Talk:Race and genetics

Quayshawn Spencer

I feel that this edit put extremely WP:UNDUE weight on Spencer. Now, he absolutely is a scholar in the field, and we could notionally cite him for a sentence or so... but a lot of his work is fairly recently-published and not, as far as I know, particularly widely-accepted. At the end of the day he's one associate professor who has published a few papers - we'd need more reason to think he's significant to devote entire massive sections to him or to frame whole areas of the topic around his structuring and rebuttals. A Google News search finds almost no coverage of him; a Google Scholar search finds a few dozen citations to his papers, at most. I mean, I may be overlooking something, but nothing I could find on him implies that he's someone we should be devoting a huge section of the article to right now - and he himself describes his own views as radical (ie. not yet widely accepted in the field.) --Aquillion (talk) 18:53, 28 April 2019 (UTC)

Widely-accepted by whom? Von Clown (talk) 10:34, 3 May 2019 (UTC)

By the rest of the academic establishment; we can mention his views, but by constantly including his response to every single point on the topic, we give the impression that he's a major figure in the field, which he absolutely is not. --Aquillion (talk) 18:28, 3 April 2021 (UTC)

Again, we shouldn't rely so heavily on one relatively obscure source; we already cite him several times (and I'm happy to cite him once or twice; I left several citations to him in place), but at the moment the entire "Objections to Racial Naturalism" section is written with his response to almost every point made, which gives the impression that he is a towering figure in the field whose opinions define the entire debate. That's WP:UNDUE. One sentence mentioning his views is fine. --Aquillion (talk) 15:22, 6 April 2021 (UTC)

I'm not always against relatively minor figures being heavily used for articles, but there has to be evidence in reliable sources that their views are being taken up by the discipline and these sources may contain counterarguments that need to be cited to avoid WP:UNDUE. Recent works by relatively minor figures are problematic since the necessary supporting reliable sources probably don't exist. I haven't looked at Spencer's work but what Aquillion says does reek of WP:UNDUE. — Charles Stewart (talk) 07:55, 20 April 2021 (UTC)

Spencer is relatively well-known among philosophers writing about the metaphysics and philosophy of science of race. One thing that I think is challenging about this topic is that the question of whether races are real is a metaphysical one, albeit one that's informed by science, analogous to debates about whether other classification schemata are real and what the nature is. (E.g., are species real? Are species kinds or particulars? Etc.) I think it's worth not giving philosophical conceptions of race too much weight in an article on race and genetics, including Spencer's, but I also think one thing that's vexing about this topic is that scientific views that often reflect a mere sociological consensus about what is ultimately biologically real are given too much weight in these debates generally (e.g., the AAPA's statement).167.244.212.246 (talk) 17:42, 14 December 2022 (UTC)

If what he says relates to the topic, why would the number of sentences matter? Certainly whatever he says has a counter-point. I just came across him because he was cited in another source: Anthony F. Peressini, Against the philosophical project of “biologizing” race.

I just read Quayshawn Spencer's paper titled: A Radical Solution to the Race Problem. Under the heading "Race and human genetic variation" on this Wiki page, he wasn't even cited there, and he writes in direct conflict to what Peressini cited him for (that newer research shows there are clear distinctions between five major racial groups, but that it is a very, very tiny difference in variance. But that small difference is enough for reliable divisions). A lot of this topic in general on Wikpedia seems to be a conflation of race and subspecies. Captchacatcher (talk) 00:06, 24 December 2023 (UTC)

Regarding the legitimacy of the 99.9 % figure

I have found absolutely no corroboration to that figure in scifinder. I checked the two putative sources, but they weren't primary sources; they were from books by anthropologists.

It doesn't make sense to dissimulate about this; diversity is a positive thing that should be heralded. If you want to diminish racism, a valiant cause, focus on the inter- Vs intra- population differences, don't parrot a patently false figure.

It's also absurd from a biological standpoint; if we were to have 99.9% genetic similarity, that would anyways be a catastrophe. We already have 3x less genetic diversity with our species' LCA; this putative figure would put as at 15x less diversity.

I don't want to change an article that isn't mine, but I hope my argument is well-understood. Genetic diversity is essential for a species (ours included) survival; we should be highlighting it as a good (and necessary) thing we want MORE of, not diminishing it. IamtheStudent (talk) 09:24, 18 January 2023 (UTC)

Who We Are and How We Got Here, David Reich, page 4:Around 99.9 percent of these letters are identical across two lined-up genomes, but in that last ~0.1 percent there are differences, reflecting mutations that accumulate over time. These mutations tell us how closely related two people are and record exquisitely precise information about the past.  Tewdar  09:50, 18 January 2023 (UTC)

Here's an article if you prefer: Single-nucleotide polymorphisms (SNP) represent genetic variation in a human population; 99.9% of the DNA sequence is identical and remaining 0.1% of DNA contains sequence variants  Tewdar  09:55, 18 January 2023 (UTC)

If you want to argue otherwise, present some reliable sources please.  Tewdar  09:58, 18 January 2023 (UTC)

I mean, there's this, but it doesn't actually give us a percentage so we'd have to do some original research to get the percentage...  Tewdar  10:43, 18 January 2023 (UTC)

I just found a source and updated it which said 99.6-99.8%, so I included with the others and set to 99.6-99.9%, based on this source after the genome project was complete: https://www.nature.com/articles/ng1438. Captchacatcher (talk) 16:23, 28 December 2023 (UTC)

Variation between racial groups based on Ancestral Interbreeding/admixture

I'm confused here, because these two lines seem to be incorrect:

"The vast majority of this genetic variation occurs within groups; very little genetic variation differentiates between groups. Crucially, the between-group genetic differences that do exist do not map onto socially recognized categories of race."

There is in fact anywhere up to 5% variation between humans, based on "socially recognised categories of race". Melanesians can have up to 5% Denisovian DNA, retained because of historical interbreeding. But Europeans don't have any Denisovian DNA, but they have up to 2% Neanderthal (because their ancestors bonked them). African people have practically 0% Neanderthal DNA... but Europeans have on average 1-2%. So between the "socially recognised categories of race, there is 1-2% variation between the African and Europeans, and 5% between Europeans and Melanesians.

Deathlibrarian (talk) 07:06, 13 September 2023 (UTC)

The degree of archaic admixture itself is not an indicator of absolute variation among modern human populations. For instance, Papuans and Europeans (both being Out-of-Africa populations) share more alleles with each other than West Africans and San (or other hunter-gatherers of southern Africa) do, contrary to what focussing on archaic admixture alone suggests. The 95% to 100% non-archaic contributions to the genome of modern human populations obviously outweigh the effects of archaic admixture when looking at absolute degrees of allele sharing.

For that matter, there are no "socially recognized categories of race" in an absolute sense. When societies create these categories, they are entirely situational and quite arbitrary, with no match in the genotype for those constructed racial categories, except for isolation-by-distance effects. –Austronesier (talk) 20:28, 13 September 2023 (UTC)

Thanks Austronesier, assuming the cultural definitions of race rather than scientific, is the difference in DNA between Europeans and Melanesians more than the traditionally suggested figure of 99.9%, and if so, does archaic admixture play a part in this? Deathlibrarian (talk) 21:46, 13 September 2023 (UTC)

When taking about percentages of admixture, we have to keep in mind that Denisovans and Neaderthals shared ca. 99% DNA (or 99 point somthing) with modern humans. This page[1] from the Smithsonian explains it nicely (see the section "Shared DNA: What Does It Mean?"). So max. 5% admixture from a gene pool that is almost identical with ours doesn't equal 5% divergence. To do rough arithmetics: Denisovans are assumed to have split from modern humans around 800kya, so a contribution of up to 5% Denisovan DNA comes equal to the genetic drift that occurs between populations that have been separated for a few tens of thousand years, which means archaic admixture produced genetic shifts that lie within the usual range among Out-of-Africa populations.

This is nicely illustrated by the fact that before archaic Denisovan admixture was detected, Papuans and Australians were thought to have split around 70kya (or earlier) from other OoA populations (predating the split between ancient West Eurasian and ancient East Asian ancestries), because the genetic distance was ascribed to separation alone. However, when taking Denisovan admixture into account, it has become clear the ancestors of Papuans and Australians were actually a sister lineage of ancient East Asians with a split date of c. 45kya. So yes, archaic admixture had an impact on genetic distances to the point that it superficially can distort our understanding of the relation between ancient ancestral lineages, but not to the point of pushing this distance beyond the boundaries of regular drift by separation.

But note that archaic admixture does not correlate with racial categories of 19/20th century physical anthropology (your suggested "mapping"): e.g. Andaman islanders, Malaysian Negritos, Philippine Negritos, Papuans and Indigenous Australians were classified as "Australoid" (or "Australo-Melanesian"; popular among scholars studying skeletal morphology). However, only Philippine Negritos, Papuans and Indigenous Australians show traces of Denisovan admixture, while Andaman islanders and Malaysian Negritos don't. Similar things can be observed among populations historically classified as Caucasian/Europoid: some have levels of Neanderthal admixture similar to the one of East Asians, while for others it is significantly lower. –Austronesier (talk) 17:58, 14 September 2023 (UTC)

Thanks so much Austronesier for taking the time to explain all that, yeah, it finally clicked that 5% admixture from a gene pool that is almost identical with ours doesn't equal 5% divergence, as that 5% isn't purely distinct non shared DNA (as you say), so the difference between a human with archaic human DNA and one that doesn't have that isn't going to be anything more that the 99.9% standard figure. However, upon further reading, if someone *theoretically* had a high enough percentage of general Neanderthal or Denisovian DNA that meant the non shared portion of that could rise above the .01% level.Deathlibrarian (talk) 05:12, 17 September 2023 (UTC)

Edit dispute

@Generalrelative I am not a fan of doing re-reverts by any means. In this one instance however, I do believe this is warranted. My edit was consistent with WP:NPOV and does not fall under WP:UNDUE. The two sources I used were already used in the specific paragraph and are both extensively cited pieces of literature on the topic. As far as I can tell, the arbitration for this article, which has apparently been contentious in the past, does not require WP:1RR or WP:0RR.

That said, I only reinstated part of my edit, which provided more greater specification of how much is "the vast majority" without challenging the "vast majority" claim. I would like to use this topic to open dialogue and see what common ground can be found. Jokojis (talk) 18:03, 26 June 2024 (UTC)

I have reverted your additions again. The text that was modified and moved into the passage that is referenced to Rosenberg et al. (2002) is not supported by Rosenberg et al. (2002). The stable text, however, is supported by the AAPA source: Socially-defined racial categories do not map precisely onto genetic patterns in our species: genetic variability within and among human groups does not follow racial lines. –Austronesier (talk) 18:49, 26 June 2024 (UTC)

Hi Austronsier. First, I want to say thank you for giving more details in your reversion.

It seems I mixed some things up, the 5% came from the "Genetics and the Shape of Dogs" article from American Scientist referenced in the 5th paragraph of the introduction, my apologies that I didn't add that citation. While the unref'd source is more than a decade newer, I do think it'll be better to go with the ~3-5% from Rosenberg et al (2002) given the higher source reputability of peer-reviewed articles, at least until any newer peer reviewed sources appear with a different estimate.

As far as the "do not" vs "don't always", the other source immediately after the sentence in question, Tishkoff, SA; Kidd, KK (2004), says in the abstract: "Although populations do cluster by broad geographic regions, which generally correspond to socially recognized races, the distribution of genetic variation is quasicontinuous in clinal patterns related to geography." This does support the "don't always" version, and is further elaborated under the subsection "Determining individual ancestry".

On a tangential note, some of the references are repeats, should we fix that?

Jokojis (talk) 19:54, 26 June 2024 (UTC)

@Jokojis: Thank you for your efforts to get an updated figure for the average genomic distance between two individuals. I see the 2001 value of 99.9% cited even in recent publications of the last few years (with an interesting twist in a page by the NIH National Human Genome Research Institute [2]). May I ask how exactly you have extracted the figure of 99.35% from the 1000 Genomes Project Consortium paper? –Austronesier (talk) 09:40, 29 June 2024 (UTC)