Hamilton's rule

Hamilton's rule is the book's framework for explaining how altruistic behavior can evolve through its effects on relatives: the benefit to a recipient counts in proportion to the likelihood that the recipient shares the relevant genes with the actor. It turns apparent self-sacrifice into a question about the relative reproductive prospects of genes, while leaving the behavior itself to be explained through imperfect rules and complex development rather than conscious calculation.

Relatedness and the threshold for altruism

The argument begins with relatedness as a measure of the chance that two individuals share a gene. A parent and child are related by one-half; brothers and sisters also average one-half, though particular pairs may share more or fewer genes by chance. The book offers a practical method for estimating relatedness: identify the most recent common ancestors, count the steps up and down the family tree between the individuals, and use that generation distance to calculate the average. This makes kinship a graded quantity rather than a simple division between relatives and strangers.

That gradient determines how much benefit to others can compensate for a cost to the actor. A suicidal act that saves five cousins would not, on average, make the responsible gene more numerous, whereas saving five brothers or ten first cousins could. The corresponding thresholds rise as relatedness falls: more than two siblings, children, or parents; more than four half-siblings, aunts, uncles, nieces, nephews, or grandparents; or more than eight first cousins. In each case the proposed gene can persist through the relatives saved, compensating for its loss in the altruist.

The example of identical twins sharpens the point: a gene carried by both makes one twin's sacrifice for the other, genetically speaking, as favorable as saving oneself. The argument does not claim that such acts are commonly observed. It notes instead that nine-banded armadillos are born as identical quadruplets and suggests that strong altruism among them would be worth investigating.

Parental care, kinship, and social categories

Parental care is presented as a special case of kin altruism, not a genetically unique relationship. An adult is, on average, as closely related to an orphaned baby sibling as to its own child. A gene influencing an older sibling to care for a younger one could therefore spread on the same relatedness basis as a gene influencing parental care. The book immediately qualifies this comparison: sibling care is much less common in nature, and further complications matter. But the genetic accounting does not privilege parenthood over siblinghood merely because parents directly pass genes to children; siblings can carry identical copies inherited from the same parents.

The text distinguishes this account from group selection and individual selection, and notes that some writers use the term kin selection for selection within families. It also warns that the terminology can confuse the issue: E. O. Wilson classifies kin selection as a special case of group selection, a use the author says risks muddling distinctions between levels of selection.

Human kinship adds another layer. People identify relatives through names, formal marriage, written records, memory, customs, and ritual; ancestor worship, family obligations, blood-feuds, and inter-clan warfare can all be interpreted in relation to kinship. These cultural understandings are not identical to genetic relatedness. The book also treats incest taboos separately from altruism: their proposed genetic advantage concerns harmful recessive genes associated with inbreeding, not assistance to relatives.

Costs, prospects, and decision models

A simple relatedness fraction is not enough to compare every act. The recipient's age and prospects affect how much future genetic benefit can result from assistance. Saving a close relative near the end of life may have less effect on the future gene pool than saving an equally related person with much of life ahead. Grandparents and grandchildren share one-quarter of their genes, but if grandchildren have greater reproductive prospects, selection may favor assistance from grandparent to grandchild more strongly than assistance in the reverse direction. The text cautions that this age pattern is not universal: in species with high infant mortality, the relative prospects may be reversed.

The book extends the calculation by analogy with life insurance. An individual risks some of its own assets for another, weighing relatedness and the recipient's prospects. More strictly, the relevant estimate is not just life expectancy but reproductive expectancy, or the general capacity to benefit one's genes in the future. Altruism can evolve when the actor's net risk is less than the recipient's net benefit multiplied by relatedness, with the risks and benefits assessed in this actuarial fashion.

No animal needs to perform this arithmetic consciously. The book compares such behavior to catching a thrown ball: the action can function as if calculations had been made without the actor knowing the mathematics. It also imagines a computer model that lists possible behaviors and scores each by adding benefits and subtracting risks, weighting effects on relatives by their relatedness and initially setting aside additional factors such as age and health. The actor's own interests receive full weight because relatedness to itself is one; a brother's receives one-half, and a first cousin's one-eighth. The model selects the action with the highest net-benefit score, even if every option is unfavorable; doing nothing can win if its score is highest. The rule describes the selection logic, not a conscious computation performed by the animal.

Imperfect recognition and the limits of the rule

In practice, relatedness is often uncertain. If two individuals might be either half-brothers or full brothers, the useful estimate is an average based on the probabilities of each relationship. The book gives a case in which they certainly share a mother but have only a one-in-ten chance of sharing a father: the effective relatedness is 0.275, combining a ten-percent chance of full brotherhood with a ninety-percent chance of half-brotherhood. The selection calculation therefore depends not on perfect knowledge but on estimates that can be wrong.

The book asks how animals might behave as if they recognized kin without needing all-wise knowledge of their ancestry. A simple rule such as favoring individuals who physically resemble oneself could, under ordinary conditions, benefit relatives statistically; changed living conditions could make the same rule misfire. The text proposes racial prejudice as a possible irrational generalization of a tendency to favor resemblance, while emphasizing the indirect and imperfect character of such a rule. Genes do not issue a conscious instruction to help relatives: they influence the development of nervous systems and behavior, and selection favors a behavioral disposition only if genes associated with it persist more successfully than alternatives.

Adoption illustrates the costs of misfiring. In herds or troops, a female may adopt an orphan, sometimes called an aunt by observers even when there is no evidence she is related to it. The book treats many such adoptions as possible errors of a built-in rule: care for an unrelated orphan can divert time and energy from the female's own kin, including future offspring. The tendency need not be perfectly selective if such cases occur too rarely for selection to alter it. A more extreme case is a bereaved monkey mother stealing another female's infant and caring for it. The author calls this a double mistake because the adopter spends resources and may also relieve the real mother of child-rearing costs. The example is presented as a challenge requiring research into its frequency, the adopter's average relatedness to the infant, and the real mother's response—not as a settled refutation of the theory.

Significance and qualifications

Hamilton's rule organizes the book's explanation of kin altruism around effects on gene survival, rather than the welfare of a species or group as a whole. Yet its formula is not a claim that a single gene independently builds a behavior. Genes interact with one another and with the environment; a behavioral gene means one that influences nervous-system development so that a behavior becomes more likely. The wider discussion stresses that individual genes have many effects, that traits arise through cooperative developmental processes, and that natural selection works through the differential survival of genes in the gene pool.

The rule is therefore both a compact selection criterion and a guide to more complicated conditions: relatedness, costs, benefits, uncertainty, and future reproductive prospects all matter. Its examples show why apparent generosity may be favored, why the same inherited tendency can fail in unusual circumstances, and why human kinship practices cannot simply be equated with genetic relationships.