The Structure of Evolutionary Theory
Fifteen hundred pages, published the year Gould died -- his magnum opus and lifetime summation.
The book is an extended argument against the adaptationist programme: not every trait is an optimised product of natural selection, and the history of life is marked by contingency, constraint, and hierarchical structure rather than smooth gradualism.
Spandrels, exaptation, punctuated equilibria, species selection -- all the threads Gould pursued across forty years converge here.
It is not an easy read, and Gould's critics found it self-indulgent, but no serious student of evolution can ignore the alternative architecture he proposed.
The parallel to how we think about organisational design -- where not every feature is adaptive, and path dependence matters enormously -- is hard to miss.
Central argument
Gould's central argument is that orthodox neo-Darwinism, with its near-exclusive emphasis on natural selection acting on individual organisms, is an incomplete and distorted account of evolutionary change. He contends that constraints (structural, developmental, historical) limit what selection can shape, that many traits are non-adaptive byproducts ('spandrels') or repurposed structures ('exaptations') rather than optimised solutions, and that evolution operates hierarchically — at the level of genes, organisms, and species simultaneously. Punctuated equilibria, his earlier thesis with Eldredge, is embedded in this larger framework: stasis and rapid change, not gradualism, characterise the actual fossil record.
Critique
The book's central tension is that Gould argues forcefully for contingency and pluralism in evolution while constructing an extraordinarily systematic, almost totalising theoretical edifice — a form of intellectual contradiction his critics, including Dawkins and Dennett, exploited with some justice. More seriously, the hierarchical selection framework, particularly species selection, remains empirically contested; the mechanisms are often asserted with rhetorical confidence that outruns the available evidence. A thoughtful reader might also note that dismantling adaptationism as a heuristic is easier than replacing it with one of comparable predictive power.
Why it matters for product
The concept of spandrels is directly usable in product retrospectives: features that accumulate in a product are often structural byproducts of architectural or organisational decisions, not intentional adaptations to user needs, yet they accrete legitimacy over time simply by existing. Gould's emphasis on path dependence and constraint is a corrective to the CPO instinct to treat every product decision as reversible and optimisable — legacy systems, inherited team topologies, and technical debt are evolutionary constraints that selection (market feedback) cannot simply dissolve. Punctuated equilibria also reframes how to interpret product metrics: long plateaus followed by discontinuous shifts are not signs of failure but may reflect the actual structure of how products and markets co-evolve.