Richard Dawkins vs. Bill Nye on race

[BH]

Late comer to the thread.

Are there enough differences between the DNA of the 3 or 4 "races" as traditionally called to argue there may be speciation taking place?

 

[ApostateAbe]

Late comer to the thread.

Are there enough differences between the DNA of the 3 or 4 "races" as traditionally called to argue there may be speciation taking place?

"Speciation" sometimes refers to the whole evolutionary process and sometimes refers only to the end of that process. At this point, the major human races would be at an intermediate point in that process, not the end of it. "F_st" is Wright's fixation index, which is a scale of genetic differentiation among populations. The F_st among the major human races is about 0.12. There is no standard F_st threshold for speciation, but the F_st between bonobos and chimpanzees (barely two different species) is about 0.5. So, maybe human races are a quarter of the way toward speciation. It does not quite follow that the same set of races will ever reach speciation in the future, as maybe they will reverse the process and homogenize through interbreeding, or maybe a different set of races will come about in the future (i.e. the rich vs. the poor).

 

[Bomb#20]

Well, it sure looks like his point is that the existing concept of race pretty obviously maps well enough to some fact-based genetic differences for 23andMe and the cops to identify someone's race by examining his or her DNA. The larger point is presumably to nominate that level of mapping accuracy as a candidate for the "map well enough" standard used to judge whether the existing concept of race is pseudoscience.

If race denialists aren't willing to accept that as the standard, then they need to either nominate a different candidate or else admit that it's the race denialists and not the rest of us who are practicing pseudoscience.

I fully agree, and you express it well.

I should add that I don't think the "map well enough for cops to do their job" candidate is an adequate standard for determining whether the conventional races are scientific -- that standard would be rather like deciding who qualifies as a high-jumper by asking people to clear a bar set two feet high. Better than no test at all, but hardly sufficient. But this isn't about what would persuade me -- I already know the conventional races have passed much harder tests than that one. No, the question is what standard would persuade a fair-minded believer in race denialism. As of now, race denialists' thinking is still at the "no test at all" stage.

 

[Bomb#20]

Late comer to the thread.

Are there enough differences between the DNA of the 3 or 4 "races" as traditionally called to argue there may be speciation taking place?

"Speciation taking place" implies the races are in the process of becoming more different. They're actually in the process of merging back together -- the high water mark of maximum racial difference was surely thousands of years ago, before decent boats, before the Silk Road, before domestication of camels. Those developments made DNA move much more rapidly from race to race than it had before. So it would make more sense to say speciation used to be taking place before the whole process was overtaken by countervailing forces.

Moreover, before you ever get new species you get new subspecies. And while there's no hard-and-fast rule for how different populations need to be for them to be considered subspecies, for the most part in animals it takes more difference for biologists to call two populations different subspecies than we see between human races. Since modern humans hadn't gotten as far as having subspecies when the process was interrupted, instead of saying speciation was taking place, the most accurate way to put it would be to say that subspeciation was taking place, but is now running in reverse. Modern races are the remnants of that process. Subspeciation is normal for animals of all species; we're no exception. There was a time when Homo sapiens had fully subspeciated -- we had Neanderthals living alongside Cro Magnons at the same time we Cro Magnons were starting to subspeciate again.

In the course of its speciation, every kind of animal we now call a species was once just a subspecies; and before that it was a population only as different from its cousins as human races are from one another now; and before that it was even less different, but still distinguishable. Such is life in a continuous world.

 

[ApostateAbe]

Late comer to the thread.

Are there enough differences between the DNA of the 3 or 4 "races" as traditionally called to argue there may be speciation taking place?

"Speciation taking place" implies the races are in the process of becoming more different. They're actually in the process of merging back together -- the high water mark of maximum racial difference was surely thousands of years ago, before decent boats, before the Silk Road, before domestication of camels. Those developments made DNA move much more rapidly from race to race than it had before. So it would make more sense to say speciation used to be taking place before the whole process was overtaken by countervailing forces.

Moreover, before you ever get new species you get new subspecies. And while there's no hard-and-fast rule for how different populations need to be for them to be considered subspecies, for the most part in animals it takes more difference for biologists to call two populations different subspecies than we see between human races. Since modern humans hadn't gotten as far as having subspecies when the process was interrupted, instead of saying speciation was taking place, the most accurate way to put it would be to say that subspeciation was taking place, but is now running in reverse. Modern races are the remnants of that process. Subspeciation is normal for animals of all species; we're no exception. There was a time when Homo sapiens had fully subspeciated -- we had Neanderthals living alongside Cro Magnons at the same time we Cro Magnons were starting to subspeciate again.

In the course of its speciation, every kind of animal we now call a species was once just a subspecies; and before that it was a population only as different from its cousins as human races are from one another now; and before that it was even less different, but still distinguishable. Such is life in a continuous world.

I wouldn't claim that "They're actually in the process of merging back together" to be established fact. I tried looking for research claiming one thing or the other, but no luck, so if you happen to have on hand research that makes such a claim then I would love to know about it. The rates of interbreeding could be causing the genetics to either drift apart, merge together, or it could be that the races are merging within each nation but drifting apart between nations. Henry Harpending claimed that the races are accelerating apart between continents, but he made that claim in an interview only. I suspect that any given rate of interbreeding would merely converge on a third admixed race given infinite time, never completely merging, but that would be just a guess.

 

[Bomb#20]

I wouldn't claim that "They're actually in the process of merging back together" to be established fact. I tried looking for research claiming one thing or the other, but no luck, so if you happen to have on hand research that makes such a claim then I would love to know about it.

No, just my own back-of-the-envelope calculations. It took some 50,000 years for the genetic distances to get as high as they are; based on how fast the mixed-race population is currently rising the distances could be close to zero in only a thousand years.

 

[BH]

Also, as far as we know, all the so called "races" can still be interfertile with each other? right?

 

[Bomb#20]

Also, as far as we know, all the so called "races" can still be interfertile with each other? right?

Of course. It takes a lot more genetic difference than our races have to cause fertility problems. Even Neanderthals were interfertile with H. sapiens sapiens.

 

[ronburgundy]

Your characterization implies that there is zero covariance between who is placed in the same or different racial sub-categories under the common approach and who would be in different or same sub-categories under a scientifically valid biological approach. IOW, the current categorization doesn't match a biological one any better than if all people were just randomly assigned a sub-category.

In more concrete terms, it assumes that two black men from Alabama are no more likely to be genetically similar to each other than either is to a white man from Alabama.

Do you think that is a defensible assumption?

If we removed information that coded for skin color, and gave geneticists the DNA of all Americans categorized as "white" or "black" under the common rubric, would they be able to do better than chance at categorizing the DNA into the 2 subgroups?

Your assumption that there would be only two 'subgroups' as a result of such a blind analysis is deeply flawed.

I am not assuming that their are 2 and only 2 "objective" categories of humans. There is no such thing as an objective number of categories of anything. There are objective differences between all instances of anything. Those objective differences can be used to reliably divide those instances into any number of categories.
Every species could be divided into as many categories as there are organism that belong to that species, or they could (with scientific validity) be all put into the same category, which is what is done when they a considered members of the same species.

What I said was that if you take only the people who currently check "black" or "white" on a question of race or ethnicity, and you sorted them based only on their DNA into 2 groups, would the match between the DNA sorted groups match the self-reported race groups better than chance?
IF yes, then that means there is some degree of biological and scientific correspondence to to the common conception of race.

Sure, it might turn out that the poorly informed guesses of our forebears happened to match reality by pure luck.
.

It isn't pure luck. The appearance factors used to categorize people as black and white Americans are empirically related to differences in ancestral history and stark differences in the amount and type of environmental variations and need for adapting to new environments. Not only did only a particular non-random subset of the gene pool leave Africa to become the ancestors of "white Amercians", but all those environmental differences would almost certainly cause differences in the probability distributions of various traits or combinations of traits, with skin color and other visible traits being only a fraction of those traits on which those.

 

[ApostateAbe]

Also, as far as we know, all the so called "races" can still be interfertile with each other? right?

Yes, though a central problem with that criterion is that there is no standard threshold for what percentage of each respective race is unable to interbreed with fertile offspring before it counts as two different species. Vincent Sarich claimed, "The largest difference for any of the human data sets is 46 percent, which comes from comparing Teita with Buriat (who live in the Lake Baikal area and speak a Mongolian language)." If this is true, then maybe there is a small percentage of the Teita tribe who can not fertily mate with a small percentage of the Buriat tribe. Could be 0%, 1%, 5%, maybe even 50%. It can not be ethically put to the test. But, whether any percentage counts as the two respective tribes being members of two different species is up to any biologist, though of course chances are they wouldn't, one way or the other. They collectively chose that any two races can't even be two different subspecies.

 

[Rhea]

Well, it sure looks like his point is that the existing concept of race pretty obviously maps well enough to some fact-based genetic differences for 23andMe and the cops to identify someone's race by examining his or her DNA.

Doesn't the use of this example suggest that the Germans and the Norwegians are different "races"? Likewise the French and the English? Because that 23 and me stuff claims to be able to tell you WHICH white country your peeps are from.

So is that actually saying ahything about "races" then, or did this statement just prove that races AREN'T an objective thing?

 

[Bomb#20]

Well, it sure looks like his point is that the existing concept of race pretty obviously maps well enough to some fact-based genetic differences for 23andMe and the cops to identify someone's race by examining his or her DNA.

Doesn't the use of this example suggest that the Germans and the Norwegians are different "races"? Likewise the French and the English?

Not really -- what it suggests is that bilby wasn't making a substantive criticism when he said the existing concept of race doesn't "map well enough" to any fact-based genetic differences. How well does something have to map in order for it to be "well enough"? Well enough for what? It evidently maps well enough for some purposes; but to know whether it implies Germans and Norwegians are different races we have to decide how well "well enough" is.

Because that 23 and me stuff claims to be able to tell you WHICH white country your peeps are from.

So is that actually saying ahything about "races" then, or did this statement just prove that races AREN'T an objective thing?

How could it have just proved that? What's the criterion for whether something is a race? Either it's the test Trausti proposed, or else it isn't, right? Well, if the criterion for a group being a race is that we can tell whether you're in that group by inspecting your DNA, then 23andMe's success doesn't prove races aren't an objective thing. It proves they are an objective thing, and Germans and Norwegians are two different races, and there are a lot more races than people thought there were.

Contrariwise, if the criterion for a group being a race isn't that we can tell whether you're in that group by inspecting your DNA, then 23andMe's success doesn't actually say anything about races, because they use the wrong test for answering race questions. So it can hardly have proved they aren't an objective thing.

So are any race denialists willing to state their criteria for a group being a race, so we can check whether there are any groups that satisfy it? Or are all objections as vague as bilby's -- so vague they're pseudoscience?