astralcodexten.com
I'm generally a fan of "blurry" definitions where something can qualify as X if it fulfills a few of many criteria. I think trying to create hard rules around blurry areas like race and culture is fool's errand, and Scott does a great job laying out how overly strict definitions can go wrong.
Jump in the discussion.
No email address required.
Notes -
Scott seems to not understand. Race is still a social construct. There are genetic variations among different populations, but this doesn't mean the categories of race are not socially constructed. Who decided we are going to define one race white and another black, based on skin color? He uses the example with Jews, but this makes no sense since their categorization of race is different from the Western categorization. These racial categories have a purpose and are useful for a variety of reasons, but he's not making a convincing point that racial categories are not socially defined. Certain racial categories are fuzzier and an American invention: whites and blacks.
Do you also dispute the wavelength basis of color? It fits in perfectly:
Put plainly, everything is a fuzzy socially-defined category, even the categories used in the hardest of hard physics. Bringing up this argument for genetics only is an isolated demand for rigor.
I am not sure what your counterargument is for. I am not disputing the existence of genetic variation in different populations. I am pointing out race (mostly Western system) is a social construct. I am not even disputing the utility of these categorizations. I am just saying Scott makes the conflation that social construct means it doesn't exist.
For your example... Color is a spectrum. It's not that one color definitely ends here and starts there. Language is a limiting factor. There are terms for different types of colors that aren't existent in Anglo Saxon. Useful categories, but doesn't mean they aren't socially defined. Socially defined doesn't necessarily mean there are no differences between a certain part of the spectrum compared to the other.
Though I understand you are just using an example, it's not a very good one since genetic differences between races aren't always clear. I am pretty sure ethnic differences are larger within one race than between different races. Way less clear than physical phenomena in hard sciences. Defining it by skin color or certain phenotypes is just one way of doing it.
/images/1709960115303152.webp
That's a very common conflation in my experience, which makes it a valid target for counterarguments.
The socially constructed definition of race includes genetic information, which means that it is a physically-grounded system instead of an arbitrary one. This puts limits on how much society can change the definitions without going off the rails.
First, differences between colors aren't clear either. Light with uniformly-random wavelengths is widely agreed to be "white". What about light with 1/3 each 450 nm, 550 nm, 650 nm? It might appear identical or different depending on the situation, so we've created the color rendering index to deal with that. What about fluorescent objects? They reflect visible light in a way that's easily-describable using standard terms, but they also create some extra by converting UV light. Category differences not being clear is completely normal, and there's nothing special about genetics in that sense.
Second, it doesn't look that bad? Look at the graphs Scott included just above your quote: they sure look like clusters to me, and the line-drawing isn't too egregious. Also remember that we're looking at a 2D projection of an N-dimensional analysis, so some more differences will show up in the later principal components.
I've heard that statement before, but I still haven't got a good explanation of what the factual claim is supposed to be. My attempts all end up in nonsense.
My first thought was "a random pair of coethnics is more genetically different than a random pair of non-coethnics humans", which seems trivially false. My second was "a random pair of coethnics is more genetically different than a random pair of archetypal members of each race", which seems like a category error for the comparison and also plausibly false (see the graph again: races have size 0.2ish, while the distances between their centers are 0.35ish).
What do you mean in a hard statistical sense by that statement?
I am very confused because it seems there is confusion between map for territory. You talking about genetic material being different between different populations and sub-populations doesn't mean that race is not socially constructed. It seems we agree that social construction of race is not useless nor does it mean that there are differences between groups of people?
A black man and a white man has more similarities in genes, just different expressions of phenotypes, which often is what is used to classify race to begin with.
My understanding is also that differences within one race are higher than between. African populations have more genetic diversity within different sub-populations as an example.
https://ksj.mit.edu/tracker-archive/new-study-confirms-africans-are-most-gen/
https://pubmed.ncbi.nlm.nih.gov/12493913/
Of course it is, just like everything else. The mere fact that a category is socially constructed is utterly unremarkable. If you want to build policies from that, then you need a stronger grounding. Is it arbitrarily socially constructed, and therefore easy to change? Are there unprincipled exceptions in the social construction, that should be rectified? Is it too complex (or variable across regions and cultures) so it can't be used to communicate clearly? If you could argue that something is a bad social construction, then I'd listen to your ideas about what a good one is.
It has uses (I'll leave whether it's net useful unanswered), and the causation flows the other way: genetic differences caused racial classification, the classification didn't affect the genes.
What is "more" referring to?
If it's a comparison of the average black+white vs. white+white, then I'd say it's plainly incorrect.
Thanks for the link. It's an interesting find, but the comparison is nigh-meaningless and shouldn't drive your decisions.
Using that same within/between comparison can lead you wildly astray: The income differences within genders are higher than between, but the wage/earnings gap is still a live issue. The life expectancy differences within countries are higher than between, but it's still used as a key indicator of progress. The temperature changes within a year are higher than between, but climate change is still concerning.
Comparing the variance between different individual things to the variance between different large aggregates will practically always say that the individuals are more distinct.
That's a neat bit of trivia, but not remarkable either. Sticking with the color example, there's more diversity in wavelengths in "Red" (130 nm) than in "Green", "Yellow", and "Orange" combined (125 nm). Groups aren't always the same size, so "African" being more diverse than "European" and "Asian" combined isn't (wouldn't be?) notable.
EDIT to add: Yes, the category is socially constructed, and yes, the tea has quantum mechanical effects inside your body.
In addition, the elision between discussing West/Central black African genetics (from which black Americans draw the majority of their ancestry) and African genetics as a whole is another common and classic tactic to distract, when the Western discourse is centered around black Americans. Africa contains populations such as North Africans who are more closely related to other Mediterranean populations (albeit with some more recent black admixture related to reasons such as the slave trade and/or Arab expansion), Ethiopians (who have significant Arab ancestry), Khoisan (who were the first to branch off from the rest of humanity), African pygmies (the second to branch off from the rest of humanity), the Malagasy (who were originally Austronesian).
It'd be like trying to poo-poo the genetic differences between birds and non-Avian dinosaurs because the majority of Archosaurian genetic diversity is found within crocodilians, pterosaurs, and non-Avian dinosaurs.
On a side note, African pygmies have long been on the wrong side of systematic replacement, genocide, rape, and even cannibalism from their West/Central black African neighbors, persisting to this day. Somehow, this phenomenon remains largely undiscussed in mainstream Western discourse.
Joke: a black African caught a pygmy and cooks him on a spit. An European walks by, sees it and says "No, you cannot eat people!". The African in surprise responds he doesn't. The Europeran points to the pygmy and asks "What's this?". -- "AAh, but it itsn't a human"
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link