The Selfish Gene

Dawkins likens unrelated organisms to parts of the environment. However, unlike a rock or a river, one survival machine can return the attacks of another survival machine. Natural selection produces survival machines that make use of the environment, including other survival machines of the same or different species. All organisms share the same environmental resources, even if sometimes remotely. The closer relatives, such as members of the same species, interact more. For example, eating and mating both require common resources for members of one species:

The logical policy for a survival machine might therefore seem to be to murder its rivals, and then, preferably, to eat them. Although murder and cannibalism do occur in nature, they are not as common as a naive interpretation of the selfish gene theory might predict (55).

Instead of fights to the death, animals often defer to bluffs, signaling surrender. Rather than altruism, Dawkins ascribes this to the costs of violence. Attacking rivals takes time and energy, it and allows yet other rivals to out-compete. Game theory can explain the costs and benefits of aggression.

John Maynard Smith develops the idea of the evolutionarily stable strategy. As populations attack or defend according to different protocols, one strategy falls into place that cannot get dislodged by other strategies. The strategy of an individual therefore depends on the strategies of others: “Since the rest of the population consists of individuals, each one trying to maximize his own success, the only strategy that persists will be one which, once evolved, cannot be bettered by any deviant individual” (57).

For example, in a population of peaceful and violent members, each can benefit its likelihood of passing on genes through the outcome of conflicts. Hawks aggressively fight, with high risk and reward. Doves bluff and flee, with low risk and reward. Through mathematical calculations, one finds that a mutant hawk among doves, or vice versa, would produce more descendants. Eventually the reproducing population stabilizes at an equilibrium with a certain distribution of aggressive versus passive strategies.

Evolutionarily stable strategies differ from group selection. The successful strategies evolve on the basis of assisting the individuals’ genes, not promoting the welfare of any population. A group selection could theoretically result in a population distribution more beneficial to all individuals. However, the incentives for any individual to violate the group outweigh the membership benefits, so the group becomes unstable and would revert to the evolutionarily stable strategy.

Humans can form more advantageous groupings than evolutionarily stable strategies, through foresight. However, incentives to violate the artificial arrangement remain. Evolutionarily stable strategies predict resultant stable gene distributions, or “stable polymorphisms” (75). Scientists start with a simple model and expand it to portray more realistic situations. For example, Maynard Smith adds strategies such as the retaliator and bully to dove and hawk. Computer simulations reveal which strategies remain stable through population dynamics, accurately reflecting biological organisms.

In animals with high-stakes gambles, such as elephant seals fighting for a harem, the stable strategy involves violent fights. However, small birds in cold climates, such as the great tit, have higher penalties for wasting time instead of from injury. In heavily armored animals, only lost time poses a risk, not injury, so they fight “wars of attrition” (75), such as staring matches until one leaves to pursue other matters. Over time, stable strategies evolve.

Strategies vary as animals adapt to each other. Rather than oscillating strategies, however, evolutionarily stable strategies settle on the average effort that a resource represents, with small random variations. Any hint of giving up, such as a flickering whisker, would give an opponent the advantage: “So natural selection would quickly penalize whisker-flickering and any analogous betrayals of future behaviour. The poker face would evolve” (62).Telling lies or the truth directly would allow the poker face to win, making the latter strategy stable.

Maynard Smith extended the model to include asymmetric warfare, when one party has more strength than the other. Contestants can differ in their size, in their valuation of the prize, and in arbitrary conditions such as who arrives first. Each affects the evolutionarily stable strategy. The stable strategy can depend on which strategies already predominate. Going against the prevailing strategy carries costs, but occasionally the conditions could shift, such that a previously ineffective strategy becomes effective. Over time, however, conditions generally stabilize on a single strategy.

In some paradoxical strategies, such as intruders generally defeating residents, thus becoming residents themselves and less likely to win fights, the strategy would defeat itself. Instead, territoriality strengthens itself and becomes the norm. These paradoxical strategies therefore do not commonly occur.

An ethologist demonstrated territoriality in stickleback fish. Two males had built nests on opposite sides of a tank. When the ethologist put the fish in separate test tubes, the fish would assume aggressive or defensive positions, depending on which nest he put the test tubes closest to. The evolutionarily stable strategy develops regardless of particular species. Size and other aspects of fighting ability produce evolutionarily stable strategies. Smaller animals can see larger animals and flee, while larger animals can pursue smaller animals.

The possibility of injury complicates strategies. It could theoretically produce a paradoxical strategy of fighting only larger animals. However, this and other paradoxical strategies generally do not occur. One exception Maynard Smith points out is the Mexican social spider, which on occasion drive each other out of their homes en masse, reversing the territoriality strategy.

Crickets form general memories of their fights. Winning crickets become more combative, losing crickets less so. Over time, a hierarchy develops. Monkeys form specific memories of which opponents have won or lost. Evolutionarily stable strategies involve deferring to victors. Hens also remember who won their individual fights, forming a pecking order. These strategies reduce fighting and increase production such as eggs. Members of different species share fewer resources and have fewer conflicts. For example, robins and great tits do not contest territory, instead residing in overlapping regions.

Dawkins describes a lion and an antelope as competing for the resource of the antelope’s body (food energy). The genes of either animal fight for the meat for its own survival machine, producing a conflict of interest. Lions do not hunt other lions, and antelopes run away instead of fighting back against lions. Evolutionarily stable strategies favor not fighting against lions. Strategies often vary according to combatant asymmetry.

A slight asymmetry gets expanded until it reaches stability, such as the large asymmetry between lions and antelopes: “I have a hunch that we may come to look back on the invention of the ESS concept as one of the most important advances in evolutionary theory since Darwin. It is applicable wherever we find conflict of interest, and that means almost everywhere” (67).

Dawkins criticizes those who refer to “social organization” as having biological advantage. Instead, he views social order as arising from selfish entities enacting evolutionarily stable strategies. He even extends this view to ecosystems as a whole. Likewise, genes have to compete amid other genes. Evolutionarily stable strategies at the genetic level could explain how complexes of genes work together, such as the sets of genes that control butterfly mimicry.

Comparing genes to oarsmen, if some oarsmen speak English and others German, then an evolutionarily stable strategy would favor English speakers congregating, and German speakers congregating separately, to ease communications. Likewise, selecting oarsmen solely on individual merit, appropriate groupings of left- and right-handers would emerge: “Selection at the low level of the single gene can give the impression of selection at some higher level” (69).

Genes evolve in the gene pool. Genes that survive reproduce more, becoming predominant. Dawkins makes more precise the notion that “good” genes build effective survival machines, by specifying that they reach stability:

The gene pool will become an evolutionarily stable set of genes, defined as a gene pool that cannot be invaded by any new gene. Most new genes that arise, either by mutation or reassortment or immigration, are quickly penalized by natural selection: the evolutionarily stable set is restored. Occasionally a new gene does succeed in invading the set: it succeeds in spreading through the gene pool. There is a transitional period of instability, terminating in a new evolutionarily stable set—a little bit of evolution has occurred (69).

Rather than a constant climb, evolution may step from stable plateau to stable plateau. Genes evolve on the basis of survival, producing coordinated transitions that appear as a unitary population. Genetic battles take place not only among animal bodies, but even among genes within a single animal body:

The vast majority of significant interactions between genes in the evolutionarily stable set—the gene pool—go on within individual bodies. These interactions are difficult to see, for they take place within cells, notably the cells of developing embryos. Well-integrated bodies exist because they are the product of an evolutionarily stable set of selfish genes (70).

A selfish gene is not simply one instance of the gene. Rather, it represents all of the copies of the same configuration of the gene. Genes thrive by making copies. They propagate by having survival machines reproduce: “The key point of this chapter is that a gene might be able to assist replicas of itself that are sitting in other bodies. If so, this would appear as individual altruism but it would be brought about by gene selfishness” (70).

One of the genes that can cause a human to become albino is recessive, requiring two copies for a person to express it: 1 in 70 humans has a single copy, while 1 in 20,000 has two copies and is albino. The albino gene could improve survival of its other copies by making its bearer act favorably towards albinos. Sacrificing one body would pay off for the gene if it saves numerous other bodies carrying copies.

However, the gene would have to produce two separate attributes: pale skin, and altruism towards others with pale skin, making such altruism unlikely. A gene could recognize copies of itself in a body performing comparable acts to what the gene’s own survival machine does, yet this would also be unlikely. Close relatives have greater likelihood of sharing genes:

It has long been clear that this must be why altruism by parents towards their young is so common. What R. A. Fisher, J. B. S. Haldane, and especially W. D. Hamilton realized, was that the same applies to other close relations-brothers and sisters, nephews and nieces, close cousins. If an individual dies in order to save ten close relatives, one copy of the kin-altruism gene may be lost, but a larger number of copies of the same gene is saved (71).

Hamilton calculated in 1964 the exact probability of close relatives sharing a gene. Certain genes may lack an obvious “good,” making them rare, yet appear often within a family. One’s sister has a 50 percent chance of sharing an otherwise rare gene. The gene came from a single copy in either parent, so any sibling has a 50 percent chance of inheriting it. Any child of a person has a 50 percent chance of inheriting a particular gene. On average, half the genes of one person will be present in a sibling.

In general, one can calculate relatedness by finding the nearest common ancestor of two people. The generation difference equals the sum of steps up and down between the two people, via the common ancestor. For example, one person has a father (one step) who is the grandfather (two steps) of another person (three total steps). Putting ½ to the power of steps yields the fraction of common genes, in this case (½)3 = ½ x ½ x ½ = ⅛.

One adds the genes from each common ancestor. For example, first cousins have two common ancestors (grandparents), each with distance 4, yielding a total relatedness of (½)4 x 2 = ⅛. A grandchild shares one common ancestor with a grandparent (the grandparent) with distance 2, yielding (½)2 x 1 = ¼.

At the distance of third cousins (1/128), shared genetics approximates random members of the population. For altruistic genes, second cousins would show a small effect (1/32), first cousins more (⅛), and siblings and parents or offspring yet more (½). Identical twins (relatedness = 1) share all genes.

Rather than a rigid distinction between family and others, genes gradually become less common the farther the distance between people. The normal rules of gene selection prevail:“We can now see that parental care is just a special case of kin altruism” (74).In actual situations, people face messier uncertainties than the theoretical calculations. For example, grandchildren often have longer expected remaining life than grandparents, so even though they share the same genetic distance, they have asymmetric interests.

Dawkins describes people metaphorically as “life-insurance underwriters” (76), gauging the expected outcomes for its genetic copies in different bodies. When the benefit of assisting a close relative exceeds the risks, altruism evolves. Calculations occur subconsciously, as when someone catches a ball without consciously calculating its trajectory:

The whole sum for any one of the alternative behaviour patterns will look like this: Net benefit of behaviour pattern = Benefit to self - Risk to self + ½ Benefit to brother - ½ Risk to brother + ½ Benefit to other brother - ½ Risk to other brother + ⅛ Benefit to first cousin - ⅛ Risk to first cousin + ½ Benefit to child - ½ Risk to child + etc. (76).

Dawkins gives the example of an animal stumbling on a patch of mushrooms. The net benefit of eating them is outweighed by the net benefit of making the food call so that relatives eat some of the mushrooms, for the genes: “I should give the food call; altruism on my part would in this case pay my selfish genes” (77).

Genes evolve approximations on the basis of past genetic survival. When environments change dramatically, these approximations fail and animals (including humans) suffer. Actual animals can usually only estimate the relatedness of each other approximately. Humans keep close track of formal kinship through records and names:

Human customs and tribal rituals commonly give great emphasis to kinship; ancestor worship is widespread, family obligations and loyalties dominate much of life. Blood-feuds and inter-clan warfare are easily interpretable in terms of Hamilton's genetic theory. Incest taboos testify to the great kinship-consciousness of man (78).

Animals, lacking formal knowledge of their relations, could statistically assist their genes by acting altruistically towards other animals they resembled or frequently encounter. Whales aid babies and injured whales, and dolphins have even aided a human. Baboons risk their lives to defend their troops. Chicks twitter when they find food, alerting other chicks. These could represent animals protecting their genes in any member they perceive as relations.

Dawkins describes adoption in animals as “a misfiring of a built-in rule. This is because the generous female is doing her own genes no good by caring for the orphan. She is wasting time and energy which she could be investing in the lives of her own kin, particularly future children of her own” (80). In some cases, female monkeys who lose offspring steal another baby to care for. Dawkins describes this as a double mistake, wasting the thief’s own time and freeing the time of the biological mother, or else evidence against the selfish gene theory.

Cuckoos and other “brood parasites” (80) lay eggs in the nests of other species, exploiting the rule that birds have evolved to care for small birds in their own nests. Because herring gulls evolved in conditions where their eggs remained in their own nests, these birds did not evolve to distinguish their eggs from other eggs, or even artificial eggs placed by a researcher. Chicks, by contrast, wander, so gulls evolved to recognize their chicks.

Guillemots lay eggs on flat rocks where they can roll, so these birds did evolve to recognize their own eggs. Group selection would predict that any hen could sit on any egg, however gene selection reveals that this would result in cheaters exploiting the system until it breaks. The evolutionarily stable strategy here is to sit on one’s eggs selfishly, not on other eggs altruistically.

The song birds whose nests cuckoos exploit have evolved in response, favoring eggs that have a particular appearance. The cuckoos evolved to have eggs ever more comparable to those of the song birds. Dawkins calls this an effective lie. The arms race results in mimicry difficult to discern, and eyes capable of discerning: “This is a good example of how natural selection can sharpen up active discrimination, in this case discrimination against another species whose members are doing their best to foil the discriminators” (81).

Naturalists estimate the relatedness of different animals. For example, lions may have average relatedness of 0.22 between random males, and 0.15 between random females: “That is to say, males within a pride are on average slightly less close than half brothers, and females slightly closer than first cousins” (82). Lions do not know these numbers, yet the figure affect how genes evolve. Animal behavior approximates expected returns in gene survival.

Altruism depends on animal estimates of relatedness, rather than on the exact underlying relatedness: “This fact is probably a key to understanding why parental care is so much more common and more devoted than brother/sister altruism in nature, and also why animals may value themselves more highly even than several brother” (82). Because of the possibility of mistakes and frauds, genes for kin altruism cannot quite compete against genes for selfish behavior.

Mothers lay eggs or give birth, therefore knowing more about their offspring than the fathers do. As such, mothers should invest more parental care than fathers do, and likewise for maternal grandmothers over paternal grandmothers:

Similarly, uncles on the mother's side should be more interested in the welfare of nephews and nieces than uncles on the father's side, and in general should be just as altruistic as aunts are. Indeed in a society with a high degree of marital infidelity, maternal uncles should be more altruistic than 'fathers' since they have more grounds for confidence in their relatedness to the child (83).

Parents are older and have more resources to care for a child than vice versa, despite their common level of relatedness: “The truth is that all examples of child-protection and parental care, and all associated bodily organs, milk-secreting glands, kangaroo pouches, and so on, are examples of the working in nature of the kin-selection principle” (84).

Dawkins distinguishes between bearing new babies versus caring for living babies. Caring for a baby calls for determining how closely related the baby is, and how much food it needs. Bearing a baby calls for determining whether to reproduce. The two possibilities compete for resources. Different mixes of caring and bearing can become evolutionarily stable strategies:

The species with which we are most familiar-mammals and birds-tend to be great carers. A decision to bear a new child is usually followed by a decision to care for it. It is because bearing and caring so often go together in practice that people have muddled the two things up. But from the point of view of the selfish genes there is, as we have seen, no distinction in principle between caring for a baby brother and caring for a baby son (85).

The group selection controversy has largely focused on child-bearing. Group selection, as a theory of “population regulation” (85), argued for individuals altruistically limiting their birth rates. Populations bearing more than two babies per couple grow exponentially. Having babies earlier in life also produces a larger population.

Dawkins illustrates with the population of Latin America. Extrapolating birth rates 500 years into the future, humans would cover the entire continent: “By 2,000 years, the mountain of people, travelling outwards at the speed of light, would have reached the edge of the known universe” (85). Realistically, disease and war and birth control limit population growth. While some humans have foresight, most animals act for the selfish gene, and would reproduce without regard for the planet. Animals generally do not reproduce without limit, however. Most die of starvation or other causes, preventing population explosions.

Group selectionists and gene selectionists agree that animals also regulate their birth rates. They disagree over whether this regulation happens for altruistic or selfish reasons. According to group selection, parents limit offspring “for the good of the species” (85). While natural selection generally favors producing more offspring, group selection could theoretically favor groups that reproduce less, preserving food. Social concepts like territoriality and dominance could be interpreted as symbolic licenses to reproduce. V.C. Wynne-Edwards even argues that large animal groupings such as flocks gather to conduct population censuses.

According to the selfish gene theory, as argued by ecologist David Lack, animals produce offspring according to their own genetic coding. For example, birds can have genes to produce 1, 2, 3, 6, 12, or some other number of eggs at a time. The genes that make more copies of themselves will produce the appropriate number of eggs. Four could be better than three. However, a bird cannot produce an unlimited number of eggs, implying some maximum. A trade-off arises between bearing and caring.

Instead of serving group interests, the eggs serve individual interests. An evolutionarily stable strategy yields the number of eggs producing the most surviving offspring. Instead of altruistically limiting eggs for the common good, animals selfishly limit eggs for their own genetic good. Offspring cost time and food to produce. These limits imply the number of offspring that natural selection produces, maximizing resources: “Individuals who have too many children are penalized, not because the whole population goes extinct, but simply because fewer of their children survive” (90).

Modern humans can have more offspring than they can feed, because of government support. Before the state, these offspring would have starved to death. Dawkins argues that a welfare state requires contraception to prevent even worse problems from overpopulation: “The welfare state is perhaps the greatest altruistic system the animal kingdom has ever known. But any altruistic system is inherently unstable, because it is open to abuse by selfish individuals, ready to exploit it” (90).

The selfish gene explains territoriality, dominance, and other social concepts as it does clutch size. Instead of working for the good of the group, animals find the strategies that improve their own genetic chances. Instead of another risky fight, losers postpone for more resources: “A seal who leaves the harem-holders unmolested is not doing it for the good of the group. He is biding his time, waiting for a more propitious moment” (91).

Overcrowding reduces birth rates. Either group selection or gene selection could explain this. In experimental mice given plentiful food, after breeding to crowded conditions, the mice reproduced less. Group selection would argue that mice sense food limits and reduce reproduction. However, the experiment would have continued providing food. The selfish gene predicts that mice would instead sense overcrowding as a predictor of future famine, reasonably calling for the reduction of births.

Likewise, group selection or gene selection could predict large gatherings of animals, for different reasons. For example, altruistic birds could gather to conduct a census to find the number of offspring to have. Alternatively, selfish birds could gauge the amount of food expected to feed their offspring, even mimicking multiple birds to intimidate other birds to have fewer offspring with competing genes. Dawkins writes that the selfish gene theory can explain group selection theories of reproduction generally:

Our conclusion from this chapter is that individual parents practise family planning, but in the sense that they optimize their birth-rates rather than restrict them for public good. They try to maximize the number of surviving children that they have, and this means having neither too many babies nor too few (93).