[ExI] Race Biology

[Rafal Smigrodzki]

On Fri, Mar 21, 2008 at 6:36 PM, Damien Broderick <thespike at satx.rr.com>
wrote:

> At 05:27 PM 3/21/2008 -0400, Rafal wrote:
>
> >The primary meaning has to do with de[s]cent, and identifiable
> >differences between groups of different descent within a potentially
> >interbreeding population. These are strictly biological facts, in
> >the sense that an automaton completely ignorant of culture would be
> >able to analyze and properly segregate organisms strictly by
> >investigating their inherited traits.
>
> By "properly" I assume you mean "reliably" or "replicably" rather
> than "appropriately." The distinction is important. What  canon of
> criteria, independent-of-human-history, does the automaton use to
> select the major traits it uses to characterize individuals into 3,
> 30 or 300 "human races"? For example, to what extent are the tens of
> thousands of traits of people called "Negroes" or "blacks"
> cross-correlated, and would the automatons rediscover this "racial
> group" as automatons have rediscovered certain theorems of set
> theory?


### Yes. You can randomly select neutral polymorphisms (i.e. variations in
coding regions of the genome which do not change protein sequence, and
therefore can be assumed for the most part to be neutral with respect to
phenotype and fitness), run a cluster analysis, and the common racial groups
pop out without any additional input. See here:
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1196372. Please
note here: Self-reported race correlates with the clustering (which was done
without any input about phenotype) in 99.85% of cases. So there is a
genetic, heritable quality to all of us, almost all of us know it, and call
it "race".

Another study using the same methodology and looking not only at the US but
at populations widely scattered across the world found clustering in five
groups (
http://www.plosgenetics.org/article/info:doi/10.1371/journal.pgen.0010070#journal-pgen-0010070-b001)
corresponding to racial groups recognized previously, as well as some
populations which form their own smaller clusters and could add to the count
of races. Lumpers and splitters can have a field day here but one way or
another, commonly recognized races can be re-discovered in complete
ignorance of culture, phenotype and geography.

-------------------------------------------------------------


> *We* know they nearly all have visibly darker skin that some
> other humans of historically different geographical descent, of
> course. Do they all have thick everted lips? No. Wide flattened
> noses? No. Kinky hair? No. Specific blood types? MHC genes? Fast
> twitch muscles? Weeeeellll--of course, those people traditionally
> labelled "black" whose ancestors lived fairly segregated lives in
> regions of Africa or Papua or Australia or India for many many
> generations will tend to share some but not all of these traits.
> Plenty of other people of different historical descent in widely
> separated parts of the planet will share other traits. What's a poor
> mindless automaton to do? Look at tooth size? Fingernail shape? Exact
> hue of liver? Height? Girth? IQ? Ah, look, here's the "race of
> Mensans" popping out! They all do great on tests, and you can tell
> them by their distinctive skin color... well, no, wait, it's true
> that many have ancestors from China and Japan, but that's also true
> of a very large number of fairly ordinary or stupid people.
>
> ### Morphometric analysis can also be done, and will in general agree with
random genotype testing.... but of course, why bother. IQ testing alone is
insufficient to determine an individual's descent, although it does provide
some clues, e.g. very high IQ makes it more slightly more likely that the
individual is Ashkenazi, or Asian.
Rafal
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20080322/88b358ab/attachment.html>