Human Races

quote:

---

The averages given show the slightly more within than
between (but nowhere near as much as quoted in other
people's posts).

---

You're talking about two different things here. The values cited in other posts concerned how different two individuals are. If you compare you and your next-door neighbor genetically, and then compare yourself to someone whose ancestry is on another continent, how much more different are the two of you in the latter case? The answer is, very little; on average you are only slightly more different from someone of a different ethnic background than you are from someone of the same background. In the case of Africans and non-Africans, in fact, a random African is likely to be more similar to a random European than he is to another random African. That result alone makes the concept of race pretty useless.What the numbers you quote here are talking about is the total number of variants seen in a population. The complete set of variants differs more between populations, because it includes lots of rare variants that are particular to a population but that contribute very little to average differences between individuals. The differences can be used to distinguish ethnic groups, but they don't contribute significantly to what is usually meant by races: if 0% of group A has an allele and 2% of group B has it, it seems kind of silly to say that the allele is a characteristic of group B and not of A.(This study actually exaggerates the differences between populations, by the way, by using very small sample sizes. Many of the alleles that were observed only in a single population would be observed in other populations if more individuals had been tested.)In the great majority of cases, genetic differences between populations are of two kinds: 1) the same alleles occur in both populations, but at different frequencies, or 2) one population has an allele that another doesn't, but at low frequency. There do exist fixed differences between populations (i.e. cases where one population all has one allele and the other all has a different one), and some of them may be biologically important, but they are rare.

quote:

---

I was simply using 15% as a stated-in-another-post within population difference figure.

---

You used it incorrectly. There is no sense in which two humans are 15% different from each other genetically.

quote:

---

Evidence tends to indicate that there is a set of genetically
determined traits that are unique to different populations.
This means that 'race' is a genetic phenomenon.

---

No, that means that different populations have different unique genetic characteristics (rarely -- most genetic characteristics are not unique to a population). Human geneticists distinguish populations all the time, for a variety of reasons. "Race", on the other hand, covers a range of meanings, some of which map very poorly onto genetics. The thread started out with the idea of a few "pure" races which could be mixed to produce everyone. You, on the other hand, cited genetic differences between Vikings and other northern Europeans as evidence for the usefulness of "race" as a concept. Those are not the same concept of race. Geneticists tend to avoid the term "race" because of its multiple meanings and its heavy non-scientific baggage.

quote:

---

Variation within a population is different to variation between
populations and so cannot be compared in the way that it
has been (IMO).

---

How are they different, and why can't they be compared? (And what do you mean by "population" here? Are Africans one population or many?)

quote:

---

Folk concepts of race are often (though not always) focussed
on observable difference -- if that difference is genetically
determined ...

---

Then you've established a classification based on a handful of genetic traits, out of thousands. Any scheme that groups together Yoruba, San, Mbuti pygmies, Andaman Islanders and Phillipine Negritos, as racial classification usually does (they're all black or negroid), is gibberish to a geneticist.

quote:

---

That being, that the logic of the interpretation of the evidence
is flawed. The two deviation figures used are not based upon
the SAME thing.

---

The figures quoted for within vs between group genetic differences are based on exactly the same measurement -- usually the heterozygosity. So I'd still like to know what you think is different about the comparison within populations and the comparison between populations.

quote:

---

If two populations (however you have defined it) have different
unique genetic traits, the surely they can be characterised
by those traits.

---

If only 1% of the members of the population have the trait, the population can be characterized by the trait. The members of the population cannot be characterized by it.

quote:

---

What do you mean by a 'unique trait' in any case. Is, for
example since it seems the main issue here, skin colour
a unique trait or a common trait?

---

Since no skin color is unique to any of the kind of groups you're talking about, it would seem not to be a unique trait.

quote:

---

quote:

---

Human geneticists distinguish populations all the time, for a variety of reasons. "Race", on the other hand, covers a range of meanings, some of which map very poorly onto genetics. The thread started out with the idea of a few "pure" races which could be mixed to produce everyone.

---

Do you view that as not logivally possible, or as unsuported/refuted
by evidence?

---

Refuted by evidence. You cannot decompose human genetic variation into a handful of ideal types.

quote:

---

quote:

---

You, on the other hand, cited genetic differences between Vikings and other northern Europeans as evidence for the usefulness of "race" as a concept. Those are not the same concept of race.

---

What makes them different?

---

By the definition used earlier in the thread, all northern Europeans are the same race. By yours they're not. More generally, the two definitions are simply different: you mean by race any genetically distinct group, while the earlier poster meant one of a small number of basic genetic types.

quote:

---

I'm sure that Vikings viewed themselves as a distinct race,
and they were certainly viewed that way by the other peoples
of Europe. One might see a tall blonde or ginger person and
say 'That's a Viking.' -- racial characterisation based upon
a handful of common traits that are uncommon in other 'races'.

---

Leaving aside the fact that "Viking" was actually an occupation, not an ethnic group, do you have any evidence that Vikings thought of non-Scandinavians as of another race? (Other than just assuming that they thought the same way you do.) Humans have a general tendency to dinstinguish their own group from others, but it is not at all obvious to me that they always use racial characteristics to do so.

quote:

---

A site with high variability within one group that is different
(even non-varying) in another is not directly compared, and yet
it is that data that is relevent to the question of race.

---

?? Of course it is. One common procedure: take a pair of chromosomes and compare them in a stretch of sequence. Count the number of differences. Then look at a bunch more pairs. The average number of differences is the measure of genetic diversity. If you do that with chromosomes from the same population, you'll find about one difference every 1200 base pairs. If you do it with chromosomes from different populations, you'll find about one difference every 1300 chromosomes.Another procedure is to find places that are vary between individuals by sequencing chromosomes from both populations. Then compare the allele frequencies at each site for the two populations. Sites that have very different frequencies translate into a large genetic distance, while sites with small differences in frequency mean a small genetic distance. Genetic distances between human populations are generally small.

quote:

---

For a population to be characterised by a trait
then it has to be consistently present. If it is, then all
individuals within the population bear the trait ... that many
human populations are the result of millenia of interaction between
groups brings populations genetically closer together.

---

If that's what you mean by a trait characterizing a population, then there are almost no traits that distinguish races. Almost all traits that are borne by all members of one population are frequent in other populations.

quote:

---

quote:

---

Since no skin color is unique to any of the kind of groups you're talking about, it would seem not to be a unique trait.

---

Which groups were those?

---

Groups that are as closely related as two populations from northwestern Europe, from your Viking example.

quote:

---

What do you mean by skin colour? (i.e. 'we all have skin that
has a colour' is different from consideration of the usual
range of skin colours exhibited within a population).

---

What did you mean by it? I wasn't the one who brought it up.

quote:

---

The opinions expressed in this thread seemed to be saying that
there was no such thing as race at a genetic level. The data
presented, as far as I can see, shows the opposite.

---

Which data? I haven't seen any data presented that supports your position, to the extent that I've figured out what your position is.

quote:

---

Viking is the name given to the Danes, Norwegians and Swedes
around the 800-1050 AD mark -- it's not what they called
themselves -- it's what we call them now.

---

You may call them that, but I certainly don't. Scandinavians of the time who were Vikings called themselves Vikings, but the rest didn't. "Viking" meant (and still means, to anyone who studies the period) a raider or pirate. If you were going off raiding, you were going aviking.

quote:

---

There is more to race than physical appearance -- and that is
captured culturally by (for example) Zulu's enslaving Bantu and
other 'inferior' tribes, or Native American tribes whose name
for themselves means 'Human Beings' with the implication that
all outside the tribe are not.

---

I thought we were talking about races as a genetic concept. What does that have to do with these distinctions?

Perhaps it would clarify matters if Peter would list a few groups that he thinks constitute races.

quote:

---

The following nicely sums up my objection to the rejection of
human races as a biological concept as argued in this thread.
I will admit though (to save time) that this paper agrees that
the link between cultural concepts of race and a meaningful
biological one is likely not there.

---

Since that's precisely what you're arguing, the paper would appear to undercut your position.

quote:

---

I tend to disagree with that aspect of the paper, simply because
the commonly held view of what makes races of humanity different
focusses on 'adaptive' differences.

---

Really? I'm not aware of any good evidence that any of the commonly used differences, apart from skin color, is the result of adaptation.I have a couple of problems with your use of this paper. First, recall that your claim was that there is a genetic basis to what we call race. These authors disagree, unless you mean something very peculiar by "we". They explicitly state that human races, in the meaning given to the term by human geneticists, in the meaning given by anthropologists, and in the popular meaning, are not real biological entities. Instead, they argue that human populations do qualify under quite a different definition of race, one that makes it equivalent to "ecotype". (Why they think it's important to make such a seemingly pointless argument I haven't got a clue.)The second problem is related to the first. The definition of race that you would like to defend is based on clusters of traits, while the authors again explicitly state that different traits do not naturally form the same group. A phylogenetically based race will tend to share a suite of traits, while a race that is characterized only by a common selective environment will not, since different selective pressures vary differently with geography.In their words, "And of course, as has already been noted, insofar as folk races are supposed to pick out populations that systematically differ from each other over a wide range of genetic and phenotypic measures, biology provides no support for the existence of such populations (and indeed, provides evidence that no such populations exist)." And, "Of course, this implies that insofar as we focus on an ecotype conception of race, there will not necessarily be a unique 'race' to which any given member of a population belongs. Any given individual may in fact belong to a number of different ecotypic races, and/or be a member of one (or more) intermediate population(s) within a (series of) clinal distribution(s)."

quote:

---

Consensus is hardly a validation of anything -- a couple
hundred years ago the concensus was that the earth was flat.

---

There's been a consensus in the West, at least, that the Earth is spherical for over two thousand years.

quote:

---

Biological lineages can be traced genetically (e.g. the Viking
study that found a correlation with the known interaction
with the Danes and the boundaries of the Dane-Law).

​

The differences are related to locally adaptive traits, and are
heritable.

---

Differences between Scandanavians and Anglo-Saxons are related to locally adaptive traits? Like what? The differences that were used to distinguish Scandanavian genetic contributions were not functional one. You've confused definitions of race again: the distinction between Scandanavians and Anglo-Saxons is a phylogenetic one, while the definition of race you've just been supporting is an adaptive one. You just used two different definitions in two sentences.

quote:

---

So distinctive differences in physiology, bone stucture, muscle
attachment, etc. are just random coincidence and not the
result of some past selective pressure?

---

Quite possibly. Also quite possibly not. Determining whether any particular trait in real-life populations is the result of selection or not is actually quite difficult, except in the most extreme cases. Most human traits are in the "we don't know" category. Even the evidence for selection for skin color is a little tenuous.

quote:

---

So a population that has been subject to the same selective
pressures will NOT share a set of traits?

---

A single population subject to the same set of selective pressures will probably share a set of traits. But different selective pressures are unlikely to have exactly the same geographical distribution, which is why the paper you quoted from talks about membership in multiple, overlapping races. Some environmental factors change over very short distances, while others remain similar over thousands of miles.

quote:

---

If there was concensus in the west for a spherical earth for
two thousand years, how is that people thought Columbus was
going to fall off the egde if he went too far?

---

They didn't. The whole flat-earth thing was a product of tendentious (and crappy) nineteenth century antireligious efforts at history. The reason people thought Columbus would fail was that he was using the wrong value for the diameter of the Earth. (The more commonly accepted value was around 25,000 miles, and was quite accurate; he was using a much smaller estimate (18,000 miles? I forget). Both estimates were made by the ancient Greeks.) There's no evidence that anybody in Europe three or four hundred years ago thought that the Earth was flat. Uneducated peasants might have thought so (they might still think so in some places, for all I know), but nobody bothered to record what peasants thought. The educated knew better. (You probably mean more than than three or four hundred years, by the way -- the first circumnavigation of the globe was completed well over 400 years ago.)I'll deal with your substantive comments latter. Right now my kids are harassing me to read to them.

quote:

---

So your rejection of the traits being adaptive is as incorrect
as my suggestion that they are.

---

Had I stated such a rejection, it would have been incorrect. I didn't, so your comment does not apply. I just reject labeling traits as adaptive until we have evidence that they are.

quote:

---

Can we positively identify ANy traits as adaptive except by
a significant amount of hypothesising?

---

We can make a good case, but only when there's enough genetic evidence. E.g. "Detecting recent positive selection in the human genome from haplotype structure." (Sabeti et al., Nature. 2002 Oct 24;419(6909):832-7.) Without doing the work, however, it's just a bunch of guesses.

quote:

---

If geographic separation subjects different groups to different
sets fo selective pressures, then the populations that are
affected by those pressures will share trait sets that populations
elsewhere do not have.

---

Could you be more specific? Identify some groups that are the sort that you think could be considered races, and identify their distinctive environmental conditions. That would give me a much clearer idea of what you mean.

quote:

---

doesn't detract from concensus being a poor
metric for a theory.

---

Sure. It's the evidence that counts, not the consensus.

quote:

---

Darwin didn't have the luxury of genetic evidence, and yet he could
still infer adaptive traits from observation of traits wrt
environment ... isn't that sufficient? If a trait provides a benefit
of some kind in the environment where it is the norm?

---

Just showing that there's a benefit is not usually enough -- and showing that a trait is beneficial is pretty difficult in itself in humans. There are cases where the circumstantial evidence is good enough that it's pretty persuasive; the persistence of lactose tolerance in populations that have historically practiced dairy-farming is a good example. I can't think of any that are markers for racial groups, however.

quote:

---

Isn't the emergence of differentiation between two populations
of the same species in different locations required by evolutionary
theory ... and it's absence an undermining of such?

---

That depends on how recently the populations have separated and how much gene flow there is between them.

quote:

---

Inuit :- live in harsh, cold environment with limited vegitation.

---

An environment they share with a number of other ethnic groups. Are they all one race, or are they many races? I would say that the Inuit are a group that has shared a similar enough environment and has been isolated enough that they might have distinctive adaptive allele. Then again, they might not -- they might share adaptive alleles with other arctic groups, for example. Is there any evidence that they do have distinctive, adaptive alleles?

quote:

---

Kalahari Bushmen :- live in harsh, desert environment some game, some
fruit/nuts/edible roots.

---

This is a better example of more typical human populations. "Bushmen" is not a category that they would use to self-identify -- it's a grouping invented by European colonists. They're confined to the Kalahari primarily because they've been pushed there by recent pressure, so it's not at all clear that there's any reason they would have unique adaptations for desert life. San-speaking groups originally inhabited all of southern Africa. They admixed heavily with agriculturalists arriving from farther north; some groups retained San languages, others northern languages, but all were genetically mixed.

quote:

---

Celts :- Originated in cold, wet hilly location unsuitable for much agriculture.

---

Really bad example. Many or most self-identified Celts today have little or no genetic inheritance from the original Celts. I'm not sure where you mean by the cold, wet hilly location they originated -- do you mean southern Germany? Or the Asian steppes, where they may have come from earlier? To the extent that they are a genetically identifiable population, they don't seem to have stuck around anywhere for very long, so it makes no sense to talk about unique environmental pressures for them.

quote:

---

Let's just ignore disease adaptations (like sickle cell) for now; probably the most frequently used racial identifier - skin color - does have regional dependence. Skin color is based on a balance between two issues: genetic damage from UV light (having extra melanin absorbs the light, and reduces the risk of damage), and vitamin-D synthesis (which melanin inhibits). In regions with more annual light intensity, darker skin is adapted for; in regions with less annual light intensity, lighter skin is.

---

Sorry, I wasn't clear -- I'd already agreed several posts back that skin color is probably adaptive (dark color quite likely, light color less so -- it's possible it just represents a relaxed constraint, not positive selection). It's not at all clear to me that it does represent a racial marker, in the sense that he's using "race" here, or at least that it represents anything but a poor marker. He's treating race as self-identified ethnic group with a distinctive set of traits, and a very large number of such groups share similar skin color.

quote:

---

Isn't it likely that light skin was selected for in
low-sunlight regions because darker skin reduces (in the
sense that you need more light to do it) Vit-D synthesis ?

​

Otherwise the 'western europeans' would range in skin tone
across the ful spectrum rather than in the lighter
end, surely.

---

More likely? Yes, I'd say so. But I don't see that it's certain. All northern peoples have straight hair, as far as I know -- would you argue that that must be the result of selection, since without selection they'd show the full range of hair shapes?

quote:

---

This implies that regionally associated, heritable traits
just emerge at random,

---

It's certainly true that many regionally associated, heritable traits emerge at random. Remember all of those nonfunctional genetic markers we were talking about? They're all heritable traits that emerged at random; many of them are regionally associated. A random trait is unlikely to be associated only with one region (assuming we're talking about fairly small regions, as you have been) and also to be very common in that region -- that would suggest either selection or an extreme bottleneck. But there aren't very many traits (that I know of) that fit that description.

quote:

---

and leads to the populations originating
in different regions came into being independently of one
another.

---

I don't know what you mean here.

quote:

---

If the population that headed out-of-Africa were a single population
then according to what has been suggested in this thread
they would be pretty homogenous.

​

That populations in different regions are observabley different
and these differences are heritable means that something
promoted particular phenotypes in the different regions.

---

Different populations still are pretty homogeneous, compared to other species; since we're dealing with our own species, we're attuned to picking out fairly minor differences. There's nothing that says heritable differences between the populations have to be the result of selection. Widely separated populations are pretty isolated genetically (until recently), and there has clearly been a significant amount of genetic drift, especially in all non-African populations. To a good approximation, populations in western Europe and eastern Asia each carry subsets of the alleles that left Africa; they just carry somewhat different subsets.Because of drift, populations will differ in their neutral variants. A signature of selection would be if in a particular trait two populations differed much more than they do on average over the rest of the genome, especially if there is independent evidence that the different phenotypes have selective advantage in the two populations. Skin color is in this category, as is lactase persistence and sickle cell trait. None of those traits is characteristic of a particular race, however, as you have defined race, and two of them aren't visible characteristics.

quote:

---

Are you saying that the ToE's dependence on differentiation
within a species depends on the above, or whether it happens
or not depends on the above.

---

I'm saying that in evolutionary theory, the genetic distance between populations depends on the time since they separated and the amount of gene flow since. Depending on those parameters, the distance can vary from almost zero to a whole bunch.

quote:

---

Inuit are genetically pre-disposed to obesity (i.e. fat deposition)
which is beneficial in a cold climate,

---

Are they more pre-disposed to obesity than many other populations, especially ones that are traditionally hunter-gatherers? Many populations are predisposed to obesity and diabetes. I'd say it's at least as likely that this is the ancestral human phenotype, and that the phenotype we often define as normal arose only in a few populations (probably agricultural ones).

quote:

---

require less calcium
in their diets (recommended calcium levels given to Inuit
children can cause hypercalciuria at significantly higher rates
than seen in non-inuit)

---

Plausible. Is there good evidence for this, though? All I can find is a single report in CMAJ that is pretty scanty evidence. (See
Low-calcium diet | CMAJ.) I wouldn't be surprised if it were true, however. As I said, the Inuit are a group in which I might expect to see some distinctive selected alleles.

quote:

---

They also possess lower incidence of
genes associated with cardio-vascular disorders, which is beneficial
if you want to survive longer with higher body-mass-index.

---

I don't think there's any evidence at all for this one. If anything, the evidence is to the contrary: "Specifically, compared to whites, both Oji-Cree and Inuit have an excess of 'deleterious alleles' from 12 candidate genes in atherosclerosis and/or diabetes." (Clin Chim Acta. 1999 Aug;286(1-2):47-61). Inuit are thought to have less cardiovasular disease (although even that is in some dispute --
see "Low incidence of cardiovascular disease among the Inuit--what is the evidence?" Atherosclerosis. 2003 Feb;166(2):351-7), but they also tradionally have a diet high in fish oil, which is thought to be protective against cardiovascular disease.

quote:

---

The intrusion of Bantu herders into their hunter-gathering lifestyle
pre-dates western interaction with the region, but is not
overly indicative of inter-breeding between the two different
tribal cultures.

---

No, but the genetic date are pretty clear:
"The data of Soodyall (1993) indicate that the Dama similarly have only ~5% Khoisan lineages but that the southern African Xhosa and Zulu may have ~25% and ~50% Khoisan lineages, respectively (also all L1d). This much higher level of assimilation is consistent with the presence of Khoisan click consonants in both languages. In southern Africa, Khoisan speakers themselves appear to have experienced high levels of assimilation of Bantu lineages: ~23% in the Vasikela !Kung (Chen et al. 2000), ~24% in the Sekele !Kung (Soodyall 1993), and ~61% in the Khwe, consistent with their similar physical appearance to southern African Bantu speakers (Chen et al. 2000)." That's from a paper on mitochondrial DNA; Y chromosome data shows a similar pattern. (References on request.)I have to go to bed now -- lot's of snow shoveling to do tomorrow. I'll write more if I have time.

quote:

---

Other selected quotes are :-
"What the eye sees as racial differences - between Europeans and Africans, for example - are mainly adaptations to climate as humans moved from one continent to another. "

---

If Cavalli-Sforza has evidence that these traits, apart from skin color, are the result of adaptations to climate, he should tell the rest of us.

quote:

---

"All Europeans are thought to be a hybrid population, with 65% Asian and 35% African genes"

​

You cannot have it both ways ... there are either genetically
observable differences that allow us to trace humanity's
origins back to Africa, or there are no significant differences
between races that can be seen in the genome.

---

No one has denied that there are genetically observable differences between populations. What we deny is that there are sets of differences that line up well enough with culturally defined races or ethnic groups to justify talking about those groups having a basis in genetics.Here's the genetic situation as I understand it. Human groups do differ genetically, on every scale that you can examine. It's trivial to distinguish between a West African and a northern European, or a Mayan and a Tamil, based on their chromosomes, for example. It gets harder to distinguish groups as they get closer geographically and historically, but it is still possible. You can distinguish African from non-Africans. You can distinguish groups from different parts of the same continent, or groups from different language families, or neighboring tribes; you can very likely distinguish neighboring villages. Some of the genetic differences are the result of natural selection, either because they represent different genetic responses to the same pressure or because selective pressures are different in different environments (including in different cultural environments). Some of it is purely random. How much phenotypic difference falls into either category is not known at present.There is almost never a clear boundary between different genotypes, but rather a slow change over distance, with different genes varying at different rates and in different geographic regions. Culturally defined groups are (or were, at least) usually confined to some geographic region, so cultural groups inevitably correlate with certain genotypes. That does not, by itself, make the cultural groups a natural way of dividing the genotypic variation; any arbitrarily assigned geographic boundaries would also show a correlation. Some (but not all) culturally defined groups do, in addition, show a higher degree of relatedness within them than a randomly assigned geographical group would; i.e. they are somewhat inbred. This is a fairly small effect for humans, since we are a highly outbred species. Populations for which it is significant tend to be small, isolated recent hunter-gatherer groups. These groups probably come closest to what you mean by races, but they generally don't have the kind of obvious distinctive physical characteristics that would be used as race markers. Many other groups are outbred enough that they aren't usefully thought of as a genetic subpopulation, and some groups are (genetically speaking) a hodge-podge of different ancestries. Overall, then, culturally defined groups are at best a very imprecise reflection of patterns of genetic variation.Even with its imprecision, group identity can still be useful: physicians will sometimes use self-identified ethnicity as a clue to guide diagnosis or treatment. That's rational, but it's a weak source of information and can easily be wrong, not just about individuals but about entire communities.None of this is to say that political/social considerations never obscure genetic realities. There are people involved in genetics (especially among those that worry about ethical and social aspects of genetics) who adopt an extreme rejection of genetic differences between human populations. For example, the International Haplotype Map Project is currently planning to map haplotypes in three populations, one African, one Asian and one European; the hope is to add more later if they prove to be needed. For those who planned this strategy, the justification for it is simple: different populations have somewhat different sets of haplotypes, and if the map is going to be medically useful for most of the world's people (which is its goal), it has to provide maps for different populations. (Fortunately, haplotypes only vary a lot on large geographic scales, so a small number of populations should be enough to describe common haplotypes for most people.) For some of the ethics people, however, the strategy amounts to the "racialization" of genetics, and is something to be condemned. I think such reactions are quite confused, but they do occur.Most working human geneticists don't react that way, however. On the other hand, they do tend to be aware that when they talk, they're talking about people, and that people (other than other geneticists) may well be listening. They therefore choose their words with care. Couching any kind of statement about human genetics in terms of the word "race" pretty much guarantees that it will be misunderstood, and that by itself an excellent reason for avoiding it.Now, if you will excuse me, I have to go finish writing a short talk in which I will explain to the steering committee of the HapMap why they should choose markers in Africans and not in Europeans.

quote:

---

All of the citations show that differentiating between different
cultural members IS possible at the level of the genome.
Even with the inter-breeding between Zulu, Bantu, and Khosian
people, from fairly narrow geographic ranges.

---

The problem is that you can make many other equally valid divisions. A Khoisan-speaking village may be closer genetically to the Bantu-speaking village next door than it is to the Khoisan-speaking village 50 miles away, but you would label the two Khoisan-speaking villages as part of the same race and the Bantu-speaking village a different race. When you do that, you demonstrate that your labels are cultural, not biological. It's not that the cultural labels never correlate with genetic relatedness; they often do -- on average, Khoisan speakers are more closely related to each other than they are to Bantu speakers. But that correlation does not mean that the cultural label is itself a genetic description. When you treat the label as if it were a biological description, which is what you're doing, you're are led into ways of thinking that may be quite mistaken. Language influences how we think, and in this case it is my professional opinion that the language you're advocating make thought less precise and more error-prone.Instead of race, think about people who live in the same city. Inhabitants of a given city are, on average, more closely related to each other than they are to people living elsewhere. Do you think it therefore makes sense to talk about a "biological basis" of place of residence? Saying someone is a Dubliner may well convey more genetic information than saying he's Irish, but the latter description automatically makes us think that it's saying something more fundamental about the person's makeup than the former. Even for geneticists who know all the caveats it's very easy to fall into essentialist thinking about groups, and be misled in the process. Using racial language only makes the problem worse.