Biology should not dispense with sexes.

It has been argued that biological sex, defined by the production of one or other type of anisogamous gametes - eggs and sperm - is "an incoherent category, one that has perhaps outlived its use." The idea of biological sex is an outmoded construct that should be 'eliminated' by scientific progress, like the four humours of medieval medicine. Furthermore, the distinction between biological males and females should be replaced by a "multivariate and nonbinary" categorization scheme or by "reproductive dimorphism", a model in which some species have two sexual morphs, which cannot be identified as male and female. Dispensing with sexes, the argument goes, would both lead to better science and support progressive social change. Here, we argue that dispensing with sexes would be a grave scientific error, depriving biology of one of its most powerful tools for explaining biological diversity. We identify the error underlying proposals to dispense with sexes. These proposals assume that the scientific value of sexual categories depends on there being some essence of femaleness shared by every female organism and which causes females to have essentially female characteristics, and likewise for males. But such essentialism, or "typological thinking", about biological categories was rightly rejected by Ernst Mayr and the other founders of the modern synthesis long ago. Essentialism plays no role in the models that explain the evolution of sexes and sex-associated phenotypes. The definition of sexes by anisogamy, however, is at the heart of these models. Hence, it is essentialism and typological thinking that we should dispense with, not biological sexes.