Mismeasuring man thirty years later

Humankind has long been regarded as naturally divided into distinct groups or races, much like other animal species. Only in the second half of the twentieth century, through the work of Frank Livingstone, Richard Lewontin and Stephen Jay Gould, the race concept began to be questioned as a useful tool for understanding human biodiversity. Recent genomic studies have shown that we are all different, and that there are geographical patterns in human genetic variation. However, these patterns do not allow one to define clusters of biologically differentiated individuals, because each human population harbors a large share of the species’ genetic diversity, and each individual genome is a mosaic of DNA fragments of different origins. These data explain why studies of human morphology never led to an agreement about the number and definition of human races, with proposed races numbering from 2 to 200; people can be clustered in many ways, but variation within clusters is always large, and most alleles are cosmopolitan, i.e. present, at variable frequencies, in all continents. Race remains an important component of our social and psychological world, but envisaging our species as subdivided in genetically-differentiated groups leads to poor evolutionary inference and to errors in clinical practice.