A class of identifiable phylogenetic birth-death models

Thousands of publications have utilized phylogenetic birth-death models to make inferences about fundamental evolutionary processes like speciation, extinction, and the spread of pathogens. A recent study has called these findings into question, by showing that a variety of different evolutionary hypotheses are consistent with any given dataset and cannot be distinguished from one another regardless of how many additional data are collected. Here, we qualify this grave result by proving that a parsimonious subset of the evolutionary hypothesis space can be inferred from data.