a wide
range of variability, differences in the kinds and/or intensity of desires
which constitute the motivational basis for religious behavior should vary
systematically with differences in family systems (including socialization
systems), as well as with the alternative, nonreligious means for their
satisfaction. The latter qualification is important. I have stressed, with
respect to the three sets of desires which have been discussed, that in the
absence of alternative institutional means, it is religion which is the means
par excellence for their satisfaction. If cognitive desires, for example, are
satisfied by science; if substantive desires are satisfied by technology; or if
expressive desires are satisfied by politics or art or magic, religion should,
by that extent, be less important for their satisfaction. In short, the
importance of religion would be expected to vary inversely with the
importance of other, projective and realistic, institutions.8
8. This does not imply, as some nineteenth-century thinkers believed, that as other institutions
assume more of the traditional functions of religion, the latter will disappear. So far, at least, there has
been no viable alternative to religion in providing a solution to the problem of ‘suffering’: and the
malaise of modern man, on the one hand, and the persistence of religion among many modern
intellectuals, on the other, seem to suggest that a viable functional alternative is yet to be discovered.
Holding other institutions constant, then, the kinds and intensity of drives
which are satisfied by religion, the means by which they are believed to be
satisfied, and the conceptions of the superhuman beings that are the agents
of satisfaction should vary with variations in childhood experiences in
which drives (and their intensity) are acquired, the means by which children
influenced parents (and surrogates) to satisfy their drives, and the degree to
which parents (and surrogates) do, in fact, satisfy them (Spiro & D’Andrade
1958).
In short, a motivational explanation of religious behavior can, in
principle, explain variability in behavior and, hence, can be tested
empirically. A great deal of culture-and-personality research, too extensive
to cite here, has been devoted to this very problem. Indeed, much of the
research concerned with the structural bases for religious cognitions has,
simultaneously, been devoted to the motivational bases for religious
behavior. Unfortunately, however, cognitive desires have received less
attention than expressive and substantive desires, possibly because we
know comparatively little about how they are acquired.

Causal and Functional Explanations

We may now return to the question with which we began this paper: is the
existence of religion to be explained causally or functionally? I have
suggested that the acquisition of religious beliefs is to be explained
causally, and that the practice of these beliefs is to be explained in terms of
motivation—which means that it is explained both causally and
functionally. Religion persists because it has functions—it does, or is
believed to, satisfy desires; but religion persists because it has causes—it is
caused by the expectation of satisfying these desires. Both are necessary,
neither is sufficient, together they are necessary and sufficient. The causes
of religious behavior are to be found in the desires by which it is motivated,
and its functions consist in the satisfaction of those desires which constitute
its motivation.
 Classical functionalist theory has, of course, tended to dismiss motivation
and other psychological variables as being outside the domain of
anthropology. Firth, for example (1956:224), writes that in the study of
ritual the anthropologist is not concerned with “… the inner state as such of
the participants,” but rather with “the kinds of social relations that are
produced or maintained.”9 If ‘inner states’ were irrelevant for an
explanation of religion, one might wish to defend the thesis that
anthropology, whose central concern is the explanation of social and
cultural institutions, should ignore them. But if by ignoring ‘inner states’,
and other psychological variables, we cannot adequately explain religion,
or, for that matter, the ‘kinds of social relations’ which it produces, we
ignore them at our peril. Let me, beginning with the first proposition, take
up each in turn.
9. I should hasten to add that, in a later publication (1959:133), Firth changed his view and stressed
the importance of the individual for a complete explanation of religion.
An explanation of religion, or of any other sociocultural variable,
consists in the specification of those conditions without which it could not
exist. If religion is what is to be explained, if religion, that is, is the
dependent variable, it can be explained only in terms of some independent
variable, some condition by which it is maintained or sustained. I have
argued that motivation is the independent variable. That religion, like other
sociocultural variables, has sociological functions—it produces or
maintains certain kinds of ‘social relations’—is undeniable; and it is one of
our tasks to study these sociological functions. Indeed, these sociological
functions may be crucial for the maintenance of society. If, for example,
social solidarity is a functional requirement of society, and if religion—as it
is frequently argued—is one of the institutions that satisfies this
requirement, religion is necessary (a cause) for the maintenance of society.
Notice, however, that if religion does produce solidarious social relations,
solidarity provides us with an explanation of society, not of religion. In this
case, in short, social solidarity does not explain religion; religion explains
social solidarity. For social solidarity is the dependent, and it is religion that
is the independent, variable. In sum, if we are interested in the ‘kinds of
social relations’ that are produced by religion, we are interested in
explaining society; and religion—it is assumed—can supply us with an
entire, or a partial, explanation.
 But the functionalist argument sometimes assumes a different form.
Religion, it is argued, not only satisfies certain functional requirements of
society, but it is a necessary condition for their satisfaction. Since society
cannot exist unless these requirements are satisfied, and since religion is a
necessary condition for their satisfaction, these requirements ‘cause’ the
existence of religion. It is not that the desire to satisfy some requirement
motivates religious behavior—the latter may be accounted for by numerous
other desires of the kind suggested, perhaps, in this paper. Rather, since
religion is a necessary condition for the satisfaction of this requirement, the
need for its satisfaction can explain the existence of religion. This
argument, I think, is implicit in most functionalist interpretations of the
Radcliffe-Brown variety. Thus Radcliffe-Brown himself (1948:324) writes
that religious ceremonies “are the means by which the society acts upon its
individual members and keeps alive in their minds a certain system of
sentiments. Without the ceremonial these sentiments would not exist, and
without them the social organization in its actual form could not exist.”
Notice, before we examine this type of explanation, that it is not a
functionalist explanation at all; it is a causal explanation in which the cause
happens to be a functional requirement. Notice, too, that despite the present
tense of the predicates, this is really an explanation of the origin, not the
persistence, of religion: the practice of religion—which is the only means
by which religion persists—is not motivated by a desire to keep these
sentiments alive, although the persistence of these sentiments is its
consequence. But these questions aside, as an explanation of religion, this
theory suffers from three defects: technical, methodological, and
theoretical. Technically, no mechanism is specified by which the need for
solidarity—and, therefore, the need for religion—gives rise to, or ‘causes’,
religion. Methodologically, it cannot explain the variability of religion—
how is it that the need to sustain social sentiments produces such a
bewildering range of religious beliefs and rituals?—and, therefore, there is
no way by which it can be tested. Theoretically, it is based on an
unwarranted functionalist assumption, which Merton (1957) has aptly
termed the assumption of ‘indispensability’. What does this mean?
For some functional requirements any number of variables may be
adequate for their satisfaction. Since none is necessary, although each may
be sufficient, they are called ‘functionally equivalent alternatives’. These
are to be distinguished from a variable which is necessary or indispensable
for the satisfaction of a requirement—without it the requirement could not
be satisfied. The cross-cultural evidence strongly suggests that there are
few, if any, indispensable variables; that, rather, for almost any sociocultural
variable there are functional equivalents. Hence, even if it is the case that
religion is the means by which Andamanese ‘sentiments’ are kept alive, it
does not follow that religion is the only means by which this end could be
accomplished, or that in other societies other institutions do not or cannot
serve the identical function.10
10. To cite but two counter-instances. First, there are the more than 150 atheistic kibbutzim in
Israel, in which religious ceremonial does not exist (Spiro 1956). Second, there are societies in which
the important social sentiments are supported by secular rather than religious institutions. It would be
difficult, for example, to discover even one sentiment important for the maintenance of capitalist
democracy which is conveyed by the paramount Christian ceremony, the Eucharist. National
celebrations, however, especially as interpreted by Warner (1959), serve this function par excellence.

In sum, from the fact that religion is a sufficient condition for the
satisfaction of a requirement, it is invalidly deduced that it is a necessary
condition. And, if it is not a necessary condition, its existence cannot be
explained by arguing that without it society could not survive.
Although it is society, rather than religion, which is explained by the
sociological consequences of religion, social anthropology—as the
comparative study of social and cultural systems—is most certainly
concerned with these functions; and psychological variables are not only
necessary for the explanation of religion, but without them certain of its
sociological functions would go unrecognized. (The classic example, of
course, is Weber’s [1930] analysis of the rise of capitalism, but it deals with
the change in, rather than the persistence of, a social system.) Thus, the
adjustive (real) function of religion, by satisfying the need for explanation,
provides a society with a common ‘behavioral environment’ which, as
Hallowell (1955) observes, satisfies a set of minimal requirements for the
existence of any society: the requirement for a common object orientation,
spatiotemporal orientation, motivational orientation, and normative
orientation. It would be difficult to conceive of the possibility of social
integration without a minimum level of such shared orientations.
The adaptive (apparent) function of religion, in satisfying the desire for
the attainment of goals, provides—as Marxism has stressed—a most
important basis for social stability. Disbelief in the efficacy of superhuman
means for the achievement of this-worldly goals could certainly become a
potential basis for social discontent and socioeconomic change. At the same
time, the (real) function of religion in reducing anxiety concerning the
attainment of goals—especially those for whose attainment available
technological skills are ineffective—and, thus, in providing a minimum
level of psychological security, serves to release energy for coping with the
reality problems of society.
Finally, the inteerative (real) function of religion, in allowing the
disguised expression of repressed motives, serves a number of sociological
functions. By providing a culturally approved means for the resolution of
inner conflict (between personal desires and cultural norms), religion (a)
reduces the probability of psychotic distortion of desires, thereby providing
a society with psychologically healthy members,11 (b) protects society from
the socially disruptive consequences of direct gratification of these
forbidden desires, (c) promotes social integration by providing a common
goal (superhuman beings) and a common means (ritual) by which the
desires may be gratified.
11. Nadel, recognizing the ‘defensive’ function of religion, writes that in providing rituals by
which forbidden impulses may be expressed, religion “… anticipates as well as canalizes the working
of psychological mechanisms, which might otherwise operate in random fashion or beyond the
control of society, in the ‘private worlds’ of neurosis and psychopathic fantasies” (1954:275).

Conclusion
It would appear from the foregoing discussion that an adequate explanation
for the persistence of religion requires both psychological and sociological
variables. If the cognitive bases for religious belief have their roots in
childhood experience, their explanation must be found in social structural
and, more specifically, family structure variables. Here religion is the
dependent variable, and family structure is the independent sociological
variable which effects religious belief by means of such intervening
psychological variables as fantasies, projections, perceptions, and the like.
If religion persists because of its gratification of desires, explanations for
the bases of religious behavior must be found in psychological and,
specifically, motivational variables. Here, again, religion is the dependent,
and motivation is the independent, variable. Since, however, motivation
consists in the intention of gratifying desires, and since desires are rooted
either in organic or in acquired drives, the motivational roots of religious
behavior can, ultimately, be found in those biological and social structural
variables, respectively, by which they are produced and/or canalized. Again,
it is the family which emerges as the crucial sociological variable. Religion,
then, is to be explained in terms of society and personality.
Many studies of religion, however, are concerned not with the
explanation of religion, but with the role of religion in the explanation of
society. Here, the explanatory task is to discover the contributions which
religion, taken as the independent variable, makes to societal integration, by
its satisfaction of sociological wants. This is an important task, central to
the main concern of anthropology, as the science of social systems. We
seriously err, however, in mistaking an explanation of society for an
explanation of religion which, in effect, means confusing the sociological
functions of religion with the bases for its performance.
In this paper, I have been concerned almost exclusively with the latter
aspect of religion. I have not, except incidentally, dealt with its sociological
functions or, what is perhaps more important, with how these are to be
measured. I have not dealt, moreover, with the problem of religious origins
because—despite the fact that numerous speculations have been proposed
(and I have my own, as well)—these are not testable. Nor have I dealt with
the problem of the cross-cultural variability in religion, except to suggest
some motivational bases for the persistence of different types of belief and
ritual. But the crucial problems—to which Max Weber has most
importantly contributed—I have not even touched upon. If, for example,
religion is centrally concerned with the problem of ‘suffering’, why is it that
explanations for suffering run such a wide gamut: violation of ethical
norms, sin of ancestors, misconduct in a previous incarnation, etc.? Or, if
religion promises redemption from suffering, how are the different types of
redemption to be explained? And, moreover, what is the explanation for the
different means by which the redemptive promise is to be achieved? These
are but a few of the central problems in the study of religion with which this
chapter, with its limited focus, has not been concerned.

Acknowledgments
Thanks are due to the author, the editor, and the American Academy of Arts
and Sciences for permission to quote the passage from “Golden Bough or
Gilded Twig?” by E. R. Leach, in Daedalus, 1961.

Note
Reprinted from Anthropological Approaches to the Study of Religion, edited by Michael Banton
(London: Tavistock Publications, 1966), pp. 85–126.

References

Durkheim, E. 1954. The elementary forms of the religious life.

Glencoe, 111.: Free Press.
Evans-Pritchard, E. E. 1954. The institutions of primitive society.
Oxford: Blackwell.
Firth, R. 1956. Elements of social organization. London: Watts; New
York: Philosophical Library.
______ 1959. Problem and assumption in an anthropological study of
religion. Journal of the Royal Anthropological Institute 89.
Fortes, M. 1959. Oedipus and Job in West African religion.
Cambridge: Cambridge University Press.
Freud, S. 1928. The future of an illusion. London: Hogarth Press.
Gerth, H. H., and Mills, C. W., eds. 1946. For Max Weber: Essays in
sociology. New York: Oxford University Press.
Goody, J. 1961. Religion and ritual: The definitional problem. British
Journal of Sociology 12.
Hallowell, A. I. 1955. Culture and experience. Philadelphia:
University of Pennsylvania Press.
______ 1956. The Rorschach technique in personality and culture
studies. In Developments in the Rorschach technique, vol. 2, ed. B.
Klopfer. Yonkers-on-Hudson, N.Y.: World Books; London: Harrap.
Hempel, C. G. 1952. Fundamentals of concept formation in empirical
science. Chicago: University of Chicago Press; London:
 Cambridge University Press.
Horton, R. 1960. A definition of religion and its uses. Journal of Royal
Anthropological Institute 90.
Kardiner, A. 1945. The psychological frontiers of society. New York:
Columbia University Press.
Leach, E. R. 1954. Political systems of highland Burma. London: Bell.
______1961. Golden bough or gilded twig? Daedalus.
Lowie, R. 1924. Primitive religion. New York: Boni & Liveright.
Merton, R. K. 1957. Social theory and social structure, rev. ed.,
Glencoe, 111.: Free Press.
Nadel, S. F. 1954. Nupe religion. Glencoe, 111.: Free Press.
Pruyser, P. 1960. Some trends in the psychology of religion. Journal of
Religion 40.
Radcliffe-Brown, A. R. 1948. The Andaman islanders. Glencoe, 111.:
Free Press.
Radin, P. 1957. Primitive religion. New York: Mayflower & Vision
Press.
Spiro, M. E. 1952. Ghosts, Ifaluk, and teleological functionalism.
American Anthropologist 45.
______1953. Ghosts: An anthropological inquiry into learning and
perception. Journal of Abnormal and Social Psychology 48.
______ 1956. Kibbutz: Venture in Utopia. Cambridge: Harvard
University Press.
______1961a. Social systems, personality, and functional analysis. In
Studying personality cross-culturally, ed. B. Kaplan. Evanston,
111.: Row, Peterson.
______1961b. An overview and a suggested reorientation. In
Psychological anthropology, ed. F. L. K. Hsu. Homewood, 111.:
Dorsey Press.
______ 1963. Causes, functions, and cross-cousin marriage: An essay
in anthropological explanation. Journal of Royal Anthropological
Institute 97.
______1964. Religion and the irrational. In Symposium on new
approaches to the study of religion, Proceedings of the American
Ethnological Society, Seattle.
Spiro, M. E., and D’Andrade, R. G. 1958. A cross-cultural study of
some supernatural beliefs. American Anthropologist 60.
Turner, V. W. 1962. Chihamba: The white spirit. Manchester:
Manchester University Press.
Warner, W. L. 1959. The living and the dead. New Haven: Yale
University Press.
Weber, M. 1930. The protestant ethic and the spirit of capitalism.
London: Allen & Unwin; New York: Scribner’s.
Whiting, J. W. M. 1961. Socialization process and personality. In F. L.
K. Hsu (ed.), op. cit.
9
Virgin Birth, Parthenogenesis, and
Physiological Paternity: An Essay in
Cultural Interpretation

Introduction
IN HIS CLASSIC report, W. E. Roth (1903:22) described the conception beliefs
of the Tully River Blacks. None assigns a procreative role to the father.
Although sexual connection as a cause of conception is not recognized among the Tully River
blacks so far as they are themselves concerned, it is admitted as true for all animals:—indeed this
idea confirms them in their belief of superiority over the brute creation. A woman begets children
because (a) she has been sitting over the fire on which she has roasted a particular species of
black bream, which must have been given to her by the prospective father, (b) she has purposely
gone a-hunting and caught a certain kind of bullfrog, (c) some man may have told her to be in an
interesting condition, or (d) she may dream of having a child put inside her.
By whichever of the above methods the child is conceived, whenever it eventually appears, the
recognized husband accepts it as his own without demur. (My emphasis.)

In 1961 Edmund Leach (1961:376) wrote “[It is not] a legitimate

inference to assert that these Australian aborigines were ignorant of the
connexion between copulation and pregnancy.” Instead,
The modern interpretation of the rituals described would be that in this society the relationship
between the woman’s child and the clansmen of the woman’s husband stems from the public
recognition of the bonds of marriage, rather than from the facts of cohabitation, which is a very
normal state of affairs.
In 19661 wrote (Spiro 1966:112) that since, on the face of it, I found this
interpretation implausible, and since Dr. Leach had not supported it (either
with data or with theory), I would hold, instead, that:
… the aborigines are indeed ignorant of physiological paternity, and that the four statements
quoted in Roth are in fact proffered explanations for conception.1

1. Since my entire contribution to the present controversy is confined to this brief statement, it
is rather intriguing to know by what modes of inference Dr. Leach was able to conclude in his
most recent contribution that (to take some examples at random):

Professor Spiro believes that explanation consists of postulating causes and ultimate origins
for the facts under observation (Leach 1967:39).
He is positively eager to believe that the aborigines were ignorant and he accepts their
assertion as a fact without investigating the evidence at all (1967:41, original emphasis).
He displays an extreme reluctance to believe that the products of aboriginal thought can be
structured in a logical way (1967:41, my emphasis).
Professor Spiro (and all the neo-Tyloreans who think like him) desperately wants to believe
… that dogma and ritual must somehow correspond to the inner psychological attitudes of the
actors concerned (1967:40, my emphasis).

As ostensible characterizations of my position, these—and every other position which Leach

imputes to me—have one thing in common. They are all false.

In 1967, “provoked” by my “critical comments,” Leach returned to the

problem of aborigine conception beliefs in his Henry Myers Lecture (Leach
1967). Stripped of its many and tedious ad hominem arguments—the
impugnations of his opponents’ scholarship,2 their scientific integrity,3 and
even their “normality,”4 the gratuitous imputation to them of motives and
attitudes,5 the obsessive imputation to them of a racist ideology6—this
lecture might be summarized in the following propositional form. 1) If a
cultural belief in Australia states that a woman conceives when a spirit-
child enters her womb, 2) it may not be inferred from this belief that the
natives are ignorant of physiological paternity, because 3) only “idiots”
would believe this conception theory, and 4) only those anthropologists who
view the natives as “childish,” “stupid,” “superstitious,” and so on, would
believe that the natives would believe in such a theory. Hence 5) although
this belief seems to be about conception, it is really not about conception at
all; rather 6) it is about patrilateral filiation. 7) Moreover, even if it is about
conception this belief must be viewed as a cultural “dogma,” i.e., as a “kind
of religious fiction,” which, though asserted, is not really believed to be
true.
2. One or two instances will suffice. Thus according to Leach, I am “quite unaware that almost
identical questions have been asked many times before” (1967:39); I am ignorant of the fact “that a
huge literature of commentary surrounds Roth’s original ethnographic report,” and so on.
3. For example, the proponents of physiological ignorance “… have shown themselves willing to
accept even the flimsiest evidence for the fact of ignorance, while evidence against, even when it is
most meticulously recorded, is repeatedly rejected …” (1967:46, original emphasis).
4. Thus, convinced in his own mind that when her informants spoke to Kaberry about birth and
pregnancy they could not have been ignorant of physiological paternity since—according to Leach—
they were really speaking of metaphysical matters, Leach comments: “That a distinguished
anthropologist [Kaberry] should once have thought otherwise displays the oddity of anthropologists
rather than the oddity of the aborigines” (Leach 1967:47). Leach’s comment is all the more gratuitous
since, contrary to the implication of this quotation, Kaberry goes out of her way (Kaberry 1939:ch. 2)
to emphasize the rational and empirical bases for aboriginal belief.
5. For example, if certain groups “have persuaded” their ethnographers that they were “ignorant of
the facts of life,” and if the ethnographer “believed what he was told it was because such belief
corresponded to his own private fantasy of the natural ignorance of savages” (1967:41). Again, the
motivation for this “fantasy” is that these ethnographers “seem to gain reassurance from supposing
that the people they study have the simple minded ignorance of small children” (1967:45).
6. This is examined below.

Before examining this argument, it is important to set one fact straight.

Although the question of whether it is possible to draw valid conclusions
about the personal beliefs of social actors from their cultural beliefs is an
important theoretical question—and one to which we shall return—Dr.
Leach begs the empirical question by posing it as the central problem in this
controversy. For contrary to Leach’s claim, Sharp and Malinowski, Kaberry
and Austin, and most of the others who contend that the natives are ignorant
of physiological paternity, do not claim it to be an inference based on the
Australian cultural belief which they recorded; they claim, rather, that—the
cultural belief aside—it is an empirical finding. And they support this claim
with a great deal of field data. Based on their interviewing, probing,
inquiring, and challenging of informants, they concluded that the personal
beliefs of the social actors corresponded to the cultural beliefs which they
had been taught. If, then, it can be assumed that the quality of Australian
fieldwork is no worse, on an average, than field work conducted in other
parts of the world, Austrlian ignorance of physiological paternity is an
anthropological finding whose empirical status—until refuted—is no worse
(and no better) than most anthropological findings. (The meaning of
“ignorance,” however, is not simple, as we shall see below.) Having
structured the controversy, then, as involving ethnographical inference
(rather than ethnographical fact7) it must be stressed that Dr. Leach nowhere
confronts, let alone refutes, this finding.
7. All “facts,’’ to be sure, are inferences. “Facts” obtained through anthropological interviews rest
on a number of inferences concerning the reliability of our interviewing instruments, the motivation
and veracity of our informants, and so on. But to the extent that this is so, the factual basis for
Australian conception beliefs is no less suspect than most other anthropological “facts.”
Given Leach’s premise, that the controversy concerns ethnographical
inference, his argument (as summarized above) contains two main theses:
the cultural belief concerning conception does not mean what it says, and,
even if it does, the natives do not believe what it says. There are, then, two
distinct issues in this controversy. First, what is the meaning of the cultural
belief? Second, whatever its meaning, can it be inferred from this belief that
the natives are ignorant of physiological paternity? Since it is impossible to
examine all of Leach’s arguments, I shall confine my discussion to those
which seem to be salient. With respect to the issue of physiological
paternity, I shall examine three of his arguments, viz.: 1) Only “idiots’’
could be ignorant of physiological paternity, and only those who view
natives as “idiots” would argue that they are ignorant of the relationship
between copulation and conception. 2) The Australian evidence supports
his thesis that the aborigines are not ignorant of this relationship. 3) The
comparative—and especially the Christological—evidence supports this
thesis. With respect to the meaning of the cultural belief, I shall examine
two of Leach’s arguments, viz.: 1) Both the Australian and 2) the
Christological evidence support the sociological meaning which he
attributes to the belief.

The Issue of Physiological Paternity

1. The Argument of Alleged Intellectual Inferiority

Although Leach’s attribution of a racist ideology to his opponents is, like all
ad hominem arguments, irrelevant to the substantive issues of this
controversy, it must nevertheless be examined because it looms so large in
his lecture: it is, he claims, the “crux” of the controversy (1967:45). Leach’s
argument is simply stated. That Australians are ignorant of physiological
paternity is not a scientific finding, but a personal prejudice based on the
conviction that primitive peoples are intellectually inferior and incapable of
logical thought.
To say that a native is ignorant [notice the escalation of the specific, “ignorant of physiological
paternity,” to the generic, “ignorant”] amounts to saying that he is childish, stupid, superstitious.
Ignorance is the opposite of rationality; it is the quality which distinguishes the savage from the
anthropologist (1967:41, original emphasis).

But surely the premise of this argument is false. None of Leach’s

opponents—neither Roth nor Malinowski, neither Kaberry nor Ashley
Montagu—holds that “ignorance is the opposite of rationality.” For, of
course, ignorance is the opposite not of rationality, but of knowledge;
irrationality, not ignorance, is the opposite of rationality. To be ignorant of
something, i.e., to have no knowledge (or false knowledge) is not
necessarily to be irrational (or, for that matter, childish, stupid,
superstitious, etc.). Indeed, if it were, Leach’s efforts to salvage their
“rationality” by insisting that the natives are aware of physiological
paternity would be in vain. For they would still be ignorant of the fact that
the earth is round, that it moves around the sun, that genes are the physical
basis of heredity, that man evolved from a lower primate, etc., etc. And not
only the Australians! On this premise, Europeans were irrational until the
nineteenth century because they were ignorant of human evolution; until the
fifteenth century, because they did not believe in a heliocentric universe;
and so on.8 In short, the premise—ignorant=irrational—is not only
semantically false, but it is culturally and historically absurd.
8. Malinowski’s statement concerning this matter is as apposite today as it was thirty-five years
ago:

I think it is rather inconsistent to get excited about the faulty knowledge of the Trobrianders when
it comes to processes of sexual fertilisation, wnile we are perfectly satisfied that they possess no
real knowledge as to processes of nutrition, or metabolism, the causes of disease and health, or
any other subject of natural history … (Malinowski 1932:ii).

If knowledge (rather than rationality) is the opposite of ignorance, the

question can then be posed: is it irrational for Australians to be ignorant of
physiological paternity and to believe, instead, that women conceive
because they are entered by spirit-children? The answer to this question
clearly depends on what we might mean by “irrational.” When we
characterize a belief as “irrational” we can mean, so far as I can tell, one or
more of the following things. 1) The belief is a conclusion invalidly
deduced from some axiom or premise; to hold the belief is irrational
because it rests on a fallacy in deductive logic. 2) The belief is an empirical
generalization which is invalidly derived from available evidence; to hold
the belief is irrational because it rests on a fallacy in inductive logic. 3) The
belief asserts some proposition which is inconsistent with, or in
contravention of, some other belief; to hold the belief is irrational because it
violates the law of contradiction. 4) The belief asserts something about the
universe which is inconsistent with, or in contravention of, reliable
knowledge; to hold the belief is irrational because it is empirically absurd
(cf. Spiro 1964).
If I am correct in assuming that these are the criteria we use for judging
the irrationality of belief, Australian ignorance of physiological paternity
can scarcely be taken as a measure of irrationality. Of all the possible
reasons—too numerous to detail here, but superbly summarized by Ashley
Montagu (1937:235–76)—which have been offered for the limitation in
their knowledge of the relationship between coitus and conception, none
includes faulty logic. Moreover, using the above criteria, for the aborigines
to believe in their spirit-children theory of conception is entirely rational: it
explains all the facts, it accounts for all the anomalies rendered inexplicable
by a procreation theory (cf. Malinowski 1932:184–94), it is contradicted by
no available knowledge, and it leads to valid deductive and inductive
conclusions (cf. Ashley Montagu, 1937:ch. 14). To be sure, this conception
theory is false; but to hold a false belief is not in itself irrational, unless it
violates the four criteria mentioned above. For, although judgments
concerning the truth of a belief are relative to the latest developments in
science, judgments concerning the rationality of holding that belief must be
relative to the state of knowledge of the society in which it is found. The
cultural heritage of any group provides its members with a Weltanschauung
which is based on (among other things) the level of knowledge found in that
society. This Weltanschauung includes a set of assumptions concerning the
nature of the world which not only comprise the major premises for any
argument or discourse, but which constitute the parameters of their
“behavioural environment” (Hal-lowell 1955: ch. 4). Hence, given their
major premises—and they have no access to alternative premises—and
given their behavioral environment (is any other available to them?) the
Australians would exhibit none of the four types of logical fallacy alluded
to above in holding their conception beliefs to be true (cf. Spiro: 1964).
Holding them does not render them childish, stupid, superstitious or idiotic;
and for anthropologists to believe that they hold these beliefs does not mean
that they view them in this manner.
Indeed, Leach’s allegation that his opponents do, in fact, take ignorance
of physiological paternity to be a sign of childish, irrational mentality is
simply untrue. Since it would be tedious to take up each of his opponents in
turn, I shall confine my comments to Ashley Montagu, whose monograph
(1937) remains the classic statement of the controversy, and who, after an
examination of the entire ethnographical literature, concludes that nowhere
in Australia can one find true knowledge not only of physiological
paternity, but of physiological maternity as well. Now, for Ashley Montagu
—whether his conclusion is right or wrong—to be included among those
who believe in the inferiority of any group is nothing short of extraordinary.
Thus, at the very beginning of his study, Ashley Montagu (1937:10–11)
dismisses any possible notion that ignorance of physiological paternity
implies
… that the Australian aboriginal is a being mentally inferior to ourselves, or that he is incapable
of our particular kind of reasoning, or that he has a pre-logical mentality, or what not. The facts
point clearly in the opposite direction, namely, that the Australian aboriginal native endowment is
quite as good as any European s, if not better. In support of the latter statement that exists a
certain amount of evidence of the weightiest kind …

He concludes his study (1937:308) with a similar statement.

The mind of the Australian is no more pre-logical than that of the modern educated man or
woman. Essentially the mind of the savage functions in exactly the same way as our own does,
the differences perceptible in the effects of that functioning are due only to the differences in the
premises upon which that functioning is based, premises which represent the logical instruments
of the native’s thought, and have their origin in categories and forms of judgment which are to
some extent different though quite as rigorously organised as our own.

Although space forbids the citation of similar statements by, for example,
Roth (1903:23), Malinowski (1932:ii sqq), Kaberry (1939: ch. 2), Spiro
(1964:109–10), etc., the conclusion is inescapable. It is not true that those
anthropologists who believe that the aborigines are ignorant of
physiological paternity either view their ignorance as childish, or view them
as intellectually inferior.
2. The Argument from the Australian Evidence

Since those who hold that the Australians are ignorant of physiological
paternity base their conclusion, according to Leach, on prejudice, an
objective examination of the ethnographical data, so he argues, supports his
thesis that the Australians are not ignorant of physiological paternity. I
submit that, with one exception, none of the data which Leach cites support
this conclusion.
Leach’s first argument is that there is a “classically established reason”
for supposing that the Tully River Blacks were not ignorant: “They freely
admitted to Roth that the cause of pregnancy in animals other than man is
copulation” (Leach 1967:40). Since that is the entire argument and the full
quotation, it is rather difficult to divine precisely what the force of this
argument might be:9 that we can best know about a people’s beliefs about
themselves from their beliefs about animals? But then Dr. Leach would
have to concede that the Trobrianders and the Arunta, for example, are
indeed ignorant of physiological paternity because they, like many other
groups, deny any procreative basis for animal pregnancy. (For the
Trobriands, cf. Malinowski 1932:162; Austin 1934:112; for the Arunta, cf.
Spencer & Gillen 1899:84, v. 2.) Moreover, people everywhere make
distinctions between themselves and animals, sometimes invidious to
themselves, sometimes invidious to the animals. In the Tully River case the
latter is the case, as Roth tells us: “This idea confirms them in their belief of
superiority over the brute creation.”
9. In order to forestall Leach’s penchant for impugning the scholarly knowledge of his opponents, I
would only comment that know the arguments of Roheim, Warner, and Thomson on the subject. It is
Leach’s that I am unaware of.
Leach’s second argument is that the data of the more recent
ethnographers, working in other parts of Australia, show that Australian
conception beliefs are cultural dogmas, “a kind of religious fiction”; hence
the “formally expressed ignorance” reflected in the dogmas does not
correspond to the attitudes of the social actors (Leach 1967:40). Ignoring
the debatable methods by which data from other parts of Australia are used
to refute the Tully River data, or (even conceding that) by which data
collected thirty to fifty years after the original report are used (considering
the variety of ensuing acculturative forces) to interpret its meaning, my
reading of these data is rather different from his.
 Of the six ethnographies which Leach says support his position Warner’s
alone (1937) can be claimed; the others are either unclear, or else they
refute it. Thus, Stanner (1933) argues that the “confused version” of
physiological paternity, which is found side by side with a “mystical
theory” among the Daly River tribes, is a result of acculturation. Meggitt,
although certainly stressing that the contemporary Walbiri (especially the
females) emphasise both spiritual and physical causes of pregnancy, is
cautious, in view of Spencer and Gillen’s data, about drawing conclusions
conerning the beliefs of former generations from contemporary data.
“Although it is clear,” he writes, “that people in the past recognized the fact
of physiological maternity, we cannot be sure to what extent they were
aware of physiological paternity” (Meggitt 1962:272). Roheim, although
arguing that the Australians are indeed aware of physiological paternity, yet
admits that following initiation, the knowledge is repressed, i.e., it is
rendered unconscious (Roheim 1932:97); on a conscious level, adults are
ignorant.
Thus far, then, Leach’s contention that his position is supported by recent
ethnographers, is rather misleading. When, however, he includes Sharp and
Thomson (1933) among those whose data allegedly support him, his
contention is not only misleading, it is untrue. Lauriston Sharp (in letters
written from the field to Ashley Montagu) states repeatedly that the Yir-
Yiront are literally ignorant of physiological paternity. Moreover, after three
years of fieldwork he expresses astonishment that their ignorance “…
should by some be considered as an indication that they must be moronic”
(Sharp, in Ashley Montagu 1937:162–65). As for Thomson, although he
reports that the Koko Ya’o believe in the procreative function of semen,
their procreative knowledge remains as “magical” as—and, from Leach’s
view, even more “irrational” than—that of the other Australian tribes. For,
while affirming physiological paternity, they strongly deny physiological
maternity: “Informants … treated contemptuously any suggestion that the
mother has any part in conception” (Thomson 1933:506).10
10. Leach concedes that the data of at least one “modern” ethnographer—Phyllis Kaberry—do not
support his thesis; but he attributes this fact to her faulty interpretation. When properly interpreted, he
argues, her data support him. Thus, he claims, Kaberry’s finding concerning the belief in a spirit-
child theory of conception means that the entry of a spirit-child into a woman is a sign of pregnancy,
not its cause. Commenting on Kaberry’s report, he writes (1967:47). “In other words, a woman
recognizes that she is pregnant when she experiences ‘morning sickness’—which is true also of
European women” (my emphasis). Now, in addition to his misreading of European ethnoscience—
European women recognize pregnancy by the stopping of their menses, an event which occurs before
morning sickness—Leach has misread Kaberry as well. For, in the passage quoted by Leach, Kaberry
writes: “Conception occurs when one of these [spirit children] enters a woman. Its presence in the
food given her by her husband makes her vomit …” (Leach 1967:47, my emphasis). I submit that
“conception occurs” does not mean “… a woman recognises that she is pregnant… Exegesis aside,
however, although Kaberry herself interprets this belief as referring to the cause of conception, not its
recognition, Leach’s only rebuttal is to state that it is not so.
Again, to Kaberry’s comment that, “Questioned on the function of intercourse natives admitted
that it prepared the way for the entry of the spirit child” (Leach 1967:47). Leach is “very puzzled as
to how anyone [i.e. Kaberry] could interpret such data as indicating ‘ignorance of physiological
paternity.’” But I am puzzled by Leach’s puzzlement. It is one thing to say that sexual intercourse
opens the vagina so that the spirit child can enter the woman—a frequently recorded belief in
Australia; it is quite another to say that it impregnates the woman, especially when the conception
belief states explicitly that only the spirit-child does that. The issue, in short, is whether intercourse is
believed to be the cause of pregnancy, and this the natives deny. As Ashley Montagu (1937:200),
summarizing all of the data on this question, remarks: [When the anthropologist] inquires whether
intercourse has any connection with pregnancy or childbirth he is in most cases informed that
intercourse serves to prepare the woman for the entry of a spirit child into her, but that this
preparation is not in itself the cause of pregnancy or of the entry of the child into the woman … In
other words, intercourse prepares the way for that factor to become operative which is the cause of
pregnancy [spirit child entry], but intercourse is not either alone or in conjunction with other factors
the cause of pregnancy.
Leach’s last argument in support of his thesis that the Tully River Blacks
believe in physiological paternity is that the alleged ignorance of
physiological paternity in the Trobriands is seriously to be questioned.
Thus, Leach claims that 1) by 1932 Malinowski had seriously qualified his
original position; 2) Powell and Austin, independently, had recorded
observations “strikingly similar” to those of Meggitt on the Walbiri; 3)
Fortune showed that Trobriand “ignorance” reflected religious dogma,
rather than the state of their knowledge. Ignoring again the methodological
problem, the fact is that all of these claims (with the exception of the one
concerning Powell) are demonstrably false.

1. It is true that in his “Special foreword,” Malinowski (1932) (good

contextualist that he is) cautions his readers to view Trobriand
conception beliefs within the total context of Trobriand culture.
Nowhere in this foreword, however, does he retract anything he had
previously written about their beliefs. More important, five years later,
Malinowski (1937:xxiv) writes:
[The verdict of Ashley Montagu] that in Australia practically universally, according to
orthodox belief, pregnancy is regarded as causally unconnected with intercourse will, I
think, remain the ultimate conclusion of science. It is supported by an irrefutable body of
solid fact.
2. Since Powell’s thesis is not available to me, I cannot comment.
Austin’s report, however, is available, and, contrary to Leach’s claim,
everything in his report explicitly and unambiguously confirms
Malinowski’s findings. The following samples will have to suffice.
“The Trobriand Island native does not know of the fertilizing agency
of the male seed. In my interviews with some fifty intelligent natives,
drawn from all parts of the Trobriand group of islands, not one
believed that male semen, or any part of it, entered the uterus” (Austin
1934:105, original emphasis). Again, when asked about precautions
taken by women to prevent conception by ridding themselves of
semen, they would answer in these terms: “Why should our girls do
this? Momona [semen] does not cause pregnancy. It is like water”
(106). Austin also reports that although most informants held to the
baloma theory of conception, a minority—“not at all considered the
prevailing opinion”—did not consider this the controlling principle.
Even they, however, did not place any credence in sexual intercourse
as a cause of pregnancy. On the contrary, for them, “… the child was
merely formed out of the mother’s blood” (107). As Austin says (108),
“In both cases, however, there was no doubt that ‘man’s contribution
towards the new life in the mother’s body’ was nil.” In short, “there
was no doubt that the father was not thought to be a contributor
towards the formation of the foetus …” (112; original emphasis).
3. From an anecdote in Fortune (1963:239), Leach (1967:48) infers that
the debate reported between the Trobrianders and the Dobuans over
the former’s conception theory “… was plainly about doctrine not
about knowledge.” Aside from the gratuitous assumption that doctrine
cannot be about knowledge, the logic of Leach’s inference eludes me.
His inference is based on the fact that when Fortune, who had
accompanied a party of Dobuans to the Trobriands, arranged for the
debate, the Dobuans became angry with him and turned their heads to
the wall. On the basis of this rather strange inference Leach then
concludes that the “evidence for [Trobriand belief being] ‘dogma’
[rather than knowledge] is very clear” (my emphasis). Fortune himself,
incidentally, whose interpretation Leach ignores, provides no support
for Leach’s interpretation. In his original report Fortune attributed the
Dobuan behaviour to the fact that the two parties had quarrelled about
their respective theories in the past and had no desire to revive the
 dispute (Fortune 1963:329). In his new preface (1963:xiv-xv) and new
appendix (1963:311–12), he suggests that the more basic reason for
Dobuan anger is to be found in their taboo against asking to hear
legends while on an outward-bound voyage or in a port of call. This
interpretation, too, is ignored by Leach.

In summary, then, the Australian data cited by Leach are, for the most
part, the reverse of what he purports them to be. 1) That Tully River Blacks
admit that sexual intercourse is the cause of animal pregnancy is true, but
its implication for the issue at hand is unclear. 2) Contrary to Leach’s
argument, the modern ethnographies cited by him refute his claims that as
cultural “dogmas” Australian conception beliefs are not taken seriously. 3)
Contrary to Leach’s argument, Malinowski’s data on the Trobriand remain
unchanged by either what he, or any of the other scholars cited by Leach
(with the possible exception of Powell) have written.

3. The Argument from the Christian Evidence

Referring to the numerous European folk tales which recount the magical
conception of gods and heroes, Leach (1967:40) observes that “some of
these stories resemble very closely indeed the account given to Roth by the
Tully River Blacks of how ordinary human births occur” (original
emphasis). He then asks, rhetorically:
But if the existence of European tales about ladies who became pregnant after eating magical fish
is not now held to imply that Europeans are, or were, ignorant of the facts of physiological
paternity, why should such stories have this implication in the case of the Tully River Blacks?

This same rhetorical question, phrased in different ways, and directed

especially to his trump card—the Virgin Birth—is repeated a number of
times: if the magical birth of heroes and gods in the religions and myths of
higher (and especially Christian) civilization does not imply ignorance of
physiological paternity in Christendom, why—except to reinforce Western
belief in the ignorance of “natives”—should such an inference be drawn
from reports of magical conception among Australians? Hence, if we do not
accord these two cases the same status it can only be, according to Dr.
Leach, because of Western prejudice, viz.: “If we believe such things, we
are devout; if others do so they are idiots” (1967:45, original emphasis).
That Leach should not have recognized the obvious fallacy in his analogy
is especially surprising since his own statement underscores the basis for
the fallacy. In the passage quoted above he stresses that he is comparing
ordinary births among the Australians with the births of gods and heroes—
extraordinary births—among Europeans.11 And there’s the rub! If—to take
Leach’s piece de resistance—the dogma of the Virgin Birth is not held to
imply that Christians are or were ignorant of physiological paternity, it is
because this cannot possibly be its implication. Far from claiming that the
virginal conception of Jesus typifies a general norm of virginal conception,
this dogma, by holding that His birth was a miracle, makes precisely the
reverse claim; it denies that the norm of procreative conception applies in
His case. Indeed, Leach’s entire argument is misconceived because, unlike
the spirit-child belief, the Virgin Birth does not assert that Jesus had no
genitor, it asserts that he had no human genitor; and unlike the Christian
dogma, the Australian belief is not about virgin births, but about
nonprocreative births. In referring to the Christian and Australian beliefs as
“virgin birth” doctrines, Leach confuses virgin birth with parthenogenesis.
Although both of these former beliefs—like parthenogenesis itself—share
the notion that conception occurs without copulation, the similarity stops
there. The Australian case is not strictly analogous to nonhuman
parthenogenesis because, paradoxically, it is nonvirgins who conceive
parthenogenetically; in the Christian case, a Virgin (the first paradox)
conceives procreatively (the second paradox).
11. For this reason, Leach’s more telling argument would have been the Western secular belief that
children are brought by storks. Like the Australian conception belief this belief explains all—not
merely extraordinary—births, and it is even more “irrational” than the former: it reflects ignorance of
physiological maternity as well as paternity. Moreover, we know that it was transmitted (to children),
though not believed (by adults).

In both cases, however, the paradox is only apparent. In Australia,

conception does not result from copulation, but from the entry of an already
formed spirit-child into the vagina. Hence, though the mother is a non
virgin, there is no genitor. In the Christian myth there is a genitor, and
conception does result from fertilization; but the mother remains a Virgin
because fertilization occurs without copulation—procreation is aural, not
vaginal; semen is spiritual, not physical; genitor is nonhuman (God), not
human. For only on the assumption that normal conceptions are nonvirginal
can the dogma claim that the virginal conception of Jesus is a miracle
(which is why His birth is not just another virgin birth, but the Virgin Birth).
And it is precisely because it is a miracle that the Virgin Birth is a dogma of
the church, i.e. an article of faith. For, like all miracles, the Virgin Birth is
not normal but abnormal, not ordinary but extraordinary, not rational but
irrational—in a word (as Tertullian put it), absurd! And to believe in the
absurd requires a leap into faith: credo quia absurdum est.
In sum, Leach’s analogical argument is invalid. Whatever else it can
imply, the Virgin Birth cannot imply ignorance of physiological paternity; it
must imply knowledge of physiological paternity.12
12. The identical logical critique renders invalid yet another of Leach’s Christological arguments.
“Theologians who debate the doctrine of transubstantiation,” Leach (1967:48) informs us, “cannot
usefully be accused of ignorance of the elementary facts of chemistry.” Not only “usefully,” but
“logically.” For, of course, it is only because these theologians are aware of the elementary facts of
chemistry that transubstantiation is, for them, a supernatural, not a natural, event—a Mystery. By the
laws of nature, it could not happen. It is precisely because it violates what they believe to be the laws
of nature—“the elementary facts of chemistry”—that transubstantiation must be an article of faith.

But Leach’s analogical argument does not stop there. If, as he argues, the
dogma of the Virgin Birth is not held to imply ignorance of physiological
paternity it is because, being a dogma, its biological message is not taken
very seriously; rather it serves to “reinforce the dogma that the Virgin’s
child is the son of God” (1967:32). Now ignoring the logical difficulty
posed by this argument—since the “dogma” that the Virgin’s child is the
son of God is one and the same dogma, it is difficult to understand how a
dogma can reinforce itself—of course (as I would prefer to put it) the
theological function of the Virgin Birth is to legitimize the Christian claim
that Jesus is the son of God. It is not clear, however, why this implies that
the dogma of the Virgin Birth does not have a biological message, or—
since it explicitly states that Jesus was not begotten of Joseph—that this
message is not taken seriously. Indeed, how else, except by denying to
Joseph a procreative role in the conception of Jesus, can the dogma then
assign that role to the third person of the Trinity? And how else can Jesus
be divine unless He is God’s—not Joseph’s—only begotten Son? In short,
were it not for its biological message—Jesus did not have a human genitor
—and were this message not taken seriously, the theological function of the
dogma of the Virgin Birth could not be served.
Having examined Leach’s arguments on behalf of his first thesis, viz.,
that from Australian conception beliefs it may not be inferred that the
natives are ignorant of physiological paternity, we may now turn to his
arguments on behalf of his second thesis, viz., that these beliefs are not
about conception but about patrilateral filiation.

The Issue of the Meaning of Australian Conception

Beliefs

1. The Argument from the Australian Evidence

According to Dr. Leach, the “modern interpretation” of the Tully River

conception beliefs is that “the relationship between the woman’s child and
the clansmen of the woman’s husband stems from the public recognition of
the bonds of marriage.” The meaning of this interpretation is unclear. Does
it refer to the cognitive meaning of the beliefs (and, if so, to their conscious
or unconscious meaning)? Or does it refer, rather, to their sociological
function (and, if so, to their manifest or latent function)? Or does it refer to
their sociological implication, i.e., to the sociological message which they
convey, not to the natives, but to the anthropologists? However, I shall
ignore the conceptual difficulties entailed by each of these possible
meanings of the interpretation, and, instead, restrict my comments to
Leach’s arguments on its behalf.
Whatever construction we might wish to place on Leach’s interpretation
it is not (to say the least) self-evident that the proper interpretation of “a
woman begets children because she may dream of a child put inside her,” is
that “the relationship between the woman’s child and the clansmen of the
woman’s husband stems from the public recognition of the bonds of
marriage.” Hence, since this belief, according to Leach’s interpretation,
does not mean what it seems to mean—i.e., it is not about conception but
about patrilateral filiation—we should like to know something about the
evidence (or theory) upon which this symbolic interpretation is based. Is
there anything in Roth’s ethnography which suggests that this is what the
natives themselves take this belief to mean? Short of that, are there any data
on Tully River symbolism from which it might be inferred that this is its
meaning? Short of that, can it be deduced from a cross-cultural theory of
symbolism that this (whether conscious or unconscious) is its meaning?
It is rather disappointing, then, to discover that Leach presents no
evidence for his interpretation. Instead he offers one argument,13 to wit: his
interpretation is not original with him—it is found in Frazer, Malinowski,
Radcliffe-Brown, Ashley Montagu, “and elsewhere.” Disappointment
rapidly changes to surprise when it is realized that even this argument (slim
as it is) is misconceived: for, in fact, the authorities cited by Leach do not
support his interpretation. All of them agree, to be sure, that in Australia,
fatherhood (and by implication, patrilateral filiation) is sociological rather
than biological; it is this, presumably, that Leach takes to be support for his
position. For them, however, this is not the meaning of the Australian
conception beliefs, but the implication of the natives’ ignorance of
physiological paternity. Since the mother’s husband is pater, and yet—being
ignorant of physiological paternity—he is not believed to be genitor, it must
be, so they argue, because fatherhood (and, by implication, patrigroup
filiation) is grounded in sociology, not in biology.
13. Although Leach professes to offer four arguments in support of this interpreta tion, three of the
four speak to the separate issue of physiological paternity (we have examined them above). If, so
Leach seems to argue, the natives can be shown to believe in physiological paternity, it then follows
that his interpretation of their conception beliefs is true. This conclusion, needless to say, does not
follow.

2. The Argument from the Christian Evidence

Leach’s real concern, however, is not with the Australian evidence, but with
the Christian analogy, the Virgin Birth. The basic message of the Virgin
Birth—and so, too, the basic message of Australian conception beliefs—is,
according to Dr. Leach, neither biological nor theological, but sociological.
Moreover, it is identical with his interpretation of the Australian conception
beliefs. Thus, from the fact that although the dogma of the Virgin Birth
asserts that God is His father, two of the synoptic Gospels (Matthew and
Luke) provide Jesus with a pedigree which places him “in the direct line of
patrilineal descent from David through Joseph” (Leach 1967:42, original
emphasis), Leach concludes:
In other words the kind of interpretation which I put on Roth’s evidence … has been orthodox
among Christians for about 1600 years. The myth, like the rite, does not distinguish knowledge
 from ignorance. It establishes categories and affirms relationships.

For Leach, then, the message of the key dogma of Christianity—the

Incarnation—and, hence, the central message of Christianity is that Jesus’s
relationship to Joseph is sociological rather than biological! Although Dr.
Leach must be admired for his consistency—in order to defend his thesis he
not only does not shrink from, but he is even “positively eager” to embrace,
its reductio ad absurdum—his logic and theology are rather less deserving
of admiration. For, on both accounts, the argument from Jesus’s Davidic
pedigree refutes, rather than supports, this thesis.
Granted that the Gospels, concerned to provide Jesus with Davidic
descent, establish his pedigree on the sociological paternity of Joseph, the
dogma of the Virgin Birth denies physiological paternity to Joseph not to
enunciate a banal sociological principle, but to proclaim an audacious
theological truth, viz.: God, in order to save mankind, became flesh—in the
person of Jesus. If, then, the dogma of the Virgin Birth claims that God is
Jesus’s genitor, it is not to announce that Joseph is His pater, but to proclaim
that He is Saviour. But Jesus is not only Saviour (=God), he is also Christ
(= Messiah); and in the context of ancient Judaism these distinct claims
created an urgent theological dilemma for Christianity. For although, qua
Saviour, Jesus cannot (being God) have a human genitor, qua Christ, He
must have a human genitor for—according to Judaism—the Messiah is a
patrilineal descendant of David. In short, the dogma of the Virgin Birth
solved one theological problem only to create another: the very criterion by
which it established the claim that Jesus is Saviour disqualified Him for the
claim that He is Messiah. Hence, even leaving aside Leach’s aberrant
interpretation of its meaning, the structural consequence of the Virgin Birth
is precisely the reverse of what he claims it to be. Rather than “establishing]
categories and affirming] relationships” so as to provide Jesus with a
Davidic pedigree, it destroyed the only recognized (Jewish) basis for such a
pedigree—genealogical relationship.
Although, contrary to Leach, the consequence of the Virgin Birth was to
deprive Jesus of the possibility of a Davidic pedigree, both Mark and Luke
—tracing his descent, generation by generation, from David through Joesph
—provide him with a Davidic genealogy. This genealogy, however, is not
only extrinsic to the dogma of the Virgin Birth, but, as all Biblical critics
(from the earliest Church Fathers) have realized, it is clearly incompatible
with it. Troubled by this obvious inconsistency and, especially, by the uses
made of it by anti-Christian writers—who, accepting the genealogy,
rejected the Virgin Birth, or who, having always rejected the Virgin Birth,
viewed the genealogy as an attempt to rescue Mary from charges of sexual
immorality—the Church Fathers (and later Christian theologians) offered a
variety of solutions for it.14 One (but only one) of these attempted solutions
proposed that, although Joseph is not His biological father, the Gospels’
genealogy (as well as Jesus’s claim to Messiahship) can be justified on the
grounds that Joseph is his sociological father. This interpretation was
proposed, however, not as the message of the genealogy, let alone as the
message of the Virgin Birth—for the Church Fathers (and for all Christians)
nothing could be more absurd than the suggestion that the function of
Jesus’s Davidic genealogy (let alone the Virgin Birth) is to affirm this
structural principle—but as a solution to the theological difficulty which it
poses. For the Biblical commentators (and for all Christians) the function of
the Davidic genealogy, however it be reconciled with the dogma of the
Virgin Birth, was (and is) clear: it legitimizes the Christian claim that Jesus
is the Messiah. In sum, the Christological argument once again refutes,
rather than sustains, Leach’s position.15
14. For summaries of the controversies surrounding this problem of Jesuss putative Davidic
genealogy, cf., “Genealogy (Christ),” in The interpreter’s dictionary of the Bible; “Genealogies of
Jesus Christ,” in Catholic Biblical encyclopedia, New Testament; “Genealogies of Jesus Christ,” in
Hastings dictionary of Christ and the gospels.
15. Because of space limitations, I shall forbear to comment on Leach’s other interpretations of the
Christian myth, especially his speculations on the type of society which finds it most congenial. They
are irrelevant to the central issue of this controversy (and set beside the historicoeconomic
interpretations of Kautsky, the historical interpretations of Harnack, the psychoanalytic
interpretations of Freud and Fromm, and the cultural interpretations of Warner, they are, to say the
least, unconvincing). Leach himself is not unaware of their irrelevance; in the midst of these
speculations he says—quite correctly—“Professor Spiro will be wholly unimpressed. He will ask:
But how can you prove that these associations of facts are relevant?” (Leach 1967:44, original
emphasis). Surprisingly, he does not attempt to indicate their relevance; indeed, he is disarmingly
frank about their irrelevance. Thus, he writes, “Well, quite frankly, I don’t claim to prove anything at
all” (1967:43); and, “How do I know that such patterns are significant? I don’t. I find them
interesting” (1967:45, original emphasis).

In summary, it may be concluded that none of Dr. Leach’s arguments

stands up under scrutiny. His spurious racist argument aside, the Australian-
based evidence is the reverse of what he purports it to be, and the
conclusions to be drawn from his Christological evidence are the reverse of
what he believes them to be. Since, then, his arguments are refuted by his
own evidence, his contentions—that the Australians are not ignorant of
physiological paternity, and that their conception dogmas enunciate a
structural principle—remain unsupported.

Alternative Models of Culture and Cultural

Interpretation

1. Leach’s Structural Interpretation

If this controversy were primarily concerned with the proper interpretation

of arcane ethnographical facts, this article would have reached its
conclusion in the previous section. But this controversy is not primarily a
disagreement over fine points of Australian ethnography (although, to be
sure, it stems from such a disagreement): over-shadowing this is a much
more important controversy, for which the ethnographical facts are merely a
medium of intellectual exchange. The latter controversy is theoretical and
methodological in character. It involves such issues as the nature of culture,
how it works, what its functions are, what explanatory variables must be
attended to in attempting to interpret any of its manifestations. The better to
deal with these issues, let us assume that the ethnographical reports are
silent on what the natives believe concerning physiological paternity; let us
assume, rather, that our only datum concerning this matter is the cultural
belief in spirit-children. On this assumption, then, let us turn from Leach’s
arguments on behalf of his interpretation, and examine, instead, its
conceptual structure, comparing it with the conceptual structure of its rival
interpretation.
Placed in the context of Leach’s lecture, it is somewhat difficult to decide
whether his interpretation—the basis for patrilateral filiation is the public
recognition of the marriage of the child’s parents—of the Australian spirit-
child belief refers to the cognitive meaning of this belief, or to its
sociological implication. Sometimes, depending on his specific arguments,
it seems to refer to the one; sometimes to the other. Nevertheless, however
it be construed, this interpretation encounters serious conceptual difficulties.
If, for example, his interpretation has reference to the cognitive meaning of
the belief, it obviously cannot refer to the meaning of its manifest content,
for the latter is patently concerned with conception. Since the manifest
content, moreover, enunciates a nonprocreative explanation for conception,
this interpretation is inconsistent with Leach’s claim that the natives are not
ignorant of physiological patebnity. Hence, if his interpretation refers to the
cognitive meaning of the belief, it must have reference to the meaning of its
latent content. If so, it poses some difficult functional and motivational
problems.
First, why is not this sociological message expressed directly? What
function can possibly be served in expressing it by means of a symbolic
circumlocution? Moreover, if structural principles (for some unknown
reason) are best expressed (or enunciated) symbolically, what function can
be served by enunciating this particular principle in the symbolism of
conception? Second, since ex hypothesi, the natives are not ignorant of
physiological paternity, what can possibly be the function of expressing this
message by means of symbolism, the meaning of whose manifest content is
the opposite of what they really believe? To say that the natives transmit,
generation upon generation, a conception belief in whose manifest meaning
they do not believe, so that they may enunciate a structural message
contained in its latent content, and in which they do believe—this can only
make sense on the assumption that the latter message, being painful, has
undergone repression, and can only be expressed by means of unconscious
symbolism. If, then, Leach’s interpretation of the belief refers to its
cognitive meaning, does he mean to say that this structural message is
indeed unconscious? If so, he must explain why such an innocuous message
is repressed.
Even if these functional problems were resolved, this interpretation,
when combined with Leach’s claim that the natives are not ignorant of
physiological paternity, raises a serious motivational problem. For observe:
if the natives are ignorant of physiological paternity, then, in default of any
knowledge concerning the biological basis for paternity, father, in this
cultural context, can only be pater. Consequently there is no alternative to a
sociological basis for patrilateral filiation: like the sociological basis for the
relationship between mother’s son and mother’s husband, the sociological
basis for the latter relationship, too, is a necessary consequence of
biological ignorance. But if, as Leach holds, the natives are not ignorant of
physiological paternity, the culturally designated basis for paternity and for
patrilateral filiation is the consequence not of default, but of choice: the
choice represents a culturally designated preference between two equally
known alternative bases—patership or genitorship (or any combined
weighting of these). Hence, if Leach holds that his postulated sociological
basis for patrilateral filiation is the structural consequence of the
sociological basis for fatherhood, then since, ex hypothesis the natives know
that father is genitor, what could possibly be their motive for rejecting
genitorship as the basis for fatherhood? (Moreover, if it is the consequence
of this rejection, the cultural belief enunciates a superfluous message; the
message necessarily follows from the rejection.) If he holds, on the other
hand, that genitorship is rejected as the basis for fatherhood in order to
establish paterhood as the basis for filiation, what could possibly be the
motive for that choice?
Perhaps, in view of these unresolved difficulties, Leach’s interpretation
refers not to the cognitive meaning of the cultural belief, but to its
sociological implication: that patrilateral filiation is based on the father’s
role as pater is not, then, the message of the belief, but its function. Hence,
on the basis of this cultural belief, the anthropologist can infer that one of
the structural features of the society in which it is embedded (and which it
reflects) is that patrilateral filiation is sociological. Aside from the fact that
we are then left without a cognitive meaning for the belief—and unless we
know its meaning how can we assess its sociological implications?—this
interpretation, like the first, is incompatible with Leach’s claim that the
natives are not ignorant of physiological paternity.
If (because its manifest content enunciates a nonprocreative explanation
for conception) the conception belief is taken to imply ignorance of
physiological paternity, father is necessarily pater, and the basis for
patrilateral filiation in such a society is necessarily sociological. (Since this
conclusion is entailed by the premise, if the premise is true, the probability
of the conclusion being true approaches certitude.) If, however, following
Leach, it is assumed that the natives are not ignorant of physiological
paternity, then (even when descent is matrilineal) how can this belief
possibly imply that in such a society patrilateral filiation is sociological? If
the natives are aware of physiological paternity, there is no basis for arguing
that the conception belief implies that fatherhood and, therefore, patrilateral
filiation are sociological. From this cultural belief alone, there is no basis
whatsoever for contending that this is its implication. (If, in fact, the
premise were true—i.e., if the natives were not ignorant of physiological
paternity—the probability of the conclusion being true would be no greater
than chance. For, when physiological paternity is known, the culturally
assigned basis for fatherhood might be either biological or sociological.)
No explanation, of course, is required to deal with all of the difficulties
raised by its critics. But any explanation which merits a serious hearing
must, at the very least, be free from internal logical contradictions. And any
explanation of culture, which is more than a private assertion, must be
consistent (if not derived from) those principles which are believed to
govern human behavior. Leach’s explanation is vulnerable on both scores.
This being so, I should like to examine its rival explanation which, I submit,
obviates both difficulties.

2. An Alternative Functional Explanation

As in the case of my examination of Leach’s explanation, I am concerned,

not with the truth of its alternative—so far as I know the evidence is lacking
for testing the truth of either—but with its conceptual structure. Is it
internally consistent? Is it consistent with what we know about human
society and personality? The present explanation rests on three
assumptions.
Since the Australian conception belief states that conception is caused by
the entry of a spirit child into the mother, and since—in the absence of any
evidence which indicates the contrary—it is gratuitous to assume that this
cultural belief does not mean what it says, it would seem not unreasonable
to assume that it enunciates a theory of conception. Since, moreover, it is a
nonprocreative theory, and since—in the absence of any evidence which
indicates the contrary—it is gratuitous to assume that the natives do not
believe what it says, it would seem not injudicious to assume that the belief
implies that the natives are ignorant of physiological paternity. On the basis
of these two assumptions, and on the basis of the further assumption that
the Australian natives are as intelligent as any other natives—the
Americans for example—the Australian conception belief is amenable to
two different but equally valid interpretations. Although sharing these three
assumptions, these interpretations disagree on the basis for the natives’
ignorance, the motive for their belief in the conception theory, and the
psychological function of the theory.
The first interpretation, deriving from Malinowski, holds that the basis
for the natives’ ignorance of physiological paternity is cultural: for a variety
of perfectly rational reasons (both logical and empirical)—and spelled out
by Malinowski and Ashley Montagu—knowledge of the causal relationship
between sexual intercourse and conception is absent from their cultural
inventory of biological knowledge. Hence, in default of a procreative
explanation, the conception belief enunciates a nonprocreative theory of
conception. This being so, the function of the message is cognitive-
explanatory, and the motivational basis for the natives’ belief in the
message is intellectual curiosity concerning an otherwise inexplicable
phenomenon, and one which is of central concern.
This interpretation poses no conceptual difficulties. The interpretation of
the message of the belief is consistent both with its manifest content and
with the inference that the natives are ignorant of physiological paternity;
and all three are consistent with the postulated function of, and the motive
for belief in, the message. From this interpretation, moreover, it may be
deduced that patrilateral filiation is sociological (without falling into the
logical traps entailed by Dr. Leach’s interpretation): since, ex hypothesi, the
natives are ignorant of physiological paternity, there is no alternative to a
sociological basis for pa-trilateral filiation. According to this interpretation,
however, the sociological basis for paternity (and hence for filiation) is
neither the message nor the function of the spirit-child theory (although it is
its sociological implication); it is, rather, the unintended (though
recognized) function of the natives’ ignorance of physiological paternity.
Although this interpretation poses no conceptual difficulties, it raises a
serious ethnographical problem. Since most peoples seem not to be ignorant
of physiological paternity, it is somewhat difficult to understand why the
Australians, too, have not discovered the causal link between sexual
intercourse and conception.16 This difficulty is the point of departure for a
possible second interpretation. Distinguishing between two kinds of
ignorance—one based on the absence of knowledge, the other based on its
rejection—this interpretation, deriving from Jones (1924), argues that the
basis for the natives’ procreative ignorance is not cultural, but
psychological; i.e., it is based not on the absence of biological knowledge,
but on its rejection. If so, the natives’ procreative ignorance is motivated—
motivated by the wish to deny physiological paternity; and the spirit-child
belief enunciates a nonprocreative theory of conception, not in default, but
in lieu, of a pro-creative explanation.17 This rather strange interpretation
must be explicated.
16. Dr. Leach, quite properly, raises this same problem. For Leach, however, the reported evidence
for such ignorance does not constitute a scientific problem, but a problem, rather, in the psychology
of science. “If,” he argues, “certain peoples have persuaded their ethnographers that they were
ignorant of the facts of life then it is because ‘ignorance’ was for these people a kind of dogma. And
if the ethnographer in question believed what he was told, it was because such belief corresponded fo
his own private fantasy of the natural ignorance of savages” (Leach 1967:41).
Since I do not share his psychological assessment of a large group of anthropologists, I do not believe
that the scientific problem is so easily dismissed. Thus, if the natives really believe that women
conceive because they are entered by spirit-children, the psychological problem is not, pace Leach,
why are they so stupid and idiotic? It is, rather, on what grounds, which they accept as evidential, do
they believe it to be true? And if, in all deference to Malinowski et al., we have reason to suspect that
the natives do not “really” believe that this is the cause of conception, then the psychological
problem is not, pace Leach, on what motivational grounds were the ethnographers “taken”? It is,
rather, on what functional or motivational grounds do the natives perpetuate a theory which, on some
level, they believe to be false?
17. That the natives are (on some level) aware of physiological paternity might be argued from a
symbolic interpretation of some of the ethnographical data. Thiis, for example, the spirit-child is
typically an ancestor; and ancestors, of course, are fathers two or three times removed. Again, the
Trobriand child, Malinowski informs us, is believed to look like his father—but never like his
mother. Trobrianders, too, are filled with embarrassment should the discussion of parental intercourse
be touched upon. More important than these symbolic interpretations, however, is the report of
Roheim (1932) that the procreative awareness which he found explicitly in young children is
“repressed” following initiation.

That a son should wish to reject knowledge of the fact that his father is
his genitor is not, of course, a strange notion in the annals of child
development. One explanation for this frequently found wish is based on
the assumption—derived from psychoanalytic theory, and supported by a
great deal of empirical evidence—that fathers are both loved and hated, and
that the latter emotion derives from one or both of the following conditions:
resentment over their punitive authority, and/or jealous rivalry for the love
(sexual and/or affectionate) of the mother. But hatred of the father leads to a
typical Oedipal conflict. On the one hand, the child, motivated by
resentment or by rivalry, wishes to harm, to be rid of, the father. On the
other hand, whether from a talion fear (“I want to harm him, therefore he
wants to harm me”) or from guilt (“Since he loves me and/or since I love
him, how can I wish to harm him?”) this wish is extremely painful. In the
absence of institutional or cultural assistance in dealing with this conflict,
the child must cope with it by his own internal resources, of which I shall
mention only two. He can repress his hatred, which is the typical (and
normal) technique found in Western society, or he can express it
symbolically by denying in fantasy that his father is his genitor. (The latter
is often accomplished, both in private fantasy as well as in hero myths, by
the substitution of grandiose fathers—gods, kings, and so on—for the real
father.) Sometimes, it should be added, rather than denying that his father is
genitor, the child denies that he had any genitor: his conception was not
caused by procreation. (For a discussion of the mechanism of denial, cf.
Freud 1946: chaps. 5–7.) Both of these defense-mechanisms resolve the
conscious conflict, but in different ways—the former by rendering the
child’s hatred unconscious, the latter by expressing it by means of a
symbolic short-circuit.
Not all conflicts, however, find resolution by the unaided, intrapsy-chic,
efforts of the individual. Many conflicts, expecially those which are shared
by social actors, are resolved by means of social institutions and cultural
beliefs. Indeed, just as culture has cognitive-explanatory functions, so it has
(among others) affective-integrative functions. My second, alternative,
interpretation of the Australian conception theory views it as having such a
function. This interpretation holds that the spirit-child theory, whatever the
basis for its origin might have been, now serves (this at least would be one
of its functions) to resolve the natives’ Oedipal conflict by providing them
with the cognitive (cultural) basis for denial. By believing in the traditional
conception theory, each Australian is implicitly, but necessarily, denying
that his father is his genitor. For since, according to this cultural theory, all
conceptions are caused by a nonprocreative method, then he too was
conceived without parental intercourse. Hence, his father is not his genitor,
although he is his mother’s husband.
This conception belief not only enables the Australian to deny that his
father is his genitor; it enables him to do so by means of a wish-fulfilling
fantasy. For since, according to this belief, conception is caused by the entry
of a spirit-child into a woman’s womb, he too was a spirit-child, and he too
was conceived by entering the womb of his mother. Hence, since he not
only came out of, but also went into, his mother’s womb, it was his entry,
not his father’s, that caused her to become pregnant. By believing in this
cultural theory, then, the boy takes his father’s place, for, according to this
theory, the son is literally the father of the man—and of the child; he is his
own genitor. By believing in this conception theory, then, the boy both
denies the painful fact that his conception was caused by the sexual
intercourse of his parents, and he fulfills his wish that he take his father’s
place.
If, then, according to the first of these functional interpretations, belief in
the conception theory is motivated by curiosity concerning the cause of
conception, it is motivated according to this second interpretation, by the
child’s pain-induced hatred of his father; and if, according to the first
interpretation, the function of belief is cognitive—it satisfies an intellectual
need—according to the second, its function is ego-integrative—it resolves
emotional conflict. For both interpretations, however, this conception belief
has the same microsociological implication, viz., it implies that father is
pater (either because, according to the first interpretation, genitorship is
unknown; or because, according to the second, it is denied); and for both,
therefore, it has the same macrosociological implication, viz., the basis for
patrilateral filiation is sociological. (For the first explanation these
implications are the unintended—but recognized—functions of ignorance
of physiological paternity; for the second they are the unintended—but
recognized—functions of denial of physiological paternity.)
According to the second interpretation, finally, belief in the conception
theory has an (unintended) interpersonal adaptive function whose
importance exceeds these (adaptively neutral) sociological functions: it
promotes a high degree of amity between father and son. Whether the
culturally assigned basis for the father-son relationship is the father’s
paterhood or his genitorship, the emotional quality of this relationship
depends on (among other things) how the son copes with his hostility to his
father. If the hostility-induced conflict is not resolved, father is, at best, the
affectively neutral mother’s husband, at worst the hated Oedipal figure. It
can be resolved, as we do, intrapsychically, by the repression of the hatred;
or it can be resolved—as suggested by this second interpretation of the
spirit-child belief—culturally, by institutionalized denial. Both techniques
of conflict resolution have their respective advantages and disadvantages.
The second technique can serve to transmute a hated genitor, not merely
into a mother’s husband—a sociological pater—but, if we can believe
Malinowski, into a warm and nurturant father—a psychological pater.
 Here, then, is a second functional interpretation of the Australian
conception belief which, like the first, is faithful to the manifest meaning of
the belief, whose internal structure creates no logical difficulties; and whose
sociological and psychological consequences pose no conceptual problems.
Which of these interpretations—if either—is true, cannot be decided from
the available data. Both are consistent with the reported data that the natives
are—or, at least, were—ignorant of physiological paternity, but they
interpret these data differently: and from the data, alone, it cannot be known
whether ignorance is—or, at least, was—based on absence of knowledge or
on denial. Theoretically, there is little to choose between them. The first
explanation is by far the more parsimonious; the second copes with the
ethnographical problem raised by procreative ignorance.
Both explanations, finally, are derived from the same model of culture, a
model whose assumptions are, if I understand him correctly, precisely the
reverse of Leach’s. In this model, to mention only those of its assumptions
that are relevant to this controversy, culture serves a number of functions—
not only sociological, but psychological (cognitive and emotional) as well;
cultural theories attend not only to the functional requirements of society,
but also to the functional requirements of social actors—hence, their
messages may relate to nature and human nature, as well as to social
structure; belief in these theories is motivated (by cognitive and/or
emotional needs)—hence their functions and meaning, alike, are often
inexplicable without an awareness of the possible motivational bases for
beliefs; finally, culture is man’s most important means of adapting to the
condition of being human—hence, since psychological tension (including
cognitive curiosity and emotional conflict) is a persistent feature of that
condition, cultural beliefs represent, among other things, the attempts of
social groups to cope with these tensions.18
18. After receiving proofs, my attention was called to still a third possible basis for ignorance of
physiological paternity (at least among the Trobrianders) in a suggestion made by Dorothy Lee. In
her ingenious reconstruction of Trobriand thought (Lee 1940), she suggests that Trobriand conceptual
categories do not allow for the perception of teleology, in any strict sense of that term; that the
cultural emphasis is on the “essential, rather than the accidental or relational” (361); that, given this
conceptual set, events are viewed as “self-contained and essentially unrelated” (360); and that serial
events are perceived to exhibit “simple sequence” rather than causal relationship. If this be so, then—
as Lee herself observes (358)—Trobriand ignorance of physiological paternity might be a function of
their (culturally based) nonteleological, cognitive set.
If her reconstruction is sound, this epistemological hypothesis—that Trobriand cognition is
concerned with things (objects, events, essences) rather than with relationships among things—is
consistent with either of the two hypotheses advanced in the text to account for ignorance of
physiological paternity. The cultural hypothesis becomes, then, a theorem which is deducible from
the epistemological axiom; and the psychological hypothesis becomes a subset of a more
comprehensive theory which would attempt to discover the psychodynamic function of a cultural
epistemology that emphasizes things over relationships.

Note
Reprinted from Man: The Journal of the Royal Anthropological Institute, vol. 3, no. 2 (June
1968):224–61. This article was written while I was a Research Fellow in the Social Science Research
Institute, University of Hawaii. It has importantly benefited from the criticisms of Robert I. Levy and
David M. Schneider.

References

Austin, L. 1934. Procreation among the Trobriand islanders. Oceania

5:102–18.
Fortune, R. F. 1963. Sorcerers of Dobu (new ed.). London: Routledge
& Kegan Paul.
Freud, A. 1946. The ego and the mechanisms of defense. New York:
International Universities Press.
Genealogies of Jesus Christ 1917. In Hastings dictionary of Christ and
the gospels. New York: Scribners.
Genealogies of Jesus Christ 1956. In Catholic Biblical encyclopedia:
New Testament. New York: Joseph T. Wagner.
Genealogy (Christ) 1962. In The interpreter’s dictionary of the Bible.
New York: Abingdon Press.
Hallowell, A. I. 1955. Culture and experience. Philadelphia: Univ. of
Pennsylvania Press.
Jones, E. 1924. Mother-right and the sexual ignorance of savages.
International Journal of Psycho-Analysis 6:109–30.
Kaberry, P. M. 1939. Aboriginal woman. London: Routledge.
Leach, E. R. 1961. Golden bough or golden twig? Daedalus (Spring)
1961:371–87.
--------- 1967. Virgin birth. Proceedings of the Royal Anthropological
Institute 1966:39–50.
Lee, D. 1940. A primitive system on values. Philosophy of Science
7:355–78.
Malinowski, B. 1932. The sexual life of savages in northwestern
Melanesia (3rd ed., with special foreword). London: Routledge.
--------- 1937. Foreword. In Coming into being among the Australian
aborigines, by M. F. Ashley Montagu. London: Routledge.
Meggitt, M. J. 1962. Desert people. Sydney: Angus & Robertson.
Montagu, M. F. Ashley 1937. Coming into being among the Australian
aborigines. London: Routledge.
Roheim, G. 1932. Psycho-analysis of primitive cultural types.
International Journal of Psycho-Analysis 13:1–224.
Roth, W. E. 1903. Superstition, magic and medicine (N. Queensl.
Ethnogr. Bull. 5). Brisbane: Vaughan.
Spencer, B. & F. J. Gillen 1899. The native tribes of Central Australia.
London.
Spiro, M. E. 1964. Religion and the irrational. In Symposium on new
approaches to the study of religion, ed. J. Helm (Proceedings of the
American Ethnological Society 1963). Seattle: Univ. of
Washington Press.
______ 1966. Religion: Problems of definition and explanation. In
Anthropological approaches to the study of religion, ed. M. Banton
(Ass. Social Anthrop. Monogr. 3). London: Tavistock Publications.
Stanner, W. E. H. 1933. The Daly River tribes: The theory of sex.
Oceania 4:26–28.
Thomson, D. F. 1933. The hero cult, initiation and totemism on Cape
York: The knowledge of physical paternity. Journal of the Royal
Anthropological Institute 63:505–10.
Warner, W. L. 1937. A black civilization. New York: Harper.
10
Whatever Happened to the Id?

There has been a strong tendency in structural and symbolic anthropology to assume that sex and
aggression are of no concern to cultural symbol systems. Even when cultural beliefs, myths, or
rituals are explicitly and preponderantly sexual or aggressive in content, they are typically
interpreted as metaphors for social structural themes. This thesis is illustrated with respect to
aggression by an analysis of Lévi-Strauss, interpretation of a Bororo myth, after which the
assumptions that structural theory makes concerning the place of aggression in cultural symbol
systems are contrasted with the opposing assumptions of psychoanalytic theory
[STRUCTURALISM, PSYCHOANALYSIS, CULTURAL SYMBOL SYSTEMS, MYTH].

I HAD INTENDED, when asked to address the subject of anthropology and

psychoanalysis for this symposium, to write on the problem of
hermeneutics, which I see as the important meeting ground for
anthropological and psychoanalytic theorizing. Despite my intention, I have
instead addressed an old war-horse, the relationship between nature and
culture. I have chosen this subject because, in reviewing the literature on
structural and symbolic anthropology in preparation for the hermeneutic
paper, I came increasingly to realize that the received opinion in many
quarters of cultural anthropology holds that the body, or its drives, or the
affects and motives to which they give rise—but most especially those
related to sex and aggression—are seldom the concern of cultural symbol
systems. If the latter appear to be concemed with sex or aggression, it is the
job of the anthropologist to uncover the reality behind the appearance.
Although examples are abundant, I can here mention only a few.
I would instance, for example, Leach’s contention (Leach 1967) that the
denial of physiological paternity in Australia and parts of Melanesia, or the
denial of a human genitor to Jesus in parts of Christendom, are statements
not about biological sex but about rules of descent. Again, I would instance
the contention of Mary Douglas that the (widely held) belief that males are
endangered by the vagina and vaginal fluids is not a belief about the “actual
relation of the sexes” but is, rather, a symbol of the hierarchical structure of
the social system (Douglas 1966:4), or that rituals concerning excreta,
breast milk, saliva, and other bodily emissions are concerned not with the
body but with the “powers and dangers credited to the social structure,” for
which the body is a symbol (1966:115); or that rituals of genital bleeding,
such as subincision rites, are concerned not with sex, or blood, or the penis
but with society: “What is being carved in human flesh [the penis] is an
image of society,” and when they are performed by tribes with moie-ties
these genital mutilations “are concerned to create a symbol of the symmetry
of the two halves of society” (1966:116).
Now, what is even more remarkable than the Leach-Douglas theory itself
is that it is widely accepted as received anthropological wisdom. It is all the
more remarkable because this counterintuitive theory is presented by its
proponents as a self-evident truth requiring no support other than that of
assertion. Consider the following examples, taken at random from Douglas.
Item: the notion that beliefs concerning sexual pollution may actually be
concerned with sexual pollution is simply “implausible” (1966:3). Item:
“we cannot possibly” take rituals concerning excreta, milk, and the like to
be in fact concerned with these bodily fluids and emissions (1966:115).
Item: “I insist” that the seeming “obsession” of the Yurok with notions of
pollution must be related to the “fluid formlessness” of their social structure
(1966:127). That is all. No argument or evidence is offered in support of
those contentions.
It used to be that one had to be a Freudian to believe that nonsexual
themes in ritual and myth might possibly be viewed as disguised
expressions of sexual concerns. It now appears that only a Freudian might
believe that undisguised sexual themes might be expressions of sexual
concerns. And not only sexual. For the wide influence of Lévi-Strauss’
theory of myth also suggests that only a Freudian might believe that
explicitly aggressive themes in myth are really concerned with aggression.
Indeed, it was my recent reading of that marvelous book The Raw and the
Cooked that finally stimulated me to change the topic of this paper from
hermeneutics to the cultural relevance of sex and aggression. Since,
however, I have dealt elsewhere (Spiro 1968a,b) with the Leach-Douglas
denial of sex as a relevant variable in belief systems, I shall concentrate
here on Lévi-Strauss’ denial of aggression in myth.
 Although most readers will doubtless recall the Bororo myth (Lévi-
Strauss 1969:35–37) that constitutes the key text for The Raw and the
Cooked, I shall briefly summarize it. One day, when the women went into
the forest to gather palms, a certain youth secretly followed his mother and
raped her. Discerning what had happened, and “anxious to avenge himself,”
the youth’s father sent his son on a series of dangerous undertakings
intended to cause his death, but the son repeatedly eluded the traps (with the
assistance of helpful animals). Finally, the father acted more directly. After
ordering his son to climb to the top of a cliff to capture some birds, he
overturned the pole by which the son had ascended, abandoning him to
death. After many tribulations the son managed, however, to return to the
village. One day, when his father was hunting, the son “full of thoughts of
revenge,” saw his opportunity. Donning some false antlers and transforming
himself into a deer, he “rushed at this father with such ferocity that he
impaled him on the horns.” He then galloped to a lake inhabited by
carnivorous fish, and dropped his father into the lake, where he was
devoured by the fish. Returning to the village, the son then “took his
revenge” on his fathers wives, including his mother. (We are not told what
form the “revenge” took, and we can only guess that it was either rape or
murder.)
According to Lévi-Strauss, that myth, the central text of the book, is an
etiological myth that explains the origin of cooked food. Conceding that “to
all intents and purposes” that theme appears to be absent from the myth,
Lévi-Strauss contends (1969:64) that it is nevertheless “concealed” in it,
and in a remarkable tour deforce, combining staggering erudition with
astonishing brilliance, he attempts in the remaining 300 pages of his book to
demonstrate the truth of this startling interpretation. Briefly, since an older
version of this myth ends with the son’s declaring that he will take revenge
on the entire clan of his father by sending cold, wind, and rain, this
indicates, so Lévi-Strauss claims, that on its surface the myth deals with the
origin of storms and rain.1 On the premise, then, that water is (on a number
of dimensions) the inversion of fire, he then proceeds to show, by means of
a succession of complex and ingenious structural analyses of a corpus of
South American myths, that this Bororo myth can be interpreted as a
transformation by inversion of a set of Ge myths that explicitly explain the
origin of cooking fire.
 1. In fact, the text itself does not support that contention. When the son, following abandonment by
his father, returns to the village, he reveals his identity to his grandmother (who had consistently
befriended him against his father) and on that night there occurred a “violent wind, accompanied by a
thunder storm which put out all the fires in the village except the grandmother’s” (Lévi-Strauss
1969:36). Now, if wind and thunderstorm already occur in the middle of the myth (and their
occurrence is not attributed to the boy’s creation), it is difficult to understand how the older version,
which ends with the boy taking his revenge by sending cold and rain, can be interpreted as their
origin. This would seem to signify, rather, that the boy had the power to control the rains, not that this
was their origin. Implicitly aware of this criticism, Lévi-Strauss defends his interpretation by noting
(1969:137) that (in their commentary) the collectors of the myth state that the natives themselves
claim that the rain that falls in the middle of the myth was the origin of rain. This hardly proves,
however, that the origin of rain is the central concern of this myth.
Indeed, to the extent that the text itself has any etiological reference, it has to do with the origin of a
certain type of aquatic plant. Thus, after the boy’s father was devoured by the carnivorous fish, “All
that remained,” the text continues, “were the bare bones which lay on the bottom of the lake, and the
lungs which floated on the surface in the form of aquatic plants, whose leaves, it is said, resemble
lungs” (1969:37).
Now, even granting the validity of that transformational hypothesis,2 I
would nevertheless have thought that violent Oedipal conflict was the
central theme of a myth that begins with a son’s raping his mother,
continues with the fathers numerous attempts to kill the son, proceeds to the
son’s brutally killing his father, and ends with his killing or raping
(depending on correct interpretation of the ambiguous term “revenge’) his
mother and stepmothers. That its central theme is taken instead to be the
origin either of cooking fire or of rain—even granting that an etiological
theme is one of its elements—is surely a remarkable reversal of figure and
ground.3 At least it seems to be remarkable until one comes to realize that
the code constructed4 by Lévi-Strauss for the translation of these myths
precludes the possibility that violence of any kind might be taken as a
mythic concern; instead, the code converts all acts of violence into
metaphors for nonviolent social structural relationships.
2. If, as I suggest in the note above, the Bororo myth is not concerned with the origin of rain, the
transformational hypothesis is invalid, since it rests on a set of putative structural relationships that
obtain between fire and water (rain), Lévi-Strauss, perhaps anticipating this problem, contends
(1969:137) that, even without the transformational argument, the Bororo myth in itself can be seen as
dealing with the origin of fire. Since the storm extinguished all the fires in the village except that of
the boy’s grandmother, the boy “becomes the master of fire, and all the inhabitants of the village
must apply to him to obtain firebrands with which to kindle the lost fire. In this sense the Bororo
myth also relates to the origin of fire, but by a process of omission.”. That interpretation seems to be
adventitious, however, because, even assuming that the boy was living with his grandmother, there is
no indication that the villagers applied to him for fire, rather than to his grandmother; indeed since he
was still uninitiated, the former hypothesis would seem unlikely. Nevertheless, fifty pages later, Lévi-
Strauss escalates his claim: “The Bororo myth creates water in order to destroy fire or, more
precisely, to allow the hero [the boy] to become the master of fire” (1969:188).
 3. It is remarkable for at least three reasons. First, the occurence of the rain is a minor (two-line)
episode in a very long myth concerned primarily with the relationship between the boy and his
parents. Second, even if we accept that the natives say that the rain in the myth was the origin of rain,
the commentators do not contend that they take the myth to have an etiological aim. Third, even the
proforma “And this is the origin of x” (often appended to a myth as a device for attributing didactic
import to an otherwise nondidactive narrative) is not found in this case.
4. I say “constructed” because it is important to emphasize that this code is not discovered by
structural analysis of the myths but is an invention of Lévi-Strauss. Although clearly ambivalent
about the issue, Lévi-Strauss admits that this may be so. Thus, in addressing the problem, he begins
by claiming that his code “has neither been invented nor brought in from without. It is inherent in
mythology itself where we simply discover its presence” (1969:12). (That is reminiscent of
Michelangelo’s modest claim that, rather than creating a sculpture ex nihilo, he merely reveals the
sculpture that is inherent in the uncarved marble—which, of course, is true, except that for n
sculptors there are n sculptures inherent in the marble.) In a later passage, however, Lévi-Strauss
concedes that that may not be the case: “It is in the last resort immaterial whether in this book the
thought processes of the South American Indians take shape through the medium of my thought, or
whether mine take place through the medium of theirs” (1969:13).

Take, for example, the son’s rape of his mother. In the first step in code
switching, that episode is classified by Lévi-Strauss as “incest,” i.e., its
sexual dimension remains but its violence dimension disappears. In the
second step the sexual dimension also disappears, for “incest” becomes a
metaphor—voila!—for the boy’s dependent attachment to his mother. Thus,
since the palms collected by the women in the myth are of the type used for
penis sheaths, that signifies that the son had reached the age of initiation,
when boys are expected to transfer from their mother’s hut to the men’s
house. In such a sociocultural context, the mother-son “incest” (the code
word for the boy’s rape of his mother) signifies the son’s objection to the
“loosening of the maternal bonds” and his “return to the mother’s bosom at
a time when other sons are about to be weaned for good” (1969:56).
As if to emphasize that this translation is not a structuralist spoof, Lévi-
Strauss clinches his argument with the following piece de resistance. Since
the adult hero of a cognate myth, the putative homologue of the boy in the
present myth, is nicknamed “Secluded,” and since, by structuralist
assumptions, the boy can be assumed to have had the same nickname, and
since a secluded boy in the sociocultural context alluded to above is one
who “refuses to be separated from female society,” it follows that the son is
“the sort of boy who, as we say, ‘clings to his mother’s apron strings’”
(1969:57).
And what about the behavior of the boy’s parents? By a similar technique
of code switching, the violent acts in which the mother and father are
respectively involved with their son signify the polar differences in their
“attitudes” toward him. Thus, having classified the son’s rape of his mother
as incest, the mothers “incestuous” relationship with the son is taken to be
indexical of her “close” attitude to the son, while the father’s “murderous”
relationship is indexical of his “remote” attitude to him (1969:138–39).
Thus it is that, by the alchemy of code switching, aggression is simply
abolished as a mythic category: the son’s sexual assault on his mother is
transmuted into his dependency upon her, while the mother’s victimization
by the son’s assault as well as the father’s assaults upon the son are
transmuted into positions on a scale of psychological distance from the son.
Lest I be misunderstood, I hasten to emphasize that I am not un-mindful
of the fact that Lévi-Strauss is not so much interested in the content of
myths, per se, as in their logical structure—“the system of interrelationships
to which we reduce them” (1969:12)—and that, since that aim is achieved
by the structural comparisons of a large corpus of myths, it must be pursued
by means of a code whose concepts are sufficiently abstract to subsume the
concrete concepts of the codes employed by the myths themselves. Since,
however, his code consists not of the formal and content-free symbols of
mathematics and symbolic logic, but of concepts only slightly more abstract
than those of the myths he analyzes, the fact that Lévi-Strauss chooses
concepts that systematically eliminate every theme of violence from these
myths indicates that this consequence is principled rather than fortuitous.
An excellent case in point is the conversion of rape (“incest”) and murder in
the Bororo myth into the attitudes “close” and “remote.”
The attitudes “close” and “remote” are attributed to the parents in the
Bororo myth in order to show that “family attitudes” is one variable in a set
of variables by which this myth can be inferred to be the inversion (and
hence the transform) of the Ge myth. In the latter myth, a young boy who is
left by his brother-in-law to die in the wilderness is rescued by a jaguar,
who then brings him to his own home (thereby becoming his “adopted
father”). The jaguar’s wife (therefore, the boy’s “adopted mother”) dislikes
the boy and tries to kill him, but, encouraged by the jaguar, the boy kills her
instead. As Lévi-Strauss sees it, the parental attitudes exhibited in the
Bororo myth are an inversion of those exhibited in the Ge myth: in the
former the mother is “close” and the father “remote,” whereas in the latter
the father is “close” and the mother “remote.” The dimension close/remote
is a perfectly good dimension for describing attitudes, but there are at least
three reasons for rejecting the assumption that this is the dimension along
which the opposite-sex parents in these myths sustain the logical relation of
binary opposition within each myth and that of identity across the myths.

(a)That the attitudes of the (adopted) Ge parents toward their son exhibit
a relation of binary opposition is self-evident: the father saves the
boy’s life, while the mother attempts to kill him. Nevertheless, even on
the dubious assumption that “close” is an appropriate term to
characterize the attitude expressed in a father’s rescue of a son’s life, to
characterize as “remote” the attitude expressed in a mother’s attempt to
kill her son is surely to distort the meaning of the text beyond
recognition. These designations not only distort the meaning of the text
but are not necessary in order to show that binary opposition in the
parents’ attitudes. Thus, for example, the dimension life-giving/life-
destroying (or some such terms) is not only faithful to the text but
denotes much more effectively the relation of opposition that
characterizes the attitudes of the Ge father and mother to their son.
(b)The dimension close/remote is even less applicable to the Bororo
parents than to the Ge parents. If “remote” which converts lethal
intentions into indifference, is an inappropriate designation for the
murderous father, “close” is entirely misplaced when applied to the
mother. For, whether we remain faithful to the text (and describe the
mother as the victim of her son’s rapacious assault) or distort the text
(and describe her as the object of the son’s “incest”), how can the
attitude “close” or any other attitude be applied either to an involuntary
victim (rape) or a passive object (incest)? If, nevertheless, the
close/remote dimension is still held to be applicable to the mother, I
would have thought that “remote” would be the more appropriate term
because, presumably, one is raped only after resisting (= “remote”)
seduction. But if the mother is “remote,” the Bororo parents cannot be
contrasted along the close/remote dimension unless we are prepared to
designate the father as “close.” Rejecting such sophistry, a relation of
binary opposition can be validly attributed to the Bororo parents only
by comparing them (as I suggested in the case of the Ge parents) along
some dimension that is faithful to the violent themes of the myth itself.
That can be done in one of two ways. Restricting the comparison to the
episodes described in the myth, we can draw a contrast between the
mother, as the victim of violent (sexual) assaults by her son, and the
 father, as the agent of violent assaults on his son. Or, going beyond the
episodes described in the myth, we can contrast the parents along the
same dimensions employed in characterizing the Ge parents: the
mother (having borne the son) is life-giving and the father (being
murderous) is life-destroying. From a structuralist point of view, and
consistent with Lévi-Strauss’ aim of showing that the Bororo myth is a
transformation of the Ge, that has the additional advantage of
demonstrating that the attitudes of the parents in the Bororo myth are
an inversion of those exhibited by the parents in the Ge myth.
(c)Contrary to Lévi-Strauss, a comparative analysis of the two myths
does not permit the conclusion that the Bororo mother and Ge father
sustain the logical relation of identity on the close/remote dimension.
For, even granting the dubious classification of incestuous rape as
“incest,” by what logical criterion is the attitude “close” attributed both
to a passive object of incest (the Bororo mother) and an active agent of
nurturance (the Ge father)? If, moreover, the classification of
incestuous rape as “incest” is invalid (as I deem it to be), the
attribution of an identity relation to the Bororo mother and Ge father
would seem to be fallacious. How can the same attitude (“close”) be
attributed to two parents of which one is the victim of a son’s
rapacious assault and the other a protector of the son against assault?
On either account, when they are compared with respect to the
close/remote dimension, it must be concluded that, rather than
sustaining a relation of identity, the relation of the Bororo mother and
Ge father is one of binary opposition. If they are to be attributed
instead with a relation of identity—and if, therefore, the parents of the
Bororo myth are to be seen as an inversion of those in the Ge myth—
that can only be accomplished by concepts that are faithful to the
violent themes found in these myths. Thus, for example, use of the
dimension violent/nonviolent (restricting the comparisons to the
episodes described in the myths themselves) or the dimension life-
giving/life-destroying (going beyond the episodes described in the
myths) makes it apparent that the opposite-sex parents sustain a
relation of binary opposition within the myths and one of identity
across the myths: hence, the relation of the parents in the Bororo myth
is an inversion of that of the parents of the Ge myth.
 By this time the reader may be wondering why I stated at the outset that I
had changed my topic from hermeneutics to the relationship between nature
and culture when, thus far, this chapter has been devoted entirely to the
interpretation of cultural symbol systems. The paradox is more apparent
than real, however. It should be evident by now that the two codes
examined (one that systematically and totally denies that sex and aggression
in ritual and myth really signify sex and aggression and one that allows that
they may indeed signify sex and aggression) may seem to differ in
hermeneutic assumptions about the role of metaphor in cultural symbol
systems, but actually differ in psychological assumptions about the
importance of sex and aggression in human nature following the transition
to culture.
What might be the assumptions of a code that interprets the Bororo myth
(which on its surface explaines why it is that a boy became a vicious
parricide) to be an origin myth, whether of cooking or of rain, a code by
which a boy’s rape of his mother signifies his dependent attachment to her,
the mother’s being the victim of his rape signifies her close attitude to him,
and the father’s attempts to murder his son signify his (the father s) remote
attitude to him? Such a code, I would suggest, can be based only on an
assumption that, in a state of culture, aggressive motives of this type are
either transcended or markedly muted so that they cannot possibly be the
real concern of myths.5 Hence it is that, if aggression seems to be the
dominant theme of the Bororo myth, its aggressive episodes must be
interpreted to mean something other than (if not the very opposite of) what
they seem to mean.
5. The same assumption underlies Douglas’ (though not necessarily Leach’s) elimination of sex as
a concern of cultural belief systems.
In making this suggestion, I am aware, of course, that Lévi-Strauss
contends (1969:12) that his interpretations of myths refer to their
“unconscious formulations.” That contention, however, does not affect the
validity of my suggestion, because, for Lévi-Strauss, the “unconscious”
meaning of a myth or a mythic episode is not its trivial meaning, but its
most fundamental meaning. Hence, when he interprets a myth of Oedipal
violence as (unconsciously) signifying the origin of cooking fire, he is not
making the trivial claim that the latter element is also to be found in the
myth; rather, he is making the important claim that this myth, which seems
to be about violence, is really about something else. Consider, then, that in
Lévi-Strauss’ code, aggression is systematically and gratuitously precluded
as the (unconscious) meaning of any of the violent episodes in these myths.
Consider, too, that alternative codes (such as those suggested above), whose
(unconscious) interpretations are faithful to the violent content of the myths,
achieve the aims of structural analysis without the dubious logic and textual
distortions often required by Lévi-Strauss’ interpretations. Consider, also,
that for Lévi-Strauss, their unconscious meanings are the fundamental
meanings of myths. Considering all of this, it seems reasonable to conclude
that Lévi-Strauss’ elimination of violence as a mythic category, although
applying to the unconscious meaning of myths, is based on the assumption
that violent motives of the type exhibited in the Bororo and Ge myths
cannot be their real concern.
That conclusion is supported by his contention (1969:164)—although he
attributes the notion to native thought—that cooking marks the transition
from nature to culture and that the Bororo myth therefore explains the
origin not merely of cooking but also (and by that very same fact) of
culture. That is not only, he argues, because of the analogy raw:cooked:
:nature:culture but also because in the Ge myths both useful and ornamental
objects are manufactured from the inedible parts of the cooked plants and
animals. That he, therefore, interprets all of the aggressive episodes in the
Bororo myth metaphorically suggests that this exegetical maneuver is a
solution to a painful dilemma. It is as if he were saying that, if this myth is
really concerned with filial rape, parracide, and filicide, the origin of
cooking cannot represent the transition to culture, because aggressive
motives of this kind, although found perhaps in a state of nature, cannot
possibly persist in a state of culture; but, on the other hand, if this myth
does not explain the origin of cooking, and if cooking does not represent the
transition to culture, the entire argument of the book would collapse. It is
this dilemma that is resolved by a code that interprets all aggression in the
myth as a metaphor.
It will come as no surprise—turning finally to the psychoanalytic
dimension of this paper—that the assumptions of Freud concerning the
relationship between aggression and culture are the polar opposite of those
of Lévi-Strauss. In Freud’s view, aggressive motives are as strong in a state
of culture as in a state of nature. Given that assumption, the themes of
parricide, filicide, and the like, although really signifying parricide and
filicide, would not in themselves disqualify the Bororo myth from being
interpreted by Freud as a myth of the origin of culture. Nevertheless, even
granting that this myth explains the origin of cooking fire, Freud would
never agree that it depicts the origin of culture, as can easily be seen from
even a cursory glance at Freud’s own origin myth. That myth (Freud
1971a), although remarkably similar to the Bororo myth (both are
concerned with violent conflicts between father and son over the sexual
possession of the wife-mother), is its structural inversion. In Freud’s myth
(which, however, he took to be history), the sons desire the father’s wives,
and since he will not share them, the sons kill the father in order to take
possession of them. Following their patricide, however, the sons experience
feelings of remorse that lead them to atone for their deed by renouncing
their claims on the women and to institute prohibitions against further
parricide, fratricide, and incest.
For Freud, then, it is not the acquisition of cooking fire, nor even the
manufacture of useful and ornamental objects attendant upon the cooking of
food, that marks the transition from nature to culture, but the institution of
norms for the control and regulation of aggression (first, within the sibling-
group and subsequently to increasingly larger groups). For Freud,
moreover, the transition to culture does not mean the transcendence of
nature. Hence, although aggressive behavior, regulated as it is by cultural
norms, may be inhibited in a state of culture, aggressive motives are not
extinguished. Indeed, it is precisely because these motives (including
parricide and filicide) persist in and are even exacerbated by culture (Freud
1971b) that the existence of norms for their regulation is a necessary
condition for the existence of culture. Necessary, although not sufficient.
For, as Freud views it, it is the internalization of these norms (as a result of
the infant’s prolonged attachments to nurturant parenting figures) that
produces the moral anxiety and guilt that, by inhibiting expression of
aggressive motives, preclude the return to a state of nature.6
6. That Freud is the (unacknowledged) spiritual descendant of Hobbes, just as Lévi-Strauss is the
(acknowledged) spiritual descendant of Rousseau, would seem to be obvious. Freud differs from
Hobbes in that the latter believed that the physical authority of the state is necessary to assure
compliance with the cultural prohibition on aggression, while Freud believed that the key to the
problem is the moral authority of the parents, internalized as the superego and reflected in moral
anxiety and guilt.
We can now see why Freud would never characterize the Bororo myth as
marking the transition from nature to culture—not because the son desired
his mother, or the father wished to kill his son, or the son wished to kill his
father, nor even because they acted upon their wishes, but because they
experienced neither moral anxiety about their desires nor guilt about their
deeds.7 Since absence of anxiety and guilt implies an absence of cultural
norms prohibiting aggressive behavior, in Freud’s view this myth could not
possibly represent a transition from nature to culture, even if it were granted
that it represents the origin of cooking.
7. Indeed, rather than experiencing guilt and, thereby, atoning for killing his father by renouncing
his claims on his mother and stepmothers, the son continues to take his “revenge” on them.
For Freud this myth would, however, represent something else. Holding
that the transition to culture, although marking the control of aggressive
behavior, does not mark the disappearance of aggressive affects and
motives. Freud assumed that aggression is a matter of important concern to
social actors in a state of culture. And since, so he further assumed, their
important concerns are inevitably represented in their fantasies—the
privately constituted fantasy of dreams and the culturally constituted
fantasy of myth and religion—it follows that the manifestly aggressive
themes in the Bororo myth, like the manifestly sexual themes in the rituals
and beliefs analyzed by Leach and Douglas, are just that: aggressive and
sexual themes. In short, for Freud, the myths, rituals, and beliefs discussed
in this paper are concerned with those very sexual and aggressive wishes
and fears that these anthropologists and their many followers, denying that
they are of concern to cultural symbol systems, interpret out of existence. It
is rather startling that this simpleminded view (which, of course, is hardly
original with Freud or restricted to Freudians) must be reiterated to an
anthropological audience in 1977.

Notes
Reprinted from American Anthropologist vol, 81, no. 1 (Mar. 1979):5–13, by permission of the
American Anthropological Association. Not for further reproduction.
This paper was prepared for, and a shorter version was read at, the 1977 annual meetings of the
American Anthropological Association, as a part of a symposium entitled “Psychological
Anthropology: a Perennial Frontier?” organized by Professor Theodore Schwartz. The preparation of
this paper was assisted by a grant from the National Institute of Mental Health.
References
Douglas, Mary. 1966. Purity and danger. New York: Praeger.
Freud, Sigmund. 1971a. Totem and taboo. In The standard edition of
the complete psychological works of Sigmund Freud, vol. 13.
London: Hogarth Press.
_____1971b. Civilization and its discontents. In The standard edition
of the complete psychological works of Sigmund Freudy vol. 21.
London: Hogarth Press.
Leach, Edmund. 1967. Virgin birth. Proceedings of the Royal
Anthropological I nstitute for 1966:39–49.
Lévi-Strauss, Claude. 1969. The raw and the cooked. New York:
Harper and Row.
Spiro, Melford E. 1968a. Review of purity and danger. American
Anthropologist 70:391–93.
_____1968b. Virgin birth, parthenogenesis, and physiological
paternity. Man 3:224–61.
11
Some Reflections on Family and
Religion in East Asia

Introduction
THE FAMILY and religion were among the core interests of anthropological
inquiry from its very inception, and they have remained among its most
perduring subjects of investigation. There are, I believe, at least two reasons
why this should have been so.
In the first place, although like the family and religion other sociocultural
systems are also universal, none is as easily recognizable across all the etic
types which are employed by anthropologists to classify the wide array of
sociocultural systems as are family and religious systems. In the case of
economic and political systems, for example, those which fall at the polar
extremes of any of the recognized typologies by which they are classified
are often identifiable as members of the same series only because of their
similarities to the intermediate types comprising the typologies. We need
only remind ourselves of the differences between nomadic-gathering
economies and industrial-bureaucratic economies, or between small,
acephalous band organizations and large-scale centralized empires in order
to grasp this rather simple point. For both comparisons it is difficult to
identify an invariant sociocultural core that cuts across all the types
comprising a formal typology, or that persists from the earliest
manifestation of an evolutionary typology to the most recent of its
manifestations.
The contrary, however, is the case in regard to the family or religion, or
so it would seem, if I am correct in claiming that the nuclear family is the
invariant core of every family system, and that the worship of superhuman
beings comprises the invariant core of every religious system. Indeed, the
founders of nineteenth century evolutionary thought in Europe were as
perplexed by the similarities between their own (Victorian) family and
religious systems and those of the non-European societies that they studied
as they were intrigued by (and sometimes contemptuous of) the differences.
The second reason, I believe, for the perduring anthropological interest in
the family and religion is that these two systems are related to one another
in a systematic relationship which holds for no two other sociocultural
systems. At first blush, this statement seems paradoxical because while the
human family marks man’s affinity with the rest of the animal kingdom—
especially the class of mammals—religion marks his uniqueness. That is,
the family (whether uniparental or biparental) is a generic mammalian
institution, and since man evolved from a mammalian (more particularly a
primate) species, it is hard to escape the assumption that the human family
is phylogenetically rooted in the family system of our prehominid ancestors.
Since religion, however, is found (as far as we can tell) in our species
alone, if religious systems are also universal it is because (as Robertson
Smith and Freud pointed out a long time ago) they are rooted in, and may
be viewed as metaphorical expressions of family (including kinship)
relations. If that is so, then religion and the family (in contrast, say, to
religion and economics or religion and politics) sustain a special
relationship with each other, the existence of religion being in large part a
function of the existence of the family.
To say that religious systems may be viewed as a metaphorical
expression of family relations is to say that while the existence of the family
may be explicable in terms of biological characteristics and needs which we
share with other mammals, religion is explicable only in terms of the
uniquely human capacity for symbolization, for it is in the symbolic process
that the privately-constituted world of fantasy, the wellspring of religious
belief, is transformed into the culturally-constituted world of religion. That
is so because the symbol creates Being (spirits and gods) out of non-Being,
and it invests words and gestures with the instrumental power that is
imputed to religious ritual. In short, religious symbols often represent the
transformation and elaboration, at the cultural level, of fantasies and
cognitions that are found at the psychological level, which in turn, are
produced by family relations at the social level.
 Although there is nothing new in this, its implications for the universal
dimensions of the family and religion have not always been spelled out by
anthropologists who, in their special concern with variation, have more
often concentrated on the cross-cultural differences in family and religious
systems than in their regularities. Although there can be no denying the
importance of these differences, their regularities are equally important, and
it is the recognition of the cross-cultural regularities in the family, viewed as
a system of social relationships, that enables us to understand its connection
with religion, viewed as a system of symbolic relationships.

Cross-Cultural Regularities in Human Family

Systems

The panhuman roots of the regularities in human family systems are not
hard to discover, for however much these systems must adapt to and are
conditioned by variations in ecology, economy, demography, the polity, and
the like, every family (and kinship) system is a response to certain
irreducible biological characteristics of human existence, among which I
would stress the following.

(1)Human reproduction is bisexual, and conception is effected by means

of sexual intercourse.
(2)Human beings are bom helpless, and they remain dependent both
physically and emotionally for a prolonged period on their caretakers.
(3)Human beings are also born instinctless, so that their caretakers attend
not only to their dependency needs but also to their need to acquire the
cultural traditions of the group into which they are born.
(4)Since relatively permanent pair-bonding, brought about both by the
absence of estrus and the need for economic cooperation, has been a
human characteristic since at least the origin of hunting, the core
caretakers are parents, together with whom children comprise a
domestic group.
(5)Dependency being the child’s prepotent need, he develops feelings of
affectionate attachment toward his caretaking parents who, to a greater
 or lesser degree, gratify that need.
(6)Gratification, however, is always relative to frustration, and caretakers
not only gratify, but they also frustrate children’s needs.

Caretakers are frustrators, willy-nilly, in a number of ways. First, since

they are agents of socialization and enculturation, they impose restrictions,
constraints, and prescriptions on their offspring which are almost always
frustrating if not downright painful. Second, since there is no
incompatibility between lactation and sex in human beings, as there is in
infrahuman mammals (in which the female does not enter estrus until her
infant is weaned), caretakers are simultaneously both parent and spouse.
Hence, since the mother, for example, is simultaneously mother and wife,
the child must share her attention and love with the husband-father. Third,
since humans are dependent for a highly prolonged period, their
dependency does not cease with the birth of a sibling—as it does in
infrahuman mammals, who either leave or are driven from the domestic
group by the time the new infant arrives—which means that the attention
and love of the mother must be shared with siblings as well as father. The
sharing of love and attention is frustrating enough for adults, as the ubiquity
of jealousy and envy indicates; for children, however, the frustration is even
stronger.
Given, then, that caretakers both gratify and frustrate children’s
dependency needs, parents are not only the first and most important objects
of their children’s affection, but they are also, together with siblings, the
first and most important objects of their hostility.
In sum, so far as their emotional texture is concerned, we would expect
that in any society the relationships among all of the dyads comprising the
family would be characterized by strong ambivalence, and that children
would develop both an Oedipus complex (Spiro 1982a) and sibling rivalry.
That these expectations also hold for East Asian families is abundantly
evidenced in the papers that have been prepared for this conference—those
at least that deal with the social relationships of the family. That these
characteristics are symbolically expressed in the religious systems of East
Asian cultures is no less evident from those reports. I shall return to these
points later in my discussion.
These general observations concerning the cross-cultural regularities in
the human family have important implications for our understanding of its
variability. For if these observations are correct, the biological
characteristics enumerated in the foregoing discussion comprise a set of
parameters, or invariant conditions, which all societies have had to cope
with in the historical development of their family systems. On the one hand,
therefore, these invariant conditions might be said to account for the cross-
cultural regularities in human family systems. On the other hand, however,
the variability that is found in these systems—variability, for example, in
the principles of recruitment to the domestic group, the classification of
kintypes, the norms which govern social relationships within the family, the
rules which determine the distribution of inherited property, and so on—
may be said to represent a limited range of institutionalized solutions to the
problems, both sociological and psychological, created by those same
invariant conditions. In sum, it might be argued that certain invariant
biological conditions (bisexual reproduction and prolonged biological
dependency) produce certain invariant sociological consequences (the
biparental family and its caretaking functions) from which there flow
certain invariant psychological consequences (ambivalence to parents and
siblings), and that these consequences lead to variable cultural responses
(norms and rules) which regulate the potentially disruptive effects of both
dimensions—love and hate—of these ambivalent relationships.
I shall now argue that the invariant conditions that account for the cross-
cultural regularities in family systems are no less important for the
understanding of family behavior than are the culturally variable rules and
norms that govern family relationships. I am not arguing, I hasten to add,
that these rules and norms are merely epiphenomena—superstructure, as
Marxists say. I am arguing, rather, that the emotional and motivational
dispositions of family actors that require the elaboration of cultural rules
and norms for their regulation continue to operate in these actors even after
they require those rules and norms, and that their behavior, therefore, is a
product (in the algebraic sense) of the simultaneous influence of both of
these determinants.
Take, for example, filial behavior. Whatever the cultural values regarding
parents might be, children’s sentiments and attitudes regarding their parents
are not formed exclusively by these culturally variable values. They are
formed as well by their invariant, albeit socially acquired emotions of the
type discussed earlier; and their filial sentiments and attitudes represent an
interaction of these two sets of determinants. To be sure, to the degree that
filial emotions conflict with filial values we would expect filial behavior to
comply not so much with the actors’ emotions as with their cultural values
which, expressed in rules and norms, govern their duties and obligations to
parents. Since, in such a case, filial emotions conflict with filial values, the
former must be repressed. Inasmuch, however, as the parents remain their
unconscious targets, these emotions are as powerful as they ever were, but
they are expressed in various disguises—some more, some less disruptive
in their social consequences.
In conclusion, if social relationships are governed by the attitudes and
sentiments an actor has toward some Alter, and if these attitudes and
sentiments are produced not only by culturally acquired values, but also by
emotional and motivational dispositions acquired by the actor in his social
experience with Alter, it is as foolish to ignore the emotional as it is to
ignore the cultural determinants of their relationship. This is especially so in
the case of family relationships, for many attitudes and sentiments which
children hold toward parents and siblings are based on conceptions of them
that are formed much before their acquisition of language—hence, before
the acquisition of the cultural values which comprise the normatively
expected conceptions of parents and siblings. That is, these attitudes and
sentiments are formed on the basis of their personal experiences with their
parents and siblings, experiences which, as was argued above, arouse
conflicting emotions of love and hate, of attachment and resentment, and
the like. Since, then, these emotions are usually reinforced by later
experiences with parents and siblings, they inevitably play a significant role
in the development of the attitudes and sentiments that inform their social
relationships with them.

Religion and the Family

How, now, to turn to the second aspect of the theme of our conference, does
the discussion of the family relate to religion? As students of religion we
can never know the superhuman beings postulated by religious belief
systems directly; we can only know them indirectly, i.e., by means of the
conceptions that religious actors have of them. Indeed, with some few
exceptions—for example, mystical experience and trance possession—the
religious actors themselves do not claim to have direct knowledge of them.
They, too, know them only indirectly—as they are represented in the
collective representations of their culture, in their own mental
representations of them, and in the rituals by which they attempt to relate to
them.
If, then, we take these three sets of data as our evidence for the
conceptions which religious actors have of superhuman beings, it seems
safe to say (on the basis of a great deal of comparative research) that these
conceptions are more or less isomorphic with the conceptions, unconscious
as well as conscious, which, as family actors, they form of their family
members, and more particularly the conceptions which as children they
form of their caretakers, usually their parents, in their personal encounters
with them. It also seems safe to say that the rituals by which they attempt to
relate to these superhuman beings express, and sometimes gratify, the
wishes that are instigated by the emotions which those caretakers arouse in
them, most especially the emotions of dependency and love, of fear and
hatred. Thus, if the child’s dependency needs, for example, are gratified by
a nurturant mother, it is not unlikely that as an adult he will worship a
mother-like superhuman being(s) from whom he anticipates the
gratification of his wish for continuing childlike dependency. Similarly, if
certain of his childhood needs are frustrated, for example, by an
authoritarian father, whom he consequently learns to fear or hate, it is not
unlikely that as an adult he will propitiate a fatherlike superhuman being(s)
so as to avoid his wrath which he fears, and/or to express his hostility to
him, something he seldom does (not at least overtly) in his relationship with
his father.
In short, I am suggesting that whatever the “objective” characteristics of
his parents might be when seen through the lens of a camera, the child
forms various mental representations of them which, given the fact that the
child’s lens is neither objective nor realistic, distort and exaggerate their
characteristics. I am suggesting, further, that these parental representations,
partly conscious, partly unconscious, constitute the anlage or conceptual
schemata for the mental representations which he later forms of
superhuman beings. I am suggesting, finally, that the psychological reality
of superhuman beings, like that of the parents, is in no way affected by their
physical reality. Thus, even if their parents have died and are no longer in
the physical world, they continue to exist for their children in the latter’s
representational world—i.e., in their mental representations of them—and it
is in the latter world that, even when they are alive, they have their
important, i.e., their psychological reality. This condition also holds, part
passu, for gods, ghosts, and ancestors, to use Jordan’s felicitous designation
for the superhuman beings of East Asia (Jordan 1972).
With this conceptual orientation to the relationship between the family
and religion, we may now examine the extent to which this schema applies
to the East Asia materials. I should also note that if much of the focus of my
discussion is on male actors, it is because the material herein has been
typically presented from that perspective.

Family Tensions in East Asia

In order to confine my discussion to reasonable boundaries, I have decided
to focus on lines of tension in the East Asian family. Since, however, filial
piety and family solidarity have always received a great deal of attention in
discussions of the East Asian family, such a focus may perhaps contribute
some new dimensions to the subject.
One might expect the following lines of tension to be most salient in the
families of East Asia. First, given the extraordinary relationship of
nurturance and dependency characteristic of the mother and the son, most
especially, so it seems, in Japan, I would expect considerable tension to
develop between the father and the son: on the father’s part because of the
wife’s obvious emotional preference for the son, on the son’s part because
his father is a most important competitor for his wish for the mother’s
exclusive attention. If nothing else, it is the father—not the son—who has a
monopoly on the mother’s sexuality. Moreover, since the father-husband is
an especially important authority figure, both for the wife and the children,
one whose jural, if not personal, authority requires obedience and respect,
one would expect that this would constitute an equally important source of
tension in the father-son relationship.
It must be noted here that although the mother-son relationship is
described in many of these papers, none of them addresses the father-son,
father-daughter, or mother-daughter relationships, nor, except in passing,
the relationship among siblings. That this disregard of the latter
relationships would not occur in a conference dealing, for example, with
South Asia or the Middle East, only serves to underscore the pivotal
emotional, though not jural, importance of the mother-son dyad in the East
Asia family system. (It also means that my comments will be incomplete
and somewhat distorted.)
Given the subordination of children to both parents, as well as the duties
and obligations that the former owe the latter (which are summed up in the
key concept of filial piety) and which continue even after their death, I
would also expect considerable tension to develop not only between father
and son, but also between mother and son. But there is an additional—and
more important—reason that I would expect tension to develop in the latter
relationship. Although the mother is extraordinarily nurturant to the son and
attentive to his needs—which, of course, leads to the loving and dependent
attachment to her that is stressed in all the papers—that very attentiveness
can be expected to lead to three types of tension.
First, the mother’s devotion, conceived as a “perfect act of unchanging
selfless sacrifice” to quote Tanaka (1984), may produce a “deep feeling of
guilt and indebtedness in the son,” and such a feeling can only lead to
profound (though probably unconscious) resentment. Second, the young
son’s intimate and persistent physical contact with the mother—he sleeps
with her, is bathed by her, and so on—most probably arouses erotic feelings
for her which, however, are necessarily frustrated, and, I assume, ultimately
repressed. It is for that reason, I would assume, that the mother-son
relationship is characterized, to quote Tanaka again, by “the continuous
presence of unresolved libidinality.” Third, the dependent attachment to the
mother, which she herself encourages, is in conflict with the child’s need for
autonomy, including psychological separation and individuation (Mahler et
al. 1975).
I would also expect considerable tension to develop in the husband-wife
relationship: on the husband’s part because of his subordinate place to the
son in the wife’s emotional life; and on the wife’s part because, as Tanaka
puts it, of the “unrecognition of sexuality in the marital relationship.” Now
if, as Tanaka also points out, the wife’s relationship to her husband
recapitulates that of the mother to her son. that is highly gratifying for the
husband, especially since he can gratify his erotic needs outside of the
marriage—hence, have his cake and eat it too—but it can only be
frustrating to the wife whose erotic needs, however much she may
sublimate them in her relationship with her son, are nevertheless frustrated
in her marital relationship in which (to quote Tanaka again) “sexuality is
very much downplayed.” Indeed, I would argue that it is precisely because
her sexual needs are frustrated in her role as wife that the woman invests
such great affect in her role as mother, her nurturant relationship to her son
being a sublimation of her frustrating erotic relationship with her husband.
I would argue, too, that the husband’s relative disinterest in his wife as an
erotic object is the last link in a feedback loop in which, having recoiled as
a boy from the incestuous implications of his attachment to his highly
affectionate mother, he marries a woman who in so many respects
represents the mother. In Japan, for example, the husband not only calls his
wife, “mother,” following the birth of their first child, but since her
relationship to him is, as Tanaka puts it, “not essentially different from her
relationship to her young children”—implying that it is little different from
his mother’s relationship to him when he was a child—he comes to perceive
her, so I would suggest, as a mother. In short, since the wife-husband
relationship and the mother-son relationship are “dangerously similar” (to
use Tanaka’s words), it is hardly surprising that the wife becomes a
nonerotic object for her husband.
The dynamics of that process are encapsulated in Freud’s pithy comment
concerning a class of males—the males of East Asia, of course, were far
from his mind—concerning whom he writes, “Where they love they do not
desire and where they desire they cannot love. They seek objects which
they do not need to love, in order to keep their sensuality away from the
objects they love” (Freud 1912:183). For such males, the wife is classified
with the class of females who, like the mother, are viewed as asexual, and
they are distinguished from the class of females (including prostitutes,
concubines, and mistresses) who are viewed as sexual. The former class,
being pure, are worthy of love; the latter, being impure, are worthy only of
sex.
I would suggest, then, that the neo-Confucianist view of marriage,
according to which, to quote Tu (1984), “mutual responsibility rather than
romantic love” ought to characterize the conjugal relationship, is more a
reflection of than a model for the actual relationship between the spouses.
In either event, for the wife to be treated by her husband, as Ch’eng I
reports his father to have treated his mother, with “full respect” and
“reverence,” or for that same wife to live with her husband in “tranquility
and correctness,” and never to be the object of “indecent liberties and
improper intimacies,”—all these quotations are taken from sources quoted
by Tu—such a wife, I would suggest, is not only sexually frustrated, but in
the context of East Asian society she is all the more resentful (perhaps
unconsciously) because though she is not the object of her husband’s
“indecent liberties and improper intimacies,” she knows that other women
—concubines, mistresses, or whatever—are.
A fourth line of tension, as I see it, develops between siblings. It seems
reasonable to assume that the birth of a new child, especially a son, means
that the elder child is, to some degree, displaced by the younger as the focal
attention of the mother. In Korea this displacement is both symbolized and
actualized in the sleeping arrangements in which, as Lee (1984) describes it,
the elder sibling is extruded from his parents’ bedroom following the birth
of a younger sibling, and is sent to sleep in the room of his paternal
grandmother. But even in Japan, where such extrusion does not occur, the
rivalry between siblings for maternal love has its effects, so that it is little
wonder that in the Kojiki myths analyzed by Sofue (1984), 12 of the 15
myths which deal with the relationship between brothers entail competition
and rivalry, and that in 8 of these 12 the rivalry culminates in fratricide.
Given such strong indications of sibling rivalry, it is little wonder, too,
that both in Japan and Korea the domestic unit comprises a stem, rather
than extended family. Moreover, although Lee contrasts the Japanese and
Korean stem family households with the extended family household in
China, in fact the situation in China (Hsu 1971; Yang 1969) is very little
different from that in Japan and Korea, and for the same reason: following
the death of the father, friction between married siblings leads to the
segmentation of the extended family and their formation of independent
households.

Family Tensions and Religion in East Asia

In the following sections I wish to examine some possible links between the
putative tensions in the East Asian family discussed in the previous section
and certain aspects of East Asian religion. Before examining these links it is
important to emphasize two points. First, I am not suggesting that religious
beliefs and rituals can be “reduced” to sociological or psychological
variables. I am suggesting, rather, that social relationships, cognitive
orientation, and motivational dispositions both inform and are reflected in
belief and ritual systems, whether sacred or secular. Second, in focusing on
their relationship between family tensions and religion, I am not suggesting
that the solidarious dimension of the family is not reflected in religion.
Rather, that dimension is not the subject of my inquiry.

Ancestor Worship and the Father

Although ancestors are most often viewed as benign, it is also the case that
sometimes they may be punitive. Fortes claimed that in East Asia, as well
as West Africa, “the feature that stands out most conspicuously in all
varieties of ancestor worship … is their punitive character” (Fortes
1977:145). Thus, in Korea, the dead (including ancestors and ghosts) are
dangerous, so Kendall (1984) remarks, “simply because they are dead …
and their touch brings illness or affliction.” This is especially true in the
case of ancestors who died with “unfulfilled desires.” Restless ancestors, as
well as ghosts and angry household gods, cause not only illness, but
financial loss and domestic strife, as well. Similarly, Lee observes that if the
ritual service for an ancestor is not performed, the ancestor spirit becomes a
wandering ghost; and although, he further observes, ghosts have no power
to punish their descendents directly, this implies, I would assume, that they
do have power to punish them indirectly.
In his treatment of ancestor worship in China (Taiwan) Suenari (1984)
does not deal with the punitive dimension of ancestors, but it is implicit in
his emphasis on (what he calls) the “economic reciprocity” characteristic of
family relationships, including that with ancestors. Like their relationship
with the gods, the Chinese relationship with their ancestors is “contractual,”
which implies that the latter’s punitive or nonpunitive action is contingent
upon the offering or withholding of gifts by their descendents. This
implication is explicit in the work of Emily Ahern on ancestor worship in
Taiwan. According to Ahern’s findings ancestors are not infrequently
blamed for such serious misfortunes as insanity, serious infirmity and death
(Ahern 1973: chap. 12).
 The situation in Japan is no different. Thus, Morioka (1984) observes that
for the lower class, at least, the function of ancestor worship is to “avert
disaster” which would be caused by ancestors if their worship were
neglected. Carmen Blacker makes the same point without restricting it,
however, to the lower class. Thus, if “the ancestral dead are not correctly
treated by their descendants, if the offerings or the obsequies necessary to
their nourishment are neglected, then with frightening suddenness their
nature will change. The kindly old grandfather, the sympathetic father, the
loving mother will turn in an instant into a vicious and capricious tyrant,
punishing the neglectful family with curses.” (Blacker 1975:47–48).
In order to understand these findings, it is important to consider other
data. First, approximately 50 percent of Japanese families, according to
Morioka, continue to practice ancestor worship even when the te system has
collapsed. Second, in Korea, according to Lee, the ancestor tablet is kept in
the ancestral shrine only until the fourth ascending generation, following
which it is buried in the grave, which implies that the ancestor remains
individuated only for a relatively short time, after which he is assimilated to
the generic class of “ancestor.”
Now although in ancestor worship, rites are performed for all one’s
ancestors, these findings suggest that the cognitively salient ancestors are
not the genealogically remote ancestors, but rather the genealogically close
and immediately dead ancestors—i.e., the parents and grandparents. The
remote ancestors, of course, are important both jurally (to establish claims
on property, to enhance the prestige of a clan line or to legitimize its rights)
and politically (to inculcate respect for authority, beginning with the family
and ending with the centralized state). But for the average individual, I
would suggest, these corporate functions are second in importance to their
“religious” functions. In the latter regard, an ancestor (like anyone else) is
cognitively salient for a religious actor only to the degree that he has a clear
and vivid mental representation of him, and the ancestors concerning whom
he has the clearest and most vivid mental representations are his deceased
parents and grandparents—those whom he himself has personally
encountered. Hence, even though in ancestor worship the actor in principle
attends to all of his ancestors, it is his immediate ancestors, especially his
parents, who, so I would suggest, he has most in mind, or whose mental
representation forms the template for his conception of the other ancestors.
Fortes (1961:187) put it most succinctly in his remark that “ancestor
worship is primarily the religious cult of deceased parents.”
These claims are supported by Morioka’s finding that (a) although
traditionally a Japanese “ancestor” is the ancestor of the te, with the
collapse of the ie in urban families, “ancestor” has increasingly come to
designate the “deceased bilateral kindred” (which most importantly means,
I would suggest, the ascendance of the mother to the status of a cognitively
salient ancestor), and that, (b) the “private” meanings of ancestor worship
have superseded its “public” meanings. Consistent with my previous
hypothesis, however, I would suggest that these private meanings were
always foremost in the worshipper’s mind (although the public meanings
were, no doubt, the important formal meanings), and that the collapse of the
ie merely permits their centrality to be acknowledged.
Even more important, however, is Morioka’s finding that although many
urban families do not own a butsudan, the rate of ownership dramatically
increases with the death of a parent, and that even in extended family
households (in which, presumably, the ie is still somewhat important) it
increases significantly in households with widows. These two findings
suggest once again that the cognitively and emotionally salient ancestors
are the immediate dead. Moreover, taking Morioka’s findings concerning
widows into account, the ancestors need not even be lineal ancestors so
long as they had been household members with whom the actor had
sustained important social relationships.
If this is so, then inasmuch as ancestors are not only revered—an
extension of filial piety—but also feared, I would suggest that both attitudes
are a function of the mental representations that, as children, the actors had
formed of their immediately deceased ancestors, most especially, but not
exclusively, the father. The latter attitude, which is the one we are
concerned with here, might be explained in the first instance by Fortes’
hypothesis that it is the “authority component” of the father that is elevated
to ancestorship. If, then, in addition to his positive feelings for him, the
child also develops negative feelings toward the father, one would expect—
given the over-arching value of filial piety—that he would probably repress
these feelings, or at least not exhibit them in overt behavior. That upon his
death the father—now an ancestor—is viewed as a potentially dangerous
figure, capable of inflicting harm on his descendants, is then susceptible of
two complementary interpretations.
 First, the repressed hostility which was felt for the father when he was
alive can now find an outlet in the culturally-constituted belief that
ancestors are potentially dangerous. Specifically, that belief allows the child
to project his erstwhile hostility toward the living father onto the dead
ancestor, thereby transforming him from an ordinarily oppressive authority
figure into a potentially dangerous one. The second interpretation is more
complex. Clinical evidence indicates that hostility toward some person may
generate death wishes (if only as an unconscious fantasy) toward him;
should that person die, the actor, given that the “omnipotence of thoughts”
is one of the characteristics of unconscious mentation, may unconsciously
experience his death as resulting from his death wishes toward him. Given,
then, that the principle of lex talionis informs not only many legal systems,
but unconscious mentation as well, the belief that the deceased father is
potentially dangerous might be explained by the unconscious conviction of
the child that the former might harm him in retaliation for the ‘harm’ that he
(the child) had inflicted on the father.

Goddesses, Religious Specialists and the Mother

Since the most notable feature of the mother-son relationship in East Asia is
the mother’s nurturance and the son’s dependence, we would expect that
dimension of the mother-son relationship to be reflected in East Asian
religion. The extraordinary nurturant-dependent nature of the mother-son
relationship in East Asia, at least in Japan, is stressed in the reports by
Tanaka and Sofue. A “good mother,” Tanaka observes, “is believed to care
for and worry about her son eternally.” Hence, the son’s dependency (amae)
on the mother persists not only over her lifetime, but even after her death
when, as an ancestress, she is still “supposed to be watching over him.” As
a measure of what he characterizes as the son’s “very strong dependency
need” in regard to the mother, Sofue points to the fact that it is the favorite
theme of Japanese popular culture. For him, therefore, this need comprises
a “mother-complex.” Whatever that expression may denote, it certainly
connotes the formation by the son of a mental representation of the mother
as extraordinarily loving and nurturant, one who can be expected to do
anything in her power in the service of his welfare.
 Such a maternal representation is too good to give up. Hence, it is little
wonder that when the mother becomes an ancestor the son continues to
expect that he can rely on her assistance. It is little wonder, too—though I
would not have predicted it—that with the introduction of Buddhism to
China, and thence to Japan, the infinitely compassionate Hindu god,
Avalokitesvara, was transformed into the goddess, Kuanyin (China) or
Kannon (Japan). That the amae relationship with the mother is transferred
to Kannon—probably the most popular deity (actually Bodhisattva) in
Japan—and that the benevolent dimension of the maternal representation is
reflected in the collective representation of Kannon (just as the authority
dimension of the paternal representation is reflected in the collective
representation of the male ancestors) can be seen in the following statement
of Teruko Furuya. (The translation is by Yohko Tsuji.)
Kannon’s concern is not directed toward heaven or a utopia, but toward this world (in which
many people still suffer). Kannon is benevolent and omnipotent. She never punishes us, nor gets
angry with us. On the contrary, consistently and promptly she answers our selfish prayers, such as
a desire to have an attractive child, a desire to pass an entrance examination for a prestigious
school, a desire to get promoted at work, and so on. She is just like an amai mother who always
listens to the desires of an indulged child.

Like the mother, Kannon is not only infinitely compassionate, but she has
another quality that the mother does not have: she is also all-powerful, as
the following quotation from Blacker (1975:94) indicates.
A man only has to think of the Bodhisattva Kannon to be saved from every conceivable calamity.
A man hurled into a fiery pit has but to think of the Kannon for the fire to be quenched. A man
floundering in an ocean of sea monsters has but to think of Kannon and he will neither sink nor
drown. A man bombarded with thunderbolts has but to think of Kannon and not a hair of his head
will be hurt. A man beset by goblins, demons, ghosts, giants, wild beasts or fearful fiery serpents
has but to think of Kannon for these creatures to vanish.

Power of this magnitude, of course, is never found in any human being but
it is found in the mental representations that a young child forms of his
parents. In the child’s eyes the parent, who literally has the power of life or
death over him, is indeed omnipotent. Hence, when the omnipotence of the
maternal representation of the Japnese child is conjoined with its
benevolence, the result—I would suggest—is a maternal representation that
is projected in the adult’s collective representation of Kannon.
 But the relationship between the religious devotee and Kannon is not the
only manifestation of the child’s amae relationship with the mother on the
religious plane. It is also manifested, as Tsuji (1980) has suggested, both in
the relationship between client and shaman (miko)—and here, I would
include Korea as well as Japan—and in that between the members and
founders (kyoso) of the new religions.
In Japan and Korea, though not in China, shamans are almost exclusively
female, and the rare male shaman performs his role as a trasvestite. In
Korea, according to Kendall, the shaman (mansin) is used to help the
household overcome the afflictions—illness, financial loss, domestic strife
—that are brought about by restless ancestors, ghosts, and angry household
gods, as well as to help young women overcome infertility. According to
Blacker, similar functions are served by the Japanese miko as well as by her
modern counterpart, the kyoso, most of whom are also female. So far as the
latter are concerned, Davis observes that the “great majority” of those who
join a new religion hope “to receive some practical benefit—cure of
disease, solution to some personal problem, support for some psychological
difficulty, etc—from their affiliation” (Tsuji 1980:ms). In short, in both
cases when faced with adversity, the adult reestablishes a dependency
relationship with a female religious specialists that as a child he had
experienced with his mother.
I might add that just as the devotees’ relationship with these female
religious specialists is best understood by reference to the family—as a
recapitulation of their early experience with the mother—the recruitment of
these women to their religious vocations is also best understood by
reference to the family—to their experience as wives in a sexually
frustrating marital relationship. This, in Korea, Kendall tells us, shamans
are usually recruited to their calling in middle age, after suffering a “run of
ill luck” as a result of possession by some god who claims them. In Japan,
Sasaki (1984) writes, there are two ways of becoming a shaman: “One is by
divine calling and the other by self-searching for shamanship.” In the
former case, the woman suffers from some mental and physical
abnormality, including visual and auditory hallucinations, trance, decrease
in appetite, severe palpitations of the heart, sleeplessness and loss of
weight. If her condition is not improved by resort to modern medical
specialists, she will visit a shaman to discover the cause of her affliction.
Should it be diagnosed as resulting from spirit possession, the most
important means for overcoming her afflictions is for her to become a
shaman herself, and to serve the spirit or god who has possessed her. (In
Okinawa, this often means agreeing to marry him.) After agreeing to
become a shaman, her “abnormality” gradually disappears.
These women, according to Blacker, exhibit in their personal histories a
“curiously uniform pattern,”
Nearly all of them in their early life betray symptoms of what could be called ‘arctic hysteria.’
They are sickly, neurotic, hysterical, odd, until a moment comes when exacerbated by suffering,
these symptoms rise to a climactic interior experience of a mystical kind. A deity, by means of a
dream or a possession, seizes them and claims them for his service. Thenceforward they are
changed characters. Their former oddity and sickliness give way to a remarkable strength and
magnetism of personality, which is conferred on them, together with various supernormal powers,
by the deity who has possessed them (Blacker 1975:129)

The characteristics of these women—which are almost identical with the

characteristics and mode of recruitment of Burmese shamans whom I
investigated in the 1960’s (Spiro 1978)—are the classical symptoms of
conversion hysteria, a condition that is typically brought on by the
repression of frustrated sexual needs. In this case, I would suggest, these
frustrated needs are symbolically gratified by means of trance possession—
i.e. by a hallucinatory experience—in which they are finally claimed by the
most potent male of all, a god. If, then, the East Asian woman, like her
counterpart in South and Southeast Asia, is often frustrated by her
unfulfilled libidinal attachment to her father, as Roy (1975) observes in the
case of India; if moreover, her unfulfilled desires continue to be frustrated
in her sexually unsatisfactory relationship with her husband as occurs, so
Tanaka and Koh suggest, both in Korea and Japan, and if, finally, the
sublimation of her repressed libidinal desires in her relationship with her
son is not entirely effective; if all this is true, then it is hardly surprising, as
Blacker observes, that the women who become shamans or founders of new
religions represent merely the tip of an iceberg. Nor is it surprising that
some of these women—usually lower class and not highly educated—
should find an outlet in these religious callings, especially since possession
by a god serves not only to gratify their frustrated libidinal needs, but their
status needs as well. From a position of subordination and relative
powerlessness in the formal social structure, they suddenly become the
medium for the gods themselves, so that their personality undergoes a
transformation of corresponding magnitude (Blacker 1975).
Buddhist Monasticism and the Parents

As a final example of the relationship between tensions in the parent-child

relationship and religion in East Asia, I wish to turn to Buddhist
monasticism. The crucial feature of Buddhist monastic recruitment is found
in the carrying out of the Buddha’s injunction that in order to achieve the
Supreme Goal of Buddhism it is necessary—in the words of the
Mahavagga Sutta—that “family men go forth from home into
homelessness.”
Now what was peculiar to Indian civilization at the time of the Buddha,
as Dutt (1962:43) observes in his magisterial history of Buddhist monks and
monasteries in ancient India, is not that India produced saints and ascetics
who renounced family and the world for a higher goal—religious
manifestations of that type were also found in other civilizations as well—
but that in India the “goers-forth” formed a community. Lancaster (1984)
also stresses this point in regard to East Asia, but with an important twist
that we shall note below. In ancient India, moreover, this community was
“recognized as such not only by the people, but also by the State”—
something which continues to be true of the Buddhist societies of Sri Lanka
and Southeast Asia (Spiro 1982b)—whereas in East Asia the Sańgha has
most frequently met strong opposition from the people and the State alike.
The State aside, opposition to the Sańgha is entirely understandable in
the case of East Asia where “leaving home” is the ultimate act of filial
impiety because as Lancaster observes, it means giving up the family name,
removing oneself from the ancestral lineage by not having children,
producing no descendants for the continuation of ancestor worship, and—
on a more mundane note—causing trouble after death because monks leave
no descendants to worship them.
It may also be remarked that, beginning with the Buddha himself,
“leaving home” has meant abandoning not only parents, but also—since
some “goers-forth” have been married when embarking upon their quest for
Enlightenment—wives and children as well. The locus classicus is the
Vessantara Jataka, the most famous Buddhist myth in Theravada Buddhist
societies (The Jataka 1957:vol. 6). The Prince Vessantara, an earlier
incarnation of the Buddha, abandoned his beloved wife and children in
order to seek Enlightenment. To attain his quest he even gave his children
as servants to a cruel Brahmin, and his wife to yet another. When his
children, beaten and oppressed by the Brahmin, managed to escape and find
their way back to Vessantara, he was filled with “dire grief’—his heart
palpitated, his mouth panted, blood fell from his eyes—until he arrived at
the insight that “All this pain comes from affection and no other cause; I
must quiet this affection, and be calm.” Having achieved that insight, he
was able to abandon his children.
Such an attitude, as Ozaki reminds us, (1984), was already found in East
Asia prior to the arrival of Buddhism, being present in Taoism as well. The
following story of Lu Hsiu Ching, which I quote from Ozaki’s chapter,
indicates that very clearly.
Lu Hsiu Ching retired from the world to the mountains where he studied. He left the mountains
for a while to look for some medicine. When he passed through his native place he stayed at his
house for a few days. At that time his daughter began to run a fever all of a sudden and fell into a
critical condition. The family pleaded with him to cure her. But Hsiu Ching left, saying: ‘Having
abandoned my family, I am in the midst of training. The house I stopped by is no different from
an inn to me.’1

1. That the pursuit of the religious life requires the rejection of family ties is, of course, not
restricted to the salvation religions of Asia. Early Christianity (as the attitude of Jesus, both to his
ties with his own family, as well as to family ties in general reveal) required an equally powerful
rejection (cf. Mark 3:31ff, Luke 9:59ff, Luke 14:26).

In both cases, Buddhist and Taoist alike, the attitude of the perfect “goer-
forth” is best described in the famous injunction of the Sutta Nipata.
“Having left son and wife, father and mother, wealth, and corn, and
relatives, the different objects of desire, let one wander alone like a
rhinoceros” (Sutta-Ntpata, verse 26 of the Khaggavisana Sutta). But, of
course, for the typical Buddhist monk, “leaving home” does not entail
wandering alone like a rhinoceros; instead, everywhere he enters a
community of like-minded “goers-forth,” a monastery.
Unless, as Lancaster perceptively observes, we recognize the “appeal of
the monastery life, it is difficult to account for the fact [that despite the
monks’ violation of the sacred duty of filial piety] Buddhist monastic
organizations thrived and became one of the most important features of the
religious, economic and social life of China, Korea and Japan.” In his
discussion of monastic recruitment, Lancaster is entirely correct in stressing
the appeal of the “pull” factors, as migration theorists call them, that attract
young men to the monastery—special dress, ritual, mystical practices, and
the like; but for the purposes of this paper, I should like to stress the “push”
factors that motivate them to “leave home.”
In attempting to understand these “push” factors it is important to stress
that when the young man “goes forth,” he does more than leave home—that
is much too passive a term to characterize this process, especially in the
societies of East Asia in which filial piety is an overriding value. Rather, he
abandons home, i.e. he actively severs his ties with his parents and siblings,
and he refrains from forming normally expectable ties with a wife and
children. It is the wish to sever the former ties and to refrain from forming
the latter which constitutes, I am suggesting, the “push” factor in his
“leaving home.” This suggestion is supported by the fact that (as has
already been noted) “going forth” does not mean wandering alone like a
rhinoceros, but rather substituting a voluntary community, based on
religiomystical ties, for an involuntary one, based on biological-kinship ties.
When it is considered, moreover, that the voluntary community, the
monastery, has many of the characteristics of the family—indeed, in China,
as Lancaster observes, the monastery became an actual family surrogate,
even including fictive father-son relationships, fictive lineage formations,
and fictive ancestral tablets—it becomes all the more obvious that it is not
living in a family-like structure as such that the “goer-forth” rejects in
“leaving home,” but rather living in his own biological family.
And make no mistake about it. “Leaving home” is a rejection of the latter
family, despite the monks’ attempts to rationalize it, and thereby cope with
the guilt induced by this act of filial impiety, by claiming that by
transferring merit to deceased parents and thereby promoting their
otherwordly welfare, the monastic vocation is in fact an expression of filial
piety. This is tellingly demonstrated, for example, in Lancaster’s data which
show that 70 percent of the Korean monks he interviewed entered the
monastery against the wishes of their parents, that they persist in their
decision despite the fact that for as many as ten years their parents begged
them to return home, that the resentment of their siblings for having to
assume the entire burden for caring for their aged parents is well-known to
them, and that their lingering guilt for having abandoned the parents is
evidenced by their resistance to discuss this matter in their interviews.
The recognition that monastic recruitment violates the norm of filial piety
—and the attendant psychological consequences of guilt, remorse, and
rationalization attendant upon this violation—is clearly evident as well in
the autobiographies of the five Korean nuns that Koh summarizes in her
paper. To be sure, the “pull” factors in the nuns’ motivation to enter the
monastery—the traumata attendant upon such experiences as the death of a
lover, marriage failure, frustrated child-lessness, the remarriage of a father,
the death of a mother, etc.—were even stronger, it is safe to say, than the
“push” factors. Nevertheless, their recognition of their violation of the duty
of filial piety, more especially since they had been importantly influenced
by the Confucian ethic, is equally evident. Thus, one nun characterized her
decision as “this unfilial act,” but then immediately rationalized the
decision by saying that as a Buddhist nun she could more effectively fulfill
her filial duties. Another nun, though her father was a Christian minister,
made the same claim. Their rejection of the family, their guilt, and their
rationalizations are all evidenced in the fact that, as Koh observes, they all
experienced “sorrow” about leaving home without their families’
permission, and yet they nevertheless carried out their decision over the
strong opposition of their families, and in the full realization that the latter
would suffer “tremendous social stigma.” Now, it may also be true, as Koh
claims, that their decision, given their “sentimental and deep attachment” to
their family members, is a measure of their self-reliance, but it is also a
measure (I would argue) of their willingness, if not wish, to reject their
families.
Since, however, the nuns’ decisions to enter the monastery were
traumatically motivated, I shall return to the monks in order to address the
problem that is by now rather obvious: what “push” factors could possibly
account for the fact that a young man, reared in a culture which places such
strong emphasis on filial piety, is nevertheless motivated to violate his filial
duties in such an extreme fashion?
The answer—or at least one of the answers—is to be found, I would
suggest, in the wish to escape the tensions that, as discussed at the
beginning of this chapter, are endemic in East Asia (but not only East Asia)
in the relationship between the boy and the other members of his family.
These tensions include the fear and resentment engendered by the father,
the incestuous and dependency anxiety aroused by the mother, and the
rivalry induced by male siblings in his family of origin. They also include
the Oedipally-induced fears concerning the formation of a sexual
relationship with a woman other than the mother, as well as the anxiety
about giving up his dependency orientation, both of which are aroused in
anticipation of establishing a family of procreation. All of these tensions, I
would submit, are experienced in some sense by most young men in East
Asia. (For a Southeast Asia parallel example, cf. Spiro 1977.)
In most cases these tensions are of a magnitude that can be handled and
overcome. In some few cases, however, they are too powerful to sustain in
continuous and ongoing relationships with members of the family—
especially since most of these tensions continue to be experienced
interminably because of the stem and extended family households of East
Asian societies. For such men the monastery is a marvelously contrived
institution which, inasmuch as it is religiously sanctioned, permits them to
avoid family relationships while at the same time providing cultural
legitimacy for their violating the duty of filial piety by interpreting the
seeming violation as motivated by a higher duty. Indeed, I would suggest
that the resort to such an extreme solution, in spite of the sacred duty of
filial piety, is convincing demonstration of how painfully those family
tensions are experienced by them.
But we don’t have to turn to those few who seek a solution in
monasticism to assess their strength even for the majority that are able to
cope with them. Thus, it is not accidental, I believe, that when they have the
chance, even those who continue to recognize the duty of filial piety seize
the opportunity to leave their homes, not for the monastery, but for the city.
And once there, rather than forming stem or extended family households,
most of them establish nuclear family households, as Tanaka and Lee have
shown for Japan and Korea respectively. It is for that reason that I disagree
with Lee’s interpretation of the “modernization of the Korean family and
religion as an extension of traditional familism.” The traditional sentiment
may still remain—after all we are still witnessing the first generation of this
phenomenon—but the difference between the persistence of the sentiment
of familism and its expression in the formation of a stem household is a
difference that, as William James puts it, makes a difference.
I wish now to conclude this chapter with the mother-son relationship, the
theme with which it began. I want to suggest that of all the tensions that
motivate home-leaving, whether it be for a celibate life in the monastery or
a married life in the city, the most important is the tension the son
experiences in his relationship with the mother. Let us consider the choice
of the monastery—because we can learn most from the more extreme case.
 Since the monastery can be viewed, as we have already seen, as a kind of
nonbiological family, it is not inaccurate to say about Buddhist monasticism
everywhere—as Lancaster says about Buddhist monasticism in China—that
the monk can “join the new group [the monastery] and break the binds of
the family system and yet find within Buddhism a re-creation of the
family.” That is not, as I said, inaccurate, but it is not entirely accurate
either, because although the monk can re-create in the monastery a
relationship with a “father,” “sons,” and (male) “siblings,” there is one
relationship that he cannot re-create, that with a “mother”! And it is that
pivotal relationship of the East Asian son, I would suggest, that the monk
especially wishes to avoid by joining the monastery. For despite its highly
pleasurable aspects, the young boy’s relationship with the mother, as I have
already stressed, has two potentially frightening dimensions, as well: a
sexual dimension, on the one hand, and a symbiotic one (Mahler et al.
1975) on the other.
Thus, if the highly attentive mother is “seductive” in her relationship
with the son, the intensity of the libidinal dimension in their relationship
can become frightening for him because of its incestuous implications.
Similarly, if, rather than being seductive, the mother is overprotective
toward him, the exaggeration of his dependency on her can become
frightening for the son because it signifies a regressive pull to the symbiotic
state of early infancy in which the psychic differentiation between self and
mother has not yet been achieved. If either alone can be frightening, the
combination can be terrifying. In becoming a monk, then, the son not only
escapes these frightening dimensions of his relationship with the mother,
but he also—because of the monastic rule of celibacy—avoids their re-
creation in a relationship with a wife. (For a more detailed analysis of these
and other motives for monastic recruitment in Southeast Asia, see Spiro
1982b). I am suggesting, then, that the monastery is attractive to those few
men for whom the relationships with mother and wife are too threatening to
sustain because it allows them to escape the former and avoid the latter.

Note
 Reprinted from Religion and Family in East Asia, edited by George DeVos and Takao Sofue. Senri
Ethnological Studies, no. 11. Presented at the Fifth International Symposium, National Museum of
Ethnology, Sept. 1981. (Osaka: National Museum of Ethnology, 1984), pp. 35–54. © 1987 The
Regents of the University of California.

References
Ahem, Emily. 1973. The cult of the dead in a Chinese village.
Stanford, Calif.: Stanford University Press.
Blacker, Carmen. 1975. The Catalpa bow. London: George Allen and
Unwin.
Dutt, Sukumar. 1962. Buddhist monks and monasteries of India.
London: George Allen and Unwin.
Fortes, Meyer. 1961. Pietas in ancestor worship. Journal of the Royal
An-thropological Institute 91:166–91.
_____ 1977. Custom and conscience in anthropological perspective.
International Review of Psychoanalysis 4:127–54.
Freud, Sigmund. 1968. On the universal tendency to debasement in the
sphere of love. In The standard edition of the complete
psychological works of Sigmund Freud, vol. 11 [1912]. London:
The Hogarth Press.
Hsu, Francis L. K. 1971. Under the ancestor’s shadow. Stanford,
Calif.: Stanford University Press.
Jordan, David K. 1972. Gods, ghosts and ancestors: Folk religion in a
Taiwanese village, Berkeley: University of California Press.
Kendall, Laurel. 1984. Korean shamanism: Women’s rites and a
Chinese comparison. In Religion and the family in East Asia, ed.
George DeVos and Takao Softie. Osaka: National Museum of
Ethnology.
Koh, Hesung Chun. 1984. Religion and socialization of women in
Korea. Ibid.
Lancaster, Lewis. 1984. Buddhism and family in East Asia. Ibid.
Lee, Kwang Kyu. 1984. Family and religion in traditional and
contemporary Korea. Ibid.
Mahler, Margaret S., Fred Pine, and Anni Bergman. 1975. The
psychological birth of the human infant. New York: Basic Books.
Morioka, Kiyomi. 1984. Ancestor worship in contemporary Japan:
Continuity and change. In Religion and the family, ed. DeVos and
Sofue.
Ozaki, Masaharu. 1984. The Taoist priesthood: From Tsai-chia to
Ch’u-chia. Ibid.
Roy, Manisha. 1975. The Oedipus complex and the Bengali family in
India (a study of father-daughter relations in Bengal). In
Psychological anthropology, ed. Thomas E. Williams. The Hague:
Mouton.
Sasaki, Kokan. Spirit possession as an indigenous religion in Japan
and Okinawa. In Religion and the family, ed. DeVos and Sofue.
Sofue, Takao. 1984. Family and interpersonal relationships in early
Japan. Ibid.
Spiro, Melford E. 1977. Kinship and marriage in Burma. Berkeley:
University of California Press.
-------- 1978. Burmese supematuralism. 2d exp. ed. Philadelphia:
I.S.H.I.
_____ 1982a. Oedipus in the Trobriands. Chicago: University of
Chicago Press.
_____ 1982b. Buddhism and society: A great tradition and its Burmese
vicissitudes. 2d exp. ed. Berkeley: University of California Press.
Suenari, Michio. 1984. The “religious family” among the Chinese of
central Taiwan. In Religion and the family, ed. DeVos and Sofue.
Tanaka, Masako. 1984. Maternal authority in the Japanese family. Ibid.
Tsuji, Yohko. 1980. Females in Japanese religion. Unpublished ms.
Tu, Wei-Ming. 1984. On neo-Confucianism and human relatedness. In
Religion and the family, ed. DeVos and Sofue.
Yang, C. K. 1969. Chinese communist society: The family and the
village. Cambridge: MIT Press.
Pali Texts (translations)
1957. The Jataka, ed. E. B. Cowell. London: Luzac and Co.
1881. Sutta-Nipata, transi. V. Fausboll. Oxford: Clarendon Press.
12
Symbolism and Functionalism in the
Anthropological Study of Religion

Introduction
FOR THE HISTORIAN and sociologist of science, the history of social science
provides many fascinating problems for inquiry, not the least of which is the
propensity of its practitioners to reject old approaches as false whenever a
new approach acquires saliency. Even Marxist and Hegelian social
scientists for whom antitheses must necessarily eventuate in some kind of
synthesis, nevertheless view the different approaches to sociocultural
inquiry as constituting binary opposites, one of which (their own) is
believed (with all the fervor of a Manichean) to represent the forces of light,
the other the forces of darkness. A relevant case is the rejection of
functionalism by the practitioners of the new symbolist approach to
religious anthropology, much as the functionalists had rejected the
evolutionary approach of the generation before them. The current symbolic
approach, it might be added, is part of the contemporary Zeitgeist;
everywhere, in politics and kinship, as well as in religion and myth,
functionalism is out, symbolism is in, motivation is out, cognition is in,
social processes are out, mental processes are in.
Roughly speaking, we can distinguish at least three different
anthropological approaches to the study of cultural symbols and symbol
systems; (a) phenomenological analysis of the philosophical meanings of
the symbols, (b) structural analysis of the logical relationships among the
symbols and (c) formal semantic analysis of their classificatory schemata.
Although these three approaches are obviously quite different from each
other, they have in common an exclusive attention to symbols and symbol
systems as such, and an avoidance of attention to the relationship between
cultural symbols and social experience. The latter is viewed (at best) as
theoretically uninteresting, and (at worst) as an intellectual sin against the
Cartesian theory of the mind held by many of them.
The previous generation of functionalists were equally dogmatic in the
opposite direction. For many of them, at least, the cognitive meaning of
cultural symbol systems was more or less ignored in favor of their
relationship either to the psychobiological needs of the social actors or to
the functional requirements of their social system. With some few
exceptions the cognitive meaning of the symbols, as such, was of little
interest.
Despite the polemical opposition between symbolic and functionalist
approaches, there is no intrinsic opposition between them. On the contrary,
I would say (to paraphrase Kant) that functionalism without symbolism is
blind, and symbolism without functionalism is lame. For, surely, the social
functions of symbol systems largely depend on their cognitive meaning, and
the meanings which symbol systems have for social actors derive from and
are related to their social context. This, at least, is the thesis I wish to
explore in this chapter, not, however, by a logical analysis of the postulates
and theorems of these respective theoretical approaches, but by a concrete
comparison of a subset of the set of symbol systems of traditional Judaism
and Theravāda Buddhism, namely, their soteriological symbols. My choice
of these two religions stems primarily from the fact that I happen to have
conducted research in two communities—an Israeli kibbutz and a Burmese
village—whose cultures are informed by these two religious traditions.
Theravāda Buddhism comprises the most important idea system (religious
or secular) in Burma; and although the kibbutz (a collective agricultural
settlement) which I studied is atheist, its founders—those who created its
institutions and established its ethos—were reared in, and internalized most
of the basic concepts and values of, orthodox Judaism.
What I hope to do, then, is to compare some basic, or core, symbols of
these religions with respect to their philosophical meaning—I shall forgo a
structural and logical analysis of the symbols—and I shall then attempt to
relate them to the cultural orientations and motivational dispositions found
in Yeigyi, a village in Upper Burma, and Kiryat Yedidim, a kibbutz in
central Israel.
The Soteriological Symbolism of Judaism and
Buddhism

Introduction

Although, at first blush, no two religions appear to be as disparate as

traditional Judaism1 and Theravāda Buddhism,2 when they are looked at
more closely their basic concepts bear a remarkable resemblance to each
other. Both religions are characterized by two sets of conceptual trinities—a
trinity of faith and a trinity of soteriologically valued action—which,
structurally viewed, are almost identical. If this proposition seems
outrageous, consider, for example, the following data. In both religions
there are three ultimately sacred symbols which are taken, in some
important sense, to be the essential instruments of salvation, and which
constitute the irreductible objects of devotion. This trinity—the trinity of
faith, as I shall term it—consists of a sacred being (Jahweh—Buddha),3 a
sacred Law (Torah—Dhamma), revealed by the sacred being, and a sacred
community (Israel—Sańgha), which, prescriptively, lives in accordance
with the Law. In the case of Buddhism, the Buddha, His Law, and His Order
of Monks (Sańgha) are the “Three Gems” in whom, daily, the Buddhist
“take(s) Refuge.” In Judaism, although there can be no god but God, “God,
Torah, and Israel,” as the tradition has it, “are one.”
1. By ‘traditional Judaism’ I refer to pre-World War I Eastern European orthodox Judaism, the
faith in which the founders of Kiryat Yedidim were raised.
2. Theravāda Buddhism, the form of Buddhism found in South and Southeast Asia, is as different
from Mahāyāna Buddhism found in Northern and Eastern Asia as, say Roman Catholicism differs
from Calvinist Protestantism. In the rest of this paper I shall use ‘Buddhism’ to refer to its
Theravādist form exclusively.
3. In this, and in all subsequent contrasted pairs of concepts, the first is Judaic, the second
Buddhist.

In any salvation religion, certain types of action are prescribed as

necessary, if not sufficient, for the achievement of its soteriological goal.
Again, Judaism and Buddhism seem to share a common trinity of
soteriologically oriented action, viz., charity (sedakah—dāna), morality
(miswah—stla), and a form of intellectual activity—study or meditation
(talmud torah—bhāvanā). For both, too, the latter type of action is the most
important; for Buddhism, charity and morality are merely stages en route to
bhāvanā; for Judaism, the study of Scripture is (as the Mishnah puts it)
“equal to all the other commandments combined.”
This structural similarity between these Judaic and Buddhist symbols
seems to pose a serious challenge to that part of functionalist theory which
maintains that there is some kind of “fit” between religious ideology, on the
one hand, and society or personality, on the other. Whether from a Weberian
point of view, in which ideology influences behavior, or from Freudian,
Manheimian, or Marxist points of view, in which ideology is a reflection,
respectively, of unconscious conflict, social structure, or economic
relations, it would not have been expected that these structurally similar
symbols would be found in communities that are so markedly different in
social structure, motivational disposition, and cultural ethos. Whereas the
kibbutz is achievement-, work-, and group-oriented, village Burma is
characterized, on each of these variables, by its polar opposite: work is not
an intrinsic value, achievement (including change) is not an important goal,
and individual concerns take precedence over group values.
Upon closer inspection, however, this challenge to the functionalist thesis
becomes somewhat less formidable. When the Buddhist and Judaic symbols
are compared in terms of their cultural and psychological meanings, their
structural similarities are seen to be more apparent than real, for such a
comparison reveals that whereas there is almost no point at which the
Buddhist symbols articulate with the secular social order (either to give it
value, on the one hand, or to provide a fulcrum by which it can be changed,
on the other), the homologous Judaic symbols make contact with the social
order at almost every point. For purposes of brevity I shall examine this
thesis only with respect to those symbols which comprise the trinity of faith
of these two religions.

Jahweh—Buddha

As traditionally depicted, Jahweh represents activity; He is a world-creating

being. The Buddha, by contrast, represents passivity; He is a contemplative,
world-negating being. Compare, for example, Michelangelo’s wrathful
Moses—for Judaism, Moses and Jahweh are, on many attributes, almost
interchangeable—with the blissful calm of the traditional Buddha image.
Jahweh is not only a world creator, He is a world transformer. He is a
Redeemer, who intercedes in the world to change it. The Buddha, on the
other hand, teaches the Way to redemption, but He himself does not
redeem. His Way, moreover, emphasizes the renunciation of the world, not
its transformation.
Concretely, Jahweh redeemed Israel from Egypt, and, after a long
struggle with a harsh desert environment and with their own
“stubbornness,” He brought them to the land of “milk and honey” which,
however, was acquired only through conquest. The Exodus, Sinai, and the
Conquest of the Land comprise, jointly, the symbols of the nuclear
historical experience of Judaism. This set of symbols implies (a) that
suffering can be overcome in this world, for (b) the world is potentially
good, but that (c) to overcome suffering and to make the world good
requires struggle.
The Buddha is almost the mirror image of Jahweh. As the Prince
Gautama, he had tasted from birth the “milk and honey” of the royal court,
only to reject it to become, in pursuit of Enlightenment, a wandering
mendicant. It is the Buddha’s renunciation of the world, and His ultimate
attainment of Enlightenment which comprise the symbols of the nuclear
historical experience of Buddhism. Milk and honey, this symbol set implies,
(a) are an obstacle to be overcome, not a goal to be desired, for (b) though
suffering can only be overcome by struggle, the struggle is not with
impedimenta of the external world, but with one’s own impulses
(specifically, a “clinging” (tanhā) to the world), and (c) the aim of this
struggle is not to change the world, for since (as we shall see) suffering is
an irreducible attribute of all sentient existence, such an aim is but yet
another snare on the way to liberation. The aim, rather, is to transcend the
seductions of the world by achieving a state of detachment. It is no
accident, surely, that monastic asceticism, which has never found a place in
normative Judaism—the Essenes were viewed as heretics—is the key
institution of Buddhism.
These differences between the Buddha and Jahweh, it should be noted,
are dramatically symbolized in their respective festivals. The important
holy days of Judaism commemorate those events which comprise Israel’s
nuclear historical experience—the redemption from slavery, the revelation
of the Commandments, and deliverance into the Promised Land. The
important Buddhist holy days commemorate the Buddha’s nuclear
experiences—His renunciation of the world and His deliverance from the
realm of desire and attachment. If, then, sacred beings are not merely
objects of devotion, but if they are also models for emulation, then, the
Buddha is a model for the denigration of (and retreat from) the world, while
Jahweh is a model for valuing the world and for attempting to change it.

Torah-Dhamma

The Buddhist Teaching, or Law (Dhamma), stresses two themes: suffering

and release from suffering. Since, according to Buddhism, suffering
(dukkha) is one of the three essential attributes of sentient existence—the
other two are impermanence and nonself—any attempt to change the world
which is based on the assumption that suffering can thereby be eliminated is
irrational. Not only can such attempts not succeed, but they have the
opposite effect; they not only increase suffering by increasing attachment
(tanhā) to the world (the ultimate cause of suffering), but such attachment,
in turn, precludes the attainment of nirvana, the only goal whose attainment
signals the extinction of suffering.
Although the Judaic Law (Torah) is not blind to the existence of
suffering, it views the latter not as an inevitable attribute of existence—after
creating the world, God surveyed His handiwork, and pronounced it as
“very good”—but as retribution for sin. By refraining from sin, i.e., by
living in accordance with the Law, suffering can be avoided. Hence,
according to the Prophets, if suffering is to be overcome, it is necessary to
change the world in compliance with the Law. Since, then, life is basically
(or, at least, potentially) good, to try to make of the world a place in which
its goods can be experienced, is rational, not (as in Buddhism) irrational.
But this contrast between Torah and Dhamma can be drawn even more
sharply. The laws of the Torah were given, as Jahweh told Israel, “so that ye
may live by them.” The Dhamma, on the contrary, was given so that
(literally) one may die by them. Life, according to Buddhism, comprises an
endless round of rebirths (all, from the lowest hell to the highest heaven,
characterized by suffering), and the Law is the means by which the wheel of
life can be brought to an end (through the achievement of nirvana).
 In short, while the Dhamma stresses the rejection and transcendance of
the world as a means to extinction of suffering, the Torah stresses the
acceptance of the “yoke of the Law”—which, in effect, means the changing
of the world—as the means to that end.
This difference between these two religions is exemplified in the
difference between the Jewish and Buddhist male initiation ceremonies.
(Neither has a female initiation.) In Burma, the initiation (shimbuy) is a
reenactment and commemoration of the Buddha’s transformation from
world-embracing prince to world-rejecting mendicant. In Judaism, the
initiation (bar mitzva) signals the acceptance of the yoke of the Law, an
ontogenetic recapitulation of a phylogenetic experience.

Israel—Sahgha

The true Buddhist is the monk. The monastic community (Sańgha) is an

elite, consisting of those few who possess the necessary spiritual
qualifications (pāramttā) to practice the Dhamma in its entirety, including
the 227 regulations that comprise the monastic Rule (Vinaya).
Distinguished from this elite is the great mass of laymen who are spiritually
qualified to follow only the Five Precepts (to refrain from lying, killing,
stealing, drunkenness, and sexual immorality). The layman can only hope
that the piety exhibited in his present birth will enable him to acquire
sufficient merit so that, in a future birth, he too will have the qualifications
for admission to the monastic order.
For Judaism, all of Israel is the elite, all are the Chosen People. Hence,
all must observe the 613 commandments of the Torah. None can be exempt
from their hereditary elitist status—and hence from this responsibility—for
all equally received the Commandments on Sinai. “You only have I known
of all the families of the earth,” Jahweh warns, and, therefore, “I will visit
upon you all your iniquities.” Indeed, it is not far-fetched to argue that Israel
is to the rest of mankind what the Sańgha is to the Buddhist laity, for while
all of Israel must observe the 613 commandments, the Gentiles need
observe the seven Noachite laws only. (These include obedience to
authority, reverence for the Divine Name, and abstinence from idolatry,
incest, murder, robbery, and eating the flesh of a living animal.) This same
theme is implicit in the prophetic notion of Israel as the Suffering Servant.
 The comparison of the respective sacred communities of Buddhism and
Judaism yields yet another important difference. In Buddhism neither
laymen nor Sańgha, whether separately or jointly, comprise a church, a
corporate group. Even within the monastic order each monk (or, at any
event, each monastery) is a law unto himself (or itself). This organizational
feature of the Sańgha parallels the ideological structure of Buddhism: each
individual must seek his own salvation. As the Master stressed in a famous
Sutta, monks must “wander alone like the rhinoceros,” or, as He put it in yet
another metaphor, they must “live as islands unto [themselves].”
For Judaism, on the other hand, the sense of corporate identity and
corporate responsibility is keen. Because of a mythical kin tie—all Israel
are “descendants of Abraham, Isaac, and Jacob”—and because of a
corporate contract—they all entered into the covenant with God—“All
Israel,” according to the Mishnah, “is responsible one for the other.” Hence,
in the liturgical confession of sins, it is not “my” sins, but “our” sins that are
confessed; and it is not for “my” sins, but for “our” sins that punishment is
due. This sense of corporate responsibility, moreover, perdures through
historical time, the sins of the ancestors being visited upon their
descendants even unto three generations.

The Soteriological Goal

Although there is no reason to expect a univocal relationship between

means and ends—many different roads, as they say, can lead to Rome—it
is, nevertheless, not surprising that the dramatic differences between the
belief systems of Buddhism and Judaism should be associated with
radically different notions of redemption. Although both religions postulate
similar conditions of proximate salvation—rebirth into some heaven-like
existence—they postulate radically different conditions of ultimate
salvation.
For Buddhism ultimate salvation consists in nirvana which, whatever else
it might connote—and scholars and laymen alike differ concerning its
meaning—signals the end of the wheel of rebirth. Since all forms of
existence entail suffering, salvation means deliverance from all thirty-one
realms of existence. In addition to the existential meaning of Buddhist
salvation, attention must also be directed to its exclusively individualistic
meaning. Nirvana is individualistic in two senses: the individual is both the
object of salvation—the individual alone can be saved—and the instrument
of salvation—the individual must save himself.
For Judaism, ultimate salvation consists in the attainment of the Kingdom
of God, which is achieved, with the Messianic Coming, at the End of Days.
Unlike the Buddhist case, however, salvation does not entail transcendence
of the physical and social worlds as we know them, but, rather, in their
transformation. In the words of the Prophet, the lion lies down with the
lamb, the crooked is made straight, the people are comforted. It will be
noted, moreover, that unlike the individualistic character of Buddhist
salvation, that of Judaism is corporate. It is the group, Israel, that is saved,
and it is the group, by its collective acceptance of the yoke of the Law, that
is the instrument of salvation.
For Buddhism, then, salvation, (a) is individualistic, and (b) it consists in
the transcendence of the world such that the normal and recognizable
ethical, social, and political—not to mention physical and biological—
categories are rendered nonexistent. For Judaism, salvation (a) is collective,
and (b) it consists in the transformation, rather than the transcendence, of
the sociopolitical world within the known spatio temporal world (however
much the latter might be modified).

Sacred Symbols and Social Behavior

Thus far, I have attempted to show that there are differences between
Buddhist and Judaic soteriologies, that each of these soteriologies is
associated with a set of three sacred symbols (a sacred being, a sacred law,
and a sacred community) which, though structurally isomorphic, are
semantically dissimilar. Although I have only touched the tip of the
symbolic iceberg, I shall forgo the temptation to explore the other
dimensions of these sacred symbols, and, instead, ask yet another question
of them. Do these symbols essentially reflect the nature of the human mind?
(Which, if such were the case, would suggest that the mind, unlike the
claims of some linguistically oriented structuralists, operates not only
according to binary oppositions, but also triadic complementaries.) And, if
this question were to be answered in the affirmative, are such symbols
created, as Lévi-Strauss once suggested about totemic symbols, because
they are good to think, in splendid isolation from any social context?
Although such hypotheses are very much in vogue today, no student,
however, of either the Burmese or the Jews could entertain these hypotheses
for long. Even ignoring the symbolic sets we have been examining—which,
if space permitted, could be shown to be related to certain common
dimensions of Jewish and Buddhist social life—the semantic differences in
the individual symbols would alone infirm these hypotheses. These
differences can be shown to be systematically related to differences in ethos
and behavior found in village Burma and the Israeli kibbutz. To be sure, the
explanations for these relationships are moot. Thus, some theorists would
argue that the relationships are causal, although they would disagree about
the direction of causality (one group contending that the ideas are prior, the
other that the societal forms are prior). Since I wish to avoid such
controversies—my own view, which is different from both, may be found in
my two books on Burmese religion—I am content to observe that such
relationships do in fact exist, and that these religious concepts serve to
express (and perhaps to sustain) the basic components of the ethos of each
culture.
Thus, the Buddhist symbol set, which denigrates the sociocultural world,
which emphasizes the renunciation of the politicoeconomic world, and
which aims at the transcendence of the spatio temporal world is found in a
society (Burma) which places little value on social and economic change,
on achievement values, or on corporate responsibility; in a society whose
cynosure is the world-renouncing monk, and whose ideal is the arhant, the
individual who, oblivious to the suffering of others, seeks and attains
salvation for himself.
The Judaic model which values the sociocultural world, and which
emphasizes the transformation of the politicoeconomic world, is found in a
society (the kibbutz) which places great value on social and economic
revolution, on achievement, and on group responsibility; in a society whose
cynosure and ideal alike is the halutz, whose personal goals are subordinate
to, and who sacrifices his needs for, the transformation of society.
Needless to say, I am not concerned here with evaluating the relative
merits of these different religious orientations and their social
concommitants, nor, qua anthropologist, am I qualified to do so. I am even
less concerned with evaluating the different character types that are
associated with these different orientations, especially since both are the
products of social and individual histories over which social actors have
little if any responsibility. I would like instead to make two additional
observations concerning the relationship between religious symbols and
social behavior. The first is stimulated by a contrast between Buddhism and
Judaism, the second by their similarity.

Religious Symbols and Social Behavior

I have already observed that the conceptions of ultimate salvation found in
Judaism and Buddhism are found side by side with entirely different
conceptions of proximate salvation. In both cases, moreover—and this is
more important—the latter conceptions have usually been the more strongly
cathected by the religious actors. If, as has often been the case, this latter
fact is ignored, we are led not only to a distorted conception of the salient
beliefs of the religious actors, but to a misleading interpretation of the
relationship between religious symbols and social action.
For Buddhism, proximate salvation consists in rebirth in any of the many
material paradises—the abodes of the devas—and, as all anthropological
fieldwork in contemporary Buddhist societies has revealed, it is this form of
salvation that the majority of their members most strongly desire. For
although the aspiration for nirvana is reiterated by the average Buddhist in
ritual formulae, the latter goal is for him more a cultural cliché than a
personal motive; his summum bonum consists not in the extinction of
desire, but in its satisfaction. And if rebirth in a heavenly abode is viewed
as beyond his reach, he is happy to settle for rebirth as a wealthy and high
status citizen of this planet, preferably of his present society. The Buddhist’s
preference for proximate salvation does not, however, require any changes
in the semantics of the soteriological symbols discussed above. Moreover,
the soteriological mechanism remains the same. Whether salvation is
viewed in ultimate or proximate terms, it is achieved as the consequence of
the impersonal law of karma, whose working is exclusively affected by
individual action, and by the merit and demerit which such action produces.
For Judaism, too, it is a heavenly existence which constitutes proximate
salvation, and although the End of Days occupies a central place in Jewish
liturgy, the post-Exilic belief in an afterlife has been at least equally
important in the soteriological concerns of post-Exilic and Diaspora Jewry.
As in the case of Buddhism, this emphasis on proximate salvation requires
little change in the semantics of the soteriological symbols of Judaism. It
does, however, signal a change in the object of salvation from a corporate to
an individualistic orientation. It is individual, not group, action which
determines one’s fate in the afterlife, and it is the individual, not the group,
that is rewarded (or punished). The mechanism of salvation, however, is the
same in both, salvation being achieved not by means of a karma-like
mechanism, but by means of divine intercession. Whether from justice or
mercy, God is influenced by human action to establish His Kingdom on
earth (ultimate salvation), or to send the individual to His heavenly abode
(proximate salvation).
In this connection, there are two psychologically interesting and
historically significant aspects of the soteriology of the Eastern European
founders of the kibbutz. The first aspect, already alluded to, is the
restoration of corporate redemption to its pre-Exilic place of centrality. This
reversal, it should be noted, occurred not only in the kibbutz movement, but
it occurred much earlier in the nonsocialist Zionist movement of Eastern
and Central Europe, as well as in the theology of Western European Reform
Judaism. Whereas the kibbutz founders, however, placed equal stress on
both the parochial (Jewish) and universal (international socialist)
dimensions of the Messianic Days, General Zionism stressed its parochial
dimension, and Reform Judaism its universal dimension. With respect to the
theoretical concern of this paper, viz., the relationship between religious
symbols and social action, this example illustrates the elementary principle
that man manipulates his cultural symbols to serve his own needs. Man is as
much the master, as the servant, of his cultural symbols.
Judaism and Buddhism, surely, are not the only religions in which
alternative and even conflicting symbol systems exist side-by-side, and
their differential appeal in different historical epochs, or to different social
groups within the same historical epoch, remains a little-explored area of
inquiry in religious research. Since these alternative symbol systems are
equally normative, it seems not unlikely that their differential appeal might
be explained by their differential cathexis by the social actors, in short, by
their personality differences. But since in any society (especially if it
constitutes a relatively endogamous breeding population), personality
differences (either among its subgroups or its historical epochs) are a
function of differences in social conditions and their attendant cognitive and
affective consequences, these personality differences are, in turn, related to
sociological differences. Since it would take us too far afield, however, to
examine the possible sociological conditions that contributed to the
renewed emphasis on corporate soteriology in nineteenth and early
twentieth century Judaism, I shall only observe that it is another illustration
of the thesis that religious symbols, no less than any others, do not spring
Athena-like from the human mind; rather, they emerge from concrete social
experiences which act upon the mind and influence its activity.
A second important aspect of the soteriology of the founders of the
kibbutz is the elimination of divine intercession for the achievement of the
soteriological goal. Kibbutz ideology, as I have already mentioned, is based
on a thoroughly naturalistic metaphysics. As in the case of Buddhism, the
soteriological aspirations of the kibbutz—if I may use a religious concept to
characterize a secular belief system—are achieved by human action alone;
its consequences, however, are not deemed to be governed by the law of
karma, but by the laws of sociology and history. Although space, again,
does not permit an examination of the conditions that might possibly
produce such thoroughgoing naturalism, this aspect of kibbutz soteriology
—the absence of divine inter-cession—leads to the final section of this
paper.

Chercher la Femme
Buddhism and Judaism not only share a common trinity of formal
soteriological symbols—a sacred person, a sacred law, and a sacred group
—but these symbols share a common attribute—none of them is feminine.
To be sure, it can be argued—and empirical support can be adduced for
such an argument—that feminist elements have been bootlegged into both
Buddhism and Judaism in the unconscious meanings of some of their other
symbols. Nevertheless, at the level of their formal and conscious meanings,
their symbols, including those that have been examined in this chapter, are
primarily (Judaism) if not exclusively (Buddhism) masculine. I say
“primarily” in the case of Judaism because Israel may be viewed as a
combined male and female symbol since its empirical referent, the people
of Israel, includes both males and females. Even this statement, however,
must be qualified by the observation that the female component in the
symbol, Israel, is only physically, but not ritually or jurally, relevant. God’s
covenant with Israel, which is the jural basis for its status as a sacred
people, is renewed in, and symbolized by, the circumcision ritual, which is
performed for males alone. That, nevertheless, physical membership in the
group is determined by the mother—by Talmudic law, the child of a Jewish
father and a non-Jewish mother is not a Jew—is one of those paradoxes
which might be variously resolved by structuralist, psychoanalytic, and
other theories which delight in paradox. Ritually, too, Israel is essentially a
masculine symbol in that, traditionally, Jewish congregational worship
requires a quorum of ten adult males. Although females, like minors, may
participate in the worship, they may not be counted in the quorum.
Buddhism is even more exclusively masculine in its symbolism than
Judaism. Not only is the historical Buddha, Gautama, a male, but all
previous Buddhas have also been, and all future Buddhas will also be,
males. Buddhahood can be attained only in a masculine form. The Sańgha,
too, is an exclusively male institution. In Judaism the physical Israel, at
least, includes both sexes, even though jurally and ritually females are
peripheral members. This is not the case even with respect to the physical
Sańgha. To be sure, in early Buddhism, females were permitted to become
monks—over the objection, it should be noted, of the Buddha—but after a
few short years they were, and they have continued to be, excluded from the
order. A woman’s aspiration to the monkhood must be deferred until her
future rebirth as a male.
To summarize, then, Judaism and Buddhism alike are characterized by a
pervasive and systematic exclusion of females and feminine symbols from
soteriological, devotional, and liturgical significance. On this dimension
Buddhism and Judaism may not only be classified together, but (among the
great religions) they are members of a class which includes Islam,
Confucianism and Protestantism, and which excludes Hinduism and
Catholicism. Like Buddhism and Judaism, the latter two religions are also
characterized by a soteriological trinity, but, unlike the former, they include
an important feminine component. Now, of course, just as Lévi-Strauss and
the structuralists can always find binary oppositions, so too I can be charged
with always finding trinities. In each of these cases, however, the trinity was
not discovered by fancy intellectual methods; rather trinities are enunciated
by the social actors themselves, and the problem consists in deciding which,
among alternative trinities, is most germane to the present discussion. Thus,
for example, both Hinduism and Catholicism exhibit one type of trinity
which is isomorphic with that of Judaism and Buddhism: a sacred being
(Brahma and God), a sacred law (Dharma and the teachings of the Church),
and a sacred group (Varna and the Church). Unanalyzed, these trinities are
as exclusively masculine as their Judaic and Buddhist counterparts. But the
Hindu “Brahma” and the Catholic “God,” it will be recalled, are generic
terms, best glossed as “divinity” or “godhead,” each of which, depending
on the interpretation, comprises three component deities or aspects of the
godhead.
These divine trinities are, of course, well known. The Hindu trinity
consists of Vishnu, Shiva, and one of the forms taken by the feminine
principle—Sakte—variously conceived as Parvati, Kali, or Durga. The
Catholic trinity consists of God, Christ, and the Virgin Mary. In Catholic
theology, to be sure, the Trinity refers to the triune God—Father, Son, and
Holy Spirit. To the anthropologist, however, who follows the lead of the
religious actor and his devotional concerns, the third member is not the
Holy Spirit, but the Holy Virgin.
Now the difference between the presence or absence of a feminine
symbol within the set of sacred symbols of a religion is not, I need
emphasize, a trivial difference; and it certainly involves much more than a
cognitive preference or lack of preference for binary oppositions (the
opposition, in this case, of male and female). For, although I have thus far
used the abstract concept, “feminine,” in referring to the Virgin and Parvati,
neither, as is well known, represents a generically conceived, abstract,
female principle. Rather, both represent a specific type of female, and that
type is not wife, or sister, or daughter, or aunt, but Mother. Both are
expressly referred to as Mother. The difference, then, between the absence
of a sacred female symbol in Judaism and Buddhism (as well as in Islam,
Protestantism and Confucianism), on the one hand, and its presence in
Catholicism and Hinduism, on the other, is the difference between the
presence or absence of a Mother Goddess. And this difference, in turn, is
related, on the one hand, to the general importance of familial symbols
within the sacred symbol system, and to the role of women in the secular
domain, on the other.
 Of the four religions we are considering here, Buddhism has most
systematically expunged all familial symbols from its sacred symbol
system. It not only excludes all mother symbols, but it also excludes all
father symbols. The Buddha is a male, but he is never designated as father.
Judaism, too, excludes all mother symbols—though the prophets had to
fight a long battle before the various Astartes were expunged from the
religious life of the ancient Hebrews—but the Jewish God is an expressly
conceived father symbol, as is the God of Islam and Protestantism.
Hinduism, on the other hand, includes both father and mother symbols.
With respect to the latter it might also be observed that the mother symbol
is not restricted to the Goddess, for we can hardly ignore the symbolism of
Mother India or of the sacred cow. It should also be added that in the form
of baby Krishna, Hinduism, like Catholicism, includes the sacred child
symbol—and why is the child a son rather than a daughter?—although
father, mother, and son are not integrated to form one sacred family. Family
symbolism, of course, is most pervasive in Catholicism. Not only does the
earthly family find its isomorphic representation in the Divine Family of
Father, Mother, and Son, but the priest is “father,” the nun is “sister,” and
the Church (as well as the nun) is a “bride” (the Bride of Christ), and so on.
Now to pursue all the ramifications of the conceptual, not to mention the
soteriological, differences in the sacred symbols of these religions would
require a monograph. Here, I am interested only in the presence or absence
of the Mother symbol. Since all human beings have mothers, why is this
symbol such a salient feature in some religions, absent in others, and almost
phobically avoided in still others? It is in their different answers to this type
of question that contemporary symbolic approaches differ from
functionalist approaches to the study of religion. The various symbolic
approaches would attempt to answer this question by reference to the
semantics, or syntax, or grammar of the symbol systems themselves, while
the various functionalist approaches would look outside of the symbol
systems for an explanation. A functionalist cannot ignore the obvious
correlations, for example, between the following array of variables in India
and China, respectively. The worship of mother goddesses in India is
associated with an almost obsessive concern with the mother, the sacredness
of the cow, the dietary importance of milk and milk products, the
voluptuousness of sacred female iconography, and the heavy-breasted
woman. The lesser importance of mother goddesses in China, on the other
hand, is associated with the relative unimportance of the mother, the lack of
concern with the cow, the rejection of milk and milk products, the absence
of female iconography, and the flat-chested woman. Given what we know
today about population genetics and the influence of cultural factors on
biological selection, the possibility of a systematic relationship among these
social, cultural and biological variables is not as farfetched as it might
seem.
But these correlations do not provide an explanation for the difference
between the presence or absence of mother symbols in different religions.
Rather, they provide a broader context or frame in which such symbols are
to be viewed, and their systematic covariation with a wide array of variables
suggests to a functionalist that differences in symbols and symbol systems
are best explained by differences in the social systems in which they are
embedded and from which, ex hypothec, they arise and derive their
meaning. He would point, for example, to differences in cultural attitudes to
sex, in the social status of women, in the roles of women as wives and
mothers, and, if he is psychoanalytically oriented, to differences in the
parent-child relationship and, especially, to the variable vicissitudes of the
Oedipus complex.
On the basis of my own work in Buddhist Burma, I would suggest that
the presence or absence of the mother goddess is systematically related to
two feminine roles, those of mother and of wife, and to the differences in a
single dimension of each of these roles, viz., the nurturance dimension of
the mother role and the dominance dimension of the wife role. Although
preliminary research (to be published separately) supports this hypothesis,
for our present purposes the specific empirical findings are less important
than the assumption underlying its construction, viz., religious symbols—
and cultural symbols in general—though created by the human mind, do not
arise from a mental tabula rasa. Although they are elaborated by
philosophical thought, and infused with theological and metaphysical
meaning, these symbols are ultimately the products of individual fantasies
which are produced by (and therefore vary with) different types of social
experience. Like private symbols they serve to express certain aspirations
and needs which arise from such experience, to resolve the conflicts that are
induced by the experience, and to integrate them in a manner which renders
them existentially meaningful. For religious actors, religious symbols are
created not only to think by, but also—and much more important—to live
by.
The Relationship Between the Presence or Absence of Mother Goddesses and Certain Female Roles
mother wife

India + -
Catholicism (Italy) + -
Ceylon + -
Burma + +
Israel + +
Islam (Arab) - -
China - -
Calvinism (Reformation) - -
Key:mother: + = high nurturance
- = low nurturance
wife: + = high dominance
- = low dominance

Note
Reprinted from Science of Religious Studies in Methodology, edited by Lauri Honko. Proceedings
of the Study Conference of the International Association for the History of Religion, Turku, Finland,
Aug. 27–31, 1973. (The Hague: Mouton, 1979), pp. 322–39.
Index

Aberle, D. F., et al., 112

Abnormality: relativism and, 145–59; religion and, 145–59
Absolutism, relativism vs., 147–51
Action systems in religion, 198
Adaptation, human, 24
Adaptive function of religion, 209, 218–19
Adjustive function of religion, 209, 219
Aggression, 1–15; sex and, 250–61
Ahern, Emily, 273
Alliance theorists, 83n
Ancestor worship, 272–75
Anger: in Ilongot culture, 41–46; Western theory of, 41–46
Anxiety, moral, 136–39
Aries, Philippe, 93
Arunta, 230
Atheism in Buddhism, 19
Augustine, St., 20
Austin, L., 225, 230, 232
Australia: Oedipal repression in, 94; paternity explanation in, 223–46, 251

Ban ton, Michael, 187n

Bardwick, Judith M., 65
Bateson, Gregory, 39
Beach, F., 115
Behavior, culture and, 34
Belief(s): culturally constituted, 163–67; defined, 162–63
Belief systems: culturally constituted, 201–5; defined, 198
Benedict, Ruth, 6, 15, 136, 145
Bengal, 91
Bettelheim, Bruno, 65, 81
Biological needs, precultural, 62
Bischof, Norbert, 74, 75n, 76, 85
Blacker, Carmen, 272, 276–78
Bolk, L., 111
Bororo, 252–59
Brown, Norman O., 97
Buddha, 288, 289–90
Buddhism, 190, 193–95; atheism in, 19; Burmese, 18–21, 277–83; in China, 275–78, 279–80;
initiation in, 291; in Japan, 275–78; religious symbols, 295–97; sacred symbols, 294–95;
 soteriology, 193, 288–93; study of, 287–301; Theravada, 151–58, 190, 193–95, 287–301; women
in, 275–78, 297–301
Burma, 18–21; Buddhism in, 294; religion in, 18–21; shamans in, 277; village character, 289;
witchcraft in, 53–54
Burmese monasticism, 151–58
Bushnell, John, 91, 97
Buxbaum, Edith, 81

Campbell, Donald T., 72

Canalization, 121, 127, 133–34
Carpenter, C. R., 88, 115
Carstairs, G. Morris, 91
Cassirer, Ernest, 112
Catholicism, 298
Causal explanation, 199–200, 216–19
Chance, Michael R. H., 85, 88
Child, I., 122
Child extrusion, 93–96
China: ancestor worship in, 272–73; Buddhism in, 275, 283; families in, 271; goddesses in, 276;
incest taboos in, 76–78; marriage in, 76–78, 84
Chomsky, Noam, 46
Christianity: family and, 279n; parthenogenesis in, 234–35
Circumcision, 97
Clitoridectomy, 97
Cognitive desires, in religion, 209
Collective representations, in religious symbol systems, 161
Confucianism, 298
Cooking, 252, 259, 260
Cultural conformity, 110n
Cultural determinism, 36–40; emotion and, 32–58; fallacies in, 24–27; generic, 45; particularistic,
40–46, 56; reason and, 32–58; relativism and, 32–58
Cultural doctrine, 163–64
Cultural frames, 37–39, 47, 50–55
Culturally constituted belief systems, 201–5
Cultural motivation: extrinsic, 117–20; internalized, 135–39; intrinsic, 120–34
Cultural propositions, rational/irrational, 50–54
Cultural relativism, 7–8, 25–27, 36–40, 147, 150–51; emotion and, 32–58; normative, 46–55; reason
and, 32–58
Culture(s): behavior and, 34; conception of, 32–36; defined, 163; global conceptions of, 22; guilt
types, 136–39; holistic conceptions of, 22; human nature and, 3–31, 250–61, incommensurability
of, 56–57; internalized, 27–30; learning of, vs. enculturation, 35–36; personality and, 23–30,
109–44; projection systems in, 9; racist arguments regarding, 7–8; sex consciousness and, 59–71;
shame types, 136–39; universal pattern, 25, 113
Culture-and-personality, 9; in structural beliefs, 166

Daly River tribes, 235

D’Andrade, Roy G., 32n, 36, 161n, 215
Darling, F., 115
Davis, quoted in Tsuji, 276
Defense mechanisms, 180–81; culturally constituted, 145–59; religious systems as, 145–60
Defensive function of religion, 209, 219n
Definition(s): analytic, 188; functional, 191; nominal, 187–88; ostensive, 189, 192; real, 187–92; in
religion, 189–97
Demarest, William J., 76
Devereux, George, 53
De Vos, George, 262n
Dhamma, 288, 290–91
Dharma, 298
Dobuans, paternity theory, 232–33
Dollard, John, 8, 117n
Douglas, Mary, 251. 252, 258n, 261
Dreams: distinctions in, 169–71; poetry and, 178–80; religion analogy, 168–71
Drive-goal connection, 121–25
Du Bois, Cora, 140
Durkheim, Emile, 5, 145, 161, 162, 177, 183, 189–92, 197, 202–4
Dutt, Sukumar, 278

East Asia: family and religion in, 262–85; family tensions in, 268–71, 271–83
Eggan, D., 134
Ego, 27
Egypt, marriage in, 83–84
Elwin, Verrier, 93
Emotion, in culture, 32–58
“Empty box,” 6–7, 23
Enculturation: as humanization, 45; learning a culture vs., 35–36
Erikson, Erik H., 65, 97, 122, 129
Essenes, 290
Evans-Pritchard, E. E., 191
Existentialism, 4
Explanation(s), 177n; causal, 199–200, 216–19; functional, 200–201, 216–19; in social anthropology,
199–201
Expressive desires, in religion, 209, 213–15

Family: Buddhist monasticism and, 278–83; in East Asia, 262–85; kibbutz as, 66–67; monasticism
and, 278–83; religion and, 262–85
Family systems: biological characteristics of, 265–66; cross-cultural regularities in, 265–68
Family tensions in East Asia, 268–71, 271–83
Fantasy as reality, 168–71
Faris, E., 140
Father: ancestor worship and, 272–75; God as, 176; religious beliefs and, 202–5
Father-son relationship, 73–103, 252–60; ancestor worship and, 274–75; in East Asian families, 269
Feuer, Lewis S., 97
Firth, Raymond, 109, 123n, 216
Fortes, Meyer, 203, 272, 274
Fortune, Reo F., 232, 233
Fox, Robin, 75, 76, 85, 93
Frazier, J. G., 76, 100–102, 210
Freud, Anna, 245
Freud, Sigmund, 8, 27–29, 65, 66, 72, 87, 135, 202–4, 239n, 259–61, 263, 270
Fromm, E., 120–21, 239n
Frustrations, 12, 13, 24
Functional analysis, 107–84; personality and, 109–44; social systems and, 109–44
Functional explanations, 200–201, 216–19
Functionalism, 107–84; religion and, 286–301; typology of, 123n
Furuya, Teruko, 275–76

Gathome-Hardy, Jonathan, 93
Ge, 253–59
Gender, 59–71
Genotypes, in culture, 26–27
Gerth, H. H., 196, 209
Gillen, F. J., 230
Girls, parenting need of, 63–71
God as genitor, 235, 238
Goddesses, 299–301; mothers and, 275–78
Goldenweiser, Alexander, 5
Goldschmidt, Walter, 161n
Goody, J., 195
Gough, E. Kathleen, 88, 97
Guilt cultures, 136–39

Hallowell, A. Irving, 8, 10, 112, 116, 134, 135, 145, 146, 150, 209n, 212, 218, 228
Harlow, H., 118
Hamack, A. von, 239n
Headhunting, Ilongot, 43–45
Hempel, C. G., 123n, 187
Henry, J., 120
Herdt, Gilbert H., 93n, 94–96
Hiatt, L. R., 93
Hinduism, 18, 194, 195, 298
Hine, R., 115
Hobbes, Thomas, 260n
Hoijer, Harry, 161n
Honko, Lauri, 286n
Hopi, 212, 213
Hopkins, Keith, 83
Horton, R., 195, 204
Hostility, 11–17, 23; father-son, 73–103
Hsu, Francis L. K., 271
Huang, Chich-Shan, 76
Humanization, enculturation as, 45
Human nature: culture and, 3–31, 250–61
Hunt, Robert C., 91
Husband-wife relationship in East Asian families, 269–71

Id, 27, 250–61

Ifaluk, 9–15, 21; religion, 130–32
Ilongot, 40–46
Immortality, 20–21
Incest, 74–76, 83–84, 87–103, 254–61
Incest taboos, 75–77, 87–103; extinction of, 90–96; outcomes of, 90–103; reaction formation against,
90; repression of, 90–96
Incestuous behavior, in the United States, 74
Incestuous wishes, in dreams, 74
India, Buddhism and, 278
Infant damnation doctrine, 38
Initiation, in Buddhism and Judaism, 291
Initiation rites, Oedipal interpretation of, 93–95
Inkeles, Alex, 134
Institution(s), 114; religion as, 197
Integrative function of religion, 209, 219
Intention in functional analysis, 126n
Interaction in religion, 197–98
Interpretation, 177n
Islam, 298
Israel, 15–17, 21
Israel—Sańgha, 288, 291–92
Italy, 91; Marian cult in, 97

Jahweh—Buddha, 288, 289–91

James, William, 103, 171–72, 282
Japan, 91, 283; ancestor worship in, 273–74; families in, 268–71; goddesses in, 275; shamans in,
276–77
Jataka, The, 279
Jesus, paternity and genealogy, 235–39, 251
Jolly, Clifford J., 85–88
Jones, Ernest, 97, 244
Jordan, David K., 47, 268
Joseph, St., as father, 235, 237–39
Judaism: initiation in, 291; religious symbols, 295–97; sacred symbols, 294–95; soteriological
symbolism, 288–93; study of, 287–301; women in, 297–301
Justice, Blair, 74
Justice, Rita, 74

Kaberry, Phyllis M., 224n, 225, 226, 229

Kachin, 204, 205
Kaffman, Mordecai, 82
Kakar, Sudhir, 91
Kalabari, 204
Kannon, goddess, 275–76
Kant, Immanuel, 287
Kaplan, Bert, 109n, 133
Kardiner, Abram, 8, 202
Kautsky, Karl, 239n
Kendall, Laurel, 272, 276, 277
Kibbutzim, 15–17, 59–71, 78–84, 218n, 294–97; character of, 289; as families, 66–67
Kiryat Yedidim, 60–71, 196
Kluckhohn, C., 134
Koh, Hesung Chun, 278, 281
Koko Ya’o, paternity theory, 231
Korea: ancestor worship in, 272–73; Buddhism in, 280–83; families in, 271, 283; shamans in, 276,
277
Kroeber, Alfred L., 48, 124
Kuangyin, goddess, 275–76
Kuhn, Thomas S., 99
Kummer, Hans, 88
Kwakiutl, 148

La Barre, Weston, 63, 111

Lancaster, Jane B., 85
Lancaster, Lewis, 278–83
Langer, S., 112
Langness, L. L., 94
Language competence, 173–75
Lasswell, Harold D., 97
Law, religious/sacred, 288, 291–92
Leach, E. R., 204, 210, 211, 221, 223–47, 251, 252, 258n, 261
“Learned,” meaning of, 113
Lee, Dorothy, 247n
Lee, Kwang Kyu, 271–73, 282
Lehrman, D., 115
Le Vine, Robert A., 32n
Levinson, D., 134
Lévi-Strauss, Claude, 45, 166, 250–61, 294, 298
Levy, Robert J., 223n
Lewin, K., 118
Lichter, S., 97
Lidz, Ruth W., 94
Lidz, Theodore, 94
Lindzey, Gardner, 74, 101
Linton, R., 114
Love, need for, 23
Lowie, R., 206n

Magic, 50–54
Mahayana Buddhism, 193, 195, 288
Mahler, Margaret, et al., 270, 283
Malinowski, Bronislaw, 72–103, 112, 225–46
Maquet, Jacques, 161n
Marian cult, 97
Marxism, 4
Matrifocal households, 88
Mead, Margaret, 22n, 112, 137
Meaning: of cultural doctrines, 163–64; meanings of, 209
Meggitt, M. J., 230
Mental representations, in religious symbol systems, 161
Merton, Robert K., 114, 123n, 126n, 135, 208, 212, 218
Mexico, 9; Marian cult in, 97
Miller, Neal A., 8, 117n
Mills, C. Wright, 196, 209
Mind, 183; defined, 161–62; religious symbols and, 167–69; theory of, 287
Monasticism, 291–92; Buddhist, 278–83; Burmese, 151–58; family and, 278–83; psychosis and,
155–59
Money, John, 76
Montagu, M. F. Ashley, 111, 226–27, 228, 231, 232
Moral anxiety, 136–39
Morioka, Kiyomi, 273, 274
Mother(s), in East Asian religion, 275–78
Mother-son relationship, 73–103, 253–61; in East Asia, 266, 282–83
Moulton, Ruth, 70
Murdock, George Peter, 48
Murphy, Gardner, 117, 121, 122, 133

Nadel, S. F., 110, 114, 136, 219n

Nagel, Ernest, 123n
Naroll, Raoul, 72
Nayar, 88
National Institute of Mental Health, 161n, 250n
National Institutes of Health, 187n
National Science Foundation, 151n
Nebraska Symposium on Motivation, 117n
Need concept, 122–23
Needs: of children, 23, 24; precultural, 62–71; in religion, 206–8; roles and, 123–25, 128–29
Newcomb, Theodore M., 116
New Guinea, 93n, 94, 96; Oedipal repression in, 94
Nirvana, 19–20, 293
Nuclear complex, 98n

Object constancy, 173–74

Obsession, 50–54
Oedipus complex, 72–106, 262–65; in Australia, 244–46; cross-cultural variability in, 88–98; in East
Asian families, 265; in Israel, 16; structure of, 88–89; in the Trobriands, 72–103; universality of,
72–103
Ojibwa, 11
Okinawa, 277
Ozaki, Masaharu, 279

Pali canon, 194

Pancultural psychological characteristics, 14, 23
Parenting need, sex differences in, 63–71
Parents: religious beliefs and, 202–5; in religious symbolism, 175–77
Parsons, Anne, 91, 97
Parsons, Talcott, 101n, 123n
Parthenogenesis, 223–49
Particularism, universalism vs., 146–47
Paternity, physiological, 223–49
Paul, Robert A., 97n
Peirce, C. S., 32, 169–70, 179
Personality: Burmese monasticism and, 153–59; culture and, 9, 23–27, 27–30, 109–44, 166; as
culture internalized, 27–30; Freudian model of, 27–30; functional analysis and, 109–44;
internalization-of-culture model, 27–30; roles and, 120–25; social systems and, 109–41; universal
pattern, 25
Personal motives, social functions and, 120–34
Phenotypes in culture, 26–27
Philosophy, anthropology and, 3
Physiological paternity, 223–49
Piers, G., 136
Poetry, dreams and, 178–80
Poole, Fitz John, 93n, 161n
Powell, H. A., 93, 232, 233
Precultural needs, 62–71
Preculture, gender and, 59–71
Primitivism, 8
Profane, the, 196
Projection systems, in culture, 9
Protestantism, 298
Pruyser, P., 206n
Psychoanalysis, eulture-personality and, 9
Psychobiological needs, precultural, 62
Psychosis, monasticism and, 155–59
Psychosocial needs, precultural, 62

Radcliffe-Brown, A. R., 119, 217

Radin, Paul, 56, 196
Rainwater, Lee, 88
Rapaport, D., 117n
Read, Kenneth E., 93n, 96
Reality, fantasy as, 168, 171
Reality-testing, 173, 175–78
Reason, in culture, 32–58
Recognition, in functional analysis, 126n Redeemer/redemption, 289–90, 292–93, 296
Reductionism, 7
Reik, Theodore, 94
Relativism: abnormality and, 145–59; absolutism vs., 147–51; cultural, 147, 150–51; cultural
determinism and, 32–58; religion and, 146–59; social, 147–50, 151
Religion: abnormality and, 145–59; adaptive function of, 209, 219; adjustive function of, 209, 218;
anthropological study of, 286–301; in Burma, 18–21; cognitive desires in, 209–11; defensive
function, 219n; definition and exploration, 187–222; distinctions in, 172; dream analogy, 168–71;
in East Asia, 262–85; explanation problems, 199–221; expressive desires in, 209, 213–15; family
and, 267–68; family tensions and, 271–83; functions of, 172–76, 209; hostility and, 13–15;
integrative function of, 209, 219; needs in, 206–8; relativism and, 146–59; substantive desires in,
209, 211–15; symbolism and functionalism in the study of, 286–301
Religious beliefs: causal and functional explanations of, 216–19; defined, 162–63; explanations of,
216–19; functions of, 208–9; practice of, 205–15; “truth” of, 201–5
Religious specialists, mothers and, 275–78
Religious symbol(s), social behavior and, 295–97
Religious symbol systems, representations in, 161–84
Religious systems, as defense mechanisms, 145–60
Renunciation, 290
Representations:
collective, 161; mental, 161; in religious symbol systems, 161–84
Repression, 79, 92, 126–34; sexual, 60–61, 67
Riesman, David, 110n
Ritual in religious interaction, 198
Rivalry, 23
Róheim, Géza, 111, 229n, 230
Roles: needs and, 123–25, 128–29; personality and, 120–25
Rosaldo, Michelle, 36, 39–42, 45, 46
Rosaldo, Renato, 43, 44
Roth, W. E., 210, 211, 223, 226, 229, 233, 236
Rothman, Stanley, 97
Rousseau, J. J., 17, 260n
Rowell, Thelma E., 85
Roy, Manisha, 91, 278
Russell, Bertrand, 38, 163

Sabra counterrevolution, 62–71

Sacred symbols, social behavior and, 288, 293–95
Salvation, 20; Judaic and Buddhist, 272–93. See also Soteriology
Sambia, 94–95
Sańgha, 278, 288, 291–92
Sapir, E., 120
Sasaki, Kokan, 277
Saulteaux, 10, 11
Saussure, Ferdinand de, 34, 170, 172
Schneider, David M., 223n
Schwartz, Theodore, 3n, 161n, 250n
Sex, aggression and, 250–61
Sex-blind environment, 59–71
Sex-role differentiation, 69–70
Sexual dimorphism, in kibbutz culture, 59–71
Sexual equality, 60, 70
Shamans, in East Asia, 277–78
Shame cultures, 136–39
Sharp, Lauriston, 225, 230, 231
Shepher, Joseph, 64, 78, 79
Shils, E., 123n
Shweder, Richard A., 32, 35, 36, 37, 39, 40, 46–48, 50, 51, 55
Sibling rivalry, in East Asian families, 265, 271
Singer, M. B., 136
Sioux, 128–31
Smith, Raymond T., 88
Smith, Robertson, 263
Social anthropology, explanations in, 199–201
Social behavior: religious symbols and, 295–97; sacred symbols and, 288, 293–95
Social conformity, 110n
Social functions: latent, 126–34; manifest, 120–25; personal motives and, 120–34
Social heritage theory, 110
Social relativism, 147–50, 151
Social-sanctions theory, 110
Social systems, 111–17; extrinsic control in, 117–20, 141; functional analysis and, 109–44;
internalized control in, 135–39, 141; intrinsic control in, 120–34, 141; learned, 113–14;
personality and, 109–44
Sofue, Takao, 262n, 271, 275
Son extrusion, 93–96
Son-father hostility, 73–103
Soteriological goals, 292–93
Soteriology: Buddhist, 287, 288–93; Judaic, 288–93, 296–97
Spencer, B., 230
Spirit-children, conception and, 225, 227, 228, 231, 234, 240, 242, 243, 244n
Spiro, Melford E., 78, 79, 93, 116, 131, 139n, 151, 162, 178, 200, 202, 208, 212, 214, 215, 252, 265,
277, 278, 282, 283
Sri Lanka, Buddhism and, 278
Stanner, W. E. H., 230
Steele, James, 153, 154
Structuralism, in cultural beliefs, 166
Structure, deep vs. surface, 26–27
Substantive desires, in religion, 209, 211–13, 214–15
Suenari, Michio, 272
Suffering: in Buddhism, 290; explanation of, 215, 220; in Judaism, 291
Superego, 2; types of, 138–39
Superhuman beings, 21, 192–93, 195; in family and religion, 267–68, 276–78; in religion, 197, 198
Superstition, religion and, 190
Sutta-Nipata, 280
Swartz, Marc, 32, 47, 161n
Symbolism: religion and, 286–301; in religion study, 286–301; soteriological, 289–93
Symbol systems: approaches to the study of, 286–87; religious, 161–84

Tallensi, 203
Talmon, Yonina, 78
Tanaka, Masako, 91, 269–70, 275, 282
Tertullian, 17, 235
Thailand, Burmese in, 18
Theravada Buddhism, 151–58, 190–95, 287–301
Theresa, St., 148
Thompson, W., 115
Thomson, D. F., 229n, 230, 231
Tiger, Lionel, 64
Tinbergen, N., 115
Tolman, E. C., 117n
Torah—Dhamma, 288, 290–91
Totemic symbols, 294
Trinities, 298–99
Trivers, Robert L., 64n
Trobrianders, 230–33, 244n, 247n
Trobriand Islands, 72–103
Truth, of religious beliefs, 201–5
Tsuji, Yohko, 275–76
Tu, Wei Ming, 271
Tully River Blacks, paternity theory, 210, 223, 229, 230, 233, 236
Turner, V. W., 205n
Tuzin, Donald F., 32, 93n, 96, 161n
Tylor, E. B., 110

United States, incestuous behavior in, 74

Universalism, particularism vs., 146–47

Value systems in religion, 198

Vessantara, 279
Virgin birth, 223–49

Walbiri, paternity theory, 230, 232

Warner, W. Lloyd, 218n, 229n, 230, 232, 239n
Weber, Max, 18, 19, 172, 196, 209, 211, 218, 220
West Africa, ancestor worship in, 272
Westermarck, Edward, 75, 76, 78, 83, 87, 101
Wheeler, W. M., 111
White, Leslie, 112
Whiting, J. W. M., 122, 203
Will to believe, 171–72
Wilson, Edward O., 76
Wilson, Monica, 93
Wissler, Clark, 25, 113
Witchcraft, 51–54
Wolf, Arthur P., 76, 77
Women: in Buddhism, 275–78, 297–301; in Judaism, 288–301

Yancey, William L., 88

Yang, C. K., 272
Yir-Yiront, paternity theory, 231
Yurok, 251

Zuckerman, S., 115

Cover