A surge in new species descriptions must be accompanied by an equal amount of healthy skepticism. Herein, we critique the current approach to species delimitation for the genus Anoura and assess the methodology used in these studies. It is not uncommon for studies committed to the delimitation of species to incur in a mismatch between their underlying epistemological perspective and the nature of species as real entities or ontological individuals. This is the reason why these studies must capitalize on the statistical paradigm to ascertain the degree of vagueness upon their particular approximation to real or purportedly real species. It is common for species to have fuzzy boundaries and numerous sources of variation. Furthermore, as multi-organismal entities, species deserve a more cautious action to their delimitation than purely verbal descriptions from the point of view of a single observer. We highlight the need for quantifiable methods that provide clear perspectives on the magnitude of overlap and variability within and among species. We argue that the delimitation of complex entities as evolutionary species must be framed under the paradigm of hypothesis testing and measurable and concrete estimates of character states. Quantitative hypothesis testing should be a requirement for the practice of systematics, taxonomy and species delimitation. Species are not mind constructs but complex ontological individuals awaiting discovery by means of precise statements of uncertainty.