Abstract
Gabora and Steel (Gabora L, Steel M. 2021 An evolutionary process without variation and selection. J. R. Soc. Interface 18, 20210334. [doi:10.1098/rsif.2021.0334]) claim that cumulative adaptive evolution is possible without natural selection, that is, without variation and competition. To support this claim, the authors modelled a theoretical process called self-other reorganization (SOR) that envisages a population of reflexively autocatalytic sets that can accumulate beneficial changes without any form of birth, death or selection, that is without population dynamics. The authors claim that despite being non-Darwinian, adaptive evolution happens in SOR, deeming it relevant to the origin of life and to cultural evolution. We analysed SOR and the claim that it implements evolution without variation and selection. We found that the authors, by design, ignore the growth and/or degradation of autocatalytic sets or their components, assuming all effects are beneficial and all entities in SOR are identical and immutable. We prove that due to these assumptions, SOR is a trivial model of horizontal percolation of beneficial effects over a static population. We implemented an extended model of SOR including more realistic assumptions to prove that accounting for any of the ignored processes inevitably leads to conventional Darwinian dynamics. Our analysis directly challenges the authors' claims, revealing evidence of an overly fragile foundation. While the best-case scenario the authors incorrectly generalize from may be mathematically valid, stripping away their unrealistic assumptions reveals that SOR does not represent real entities (e.g. protocells) but rather models the triviality that fast horizontal diffusion of effects can effectively equalize a population. Adaptation in SOR is solely because the authors only consider beneficial effects. The omission of death/growth dynamics and maladaptive effects renders SOR unrealistic and its relevance to evolution, cultural or biological, questionable.