When evolvability meets constraint: reflections on empirical fitness landscapes after Weinreich et al. (2006).

In 2006, a manuscript appeared with a provocative title: "Darwinian evolution can follow only very few mutational paths to fitter proteins." Twenty years on, I argue that it was a landmark study, one that promoted a tractable model system-the empirical fitness landscape-for theoretical questions in population and evolutionary genetics, including the dialectic between evolvability and constraint. Drawing on a reanalysis of the original data and a synthesis of the subsequent literature, I survey areas of evolutionary genetics that have developed in its wake. Its deepest legacy, I argue, lies less in its headline finding than in how it bridged canonical population-genetic theory with experimental practice, helping inspire a research program that interrogates the mechanistic bases of adaptive evolution and remains in active conversation with applications spanning biomedicine, biological engineering, and artificial life.