: Why can’t DNA tests and Genetics determine race & ethnicity when it can tell you where your ancestors are from, and if not what does?
In: 0
Because race and ethnicity are social constructs and not biologically determined.
That doesn’t mean they aren’t real and have no effect in our society, they very much do. But race and ethnicity are not biologically defined.
Genetics might say your genotype is 17% African. Does that make you “black” or “not black”? The answer to that question is not scientific, and most racial identity advocates say it’s more important how you “identify” than what the genetics say.
Until race or ethnicity are completely redefined, in ways that seem completely the opposite of current understandings, DNA isn’t going to be the answer to racial or ethnic questions.
They can and can’t. There are most certainly differences between people’s genomes that correlate to ancestry in particular places around the world.
The more arbitrary social construct part comes in when you start trying to label these differences. For instance, how many polymorphisms does a person need to be viewed as belonging to some specific ethnic group?
It’s the difference between saying “you probably had ancestors from [region]”, which is fairly clear-cut, vs “you probably had ancestors from [culture]”, because ethnicity includes a major social component.
The same reason we can’t make machines that “tell you” a color, you can only get wavelengths, a measurement that we interpret a certain way. Red or blue mean nothing mechanically, that’s emergent to us. Ethnicity and race is the same way, but genes are even more complicated. There is no such thing as a discreet race or ethnicity, there’s groupings and proportions of genes we have decided to interpret that way.
Because our DNA is ancient. You might have markers from Asia, but that just means somewhere in your ancestry that DNA was introduced. Doesn’t mean it was recent enough to affect your outward appearance or where you grew up (both of which build your personal identity of race/ethnicity)
A lot of responses have touched on the societal reasons it is impossible to empirically determine race or ethnicity from DNA so I thought I would also touch on how the science works and how ancestry results can be misleading.
The ancestry percentages that you receive from a test like 23andMe are not a percent out of 100. If your results say that you have “60% European ancestry” they aren’t saying that 60% of your DNA is European. They are calculating the likelihood that your genome came from a specific region. So the percentages are actually saying “there is a 60% chance that your genome comes from Europe”.
To calculate these percentages they look at specific regions of the genome that are known to have variation between people (99.9% of you DNA is the same as everyone else so there are relatively few places where variation occurs). For instance, a specific location in your DNA might read A – **T** – **C** and mine might read A – **T** – **A.** It might just so happen that having a **T** in the second spot is more common in people from Europe and having a **C** in the 3rd spot is more common in people from Asia. So 23andMe would say that your ancestry is 50% European and 50% Asian. That does not mean that you are half European and half Asian, it means that there is a 50/50 likelihood that your DNA is European OR Asian (the actual math is way more complicated, if interested look up [Random Match Probability](https://dnaconsultants.com/random-match-probability/) and [likelihood ratios](https://www.forensicmag.com/3425-Featured-Article-List/576342-How-to-Use-the-Likelihood-Ratio/)).
This is similar to how DNA “matches” are calculated in a forensic setting.
Source: am Forensic Biologist
Imagine you are tasked with reading thousands of books written in English, you don’t speak English or know anything about nationalities, but you are cataloging and counting the words.
You can realize pretty quickly there is a whole segment of English books that use the phrase “Ayuh” so you lump those into category A, whatever the means. Another segment of books uses the term “colour” repeatedly, so that’s category B. A final group of books contains the phrase “smth” so that’s category C.
Now an analysis can see “Colour” and understand that’s probably books from English, Indian, Pakistani, or Australian writers so that’s a bit of a broad swathe of “nationalities”. It’s hard to peg this one down specifically.
The “Ayuh” phrase is actually determined to be very specific to the American states of Maine, New Hampshire, and Vermont. So that’s very specific.
Finally the “smth” phrase is the hardest to place, it pops up a lot in Central Europe, South American, Asia, and Africa and the analyst can only determine it’s a phrase taught in English-as-a-second language classes as an abbreviation for “something”. So it’s super broad and the hardest to lock down to a specific place.
The genetic testing works similar to the above. You can just look for patterns, even without knowing the *meaning* of the patterns and then rely on analysis to link a pattern to a source.
Latest Answers