Computer Science Meets Evolutionary Biology: Pure Possible Processes and the Issue of Gradualism

This paper investigates the relations between biological evolution and computer simulations of evolving entities through natural selection. It argues that what is proper to algorithmic evolution is that the selective dynamics of one modeled entity – for ex. genes, or species – is happening in the simulation with no immediate entangling with other levels of a hierarchy, unlike in biological evolution, where all the levels of the biological hierarchies are present together and their selective dynamics are entangled. This amounts computer simulation to propose “pure possible processes” of evolution, i.e. processes for which we know what kind and level of selection is at work. Algorithmic investigation therefore suggests processes as candidate explanations for the patterns of evolution we see out there. First, this fact allows one to solve issues which have been recently raised about the validation problem for simulation; second, in those conditions computer science is also likely to suggest new kinds of evolutionary processes whose outcomes would be discontinuous patterns of evolution. Drawing on recent work by Richard Watson, the last section of the paper will therefore consider how the longstanding issue of gradualism vs. discontinuities in evolutionary theory can be reassessed on the grounds of new insights provided by simulations like genetic algorithms. In conclusion I qualify the strong AL thesis according to which evolution by natural selection can be conceived of as an algorithm, and evolutionary biology as a branch of a general science of such algorithm.