Source
Hamilton (1964), “The Genetical Evolution of Social Behaviour. II”, LIT-506 (Journal of Theoretical Biology 7: 17–52), read in full in NOTE-410. It rests on Part I (LIT-497, NOTE-401). Fractions garbled in extraction are taken from the prose.
What was actually shown
The first principle (§3). “The social behaviour of a species evolves in such a way that in each distinct behaviour-evoking situation the individual will seem to value his neighbours’ fitness against his own according to the coefficients of relationship appropriate to that situation.” It is offered as a “generalized unrigorous statement” of Part I. The model is not extended mathematically.
The second principle (§4). The situations a species discriminates “tend to evolve and multiply in such a way that the coefficients of relationship involved in each situation become more nearly determinate”. Herring gulls accept strange chicks for the first days and reject them by the end of the first week, when chicks begin to wander. Ground-nesting Sooty terns learn their young within four days, and tree-nesting Noddies never do. Strangers are expected to be attacked or “relegated to the lowest social rank”.
Hymenoptera (§5). Full sisters by a singly mated mother are related by 3/4, more than a mother to a daughter. That predisposes workers to altruism. Multiple mating lowers it: with n equally contributing drones the mean sister relationship is ¼ + 1/(2n), about 0.31 for the honeybee’s n ≈ 8 (NOTE-410 computes this from the formula the prose describes).
Fights (§6). Sibs should fight to the death only if the winner gains more than 50%. Hymenopteran full sisters need more than 75%. Gregarious parasitoid larvae, which are siblings or clones, do not fight, and solitary ones kill rivals.
What could have come out otherwise. The natural history could have shown no link between relatedness and these behaviours. The author reports cases where it did not fit and explains them away (§7).
What this does not say
- That the generalization is proved. It is an analogy with the classical theory, “about as good” as that theory’s own principles (§2).
- That haplodiploidy explains hymenopteran sociality. The author reports multiple mating in Apis and Atta and Michener’s Australian halictines as “against our hypothesis”, and calls the evidence “somewhat contrary”.
- That anomalies were tested. §7 handles each by error or misinterpretation, with no criterion that could count against the principles (NOTE-410, C5).
- Anything about display or deference in contests. §6 concerns restraint between relatives, not signalling.
- Anything about morality. Fisher’s attempt to explain “the heroic ideal in barbaric human societies” is reported and not pursued.
Connections
- It extends Part I (THEORY-408), whose model it generalizes.
- Maynard Smith & Price (LIT-505, THEORY-466) explain restraint in fights between unrelated animals by retaliation. This explains it between relatives by relatedness. Their remark that very dangerous weapons “would be opposed by kin selection” is where the two meet.
- Trivers (LIT-510, THEORY-663) disputes the kin explanation of alarm calls given here and offers a self-interested one.
- The second principle is the biological root of the in-group partiality that morality-as-cooperation files under family and group loyalty (LIT-503). Hamilton’s mechanism is relatedness, though, and that account’s group domain rests on shared payoffs among non-kin (NOTE-410).
- The clones-and-colonies section is an early case of the units-of-selection question that Lloyd’s entry (LIT-160) treats.
Note of 2026-10-29: what holds it back
Read in full (NOTE-410), but its promote_when is unmet. The record holds no comparative test with relatedness measured and predictions fixed in advance, whether of kin discrimination, of escalation against measured r, or of worker altruism against measured sister relatedness. Hamilton’s evidence is anecdotal, and he reports multiple mating and the Australian halictines as running against the hymenopteran hypothesis.