Open Book Publishers logo Open Access logo
  • button
  • button
  • button
GO TO...
book cover

7. Avoiding Inbreeding and Incest1

© 2017 Patrick Bateson, CC BY-NC-ND 4.0

Finding a compatible partner is an important part of reproductive behaviour in many animals in which mates are chosen carefully. Members of different species do not make good mates. At the other pole, too much inbreeding can also reduce reproductive success.

Inbred animals are more likely carry some damaging genes. Most potentially harmful genes are recessive and are therefore harmless when they are paired with a dissimilar gene 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 inbreeding. The genetic costs of inbreeding arising from the expression of damaging recessive genes are the ones that people usually worry about.2 Recessive genes are less of a problem in mammals than they are in birds because mammals generally move around less and they may live in quite highly inbred groups where the harmful alleles have been purged. The most important biological cost of excessive inbreeding is that it negates the benefits of the genetic variation generated by sexual reproduction. If an animal inbreeds too much, it might as well make copies of itself without the effort and trouble of courtship and mating.

On the other side, excessive outbreeding also has costs. For a start, excessive 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 characteristics. So too are jaw size and shape, as may be seen in many paintings of the Hapsburg family, scattered in the museums around the world. The Dürer painting of the Holy Roman Emperor Maximilian I reveals the large Hapsburg jaw, which was even more pronounced in his highly inbred great-great-great-grandson, Philip IV of Spain, shown in the painting by Velasquez. The potential problem arising from too much outbreeding is that the inheritance 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 probably 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 illustrates the more general cost of too much outbreeding.

Two members of the Hapsburg family separated by five generations. Albrecht Dürer, Portrait of Maximilian I, Holy Roman Emperor (1519), Kunsthistorisches Museum, Vienna. Wikimedia,ürer_-_Portrait_of_Maximilian_I_-_Google_Art_Project.jpg, Public Domain. Diego Velázquez, Portrait of Philip IV (1656), National Gallery, London. Wikimedia,, Public Domain.

With costs accruing to both inbreeding and outbreeding, Darwinian evolution is presumed to have operated on mechanisms involved in mate choice to minimise both. The outcome of Darwinian evolution is a preference for a mate that is not too closely related and not too distantly related.3 Hybrid vigour is so dramatic when it occurs that it seems to argue against the view that marginal outbreeding is beneficial. Such hybrids are usually infertile and the original parents will have few or no grand-offspring. This means that tests of outbreeding depression should be made in the natural environment where the benefits of co-adapted gene complexes can be revealed.

In experiments with the mountain delphinium the largest number of seedlings was produced by crosses between plants that were ten metres apart. Plants that were self pollinated and those that were crossed with plants 1000 metres away, both gave rise to significantly smaller numbers of seedlings.4

This delphinium produces most seeds when crossed with plants 10 metres away from it. Delphinium nuttallianum. Photo by Walter Siegmund (2009), Wikipedia,, CC BY 3.0.

In animals direct evidence for the genetic costs of outbreeding is still relatively slender, although the examples are starting to multiply. Some of the studies of humans suggest that fecundity is related to the similarity between spouses. For example, in one study the more alike human couples were on 17 out of 19 measures of the body (such as forearm length, height and ear length), the more children they had. Although most correlations were positive, each correlation coefficient was low and could, of course, be explained by shared associations between the measures and a third variable such as social class. In a study of Icelandic partners, those who were distant cousins had more grandchildren than those who were more closely or more distantly related. The best outcome in terms of biological adaptiveness is to have the most grandchildren when other things such as social status are equal.5

Not all the costs of outbreeding too much are genetic in character. If and when they operate, these non-genetic costs further complicate the interpretation of laboratory breeding experiments aimed at settling whether or not outbreeding can be costly. For instance, two unrelated individuals in the laboratory may produce many offspring and, because they share adjoining cages and therefore common antibodies, they incur no cost from being exposed to pathogens carried by the other. It could be a different story in the natural environment. Similarly, the advantage of using skills acquired for dealing with the local environment could counteract the genetic advantage of moving into another area prior to breeding. Yet the non-genetic advantage could not be assessed in a laboratory experiment. Nor could the various costs when moving away from the natal area such as increased risks from predation. Some caution is needed before jumping to conclusions about just where the balance between inbreeding and outbreeding is likely to be struck.

If an animal does best by choosing a mate that is neither a close relative nor totally unrelated to it, what mechanisms could it use? Of the various types of explanation that have been offered, two are likely to be important. The first proposes that prior to mating, members of one sex move away from the area where they were hatched or born.6 Providing they do not move too far, their mates are likely to bear some genetic relationship to themselves and so optimal outbreeding could be achieved. The second possibility is that animals are able to recognise close kin and, on the assumption that physical appearance is a measure of genotypic similarity, choosing a mate that looks, sounds or smells a bit different but not too different from close kin will result in optimal outbreeding. These two explanations are not mutually exclusive. Some species could employ both mechanisms. The evidence suggests that both are found in the animal kingdom.

First, in many species of bird and mammal one sex moves out of the natal area prior to breeding. In most species of birds, females move away although exceptions exist such as the snow goose in which the male is unlikely to return to the natal area. In most species of mammal the males are more likely to move away but here again exceptions are known such as the chimpanzee in which the females move away from their natal group. The costs of travel can be quite considerable and usually the distance travelled by the sex that moves is not great. The net effect of restricted movements in one direction and returns by offspring in the next generation could be an overall population that was quite highly inbred. An important question remains whether such a system would be sufficiently finely tuned to preserve the optimal balance between inbreeding and outbreeding.

Recognition of kin, the other suggested mechanism for optimal outbreeding, could be accomplished in one or two ways. One possibility is that the genes that influence an animal’s external appearance also directly influence its ability to recognise another animal very much like itself without the involvement of any learning process. More plausibly, the animal learns the characteristics of close kin, or failing that of itself, and can then recognise novel individuals that are similar to kin. In Japanese quail a first-cousin is preferred by both sexes over siblings and also over unrelated birds.7

Natural experiments have been performed unwittingly on humans. The most comprehensive evidence has come from the marriage statistics from Taiwan in the late nineteenth and early twentieth century, when Taiwan was under Japanese control. The Japanese kept detailed records for the births, marriages and deaths of everyone on the island. As in many other parts of South East Asia, marriages were arranged, and they 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 marriage, the wife-to-be was adopted as a young girl into the family of her future husband. In minor-type marriages, therefore, the partners grew up together like siblings. 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.8 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. 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.

In the past Israeli kibbutzniks who grew up together like siblings rarely married each other. The few 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.

Early experience and sexual attraction

Neither of the evidence from Taiwan or Israel means that the learning process that affects adult sexual preferences is completed early in life. If children grow up together and consequently see a lot of each other, they revise the ways in which they recognise 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.9 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.

How could finely tuned sexual preferences arise from early experience? The responsiveness to the familiar could be reduced by mere exposure, and consequently individuals that differed slightly from the known standards would be most attractive. A simple way of producing a finely tuned preference displaced away from the familiar could be to superimpose habituation on imprinting. The net effect of superimposing one learning process on the other would be to produce a sharply peaked preference for something a bit different from the familiar. The evolutionary benefits and costs of inbreeding are likely to be numerous and to vary in importance from one species to the next. Therefore the precise balance will almost certainly differ between species. Even within a species, the balance is likely to depend on local conditions and on how inbred a population has become. Finally, the sexes may differ, especially when one sex is likely to have more matings than the other. Nonetheless, a general point remains. When choosing a mate an animal may have to pay careful attention, among other things, to similarities between its proposed partner and close kin.

Incest taboos

Many authors have suggested that individuals may derive reproductive success from incest taboos.10 Those individuals who impose the prohibitions do not derive immediate personal benefits from them. 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. 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.

Whether or not people are aware of the effects of inbreeding is another issue. In many cultures they are.11 Awareness of the ill-effects of inbreeding would be best translated into the conviction that the aware individual should not have children with his or her sibling. Nothing more is required of Darwinian evolution. The awareness does not immediately translate into a conviction that others should be stopped from having children with their siblings.

Did inbreeding avoidance and incest taboos evolve by similar mechanisms or do they have a common utility in modern life? Incest taboos need not necessarily serve the same function as the inhibitions derived from early experience. What other mechanism for the cultural evolution of incest taboos should be entertained? Humans might often have an inclination to prevent other people behaving in ways in which 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 them disturbing. Similarly the moral repugnance that many people show for homosexuality between consenting adults is often a violent one — in some societies homosexuality may be punished by death. In the same way, those who were known to have had sexual intercourse with close kin were discriminated against. People who had grown up with kin of the opposite sex were generally not attracted to those individuals, and disapproved when they discovered others who were. On this view, the incest taboo was nothing to do with society not wanting to look after the cognitively challenged offspring of inbreeding, since in many cases they had no idea that inbreeding was the cause. Rather, the disapproval was about suppressing abnormal behaviour, which is potentially disruptive in small societies. Such conformity looks harsh to modern eyes, even though plenty of examples of it are found in contemporary life.

When so much depended on unity of action in the environment in which humans evolved, wayward behaviour could have destructive consequences for everybody. It is not difficult to see why conformity should have become a powerful trait in human social behaviour.12 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, non-kin as well as kin, were likely to be familiar from early life.

In the Anglican Book of Common Prayer there is a table of Kindred and Affinity ‘wherein whosoever are related are forbidden by the Church of England to marry together’. A man may not marry his mother, sister or daughter and a variety of other genetically related individuals. The restrictions for a woman are reciprocal and also, among others, she must not marry an uncle, nephew, grandparent or grandchild. The list continues with the following exclusions: a man may not marry his wife’s father’s mother or his daughter’s son’s wife. At least six of the 25 types of relationship that preclude marriage involve no genetic link. The Church of England did not worry about marriages between first cousins. Other cultures do, but here again striking inconsistencies are found. In a great many cultures marriages between first cousins who have parents of the same sex are forbidden whereas marriages between cousins who have parents of different sexes are not only allowed but, in many cases, actively encouraged.

If these ideas are correct, human incest taboos did not arise historically from a deliberate intention to avoid the biological costs of inbreeding. Rather, in the course of history, two separate mechanisms appeared. One was a developmental process concerned with striking an optimal balance between inbreeding and outbreeding when choosing a mate. The other was concerned with social conformity.13 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 which could be transmitted from generation to generation, first by word of mouth and later in written form.


While the incest taboo is not likely to be a result of Darwinian evolution, the preference for a mate slightly different from close kin undoubtedly is. Other factors also influence mate choice, particularly in humans. The attractiveness of a potential mate, the resources he or she might hold, and so forth also play a part. But in terms of producing grand-offspring, optimal outbreeding is important. The mating preference appears to be well designed to achieve that except in modern human society where having a lot of grandchildren is not usually important.

1 This chapter is drawn, with permission, from: Bateson, P. (1983), Optimal outbreeding. In: P. Bateson (ed.), Mate Choice. Cambridge: Cambridge University Press, pp. 257–277.

2 Across Asia, the effects of genetic disorders are becoming increasingly obvious. This change is especially important among the children of couples who have married cousins and is also found in migrant communities resident in North America, Western Europe and Australasia who continue the tradition of close kin marriages (Bittles, A.H. (2003), Consanguineous marriage and childhood health. Developmental Medicine 45.8, 571–576, The unfortunate consequence arises from practicing first cousin marriages generation after generation. In relatively inbred animal populations the preferred genetic distance between mates may be greater than in more outbred populations thereby offsetting the effects of mating with close kin (see Bateson, P. (1983), Mate Choice).

3 The idea of balance between inbreeding and outbreeding was first suggested by Wright, S. (1933), The roles of mutations, inbreeding, crossbreeding and selection in evolution. Proc. VIth Internat. Congr. Genetics 1, 356–366, available at

4 Price, M.V. & Waser, N.M. (1979), Pollen dispersal and optimal outcrossing in Delphinium nelsonii. Nature, 277.5694, 294–297. Independently of the use of ‘optimal outbreeeding’ (see Bateson, P. (1978), Sexual imprinting and optimal outbreeding. Nature, 273.5664, 659–660, they used the term ‘optimal outcrossing’.

5 Helgasson, A. et al. (2008), An association between the kinship and fertility of human couples. Science 319.5864, 813–816,

6 Greenwood, P.J. (1980), Mating systems, philopatry and dispersal in birds and mammals. Anim. Behav. 28.4, 1140–1162,

7 Bateson, P. (1982), Preferences for cousins in Japanese quail. Nature 295.5846, 236–237,

8 Wolf, A.P. (1994), Sexual Attraction and Childhood Association. Stanford: Stanford University Press.

9 Maccoby, E.E. (1990), Gender and relationships: a developmental account. Amer. Psychol. 45.4, 513–520,

10 Many authors have linked avoidance of inbreeding with the incest taboo. One of the most prominent is Wilson, E.O. (1998), Consilience: The Unity of Knowledge. New York: Alfred A. Knopf.

11 Durham, W.H. (1991), Coevolution: Genes, Culture and Human Diversity. Stanford: Stanford University Press.

12 Westermarck, E. (1891), The History of Human Marriage. London: Macmillan.

13 Westermarck’s ideas are discussed at length in the chapters in Wolf, A.P. & Durham, W.H. (eds.) (2004), Inbreeding, Incest and the Incest Taboo. Stanford: Stanford University Press.