[atlantic]

Let me give you an example. Here's a concept: countries. You've got the US and Mexico, and there is a border between them. You could probably find an infinite number of points right on the border, that can't be said with certainty to belong to either country. And based on those outliers, you could argue that the whole concept of country is invalid, and that countries and borders should be abolished. You see where I'm going with this?

[whelp_24]

This is an ironic example. Countries are a social/political construct. Whether a country even exists (Taiwan, micronations, various other countries) depends on who you ask, as does the boundaries of the border. Countries do not exist in any physical sense. Indeed countries and race are very tied to each other and exist about equally. Even if you pretend that culture is a physical thing, you'll find that cultures don't align with the borders of Nation-states properly either.

Society exists, and it has (often serious) repercussions but it isn't an inherent physical property of geography or genetics.

[atlantic]

I think you're missing the point. The question isn't whether concepts refer to existing things or not, but whether they are valid and useful.

Nothing general ever exists - only the particular. The number two doesn't exist, although you can find twos in nature. Basic Aristotle. Similarly, race doesn't exist, but you can find instances of races in nature.

That's why arguing that a concept is invalid because it is "only" a social construct is disingenuous. Concepts are always constructs; they are never "natural" in themselves. What determines their validity is their usefulness, which is related to how closely they mirror the structure of human experience.