ImBack
Wizard
Joined: 28 Jun 2006
{Posts: 587 }
|
 Posted: Tue 26 Aug 2008 06:17 Post subject: DOES RACE EXIST ACCORDING TO FST STRUCTURED GROUPING? |
 |
|
Hey Frank,
I wanted to continue. I recently lost a debate to someone regarding race because they showed me the rosenberg study and a few others which, I must admit, were much more powerfull then anything i had up my sleeve. Now, are you telling me I ought to have won?
| Quote: |
FRANK SAID
In short, even if you divide our species into millions of tiny "races," you will always find that there is more variation between sub-groups within each of those millions of "races" than there is between the "races" themselves.
The exception? You will reach the holy grail of "racial" definition if you divide humankind into 6.5 billion "races" of one individual each. Only then does inter-group variation exceed intra-group variation. The variation between different individuals is greater than the variation among the cells of each individual.
In conclusion, you can show objectively, replicably, mathematically, that our species comprises only one "race" of 6.5 billion individuals. Or you can show that we comprise 6.5 billion "races" of one individual each. But no one has ever found anything in between. |
Okay, I of course would agree with you but for the problem that as I understood these recent studies they did not find APPRECIABLE difference beyond 2% FST. So if you keep subdividing you get down to the base difference in proportion of variance between individuals in the population which is 2% or so for most of the populations studied (I think). Theres no point in trying to subdivide further because you won't be able to structure any meaningful population on geographic grounds because at that level genetic variation is I suppose, almost random within the larger group.
What I am getting at, is that while it's true you can keep subdividing past 5%, you can only do so down to about 1-2% and that's it. So you would never make it down to 6.5 billion individual races if what I am saying is correct. Here is a summary of my points.
1. Even with HUGE samples structuring programs can group humans into populations based on genetic distance between them.
2. Each grouped population has a base level of natural variance between individuals so that once you've divided them into groups with this FST you cannot proceed further.
3. The fact that there is greater variation within groups than between groups, even if it were true down to 6.5 billion races, (which 1-2 deny) obscures the important issue that there are still LARGE differences between some geographically defined populations, and only SMALL differences between their subdivisions.
4. Points 1-2 are sufficient to establish that races of a sort exist, IF I AM RIGHT ABOUT POINT 2. Points 1 and 3 are sufficient if I am right about point 3.
I don't see how it can be argued otherwise so long as I am correct about either points 2 or 3. What do you think? |
|
'
FAQ
Search
Register
Log in
