> even in the olden days of pseudoscience on race, obviously people didn't believe that races couldn't interbreed, as they observed such themselves. They did h

> even in the olden days of pseudoscience on race, obviously people didn't believe that races couldn't interbreed, as they observed such themselves. They did however believe that there were "purebreeds" of races; that you could objectively categorize people into various races. Mixing races wouldn't create new races, it would only create a mixture of races. As an analogy, you can create metallic alloys from different metallic elements. That won't create new metallic elements though, it will just create new mixtures of the pure elements. The strongest example of this view is Polygenism. Before Darwin, people were arguing about whether there was a single origin for people or multiple origins corresponding to the assumed races. This https://en.wikipedia.org/wiki/Polygenism#Scientific_polygenism https://en.wikipedia.org/wiki/Polygenism#Scientific_polygeni... The person you're responding to is trying to say that humans are a single species, and thus we can't be different races (species). But that is obvious in the modern context and isn't really what people are saying when they talk about race. Today, the argument is over: what are racial categories if they aren't species, and given a definition of race what does it actually tell us? > If we use the word race for human being it can only mean cultural or other such societal concepts meaning it is a social construction. > The fact remains that if you run PCA and clustering on genetic data ... you will get groups that generally correspond with Caucasian/SS African/East Asian/Amerindian The person you're responding to is being very imprecise. There is no one definition of race. It a squishy class of "categorizations I can define which separate europeans, amerindians, africans, and asians". When we create a racial categorization, we are flattening these various categorizations into a single categorization. It is a social construction to choose the most important categories as "europeans, amerindians, africans, and asians" regardless of the categorization you're using. As an example of the limitations of defining race from genetic clusters, Think of Obama in the U.S. He has a White mother and a Black father, but can you honestly say he is treated as equally White and Black in the U.S.? No, he is treated as predominately Black, even though the genetic view disagrees. You could say that the social definition is wrong and the genetic definition is correct, but you could equally say that your genetic definition fails to describe the social racial categorization.