5 MAY 2017 • VOL 356 ISSUE 6337 485 SCIENCE sciencemag.org PHOTO: HERO IMAGES/GETTY IMAGES By Harold P. de Vladar 1 and Eörs Szathmáry 1,2 W hy do worker bees give up their own reproduction in favor of other offspring of the queen? Does this make sense from a Darwin- ian point of view, which prescribes maximization of reproductive suc- cess? Ever since Darwin, evolutionary bi- ologists have time and again revisited this problem of how social behavior evolved. There must be some benefit to the donor in terms of fitness, otherwise the trait would vanish. However, how to evaluate this fitness benefit remains controversial because confu- sion about which models to use abounds. Most experts in evolution agree that the first detailed, and essentially correct, expla- nation for altruistic behavior was given by Hamilton in 1964 (1). Hamilton aimed to de- termine the conditions that allow altruistic behavior to spread. The altruist (the worker bee) cooperates by giving a benefit b to the recipient (other offspring of the queen) at a cost c to itself; both b and c are measured in terms of fitness, e.g., the expected num- ber of offspring. One can trivially guess that b > c must hold, but this is not enough. Ham- ilton’s insight was that relatedness (degree of kinship) r between donor and recipient must enter the equation. Thus, Hamilton’s rule (HR) is br > c. HR is derived from what many think in turn to be maximized in evolution: the inclusive fitness of an organism. Inclusive fitness underlies kin selection, namely, it is the number of offspring equivalents in the following sense: An actor is causally res- ponsible for some fitness contributions to all other individuals whom it has helped, at the cost of lowering its own fitness by c. Inclu- sive fitness is the weighted sum of all these additive contributions, with the weights being the relatedness values between the actor and the recipients. This approach has been applied far beyond social insects. For example, all somatic cells in a human body are maximally related to each other; this is how reproductive division of labor could evolve, so that only the germ cells make it to next generation. In the eyes of many, HR comes close to what physicists would call a natural law. However, Nowak et al. have argued that HR “almost never holds” (2). This extraordinary claim has created much debate. Four recent insightful papers (3–6) shed more light on the evolution of cooperation, kin selection, and the role of relatedness in the evolution of cooperation. They indicate that HR has no fundamental role unless interpreted in causal terms (3, 6), tackle issues associated with the notion of inclusive fitness and the role of relatedness in general (4), and reveal that group selection cannot always be re- duced to kin selection, and vice versa (5, 6). As Birch and Okasha (3) explain, re- searchers have inadvertently been arguing about different versions of HR that are not directly comparable. The authors introduce a distinction between different versions, of which we consider only two: the special case (HRS) and the general case (HRG). These versions both use the same form of HR (br > c), but each has a different interpreta- tion of the parameters b and c. In HRS, costs and benefits stem from the payoff matrix describing evolutionary encounters: who gets how much in terms of fitness upon meeting with partners of the same or other types, exactly as in Nowak et al.’s work (2). In contrast, HRG obtains b and c values by applying a statistical approach to the full model of a population. If the payoffs are additive, as in Ham- ilton’s original paper, then HRS is exact. Because additive payoffs are islands in an ocean of nonadditive ones, one may say that HR “never holds.” If the payoffs are nonaddi- tive then one can always design a particular EVOLUTIONARY BIOLOGY Beyond Hamilton’s rule A broader view of how relatedness affects the evolution of altruism is emerging 1 Institute of Advanced Studies Kőszeg, 14 Chernel utca, Kőszeg, H-9730 Hungary. 2 MTA Centre for Ecological Research, Evolutionary Systems Research Group, Klebelsberg Kuno u. 3, Tihany, H-8237 Hungary. Email: harold.vladar@parmenides-foundation.org; szathmary.eors@gmail.com PERSPECTIVES Unrelated individuals voluntarily collect garbage, adhering to the social norm that it is desirable to keep the environment clean. Recent research investigates how such norms and behaviors spread in populations. Published by AAAS on May 5, 2017 http://science.sciencemag.org/ Downloaded from