Situer l’analyse phylogénétique entre les sciences historiques et expérimentales

Abstract

This article offers a conceptual study of a scientific practice. Phylogenetic analysis is one of the main methods of evolutionary biology as it allows us to deduce the evolutionary relationships between different species or organisms. In current practice, these analyses are often performed with the help of molecular data, resulting in what are called molecular phylogenies. A major question is what is the best way to characterize this practice? We begin with a presentation of this method articulated in four stages: (1) identification; (2) alignment of homologous sequences (stemming from a common ancestor); (3) construction (4) interpretation of phylogenetic trees. We show that phylogenetic analyses are not experiments, and therefore do not belong to the “laboratory style”, as defined by Hacking. This practice does not correspond to a typical practice from the historical sciences, as described by Cleland. Although there is only a partial correspondence between phylogenetic analysis and these conceptualizations, we argue that each confrontation sheds light on distinct aspects of this practice. We also question the idea of a strong methodological separation between experimental sciences and historical sciences.