Abstract

Universal Darwinism correctly identified that selection dynamics appear across domains—biological, cultural, epistemic, technological. But it inherited from biology an assumption that has limited its generality: that selection requires replication. This assumption generates persistent puzzles about what counts as a "replicator" and whether replication is truly necessary for evolutionary dynamics. We propose that the fundamental mechanism underlying selection is differential propagation under constraints—not replication. Replication is one implementation of this mechanism, not a requirement for it. This reframing dissolves the replicator problem, extends evolutionary logic to domains where copying is absent or fuzzy, and reveals selection as a consequence of constraint satisfaction dynamics formalized through free-energy minimization. The implications extend from physics through culture: wherever patterns exist in possibility space and constraints define viability, selection occurs. Replication accelerates and channels this process but does not constitute it.