[fdsaaf]

Race absolutely correlates with genetic clusters. Look at this map, [1], and say that there's no correlation between genetic markers and continental origin. Forensic anthropologists can identify race very well using nothing but bone structure.

The idea that race is scientifically invalid is complete nonsense. You can argue that race doesn't correlate with phenotypical characteristics that we care about (okay, but it does though), but arguing that race itself is just in our heads is nonsense.

Is it any wonder that more and more people distrust science on issues like evolution, climate change, autism, and so on when scientists claim that you can't really tell whether someone is white, black, or asian?

[1] https://en.wikipedia.org/wiki/File:The_history_and_geography...

[grardb]

> The idea that race is scientifically invalid is complete nonsense. You can argue that race doesn't correlate with phenotypical characteristics that we care about (okay, but it does though), but arguing that race itself is just in our heads is nonsense.

You couldn't be more wrong.

My mother is Puerto Rican, and my father is mostly Irish (with some ancestors from other European countries). My 23AndMe results show that all of father's genes are from Europe, and my mom's come from Europe (mostly Spain, but with a tiny bit of Ashkenazi), Africa, and the Caribbean (Native Americans, likely the Tainos).

My genetic makeup is roughly 7% Native American, 14% African, and most of the rest is European; some percentage is inconclusive.

You may be able to guess that I mostly look white, and you'd be correct. However, with my mother being roughly half non-white and half white, what would you guess? I'll tell you right now that I wouldn't guess that she's from Europe, but she does have relatively light skin. Her sister, however, could easily pass for a black person.

To think that your phenotype can be determined just by looking at your genes is ridiculous, and it's clear to me that you have very little experience in diverse environments. Things like skin and hair color are complex and are determined by multiple alleles, and there's no way to know which ones are going to be dominant.

Even aside from genetics, the social construct of race is even more complex, and I challenge you to do more reading on why that's the case.

Edit: To add more fuel to this, most of my girlfriend's genes come from China and surrounding areas in the Southeast. If we have a child, what will they look like?

Roughly 3.5% Native American, 7% African, 39.5% European, and 50% East Asian. Tell me, what race will they be? What about their phenotype? There's simply no way of knowing. Scientifically, race just does not exist.

[fdsaaf]

Why would it be surprising that the child of someone who is white and someone who is half white might go on to appear mostly white? You yourself mention that your genetic makeup (presumably from 23andme or something) also suggests a "mostly white" appearance. In this case, there's a pretty good match between phenotype and genotype.

You absolutely can predict someone's phenotype from her genotype, roughly speaking. In cases of recent admixture, the exact gene expression can be uncertain. Just look at Mendel and his peas! Even this variable heredity has limits. There was a zero percent chance that you'd end up looking like a typical Japanese person.

I don't understand your point. I'm not claiming that the classic races are fixed for all time. I'm not suggesting that we can't arrange genes in new combinations. I am claiming that people today cluster in certain historically-contingent ways and that these clusters reflect the everyday understanding of race.

[grardb]

It seems like you don't understand my point because you've ignored the fact that my mother and her sister look they belong to different races, despite having roughly the same genetic makeup (about half white and half non-white). My result is not surprising; theirs are, however.

As another commenter said, what about Obama? I've met people who are half black and half white and end up looking like white people. Obama doesn't look like someone from Africa, nor does he look like someone from Europe. Because of what society determines as being white, most people would say, "he looks more black," which isn't really true, but regardless, you likely would not be able to tell just by looking at his genetic makeup. Many people in the same circumstance look wildly different from him.

If your argument is that when someone's genes all come from one area, then you can take a guess at what they might look like, then sure. But to apply that blanket statement to the entire world is just not the case. There are parts of the world where those lines get very blurred (e.g. Turkey, Nepal, Afghanistan), and plenty of people who are of "mixed descent" with different phenotypes from you would predict from their genes.

[ptaipale]

I don't see it as very surprising that your mother and her sister look like they belong to different races. That's how genetics works in case of individuals; siblings can get different sets of dominant and recessive alleles from their parents.

In terms of population, things then average out. There are indeed parts of the world where things are very blurred because of extensive human interaction across populations. If every place in word were like that, then there indeed wouldn't be what we call races. But most of the world is not quite like that, and particularly, in the course of history it hasn't been.

[mikeash]

Race and "continental origin" are only vaguely related.

For example: what race is the current President of the United States? Most would say "black" (or "African American" if they're trying to be politically correct). And yet his ancestry is half European.

This is hardly an isolated example. The average African American has about one quarter European ancestry. About 10% of African Americans have a majority European ancestry.

Yes, you can use genetics to divide up humans into related groups. But those groups will not match with how we divide up races.

[ZeroGravitas]

Could you list what races you think exist, and some kind of paper that establishes a scientific method for where the dividing lines in the genetic gradients are drawn and why such a line needs to be drawn at all? That is, going from DNA to races, and not vice versa. Is there still an Irish race for example?

Also, is it worth pointing out that the people who don't believe in climate change, think vaccines cause autism, think evolution and fossils are a hoax etc. could make your exact claim about how scientists not coming clean on their pet subject and rather intentionally misrepresenting the facts to the public for whatever nefarious reason makes it their own fault that no-one believes in science.

How do you convince yourself you're not one of them and therefore actually the cause of the very problem you lambast scientists for? I'm sure they have books with nice diagrams proving them correct too.

[fdsaaf]

> Could you list what races you think exist, and some kind of paper that establishes a scientific method for where the dividing lines in the genetic gradients are drawn and why such a line needs to be drawn at all?

Let's define our terms. By "race", I mean the classic continental groupings of people whose ancestors come from areas centered on Europe and the near east, east Asia, and Sub-Saharan Africa. When we say "race", we're talking about that category to which people readily self identify and to which we can easily assign others. When we say "race exists", it means that these categories are not arbitrary.

More generally, they're big ancestor clusters. You can see the clusters yourself if you take genome corpuses and run principal component analysis (or other grouping algorithms) on them. If you select k=3, the classic continental races come out of the data.

I'm not sure how much more real a taxonomic classification scheme can get. We use the same genomic approach for organizing the rest of life.

With more groupings, you get finer-grained population clusters nested inside the larger ones. If you look for enough clusters, you start seeing an "Irish" grouping. Can we agree that, say, the Irish, the Italians, and the Slavs are distinct hereditary groups? Can we agree that they're more similar to each other than, say, the Irish are to the Pygmies?

You could, in principle, put everyone into her own cluster. Sure, at k=7e9, race doesn't exist. But that's not a very useful classification scheme, because it ignores the reality that there are high-level classifications that we can see with our own eyes.

> How do you convince yourself you're not one of them and therefore actually the cause of the very problem you lambast scientists for?

That's an excellent question. Epistemology is hard. The best we can do is try to explain observations using the best-predicting theories we can find. I reject the "race does not exist" theory because it fails to explain observable facts about the world. This theory requires, in order to explain our observations, elaborate systems of oppression. It's full of epicycles. Even so, it fails to predict the result of studies like the Minnesota Transracial Adoption Study.

The "race corresponds to allele clusters" theory makes better predictions. It explains heritable and persistent differences in measurable characteristics. It agrees with genomic observations. It requires no hidden assumptions. This latter theory isn't politically correct, but this status can't affect its truth value. We delude ourselves about things all the time.

Look: the traditional continental race classification scheme is a crude folk theory. It's very embarrassing for science when even a crude folk theory makes better predictions than the best theory to come out of the academy.

Eppur si muove.

[ZeroGravitas]

So you seem to agree more with the "race is a social construct" people than it first appears since all the boundaries you talk about are fuzzy and arbitrary, you even suggest that people must be able to distinguish them with their own senses for them to be meaningful, and that if it's not a useful tool to humans then it's pointless. All of which sounds very social construct-ish to me.

Possibly you're just talking past each other, and don't fundamentally disagree at all.

[fdsaaf]

When you take a big corpus of genomes and split them using impartial mathematics into similarity-clusters, you get clusters that almost perfectly match the continent-scale races that people self-report. That's not "arbitrary". Furthermore, people in these groups differ in characteristics that are measurable and important.

When I say that "race exists", I mean that there are real differences between the k={3,4,5} groupings and that almost always tell someone's group affinity by sight alone. The existence of marginal cases doesn't somehow invalidate the reality and utility of the high-level groups.

When people say "race doesn't exist", the general public reads that as "there are no differences on average between people from the various continents", and this claim is not only false, but it's so false that one doesn't need sophisticated instruments to tell.

I believe that this confusion is deliberate and is part of a misguided attempt to eliminate bigotry by delegitimizing the classification schemes upon which bigotry is built. This strategy is doomed, because you can't take away people's eyes and ears.