Are Meds and Nordics Different Sub-Races?

A look at neutral and "effective" genetic variation

To a European eye, it’s usually not hard to guess where someone is from. Within Europe we intuitively distinguish “types”: darker, shorter, Mediterranean-looking people; taller, paler, blond or reddish Nordics; Central Europeans somewhere in between. Skin, eye and hair colour, body size and facial features all vary enough that we feel we can often tell an Italian from a Swede at a glance.

The same happens within East Asia. Many people can see a difference between, say, northern Chinese or Koreans and Southeast Asians, even if they couldn’t necessarily label each group correctly. By contrast, telling apart members of a different major race or continental cluster is much harder: a European will be much worse at distinguishing between different African ethnic groups than between Italians and Norwegians.

Some readers asked me whether these visible differences reflect deeper genetic clusters. Do Northern and Southern Europeans form two distinct “sub-races”? What about Northeast Asians versus Southeast Asians? If we take the genetic data seriously, do we see two clean clusters in each case, or just a gradient of variation inside a larger Eurasian cluster?

In this post I try to answer that question using the same Fst-based clustering approach I used in my previous analyses. Instead of relying on impressions from faces and phenotypes, I look at genome-wide Fst within and between four macro-groups – Northern Europe, Southern Europe, Northeast Asia and Southeast Asia – and I ask a very strict question: do the between-group differences cleanly exceed all the within-group differences, or not?

I will also examine differences that are substantively meaningful, focusing on genetic variation known to affect phenotypic traits such as body size and personality.