Unto Others: The Evolution and Psychology of Unselfish Behavior
Source: https://archive.org/details/untoothersevolut0000sobe ↗
The most rigorous contemporary defence of group selection — the idea that natural selection can operate on groups, not just individuals or genes.
Sober and Wilson dismantle the orthodoxy that had dismissed group selection since the 1960s, showing that the theoretical and empirical case is far stronger than the gene-centric mainstream acknowledged.
The book is central to the debate about evolutionary altruism: how genuinely unselfish behaviour can evolve and persist.
Written with philosophical care but engaging real biological data, it reopened a question most biologists considered settled.
For anyone interested in cooperation, organisations, or collective behaviour, this is the foundational evolutionary argument.
Where Sober and Wilson open the debate, Okasha later formalises it.
Central argument
Sober and Wilson rehabilitate group selection — the theory that natural selection operates not only on individual organisms or genes but on groups as units — arguing that the gene-centric consensus that displaced it since the 1960s rested on flawed reasoning rather than decisive evidence. Their central thesis is that genuinely altruistic behaviour, where an individual incurs a real fitness cost to benefit others, can evolve and stabilise when between-group competition outweighs within-group competition. They demonstrate this through both philosophical analysis of multilevel selection theory and a survey of empirical cases across species, making the case that dismissing group selection was a premature theoretical closure rather than a scientific verdict.
Critique
The book's most persistent tension is that multilevel selection and gene-level selection are often mathematically equivalent — a point critics like Maynard Smith and Dawkins pressed hard — which raises the question of whether Sober and Wilson are offering a genuinely superior causal account or a reframing that purchases intuitive clarity at the cost of theoretical parsimony. If the two frameworks make identical predictions, the choice between them risks becoming metaphysical rather than empirical, and the book does not fully resolve when multilevel selection is doing distinct explanatory work versus when it is simply a more sociologically appealing description of the same dynamics.
Why it matters for product
The book's core structural insight — that what is optimal for individuals within a group can systematically undermine what makes the group competitive against other groups — maps directly onto how CPOs should think about incentive design across autonomous product teams: local optimisation on individual team metrics (velocity, OKR attainment) can erode the collective coherence that determines whether the product organisation outperforms competitors. More concretely, Sober and Wilson's argument that altruistic behaviour stabilises only when between-group selection pressure is strong enough gives a principled basis for when to tighten cross-team accountability structures and shared outcome metrics, rather than relying on culture alone to prevent internal defection.