inbreeding, incest,

and the incest taboo

This page intentionally left blank
Inbreeding, Incest,
and the Incest Taboo
t h e s tat e o f k n ow l e d g e
at t h e t u r n o f t h e c e n t u ry

Edited by Arthur P. Wolf

and William H. Durham

stanford university press

stanford, california 2005
Stanford University Press
Stanford, California
© 2004 by the Board of Trustees of the
Leland Stanford Junior University.
All rights reserved.
No part of this book may be reproduced or transmitted
in any form or by any means, electronic or mechanical,
including photocopying and recording, or in any
information storage or retrieval system without the
prior written permission of Stanford University Press.
Printed in the United States of America on acid-free,
archival-quality paper
Library of Congress Cataloging-in-Publication Data
Inbreeding, incest, and the incest taboo :
the state of knowledge at the turn of the century /
edited by Arthur P. Wolf and William H. Durham.
p. cm.
Papers presented at a conference held at Stanford
University, Feb. 24–25, 2000.
Includes bibliographical references and index.
isbn 0-8047-4596-X (cloth : alk. paper)
1. Incest. 2. Consanguinity. 3. Inbreeding.
4. Taboo. I. Wolf, Arthur P. II. Durham, William H.
gn480.3.i53 2005
306.877—dc22 2004003967
Original Printing 2005
Last figure below indicates year of this printing:
14 13 12 11 10 09 08 07 06 05
Typeset by James P. Brommer in 10/12.5 Sabon
Contents

List of Tables and Figures vii

Introduction 1
arthur p. wolf
1. Inbreeding Avoidance and Incest Taboos 24
patrick bateson
2. Genetic Aspects of Inbreeding and Incest 38
alan h. bittles
3. Inbreeding Avoidance in Primates 61
anne pusey
4. Explaining the Westermarck Effect, or,
What Did Natural Selection Select For? 76
arthur p. wolf
5. Ancient Egyptian Sibling Marriage and
the Westermarck Effect 93
walter scheidel
6. From Genes to Incest Taboos: The Crucial Step 109
neven sesardic
7. Assessing the Gaps in Westermarck’s Theory 121
william h. durham
8. Refining the Incest Taboo: With Considerable
Help from Bronislaw Malinowski 139
hill gates
9. Evolutionary Thought and the Current
Clinical Understanding of Incest 161
mark t. erickson
10. The Incest Taboo as Darwinian Natural Right 190
larry arnhart

List of Contributors 219

Index 221
This page intentionally left blank
Tables and Figures

Tables

2.1 Major Types of Consanguineous Relationship 39

2.2 Levels of Death and Defect Reported in Four
Studies of Incest 50
3.1 Avoidance of Sexual Activity with Relatives 64
4.1 General Fertility by Wife’s Age at First Association 80
4.2 Fertility/ Divorce Index by Wife’s Age at First
Association 81
4.3 General Fertility by Husband’s Age at First
Association When Wife’s Age at First Association
Is Zero 84
4.4 Fertility/ Divorce Index by Husband’s Age at First
Association When Wife’s Age at First Association
Is Zero 85
4.5 General Fertility by Wife’s Age at First Association
When Wife Is Adopted Before Husband’s Birth 87
4.6 Fertility/ Divorce Index by Wife’s Age at First
Association When Wife Is Adopted Before
Husband’s Birth 87
4.7 Comparison of the Consequences of Early Association
When the Wife Is an Infant (Tables 4.3 and 4.4) and
the Husband Is an Infant (Tables 4.5 and 4.6) 87
5.1 Sibling Couples Attested in Papyrus Documents
from Roman Egypt (first to third centuries ad) 95
5.2 Marital Fertility in Roman Egypt (mean number
of births per maternal age cohort) 98
5.3 Cases of Divorce by Spousal Age Gap 101
5.4 Birth Intervals of Offspring of Prolific Sibling
Couples Separated by Fewer Than Six Years of Age 102
9.1 Estimated Prevalence of Incest, 1953–1996 164
viii Tables and Figures

Figures

2.1 Map of the United States Indicating States in Which

First-Cousin Marriages Are Legal and Those in Which
They Are Civil or Criminal Offenses 42
3.1 Copulations Between Mothers and Sons and Sisters
and Brothers in the Gombe Chimpanzees 69
4.1 General Fertility by Wife’s Age at First Association 80
4.2 Fertility/ Divorce by Wife’s Age at First Association 81
4.3 General Fertility by Husband’s Age at First Association 84
4.4 Fertility/ Divorce by Husband’s Age at First Association 85
6.1 Aversion and Prohibition: Mismatch 113
6.2 Aversion and Prohibition: No Mismatch 115
6.3 The Expression Hypothesis and the Aversion
Hypothesis: An Epistemological Tension 118
inbreeding, incest,
and the incest taboo
This page intentionally left blank
 Introduction

Arthur P. Wolf

The chapters of this volume are the fruits of a conference held at

Stanford University, February 24–25, 2000. The conference was one of a
series held to inaugurate the university’s new Department of Anthropolog-
ical Sciences. This was a happy—and not entirely fortuitous—conjunction
of topic and occasion. The new department takes as its primary subject the
relationship between biology and culture, and no point of contact between
the two has generated more controversy than the relationship between in-
breeding and the incest taboo.
This was not the first time scholars representing the biological and social
sciences gathered at Stanford to discuss what is commonly called “the in-
cest problem.” In the spring of 1956, seven eminent researchers—David F.
Aberle, Urie Bronfenbrenner, Eckhard H. Hess, Daniel R. Miller, David M.
Schneider, and James N. Spuhler—organized a workshop at the Center for
Advanced Study in the Behavioral Sciences “to consider the problem of the
origins of the incest taboo.”1 Thus we have an appropriate and well-defined
baseline to assess the development of the topic in the second half of the
twentieth century. The changes are profound and will be readily apparent
to anyone who compares Aberle et al.’s arguments with those developed by
Patrick Bateson in Chapter 1 of this volume. I think these changes point to
the direction anthropology is likely to take—and ought to take—in the new
millennium.
In 1878, Mark Twain, traveling up the Rhine on a barge, came to a
small town perched on “an instantaneous hill—a hill two hundred and
fifty or three hundred feet high, and round as a bowl.” It was Dilsberg,
whose 700 inhabitants, Twain learned, were all “blood-kin to each other”
and “have been blood-kin to each other for fifteen hundred years.” The re-
sult, according to the captain of the barge, was that “for ages Dilsberg has
been a thriving and diligent idiot factory.” When, after a visit to Dilsberg,
2 wolf

Twain noted that he saw no idiots there, the captain explained that this
was “Because of late the government has taken to lugging them off to asy-
lums and otherwheres.” Twain comments: “The captain probably imag-
ined all this, as modern science denies that the intermarrying of relatives
deteriorates the stock.”2
This remained the majority opinion of “modern science” for the next
seventy-five years, as is evident in the reaction to Edward Westermarck’s
suggestion that “the psychical cause” of the incest taboo “has a biological
foundation in injurious consequences following unions of the nearest blood
relatives.”3 Although Westermarck could quote in support of his suggestion
the opinions of both Charles Darwin and Alfred Wallace,4 he was widely
accused of ignoring the findings of modern science. Lord Raglan faulted
him for having assumed that inbreeding is harmful “in the face of all the
evidence”5; Bronislaw Malinowski argued against him that “biologists are
in agreement that there is no detrimental effect produced upon the species
by incestuous unions”6; and Robert Briffault claimed that “there is not in
the records of breeding from domesticated animals a single fact . . . which
indicates, much less evidences, that inbreeding, even the closest, is itself
productive of evil effects.”7
In 1934 Westermarck rebutted these criticisms in detail, but his argu-
ments were universally ignored. Writing fifteen years later, Leslie White had
no doubts whatsoever about the relationship between inbreeding and the
incest taboo. There is none. The theory that incest was prohibited because
inbreeding causes biological deterioration is “so plausible as to seem self-
evident, but it is wrong for all that. . . . Inbreeding as such does not cause
degeneration; the testimony of biologists is conclusive on this point.”8 Ac-
cording to White, inbreeding can only intensify the inheritance of traits,
good or bad. If Dilsberg was an idiot factory, it is only because the founders
were inclined toward idiocy. In societies where brother-sister marriage is
permitted in the ruling family, “we may find excellence. Cleopatra was the
offspring of brother-sister marriages continued through several generations
and she was ‘not only handsome, vigorous, intellectual, but also prolific . . .
as perfect a specimen of the human race as could be found in any age or
class of society.’”9
Claude Lévi-Strauss, writing only a year after White, reached the same
conclusion. He acknowledged E. M. East’s work on maize and his view that
“because objectionable recessive traits are as common in the human race as
they are in maize,”10 folk beliefs about the injurious effects of inbreeding
are largely justified. But he then argued that “East’s work has indirectly es-
tablished that these supposed dangers would never have appeared if man-
kind had been endogamous from the beginning.” His conclusion was that
“the temporary danger of exogamous unions, supposing such a danger to
 Introduction 3

exist, obviously stems from an exogamous or pangenetic tradition, but it

cannot be the cause of this tradition.”11 Lévi-Strauss’s preferred authority
was George Dahlberg, who concluded that “as far as heredity is concerned
these inhibitions [i.e., the incest taboos] do not seem to be justified.”12
When Aberle et al. met at Stanford in 1956, they too appear to have
taken the position that inbreeding is not necessarily deleterious. But by the
time they published the results of their deliberations in 1963, they had
changed their minds because of “new information” “which appeared after
the 1956 argument had been developed.”13 This new information was, first,
the finding that “the ratio of deleterious and lethal recessive genes to selec-
tively advantageous genes is very high indeed,” and, second, the finding that
“the percentage of individuals homozygous for lethal or deleterious reces-
sives rises sharply” as degree of relatedness increases. This led to the pub-
lished conclusion that “the biological advantages of the familial incest taboo
cannot be ignored.” “Close inbreeding of an animal like man has definite bi-
ological disadvantages, and these disadvantages are far more evident as re-
spects the mating of primary relatives than as respects other matings.”14 In
other words, Westermarck (and Twain’s captain) were right all along.
The present state of our knowledge of the consequences of inbreeding is
summarized in this volume by Alan H. Bittles in Chapter 2. For Bittles, the
question is not whether inbreeding is injurious; it is how injurious. Bittles’s
concern is that the rarity of sibling and parent-child unions combined with
their being tabooed may have led researchers and makers of social policy to
overestimate the dangers of close inbreeding. One possibility is that “a rig-
orous examination, including determination of paternity, may only be ini-
tiated if a child shows symptoms of physical and /or intellectual handicap.”
Another is that studies of familial incest do not have adequate controls for
“the potentially adverse effects of nongenetic variables, such as very young
or advanced maternal and paternal ages, and unsuccessful attempted inter-
ruption of the pregnancy.”
Bittles’s solution to these problems is to use cousin and uncle-niece unions
to estimate the dangers of sibling and parent-child unions. The advantage of
this method is that in many parts of the world—most notably Japan, South
India, and Pakistan—cousin and uncle-niece unions are both legal and com-
mon. Consequently, we now have numerous studies (many conducted by
Bittles himself) of the biological consequences of such marriages. Estimates
based on these studies put the excess death rate among the progeny of sib-
ling and parent-child unions at 16 to 20 percent, and the excess morbidity
rate for such progeny at 6 to 16 percent. This suggests a total death and ma-
jor disabilities rate of somewhere between 22 and 36 percent.
Although the “new information” that changed Aberle et al.’s minds indi-
cated that inbreeding was at least as dangerous as Bittles’s estimates imply,
4 wolf

they rejected the possibility that primitive man recognized this danger.15
This was a consequential decision, because it denied them a simple expla-
nation of the incest taboo. They could not argue that it is a consciously in-
stituted prophylactic. But how, then, did the taboo come into being? Proba-
bly because Eckhard H. Hess was a member of their workshop, Aberle et
al. seriously considered the possibility that it expresses a disposition found
in other species. Hess had shown that so long as they are reared together,
Canada geese from the same brood never mate.
Experimental work . . . indicates that this fastidious behavior is the result of sexual
imprinting. It is necessary to emphasize that the reaction persists without external
sanctions. The luckless breeder who takes a male and female from the same brood
to raise geese is doomed to disappointment: the pair will not mate even if no other
partners are available. If, however, two members of the same brood are separated
before hatching occurs and are subsequently re-introduced to each other, having
been raised in different families, they may become mates.16

Although Aberle et al. failed to make the connection, this is evidence for
what has long been known as “the Westermarck hypothesis.” Westermarck
argued that close inbreeding is injurious, but he did not argue (as Aberle et
al. mistakenly imply) that recognition of the fact led to the incest taboo. In-
stead, he argued that the deleterious consequences of inbreeding have se-
lected for an innate tendency to develop an aversion to sexual relations
with childhood associates. This tendency, not recognition of the dangers of
inbreeding, was the source of the incest taboo. As he put it in 1934 in what
turned out to be his last words on the subject,
I must confess that the attempts to prove the harmlessness of even the closest in-
breeding have not shaken my opinion that there is convincing evidence to the con-
trary. And here I find, as before, a satisfactory explanation of the want of inclina-
tion for, and consequent aversion to, sexual intercourse between persons who from
childhood have lived together in that close intimacy which characterises the mutual
relations of the nearest kindred. We may assume that in this, as in other cases, nat-
ural selection has operated, and by eliminating destructive tendencies and preserv-
ing useful variations has moulded the sexual instinct so as to meet the requirements
of species.17

Might it not be, then, that Westermarck was right about the effects of
early association as well as the dangers of inbreeding? Indeed, might it not
be that Alfred Wallace was right in thinking that Westermarck had “solved
the [incest] problem”?18 Aberle et al. do not give the possibility a moment’s
consideration. They mention what they call “the indifference theory . . .
only for the sake of completeness”: “The indifference theory has both log-
ical and empirical difficulties. It is hard to see why what is naturally repug-
nant should be tabooed, and the evidence for sexual attraction among kins-
 Introduction 5

men is quite adequate for rejecting the theory. We mention it only for the
sake of completeness.”19
The “logical difficulties” refer to Sir James Frazer’s claim that the exis-
tence of the incest taboo is alone adequate to prove that Westermarck was
wrong. In Sir James’s words,
It is not easy to see why any deep human instinct should need to be reinforced by
law. There is no law commanding men to eat or drink or forbidding them to put
their hands in the fire. Men eat and drink and keep their hands out of the fire in-
stinctively for fear of natural not legal penalties. . . . The law only forbids men to
do what their instincts incline them to do; what nature itself prohibits and punishes,
it would be superfluous for the law to prohibit and punish. Accordingly, we may al-
ways safely assume that crimes forbidden by law are crimes that many men have a
natural propensity to commit. If there was no such propensity there would be no
such crimes, and if no such crimes were committed what need to forbid them? In-
stead of assuming, therefore, from the legal prohibition of incest that there is a nat-
ural aversion to incest, we ought rather to assume that there is a natural instinct in
favour of it, and that if the law represses it, as it represses other natural instincts, it
does so because civilized men have come to the conclusion that the satisfaction of
these natural instincts is detrimental to the general interests of society.20

This argument has been repeated, mantralike, by Westermarck’s many crit-

ics. It was quoted in full by Sigmund Freud as early as 1911 and by Mau-
rice Godelier as late 1989.21 It was noted by critics as diverse in their views
as George Peter Murdock, Claude Lévi-Strauss, Marshall Sahlins, and
Richard Lewontin. It should be entered as exhibit 1 by anyone arguing that
twentieth-century social thought was biased toward what I call functional-
ist fundamentalism.
The empirical difficulty Aberle et al. have in mind is Freud’s claim that
“psychoanalytic investigations have shown beyond the possibility of doubt
that an incestuous love choice is in fact the first and regular one.”22 This is
another anti-Westermarck mantra. It is repeated by A. L. Kroeber, Brenda
Seligman, Leslie White, Alexander Goldenweiser, Claude Lévi-Strauss, and
Marvin Harris, to name only the best-known of the many authors who
quoted Freud against Westermarck. It is an essential part of functionalist
fundamentalism because the view that human nature is selfish and unruly
(perhaps even sinful) is necessary to the view that laws exist because they
are needed.
Although Westermarck cited evidence suggesting that early association in-
hibits sexual attraction among many mammals (including horses and dogs),
Aberle et al. ignored this and the experience of generations of animal breed-
ers. It appears that they cited Hess only because it would have been impolite
to ignore the work of an eminent colleague. Their unqualified conclusion
was that “there is no evidence to suggest that asexual imprinting occurs
6 wolf

among mammals.” Asexual imprinting “does not seem to occur in man, the
apes, the monkeys, or even in more remote mammalian species.” We have
therefore to assume that “this adaptive device was simply not available—not
a part of the genetic equipment of man’s ancestors or relatives.”23
Although Aberle et al.’s conclusion regarding “more remote mammalian
species” was unjustified, they can be excused for concluding that sexual im-
printing does not seem to occur among the apes and monkeys. They wrote
before primatology was an established research field. The difference this has
made is dramatically summarized in Chapter 3, by Anne Pusey. After briefly
reviewing evidence suggesting that inbreeding is injurious for most mam-
mals (and more so in the wild than in captivity), Pusey catalogs a wealth of
evidence indicating that something like asexual imprinting is found among
our nearest relatives—rhesus macaques, baboons, gorillas, bonobos, and
chimpanzees. Field and laboratory studies of “nonhuman primates provide
abundant evidence for an inhibition of sexual behavior among closely re-
lated adults,” and “the primate data support Westermarck’s theory that fa-
miliarity during immaturity is a major reason for this avoidance.” In several
species immature mates do engage related females sexually, “but [this be-
havior] stops before the risk of conception.”
The incest taboo posed a nearly impossible task for the functionalist fun-
damentalists. Rejecting Westermarck in favor of Freud, they had to find a
supranatural source for the taboo. Their solution was to resurrect and re-
model Edward Burnett Tylor’s 1889 suggestion that “among tribes of low
culture there is but one means of keeping up permanent alliances, and that
is by means of intermarriage.”24 The essential first step in the argument was
to insist that the incest taboo is only a way of implementing exogamy.
“Nuer say that marriage to persons standing in certain relationships is for-
bidden because it is rual, incestuous,” but E. E. Evans-Pritchard argued that
“we may reverse this statement and say that sexual relations with persons
standing in these relationships are considered incestuous because it would
be a breach of the marriage prohibitions to marry them. I would hold that
the incest taboo can only be understood by reference to the marriage prohi-
bitions, and that these prohibitions must be viewed in the light of their
function in the Nuer kinship system and in their whole social structure.”25
Putting exogamy before the incest taboo led to the remarkable conclu-
sion that the incest taboo is the means by which human beings transcended
their animal nature. For Leslie White and Claude Lévi-Strauss, this made
the incest taboo the passage between nature and culture. I put their formu-
lations side by side to show how two authors who shared little else reached
the same conclusion about the origins of the taboo. The similarity is evi-
dence that they were responding to intellectual trends larger than them-
selves. First, Leslie White:
 Introduction 7

In the primate order . . . the social relationships between mates, parents and chil-
dren, and among siblings antedates articulate speech and cooperation. They are
strong as well as primary. And, just as the earliest cooperative group was built upon
these social ties, so would a subsequent extension of mutual aid have to reckon
with them. At this point we run squarely against the tendency to mate with an inti-
mate associate. Cooperation between families cannot be established if parent mar-
ries child; and brother, sister. A way must be found to overcome this centripetal ten-
dency with a centrifugal force. This way was found in the definition and
prohibition of incest. If persons were forbidden to marry their parents or siblings
they would be compelled to marry into some other family—or remain celibate,
which is contrary to the nature of primates. The leap was taken; a way was found
to unite families with one another, and social evolution as a human affair was
launched upon its career. It would be difficult to exaggerate the significance of this
step. Unless some way had been found to establish strong and enduring ties be-
tween families, social evolution could have gone no further on the human level than
among the anthropoids.26

And then Claude Lévi-Strauss:

It will never be sufficiently emphasized that, if social organization had a beginning,
this could only have consisted in the incest prohibition since, as we have just shown,
the incest prohibition is, in fact, a kind of remodeling of the biological conditions of
mating and procreation (which know no rule, as can be seen from observing animal
life) compelling them to become perpetuated only in an artificial framework of ta-
boos and obligations. It is there, and only there, that we find a passage from nature
to culture, from animal to human life, and that we are in position to understand the
very essence of their articulation.
As Tylor has shown almost a century ago, the ultimate explanation is probably
that mankind has understood very early that, in order to free itself from a wild strug-
gle for existence, it was confronted with the very simple choice of “either marrying-
out or being killed-out.” The alternative was between biological families living in
juxtaposition and endeavoring to remain closed, self-perpetuating units, over-ridden
by their fears, hatreds, and ignorances, and the systematic establishment, through
the incest prohibition, of links of intermarriage between them, thus succeeding to
build, out of the artificial bonds of affinity, a true human society, despite, and even
in contradiction with, the isolating influence of consanguinity.27

Although Les Structures élémentaires de la parenté appeared in 1950

(and was summarized in English in an article published in 1956), Aberle et
al. do not refer to Lévi-Strauss, but they do devote substantial space to
White’s version of what they call “the social and cultural systems theory”
(which I prefer to call “group alliance theory”). In their view, “it is clear
that the advantages postulated by White exist, and that, given a tendency
to choose the most easily available mate, a complete prohibition on famil-
ial sexual relations is the simplest device for forcing ties between fami-
lies.” But they were not functionalists of the fundamentalist variety. They
8 wolf

worried “that [White’s] theory seems to assert that because the shift was
advantageous, it came into being,” and they worried that to come into be-
ing, the shift would “require a movement in opposition to certain strong
trends.”28
It requires the elimination of some younger members from the family, in spite of emo-
tional attachments, and entrusting these members to groups where stable relation-
ships do not yet exist. It also requires that primitive man understand the advantages
of exchange—or else must assume that familial exogamy and the familial taboo
arose as a chance “mutation” and survived because of their adaptive character.29

But what, then, could have motivated the change?

Aberle et al.’s premises put them in an awkward position. They accepted
the evidence indicating that inbreeding is injurious, but they also accepted
Freud’s claim that human beings are naturally inclined to mate and marry
within the family. Moreover, they rejected the idea that “primitive men”
(whoever they might be) would have recognized either the disadvantages of
inbreeding or the advantages of alliances. Having gotten themselves into
this position by accepting Freud, they naturally turned to Freud to solve
their problem. After his famous reconstruction of what he called the “emo-
tional motive” for the incest taboo, Freud went on to argue that “it has a
practical basis as well.”
Sexual desires do not unite men but divide them. Though the brothers had banded to-
gether in order to overcome their father, they were all one another’s rivals in regard
to the women. Each of them would have wished, like his father, to have all the
women to himself. The new organization would have collapsed in a struggle of all
against all, for none of them was of such overmastering strength as to be able to take
on his father’s part with success. Thus the brothers had no alternative, if they were
to live together, but—not, perhaps until they had passed through many dangerous
crises—to institute the law against incest, by which they renounced the women
whom they desired and who had been their chief motive for dispatching their father.
In this way they rescued the organization which had made them strong.30

Citing Freud as the originator of what they call “the family theory” (and I
call “group harmony theory”), Aberle et al. argued that the incest taboo was
instituted to maintain order in the family. This was possible for an animal
with “language and limited culture” because domestic strife would “be ob-
servable as a pressing problem, on a day-to-day basis, and the source of the
problem in sexual competition would be equally evident.” They recognized
that the problem might be solved by regulating sex rather than by eliminat-
ing it, but argued that this solution would not survive over time because it
would not solve the genetic problem posed by the dangers of inbreeding.31
Although Aberle et al. avoided many of the mistakes made by White
and Lévi-Strauss, they ended up with a story that is no more convincing. It
 Introduction 9

is, as Hill Gates asserts in Chapter 8 of this volume, the result of anthro-
pology’s “embrace . . . of Sigmund Freud’s Oedipus complex and its odd al-
liance with contract theory.” The incest taboo is taken as promising, if not
utopia, at least a healthier, more orderly existence. But human nature, in-
cestuously inclined, stands in the way. A means must be found to overcome
it. Human beings must see that to fully realize their potential they have to
repress their selfish sexual interests. It is a particularly attractive parable
because everyone knows that the story has a happy ending. Somehow,
somewhere, for some reason or other, our ancestors saw the light and made
the necessary sacrifices. Rationality triumphed. All good men agreed that
they would forgo their sisters and exchange them for wives.
The antihero of this heroic tale is Edward Westermarck. If early associa-
tion were to be found to inhibit sexual attraction among humans beings as
well as among geese, the plot would lose its dramatic motive and much of
its appeal. Our ancestors would not have had to repress their natural incli-
nations to harvest the advantages of outbreeding and exogamy. They
would be guaranteed by their natural inclinations. It is, then, ironic that
even before Aberle et al.’s version of the story appeared in print, the Frazer/
Freud tide had turned. In 1962 Robin Fox published an essay in which he
argued that reaction to the possibility of sex among persons who have ex-
perienced close bodily contact as children “varies from ‘disgusting’ or ‘un-
thinkable’ to ‘indifferent.’” “It is the reaction of indifference that we find
most interesting, and most neglected, due to the facile rejection of Wester-
marck’s observation.”32
Fox’s bellwether essay was followed two years later by Yonina Talmon’s
study of sexual relations among children reared together in two Israeli kib-
butzim, four years later by my first report of the sexual consequences of mi-
nor marriages in Taiwan, and less than a decade later by Joseph Shepher’s
survey of a large sample of marriages in Israel. All three studies documented
“a lack of inclination for . . . sexual relations between persons who have
lived together in a long-continued relationship from a period of life when
the actions of sexual desire, in its acuter forms at least, is naturally out of
the question.”33
A Westermarck revival was under way. In the years since, it has amassed
evidence that leaves little doubt that Aberle et al. erred in concluding that
asexual imprinting does not have an analogue among humans. The only
ethnographic case that could ever be mustered in support of their conclu-
sion—brother-sister marriage in Roman Egypt—is nullified by Walter Schei-
del in Chapter 5 of this volume. After carefully reexamining the forty-six
known cases of sibling and half-sibling marriages, he concludes that “all in
all, there is nothing to show that as far as the correlation of early childhood
association and sexual inhibition is concerned, the evidence for Roman
10 wolf

Egyptian sibling marriage deviates significantly from the pattern derived from
the Chinese data on ‘minor marriages’ and other information on the dem-
ographic context of incestuous behavior and incest avoidance in humans.”
Thus, the second half of the twentieth century saw two major changes in
the assumptions researchers bring to the incest problem. The first is that
close inbreeding is injurious. Denied by White in 1949, by Lévi-Strauss in
1950, and probably by Aberle et al. in 1956, the position advocated by
Westermarck since 1890 was well on its way to general acceptance by 1963.
The second and equally important change was the discovery that among
most mammals and all the primates—including, most definitely, humans—
early association inhibits sexual attraction. Again Westermarck was proved
right. Thus the man who was mentioned in 1963 “only for the sake of
completeness” enters the twenty-first century as almost the only man worth
mentioning.
Recognition of the importance of these changes is what unites the chap-
ters in this volume, but they are not the focal subjects of the volume. There
are three problems raised by the discovery that inbreeding is injurious and
early association inhibiting. I call them the mediation problem (A), the
representation problem (B), and the localization problem (C). The media-
tion problem is, How are the deleterious consequences of inbreeding and
the inhibiting nature of early association related? Most of the contributors
to the volume are willing to assume that they are linked as cause and ef-
fect, but William Durham, volume editor and author of Chapter 7, de-
murs. He worries that “the Westermarck effect” may not be an adapta-
tion. “What is lacking,” in his view, “is conclusive evidence to show that
the aversion was specifically shaped over time by genetic selection for the
function it now performs.”
I think doubts of this kind arise because we still do not know how the
Westermarck effect is effected. In other words, we still do not know what
causes us to respond to early association with an enduring aversion. I call
the search for this cause the mediation problem because I am confident that
whatever it is, it is a product of the dangers of inbreeding. It is what this se-
lective force selected for. My hope is that when the mediation problem is
solved the solution will convince skeptics like Durham that “in this case, as
in other cases, natural selection has operated . . . so as to meet the require-
ments of the species.”
When we solve the mediation problem (A), we will know why people
avoid incest, but we will still not have an answer to the question that Bate-
son puts at “the heart of the matter”: “What relations, if any, can be found
between the avoidance of inbreeding and the incest taboo?” In other words,
we will still have to solve the representation problem (B). Generally speak-
ing, this is the problem of how the loves and hopes and fears and phobias
 Introduction 11

of individuals give rise to norms, if they do. Bernard Williams (who was
the first to use the term representation problem) puts it this way:
It is the notion of a norm that perhaps gives rise to the central representation prob-
lem. . . . The most, it seems, that a genetically acquired character could yield would
be an inhibition against behaviours of a certain kind; what relation could that have
to a socially sanctioned prohibition? Indeed, if the inhibition exists, what need
could there be for such a prohibition? If a prohibitionary norm is to be part of the
“extended phenotype” of the species, how could we conceive, starting from an in-
hibition, that this should come about?34

As I analyze it, the representation problem consists of a cluster of three

related problems concerning the relationship between individual inclina-
tions and social regulations. I call them the externalization problem (B1),
the expression problem (B2), and the moralization problem (B3). The ex-
ternalization problem concerns the fact that the incest taboo is not a mat-
ter of self-regulation. It is a matter of public condemnation. The fact that
early association inhibits sexual attraction explains why most people avoid
sexual relations with their parents and siblings, but it does not explain why
they condemn other people for having sexual relations with their parent or
their sibling.
Although it is implied by his general statement, Williams did not address
the externalization problem (B1). For him, the core of the representation
problem is what I call the expression problem (B2). And, in his view, it is an
insoluble problem. For Williams there is, as Neven Sesardic puts it in Chap-
ter 6, a “transcendental obstacle” blocking all movement from biologically
based inhibitions to socially sanctioned prohibitions. I call it “the expression
problem” because it concerns the conceptual content of the incest taboo.
The argument is that the aversion aroused by early association cannot pos-
sibly explain the taboo because they are about different things. One is about
“those [people] that one is brought up with”; the other is about “marriages
that could constitute close inbreeding.” In Williams’s own words,
Not only does extra conceptual content have to be introduced to characterize the
human prohibition, but also the introduction of that content stands in conflict with
the proposed explanation of it. There are no sanctions against marrying those that
one is brought up with (as such); the sanction is against marriages which would
constitute close inbreeding. The conceptual content of the prohibition is thus dif-
ferent from the content that occurs in the description of the inhibition. It indeed re-
lates to the suggested function of that inhibition, but that fact will not explain how
the prohibition which is explicitly against inbreeding will have arisen. It certainly
does not represent a mere “raising to consciousness” of the inhibition.

Although it is also implied by his general statement of the representation

problem (B), Williams fails to address what I call the moralization problem
12 wolf

(B3). This is surprising given his interest in morality, because the problem
arises from the fact that universally the incest taboo is represented as having
strongly felt moral content. Not only do people disapprove of incest, their
disapproval is accepted as morally motivated. A solution to what I call the
externalization problem would explain why people condemn incest, but it
would not explain why such condemnation elicits universal approbation.
Neither the mediation problem (A) nor the representation problem (B)
were seriously discussed until the end of the twentieth century. Until the
early 1970s this was because most researchers accepted White and Lévi-
Strauss’s contention that the source of the incest taboo was supranatural;
afterward, it was because the succeeding generation of cultural relativists
decided there was no incest taboo. Their reasons were succinctly stated by
Rodney Needham as early as 1971. “I conclude,” Needham wrote, “that
‘incest’ is a mistaken sociological concept and not a universal.” There were,
in Needham’s view, two reasons for this conclusion:
The first is the wide and variable range of statuses to which the prohibitions apply.
The scope of application is in each case an integral feature of the social system, and
in some sense a function of it; i.e., the complex of prohibitions in a society cannot
be comprehended except by a systematic purview of the institutions with which
they are implicated. By this account of the matter there are as many different kinds
of incest prohibitions as there are discriminable social systems.
The second consideration is that the incest prohibitions are in part moral in-
junctions; they are expressions of indigenous ethical doctrines and, whether or not
they are touched with a peculiar emotional quality, they have cultural meanings
which no attempt at explanation can reasonably neglect.35

Needham’s arguments were quickly seconded by David M. Schneider,

Peter Riviere, and Roy Wagner. In fact, arguments that were original with
Needham in 1971 were orthodoxy by the end of the decade. Thus, my pre-
ceding account of the fate of Aberle et al. is clearly “a presentist” version
of history. What I have called a Westermarck revival was only a minor
countercurrent in a tide drawn by the view that there is no incest taboo at
all, only clusters of cultural particulars. The history of the period would be
better represented by focusing on David M. Schneider than on Edward
Westermarck. Although Schneider was a member of the Aberle et al. group
(and, I was once told, drafted the report they published), he soon followed
Needham’s lead and abandoned the incest taboo as “a mistaken sociologi-
cal concept.” In a paper published in 1976 his “main point” with respect to
the problem of incest was “to stop looking for causal explanations of ori-
gin, functional explanations of maintenance and to start looking at it as a
problem in meaning in its cultural context.”36
Now that the high tide of cultural relativism is receding, we can see that
though it tried, it did not succeed in dissolving all social phenomena into
 Introduction 13

their manifest qualities. There is, contra Needham and those who followed
his lead, a universal incest taboo. But to insist on this does not deny that
“there are as many different kinds of incest prohibitions as there are dis-
criminable social systems.” Needham was right about this. He was justified
in claiming that “the range of structures to which the [incest] prohibitions
apply” is “wide and variable,” and he was justified in claiming that incest
taboos are moral injunctions that express what he calls “indigenous ethical
doctrines.”
There is, then, a third problem to be faced—in addition to the media-
tion problem (A) and the representation problem (B). This is what I call
the localization problem (C). In general terms, it is the problem of how
and why a universal tendency is implemented in diverse ways. The prob-
lem is represented in Chapter 8 of this volume by Hill Gates’s reanalysis of
Bronislaw Malinowski’s account of the incest taboo in the Trobriand Is-
lands. Why do the Trobrianders disapprove of sexual relations among all
matrilineally related boys and girls as well as among siblings? Why does
the strength of this disapproval diminish as the matrilineal connection be-
comes more remote? Why do the Trobrianders regard sibling incest as par-
ticularly horrifying? Why do they not include parental incest in the same
class of delicts as maternal incest and sibling incest? Why do some Trobri-
and clans tolerate incest on the part of what Malinowski calls their “chiefs”?
And so forth.
Although they rarely appear in the service of an attempt to link the dan-
gers of inbreeding or the effects of early association to the incest taboo of a
particular society, arguments addressed to the localization problem are
common. Indeed, most of what anthropologists have written about the in-
cest taboo belongs to this genre. The mediation problem and the represen-
tation problem, in sharp contrast, have no established place in the anthro-
pological literature. Before 1983 only Westermarck had addressed any part
of the representation problem, and before 1989 no one, not even Wester-
marck, had addressed the mediation problem. Thus the chapters in this
volume mark a sharp turn in the direction of scholarly interests. More than
half are primarily concerned with one aspect or another of either the medi-
ation problem or the representation problem.
The mediation problem (A) is addressed most directly by Mark Erick-
son, in Chapter 9, and myself, in Chapter 4 (though Anne Pusey comments
on it at the end of Chapter 3). The argument we make is a version of a hy-
pothesis first advanced by Erickson in 1989. In Chapter 9 he puts the argu-
ment in historical context by noting that while Westermarck did not pro-
pose “a psychology of incest avoidance,” Freud did. This was not an easy
task for Freud, because it involved resolving a dilemma created by his as-
sumption that “all social bonds are ultimately sexual.” If all social bonds
14 wolf

are ultimately sexual, how could it be that “sexuality is inhibited among

kin”? The answer is, of course, the deus ex machina of the Oedipal drama
—fear of castration. Discovering that his sister has no penis, every little boy
fears that he will suffer the same fate if he persists in wanting his mother.
And thus, if one can accept that fear of castration is universal, a natural
preference for incest is turned into incest avoidance.
Erickson and I avoid Freud’s dilemma—and his desperate resolution—
by arguing that a distinction must be made between sexual bonds and
asexual bonds. The bonds an infant forms with the mother and other care-
takers are fundamentally different from those formed between adult sexual
partners. Infant /caretaker bonds are inherently contrasexual. Our argu-
ment is inspired by John Bowlby’s account of what he calls “attachment
behavior.” What we add is the thought that Bowlby’s attachment and
Westermarck’s aversion are two aspects of the bonds formed in infancy and
early childhood. It could not be otherwise, because without the aversion,
childhood attachments would lead to incest, and without attachment,
childhood aversion would lead to abandonment. To evolve at all, the two
dispositions had to evolve together.
Thus part of our solution to the mediation problem (A) is to suggest that
what natural selection selected for is a universal disposition to form con-
trasexual attachments to those persons by whom and with whom one is
reared. Putting together the evidence presented in this volume with that re-
viewed in a recent handbook of attachment behavior,37 this part of our ar-
gument appears well founded. The problem is how to account for the atti-
tude of parents and other caretakers. We have to assume that they are
somehow inhibited by association with their children. But how? Bowlby
suggested that attachment behavior naturally elicits what he calls “caretak-
ing.” Thus the simplest solution to our problem would be to suggest that
caretaking, like attachment, is inherently contrasexual, but not all the evi-
dence now available is encouraging. The evidence cataloged in Chapter 9
suggests that caretaking is not necessarily contrasexual or that attachment
is not always successful in eliciting caretaking.
The representation problem (B) is addressed in this volume by Patrick
Bateson (Chapter 1), William Durham (Chapter 7), and Neven Sesardic
(Chapter 6). I will consider first that part of the problem I call the exter-
nalization problem (B1). Why do we condemn others for having sex with
their relatives? What has this to do with our not being interested in having
sex with our relatives? Westermarck’s answer was simple (at least as long
as we allow Smithian sympathy as a universal human attribute). Sexual
aversion of the kind aroused by early association is painful, but such pain
is not part of our everyday feelings toward our close relatives. Our every-
day attitude toward their sexuality is a comfortable indifference. Our aver-
 Introduction 15

sion is aroused only when “the idea of sexual relations with a near relative
occupies the mind with sufficient intensity and a desire fails to appear.”38
The reason we condemn other people for having sex with their relatives is
because it does just this to us. We condemn them because by arousing our
aversion their behavior causes us pain.
William Durham and I teach together and often debate in class the mer-
its of Westermarck’s solution to the externalization problem. I recommend
it; he rejects it. Ironically, as he sees it, the three studies that have done the
most to confirm Westermarck’s account of why we avoid incest all discon-
firm his account of why we condemn incest. In Taiwan, Israel, and Leba-
non, children who are not siblings were commonly reared together as inti-
mately as if they were siblings. In all three cases the result was, as
Westermarck predicted, an aversion to sexual relations as adults. But in no
one of these cases did the aversion aroused by early association produce a
tendency to condemn sexual relations between the co-reared children. The
children reared together in Israeli kibbutzim were encouraged to marry,
and the children reared together in Taiwan and Lebanon were condemned
if they refused to marry.
Durham offers in place of Westermarck’s solution to the externalization
problem a solution of his own. This is a more sophisticated version of the
view that Aberle et al. rejected when, accepting the dangers of inbreeding,
they denied that primitive man could have recognized these dangers. Dur-
ham argues that in so doing they overlooked the evidence preserved in the
origin myths of many societies. This evidence says that the deleterious con-
sequences of inbreeding were widely recognized in prehistoric times. The
incest taboo is not, as Westermarck would have it, an unintended conse-
quence of our emotional constitution; it is a consciously implemented solu-
tion to a recognized problem. Durham does not deny Westermarck’s claims
with regard to the consequences of early association. He even agrees that
this is the primary reason humans avoid incest. What he denies is that the
“social fact” we call the incest taboo is largely a product of the aversion
aroused by early association. Thus, for Durham, what Williams calls the
representation problem is not a problem at all. The incest taboo is not a
representation. It is a creation.
Durham’s position is best seen in contrast to the positions taken by Bate-
son (Chapter 1) and Gates (Chapter 8). Although Durham is reluctant to
attribute the inhibiting effects of early association to the dangers of in-
breeding, he is happy to attribute the incest taboo to these dangers. Bate-
son and Gates, in contrast, seem willing to assume that the dangers of in-
breeding account for the inhibiting effects of early association, but they are
not willing to assume that these dangers account for the incest taboo. “In
summary,” Bateson writes, “I suggest that it is unlikely that inbreeding
16 wolf

avoidance and incest taboos evolved by similar mechanisms or even have a

common utility in modern life.”
In Chapter 6, Neven Sesardic challenges Durham’s rejection of Wester-
marck’s solution to the externalization problem by suggesting that Durham
is taking unfair advantage of an “epistemological tension” in Westermarck’s
account of the incest taboo. Westermarck argues that early association
arouses a sexual aversion that gives rise to the incest taboo. His critics argue
that the aversion he identifies is a result of the incest taboo, not its source.
Thus, to make his case, Westermarck must show that even in the absence of
an incest taboo, early association arouses an aversion. Consequently, every
case Westermarck can muster in support of his aversion hypothesis will in-
evitably challenge his solution to the externalization problem. “It is,” as
Sesardic puts it, “a zero-sum game; what the theory gains by collecting evi-
dence in favor of the [first hypothesis] it automatically loses on the other
front because the same empirical data chip away at the [second hypothe-
sis].” Durham responds (in Chapter 7) by arguing that what is significant
about these cases is not simply that they involve socially acceptable unions
between childhood associates. It is that in all three cases they are the mar-
riages most parents prefer.
Chapter 6 also takes up Bernard Williams’s treatment of what I have
called the expression problem (B2). Williams argued that Westermarck was
attempting the impossible in trying to derive the incest taboo from the
aversion aroused by early association. This is impossible because the con-
tent of the aversion and the content of the taboo are not the same. The
aversion is about sexual relations with the people with whom one is reared,
while the taboo is about marriages between people who are classified as
close relatives.
Sesardic argues, in defense of Westermarck, that Williams’s argument
only “looks persuasive because it trades on a crucial ambiguity.” The fact
that childhood association arouses an aversion to sexual relations does not
necessarily mean that the people affected experience the aversion as being
directed to their childhood associates qua associates. We must distinguish
the cause of the aversion and the subjective experience of the aversion.
When we do so, Williams’s objection evaporates. It is not only possible, but
also highly likely because it would be culturally encouraged, that siblings
who are reared together experience the aversion aroused by their early as-
sociation in terms of kinship rather than in terms of association. Thus, it is
also highly likely that there is rarely a mismatch between the experience of
childhood association and the content of the incest taboo.
There is, in my view, a relationship between Sesardic’s critique of Dur-
ham and his critique of Williams. He could have employed against Durham
the same argument he employs against Williams. Both confuse the cause of
 Introduction 17

sexual aversion and the experience of that aversion. In all three of the soci-
eties Durham cites to refute Westermarck, the great majority of the people
with whom one associates as a child are parents and siblings. Thus it is
likely that the aversions aroused by childhood association are typically ex-
perienced in kinship terms. There is, then, no reason to expect that mar-
riages involving childhood associates who are not siblings will elicit disap-
proval. They lack what it takes to turn a comfortable indifference into a
painful aversion.
Most authors recognize that the incest taboo has what George Peter
Murdock called “a peculiar emotional intensity,”39 but many do not recog-
nize that it also has a peculiar moral intensity. The result is that what I call
the moralization problem has been neglected. In fact, the only thorough
treatment of the problem is in Westermarck’s Origin and Development of
Moral Ideas. His solution to the problem is deceptively simple. It is prem-
ised on the view that “the moral concepts, which form the predicates of
moral judgements, are ultimately based on moral emotions, that they are es-
sentially generalizations of tendencies in certain phenomena to call forth in-
dignation or approval.” What distinguishes the moral emotions from other
emotions is “their disinterestedness, apparent impartiality, and flavour of
generality.”40 Thus, what makes disapproval of incest moral is the fact that
the disapproval is general and does not appear to serve any selfish interest.
In sum, it is moral because it is generally approved disapproval.
In Chapter 10, Larry Arnhart points out that even those social and bio-
logical scientists who defend Westermarck’s explanation of incest avoid-
ance reject his solution to the moralization problem. With the notable ex-
ception of E. O. Wilson, they cannot accept the possibility that moral
concepts “are ultimately based on moral emotions.” Even the evolutionary
psychologists who take Westermarck’s aversion hypothesis as paradigmatic
reject his evolutionary approach to ethics as violating a fundamental fact /
value dichotomy. For David Buss and Steven Pinker, as for Callicles and
Kant, is is is and ought is another thing.
Arnhart contrasts “the ‘transcendentalist’ claim that ethics is rooted in
absolute standards that exist outside of the human mind” with “the ‘em-
piricist’ claim that ethics is rooted in natural human inclinations.” The con-
trast is neatly illustrated by the difference between Francis Hutcheson and
Bernard Mandeville on the subject of the incest taboo. Hutcheson, in the
empiricist tradition, argued that the incest taboo shows that we are all pos-
sessed of an innate moral sense. “Had we no moral Sense natural to us, we
should only look upon Incest as hurtful to ourselves, and shun it, and never
hate other incestuous Persons, more than we do a broken Merchant.”41
Mandeville, in the transcendentalist tradition, emphatically denied the ex-
istence of an innate moral sense. He agreed that “incestuous alliances are
18 wolf

abominable; but it is certain that, whatever Horror we conceive at the

thought of them, there is nothing in Nature repugnant against them, but
what is built upon Mode and Custom.”42
Arnhart’s contrasts help us to understand why Westermarck’s solution to
the moralization problem was neglected. The reason is that just as Wester-
marck was a thoroughgoing Darwinian, so also was he—ipso facto, I would
say—a thoroughgoing empiricist. Recognizing the moral content of the in-
cest taboo, he refused to separate it from the taboo’s psychological roots
and its biological origins. Instead, he attributed the moral content of the
taboo to the same sources as the aversion responsible for incest avoidance.
This violated the fact /value dichotomy by turning an ought into an is. The
argument would have appealed to Francis Hutcheson. It did not appeal to
the transcendentalists who dominated twentieth-century social thought.
It is necessary to partition the incest problem and name the parts lest we
lose sight of its complexity and be tempted by partial solutions. We must
always remember, however, that solutions to one part of the problem may
have implications for our understanding of other parts. This is illustrated
over and over again in this volume. The implications of Bittles’s concern
that the immediate effects of inbreeding have been exaggerated are not lim-
ited to the problem of how incest avoidance originated. They are also rele-
vant to the problem of why incest avoidance was mandated. The less visi-
ble the effects of inbreeding among a couple’s children, the more likely it is
that Westermarck was right in seeking the origins of the incest taboo
among the passions. The more visible the effects of inbreeding among a
couple’s children, the more likely it is that Durham is right in arguing that
rationality played a critical role.
One of the questions I take up in my own chapter is whether men are as
sensitive to the inhibiting effects of early association as women are. Have-
lock Ellis suggested that they are not. I offer evidence indicating that he
was wrong. Which of us is right will affect not only the solution to the me-
diation problem but, even more critically, the solution to the representation
problem. It is one thing to derive a prohibition that applies to both sexes
from an aversion felt by both, quite another to derive a prohibition that ap-
plies to both sexes from one felt by only one sex, particularly if the inhib-
ited sex is also the socially subordinate sex.
Hill Gates concludes Chapter 8 by noting that “under conditions not
fully mapped out, but surely recurrently, our innate alertness to the emo-
tional complexity of incest was seized upon and turned to precise cultural
ends, until something better came along.” She is pointing to the possibility
of relationships between the mediation problem and the localization prob-
lem. Mark Erickson and I argue that it is not coincidental that we sexually
avoid those persons to whom we are most strongly attached as children.
 Introduction 19

Avoidance and attachment have evolved together, and together from the
emotional core of the parent-child relationship. Thus, from our point of
view, it is easy to see why secular and religious leaders who set themselves
up as “father and mother of the people” extend the scope of the incest ta-
boo and enact severe sanctions against incest. Sanctioning incest fits emo-
tionally—and thereby helps justify politically—a paternalistic stance.
Their published report suggests that, meeting in 1956, Aberle et al. did
not concern themselves with the practical or policy aspects of the incest
problem. In my view, this was not because there was not yet an “incest epi-
demic,” or that, if there was, it had not yet been diagnosed. It was because
in 1956 incest was considered, at worst, a legal or ethical problem. In the
eyes of scientists (social and biological alike), it did not have a medical or
psychopathological aspect. Incest was a social matter without serious im-
plications for either physical or mental health. It could be safely left in the
hands of anthropologists with no interest in either biology or psychology.
The chapters in this volume show how Westermarck’s return has changed
all this. Bittles’s interest in calculating as precisely as possible the level of de-
fects among children of consanguineous unions is not entirely academic. It is
motivated by practical concerns and has clear policy implications. Although
the frequency of cousin and uncle-niece marriages has declined worldwide,
it is still high in parts of Asia and Africa and in many immigrant communi-
ties in Europe. And there would be reason for concern even if the frequency
of consanguineous marriages was lower than it actually is. An important
part of Bittles’s argument is that as infant mortality and deaths owing to in-
fectious diseases decline, genetic disorders will constitute an ever larger pro-
portion of the medical problems people experience. An inevitable result will
be that the problems produced by consanguineous unions will be ever more
obvious and ever more likely to stigmatize those people whose customs en-
courage such unions. He is particularly concerned about the effects of these
changes in Europe, where these people are immigrant minorities.
Although their importance was long obscured by the mistaken views of
Twain’s “modern science,” questions about the medical consequences of in-
cest have a long history. They were, as Bittles observes, the subject of in-
tense debate in the 1850s and 1860s, leading to Sir John Lubbock’s attempt
to have questions about cousin marriages included in the 1871 census.
What is new from a practical perspective is concern about the psycho-
pathological consequences of incest. Mark Erickson begins Chapter 9 by
noting that “a distinct recollection of my psychiatric training in the 1980s
is that of a group of patients, mostly female, who presented such a bewil-
dering array of symptoms as to defy diagnosis.” In 1980 it had not yet been
discovered that they were victims of incest.
Although we are a long way from understanding why incest predisposes
20 wolf

people to psychopathologies, it now appears certain that it does. Erickson

mentions, among the many problems of incest victims, major depression, al-
cohol and drug dependence, self-mutilation and suicide, and a wide range of
stress-related illnesses. The evidence of marriages worldwide rule out the
possibility that the cause of such a panoply of disorders is the experience of
sex with an older man. What is critical is that the older man is a father or
older brother. But why does this matter? What makes sex with a father or
an older brother so disturbing? The answer may be implicit in the solution
Erickson and I offer to the mediation problem. It is the symbiotic relation-
ship of incest avoidance and attachment behavior. The evolutionary purpose
of attachment is to elicit caretaking. This is its innate promise. Consequently,
sexual advances by a father or brother are disturbing because they deny the
promise of caretaking. They threaten abandonment. Victims of paternal in-
cest commonly complain that their father “betrayed them” or “violated their
trust.” What they are saying, we suggest, is that a person who had promised
care behaved in a way that denied having ever made such a promise.
Whether or not the specifics of this hypothesis prove correct, it is a step
in the right direction because it assumes that incest is unnatural. Incest is,
as Erickson puts it, “a kinship pathology.” So long as social scientists ac-
cepted the mid-twentieth-century view that human beings were naturally
incestuous, the problems Erickson lists could not be reasonably attributed
to incest. Incest was, in the twentieth-century view, a purely social delict. It
could arouse, at worst, feelings of guilt or shame, and then only if the delict
were discovered. It did not have the emotional stuff needed to fuel a major
pathology.
Erickson concludes Chapter 9 by putting his subject in historical per-
spective: “During most of the twentieth century, social scientists believed
incest was common in nature. Among humans, incest was thought to be
rare because of cultural taboos.” Ironically, Erickson suggests, “this view
has been turned on its head. Incest is now known to be rare in nature, and
we must seriously ask if human incest has become more, not less common,
because of cultural influences.” Thus, in Erickson’s view, the twentieth cen-
tury has seen a complete revolution in the relationship between culture, in-
cest, and human welfare. Culture has gone from what saves us from our in-
cestuous inclinations to what exposes us to the dangers of incest. What
early in the century was seen as a need to repress our innate inclinations is
now seen as a need to recover them.
What will we see in the twenty-first century? In Chapter 10 Larry Arn-
hart suggests that “we will see great advances in the biological understand-
ing of human nature. He continues:
This will force us to think about whether biological science can explain that most
distinctive trait of our humanity—our moral sense of right and wrong. Some peo-
 Introduction 21

ple will argue that our moral experience transcends our biological nature. Others
will argue that we should be able to explain our morality as an expression of our
biological nature. How we decide that debate might be decisively influenced by
whether we accept or reject Edward Westermarck’s Darwinian theory of the incest
taboo as a natural expression of human moral emotions.”

I agree with Arnhart’s assumption that the incest problem will be as hotly
debated in the twenty-first century as it was in the twentieth century, and I
agree with his prediction that the focal point of the debate will shift from
the question of why we avoid incest to the question of why we condemn in-
cest. I would like to add only the prediction that with this shift, the debate
will become even more intense. I say this because in my view the underlying
question that sustains interest in the incest problem is the degree to which
we have managed to transcend our animal origins. Almost completely, ar-
gued the mid-twentieth-century transcendentalists, claiming the incest taboo
as evidence that we are capable of overcoming, even remaking our animal
nature. Then came the Westermarck revival showing that we avoid incest
for natural, not cultural reasons. Now all that remains of the transcenden-
talists’ case is the fact that we do not just “look upon Incest as hurtful to
ourselves, and shun it.” We “hate other incestuous Persons.” This is the
is/ought barrier, the transcendentalists’ last line of defense. They will contest
any attempt to breach it and do so as ardent champions of human dignity.

notes
1. David F. Aberle, Urie Bronfenbrenner, Eckhard H. Hess, Daniel R. Miller,
David M. Schneider, and James N. Sphuhler, “The incest taboo and the mating
patterns of animals,” American Anthropologist, vol. 65 (1963), p. 253.
2. Mark Twain, A Tramp Abroad (1880), (Oxford: Oxford University Press,
1996), p. 173.
3. Edward Westermarck, “Recent theories of exogamy,” in Three Essays on
Sex and Marriage (London: Macmillan, 1934), p. 147.
4. See Arthur P. Wolf, Sexual Attraction and Childhood Association: A Chinese
Brief for Edward Westermarck (Stanford, Calif.: Stanford University Press, 1995),
pp. 2–8.
5. Lord Raglan, Jocasta’s Crime (London, 1933), p. 16.
6. Bronislaw Malinowski, Sex and Repression in Savage Society (London,
1927), p. 243.
7. Robert Briffault, The Mothers (London, 1927), vol. 1, p. 215.
8. Leslie White, “The definition and prohibition of incest,” American Anthro-
pologist, vol. 50, part I (1948), p. 417.
9. Ibid.
10. E. M. East, Inbreeding and Human Affairs (New York, 1938), p. 156.
22 wolf

11. Claude Lévi-Strauss, The Elementary Structures of Kinship (Boston:

Beacon Press, 1969), trans. James Harle Bell, John Richard von Strummer, and
Rodney Needham, pp. 14–15.
12. G. Dahlberg, “On rare defects in human populations with particular
regard to inbreeding and isolate effects,” Proceedings of the Royal Society of
Edinburgh, vol. 58 (1937–38), p. 224.
13. Aberle et al., “Incest taboo,” p. 256.
14. Ibid., pp. 256–57.
15. Ibid., p. 257.
16. Ibid., p. 259.
17. Westermarck, “Recent theories of exogamy,” pp. 158–59.
18. Alfred Wallace, K. Rob, and V. Wikman, “Letters from Edward B. Tylor
and Alfred Russell Wallace to Edward Westermarck,” Acta Academia Aboensis
Humanioria, vol. 13, no. 8 (1940), p. 18.
19. Ibid., p. 258.
20. J. G. Frazer, Totemism and Exogamy (London, 1910), vol. 4, pp. 97–98.
21. See Maurice Godelier, “Incest taboo and the evolution of society,” in
Evolution and Its Influence, ed. Alan Grafen (Oxford: Clarendon Press, 1989),
p. 69.
22. Sigmund Freud, A General Introduction to Psychoanalysis (1920), trans.
Joan Riviere (New York: Pocket Books, 1953), pp. 220–21.
23. Ibid., p. 261.
24. Edward B. Tylor, “On a method of investigating the development of
institutions; applied to laws of marriage and descent,” Journal of the Royal
Anthropological Institute of Great Britain and Ireland, vol. 18 (1889), p. 267.
25. E. E. Evans-Pritchard, Kinship and Marriage Among the Nuer (London:
Oxford University Press, 1951), pp. 43–44.
26. White, “Definition and prohibition,” p. 425.
27. Claude Lévi-Strauss, “The family,” in Man, Culture, and Society, ed.
Harry L. Shapiro (New York: Oxford University Press, 1960), p. 278.
28. Aberle et al., “Incest taboo,” p. 258.
29. Ibid.
30. Sigmund Freud, Totem and Taboo, trans. James Strachey (New York:
W. W. Norton, 1950), p. 144.
31. Aberle et al., “Incest taboo,” p. 262.
32. Robin Fox, “Sibling incest,” British Journal of Sociology, vol. 13 (1962),
p. 147.
33. Edward Westermarck, A Short History of Human Marriage (London,
1926), p. 86.
34. Bernard Williams, “Evolution and ethics,” in Evolution from Molecules to
Men, ed. D. S. Bendall (Cambridge: Cambridge University Press, 1983), p. 560.
35. Rodney Needham, “Remarks on the analysis of kinship and marriage,”
in Rethinking Kinship and Marriage (London: Tavistock, 1971), pp. 25–26.
36. David M. Schneider, “The meaning of incest,” Journal of the Polynesian
Society, vol. 85 (1976), p. 168.
 Introduction 23

37. Jude Cassidy and Phillip R. Shaver, eds., Handbook of Attachment Behav-
ior (New York: Guildford Press, 1999).
38. Westermarck, History of Human Marriage, vol. 2, p. 197 or 214.
39. George Peter Murdock, Social Structure (New York: Macmillan, 1949),
p. 288.
40. Edward Westermarck, The Origin and Development of Moral Ideas
(London: Macmillan, 1906–8), vol. 2, pp. 738–39.
41. Francis Hutcheson, An Inquiry into the Original of Our Ideas of Beauty
and Virtue (London: J. Darby for Will and John Smith, 1725), p. 192.
42. Bernard Mandeville, The Fable of the Bees (Oxford: Clarendon Press,
1924), vol. 1, p. 331.
1 Inbreeding Avoidance and Incest Taboos

Patrick Bateson

I have never much liked the way some of my colleagues in the bio-
logical sciences have applied terms such as rape or marriage to animals. I
appreciate that this is sometimes done to lighten the normally dull language
of scientific discourse. However, these terms have established usage in hu-
man institutions with all their associated rights and responsibilities of indi-
viduals and culturally transmitted rules on what people can and cannot do.
Problems of communication between disciplines are compounded when,
having found some descriptive similarities between animals and humans
and having investigated the animal cases, biologists or their popularizers
use the animal findings to “explain” human behavior. Such arguments rely
on a succession of puns, which are usually unconscious, but which are es-
pecially unfunny to those social scientists who feel threatened by an appar-
ent takeover bid of the biologists.
I believe that incest should be restricted to human social behavior where
culturally transmitted proscriptions limit sexual contact and marriage with
close kin (and others who might be deemed to be close kin). Inbreeding
avoidance should be used for behavior that makes matings with close kin
less probable in both humans and nonhuman animals. This separation then
leaves open the question of whether these behaviors have evolved for simi-
lar reasons and whether the two phenomena have similar current functions.
This chapter briefly reviews the evidence that people unconsciously
choose mates who are a bit different from those individuals who are famil-
iar from early life but not too different. In a biological context this is often
referred to as optimal outbreeding.1 Why did it evolve? The question in-

Generous financial support from the Australian Research Council and the benefits
of ongoing collaboration with many colleagues in Australia, India, Pakistan, the
United Kingdom, and the United States is gratefully acknowledged.
 Inbreeding Avoidance and Incest Taboos 25

vites examination of the concept of adaptation and the role of Darwinian

evolution in generating such adaptations. Since evolution is thought to in-
volve changes in genes, it is necessary to be clear about the role of genes in
an individual’s development. When development is considered, a quite dif-
ferent set of issues is raised. These need to be considered in relation to the
formation of mating preferences. Finally, it is necessary to come to the
heart of the matter: what relations, if any, can be found between the avoid-
ance of inbreeding and incest taboos?

Optimal Outbreeding

The biological costs of inbreeding are evident enough in other ani-

mals.2 They are particularly obvious in birds. If a male bird is mated with
his sister, and their offspring are mated together, and so on for several gen-
erations, the line of descendants usually dies out fairly quickly. This hap-
pens because some damaging genes are more likely to be expressed in in-
bred animals. Some potentially harmful genes are recessive and therefore
harmless when they are paired with a dissimilar gene, but they become
damaging in their effects when combined with an identical gene. They are
more likely to be paired with an identical recessive gene as a result of in-
breeding. The presence of such genes is a consequence of the mobility of
the birds and the low probability that they will mate with a bird of the op-
posite sex that is genetically similar to them. Over time, the recessive genes
have accumulated in the genome because they are normally suppressed by
their dominant partner gene.
The genetic costs of inbreeding arising from the expression of damaging
recessive genes are the ones that people usually worry about. However, re-
cessive genes are less important in mammals than they are in birds because
mammals generally move around less and may live in quite highly inbred
groups. The most important biological cost of excessive inbreeding is that
it negates the benefits of the genetic variation generated by sexual repro-
duction. If an animal inbreeds too much, it might as well make many cop-
ies of itself without the effort and trouble of courtship and mating.
On the other side, excessive outbreeding also has costs.3 For a start, ex-
cessive outbreeding disrupts the relation between parts of the body that
need to be well adapted to each other. The point is illustrated by human
teeth and jaws. The size and shape of teeth are strongly inherited charac-
teristics. So too are jaw size and shape, as may be seen in the famous paint-
ings of the Hapsburg family, scattered around the museums of the world.
The Dürer painting of the Holy Roman Emperor Maximilian I reveals the
large Hapsburg jaw, which remained as pronounced in his great-great-
26 bateson

great-grandson, Philip IV of Spain, shown in the painting by Velasquez.

The potential problem arising from too much outbreeding is that the in-
heritance of teeth and jaw sizes are not correlated. A woman with small
jaws and small teeth who had a child by a man with big jaws and big teeth
lays down trouble for her grandchildren, some of whom may inherit small
jaws and big teeth. In a world without dentists, ill-fitting teeth were proba-
bly a serious cause of mortality. This example of mismatching, which is one
of many that may arise in the complex integration of the body, simply illus-
trates the more general cost of outbreeding too much.
Some of the evolutionary pressures on mate choice arose from too much
inbreeding, on the one hand, and from too much outbreeding on the other.
A preference for an individual somewhat like close kin will minimize the
opposing ill effects of breeding with individuals who are genetically too dif-
ferent. A sexual preference for individuals who are a bit different from
close kin strikes a balance between the biological costs of inbreeding and
those of outbreeding.4
The suggestion is that individuals had greatest reproductive success if
they mated with a partner who was somewhat similar to themselves, but
not too similar. The hypothesis has gathered considerable empirical sup-
port from studies of animals.5 Japanese quail, for example, prefer mates
that are first or second cousins, when given a choice in laboratory experi-
ments.6 If they have been reared with unrelated individuals, the quail pre-
fer mates that are a bit different from these familiar individuals. In hu-
mans a great mass of data shows that freely chosen human spouses are
more like each other than would be expected on a chance basis. Similari-
ties are not only social and psychological but also found in measures of
body dimensions such as length of earlobe.7
Humans choose partners somewhat like themselves.8 At the same time,
people prefer sexual partners who look slightly different from individuals
with whom they have grown up. Faces are perceived as more attractive if
some of the facial features are exaggerated by caricaturing the image so
that it differs from the average.9 Most people are attracted to faces that are
distinctive and depart from the average. When the faces of individuals who
were perceived as being attractive were averaged, this composite was pre-
ferred over the average of all faces.
Natural experiments have been performed unwittingly on human be-
ings. Famously, Israeli kibbutzniks grow up together like siblings and rarely
marry each other.10 The most comprehensive evidence has come from Ar-
thur Wolf’s (1994) study of the marriage statistics from Taiwan in the late
nineteenth and early twentieth century, when Taiwan was under Japanese
control.11 The Japanese kept detailed records for the births, marriages, and
 Inbreeding Avoidance and Incest Taboos 27

deaths of everyone on the island. As in many other parts of China, mar-

riages were arranged and occurred mainly and most interestingly in two
forms. The major type of marriage was the conventional one in which the
partners first met each other when adolescent. In the minor type of mar-
riage, the wife-to-be was adopted as a young girl into the family of her fu-
ture husband. In minor-type marriages, therefore, the partners grew up to-
gether like siblings. In this sense they were like the quail in the laboratory
experiment, having been reared with an individual of the opposite sex to
whom they were not genetically related. Later in life their sexual interest in
their partner was assessed in terms of divorce, marital fidelity, and the
number of children produced. By all these measures, the minor marriages
were conspicuously less successful than the major marriages. Typically, the
young couples who had grown up together from an early age, like brother
and sister, were not much interested in each other sexually when the time
came for their marriage to be consummated.
While a great deal is still unknown about sexual preferences in both an-
imals and humans, the similarities are quite striking. The processes in-
volved in preference of humans for slight novelty have been subject to Dar-
winian evolution. However, acceptance of this point has to be tempered by
an awareness that mate choice is influenced by many qualities that are be-
yond the scope of this chapter.12

Adaptation and Darwinism

Unconscious preferences for slight novelty are seen by biologists as

being adaptive in the sense that they serve to enhance the reproductive suc-
cess of the individual who acts on those preferences. Darwinism has gener-
ated much distrust in the social sciences because it seemed to spawn such
strange and, indeed, wicked social theories. The reason why biologists like
me are still greatly enamored of the Grand Old Man is because he provided
what is probably the only coherent and systematic explanation for adapta-
tion—the match between the characteristics of organisms and the worlds in
which they live.
Complicated things found in biology have the appearance of having
been designed for a purpose. In the early nineteenth century a famous the-
ologian, William Paley, put it this way: “It is the suitableness of these parts
to one another; first, in the succession and order in which they act; and,
secondly, with a view to the effect finally produced.”13 Paley took this as
proof of the existence of God. Darwin provided us with an explanation of
how it came about. When individual differences are inherited, those indi-
28 bateson

viduals that are better adapted than others are more likely to survive and
reproduce and then have offspring that share their adaptations.
The perception that behavior is designed springs from the relations be-
tween the behavior, the circumstances in which it is expressed, and the re-
sulting consequences. The closeness of the perceived match between the tool
and the job for which it is required is relative. In human design, the best that
one person can do will be exceeded by somebody with superior technology.
If you were on a picnic with a bottle of wine but no corkscrew, one of your
companions might use a strong stick to push the cork into the bottle. If you
had never seen this done before, you might be impressed by the selection of
a rigid tool small enough to get inside the neck of the bottle. The tool would
be an adaptation of a kind. Tools that are better adapted to the job of re-
moving corks from wine bottles are available, of course, and an astonishing
array of devices have been invented. One ingenious solution involved a
pump and a hollow needle with a hole near the pointed end; the needle was
pushed through the cork and air was pumped into the bottle, forcing the
cork out. Sometimes, however, the bottle exploded and this tool quickly be-
came extinct. As with human tools, what is perceived as good biological de-
sign may be superseded by an even better design, or the same solution may
be achieved in different ways.
Among those who spin stories about biological design, a favorite figure
of fun is an American artist, Gerald Thayer.14 He argued that the purpose
of the plumage of all birds is to make detection by enemies difficult. Some
of the undoubtedly beautiful illustrations were convincing examples of the
principles of camouflage. However, among other celebrated examples, such
as pink flamingos concealed in front of the pink evening sky, was a paint-
ing of a peacock with its resplendent tail stretched flat and matching the
surrounding leaves and grass. The function of the tail was to make the bird
difficult to see! Ludicrous attributions of function to biological structures
and behavior have been likened to Rudyard Kipling’s Just-So stories of
how, for example, the leopard got his spots.15 However, the teasing is not
wholly justified. Stories about current function are not about how the leop-
ard got his spots, but what the spots do for the leopard now. That question
is testable by observation and experiment.
Not every speculation about the current use of a behavior pattern is
equally acceptable. Both logic and factual knowledge can be used to decide
between competing claims. Superficially attractive ideas are quickly dis-
carded when the animal is studied in its natural environment. The peacock
raises his enormous tail in the presence of females, and he molts the cum-
bersome feathers as soon as the spring breeding season ends. If Thayer had
been correct about the tail feathers being used as camouflage, the peacock
should never raise them conspicuously and he should keep them year round.
 Inbreeding Avoidance and Incest Taboos 29

Genes and Development

A Darwinian account tells us nothing about how those sexual pref-

erences develop in the individual. It is true that for the Darwinian evolu-
tionary mechanism to work, something must be inherited. But even if a sin-
gle gene provides the basis for the distinctive beneficial character of the
individual, a single gene is not sufficient for the development of the charac-
ter. This is where we get to the heart of a very lively debate in biology.
Scientists collaborating on the Human Genome Project have elucidated
nearly all the DNA sequences of all the genes on all twenty-three pairs of
chromosomes found in every human cell. It is a staggering achievement.
But the excitement about what is being done should be greatly moderated.
“The Book of Life,” as one leading scientist called it, will not provide the
complete story about human nature.
The human genome is like a cook’s larder list.16 Working out all the dishes
that cooks might make from the ingredients available to them is another
matter. If you want to understand what happens in the lifelong process
from conception to death, you must study the process. The starting points
of development include the genes. But they also include factors external to
the genome, and of course, the social and physical conditions in which the
individual grows up are crucial.
The language of a gene “for” a particular characteristic is exceedingly
muddling to the nonscientist—and, if the truth be told, to many scientists
as well. What the scientists mean (or should mean) is that a genetic differ-
ence between two groups is associated with a difference in a characteristic.
They know perfectly well that other things are important and that, even in
constant environmental conditions, the outcome depends on a combination
of many genes. Particular combinations of genes have particular effects,
and a gene that fits into one combination may not fit into another. Unfor-
tunately, the language of a gene “for” a characteristic has a way of seduc-
ing the scientists themselves into believing their own sound bites. The lan-
guage rests on a profound misunderstanding.
While genes obviously matter, even a cursory glance at humanity reveals
the enormous importance of each person’s experience, upbringing, and cul-
ture. Nobody could seriously doubt the remarkable human capacity for
learning from personal experience and from others. It is obvious that expe-
rience, education, and culture make a big difference, whatever an individ-
ual’s genetic inheritance. Individuals are not like the dry Japanese paper
flowers that are simply put into water to open out.
The notion that genes are simply blueprints for an individual human is
hopelessly misleading. In a blueprint, the mapping works both ways: start-
ing from a finished house, the room can be found on the blueprint, just as
30 bateson

the room’s position is determined by the blueprint during the building

process. This straightforward mapping is not true for genes and individual
human behavior patterns, in either direction.
Genes do not make behavior patterns or physical attributes. Genes make
proteins. Each human has about 30,000 genes, each of which is an inher-
ited molecular strand (or set of strands) that may be translated into a pro-
tein molecule (or part of one). The proteins are crucial collectively to the
functioning of each cell in the body. Some proteins are enzymes, controlling
biochemical reactions, while others form the physical structures of the cell.
These protein products of genes work not in isolation but in a cellular en-
vironment created by the conditions of the local environment and by the
expression of other genes. Each gene product interacts with many other
gene products.
Any characteristic of an individual, such as a behavior pattern or psy-
chological attribute, is affected by many different genes, each of which con-
tributes to the variation between individuals. In an analogous way, many
different design features of a motor car contribute to a particular charac-
teristic such as its maximum speed. A particular component such as the
system for delivering fuel to the cylinders may affect many different aspects
of the car’s performance, such as its top speed, acceleration, and fuel con-
sumption. A broken wire can cause a car to break down, but this does not
mean that the wire by itself is responsible for making the car move.
The image of a genetic blueprint also fails because it is too static, too sug-
gestive that adult organisms are merely expanded versions of the fertilized
egg. In reality, developing organisms are dynamic systems that play an active
role in their own development. To some extent each individual chooses and
shapes its own physical and social environment. This can have interesting
consequences. People who differ in ways that relate to differences in their
genes may also pick certain physical and social environments in which to
live. This process has been given the name “niche-picking.”17 It means that
individuals with different characteristics, some of which reflect differences
in their genes, end up by their own actions experiencing the world in dif-
ferent ways.
Environmental and inherited factors often work together to produce
much larger overall effects than when either factor is present on its own.
The often uncanny similarities between identical twins provide striking ev-
idence for the importance of genes in shaping physical and behavioral char-
acteristics. But one surprising finding to emerge from studies of identical
twins is that twins reared apart are sometimes more like each other than
those reared together.18 To put it another way, rearing two genetically iden-
tical individuals in the same environment can make them less similar. This
fact pleases neither the extreme environmental determinist nor the extreme
genetic determinist.
 Inbreeding Avoidance and Incest Taboos 31

The environmental determinist supposes that twins reared apart must

have different experiences and should therefore be more dissimilar in their
behavior than twins who grew up together in the same environment. The
genetic determinist does not expect to find any behavioral differences be-
tween genetically identical twins reared together; if they have had the same
genes and the same environment, then how can they be different? Of course,
one twin provides a social environment for the other and siblings often
hate to do what the other one is doing.
A single developmental ingredient, such as a gene or a particular form of
experience, might produce an effect on behavior, but this knowledge gives
only a feeble insight into developmental processes. The best that can be
said of the nature/nurture split is that it provides a framework for uncov-
ering a few of the genetic and environmental ingredients that generate dif-
ferences between people. At worst, it satisfies a demand for simplicity in
ways that are fundamentally misleading.
The processes involved in behavioral and psychological development
have certain metaphorical similarities to cooking, to which I have already
alluded. Both the raw ingredients and the manner in which they are com-
bined are important. Timing also matters. In the cooking analogy, the raw
ingredients represent the many genetic and environmental influences, while
cooking represents the biological and psychological processes of develop-
ment. Nobody expects to find all the separate ingredients represented as dis-
crete, identifiable components in a soufflé. Similarly, nobody should expect
to find a simple correspondence between a particular gene (or a particular
experience) and particular aspects of an individual’s behavior or personality.

The Development of Mating Preferences

Returning to mating preferences, how do they develop? A preference

for faces that are a bit different, but not too different, from a familiar stan-
dard is relevant to mate choice in other species. Animals of many species
tend to avoid mating with individuals who are very close kin, such as sib-
lings, but they do sometimes prefer to mate with more distant relatives. The
developmental process involved was first made famous by Konrad Lorenz
and called “sexual imprinting.”19 This is coupled with some habituation to
the very familiar, which offsets the preference to the slightly novel.20
The few Israeli kibbutzniks who chose to marry within their peer group
were usually those who had entered the kibbutz after the age of six and
therefore had not grown up with their future spouses.21 In Taiwan, girls
who were adopted into families before the age of three and then married
their adopted “brother” had a lower fertility than girls adopted later.22 Nei-
ther of these findings means that the learning process that affects adult sex-
32 bateson

ual preferences is completed early in life. If children grow up together and,

as a result, see a lot of each other, they revise the ways in which they rec-
ognize each other; this goes on until they are sexually mature. By the time
they are three, children are highly conscious of their own sex and are much
less likely to play with somebody of the opposite sex, particularly a child
who is not well known to them.23 It seems plausible then that a girl who is
adopted when over three will be viewed as a stranger by the boy and treated
differently from a girl who is adopted when younger.
Two developmental explanations have been proposed for the age-
dependent effect. The first suggests that all the information that will be
used is gathered within a period when the brain is most easily affected by
such experience.24 This is a classical critical period hypothesis used at one
time for behavioral imprinting in birds but now largely discarded. It sup-
poses that the adult can remember the faces of its siblings and generalize
from their childlike characteristics to their adult form. As far as I know,
this possibility has never been tested. I don’t find it especially plausible, but
in its defense, it should be remembered that algorithms have been devised
for aging pictures of children so that those who have been kidnapped early
in life might be recognized many years later. If computers can do it, then
perhaps humans can too.
I prefer a rather different hypothesis, which is that while the process of
learning starts most readily at a relatively early stage in development, the
representations of familiar faces are updated by continuing close contact.25
A lot hinges on the type of relationship the couple have when they are
young. If they play together and, as a result, see a lot of each other, the in-
difference is likely to be greater. By the time they are three, children are
highly conscious of their gender and are much less likely to play with a
member of the opposite sex, particularly a strange member.26 It seems plau-
sible that a girl adopted when over three will be seen as a stranger by the
boy and treated very differently from girls who are adopted when younger.
The idea that familiarity of a certain kind does reduce sexual attractive-
ness may be applied rewardingly to explain one striking feature of divorce
statistics. For instance, in British women who married before the age of
twenty, the proportion of marriages that ended in divorce has been ap-
proximately double that of the marriages of women who married between
twenty and twenty-four.27 The difference is found at any time between four
and twenty-five years after marriage. Many factors, such as differences be-
tween social classes in attitudes toward marriage, could explain or con-
tribute to explaining the difference. However, early marriages may involve
a great deal of intimacy but relatively little sexual satisfaction. Indeed, peo-
ple often report that their early sex lives were relatively unrewarding. If the
effects of habituation are not powerfully offset by rewarding sexual expe-
 Inbreeding Avoidance and Incest Taboos 33

rience, the partner may lose his or her attractiveness and become the equiv-
alent of a sibling.

Evolution and Function of Mating Preferences

The presumption made by biologists is that inbreeding avoidance

evolved because those individuals that did it were more likely to have greater
reproductive success than those who did not. Since mate choice is affected by
early experience in a great many birds and mammals, with a preference for
types that are familiar from early life, it would not be a big step to superim-
pose on such a mechanism a reduced preference for those individuals that
are very familiar. Two well-known learning processes, behavioral imprinting
and long-term habituation, are able to generate a preference for individuals
who are a bit different but not too different from those individuals who are
familiar from early life.28 If, as the evidence strongly suggests, inbreeding and
excessive outbreeding carry biological costs in the form of reduced repro-
ductive success, then the activation of both processes in the development of
sexual preferences would have been favored. It does not matter for the evo-
lutionary argument whether behavioral imprinting came before, after, or at
the same time as the habituation of sexual preferences. Just how the early ex-
perience is mediated through the rest of childhood to affect adult sexual
preferences remains unresolved. But uncertainty about the developmental ar-
gument does not detract from the evolutionary hypothesis that it was more
beneficial to base sexual inhibitions on early experience rather than later ex-
perience. Those individuals who set the standards determining with whom
they should not mate when they were likely to be surrounded by family were
fittest. They were more likely to obtain biological advantage than those who
started the process later.
What about incest taboos? Differences between cultures in the close rela-
tives (or presumed close relatives) that are prohibited as targets of sexual in-
terest are not genetically inherited. Many authors have suggested that indi-
viduals may derive reproductive success from the taboos.29 However, those
individuals who impose the prohibitions do not derive immediate personal
benefits from their behavior. Others derive those direct fitness benefits. Social
benefits may be derived because the group does not have to pay the costs of
caring for individuals who in various ways are less fit. But note here that an
attempt to mount a purely eugenic argument would be confused because the
maladaptive genes expressed when inbreeding is common are not removed
from the population by preventing inbreeding. Indeed, inbreeding is the best
way of getting rid of those genes in the long run. The social benefits of incest
taboos may therefore be seen as equivalent to those modern laws that re-
quire the wearing of seat belts in cars.
34 bateson

Whether people are aware of the effects of inbreeding is another issue.

William Durham has presented persuasive evidence that in many cultures
people are aware.30 He also documents cases where they are not, and we
are left with the possibility that either they were aware in the past or incest
taboos were driven by other pressures. Awareness of the ill effects of in-
breeding would be best translated into the conviction that the aware indi-
vidual should not have children with his or her sibling. Nothing more is
required of Darwinian evolution. The awareness does not immediately trans-
late into a conviction that others should be stopped from having children
with their siblings. Or if it does, it is driven by a different utility that in-
volves group selection.
Differences of outcome raise problems for an argument developed by
E. O. Wilson. He wrote,
By translating the Westermarck effect into incest taboos, humans appear to pass
from pure instinct to pure rational choice. . . . I suggest that rational choice is the
casting about among alternative mental scenarios to hit upon the ones which, in a
given context, satisfy the strongest epigenetic rules. It is these rules and this hierar-
chy of their relative strengths by which human beings have successfully survived
and reproduced for hundreds of millennia. The incest avoidance case may illustrate
the manner in which the coevolution of genes and culture has woven not just part
but all of the rich fabric of human social behavior.31

Implicit in this argument is that the incest taboo is serving the same func-
tion for the individual as inbreeding avoidance. The same point arises in a
formal genes-culture coevolutionary model developed by Aoki and Feld-
man.32 Further, the model assumes random mating by those who don’t
have the postulated “avoid sibling” behavior pattern. Since such random
mating does not occur, the assumption renders the model questionable.
In summary, then, I suggest that it is unlikely that inbreeding avoidance
and incest taboos evolved by similar mechanisms or even have a common
utility in modern life. I fully accept the argument in favor of having both
belt and braces (see Chapter 6). Redundant mechanisms are well known in
biology, those used in navigation by birds being a famous example. Even
so, incest taboos need not necessarily serve the same function as the inhibi-
tions derived from early experience. If that much is accepted, what other
mechanism for the cultural evolution of incest taboos should be enter-
tained? Like Wolf, I think that a strong case can be made for the hypothesis
advanced by Edward Westermarck.33
Westermarck suggested that humans have an inclination to prevent
other people from behaving in ways they would not themselves behave. On
this view, left-handers were in the past forced to adopt the habits of right-
handers because the right-handers found left-handers disturbing. In the
same way, those who were known to have had sexual intercourse with
 Inbreeding Avoidance and Incest Taboos 35

close kin were discriminated against. People who had grown up with kin of
the opposite sex were generally not attracted to those individuals and dis-
approved when they discovered others who were. It had nothing to do with
society not wanting to look after the half-witted children of inbreeding,
since in many cases they had no idea that inbreeding was the cause. Rather,
the disapproval was about suppressing abnormal behavior, which is poten-
tially disruptive in small societies.
Such conformity looks harsh to modern eyes, even though we have
plenty of examples of it in contemporary life. However, when so much de-
pended on unity of action in the environment in which humans evolved,
wayward behavior could have destructive consequences for everybody. It is
not difficult to see why conformity should have become a powerful trait in
human social behavior. Once in place, the desire for conformity, on the one
hand, and the reluctance to inbreed, on the other, would have combined to
generate social disapproval of inbreeding. The emergence of incest taboos
would take on different forms, depending on which sorts of people, nonkin
as well as kin, were likely to be familiar from early life.
If these ideas are correct, human incest taboos did not arise historically
from deliberate intention to avoid the biological costs of inbreeding. Rather,
in the course of biological evolution, two separate mechanisms appeared.
One was a developmental process concerned with striking an optimal bal-
ance between inbreeding and outbreeding when choosing a mate. The other
was concerned with social conformity. When these two propensities were
put together, the result was social disapproval of those who chose partners
from within their close family. When social disapproval was combined with
language, verbal rules appeared that could be transmitted from generation
to generation, first by word of mouth and later in written form.

Conclusion

I believe that the divide between the biological sciences and the so-
cial sciences has narrowed to the point where real dialogue occurs. At-
tempts at a takeover by sociobiologists and, more recently, evolutionary
psychologists set this process back, all the more so because they persuaded
some social scientists to change their faith and preach with the zeal of the
recently converted.
Nevertheless, I believe that we are getting closer. The biologists have
come to understand that their own thinking is affected by where they have
come from, and the social scientists have started to understand how evi-
dence changes the way they think. Darwinism is no longer the threat it
once seemed. It offers explanations that are quite different in character
from those provided by studies of how an individual develops. The days
36 bateson

of both genetic and environmental determinism are numbered. The center

of intellectual activity is now focused on process and how the individual
develops.

notes
1. P. Bateson, “Optimal outbreeding,” in Mate Choice, ed. P. Bateson (Cam-
bridge: Cambridge University Press, 1983), pp. 257–77.
2. Ibid.
3. Ibid.
4. Ibid.
5. Ibid.
6. P. Bateson, “Preferences for cousins in Japanese quail,” Nature, vol. 295
(1982), pp. 236–37.
7. William Durham, Coevolution: Genes, Culture, and Human Diversity
(Stanford, Calif.: Stanford University Press, 1991).
8. Bateson, “Optimal outbreeding.”
9. D. I. Perrett, K. A. May, and S. Yoshikawa, “Facial shape and judgments of
female attractiveness,” Nature, vol. 368 (1994), pp. 239–42.
10. J. Shepher, “Mate selection among second generation kibbutz adolescents
and adults: Incest avoidance and negative imprinting,” Archives of Sexual Behav-
ior, vol. 1 (1971), pp. 293–307.
11. Arthur P. Wolf, Sexual Attraction and Childhood Association: A Chinese
Brief for Edward Westermarck (Stanford, Calif.: Stanford University Press, 1995).
12. P. Bateson and P. Martin, Design for a Life: How Behavior and Personality
Develop (New York: Simon and Schuster, 2000).
13. W. Paley, Natural Theology (London, Faulder, 1802).
14. Gerald H. Thayer, Concealing-Coloration in the Animal Kingdom (New
York: MacMillan, 1909).
15. J. S. Gould and R. C. Lewontin, “The spandrels of San Marco and the
Panglossian program: A critique of the adaptationist program,” Proceedings of
the Royal Society of London, B, vol. 250 (1979), pp. 281–88.
16. Bateson and Martin, Design for a Life.
17. Ibid.
18. J. Shields, Monozygotic Twins Brought Up Apart and Brought Up
Together (Oxford: Oxford University Press, 1962).
19. Konrad Lorenz, “Der Kumpan in der Umwelt des Vogels,” Journal für
Ornithologie, vol. 83 (1935), pp. 137–213, 289–413.
20. Bateson, “Optimal outbreeding.”
21. Shepher, “Mate selection.”
22. Wolf, Sexual Attraction and Childhood Association.
23. Eleanor E. Maccoby, “Gender and gender relationships,” American
Psychologist, vol. 24 (1990), pp. 523–520.
24. Wolf, Sexual Attraction and Childhood Association.
 Inbreeding Avoidance and Incest Taboos 37

25. P. Bateson, “Rules for changing the rules,” in Evolution from Molecules
to Men, ed. D. S. Bendall (Cambridge: Cambridge University Press, 1983),
pp. 483–507.
26. Maccoby, “Gender and gender relationships.”
27. Bateson, “Rules for changing the rules.”
28. Bateson, “Optimal outbreeding.”
29. Durham, Coevolution.
30. Ibid.
31. E. O. Wilson, Consilience: The Unity of Knowledge (New York: Knopf,
1998).
32. K. Aoki and M. W. Feldman, “A gene-culture coevolutionary model for
brother-sister mating,” Proceedings of the National Academy of Sciences, USA,
vol. 94, no. 13 (1997), pp. 46–50.
33. Edward Westermarck, The History of Human Marriage (London:
Macmillan, 1891).
2 Genetic Aspects of Inbreeding and Incest

Alan H. Bittles

When referring to humans, the term inbreeding is used to describe

unions between couples known to share at least one common ancestor.
While now rare in Western societies, marriages between close biological
kin are preferential in many parts of the world, including north and sub-
Saharan Africa, the Middle East, Central Asia, and much of the Indian
subcontinent.1 Although the rates and types of inbred union may vary ac-
cording to religious and societal norms, marriages between first cousins are
especially common; for example, in Pakistan they currently account for ap-
proximately 50 percent of all marital unions.2
An incestuous relationship is a union between biological relatives that is
genetically closer than permissible under prevailing civil legislation. Most
commonly, incest is defined as sexual intercourse between persons defined
as first-degree relatives, that is, father-daughter, mother-son, or brother-
sister. However, in some countries, such as Scotland, the definition includes
half-sib and uncle-niece unions, where the partners have 25 percent of
shared genes.3
In all forms of consanguineous union the partners share genes inherited
from one or more common ancestors; for example, in first-cousin mar-
riages the spouses are predicted to have 12.5 percent of their genes in com-
mon. This means that on average their progeny will be homozygous at 6.25
percent of gene loci; that is, they will have inherited identical gene copies
from each parent at these sites in their genome. As shown in Table 2.1, this
is conventionally expressed as the coefficient of inbreeding (F), which for
first-cousin offspring is 0.0625, whereas in matings between first-degree
relatives the couple has 50 percent of their genes in common, with an
equivalent F value in their progeny of 0.25. Irrespective of the level of in-
breeding, if the same mutant gene is inherited from both parents, the indi-
vidual will express the disorder, prenatally, at birth, or later in life, depend-
 Genetic Aspects of Inbreeding and Incest 39

table 2.1
Major Types of Consanguineous Relationship
Fraction Coefficient of
Genetic of Genes Inbreeding (F)
Family Relationship Relationship in Common in Progeny

Incestuous First degree 1/2 0.25

Uncle-niece Second degree 1/4 0.125
Double first cousin 1/4 0.125
First cousin Third degree 1/8 0.0625
First cousin once removed Fourth degree 1/16 0.0313
Double second cousin 1/16 0.0313
Second cousin Fifth degree 1/32 0.0156

ing on the nature and site of the mutation, thus contributing to the phe-
nomenon of inbreeding depression.

Attitudes Toward Consanguineous Marriage

Within Different Religions

Many examples of consanguineous unions are cited in the biblical

texts, with Abraham and Sarah identified as half-brother and sister (Genesis
20: 12), and Amran and Jochebed, the parents of Aaron and Moses, related
as nephew and aunt (Exodus 6: 20). At a later date the permitted degrees of
marital relationships between biological kin were extensively defined, with
marriages up to and including uncle-niece (but not aunt-nephew) permitted
(Leviticus 18: 7–18).
In attempting to assess the rationale underpinning these regulations, it is
important to acknowledge that many early societies possessed a quite so-
phisticated knowledge of potentially fatal inherited disorders. For example,
the X-linked recessive mode of inheritance of hemophilia A appears to be
recognized in the Talmud and other Jewish religious texts dating back to
the second to fifth centuries ad. Dispensation for the normally obligatory
male circumcision on the eighth day of life could be granted under two sets
of circumstances:
1. If a woman had given birth to two or three sons who had died
following circumcision, any future male children she might bear would
be excused circumcision even if the pregnancy was with a different
husband.
2. Where the sons of three sisters had died following circumcision,
all male children born to other female siblings would also be excused
circumcision.4
40 bittles

According to the Venerable Bede, an early ruling on consanguinity within

Christianity was given by Pope Gregory I to St. Augustine, the first arch-
bishop of Canterbury, in approximately ad 597.5 Citing Leviticus 18: 6, the
Pope advised that marriages between consanguineous spouses did not result
in children and that sacred law forbade a man to “uncover the nakedness of
his kindred.” For members of the Latin Church, the effect of this ruling was
to prohibit marriage with a biological relative, usually up to and including
third cousins, although because of different methods of calculating the de-
grees of biological relationship there was some confusion as to exactly
which degrees of consanguineous union were permitted or prohibited. This
problem was solved in the canon issued by Pope Alexander II in ad 1076,
which resulted in a shift in the formal method of consanguinity classifica-
tion from the Roman to the Germanic system. For example, third-cousin
marriages, which according to the Roman system were of the eighth degree,
became fourth degree under the Germanic classification.6
The formal proscription of consanguineous unions was confirmed in
1215 by the Fourth Lateran Council and generally remained in force within
the Latin church until 1917, when it was limited to unions between couples
related as first cousins or closer. Dispensation could, however, be granted at
Diocesan level for related couples who wished to marry within the prohib-
ited degrees of consanguinity, albeit with payment of an appropriate bene-
faction to the church.7
Similar strict rules governing consanguineous marriage continue to be
applied by the Christian Orthodox Church. Among the Reformed Protes-
tant denominations, the existing biblical guidelines (Leviticus 18: 7–18)
were generally adopted, although the closest form of approved union usu-
ally has been between first cousins. Somewhat paradoxically, the highest
rates of consanguineous unions historically recorded in Europe appear to
have been in the southern Roman Catholic countries, where consanguinity
was subject to church sanction, rather than the Protestant countries of
northern and northwestern Europe.8
Among other major world religions—Buddhism, Islam, and Judaism, and
in smaller religious communities such as the Zoroastrians/ Parsis—attitudes
toward close kin marriage are generally favorable or neutral.9 Cross-cousin
marriages also are permitted in the Chinese Taoist /Confucian tradition,10
and in rural Han communities in the period 1949–67, 0.7 percent to 1.2 per-
cent of marriages were between first cousins.11 While a generally tolerant ap-
proach was adopted toward marriage between biological relatives at the
inception of the People’s Republic of China,12 under the terms of the 1981
Marriage Act unions between first cousins became illegal. In the face of this
prohibition, it must be assumed that their prevalence has since declined, at
least in the majority Han population.
 Genetic Aspects of Inbreeding and Incest 41

The situation within Hinduism is much more complicated and embodies

three sets of regulations, one of which promotes endogamy whereas the
other two require exogamy. At the uppermost level of differentiation, all
Hindu communities are structured into a number of hierarchical groups,
that is, varnas, castes, and subcastes, each of which has its own set of
rights, duties, and privileges.13 There are four varnas—the Brahmins, Ksha-
triyas, Vaishyas, and Sudras—and each varna is subdivided into castes
(jati), with caste membership hereditary and basically inviolable. There-
fore, with few exceptions, caste endogamy is obligatory. Castes are further
subdivided into gotras, which are transmitted through the male line and are
exogamous insofar as marriages between the children of two brothers are
prohibited.14 In addition, according to sapinda regulation, persons related
to each other within a certain number of generations on the paternal and
maternal sides of a pedigree are forbidden to marry.15
The net result is that virtually all Hindu marriages continue to be con-
tracted within caste boundaries, and gotra regulations also are observed on
a near-universal basis. But at the sapinda level, major differences are ob-
served between the peoples of North and those of South India. Among the
Aryan Hindus of North India, pedigrees are examined over an average of
seven generations for males and five generations for females to ensure avoid-
ance of a consanguineous union.16 By comparison, uncle-niece marriage
and first-cousin unions between a man and his maternal uncle’s daughter
(mother’s brother’s daughter) have a long tradition in South India,17 where
they continue to be preferential, especially in the states of Andhra Pradesh,
Karnataka, and Tamil Nadu.18

The Development of Civil Legislation on Consanguineous

Marriage in Western Societies

Prior to the mid-nineteenth century, first-cousin marriages were com-

monly contracted in many Western societies. From the 1850s onward, an
often heated debate was conducted within the scientific and medical com-
munities of North America and Western Europe over the biological effects
of close kin marriage.19 Charles Darwin, who had married his first cousin
Emma Wedgwood, became a leading protagonist in the controversy. Through
the intervention of Sir John Lubbock, Darwin attempted to persuade the
Parliament of Great Britain and Ireland that the prevalence of first-cousin
marriages should be determined in the 1871 census. The proposal was de-
nied by a parliamentary select committee, largely on the grounds that the
subject was of great sensitivity and any formal inquiry would be unaccept-
ably intrusive. For example, according to the report in Hansard, committee
42 bittles

member Gathorne Hardy stated that “he did not see the desirability of
holding up families where such marriages had taken place, and the children
being anatomised for the benefit of science.”20 Likewise, in the opinion of
his colleague Mr. Locke, “This was a piece of the grossest cruelty ever
thought of.” Darwin’s marriage to his cousin had produced eleven chil-
dren, and he displayed considerable annoyance toward the politicians’ re-
buttal, observing in The Descent of Man, “When the principles of breeding
and of inheritance are better understood, we shall not hear ignorant mem-
bers of our legislature rejecting with scorn a plan for ascertaining by an
easy method whether or not consanguineous marriages are injurious to
man.”21
To overcome the resulting stalemate, Darwin’s son G. H. Darwin de-
vised a method of assessing the prevalence of first-cousin unions using sur-
name analysis.22 On presentation of his findings to the Statistical Society in
London, the younger Darwin’s efforts were praised by Francis Galton, who
was half-cousin to Charles Darwin, for the success he had achieved in at
least partially correcting “an exaggerated opinion which was current as to
the evil resulting from first-cousin marriages.”
Within Europe the debate on consanguinity resulted in few if any changes
in legislation. However, in the United States the eventual outcome was the
passage of laws at state level to control, and in many cases to ban, first-
cousin unions, even though much of the information on which these laws
were based was of questionable biological validity. Despite a unanimous
recommendation by the National Conference of Commissioners on Uni-
form State Laws that first-cousin marriages be freely permitted in the United
States,23 such marriages remain illegal in twenty-two states and are a crim-
inal offense in eight others, as shown in Figure 2.1.24 In some states the tra-
ditions of specific communities were recognized, and so, for example,
uncle-niece marriages are permissible in the Jewish community of Rhode Is-
land, while in Colorado Native Americans may marry their stepchildren.25
Wisconsin provides an interesting example of a state where legislators had
apparently attempted to permit first-cousin unions while circumventing
possible adverse biological outcomes; first-cousin marriages are permitted
if one or both partners are infertile or the female is over fifty-five years of
age and thus presumed to be postmenopausal.26

Societies in Which Consanguineous Unions

Are Preferential

As indicated in the introductory paragraph, the current legal situa-

tion in the United States differs from that in many other parts of the world.
 Genetic Aspects of Inbreeding and Incest 43

figure 2.1. Map of the United States Indicating States in Which First-Cousin
Marriages Are Legal and Those in Which They Are Civil or Criminal Offenses

Preferential consanguineous marriage is mainly explained in social and eco-

nomic terms, and the reasons given include the strengthening of family re-
lationships and the maintenance of family property, including landholdings.
Prenuptial arrangements also are greatly simplified, and the security of mar-
rying a partner whose entire family background is known is perceived as a
major benefit in ensuring the success and stability of the union, with lower
divorce rates among consanguineous couples.27 In general terms, the high-
est rates of consanguineous union are reported in rural areas among the
poorest and least educated sections of society, although landowning fami-
lies also arrange intrafamilial marriages to ensure the maintenance of their
landholdings.28
Variations in the specific types of marriage contracted—for example,
with first-cousin unions between a man and his father’s brother’s daughter
preferred in Arab Muslim communities,29 as opposed to mother’s-brother’s-
daughter marriage in such disparate populations as the Dravidian Hindu
states of South India,30 the Han Chinese,31 and the Tuareg of North Af-
rica32—indicate additional customary influences. Equally, while uncle-niece
unions (F = 0.125) are practiced within Judaism, they are proscribed by the
Koran, even though Muslims are permitted double first-cousin marriages,
which represent the same genetic distance.
44 bittles

The Influence of Consanguinity on Reproductive

Behavior and Success

Based on data collected from the Hutterites, a highly endogamous

Anabaptist sect resident in South Dakota, it has been claimed that mate
choice is influenced by HLA haplotypes, with a lower-than-expected inci-
dence of HLA-haplotype matches.33 These studies, which appear to impli-
cate HLA-associated human odor preferences, are preliminary and some-
what controversial, and it has been suggested that the underlying biological
mechanism may operate effectively only in communities with very limited
HLA repertoires, such as the Hutterites themselves.34 If this is correct, then
similar preconditions could potentially exist in many highly inbred commun-
ities. Nevertheless, the relevance of the findings to the majority of present-
day consanguineous unions is probably marginal, since in populations
where consanguinity is highly preferential, such as the Middle East, South
India, and Pakistan, marriages are customarily arranged by the parents of
the couple. Thus the potential degree of autonomy in partner choice is very
limited.
Enhanced genetic compatibility would be expected between mother and
fetus in marriages between close biological kin, resulting in an advanta-
geous pregnancy outcome because of decreased rates of ABO and more es-
pecially Rh incompatibility35 and preeclamptic toxemia.36 Conversely, ac-
cording to the fetal allograft concept,37 antigenic disparity between mother
and fetus is beneficial to fetal development.38 The hypothesis has been ex-
tensively examined in human populations, especially by reference to women
who have experienced primary recurrent spontaneous abortions with all
conceptions lost. In some studies there was a significant positive association
between parental allele sharing at HLA loci, principally involving HLA-
DR, -DQ, and -B alleles, whereas in others none could be detected.39
Even if a positive association does exist between infertility and parental
HLA allele sharing, this association is far from complete, as successful
pregnancies have been described in which the HLA haplotypes of the
mother and offspring were identical.40 Furthermore, in the highly inbred
Hutterite community, couples who shared HLA-DR antigens had a median
completed family size of 6.5, as opposed to 9.0 among those with no HLA
alleles in common.41 Accordingly, it has been proposed that the association
between parental HLA compatibility and recurrent spontaneous abortion
may best be explained in terms of the expression of deleterious recessive
genes located in the HLA-DQ-DR-B region of the major histocompatibility
complex.42
The validity of each of these hypotheses, and their direct relevance to
highly inbred unions, has to be judged against the very high spontaneous
 Genetic Aspects of Inbreeding and Incest 45

abortion rates in humans, with on average some 40 percent of all post-

implantation conceptions lost.43 Reduced levels of pathological sterility
have been reported in a number of different inbred populations,44 and
other indirect indicators of fetal survival, such as rates of multiple births
and the secondary sex ratio, also have failed to show an adverse inbreed-
ing effect.45
Conversely, there is preliminary evidence that consanguineous couples
enrolled in assisted reproductive technology programs may have higher mis-
carriage rates,46 and studies conducted in the Hutterite community indicate
somewhat longer birth intervals in the more inbred women, possibly owing
either to lower conception rates or higher peri-implantation losses.47 In ad-
dition, DNA-based investigations among highly inbred communities in Pak-
istan have suggested that there may be selection against homozygosity at
specific gene loci involved in embryonic and early fetal development.48
Assessed in terms of completed family size, a recent meta-analysis was
unable to identify any overall adverse effect of inbreeding on fertility, and
in a large majority of the constituent studies consanguineous couples actu-
ally had more children.49 The data did, however, suggest that the greater
number of children born to related couples in part reflected reproductive
compensation for children dying at a young age.

Estimating the Biological Outcomes of

Consanguineous Marriages

As indicated in the introduction, the coefficient of inbreeding (F) is a

numerical estimate of the degree of inbreeding of an individual. Of course in
many communities there is a long continuous history of consanguineous
unions, and so the cumulative level of inbreeding may be significantly higher
than the value calculated for a single generation. In such populations a cor-
rection can be applied to account for the effects of ancestral inbreeding, us-
ing this formula:
F = ∑ (½)n (1 + FA)
where FA is the ancestor’s inbreeding coefficient, n is the number of individ-
uals in the path connecting the parents of the individual, and the summa-
tion (∑) is taken over each path in the pedigree that goes through a com-
mon ancestor.
The first structured study into the biological effects of inbreeding was
organized by Samuel Bemiss of Louisville, Kentucky, who collated reports
forwarded by medical colleagues on the outcomes of unions ranging from
incest to third-cousin marriages.50 Since that time, many additional studies
46 bittles

have been undertaken, based on a variety of sampling techniques including

pedigree analysis, household surveys, questionnaires administered to hos-
pital in- and outpatients, Roman Catholic dispensation records, and iso-
nymy (surname analysis). In the populations of Dravidian South India,
where uncle-niece and first-cousin unions are preferential and jointly ac-
count for some 30 percent of marriages, unions beyond second cousins
(F < 0.0156) are of very limited biological significance.51 By comparison, in
countries such as Finland, where consanguineous unions generally are
rare, biologically remote relationships may result in a moderate level of cu-
mulative inbreeding through time and may be of potential clinical rele-
vance.52 This situation also applies in endogamous communities in which
close cousin marriage has been proscribed on religious grounds but mar-
riages between couples related to a lesser degree are frequently contracted,
an example being the Roman Catholic community on the isle of Eriskay in
Scotland.53
In an attempt to rationalize the results of inbreeding surveys into a
meaningful and reproducible format, a method was devised to calculate the
numbers of lethal gene equivalents in a community or population by com-
paring death rates in the progeny of consanguineous and that of unrelated
couples. Lethal gene equivalents are the numbers of detrimental recessive
genes carried by an individual in the heterozygous state, which, if homo-
zygous, would result in death. In a population, the number of lethal gene
equivalents can be calculated according to this formula:
–loge S = A + BF
where S is the proportion of survivors in the study population, A measures
all deaths that occur under random mating, B represents all deaths caused
by the expression of recessive genes via inbreeding, and F is the coefficient
of inbreeding.54 By plotting a weighted regression of the log proportion of
survivors (S) at different levels of inbreeding (F), A can be determined from
the intercept on the y-axis at zero inbreeding (F = 0), and B (the number of
lethal gene equivalents) is given by the slope of the regression. A number of
theoretical and methodological limitations have been identified;55 however,
the regression technique offers a simple and convenient method to assess
and compare data on the effects of inbreeding in different populations.
Through time estimates of the numbers of lethal gene equivalents per
population have been revised downward,56 in part because of better sam-
pling techniques and the recognition that earlier surveys may have pro-
duced spuriously high values owing to inadequate control for nongenetic
variables, including socioeconomic status. Other, more direct methods have
also been used, and in a study that compared deaths in the offspring of
first-cousin and unrelated unions from approximately six months gestation
 Genetic Aspects of Inbreeding and Incest 47

to ten years of age, mortality was 4.4 percent higher in the consanguineous
group, equivalent to 1.4 lethal gene equivalents per zygote.57

The Phenomenon of Incest Avoidance

With a few notable exceptions, incest avoidance appears to be com-

mon to virtually all societies. Even in contemporary fiction—for example,
Ian McEwan’s Cement Garden and A. S. Byatt’s Angels and Insects—any
breach of this accepted norm evokes outrage.58 Although a genetic ratio-
nale cannot be dismissed, the various explanations given for this type of re-
sponse are mainly social in origin, including Freudian guilt, and theories as-
cribed to Lévi-Strauss and others that the incest taboo serves both to
maintain the existence of the family and to encourage the establishment of
affinal relations with other kin groups.
According to the Westermarck hypothesis, incest avoidance can be ex-
plained in terms of negative imprinting against close associates of early
childhood.59 Studies conducted on individuals raised under mixed-sex
child-rearing regimes in Israeli kibbutzim have been cited in support of the
negative imprinting theory, since they rarely marry or enter into sexual re-
lationships.60 More convincingly, a detailed investigation into the practice
of sim-pua marriage in Taiwan, whereby a girl is adopted with the expec-
tation that she will marry a son of the adoptive family, clearly showed both
lower mean fertility and lesser marital stability.61 (For further discussion on
this topic, see Chapter 4.)

Historical Examples of Dynastic and

Nondynastic Incest

There have been a number of societies in which dynastic incest was

practiced over multiple generations, including Egypt during the eighteenth
and nineteenth dynasties (prior to 332 bc) and the Ptolemaic (323–30 bc)
and Roman (30 bc–ad 324) periods, Zoroastrian Iran, the Incas, and the
royal families of Hawaii.62 In Pharaonic Egypt it was believed that the
royal bloodline would be maintained and strengthened through brother-
sister unions, although the Pharaoh also had ready access to other, unre-
lated brides and concubines. Increasingly, information has emerged on in-
cestuous marriages outside these ruling classes in which marriage contracts
were written and the bride brought her own dowry.63 For example, in Ro-
man Egypt, full brother-sister unions accounted for 19.6 percent of mar-
riages in the city of Arsinoe, with a further 3.9 percent of marriages be-
tween half siblings (see Chapter 5).64
48 bittles

The rationale for incestuous unions among Zoroastrians appears to

have been based on religious principles and to have been associated with
their attitudes toward marriage, which was regarded as divinely favored
and almost akin to a religious duty. While unknown in the Avesta, xvaét-
vadatha, or xvétxvét, (usually translated as next-of-kin marriage) was first
discussed in the Pahlavi texts of the sixth to ninth centuries ad,65 with all
three types of incestuous union—father-daughter, mother-son, and brother-
sister—advocated. Xvétxvét was described as being of special religious
merit, being the ninth of thirty-nine ways of gaining entry to heaven, and
its practice was accepted as a means of expiating mortal sin.
It has been calculated that in early human societies, with restricted num-
bers of potential marriage partners, an incest taboo could result in consid-
erable demographic costs to the community.66 The practical desirability of
incestuous unions on financial grounds, and to prevent the subdivision of
landholdings, was later considered by the Christian Cathar community of
Occitania in the Middle Ages,67 a trait that probably contributed to its sup-
pression by the Latin church. Interestingly, in the light of contemporary in-
formation on the adverse health effects of incestuous unions described be-
low, the genealogies of the eighteenth- and nineteenth-dynasty Egyptian
kings have given no indication of reduced reproductive capacity, and there
was little or no recorded evidence of physical or possible mental defects in
the mummies of the royal brother-sister offspring.68

The Prevalence and Types of Incestuous

Unions in Contemporary Societies

There is the conundrum that, despite the claimed near-universality of

incest avoidance, an extensive and growing literature exists on the occur-
rence of incest in contemporary Western societies. These reports most com-
monly have emanated from case studies conducted on persons examined be-
cause of intellectual handicap,69 consultations with patients in psychiatric
clinics,70 or individual cases of young children diagnosed with a rare inher-
ited disorder,71 thus leading to the general conclusion that incestuous rela-
tionships are highly detrimental to those classified as victims. There have
been occasional reports suggesting that the outcome need not necessarily be
psychologically damaging, even when a long-term father-daughter relation-
ship had commenced while the female was still prepubertal.72 As incest is
both illegal and widely regarded as morally reprehensible, it would be diffi-
cult to ascertain how commonplace this latter response may be, without ex-
posing individuals to the risk of self-incrimination (for further discussion on
this topic, see Chapter 9).
 Genetic Aspects of Inbreeding and Incest 49

The Biological Outcomes of Incest

Given the strongly judgmental societal attitudes, direct assessment of

the biological outcomes of incest may be subject to significant ascertain-
ment bias, as a rigorous examination, including determination of paternity,
may be initiated only if a child shows symptoms of physical or intellectual
handicap, or both. The net result is a marked lack of data, which probably
does not reflect the actual numbers of incestuous pregnancies that success-
fully proceed to term. Problems also arise in attempting to control for the
potentially adverse effects of nongenetic variables, such as very young or
advanced maternal and paternal ages, parental disease, and unsuccessful at-
tempted interruption of the pregnancy.
These points are illustrated in Table 2.2, which contains outcome infor-
mation from the four most comprehensive studies of incest that have been
reported to date. In three of the studies the data were collected prospec-
tively,73 and in the fourth the information was collected from early in post-
natal life.74 In the studies there were equal numbers of father-daughter and
sib incest (for each, n = 96), and one mother-son offspring. This is some-
what surprising, as most commentators on the subject have considered
father-daughter relationships to be the most common form of incest.
The first and obvious observation is that very few cases of children born
to incestuous unions have been systematically investigated, surprisingly so,
since the topic features so prominently in the social and behavioral litera-
ture. The brief report by C. O. Carter is severely restricted by a lack of in-
formation on the physical and mental status of the parents and their so-
cioeconomic background, and by a lack of controls.75 In one of the studies,
strenuous efforts were made to recruit matched nonincestuous controls,
and the small numbers are indicative of the difficulties encountered by the
investigators in this task.76
The numerically largest study was that of E. Seemanová, retrospectively
conducted in the former Czechoslovakia on incestuous births reported be-
tween 1933 and 1970.77 The method of control employed is at first sight
very convincing, with the outcomes of incestuous matings compared with
those of pregnancies undertaken by a subset of the same women with unre-
lated partners. Unfortunately, the methodology could not overcome the fact
that the physical and mental states of this latter group of women, just 33
percent of the original total, was significantly superior to the remaining
majority of those studied.78 For example, of the 141 females in incestuous
unions, twenty were intellectually handicapped (of whom two additionally
were deaf-mutes, two had congenital syphilis, and two were epileptics), a
further two were deaf-mutes and three were schizophrenic. By comparison,
among the forty-six control mothers, that is, those who also had pregnan-
 table 2.2
Levels of Death and Defect Reported in Four Studies of Incest
Others,
Congenital Nonspecific Including
Autosomal Malformations/ Severe Mild
Country No. Follow-up Recessive Sudden Infant Intellectual Intellectual
of Origin Studied (yr.) Disorders Deaths Handicap Handicap Normal Source

United States 18 0.5 2 4 0 5 7 Adams and Neel 1967

United Kingdom 13 4–6 2 1 1 4 5 Carter 1967
Czechoslovakia 161 1–37 20 21 24 18 78 Seemanová 1971
Canada 21 0.5–1.9 1 8 0 4 8 Baird and McGillivray 1982
Totals 213 25 34 25 31 98
(11.7%) (16.0%) (11.7%) (14.6%) (46.0%)

source: C. O. Carter, “Risk to offspring of incest,” The Lancet, vol. 289 (1967), p. 436.
 Genetic Aspects of Inbreeding and Incest 51

cies with unrelated males, only two were intellectually impaired (one addi-
tionally being a deaf-mute) and two others were deaf-mutes. Similarly, of
the 138 fathers in the incestuous unions, eight were intellectually handi-
capped, thirteen were chronic alcoholics, two had syphilis, and four had
subsequently committed suicide. Whereas in the nonincestuous control group
of fifty-two fathers, none were intellectually handicapped, two were chronic
alcoholics, and there was one case of polydactyly.
In such a relatively small-scale study, the actions of a single individual also
can greatly influence the overall collated results. According to Seemanová,
one alcoholic individual had fathered five children with three different daugh-
ters; each of these five children had been diagnosed with varying degrees and
types of abnormality, and three had died within the first ten days of life.79
The very young age of many of the mothers was a further factor that
may have adversely affected the viability and health of the incestuous prog-
eny. The mean and modal maternal ages at time of birth in the father-
daughter matings were 18.9 and 16 years, and 19.9 and 14 years in the
brother-sister matings, versus 24.9 and 21 years in the married control
group. Given the extended time period over which the study was conducted
and the marked negative secular trend with regard to menarche in Europe
during the middle decades of the twentieth century,80 it seems probable that
many of these females had just entered their reproductive phase when the
pregnancy commenced.
Conceptions initiated within two years of menarche may be associated
with gynecological immaturity and incomplete pelvic growth, and hence an
adverse pregnancy outcome. For example, in the United States pregnancies
among black females of fourteen years of age or younger resulted in adverse
maternal health outcomes, including acute toxemia, uterine dysfunction, and
one-day fever, and their progeny had elevated rates of cardiovascular anom-
alies.81 Likewise, the progeny of white females aged thirteen to fifteen years
exhibited lower birth weight, prematurity, and small size for gestational
age.82 Some of these differences may have been social in origin and may
have reflected the disadvantaged circumstances of the mothers prior to and
during pregnancy.83 However, pregnancies in very young women have been
shown to exhibit increased rates of chromosomal anomalies,84 congenital
abnormalities,85 and neonatal and postneonatal mortality.86
With these factors in mind, while unavoidable, the age discrepancies be-
tween the mothers in incestuous unions and those in unrelated unions may
cause significant difficulties in comparing the test and control groups. In-
deed, if the data are censored to exclude physical and mental abnormalities
among the male and female parents, and major disparities with respect to
young and advanced maternal age, few differences remain in the overall
health outcomes recorded for each group.
52 bittles

Combining the cases of specific autosomal recessive disorders, major

congenital malformations, and nonspecific severe intellectual handicap, 64
out of 213 (39.4 percent) of the progeny of incestuous unions were ad-
versely affected and 30 (14.1 percent) of these individuals had died. In the
two studies for which nonconsanguineous reference groups were available,
9 out of 113 control children died or a serious defect was diagnosed (8.0
percent). Thus the mean excess level of death and severe defect in the off-
spring of incestuous unions was 31.4 percent, not all of which would nec-
essarily be genetic in origin.
How should these four data sets be assessed? Clearly, the composite re-
sults indicate a high rate of physical and mental abnormality. The collection
of data on this subject is, however, extremely difficult and, as acknowledged
by several of the authors, control for nongenetic variables in the incestuous
unions may have been incomplete. Attempted illegal abortions were reported
in a number of incestuous pregnancies in the earlier studies, which also may
have resulted in nonfatal damage to the developing fetus.87 Given these var-
ious circumstances, the average level of reported death and severe defect
(39.4 percent) can perhaps best be considered as an upper bound estimate,
but one that may well be inflated by a variety of important nongenetic fac-
tors, such as the following:
Young maternal age
Advanced paternal and maternal ages
Physical or mental abnormality, or both, in one or both parents
Low socioeconomic status
Specific adverse environmental and social effects, including
attempted abortion
The second major method involves calculation of the probable levels of
defect in the progeny of incestuous matings, based on information gained
from legal consanguineous unions. For example, it is assumed that as the
progeny of an incestuous mating have a coefficient of inbreeding of F =
0.25, their levels of disease and disability will be approximately four times
that observed in first-cousin offspring (F = 0.0625). The main advantage of
this approach is that data on first-cousin unions are more plentiful, and they
are less likely to be subject to ascertainment bias. The potential disadvan-
tage is that the relationship between the degree of inbreeding and rates of
disease and disability may be curvilinear in nature, with disproportionate in-
creases in the prevalence of deaths and defects at closer levels of inbreeding.
This situation could arise if incest was associated with increased rates of
conditional lethals, genes that are expressed only under especially stressful
circumstances, as might be encountered in an incestuous pregnancy.
Two lines of evidence suggest that conditional lethals may not form a
 Genetic Aspects of Inbreeding and Incest 53

significant component of the inbreeding load in the progeny of incestuous

unions. The first is based on the study of survival rates at first-cousin level,
comparing the progeny of first-cousin and nonconsanguineous unions and
with the data plotted according to the standard equation y = a + bx. If con-
ditional lethals were expressed in close kin unions, then one would expect
either a regression term (b) significantly greater than 1.0 or a dispropor-
tionately greater effect of inbreeding in populations with higher infant and
childhood death rates. In fact, the estimate for the regression line was 1.02
± 0.09, with a coefficient of determination (R2) of 0.80, even though the
levels of background mortality ranged from 3.3 percent to 39.5 percent.88
The second complementary piece of evidence is based on a follow-up
study, conducted in a similar manner on a subset of twenty-one of these
populations for which data were available at F = 0.125, that is, the progeny
of uncle-niece or double-first-cousin marriages. In many cases the sample
sizes were small, reflecting the lower frequency of these marriages, and as a
result R2 = 0.49. Nonetheless, mean excess mortality was 9.6 percent (stan-
dard error 4.7), with a value for the regression line of 0.95 ± 0.21, and thus
broadly comparable in effect to the figure of 4.4 percent derived for first-
cousin progeny (F = 0.0625).
If the excess mortality rate of 4.4 percent derived for deaths among first-
cousin offspring89 is applied as the basis for estimating lethal defects in an
incestuous pregnancy, the net result would suggest approximately 16 per-
cent to 20 percent excess deaths at F = 0.25. Based on the Japanese data of
W. J. Schull and J. V. Neel,90 excess morbidity among incestuous offspring
would predictably range from 6 percent to 16 percent, thus suggesting a to-
tal excess rate for death and major disability of between 22 percent and 36
percent, not all of which would necessarily be genetic in origin.

Discussion and Conclusions

The data currently available on the outcomes of both legal consan-

guineous marriages and incestuous relationships merit further examination
and updating, and in evolutionary terms due allowance for the possible
beneficial effects of kin altruism should be included (for further discussion
on this topic, see Chapter 1). Since the classic studies by Schull and Neel in
Japan, no comparably detailed investigation has been undertaken, despite
the remarkable epidemiological transitions that have occurred during the
last two generations.91 Across Asia, the effects of genetic disorders are be-
coming increasingly obvious as deaths owing to infectious diseases decline
and early childhood mortality is replaced by morbidity. This change is es-
pecially important among couples in consanguineous unions, where higher
rates of genetic disorders with both autosomal recessive and polygenic
54 bittles

modes of inheritance can be expected.92 It is likely to be of greatest imme-

diate significance within migrant communities from Africa and Asia now
resident in North America, Western Europe, and Australasia, who have
chosen to continue their tradition of close kin marriage.
With respect to incest, it seems remarkable that the voluminous social
and behavioral literature has been so inadequately matched in biological
terms. This raises the question as to whether, from a genetic perspective, in-
cest should most appropriately be regarded as an extreme point on the
spectrum of consanguinity or as a separate and distinct entity that mainly
relies for its significance on social approbation. Irrespective of whichever
stance is adopted, comprehensive studies to identify the precise nature, ex-
tent, and timing of any genetic defects that may arise are needed, especially
since in a proportion of cases the children of incestuous unions are placed
for adoption.93
The advent of DNA-based analytical methods could greatly facilitate this
process and provide results of previously unparalleled insight and sophisti-
cation.94 Indeed, there have been cases where the investigation of a severe au-
tosomal recessive disorder has revealed a previously unsuspected incestuous
relationship.95 Given the social stigma associated with incest, continuing dif-
ficulties can be expected in obtaining unbiased data sets with sufficiently large
numbers of subjects, especially since in many Western countries therapeutic
abortion may be offered in known or suspected cases of incest.96 The degree
to which this latter option is acceptable depends on the religious beliefs and
possibly the ethnic origins of the individuals concerned, which in turn could
result in overrepresentation of certain groups in surveys of incest.97 For the
foreseeable future, the data sets summarized in Table 2.2 may remain our
best, if imperfect, direct estimates of the genetic outcomes of incest.
At the same time, changes have been introduced into civil legislation
governing permitted biological relationships within marriage. In the Neth-
erlands, special permission may be granted by the Ministry of Justice for a
male or female to marry their adoptive sibling, and under the terms of the
Swedish Marriage Law of 1987, half-sibs may marry, once again subject to
special governmental approval. As in the past, there is a more proscriptive
general attitude in the United States, exemplified during 1999–2000 by the
introduction into the Maryland state legislature of an ultimately unsuccess-
ful bill to ban first-cousin marriages. No specific guiding principle is dis-
cernible in these apparently contradictory approaches, and in the case of
the Maryland bill, the empirical biological data presented in support of the
legislative change were dated and of dubious merit. While this is clearly an
area that merits future informed attention, in the short term agreement re-
mains improbable at the international level, and within the United States
even between individual states.
 Genetic Aspects of Inbreeding and Incest 55

notes
1. A. H. Bittles, Empirical Estimates of the Prevalence of Consanguineous
Marriage in Contemporary Societies. Working Report no. 74, Morrison Institute
for Population and Resource Studies (Stanford, Calif.: Stanford University, 1998);
A. H. Bittles, “Consanguinity,” http://www.consang.net (2003).
2. R. Hussain and A. H. Bittles, “The prevalence and demographic characteris-
tics of consanguineous marriage in Pakistan,” Journal of Biosocial Science, vol. 30
(1998), pp. 261–79.
3. C. H. W. Gane and C. N. Stoddart, A Casebook on Scottish Criminal Law
(Edinburgh: Green, 1988), pp. 674–77.
4. F. Rosner, “Hemophilia in the Talmud and Rabbinic writings,” Annals of
Internal Medicine, vol. 70 (1969), pp. 833–37.
5. Bede, The Ecclesiastical History of the English People (731: reprint, Lon-
don: Penguin Books, 1990), pp. 79–81.
6. J. Goody, The Development of the Family and Marriage in Europe (Cam-
bridge: Cambridge University Press, 1985), pp. 134–36.
7. Ibid., pp. 136–39.
8. J. M. McCullough and D. H. O’Rourke, “Geographical distribution of con-
sanguinity in Europe,” Annals of Human Biology, vol. 13 (1986), pp. 359–68.
9. A. H. Bittles, H. S. Savithri, H. S. Venkatesha Murthy, W. Wang, J. Cahill,
G. Baskaran, and N. Appaji Rao, “Consanguineous marriage: A familiar story full
of surprises,” in Ethnicity and Health, ed. H. Macbeth and P. Sherry (London:
Taylor and Francis, 2001), pp. 68–78.
10. H.-T. Fei, Peasant Life in China (New York: Dutton, 1939), p. 50; W. R.
Geddes, Peasant Life in Communist China (Ithaca, N.Y.: Cornell University Press,
1963), p. 28; E. Cooper, “Cousin marriage in rural China: More and less than
generalized exchange,” American Ethnologist, vol. 20 (1993), pp. 758–80.
11. R. B. Du, Z. L. Zhao, L. J. Xu, Y. F. Wang, W. Y. Cui, Z. R. Mao, et al.,
“Percentages and types of consanguineous marriages of different nationalities and
regions in China,” National Medical Journal of China, vol. 61 (1981), pp. 723–28
[In Chinese].
12. W. Muller-Freienfels, “Soviet family law and comparative Chinese devel-
opments,” in Chinese Family Law and Social Change, ed. D. Buxbaum (Seattle:
University of Washington Press, 1978), pp. 323–99.
13. P. S. S. Rao, “Inbreeding in India: Concepts and consequences,” in The
People of India, ed. J. R. Lukacs (New York: Plenum, 1984), pp. 239–68.
14. L. D. Sanghvi, “Inbreeding in India,” Eugenics Quarterly, vol. 13 (1966),
pp. 291–301.
15. K. M. Kapadia, Marriage and Family in India, 2nd ed. (Calcutta: Oxford
University Press, 1958), pp. 117–37.
16. Ibid.
17. Ibid.; K. A. N. Sastri, A History of South India from Prehistoric Times to
the Fall of Vijayanagar, 4th ed. (Madras: Oxford University Press, 1976), p. 66.
18. A. H. Bittles, J. M. Coble, and N. Appaji Rao, “Trends in consanguineous
marriages in Karnataka, South India,” Journal of Biosocial Science, vol. 25 (1993),
56 bittles

pp. 111–16; International Institute of Population Sciences, National Family and

Health Survey, India (Bombay, IIPS, 1995).
19. A. H. Bittles, “The basis of Western attitudes towards consanguineous
marriage,” Development Medicine and Child Neurology, vol. 45 (2003),
pp. 135–38.
20. HMSO (Her Majesty’s Stationary Office), Parliamentary Debates, Third
Series (London: Her Majesty’s Stationary Office, 1870), pp. 1006–10.
21. C. Darwin, The Descent of Man, vol. 2 (London: John Murray, 1871),
p. 403.
22. G. H. Darwin, “Marriages between first cousins in England and Wales
and their effects,” Journal of the Statistical Society, vol. 38 (1875), pp. 153–84;
G. H. Darwin, “Note on the marriage of first cousins,” Journal of the Statistical
Society, vol. 38 (1875), pp. 344–48.
23. National Conferences of Commissioners, Handbook on Uniform State
Laws and Proceedings of the Annual Conference Meeting in Its Seventy-ninth
Year (Baltimore, Md.: Port City Press, 1970).
24. M. Ottenheimer, “Lewis Henry Morgan and the prohibition of cousin
marriage in the United States,” Journal of Family History, vol. 15 (1990),
pp. 325–34; M. Ottenheimer, Forbidden Relatives: The American Myth of
Cousin Marriage (Urbana: University of Illinois Press, 1996), pp. 19–41.
25. C. S. Bratt, “Incest statutes and the fundamental right of marriage: Is
Oedipus free to marry?” Family Law Quarterly, vol. 18 (1984), pp. 257–309.
26. M. G. Farrow and R. C. Juberg, “Genetics and laws prohibiting marriage
in the United States,” Journal of the American Medical Association, vol. 209
(1969), pp. 534–38; C. S. Bratt, “Incest statutes and the fundamental right of
marriage.”
27. A. H. Bittles, “The role and significance of consanguinity as a demo-
graphic variable,” Population and Development Review, vol. 20 (1994), pp.
561–84; R. Hussain, “Community perceptions of reasons for preference for
consanguineous marriages in Pakistan,” Journal of Biosocial Science, vol. 31
(1999), pp. 449–61.
28. Bittles, “The role and significance of consanguinity.”
29. M. Klat, “Endogamy in the Arab world,” in Genetic Disorders Among
Arab Populations, ed. A. S. Teebi and T. I. Farag (New York: Oxford University
Press, 1997), pp. 63–80.
30. P. S. S. Rao and S. G. Inbaraj, “Inbreeding in Tamil Nadu, South India,”
Social Biology, vol. 24 (1977), pp. 281–88.
31. L. Wu, “Investigations of consanguineous marriages among 30 Chinese
ethnic groups,” Heredity and Disease, vol. 4 (1987), 163–66 [in Chinese].
32. L. Degos, A. Colombani, A. Chaventre, B. Bengston, and A. Jacquard,
“Selective pressure on HL-A polymorphism,” Nature, vol. 249 (1974), pp. 62–63.
33. C. Ober, L. R. Weitkamp, N. Cox, D. Dytch, D. Kostyu, and S. Elias,
“HLA and mate choice in humans,” American Journal of Human Genetics,
vol. 16 (1997), pp. 497–504.
34. D. D. Kostyu, C. L. Ober, D. V. Dawson, M. Ghanayem, S. Elias, and
A. O. Martin, “Genetic analysis of HLA in the U.S. Schmiedenleut Hutterites,”
 Genetic Aspects of Inbreeding and Incest 57

American Journal of Human Genetics, vol. 45 (1989), pp. 262–69; Ober et. al,
“HLA and mate choice.”
35. C. Stern and D. R. Charles, “The Rhesus gene and the effect of consan-
guinity,” Science, vol. 101 (1945), pp. 305–7.
36. A. C. Stevenson, B. C. C. Davison, B. Say, S. Ustuoplu, D. Liya, M. Abul-
Einem, and H. K. Toppozada, “Contributions of fetal/maternal incompatibility to
aetiology of pre-eclamptic toxaemia,” The Lancet, vol. 298 (1971), pp. 1286–89;
A. C. Stevenson, B. Say, S. Ustaoglu, and Z. Durmas, “Aspects of pre-eclamptic
toxaemia of pregnancy, consanguinity, and twinning in Ankara,” Journal of Med-
ical Genetics, vol. 13 (1976), pp. 1–8.
37. B. Clarke and D. R. S. Kirby, “Maintenance of histocompatibility poly-
morphisms,” Nature, vol. 211 (1966), 999–1000.
38. M. Adinolfi, “Recurrent habitual abortion, HLA sharing, and deliberate
immunization with partner’s cells: A controversial topic,” Human Reproduction,
vol. 1 (1986), pp. 45–48; C. Ober, “HLA and pregnancy: The paradox of the fetal
allograft,” American Journal of Human Genetics, vol. 62 (1998), pp. 1–5.
39. A. H. Bittles and P. Matson, “Genetic influences on human fertility,” in
Infertility in the Modern World: Present and Future Prospects, ed. G. R. Bentley
and C. G. N. Mascie-Talyor (Cambridge: Cambridge University Press, 2000),
pp. 46–81.
40. D. C. Kilpatrick, “A case of materno-fetal histocompatibility—implications
for leucocyte transfusion treatment of recurrent abortions,” Scottish Medical Jour-
nal, vol. 29 (1984), 110–12; J. R. Oksenberg, E. Persitz, A. Amat, and C. Brautbar,
“Maternal-paternal histocompatibility: Lack of association with habitual abortion,”
Fertility and Sterility, vol. 42 (1984), pp. 389–95.
41. C. Ober, S. Elias, E. O’Brien, D. D. Kostyu, W. W. Hauck, and A. Bombard,
“HLA sharing and fertility in Hutterite couples: Evidence for prenatal selection
against compatible fetuses,” American Journal of Reproductive Immunology and
Microbiology, vol. 18 (1988), pp. 111–15.
42. K. Jin, H. N. Ho, T. Speed, and T. J. Gill, “Reproductive failure and the
major histocompatibility complex,” American Journal of Human Genetics, vol. 46
(1995), pp. 1456–67.
43. Bittles and Matson, “Genetic influences on human fertility.”
44. Y. Yanase, N. Fujiki, Y. Handa, M. Yamaguchi, Y. Kishimato, T. Furusho,
et al., “Genetic studies on inbreeding in some Japanese populations. XII. Studies
of isolated populations,” Japanese Journal of Human Genetics, vol. 17 (1973),
pp. 332–36; P. S. S. Rao and S. G. Inbaraj, “Inbreeding effects on human repro-
duction in Tamil Nadu of South India,” Annals of Human Genetics, vol. 41
(1977), pp. 87–98; M. Edmond and M. De Braekeleer, “Inbreeding effects on
fertility and sterility: A case-control study in Saguenay-Lac-Saint-Jean (Quebec,
Canada) based on a population registry,” Annals of Human Biology, vol. 20
(1993), pp. 545–55.
45. A. H. Bittles, A. Radha Rama Devi, and N. Appaji Rao, “Consanguinity,
twinning, and secondary sex ratio in the population of Karnataka, South India,”
Annals of Human Biology, vol. 15 (1988), pp. 455–60.
46. P. E. Egbase, M. Al-Sharhan, S. Al-Othman, M. Al-Mutawa, and J. G.
58 bittles

Grudzinskas, “Outcome of assisted reproduction technology in infertile couples

of consanguineous marriage,” Journal of Assisted Reproduction and Genetics,
vol. 13 (1996), pp. 279–81.
47. C. Ober, T. Hyslop, and W. W. Hauck, “Inbreeding effects on fertility in
humans: Evidence for reproductive compensation,” American Journal of Human
Genetics, vol. 64 (1999), pp. 225–31.
48. W. Wang, S. G. Sullivan, S. Amed, L. A. Zhivotovsky, and A. H. Bittles,
“A genome-based study of consanguinity in three co-resident endogamous Pak-
istan communities,” Annals of Human Genetics, vol. 64 (2000), pp. 41–49.
49. A. H. Bittles, J. C. Grant, S. G. Sullivan, and R. Hussain, “Does inbreed-
ing lead to decreased human fertility?” Annals of Human Biology, vol. 29 (2002),
pp. 111–31.
50. S. M. Bemiss, “Report on influence of marriages of consanguinity upon
offspring,” Transactions of the American Medical Association, vol. 11 (1858),
319–425.
51. A. H. Bittles, W. Mason, J. Greene, and N. Appaji Rao, “Reproductive
behavior and health in consanguineous marriages,” Science, vol. 252 (1991),
pp. 789–94.
52. E. O’Brien, L. B. Jorde, B. Ronnlof, J. O. Fellman, and A. W. Erickson,
“Founder effect and genetic disease in Sottunga, Finland,” American Journal of
Physical Anthropology, vol. 77 (1988), pp. 335–46.
53. A. P. Robinson, “Inbreeding as measured by dispensations and isonymy on
a small Hebridean island, Eriskay,” Human Biology, vol. 55 (1983), pp. 289–95.
54. Newton E. Morton, J. F. Crow, and H. J. Muller, “An estimate of the
mutational damage in man from data on consanguineous marriages,” Proceedings
of the National Academy of Sciences, USA, vol. 42 (1956), pp. 855–63.
55. E. Makov and A. H. Bittles, “On the choice of mathematical models for
the estimation of lethal gene equivalents in man,” Heredity, vol. 57 (1986),
pp. 377–80.
56. A. H. Bittles and E. Makov, “Inbreeding in human populations: Assess-
ment of the costs,” in Mating Patterns, ed. C. G. N. Mascie-Taylor and A. J. Boyce
(Cambridge: Cambridge University Press, 1988), pp. 153–67.
57. A. H. Bittles and J. V. Neel, “The costs of inbreeding and their impli-
cations for variations at the DNA level,” Nature Genetics, vol. 8 (1994),
pp. 117–21.
58. Ian McEwan, Cement Garden (London: Jonathan Cape, 1980); A. S. Byatt,
Angels and Insects (London: Chatto and Windus, 1992).
59. P. L. van den Berghe, “Human inbreeding avoidance: Culture in nature,”
Behavioral and Brain Sciences, vol. 6 (1983), pp. 91–102.
60. Y. Talmon, “The family in a revolutionary movement: The case of the
kibbutz in Israel,” in Comparative Family Systems, ed. M. F. Nimkoff (Boston:
Houghton Mifflin, 1965), pp. 259–86; M. E. Spiro, Children of the Kibbutz
(New York: Schocken Books, 1969), pp. 326–35, 347–50.
61. A. P. Wolf, Sexual Attraction and Childhood Association: A Chinese Brief
for Edward Westermarck (Stanford, Calif.: Stanford University Press, 1995).
62. R. Middleton, “Brother-sister and father-daughter marriage in Ancient
 Genetic Aspects of Inbreeding and Incest 59

Egypt,” American Sociological Review, vol. 27 (1962), pp. 603–11; B. D. Shaw,

“Explaining incest: brother-sister marriage in Graeco-Roman Egypt,” Man (N.S.),
vol. 27 (1992), pp. 267–99.
63. K. Hopkins, “Brother-sister marriage in Roman Egypt,” Comparative
Studies in Social History, vol. 22 (1980), pp. 303–54; W. Scheidel, “Brother-sister
and parent-child marriage outside royal families in ancient Egypt and Iran: A chal-
lenge to the sociobiological view of incest,” Ethnology and Sociobiology, vol. 17
(1996), pp. 319–40.
64. W. Scheidel, “Brother-sister marriage in Roman Egypt,” Journal of Bio-
social Science, vol. 29 (1997), 361–71.
65. L. H. Gray, “Iranian marriages,” in Encyclopedia of Religion and Ethics,
vol. 8, ed. J. Hastings (Edinburgh: Clark, 1915), pp. 455–59.
66. E. A. Hammel, C. K. McDaniel, and K. W. Wachter, “Demographic con-
sequences of incest tabus: A microsimulation analysis,” Science, vol. 205, (1979),
pp. 972–77.
67. E. Le Roy Ladurie, Montaillou: Cathars and Catholics in a French Village,
1294–1324 (London: Penguin, 1980), pp. 36, 52, 179ff.
68. M. A. Ruffer, “On the physical effects of consanguineous marriages in
the royal families of Ancient Egypt,” in Studies in the Paleopathology of Egypt,
ed. L. R. Moodie (Chicago: University of Chicago Press, 1922), pp. 322–66.
69. J. Jancar and S. J. Johnston, “Incest and mental handicap,” Journal of
Mental Deficiency Research, vol. 34 (1990), pp. 483–90.
70. M. Virkkunen, “Incest offenders and alchoholism,” Medicine, Science, and
the Law, vol. 14 (1974), pp. 124–28; A. A. Rosenfeld, “Incidence of a history of
incest among 18 female psychiatric patients,” American Journal of Psychiatry,
vol. 136 (1979), pp. 191–95.
71. B. H. Bulkley and G. M. Hutchins, “Pompe’s disease presenting as hyper-
trophic myocardiopathy with Wolff-Parkinson-White syndrome,” American Heart
Journal, vol. 96 (1978), pp. 246–52; C. Garrett and H. Tripp, “Unknown syn-
drome: Mental retardation and postaxial polydactyly, congenital absence of hair,
severe seborrhoeic dermatitis, and Perthes’ disease of the hip,” Journal of Medical
Genetics, vol. 25 (1988), pp. 270–72; Eva Seemanová, “A study of incestuous
matings,” Human Heredity, vol. 21 (1971), pp. 108–28; J. Delanghe, H. Vlam-
inck, D. Bernard, E. Robbereacht, M. Paret, M. De Buyzere, et al., “Absence
of serum alanine aminotransferase activity,” Clinical Chemistry, vol. 43 (1997),
pp. 1665–67.
72. N. Lukianowicz, “Incest. I Paternal incest. II Other types of incest,”
British Journal of Psychiatry, vol. 120 (1972), pp. 301–13.
73. Morton S. Adams and James V. Neel, “Children of incest,” Pediatrics,
vol. 40 (1967), pp. 55–62; C. O. Carter, “Risk to offspring of incest,” The Lancet,
vol. 289 (1967), p. 436; P. A. Baird and B. McGillivray, “Children of incest,”
Journal of Pediatrics, vol. 101 (1982), pp. 854–57.
74. Seemanová, “A study of incestuous matings.”
75. Carter, “Risk to offspring.”
76. Adams and Neel, “Children of incest.”
77. Seemanová, “A study of incestuous matings.”
60 bittles

78. A. H. Bittles, “Incest re-assessed,” Nature, vol. 280 (1979), p. 107.

79. Seemanová, “A study of incestuous matings.”
80. J. M. Tanner, Growth at Adolescence (Oxford: Blackwell, 1962).
81. J. B. Coates, “Obstetrics in the very young adolescent,” American Journal
of Obstetrics and Gynecology, vol. 108 (1970), pp. 68–72.
82. A. M. Fraser, J. E. Brockert, and R. H. Ward, “Association of young
maternal age with adverse reproductive outcomes,” New England Journal of
Medicine, vol. 332 (1995), pp. 113–17.
83. R. L. Goldenberg and L. V. Klerman, “Adolescent pregnancy—another
look,” New England Journal of Medicine, vol. 332 (1995), pp. 1161–62.
84. D. H. Carr, “Chromosomes and abortion,” Advances in Human Genetics,
vol. 2 (1971), pp. 210–57.
85. C. H. Hendricks, “Congenital malformations: Analysis of the 1953 Ohio
records,” Obstetrics and Gynecology, vol. 6 (1955), pp. 592–98.
86. J. Yerushalmy, “Neonatal mortality by order of birth and age of parents,”
American Journal of Hygiene, vol. 28 (1938), pp. 244–70; R. R. Gordon and
R. Sunderland, “Maternal age, illegitimacy, and postneonatal mortality,” British
Medical Journal, 297 (1988), p. 774.
87. Adams and Neel, “Children of incest;” Seemanová, “A study of incestuous
matings.”
88. Bittles and Neel, “The costs of inbreeding.”
89. Ibid.
90. William J. Schull and James V. Neel, The Effects of Inbreeding on Japanese
Children (New York: Harper and Row, 1965).
91. Ibid.
92. A. H. Bittles, “Consanguinity and its relevance to clinical genetics,”
Clinical Genetics, vol. 60 (2001), pp. 89–98; A. H. Bittles, “Consanguineous
marriages and childhood health,” Developmental Medicine, vol. 45 (2003),
pp. 571–76.
93. J. G. Hall, “Children of incest: When to suspect and how to evaluate,”
American Journal of Diseases of Children, vol. 132 (1978), p. 1045 (letter); Baird
and McGillivray, “Children of incest.”
94. J. H. Edwards, “Evidence of incest based on homozygosity,” Annals of
Human Genetics, vol. 52 (1988), pp. 351–53; R. E. Wenk, F. A. Chiafari, and
T. Houtz, “Incest diagnosis by comparison of alleles of mother and offspring at
highly heterozygous loci,” Transfusion, vol. 34 (1994), pp. 172–75.
95. R. A. Wells, B. Wonke, and S. L. Thein, “Prediction of consanguinity
using human DNA fingerprints,” Journal of Medical Genetics, vol. 25 (1988),
pp. 660–62.
96. S. Bundey, “The child of an incestuous union,” in Medical Aspects of
Adoption and Foster Care, ed. S. Wolkind (London: Spastics International
Medical Publications, 1979), pp. 36–41.
97. Baird and McGillivray, “Children of incest.”
3 Inbreeding Avoidance in Primates

Anne Pusey

Anthropologists interested in biological explanations for incest avoid-

ance and incest taboos have long looked to primate and other animal stud-
ies for support for their theories.1 If primates, our closest living ancestors,
avoid close inbreeding, this finding provides evidence that this behavior pre-
ceded human cultural practices. The study of the mechanisms involved in
inbreeding avoidance can also be augmented by experimentation in animals
in a way that is not possible in humans. Edward Westermarck believed that
incest avoidance is a naturally selected adaptation to avoid the deleterious
consequences of inbreeding depression, and he proposed that the mecha-
nism was that individuals raised together in infancy (usually kin) develop
an aversion to sexual relationships with each other as adults.2 Developing
this idea, Arthur Wolf (Chapter 4) and Mark Erickson (Chapter 9) have
proposed that it is in particular the caretaking, attachment relationships be-
tween these individuals that preclude sexual behavior later on. On the other
hand, William Durham (Chapter 7) leans toward Roger Burton’s recogni-
tion and attribution theory, which proposes that people avoid incest be-
cause they are aware of its deleterious consequences. In this chapter, I re-
view evidence for inbreeding avoidance in primates and discuss the extent
to which recent research on primates enables us to distinguish between the
various mechanisms that have been proposed to underlie this behavior.

Inbreeding Depression

Alan Bittles (Chapter 2) discusses the reasons for inbreeding depres-

sion and reviews evidence for this in human populations. It is very difficult
to gather evidence of inbreeding depression in wild populations of animals
because, on the whole, close relatives never breed, but in those rare cases in
62 pusey

which they do, inbreeding depression is often severe.3 Data on inbreeding

in mammals kept in captivity are more extensive and show that one com-
ponent of fitness, juvenile survival, is depressed on average by about 30
percent in the offspring of first-order relatives (parent-offspring or full-sib
pairs).4 Moreover, there is evidence from a variety of species that inbreed-
ing depression tends to be more severe under natural conditions.5

Mechanisms of Inbreeding Avoidance in Primates

There are now many long-term studies of individually identified pri-

mates whose pedigrees are known, and all these show that breeding be-
tween first-order relatives almost never occurs.6 This is for at least two rea-
sons. First, patterns of dispersal of individuals from their natal groups are
such that close relatives rarely reside in the same group or area as adults.
Second, when close relatives do reside in the same group as adults, they
usually do not mate.

Sex-Biased Dispersal

In a twenty-seven-year study of the olive baboons at Gombe Na-

tional Park, Tanzania, eighty-nine of eighty-nine males that reached adult-
hood dispersed from their natal troops and joined other troops as young
adults, while all females remained in their natal group.7 Similar dispersal of
males and not females also occurs in a variety of other species of primates,
including many species of macaques, capuchin monkeys, and some lemurs.8
In a few primate species, including chimpanzees, spider monkeys,9 muri-
quis,10 and some populations of red colobus, we find the opposite pattern,
in which males remain in their natal groups while most or all females join
other groups before breeding.11 This striking pattern of sex-biased disper-
sal is also widespread in other animal groups.12 Because such dispersal ef-
fectively separates all close relatives of the opposite sex, with the occasional
exception of individuals of the dispersing sex who are still in the same group
when their opposite-sexed philopatric offspring reach maturity, most biol-
ogists believe that this pattern has evolved at least partly as an inbreeding-
avoidance mechanism. Such complete dispersal of one sex is difficult to ex-
plain by other factors such as intrasexual competition for resources.
Proximate causes of this dispersal also lend credence to the idea that it oc-
curs to avoid inbreeding. In several species, members of the dispersing sex
transfer to other groups during the mating season or when sexually receptive
(if female), and they show greater sexual attraction to mates in other groups
than they do to opposite-sexed members of their own groups.13 Against the
 Inbreeding Avoidance in Primates 63

alternative hypothesis that competition completely accounts for these disper-

sal patterns,14 immigrants in several species are not driven out of their natal
group and actually receive more aggression in the new group.15

Avoidance of Sexual Behavior Between Close

Relatives Residing Together

Even though dispersal greatly reduces the chances of close inbreed-

ing, there are sometimes circumstances in which close relatives do remain
for at least a time in proximity as adults. Male baboons start mating and
are often fertile before they leave their natal troop, and in provisioned and
captive groups of macaques, males sometimes remain in their natal group
as adults. Also, in chimpanzees, some females are still alive when their sons
become mature, and in both gorillas and chimpanzees, young females
sometimes start mating in their natal group when their fathers are still alive
and adult brothers are present. We now know from long-term studies that
although these individuals are sexually active, they often avoid mating with
relatives. Table 3.1 lists long-term studies where pedigrees are known and
in which the frequency of mating between close relatives has been investi-
gated. In most species in which adult sons are observed with maternal rela-
tives, sexual activity between these males and their mothers is negligible,
and it is also greatly reduced between maternal siblings. Significant reduc-
tion of sexual activity has also been detected among individuals as distantly
related as maternal first cousins in some species where these relationships
are known. In some cases—ring-tailed lemurs,16 Barbary macaques,17 and
rhesus monkeys18—paternity analysis has confirmed that matrilineally in-
bred offspring are seldom produced.
Avoidance of sex with paternal relatives is much less well studied, mostly
because in most primate species females mate with several males in the
group, so paternal relationships cannot be determined by observational
data alone. However, in gorillas, the silverback is often the only breeding
male, and very little mating is seen between fathers and daughters. In ba-
boons, Craig Packer found that sexual activity was reduced between fe-
males and males that had been present in the troop when the female was
born (and were thus potential fathers),19 and in chimpanzees some females
avoided males old enough to be their fathers,20 while others did not.21
Recently, analysis of DNA has allowed precise identification of paternal
relatives in some species. A detailed study of a large population of captive
Barbary macaques living in an outdoor enclosure found no inhibition of
mating between paternal siblings or daughters and their fathers compared to
mating with unrelated individuals.22 Similarly, in a captive group of rhesus
 table 3.1
Avoidance of Sexual Activity with Relatives
Other
Maternal Maternal Paternal
Species Mother Siblings Relatives Father Siblings

Ring-tailed lemurs1 + + –
Muriquis2 +
Marmosets3 + +
Vervet monkeys4 + + +
Japanese macaques5 + + +
Rhesus macaques6 + + + + –
Barbary macaques7 + + + – –
Stumptail macaques8 + +
Olive baboons9 + + +
Yellow baboons10 + + + +
Chimpanzees11 + + +/-
Gorillas12 +

note: + signifies that inhibition of mating occurs, - that it does not; +/- means that inhibition occurs
between some pairs but not others. Cells remain blank if the frequency of mating in this category has
not been measured in this species.
sources:
1
M. E. Pereira and M. L. Weiss, “Female mate choice, male migration, and the threat of infanticide
in ringtailed lemurs,” Behavioral Ecology and Sociobiology, vol. 28 (1991), pp. 141–52
2
K. B. Strier, “Mate preferences of wild muriqui monkeys (Barchyteles arachnoides): Reproductive
and social correlates,” Folia Primatologica, vol. 68 (1997), pp. 120–33.
3
D. H. Abbott, “Behavioral and physiological suppression of fertility in subordinate marmoset
monkeys,” American Journal of Primatology, vol. 6 (1984), pp. 169–86; J. V. Baker, D. H. Abbott,
and W. Saltzman, “Social determinants of reproductive failure in male common marmosets housed
with their natal family, Animal Behaviour, vol. 58 (1999), pp. 501–13; J. A. French, “Proximate
regulation of singular breeding in Callitricid primates,” in Cooperative Breeding in Mammals,
ed. N. G. Solomon and J. F. French (Cambridge: Cambridge University Press, 1997), pp. 34–75.
4
C. A. Bramblett, “Incest avoidance in socially living veret monkeys,” American Journal of Primatol-
ogy, vol. 63 (1983), p. 176; Dorothy L. Cheney and Robert M. Seyfarth, personal communication.
5
K. Tokuda, “A study on the sexual behavior in the Japanese monkey troop,” Primates, vol. 3
(1961–62), pp. 1–40; T. Enomoto, “The sexual behavior of Japanese monkeys,” Journal of Human
Evolution, vol. 3 (1974), pp. 351–72; Y. Takahata, “The socio-sexual behavior of Japanese monkeys,”
Zeitschrift fur Tierpsychogie, vol. 59 (1982), pp. 89–108.
6
D. S. Sade, “Inhibition of mother-son mating among free-ranging rhesus monkeys,” Science and
Psychoanalysis, vol. 12 (1968), pp. 18–38; D. G. Smith, “Inbreeding in three captive groups of rhesus
macaques,” American Journal of Physical Anthropology, vol. 58 (1982), pp. 447–51; D. G. Smith,
“Avoidance of close consanguineous inbreeding in captive groups of rhesus macaques,” American
Journal of Primatology, vol. 35 (1995), pp. 31–40; B. Chapais, “Male dominance and reproductive
activity in rhesus monkeys,” in Primate Social Relationships, ed. R. A. Hinde (Oxford: Blackwell,
1983), pp. 267–71; J. H. Manson and S. E. Perry, “Inbreeding avoidance in rhesus macaques: Whose
choice? American Journal of Physical Anthropology, vol. 90 (1993), pp. 335–44.
7
A. Paul and J. Kuester, “Intergroup transfer and incest avoidance in semi-free-ranging Barbary
macaques (Macaca sylvanus) at Salem (FRG),” American Journal of Primatology, vol. 8 (1985),
pp. 317–22; J. Kuester, A. Paul, and J. Arnemann, “Kinship, familiarity, and mating avoidance in
Barbary macaques, Macaca sylvanus,” Animal Behaviour, vol. 48 (1994), pp. 1183–94.
8
R. Daniel Murray and E. O. Smith, “The role of dominance and intrafamilial bonding in the avoid-
ance of close inbreeding,” Journal of Human Evolution, vol. 12 (1983), pp. 481–86.
9
Craig Packer, “Inter-troop transfer and inbreeding avoidance in Papio anubis,” Animal Behaviour,
vol. 27 (1979), pp. 1–36; L. M. Scott, “Reproductive behavior in adolescent female baboons (Papio
anubis),” in Female Primates: Studies by Women Primatologists, ed. M. Small (New York: Alan Liss,
1984), pp. 77–100; B. B. Smuts, Sex and Friendship in Baboons (New York: Aldine, 1985).
10
S. C. Alberts and J. Altmann, “Balancing costs and opportunities: Dispersal in male baboons,”
American Naturalist, vol. 145 (1995), pp. 179–306; S. C. Alberts, “Paternal kin discrimination in wild
baboons,” Proceedings of the Royal Society, London, Series B, vol. 266 (1999), pp. 1501–6.
11
C. E. G. Tutin, “Sexual behaviour and mating patterns in a community of wild chimpanzees
(Pan troglodytes schweinfurthii). Ph.D. thesis, 1975. Zoology. Edinburgh, University of Edinburgh;
A. E. Pusey, “Inbreeding avoidance in chimpanzees,” Animal Behaviour, vol. 28 (1980), p. 543;
J. Goodall, The Chimpanzees of Gombe (Cambridge: Harvard University Press, 1986).
12
K. J. Stewart and A. H. Harcourt, “Gorillas: Variation in female relationships,” in Primate Societies,
ed. B. B. Smuts, D. L. Cheney, R. M. Seyfarth, T. T. Struhsaker, and R. W. Wrangham (Chicago: Uni-
versity of Chicago Press, 1987), pp. 358–69.
 Inbreeding Avoidance in Primates 65

macaques, David G. Smith found that, in general, inbreeding between patri-

lineal relatives was not different from expected were mating random.23
However, father-daughter inbreeding was less frequent than expected. In
yellow baboons, Susan Alberts found that although consorting between pa-
ternal siblings occurred as often as between nonkin, levels of sexual activity
and affiliative behavior were lower during consortships between paternal
siblings.24 Finally, in chimpanzees, two females mated as often as expected
with their fathers, while one avoided her father at high rates.25

Mechanisms of Kin Recognition

Biologists have proposed two ways that individuals might recognize

kin.26 First, they may use phenotype matching, in which the individual
matches cues from other individuals, such as odor, to cues from themselves
or their relatives and uses the degree of similarity to determine the degree
of kinship. This mechanism has been demonstrated in rodents. For exam-
ple, inbred mice that differ only in genes at the major histocompatibility
complex (MHC) prefer to mate with individuals that differ from them-
selves in MHC haplotype. MHC genes influence odors in the urine, and
this is the cue that the mice match.27 Cross-fostering experiments show that
mice actually match the odor to odors of their parents and choose mates
that differ from their parents. Evidence of phenotype matching has also
been found in several other species of rodents, crickets, and tadpoles.28
The second mechanism, often labeled association or familiarity, is when
individuals treat as kin others with whom they had a close association in
immaturity. This is the mechanism that Westermarck first proposed to ex-
plain incest avoidance.29 Association has been shown to be important in
mate choice in a variety of species of mammals, and experiments in which
degree of familiarity is manipulated show that it often has a stronger effect
than real kinship.30 For example, cross-fostering experiments in various
species of rodents in which relatives are raised apart and nonrelatives are
raised together show that individuals raised apart prefer each other as mates
over those reared together, regardless of relatedness.31
Among primates, the patterns of inbreeding avoidance described above
are consistent with familiarity being the primary mechanism of kin recogni-
tion. In all primates, mother-offspring bonds are close and long-lasting, with
the result that maternal siblings are often reared in close proximity to one
another. In species that show female philopatry, females usually have closer
and more affiliative relationships with their close adult female kin. Therefore
the fact that mating is strongly inhibited among all kinds of maternal kin in
several species is consistent with the Westermarck effect.32
In contrast to the consistency of the mother-offspring bond across pri-
66 pusey

mate species, close bonds between fathers and daughters occur only in
some species. In monogamous species like marmosets, or species like goril-
las in which one male has exclusive reproductive access to group females
for periods beyond their daughters’ age of maturity, father-daughter bonds
are strong. In these, father-daughter mating is strongly inhibited (Table
3.1). In other species such as baboons and wild macaques, males do not of-
ten remain in the group until their daughters mature, and when they do,
there can be great variation in the strength of their relationships with their
potential offspring in their first few years of life.
In Barbary macaques, males form close caretaking relationships with
several infants in the troop, during which they carry them around and pro-
tect them. On the basis of the Westermarck effect, we might expect females
that were cared for by some males and not others to avoid males that cared
for them regardless of kinship, and this is indeed what is observed in Bar-
bary macaques.33 While there was no evidence that daughters avoided their
fathers (as determined from DNA), mating between females and their pre-
vious caretakers was inhibited, and the strength of the inhibition depended
on the length of the caretaking relationship. Packer’s finding, that females
were more likely to avoid consorting and mating with males that had been
present at the time of their birth than males that arrived in the troop after
their birth, is also consistent with the Westermarck effect.34 In chimpan-
zees, females usually leave their natal community, but some stay. Chim-
panzee males do not obviously form close relationships with particular in-
fants, although they are frequently tolerant and friendly to most infants.35
At Gombe, of three females that mated in their community while their fa-
thers were still alive, two did not obviously avoid mating with their fathers,
but one did.36 Future research should determine whether there were differ-
ences in early association between the male and female in these cases.
Data on paternal siblings are even scarcer than on fathers and daugh-
ters. While the studies of Barbary macaques and rhesus macaques found no
significant avoidance of sexual activity between patrilineal siblings,37 Al-
berts’s baboon study found subtle differences, some of which are consistent
with the Westermarck effect.38 Alberts compared paternal siblings and
nonkin on two measures: (1) the rates of consorting activity, when a male
maintains a close, guarding relationship with a female during her period of
receptivity, and (2) the level of cohesiveness including sexual behavior
within consorting pairs. She found that, overall, paternal siblings consorted
as often as nonkin did but had lower levels of cohesiveness. She reasoned
that one way that paternal siblings might recognize each other is through
association. In this population, because the alpha male fathers most of the
offspring conceived during his tenure, age cohorts are likely to be paternal
kin. Age cohorts also spend more time together than individuals born more
 Inbreeding Avoidance in Primates 67

than two years apart. Alberts found that individuals within the same age
cohort consorted at lower rates than others, regardless of kinship, and also
showed lower cohesiveness. She found that on the measure of consortship
cohesiveness the three pairs scoring lowest were from the same age cohort
and that one pair of baboons that was unrelated but born within two years
of each other scored the lowest among unrelated pairs and the highest
among age-cohort members. She concluded on the basis of this and the fact
that cohesiveness is low between paternal siblings that are not in the same
age cohort, that although association is clearly important, phenotype
matching may also play a role in the recognition of paternal siblings.
Although systematic rearing experiments have not been done in primates
to distinguish between the importance of relatedness per se and familiarity
in mating avoidance, some anecdotal evidence supports the importance of
the Westermarck effect. Unrelated chimpanzees raised together mated very
little with each other and preferred unfamiliar individuals as mates.39 Also,
a male Japanese macaque avoided mating with an unrelated female who
had been a close associate of his mother, although he mated readily with un-
familiar females of the same age.40

The Ontogeny of Avoidance of Sexual

Activity Between Relatives

In many primate species, elements of sexual behavior appear during

immaturity. Therefore it is of interest to examine the extent to which such
behavior is directed toward relatives compared to nonrelatives and when
inhibition between relatives first appears. Data are still very incomplete on
this point. Young male macaques,41 baboons,42 and chimpanzees43 get penile
erections in the first year of life, and as infants and juveniles they often
thrust on other immature individuals during play or on others including
adults during play and other social contact. In species in which females ex-
hibit sexual swellings, immature males sometimes mount, achieve intromis-
sion, and thrust on swollen, semiswollen, and sometimes unswollen females
in a similar way to adult males several years before they are capable of ejac-
ulation. Although immature females in these species may be mounted and
thrust upon by immature males, no intromission appears to occur until the
females reach adolescence and exhibit sexual swellings or other forms of
sexual receptivity.
How much of this sexual behavior by immature males is directed at rel-
atives? Immature sons have been reported to copulate with their mothers in
several species. One study found that among captive Japanese macaques,
infant males copulated with their mothers, but that juvenile and adolescent
68 pusey

males rarely did so and never achieved ejaculation.44 Male rhesus monkeys
reach puberty at the age of about four years.45 One study found that most
males aged three to five years showed no copulatory behavior, but those
that did mated almost exclusively with their mothers and maternal sis-
ters.46 However, they did not show the typical consorting behavior in which
a male follows the female, grooms her, and has repeated mount series.
Males older than five years that stayed in their natal group mated with
other females, but not with their mother or sisters. Two infant male ba-
boons were seen to copulate with their mothers,47 and a juvenile male was
observed to copulate several times with his cycling sister, but mating be-
tween postpubertal males and their female relatives was very rare.48 In
Packer’s detailed study of the mating behavior of postpubertal males, three
males had mothers or sisters that were sexually receptive but never con-
sorted them.49 One male achieved a complete copulation with his pregnant
mother, and on another occasion mounted his mother without intromis-
sion when her sexual swelling had just deflated. Another male mounted his
lactating mother. In these cases the mother made aggressive vocalizations
and terminated the mount.
The most detailed information on the incidence of mating with relatives
by immature males comes from chimpanzees.50 Infant males as young as
three years old are able to copulate and thrust with intromission on swol-
len females when these present their swellings to them, but the males do
not reach puberty for many years after this. Mothers in the chimpanzees of
Gombe National Park, Tanzania, generally do not resume sexual cycles un-
til their infants are three to six years old, although one or two have re-
sumed cycles within the first year of their infant’s life. Males in this popu-
lation are not capable of ejaculation until they are nine to ten years old. A
recent study found that while males of one to two years did not copulate
with their swollen mothers, males of three to six years usually copulated
quite frequently with their mothers, accounting for about 5 to 7 percent
(range 0 to 29 percent) of their mother’s total copulations with males when
she was swollen (Fig 3.1a).51 One context in which such copulations oc-
curred was when mothers presented to their distressed infants during
weaning.52 However, many copulations with the mother by males of this
age occurred when both male and mother were calm.53 From the age of
seven years onward, most males never copulated with their mothers (Fig
3.1a). However, two of five adult males, Goblin and Frodo, did mate with
their mothers several times after the ages of fifteen when they had attained
high rank among the males. In these cases the mother usually screamed and
strenuously tried to resist her son’s advances.54 In a captive group, a five-
year-old juvenile male frequently copulated with swollen females, showing
typical male courtship to them beforehand.55 He also occasionally copu-
lated with his mother, even though she had not yet resumed sexual cycles.
 (a)
0.20

0.15
Proportion of Mother’s Copulations

0.10

0.05

0.00

1 10 20 30
Male Age

(b)
0.20

0.15
Proportion of Sister’s Copulations

0.10

0.05

0.00

1 10 20 30
Male Age

figure 3.1. Copulations Between Mothers and Sons and Sisters and Brothers
in the Gombe Chimpanzees. (a) Mean (SE) proportion of mothers’ copulations
contributed by sons of each age (N = 25 mother-son pairs, with some pairs
observed over several years). (b) Mean (SE) proportion of sisters’ copulations
contributed by brothers of each age (N = 22 sister-brother pairs, with some
pairs observed over several years). (Data from Pusey, Schumacher Stankey, and
Goodall, in preparation.)
70 pusey

These copulations generally took place when he appeared tired and de-
pressed, and they were unusual in form. He showed no courtship, the
mother did not stand or crouch for her son, and his intromission and
thrusting were much more slowly paced and of longer duration than usual.
If she prevented these attempts by covering her genitalia, the male “hoo
whimpered” and even threw temper tantrums.56
At Gombe, males of three to four years mated with their mature sisters,
accounting for about 4 percent of their sisters’ total copulations, but be-
tween the ages of five and fifteen years, males rarely mated with their sis-
ters (Fig 3.1b). In those cases in which females remained in their natal
group as adults, some brothers aged sixteen to thirty-three years mated
with their sisters at quite high rates, although the sisters often resisted,
while other brothers never did so (Fig 3.1b).57
Taken as a whole, the data from all these species suggest that males mate
with their mothers and, sometimes, maternal sisters with impunity as in-
fants, but that inhibition of such activity sets in before or at puberty. More
detailed observations are required to determine whether this inhibition is due
to an intrinsic change in the male or is triggered by an increase in resistance
from the female. Jane Goodall described how one female chimpanzee al-
lowed her three-year-old infant brother to copulate while she was getting
adolescent swellings but prevented him from doing so once she began mat-
ing with adult males.58 However, he was observed copulating with her sev-
eral times after this.59
As well as sexual behavior, infant and juvenile primates exhibit most
other patterns of adult behavior, including aggressive behavior. Although
these patterns are sometimes performed in an adultlike context, they are of-
ten performed during play.60 Perhaps the sexual behavior of immature indi-
viduals is best regarded as play and practice, and the tolerance of female
relatives during sexual and other social interactions with these males as in-
vestment in improving the social skills of their kin. In all species, copula-
tion among maternal relatives virtually ceases before the male is fully fer-
tile, so the female is at little risk of inbreeding.

The Attachment Theory

How do the primate data speak to the idea that it is attachment and
caretaking relationships that preclude sexual behavior in adulthood, rather
than other kinds of familiarity? Certainly the relationship of offspring and
their mothers is one of attachment and caretaking. This is also likely to be
true of relationships between infants and their older siblings. In most pri-
mates, females show many caretaking behaviors toward their young sib-
 Inbreeding Avoidance in Primates 71

lings, and in at least some, males do so too. In gorillas and some species liv-
ing in multimale groups, males direct caretaking behavior to infants, and
sexual behavior is least common among pairs that were involved in such
relationships.
The inhibition of sexual behavior between age cohort mates in baboons
is less easily explained by this idea.61 These individuals are close to the
same age and are likely to participate in playful rather than caretaking re-
lationships during immaturity. Finally, the unusual case of hamadryas ba-
boons raises some problems for the idea that caretaking behavior precludes
later sexual behavior. Hamadryas baboons form one-male units within
larger bands in which an adult male is closely bonded to several adult fe-
males with whom he mates.62 One way in which new units form is for a
young adult male to coax an infant female away from her mother and
eventually keep her as a mate.63 During this gradual process the male car-
ries and cares for the infant. At first he directs no sexual behavior toward
her, but later he may attempt to mate. Females in this species have sexual
swellings, and males have been observed to try to mate with females before
they have started swelling. In several cases, such females have produced
precocious sexual swellings at earlier ages than females that remain with
their natal unit, and begin sexual behavior earlier. The only way to recon-
cile these observations with the theory that it is attachment in immaturity
that precludes sexual behavior later on would be to propose that the sensi-
tive period for this to be effective occurs very early in infancy and is al-
ready past for the female by the time a male starts to adopt her.

Conclusion

Nonhuman primates provide abundant evidence for an inhibition of

sexual behavior among closely related adults. This finding is consistent
with the idea that inbreeding avoidance behavior is a naturally selected be-
havior that was already present among animals before humans evolved.
The primate data support Westermarck’s theory that familiarity during im-
maturity is a major reason for this avoidance. Mating among adults is most
inhibited among maternal relatives in species in which these have close as-
sociations. The extent to which mating is inhibited among close paternal
relatives is more variable and appears to depend largely, though perhaps
not completely, on the closeness of association during immaturity. It is also
becoming clear that extensive sexual behavior by immature males with
close female relatives sometimes occurs before sexual maturity in several
nonhuman primate species, but it stops before the risk of conception. More
study is required to understand the proximate causes for the waning of this
72 pusey

behavior. The primate data are less clear about the kind of familiarity in
immaturity that is necessary to prevent mating among adults. Although
some relationships, such as those between mother and son, and between
older and younger siblings, fit the idea that attachment and caretaking re-
lationships are important, inhibition of sexual behavior between peers of
the same age in baboons does not.

notes
1. Pierre L. van den Berghe, “Human inbreeding avoidance,” Behavioral and
Brain Sciences, vol. 6 (1982), pp. 91–124.
2. Edward Westermarck, The History of Human Marriage (London: Macmil-
lan, 1891).
3. Anne Pusey and Marisa Wolf, “Inbreeding avoidance in animals, Trends in
Ecology and Evolution, vol. 11 (1996), pp. 201–6.
4. K. Ralls, J. D. Ballou, and A. Templeton, “Estimates of lethal equivalents
and the costs of inbreeding in mammals,” Conservation Biology, vol. 2 (1988),
pp. 185–93.
5. Pusey and Wolf, “Inbreeding avoidance in animals.”
6. Anne Pusey, “Mechanisms of inbreeding avoidance in nonhuman primates,”
in Pedophilia: Biosocial Dimensions, ed. J. R. Feierman (New York: Springer-
Verlag, 1990), pp. 201–20.
7. Craig Packer, D. A. Collins, A. Sindimwo, and Jane Goodall, “Reproductive
constraints on aggressive competition in female baboons,” Nature, vol. 373
(1995), pp. 60–63.
8. Anne Pusey and Craig Packer, “Dispersal and philopatry,” in Primate
Societies, ed. B. B. Smuts, D. L. Cheney, R. M. Seyfarth, T. T. Struhsaker,
and R. W. Wrangham (Chicago: University of Chicago Press, 1987),
pp. 250–66.
9. M. M. Symington, “Sex ratio and maternal rank in wild spider monkeys:
When daughters disperse,” Behavioral Ecology and Sociobiology, vol. 20 (1987),
pp. 333–35.
10. K. B. Strier, “Mate preferences of wild muriqui monkeys (Barchyteles
arachnoides): Reproductive and social correlates,” Folia Primatologica, vol. 68
(1997), pp. 120–33.
11. Pusey and Packer, “Dispersal and philopatry.”
12. Paul J. Greenwood, “Mating systems, philopatry, and dispersal in birds
and mammals,” Animal Behaviour, vol. 28 (1980), pp. 1140–62.
13. Pusey and Packer, “Dispersal and philopatry.”
14. Jim Moore and Rauf Ali, “Are dispersal and inbreeding related?” Animal
Behaviour, vol. 32 (1984), pp. 94–112.
15. Craig Packer, “Dispersal and inbreeding avoidance,” Animal Behaviour,
vol. 33 (1985), pp. 676–78; Anne Pusey, “Sex-biased dispersal and inbreeding
 Inbreeding Avoidance in Primates 73

avoidance in birds and mammals,” Trends in Ecology and Evolution, vol. 2

(1987), pp. 295–99.
16. M. E. Pereira and M. L. Weiss, “Female mate choice, male migration, and
the threat of infanticide in ringtailed lemurs,” Behavioral Ecology and Sociobiol-
ogy, vol. 28 (1991), pp. 141–52.
17. Jutta Kuester, Andreas Paul, and J. Arnemann, “Kinship, familiarity, and
mating avoidance in Barbary macaques, Macaca sylvanus,” Animal Behaviour,
vol. 48 (1994), pp. 1183–94.
18. David G. Smith, “Avoidance of close consanguineous inbreeding in captive
groups of rhesus monkeys,” American Journal of Primatology, vol. 35 (1995),
pp. 31–40.
19. Craig Packer, “Inter-troop transfer and inbreeding avoidance in Papio
anubis,” Animal Behaviour, vol. 27 (1979), pp. 1–36.
20. Caroline E. G. Tutin, “Sexual behaviour and mating patterns in a commu-
nity of wild chimpanzees (Pan troglodytes schweinfurthii). Ph.D. thesis, 1975.
Zoology. Edinburgh, University of Edinburgh; Anne Pusey, “Inbreeding avoidance
in chimpanzees,” Animal Behaviour, vol. 28 (1980), p. 543.
21. Jane Goodall, The Chimpanzees of Gombe (Cambridge: Harvard Univer-
sity Press, 1986).
22. Kuester, Paul, and Arnemann, “Kinship, familiarity, and mating avoidance.”
23. Smith, “Avoidance of close consanguineous inbreeding.”
24. Susan Alberts, “Paternal kin discrimination in wild baboons,” Proceedings
of the Royal Society, London, Series B., vol. 266 (1999), pp. 1501–6.
25. Anne Pusey, J. Schumacher Stankey, and Jane Goodall, “Age changes in
frequency of mating between consanguineous chimpanzees,” in preparation.
26. W. G. Holmes and P. W. Sherman, “Kin recognition in animals,” American
Scientist, vol. 71 (1983), pp. 46–55.
27. J. L. Brown and A. Eklund, “Kin recognition and the major histocompati-
bility complex: An integrative review,” American Naturalist, vol. 143 (1994),
pp. 435–61.
28. David J. C. Fletcher and Charles D. Michener, eds., Kin Recognition in
Animals (New York: Wiley, 1987); P. G. Hepper, Kin Recognition (Cambridge:
Cambridge University Press, 1991); P. W. Sherman, H. K. Reeve, and D. W.
Pfennig, “Recognition systems,” in Behavioural Ecology, ed. J. R. Krebs and
N. B. Davies (Oxford: Blackwell, 1997).
29. Westermarck, The History of Human Marriage.
30. Pusey and Wolf, “Inbreeding avoidance in animals.”
31. D. A. Dewsbury, “Kin discrimination and reproductive behavior in muroid
rodents,” Behavior Genetics, vol. 18 (1988), pp. 525–36.
32. Jeffrey R. Walters, “Kin recognition in nonhuman primates,” in Kin
Recognition in Animals, ed. David J. C. Fletcher and Charles D. Michener (New
York: Wiley, 1987), pp. 359–93.
33. Kuester, Paul, and Arnemann, “Kinship, familiarity, and mating avoidance.”
34. Packer, “Inter-troop transfer and inbreeding avoidance.”
35. Jane Goodall, The Chimpanzees of Gombe.
74 pusey

36. Pusey, Schumacher Stankey, and Goodall, “Age changes in frequency of

mating.”
37. Kuester, Paul, and Arnemann, “Kinship, familiarity, and mating avoid-
ance”; Smith, “Avoidance of close consanguineous inbreeding.”
38. Alberts, “Paternal kin discrimination in wild baboons.”
39. C. L. Coe, A. C. Connolly, H. C. Kremer, and S. Levine, “Reproductive
development and behavior of captive female chimpanzees,” Primates, vol. 20
(1979), pp. 571–82.
40. N. Itoigawa, K. Negamaya, and K. Kondo, “Experimental study of sexual
behavior between mother and son in Japanese monkeys,” Primates, vol. 22
(1981), pp. 494–502.
41. J. P. Hanby and C. E. Brown, “The development of sociosexual behaviors
in Japanese macaques, Macaca fuscata,” Behaviour, vol. 49 (1974), pp. 152–96.
42. L. T. Nash, “The development of the mother-infant relationship in wild
baboons (Papio anubis),” Animal Behaviour, vol. 26 (1978), pp. 749–59.
43. Jane Goodall, “The behaviour of free-living chimpanzees of the Gombe
Stream Reserve,” Animal Behavior Monographs, vol. 1, pp. 161–311; Jane
Goodall, The Chimpanzees of Gombe; Anne Pusey, “Behavioural changes at
adolescence in chimpanzees,” Behaviour, vol. 115 (1990), p. 203; E. S. Savage
and C. Malik, “Play and socio-sexual behaviour in a captive chimpanzee (Pan
Troglodytes),” Behaviour, vol. 60 (1977), pp. 179–94.
44. Hanby and Brown, “The development of sociosexual behaviors.”
45. J. Colvin, “Influences of the social situation on male emigration,” in
Primate Social Relationships, ed. R. A. Hinde (Sunderland, Mass.: Sinauer, 1983),
pp. 160–71.
46. Elizabeth A. Missakian, “Genealogical mating activity in free-ranging
groups of rhesus monkeys (Macaca mulatta) on Cayo Santiago,” Behaviour,
vol. 45 (1973), pp. 224–40.
47. Nash, “Development of the mother-infant relationship.”
48. Packer, “Inter-troop transfer.”
49. Ibid.
50. Savage and Malik, “Play and socio-sexual behaviour.”
51. Pusey, Schumacher Stankey, and Goodall, “Age changes in frequency of
mating.”
52. C. B. Clark, “A preliminary report on weaning among chimpanzees
of Gombe National Park, Tanzania,” in Primate Bio-social Development,
ed. S. Chevalier-Skolnikoff and F. E. Poirier (New York: Garland, 1977),
pp. 235–60.
53. Pusey, Schumacher Stankey, and Goodall, “Age changes in frequency
of mating.”
54. Ibid.
55. Savage and Malik, “Play and socio-sexual behaviour.”
56. Ibid.
57. Pusey, Schumacher Stankey, and Goodall, “Age changes in frequency
of mating.”
 Inbreeding Avoidance in Primates 75

58. Goodall, “The behaviour of free-living chimpanzees.”

59. Pusey, Schumacher Stankey, and Goodall, “Age changes in frequency of
mating.”
60. Jeffrey R. Walters, “Transition to adulthood,” in Primate Societies,
ed. B. B. Smuts, D. L. Cheney, R. M. Seyfarth, T. T. Struhsaker, and R. W.
Wrangham (Chicago: University of Chicago Press, 1987), pp. 358–69.
61. Alberts, “Paternal kin discrimination.”
62. Hans Kummer, Social Organization of the Hamadryas Baboons (Chicago:
University of Chicago Press, 1968).
63. H. Sigg, A. Stobla, J. J. Abegglen, and V. Dasser, “Life history of
hamadryas baboons,” Primates, vol. 23 (1982), pp. 473–87; J. J. Abegglen,
On Socialization in Hamadryas Baboons (Cranbury, New Jersey, 1984).
4 Explaining the Westermarck Effect
o r , w h at d i d n at u r a l s e l e c t i o n s e l e c t f o r ?

Arthur P. Wolf

You are thinking of your sons—but do not you know that

of all things upon earth, that is the least likely to happen;
brought up, as they would be, always together like brothers
and sisters? It is morally impossible. I never knew an
instance of it. It is, in fact, the only sure way of providing
against the connection.
—Jane Austen, Mansfield Park

This is the argument with which Mrs. Norris persuaded her brother-
in-law, Sir Thomas, to invite his niece, Fanny Price, to live at Mansfield
Park. Mrs. Norris was being meddlesome, as usual, and her argument was
self-serving. But was she right? Is “that” the thing least likely to happen
when male and female children are brought up together? Is early associa-
tion a good way, if not necessarily the only good way, of ensuring against a
later “connection”?
Until the mid-twentieth century, custom in most of China (as well, I
might add, as in most of Korea) gave families a choice of how to acquire
wives for their sons. One way was to wait until the son was fully grown and
then arrange a marriage with a young adult who would come to live with
her husband and his parents. In this case, which I call “major marriage,” the
young couple did not ordinarily meet until the day of their wedding. The
alternative was what I call “minor marriage.”1 In this case the family
“adopted” (or bought) a girl and raised her as a daughter-in-law. Many of
these girls were taken in as infants and nursed by their future mother-in-law.
It was common practice for a woman who bore a son and then a daughter
to give her daughter away and raise her son’s wife in her daughter’s place.2
In Taiwan these girls were called sim-pua, “little daughters-in-law.”

Much of the argument of this paper and some of the data appear in a paper entitled
“Reformulating (Yet Again) the Westermarck Hypothesis, or Was Dr. Ellis Right?”
 Explaining the Westermarck Effect 77

Although the great majority of girls adopted as sim-pua were taken by

couples who already had a son for the girl to marry, there were exceptions.
Prompted by the belief that adopting a girl would enhance the wife’s
chances of bearing a son—or by anxiety about finding a girl to adopt in a
tight adoption market—some couples adopted a little daughter-in-law be-
fore producing a son.3 The result was that about 5 percent of the women in
minor marriages were present to witness their husband’s birth. We will see
later that they are the key to answering an important part of the question
raised by Mrs. Norris’s meddling.
When I began field research in northern Taiwan in the late 1950s, I
found that nearly half of the women over thirty years of age had been mar-
ried in the minor fashion. It was what I saw of their lives that alerted me to
the significance of this form of marriage, but the evidence I present in this
chapter is largely drawn from household registers compiled in Taiwan by
the Japanese colonial government. In 1905 the Japanese police interviewed
every person on the island and recorded, among other things, their birth
date, their adoption date if adopted, and the form and date of their current
marriage. After that, household and village heads were required to report
to the police all vital events—births, deaths, marriages, divorces—within
ten days of their occurrence.4 My previous work shows that the great ma-
jority of this information is highly reliable. The only exceptions are events
in the lives of people who were already elderly when the police interviewed
them in 1905. In this chapter, I will avoid the problems this creates by con-
fining my analysis to the marriages of women born after 1890.
My work to date—which now spans more than forty years—has estab-
lished three points relevant to the question of whether that is unlikely when
children are brought up “always together.” The first two are based on the
reconstruction of upwards of 20,000 marriages. The first point is that when
women in minor marriages were adopted at an early age, their fertility was
40 percent lower than that of women in major marriages.5 The second point
is that when women in minor marriages were adopted early, their chances of
experiencing divorce were three times higher than those of women in major
marriages.6 My third point is based on interviews concerning extramarital
sexual relations among 551 women. My finding is that according to their
relatives and neighbors, women who married a childhood associate were
more than twice as likely to seek sexual satisfaction outside of marriage
than women who married a stranger.7
When I began publishing this evidence in the early 1960s, the great ma-
jority of social scientists were convinced that Mrs. Norris was wrong. With
very few exceptions, they accepted Sigmund Freud’s contention that “an in-
cestuous love-choice is in fact the first and regular one, and that it is only
later that any opposition is manifested towards it, which is not to be sought
78 wolf

in the psychology of the individual.”8 This attitude changed remarkably in

the following thirty years. Studies of couples reared together in Israel,
Lebanon, and Taiwan convinced all but the most conservative Freudians
(and unreconstructed social constructionists) that Mrs. Norris was right.9
The man who had listened to the Mrs. Norrises of the world reemerged as
the authority on the subject of incest avoidance. It is now generally accepted
that Edward Westermarck was right when, in 1895, he argued that “there is
a remarkable absence of erotic feelings between people living closely to-
gether from childhood.”10
The problem now is why? Why does early association inhibit sexual at-
traction? The evidence reviewed in this volume by Alan H. Bittles (Chapter
2) and Anne Pusey (Chapter 3) argues that from an evolutionary perspec-
tive, the most likely answer is that the dangers of close inbreeding selected
for something that causes us to respond to early association with an endur-
ing sexual aversion. But what is this something? What did natural selection
select for? Mrs. Norris insists that even if Fanny Price turns out “to have
the beauty of an angel,” she “will never be more to [her cousins] than a sis-
ter.”11 But why? Why would growing up together cause two young men to
ignore a beautiful young woman?
In this case, as in most cases of the kind, there is, as Karl Popper put it,
“the problem of formulating the problem.”12 I think the most promising
formulation is to ask what it is about “being always together” that matters.
In other words, what conditions are necessary to produce “the Wester-
marck effect”? Thus my strategy will be to try to answer three questions
that Westermarck neglected. The first is, what does childhood mean in his
famous hypothesis? How early do males and females have to meet to qual-
ify as “living closely together from childhood”? The second is, what does
closely mean in the phrase “living closely together”? What kind of associa-
tion is necessary to produce sexual aversion? And the third is, are males
and females equally sensitive to the effects of early association? In a review
of the fifth edition of Westermarck’s The History of Human Marriage, the
famous sexologist Havelock Ellis suggested that the inhibition is “probably
more clearly marked in the female” than in the male.13 Westermarck’s only
response was to note that “Dr. Ellis may be right.”14 But was he?
In this chapter I will pursue these questions by examining more closely
than I have previously variation in sexual attraction among minor mar-
riages. My measures of sexual attraction are general marital fertility and an
especially constructed index I call the fertility/divorce index. General mari-
tal fertility is calculated by dividing the number of births among married
women aged fifteen to forty-five by the number of years they were married
during these ages. My fertility/divorce index is general marital fertility ad-
justed to take account of the number of divorces experienced by the women
 Explaining the Westermarck Effect 79

in question. The adjustment consists of subtracting from the numerator of

the marital fertility rate five births for every divorce. The number five is ar-
bitrary. I have chosen it as a rough estimate of the number of children
women who were divorced would have borne if they had not been divorced.
The fertility/divorce index is intended to capture in one number the two
clearest quantifiable manifestations of the sexual aversion aroused by early
association. Divorce alone is an unreliable measure because it was stigma-
tized—to the point of being regarded as polluting—and therefore infre-
quent. We have to assume that the great majority of the couples affected by
early association never considered divorce as the solution to an unhappy
relationship. My reason for creating an index that takes it into account is
the possibility that the few couples who did risk divorce were those who
were the most sensitive to the inhibiting effects of early association. The
fertility/divorce index is best considered an attempt to estimate what the
fertility rate would have been if there were no divorce.
The marriages out of which my tables and graphs were constructed in-
clude all the minor marriages contracted by women born in the years 1890
to 1920 in eighty-five villages and neighborhoods (including the neighbor-
hoods that constitute two small towns). Twenty-eight of these communities
were included in the analyses presented in my 1995 book and are located in
northwestern Taiwan (in the southwestern corner of the Taipei Basin and
in the hills overlooking Hsin-chu City). The remainder are additions to my
database and are located in the Pescadores Islands and in northeastern Tai-
wan (on the coast south of Keelung City and on the Ilan Plain).
I will begin with what childhood means—or should mean—in the Wester-
marck hypothesis. This is not simply a question of when childhood ends. It
is also a question about the relative impact of the ages included. One possi-
bility is that the Westermarck effect is primarily a product of association
during a particular developmental phase. The other is that it is largely a
function of the number of years of association prior to puberty. To address
the matter, I have plotted my two indexes of sexual attraction against the age
at which women in minor marriages first met their future husband. The plot-
ted points include all the data available except for marriages in which the
wife was adopted before the husband’s birth and marriages in which the hus-
band was eight or more years older than the wife. We will see later why
these two were excluded.
Table 4.1 and Figure 4.1 display in numerical and visual form the rela-
tionship between the fertility of minor marriages and the wife’s age at adop-
tion. It is important to note that the data are heavily concentrated at age
zero and diminish steadily as age rises. This is because Taiwanese women
preferred to adopt their son’s wife as early as possible. The result for us is
that above age four the plotted relationship is somewhat irregular. But if we
 300

290

280

270

260

250
General Fertility

240

230

220

210

200

190

180

170
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Wife's Age at First Association

figure 4.1. General Fertility by Wife’s Age at First Association

table 4.1
General Fertility by Wife’s Age at First Association
Wife’s Age at General Fertility
First Association Years of Marriage Rate

0 28,873 180
1 5,514 188
2 3,698 208
3 2,651 229
4 1,927 234
5 2,560 223
6 1,886 236
7 1,473 249
8 1,867 239
9 1,834 264
10 1,930 254
11 1,765 267
12 1,743 266
13 1,335 246
14 1,042 }
290 268
 270
260
250
240
230
220
Fertility/Divorce Index

210
200
190
180
170
160
150
140
130
120
110
100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Wife's Age at First Association

figure 4.2. Fertility/ Divorce Index by Wife's Age at First Association

table 4.2
Fertility/Divorce Index by Wife's Age at First Association
Wife’s Age at Fertility/Divorce
First Association Years of Marriage Index

0 28,873 109
1 5,514 128
2 3,698 160
3 2,651 166
4 1,927 172
5 2,560 182
6 1,886 191
7 1,473 213
8 1,867 217
9 1,834 245
10 1,930 203
11 1,765 236
12 1,743 238
13 1,335 230
14 1,042 }
261 245
82 wolf

average the figures for ages thirteen and fourteen (which is, I think, justified
by the small denominators), the trend is sharply upward for every two-year
period before age nine, which marks the beginning of a high, fairly level
plateau. We must note, however, that the rise up to age three is far steeper
than the rise from age three to age nine. The average rise per year is 16.3
units for ages zero to two, as compared with 5.8 units for ages three to nine.
The data displayed in Table 4.2 and Figure 4.2 surprised me. I did not
expect adjusting for divorce to make such a large difference. Not only is
the distance between the lowest and highest points in Figure 4.2 54.6 per-
cent greater than in Figure 4.1, the relationship between the plotted vari-
ables is more regular. There are no reversals before age nine. Nonetheless,
our two indexes of sexual dissatisfaction present broadly similar profiles.
They rise sharply in early childhood, modulate, and then level out well be-
fore adolescence. The only significant difference is that the fertility/divorce
index locates the end of the early childhood rise a year earlier than the gen-
eral fertility index.
Although more data are needed to stabilize the trends at later ages, that
shown is sufficient to provide a better-than-tentative answer to our first
question. What does childhood mean in the Westermarck hypothesis? How
early do males and females have to meet to qualify as “living closely to-
gether from childhood”? The crude answer is “before age ten.” The more
refined answer is that while every year of association before age ten adds to
the sum of the Westermarck effect, association beginning before age three
is particularly potent. There is—to use Patrick Bateson’s language—a sen-
sitive period and a very sensitive period.15
The data presented above say that the earlier a girl destined to marry in
the minor fashion was adopted, the less satisfied she was with the relation-
ship created by her marriage. But what about her husband? Wasn’t his age
when he first met his future wife a factor? Girls adopted as infants were
most often matched with a boy born two or three years previously, but
many were matched with boys born six or seven years previously and oth-
ers with boys born after their arrival. Might it be, then, that we need to at-
tend to the husband’s age at first association as well as the wife’s? It all de-
pends on whether Dr. Ellis was right. If the “instinct” was really “more
marked in the female” than in the male, it might not matter very much
how old a boy was when his future wife was adopted.
One of my reasons for expanding my database beyond that employed in
my 1995 book was to be able to address this question. A large database is
required because to examine the effect of the husband’s age at first associa-
tion, one has to control on the wife’s age. Even with the incremental data I
have to confine my analysis to marriages in which the wife was adopted as
an infant (i.e., before age one). The data I use include all minor marriages
 Explaining the Westermarck Effect 83

in which the wife was adopted as an infant except those cases in which she
was adopted before her husband was born. Again I leave these special cases
until later.
The relationships between my two indexes and the husband’s age at his
wife’s adoption are shown in Tables 4.3 and 4.4 and Figures 4.3 and 4.4.
Again the profiles presented are broadly similar. Both indexes decline from
age zero to age one, remain low until age seven, rise sharply at age eight,
and then decline sharply at age twelve. I will ignore for the time being both
the early and late declines because the figures are not well grounded. Thus
what the evidence says is that so long as the husband is not eight or more
years older than his wife, his age when he first meets her does not matter.
The rise in both indexes if he is eight or more years older is most likely due
to the couple’s not “living closely together from childhood.” A man eight or
more years older than his wife was nine or more when she joined his fam-
ily. He would already be spending most of his waking hours studying or
working rather than playing with the nursing infant destined to be his wife.
Must we conclude, then, that Dr. Ellis was indeed right? If we assume,
as many anthropologists and most biologists do assume, that females invest
far more in their offspring than males, parental investment theory suggests
that he should be right.16 But is this assumption correct? It is certainly the
case that when the wife is the younger partner in a minor marriage, the tra-
jectory of my indexes is largely controlled by her age at first association.
The husband’s age does not matter as long as he is still a small child. But
what if the wife were the older partner? It is this question that makes im-
portant those marriages in which the wife was adopted before the hus-
band’s birth. They allow us to see what happened when the husband was a
nursing infant at first association and the wife a small child.
The relevant data are displayed in Tables 4.5 and 4.6. The figures shown
there are irregular because their base is small, but they all support the same
conclusion. They say that when the husband is the infant at first associa-
tion the aversion is as strong as when the wife is the infant. The critical
comparison is between the data in Tables 4.3 and 4.4 (where the wife is an
infant and the husband is a small child) and that in Tables 4.5 and 4.6
(where the husband is an infant and the wife is a small child). To make the
comparison easier, I have created Table 4.7, in which the two sets of data
are shown side by side. The obvious fact is that early association on the
husband’s part is as consequential as early association on the wife’s part. It
is particularly worth noting that when an infant boy is matched with a
three- to six-year-old girl, fertility is even lower than when an infant girl is
matched with a three- to six-year-old boy.
Suppose for a moment that the figures in Tables 4.5 and 4.6 rose steadily
with age, as is the case with the figures in Tables 4.1 and 4.2. One could
 230

220

210

200
General Fertility

190

180

170

160

150

140
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Husband's Age at First Association

figure 4.3. General Fertility by Husband’s Age at First Association

table 4.3
General Fertility by Husband’s Age at First Association
When Wife’s Age at First Association Is Zero
Husband’s Age at General Fertility
First Association Years of Marriage Rate

0 ,554 212
1 1,911 178
2 6,146 181
3 6,725 182
4 5,397 173
5 3,629 182
6 2,339 175
7 2,084 192
8 1,390 224
9
10
11
12
13
,451
,558
,307
,317
—
180

150
151
—
}
203 182

14 — —
 200

190

180

170
Fertility/Divorce Index

160

150

140

130

120

110

100

90
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Husband's Age at First Association

figure 4.4. Fertility/ Divorce Index by Husband's Age at First Association

table 4.4
Fertility/Divorce Index by Husband’s Age at First
Association When Wife’s Age at First Association Is Zero
Husband’s Age at Fertility/Divorce
First Association Years of Marriage Index

0 ,554 130
1 1,911 121
2 6,146 111
3 6,725 99
4 5,397 100
5 3,629 125
6 2,339 117
7 2,084 127
8 1,390 184
9
10
11
12
13
,451
,558
,307
,317
—
149

134
120
—
}
194 156

14 — —
86 wolf

then argue that these figures are low only because females are sensitive to
early association with a younger male as well as early association with an
older male. In other words, one could argue that Dr. Ellis was right. But in
fact there is no significant trend among the figures in either table. The rea-
son can only be because the source of the problem in these marriages is the
male. His age does not vary, and so the indexes do not vary. Consequently, I
conclude that Dr. Ellis was wrong. Males and females are equally sensitive
to the sexually inhibiting effects of early association. What really matters is
whether the male or the female is the younger partner. An interesting—and,
I think, important—implication is that we have either to reject parental in-
vestment theory or to conclude that our male ancestors made a much larger
contribution to their offspring’s survival than is commonly assumed.
I skipped the second of the questions raised above in order to pursue the
implications of Tables 4.3 and 4.4 for the question of a sex difference. The
question skipped was, What does closely mean in the phrase “living closely
together”? All I have to contribute to the answer at this point is the data re-
ported in Tables 4.3 and 4.4. The indexes in both tables indicate that a mi-
nor marriage is less likely to manifest sexual aversion if the husband is
eight or more years older than the wife. This suggests that closely means in-
tense interaction of the kind commonly found among children who are
close in age and who regularly play together. Examination of the composi-
tion of the sibling sets in which the parties to minor marriages were reared
might prove the point. The problem is that it would require a database sev-
eral times the size of the one on which this chapter rests.
Contrary to Mrs. Norris’s confident prediction, Fanny Price fell in love
with the younger of her two cousins, Edmund. He ignored her for years in
favor of Mary Crawford, but in the end he ceased “to care about Miss
Crawford, and became as anxious to marry Fanny, as Fanny herself could
desire.”17 So once again Mrs. Norris’s advice turned out to be bad advice,
but we can now see that though misapplied, the general principle on which
she based herself was sound. When she moved to Mansfield Park, Fanny
Price was “just ten years old.”18 She had just reached the age at which as-
sociation with the opposite sex is no longer inhibiting. Mrs. Norris’s “that”
was to be expected.
We need, then, to revise the Westermarck hypothesis. Age at first associa-
tion—and particularly the age of the younger member of any pair of poten-
tial partners—is far more important than previously realized. My revision—
phrased to remind us of the author of the original hypothesis—reads, “There
is a remarkable absence of erotic feelings between people who live together
and play together before age ten. The absence is particularly marked among
couples brought together before age three, and, for any given couple, largely
depends on the age of the younger partner when they first meet.”
 table 4.5
General Fertility by Wife’s Age at First Association
When Wife Is Adopted Before Husband’s Birth
Wife’s Age at General Fertility
First Association Years of Marriage Rate

0 ,213 188
1 ,822 202
2 1,949 200
3 1,508 167
4 1,151 177
5 ,470 147
6 ,340 168

table 4.6
Fertility/Divorce Index by Wife’s Age at First Association
When Wife Is Adopted Before Husband’s Birth
Wife’s Age at Fertility/Divorce
First Association Years of Marriage Index

0 ,213 127
1 ,822 128
2 1,949 131
3 1,508 83
4 1,151 109
5 ,470 85
6 ,340 50

table 4.7
Comparison of the Consequences of Early Association
When the Wife Is an Infant (Tables 4.3 and 4.4)
and the Husband Is an Infant (Tables 4.5 and 4.6)
General Fertility Fertility/Divorce Index

Age of Spouse at Wife Is an Husband Is an Wife Is an Husband Is an

First Association Infant Infant Infant Infant

0 ,212 188 130 127

1 ,178 202 121 128
2 181 200 111 131
3 182 167 99 83
4 173 177 100 109
5 ,182 147 125 85
6 ,175 168 117 50
88 wolf

How, then, are we to explain the reformulated Westermarck effect? Why

does association before age ten inhibit sexual attraction? Why is associa-
tion during the first two or three years of life particularly inhibiting? Why
is it that the strength of the inhibition depends on the age of the younger
partner rather than on the age of both partners? In sum, what is it that nat-
ural selection selected for that produces these effects?
Answering these questions will not be easy because it requires facing up
to a dilemma that I have so far avoided mentioning. It stems from the fact
that when the husband in a minor marriage is the older partner, his age at
first association does not influence the intensity of the aversion as long as
he is still a child. The intensity depends entirely on the age of the wife at
first association. This implies that the aversion is largely located on her side
of the relationship. But how, then, could the fertility of minor marriages be
so low and the divorce rate so high? My data come from China, where, in
the words of Friar Domingo Navarette, “the wives are half slaves and their
subjection is extraordinary.”19 How could wives who were “half slaves”
frustrate their husbands’ sexual drive and their desire for a houseful of sons
and grandsons? How could women whose subjection was “extraordinary”
get a divorce when divorce was regarded as a social disaster? I think that
anyone who knows the traditional Chinese family will agree that they
couldn’t. The aversion must have been felt, acutely, by the husband as well
as by the wife. But how could this be, if the strength of the aversion de-
pends on the wife’s age at first association and not on the husband’s age?
This is the dilemma. Solving it is one of the feats the Hercules of incest
avoidance must perform. Sadly, the field presently lacks a credible Her-
cules, but there are three promising candidates.
In a recent book one of my Stanford colleagues, Eleanor Maccoby, argues
that “there is a powerful tendency for children to segregate themselves by
gender in childhood and to play more compatibly with same-sex partners.”
This “drift into same-sex groups” is found in all societies and “begins to
show itself in the third year of life . . . and progressively strengthens until it
is strong indeed by middle childhood.”20 “It may be, then,” Maccoby sug-
gests, noting the reaction to minor marriages, “that children’s spontaneous
avoidance of cross-sex others who are not kin serves the biological function
of keeping these others within the pool of potential mates.”21
The timing Maccoby describes fits neatly with the age-dependent trends
displayed in Tables 4.1 and 4.2 (and, more dramatically, in Figures 4.1 and
4.2). It also fits with the fact that when the wife is the younger partner, the
strength of the aversion in minor marriages depends on the age at which
she joins her future husband’s family. Being the younger of the two, she
would be the last to join an exclusive same-sex group. Consequently, her
age, but not her husband’s age, would influence how frequently the couple
 Explaining the Westermarck Effect 89

interacted as children. The argument does not explain the Westermarck ef-
fect, but it does explain why the effect is age-related and why, when the fe-
male is younger than the male, her age seems to matter more than his.
The argument suggested by Maccoby is entirely compatible with West-
ermarck’s emphasis on “living closely together from childhood.” A more
radical approach to the problem of incest avoidance has been suggested to
me by another of my Stanford colleagues, Hill Gates, who, as an anthro-
pologist with a special interest in Taiwan, knows minor marriages at first
hand.22 Gates’s argument is radical in suggesting that what produces the
Westermarck effect may not be some aspect of playing together, eating to-
gether, and sleeping together. It may be something far more easily identi-
fied. It may be having been breastfed by the same woman.
Gates’s argument begins with the well-known finding that both males
and females prefer as sexual partners persons whose major histocompati-
bility complex (MHC) is different from their own.23 The reason, it is ar-
gued, is that they smell like relatives. Might it not be, then, that the dangers
of inbreeding have selected for the ability to identify relatives by their odor
and avoid them? The problem with this elegant solution to the problem of
incest avoidance is that it is contradicted by all the evidence presented
above. Couples married in the minor fashion were not relatives and there-
fore should not have been olfactorily obstructed. In a recent article Mark
Schneider and Lewellyn Hendrix have tried to salvage the MHC hypothe-
sis by suggesting that we learn the odors of the people with whom we are
reared and avoid them because they are probably relatives.24 Gates avoids
the complications this introduces by arguing that because the development
of MHC is strongly influenced by breastfeeding, children who are breastfed
by the same women tend to smell alike even if they are not siblings.
The great advantage of Gates’s hypothesis is that it resolves the minor
marriage dilemma. Taiwanese men were always put off by girls their mother
nursed for the simple reason that they had been nursed by the same woman.
Another advantage of the hypothesis is that it explains why association dur-
ing the first two years of life was so much more potent than association dur-
ing later years. This is because girls adopted at an early age were often
nursed by their future mother-in-law, while those adopted later were not.
The one problem with the hypothesis is that it does not explain why associ-
ation beginning after age three has any impact at all.
The third candidate is the one I nominated in my 1995 book, Sexual At-
traction and Childhood Association.25 It begins with John Bowlby’s account
of what he calls “attachment behavior.”26 Bowlby defines this “as any form
of behavior that results in a person’s attaining proximity to some other dif-
ferentiated and preferred individual, usually conceived of as stronger and /or
wiser.” The behavior includes clinging, crying, calling, greeting, and smiling.
90 wolf

It is evident from six months onward, “when an infant shows by his behav-
ior that he discriminates sharply between his mother-figure, a few other fa-
miliar people, and everyone else.” It reaches its “maximum during the sec-
ond and third year of life and then diminishes slowly.”27
Bowlby believes attachment behavior is one of three basic components
of human nature. The second is “the urge to explore the environment, to
play and to take part in varied activities with peers,” and the third is care-
giving, which Bowlby characterizes as “the prime role of parents and com-
plementary to attachment.”28 Just as human beings are born with a ten-
dency to seek and maintain contact with persons who are better able to
cope, so also they are born with an innate tendency to succor and support
other human beings who are not yet able to cope. This is, in Bowlby’s view,
“readily understood since it serves to promote the survival of offspring and
thus the individual’s own genes.”29
My solution to the dilemma posed by minor marriages is to combine
Bowlby’s argument with Westermarck’s. The selection forces favoring the
dispositions underlying attachment and caregiving push us ever closer to
our genetic relatives and thus expose us ever more acutely to the dangers of
inbreeding, while the selection forces favoring the dispositions underlying
incest avoidance push us ever further from our genetic relatives and thus
make us ever more vulnerable to the dangers of isolation. Consequently, if
the advantage to be gained by strengthening one set of dispositions is to re-
sult in a net genetic gain, the other set must be strengthened at the same
time. To evolve at all, the two sets of dispositions had to evolve together.
The fact that attachments and sexual aversions both form more readily be-
fore age three than after is not coincidental. They are the same thing.30
What is proposed, then, is that “little daughters-in-law” taken before
age three attached themselves to their future husband because he was older
and appeared “stronger and /or wiser.” This behavior elicited caregiving in
return and thereby created an asexual relationship because having evolved
together with incest avoidance, attachment and caregiving are inherently
contrasexual. The reason the fertility of minor marriages varies with the
wife’s age and not the husband’s is simply because in most cases the wife is
the younger partner and thus the one who does or does not form an at-
tachment. The location of the aversion moves from the female side to the
male side when the husband is the younger partner.
These hypotheses need to be tested and can be tested. The Taiwan house-
hold registers contain all the information needed to reconstruct the exact
composition of households and even neighborhoods. Thus it would be pos-
sible to determine the availability of same-sex playmates and the likelihood
that couples in minor marriages had grown up separated by membership in
same-sex groups. The registers also record the removal of infants by death
 Explaining the Westermarck Effect 91

and adoption—even when the infant is only a few days old. Thus one could
determine whether a woman was lactating when she adopted a little daugh-
ter-in-law and thereby estimate the chances of her having nursed her son’s
wife. The problem is that analyzing a sufficiently large number of household
registers would be a huge task. Even Hercules would have hesitated.

notes
1. The relative frequency of these two forms of marriage in Taiwan is
documented in Arthur P. Wolf and Chuang Ying-chang, “Marriage in Taiwan,
1881–1905: An example of regional diversity,” Journal of Asian Studies, vol. 54
(1995), pp. 781–95.
2. A detailed account of major and minor marriages as institutions is available
in Arthur P. Wolf and Chieh-shan Huang, Marriage and Adoption in China,
1845–1945 (Stanford, Calif.: Stanford University Press, 1980).
3. See Wolf and Huang, Marriage and Adoption, chapter 18, pp. 242–50.
4. The registers are described in detail in Wolf and Huang, Marriage and
Adoption, chapter 3, pp. 16–33.
5. See Arthur P. Wolf, Sexual Attraction and Childhood Association: A Chinese
Brief for Edward Westermarck (Stanford, Calif.: Stanford University Press, 1995),
chapters 7 and 12, pp. 115–34 and 198–213.
6. Ibid., chapters 6 and 11, pp. 98–114 and 181–97.
7. Ibid., chapters 5 and 10, pp. 79–97 and 166–80.
8. Sigmund Freud, A General Introduction to Psychoanalysis (1920), trans.
Joan Riviere (New York: Pocket Books, 1953), pp. 220–21.
9. See, most importantly, Yonina Talmon, “Mate selection in collective settle-
ments,” American Sociological Review, vol. 29 (1964), pp. 491–508; Joseph
Shepher, “Mate selection among second-generation kibbutz adolescents: Incest
avoidance and negative imprinting,” Archives of Sexual Behavior, vol. 1 (1971),
pp. 293–307; Justine McCabe, “FBD marriage: Further support for the Wester-
marck hypothesis,” American Anthropologist, vol. 85 (1983), pp. 50–69; and
Wolf, Sexual Attraction.
10. Edward Westermarck, A Short History of Human Marriage (London:
Macmillan and Co., 1926), p. 80.
11. Jane Austen, Mansfield Park (1814) (New York: Penguin Books, 2000),
p. 8.
12. Karl Popper, Unended Quest: An Intellectual Autobiography (La Salle,
Illinois: Open Court Publishing, 1976), pp. 134–35.
13. Havelock Ellis, Views and Reviews (London: Desmond Harmsworth,
1932), p. 168. First published in the Birth Control Review (New York), September
1928.
14. Edward Westermarck, “Recent theories of exogamy,” Sociological Review,
vol. 26, no. 1 (1934), p. 36.
15. Patrick Bateson and Robert A. Hinde, “Developmental changes in sensitiv-
92 wolf

ity to experience,” in Sensitive Periods in Development, ed. Marc H. Bornstein

(Hillsdale, N.J.: Lawrence Erlbaum, 1987), pp. 19–20.
16. See Robert L. Trivers, “Parental investment and sexual selection,” in
Sexual Selection and the Descent of Man, ed. Bernard Campbell (London: Aldine,
1972), pp. 136–79.
17. Austen, Mansfield Park, p. 387.
18. Ibid., p. 12.
19. The Travels and Controversies of Friar Domingo Navarette, ed. J. S.
Cummins, Hakluyt Society, ed. ser., no. 118 (Cambridge: Cambridge University
Press, 1960), vol. 1, p. 161.
20. Eleanor E. Maccoby, The Two Sexes (Cambridge: Harvard University
Press, 1998), pp. 29–30.
21. Ibid., p. 94.
22. Personal communication during dinnertime conversation. Hill Gates is
my wife.
23. See C. Wedekind, T. Seebeck, F. Beetens, and A. J. Paeike, “MHC-
dependent mate preferences in humans,” Proceedings of the Royal Society of
London, Series B, vol. 260 (1995), pp. 345–49; and C. Wedekind and S. Furi,
“Body odour preferences in men and women: Do they aim for specific MHC
combinations or simply heterozygosity?” Proceedings of the Royal Society of
London, Series B, vol. 264 (1997), pp. 1471–79.
24. Mark A. Schneider and Lewellyn Hendrix, “Olfactory sexual inhibition
and the Westermarck effect,” Human Nature, vol. 11, no. 1 (2000), pp. 65–91.
25. See Wolf, Sexual Attraction, chapter 28, pp. 463–75. My argument was
anticipated by Mark Erickson in articles published several years previously.
See Mark T. Erickson, “Incest avoidance and familial bonding,” Journal of
Anthropological Research, vol. 45., no. 3 (1989), pp. 270–80; and “Rethinking
Oedipus: An evolutionary perspective on incest avoidance,” American Journal of
Psychiatry, vol. 150, no. 1 (1993), pp. 400–413.
26. Essentially the same explanation was developed independently by Mark T.
Erickson. See Erickson, “Incest avoidance and familial bonding.”
27. John Bowlby, “Attachment theory, separation anxiety, and mourning,” in
American Handbook of Psychiatry, ed. David A. Hamburg and H. K. H. Brodie
(New York: Basic Books, 1975), vol. 6, p. 292.
28. John Bowlby, A Secure Base (New York: Basic Books, 1988), pp. 163–65.
29. Ibid., p. 165.
30. This argument is developed in detail in Wolf, Sexual Attraction, chapter 28,
pp. 439–75.
5 Ancient Egyptian Sibling Marriage
and the Westermarck Effect

Walter Scheidel

Under Roman rule, all residents of Egypt were required to partici-

pate in a periodic census held every fourteen years. The household was the
basic unit of registration; each head of a household had to file a return list-
ing all of its members with their names, ages, and kinship affiliations. A
minute fraction of the millions of census declarations ever filed has sur-
vived on papyrus. So far, close to 300 of them have been published and
subjected to demographic analysis.1 Of 121 current marriages documented
in these records, twenty are between full siblings and four between half-
siblings. Another three sibling unions are known to have ended in divorce.
Other kinds of papyrus texts outside this corpus contain references to thir-
teen further sibling couples, one of them divorced. All these unions were
(legally) monogamous.
Sibling marriage, attested for the second and the early third centuries ad,
appears to have been more prevalent in urban settings than in the villages.
In the most amply documented location, the district capital of Arsinoe in the
Fayum Oasis (southwest of modern Cairo), seventeen of forty-six known
unions, or 37 percent, are between full siblings. Owing to the limited avail-
ability of suitable sibling-spouses in any particular family and a strong pref-
erence for younger wives, the observed incidence approaches the feasible
maximum. In this environment, sibling marriage was a cultural norm rather
than merely an acceptable option.2 This invalidates Arthur Wolf’s claim,
based on an inadequate sample of evidence, that according to the Egyptian
data, “even in the absence of an incest taboo, children who are raised to-
gether rarely marry.”3
Brother-sister marriage in Roman Egypt has long been noted as a con-
spicuous exception to near-universal taboos against regular sexual and mar-
ital relations between very closely related kin. Traditionally, this phenome-
non has been discussed from a cultural-constructivist perspective.4 In a
series of interlocking studies, I have sought to redress the balance by analyz-
94 scheidel

ing the available data within a biosocial framework. Rather than guessing at
possible explanations, I have focused on the probable repercussions of and
constraints on this peculiar custom, drawing on comparative evidence for
inbreeding depression and avoidance behavior.5 As a result, Roman Egypt-
ian sibling marriage has now finally entered the debate on the biological ba-
sis of human incest taboos.6
It is easy to show that the most intensely incestuous segments of the
Egyptian population, such as the residents of Arsinoe, were more heavily
inbred than any other known human population.7 In the absence of direct
or indirect textual or physical evidence, the scale of resultant inbreeding de-
pression is uncertain but likely to have been substantial (see below). Ex-
plicit references to an aversion against sibling marriage are also missing
from the record (such as private letters on papyrus or literary accounts). In
this regard, the Egyptian case differs from Zoroastrian “close-kin mar-
riage” between parents and children and brothers and sisters, practiced in
Iran and the Near East in antiquity and the Middle Ages, perhaps primarily
in polygamous circles. Glorifying incest as exceptionally meritorious, the
Zoroastrian spiritual authorities expressly considered it challenging even
for devoted insiders and envisioned only infrequent sexual encounters of
this nature. Instinctive reluctance was reported by sympathetic insiders and
hostile outsiders alike.8
Comparative evidence sheds some light on the probable psychological ef-
fects of Egyptian sibling marriage. Research on modern populations has es-
tablished a “remarkable absence of erotic feelings between people living
closely together from childhood”9 and consequent aversion against repro-
ductive sexual intercourse at mature ages, known as the “Westermarck ef-
fect.” In his work on the custom of “minor marriage” in Taiwan—in which
the parents of a newborn son adopted (or bought) a young girl, often an in-
fant, to be raised together with their son as his future wife—Wolf found that
the marital fertility of girls who had been adopted at an early age was 40
percent lower than that of wives who had been raised by their own families,
and also that “minor marriages” with a small age gap between spouses were
more than twice as likely to involve adultery and three times as likely to end
in divorce than other unions. Wolf shows that the intensity of early child-
hood association was the critical variable mediating marital success.10 In this
chapter, I will assess the significance of this factor in the context of Egyptian
sibling marriage.

Data

Table 5.1 summarizes the vital statistics of all Roman Egyptian sib-
ling couples that are currently known. The majority of cases (twenty-seven
 table 5.1
Sibling Couples Attested in Papyrus Documents from Roman Egypt
(first to third centuries AD)
Offspring Residing
Couples with Parents

Age Gap Age/Husband Age/Wife Ages

Case (years) (years) (years) Number (years)

Full siblings (census returns)

1 ,2 31 29 1 8
2 3 21 18 1 ?
3 3 46 43 6 or 7 20, 16, 14, 10, 7, 3, 2(?)
4 4 [4]4 40 2 12, 8
5 4 46 42 5 12, 10, 8, 6, 2
6 <6 5[-] 54 8 29, 26, 23, ?, 17, ?, 9, 7
7 6 36 30 0
8 8 21 13 0
9 8 21 13 0
10 8 22 14 0
11 8 44 52 1 8
12 ? (0–9) 20 2[-] 0
13 10 48 38 1 ?
14 17 34 17 1 1
15 ? 21 ? 2 >1, <1
16 ? 29 ? 2 1, 1 (twins)
17 ? ? 35 2 ?
18 ? 40 ? 1 2
19 ? 73 ? 2 40, ?
20 ? ? ? 3 23, ?, ?
21* ? ? ? 2 ?, 10
22 ? ? ? 2 ?, ?
Full siblings (other documents)
23 ,1 ? ? 1 known ?
24 3 ? ? 2 known ?
25* 4 54 50 ? ?
26 8 43 35 1 known 22
27 ? ? ? 2 known ?
28 ? ? ? 1 known ~14
29 ? ? ? 1 known 13
30 ? ? ? 1 known 4
31 ? ? ? 1 suggested ?
Half-siblings (census returns)
32 ,2 31 29 4 14, 6, 3, 2(?)
33* 4 22 26 3 ?
34 12 30 18 0
35 22(?) [6?]3 41 4 21, 13, 11, ?
36 ? ? ? 2 ?, 14
Full or half-siblings (other documents)
37 ,? ? ? 1 known
Half-siblings (other documents)
38 ,? ? ? 1 known 8
39 ? ? ? 2 known
40* ? ? ? ?

key: * divorced; ? missing information; (?) uncertain reading; [ ] missing/illegible numeral

sources: Cases 1–22 and 32–36: Roger S. Bagnall and Bruce W. Frier, The Demography
of Roman Egypt (Cambridge: Cambridge University Press, 1994); cases 23–24, 27:
P. Amherst 75; case 25: P. Kron. 52; case 26: BGU 183 (P. 6867); cases 28, 37: P. Tebt.
32 0; case 29: P. Oxy. 43.3 096; case 3 0: WChr. 211; case 31: P. Duk. inv. 491; case 38:
P. Pet. 1–2; case 39: P. Oxy. 43.3137; case 4 0: P. Strasb. 768.
96 scheidel

of forty) are attested in census returns. Thanks to the standardized format

of the census declarations and their objective to record all coresident kin in
a given household, the demographic information contained in these docu-
ments is often more complete and more fully contextualized than that pre-
served in the other texts. Even so, each census return gives only a snapshot
of household composition at a particular moment in time. Thus, no com-
pleted life histories are available; while young couples would have pro-
duced additional offspring later on, some of the children of older couples
may already have left the parental home at the time of registration and re-
main invisible. Deceased progeny is not recorded.
In view of very high levels of mortality—most likely of the order of 30
percent up to age one, 45 percent up to age five, and 55 percent up to age
twenty (and probably even more in those cases in which inbreeding depres-
sion needs to be taken into account; see below)—the discrepancy between
the number of children ever born to a couple and the number of their chil-
dren who were alive at any given time must have been substantial. As a
consequence, the census records cannot be used to measure the total mari-
tal fertility rate (TMFR). Other texts that mention brother-sister marriage
are frequently concerned with one particular child (listed as “known” in
Table 5.1) and omit others that may have existed, as well as information
on parental age. In a number of instances, the papyrus texts are so poorly
preserved that some entries are fragmentary or difficult to decipher; these
problems are indicated in Table 5.1.
In the census returns in general, illegitimate children are usually identified
as such, yet there are no references to the (acknowledged) illegitimacy or
adoption of sibling-spouses or their descendants. Although unknown or
concealed adulterous or extramarital conceptions may sometimes have
occurred, there is nothing to suggest that sibling-spouses were frequently bi-
ologically unrelated to one another and /or their putative children. No prac-
tice comparable to Chinese “minor marriage” is known from Roman Egypt.
In fact, individual full sibling-spouses are repeatedly listed as “brother/sis-
ter from the same father and the same mother.”11 With a single exception
(case 37), this precision enables us to distinguish between couples of full and
half-siblings.
By the standards of other ancient age records or census counts in mod-
ern developing countries, the degree of age rounding (defined as preference
for numerals ending in multiples of five) in the census returns is remarkably
low, and particularly modest in those urban locales that produced much of
the relevant texts.12 The age data in Table 5.1 show no significant deviation
in favor of multiples of five (thirteen of eighty-five final digits, or 15.3 per-
cent, consistent with the expected rate of 20 percent). For this reason, the
recorded age gaps between siblings may generally be accepted at face value.
 Ancient Egyptian Sibling Marriage 97

Birth Intervals and Spousal Age Difference

The observation that early childhood association curbs sexual at-

traction and may even trigger sexual aversion at mature ages provides an
obvious starting point for a reappraisal of brother-sister marriage. I there-
fore agree with Lewellyn Hendrix and Mark Schneider that “a crucial con-
sideration would be how intimate siblings were in Graeco-Roman Egypt
during the sensitization period of early childhood.”13 However, Seymour
Parker is mistaken in claiming that “there are no data on the closeness of
opposite-sex sibling socialization during the early years.”14 The recorded
age gaps between sibling spouses shed some light on the likely level of vari-
ation in the intensity of early childhood association in this group. Two dif-
ferent ways of calculating the average age difference between full sibling
spouses produce very similar results.
The average length of birth intervals can be deduced from the probable
fertility schedule of married women in Roman Egypt. The general age dis-
tribution in the census returns as well as comparative evidence from early-
twentieth-century Egypt and comparable historical populations points to a
mean life expectancy at birth of between twenty and twenty-five years.15
Again drawing on the census returns, Bruce Frier was able to show that the
attested birthing schedule is fully consistent with a natural fertility regime,
characterized by the absence of family limitation achieved through stop-
ping behavior; thus, birth intervals were relatively long but childbearing
continued into the forties, gradually decreasing with diminishing female fe-
cundity.16 Positing a mean life expectancy at birth of about twenty-two to
twenty-five years, this observation allowed Frier to reconstruct two ver-
sions of the Roman Egyptian marital fertility schedule, one reflecting the
raw data and the other based on an idealized Gompertz curve fitted to the
documented maternal age distribution of childbearing (Table 5.2).
According to the model pattern, a married woman at the peak of her fe-
cundity on average gave birth every 3 years. This rate slowed to one birth
every 3.5 years in her thirties before dropping sharply in her forties. Ap-
proximately seven out of eight children born in wedlock were born to
women aged fifteen to thirty-nine and were on average separated by 3.4
years. Two qualifications deserve attention. First, divorce and the death of
husbands would sometimes interrupt periods of high fertility, thereby in-
creasing the average actual distance between births. The total fertility rate
(TFR) was close to six children per woman surviving to menopause, fully
30 percent lower than TMFR in the model. Second, and even more impor-
tant, high levels of prereproductive mortality would, on average, greatly
raise the age difference between siblings surviving to sexual maturity. If
only half of all newborns could hope to survive into their early teens, as
98 scheidel

table 5.2
Marital Fertility in Roman Egypt
(mean number of births per maternal age cohort)
Age Attested Gompertz Model

12–14 ,0.066 0.069

15–19 1.16 1.245
20–24 1.715 1.665
25–29 1.835 1.625
30–34 1.465 1.495
35–39 1.09 1.31
40– 44 1.095 0.83
45– 49 0.67 0.185
Total 9.096 8.424

source: Adapted from Bruce W. Frier, “Natural Fertility and

Family Limitation in Roman Marriage,” Classical Philology 89
(1994), 325, table 1.

seems likely, the mean age gaps between these individuals would have been
twice as wide as the average birth intervals. For children born to women
aged fifteen to thirty-nine, the resultant age difference would, on average,
have amounted to approximately seven years.
This schematic calculation tallies well with the attested age gaps be-
tween sibling spouses in Table 5.1. The mean age difference for sixteen
couples of full siblings is 6.06 years, while the median (including case 6) is
6 years. If four half-sibling couples are included, the mean rises to 6.85
years. These observed rates are fully consistent with the estimate generated
by the model fertility schedule.
Five papyrus documents report brother-sister marriage in two successive
generations, and in one case, the practice continued across three genera-
tions.17 In those instances, the probable impact of inbreeding depression also
needs to be taken into account. Frier’s fertility schedule is predicated on the
assumption of zero or marginal natural population growth. However, ow-
ing to the likely impact of inbreeding depression (expressed in higher rates
of fetal loss, child mortality, and disability), it remains unclear whether in-
cestuous families were able to reproduce at full replacement level. If they
did, they may on average have needed more live births to produce the same
number of mature offspring than nonkin couples. In an earlier study, I ten-
tatively calculated the required increase in marital fertility, which amounts
to 19 percent for first-generation sibling-spouses, 30 percent in the second
generation, and 42 percent for the last of three successive sibling matings.18
For first-generation sibling couples (arguably the most common variety),
this boosts TMFR from 8.4 to 10 and reduces average birth intervals ac-
cordingly, to about 2.9 years in the fifteen- to thirty-nine-year maternal age
 Ancient Egyptian Sibling Marriage 99

bracket. At the same time, owing to higher prereproductive wastage, the

mean age difference between mature siblings remains unchanged. Alterna-
tively, it is very well possible that, on average, incestuous couples could not
reproduce at full replacement level. In that case, mean birth intervals may
have been closer to the norm, while the resultant average age gaps between
surviving siblings would have exceeded seven years.
The overall incidence of second-generation sibling marriage remains ob-
scure. On average, no more than about 60 percent of all brothers had a sis-
ter of marriageable age, and owing to a strong preference for younger wives
(in sixteen of eighteen sufficiently well-documented couples in Table 5.1, the
brother-husband is older than the sister-wife), the actual incidence of sibling
marriage in any given generation must have been closer to 30 percent. Thus,
any increase of mean spousal age difference through inbreeding depression
was unlikely to affect more than a relatively small minority of all sibling
couples. We may conclude that the average age gap between full sibling
spouses must have been of the order of seven years.
This overall average necessarily conceals considerable variation. In our
data set, nine of twenty-one sibling couples, or 43 percent, are separated by
eight or more years of age. Despite an emerging consensus on the relevance
of early childhood cohabitation and cosocialization for mating preferences
in adulthood, there is some disagreement as to the precise length of the sen-
sitization period. Joseph Shepher reckoned with mature avoidance caused
by close contact during the first six years of life.19 The most detailed stud-
ies, Wolf’s work on “minor marriages,” show a strong sensitization effect
during the first three years followed by a steady decrease. When a future
husband was eight years or older at the time when a newborn girl was
adopted as his future wife, the impact of their belated association on sub-
sequent marital success (measured by fertility and divorce rates) was fairly
negligible.20 This configuration is equivalent to the relationship between a
brother-husband and a sister-wife who was eight or more years younger
than he. In their analysis of modern incest, Irene Bevc and Irwin Silverman
find that full genital—that is, potentially reproductive—intercourse between
siblings today is strongly associated with prolonged separation during the
first three years of life.21
All these studies agree that the effects of early childhood association on
later mating behavior are strongest for the first few years of life and mar-
ginal if sibling age difference exceeds six to eight years. Hence, in almost
half of all cases, the Egyptian sibling spouses were too far apart in age to
have been exposed to intensive sensitization. The Chinese and other com-
parative data indicate that under these circumstances (which include strong
cultural expectations favoring incest), instinctive aversion to reproductive
relations would have been weak or altogether missing and that the success
100 scheidel

of such sibling unions would have been unimpaired by the Westermarck ef-
fect. Moreover, the consistent preference for younger sister-wives must have
limited close contact between older brothers and significantly younger sis-
ters; while older girls may have been expected to care for infant brothers,
the reverse scenario is considerably less likely.

Conjugal Dissolution

The Chinese data for “minor marriages” suggest that divorce rates
were a function of the degree of early childhood association between future
spouses.22 Thus, the likelihood of conjugal dissolution was highest when
the wife had been adopted between ages zero and four, and negligible after
age nine, while for men, it was highest when the future spouse had been
adopted during the first six years of the boy’s life.23 This raises the question
of whether a similar trend can be discerned in Roman Egypt. Strictly speak-
ing, the age at adoption in “minor marriages” is not fully equivalent to age
difference between biological sibling-spouses; in the Chinese context, both
future partners would sometimes initially spend some time on their own
before the girl was introduced into her new family. In one Chinese sample,
44.5 percent of these girls were adopted before age one and 64.3 percent
before age three.24 Thus, while age at adoption is generally indicative of the
intensity of early childhood association, it can serve only as a rough proxy
for age difference between siblings as observed in the Egyptian data.
Nevertheless, even allowing for this discrepancy, the Chinese evidence
would seem to predict elevated rates of marital failure among Egyptian sib-
ling couples close in age, especially among those who were separated by no
more than four to six years. On the face of it, the data are consistent with
this prediction (Table 5.3).
The average divorce rate for nonsibling couples in the census returns is
11.8 percent (13 of 110).25 By contrast, 30 percent of known unions be-
tween siblings who were close in age ended in divorce. Owing to the exigu-
ous size of the sample, the significance of this deviation from the putative
mean remains very weak (p<0.17, z-test). However, two qualifications are
in order. The final divorce rate remains unknown even for these few cou-
ples. The lack of completed life histories means that the available data sys-
tematically underestimate the actual incidence of separation. This raises the
possibility that complete life histories would reveal a more significant cor-
relation between age difference and divorce. Besides, the suggestive match
between the patterns of divorce in the scant Egyptian and the much more
numerous Chinese data would make it seem rash to dismiss the findings de-
rived from the smaller sample out of hand.
 Ancient Egyptian Sibling Marriage 101

table 5.3
Cases of Divorce by Spousal Age Gap
Age Gap Between Spouses

1–4 Years 6+ Years

All couples ,10 10

Divorced couples 3 0

source: Table 5.1.

Cross-Fostering

The data discussed so far are consistent with comparative evidence

and general predictions derived from it; thus, Egyptian sibling marriage may
have worked reasonably well when spousal age difference was considerable
but less so if spouses were close in age. However, several prolific sibling cou-
ples with a modest age difference diverge from this inherently plausible pat-
tern (cases 3, 5, 6, 32). The Chinese data as recalculated by Wolf show a
strong correlation between age at adoption (our rough proxy for age differ-
ence) and marital fertility even if divorce is controlled for; close association
in the first few years of life results in significantly reduced reproductive suc-
cess at mature ages.26 The four Egyptian instances confound our expectations
that the same ought to have been true of genuine sibling marriage. Several
explanations are possible. The paucity of data may be an obvious answer; in
a sample this small, some random anomalies would be unsurprising. Perhaps
more important, the Chinese evidence leaves no doubt that although the im-
pact of early childhood association on fertility was significant, it was also
limited in scale. In many cases, inhibitions were to some extent overcome by
parental coercion and social expectations.27 As a consequence, above-average
fertility in some sibling couples need not be particularly noteworthy.
For all that, the observed age gaps (and implied birth intervals) in these
successful families may provide a clue to a more satisfying answer that takes
full account of the mechanisms of early childhood association (Table 5.4).
The three most prolific couples in this group, all of them consisting of full
siblings, list living offspring with a total of thirteen known age gaps between
siblings, ranging from one to four years. (The two unknown intervals in
case 6 are likely to fall in the same range.) Some of these intervals seem
short for two reasons. First, we have to allow for infant and early childhood
mortality; some of the actual birth intervals may have been shorter than the
observed age gaps between living siblings. (The first interval in case 32 is an
obvious candidate.) Second, birth intervals of one or two years are hard to
reconcile with ancient Egyptian nursing practices. Judging by the literary
102 scheidel

table 5.4
Birth Intervals of Offspring of Prolific Sibling Couples Separated
by Fewer Than Six Years of Age (in years)
Couples

Birth Interval Case 3 Case 5 Case 6 Case 32 All

1 1(?) — — 1(?) 2(?)

2 1 3 1 — 5
3 1 — 2 1 4
4 3 1 — — 4
<6 — — 1 — 1
<8 — — 1 — 1
8 — — — 1 1

source: Table 5.1.

tradition, children were to be breast-fed for three years.28 The papyrological

record from the Roman period includes forty wet-nursing contracts in
which unrelated women (often slaves) are hired to breast-feed an infant (of-
ten a slave or foundling, sometimes the child of the employer) for a specified
period of time. The term of employment is known in twenty-eight cases; 2
and 2.5 years are the most common periods, while 3.5 years is the longest.
Terms for freeborn nurslings range from 2 to 3 years.29
“It is now well-established that breastfeeding can exert a powerful con-
traceptive effect, potentially delaying the return of full fecundity during the
postpartum period by two years or more.”30 The evidence cited here shows
that freeborn Egyptian children were regularly supposed to be breast-fed
for up to three years. In addition to this strong cultural preference for pro-
longed breast-feeding, other factors contributed to the comparatively long
birth intervals suggested by Table 5.2. Thus, endemic ill health, parasitism,
and high mortality at all ages, even among the propertied classes, must
have resulted in correspondingly elevated rates of fetal loss and temporary
sterility,31 as did inbreeding depression in those couples who were them-
selves the issue of brother-sister unions (see above).
There is no direct evidence that any known sibling-spouse had been cared
for by a wet nurse. Even so, the observed combination of extended breast-
feeding, short age gaps between living siblings (with potentially even shorter
birth intervals), and the apparent popularity of wet-nursing arrangements
suggests that some of the children of the most prolific sibling couples may
have been nursed by women other than their own mothers. At the very least,
intervals of one or two years as documented in Table 5.4 strongly support
this assumption. This in turn raises the possibility that such couples may
have followed the example of their own parents and that some sibling-
spouses had themselves been in the care of wet nurses. After all, age gaps of
 Ancient Egyptian Sibling Marriage 103

one or two years are likewise attested for some sibling couples (cases 1, 23,
and 32).
The available evidence for the technical details of wet-nursing arrange-
ments is somewhat ambiguous. In a number of cases, the contracts obligate
the wet nurse to raise the nursling in her own home (or that of her owner
if she was a slave).32 This clause, however, is found only in agreements pro-
viding for the nursing of slaves and foundlings and does not appear in any
known contract for freeborn infants. Instead, in one text of the latter cate-
gory, the wet nurse expressly promises her employer, “I will stay day and
night in your house together with the baby.” The “day and night” stipula-
tion is also found in another contract for a free child. A poorly preserved,
lacunose contract for a third free nursling appears to give the wet nurse an
option of whether to follow her employer beyond his regular residence,
which implies that she would ordinarily stay there.33
An infant’s transfer to the home of a wet nurse would have sheltered the
child from early association with siblings and other family members. How-
ever, the contracts suggest that when their own children were involved, par-
ents preferred to hire a live-in wet nurse. While it is impossible to be sure
that future sibling-spouses were never kept in the homes of unrelated wet
nurses, there is no positive evidence for this. In any case, the “day and night”
stipulation makes it likely that a live-in wet nurse acted as a child minder be-
yond the actual feeding process and would have been a major focus of at-
tachment for the child. The same would have been true of lactating house-
hold slaves who took care of their owners’ children, a practice that was
common in Rome and must also have existed in affluent Egyptian circles.34
Recent work on olfactorily mediated sensitization in early childhood
raises the possibility that wet-nursing arrangements of this kind might have
interfered with the development of inhibitions to sexual intercourse with
close relatives at mature ages. It is now well known that various animal
species including humans are highly sensitive to information, conveyed by
body odor, on genetic variation in the major histocompatibility complex
(MHC, known as HLA in humans). The MHC, a highly polymorphic group
of genes, serves as a matching system used by the immune system to dis-
criminate between self and others and is therefore also instrumental in kin
recognition.35 Olfactory perception of minute genetic differences in MHC-
type can be shown to influence mating preferences in animals and humans.36
This mechanism frequently—though not invariably—guides individuals to-
ward mates with MHC types that are different from their own. In such
cases, it is heterozygosity as such that is favored rather than any particular
MHC type.37 This generalized preference for heterozygosity may enhance
immunological resistance to pathogens.38 At the same time, it also favors
outbreeding and may therefore have evolved to avoid inbreeding.39
104 scheidel

Humans can be shown to prefer the body odor of potential mates with
different MHC types.40 Despite these preferences, evidence for disassorta-
tive mating (in the form of marriage) has so far remained ambiguous.41 It
may be possible to explain the absence of disassortative mating practices
with reference to a lack of free mate choice or the very real possibility that
MHC preference is based on (less precise) familial imprinting rather than
self-reference.42
Preference for different MHC strains is the result of an early learning
process. At the very least, associate reference (in which infants are sensi-
tized to their own MHC type by being sensitized to the body odor of those
around them, who are most likely to be close kin) plays a significant role
alongside self reference (in which MHC preference would be directly deter-
mined by one’s own body odor). The importance of associate reference was
demonstrated when laboratory mice that had been removed from their na-
tal litter after birth, fostered by an unrelated female, and then returned to
their own group exhibited avoidance of the MHC type of their foster par-
ents but not of their biological kin.43 The same effect could be shown in
wild-derived mice in seminatural conditions.44
I know of no empirical evidence for the impact of cross-fostering on the
sensitization of humans. In view of other structural similarities with regard
to MHC-mediated preferences across different mammalian species including
humans, it seems at least possible that human infants might react similarly
to the way mice react. In pseudo-sibling relationships, such as Chinese “mi-
nor marriage,” exposure to the body odor of biologically unrelated mem-
bers of the adoptive family can be expected to trigger subsequent avoidance
of mates from within that group and so help account for the Westermarck
effect.45 On occasion, Chinese mothers were known to breast-feed their
adopted daughters-in-law; a girl nursed by the mother of a biologically un-
related brother-spouse may thus have been sensitized against later sexual re-
lations with that male.46
In Roman Egypt, the opposite effect may have occurred. Several years of
regular exposure to the breast milk and the breast and axillary odor of an
unrelated wet nurse may have sensitized small children to an MHC type
other than their own and thereby reduced their inhibitions against sexual re-
lations with their own kin at mature ages. (In this scenario, it would not
matter whether two sibling-spouses had been nursed by the same stranger or
by different women.) The overall impact of this sensitization relative to con-
current sensitization to coresident siblings remains open to debate. If future
sibling-spouses had been physically removed from their natal families for the
first few years of their lives, the dominance of their sensitization to nonkin
MHC-types could hardly be doubted. As it is, it remains uncertain whether
breast-feeding and nursing were more potent elements of early childhood
 Ancient Egyptian Sibling Marriage 105

sensitization than contact with coresident siblings and parents. Further stud-
ies on the effects of cross-fostering are required to shed light on this issue.

Conclusion

We cannot say why so many couples in Roman Egypt arranged mar-

riages between their own children. Evidence of religious motivation, which
is richly available for Zoroastrian close-kin marriage, is lacking here.47
Concern about the preservation of family property and family privilege has
been mooted as a possible motive but could equally well have been allevi-
ated by the unexceptional custom of first-cousin marriage. (Unions of this
type also appear in the census returns but can be identified only if a cou-
ple’s sibling parents resided in the same household.) Both status enhance-
ment via religious devotion and status preservation via “closed-family mar-
riages” would constitute plausible proximate mechanisms to raise fitness
even if they were bound to be maladaptive in the long run.
As I have tried to show, it may be easier to explain the actual workings
of this custom in Darwinian terms. There is no sign that a short-lived me-
metic mutation such as monogamous sibling marriage rendered evolved
constraints immaterial. Rather, its temporary success appears to have been
both facilitated and limited by a combination of different factors. In a sub-
stantial proportion of all cases, probably close to one-half, considerable age
differences between the spouses would have reduced or removed inhibi-
tions to sexual intercourse and reproduction. Some couples who were close
in age may conceivably have been sensitized to the body odor of unrelated
wet nurses and might consequently have been spared strong feelings of sex-
ual aversion at mature ages. In still other cases, sibling-spouses who were
close in age and had been sensitized to their own kin may well have experi-
enced elevated rates of conjugal dissolution. All in all, there is nothing to
show that as far as the correlation between early childhood association and
sexual inhibition is concerned, the evidence for Roman Egyptian sibling
marriage deviates significantly from the pattern derived from the Chinese
data on “minor marriages” and other information on the demographic
context of incestuous behavior and incest avoidance in humans.

notes
1. Roger S. Bagnall and Bruce W. Frier, The Demography of Roman Egypt
(Cambridge: Cambridge University Press, 1994).
2. Walter Scheidel, “Brother-Sister Marriage in Roman Egypt,” Journal of
Biosocial Science, vol. 29 (1997).
106 scheidel

3. Arthur Wolf, Sexual Attraction and Childhood Association: A Chinese

Brief for Edward Westermarck (Stanford, Calif.: Stanford University Press, 1995),
p. 433.
4. For example, Keith Hopkins, “Brother-Sister Marriage in Roman Egypt,”
Comparative Studies in Society and History, vol. 22 (1980), pp. 303–54; Brent D.
Shaw, “Explaining Incest: Brother-Sister Marriage in Greco-Roman Egypt,” Man,
vol. 27 (1992), pp. 267–99; S. Parker, “Full Brother-Sister Marriage in Roman
Egypt: Another Look,” Cultural Anthropology, vol. 11 (1996), pp. 362–76.
5. Walter Scheidel, Measuring Sex, Age, and Death in the Roman Empire:
Explorations in Ancient Demography (Ann Arbor, Mich.: Journal of Roman
Archaeology, 1994), pp. 9–51; Scheidel, “Brother-Sister and Parent-Child Mar-
riage Outside Royal Families in Ancient Egypt and Iran: A Challenge to the
Sociobiological View of Incest Avoidance?” Ethology and Sociobiology, vol. 17
(1996), pp. 319–40; Scheidel, “Brother-Sister Marriage in Roman Egypt.”
6. L. Hendrix and M. A. Schneider, “Assumptions on Sex and Society in the
Biosocial Theory of Incest,” Cross-Cultural Research, vol. 33 (1999), pp. 193–218;
M. A. Schneider and L. Hendrix, “Olfactory Sexual Inhibition and the Wester-
marck Effect,” Human Nature, vol. 11 (2000), pp. 65–91.
7. Scheidel, “Brother-Sister Marriage in Roman Egypt.”
8. Scheidel, “Brother-Sister and Parent-Child Marriage in Ancient Egypt and
Iran”; “Brother-Sister and Parent-Child Marriage in Premodern Societies,” in
Human Mate Choice and Prehistoric Marital Networks, ed. Kenichi Aoki and
Takeru Akazawa (Kyoto: International Research Center for Japanese Studies,
2002), pp. 33–47.
9. Edward Westermarck, A Short History of Human Marriage (London:
Macmillan, 1926), p. 80.
10. Wolf, Sexual Attraction; and Chapter 4 of this volume.
11. The most recently published text (case 31) documents the first known
instance of marriage between twins: N. Gonis, “Incestuous Twins in the City
of Arsinoe,” Zeitschrift für Papyrologie und Epigraphik, vol. 133 (2000),
pp. 197–98.
12. Scheidel, Measuring Sex, Age, and Death, pp. 53–91.
13. Hendrix and Schneider, “Assumptions on Sex and Society,” p. 213.
14. Parker, “Full Brother-Sister Marriage,” pp. 373–74.
15. Bagnall and Frier, Demography of Roman Egypt, 75–110; Walter Scheidel,
Death on the Nile: Disease and the Demography of Roman Egypt (Leiden,
Netherlands: Brill, 2001), pp. 118–80.
16. Bruce W. Frier, “Natural Fertility and Family Limitation in Roman
Marriage,” Classical Philology, vol. 89 (1994), pp. 318–33.
17. Scheidel, Measuring Sex, Age, and Death, pp. 11–12.
18. Ibid., 25, table 1.11, based on K. Ralls, J. D. Ballou, and A. Templeton,
“Estimates of Lethal Gene Equivalents and the Cost of Inbreeding in Mammals,”
Conservation Biology, vol. 2 (1988), pp. 185–93; Alan H. Bittles and James V.
Neel, “The Costs of Human Inbreeding and Their Implications for Variations at
the DNA Level,” Nature Genetics, vol. 8 (1994), pp. 117–21.
 Ancient Egyptian Sibling Marriage 107

19. Joseph Shepher, Incest: A Biosocial View (New York: Academic Press,
1983), pp. 51–62.
20. Wolf, Sexual Attraction; and Chapter 4 of this volume.
21. I. Bevc and I. Silverman, “Early Separation and Sibling Incest: A Test of
the Revised Westermarck Theory,” Evolution and Human Behavior, vol. 21
(2000), pp. 151–61; cf. already their “Early Proximity and Intimacy Between
Siblings and Incestuous Behavior: A Test of the Westermarck Hypothesis,”
Ethology and Sociobiology, vol. 14 (1993), pp. 171–81.
22. Wolf, Sexual Attraction; chap. 11.
23. Ibid., p. 197; and Chapter 4 of this volume.
24. Wolf, Sexual Attraction, p. 170.
25. Bagnall and Frier, Demography of Roman Egypt, p. 123.
26. Wolf, Chapter 4 of this volume.
27. Hendrix and Schneider, “Assumptions on Sex and Society,” p. 202.
28. Joyce Tyldesley, Daughters of Isis: Women of Ancient Egypt (London:
Penguin Books, 1995), p. 69.
29. Madriadele Manca Masciadri and Orsolina Montevecchi, I contratti di
baliatico (Milan: private printing, 1984), pp. 32–35.
30. James W. Woods, Dynamics of Human Reproduction: Biology, Biometry,
Demography (New York: Aldine de Gruyter, 1994), p. 370.
31. Scheidel, Death on the Nile.
32. Masciadri and Montevecchi, I contratti, p. 24.
33. Ibid., pp. 43, 149, 146.
34. Keith R. Bradley, “Wet-Nursing at Rome: A Study in Social Relations,” in
The Family in Ancient Rome: New Perspectives, ed. Beryl Rawson (Ithaca, N.Y.:
Cornell University Press, 1984), pp. 201–29.
35. J. L. Brown and A. Eklund, “Kin Recognition and the Major Histo-
compatibility Complex: An Integrative Review,” American Naturalist, vol. 143
(1994), pp. 435–61.
36. D. Penn and W. Potts, “How Do Major Histocompatibility Genes
Influence Odor and Mating Preferences?” Advances in Immunology, vol. 69
(1998), pp. 411–36.
37. C. Wedekind and S. Füri, “Body Odour Preferences in Men and Women:
Do They Aim for Specific MHC Combinations or Simply Heterozygosity?”
Proceedings of the Royal Society of London, Series B—Biological Sciences,
vol. 264 (1997), pp. 1471–79.
38. V. Apanius, D. Penn, L. R. Ruff, and W. K. Potts, “The Nature of Selection
on the Major Histocompatibility Complex,” Critical Reviews in Immunology,
vol. 17 (1997), pp. 179–224.
39. D. Penn and W. Potts, “The Evolution of Mating Preferences and Major
Histocompatibility Complex Genes,” American Naturalist, vol. 153 (1999),
pp. 145–64.
40. C. Wedekind, T. Seebeck, F. Bettens, and A. Paepke, “MHC-Dependent
Mate Preferences in Humans,” Proceedings of the Royal Society of London,
Series B—Biological Sciences, vol. 260 (1995), pp. 245–49; Wedekind and Füri,
“Body Odour Preferences.”
108 scheidel

41. For positive results, see C. Ober, L. R. Weitkamp, N. Cox, H. Dytch,

D. Kostyu, and S. Elias, “HLA and Mate Choice in Humans,” American Journal
of Human Genetics, vol. 61 (1997), pp. 497–504 (and cf. E. Genin, C. Ober,
L. Weitkamp, and G. Thomson, “A Robust Test for Assortative Mating,”
European Journal of Human Genetics, vol. 8 [2000], pp. 119–24). For negative
findings, see P. W. Hedrick and F. L. Black, “HLA and Mate Selection: No
Evidence in South Amerindians,” American Journal of Human Genetics, vol. 61
(1997), pp. 505–11; Y. Ihara, K. Aoki, K. Tokunaga, K. Takahashi, and T. Juji,
“HLA and Human Mate Choice: Tests on Japanese Couples,” Anthropological
Science, vol. 108 (2000), pp. 199–214; cf. also K. Jin, T. P. Speed, and G. Thom-
son, “Tests for Random Mating for a Highly Polymorphic Locus: Application of
HLA Data,” Biometrics, vol. 51 (1995), pp. 1064–76.
42. Cf. Penn and Potts, “Major Histocompatibility Complex Genes,” p. 416.
43. K. Yamazaki, G. K. Beauchamp, K. Kupniewski, J. Bard, L. Thomas, and
E. A. Boyse, “Familial Imprinting Determines H-2 Selective Mating Preferences,”
Science, vol. 240 (1988), pp. 1331–32; A. Eklund, “The Effect of Early Experience
on MHC-Based Mate Preferences in Two B10.W Strains of Mice (Mus domesti-
cus),” Behavior Genetics, vol. 27 (1997), pp. 223–29. Cf. also K. F. Arcaro and
A. Eklund, “A Review of MHC-Based Mate Preferences and Fostering Experi-
ments in Two Congenic Strains of Mice,” Genetica, vol. 104 (1998), pp. 241–44.
44. D. Penn and W. Potts, “MHC-Disassortative Mating Preferences Revised
by Cross-Fostering,” Proceedings of the Royal Society of London, Series B—
Biological Sciences, vol. 265 (1998), pp. 1299–1306.
45. Schneider and Hendrix, “Olfactory Sexual Inhibition.”
46. Wolf, Sexual Attraction, p. 183; H. Gates in Chapter 8 of this volume.
47. Scheidel, “Brother-Sister and Parent-Child Marriage in Ancient Egypt
and Iran”; “Brother-Sister and Parent-Child Marriage in Premodern Societies.”
6 From Genes to Incest Taboos
t h e c ru c i a l s t e p

Neven Sesardic

Not in Our Genes?

Today the idea that an evolutionary approach may be fruitful for re-
search in the social sciences is being passionately defended by some and no
less passionately contested by others. The resistance to Darwinism comes
mainly in two distinct varieties. The first type of criticism is based on empir-
ical or methodological objections against the current attempts to use evolu-
tionary considerations to throw some light on social science explananda.
The other line of opposition, however, is much harder to pin down and dis-
cuss because it is fueled more by rhetoric than by argument. It defines itself,
rather vaguely, as a fight against “biological reductionism” and “genetic de-
terminism” and is often accompanied by slight (or not so slight) ideological
overtones. In this chapter, I will deal only with the former (methodological)
kind of criticism. But since I don’t want to leave the latter, hazily antireduc-
tionist source of opposition to biology without comments, and since I don’t
know how to approach it in a serious way, let me wiggle out by presenting
to you a rhymed parody, “Gene-mania,” that captures some of the more
ideological criticism’s characteristic flavor:
gene-mania
Who today is not sick and tired
Of all those guys so gene-inspired?
They find a gene for every this or that:
For being gay, smart, alcoholic, fat . . .
We have to stop that madness. Take no offense,
But this approach doesn’t make much sense.
True, some fools thought after Watson-Crick
That genes could really do the trick.
What they sought they did not find—
Those ill-fated biologists of the mind.
110 sesardic

Loonies still insisting on double helix

Must be cured of that idée fixe.
Now we know better, we’ve been instructed:
Human phenotypes are socially constructed.
“Darwinism, yes,” we exclaim with glee,
Adding a proviso: “Only to some degree!”
It is quite all right for a fly or bird,
There the gene-talk is not at all absurd.
But, wait, don’t rush to generalize
From these creatures of smaller size.
Looking at Drosophila melanogaster,
Please, don’t read too much into it, buster.
Well, they rub their genitals, no doubt,
But that’s nothing to get hot about;
You must be completely off your tracks
If you think that what they have is—sex.
Your anthropomorphism and sex obsession
Deserve of course our full compassion;
Yet for the prejudices so amazing
You badly need some consciousness-raising.
Besides, to tell you frankly, there’s another thing,
It all too much smacks of—hmm, the right wing.
Hence in Boston, Stanford or Minneapolis
We might well need the thought police.
Why? Because all this genetic chitter-chatter
Is certainly not a laughing matter.
So be responsible, mind what you say.
Danger! You are talking DNA!

Incest as a Test Case

To what extent is evolutionary biology relevant for social science?

The battle over this question is being fought on many fronts, and the dis-
cussion encompasses a number of specific topics. The issue of incest taboos
stands out though. It is typically regarded as the critical case for evaluating
theoretical aspirations of sociobiology and evolutionary psychology. For ex-
ample, both in his book Consilience and in several subsequent interviews,
Edward O. Wilson illustrates the success of sociobiology with the example
of Westermarck’s biological account of incest prohibitions.1 Richard Daw-
kins also confirms the special place that this theme occupies in contempo-
rary debates about biology and culture: “Although I usually resist the temp-
tation to indulge in simple ‘selfish gene’ explanations of the social behavior
of humans and other domestic animals, inbreeding avoidance is the one
case for which I feel reasonably confident. ‘The social science orthodoxy’
 From Genes to Incest Taboos 111

has always seemed to me particularly daft in this area.”2 John Maynard

Smith echoes Dawkins’s thought: “To me, the most interesting question is
how far evolutionary biology can contribute to the human sciences. As I
have explained, I am a doubter. But I have been wrong on this issue before.
Ten years ago I regarded incest avoidance as an entirely cultural phenome-
non; only a bigot could hold this view today.”3

Three Claims About the Aversion

In an earlier publication I suggested that Westermarck’s theory about

incest is best broken into three component claims:
1. A sexual aversion tends to develop between those raised together in
early childhood.
2. This aversion is an evolutionary adaptation (serving as a barrier to
inbreeding depression).
3. The aversion causes (expresses itself as) the incest prohibition.4
Bill Durham has usefully dubbed these three claims (1) the aversion hy-
pothesis, (2) the adaptation hypothesis, and (3) the expression hypothesis.5
Each of these claims says something about the sexual aversion: (1) The
adaptation hypothesis says that the sexual aversion exists; (2) the adapta-
tion hypothesis says that the cause of the sexual aversion is natural selec-
tion; and (3) the expression hypothesis says that the effect of the sexual
aversion is the social prohibition of incest. I agree with Durham that the
aversion hypothesis has the strongest empirical support and that the ex-
pression hypothesis is the weakest link in Westermarck’s line of argument.
One reason for problems with the expression hypothesis is that a lot of pre-
liminary analysis and clarification has to be done before one can embark
on the straightforward task of empirical evaluation or hypothesis testing.
The complexity of these anthropological issues naturally attracts the visi-
tors whose professed job is precisely to untangle difficult conceptual and
methodological puzzles: philosophers. But scientists do not always warmly
welcome philosophers to join the discussion. And, as it happens, this in-
hospitableness is not wholly unjustified.
Steven Weinberg once wrote, “The insights of philosophers have occa-
sionally benefited physicists, but generally in a negative fashion—by pro-
tecting them from the preconceptions of other philosophers.”6 In the debate
about incest as well, it seems to me that, unfortunately, some philosophical
contributions have clouded the issues instead of elucidating them. I hope,
however, that the analysis undertaken in the next section will be recognized
as containing more than just this purely negative result (the criticism of other
philosophers’ views).
112 sesardic

Inhibition and Prohibition: A Difference in Content?

The crucial idea in Westermarck’s account of incest taboos is that it

is the biological inhibition that gives rise to a cultural prohibition. It is
quite clear that this hypothesis is fraught with difficulties and that many
questions have to be answered before the argument that biological inhibi-
tion leads to cultural prohibition becomes persuasive. But some thinkers
want to go much further than just pointing to that explanatory gap; they
claim that a very general reason blocks in advance any transition from bio-
logical inhibition to cultural prohibition and that we should acknowledge
that even without looking into empirical issues. Bernard Williams refers to
this transcendental obstacle as the “representation problem.”
Not only does extra conceptual content have to be introduced to characterize the
human prohibition, but also the introduction of that content stands in conflict with
the proposed biological explanation of it. . . . There are no sanctions against marry-
ing those that one is brought up with (as such); the sanction is against marriages
which would constitute close in-breeding. The conceptual content of the prohibition
is thus different from the content that occurs in the description of the inhibition. It
indeed relates to the suggested function of that inhibition, but that fact will not ex-
plain how the prohibition which is explicitly against in-breeding will have arisen. It
certainly does not represent a mere “raising to consciousness” of the inhibition.7

Although Williams’s idea reduces to a simple conceptual point, it still

promises to settle an empirical issue; it is offered as a reason why we are
entitled to dismiss a scenario in which cultural prohibition is seen as a mere
manifestation of biological inhibition. Despite being quite influential and
even making an impact on some scholars conducting empirical research on
incest, this piece of a priori anthropology is fundamentally flawed. Since I
have criticized it extensively elsewhere, let me here explain my reason for
disagreeing with Williams more briefly and in a slightly different form.
Westermarck’s basic idea is that for evolutionary reasons (i.e., the danger
of inbreeding) a sexual aversion tends to develop between those who hap-
pened to be close associates in the crucial period of early childhood. Why
does an aversion between childhood associates exist? Well, the hypothesis is
that during the long time that natural selection has molded human psychol-
ogy, close childhood associates were in fact almost always siblings. One ob-
vious reason why it would be difficult for the aversion to be focused directly
on siblings is that the concept of sibling is a fairly complicated social cate-
gory—apparently an unlikely kind of object to which a genetically pro-
duced aversion might be immediately attached. Therefore, if there is a co-
extensiveness (or very strong correlation) between these two properties,
childhood associates (CA) and siblings (S), then natural selection might
have found it easier, although no less effective as an instrument against in-
 From Genes to Incest Taboos 113

S S
Prohibition
CA CA

S S
Aversion
CA CA

figure 6.1. Aversion and Prohibition: Mismatch

breeding, to instill a sexual aversion between childhood associates (and not

between siblings as such). Of course, the assumed strong correlation be-
tween CA and S would mean that as a consequence there would be a sexual
aversion between siblings as well, but still, sensu stricto, it would only be
correct to say that, in causal terms, what was producing the aversion was
the characteristic CA, not S! This is readily confirmed in a pair of counter-
factual or presumably rare exceptional situations in which the correlation
between CA and S is violated: (1) As long as the CA-relation obtains, the
aversion would still be there, even without S; (2) despite the presence of S,
the aversion would not be there if the S in question failed to be CA.
On the basis of this (correct) description of the situation, Williams
mounts an attack on the suggestion that the aversion could produce the in-
cest prohibition. In Figure 6.1 the contrast between the aversion and the
prohibition comes to the fore. Typically, two individuals of the opposite sex
are both childhood associates (CA) and siblings (S), but the reasons they
are kept apart are different in the two cases (aversion and prohibition).
Although as a rule it is the same individuals that are sexually kept away
from one another by the aversion and by the prohibition, Williams wants
us to notice that the forces of separation in these two cases are essentially
different. The aversion is produced by a force opposing a CA-relationship,
whereas what the prohibition condemns is just an S-relationship. Ergo, the
argument goes, the prohibition cannot be a mere expression of the aversion
because the two are directed to two entirely different aspects. The fact that
CA-relationships are usually, or perhaps always, also S-relationships (and
vice versa) is neither here nor there. Surely even after we concede that the
two characteristics are correlated, it remains quite unclear by what kind of
transformation the content of the prohibition that speaks only about S
could be suddenly obtained from the mere aversion toward CA.
Williams’s argument looks persuasive because it trades on a crucial am-
114 sesardic

biguity. The statement that the aversion is directed toward CA (childhood

associates) can actually mean two things. First, in the causal sense, it can
mean that the characteristic that is actually producing the aversion is CA.
Second, in the subjective sense, it can mean that people who have the aver-
sion experience it as being directed toward CA. I would like to point out
that there is no necessary connection between the two senses; they are dis-
tinct and separable. In particular, if the aversion is directed toward CA in
the causal sense, it by no means follows that the subjects would also expe-
rience it as being directed toward CA. They may well believe (wrongly)
that the aversion is, say, generated by S.
The causal sense is relevant for the natural selection scenario. And yes,
Westermarck’s theory does contain a claim (the adaptation hypothesis)
that the sexual aversion is objectively produced by being triggered by CA-
aspect. The expression hypothesis, on the other hand, says that the prohi-
bition is a manifestation of the inhibition. But here, in the context of the
expression hypothesis, the content of the inhibition (that gives rise to the
prohibition) is determined by the subjective sense of “being directed to,”
and not the causal sense! The fact that in the causal sense the aversion is
indeed directed to CA (in contrast to the prohibition, which is directed to
S) does not imply that there is a “mismatch” between the two contents,
simply because the causal sense of the aversion’s “being directed to” does
not speak about content at all. It only speaks about which property is ob-
jectively causing the aversion. The content emerges only at a later stage
when we ask how the people having the aversion experience it subjectively,
that is, in terms of which property they conceptualize their own aversion.
So it seems that Figure 6.1 was an oversimplification. It should be replaced
by Figure 6.2.
Now we see that the fact that the aversion objectively picks out child-
hood associates as its object does not necessarily signify that its content is
discrepant from the content of the prohibition (which forbids sex with sib-
lings). For the content of the aversion is not fixed by such objective matters
as the question about what is causing what. On the contrary, it essentially
depends on how the aversion looks “from the inside” to those people who
have it. And for all that matters, there is no guarantee here at all that there
will be an accord between the objective and the subjective. In fact, given
that natural selection found it easier to achieve its end (decreasing the
probability of sex between siblings) by directing the aversion toward the
more accessible and conveniently correlated proxy (childhood associates),
it is perfectly possible that people, too, picking out environmental cues to
make sense to themselves about their own aversion, also choose the line of
least resistance and simply reach for the more meaningful, social category
like brother or sister, rather than a seemingly irrelevant and queer charac-
 From Genes to Incest Taboos 115

S S
Prohibition
CA CA

S S Aversion
(causal
CA CA picture)

S S Aversion
(subjective
CA CA side)

figure 6.2. Aversion and Prohibition: No Mismatch

teristic like someone with whom I spent the first years of my childhood.
Notice that I say that this is possible, not true. My concern here is not the
truth, but the criticism of Williams’s “impossibility proof.” I tried to show
that the transcendental obstacle he placed in the path of the Westermarck-
ian account can be removed.
Williams hoped to demonstrate on very general grounds that the biolog-
ical aversion couldn’t produce the cultural prohibition. But his argument is
based on the unjustified assumption that the content of the inhibition has
to be presented to the subjects in terms of that property that is in fact
causally operative in producing that inhibition. Once we realize that this is
not necessary at all, the argument breaks down. The inhibition may well be
subjectively presented differently (in a way that does not reflect its actual
causal origin) and then, distorted in the “right” way, it may well match in
content the corresponding cultural prohibition.

Westermarck’s Argument as a Zero-Sum Game

There is another, again very general argument that is proposed as

undermining the theory that the aversion gives rise to the prohibition (the
expression hypothesis).
To make matters worse, there is little or no direct support for Westermarck’s
“moral disapproval” step. First, of the three studies already cited as supporting
116 sesardic

Westermarck’s argument about intimacy and aversion, not one shows evidence for
a moral disapproval expressive of that emotion. . . . Any one of these cases might
be taken as exceptional, and therefore dismissed from concern. But it is hard to ac-
cept that argument for all three, especially since the most familiar of all potential
partners are not prohibited. In effect, the available evidence says, “Aversion, yes—
moral disapproval, no.”8

This is not a conceptual or transcendental argument. Durham’s objec-

tion is based on probability. As he correctly notes, the existence of the
Westermarckian sexual aversion between close childhood associates gets
empirical support basically from three studies conducted in three different
environments (Israel, Taiwan, and Lebanon). In all these cases there were
clear signs of the lack (or decrease) of sexual attraction between childhood
associates, although sexual contacts between the individuals in question
were not prohibited but even encouraged. Durham then stresses one thing.
These three studies are the best available evidence for the existence of the
aversion. But in none of the three cases did the aversion lead to the prohi-
bition. Why is that? He suggests that if the prohibition failed to emerge in
only one of those situations, then it could still be reasonable to cling to the
expression hypothesis and say that something exceptional occurred in that
single instance that disrupted the usual tendency of the aversion to produce
the prohibition. But if the same thing happened in all the three cases (as it
did), he argues that this should make us strongly doubt, on purely induc-
tive grounds, that there is any causal connection between the aversion and
the prohibition.
I disagree. In my opinion, even the three repeated instances of “aversion,
yes—moral disapproval, no!” do not justify skepticism with respect to the
expression hypothesis. Contrary to Durham, I do not think that the three-
fold presence of aversion-sans-disapproval points toward the conclusion
that the former does not produce the latter. True, the fact that in those
three quite vital cases for the evaluation of Westermarck’s theory the aver-
sion is not accompanied by moral disapproval cannot be dismissed as a
mere coincidence (or “a few exceptions that prove the rule”). The absence
of moral disapproval needs to be explained. One reason why the aversion
is there, but moral disapproval is not, may indeed be that there is no in-
trinsic connection between the two and that in those other cases when they
happen to be found together they are actually the result of different and
largely independent causal processes. This is Durham’s tack: if A is not fol-
lowed by B in the three key cases, this undercuts (at least to some extent)
the hypothesis that A produces B. But there is another way to understand
why the aversion appears without moral disapproval in those three test
cases. Maybe the absence of moral disapproval does not indicate that there
is no general causal connection (aversion taboo); perhaps it’s just a
 From Genes to Incest Taboos 117

sign that we are here dealing with a particular kind of situation in which,
by a very special logic of hypothesis testing, the aversion could not have
produced the taboo. Let me explain.
Westermarck’s aversion hypothesis asserts that there is an inborn sexual
aversion that spontaneously develops between siblings, qua close childhood
associates. The hypothesis faces an obvious difficulty. Namely, since as a
matter of fact the aversion is usually correlated with the corresponding cul-
tural taboo against sexual contacts between siblings, how do we know that
the aversion is not simply a consequence of the taboo (rather than being an
inborn, biologically mediated inhibition)? Well, fortunately for the aversion
hypothesis, there are three extensively researched cases in which there is no
taboo but the aversion still seems to be there.9 These cases constitute critical
empirical support for the aversion hypothesis. But, somewhat mischievously,
these same cases undermine the expression hypothesis. To spell it out, on
one hand, if the correlation between the aversion and taboo is occasionally
broken, this is a welcome result for the aversion hypothesis because the
aversion occurring without the taboo shows that the aversion stands by it-
self and is not just a side effect of the taboo. But on the other hand, the aver-
sion occurring without the taboo is bad news for the expression hypothesis;
it is weakening the hypothesis because what the hypothesis basically says is
that, other things being equal, the aversion leads to the taboo. So, Wester-
marck’s global theory about incest has two components, the aversion hy-
pothesis and the expression hypothesis, that pull empirical evidence in the
opposite directions. For this reason, his global theory is in a strange episte-
mological predicament in that, under the circumstances, it just cannot re-
ceive full empirical confirmation. It is a zero-sum game; what the theory
gains by collecting evidence in favor of the aversion hypothesis it automati-
cally loses on the other front because the very same empirical data chip away
at the expression hypothesis.
Because of the zero-sum nature of Westermarck’s argument, we should
not be much impressed by the fact that in every single case that supports
the existence of the biological aversion the taboo is missing. This is dictated
by the logic of the situation.
Referring to Figure 6.3 and the two circles, the area of intersection rep-
resents the presence of both the aversion and the taboo. These cases are
consistent with the expression hypothesis, but they cannot provide evi-
dence for the aversion hypothesis. If all the data were in that area, the aver-
sion hypothesis would be considerably weakened because then it would
make sense to hypothesize that the aversion is produced by the taboo and
is not an independent psychological phenomenon. However, the area where
aversion occurs without the taboo sends an opposite epistemological mes-
sage. The data points located in that section support the theory that the
118 sesardic

Aversion Taboo

aversion aversion
without and
taboo taboo
Supporting Congruent with
the aversion the expression
hypothesis hypothesis
but weakening but weakening
the expression the aversion
hypothesis hypothesis

figure 6.3. The Expression Hypothesis and the Aversion Hypothesis:

An Epistemological Tension

aversion is an autonomous event and not a mere offshoot of the taboo. But
at the same time this break of connection between the aversion and the
taboo creates a problem for the expression hypothesis, which states that
the aversion regularly brings about the taboo.
So, there is nothing puzzling about the fact that all three cases support-
ing the aversion hypothesis speak against the expression hypothesis. In-
stead of being regarded as a consilience of independent cases strongly
pointing toward the probable falsity of the expression hypothesis, this fact
is better seen as just reflecting the logical peculiarity of Westermarck’s the-
ory. Although the two components of his theory, the aversion hypothesis
and the expression hypothesis, are perfectly compatible and mutually con-
sistent, there is an epistemological tension between them in that, at least
at the present stage of theory testing, empirical evidence cannot support
the aversion hypothesis without raising some doubts about the expression
hypothesis.
For the purpose of illustration, let me give another example that comes
from an entirely different context but which, analogously, exhibits the same
kind of epistemological tension. Take the following hypothesis: “Many peo-
 From Genes to Incest Taboos 119

ple who approach the state of clinical death have the same strange experi-
ence: it appears to them that they perceive the world from a location out-
side their own bodies.” Let’s call this hypothesis NDE (near-death experi-
ence). Take now another hypothesis: “Usually, a near-death experience is
immediately followed by death.” Let’s call this latter hypothesis DSA (death-
soon-afterward).
It is quite clear that, from a logical point of view, NDE and DSA are
fully consistent with one another. But here again, as in the case of Wester-
marck’s two hypotheses, there is an epistemological strain. For, if the evi-
dence in favor of DSA becomes too strong, then as a direct consequence of
that, NDE will lose much of its empirical support. Namely, if the regularity
postulated by DSA were perfect and exceptionless, and if, accordingly, all
cases of NDE were immediately followed by death, there would be simply
no one left to tell us about the existence of NDE! On the other hand, and
this is the key point, if there happened to be some cases where, contrary to
the general tendency expressed in DSA, the subjects of NDE survived and
were able to report later about their experiences, it would be wrong to use
this to attack DSA, along the lines reminiscent of Durham’s argument in
connection with Westermarck. That is, it would be wrong to say, “Of all
the cases supporting NDE, not one shows evidence for DSA. Any one of
these cases might be taken as exceptional, and therefore dismissed from
concern. But it is hard to accept that argument for all of them.”
True, all the cases supporting NDE undermine DSA, or at least weaken
it to some extent. But this, in itself, should not be interpreted as a con-
silience of several independent instances, pointing to the probable falsity of
DSA. Rather, the fact that all the cases corroborating NDE do clash with
DSA is better regarded as just trivially following from the zero-sum logic
of confirmation involving these two hypotheses. To put it differently, if
from the outset it were quite open whether the data supporting NDE will
happen to be in accord with DSA or not, and if it then turned out that all
of them were aligning themselves in the opposition to DSA, this could in-
deed be reasonably taken as accumulation of important negative evidence
against DSA. But this is not how things are. Actually, we know well in ad-
vance that any confirmation of NDE (i.e., an ex post facto first-person re-
port about NDE) will inevitably be a counterexample to the regularity as-
serted in DSA (“NDE is immediately followed by death”). For this reason
it is misconceived to ask, “Why is it that not just one or two cases con-
firming NDE undermine DSA, but all of them do?” The answer is, It could
be no other way. The data speak here with one voice because of the episte-
mological peculiarity of the situation. This is not a probabilistic indication
of anything.
120 sesardic

Conclusion

I tried to show that Westermarck’s theory about incest is not threat-

ened by the two very general methodological objections directed against it.
But I hope that amid the critical tones, which dominate this chapter, a more
positive message will be recognized too. Namely, the epistemological
scrutiny of Westermarck’s views that is here undertaken reveals a theoreti-
cal edifice of great complexity and conceptual sophistication. Although to-
day the main efforts seem to be focused on the attempts to evaluate empir-
ical evidence that will ultimately decide the conflict between the biological
account of incest taboos and its rivals, at those moments when we are
forced to stand back and inspect the logical structure of Westermarck’s the-
ory, we realize that some of its implications, interconnections, and “forking
paths” have yet to be fully explored.

notes
1. Edward O. Wilson, Consilience (New York: Knopf, 1998); Edward O.
Wilson, “Resuming the Enlightenment Quest,” Wilson Quarterly (1998).
2. Richard Dawkins, “Opportunity Costs of Inbreeding,” Behavioral and
Brain Sciences, vol. 6 (1983): pp. 105–6.
3. John Maynard Smith, “Constraints on Human Behavior,” Nature, vol. 276
(1978): p. 121.
4. Neven Sesardic, “From Biological Inhibitions to Cultural Prohibitions:
How Not to Refute Edward Westermarck,” Biology and Philosophy, vol. 13
(1998): p. 224.
5. See William H. Durham, Chapter 7 in this volume.
6. Steven Weinberg, “Against Philosophy,” in Dreams of a Final Theory
(London: Hutchinson, 1993), p. 132.
7. Bernard Williams, “Evolution, Ethics, and the Representation Problem,”
in Making Sense of Humanity and Other Philosophical Papers, 1982–1993
(Cambridge: Cambridge University Press, 1995), pp. 105–6.
8. William H. Durham, Coevolution (Stanford, Calif.: Stanford University
Press, 1990), p. 323.
9. However, for serious methodological doubts about one of these cases
(the alleged evidence from Israeli kibbutzim), see Jonathan Hartung, “Review of
Incest: A Biosocial View,” American Journal of Physical Anthropology, vol. 67,
no. 2 (1985): pp. 169–71.
7 Assessing the Gaps in
Westermarck’s Theory

William H. Durham

The last two decades of the twentieth century were just as kind to Ed-
ward Westermarck as the first two decades were harsh. The skepticism and
dismissal that plagued Westermarck’s incest theory in the wake of early cri-
tiques by Sigmund Freud and Sir James Frazer have given way to a recent
groundswell of empirical validation and approbation. Today, Westermarck’s
theory is often held up as paradigmatic of current understandings of the re-
lationship between genes and culture in human evolution. Frans de Waal,
for example, writes that the Westermarck effect—that is, the absence of sex-
ual interest between adults who were reared together as young children—
serves as “a showcase” of new Darwinian approaches to human behavior.1
Other authors claim that sibling incest avoidance not only vindicates West-
ermarck but also shows that “a tight and formal connection can be made
between biological evolution and cultural change.”2 And in his recent trea-
tise on the unity of knowledge, E. O. Wilson notes that Westermarck’s ar-
gument is, simply, “the current explanation” of incest avoidance.3 Outside
observers of this changing tide could well be given the impression that West-
ermarck’s theory has triumphed of late over all alternative explanations for
the incest taboo and that we are dealing at the turn of the century with an
open and shut case for Westermarck.
Such a conclusion would be grossly misleading. In the rush to vindicate
Westermarck, touched off by careful documentation of the Westermarck ef-
fect (described by Patrick Bateson in Chapter 1 and Arthur Wolf in Chap-
ter 4), it has proved easy to forget that Westermarck’s is a complex theory
containing at least three main hypotheses. Empirical support for one of the
pieces of that theory, even if it is the crucial starting piece, is not the same
as support for all three. For Westermarck to sweep into the new millen-
nium with full validation, we would need equivalent evidence to support
all three main hypotheses:
122 durham

1. The aversion hypothesis, or “Westermarck effect”—namely, the

proposition that “an innate aversion to sexual intercourse [develops]
between persons living very closely together from early youth”
(Westermarck 1891: 320).
2. The adaptation hypothesis—the proposal that the aversion is an
evolutionary adaptation, shaped by natural selection specifically for
the observed function of reducing inbreeding, and not some incidental
by-product of the nervous system (and thus an “exaptation” in
contemporary evolutionary parlance).
3. The expression hypothesis—the claim that the aversion directly
causes the incest prohibitions of human societies; in Westermarck’s
words, “aversions which are generally felt readily lead to moral dis-
approval,” and do so by “displaying” or “expressing” themselves “in
custom and law as a prohibition of intercourse between near kin.”4
There are probably many reasons why the first hypothesis on this list has
received the lion’s share of recent attention. For one thing, it flies in the face
of the long-standing Freudian position on sibling attraction and has thus re-
quired thorough documentation even to budge the skeptics. As Arthur Wolf
notes (in the Introduction to this volume), hypothesis one is thus a direct
challenge to the received wisdom of earlier generations on the subject of in-
cest. For another thing, the Westermarck effect lends itself fairly readily to
empirical testing. A number of researchers (including Wolf, Joseph Shepher,
and Justine McCabe) have investigated social settings offering some form of
“natural experiment” in which cosocialized children are both reared to-
gether and allowed—indeed encouraged—to have sexual relations and /or
marry.5 Hypotheses two and three are much harder to get at in this way, and
there has been much less concerted effort to do so. Understandably, number
one carries the day.
But I remain concerned about hypotheses two and three. Before we usher
in Westermarck’s theory as the new crown prince of incest theory, and espe-
cially before we build from it toward a full-blown “biological science of
ethics” (see Chapter 10), these two hypotheses seem to me to warrant the
same kind of scrutiny given hypothesis one. Number two concerns me be-
cause the Westermarck effect is increasingly assumed to be an evolutionary
adaptation, an assumption that renders it eligible for the “just so story” cri-
tique.6 Its position is especially precarious because there are, at first glance
anyway, other arguments—habituation among them (see Chapter 1)—to ac-
count for the deeroticization of cosocialized children. What is lacking is con-
clusive evidence to show that the aversion was specifically shaped over time
by genetic selection for the function it now performs. While we await such
evidence, it would be prudent not to rush the crown prince to the throne.
Parallel concerns apply to the expression hypothesis, but even more so.
 Assessing Gaps in Westermarck’s Theory 123

For example, one could well argue that it doesn’t really matter if Wester-
marck was right about the aversion being an evolutionary adaptation. As
long as the nervous system delivers the effect, however it is achieved, West-
ermarck could still be correct in arguing that the aversion is the basis for the
incest taboos of humanity. The crown would still be his. But the situation is
not the same for the expression hypothesis. Westermarck could well be right
about hypotheses one and two and still be fully wrong about what causes
the incest taboos of humanity. To date, we have little more than mere argu-
ments for the expression hypothesis. As rich with reason and logic as they
may be, arguments are not the same as data. What is needed is empirical ev-
idence that the internal, individual reaction of aversion (whether adaptation
or not) “displays itself in custom and law as a prohibition to intercourse be-
tween near kin.”7 We need convincing demonstration that “the law” exists
because it “expresses the general feelings of the community and punishes
acts that shock them.”8 I argue that the crown should be held in reserve un-
til these data are delivered.
In the remainder of this chapter, I would simply like to expand on these
concerns about hypothesis three and the reasons they persuade me toward
caution today, in the face of what otherwise looks like a headlong rush for
Westermarck.

The Evidence Problem

In Westermarck’s formulation, the incest taboos of humanity are a

direct social manifestation of the aversion—they are the aversion as it “ex-
presses itself in custom and law.” The causal chain is straightforward and
direct; the incest taboo culturally encodes the aversion people feel and it
punishes others—at least it threatens to—if they carry out acts against
which most members of the community are averse. To Westermarck, the
aversion is thus the basis for the moral disapproval of incest within any
given community; the reason people oppose and prohibit sexual acts be-
tween certain categories of kin is the discomfort they feel when others carry
out those acts. In parallel to the “Westermarck effect,” let me call this the
“Westermarck process”: the hypothesis that people actively condemn in
others the sexual actions toward kin that arouse the aversion in them.
At first glance, one would think it ought to be fairly easy and straightfor-
ward to test the Westermarck process against empirical data. The questions
seem simple enough: Do people actively oppose the sexual acts of others
that make them feel aversive discomfort? Does a given community’s disap-
proval of incest spring from these individual reactions? Does this aversion-
based disapproval shape and sustain the incest taboos of humanity? In prac-
tice, however, tests have proved difficult; certainly there have not yet been
124 durham

enough efforts in this direction. There are, I think, only three things we can
confidently say these days about testing the Westermarck process.
First, as Neven Sesardic has pointed out, not every aversion gives rise to a
prohibition, so there must be something special about the Westermarckian
aversion.9 In fact, human societies appear to harbor quite a number of aver-
sions that remain unmatched by prohibitions: the aversion to eating insects
alive;10 the aversion to drinking fresh milk among nondairying peoples;11 a
widespread aversion to bitter, poisonous substances; a purported aversion to
left-handers by right-handers;12 and so on. If it is truly valid, the Wester-
marck process must have something special about it that does not pertain to
these other aversions. Wolf (1995 and Chapter 4 of this volume), building
on Westermarck’s (1906–8) own arguments about “moral ideas”, suggests
that what is special here is the combination of disinterestedness, apparent
impartiality, and the “flavor of generality.” As Wolf argues,
It is in the nature of inhibitions of this kind to arouse moral disapproval. The inhi-
bition is activated as an aversion when a person expresses sexual interest in an ob-
ject belonging to the same class as those objects toward which most people are in-
different; the pain [or discomfort] produced by the thought of sexual relations with
a sexless object is expressed as disapproval of the person responsible for raising the
possibility; and this disapproval is accepted as moral because, conferring no obvi-
ous advantage, it appears to be disinterested, and, being the reaction of the majority
of the community, it qualifies as impartial and general.13

In other words, inbreeding aversion is a special kind of aversion, and the

Westermarck process is but one example of a more general process behind
the social emergence of moral ideas. It is a lovely and provocative idea, es-
pecially if it succeeded in getting everyone to believe that the expression hy-
pothesis needs no testing because it is merely a special case of an accepted
general phenomenon. The problem is that the argument is based on a West-
ermarck publication from 1906–7 that shows clear signs of “having been
modeled after his [own] earlier theory about incest prohibition.”14 In short,
Westermarck’s whole line of reasoning behind “the origin and development
of moral ideas” still remains to be tested, in general and in the case of the
incest taboo.
A second point to be made is that cross-cultural tests of the Westermarck
process have consistently failed to produce much evidence. Westermarck
was himself the first to try with a test of a corollary argument: namely, the
argument that the range of prohibited relatives in a given society should
closely parallel the group of relatives typically reared together from an early
age. He claimed,
Facts show that the extent to which relatives are not allowed to intermarry is nearly
connected [tightly correlated] with their close living together. Generally speaking,
 Assessing Gaps in Westermarck’s Theory 125

the prohibited degrees are extended much further . . . [in societies where people]
live, not in separate [nuclear] families, but in large households or communities, all
the members of which dwell in close contact with each other. . . . On the other
hand, where families live more separately, such extensive prohibitions to close in-
termarrying do not generally exist.15

In an extension of this reasoning, Westermarck even argued that coresi-

dence and “local relationships” explain why incest prohibitions are “very
often one-sided, applying more extensively either to the relations on the fa-
ther’s side or to those on the mother’s, according as descent is reckoned
through men or women.”16 He offered the general conclusion that “an
abundance of ethnographical facts . . . prove that it is not, in the first place,
by degrees of consanguinity, but by the close living together that pro-
hibitory laws against intermarriage are determined.”17
It was a noble thought, a logical corollary of the Westermarck process,
and a valiant attempt at empirical validation. But it was also wrong. To
Westermarck’s credit, later editions of his book set the record straight; he
dropped the preceding conclusion and noted in its place that taboos “fre-
quently refer to the marriage of kindred alone,” not to persons reared in
propinquity.18 Still in hopes of reconciling this finding with his theory, these
later editions made explicit a couple of additional steps in Westermarck’s
chain of logic. First, he argued that the aversion to sex with close childhood
associates is “interpreted” by people as a feeling against sex with kin. This
“transition” occurs, says Westermarck, because the aversion is supposedly
one of “an immense group of facts which, though ultimately depending
upon close living together, have been interpreted in terms of kinship.”19 Sec-
ond, he added to this transition his famous “law of association,”20 which is
variously seen by scholars today either as a brilliant, symbolic extension or
as one of the more interesting and problematic fudge factors in all of social
science. By this law, the “feelings of intimacy and kinship” from close child-
hood association are transferable to other, nonintimate acquaintances. As a
result, incest taboos, “though in the first place associated with kinship be-
cause near relatives normally live together, have come to include relatives
who do not live together.”21 The transfer is facilitated, says Westermarck,
by symbols—especially names—that have “come to stand” for consan-
guinity. This certainly does help to generalize the Westermarck effect be-
yond the limits of childhood association. But it also makes the argument
much more difficult to test. It becomes something of an open question
whether the resulting prediction—that “the extent of the prohibited degrees
is closely associated with social intimacy, whether combined with actual liv-
ing together or not”22—is even falsifiable.
Meanwhile, a couple of other attempts at cross-cultural tests of the West-
ermarck process have also failed to produce empirical support. The first of
126 durham

these was an ambitious study of first-cousin marriage practices in a global

sample of some 800 populations.23 Reviewed in more detail elsewhere,24
Ember’s study looked for evidence in this sample of societies that childhood
association among cousins was correlated with their cultural prohibition as
marriage partners. His measures of childhood association were, first, com-
munity endogamy (that is, rules about marrying within one’s natal commu-
nity) and, second, community size, on the reasonable argument that cousins
are more likely to be reared in close association when they inhabit the same
village, especially if it is small (i.e., less than 400 members). Granted that
these measures of propinquity were, at best, indirect, the results were still
challenging to Westermarck. First-cousin marriages were actually prohibited
with greater frequency where cousins are routinely precluded from child-
hood intimacy by community exogamy. Moreover, the smaller endogamous
communities (presumably conducive to greater social intimacy) failed to
show a significantly higher proportion of cousin prohibitions than did en-
dogamous communities of a larger size. Ember concluded that the child-
hood association of cousins has little value in predicting their prohibition as
marriage partners.
In a much less ambitious study in the 1980s, I attempted a retest of Em-
ber’s findings in a world probability sample of sixty societies25—a sample
with a higher proportion of independent cases (less of “Galton’s problem”)
than Ember’s much larger sample.26 I also used a measure of the kinship ex-
tension of incest taboos that is far more refined than Ember’s “yes” or “no”
on cousin marriage prohibition.27 To make a long story short, despite my
best efforts, these modifications did little to change the conclusions reached
by Ember. I, too, found that locally endogamous populations, with presum-
ably more opportunity for close relatives to share intimate childhood asso-
ciation, do tend also to be “more accepting of close relatives as sexual part-
ners than do their locally exogamous counterparts.”28 And like Ember, I also
found that small endogamous communities (in which more relatives might
be expected to grow up in close childhood association) tend to have the
least extensive incest taboos. At the same time, ethnographic details in both
studies, Ember’s and mine, make it abundantly clear that there generally is
strong moral disapproval of incest within the societies studied, albeit vari-
ably expressed. There is just no evidence from either analysis that the moral
disapproval of these communities is based on Westermarck’s aversion. By
the same token, both studies did find modest support for an alternative
source of moral community in the study populations, an alternative source
that will be discussed further below. In short, existing cross-cultural studies,
including Westermarck’s own efforts, provide little or no support for the hy-
pothetical Westermarck process.
The third and final point to be made about testing the Westermarck pro-
 Assessing Gaps in Westermarck’s Theory 127

cess concerns Sesardic’s critique (Chapter 6) of my earlier discussion of

three special cases—minor marriages in Taiwan, intra-Kibbutzim marriages
in Israel, and patrilateral parallel cousins in Lebanon—from which we get
the best data for Westermarck’s aversion.29 Upon reviewing each case, I
noted that while all three support “Westermarck’s argument about inti-
macy and aversion, not one shows evidence for a moral disapproval ex-
pressive of that emotion.”30 Sesardic takes me to task for making so much
of this point, since we already know that the aversion could not have pro-
duced a taboo in these cases, or the cases wouldn’t exist. He has a good
point; it is only possible for us to see and document the aversion in cases
where there is no taboo against the aversion-producing relationships in the
first place (if there were a taboo, as this older logic goes, any aversion
might instead be a product of the taboo). Says Sesardic, if you’ve chosen
these cases because the relationships studied are free of taboos, then of
course you can’t then expect the same cases to show evidence for the West-
ermarck process! They were chosen precisely because they have no taboo
for the relationships concerned.
I accept Sesardic’s criticism and agree that the cases can’t be expected to
show both a taboo-free aversion and an aversion-based incest taboo. That
said, the cases remain interesting to this discussion not simply because the
absence of a taboo makes them exceptional. They also stand out because the
aversive couples are actually socially desired marriage partners. To my
mind, it is the positive social valence to these unions that makes them inter-
esting, not the (logically necessary) absence of a negative valence. In the case
of minor marriage in Taiwan, it is clear from Wolf’s accounts that the par-
ents of cosocialized couples have a lot invested and eagerly desire the mar-
riages to work, even to the point of brandishing canes outside bedroom
doors of reluctant newlyweds. In the case of age-mates in the Kibbutzim,
says Shepher, the marriages found to be aversive were actually “preferred by
parents and other members of the kibbutz.”31 And in the case of Lebanon,
McCabe makes it clear that “one of the most salient features of marriage in
the Arab Middle East is the preference for a man to marry his patrilateral
parallel cousin,” despite the aversion the two feel.32 In all three cases, the
aversion is accompanied by some preference for the aversive union. Sesardic
might well come back with the reply that, yes, they almost had to be pre-
ferred forms of marriage in order for them to reach frequencies adequate for
scholars to be able to measure aversion effects. But look at what else that
tells us. The existence of social preferences for marriage between aversive in-
dividuals tells us that the hypothesized Westermarck process can be overrid-
den by other, everyday social phenomena. In at least three well-documented
cases, any aversion-based disapproval is overwhelmed and turned right
around, such that aversive unions are coupled with community moral ap-
128 durham

proval. Now, Wolf might well come back and add that this is not all that
surprising, because aversions developing out of childhood association may
be “typically experienced in kinship terms” (from the Introduction to this
volume). In both Taiwan and Israel, the childhood associates are not kin
and thus may be exempt from a disapproval experienced in kinship terms.
But what about Lebanon? In Lebanon, the aversive individuals are certainly
kin (first cousins). What is more, in Lebanon, the turnaround is nothing
new. For hundreds of years, maybe even thousands, the aversive unions of
close cosocialized kin have met with enduring community approval. To
quote Westermarck once again, the aversion here simply does not “display
itself in custom and law as a prohibition to intercourse between near kin.”
To my mind, all three exceptional cases, and especially the Lebanese exam-
ple, raise the question of the relative strength of aversion-based disapproval
compared to other ordinary social forces in human communities. If rou-
tinely overridden in Lebanon, are we to suppose it is strong enough in all
other times and places to deliver the nearly ubiquitous incest taboos of hu-
manity? I remain to be convinced.

The Burton Problem

But there is one other problem, perhaps more serious than evidence
issues, suggesting that it may still be premature to hand Westermarck the
incest crown. This is simply the problem of an alternative hypothesis that
both (1) links the Westermarck effect with incest taboos in another way,
and (2) garners enough empirical support of its own to be difficult to dis-
miss out of hand. Certainly the alternative has more supporting cases than
there are with clear evidence for the Westermarck process. I call it the
“Burton problem” because psychiatrist Roger Burton was, I believe, the
first author to formulate the argument parallel to Westermarck’s theory, be-
ginning with the aversion but running right on through to moral commu-
nities that sustain incest taboos primarily for other reasons.33 I confess at
the outset that I was struck many years ago by both the simplicity of the
argument and its base of empirical support. I was persuaded at that time to
carry out a related analysis of the Burton argument, and thus I do not pre-
tend to be nonpartisan in this debate.34 However, I also do not claim that
the Burton problem is the only conceivable alternative explanation to West-
ermarck’s. Other alternatives may well surface in the years ahead, perhaps
more worthy than any of our current “short list” of theories. That fact
should surely be borne in mind before we rush the crown out to Wester-
marck, or to his competition for that matter.
The Burton problem, then, is basically the problem (from Westermarck’s
 Assessing Gaps in Westermarck’s Theory 129

vantage point) of an alternative explanation linking aversion and the incest

taboos of humanity. The alternative occurred to Burton in the course of a
cross-cultural study on a different topic, resistance to temptation, in which
he was given “the impression that the most common reason given in both
primitive and modern societies for the incest taboo is that [inbreeding]
produces bad stock.”35 It occurred to Burton that there may well be some-
thing more than coincidence or superstition to these reports and that the
local observation and experience of genetic effects of inbreeding could well
affect the scope (i.e., the extension beyond the nuclear family) and persis-
tence of human incest taboos.
To Burton’s credit, he realized that it was precisely in societies where
Westermarck’s aversion kept childhood associates at bay that the reduced
frequency of inbred unions would allow for the evolutionary accumulation
of deleterious genes in a population.36 In a sense, Westermarck’s starting
hypothesis is what set up Burton for a rival theory of four parts:
1. Westermarck’s aversion would guarantee that inbreeding was gener-
ally rare in local gene pools but commonly deleterious and visible when
it occurred. Just how deleterious and visible remains a matter of some
debate (see Chapter 2, for example), but Burton reasonably assumed that
the additional mortality and morbidity from inbreeding among close
relatives (first cousins and closer, say) would commonly be enough for
people to notice their damaging effects over the sweep of time.
2. According to Burton, local peoples would then, over the years,
come to recognize the abnormal, undesirable consequences of inbred
unions, despite the absence of a Western scientific understanding of their
genetic causes. Eventually, they would come to see the effects both as
detrimental and as distinctive from other afflictions caused by accident
or contagion in their population.
3. Once recognized, people would commonly attribute these special
harmful consequences to the displeasure of the supernatural—a warning,
essentially, that the mating that produced the deformity (or deformities)
was a transgression of normal order (maintained by the aversion) and
was being punished by supernatural agents. People would see inbreeding
as divinely punished, which would no doubt contribute to the special
affect it is so often accorded.
4. Fear of additional retribution from the supernatural and fear that
multiplication of these infractions and their consequences could bring
harm to all—at the hands of nature or hostile enemies, for example, if
not at the hands of deities themselves—would promote moral disapproval
of the responsible act (or acts) within the group and would prompt
continuing prohibition of that which displeased the supernatural.
130 durham

As formulated by Burton, the theory had much to recommend it. It gave

Westermarck’s aversion an appropriate and prominent role; it fit nicely
with emerging genetic theory for inbreeding effects, while also not assum-
ing that local people had any knowledge of Western genetics; it accommo-
dated a wide range of local beliefs, assuming only that supernaturals of
some kind had the power to deform fetuses and /or prevent newborns from
thriving; it carries a strong built-in motivation for individual compliance (a
fearful threat of supernatural sanction); and contra Bateson (Chapter 1) it
does propel a conviction that others should be prevented from violation, or
sanctioned after the fact, lest the community suffer further disasters at the
hand of supernatural forces. Moreover, Burton’s argument immediately ex-
plains why Westermarck had, himself, come upon so many, many “folk
theories” of inbreeding effects. In fact, Westermarck’s original list of folk
theories—five pages of cases compiled innocently in support of his claim
that inbreeding could have caused the aversion to evolve by natural selec-
tion37—remains in my view the single best compilation of data for the Bur-
ton theory.38
But even with all these positive features, I was struck early on by several
limitations to Burton’s original formulation. The argument needed expand-
ing beyond simply the genetic effects of inbreeding to include any repro-
ductive impairment or negative “fitness effects” stemming from sexual rela-
tions with kin, including the psychological consequences of sexual trauma
and any social stigma affecting inbred families.39 In the right circumstances,
these additional sources of reproductive impairment could well contribute
to the locally recognized consequences of inbreeding, as indeed could the
Westermarck effect itself (through lowered fertility, etc.). But also needed
was a broadening of Burton’s notion of attribution and supernatural pun-
ishment. Rather than assume, as Burton did, that locals would always think
“the gods” were punishing close inbreeding via its reproductive effects, for
indeed some societies might well have gods lacking in such powers, it made
more sense simply to leave as an open matter the cultural interpretation of
inbreeding effects, as long it is locally meaningful, widely convincing, and
negatively valued. The resulting hypotheses, (1) that local peoples recognize
at least some adverse consequences to inbreeding, and (2) that they cultur-
ally interpret them as harmful and potentially threatening to everyone, still
constitute a form of “bad stock” argument. But they are more inclusive and
realistic than Burton’s original suggestions. And they are also more readily
incorporated into an “optimal outbreeding” framework, in which any such
perceived “inbreeding costs” might be balanced against prevailing “out-
breeding costs” in the eyes of the locals.40
 Assessing Gaps in Westermarck’s Theory 131

Empirical Support for Burton’s Theory

Once strengthened by these and other modifications, the Burton ar-

gument seemed to warrant testing in an unbiased sample and so, in an ear-
lier study, I tested its predictions against data from the same equiprobability
sample of sixty world societies mentioned earlier.41 The sample is admittedly
tiny for any attempt at global generalization, but that fact made it possible
to carefully scrutinize, with the help of Stanford student assistants, the large
(though partial) ethnographic database available for each society of the
sample—the Human Relations Area Files.42 To make another long story
short, I found that the HRAF data for forty-two of the sixty societies con-
tained at least one reported consequence to incestuous congress, including
social and supernatural sanctions. But more important, I found that twenty-
three of these forty-two cases—54.8 percent—featured some recognition of
inbreeding effects. Moreover, in a number of those twenty-three cases, in-
cluding the Tikopia of Polynesia and Eastern Toradja of Sulawesi, there was
enough ethnographic material to confirm that both recognition and cultural
interpretation were clearly at work.
Consider the case of the Eastern Toradja, a population of dry rice culti-
vators who inhabit the mountainous interior of Sulawesi and are distant
linguistic relatives of the Trobrianders discussed by Hill Gates (Chapter 8).
As informants explained to Dutch observers Albert Kruyt and Nicolaus
Adriani (missionary and linguist respectively, whose combined work in the
region ranged from 1890 to the 1940s), the Toradja incest taboo stems
from a specific physical incompatibility:
It can happen that a man and a woman have physical characteristics (oea ngkoro)
that come into conflict with each other when they marry. The harmful influence of
this will manifest itself in the children born from such as marriage: they will be
weak, sickly, or idiotic and quickly die. Therefore a marriage with a person too
closely related is considered unsuitable (bare’e raposioea, literally, not correspond-
ing in physical characteristics); their children are then not healthy and do not live
long.43

In other words, Toradja informants cite the harmful consequences in off-

spring as the reason inbreeding is prohibited, consequences they interpret
as evidence of a “conflict of physical characteristics.” Not only do these
consequences closely resemble the expected genetic effects, but there is also
(an accurate) recognition that the effects are less likely at greater kinship
distances among consanguines. Adriani and Kruyt found that many Torad-
jans “object to a marriage between cousins out of fear that such a union
will remain childless; or that the children resulting from it will not have vi-
ability, [or] will be idiotic or crippled.”44 In contrast, when the blood rela-
132 durham

tionship is more distant, marriages can still be arranged provided that cer-
tain rites of expiation are also observed, in which case (also accurately)
children are likely to be healthy and thrive.
The Toradja case thus seems a good general fit with the expanded Bur-
ton argument. Close inbreeding is “unsuitable” because of its reproductive
consequences. It is not in the first instance that the gods are angry; rather,
it is a simple physical incompatibility of characteristics. Later on, the gods
may indeed manifest displeasure in the form of widespread drought or
damaging rains, according to Adriani and Kruyt,45 but the telltale signal of
the problem, the reason inbreeding is considered “unsuitable,” is incom-
patibility. (An alternative hypothesis that suggests itself—that this “bad
stock” theory represents simply a diffusion of Dutch understandings of in-
breeding—is readily rejected by clear evidence that indigenous terms for
“incompatible characteristics” and related phenomena greatly predate the
arrival of the Dutch in 1905.) So one might reasonably ask, Why does in-
cest matter to others in a village, beyond the immediate affected parents
and close kin? Here again the Toradjan case seems exemplary even when
there is no drought; prior to pacification in the 1900s, there was almost in-
cessant concern over the economic and military vitality of one’s village
given the prevalence of headhunting in the region. Notes Robert Lagace,
“Until European contact, the Eastern Toradja lived in a state of essential
village autonomy and semiperpetual hostility with each other. . . . Inter-
village raiding served to provide scalps necessary for many rituals and to
pacify spirits (anitu) who would otherwise feed on the Eastern Toradja.”46
In this environment, a village rose or fell with the strength of its able-
bodied warriors and workers.
But there is still one further observation by Toradjan chiefs that is, to
me, the most convincing part of all in regard to the Burton argument. Ac-
cording to Adriani and Kruyt (1951, 2: 56–57) again,
There are many stories about men and women who supposedly [had sex with] a
tree spirit [bela]. There are women who say they are visited each night by a bela.
Many become pregnant from such intercourse. . . . The children who are said to re-
sult from [sex] with a spirit are described as “light of skin with blue eyes and white
hair.” This refers to albino children, who are rarely found among the Toradja. But
other characteristics of children whose father is supposed to have been a spirit are
mentioned. One of them was hairy, “like a monkey;” another had the nose and eye-
brows of a monkey. . . . So-called spirit children do not live long, because their fa-
thers [the bela, are said to] take them in order to bring them up themselves.47

Significantly, the authors report that “more than one Toradja chief assured
us that women claimed to have been made pregnant by a tree spirit when
their condition was the result of intercourse with a member of her kin group
whom she was not permitted to marry.” Not only does this local interpreta-
 Assessing Gaps in Westermarck’s Theory 133

tion provide explanation for the early death of inbred children, but the data
also show that chiefs take physical deformity of offspring as diagnostic of il-
licit incest. They, at least, are not fooled by the alibi.
In summary, data from the Toradja and twenty-two other cases in the
global sample offer striking if partial support for a different theory linking
Westermarck’s aversion to the prohibition against incest. The cases are fully
complementary with Westermarck’s own list of folk theories mentioned
above, with the additional advantage of not being gleaned from the litera-
ture to make the particular point. (Indeed, they are joined by a further ex-
ample from his separate study of marriage in Morocco: “It was also the
opinion of the ancient Arabs that the children of marriages between rela-
tives are weakly and lean. Thus a poet [wrote] ‘ . . . the seed of relations
brings forth feeble fruit.’”)48 But with only 54.8 percent of the pertinent so-
cieties recognizing inbreeding effects, these findings are best seen as simply
provocative, far from a full-blown empirical verification of Burton’s alterna-
tive explanation. Still, they should at least give us pause. They show that
non-Western human populations do commonly recognize the deleterious ef-
fects of inbreeding and interpret them as a negative sign or warning that
such unions, for a variety of reasons, invite these effects. At the same time,
bear in mind that the other, nonconfirming cases of the “Sixty Cultures”
sample warrant careful interpretation; they specifically do not mean that
these other peoples fail to recognize and culturally interpret inbreeding ef-
fects. All they mean is that there is no report of such phenomena in the (ad-
mittedly spotty) HRAF database of this study. On balance, then, these data
support the conclusion that many societies—including Western ones—have
incest taboos that are derived fully or in part from the observation of in-
breeding effects.49
This conclusion has, in my view, a number of implications for contem-
porary scholarship in the general area of incest theory. First, it is definitely
a bit early to suggest that incest taboos “had nothing to do with society not
wanting to look after the half-witted children of inbreeding, since in many
cases they had no idea that inbreeding was the cause.”50 If we may trust an
admittedly tiny probability sample, then the conclusion here implies that in
more than half of all societies known to anthropology incest taboos have
much to do with people seeking to avoid sickly, weak, and half-witted chil-
dren. No doubt some analysts will say that “more than half of all societies
known to anthropology” is still disappointing, given the near universality
of the incest taboo (on universality, see Chapter 5). To this argument, it
must again be noted that Burton’s theory has far more supporting evidence
than exists for the Westermarck process.
Second, the conclusion here implies that it is time for social science to
move beyond a lingering prejudice about the observational powers of non-
134 durham

Western peoples. Such a prejudice was bluntly articulated, for example, in

1948 by anthropologist Leslie White, who remained unconvinced by even
the genetic arguments of his time:
But suppose that inbreeding did produce inferior offspring, are we to suppose that ig-
norant, magic-ridden savages could have established this correlation without rather
refined statistical techniques? How could they have isolated the factor of inbreeding
from numerous others such as genetics [sic], nutrition, illnesses of mother and infant,
etc., without some sort of medical criteria and measurements—even though crude—
and without even the rudiments of statistics.51

I believe the data presented here suggest that White’s “ignorant, magic-rid-
den savages” did frequently figure it out, even without the advantages of
Western medicine and statistics. That said, let me emphasize that I have no
quarrel with the suggestion by Gates (Chapter 8) and others to the effect
that “hypertrophied cultural logic” about good, healthy offspring and the
conditions that cause them, rather than the experience of the negative con-
sequences of inbreeding, may have initially provoked the cultural interpre-
tations of proper and improper mating. I am not sure it matters which
came first really, and I doubt we’ll ever know for sure. All that matters is
that the connection be made somehow, convincing people that “although
X, Y, and Z promote healthy offspring, inbreeding does not.” What mat-
ters in the history of incest taboos, I submit, is this linkage, no matter
which way around it is culturally constructed.
Finally, on a related point, I am not suggesting that inbreeding beliefs are
free of superstition or of “added” or “piled on” consequences. As Wester-
marck was himself first to point out, all manner of things—“epidemics,
earthquakes, sterility of women, plants or animals, or other calamities”—
get heaped onto the pile of purported incest consequences in diverse soci-
eties, adding in the local purview to its feared, harmful effects.52 Data avail-
able for the sixty societies described here are no exception. However, the
data summarized above do call into question two of Westermarck’s corol-
laries to that observation. First, they question his assertion that, among non-
Western peoples, the incest taboos, “especially those related to the nearest
relatives, are so strictly observed that no genuine knowledge could possibly
be based on the few cases in which they are transgressed.”53 In many cases
of the “Sixty Cultures” study described here, Westermarck is directly con-
travened on the topic, sometimes as explicitly as this:
In a conversation which I had with [an informant] on the subject of incest he
brought up the matter of supernatural sanction as follows. “Some brothers and sis-
ters,” he said, “who have the one father but different mothers, will join together,
will embrace each other. When such brother and sister have begotten their children,
these keep on dying, dying, dying, and the labor is of wailing, wailing, wailing for
 Assessing Gaps in Westermarck’s Theory 135

the children who do not exist long, but die off . . . ” That the offspring of such in-
cestuous marriages do die in this way he was prepared to support by definite evi-
dence, like other informants. “The observation of it in this land is finished,” he
said, meaning that there were cases to hand, known to everyone, which formed the
empirical basis for the common opinion.54

Second, these data make me think Westermarck doth protest too much:
“Even if we had a right to make such an assumption, even if savage men
everywhere had discerned that children born of marriage between closely re-
lated persons are not so sound and vigorous as others, we could certainly
not be sure that they everywhere would have allowed this knowledge to
check their passions.”55 The issue here is not whether knowledge checks pas-
sions, although I am sure that makes a lovely debate, or as Gates (Chapter 8)
so aptly points out, over “who wins,” reason or passion. The argument is
over the taboo itself, a cultural product, and the way it is shaped by both
passion and reason. Despite Westermarck’s protestations and those of some
supporters since, there remains little or no basis for rejecting the Burton hy-
pothesis, particularly in its expanded form. It remains possible that Wester-
marck’s aversion can be causally related to the incest taboos of human soci-
eties in ways that bypass Westermarck’s process. It seems to me that this
implication alone calls for more serious attention to the actual reasons that
close inbreeding is condemned in human populations.

Conclusion

Ironically, after presenting these and related ideas at the Stanford

conference, I actually came away wondering if we have been emphasizing
the wrong question all along about incest taboos. The question we seem to
have been asking is, What explains the nearly ubiquitous incest taboos of
human societies? It occurs to me now, and is at least worth putting forward
for further reflection and discussion, that this question really parses into
two separable questions: (1) How did the incest taboos of humanity origi-
nate? and (2) What forces sustain and shape incest taboos in the diverse
human populations of the ethnographic record? The first of these questions
is very difficult, if not impossible, to answer with our current analytical
tools and data. It may well be that we simply can’t get at origins, and thus
any attempt to “pit” Westermarck versus Burton on this question may be
completely in vain. I question whether we’ll ever, in the foreseeable future,
know enough to fairly test alternative hypotheses about origin.
The second question, in contrast, seems like a much more tractable topic.
It, too, might be separated into two distinct questions, namely, What pro-
cess (or processes) or force (or forces) are most important in sustaining in-
136 durham

cest prohibitions in human populations? And what processes operate to

shape existing taboos, governing their extensions and limits as well as their
sanctions? The Westermarck theory is certainly one good candidate for pro-
viding answers to these questions. However, if I have been successful in
these remarks, the point will have been made that it is not alone in the can-
didate pool. Still to be eliminated, or perhaps somehow incorporated, is an
interesting rival hypothesis—if nothing else a foil to today’s more popular
Westermarck arguments. The rival holds that in most, or maybe all, human
populations of the ethnographic record, incest taboos are sustained and
shaped primarily by a local moral disapproval that is expressed in terms of
(what Western science calls) “inbreeding effects” as they are locally and
meaningfully interpreted.
Could the incest taboos also have originated in this way, by early recog-
nition of inbreeding effects and appropriate cultural interpretations in a va-
riety of ancestral populations? Yes, I think they could have, especially if one
regards human cultural history as a frequently branching tree (such that not
all populations had to recognize and interpret inbreeding effects indepen-
dently and from scratch). But I also think it will be a long time, if ever, be-
fore we can say anything like that with confidence. Could the incest taboos
have originated some other way, such as by Westermarck’s arguments, and
then later had recognition and interpretations of inbreeding effects added
on? Yes, this is also quite possible and could well bring to human history
certain advantages to both theories. For example, the automatic “innate ac-
tion” of Westermarck’s process could later have been joined by the more
cultural, interpretive action of the Burton theory, with the advantage I
should think of a stronger, more locally meaningful cultural reinforcement
of the taboos. Yes, I think a variety of historical scenarios are possible,
many of them potentially integrating Burton and Westermarck components.
But I also think it will be difficult to get clarity and resolution when dealing
with questions of origin. Instead, the take-away points I hope to have made
in this chapter are two: first, the case for Westermarck is by no means open
and shut as we begin the new millennium, and second, whenever it oc-
curred, the recognition of inbreeding effects and their local cultural inter-
pretation by human observers added powerful cultural evolutionary forces
to the sustaining and shaping of incest taboos. They added nothing less
than symbolism and meaning to the moral emotions.

notes
1. F. B. M. de Waal, “The End of Nature Versus Nurture,” Scientific
American, vol. 281 (1999), pp. 94–99.
 Assessing Gaps in Westermarck’s Theory 137

2. M. Ruse and E. O. Wilson, “Moral philosophy as applied science,” Philoso-

phy, vol. 61 (1986), p. 184. See also C. J. Lumsden and E. O. Wilson, Genes,
Mind, and Culture: The Coevolutionary Process (Cambridge: Harvard University
Press, 1981.
3. E. O. Wilson, Consilience: The Unity of Knowledge (New York: Alfred
Knopf, 1998), p. 173.
4. Edward Westermarck, The History of Human Marriage, 5th ed., vol. 2
(New York: Allerton, 1922), pp. 198 and 193, respectively.
5. See review in William Durham, Coevolution: Genes, Culture, and Human
Diversity (Stanford, Calif.: Stanford University Press, 1991), chapter 3.
6. See, for example, S. J. Gould, “Sociobiology: The art of storytelling,”
New Scientist, vol. 80 (1978), pp. 530–33; S. J. Gould and R. C. Lewontin, “The
spandrels of San Marco and the Panglossian paradigm: A critique of the adapta-
tionist programme,” Proceedings of the Royal Society of London, Series B,
vol. 205 (1979), pp. 581–98.
7. Westermarck, The History of Human Marriage, 5th ed., vol. 2, p. 193.
8. Ibid., p. 204.
9. Neven Sesardic, “How not to refute Westermarck,” Biology and
Philosophy, vol. 13 (1998), pp. 413–26.
10. Ibid., though not in all societies—see P. Menzel and F. D’Aluisio, Man
Eating Bugs: The Art and Science of Eating Insects (Berkeley, Calif.: Ten Speed
Press, 1998).
11. Durham, Coevolution, chapter 5.
12. Patrick Bateson and Paul Martin, Design for a Life: How Behavior and
Personality Develop (New York: Simon and Schuster, 2000), pp. 216–17.
13. Arthur P. Wolf, Sexual Attraction and Childhood Association: A Chinese
Brief for Edward Westermarck (Stanford, Calif.: Stanford University Press, 1995),
p. 512.
14. Durham, Coevolution, p. 323.
15. Edward Westermarck, The History of Human Marriage, 1st ed. (London:
Macmillan, 1891), pp. 324–27.
16. Ibid., pp. 329–30.
17. Ibid., p. 321.
18. Westermarck, The History of Human Marriage, 5th ed., vol. 2, p. 207.
19. Ibid., p. 205.
20. Ibid., p. 216.
21. Ibid., p. 214.
22. Ibid.
23. Melvin Ember, “On the origin and extension of the incest taboo,” Behav-
ior Science Research, vol. 10 (1975), pp. 249–81.
24. Durham, Coevolution, pp. 341–44.
25. See Robert O. Lagace, ed., Sixty Cultures: A Guide to the HRAF Probabil-
ity Sample Files, Part A (New Haven, Conn.: Human Relations Area Files, 1977).
26. Durham, Coevolution, chapter 6.
27. For details see ibid., pp. 344–57.
28. Ibid., p. 353.
138 durham

29. Ibid., pp. 310–14.

30. Ibid., p. 323.
31. Joseph Shepher, Incest: The Biosocial View (New York: Academic Press,
1983), p. 60.
32. Justin McCabe, “FBD marriage: Further support for the Westermarck
hypothesis of the incest taboo,” American Anthropologist, vol. 85 (1983), p. 50.
33. Roger Burton, “Folk theory and the incest taboo,” Ethos, vol. 1 (1973),
pp. 504–16.
34. Durham, Coevolution, pp. 324ff.
35. Burton, “Folk theory and the incest taboo,” p. 505.
36. For discussion of the genetic theory behind this point, see W. M. Shields,
Philopatry, Inbreeding, and the Evolution of Sex (Albany, N.Y.: State University of
New York Press, 1983), chapters 3 and 4.
37. Westermarck, The History of Human Marriage, 5th ed., vol. 2, pp. 170–74.
38. Durham, Coevolution, p. 324.
39. With regards to sexual trauma, see D. Wilner, “Definition and violation:
Incest and incest taboo, Man (n.s.), vol. 18 (1983), pp. 134–59; and with regards
to social stigma, see S. Yokoyama, “Social selection and evolution of human dis-
eases,” American Journal of Physical Anthropology, vol. 62 (1983), pp. 61–66.
40. See Patrick Bateson, “Optimal outbreeding,” in Mate Choice, ed. Patrick
Bateson (Cambridge: Cambridge University Press, 1983).
41. For details see Durham, Coevolution, pp. 329ff.
42. See Lagace, ed., Sixty Cultures.
43. N. Adriani and A. Kruyt, The Bare’e-speaking Toradja of Central Celebes
(The East Toradja), vol. 2 (Amsterdam: N. V. Noord-Hollandsche Uitgevers
Maatschappij, 1951). English translation in Human Relations Area Files, Section
OG11, “Toradja.”
44. Ibid., p. 284.
45. Ibid., p. 271.
46. See Lagace, Sixty Cultures, p. 411.
47. Adriani and Kruyt, The Bare’e-speaking Toradja, pp. 56–57.
48. Edward Westermarck, Marriage Ceremonies in Morocco (London,
Macmillan, 1914), p. 55.
49. See, for example, Jack Goody on “The Letter of Gregory” behind the lim-
its of the Catholic prohibition. Jack Goody, The Development of the Family and
Marriage in Europe (Cambridge: Cambridge University Press, 1983), pp. 35–36.
50. Bateson and Martin, Design for a Life, p. 217, emphasis added.
51. Leslie A. White, “The definition and prohibtion of incest,” American
Anthropologist vol. 40 (1948), p. 418.
52. Westermarck, The History of Human Marriage, 5th ed., vol. 2, p. 178.
53. Ibid.
54. Raymond Firth, We, The Tikopia: A Sociological Study of Kinship in
Primitive Polynesia, reprint edition, abridged by the author (Stanford, Calif.:
Stanford University Press, 1983), pp. 288–89.
55. Westermarck, The History of Human Marriage, 5th ed., vol. 2, p. 178.
8 Refining the Incest Taboo
w i t h c o n s i d e r a b l e h e l p f ro m
bronislaw malinowski

Hill Gates

...
Ikatupwo’i inala: “Mtage luguta?”
She asks her mother: “Indeed my brother?”
Kawalaga: “O latugwa boge inagowasi!”
Her speech: “O my children already they are mad!”
...
Ilikwo dabela, iseyemwo.
She unties her fiber skirt, she puts it down.
Ivayayri namwadu. . . . Iloki luleta.
She follows the shore naked. . . . She goes to her brother.
...
Ibokavili . . . iyousi, ikanarise wala obwarita.
She chases, she takes hold, they lie down right in the sea.
Ikanukwenusi, ikammaynagwasi, ivino’asi.
They lie, they go to shore, they finish.
...
Gala ikamkwamsi, gala imomomsi, u’ula ikarigasi.
They neither eat nor drink, and so they die.

—From the Trobriand Myth of the Origin of a Love Magic

Strong Enough to Incite Incest (Bronislaw Malinowski,
The Sexual Life of Savages in North-Western Melanesia)

The incest taboo is a phrase that delivers an emotional and semantic

double blow. The two words, each problematic enough alone, together be-
fog the analytic impulse with a messy mix of biological, gothic literary, and
arcane religious implications. My task here is to decompose that complex-
ity, demystify some of its historically accrued sensationalism, and ground it
empirically—without denying its genuine experiential explosiveness. What
do we mean, and what have we meant, by the incest taboo?
To answer, I briefly explore the embrace by culture-oriented twentieth-
century anthropologists of Sigmund Freud’s Oedipus complex and its odd
140 gates

alliance with contract theory. At present, our accumulating knowledge of

human biology aligns better with evidence from an earlier phase in ethnog-
raphy, epitomized by the work of Bronislaw Malinowski. His account of
Trobriand incest taboos emphasized local sociocultural influences on a
common human nature and thus anticipates the late-twentieth-century con-
clusions of Arthur Wolf. Edward Westermarck, Malinowski, and Wolf
would agree that incest taboos are rationalized expressions of innate dis-
taste, with the core aversion locally embroidered to fit cultural context.
To further refine our understanding of incest taboos, we may consider
how a universal sexual aversion within the nuclear family has been har-
nessed to social use in different political-economic clusterings of incest ta-
boos. Here, as illustration, I isolate brother-sister incest as a possible factor
in cultural evolution.1
In Totem and Taboo, Sigmund Freud did two things. He wrapped the
term incest taboo in his era’s uneasy religiosity and Lamarckian biology, and
he incautiously accepted an obsolescent and dismissive view of the intellec-
tual achievements of “primitive man.” For Freud, the taboo was a prohibi-
tion invented in a remote evolutionary Dreamtime, when men “forbade
themselves” sex with their mothers and sisters. Necessary for the emergence
of civilization or culture, and absorbed as an inherent part of our human
nature, the taboo was recapitulated in each individual as the price of sexual
and social maturity. Freud thus linked two currently fascinating topics—hu-
man evolutionary origins and family conflict—while legitimating open dis-
cussion of individual sexual development. The rapid adoption of Oedipal
explanation into Western popular thought probably owed as much to eth-
nographic ignorance and prejudice as to the delightful liberation of talking
trash at dinner tables.
The curious connections Freud drew in Totem and Taboo immediately
attracted support among anthropologists. This is perhaps because of a con-
venient intellectual slippage: between Freud’s grand vision of humans con-
structing full humanity—person by person, people by people—through acts
of will, and the sociology of knowledge that vivified post–World War I
Western European thought.2 Throughout the twentieth century, emphasis
on the socially constructed character of human nature has dominated
liberal-to-progressive social analysis. The need for such emphasis to counter
rationales for customary race and gender biases will be obvious to readers
of this volume.
Anglophone anthropologists especially committed themselves to such lib-
eral positions, finding Freud, for all his increasingly obvious ethnographic
naïveté, a charismatic ally. They ignored the inherent biologism of his posi-
tion and recast his incest taboo as an eternally reinvented social contract
limiting sexual conflict and binding vulnerable families into defensive kin-
 Refining the Incest Taboo 141

ship networks—as E. B. Tylor and Sir James Frazer had framed it previ-
ously. In the words of Claude Lévi-Strauss, “we are of the opinion that the
prohibition of incest provides sufficient guarantee that a network of alli-
ances, resulting in all other respects from free choices, will not compromise
social cohesion.”3 Leslie White, Marvin Harris, Maurice Godelier, and many
others would later concur.
According to this mainstream position, evolving humans eschewed the
convenience (and perhaps desirability) of mating with close kin in order to
force the formation of wider social networks through exogamy. Especially
when combined with the Freudian premise of nuclear family attraction, this
conquest of nature gave rise to “civilization” or “culture”—a new, human
domain of experience that transcended all previous mammalian constraints.
In the twentieth century, most anthropologists viewed the willed passage
via the incest taboo from animal to human as the very essence of culture in
its general evolutionary sense. Once across the evolutionary bridge of re-
nouncing incest, humans became free, socially plastic agents in a myriad of
specific cultures—localized, contingent, historical, all capable of transcend-
ing our animal behavioral inheritance.
The confusion in anthropological discourse that arises from the confla-
tion of two distinct uses of the term culture would be hard to overestimate.
In an essay titled “Evolution: Specific and General,” Marshall Sahlins dis-
cussed both biological and cultural evolution in these terms: “The distinc-
tion has long existed in the literature of evolutionary anthropology. E. B. Ty-
lor, . . . (1871), laid out the study of cultural evolution both “stage by stage”
as well as “along its many lines.” Sahlins continues: “General cultural evo-
lution . . . is passage from less to greater energy transformation, lower to
higher levels of integration, and less to greater all-round adaptability. Spe-
cific evolution is the phylogenetic, ramifying, historic passage of culture
along its many lines, the adaptive modifications of particular cultures.”4 The
later Marshall Sahlins and similarly relativist anthropologists now repudiate
“general cultural evolution” to describe the differences between egalitarian,
ranked, and state societies as ways of life or “stages.” But none would dis-
agree, I suppose, that the transition to fully modern humans exhibiting the
capacity for symbolic thinking—“culture”—represents such a general evo-
lutionary move. It is helpful, then, to retain Sahlins’s distinction between
culture general and culture specific. It flags the difference between culture as
a uniquely developed human capacity to transmit information and cultures
as localized and historically contingent products of this capacity.
If biological factors cannot explain cultural difference—as they surely
cannot—it has also been assumed that biology can teach us nothing about
our species as a whole. It is deemed useless (or racist or sexist) to investigate
possible genetic and developmental constraints or unities of a species nature.
142 gates

Further, if a common human nature operated on our varied cultures, we

should see a set of human universals to class with the incest taboo. But, cul-
turologists emphasize, most nineteenth-century cultural universals have been
successfully deconstructed. Feminists have purged scientific thought of many
common fantasies; students of comparative religion and race have dissolved
others. Even kinship, marriage, and the nuclear family, important contenders
for universality in our species, have been vigorously rejected in some quar-
ters.5 That the incest taboo is a willed override of nature by culture, however,
is generally agreed—often in the most amazingly muddled terms.6
This consensus is increasingly untenable. While human aversion to in-
cest surely has consequences for social organization and individual matu-
ration, to take these consequences as causes is to commit one of the most
frequently and most legitimately condemned sins of functionalist interpre-
tation. The line that contract theory draws between “us”—humans—and
“them”—the rest of nature—inevitably seems less sharp to thinkers who
know how many of their genes they share with a fruit fly.
After a short twentieth century of Freudian obfuscation and overempha-
sis on culture, research on inbreeding, incest, and the incest taboo is achiev-
ing a new clarity and reliability. Rearguard defenses of their exclusively cul-
tural origin are still mounted. By the twenty-first century, however, they can
be supported only by stubborn citation of out-of-date material.7 Evidence
from nonhuman behavior, especially among primates, as well as from large
and well-documented human communities has returned debate—as this vol-
ume demonstrates—to the serious consideration of evolutionary factors. The
insight of Darwinian anthropologist Edward Westermarck, long hidden un-
der the Freudian bushel, now has empirical support (see Chapter 4). A focus
on human nature by the anti-Freudian anthropologist Bronislaw Malinow-
ski, derided since the 1930s as psychologistic naïveté, now appears prescient.

Evaluating Ethnographic Evidence

Anthropological generalizations can be sustained only when they

meet the test of evidence got by two very different methods, each flawed in
a different way. The ethnographic case study must be thorough, and it will
always require the most exacting contextualization: Where? When? As in-
terpreted by whom? For what audience? Cross-cultural comparisons are
vexed instead by the “unit problem”: What groups of people constitute dis-
tinctive sociocultural entities? In our calculations with these units, how
should we weigh two thousand Tikopia against a billion Chinese? The dif-
ficulties are obvious, generating argument that sometimes descends to aca-
demic brawling.8 In the United States especially, the influence of British
 Refining the Incest Taboo 143

structural-functionalists waned rapidly after World War II, and the Marx-
ist and ecological anthropology of the 1960s and 1970s flourished even more
briefly. Almost by default, much twentieth-century ethnography has been
shaped by research programs with a strong culturalist bent. By the end of
the century, even word-wizard Clifford Geertz deplored this tendency while
comparing work before the cultural turn with that of his own sorcerer’s ap-
prentices: “All the human sciences are promiscuous, inconstant, and ill-
defined, but cultural anthropology abuses the privilege.”9
Fortunately, earlier, more holistic, and more comparative anthropolo-
gists have left us bodies of work sufficiently rich in detail and wide-angled
in viewpoint to extricate us from this intellectual cul-de-sac. We turn to
what is arguably the best single ethnographic description of an incest taboo
complex, embedded in Bronislaw Malinowski’s investigations of Trobriand
Island life during World War I, and then to insights from quantified, com-
parative studies.
Most of Malinowski’s observations of and conclusions about Trobriand
society and culture were published in five detailed and respectfully written
books.10 His many other publications included materials about the Tro-
brianders and their near neighbors, especially Magic, Science, and Reli-
gion; his methodology is made transparent in an unvarnished account of
his Trobriand fieldwork—A Diary in the Strict Sense of the Term—that ap-
peared posthumously.11 In Sex and Repression in Savage Society, he repu-
diated a transiently felt Freudianism, admitting himself to having been one
of “many fools” deeply impressed by the promise of psychoanalysis and its
Oedipal core.12 These he replaced with a full and nuanced discussion of
Trobriand incest that emphasized local sociocultural influences on a com-
mon human nature.
This mass of material has survived anticolonialist and feminist critique
better than most ethnography of its time. J. P. Singh Uberoi and Marshall
Sahlins expanded on but did not dispute Malinowski’s interpretation of the
region’s political economy as one where vigorous staple markets were elab-
orately constrained by ritual gift exchange among local leaders.13 Annette
Weiner researched Trobriand women’s work in the 1970s with greater gen-
der-attentiveness, adding much to our understanding of the overall system.14
She might easily have assumed what E. P. Thompson called the “enormous
condescension of the young” toward her predecessor. Instead, she rejected
the argument “that ethnographic writing can never be more than a kind of
fictional account of an author’s experiences. Although Malinowski and I
were in the Trobriands at vastly different historical moments and there also
are many areas in which our analyses differ, a large part of what we learned
in the field was similar.”15
Weiner has relieved us of another academic anxiety by commenting on
144 gates

the effects of colonialism during Malinowski’s research. As Trobrianders

sometimes told Malinowski, times had changed with the coming of foreign
administration, legal codes, and plantation labor.16 Fifty years later, Weiner
noted the surprising stability of Trobriand practice.17
For myself, I should like to see a thoughtful revision of the kinship/gen-
der/inheritance/status nexus that often puzzled Malinowski as a contradic-
tion between matriliny and father-right. Since his time, anthropologists have
developed more sophisticated interpretations of ranking, lineality, and gen-
der roles. David Labby has given us an exemplary reconception of Oceanian
social process that can illuminate Trobriand experience;18 Karen Sacks’s
Africanist perspective is also provocative.19 True to form, however, Mali-
nowski himself collected enough data to substantiate such models, so very
different from his own.
Perhaps the most valuable quality of Malinowski’s ethnography is its
recognition of both the importance and the striking limitations of local texts.
Getting out of the armchair, off the guesthouse verandah, and beyond the
simplicity of symbolic systems was an intellectual as well as a methodologi-
cal breakthrough. Insisting on attention to enacted as well as to enunciated
meaning, Malinowski observed:
[Indigenous] statements show us the polished surface of custom which is invariably
presented to the inquisitive stranger; direct knowledge of native life reveals the un-
derlying strata of human conduct, moulded, it is true, by the rigid surface of cus-
tom, but still more deeply influenced by the smouldering fires of human nature. The
smoothness and uniformity, which the mere verbal statements suggest as the only
shape of human conduct, disappears with a better knowledge of cultural reality.20

He imputes no deceit to his respondents. A Trobriander “simply does what

any self-respecting and conventional member of a well-ordered society
would do”—that is, he summarizes reality with norms and ideals;21 “native
peoples” have as much difficulty as anyone in describing their reality to out-
siders. “For in actual life rules are never entirely conformed to, and it re-
mains, as the most difficult but indispensable part of the ethnographer’s
work, to ascertain the extent and mechanism of the deviations.”22 Ethnog-
raphers who rely principally on direct interrogation get bad data. “Such ma-
terial [leads] to the anthropological doctrine of the impeccability of native
races, of their immanent legality, and inherent and automatic subservience
to custom.”23 From Malinowski, we learned that only researchers with local
experience, full command of the relevant language, and his own generous
comprehension of our common humanity could hope to capture and trans-
mit the reality of “the,” or even “an” incest taboo. One can only imagine
his salty response to the notion that culture might differentially construct,
and thus negate, such common human emotions as attachment and loss.
 Refining the Incest Taboo 145

Incest Among Trobriand Islanders

One among many peoples who inhabited the horticultural, fishing,

artisan, and trading region of northeast New Guinea and its offshore is-
lands, Trobrianders operated a complex economy in which both women
and men played necessary roles.24 Matrilineal in clan descent and property
inheritance, the key adult female/adult male dyad was not wife and husband,
but (as in matriliny generally) the brother/sister pair. Jointly inheriting var-
ious properties from their mother and mother’s brother (a clan brother if
not a uterine one), a sister and brother were jointly responsible for rearing
and socializing the next generation, and for passing to them land rights, ca-
noe shares, and other properties inherited through a line of women. At har-
vest, a man (and his wife or wives) carried the better half of his harvested
yams, the staple food, to his sister’s household for its use.25 In return,
women produced quantities of banana leaf bundles without which their
matrilineage brothers could not maintain political competitiveness and meet
kin obligations.26 A mother’s brother fed his sister’s children—his heirs—
while she herself worked to guarantee their future rights.
A boy’s mother’s brother was expected to teach him rituals that “be-
longed” to his matrilineage. A boy’s biological father—usually his mother’s
husband—often sentimentally but illicitly shared such secrets and other
property with the sons he reared and loved, stealing them from his proper
heirs—his sister’s children. A boy’s “mother’s brother assumed a gradually
increasing authority over him, requiring his services, helping him in some
things, granting or withholding his permission to carry out certain actions;
while the father’s authority and counsel become less and less important.”27
Trobriand brothers and sisters were life partners with shared interest in their
young heirs, much common property, and continuing mutual attachment.
At the same time, any hint of sexual connection between them was
sharply sanctioned.
Brother and sister thus grow up in a strange sort of domestic proximity: in close con-
tact, and yet without any personal or intimate communication; near to each other in
space, near by rules of kinship and common interest; and yet, as regards personality,
always hidden and mysterious. They must not even look at each other, they must
never exchange any light [i.e., sexual] remarks, never share their feelings and ideas.
And as age advances and the other sex becomes more and more associated with
love-making, the brother and sister taboo becomes increasingly stringent.28

Even in childhood, expression of affection between them was punished.29

Like Freud, Trobrianders appear to have conflated nonsexual attachment
among close childhood associates with sexually based bonding; Malinow-
ski, as we shall see, did not.
146 gates

Malinowski described the prohibition of brother-sister sexual relations

as “the supreme taboo of the Trobriander,” “the prototype of all that is
ethically wrong and horrible to the native. It is the first moral rule seriously
impressed in the individual’s life, and the only one which is enforced to the
full by all the machinery of social and moral sanctions.”30 Because human
beings came into the world as a brother-sister pair between whom sex was
forbidden, Trobrianders normatively described conception as a result of
connection between a woman and her ancestral spirits. The absence of a
male contribution to pregnancy was a logical necessity in the Trobriand
myth of origin, and a position that many Trobrianders held to be factual.31
From all directions, their culture instructed Trobrianders to cherish an inti-
mate but completely nonsexual bond between sister and brother.
A sibling without his or her opposite number was disadvantaged in such
a domestic economy. Among Trobrianders, as among other lineally orga-
nized peoples, the extension of kinship throughout the clan mitigated such
difficulties. Matrilineally related boys and girls were classificatory brothers
and sisters to each other, falling under the incest taboo.
Among Trobrianders, Malinowski expanded,
Sexual intercourse and marriage are not allowed within the same totemic clan.
They are more emphatically forbidden within a sub-division of the clan, common
membership in which means real kinship. And the taboo is stricter yet between two
people who can trace a common descent genealogically. Yet the natives have only
one word, suvasova, to designate all these degrees of exogamous taboo. Also, in le-
gal and formal fiction, the natives would maintain that all exogamous taboos,
whether of clan, sub-clan or proven kinship, were equally binding.32

Yet a man’s marriage with his father’s sister’s daughter—or even a father’s
sister—was common, at least in some subclans.33 Prohibition and moral
disapproval of clan mating shaded off as the matrilineal connection be-
came more remote; incestuous affairs and even marriages occurred on the
fringes.34
On the surface we have one word, suvasova, one clan kinship, one punishment, one
sense of right and wrong. In reality we have the distinction between marriage and
mere intercourse, between clan and sub-clan . . . , between genealogical kinship and
mere community of sub-clan, between the own sister and the classificatory sisters.
We have also to distinguish between direct enforcement by public opinion, and by
supernatural sanctions, neither of which works in a simple or infallible manner.35

The idea of son-mother incest provoked real distaste, and Malinowski

explored this topic with considerable care.
The nearest female of the previous generation, the mother, is also surrounded by a
taboo, which is coloured, however, by a somewhat different emotional reaction [from
that elicited by brother-sister incest]. Incest with her is regarded with real horror, but
 Refining the Incest Taboo 147

both the mechanism by which this taboo is brought home and the way in which it is
regarded are essentially distinct from the brother-sister taboo. The mother stands in
a close bodily relation to her child in its earliest years, and from this position she re-
cedes, though only gradually, as he grows up. As we know, weaning takes place late,
and children, both male and female, are allowed to cuddle in their mother’s arms and
to embrace her whenever they like.36

Sons need not conceal their sexual activities from their mothers (or fa-
thers); and
since normal erotic impulses find an easy outlet [with peers], tenderness towards the
mother and bodily attachment to her are naturally drained of their stronger sensu-
ous elements. Incestuous inclinations towards the mother are regarded as highly
reprehensible, as unnatural and immoral, but there is not the same feeling of horror
and fear as towards brother and sister incest. When speaking with the natives of
maternal incest, the inquirer finds neither the rigid suspense nor the emotional re-
actions which are always evoked by any allusion to brother and sister relations.
They would discuss the possibility without being shocked, but it was clear that they
regarded incest with the mother as almost impossible. I would not affirm that such
incest has never occurred, but certainly I have obtained no concrete data, and the
very fact that no case survives in memory or in tradition shows that the natives take
relatively little interest in it.37

Because they are members of the same household and because his daugh-
ter is his wife’s nearest kinswoman, intercourse between father and daugh-
ter or stepdaughter is taboo.38 “We do not sleep with [a daughter],” said
Trobrianders, “because [the father] fondles [and] takes [her] into his arms”
as a child.39 It occurred,40 but reality was complicated by a kin terminology
that lumped father and father’s sister’s sons—including classificatory ones.
If a true father-daughter relationship came to light, it might shame the man
into suicide, but it was not categorized as suvasova, “exogamy breaking,”
and did not result in disease.41 The possibility that father-daughter (or any)
incest might be punished by imperfect or sickly offspring is nowhere men-
tioned; a father who is not a biological progenitor fits poorly into any argu-
ment from indigenous assessment of inbreeding depression.
When discussing taboos in general, Malinowski distinguished among “the
genuine taboos with supernatural sanction, the clear prohibitions without
supernatural sanction, and prohibitions of acts which must not be done be-
cause they are [so transparently] shameful, disgusting, or else dangerous.”42
The breach of the suvasova taboo entails a “supernatural” penalty: in-
festation by an insect “spontaneously generated by the actual breach of ex-
ogamy,” which covers the skin with sores and produces pains and discom-
fort throughout the body.43 Trobrianders can wax eloquent on this not
necessarily fatal ailment: “As the natives put it: ‘We find maggots in a
corpse. How do they come? Ivagi wala—it just makes them. In the same
148 gates

way the insect is made in the body of the . . . exogamy breaker. This insect
wriggles round like a small snake; it goes round and round; it makes the
eyes swollen, the face swollen, the belly swollen, like in [dropsy].’”44
This etiology calls into question Malinowski’s use of the term supernat-
ural, which we may assume Trobrianders would not use to describe maggots
coming from a corpse any more than my own premicroscope ancestors did
for the same phenomenon. Throughout his career, Malinowski struggled to
escape the inherent ethnocentrism of the Western concepts magic, religion,
and science. Usually careful about the difficulty of drawing a boundary be-
tween “natural” and “supernatural” in Trobriand ontology, here he fails to
resolve a significant hermeneutic difficulty. Incest dermatitis is a “supernat-
ural” but not an agentive punishment; it is in the nature of things.45
Conveniently, the illness “entailed” by exogamy-breaking has “a per-
fectly well established remedy against any pathological consequences of this
trespass, a remedy considered practically infallible, if properly executed. . . .
[This is] a system of magic consisting of spells and rites performed over wa-
ter, herbs, and stones, which when correctly carried out, is completely effi-
cient in undoing the bad results of clan incest.”46
For Trobrianders, as among ourselves, Malinowski denied a “slavish ad-
herence to tradition,” even about incest, insisting on the full and self-
contradictory humanity of all people.47
[He] was told that if a man came by chance upon his sister and her sweetheart
while they were making love, all three would have to commit lo’u (suicide by
jumping from a coco-nut palm). This is obviously an exaggeration which expresses
the ideal and not the reality: if such a mishap occurred the brother would most
likely pretend to himself, and to them, that he had had seen nothing, and would
discreetly disappear.48

What human sanctions were invoked against the incestuous?

The only socially acknowledged “brother/sister” incest case occurring
during his fieldwork is described in Crime and Custom in Savage Society.49
A boy had an affair with his mother’s sister’s daughter, a fact “known and
generally disapproved of.” Her previous boyfriend, jealous, accused him in
public of incest, shaming him beyond endurance. The following morning,
the male exogamy-breaker dressed and ornamented himself, climbed a co-
conut tree, and bade farewell to his clansmen. Accusing his enemy of dri-
ving him to his death, he called on his clansmen to avenge him, leaped, and
died. Taking up his cause, his relatives wounded the rival and held the cus-
tomary mortuary ceremonies for him. His cousin-sweetheart, we learn else-
where, married, lived happily with her husband, and can be seen in the
frontispiece of Sexual Life of Savages.50
The Malasi, a clan with very highly ranked subgroups, were unusual, as-
 Refining the Incest Taboo 149

sociated in fact and story with intraclan incest. Malinowski made no men-
tion of social sanctions or suicide among them. I will return to the un-
orthodox Malasi toward the end of this chapter.
Does such loose fit between norm and practice in incest sanctions simply
indicate that Trobrianders never directly punished breakers of social norms?
No. Malinowski collected accounts of a man speared for sorcery, “a few
cases” of killing of adulterers in flagrante delicto. Insulting a chief risked
this punishment.51
Self-punishment through suicide did occur; Malinowski heard of several
people who had jumped from palms and several others who had poisoned
themselves, although none for reasons of incest.52 In a “well-known” case, a
man caught having sex with an animal was expected to commit suicide, but
he did not. “The culprit . . . has lived down his shame. He leads a happy ex-
istence in Sinaketa, where I had the pleasure of meeting him, and having a
long conversation with him.”53
From his careful investigation of a subject of great interest to both him-
self and his informants, I can only conclude that even first-degree incest
rarely resulted in serious perceivable harm to the islanders. Although incest
evoked strong negative emotions, they knew it to be less dangerous than
adultery or lèse-majesté. Malinowski’s observations led him to believe that
sexual attraction within the nuclear family was unusual; that when it oc-
curred, it had only minor and occasional consequences; that sexual attrac-
tion was common, and marriage not uncommon, among relatives at a few
removes; and that culturally defined incest at some remove was seen as trou-
bling rather than reprehensible, unlikely to result in retribution from any
quarter. The Trobrianders’ incest taboo had been shaped by the local politi-
cal economy around a “natural” core of what Patrick Bateson has identified
as optimal outbreeding.54

Reason Versus Passion

Malinowski ended his career still emphasizing that human beings, by

nature, had needs that social organization and culture must to some degree
accommodate. Everything he wrote urges us to believe that anyone could
correctly interpret behavior in other societies if we understood how their
local peculiarities intersected with our shared human nature. By the time
Raymond Firth published We, the Tikopia in 1936, such “psychologizing”
—necessarily biological at base—had largely been abandoned by anthro-
pologists, however. Even though Firth’s volume had a foreword from Ma-
linowski, Firth himself was convinced that the ultimate explanation for in-
cest taboos was not biological but sociological.
150 gates

The incest-exogamy attitudes may not be reducible to a simple formula. I am pre-

pared to see it shown that the incest situation varies according to the social structure
of each community, that it has little to do with the prevention of sex relations as
such, but that its real correlation is to be found in the maintenance of institutional
forms in the society as a whole, and of the specific interest of groups in particular. . . .
. . . where interest of rank or property steps in, the incest prohibition is likely to
melt away.55

In the late twentieth century, even this relatively grounded sociological ap-
proach was being swamped by a powerful idealist trend among anthropolo-
gists. Annette Weiner chided Malinowski because “his functionalist theories
obscured the subtleties and the significance of social action. His interest was
in the cause and effect of certain actions and activities rather than in the cul-
tural meanings that Trobrianders give.” Having completely dismissed the bi-
ological premise of Malinowski’s frequent allusion to causal emotions, she
assumed that Malinowski was arguing for an absolute rationalism.56
The incest taboo stood through the twentieth century as the last of an-
thropology’s agreed-on human universals, one thought by Tylor, Firth, Lévi-
Strauss, and many others to require no biological basis. Anthropologists
turned to the social contract outlined by Firth or to complete cultural con-
structionism as favored by Weiner at the expense of natural sentiments.
Scottish Enlightenment Passion with its Humean acceptance of affect disap-
peared behind French Enlightenment Reason.
In the light of late-twentieth-century findings, however, the disjunction
between these two positions grows increasingly uncertain. Each depends,
ultimately, on a recognized or unrecognized base in our unique human bi-
ology. To explain incest taboos, contractarians like Lévi-Strauss assume
evolved, species-specific cognition; Westermarck and Malinowski assume
the importance of our kind’s unusually rich emotions.
Contractarians explain that the incest taboo is invented when, perceiv-
ing the advantages of out-marriage, men exchange their close female kin to
make allies. (In our enlightened times, contractualists might grant women
the rationality to see such advantages for themselves. Let us assume so,
sidestepping a potential red herring.) From such a position, the role of cog-
nition in a rational allocation of resources is thrown into especial relief. In
the deep past, we became human through exercising a new intellectual ca-
pacity that our nonhuman ancestors did not have.
A more convincing, noncontractarian argument from human rationality
has been developed by William H. Durham. Durham approaches anthro-
pological questions through a dialectical relationship—a coevolution—be-
tween biology and culture. Accepting the existence of incest-aversive emo-
tion, he asks, “What is the relationship of incest taboos to the phenomenon
of inbreeding depression? and What is their relationship to the sexual aver-
 Refining the Incest Taboo 151

sion [that is now known to develop] from childhood association?”57 He an-

swers that we have coevolved cultural incest prohibitions along with a
genetically based aversion because we perceive a relationship between in-
breeding and inbreeding depression. Durham takes particular note of the
appearance in norms and folklore of a connection between incest and such
visible consequences of inbreeding as weak and defective offspring.
Proposing a cost-benefit model derived from optimal outbreeding the-
ory, he assumes that “the individual-level costs and benefits of inbreeding
and outbreeding can be measured on a scale of individual inclusive fit-
ness.”58 With this model, he predicts that incest taboos, as memetic self-
replicators, “will have high cultural fitness in any population where there
are substantial reproductive costs to inbreeding.”59 “Although the recogni-
tion and interpretation of reproductive impairment are not logically re-
quired for taboos to attain high [fitness], they are nevertheless predicted to
provide a major force of cultural selection in most populations.”60
Durham then applies his model to a sixty-culture sample from the Hu-
man Relations Area Files. He concludes from these comparisons that incest
taboos are genetically adaptive and that they “may be viewed as memes
that have been kept ‘alive’ in cultural pools largely through the positive cul-
tural evaluation of their consequences.”61 Rational actors perceive the in-
breeding depression that may result from incest and construct the negative
cultural evaluation of inbreeding that we call an incest taboo.62
Durham’s argument appears to require “that value-guided cultural se-
lection has been the principal mechanism of change in the evolutionary de-
scent of incest taboos.”63 Necessarily, if not sufficiently, cognition and ra-
tionality more than aversive affect are the aspects of human nature that
shape incest taboos. Like Adam and Eve and Lévi-Strauss, we achieve full
humanity through choice.
Edward Westermarck’s hypothesis and Arthur Wolf’s tests of it are, as
this volume illustrates, very differently based. In Wolf’s account of the
Westermarck effect, the core incest aversion is rooted in affect that unfolds
as Bowlbian attachment behavior in normal infants; a similar conclusion
was reached by the psychiatrist Mark Erickson.64 Attachment, essential to
infant survival, precludes the later unfolding of mature sexual attraction
for the object of attachment and even renders most people averse “when
the act is thought of.” A developmentally nuanced sexual aversion for for-
mer caregivers is a human universal, as is the tendency to generalize indi-
vidual repugnance to moral disapproval. Incest taboos thus are rationalized
expressions of innate distaste, with the core aversion locally embroidered
to fit cultural context.
Comparing Durham’s with Wolf’s approach to this problem illustrates
an important emerging tendency in evolutionary anthropology. We see the
152 gates

formation of two camps: those like Durham, who emphasize intelligent

perception, rational choice, the partly intentional construction of culture,
and a fairly clear line between ourselves and all other species; and those
like Wolf, who stress urgent affect, contextualized post facto rationaliza-
tion of an aversion, the interactions of sociality, and the strong persistence
of primate traits. Once put so plainly, this sharp dichotomy is instantly sus-
pect. Remembering the mess that generations of “nature versus nurture”
thinkers made of social theory, one hopes that a fruitless polarization of
reason and passion can be avoided in this century.

Brother-Sister Incest as a Mechanism

of Cultural Evolution

Each locally elaborated incest taboo may thus be a distinct creation of

reasoned minds working in concert with a unifying heritage of resistant emo-
tion. The aversion is not “the incest taboo.” Societies construct many incest
taboos; each will be distinct, the product of a line of dizzying contingency.
Still, human experience is not a chaos of infinite possibility.
Beyond strong hints of the aversion itself, what elements of well-
described incest taboos can be seen to cluster in informative ways? Do incest
taboos and taboo-breakings point us toward some analytical middle ground
between an exaggerated universality and a promiscuous particularity of ex-
planation? Can we learn something broader about the nature and evolution
of societies by focusing on this vivid commonality?
Well-grounded efforts toward the evolutionary categorization of soci-
eties are older than Lewis Henry Morgan and returned during the 1960s
and 1970s. Jack Goody pioneered a comparative structural analysis of in-
cest prohibition by pairing it with bans on adultery.65 Marxist and ecologi-
cal thinking revisited cultural evolution and infused anthropology with
resurgent respect for what Marvin Harris saw as the materials from which
all our cultures are constructed: “guts, sex, energy, wind, rain, and other
palpable and ordinary phenomena.”66
Now, as materialism returns to anthropological theory, Stevan Harrell’s
highly original attempt to find order in the spectrum of human families
builds especially on the work of Morton Fried and Jack Goody and res-
onates with many of Malinowski’s and Westermarck’s concerns.67 Like most
theorists who keep one foot firmly on material conditions, Harrell clusters
societies by political-economic criteria. He is clear on the direction of influ-
ence between family form and wider institutions: “On the whole, it is
change in the larger social organization that brings about family change
rather than the other way round. . . . [Conditions for organizing families]
 Refining the Incest Taboo 153

are set by the nature of the social system in which the family system is em-
bedded, though the family system may play a small part in determining the
nature of that larger system.68
Harrell’s powerful typology of social systems specifies the contexts
within which human families take their relatively limited forms. Harrell
pays close attention to kinship in ranked societies: those with marked dif-
ferentiation in public positions that is nevertheless insufficient to restrict ac-
cess to subsistence/reproduction goods, and where those defined as chiefs
are exempted from ordinary subsistence labor.69 Harrell’s emphasis on
ranking suggests an approach to cultural difference in rationalizations of in-
cest aversion, especially when the general tendency of aversion to brother-
sister marriage is reversed to favor such marriages.
Brother-sister mating appears as by far the commonest, if not the only,
form of mating within the nuclear family that has been linked with specific
political-economic form. In an archaeological and ethnohistorical study of
the chiefly Calusa of early colonial southwest Florida, John Goggin and
William Sturtevant list most of the known societies that accepted brother-
sister marriage; thirty-six of the forty-two distinguish either by rank (in
Fried’s terminology) or by class (in state societies).70
Some state societies are also known for sibling marriage, especially in rul-
ing houses. It is suggestive that no well-developed bureaucratic state is
known to have urged its royalty to marry incestuously, with one exception:
Roman Egypt.71 Apart from Roman Egypt, Goggin and Sturtevant list only
small, newly formed, or recently barbarian-conquered states (Bali, Cambo-
dia, the Hittite polity, Incan Andes, Java, Korea, Sinhala, and Thailand).
Most of the rest are complex chiefdoms like Hawai’i, structurally perched to
tip over into the class rigidities of state-ness. Strong chiefdoms, like small,
aristocratic kingdoms were heavily dependent for power-creation on what
Clifford Geertz called “theater state” tactics: flashy monumentalism, spec-
tacular public ritual (often with human sacrifice), and the lavishly detailed
apotheosis of rulers.72 This repertory of cultural flamboyance accords well
with the shock value of royal incest. Brother-kings married sister-queens, or
brother-chiefs their sister-chiefs, flouting a prohibition commoners will have
felt to be natural. Lords and ladies of the earth flaunted superhuman invul-
nerability, constructing auras of power by haughty taboo-breaking. They
did so especially when they had not yet invented the administrative and
communications systems to enforce their rule by less colorful means. Such
families had the power to override aversion and enforce such marriages even
on unwilling couples.
Polynesian peoples developed a number of high-chiefdom /almost-state
societies in which brother-sister marriage for the high-ranking was politi-
cally salient. Computer simulation of Tongan chiefly marriage appears to
154 gates

confirm that brother-sister marriage influenced social structure by signifi-

cantly increasing
all indicators of stratification, including the status variability that existed among
chiefly lines and the status differences that developed among chiefly individuals.
The results of the simulation suggested further that marriage rules might have been
necessary to the system in order for stratification to occur because, without the
presence of endogamous customs, the trajectories of chiefly statuses converged.73

Where conditions for ranking were auspicious, a family that risked invert-
ing taboo and ignoring aversion set in train a snowballing status improve-
ment for its descendants and an evolutionary leap in social complexity for
its society. Somewhere in the history of all early states we might expect to
find royal brother-sister incest lurking.

Return to the Trobriands

Malinowski too had something to teach us about the possible role of

brother-sister incest in the evolution of political-economic forms. Through-
out his work, the great ethnographer called Trobriand leaders “chiefs”—
guya’u—stressing the ranking between chiefly and nonchiefly subclans.
Writing before the term chief stabilized at Fried’s definition, Malinowski
followed the discourse of the day that labeled any “native leader” as a
chief. He viewed a man like To’uluwa of Omarakana as holding a position
equivalent, or nearly so, to those of the chiefs of the smaller Polynesian so-
cieties—perhaps the Tikopia leader, the Ariki Kafika.
This usage was disputed by J. P. Singh Uberoi, who concluded that Tro-
briand guya’u more closely resembled the Melanesian “big man,” a non-
heritable position dependent on continuous political-economic competi-
tion.74 It can easily be shown that Oceanic peoples invented many local
leadership forms that fit uneasily into the typologies of 1960s evolutionary
approaches. Trobriand guya’u seem to me to be one of them. What Mali-
nowski saw in the 1920s may well have been a system of leadership that
lucky and clever men could tip from classical Melanesian nonheritable
power to something much more like enduring Polynesian chiefly status for
a descent group—at least for a time. Malinowski emphasizes (though he
does not identify) one mechanism in such a process: men’s inclination to
dispossess their proper heirs—their sisters’ sons—of inheritances in favor
of their wives’ sons, whom they had helped rear and come to love.
In such a labile situation (made somewhat more so, surely, by the colo-
nial setting), were high-status Trobrianders experimenting with brother-
sister incest as another such mechanism, a shockingly visible index of
power? The possibility is almost documented.
 Refining the Incest Taboo 155

The Malasi clan claims to rank higher than all others, and among them,
the members of subclan Tabalu take first place: “they are the real chiefs, ac-
knowledged to be of supreme rank, not by the Trobriands only, but by the
adjoining areas as well.”75 They prefer to marry within other Malasi sub-
clans, often with male paternal first cousins, a relationship “regarded some-
what askance” by most Trobrianders.76 Interestingly, in view of Small’s
findings about the polarizing effect of in-marriage, the Malasi also includes
a subclan whose members “were the most despised in the entire region (de-
spite their superior industry and artistry), forced into endogamy with their
own kind.”77 We are reminded of the castelike and maritally endogamous
sociopolitical relations in Micronesia between high-ranking Yapese and the
lowly, industrious Ulithi, who supplied Yap with fine handicrafts in return
for essential raw materials.78
Malinowski also tells us,
Of the four [clans], the Malasi have the reputation of being the most persistent ex-
ogamy-breakers and committers of incest. All the incestuous marriages on record
have happened within this clan; and I was told that this was not an accident but
that only the Malasi and no other clan will tolerate such marriages. The myth of in-
cest . . . is associated with the Malasi, and so also is the magic of love and the magic
to frustrate incest disease.79
One of the previous paramount chiefs, Purayasi, was known to have lived with his
sister; and another one, Numakala, is also strongly suspected by history of this
felony. They, of course, belonged to the Malasi clan; and there can be no doubt that
with them, as with so many other dynasties and famous rulers, the feeling of power,
of being above the law, served as a shield from the usual penalties. And, as histori-
cal figures, they and their doings would not so easily lapse into oblivion as in the
case of commoners.80

When Stevan Harrell defines historically derived political-economic types,

he argues for the necessity of families to adapt to social and ecological
structures over which family strategies have little leverage. Erickson has
pointed out (personal communication, March 8, 2001) that using the
theater-state tactic of brother-sister incest, a small, powerful family might
well “strengthen its political hold over the larger group of families.” Under
the right circumstances, agency operates, with change in family practice
triggering political-economic transformation.
We may now conclude that although incest taboos vary widely, they are
necessarily responsive to an evolutionarily driven, biologically based aver-
sion for associates of the first few years of life, who are usually members of
the nuclear family. At considerable human cost, the aversion may be over-
ridden.81 Local adaptation to material constraints and cultural repertory
will result in considerable variation in incest taboos. Taboos vary greatly
(though not infinitely) in both their cultural conformity to the aversion and
156 gates

in their occasional dramatic inversion of it in normative brother-sister mar-

riage. Such marriages occur almost exclusively in societies where ranking is
sufficiently developed to exempt high-status people from most manual la-
bor, as in chiefdoms, or to give them class rights over society’s basic means
of production, as in states. And they are unlikely to occur where state-
building had created sufficient mundane power to relieve the ruling class of
most of its supernatural aura. Under conditions not fully mapped out, but
surely recurrent in human history, our innate alertness to the emotional
complexity of incest can be turned to precise political ends, until something
more dependable comes along.

notes
1. My thanks to Larry Arnhart, Patrick Bateson, Alan Bittles, and Mark
Erickson for improving comments on my earlier draft.
2. Perry Anderson, Considerations on Western Marxism (London: New Left
Books, 1976).
3. Claude Lévi-Strauss, The Elementary Structures of Kinship, trans. James
Harle Bell and John Richard von Strurmer, and ed. Rodney Needham (Boston:
Beacon Press, 1969), p. xxxix.
4. Marshall Sahlins, “Evolution: Specific and general,” in Evolution and
Culture, ed. Marshall Sahlins and Elman R. Service (Ann Arbor: University of
Michigan Press, 1960), pp. 12, 38.
5. For example, David Schneider, American Kinship: A Cultural Account,
2nd ed. (Chicago: University of Chicago Press, 1980).
6. David Schneider, “The meaning of incest,” Journal of the Polynesian
Society, special issue on Incest Prohibitions in Micronesia and Polynesia (1976),
pp. 149–70.
7. For example, Marvin Harris, Our Kind: Who We Are, Where We Come
From, Where We Are Going (New York: Harper and Row, 1989). Cited in Paul R.
Ehrlich, Human Natures: Genes, Cultures, and the Human Prospect (Washington,
D. C., 2000), p. 201.
8. See Adam Kuper, Culture: The Anthropologists’ Account (Cambridge: Har-
vard University Press, 1999).
9. Clifford Geertz, “Deep hanging out,” New York Review of Books, October
22, 1998.
10. Bronislaw Malinowski, Argonauts of the Western Pacific: An Account of
Native Enterprise and Adventure in the Archipelagoes of Melanesian New Guinea
(London: Routledge and Kegan Paul, 1922); Crime and Custom in Savage Society
(London: Kegan Paul, Trench, Trubner & Co., 1926); Sex and Repression in Sav-
age Society (London: Kegan Paul, 1927); The Sexual Life of Savages in North-
Western Melanesia: An Ethnographic Account of Courtship, Marriage, and Family
Life Among the Natives of the Trobriand Islands, British New Guinea (London:
 Refining the Incest Taboo 157

George Routledge & Sons, 1929); Coral Gardens and Their Magic, 2 vols. (Lon-
don: George Allen & Unwin, 1935).
11. Bronislaw Malinowski, Magic, Science, and Religion and Other Essays,
ed. Robert Redfield (Glencoe, Ill.: The Free Press, 1948); A Diary in the Strict
Sense of the Term (New York: Harcourt, Brace, and World, 1967).
12. Malinowski, Sex and Repression, p. vii.
13. J. P. Singh Uberoi, Politics of the Kula Ring: An Analysis of the Findings
of Bronislaw Malinowski (Manchester: University of Manchester Press, 1962);
Marshall Sahlins, Stone Age Economics (Chicago: Aldine, 1972).
14. Annette B. Weiner, Women of Value, Men of Renown: New Perspectives
in Trobriand Exchange (Austin: University of Texas Press, 1976); “Stability in
banana leaves: Colonialism, economics, and Trobriand women,” in Women and
Colonialization: Anthropological Perspectives (New York: J. F. Bergin, 1980),
pp. 270–93; The Trobrianders of Papua New Guinea (New York: Holt Rinehart,
and Winston, 1988).
15. Weiner, The Trobrianders, p. 5.
16. For example, Malinowski, Argonauts, p. 399; Crime and Custom,
pp. 116–17.
17. Weiner, The Trobrianders, pp. 25–31.
18. David Labby, The Demystification of Yap (Chicago: University of Chicago
Press, 1976).
19. Karen Sacks, Sisters and Wives: The Past and Future of Sexual Equality
(Urbana: University of Illinois Press, 1982).
20. Malinowski, The Sexual Life of Savages, pp. 425–26.
21. Ibid., p. 426.
22. Ibid., pp. 428–29.
23. Ibid.
24. Malinowski, Argonauts.
25. These are true yams (Dioscorea spp., an ancient regional cultigen), not
sweet potatoes (Ipomoea spp., a post-Magellanic introduction). The continued
use of yams in a major ritual of interdependence long after the adoption of many
other staples suggests considerable age for this complex of sibling provisioning.
26. Weiner, The Trobrianders of Papua New Guinea, pp. 119–23.
27. Malinowski, The Sexual Life of Savages, p. 7.
28. Ibid., p. 440.
29. Ibid., p. 520.
30. Ibid., p. 519.
31. Even though Trobrianders passed on this shared myth, to assume that
each individual “believed” literally in it is to attribute the primitive mass mentality
among them that Malinowski was at pains to condemn and disprove.
32. Malinowski, The Sexual Life of Savages, p. 384.
33. Ibid., pp. 447, 450. These subclans are almost certainly those of the un-
usual Malasi clan, of whom more will be said below.
34. Ibid., p. 430.
35. Ibid., p. 433.
36. Ibid., p. 440.
158 gates

37. Ibid., p. 441.

38. Ibid., pp. 384, 447.
39. Ibid., p. 446.
40. Ibid., pp. 74, 265, 445–48.
41. Ibid., p. 447.
42. Ibid., p. 390.
43. Ibid., p. 389.
44. Ibid., p. 424.
45. Ibid., p. 389.
46. Malinowski, Crime and Custom, p. 80.
47. Ibid., p. 81.
48. Malinowski, The Sexual Life of Savages, p. 439.
49. Malinowski, Crime and Custom, pp. 77–83.
50. Malinowski, The Sexual Life of Savages, p. 476.
51. Malinowski, Crime and Custom, pp. 118–19, 92.
52. Ibid., pp. 93–97.
53. Malinowski, The Sexual Life of Savages, p. 399.
54. Patrick Bateson, “Optimal outbreeding,” in Mate Choice, ed. Patrick Bate-
son (Cambridge: Cambridge University Press, 1983). Also Chapter 1 of this volume.
55. Raymond Firth, We, the Tikopia: A Sociological Study of Kinship in Primi-
tive Polynesia (London: Allen and Unwin, 1936), p. 340.
56. Weiner, The Trobrianders, pp. 8–9.
57. William Durham, Coevolution: Genes, Culture, and Human Diversity
(Stanford, Calif.: Stanford University Press, 1991), p. 315.
58. Ibid., p. 333.
59. Ibid., p. 338.
60. Ibid.
61. Ibid., p. 357.
62. Durham is struck by the wide occurrence of the idea that incest results
in defective children; this perception informs decision-making individuals, who
then make culture of it. I find this a weak argument, not only for the traditional
reasons (would small, sparsely distributed, sexually generous, and heavily inbred
populations been able to interpret the genetic consequences of their matings?).
Durham’s interpretation may be a particularly good example of a problem that
haunts thinkers from contracepting societies: we assume all too easily a distinction
between recreational and reproductive heterosexual intercourse. Through most of
human experience, children were an expected consequence of such intercourse, a
part of the sexual experience for women at least. Having sex and making babies
(even when the participants would rather not make a baby) were inextricably
linked. Ethnography rapidly piles up examples of how various peoples contextual-
ized intercourse (in the yam garden, often, after eating particular foods, according
to ritual calendars) in order to ensure good-quality offspring. It is only a short
conceptual step to predicting that “bad,” incestuous sex would result in not-good
children. I suggest that hypertrophied cultural logic rather than assessment of
actual occurrence of inbreeding depression prompted the connection so often
found in Durham’s sample.
 Refining the Incest Taboo 159

63. Durham, Coevolution, p. 359.

64. Mark Erickson, “Incest avoidance and familial bonding,” Journal of
Anthropological Research, vol. 45 (1989), pp. 280–95; “Rethinking Oedipus:
An evolutionary perspective on incest avoidance,” American Journal of
Psychiatry, vol. 150 (1993), pp. 413–30; and Chapter 9 of this volume.
65. Jack Goody, “A comparative approach to incest and adultery,” in Com-
parative Studies in Kinship, ed. Jack Goody (Stanford, Calif.: Stanford University
Press, 1969), chapter 2.
66. Marvin Harris, Cows, Pigs, Wars, and Witches: The Riddle of Culture
(New York: Random House, 1974), p. 5. Harris traveled a paradigmatic intellec-
tual road from a truly vulgar materialism to a kind of Marxism that he labeled
cultural materialism. To the end of his life, however, he sought to exclude human
biology from this dialectic.
67. Morton H. Fried, The Evolution of Political Society: An Essay in Political
Anthropology (New York: Random House, 1967); Jack Goody, Production and
Reproduction (Cambridge: Cambridge University Press, 1976); Jack Goody, The
Oriental, the Ancient, and the Primitive: Systems of Family and Marriage in the
Pre-Industrial Societies of Eurasia (Cambridge: Cambridge University Press, 1990).
68. Stevan Harrell, Human Families (Boulder, Colorado: Westview Press,
1997), p. 27.
69. Ibid., pp. 129–30.
70. John M. Goggin and William C. Sturtevant, “The Calusa: A stratified,
nonagricutlural society (with notes on sibling marriage),” in Explorations in
Cultural Anthropology, ed. Ward H. Goodenough (New York: McGraw-Hill,
1964), pp. 179–220.
71. Ibid., pp. 204–6. Goggin and Sturtevant list eight “societies where (some
type of) sibling marriage is not defined as incest for anyone” (p. 204). One of
these at least—Hoklo—is easily eliminated; south Chinese minor marriages were
made between adoptive, not biological, brothers and sisters. Of the others, the
best known is that of Roman-period Egyptian commoners among whom full-
brother-sister marriage was normative. Evidence for this society is strong but
sparse. Its exceptionality is discussed by Wolf, who computes that, of 113 cases,
only a dozen are demonstrably full siblings who can most reliably be assumed to
have been reared together. The complexities of women’s legal status as Hellenized
Egypt came under Roman rule suggest that strong pressure for sibling marriage
may have emerged in a conjuncture of status-climbing, ethnic, property-right, and
marriage-law variants. See Keith Hopkins, “Brother-sister marriage in Roman
Egypt,” Comparative Studies in Society and History, vol. 22 (1980), pp. 303–54;
Sarah B. Pomeroy, Women in Hellenistic Egypt: From Alexander to Cleopatra
(Detroit: Wayne State University Press, 1990); and Chapter 5 of this volume.
72. Clifford Geertz, Nagara: The Theater State in Nineteenth Century Bali
(Princeton, N.J.: Princeton University Press, 1980.
73. Cathy Small, “The political impact of marriage in a virtual Polynesian
society,” in Dynamics in Human and Primate Societies: Agent-Based Modeling
of Social and Spatial Processes, ed. Timothy A. Kohler and George J. Gumerman
(New York: Oxford University Press, 2000), p. 243.
160 gates

74. Uberoi, Politics of the Kula Ring, pp. 31–36.

75. Malinowski, The Sexual Life of Savages, p. 420.
76. Ibid., pp. 385 and 447.
77. Ibid., p. 420.
78. William A. Lessa, Ulithi: A Micronesian Design for Living (New York:
Holt, Rinehart, & Winston, 1966).
79. Malinowski, The Sexual Life of Savages, pp. 432, 456–74.
80. Ibid., pp. 474–75.
81. Arthur P. Wolf and Chieh-shan Huang, Marriage and Adoption in China,
1845–1945 (Stanford, Calif.: Stanford University Press, 1980); Arthur P. Wolf,
Sexual Attraction and Childhood Association: A Chinese Brief for Edward
Westermarck (Stanford, Calif.: Stanford University Press, 1995); and Chapter 4
of this volume.
9 Evolutionary Thought and the Current
Clinical Understanding of Incest

Mark T. Erickson

A distinct recollection of my psychiatric training in the 1980s is of a

group of patients whose bewildering array of symptoms defied diagnosis.
The records of these patients, who were mostly women, showed previous
psychiatrists were equally puzzled. About this time surveys revealed the
prevalence of incest to be far greater than imagined. This discovery gener-
ated a rapidly expanding research into the effects of incest.1 Studies docu-
mented that incest, and frequently associated childhood neglect and physical
abuse, predisposed individuals to a myriad of difficulties including posttrau-
matic stress disorder, chronic depression, alcohol and drug dependence, self-
mutilation, suicide attempts, borderline personality disorder, anxiety disor-
ders, somatoform disorders, bulimia nervosa, and dissociation.2
Patients who had been puzzling to clinicians, and to themselves, became
more understandable. A fortunate consequence is that treatment of those
suffering the effects of incest, and other trauma, has improved immeasur-
ably.3 Ironically, as clinicians found the prevalence of human incest to be
greater than had been believed, biologists discovered that incest is rare in
nature.4 Further, anthropologists convincingly demonstrated that humans
inherit an evolved propensity to avoid incest.5
Incest avoidance presumably evolved, in both human and nonhuman
species, because of the harmful effects of close inbreeding.6 Incest avoid-
ance is not hardwired, or present at birth, but rather depends on close as-
sociation between kin from early life. It is susceptible to disruption. Species
that naturally avoid incest, including humans, are far more likely to engage
in incest if early association is interfered with.7
The central purpose of this chapter is to reconcile the biology of incest
avoidance with our current clinical understanding of incest. To begin, I
briefly review literature on the biology of incest avoidance (more extensive
reviews are found throughout this volume). Clinical research on incest will
then be discussed within this broadened context. Examined in this manner,
162 erickson

an integrative model of incest avoidance, as a biological adaptation, and in-

cest, a pathological manifestation, emerges. The implications of this hy-
pothesis are discussed. Many clinicians believe we are in the midst of an
epidemic of incest. Although this assertion is difficult to prove, there is rea-
son to believe that cultural practices, nonexistent in our evolutionary past,
may disrupt our natural propensity for incest avoidance. This may increase
incest prevalence—particularly the incestuous abuse of children.

Evolution and Incest Avoidance: A Brief View

For most of the twentieth century, it was widely accepted that a cul-
tural rule, an incest taboo, was essential for inhibiting incest.8 This view
rested on the assumption that animals and precultural humans mated in-
cestuously. Freud took this notion to its logical conclusion when he argued
that humans are, by nature, incestuous. He proposed that repression of in-
cestuous impulses created a universal neurosis, unique to our species. Freud
called this neurosis the Oedipus complex.
Edward Westermarck, a Finnish anthropologist and contemporary of
Freud, presented a strikingly different hypothesis.9 Aware of the harmful ef-
fects of close inbreeding, Westermarck believed that through natural selec-
tion humans had acquired an aversion to incest. Westermarck, presciently,
hypothesized that close association from early life established a later pro-
pensity for incest avoidance. Because children are raised in close proximity
to parents, and siblings, in virtually all traditional cultures, his hypothesis
was plausible.
Incest was not studied in nature until the 1960s. Given the long-held be-
lief that it was common, primatologists studying rhesus monkeys were
openly surprised when mother-son incest was found to be rare.10 It has since
been shown that incest is rare in other primate, mammalian, avian, amphib-
ian, and even insect species.11 With few exceptions, incest is uncommon
throughout the animal kingdom.12
Consistent with Westermarck’s hypothesis, early association between
kin is essential to establishing incest avoidance in animal species.13 Prairie
voles, for example, rarely mate incestuously when reared naturally with
siblings. If, however, siblings are separated at birth into foster litters, as
adults they sexually avoid unrelated foster siblings but mate incestuously
with unfamiliar biological siblings.14
The major obstacle to studying incest avoidance in humans is the incest
taboo. Virtually all societies share this ostensibly culturally constructed
rule, making it difficult to isolate biological influences.15 To test Wester-
marck’s hypothesis, it was necessary to find circumstances of early associa-
 Evolutionary Thought and Clinical Understanding 163

tion without a later taboo on sexual affiliation. Two particularly useful test
cases were found, one in Israel and the other in rural Taiwan.16
On Israeli communal farms, or kibbutzim, children of the same age were,
until recent decades, raised together in a children’s house from shortly after
birth through high school graduation. Aside from visits with parents in the
evening, children remained together both day and night. As they matured,
no cultural rule opposed their sexual affiliation. If anything, peers were en-
couraged to date and marry. Because of early association, Westermarck’s
hypothesis predicts that peers should be sexually avoidant. This is what has
been observed. Sexual relationships, whether through dating or marriage,
are extremely uncommon between cosocialized peers.17
The most complete test of Westermarck’s hypothesis is occurring in Tai-
wan, where Stanford anthropologist Arthur Wolf studies simpua, or “mi-
nor” marriage.18 In minor marriage, the bride, or simpua, is usually be-
trothed in infancy and raised with her future husband in the groom’s home.
The couple is married in their midteens. Again, because of early association,
Westermarck would predict that minor marriage couples will be averse to
sexual affiliation, despite countervailing cultural pressures. Consistent with
this prediction, Wolf has found that minor marriage couples have a much
higher divorce rate, engage in adultery more frequently, and also have fewer
children per year of marriage than couples in other arranged marriages.
Wolf has meticulously ruled out alternative explanations of his data and
concludes that Westermarck’s hypothesis best comprehends the Taiwan find-
ings.19 Wolf’s study indicates that the Westermarck effect, or incest avoid-
ance, is dependent on close association during a sensitive period of about
the first three years of life.20
As the twentieth century has come to a close, Westermarck’s hypothesis
has far more objective support than any alternative model. Incest avoidance
studies provide a remarkably thorough test of this evolutionary hypothe-
sis.21 Nevertheless, clinicians are well aware that early association alone is
not sufficient to establish incest avoidance. To the contrary, most incest oc-
curs despite association. Given early association, are there definable influ-
ences that disrupt our natural propensity for incest avoidance? It is at this
juncture where clinical research provides insights not found elsewhere.

Incest: Current Clinical Findings

prevalence

Historically, estimates of the frequency of incest have varied to an

extraordinary degree. In 1953, Alfred Kinsey et al. reported the prevalence
of incest within the United States to be about one case per hundred.22 In
 table 9.1
Estimated Prevalence of Incest, 1953–1996
Year of Estimated Prevalence
Study Publication of Incest (percent)

A. C. Kinsey et al., Sexual Behavior in the Human 1953 1.0 (all types)
Female (Philadelphia: W. B. Saunders).

S. K. Weinberg, Incest Behavior (New York: 1955 0.0001 (all types)

Citadel).

D. Finkelhor, “Sex among siblings: A survey 1980 3.2 (sibling)a

of prevalence, variety, and effects,” Archives
of Sexual Behavior, vol. 9, pp. 171–94.

D. E. H. Russell, “The incidence and prevalence 1983 2.9 (father-daughter)

of intrafamilial and extrafamilial sexual 2.2 (sibling)
abuse of female children, Child Abuse
and Neglect, vol. 7, pp. 133–46.

G. Wyatt, “Sexual abuse of Afro-American 1985 1.6 (father-daughter)

and white-American women in childhood,”
Child Abuse and Neglect, vol. 9, pp. 507–19.

A. W. Baker and S. P. Duncan, “Child sexual 1985 1.3 (all types)a

abuse: A study of prevalences in Great Britain,”
Child Abuse and Neglect, vol. 9, pp. 457–67.

D. Finkelhor et al., “Sexual abuse in a national 1990 0.8 (all types)a

survey of adult men and women: Prevalence,
characteristics, and risk factors,” Child Abuse
and Neglect, vol. 14, pp. 19–28.

D. Finkelhor and J. Dziuba-Leatherman, 1994 0.9 (all types)a

“Children as victims of violence: A national
survey,” Pediatrics, vol. 94, pp. 413–20.

H. Sariola and A. Uutela, “The prevalence 1996 0.2 (father-daughter)

and context of sexual abuse in Finland,”
Child Abuse and Neglect, vol. 20, pp. 843–50.

D. S. Halperin et al., “Prevalence of child 1996 0.9 (all types)b

sexual abuse among adolescents in Geneva:
Results of a cross sectional survey,” British
Medical Journal, vol. 312, pp. 1326–29.
a
In Finkelhor’s 1980 study, 13 percent of the entire sample of 796 undergraduate
students reported some type of sexual experience with a sibling (15 percent of females
and 10 percent of males). Most of these experiences were regarded as childhood sex
play. About 25 percent of this group, or a prevalence of about 3.2 percent, had experi-
ences regarded as exploitative because force was used or because there was a large age
difference between siblings. In this study and those by Baker and Duncan and by
Finkelhor et al., the precise degree of relatedness was not specified. These data may
include incest involving kin outside the immediate family or involving step-parents or
step-siblings.
b
All data cited from Halperin et al., “Prevalence of Child Sexual Abuse,” are unpublished
findings that were derived from the authors’ 1996 study.
 Evolutionary Thought and Clinical Understanding 165

1955, Kirson Weinberg estimated the incidence at one case per million.23
This estimate persisted and was cited as recently as 1975 in a widely used
textbook of psychiatry.24
Since 1980 the prevalence of incest has been studied extensively. Table
9.1 summarizes the largest studies; combined, 17,045 individuals were sur-
veyed (9,391 females and 7,654 males). As best possible, only incest within
the nuclear family is included in Table 9.1. Step-fathers, uncles, cousins,
and other nonimmediate kin were excluded. The reason for this exclusion
was to find an approximation of the prevalence of incest, despite presumed
early association, thereby providing a measure of how often the Wester-
marck effect, or incest avoidance, fails. It must be noted, however, that
step-fathers, who are unlikely to be in close early association with step-
daughters, are much more likely to perpetrate incest than biological fa-
thers.25 Further, the prevalence of early sexual abuse, from all possible
sources, has been found to be far greater than imagined. Russell, for exam-
ple, found that 28 percent of women, from a random sample, reported
some form of sexual abuse before the age of fourteen.26 Although child sex-
ual abuse in general is not the focus of this chapter, this issue will be con-
sidered at the conclusion of the chapter.
The primary intent of prevalence surveys has been to determine the fre-
quency of intrafamilial sexual encounters of any kind. The definition of in-
cest therefore tends to be rather broad, often including everything from ex-
hibitionism, to fondling, to intercourse. Of the studies listed in Table 9.1,
five took place in the United States, one in Britain, one in Finland, and one
in Switzerland. Prevalence studies are often very different in design. For ex-
ample, some researchers surveyed adolescents at school with a pencil and
paper questionnaire.27 Others surveyed adults in a direct, face-to-face in-
terview.28 Telephone surveys were also used.29 Despite methodological dif-
ferences, the prevalence findings are in far closer agreement than earlier es-
timates.30 Given the exceptionally sensitive nature of this subject it may be
difficult to get more accurate findings.
Some surveys sought the prevalence of the most severe forms of incest.
Daniel Halperin et al. found the prevalence of incest in which penetration
occurred to be 0.3 percent.31 Diana Russell defined the most severe form of
incest as completed or attempted intercourse or oral sex.32 Using this defi-
nition, she found a prevalence of 0.8 percent for both father-daughter and
brother-sister incest. Anthony Baker and Sylvia Duncan found the preva-
lence of sexual intercourse “with a blood relative” to be 0.25 percent.33
“Blood relative” was not clearly defined, so this finding may include per-
sons beyond the immediate family. There are no studies that specifically
examine the prevalence of mother-son incest in a general population. Mc-
Carty, however, found that 4 percent of a population of convicted sex of-
166 erickson

fenders in a metropolitan treatment program were mothers who had sexu-

ally abused offspring.34
Several studies have examined the frequency of pregnancy caused by in-
cest. Within the United States, estimated rates have varied from 1 percent
to 20 percent. There appears to have been a decrease in incest pregnancy in
recent decades as birth control has become more widely available.35
Incest prevalence data from non-Western cultures is virtually nonexis-
tent. David Finkelhor has reviewed difficult-to-obtain data from twenty-
one countries (i.e., unpublished or published in difficult-to-obtain sources).
He found that across cultures females are most often the victims of incest,
males are more typically perpetrators, and a history of sexual abuse is fre-
quently associated with subsequent mental health impairment. No infor-
mation on prevalence from non-Western cultures is given.36

incest with and without early association

To date, with few exceptions, clinicians have not used findings on the
biology of incest avoidance to conceptualize clinical studies of incest.37 The
robust support for Westermarck’s hypothesis indicates that the presence, or
lack, of early association should be particularly informative. Questions not
often considered by clinicians are raised. For example, incest taboos not-
withstanding, does the lack of early association between kin make incest
more likely? Is incest without early association phenomenologically distinct
from incest with early association? Does incest with early association have
recurrent characteristics? If so, do these characteristics suggest why the
Westermarck effect is not invariable? Taken as a whole, clinical findings sug-
gest answers to all of these questions.
Violations of the Westermarck Effect:
Incest with Early Association
Incestuous families often present a facade of respectability and may
be overtly conventional to a fault.38 With closer inspection, however, their
apparent well-being is illusory. Clinicians have repeatedly found that in-
trafamilial behavior of incest families is pervasively disturbed.39 In a con-
trolled study, Philip Madonna, Susan Van Scoyk, and David P. Jones found
that on a standardized family evaluation scale incest families tended to rate
in the severely dysfunctional range. Among nonincest control families,
there were clear boundaries between individuals, allowing for appropriate
intimacy; these boundaries were far less distinct in incest families. Incest
families were inefficient at resolving conflict. They were lacking in empa-
thy. Parents tended to be neglectful, emotionally unavailable, and unable to
support autonomy in offspring.40 Alcoholism, drug abuse, and marital dis-
 Evolutionary Thought and Clinical Understanding 167

cord are more common in incest families.41 Incest families tend to have a
greater-than-average number of children.42
In father-daughter incest, fathers not infrequently offer their daughter
gifts and special privileges to gain favor. This special relationship may be
the daughter’s only source of affection. When fathers pursue incest, how-
ever, the daughter’s experience is almost invariably one of fear, disgust, dis-
belief, confusion, anger, and shame.43 In her study of father-daughter incest,
Patricia Phelan found that none of the daughters initiated the activity or
enjoyed what happened.44 Given early association, it is only on rare occa-
sions that incest appears to be emotionally acceptable to a daughter.45
Mothers, consciously or unconsciously, are often complicit with father-
daughter incest. In subtle ways a mother may encourage, or at least not dis-
courage, her husband’s incestuous behavior. Daughters who confide in their
mothers are usually bitterly disappointed. Mothers, even when fully aware
of paternal incest, often do not defend their daughters. Not surprisingly,
victims of father-daughter incest have overwhelmingly negative images of
their mothers. In Judith Herman’s study, thirty-nine of forty daughters who
were victims of father-daughter incest had extremely negative images of
their mothers, describing them as cold, indifferent, and ungiving. The only
exception was a daughter who had lost her mother in early childhood.46
Because mother-child incest is so uncommon, information had been ex-
tremely limited, consisting of little more than case reports until the studies
of Loretta McCarty and Kathleen Faller.47 Combined, these authors gath-
ered data on sixty mothers convicted of sexual abuse of offspring. Both
found patterns of extreme pathology within the family. Neglect and physi-
cal abuse often accompanied maternal incest.
Some writers have argued that social isolation and poverty may be criti-
cal factors underlying incest.48 Other studies contradict this notion.49 It has
been proposed that male dominance in a paternalistic society lies at the root
of incestuous behavior.50 More recent studies do not support this view.51
The most salient influence on incest behavior may be found in the child-
hood attachment experience of parents of incest families.52 Incestuous fa-
thers, for example, typically describe their childhood as filled with rejec-
tion, neglect, and physical and /or sexual abuse. Parental absence because
of death or abandonment is also common.53 In father-daughter incest fam-
ilies, the mother’s early experience is similarly bleak. She is likely to have
had an emotionally deprived childhood characterized by rejection and hos-
tility. A history of childhood sexual abuse is frequently found.54
McCarty presents the most extensive information on the childhood of in-
cestuous mothers. In her study nearly all mothers described an unremittingly
bleak childhood using adjectives such as “rough” or “horrible” to describe
168 erickson

their upbringing. Physical and /or sexual abuse were extremely common (95
percent) during the childhood of these mothers.55
Similar developmental conditions are observed in sibling incest. The child-
hood of the offender, most frequently an older brother, and the victim, usu-
ally a younger sister, is typically overwhelmingly grim. Mothers are de-
scribed as “emotionally absent,” “distant,” “inaccessible,” or “neglectful.”56
Likewise, fathers are often absent, through death or by abandonment or di-
vorce. If present, the father is usually emotionally distant.57 Naomi Adler
and Joseph Schultz found that 92 percent of boys who perpetrated incest
had been physically abused by one or both parents.58 Frequently, such boys
have also been sexually abused.59 The personal relationship between an of-
fending brother and his sister has been described as nonexistent except for
incest and physical abuse.60
A particularly appalling finding is that incest is often initiated very early
in the victim’s life. Father-daughter incest may begin when the daughter is
six years of age or younger.61 The average age of onset in father-daughter
incest is about eight to nine years.62 In a study of mother-child incest, the
mean age of victims at assessment was 6.4 years.63 In sibling incest the
brother typically initiates sexual abuse when he is between eleven and four-
teen years of age. The mean age of the sister at onset is about seven years.64
The early onset of much of human incest does not appear to be linked to
fixated pedophilia. Perpetrators of incest rarely limit their sexual attention
to children.65
Incest inflicted on children, and more broadly, child sexual abuse, ap-
pears to be a uniquely human behavior. This has not been observed in other
primate species.66 Given that the conditions for establishing incest avoid-
ance, association during a sensitive period, appears to be similar across
mammalian species, this variation is cause for concern. Incest perpetrated
on children may reflect a biological peculiarity of our species. A more plau-
sible explanation is that this propensity is due to cultural influences, rare or
nonexistent in our evolutionary past, which interfere with the development
of incest avoidance. The incest avoidance adaptation can be easily disrupted
in animal species by artificial intrusion.67 There is little reason to believe we
are an exception.
Incest Following Early Separation:
A Phenomenologically Distinct Entity
In his large study, Weinberg expressed particular interest in six pairs
of incestuous siblings.68 Each sibling of all six pairs desired the incestuous
relationship. There was no evidence of coercion on the part of the brother
as is usually the case in sibling incest. Though aware of an incest taboo,
these siblings were largely indifferent to this injunction. They lacked ap-
 Evolutionary Thought and Clinical Understanding 169

parent feelings of guilt concerning their relationship. Relationships often

began quickly and were passionate. Three of the couples eventually mar-
ried. Remarkably, every pair had been separated from early childhood and
were only later reunited.
In 1975 Britain enacted the Access to Birth Records Act. This act enables
adopted individuals over the age of eighteen to trace biological kin. An
unanticipated complication of the Birth Records Act has been the erotic
feelings frequently experienced by reunited kin. A study by Maurice Green-
berg and Roland Littlewood suggests that incest may frequently occur.69
They estimate that over 50 percent of reunited kin experience strong sexual
feelings. The following examples from their study are illustrative:
(A 22 year old civil servant meets his mother) “We meet, smile, kiss. . . . I notice her
nose, brow, deep set eyes; I am constantly looking for similarities. . . . She said ‘I’ve
got to touch you’ and touched my face with her hand. It felt nice. . . . Monday we
dressed up for each other. . . . She looked gorgeous, I felt teased, and she admitted
she was teasing me. . . . We kissed tinkering on the edge.”
(A 35 year old nurse meets her biological father) “It developed very quickly, we
hugged and kissed a lot that first weekend. His skin felt like mine and he smelled
like me. I had a sexual dream about him, wanting it. I just thought it was crazy but
discovered he was open to it.”

Many similar anecdotal reports exist. “Susan,” who was adopted at birth,
sought her biological parents at age twenty-two. Within months she located
her father. When asked about her reunion she commented, “There was an
immediate sense of recognition. He looked very like me . . . the face, the ges-
tures. . . . I knew I felt attracted to him which really scared me. I never talked
to him about it but I believe he felt the same.”70
Allen and Patty Muth, brother and sister, had an incestuous relationship
that spanned several years and resulted in the birth of four children. They
moved from one region of the United States to another, staying ahead of
criminal charges. The couple was eventually apprehended in Wisconsin,
convicted, and sentenced for the felony of incest. Mentioned virtually as an
aside in the popular press, was that the couple did not meet until Patty, the
younger sibling, was eighteen years of age.71
Victoria Pittorino found her brother, David Goddu, through birth records
of the State of Massachusetts. They were separated by adoption when Vic-
toria was three and David was one year old. Twenty years later they re-
united. Ms. Pittorino described her response as “love at first sight.” The at-
traction was mutual, and the couple married within weeks of reunion. Later,
they were arrested and convicted of incest under Massachusetts statutes that
originated in 1695.72
“Jackie,” an adoptee, found her biological brother. She related, “I felt ir-
170 erickson

resistibly, magnetically drawn. . . . My body reaction was to move toward

him. I wanted eye contact. I liked what I saw because he was like me.”73
At the age of twenty-five, the poet Byron entered what has been de-
scribed as a torrid affair with his half-sister Augusta Leigh. Those who saw
them together were struck by their resemblance. They met for the first time
in 1801, when Byron was fourteen and Augusta seventeen.74
Perhaps the most vivid description of father-daughter incest is found in
the diaries of writer Anaïs Nin. Although not completely separated from
her father, concert pianist Joaquin Nin, their early contact was significantly
limited because Joaquin was frequently on tour. When at home, he with-
drew from the life of his family. When Joaquin did notice Anaïs at all, it
was usually to call her the “ugly little girl.”75 When Anaïs was ten years
old, Joaquin Nin abandoned his family altogether. Twenty years later,
Anaïs and her father reunited. “But you are like me,” said Joaquin, who
then commented on the similarities of their hands and hair.76 They meet
again, at which time Joaquin states, “You are the synthesis of all women I
have loved. . . . I don’t feel towards you as if you were my daughter.” Anaïs
replied, “I don’t feel as if you were my father.” He responded, “I have fi-
nally met the woman of my life and it is my daughter. . . . I’m in love with
my own daughter.” Anaïs replied, “Everything you feel is reciprocal.”77
A recurrent story line in poems, novels, and plays in which incest is a
theme is one of early separation.78 Examples include The Book Bag by Som-
erset Maugham, Pierre, or the Ambiguities by Herman Melville, and The
Caryatids by Isak Dinesen. The myth of Oedipus is the best-known exam-
ple. Oedipus is separated from his mother, Jocasta, at birth. Many years
later they unknowingly reunite incestuously. Clinical and anthropological
findings now illuminate this myth in a way not anticipated by Freud. Oedi-
pus portrays the literal truth that early separation undermines a natural
adaptation for incest avoidance.
As predicted by Westermarck’s hypothesis, clinical studies show that in-
cest is far more likely if kin are separated early in life. It is, virtually, only
in this circumstance that incest may be mutually desired and eventuate in
marriage.79 Taboos appear to have limited influence. By contrast, given
early association, incest is rarely, if ever, mutually desired. It is perpetrated
coercively, by fathers or brothers, and experienced as being intensely aver-
sive by daughters or sisters.
One clinical finding not predicted by Westermarck is the extraordinary
fascination reunited kin show for each other. They often describe an imme-
diate sense of recognition. They notice that they smell alike, that their ges-
tures are similar, that they “resemble each other in ways that transcend
physical similarities.”80 The basis of this unusually intense attraction is far
from clear, but it may derive from a tendency, observed in many species, to
 Evolutionary Thought and Clinical Understanding 171

respond preferentially to similarities between self and others. Biologists call

this “phenotypic matching.”81
Many species, including humans, quickly learn to identify familial char-
acteristics; at birth, mothers typically adopt an en face orientation for close
inspection of their infant. Within a very short time, they can reliably iden-
tify their offspring through visual, auditory, olfactory, or tactile cues.82 Sim-
ilarly, fathers appear to have a remarkable ability to identify offspring soon
after birth. With an average of less than seven hours of postnatal contact,
blindfolded fathers were able to recognize offspring by touch of the infant’s
hand alone.83 Infants also quickly develop a preference for the parent’s phe-
notype. Two-week-old breast-feeding infants, for example, are capable of
recognizing their mother by olfactory cues alone.84 Under normal develop-
mental circumstances, early association generates a later sexual preference
for someone who is rather different from close kin.85 Separated kin, in con-
trast, may be extraordinarily intrigued by their similarities because of an
inherent tendency for phenotypic comparison or matching.86 Because of
their lack of early association, however, a sense of kinship and incest avoid-
ance are not established. Joaquin and Anaïs Nin, for example, recognized
that they did not “feel” as though they were father and daughter, yet ex-
pressed intense interest in their physical similarities and engaged in incest.
A second clinical finding not anticipated by Westermarck’s hypothesis is
the very high frequency of abuse and neglect during the childhood of the
parents of incest families. This suggests that the propensity for later incest
may be influenced very early in life by the quality of attachment relation-
ships. This notion will be developed in the next section.

Incest and the Psychopathologies of Kinship

All biological adaptations are susceptible to pathology. Bones break,

the blood supply to the heart becomes occluded, immune systems attack
the bodies they were designed to protect. Incest avoidance, a biological
adaptation, is no exception.
Understanding normal development of any adaptation invariably leads
to a better understanding of pathology. Westermarck’s hypothesis says lit-
tle about the development of incest avoidance aside from the necessity of
early association. Can more be said? Incest avoidance might, for example,
develop independently with no connection to other classes of social behav-
ior. If this were the case, however, incest should be observed as an isolated
event in otherwise healthy families. Clinical studies are clearly contradic-
tory. Incest occurs, overwhelmingly, in grossly disturbed families in which
neglect, abandonment, and physical abuse are also common. Incest may
172 erickson

then reflect pathology in an adaptation that broadly modulates adaptive

behaviors of kinship.

evolution and kinship

For Freud and many of his contemporaries, kinship was a uniquely

human phenomenon resting precariously on a cultural taboo that restrained
incestuous and aggressive instincts. A revolutionary change in our under-
standing of kinship and its evolutionary underpinnings began in the 1960s
with the work of biologists W. D. Hamilton, John Maynard Smith, and
G. C. Williams.87 A useful, if somewhat oversimplified, summation of their
argument is that natural selection maximizes the ability of individual organ-
isms, not species, to gain genetic representation in future generations.88 Be-
cause individuals are genetically more similar to kin than nonkin, this notion
predicts that individual organisms will maximize their genetic representation
by showing preference to kin.
Altruistic behavior is the most obvious class of behavior predicted to oc-
cur between close kin. Parents who, for example, show a preference for
their own offspring in providing food, shelter, and so forth should more ef-
fectively maximize their genetic representation than parents who show no
such preference. Attachment behaviors of the young are predicted to be
kin-directed because kin should be most inclined to respond altruistically
to offspring. Sexual (incest) avoidance is a third class of behavior that
should be kin-directed.
Behavioral studies in recent decades have revealed that the phylogenetic
roots of kinship go far deeper than believed by Freud. From the kin-
directed warning calls of Belding ground squirrels, to exclusive maternal
preference for offspring in California sea lions, to nepotism among rhesus
monkeys, studies have repeatedly shown that kinship powerfully predicts
patterns of social behavior.89 Altruistic behaviors have been found to occur
predominantly between immediate kin.90 British psychiatrist John Bowlby
found that in humans attachment bonds are, across cultures, directed pri-
marily to biological parents.91 Incest avoidance, as previously discussed,
has been documented in a vast range of species including humans. The dis-
covery that social organisms, from insects to primates, respond preferen-
tially and nonsexually to kin raises questions which are ultimately of great
clinical importance. If kinship is not a cultural construct, but rather has an
evolutionary history that long precedes societal taboos and our species, then
how do kin recognize one another? How do the distinctive patterns of kin-
ship behavior develop?
Patrick Bateson of Cambridge University initially proposed that sensitive
periods, long described by ethologists, could play an essential role in the de-
 Evolutionary Thought and Clinical Understanding 173

velopment of kin-recognition.92 Indeed, association during sensitive periods

now appears to be the most common mechanism by which kin come to rec-
ognize one another.93 This process is overtly simple. Association during an
early sensitive period canalizes stable social bonds from child to parent,
parent to child, and sibling to sibling. Over a century ago Westermarck pro-
posed that early association was the foundation of human incest avoidance.
It now seems that early association functions far more broadly, canalizing
different classes of kin-directed behavior, including attachment, altruistic
behavior, and incest avoidance, across many species.

incest avoidance and kinship

Taking development one step further raises the question of whether

different classes of kin-directed behaviors function independently or, con-
versely, are integrated into an overarching system. For reasons both theo-
retical and empirical, Arthur Wolf and I have hypothesized that kin-
directed behaviors are either integrated or develop so closely that, for all
intents, they are functionally integrated.94 This proposal significantly ex-
pands upon Westermarck’s hypothesis and leads to new, testable predic-
tions about the relationship between incest avoidance, attachment, and
kin-directed altruism.
Theoretically, if evolutionary forces increase the social interdependence
of kin, thereby prolonging affiliation, the adaptive significance of incest
avoidance increases accordingly. The coevolution of attachment, kin-
directed altruism, and incest avoidance therefore seems probable.95 To give
two suggestive examples, male pilot whales swim in their mother’s group,
or pod, throughout their adult life. There they remain in close proximity to
their mother and other kin, including mature sisters. Males leave their pod
only transiently, to mate. Lifelong association notwithstanding, genetic
studies of entire pods show that incest is rare, if not nonexistent.96 Male
bonobo chimpanzees develop particularly strong attachments to their mother
and maintain close ties throughout their lives. Mature mother-son pairs mi-
grate together and are mutually supportive. Despite ongoing close associa-
tion, and the active sexual lives of bonobos, incest between mother-son
pairs is rare.97
A second theoretical point comes from the work of French geneticist
Francois Jacob.98 Jacob, in a classic lecture given at Berkeley, observed that
“evolution does not produce novelties from scratch” but instead utilizes
what already exists, transforming this into more elaborate systems. Ac-
cording to Jacob, it would be unlikely that separate mechanisms of kin
recognition evolved for each class of familial behavior. It is more probable
that once an adaptation for kin recognition evolved (association), this was
174 erickson

utilized by natural selection to canalize different classes of kin-directed be-

havior, including attachment, kin-directed altruism, and incest avoidance.
A third reason to suspect that kin-directed behaviors are integrated comes
from attachment research. Secure attachment develops when parents are re-
sponsive to an infant’s needs.99 Conversely, insecure attachment emerges
when parents are unresponsive. Individuals who develop secure attachments
in their own childhood (as judged by the Adult Attachment Inventory) tend
as adults to provide responsive parenting and have securely attached off-
spring (as measured by the Ainsworth Strange Situation). Conversely, par-
ents who had insecure attachments in their childhood are usually less re-
sponsive to offspring, who in turn become insecurely attached.100 It is
important to recall that the early childhood experience of incestuous fathers,
mothers, and siblings is marked by neglect, abandonment, and physical and
sexual abuse. These are the conditions that lead to highly insecure attach-
ment. Thus, the very conditions that contribute to insecure attachment in
childhood appear to be linked to a later propensity for unresponsive parent-
ing and incestuous abuse.

the psychology of incest avoidance

Westermarck did not propose a psychology of incest avoidance. Freud,

of course, did, and his Oedipal psychology was based on the assumption
that all social bonds are ultimately sexual. This assumption created the
dilemma of explaining how sexuality is inhibited among kin. Freud argued
that an aversive experience, childhood castration fear, was essential. De-
cades later, with the discovery of a biological basis for incest avoidance, a
renewed interest in the psychology of incest avoidance emerged. Initially, au-
thors continued within the Freudian paradigm.
Anthropologist Robin Fox, for example, proposed that childhood sexual
play between siblings is frustrating because of an inability of children to
reach orgasm.101 He argued that this frustrating sexual play would act as a
negative reinforcement to later incest among siblings. Fox’s hypothesis has
several limitations, including a lack of evidence that childhood sexual play
is frustrating and evidence that children can indeed achieve orgasm.102 In a
similar vein, anthropologist William Demarest proposed that punishment in
the childhood home acts to aversively condition incestuous behavior.103 Al-
though severe early punishment can inhibit later sexual behavior in ani-
mals, the effect is general, not specific, as would be needed for adaptive in-
cest avoidance.104 Beyond these objections are clinical findings showing that
incest is most prevalent in families where aversive experiences are particu-
larly common—the opposite of what this paradigm would seem to predict.
Psychodynamically oriented authors in disagreement with the Oedipal
 Evolutionary Thought and Clinical Understanding 175

model have argued, or inferred, that under normal developmental condi-

tions a nonsexual form of affiliation exists between kin.105 These authors,
however, lacked a coherent explanation of how nonsexual affiliation could
emerge, and formal hypotheses were not put forth.
A crucial shift in conceptualizing the psychology of incest avoidance be-
gins with the recognition that a distinct “familial” affiliation can evolve be-
cause selectional forces shaping kin-directed behaviors (altruism, attach-
ment, incest avoidance) are quite different from those shaping sexual or
pair bonding.106 Distinguishing familial and sexual affiliation resolves the
difficulty of explaining incest avoidance as a product of aversive experience
and psychological repression. Rather, incest avoidance begins, psychologi-
cally, with experiences that establish familial affiliation. Attachment stud-
ies suggest this experience is a responsive childhood milieu, which, in turn,
elicits secure attachment.
The kibbutz peer findings are illustrative. Peers develop early mutual
attachments that appear to be quite secure.107 These attachments remain
stable from childhood into adulthood, as evidenced by the intimate, re-
sponsive, and nonsexual relationships that endure between peers.108 This
“familial bond” between kibbutz peers develops at an unconscious level, as
it does in other species. When kibbutz peers are asked why they are not
mutually sexually attracted, they commonly reply that they experience each
other as brother and sister. Words such as brother, sister, filial, paternal,
and maternal connote feelings usually bestowed on family members. In us-
ing such terms, peers reveal how language points to the existence of a fa-
milial type of social affiliation that is biologically and psychologically dis-
tinct from sexual affiliation.
Conversely, early abuse and neglect undermine the propensity for adap-
tive familial affiliation at an equally unconscious level. Allen Sroufe and
Mary Ward videotaped the interactions of 173 mothers with their toddlers.
Most mothers interacted with warmth, empathy, and affection, but about 9
percent of the mothers were observed to be “seductive” or sexually inap-
propriate with their child, even while being videotaped. A mother might, for
example, ask her child for a kiss and then, if refused, physically force a
“passionate” kiss on the child’s lips. Seductive mothers were also more
likely to physically abuse or strike their child. A subset of all mothers were
interviewed about their own childhood. Mothers whose childhood history
was suggestive of incest were more likely to behave in a seductive and phys-
ically abusive way with their children. The authors note that while “affec-
tionate” mothers were responsive to offspring, “seductive” mothers were
noticeably less responsive and often interrupted the flow of the child’s be-
havior.109 Clare Haynes-Seman and Richard Krugman report similar obser-
vations. Parents who had reported a childhood history of neglect or abuse
176 erickson

tended to exhibit sexualized interactions with their infants even while aware
the session was being videotaped. These studies suggest that confusion
about appropriate parental care and sexual boundaries is, to an extraordi-
nary degree, unconscious.110 The basis of this boundary confusion appears
to lie in insecure childhood attachment.
To give one last example, Herman describes a group of fathers who,
while not overtly incestuous, sexualized the relationship with their daugh-
ters.111 These fathers might present their daughters with “sexy underwear,”
frequently talk about sex with their daughters and so forth. Emotionally,
such families tended to be tense and cold. Physical displays of affection
were uncommon and uncomfortable. The relationship between daughter
and father existed not in secure affection but in a milieu of distrust. The at-
tachment hypothesis of incest avoidance would predict that seductive fa-
thers did not develop an adequately secure attachment during their own
childhood. They experience unconscious boundary confusion between sex-
ual and familial affiliation but less so than overtly incestuous fathers. This
prediction could be evaluated with instruments such as the extensively used
Adult Attachment Inventory.112
Conceptualized in this way, the extraordinary psychological virulence of
incest becomes more understandable. Familial and sexual affiliation evolved
into biologically distinct entities because each is adaptive within a different
social context. Disrupting normal development, through incest, physical
abuse, or neglect, undermines adaptive social functioning at its most basic
level. Later in life, secure, intimate relationships, whether with one’s chil-
dren or a spouse, become extraordinarily difficult to achieve. It is small
wonder that borderline personality disorder, major depression, anxiety dis-
orders, drug and alcohol dependence, and chronic suicidality are frequently
observed sequelae.
A discussion of the psychology of incest avoidance would not be com-
plete without mentioning “infantile sexuality.” Freud placed great empha-
sis on this subject. Yet even within psychoanalytic circles the scientific va-
lidity of Freud’s notions has been openly questioned.113 Early sexual play is
observed in other primate species. Infant and juvenile chimpanzees, for ex-
ample, will mount their mother.114 The significance of this behavior is un-
known—it might facilitate the development of adult sexual behavior or
have a soothing function.115 There is remarkably little controlled research
on early sexual play in humans. Although most childhood sexual play is
probably within a developmental norm for our species, attachment studies
suggest that the quality of early attachment directly influences early sexual-
ity. Sroufe et al., for example, found that insecure attachment predisposed
to precocious, sexualized behavior and boundary violations in preadoles-
cence. Secure attachment was not associated with such behavior.116
 Evolutionary Thought and Clinical Understanding 177

One final point must be made concerning the boundaries between famil-
ial and sexual affiliation. Natural selection opposes not only close inbreeding
but also excessive outbreeding.117 In light of this, one might expect to find in-
dividuals preferring mates who are neither too similar nor too different from
themselves. Consistent with this, Bateson has shown that Japanese quail are
most sexually attracted to birds whose feather coloration is somewhat, but
not dramatically, different from that of immediate kin.118 Quail “sexually
imprint” on the coloration of immediate kin and use this, comparatively, for
mate selection. Claus Wedekind et al. provide evidence suggestive of a simi-
lar process in humans. In this case, body odor, as assessed by differences in
the major histocompatibility complex (MHC), influenced sexual preference.
Females found the odor of a male more attractive if his MHC, and hence
body odor, was relatively different from her own.119 Recent findings from
cross-cultural research and endocrinology suggest the propensity for sexual
affiliation develops much later than that for familial bonding.120
To summarize, human incest avoidance appears to depend on at least
two factors: (1) early association between kin, and (2) adequately secure
childhood attachment. These two factors establish the propensity for adap-
tive familial affiliation. The powerful influence of association during a sen-
sitive period is documented by Wolf’s simpua marriage research in Taiwan,
the Israeli kibbutz peer studies, and clinical findings on incest following
early separation.121 It is further supported by animal studies showing that
incest avoidance in other species depends critically on early association.122
The importance of adequately nurturant parenting is suggested by the
developmental histories of incest families. Clinical research has repeatedly
found that the parents of incest families were, themselves, neglected, abused,
or abandoned as children.123 This is not to say that an abusive, neglectful
childhood invariably leads to later incestuous behavior. Many, perhaps most,
individuals survive harsh childhoods and go on to be good parents. Rather,
it appears that early abuse and neglect are potent factors for disrupting early
attachment and later propensities for parental care and incest avoidance.

do modern cultural practices

make incest more common?

Many clinicians question whether we are in the midst of an epidemic

of incest. If even marginally true, this is cause for concern given the perva-
sively harmful effects of incest. Prevalence studies provide no answer. All re-
search is too recent to offer historical measure. There is, however, a different
approach to this concern. As mentioned, a unique and particularly troubling
characteristic of human incest is that it is frequently perpetrated on children.
This behavior is not observed in other primate species.124 Incestuous abuse of
children, and child sexual abuse in general, may be a product of modern cul-
178 erickson

tural practices. These cultural practices, novel to our evolutionary past, may
severely disrupt early attachment and, consequently, the development of un-
conscious, adaptive boundaries between familial and sexual affiliation.
The human genome has changed little since our species emerged 100,000
to 150,000 years ago. We remain nearly identical to our Paleolithic ances-
tors yet live in a dramatically different world. This mismatch between our
evolved biology and the modern world has, almost certainly, altered pat-
terns of human pathology. The current diet of most Westerners, for exam-
ple, includes a much higher intake of salt than in hunter-gatherer cultures.
Cultures whose diet is free of added salt have low blood pressure, by Amer-
ican standards, and do not show an increase in blood pressure with age as
is common in industrialized society. Hypertension is rare.125 Analogously,
cultural practices in Western industrialized societies may interfere with
early attachment to such an extent that behaviors such as child-directed in-
cest manifest.
Until the twentieth century virtually all babies were delivered in the
home. By 1970, nearly all births in the United States occurred in hospitals.
Many clinicians and mothers have questioned whether modern birthing
practices interfere with the mother-infant relationship at its earliest stages.
Studies now show that simply providing an undivided source of emotional
support for a mother during labor and delivery can dramatically reduce the
Cesarean section rate. Women who receive emotional support tend to
spend more time stroking, smiling at, and talking to their babies immedi-
ately following delivery than women who receive routine hospital care.
Weeks later, supported mothers are observed to spend more time with and
have more positive feelings for their infant.126 In a randomized study, Susan
O’Connor et al. found that mothers who “roomed in” with their infant, in
the hospital during the first two days postpartum were at follow-up, twelve
to twenty-one months later, significantly less likely to have abused or ne-
glected their infant than mothers who received routine postpartum care.127
Although the studies of O’Connor and others have been controversial, re-
cent research reveals similar findings in other primate species.128 A review
of the literature by Dario Maestripieri suggests the existence of a postpar-
tum sensitive period of heightened “maternal motivation,” which, in effect,
canalizes a maternal bond to offspring.129 Separation from offspring during
this period increased the probability of maternal rejection among other
primates. This effect may be mediated, in humans and other primates, by
hormones such as oxytocin and prolactin.130
Mammalian breast-feeding has been molded by more than 65 million
years of evolution. In addition to its nutritional function, breast-feeding
may also affect emotional development. Primate cross-species comparisons
suggest that human children have evolved to receive benefits from breast-
 Evolutionary Thought and Clinical Understanding 179

feeding for an absolute minimum of two and a half years to an upper limit
of seven years.131 In the United States, many women do not nurse at all,
and physicians often consider six months to be extended breast-feeding.
Modern breast-feeding practices undoubtedly reduce physical if not emo-
tional intimacy between mother and child.
In Western industrialized countries, infant crying tends to be prolonged.
The duration of crying bouts is shorter in cultures such as the !Kung San of
Botswana, where infants are carried virtually continuously in a sling. !Kung
San mothers, by experimental measures, have been rated as much more re-
sponsive than Western mothers. Infants are nursed on demand rather than
by “schedule.” This indulgent form of caregiving by the !Kung San is prob-
ably typical for our species and other higher primates. Frequent short cries
elicit a positive, communicative response. Repeated many times a day, these
may engender in the infant a secure, confident attachment with mother.132
In our evolutionary past, parents slept in immediate proximity to infants
and young offspring. This practice not only provided protection but also
appears to promote normal physiological development in infants and may
reduce pathology such as sudden infant death syndrome.133 In the vast ma-
jority of non-Western cultures, various forms of parent-child cosleeping are
the predominant arrangement throughout the first few years of life. The
practice of placing infants and young children into separate bedrooms is an
entirely anomalous cultural intrusion. This practice obviously decreases the
amount of close association between parent and child. Apropos of this,
John Forbes and David Weiss found that, “contrary to expectations,” chil-
dren who coslept with parents were less likely to have been treated in a
mental health clinic for emotional and behavioral problems.134
Diminished involvement of the father, or stepfather, in the caretaking of
children is associated with increased risk of paternal incest. Hilda Parker
and Seymour Parker found that sexually abusive fathers were less likely to
have been involved in the early care of their daughters.135 Linda Williams
and David Finkelhor similarly found that low involvement by fathers in
caretaking of offspring was a risk factor for incest. They note that the act
of taking care of a child may evoke feelings of nurturance that are incom-
patible with incestuous abuse.136
Early association is crucial for the development of adaptive familial
bonds in many species. Humans are no exception. The examples just men-
tioned represent only a few of the ways in which the quality and quantity
of early kin association may be disrupted by cultural practices that have no
precedent in our evolutionary past. The combined effects of such practices
may manifest as severe pathologies of kinship. Incest perpetrated on chil-
dren may represent the most extreme example.
During most of the twentieth century, social scientists believed incest
180 erickson

was common in nature. Among humans, incest was thought to be rare be-
cause of cultural taboos. As the twenty-first century begins, this view has
been, figuratively, turned on its head. Incest is now known to be rare in na-
ture, and we must ask if human incest has become more, not less common
because of cultural influences. If pathologies of kinship exist, antidotes will
be found. As with medical science in general, the speed of discovery will
depend on how findings from nonclinical disciplines are used to inform
clinical research. Efficient progress requires open communication between
the varied disciplines of anthropology, psychiatry, psychology, ethology, be-
havioral endocrinology, and all other fields that contribute to our under-
standing of human kinship and familial relations.

notes
1. Erna Olafson, David L. Corwin, and Roland C. Summit, “Modern history
of child sexual abuse awareness: Cycles of discovery and suppression,” Child
Abuse and Neglect, vol. 17 (1993), pp. 7–24.
2. Arthur J. Barsky, Carol Wood, Maria C. Barnett, and Paul D. Clay, “Histo-
ries of childhood trauma in adult hypochondriacal patients,” American Journal of
Psychiatry, vol. 151 (1994), pp. 397–401; James A. Chu and Diana L. Dill, “Dis-
sociative symptoms in relation to childhood physical and sexual abuse,” American
Journal of Psychiatry, vol. 147 (1990), pp. 887–92; Kenneth S. Kendler, Cynthia
M. Bulik, Judy Silberg, John M. Hettema, and John Myers, “Childhood sexual
abuse and adult psychiatric and substance use disorders in women,” Archives of
General Psychiatry, vol. 57 (2000), pp. 953–59; Pamela S. Ludolph, Drew Westen,
Barbara Misle, Anne Jackson, Jean Wixom, and F. Charles Wiss, “Borderline diag-
nosis in adolescents: Symptoms and developmental history,” American Journal
of Psychiatry, vol. 147 (1990), pp. 470–76; James Morrison, “Childhood sexual
histories of women with somatization disorder,” American Journal of Psychiatry,
vol. 146 (1989), pp. 239–41; Susan M. Ogata, Kenneth R. Silk, Sonya Goodrick,
Naomi E. Lohr, Drew Westen, and Elizabeth M. Hill, “Childhood sexual and
physical abuse in adult patients with borderline personality disorders,” American
Journal of Psychiatry, vol. 147 (1990), pp. 1008–13; Elizabeth F. Pribor and
Stephen H. Dinwiddie, “Psychiatric correlates of incest in childhood,” American
Journal of Psychiatry, vol. 149 (1992), pp. 52–56; Elizabeth F. Pribor, Sean H.
Yutzy, J. Todd Dean, and Richard D. Wetzel, “Briquet’s syndrome, dissociation,
and abuse,” American Journal of Psychiatry, vol. 150 (1993), pp. 1507–11;
Donna H. Schetky, “A review of the literature on the long-term effects of child-
hood sexual abuse,” in Incest-Related Syndromes of Adult Psychopathology,
ed. Richard P. Kluft (Washington, D.C.: American Psychiatric Press, 1990),
pp. 35–54; Steven L. Shearer, Charles P. Peters, Miles S. Quaytman, and Richard
L. Ogden, “Frequency and correlates of childhood sexual and physical abuse
histories in adult female borderline patients,” American Journal of Psychiatry,
 Evolutionary Thought and Clinical Understanding 181

vol. 147 (1990), pp. 214–16; Michael H. Stone, “Incest in the borderline patient,”
in Incest Related Syndromes of Adult Psychopathology, ed. Richard P. Kluft
(Washington, D.C.: American Psychiatric Press, 1990), pp. 183–204; Bessel A.
van der Kolk, Christopher Perry, and Judith L. Herman, “Childhood origins
of self-destructive behavior,” American Journal of Psychiatry, vol. 148 (1991),
pp. 1665–71; Edward D. Walker, Wayne J. Katon, Kathleen Neraas, Ron P.
Jemelka, and Donna Massoth, “Dissociation in women with chronic pelvic
pain,” American Journal of Psychiatry, vol. 149 (1992), pp. 534–37; Edward D.
Walker, Wayne J. Katon, Peter P. Roy-Byrne, Ron J. Jemetka, Jean Russo, et al.,
“Histories of sexual victimization in patients with irritable bowel syndrome,”
American Journal of Psychiatry, vol. 150 (1993), pp. 1502–6.
3. See Jon Allen, Coping with Trauma: A Guide to Self Understanding (Wash-
ington, D.C.: American Psychiatric Association Press, 1995); Jon Allen, Traumatic
Relationships and Serious Mental Disorders (New York: Wiley, 2001); Judith L.
Herman, Trauma and Recovery (New York: Basic Books, 1992).
4. See Anne E. Pusey and Marisa Wolf, “Inbreeding avoidance in animals,”
Trends in Evolutionary Ecology, vol. 11 (1996), pp. 201–6; and also Chapter 3
of this volume.
5. Joseph Shepher, “Mate selection among second-generation kibbutz adoles-
cents and adults: Incest avoidance and negative imprinting,” Archives of Sexual
Behavior, vol. 1 (1971), pp. 293–307; Arthur P. Wolf, “Childhood association,
sexual attraction, and the incest taboo: A Chinese case,” American Anthropologist,
vol. 68 (1966), pp. 883–98; Arthur P. Wolf, “Childhood association and sexual at-
traction: A further test of the Westermarck hypothesis,” American Anthropologist,
vol. 72 (1970), pp. 503–15; Arthur P. Wolf, Sexual Attraction and Childhood
Association: A Chinese Brief for Edward Westermarck (Stanford, Calif.: Stanford
University Press, 1995). See also Chapter 4 of this volume.
6. See Chapter 2 of this volume and Pusey and Wolf, “Inbreeding avoidance in
animals,” p. 201.
7. See Leah Gavish, Joyce E. Hofmann, and Lowell L. Getz, “Sibling recogni-
tion in the prairie vole, Microtus ochrogaster,” Animal Behavior, vol. 23 (1984),
pp. 362–66; Maurice Greenberg and Roland Littlewood, “Post-adoption incest
and phenotypic matching: Experience, personal meanings, and biosocial implica-
tions,” British Journal of Medical Psychology, vol. 68 (1995), pp. 29–44.
8. Sigmund Freud, “Totem and Taboo,” in Complete Psychological Works,
standard edition, vol. 13 (London: Hogarth Press, 1953); Claude Lévi-Strauss,
“The family,” in Man, Culture, and Society, ed. H. L. Shapiro (London: Oxford
University Press, 1956); 278; Leslie A. White, “The definition and prohibition of
incest,” American Anthropologist, vol. 50 (1948), pp. 416–35.
9. Edward Westermarck, The History of Human Marriage, vol. 2 (New York:
Allerton Press, 1922).
10. Donald S. Sade, “Inhibition of son-mother mating among free ranging
rhesus monkeys,” Science and Psychoanalysis, vol. 12 (1968), pp. 18–38.
11. See Chapter 3 of this volume and Pusey and Wolf, “Inbreeding avoidance
in animals,” pp. 202–5.
12. See Hudson K. Reeve, David F. Westneat, William A. Noon, Paul Sherman,
182 erickson

and C. F. Aquado, “DNA ‘fingerprinting’ reveals high levels of inbreeding colonies

of eusocial naked mole-rats,” Proceedings of the National Academy of Science,
USA, vol. 87 (1990), pp. 2496–500.
13. Greta Agren, “Incest avoidance and bonding between siblings in gerbils,”
Behavioral Ecology and Sociobiology, vol. 14 (1984), pp. 161–69; Gavish, Hof-
mann, and Getz, “Sibling recognition in the prairie vole”; Dustin Penn and Wil-
liam Potts, “MHC-disassortative mating preferences reversed by cross-fostering,”
Proceedings of the Royal Society of London B, vol. 265 (1998), pp. 1299–306;
Kunio Yamazaki, Gary B. Beauchamp, D. Kupriewski, Judith Bard, L. Thomas,
and Edward A. Boyce, “Familial imprinting determines H-2 selective mating
preferences,” Science, vol. 240 (1988), pp. 1331–32.
14. Gavish, Hofmann, and Getz, “Sibling recognition.”
15. See Chapter 7 of this volume. Contrast with Wolf, Sexual Attraction,
pp. 497–515.
16. Bruno Bettelheim, Children of the Dream (New York: Macmillan, 1969);
Albert I. Rabin, Growing Up on a Kibbutz (New York: Springer, 1965); Shepher,
“Mate selection”; Melford E. Spiro, Children of the Kibbutz (New York: Schocken
Books, 1965); Yonina Talmon, “Mate selection in collective settlements,” Ameri-
can Sociological Review, vol. 29 (1964) pp. 491–508. See Chapter 4 of this vol-
ume and Wolf, Sexual Attraction.
17. Shepher, “Mate selection.”
18. See Chapter 4 of this volume and Wolf, Sexual Attraction.
19. Wolf, Sexual Attraction, pp. 264–375.
20. Ibid., pp. 204–13.
21. Edward O. Wilson, Consilience: The Unity of Knowledge (New York:
Alfred E. Knopf, 1998), pp. 173–76.
22. Alfred C. Kinsey, Wardell B. Pomeroy, Clyde E. Martin, and Paul H.
Gebhard, Sexual Behavior in the Human Female (Philadelphia: W. B. Saunders,
1953).
23. S. Kirson Weinberg, Incest Behavior (New York: Citadel, 1955), pp. 38–39.
24. A. M. Freedman, Howard I. Kaplan, and Benjamin J. Sadock, Comprehen-
sive Textbook of Psychiatry, 2nd ed. (Baltimore: Williams and Wilkins, 1975), as
cited in Herman, Father-Daughter Incest.
25. Diana E. H. Russell, “Intrafamilial and extrafamilial sexual abuse of
female children,” Child Abuse and Neglect, vol. 7 (1983), pp. 133–46.
26. Ibid.
27. Daniel S. Halperin, Paul Bouvier, Philip D. Jaffe, Roger-Luc Mounoud,
Claus H. Pawlak, Jerome Laederach, Helene R. Wickey, and Florence Astie,
“Prevalence of child sexual abuse among adolescents in Geneva: Results of a cross
sectional survey,” British Medical Journal, vol. 312 (1996), p. 1326.
28. See Russell, “Intrafamilial and extrafamilial sexual abuse,” p. 135.
29. See David Finkelhor, Gerald Hotaling, I. A. Lewis, and Christine Smith,
“Sexual abuse in a national survey of adult men and women: Prevalence, charac-
teristics, and risk factors,” Child Abuse and Neglect, vol. 14 (1990), p. 20.
30. In Finkelhor’s 1980 study, 13 percent of the entire sample of 796 under-
graduate students reported some type of sexual experience with a sibling (15 per-
 Evolutionary Thought and Clinical Understanding 183

cent of females and 10 percent of males). Most of these experiences were regarded
as childhood sex play. About 25 percent of this group, or a prevalence of about
3.2 percent, had experiences regarded as exploitative because force was used or
because there was a large age difference between siblings. In this study and those
by Baker and Ducan and Finkelhor, Hotaling, Lewis, and Smith the precise degree
of relatedness was not specified. These data may include incest involving kin out-
side the immediate family or involving step-parents or step-siblings. See David
Finkelhor, “Sex among siblings: A survey of prevalence, variety, and effects,”
Archives of Sexual Behavior, vol. 9 (1980), pp. 171–94; Anthony W. Baker and
Sylvia P. Duncan, “Child sexual abuse: A study of prevalences in Great Britain,”
Child Abuse and Neglect, 9 (1985), 457–67; and Finkelhor, Hotaling, Lewis,
and Smith, “Sexual abuse in a national survey.”
31. The data cited from Halperin et al.’s 1996 Geneva study are unpublished
findings. I am grateful to Paul Bouvier and Daniel Halperin of the Service De
Santé De La Jeunesse of Geneva, Switzerland, for their generosity in providing
these data.
32. Russell, “Intrafamilial and extrafamilial sexual abuse,” p. 141.
33. Baker and Duncan, “Child Sexual Abuse,” p. 461.
34. Loretta M. McCarty, “Mother-child incest: Characteristics of the offender,”
Child Welfare, vol. 65 (1986), pp. 447–58.
35. For a review see L. Roybal and Jean Goodwin, “The incest pregnancy,”
in Sexual Abuse: Incest Families and Their Victims, ed. Jean Goodwin (Chicago:
New Year Medical Publisher, 1989.
36. David Finkelhor, “The international epidemiology of child sexual abuse,”
Child Abuse and Neglect, vol. 18 (1949), pp. 409–17.
37. See Linda M. Williams and David Finkelhor, “Parental care-giving and
incest: Test of a biosocial model,” American Journal of Orthopsychiatry, vol. 65
(1995), pp. 101–13; and also Hilda Parker and Seymour Parker, “Father-daughter
sexual abuse: An emerging perspective,” American Journal of Orthopsychiatry,
vol.56 (1986), pp. 532–49.
38. Herman, Father-Daughter Incest, p. 71.
39. Hector Cavallin, “Incestuous fathers: A clinical report,” American Journ
al of Psychiatry, vol. 122 (1966), pp. 1132–38; Herman, Father-Daughter Incest,
pp. 67–125; Marisa Laviola, “Effects of older brother–younger sister incest:
A study of dynamics of 17 cases,” Child Abuse and Neglect, vol. 16 (1992),
pp. 409–21; Noel Lustig, John W. Dresser, Seth W. Spellman, and Thomas B.
Murray, “Incest: A family group survival pattern,” Archives of General Psychia-
try, vol. 14 (1966), pp. 31–40; Philip G. Madonna, Susan Van Scoyk, and David P.
Jones, “Family interactions within incest and non-incest families,” American Jour-
nal of Psychiatry, vol. 148 (1991), pp. 46–49; McCarty, “Mother-child incest”;
Marcellina Mian, Peter Marton, Deborah LeBaron, and David Birdwistle, “Famil-
ial risk factors associated with intrafamilial and extrafamilial sexual abuse of three
to five year old girls,” Canadian Journal of Psychiatry, vol. 39 (1994), pp. 348–53;
Holly Smith and Edie Israel, “Sibling incest: A study of the dynamics of 25 cases,”
Child Abuse and Neglect, vol. 11 (1987), pp. 101–8.
40. Madonna, Van Scoyk, and Jones, “Family interactions,” pp. 47–48.
184 erickson

41. Kathleen C. Faller, “Women who sexually abuse children,” Violence

and Victims, vol. 2 (1987), pp. 263–75; Herman, Father-Daughter Incest, p. 76;
McCarty, “Mother-child incest,” p. 452; Mian, Marton, LeBaron, and Birdwistle,
“Familial risk factors,” p. 350; Heikki Sariola and Antti Uutela, “The prevalence
and context of sexual abuse in Finland,” Child Abuse and Neglect, vol. 20 (1996)
p. 846.
42. Cavallin, “Incestuous fathers,” p. 1134; Herman, Father-Daughter Incest,
p. 78; Laviola, “Effects of older brother–younger sister incest,” p. 414; Narcyz
Lukianowicz, “Incest I: Paternal incest,” British Journal of Psychiatry, vol. 120
(1972), pp. 301–13; Jane M. Rudd and Sharon D. Herzberger, “Brother-sister
incest and father-daughter incest: A comparison of characteristics and conse-
quences,” Child Abuse and Neglect, vol. 9 (1999), pp. 915–28.
43. Herman, Father-Daughter Incest, p. 86; Sariola and Uutela, “Sexual abuse
in Finland,” p. 848.
44. Patricia Phelan, “Incest and its meaning: The perspectives of fathers and
daughters,” Child Abuse and Neglect, vol. 19 (1995), pp. 7–24.
45. See Lukianowicz, “Incest I,” p. 303.
46. Herman, Father-Daughter Incest, pp. 80–81.
47. McCarty, “Mother-child incest”; Faller, “Women who sexually abuse
children.”
48. Lukianowicz, “Incest I,” pp. 303–7; Lustig Dresser, Spellman, and Murray,
“Incest,” p. 38; Mian, Marton, LeBaron, and Birdwistle, “Familial risk factors,”
p. 352.
49. Baker and Duncan, “Child Sexual Abuse,” p. 459; Cavallin, “Incestuous
Fathers,” p. 1132; Herman, Father-Daughter Incest, pp. 67–72.
50. Herman, Father-Daughter Incest, pp. 50–63.
51. Madonna, Van Scoyk, and Jones, “Family interactions,” pp. 47–48; David
W. Smith and Benjamin E. Saunders, “Personality characteristics of father/daugh-
ter perpetuators and non-offending mothers in incest families: Individual and
dyadic analysis,” Child Abuse and Neglect, vol. 19 (1995), pp. 607–17.
52. Mark T. Erickson, “Incest avoidance and familial bonding,” Journal of
Anthropological Research, vol. 45 (1989), pp. 267–91; Mark T Erickson, “Re-
thinking Oedipus: An evolutionary perspective on incest avoidance,” American
Journal of Psychiatry, vol. 150 (1993), pp. 411–16; Mark T. Erickson, “Incest
avoidance: Clinical implications of the evolutionary perspective,” in Evolutionary
Medicine, ed. Wenda R. Trevathan, E. O. Smith, and James J. McKenna (New
York: Oxford University Press, 1991), pp. 165–81; Arthur P. Wolf, “Westermarck
redivivus,” Annual Review of Anthropology, vol. 22 (1993), pp. 157–75; Wolf,
Sexual Attraction, pp. 463–75.
53. Cavallin, “Incestuous fathers,” p. 1134; Parker and Parker, “Father-
daughter sexual abuse”; Blair Justice and Rita Justice, Broken Taboo (New
York: Human Sciences Press, 1979), p. 63; Irving Kaufman, Alice L. Peck, and
Consuela Tagiuri, “The family constellation and overt incestuous relations
between father and daughter,” American Journal of Orthopsychiatry, vol. 24
(1954), pp. 266–77; Lustig, Dresser, Spellman, and Murray, “Incest,” p. 33;
David L. Raphling, Bob L. Carpenter, and Allen Davis, “Incest: A genealogical
 Evolutionary Thought and Clinical Understanding 185

study,” Archives of General Psychiatry, vol. 16 (1967), pp. 505–11; Mian,

Marton, LeBaron, and Birdwistle, “Familial risk factors,” p. 350; Patricia K.
Mrazek, “The nature of incest: A review of contributing factors,” in Sexually
Abused Children and Their Families, ed. Patricia K. Mrazek and Charles H.
Kempe (Elmsford, N. Y.: Pergamon Press, 1981); Williams and Finkelhor,
“Parental care-giving and incest,” p. 106.
54. Justice and Justice, Broken Taboo, p. 63; Mian, Marton, LeBaron, and
Birdwistle, “Familial risk factors,” p. 350; Lustig, Dresser, Spellman, and Murray,
“Incest,” p. 34; Mrazek, “The nature of incest.”
55. McCarty, “Mother-child incest,” pp. 448–50.
56. Laviola, “Effects of older brother–younger sister incest,” pp. 414–15;
Rudd and Herzberger, “Brother-sister incest and father-daughter incest,” p. 919;
Smith and Israel, “Sibling incest.”
57. Rudd and Herzberger, “Brother-sister incest and father-daughter incest,”
p. 919; Smith and Israel, “Sibling incest.”
58. Naomi A. Adler and Joseph Schultz, “Sibling incest offenders,” Child
Abuse and Neglect, vol. 19 (1995), pp. 811–19.
59. Smith and Israel, “Sibling incest,” p. 104.
60. Laviola, “Effects of older brother-younger sister incest,” p. 415.
61. Kaufman, Peck, and Tagiuri, “The family constellation,” p. 266;
Lukianowicz, “Incest I,” p. 304; Phelan, “Incest and its meaning,” p. 9; Sariola
and Uutela, “Sexual abuse in Finland,” p. 846; Gail Wyatt, “Sexual abuse of
Afro-American and white-American women in childhood,” Child Abuse and
Neglect, vol. 9 (1985), p. 514.
62. Lukianowicz, “Incest I,” p. 304; Herman, Father-Daughter Incest, p. 83;
Phelan, “Incest and its meaning,” p. 9.
63. Faller, “Women who sexually abuse children,” p. 271.
64. Adler and Schultz, “Sibling incest offenders,” p. 815.
65. Roy R. Frenzel and Reuben A. Lang, “Identifying sexual preferences in
intra-familial and extra-familial child sexual abusers,” Annals of Sex Research,
vol. 2 (1989), pp. 255–75.
66. Anne E. Pusey, personal communication, September 19, 2000.
67. Gavish, Hofmann, and Getz, “Sibling recognition in the prairie vole.”
68. Weinberg, Incest Behavior, p. 78.
69. Greenberg and Littlewood, “Post-adoption incest,” pp. 29–44 (for quotes
see p. 35).
70. Sarah Strickland, “A love that does not know its true name,” The Inde-
pendent on Sunday (London), January 3, 1993, p. 20.
71. Daniel Voll, “An American family,” Esquire, July, 1998, pp. 122–45.
72. See Time magazine, July 2, 1979, and the New York Times, August 2, 1979.
73. Barbara J. Lifton, Lost and Found: The Adoption Experience (New York:
Dial Press, 1979), p. 165.
74. Benita Eissler, Byron: Child of Passion, Fool of Fame (New York: Alfred
Knopf, 1999).
75. Deirdre Bair, Anaïs Nin: A Biography (New York: Putnam and Sons,
1995), pp. 15–17.
186 erickson

76. Ibid., p. 170.

77. Ibid., p. 173.
78. D. W. Cory and R. E. L. Masters, Violations of Taboo: Incest of the Great
Literature of Past and Present (New York: Julian Press, 1963); Otto Rank, The
Incest Theme in Literature and Legend, trans. Gregory C. Richter (Baltimore:
Johns Hopkins University Press, 1992).
79. To my knowledge, the only documentation of exceptions to marriage be-
tween immediate kin, associated from early life, come from ancient sources such
as the reports of brother-sister marriage in Roman Egypt, circa A.D. 20–258. A
degree of skepticism is warranted. Direct contact with these couples is, of course,
impossible. The importance of direct contact or observation has been made
abundantly apparent in studies of incest avoidance. It was thought that incest
was common in nature until careful observational studies revealed otherwise.
In Taiwan a number of couples who had married in the minor fashion told Wolf
that they had not been sexually active despite years of marriage. Their children
were conceived in extramarital affairs but registered in the family name. It is im-
possible to rule out an explanation of this sort for suggestive findings from antiq-
uity. See, for example, Keith Hopkins, “Brother-sister marriage in Roman Egypt,”
Comparative studies in Society and History, vol. 22 (1980), pp. 303–47, and
Chapter 5 in this volume. See also Wolf, Sexual Attraction, pp. 427–29.
80. Kathleen Harrison, The Kiss: A Memoir (New York: Random House,
1997), p. 56.
81. See Paul W. Sherman and Warren G. Holmes, “Kin-recognition: Issues
and evidence,” in Experimental Behavioral Ecology and Sociobiology, ed. Bert
Hölldobler and Martin Lindauer (Sunderland, Mass.: Sinauer, 1985).
82. See Richard H. Porter, “Mutual mother-infant-recognition in humans,”
in Kin Recognition, ed. P. G. Hopper (Cambridge: Cambridge University Press,
1991).
83. Marsha Kaitz, Shimon Shuri, Shai Danziger, Ziva Herskko, and Arthur I.
Eidelman, “Fathers can also recognize their newborns by touch,” Infant Behavior
and Development, vol. 17 (1994), pp. 205–7.
84. Marsha Kaitz, A. Good, A. M. Rokem, and A. I. Eidelman, “Mothers’
recognition of their newborn by olfactory cues,” Development Psychology, vol. 20
(1988), pp. 582–91.
85. Patrick Bateson, “Sexual imprinting and optimal outbreeding,” Nature,
vol. 273 (1978), 659–60.
86.Greenberg and Littlewood, “Post-adoption incest,” p. 41–43.
87. William D. Hamilton, “Genetical evolution of social behavior,” Journal
of Theoretical Biology, vol. 7 (1964), pp. 1–52; John Maynard Smith, “Group
selection and kin selection,” Nature, vol. 201 (1964), pp. 1145–47; George C.
Williams, Adaptation and Natural Selection: A Critique of Some Current Evolu-
tionary Thought (Princeton, N.J.: Princeton University Press, 1966).
88. George C. Williams and Randolph M. Nesse, “Dawn of Darwinian medi-
cine,” Quarterly Review of Biology, vol. 66 (1991), pp. 1–22.
89. Paul W. Sherman, “Nepotism and the evolution of alarm calls,” Science,
vol. 197 (1977), pp. 1246–53; Evelyn B. Hangii and Ronald J. Schusterman, “Kin
 Evolutionary Thought and Clinical Understanding 187

recognition in captive California sea lions,” Journal of Comparative Psychology,

vol. 104 (1990), pp. 368–72; D. R. Meikle and S. H. Vessey, “Nepotism among
rhesus monkey brothers,” Nature, vol. 294 (1984), pp. 160–61.
90. Robert Axelrod and William D. Hamilton, “The evolution of coopera-
tion,” Science, vol. 211 (1981), pp. 1390–96.
91. John Bowlby, Attachment and Loss, vol. I: Attachment (New York: Basic
Books, 1969) p. 305.
92. Patrick Bateson, “How do sensitive periods arise and what are they for?”
Animal Behavior, vol. 27 (1979), pp. 470–86.
93. Sherman and Holmes, “Kin-recognition.”
94. Wolf, “Westermarck redivivus,” p. 168; Wolf, Sexual Attraction, pp. 463–
75; Erickson, “Incest avoidance and familial bonding,” pp. 278–80; Erickson,
“Rethinking Oedipus,” p. 413; Erickson, “Incest Avoidance.”
95. Wolf, Sexual Attraction, p. 471.
96. Bill Amos, Christian Schlotterer, and Deihard Tautz, “Social structure of
pilot whales revealed by DNA profiling,” Science, vol. 260 (1993), pp. 670–72.
97. Frans de Waal and Frans Lanting, Bonobo: The Forgotten Ape (Berkeley:
University of California Press, 1997), pp. 115–17.
98. Francois Jacob, “Evolution and tinkering,” Science, vol. 196 (1977),
pp. 1161–66.
99. Mary D. S. Ainsworth, Infancy in Uganda: Infant Care in the Growth of
Love (Baltimore: Johns Hopkins University Press, 1967); Russell A. Isabella,
“Origins of attachment: Maternal interactive behavior across the first year,”
Child Development, vol. 64 (1993), pp. 605–21; Russell A. Isabella and Jay
Belsky, “Interactional synchrony and the origins of infant-mother attachment,”
Child Development, vol. 62 (1991), pp. 373–84; Philip B. Smith and David R.
Pederson, “Maternal sensitivity and patterns of infant-mother attachment,”
Child Development, vol. 59 (1988), pp. 1097–101.
100. Howard Steele, Miriam Steele, and Peter Fonagy, “Association among
attachment classifications of mothers, fathers, and their infants,” Child Develop-
ment, vol. 67 (1996), pp. 541–55. See also Marius H. van Ijzendoorn, “Adult
attachment representations, parental responsiveness, and infant attachment:
A meta-analysis on the predictive validity of the adult attachment interview,”
Psychological Bulletin, vol. 117 (1995), pp. 387–403.
101. Robin Fox, “Sibling incest,” British Journal of Sociology, vol. 13 (1962),
pp. 128–50.
102. Meyer Fortes, Web of Kinship Among the Tallensi (London: Oxford
University Press, 1957), p. 251; Kinsey, Pomeroy, Martin, and Gebhard, Sexual
Behavior in the Human Female, p. 104.
103. William J. Demarest, “Incest avoidance among human and nonhuman
primates,” in Primate Social Behavior, ed. Susan Chevalier-Skolnikoff and
E. F. Poirer (New York: Garland Press, 1977), pp. 323–42.
104. Frank A. Beach, “Instinctive behavior: Reproductive activities,” in
Handbook of Experimental Psychology, ed. S. S. Stevens (New York: Wiley,
1951), p. 408.
105. Howard A. Bacal and Kenneth M. Newman, Theories of Object Rela-
188 erickson

tions to Self Psychology (New York: Columbia University Press, 1990), p. 12;
George A. DeVos, “Affective dissonance and primary socialization: Implications
for a theory of incest avoidance,” Ethos, vol. 3 (1975), pp. 165–82; Takeo Doi,
“The concept of amae and its psychoanalytic implications,” International Review
of Psychoanalysis, vol. 16 (1989), pp. 349–54; Sandor Ferenczi, “Confusion of
tongues between adult and the child,” International Journal of Psychoanalysis,
vol. 30 (1949), pp. 225–30; Heinz Kohut, How Does Analysis Cure? ed. A. Gold-
berg and P. Stepansky (Chicago: University of Chicago Press, 1984), p. 3–12; Ian
D. Suttie, Origins of Love and Hate (London: Kegan Paul, 1935), pp. 80–96.
106. Erickson, “Incest avoidance and familial bonding,” pp. 278–80; Erick-
son, “Rethinking Oedipus,” p. 413; Wolf, “Westermarck redivivus,” p. 168; Wolf,
Sexual Attraction, pp. 463–75.
107. See Rabin, Growing Up on a Kibbutz, p. 23.
108. See Shepher, “Mate selection,” p. 296.
109. Allen L. Sroufe and Mary J. Ward, “Seductive behaviors of mothers of
toddlers: Occurrence, correlates, and family origins,” Child Development, vol. 51
(1980), pp. 1222–29.
110. Clare Haynes-Seman and Richard Krugman, “Sexualized attention:
Normal interaction or precursor to sexual abuse?” American Journal of Ortho-
psychiatry, vol. 59 (1989), pp. 238–45.
111. Herman, Father-Daughter Incest, pp. 109–25.
112. Mary Main and Ruth Goldwyn, Adult Attachment Classification System:
Version 5. Unpublished manuscript, University of California, Berkeley, 1991.
113. Paul Chodoff, “A critique of Freud’s theory of infantile sexuality,” Ameri-
can Journal of Psychiatry, vol. 123 (1966), pp. 507–18.
114. See Chapter 3 of this volume.
115. de Waal and Lanting, Bonobo, p. 117.
116. Allen L. Sroufe, Christopher Bennett, Michelle Englund, and Shmuel
Shulman, “The significance of gender boundaries in preadolescence: Contempo-
rary correlates and antecedents of boundary violation and maintenance,” Child
Development, vol. 64 (1993), pp. 455–676.
117. Bateson, “Optimal outbreeding.”
118. Patrick Bateson, “Preferences for cousins in Japanese quail,” Nature,
vol. 295 (1982), pp. 236–37.
119. Claus Wedekind, Thomas Seebeck, Florence Bettens, and Alexander J.
Paepke, “MHC dependent mate preferences in humans,” Proceedings of the Royal
Society of London B, vol. 260 (1995), pp. 245–49.
120. Gilbert Herdt and Martha McClintock, “The magical age of 10,”
Archives of Sexual Behavior, vol. 29 (2000), pp. 587–606.
121. See Chapter 4 of this volume; Shepher, “Mate selection”; and Greenberg
and Littlewood, “Post-adoption incest,” pp. 29–44.
122. See Gavish, Hofmann, and Getz, “Sibling recognition in the prairie vole.”
123. See Adler and Schultz, “Sibling incest offenders”; McCarty, “Mother-
child incest”; Williams and Finkelhor, “Parental care-giving and incest.”
124. Anne E. Pusey, personal communication, September 19, 2000.
125. S. Boyd Eaton, S. Boyd Eaton III, and Melvin J. Konner, “Paleolithic nu-
 Evolutionary Thought and Clinical Understanding 189

trition revisited,” in Evolutionary Medicine, ed. Wenda R. Trevathan, E. O. Smith,

and James J. McKenna (New York: Oxford University Press, 1999), pp. 313–32.
126. Wenda R. Trevathan, “Evolutionary obstetrics,” in Evolutionary Medi-
cine, ed. Wenda R. Trevathan, E. O. Smith, and James J. McKenna (New York:
Oxford University Press, 1999), pp. 183–208.
127. Susan M. O’Connor, Peter M. Vietze, John B. Hopkins, and William A.
Altemeier, “Post-partum extended maternal-infant contact: Subsequent mothering
and child health,” Pediatric Research, vol. 11 (1977).
128. Michael Lamb, “Early mother-neonate contact and the mother-child
relationship,” Journal of Child Psychology and Psychiatry, vol. 24 (1983),
pp. 487–94.
129. Dario Maestripieri, “Is there mother-infant bonding in primates?”
Development Review, vol. 21 (2001), pp. 93–201.
130. Ibid., p. 113; Eva Nissen, Gunilla Lilia, Ann-Marie Widstrom, and
Kerstin Uvnas-Moberg, “Elevation of oxytocin levels early in post-partum
women,” Acta Obstetricia et Gynecologia Scandanavica, vol. 74 (1995),
pp. 530–33.
131. Katherine A. Dettwyler, “A time to wean,” Breastfeeding Abstracts,
vol. 14 (1994), pp. 3–4.
132. Ronald G. Barr, “Infant crying behavior and colic: An interpretation
in evolutionary perspective,” in Evolutionary Medicine, ed. Wenda R. Trevathan,
E. O. Smith, and James J. McKenna (New York: Oxford University Press, 1999),
pp. 27–52.
133. James J. McKenna, Sarah Mosko, and Chris Richard, “Breastfeeding and
mother-infant co-sleeping in relation to SIDS prevention,” in Evolutionary Medi-
cine, ed. Wenda R. Trevathan, E. O. Smith, and James J. McKenna (New York:
Oxford University Press, 1999), pp. 53–74; James J. McKenna, Evelyn B. Thoman,
Thomas F. Anders, Abraham Sadeh, Vicki L. Schechtman, and Steven F. Glotzbach,
“Infant-parent co-sleeping in an evolutionary perspective: Implications for under-
standing infant sleep development and sudden infant death syndrome,” Sleep,
vol. 16 (1993), pp. 263–82.
134. John Forbes and David S. Weiss, Cosleeping habits of military children,”
Military Medicine, vol. 157 (1992), pp. 196–200.
135. Parker and Parker, “Father-daughter sexual abuse.”
136. Williams and Finkelhor, “Parental care-giving and incest.”
10 The Incest Taboo as
Darwinian Natural Right

Larry Arnhart

In the twenty-first century, we will see great advances in the biological

understanding of human nature. This will force us to think about whether
biological science can explain that most distinctive trait of our humanity—
our moral sense of right and wrong. Some people will argue that our moral
experience transcends our biological nature. Others will argue that we
should be able to explain our morality as an expression of our biological na-
ture. How we decide that debate might be decisively influenced by whether
we accept or reject Edward Westermarck’s Darwinian theory of the incest
taboo as a natural expression of human moral emotions.
Westermarck contended that incest is wrong because most human be-
ings are naturally inclined to learn an emotional aversion to incest. The in-
cest taboo varies in its details because different societies have different sys-
tems of kinship. But generally across all societies incestuous relations with
one’s siblings, parents, or children is regarded as wrong because it elicits
moral emotions of disapproval. Westermarck inferred that since close in-
breeding tended to have deleterious effects on offspring that would lower
their fitness for survival and reproduction, natural selection would have fa-
vored the natural tendency to feel a sexual aversion toward those with
whom one had been reared in early childhood, and this emotional aversion
toward incest would tend to be expressed as social disapproval through an
incest taboo. Consequently, there is no universal, cosmic imperative of rea-
son that dictates that incest is wrong. Rather, incest is wrong for a particu-
lar species of animal that has a natural propensity to develop strong moral
emotions against incest. Even within the human species, there is variation
in this propensity, such that some individuals do not feel an aversion to in-
cest, but they will be treated by the rest of us as moral strangers. Wester-
marck thought this illustrated the general character of ethics as rooted in
the moral emotions of the human animal as shaped by natural selection in
evolutionary history.1
 Incest Taboo as Darwinian Natural Right 191

The importance of Westermarck’s theory for biological explanations of

ethics is evident in Edward O. Wilson’s work and the continuing contro-
versy over his work. Harvard University Press recently published the twenty-
fifth anniversary edition of Wilson’s Sociobiology. From its first publication
in 1975, that book has stirred a dispute over his claim that ethics is rooted
in human biology. Our deepest intuitions of right and wrong, he asserted on
the first page of the book, are guided by the emotional control centers of the
brain, which evolved by natural selection to help the human animal exploit
opportunities and avoid threats in the natural environment.2 In 1998 Wil-
son’s book Consilience renewed the controversy as he continued to argue
for explaining ethics through the biology of the moral sentiments.3 In recent
decades, the dispute over Wilson’s intellectual project has become one of the
most prominent public debates over the application of modern science to
the study of human social life.4 Since morality is ultimately rooted in the
moral sentiments of human nature, Wilson claims, a natural science of
morality would require a biology of the moral sentiments.5 Throughout his
writings, Wilson has used Westermarck’s Darwinian theory of the incest ta-
boo as the prime example of how biology can explain the moral sentiments.6
Human nature is not a product of genes alone or of culture alone, Wilson
insists. Rather, human nature is constituted by “the epigenetic rules, the
hereditary regularities of mental development that bias cultural evolution in
one direction as opposed to another, and thus connect the genes to culture.”7
The biology of the moral sentiments would be the study of the “epigenetic
rules” of moral experience as shaped by the complex interaction of genetic
propensities and cultural learning. And for Wilson, Westermarck’s theory of
the incest taboo is the best example of this.
In the debate over the ultimate grounds of ethics, Wilson sees two fun-
damentally opposed positions. The “transcendentalists” claim that ethics is
rooted in absolute standards that exist outside the human mind, while the
“empiricists” claim that ethics is rooted in natural human inclinations.8
Wilson defends the empiricist position, which he thinks can be traced back
to Aristotle’s Nicomachean Ethics. He rejects the transcendentalist posi-
tion, which he thinks is evident in the ethical writings of philosophers like
Immanuel Kant and John Rawls.9
It is remarkable that so few people accept Wilson’s biological view of
ethics as rooted in natural moral sentiments. It is hardly surprising that his
most fervent critics reject his account of ethics as confirming their charge of
crude biological reductionism and determinism. But it is surprising that
even those who generally agree with him about applying Darwinian theory
to human behavior reject his Darwinian view of ethics. All of Wilson’s in-
tellectual enemies and almost all of his intellectual friends seem to agree
that there is an absolute separation between natural facts and moral values,
192 arnhart

between is and ought, so that a biological science of ethics would commit

the “naturalistic fallacy” by inferring a moral ought from a natural is.
Professional philosophers—like Thomas Nagel, for example—reject any
Darwinian account of ethics for failing to see that ethics is “an autono-
mous theoretical subject” that manifests the “moral point of view” as at-
tained by the exercise of human reason in transcending human biology.10
Even the proponents of “evolutionary psychology,” who are generally sup-
portive of Wilson’s position, reject any biological science of ethics as vio-
lating the fact-value dichotomy. David Buss, for example, in his Darwinian
account of human mating, warns against the “naturalistic fallacy,” which,
he says, “confuses a scientific description of human behavior with a moral
prescription for that behavior.” He insists that “judgments of what should
exist rest with people’s value systems, not with science.”11 Steven Pinker ex-
plains almost everything about human behavior in Darwinian terms, but
even he concludes that there is no scientific explanation for ethics. “How
did ought emerge from a universe of particles and planets, genes and bod-
ies?” Natural science cannot answer that question, Pinker asserts, because
our human ability to reason from a moral ought transcends our natural in-
clinations.12 When Wilson delivered the keynote address to the 1996 an-
nual meeting of the Human Behavior and Evolution Society, which is the
principal organization for scholars applying Darwinian views of human na-
ture to the study of human social behavior, he spoke of the need for a unity
of the sciences that would promote a biological science of ethics. Many of
those in the audience were shocked by his denial of the fact-value dichot-
omy and his refusal to accept the transcendent autonomy of ethics.13 It seems
that almost everyone is a “transcendentalist” in viewing ethics as manifest-
ing transcendent values, and almost no one is persuaded by Wilson’s “em-
piricist” view of ethics as rooted in natural moral emotions. Almost every-
one agrees that because of the dichotomy between is and ought, reasoning
about the moral ought belongs to a transcendent realm of human freedom
beyond the empirical world of nature.
When I first read Wilson’s Sociobiology in 1975, I too rejected his bio-
logical explanation of ethics as being too crudely reductionistic in its ap-
peal to mere emotion as the ultimate ground of ethical experience. At the
time, I was a graduate student at the University of Chicago writing a dis-
sertation on Aristotle’s Rhetoric, and I noticed that Aristotle invoked a nat-
ural moral sense as expressed in moral emotions such as anger, indignation,
shame, kindness, and pity, which made me think that Wilson’s reliance on
the moral emotions might be more defensible that I had initially believed.14
I also became interested in Aristotle’s biological writing and in how his bi-
ological reasoning influenced his ethical philosophy. Like Wilson, Aristotle
explained the natural sociality of human beings by comparing them with
 Incest Taboo as Darwinian Natural Right 193

other social animals such as the social insects. When Aristotle spoke of
“natural right”—natural standards of right and wrong—he appealed to the
biological propensities of human nature. I also saw that Aristotelian phi-
losophers like Thomas Aquinas spoke of “natural right” or “natural law”
as “that which nature has taught all animals.” Eventually, I changed my
mind about Wilson’s argument and concluded that his Darwinian explana-
tion of ethics could be defended as a modern biological restatement of a
tradition of ethical naturalism that began with Aristotle.
Now, I argue that a Darwinian science of human nature can explain
ethics as conforming to what I call “Darwinian natural right.”15 I also
agree with Wilson that Westermarck’s theory of the incest taboo is the best
illustration of how such a natural science of ethics might work. A Darwin-
ian view of human nature can support a form of ethical naturalism that is
compatible with the philosophic tradition of reasoning about “natural right”
or “natural law,” a tradition that stretches from Plato and Aristotle in an-
cient Greece to Thomas Aquinas in the Middle Ages and then to David
Hume and Adam Smith in the early modern world. Furthermore, through-
out that tradition, the incest taboo is one of the prime examples of how
ethics might be rooted in the moral emotions of human nature. The great
achievement of Westermarck’s theory of incest is how it sustains that philo-
sophic tradition by providing a Darwinian explanation of the natural moral
emotions associated with the incest taboo. Wilson is right, therefore, in see-
ing that recent research confirming Westermarck’s theory provides power-
ful support for a naturalistic view of ethics as founded in the moral senti-
ments of human nature.

The Natural Right Tradition

Among the Greek sophists, it was common to contrast physis and

nomos, “nature” and “convention.” (Today we would speak of this as the
dichotomy between “nature” and “nurture.”) The sophists often asserted
that the social norms of human conduct are rooted in “convention” rather
than “nature.” Aristotle thought Callicles in Plato’s Gorgias was speaking
for the sophists generally in arguing that “nature and convention are op-
posites, and justice is a noble thing according to convention, but not so ac-
cording to nature.”16 By nature, the sophists seemed to say, human beings
are too selfish and competitive to live together in civilized communities,
and therefore social order requires the invention of social conventions or
laws that repress the otherwise unruly desires of human nature.17 If there is
any natural standard of right at all, Callicles declared, it would dictate the
rule of the stronger over the weaker, but this is exactly what conventional
morality denies.
194 arnhart

Against the sophists, Plato, Aristotle, and the other Socratic philosophers
argued that justice was not merely conventional, because one could discover
norms that were just or right by nature. The Socratic philosophers thus be-
gan a tradition of reasoning about “natural right” or “natural law.” As the
very term “natural law” suggests, this tradition tries to overcome the so-
phistic antinomy of “nature” versus “convention,” because the thought is
that some conventions are “according to nature” and others are “contrary
to nature.” But how are we to understand the complex interaction of nature
and convention? And how are we to distinguish the natural conventions
from the unnatural conventions? If social order is rooted in a universal hu-
man nature, how can we explain the apparent diversity of social conven-
tions across different societies? I believe that a Darwinian understanding of
the biological nature of social order helps to answer these questions, and
thus it vindicates the Socratic tradition of ethical naturalism against the so-
phistic tradition of ethical conventionalism.
In Plato’s Laws, Socrates speaks of the avoidance of incest as an “un-
written law” that is so strong that “among the many there isn’t the slightest
desire for this sort of intercourse.” The strength of this law arises from the
fact that everyone from the moment of birth hears incest condemned as
“hateful to the gods and the most shameful of shameful things.” Plato
leaves it unclear, however, why this sacred taboo arises in the first place.
The discussion of incest avoidance in Plato’s dialogue arises in the context
of devising laws for a good city that will be “according to nature.” But
there is no explicit discussion of whether—and if so, how—the law of in-
cest avoidance might have natural causes.18 In Xenophon’s Memorabilia,
however, Socrates identifies the “unwritten laws” legislated by the gods as
laws that could not be disobeyed without natural penalties. He speaks of
the incest taboo as one of those “unwritten laws,” because those commit-
ting incest would tend to produce defective offspring.19
Aristotle developed the biological basis of this natural law in his biolog-
ical writings. Even his works of ethical and political philosophy show a bi-
ological view of human nature, so that some scholars have concluded that
for Aristotle, “ethics and politics are in a way biological sciences.”20 The
sophists had argued that all justice was conventional and not natural, be-
cause while nature is invariable, the rules of justice are variable. But Aris-
totle saw that for biological phenomena, natural law is that which happens
“for the most part” but not always. So, for example, there is a natural pro-
pensity among most people for the right hand to be stronger, although some
people are left-handed, and some can be habituated to be ambidextrous.
Similarly, what is naturally right conforms to the natural propensities or in-
clinations of human beings, but this will vary according to the variable cir-
cumstances of action.21
 Incest Taboo as Darwinian Natural Right 195

Thus did the Socratic philosophers of ancient Greece start a tradition of

reasoning about “unwritten laws” that might be “according to nature,” in
which the incest taboo was one example of a natural propensity of human
biological nature that was enforced by the natural penalties of disobedi-
ence. The idea of “natural law” was elaborated by Stoic philosophers in
ancient Rome. Then, in the Middle Ages, Thomas Aquinas developed the
idea in its fullest form. Wilson dismisses the Thomistic tradition of natural
law as “transcendentalist,” because the natural law is viewed as expressing
God’s will, and thus it is independent of the human mind.22 But I think
Wilson is mistaken. Aquinas’s natural law is actually an “empiricist” posi-
tion, at least insofar as Aquinas speaks of natural law as rooted in natural
inclinations that are instinctive to human beings. Although Aquinas be-
lieves that God is the ultimate cause of everything in nature, the natural in-
clinations that constitute the natural law can be known by natural human
experience even without religious faith.
Aquinas condemns incest as contrary to the “natural law” (lex naturalis)
or “natural right” (ius naturale) that governs sexual mating and familial
bonding. Aquinas indicates the biological character of the natural moral law
when he quotes the ancient Roman jurist Ulpian as declaring that “natural
right is that which nature has taught all animals.” Natural law is rooted in
“natural inclination” (inclinatio naturae) or “natural instinct” (instinctu na-
turae), and therefore “all those things to which a human being has a natural
inclination, reason naturally apprehends as good, and consequently as ob-
jects to be pursued, and the contraries of these as evil and to be avoided.”
Natural law for human beings requires reason, but pure reason could not
move us to act if it were not linked somehow to inclination or instinct (that
is, some affective or emotional impulse) as the motivation to action.
Aquinas believes that human beings, like some other animals, are natu-
rally inclined to sexual union and parental care.23 Marriage belongs to nat-
ural law because it serves two natural inclinations—the care of offspring
and conjugal bonding. Aquinas can judge the naturalness of marital ar-
rangements, therefore, by how well they promote those two natural ends.24
He speaks of the human disposition to marriage as a “natural instinct of the
human species.”25 The primary natural end of marriage is to secure the
parental care of children. The secondary natural end is to secure the conju-
gal bonding of male and female for a sexual division of labor in the house-
hold. Among some animals, Aquinas observes, the female can care properly
for her offspring on her own, and thus there is no natural need for any en-
during bond between male and female. But for those animals whose off-
spring require care from both parents, nature implants an inclination for
male and female to stay together to provide the necessary parental care.26
Just as is the case for those animals whose offspring could not survive or
196 arnhart

develop normally without parental care, human offspring depend upon par-
ents for their existence, their nourishment, and their education. To secure
this natural end, nature instills in human beings, as it does in other animals,
natural desires for sexual coupling and parental care. Even if they do not
have children, however, men and women naturally desire marital union be-
cause, not being self-sufficient, they seek the conjugal friendship of husband
and wife sharing in household life.
Marriage as constituted by customary or legal rules is uniquely human,
because such rules require a cognitive capacity for conceptual reasoning
that no other animals have. But even so, such rules provide formal struc-
ture to natural desires that are ultimately rooted in the animal nature of
human beings.
Aquinas gives at least five reasons why incestuous marriage would be
contrary to natural law.27 First, as Pope Gregory I said, “we have learned by
experience that the children of such a union cannot thrive.” (William
Durham cites this statement from Pope Gregory as evidence that the Cath-
olic prohibition of incest was originally based on some recognition of the
physical abnormalities that come from inbreeding.)28 Second, incest would
disrupt familial relationships by impeding the reverence that children owe
to their parents. Third, incest would disrupt the family by widening the
range of sexual desire to include all of the kin within the household. Fourth,
incest would discourage the friendly alliances that come from marrying out
of one’s own group. Fifth, human beings have a “natural abhorrence” of in-
cest, an abhorrence that is even shown by some other animals. To support
this last point, Aquinas cites Aristotle’s claim that among camels and
horses, sons naturally abhor copulation with their mothers.29 Thus does
Aquinas anticipate in some manner most of the theories developed by mod-
ern scholars to explain the universality of the incest taboo.30

Hobbes and Smith

The modern break with the Aristotelian and Thomistic tradition of

natural law as rooted in the biology of human nature began in the seven-
teenth century with Thomas Hobbes. Aristotle and Aquinas had claimed
that human beings are by nature social and political animals. Hobbes denied
this claim and asserted that social and political order is an utterly artificial
human construction. For Aristotle and Aquinas, as belonging to the Socratic
tradition of ethical naturalism, moral and political order was rooted in bio-
logical nature. But for Hobbes, as belonging to the sophistic tradition of eth-
ical conventionalism, social order required a conquest of nature by which
human beings transcend their animal nature. What Hobbes identified as the
 Incest Taboo as Darwinian Natural Right 197

“laws of nature” that should govern human social conduct were actually
“laws of reason” by which human beings contrive by rational artifice to es-
cape the disorder that ensues from following their natural inclinations.31
Hobbes assumes a radical separation between animal societies as founded
on natural instinct and human societies as founded on social learning. Hu-
man beings cannot be political animals by nature, Hobbes believes, because
“man is made fit for society not by nature but by education.”32 Against
Aristotle and Aquinas, Hobbes argued that this dependence of human social
order on artifice and learning meant that human beings were not at all like
the naturally social animals (such as bees and ants).33 Despite the monism of
Hobbes’s materialism, in which he seems to think everything is ultimately
reducible to matter in motion, his political teaching presupposes a dualistic
opposition between animal nature and human will: in creating political or-
der, human beings transcend and conquer nature.34
This Hobbesian dualism is developed by Immanuel Kant in the eigh-
teenth century in formulating the modern concept of culture.35 Culture be-
comes that uniquely human realm of artifice in which human beings escape
their natural animality to express their rational humanity as the only beings
who have a “supersensible faculty” for moral will. Through culture, human
beings free themselves from the laws of nature. (This Kantian notion of cul-
ture as a self-contained, autonomous sphere of uniquely human meaning
eventually became the fundamental idea for cultural anthropology.)36
Opposing the Hobbesian claim that human beings were naturally aso-
cial and amoral, Francis Hutcheson and other Scottish philosophers of the
eighteenth century argued that human beings were endowed with the nat-
ural instincts of social animals, and this natural sociality supported a nat-
ural moral law as expressed in the natural moral sense. Hutcheson’s theory
of the moral sense revives the Thomistic conception of natural law as
founded in the inclinations or instincts of human nature.37
Against Hobbes, Lord Shaftesbury asserted that there was a natural
moral sense. Bernard Mandeville then responded to Shaftesbury by con-
tending that morality was a matter of custom or convention. To illustrate
this, he claimed that incestuous marriages are customary in some societies,
and “there is nothing in nature repugnant against them, but what is built
upon mode and custom.”38 Challenging this claim, Hutcheson insisted that
the abhorrence of incest did indeed express a natural moral sense.39
As a student of Hutcheson’s at the University of Glasgow, Adam Smith
developed a similar position.40 Smith agreed with Aquinas in condemning
incestuous marriage as “shocking and contrary to nature.”41 It is contrary
to nature, Smith explained, because the natural affections of familial at-
tachment between parents and children or between siblings is contrary to
the natural affections of sexual mating between husband and wife. There
198 arnhart

is, then, a natural tendency for most human beings to feel incest to be
“shocking and abominable.” (Similarly, proponents of Westermarck’s the-
ory of incest avoidance have argued that the “familial attraction” that bonds
parents and children is evolutionarily distinct from the “sexual attraction”
that bonds husband and wife, so that secure familial bonding excludes sex-
ual bonding.42)
But while the marriage of parent to a child or of siblings to one another
would be universally contrary to nature, Smith believed, the other rules for
avoiding incest can vary depending on the variable rules of kinship as de-
termined by custom. For example, prohibiting a man from marrying his de-
ceased wife’s sister, because the wife’s sister is considered to be the hus-
band’s sister, is a rule of custom rather than of nature. (The passage in 1907
of a British law allowing a man to marry his deceased wife’s sister was pre-
ceded by sixty-five years of intense Parliamentary debate.43) Thus, the nat-
ural propensity to abhor incest will be most strongly expressed in response
to relations within the immediate family circle of parents and children, but
there will be variation in how far the incest taboo extends beyond the nu-
clear family.
While Smith spoke repeatedly of nature as instilling those moral senti-
ments that would promote the survival and propagation of human beings
as social animals, he could not explain exactly how it was that nature could
shape the human animal in this way. Such an explanation was later pro-
vided by Charles Darwin and elaborated by Edward Westermarck.

Darwin

Adopting Smith’s theory of the natural moral sentiments, Darwin

then went beyond Smith in showing how the moral sense could have arisen
in human nature as a product of natural selection.44 Darwin agreed with
Kant and other writers “who maintain that of all the differences between
man and the lower animals, the moral sense or conscience is by far the most
important.” But while Kant had written about the human sense of duty or
“ought” as showing us “man as belonging to two worlds”—the empirical
world of natural causes and the transcendent world of moral freedom—
Darwin suggested that human morality could be studied “exclusively from
the side of natural history.”45 In developing his evolutionary theory of mor-
ality, Darwin adopted a Smithian naturalism rather than a Kantian dualism.
Like Smith, Darwin saw nature as the comprehensive whole of which hu-
man beings are a part. And thus he rejected Kant’s dualistic separation be-
tween the “phenomenal” realm of causal nature and the “noumenal” realm
of human freedom.
 Incest Taboo as Darwinian Natural Right 199

Like Aquinas and Smith, Darwin recognized the abhorrence of incest as

a universal trait of human societies. And although he did not elaborate his
explanation, he suggested that avoiding incest would have the advantage of
preventing the bad effects of interbreeding in the same family.46 He specu-
lated that the feeling of sexual aversion toward those with whom one had
been raised in early childhood might have been favored by natural selection
in human evolutionary history.47 This is the idea subsequently developed by
Westermarck.

Westermarck

Westermarck first won international recognition with the publica-

tion in 1889 of his History of Human Marriage, a massive survey of the
subject, in which he explained the desires for family life and sexual mating
as founded in moral emotions that had been shaped by natural selection as
part of the biological nature of human beings. In later writings—such as
The Origin and Development of the Moral Ideas (1906) and Ethical Rela-
tivity (1932)—he elaborated a general theory of ethics as an expression of
the moral emotions.48 He combined ideas from Smith and Darwin with the
comparative data of social behavior as collected by anthropologists and so-
ciologists to defend a naturalistic theory of ethics as rooted in the natural
moral sentiments. He thus provided an empirically grounded scientific the-
ory of human nature supporting the tradition of natural law reasoning.49
In agreement with Aquinas, Smith, and Darwin, Westermarck argues
that human marriage is natural because it satisfies some of the deepest in-
clinations of human nature. Against those who believe that primitive hu-
man beings lived in a state of complete promiscuity with no enduring tie
between male and female, Westermarck contends that conjugal bonding
has always been a natural instinct for human beings, because it was fa-
vored by natural selection to provide parental care for offspring that could
not survive without such care.50 This natural inclination to conjugal bond-
ing gave rise to habits, customs, and institutions that sanctioned marriage
as an enduring union of parents and children. As social animals with a
highly developed intellect, human beings feel moral indignation toward
men who abandon their wives and children, and this moral emotion of dis-
approval is expressed in customary and legal rules that enforce the duties
and rights of spouses, parents, and children.
Much of Westermarck’s work on human marriage is about the great
variability in marital practices across different societies and different his-
torical periods—ranging from monogamy to polygyny to polyandry. Even
in this variation, however, he sees regularities that manifest the universal
200 arnhart

nature of human marital emotions, and they are the same regularities seen
by Aquinas, Smith, and Darwin.51 Monogamy is practiced in all societies,
and in some it is the only permissible form of marriage. Monogamy is uni-
versal because it satisfies the human instincts for conjugal bonding and
parental care. Although polygyny has been common in many societies, be-
cause many men have a natural desire for multiple mates, the co-wives are
naturally inclined to sexual jealousy, which creates conflicts that are diffi-
cult to manage. Polyandry is the rarest form of marriage because the in-
tense jealousy of men makes it almost impossible for them to share a wife.
When polyandry is practiced, it seems to be a response to unusual circum-
stances, such as a low population of women in proportion to men, or harsh
economic conditions that force brothers to share a wife so that their fam-
ily’s property is not divided.52
The most famous part of Westermarck’s study of marriage is his theory
of the incest taboo.53 Like Plato, Aquinas, Smith, and Darwin, Westermarck
sees that incest is almost universally condemned as a morally abhorrent vio-
lation of nature. All societies prohibit mothers from marrying their sons,
and fathers from marrying their daughters. And with few exceptions, all so-
cieties prohibit marriages of brothers and sisters who are children of the
same parents. Westermarck observes that the many theories offered to ex-
plain this are unsatisfactory. As the best summary of his reasoning, the fol-
lowing passage from Westermarck’s Ethical Relativity deserves to be quoted
at length.
The theories in question imply that the home is kept free from incestuous intercourse
by law, custom, or education. But even if social prohibitions might prevent unions
between the nearest relatives, they could not prevent the desire for such unions. The
sexual instinct can hardly be changed by prescriptions; I doubt whether all laws
against homosexual intercourse, even the most draconic, have ever been able to ex-
tinguish the peculiar desire of anyone born with homosexual tendencies. Neverthe-
less, our laws against incest are scarcely felt as a restraint upon individual feelings.
And the simple reason for this is that in normal cases there is no desire for the acts
which they forbid. Generally speaking, there is a remarkable absence of erotic feel-
ings between persons living closely together from childhood; among the lower ani-
mals, also, there are indications that the pairing instinct fails to be stimulated by
companions and seeks strangers for its gratifications. . . . Plato showed a sharper eye
for the problem of incest in his observation that an unwritten law defends as suffi-
ciently as possible parents from incestuous intercourse with their children and broth-
ers from intercourse with their sisters, and that the thought of such a thing does not
enter at all into the minds of most of them.
Sexual indifference, however, is not by itself sufficient to account for exogamous
prohibitions. But such indifference is very generally combined with sexual aversion
when the act is thought of; indeed, I believe that this is normally the case whenever
the idea of sexual intercourse occupies the mind with sufficient intensity and a de-
 Incest Taboo as Darwinian Natural Right 201

sire fails to appear. . . . Aversions which are generally felt readily lead to moral dis-
approval and prohibitory customs and laws. This I take to be the fundamental
cause of the exogamous prohibitions. Persons who have been living together from
childhood are as a rule near relatives. Hence their aversion to sexual relations with
one another displays itself in custom and law as a prohibition of intercourse be-
tween near kin.54

Thus, like Plato and the other Socratic philosophers, Westermarck does
not think the incest taboo can be explained simply as a product of “law,
custom, or education,” because “in normal cases” there is a natural aver-
sion to incest that constitutes “an unwritten law.” This natural aversion is
then expressed as a legal or customary prohibition against incest. And yet
this natural aversion and its expression in law and custom are not natural
necessities that hold in every case but natural propensities that hold in most
cases. The incest taboo, like any moral rule, is a generalization of natural
emotions that hold sway “normally” in the minds of “most” people in re-
sponse to circumstances that occur “as a rule.” As Aristotle argued, what
is naturally right is variable, but it is still natural insofar as it expresses nat-
ural propensities of the human animal diversely expressed in human cus-
tom and law.
Westermarck’s Darwinian theory for explaining this can be stated in
three propositions.55 First, inbreeding tends to produce physical and men-
tal deficiencies in the offspring that lower their fitness in the Darwinian
struggle for existence. Second, as a result of the deleterious effects of in-
breeding, natural selection has favored the mental disposition to feel an
aversion to sexual mating with those with whom one has been intimately
associated from early childhood. Third, this natural aversion to incest has
inclined human beings to feel moral disapproval for incest, and this moral
emotion has been expressed culturally as an incest taboo.
Westermarck’s view of incest illustrates his general account of ethics as
rooted in natural emotions shaped by natural selection in human evolu-
tionary history. The avoidance of incest works through an emotional aver-
sion favored by natural selection. Because this emotion tends to be shared
by most human beings, it gives rise to moral emotions of disapproval that
are expressed in customary and legal rules that prohibit incest. These cus-
tomary and legal rules are culturally variable in their specific details, but
the cultural rules are grounded in an emotional propensity of human na-
ture that is universal.
Westermarck’s theory of incest was rejected by Sigmund Freud and oth-
ers who believed that the incest taboo shows how moral rules arise as cul-
tural inventions that suppress the immoral emotions of human nature.56
Freud was a Hobbesian, in the sophistical tradition of ethical convention-
alism, who saw human beings as so naturally selfish in their emotions that
202 arnhart

they could not live together in civilized societies unless they created cultural
rules to subdue their natural inclinations. Freud thought that civilization
required a rational rule of law to conquer “man’s natural aggressive in-
stinct, the hostility of each against all and of all against each.”57 Human
beings must deny their animal nature through the moral imperatives of hu-
man culture as an autonomous realm of human rationality set apart from
nature. The incest taboo was the most momentous manifestation of this
human denial of nature, Freud insisted, because it was “the most drastic
mutilation which man’s erotic life has in all time experienced.”58 The incest
taboo illustrates the general character of ethics as the rule of the “cultural
super-ego” in demanding the renunciation of natural inclinations.59 Ilham
Dilman, in his book Freud and Human Nature, comments on the remark-
able similarity of Freud’s view of human nature to that taken by the sophist
Callicles in Plato’s Gorgias.60
Claude Lévi-Strauss spoke for the many social scientists who adopted this
Freudian version of the sophistic position when he described the transcen-
dent character of the incest taboo: “Before it, culture is still non-existent;
with it, nature’s sovereignty over man is ended. The prohibition of incest is
where nature transcends itself.”61 By contrast, Westermarck believed that the
incest taboo shows how moral rules arise as cultural practices that express
human nature in manifesting the moral emotions. Like Plato, Aristotle,
Aquinas, Smith, and Darwin, Westermarck saw human beings as naturally
social animals with the natural emotions that fitted them for social life.
The debate between Westermarck and Freud over the origins of the in-
cest taboo manifests a fundamental debate in the social sciences. For West-
ermarck, as representing the Socratic natural law tradition in social theory,
human culture arises as the cultivation of human nature. For Freud, as rep-
resenting the sophistic tradition in social theory, human culture arises as
the conquest of human nature.
For Westermarck the incest taboo illustrates how social order arises from
the complex interaction of nature and convention: the incest taboo is a so-
cial convention that expresses the human nature of the moral emotions. As
a social convention, the incest taboo will vary across societies with diverse
kinship systems. But as an expression of natural emotions, the incest taboo
will show a natural propensity for most people to learn a sexual aversion
for those with whom they have been reared from early childhood.
If Westermarck’s ethical theory really is founded on an empirical science
of human nature, as he suggests, then it should be subject to empirical con-
firmation or falsification. The debate over his theory of the incest taboo il-
lustrates how his claims might be tested by scientific research. Beginning
with Freud, Westermarck’s theory has been dismissed with two criticisms.
First, the occurrence of incest in all societies seems to indicate that there is
 Incest Taboo as Darwinian Natural Right 203

no natural resistance to it. Second, if the taboo were natural, there would
seem to be no need for cultural rules to enforce the taboo.62 But these criti-
cisms assume a simple dichotomy between fixed instinct and flexible cul-
ture that Westermarck denies. According to Westermarck, the instinctive
propensity to incest avoidance is a tendency to learn sexual aversion when
certain conditions are satisfied: most human beings are inclined to feel sex-
ual aversion toward those with whom they have been reared from early in-
fancy. Westermarck predicts, therefore, that most human beings raised in
the familial environment typical for human beings will feel a strong aver-
sion to incestuous relationships. But he also predicts that in some circum-
stances, some human beings will not acquire this aversion. For example,
father-daughter incest is more likely to occur when the fathers have been
separated from their daughters during their early rearing. Furthermore, he
predicts that because of the natural variability in human emotional tem-
peraments, a few human beings will not develop the aversion to incest that
is normal for most people, and these deviant individuals will provoke a
deep disgust from others. Because of this circumstantial and temperamen-
tal variability, human communities will develop cultural practices to en-
force an incest taboo expressing the general feelings of the community in
condemning those few who are inclined to commit incest.63
The incest taboo thus illustrates how social order arises from the com-
plex interaction of nature and convention: the incest taboo is a social con-
vention that expresses the human nature of the moral emotions. As a social
convention, the incest taboo will vary across societies with diverse kinship
systems. But as an expression of natural emotions, the incest taboo will
show a natural propensity for most people to learn a sexual aversion for
those with whom they have been reared from early childhood.
Arthur Wolf has indicated, in his survey of the scientific study of incest,
that Westermarck’s predictions seem to have been confirmed by the evi-
dence. Wolf’s special contribution to this research is his study of marriage
in China. In parts of China, there were once three forms of marriage. In the
“major” form of marriage, the bride went to live with her husband’s fam-
ily on the day of the wedding. In the “minor” form, a girl in infancy would
join the family of her future husband as a simpua, or “little daughter-in-
law,” but she would not be married until she reached sexual maturity years
later. In the “uxorilocal” form, the husband would submit to the authority
of his father-in-law. From his meticulous study of marriage records in Tai-
wan, Wolf concluded that people in minor marriages showed far more sex-
ual dissatisfaction than those in the other forms of marriage. They tended
to produce more divorces, more adultery, and fewer children. He saw this
observable behavior as showing that having been reared together in the
same family from early infancy (age three or earlier), these spouses felt the
204 arnhart

sort of sexual aversion to one another that would be predicted by Wester-

marck’s hypothesis. Although they were not genetically related as brother
and sister, they displayed the same emotional discomfort with sexual union
that brothers and sisters typically feel toward one another. Natural selec-
tion has endowed us with a natural instinct to learn an emotional aversion
to sexual mating with those with whom we have been intimately associated
in our early years of rearing, because in the circumstances of evolutionary
history this would avoid the deleterious consequences of breeding with
close kin. This same natural propensity will produce such an aversion even
when the people with whom we have been reared are not our genetic kin.
The experience from the Israeli kibbutzim shows the same pattern.64 In
kibbutzim with collective child-rearing, children from different families
grew up together from earliest infancy. Although not biologically related,
they lived with one another as if they were siblings. And although they
were permitted to marry, they never did, because they felt no sexual attrac-
tion to one another. As predicted by Westermarck, early childhood associa-
tion inhibited sexual attraction.
Freud and other critics of Westermarck assumed that human beings were
the only animals that avoid incest, and thus it seemed that the incest taboo
must be a uniquely human cultural invention by which human beings sub-
due their animal emotions. Westermarck believed that many wild animals
are naturally inclined to avoid inbreeding.65 Recent evidence from animal
behavior studies seems to support Westermarck. For example, monkeys and
apes show a tendency to avoid incest that is similar to that displayed by hu-
man beings.66 As with human beings, incest does occur among other pri-
mates, but it is unusual, and it seems to arise only among exceptional indi-
viduals with abnormal temperaments or among those who have no close
bonding with their kin.67 Among chimpanzees, the most common mecha-
nism for avoiding inbreeding is that the females leave their natal group be-
fore breeding and join another group. (For most other primate species, the
males leave before breeding.) Some chimpanzee females, however, do not
leave their natal group; and all females live in the same group as their adult
sons. While there is some mating of immature individuals with immature or
adult relatives, mating of related adults is extremely rare. Mating is most
strongly inhibited between mothers and sons and between maternal siblings.
It seems, then, that chimpanzees have a natural propensity to avoid incest.68
Since chimpanzees are genetically closer to human beings than is any other
living species, it seems likely that incest avoidance among human beings
arises from a genetic propensity derived from a common ancestor.
Westermarck surveyed the biological research of his time suggesting that
inbreeding tended to produce high rates of infant mortality and of mental
and physical defects.69 Modern genetic research confirms this conclusion.
 Incest Taboo as Darwinian Natural Right 205

Inbreeding increases the probability that deleterious recessive alleles in a

population will be expressed, because any allele is more likely to be inher-
ited simultaneously from both the paternal and maternal lines of a geneal-
ogy. This probability of producing genetically defective offspring increases
in direct proportion to the closeness of the genetic relationship between
two inbreeding individuals.70
But although inbreeding is costly, so is outbreeding. For example, trav-
eling to a strange group to search for a mate might be costly. Consequently,
some animals might be naturally disposed to seek a balance between in-
breeding and outbreeding by choosing mates that differ somewhat but not
too much from kin.71
To me all of this evidence provides convincing support for Wester-
marck’s theory of the incest taboo. But it also seems to me that Wester-
marck’s theory is compatible with the partial truth of the other major theo-
ries for explaining the incest taboo. Once the Westermarckian mechanism
for learning incest avoidance is implanted in human nature by natural se-
lection, then human beings might learn by experience that avoiding incest
has many social benefits, and the recognition of those benefits might then
provide utilitarian reasons for reinforcing the incest taboo. Incest avoid-
ance mitigates sexual competition within the family. It promotes coopera-
tion between kinship groups by encouraging intermarriage. And it avoids
the harmful effects of inbreeding. There is some truth, therefore, to those
theories of the incest taboo that emphasize these social benefits as likely
motivations for human beings to deliberately enforce the taboo. But this is
fully compatible with Westermarck’s explanation for how the propensity to
learn incest avoidance originally emerged in human evolutionary history.
“Incest is considered harmful because it is disapproved of,” Westermarck
observes, “and it is not in the first place disapproved of because it is con-
sidered harmful.”72 This illustrates Adam Smith’s claim that considerations
of social utility influence our moral sentiments, although such utilitarian
considerations cannot alone explain those sentiments.73
Even Freud’s psychoanalytic theory might be partially true in a way that
is compatible with Westermarck’s theory, because if many of the patients
that Freud saw had been reared by nurses in isolation from their parents,
then Westermarck would predict that they would be inclined to develop in-
cestuous urges that would need to be repressed. After all, the myth of Oedi-
pus that so impressed Freud is the story of a man who had been separated
from his parents at birth, who therefore did not experience the childhood
familiarity with his mother that Westermarck believed was necessary to in-
still a sexual aversion to one’s mother.74
The evidence for Westermarck’s theory of the incest taboo comes from a
variety of intellectual disciplines—sociology, anthropology, primatology,
206 arnhart

genetics, and evolutionary biology. In this way, it provides Edward Wilson

with one of his best examples of what he calls “consilience”—the search
for the unity of knowledge based on the idea that nature is governed by a
seamless web of causal laws that cross the traditional disciplines of study, a
search that began, Wilson thinks, with the ancient Greek philosophers.

Objections and Replies

Many scholars reject this idea of a comprehensive science of natural

law that could explain human ethics as part of nature. In many cases, the
objections to Wilson’s Darwinian ethics are motivated by this fundamental
rejection of natural law in ethics. For example, when Thomas Nagel argues
that ethics as “an autonomous theoretical subject” cannot be explained by
Wilson’s sociobiology, he casually dismisses natural law theories because
“they no longer have sophisticated adherents.”75 In fact, most of the objec-
tions to Darwinian ethics are restatements of the arguments John Stuart
Mill used in his essay “Nature” against the idea of a natural moral law.76
But it seems to me that the evidence and arguments for a Westermarckian
theory of the incest taboo as rooted in the moral emotions of human nature
suggest answers to the most common objections.
A survey of some of the recent critiques of Darwinian ethics—such as
Paul Lawrence Farber’s book The Temptations of Evolutionary Ethics and
Peter Woolcock’s essay “The Case Against Evolutionary Ethics Today”—
would show that four objections are particularly prominent.77 The most
frequent objection to any ethical naturalism is that moral values cannot be
derived from natural facts. This fact-value dichotomy is generally attrib-
uted to David Hume, who is said to have shown that there must be a radi-
cal separation between questions of what is or is not the case, which belong
to the realm of nature, and questions of what ought or ought not to be
done, which belong to the realm of morality. Because of this dichotomy, it
seems to be a logical fallacy to infer a moral ought from a natural is. G. E.
Moore called this “the naturalistic fallacy.”78 According to those who insist
on such a dichotomy, natural science can describe the way things are, but it
cannot prescribe the way things ought to be. A Darwinian science of hu-
man nature might describe the biological factors influencing human moti-
vation, but it could not prescribe norms of proper human conduct without
invoking moral standards that transcend the facts of human biology. For
example, it would be proper for biologists to investigate the biological psy-
chology of human sexuality, but for them to infer from the biological facts
of human sexual motivation that some kinds of sexual conduct were
morally better or worse than others would be fallacious. So, for instance,
 Incest Taboo as Darwinian Natural Right 207

biological scientists cannot—based on their scientific knowledge alone—

judge whether incest is right or wrong.
The second common objection to Darwinian ethical naturalism is that it
fails to recognize the primary role of reason in discovering objective princi-
ples for moral judgment. Darwinian ethics explains ethics as rooted in nat-
ural desires or emotions. But the critics argue that ethics requires the con-
trol of the desires or emotions by reason. The uniqueness of human ethics
seems to stem from the fact that human beings are the only animals who
can use reason to formulate and enforce objective rules of conduct that
transcend the subjective desires that govern the behavior of other animals.
In the attempt to ground ethics in animal emotions or desires, Darwinian
ethical theory cannot provide any rational foundation for ethics. Darwin-
ian theorists explain human natural propensities to altruistic behavior, but
this does not explain the distinctly moral character of human social behav-
ior as guided by rational principles of obligation.
The third objection is that the great variability in human social behavior
in different societies and in different historical periods denies the universal
human nature postulated by Darwinian ethics. The incest taboo seems to il-
lustrate this very well. Just within the Anglo-American world, there is great
variation in the customary, religious, and legal rules of incest.79 Until the six-
teenth century, the canon law of the Catholic Church as applied to England
prohibited marriage to a long list of both consanguineous and affinal rela-
tives. Then, beginning in the Reformation, the list of prohibited relatives
was reduced drastically. In the twentieth century, the list was reduced even
further by eliminating most affinal relatives. In the United States, there has
been similar variation. For example, early in American history, marriage of
cousins was permitted (as it was in England). But beginning in the middle of
the nineteenth century, many states changed their laws to prohibit cousin
marriages. Now, however, the majority of the states allow cousin marriages.
If the incest taboo is governed by a biologically natural propensity, as West-
ermarck claims, then how could one explain such variation?
The fourth objection is that Darwinian reasoning about ethics assumes a
biological determinism that cannot account for the importance of human
culture and social learning in shaping ethical conduct in ways that transcend
biological nature. Ethical rules are not genetically determined in human be-
ings. They arise from the habits and customs of human society. They are not
naturally innate but socially acquired. Consequently, the cultural diversity of
human beings creates a great diversity in moral rules that cannot be ex-
plained by reference to the biological universals of human nature.
In response to the first objection, I would accept Hume’s “empiricist”
version of the fact-value dichotomy but not Kant’s “transcendentalist” ver-
sion. The Westermarckian ethical theorist would rightly agree with the
208 arnhart

Humean understanding of the fact-value dichotomy if that means that

moral judgments cannot be purely logical deductions from information
about the world. But the Westermarckian theorist would disagree with the
Kantian separation between the “phenomenal” realm of natural causality
and the “noumenal” realm of moral freedom. The critics of Darwinian
ethics often assume a Kantian notion of the “moral ought” or “the moral
point of view” as utterly transcending the observable world of nature, but
they rarely defend the metaphysical dualism implicit in this assumption, be-
cause it is hard to make such a radical dualism plausible.80
Hume distinguishes is and ought in order to show that moral assess-
ments are derived not from pure reason alone but from moral emotions.
Yet far from denying that moral judgments are judgments of fact, Hume
rightly claims that moral judgments are accurate when they correctly re-
port what our moral judgments would be in a given set of circumstances.
Correct moral judgments are factual judgments about the species-typical
pattern of moral sentiments in specified circumstances.
Hume compares moral judgments to judgments of secondary qualities
such as color.81 My judgment that this tomato is red is true if the object is
so constituted as to induce the impression of red in normally sighted hu-
man beings viewing it under standard conditions. Similarly, my judgment
that this person is morally praiseworthy is true if the person’s conduct is
such as to induce the sentiment of approbation in normal human beings
under standard conditions. Just as an object can appear red to me when in
fact it is not, so a person can appear praiseworthy to me when in fact he is
not. The moral judgment whether some conduct would give to a normal
spectator under standard conditions a moral sentiment of approbation is,
Hume insists, “a plain matter of fact.”82 The moral sentiment itself, how-
ever, is a feeling or passion rooted in human nature that cannot be pro-
duced by reason alone. As Robert McShea has argued, moral judgments
based on the facts of the moral emotions are not fallacious as long as they
claim only that “for a particular intelligent species certain feelings are pre-
dictably aroused by certain facts and the experience of such feelings is the
only basis on which we can make evaluative judgments.”83
By contrast, Kant’s version of the fact-value dichotomy separates reality
into two metaphysical realms. Judging what is the case belongs to the “phe-
nomenal” realm of nature, but judging what ought to be the case belongs to
the “noumenal” realm of freedom. As moral agents, we obey categorical im-
peratives of what ought to be that are beyond the causal laws of nature.84
“When we have the course of nature alone in view,” Kant explains, “‘ought’
has no meaning whatsoever.”85
Kant’s separation of is and ought treats morality as an autonomous
realm of human experience governed by its own internal logic with no ref-
 Incest Taboo as Darwinian Natural Right 209

erence to anything in human nature such as natural desires or interests.

Contemporary philosophers follow Kant when they speak of “the moral
point of view” as an autonomous realm of thinking that transcends factual
reasoning about human nature. Kant does this because he accepts the so-
phistic and Hobbesian view of human nature. Since human beings are by
nature selfish, asocial animals, they cannot live together in peace unless
they conquer their natural inclinations by willing submission to moral rules
devised by reason to pacify their selfish conflicts.86
This contrast between Humean empiricism and Kantian transcendental-
ism arises in the debate between Westermarck and Freud over the incest ta-
boo. If the evidence supports Westermarck’s theory of the incest taboo as
opposed to Freud’s, that would suggest that a moral rule like the incest ta-
boo can be rooted in the natural propensities of human nature, and this
would also suggest that the logical distinction between natural facts and
moral values is not an absolute separation. Logical reasoning about the
facts of sexual mating cannot by itself dictate a value judgment that incest
is wrong, and that is what Hume means when he distinguishes is and ought.
But if it belongs to the factual nature of most human beings that they have
a natural propensity to acquire a feeling that incest is wrong, and if the gen-
eralization of that feeling across society supports a social prohibition of in-
cest, then the value judgment that incest is wrong is indeed rooted in the
facts of human nature.
The appeal to human nature in condemning social practices that violate
the biological propensities for learning incest avoidance is evident in Arthur
Wolf’s defense of Westermarck’s theory. Wolf says that the Chinese practice
of “minor” marriages inflicted “wounds” on the people forced to marry
against their will, because it was “contrary to their natural inclinations.”
This social practice was “abnormal,” “maladaptive,” “painful,” “cruel,”
“oppressive,” and “unnatural.” Wolf insists: “It will be hard to find a more
striking example of the fact that while culture can overwhelm natural ten-
dencies, it is not capable of protecting the individuals involved from the
suffering entailed.”87 Wolf acknowledges that a few individuals do not
show the “Westermarck effect”: some people who have been raised to-
gether in the same family from infancy develop no inhibition to sexual re-
lations with one another. But he explains these exceptional individuals as
being similar to psychopaths who suffer from an abnormal poverty of emo-
tions so that they do not feel, or do not feel very deeply, the moral emo-
tions typical for most human beings.88 When Wolf thus appeals to the “nat-
ural inclinations” of emotionally “normal” people as setting standards for
judging social practices as good or bad, he is implicitly appealing to the
kind of natural law reasoning developed by Aquinas and others in the So-
cratic tradition of ethical naturalism.
210 arnhart

When Wolf pronounces his moral condemnation of the practice of “mi-

nor” marriages in China, he assumes that most of his readers will agree
with him. His masterful survey of the evidence and arguments for Wester-
marck’s theory of incest avoidance offers inductive and deductive reasoning
that will help to win his readers’ agreement. But he cannot persuade them
to accept his moral judgment purely through logical inference. He must as-
sume that most of his readers have the typical emotional profile that will
incline them to feel sympathy for the suffering of people forced into mar-
riages made unhappy by their natural sexual aversion. Pure reason by itself
could not elicit a moral judgment against “minor” marriages if such moral
emotions were not brought into play. Readers who lack such emotions
would not be moved to accept Wolf’s moral judgment. Thus the moral ap-
peal of Wolf’s book—as combining logical reasoning with moral emotions
—illustrates Westermarck’s general claim about the primacy of natural
emotions in moral experience.
This also suggests the proper response to the second objection to Dar-
winian ethical naturalism concerning the importance of reason in morality.
As I indicated earlier, Westermarck—like others in the Socratic tradition—
recognized the importance of reason in eliciting, directing, and generalizing
the moral emotions. But he rightly denied that reason by itself could moti-
vate ethical judgment. The incest taboo illustrates this very well. As long as
the natural emotional aversion to incest creates the feeling that incest is
wrong, reason can elicit, direct, and generalize that feeling in ways that will
determine the customary and legal expression of the incest taboo. For ex-
ample, human reason will work out elaborate systems of kinship classifica-
tions that will determine the details of whom one can and cannot marry. But
if there were no original feeling in most human beings that incest is wrong,
reason by itself by purely logical means could not judge that incest is wrong.
As indicated by the third objection, the variability in ethical rules such
as the incest taboo is interpreted by many critics of Darwinian ethics as ev-
idence that there are no universal ethical propensities rooted in human bio-
logical nature. But in fact, Westermarck’s Darwinian account of the incest
taboo illustrates how a Darwinian ethics can allow for both variability and
universality.
Westermarck suggests that the incest taboo can expand beyond the nu-
clear family to embrace a wider circle of prohibited marriage partners or
contract to cover a smaller circle, and this expansion or contraction will
depend on the circumstances of familial living arrangements and kinship
systems.89 Originally, among the earliest human ancestors, Westermarck in-
fers, the family consisted of parents and their children. The harmful conse-
quences of inbreeding among such close relatives were such that natural se-
lection would have shaped the sexual instinct so that people would tend to
 Incest Taboo as Darwinian Natural Right 211

feel a sexual aversion toward those close relatives with whom they had
lived from childhood. But once this natural propensity was acquired, it
could be expressed as a sexual aversion toward remote relatives or even un-
related people who lived together in familial groups. Finally, this natural
propensity could then lead “through an association of ideas and feelings”
to the prohibition of sexual intercourse and marriage between people de-
fined as kin who did not live together at all. Westermarck thus recognizes
that the incest taboo varies greatly across societies and across time because
of the variation in family life and kinship classification. But the variation
shows an underlying regularity that manifests a natural biological ten-
dency. For despite the variation in the incest rules for those outside the nu-
clear family of parents and children, the rule prohibiting sexual mating be-
tween parent and child and between siblings is almost universal. The rare
cases of where the marriage of parents and children or of siblings is per-
mitted are unusual exceptions that only highlight the universal rule. Any
survey of the anthropological evidence will show that societies generally
prohibit marriage between those people who typically live in the same
household, which typically includes parents and children.
As rightly indicated by the fourth objection to Darwinian ethics, human
reason gives us a flexibility in our capacity for symbolism and social learn-
ing that cannot be strictly determined by our genetic nature. But just as the
Greek sophists invoked a false antinomy of nature versus convention, so do
critics of Darwinian ethics such as Stephen Jay Gould insist on a false antin-
omy—biological potentiality versus biological determinism—that ignores
the importance of biological propensity. Gould explains that his criticism of
Edward O. Wilson “does not invoke a nonbiological ‘environmentalism’: it
merely pits the concept of biological potentiality—a brain capable of the full
range of human behaviors and rigidly predisposed toward none—against
the idea of biological determinism—specific genes for specific behavior.”90
Explaining human social behavior through biological determinism is ut-
terly implausible if that means claiming that specific genes determine spe-
cific behavior with no flexibility. Therefore, Gould asserts, we must accept
the only alternative idea—“biological potentiality—a brain capable of the
full range of human behaviors and rigidly predisposed toward none.” But
like the sophistical rhetoricians, Gould uses the word “rigidly” here to ob-
scure a third alternative: the idea of biological predisposition or propensity
as something more than a mere potentiality and yet something less than a
rigid determinism.
With respect to sexual mating, for example, human beings have a biolog-
ical potentiality for a wide range of behaviors—including celibacy, pro-
miscuity, monogamy, polygyny, and polyandry. Gould would have a strong
argument in claiming that there are no specific genes that absolutely deter-
212 arnhart

mine or “rigidly” predispose us to one of these behaviors. But he would be

wrong to assume from this that we have an indifferent potentiality for any
of these. Although we have a potential for choosing complete celibacy, most
human beings find this too difficult because it denies our strong propensity
or desire for sexual mating. Promiscuity is easier because it caters to our
sexual propensities. Polyandrous marriage seems to be a very weak poten-
tiality for human beings because the intense sexual jealousy of males inclines
them against it except in unusual circumstances. In contrast to polyandry,
monogamous mating has been universal to all human societies, and polygy-
nous mating has been common, because they satisfy biological desires. An
understanding of biological propensities can explain why celibacy is diffi-
cult, promiscuity is easy, polyandry is rare, monogamy is universal, and
polygyny is common, although none of these behaviors is “rigidly” deter-
mined by specific genes. As Wilson would say, this pattern of human behav-
ior manifests the “epigenetic rules” of gene-culture coevolution. Our nature
predisposes us to favor some mating behavior over others, although the spe-
cific expression of our mating behavior will reflect the variable conditions of
physical environment, social circumstances, and individual temperament.
Similarly, the incest taboo manifests a natural propensity of human bio-
logical nature, although this taboo is not “rigidly” determined by specific
genes. Most human beings are naturally inclined to learn a sexual aversion
to those with whom they have lived in early childhood. This natural aver-
sion will then tend to favor a universal taboo against incestuous relations
between parents and children or between siblings. But the learning of this
aversion will vary greatly in response to early childhood experiences. It will
also vary in that a few individuals will not learn this aversion as easily as
most others. Moreover, as we have seen, the extension of the taboo beyond
the nuclear family is highly variable in response to varying kinship systems.
Westermarck and Wilson emphasize this variability, and therefore they can-
not properly be accused of a rigid biological determinism. But they also see
a regularity in the incest taboo that reflects natural propensities. To talk as
Gould does of an indifferent “potentiality” would acknowledge the cul-
tural diversity in the incest taboo, but without explaining the underlying
regularity in the taboo that Westermarck and Wilson explain as a propen-
sity of human nature.

Conclusion

Westermarck’s theory of the incest taboo explains that moral rule as

a social convention that expresses the emotional propensities of human na-
ture as shaped by natural selection in evolutionary history. That theory il-
lustrates how a Darwinian science of human nature can support a natural-
 Incest Taboo as Darwinian Natural Right 213

istic understanding of ethics as governed by natural causal laws. This pro-

vides a modern biological foundation for a tradition of ethical naturalism
that begins with the Socratic philosophers of ancient Greece and extends to
Thomas Aquinas in the Middle Ages and then to David Hume and Adam
Smith in the early modern period. E. O. Wilson’s view of ethics as rooted in
the natural moral sentiments continues that tradition.
The fundamental alternative to such ethical naturalism is the ethical con-
ventionalism of the sophistic tradition that stretches from Callicles to
Hobbes and then to Freud and finally to contemporary theorists like Gould.
According to this tradition, ethics requires a transcendence of the natural
world in which we enter an autonomous realm of ethical conventions cre-
ated by human reason. According to this view, there can be no natural sci-
ence of ethics.
The scientific evidence and arguments for Westermarck’s theory of the
incest taboo show, by contrast, how the study of ethics could become an
empirical science. This also suggests the possibility of a grand unification of
all scientific knowledge in which ethics would become part of a science of
human nature embedded within a general science of nature as a whole. A
Darwinian science of ethics as natural right would thus allow us to under-
stand our human nature as part of the natural cosmos.

notes
1. Edward Westermarck, Ethical Relativity (London: Kegan Paul, Trench,
Trubner & Company, 1932), pp. 246–50, 288–89.
2. Edward O. Wilson, Sociobiology: The New Synthesis (Cambridge: Harvard
University Press, 1975), p. 1.
3. Edward O. Wilson, Consilience: The Unity of Knowledge (New York:
Alfred A. Knopf, 1998).
4. See Ullica Segerstrale, Defenders of the Truth: The Battle for Science in the
Sociobiology Debate and Beyond (Oxford: Oxford University Press, 2000).
5. Wilson, Consilience, pp. 238–56.
6. Edward Wilson, On Human Nature (Cambridge: Harvard University Press,
1978), pp. 37–39, 68–69, 229; Charles J. Lumsden and Edward O. Wilson, Genes,
Mind, and Culture: The Coevolutionary Process (Cambridge: Harvard University
Press, 1981), pp. 37, 71, 85–86, 147–58, 238, 357; Charles J. Lumsden and Ed-
ward O. Wilson, Promethean Fire: Reflections on the Origin of Mind (Cambridge:
Harvard University Press, 1983), pp. 64–65, 115, 119, 124–27, 133–38, 175–80;
Michael Ruse and Edward O. Wilson, “Moral Philosophy as Applied Science,”
Philosophy, vol. 61 (Summer 1986): pp. 173–92; Wilson, Consilience, pp. 173–80.
7. Wilson, Consilience, p. 164.
8. Ibid., pp. 238–39.
9. Ibid., pp. 248–49.
214 arnhart

10. Thomas Nagel, “Ethics as an Autonomous Theoretical Subject,” in Gunther

S. Stent, ed., Morality as a Biological Phenomenon (Berkeley: University of Califor-
nia Press, 1978), pp. 196–205.
11. David Buss, The Evolution of Desire: Strategies of Human Mating (New
York: Basic Books, 1994), pp. 16–17.
12. Steven Pinker, How the Mind Works (New York: Norton, 1997), pp. 559,
561.
13. See Segerstrale, Defenders of the Truth, pp. 362–64.
14. Larry Arnhart, Aristotle on Political Reasoning: A Commentary on the
“Rhetoric” (DeKalb: Northern Illinois University Press, 1981), pp. 118–34.
15. See Larry Arnhart, Darwinian Natural Right: The Biological Ethics of
Human Nature (Albany: State University of New York Press, 1998).
16. Aristotle, Sophistical Refutations, 173a.
17. The sophistic tradition in ancient Greece is complex. For a good survey
of its intellectual complexity, see G. B. Kerferd, The Sophistic Movement (Cam-
bridge: Cambridge University Press, 1981). For the argument that modern ideas
of cultural relativism and social constructionism continue the sophistic tradition,
see Mark Backman, Sophistication: Rhetoric and the Rise of Self-Consciousness
(Woodbridge, Conn.: Ox Bow Press, 1991).
18. Plato, Laws, trans. Thomas Pangle (New York: Basic Books, 1980),
838a–39b.
19. Xenophon, Memorabilia, IV.iv.19–23.
20. Stephen Salkever, Finding the Mean: Theory and Practice in Aristotelian
Political Philosophy (Princeton, N.J.: Princeton University Press, 1990), p. 115.
21. Aristotle, Nicomachean Ethics, 1134b18–35a5; Thomas Aquinas,
Commentary on Aristotle’s “Nicomachean Ethics”, lec. 12.
22. Wilson, Consilience, p. 239.
23. Thomas Aquinas, Summa Theologica, I–II, q. 94, a. 2.
24. Aquinas, Summa Theologica, Suppl., q. 41, a. 1; q. 65, a. 1.
25. Aquinas, Summa Contra Gentiles, bk. 3, chap. 123.
26. Aquinas, Summa Contra Gentiles, bk. 3, chaps. 122–23.
27. Aquinas, Summa Theologica, II–II, q. 154, a. 9; Suppl., q. 54, a. 10;
Summa Contra Gentiles, bk. 3, chap. 125.
28. William Durham, Coevolution: Genes, Culture, and Human Diversity
(Stanford, Calif.: Stanford University Press, 1991), p. 331.
29. Aristotle, History of Animals, 630b31–31a7.
30. For surveys of the theories to explain the incest taboo, see Sybil Wolfram,
In-Laws and Outlaws: Kinship and Marriage in England (London: Croom Helm,
1987), pp. 161–85; Carol R. Ember and Melvin Ember, Anthropology, 7th ed.
(Englewood Cliffs, N.J.: Prentice-Hall, 1993), pp. 316–19; Durham, Coevolution,
pp. 286–360; and Arthur Wolf, Sexual Attraction and Childhood Association:
A Chinese Brief for Edward Westermarck (Stanford, Calif.: Stanford University
Press, 1995), pp. 1–19, 498–515.
31. Thomas Hobbes, The Leviathan, chaps. 14–15.
32. Thomas Hobbes, De Cive, chap. 1.
33. Hobbes, Leviathan, chap. 17; Hobbes, De Cive, chap. 5, par. 5; Hobbes,
De Homine, chap. 10.
 Incest Taboo as Darwinian Natural Right 215

34. On Hobbes’s dualism, see Leo Strauss, The Political Philosophy of

Hobbes, trans. Elsa M. Sinclair (Chicago: University of Chicago Press, 1952),
pp. 7–9, 168–70.
35. Immanuel Kant, “Speculative Beginning of Human History,” in “Perpetual
Peace” and Other Essays, trans. Ted Humphrey (Indianapolis: Hackett Publishing
Company, 1983); Immanuel Kant, Critique of Judgment, trans. Werner S. Pluhar
(Indianapolis: Hackett Publishing Company, 1987), secs. 83–84.
36. See Adam Kuper, Culture: The Anthropologists’ Account (Cambridge:
Harvard University Press, 1999).
37. See Robert A. Greene, “Instinct of Nature: Natural Law, Synderesis,
and the Moral Sense,” Journal of the History of Ideas 58 (April 1997):
pp. 173–94.
38. Bernard Mandeville, The Fable of the Bees, 2 vols. (Oxford: Oxford
University Press, 1924), vol. 1: pp. 330–31.
39. Francis Hutcheson, An Inquiry into the Original of Our Ideas of Beauty
and Virtue (London: J. Darby, 1725), p. 192.
40. Adam Smith, The Theory of Moral Sentiments, ed. D. D. Raphael and
A. L. Macfie (Oxford: Oxford University Press, 1976), pp. 300–304, 321–27;
Ian Simpson Ross, The Life of Adam Smith (Oxford: Oxford University Press,
1995), pp. 48–59, 159–64.
41. Smith, Theory of Moral Sentiments, pp. 163–66, 446–47.
42. See Wolf, Sexual Attraction and Childhood Association, pp. 463–75; and
Mark T. Erickson, “Rethinking Oedipus: An Evolutionary Perspective of Incest
Avoidance,” American Journal of Psychiatry, vol. 150 (1993): pp. 411–16.
43. See Wolfram, In-Laws and Outlaws, pp. 30–40.
44. Charles Darwin, The Descent of Man, 2 vols. (London: John Murray,
1871), vol. 1: pp. 70–106, vol. 2: pp. 390–94.
45. Darwin, Descent, 1:70–71; Immanuel Kant, The Critique of Practical
Reason, trans. Lewis White Beck (Indianapolis: Bobbs-Merrill, 1956), p. 90.
46. Charles Darwin, The Origin of Species and The Descent of Man (New
York: Random House, Modern Library, 1936), pp. 485–86, 896.
47. Charles Darwin, The Variation of Animals and Plants Under Domestica-
tion, 2 vols. (New York: Appleton, 1897), vol. 2: p. 104.
48. Edward Westermarck, The History of Human Marriage, 3 vols., 5th ed.
(New York: Allerton Book Company, 1922); Edward Westermarck, The Origin
and Development of the Moral Ideas, 2 vols. (London: Macmillan Company,
1906); Westermarck, Ethical Relativity.
49. For an example of how a Westermarckian anthropological naturalism
can support Thomistic natural law, see Thomas Davitt, The Basic Values in Law
(Milwaukee, Wis.: Marquette University Press, 1978).
50. Westermarck, Marriage, vol. 1: pp. 22, 38, 53, 71, vol. 3: p. 365.
51. Westermarck, Marriage, vol. 3: pp. 104, 107, 206, 221–22; Westermarck,
Moral Ideas, vol. 2: pp. 387–92.
52. For a study of Tibetan fraternal polyandry that largely confirms Wester-
marck, see Durham, Coevolution, pp. 42–102.
53. Westermarck, Marriage, vol. 2: pp. 82–239; Westermarck, Moral Ideas,
vol. 2: pp. 364–81.
216 arnhart

54. Westermarck, Ethical Relativity, pp. 248–49; Westermarck, Marriage,

vol. 2: pp. 192, 197, 214.
55. For this formulation, see Wolf, Sexual Attraction and Childhood Associa-
tion, p. 506.
56. Sigmund Freud, Totem and Taboo, in The Standard Edition of the
Complete Psychological Works of Sigmund Freud, vol. 13, trans. James Strachey
(London: Hogarth Press, 1955).
57. Sigmund Freud, Civilization and Its Discontents, trans. James Strachey
(New York: Norton, 1961), pp. 42, 58, 62, 69, 87.
58. Ibid., p. 51.
59. Ibid., pp. 89–91.
60. Ilham Dilman, Freud and Human Nature (Oxford: Basil Blackwell, 1983).
61. Claude Lévi-Strauss, The Elementary Structures of Kinship (Boston:
Beacon Press, 1969), pp. 30–31.
62. See Richard Lewontin, Steven Rose, and Leon Kamin, Not in Our Genes:
Biology, Ideology, and Human Nature (New York: Pantheon, 1984), p. 137; and
Philip Kitcher, Vaulting Ambition: Sociobiology and the Quest for Human Nature
(Cambridge: MIT Press, 1985), pp. 280, 348.
63. Westermarck, Marriage, vol. 2: pp. 82, 88, 192, 201–3.
64. See Joseph Shepher, Incest: A Biosocial View (New York: Academic Press,
1983); and Wolf, Sexual Attraction and Childhood Association, pp. 435–38.
65. Westermarck, Marriage, vol. 2: pp. 223–24.
66. See Wolf, Sexual Attraction and Childhood Association, pp. 388–422;
Anne Pusey, “Inbreeding Avoidance in Chimpanzees,” Animal Behavior, vol. 28
(1980): pp. 543–52; Jane Goodall, The Chimpanzees of Gombe (Cambridge:
Harvard University Press, 1986), pp. 466–71; Anne Pusey and Marisa Wolf,
“Inbreeding Avoidance in Animals,” Trends in Ecology and Evolution, vol. 11
(May 1996): pp. 201–6.
67. From her observations of wild chimpanzees, Jane Goodall reports, “Copu-
lations between fathers and daughters and between paternal siblings are unlikely
to be inhibited, for the individuals concerned do not ‘know’ their relationship.
There is no close bonding between them, and they do not achieve the high level
of familiarity that presumably underlies incest avoidance between mothers and
sons and maternal siblings” (Chimpanzees of Gombe, 469).
68. See Anne Pusey, Chapter 3 in this volume.
69. Westermarck, Marriage, vol. 2: pp. 224–36.
70. See Durham, Coevolution, pp. 293–309.
71. See Patrick Bateson, “Optimal Outbreeding,” in Mate Choice, ed. Patrick
Bateson (Cambridge: Cambridge University Press, 1983), pp. 257–77.
72. Westermarck, Marriage, vol. 2: p. 182.
73. Smith, Theory of Moral Sentiments, pp. 179–93, 326.
74. The compatibility of Freud and Westermarck is argued by Robin Fox in
The Red Lamp of Incest (Notre Dame, Ind.: University of Notre Dame Press, 1983).
75. Thomas Nagel, “Ethics as an Autonomous Theoretical Subject,” in Gunther
Stent, ed., Morality as a Biological Phenomenon: The Presuppositions of Socio-
biological Research (Berkeley: University of California Press, 1980), p. 200.
 Incest Taboo as Darwinian Natural Right 217

76. John Stuart Mill, “Nature,” in Collected Works of John Stuart Mill, ed.
J. M. Robson (Toronto: University of Toronto Press, 1969), vol. 10: pp. 373–402.
77. Paul Lawrence Farber, The Temptations of Evolutionary Ethics (Berkeley:
University of California Press, 1998); Peter G. Woolcock, “The Case Against
Evolutionary Ethics Today,” in Jane Maienschein and Michael Ruse, eds., Biology
and the Foundation of Ethics (Cambridge: Cambridge University Press, 1999),
pp. 276–306.
78. G. E. Moore, Principia Ethica (Cambridge: Cambridge University Press,
1903).
79. See Jack Goody, The Development of the Family and Marriage in Europe
(Cambridge: Cambridge University Press, 1983); Wolfram, In-Laws and Outlaws;
Carolyn S. Bratt, “Incest Statutes and the Fundamental Right to Marriage: Is
Oedipus Free to Marry?” Family Law Quarterly, vol. 18 (1984): pp. 257–309; and
Martin Ottenheimer, “Lewis Henry Morgan and the Prohibition of Cousin Mar-
riage in the United States,” Journal of Family History, vol. 15 (1990): pp. 325–34.
For a vigorous argument for allowing cousin marriage in the United States, see
Martin Ottenheimer, Forbidden Relatives: The American Myth of Cousin Mar-
riage (Urbana: University of Illinois Press, 1996).
80. See, for example, Nagel, “Ethics as an Autonomous Theoretical Subject,”
and Woolcock, “Case Against Evolutionary Ethics Today.”
81. David Hume, A Treatise of Human Nature, ed. L. A. Selby-Bigge (Oxford:
Oxford University Press, 1888), p. 469; Hume, Essays, ed. Eugene F. Miller (Indi-
anapolis: Liberty Classics, 1985), pp. 233–34.
82. David Hume, Enquiries Concerning the Human Understanding and
Concerning the Principles of Morals (Oxford: Clarendon Press, 1902), p. 289.
83. Robert McShea, Morality and Human Nature (Philadelphia: Temple
University Press, 1990), p. 226.
84. Immanuel Kant, Critique of Practical Reason, trans. Lewis White Beck
(Indianapolis: Bobbs-Merrill, 1956), pp. 4–5, 18, 30–31, 99, 163–64; Kant, Foun-
dations of the Metaphysics of Morals, trans. Lewis White Beck (Indianapolis:
Bobbs-Merrill, 1959), pp. 4, 30, 44–45, 67–74, 80; Kant, Critique of Pure Reason,
trans. Norman Kemp Smith (New York: St. Martin’s Press, 1965), pp. 465,
472–79, 526; Kant, Critique of Judgment, pp. 286–87.
85. Kant, Critique of Pure Reason, p. 473.
86. Immanuel Kant, “Perpetual Peace” and Other Essays, pp. 87, 110–13,
115–17, 124–25; Kant, Critique of Judgment, pp. 317–21.
87. Wolf, Sexual Attraction and Childhood Association, pp. ix, 180, 217,
277–79, 328, 422.
88. See Wolf, Sexual Attraction and Childhood Association, pp. 259–63. For
an account of psychopaths as people whose emotional poverty makes them
“moral strangers,” see Arnhart, Darwinian Natural Right, pp. 211–30.
89. Westermarck, Marriage, vol. 2: pp. 206–7, 214–18, 236–37.
90. Stephen Jay Gould, Ever Since Darwin: Reflections in Natural History
(New York: Norton, 1977), pp. 257–58.
This page intentionally left blank
Contributors

l a r ry a r n h a r t is Professor of Political Science at Northern Illinois

University. He is the author of Aristotle on Political Reasoning: A
Commentary on the “Rhetoric”; Political Questions: Political Philosophy
from Plato to Rawls; and Darwinian Natural Right: The Biological Ethics
of Human Nature. He is the associate editor of The Encyclopedia of
Science, Technology, and Ethics.
pat r i c k bat e s o n is Professor of Ethology at the University of
Cambridge. He was Provost of King’s College, Cambridge, for fifteen
years and Biological Secretary of the Royal Society of London for five
years. He edited, among other books, Mate Choice and is coauthor with
Paul Martin of Design for a Life: How Behavior and Personality Develop.
a l a n h . b i t t l e s is Foundation Professor of Human Biology and
Director of the Centre for Human Genetics at Edith Cowan University.
During the last twenty-five years he has organized and conducted major
studies into the prevalence and effects of consanguineous marriage,
principally in South and Southeast Asia and North Africa. He has been
editor of Annals of Human Biology since 1998.
w i l l i a m h . du r h a m is Professor and Chair of Anthropological
Sciences, Bing Professor in Human Biology, and Yang and Yamazaki
University Fellow at Stanford University. He is author of Scarcity and
Survival in Central America and Coevolution: Genes, Culture and Human
Diversity and, since 1992, editor of the Annual Review of Anthropology.
m a r k t. e r i c k s o n is Assistant Clinical Professor, University of
Washington School of Medicine. He works and teaches at the Alaska
Psychiatric Institute in Anchorage.
h i l l g at e s is Professor Emerita at Central Michigan University, has
taught at Johns Hopkins and Stanford, and is now an independent
scholar of Taiwan /China political economy and gender. Her principal
publications include Chinese Working Class Lives: Getting By in Taiwan,
China’s Motor: A Thousand Years of Petty Capitalism, and Looking for
Chengdu: A Woman’s Adventures in China.
220 Contributors

a n n e p u s e y is Professor of Ecology, Evolution, and Behavior and

a Distinguished McKnight University Professor at the University of
Minnesota. She was appointed Executive Director of the Jane Goodall
Institute’s Primate Research Programs in 2003 and has written over
fifty scientific papers.
wa lt e r s c h e i d e l is Professor of Classics at Stanford University.
He is author of Measuring Sex, Age, and Death in the Roman Empire
and Death on the Nile: Disease and the Demography of Roman Egypt,
editor of Debating Roman Demography, and coeditor of several volumes
on the ancient economy and ancient empires.
n e v e n s e sa r d i c is Associate Professor in the Department of
Philosophy at Lingnan University, Hong Kong. He has published papers
in journals like Philosophy of Science, British Journal for the Philosophy
of Science, Biology, and Philosophy, and Ethics. He is currently finishing
a book, Making Sense of Heritability, which will be published by
Cambridge University Press.
a r t h u r p. wo l f is Professor of Anthropological Sciences and David
and Lucile Packard Foundation Professor in Human Biology at Stanford
University. He is editor of Religion and Ritual in Chinese Society and
Family and Population in East Asia; coauthor with Chieh-shan Huang
of Marriage and Adoption in Chinese Society, 1845–1945; and author
of Sexual Attraction and Childhood Association: A Chinese Brief for
Edward Westermarck.
Index

Aaron, 39 Barbary macaques, 63, 64, 66, 73

Aberle, David F., 1, 3, 4, 5, 6, 7, 8, 9, 10, Bateson, Patrick, 1, 14–15, 24, 36–37,
12, 15, 19, 21, 22 82, 91, 149, 156, 158, 172, 177,
Access to Birth Records Act, 169 186–188, 216, 219
Adaptation, 25–28, 36, 61, 111–112, Bemiss, Samuel, 45, 58
114, 155, 162, 168, 170–173, 186 Bevc, Irene, 99, 107
Adaptation hypothesis, 111, 114, Biological evolution, 35, 121
122–123 Biological inhibition, 112, 120
Adriani, Nicolaus, 131, 132, 138 Biological kin, 38–39, 44, 104, 169
Adult Attachment Inventory, 174, 176 Biologically mediated inhibition, 117
Adultery, 94, 149, 152, 159, 163, 203, Biological reductionism, 109, 191
207 Birth Records Act, 169
Africa, 19, 38, 54, 219 Bittles, Alan H., 3, 18–19, 38, 55–58,
Alberts, Susan, 64–67, 73–75 60–61, 78, 106, 156, 219
Altruism, 53, 172–173 Bonobo chimpanzees, 6, 173, 187, 188
Ainsworth Strange Situation, 174 The Book Bag, 170
Amran, 39 Botswana, 179
Anabaptists, 44 Bowlby, John, 14, 89–90, 92, 172, 187
Andhra Pradesh, 41 Brahmins, 41
Angels and Insects, 47, 58 Breastfeeding, 89, 102–105, 178–179.
Anglophone anthropologists, 140 See also Kin recognition, relation to
Aquinas, Thomas, 193, 195–202, 209, olfaction
213–214 Briffault, Robert, 2, 21
Arabs, 43, 56, 127 Britain, 22, 41, 164, 165, 169, 183, 220
Ariki Kafika, 154 British law, 198
Arnhart, Larry, 17, 18, 21, 156, 214, British structural-functionalists, 142–143
217, 219 Bronfenbrenner, Uri, 1, 21
Arsinoe, 47, 93, 94, 106 Brother-sister marriage, 2, 9, 47, 93–105,
Aristotle, 191–194, 196–197, 202, 214, 146, 152–54. See also Consanguineous
219–220 marriage
Asexual imprinting, 5–6, 31, 33, 46. Buddhism, 40
See also Westermarck hypothesis, Burton, Roger, 61, 128–133, 135–136,
substance of 138, 220
Asia, 19, 38, 53–54, 91, 159, 219–220 Buss, David, 17, 192, 214
Attachment, 14, 18–19, 20, 61, 70–71, Byatt, A. S., 47, 58
89–90, 145, 151, 172–177 Byron, Lord, 170, 185
Augustine, St., 40
Aversion hypothesis, 117, 118, 122, 127, Cairo, 93
206. See also Westermarck hypothesis, California sea lions, 172, 187
substance of Callicles, 17, 193, 202, 213
Avesta, 48 Calusa, 153, 159
Canada geese, 4
Baboons, 6, 62–67, 71–75 Capuchin monkeys, 62
Baker, Anthony, 164–165, 183–184 Carter, C. O., 49, 50, 59
222 Index

Catholic church, 207 mating frequency of, 42; in South

Cement Garden, 47, 58 India, 41, 43, 46; among Tuareg, 43;
Center for Advanced Study in the Behav- in U.S. states, 42–43 (fig), 54; used
ioral Sciences, 1 to estimate danger of inbreeding, 3,
Central Asia, 38 52–53; used to test Westermarck hy-
Cesarean section, 178 pothesis, 125–126; in Wisconsin, 42.
Childhood association: and adultery in See also Consanguineous marriage
minor marriages, 77, 94; age at as de- Crawford, Mary, 86
terminant of sexual attraction, 79–82 Crime and Custom in Savage Society,
(tabs), 99; and birth intervals in Ro- 148, 156
man Egypt, 97; and divorce in Tai- Critical period, 32. See also Sensitive
wanese minor marriages, 77–82 (tabs), period
94; and fertility in Taiwanese minor Cross-fostering, 101–105
marriages, 77–82 (tabs), 94; and likeli- Cultural prohibitions, 112, 115, 120, 126
hood of incest, 169–172; meaning of Culture, 141, 197. See also Incest prohi-
in Westermarck hypothesis, 78–82; and bitions, culture and
scope of incest prohibitions, 124–147; Czechoslovakia, 49–50
and sexual attraction (see Westermarck
hypothesis, evidence for); and sexual Dahlberg, George, 3, 22
relations among chimpanzees, 68–71; Darwin, Charles, 2, 18, 21, 24, 27, 28,
and sexual segregation, 32, 88–89. 40, 41, 202–204
See also Westermarck hypothesis, Darwin, G. H., 25, 27, 29, 34–35, 41–42,
substance of 56, 105, 109–110, 121, 142, 186, 190–
Chimpanzees, viii, 6, 62–70, 73–74, 173, 203, 205–215, 217, 219
176, 204, 216, 220 Dawkins, Richard, 111, 120
China, 27, 40, 55, 76, 88, 91, 160, 203, Demarest, William, 174, 187
210, 219 Dilman, Ilham, 202, 216
Christian Cathar community, 48 Dilsberg, 1, 2
Christianity, 40 Dinesen, Isak, 170
Christian Orthodox Church, 40 Divorce: and age at marriage in Great
Churches, 40, 48, 207 Britain, 32–33; among consanguineous
Cleopatra, 2, 159 couples, 43; among sibling marriages
Coevolution, 34, 150, 173, 212 in Roman Egypt, 100, 105; among Tai-
Colorado, 42, 159 wanese minor marriages, 27, 77–88
Confucian tradition, 40 DNA, 29, 45, 54, 58, 60, 63, 66, 106,
Consanguineous marriage: among Hut- 110, 187
terites, 44–45; attitude toward in world Dravidian India, 43, 46
religions, 40–41; estimating biological Durham, William H., 10, 14–18, 34,
outcomes of, 45–47; examples of in 36–37, 61, 111, 116, 119–122, 124,
biblical texts, 39; influence on repro- 126, 128, 130, 134, 136–138, 150–
ductive success, 44–45; major types of, 152, 158–159, 196, 214–216, 219
39 (tab); medical consequences of, 19, Dynastic incest, 2, 47, 153
53–54; preference for in some societies,
42–43, 94, 127–128; in U.S. state laws, East, E. M., 2
42–43 (fig), 54. See also Brother-sister Eastern Toradja, 131, 132
marriage; Cousin marriage; Inbreeding, Ecological anthropology, 143
evidence of deleterious consequences Egypt: affluence of, 103; Graeco-Roman,
of; Uncle-niece marriage 97; kings of, 47; nursing practices in,
Cousin marriage, 146; in Arab Muslim 101; Roman, 9, 46, 93, 94, 96, 97, 100,
communities, 43; in China, 40; in 104, 105; Pharaonic, 47; Ptolemaic, 47;
Europe, 40; methods of classifying sibling marriage in, 94, 99, 101
by Latin church, 40; methods of esti- Ellis, Havelock, 18, 76, 78, 82, 83, 86
 Index 223

Ember, Melvin, 126, 137, 214 God, 27, 195

Environmental determinism, 36 Goddu, David, 169
Erickson, Mark T., 13, 18–20, 58, 61, Godelier, Maurice, 5, 22, 141
92, 151, 155–156, 159, 184, 186–188, Goggin, John, 153, 159
215, 219 Goldenweiser, Alexander, 5
Ethical naturalism, 193–196, 206–211 Gombe National Park, 68, 74
Ethical Relativity, 200, 213, 215, 216 Goodall, Jane, 64, 69, 70, 72–75, 216,
Europe, 19, 40–42, 51, 54–55, 138, 217 220
Evolution, 7, 22, 25, 27, 33–35, 37, 64, Goody, Jack, 55, 138, 152, 159, 217
72, 107, 121, 138, 140, 141, 152, 154, Gorgias, 193, 202
162, 172, 178, 186–187, 191–192, Gorillas, 6, 63–64, 71
214, 216, 220 Gould, Stephen Jay, 36, 137, 211–213,
Evolutionary biology, 110, 111, 206 217
Evolutionary psychology, 110 Great Britain, 22, 41, 164, 183
Exogamy, 6, 8–9, 21–22, 41, 91. 126, Greek philosophers, 193, 206, 211
141, 148, 150 Greenberg, Maurice, 169, 181, 185, 186,
Expression hypothesis, 111, 114–118, 188
122–137 Gregory, Pope, 40, 138, 186, 196

Faller, Kathleen, 167, 184–185 Habituation, 31, 33, 122. See also West-
Farber, Paul Lawrence, 206, 217 ermarck hypothesis, substance of
Fayum Oasis, 93 Halperin, Daniel, 164–165, 182, 183
Fertility: of brother-sister marriage in Hamilton, William D., 172, 186–187
Roman Egypt, 48, 91–100 (tab); of Han, 40, 43
consanguineous unions, 44–45; of Hansar, 41
Taiwanese minor marriages, 31, 77–78 Hapsburgs, 25
(tabs), 94 Hardy, Gathorne, 42
Finkelhor, David, 164, 166, 179, 182–183 Harrell, Stevan, 152–153, 155, 159
185, 188–189 Harris, Marvin, 5, 141, 152, 156, 159,
Finland, 46, 58, 164–165, 184–185 186
Firth, Raymond, 138, 149–150, 158 Hawaii, 47
Florida, 153 Haynes-Seman, Clare, 175, 188
Fourth Lateran Council, 40 Hendrix, Lewellyn, 89, 92, 97, 106, 108
Fox, Robin, 9, 22, 174, 187, 216 Hercules, 88, 91
Frazer, Sir James, 5, 9, 22, 121, 141 Herman, Judith, 164, 167, 176, 181–188
French Enlightenment, 150 Hess, Eckhard H., 1, 4, 5, 21
Freud, Sigmund, 5–6, 8– 9, 13–14, 22, Hinduism, 41
47, 78, 91, 121–122, 139–144, 145, Hittites, 153
162, 170, 172, 174, 176, 181, 188, Hobbes, Thomas, 196–197, 201, 209,
201–205, 209, 213, 216 213–215
Freud and Human Nature, 202, 216 Human Behavior and Evolution Society,
Fried, Morton, 152–154, 159 192
Frier, Bruce, 95, 97–98, 105–107 Human Genome Project, 29
Human nature, 20–21, 179–180,
Galton, Francis, 42, 126 193–196. See also Incest prohibitions
Gates, Hill, 9, 13, 15, 18, 25, 89, 92, 108, and human nature
131, 134–135, 139, 205, 219 Human Relations Area Files, 131, 137–
Geertz, Clifford, 143, 156, 159 138
Genes, 3, 25, 29–39, 44, 46, 52, 65, 90, Human social behavior, 24, 34–35, 192,
103, 107–109, 121, 129, 137, 142, 156, 207, 211
158, 191–192, 211–214, 216, 219 Hume, David, 150, 193, 206–209, 213,
Germanic system, 40 217
224 Index

Hutcheson, Francis, 17–18, 23, 197, 215 123–128; psychopathology of, 19–2,
Hutterites, 44–45, 56 48, 161; punishment for, 147–149
Incest, causes of, 168–178; early abuse,
Inbreeding, definition of, 24, 38 172, 176; insecure childhood attach-
Inbreeding, deleterious consequences of: ment, 176; patriarchy, 168; poverty,
accepted by Aberle et al., 3, 8; argued 168
for by Westermarck, 2, 10, 130, 204– Incest, definition of, 12, 24, 38, 139, 165.
205; change in view of, 10; Darwin’s See also Inbreeding, definition of
interest in, 43–44; debated in parlia- Incest, forms of: brother-sister, 164–165
ment in 1870s, 41–42; denied by incest (tab); father-daughter, 164–165 (tab);
scholars, 2–3; genetic reasons for, 25, mother-son, 146–147, 167
38–39; medical implications, 19, 53– Incest avoidance, attribution to: child-
54; noted by Twain, 1–2; problem of hood association (see Westermarck hy-
measuring magnitude of, 3, 18, 19, pothesis, substance of); childhood pun-
49–53; relation to incest taboo, 15, ishment, 174; cultural norms, 12–13,
128–135; relation to incest taboo ques- 163; secure childhood attachment, 14,
tioned, 10–11, 15–16, 25, 34. See also 61, 71–72, 151, 176–177; sexual frus-
Inbreeding, evidence of deleterious con- tration, 174; sexual segregation, 88–89;
sequences of similar major histocompatibility com-
Inbreeding, evidence of deleterious conse- plexes, 89, 103–105. See also Incest
quences of: case of the Hapsburg jaw, prohibitions, attribution to
25–56; excess mortality, 53; excess Incest prohibitions, attribution to: cul-
morbidity, 53; from birds, 25; from tural injunctions, 12–13; childhood
captive mammals, 62; from close-kin association (see Westermarck theory);
marriages in Roman Egypt, 94; from dangers of inbreeding, 4, 129–135;
incestuous unions, 49–52 (tab); from desire for reproductive success, 33;
legal consanguineous unions, 52–53 dislike of abnormal behavior, 34–
Inbreeding, recognition of deleterious 35; fear of supernatural, 129; need
consequences by laypeople, 1–2, 4, 8, for family harmony, 8, 196; need
15, 34, 129–135, 150–151 for group alliances, 7–9, 140–141,
Inbreeding avoidance, attribution to: at- 150, 196; pain aroused by sight of
tachment, 70–71; early association (see incest, 14–15; recognition of danger
Westermarck hypothesis, substance of); of inbreeding, 129–137, 150–152,
sex-biased dispersal, 62–63 196. See also Incest avoidance, attribu-
Inbreeding avoidance, evidence for: in tion to
baboons, 6, 62–63, 65; in quail and Incest prohibitions, culture and, 12–13,
other birds, 25–26, 177; in bonobos, 20, 142, 178–180, 202
6, 177; in chimpanzees, 6, 62–63, 208; Incest prohibitions, explanation proposed
in gorillas, 6, 63–64; in lemurs, 63; by: Aberle et al., 8; Aquinas, 196;
in macaques, 6, 63, 65; in mammals Burton, 128–130; Darwin, 198–199;
generally, 5–6, 62, 163; in non-human Durham, 130–135, 150–152; Evans-
primates generally, 6, 72, 163; relation Prichard, 6; Firth, 150; Freud, 8, 140,
to incest taboo questioned, 25, 34 145; Lévi-Strauss, 7, 141; Needham,
Inbreeding depression. See Inbreeding, 12; Plato, 194; Tylor, 6, 141; White,
deleterious consequences of 6–7, 141. See also Westermarck incest
Incas, 47, 153 theory, substance of
Incest: age at initiation, 168; as cultural Incest prohibitions, instances of: in East-
concept, 12–13 (see also Incest prohibi- ern Toradja, 131–134; in Nuer, 6; in
tions, instances of); disapproval of by Tikopia, 135, 149–150; in Trobriand
third party, 11–14, 33–34, 115–119, Islands, 131, 143–150, 154–156
123–128; frequency of, 48, 161, 163– Incest prohibitions, morality and, 10–13,
168 (tab), 177; moral disapproval of, 115–119, 123–128, 190–213
 Index 225

Incest prohibitions, political aspects of, Kant, Immanuel, 17, 191, 197–198,
19, 152–156 207–209, 215, 217
Incest prohibitions, problems in explain- Karnataka, 41, 55, 57
ing: disapproval of by third party, Kentucky, 45
10–13, 115–119, 123–128; expression Kibbutzim, 9, 26, 31, 46, 78, 164,
in kinship terms (see Representation 175–176, 178, 208
problem); moral quality, 10–11, 13, Kin recognition, 65–77, 171; relation to
115–119, 123–128, 210–211; relation olfaction, 103, 171
to incest avoidance, 111–119; relation Kinsey, Alfred, 163–164, 182, 187
to dangers of inbreeding, 10–13, 15– Kinship and incest avoidance, 173–174
16, 25, 115–119, 128–130, 133–135; Kipling, Rudyard, 28
variability across societies, 207, 210 Koran, 43
(see also Localization problem); why Korea, 76, 153
needed if childhood association inhibits Kroeber, A. L., 5
sexual attraction, 5 Krugman, Richard, 175, 188
Incest prohibitions, scope of, 13, Kruyt, Albert, 131–132, 138
210–215 !Kung San, 179
Incest prohibitions, significance in con-
temporary debates, 110–111, 121, Labby, David, 144, 157
151–152, 190–191, 212–213 Lagace, Robert, 132, 137–138
Incest prohibitions, violations of, 13, 47, Latin Church, 40, 48
147–149 Lebanon, 15, 78, 116, 127–128
Incest prohibitions and human nature, Leigh, Augusta, 170
views of: Freud, 5, 140, 201–202; Lemurs, 62–64, 73
Hutcheson, 17, 197; Lévi-Strauss, 7, Lévi-Strauss, Claude, 7–8, 10, 12, 22, 47,
202; Malinowski, 140; Mandeville, 17, 141, 150–151, 156, 181, 202, 216
197; Smith, 197–198; Westermarck, 9, Lewontin, Richard, 5, 36, 137, 216
199–206; White, 6–7 Littlewood, Roland, 169, 181, 185–186,
Incest taboo. See Incest prohibitions entries 188
Incestuous child abuse, 163–168, 177 Localization problem, 10, 12–13, 211
Incestuous unions. See Consanguineous Locke, John, 42
marriage Louisville, 45
India, 3, 24, 38, 41–44, 46, 55–57, 108, Lubbock, Sir John, 19, 41
215, 217
Indifference theory. See Westermarck Macaques, 6, 62–67, 73–74
hypothesis, substance of Maccoby, Eleanor, 36, 37, 88, 89, 92
Ireland, 22, 41 Madonna, Philip, 166, 183–184
Iran, 47, 59, 94, 106, 108 Maestripieri, Dario, 178, 189
Islam, 40 Major histocompatibility complex
Isle of Eriskay, 46 (MHC), 65, 89, 103, 177
Isonymy, 48 Malasi, 148–149, 155
Israel, 9, 15, 26, 31, 47, 58, 78, 116, 120, Malinowski, Bronislaw, 2, 13, 140,
127–128, 163, 177, 183, 185, 204. See 142–152, 154
also Kibbutzim Mandeville, Bernard, 17, 201
Mansfield Park, 76, 86
Japan, 3, 26, 29, 53, 57, 60, 77, 106, 108 Marriage: Aquinas on, 195–196; age at
Japanese macaques, 64, 67, 74 and divorce in Great Britain, 32–33;
Japanese quail, 26, 36, 177, 188 forms of in Roman Egypt, 9, 93; forms
Java, 153 of in Taiwan, 27, 76–77, 203, 209; by
Jocasta, 21, 170 incestuous couples, 171; natural incli-
Jochebed, 39 nation toward, 199; Westermarck on,
Judaism, 40, 43 199–200
226 Index

Marriage Act, 40 Nicomachean Ethics, 191, 214

Marriage prohibition, 6, 126 Nin, Anaïs, 170–171, 185
Marxism, 152 Nin, Joaquin, 170, 172
Massachusetts, 169 North America, 41, 54
Mating preferences, 26–27, 31–34, 43– Nuclear family, 129, 140–142, 149, 153,
45. See also Consanguineous marriage, 155, 165, 212
preference for in some societies Nuer, 6, 22
Matriliny, 144–145 Numakala, 155
Maugham, Somerset, 170
Maximilian I, 25 Occitania, 48
McCabe, Justine, 91, 122, 127, 138 O’Connor, Susan, 178, 189
McCarty, Loretta, 167, 183–185, 188 Oedipus complex, 9, 14, 140, 162, 174,
McEwan, Ian, 47, 58 205
McShea, Robert, 208, 217 Olfaction. See under Kin recognition
Mediation problem, 10, 12–14, 18, 20 Omarakana, 154
Melville, Herman, 170 Optimal outbreeding, 24–27, 130, 149,
Memorabilia, 194, 214 161, 177, 205
Middle Ages, 48, 94, 193, 195, 213
Middle East, 38, 44, 127 Packer, Craig, 63–64, 66, 68, 72–74
Mill, John Stuart, 206, 217 Pakistan, 3, 24, 38, 44, 55–56
Miller, Daniel R., 1, 21, 217 Paley, William, 27, 36
Ministry of Justice, 54 Parker, Hilda, 97, 106, 179, 183–184,
Minor marriage, 83, 88–89, 127, 163 189
Modern science, 2, 191 Parker, Seymour, 179, 183–184, 189
Monogamy, 199, 200, 211–212 People’s Republic of China, 40
Moore, G. E., 206, 217 Phelan, Patricia, 167, 184–185
Moral emotions, 17, 21, 124, 136, Philip IV, 26
190–193 Pierre, or the Ambiguities, 170
Morality. See Incest prohibitions, moral- Pinker, Steven, 17, 192, 214
ity and Pittorino, Victoria, 169
Morbidity, result of inbreeding, 3, 19, 53 Plato, 193, 194, 200–202, 214, 219
Morgan, Lewis Henry, 56, 152, 217 Polyandry, 199–200, 211–212, 215
Morocco, 133, 138 Polydactyly, 51, 59
Mortality, result of inbreeding, 3, 19, 53 Polynesia, 131, 138, 156, 158
Moses, 39 Popper, Karl, 78, 91
Mummies, 48 Price, Fanny, 76, 86, 140
Murdock, George Peter, 5, 17, 23 Primates, 6–7, 10, 61–67, 71–75, 142,
Muslim, 43 172, 178–179, 187, 189
Muth, Allen, 169 Protestants, 40
Muth, Patty, 169 Purayasi, 155
Pusey, Anne, 6, 13, 72–75, 78, 181, 185,
Nagel, Thomas, 192, 206, 214, 216–217 188, 216, 220
National Conference of Commissioners
on Uniform State Laws, 42 Raglan, Lord, 2, 21
Native Americans, 42 Red colobus, 62
Natural right tradition. See Ethical Reformation, 207
naturalism Religion, 59, 142, 148, 220
Navarette, Friar Domingo, 88, 92 Representation problem, 10–11, 14,
Near East, 94 16–17, 112–115
Needham, Rodney, 12–13, 2, 156 Rhesus monkeys, 187
Neel, James V., 50, 53, 58–60, 106 Rhetoric, 196
New Guinea, 145, 156–157 Rh incompatibility, 44
 Index 227

Riviere, Peter, 12, 22, 91 Social relationships, 7, 64, 74

Roman Catholics, 40, 46 Socrates, 194
Rome, 103, 107, 195 Sophists, 193–194, 211
Russell, Diana, 22, 164–165, 182–183, South Dakota, 44
187 Southeast Asia, 219
South India, 44, 46, 55–57
Sahlins, Marshall, 5, 141, 143, 156, 157 Spider monkeys, 62, 72
Scheidel, Walter, 59, 93, 105–108, 220 Spuhler, James N., 1
Schneider, David M., 1, 12, 21, 22 Sroufe, Allen, 175–176, 188
Schneider, Mark, 89, 92, 97, 106–108, Statistical Society, London, 42
156 Les Structures élémentaires de la
Schull, William J., 53, 60 parenté, 7
Scotland, 38, 46 Sturtevant, William, 153, 159
Scottish Enlightenment, 150 Sudras, 41
Seligman, Brenda, 5 Sulawesi, 131
Sensitive marriage. See Mating preferences Swedish Marriage Law, 54
Sensitive period, 172–173, 178. See also
Critical period Taiwan, 76–79, 89–91, 94, 116,
Sensitivity to childhood association, sex 127–128, 163, 177, 186, 219
differences in, 18, 78, 82–86 Talmon, Yonia, 9, 58, 91, 182
Sesardic, Neven, 11, 14, 16, 109, 120, Talmud, 39, 55
124, 127, 137, 220 Tamil Nadu, 41, 56–57
Sex-biased dispersal, 62–63, 208 Tanzania, 62, 68, 74
Sexual abuse, 164–168, 174, 177, 180, Thayer, Gerald, 28, 36
182–184, 188–189 Tikopia, 131, 138, 142, 149, 154, 158
Sexual attraction, determinants of among Tongan marriage, 153–154
childhood associates: age at first associ- Toradja, 131–133, 138
ation, 79–82 (tabs), 99; differences in Totem and Taboo, 22, 140, 216, 220
age, 83, 86, 97–100; social preference, To’uluwa, 154
127–128; relative age, 83–86 (tabs). Trobrianders , 13, 131, 140–150, 155,
See also Sensitivity to childhood 157, 158
association Tuareg, 43
Sexual Attraction and Childhood Associ- Twain, Mark, 1, 2, 3, 19, 21
ation, 21, 36, 58 91, 106, 137, 160, Twentieth century, 1, 5, 10, 12, 18,
181, 214, 215–217, 220 20–21, 26, 51, 76, 97, 121, 139–143,
Sexual aversion, 14, 16, 17, 78, 79, 86, 150, 162–163, 178–179, 207, 220
97, 111–117, 140, 151, 190, 199, 200, Tylor, Edward Burnett, 6, 7, 22, 141, 150
202–205, 210–212
Sexual Life of Savages, 139, 148, Uberoi, J. P. Singh, 143, 154, 157, 160
156–158, 160 Ulpian, 195
Sexual preferences. See Mating preferences Uncle-niece marriage, 3, 19, 39, 43, 46.
Sexual segregation, 32, 88–89 See also Consanguineous marriage
Shaftesbury, Lord, 197 United States, 24, 42–43, 50–51, 54, 56,
Shepher, Joseph, 9, 36, 91, 99, 107, 122, 142, 163, 165–166, 169, 178–179, 207,
127, 128, 181–182, 188, 216 217
Sibling marriage. See Brother-sister Utopia, 9
marriage
Silverman, Irwin, 99, 107 Vaishyas, 41
Sinhala, 153 Van Scoyk, Susan, 166, 183–184
Smith, Adam, 197, 201–204, 209, 217 Varnas, 41
Smith, David G., 65 Velásquez, 26
Smith, John Maynard, 111, 172 Venerable Bede, 40
228 Index

Wagner, Roy, 12 Westermarck hypothesis, substance of, 4,

Wallace, Alfred, 2, 4, 22 47, 62, 65, 78, 86, 94, 112, 162, 199.
Ward, Mary, 60, 159, 175, 188 See also Westermarck incest theory,
We the Tikopia, 138, 149, 158 substance of
Wedekind, Claus, 92, 107, 177, 188 Westermarck incest theory, criticized by:
Wedgwood, Emma, 41 Aberle et al., 4–5; Durham, 115–119,
Weinberg, Steven, 111, 120 122–237; Frazer, 5, 212; Freud, 5, 121,
Weinberg, Kirson, 164–165, 168, 182, 201–204; Williams, 112–115
185 Westermarck incest theory, substance of,
Weiner, Annette, 143–144, 150, 157–158 2, 4, 14–15, 17, 25, 34, 111, 123–125,
Weiss, David, 179, 189 163, 194, 200–207. See also Expression
Westermarck, Edward, 2–6, 9–10, 12–13, hypothesis; Representation problem;
15–18, 20–21, 34, 61, 65–67, 71–73, Westermarck hypothesis, substance of
76–79, 82–83, 86, 88–93, 100, 104, Westermarck incest theory, viewed favor-
106–107, 110–112, 114–140, 142, ably by: Bateson, 34; Fox, 9; Wilson,
150–152, 160, 162–163, 166, 170– 34, 121, 195, 210
171, 173–174, 181, 190–191, 193, Westermarck process, 123–128, 133
198–199, 200–205, 209–217, 220 Western Europe, 41, 54
Westermarck effect, 34, 94, 100, 104, Wet nursing, 103–105
121–123, 125, 128, 130, 151, 209; as White, Leslie, 2, 5–8, 10, 12, 21–22, 134,
adaptation, 10, 122; associated with 138, 141, 181
breastfeeding (see Breastfeeding); con- Williams, Bernard, 11, 15, 16, 22,
ditions necessary for, 78–91, 163; in 112–113, 115, 120, 179, 182–183
non-human animals, 65–67; violations Williams, George C., 172, 186
of, 165–168. See also Aversion hypoth- Williams, Linda, 179
esis; Mediation problem Wilson, E. O., 17, 34, 37, 110, 120–121,
Westermarck hypothesis, evidence for, 137, 182, 191–195, 206, 211–214
122; among Canada geese, 4; effect of Wisconsin, 42, 169
early separation on likelihood of incest, Wolf, Arthur P., 21–22, 26, 34–36, 58–59,
170; among kibbutzim, 9, 26, 31, 46, 61, 72–73, 91–94 99 101, 106– 108,
78, 127, 163, 177, 204; in Lebanon, 121–122, 124, 127–128, 137, 140,
78, 122, 127; among non-human 151–152, 159, 160, 163, 173, 177–
primates, 6, 9, 63–65 (tab), 72, 204; 178, 181–182, 184, 186–188, 203,
in quail, 26–27, 177; in Roman Egypt, 209–210, 214–217, 220
9–10, 93–105; from Taiwanese minor Woolcock, Peter, 206, 217, 220
marriages, 9, 26–27, 31–32, 47, 76– World War I, 140, 143
91, 94, 127, 163, 177, 203–204, World War II, 143
210. See also Inbreeding avoidance,
evidence for Xenophon, 194, 214
Westermarck hypothesis, rejected by:
Aberle et al., 5–6; Freud, 5; studies Zoroastrians, 40, 47–48, 105; close-kin
of cousin marriages, 125–126 marriage among, 47–48, 94, 105