Comment by chiffonade | Hacker News Reader

Comment by chiffonade | Hacker News Reader

Parent Full thread

> And I'm not sure what you mean by East Asians having a higher but less creative IQ?

See, when racists use "average IQ" as a weapon against blacks and mexicans, they have to somehow justify their 2nd place to the Asians or else they will not feel comfortable using these "studies" as proof. They just say "oh, well they're less creative, so you know, white people are really still at the top", and everything is right again in the world.

Did you see what they did there? They took valid objective metrics - statistics - and disqualified the positive outcome for a selected class of people with something totally subjective - creativity. This is classic racism. It's a nonsense argument posited by younger people who really haven't had much life experience, and have not met many people who aren't of their same race.

Personally I don't believe in average anything when it comes to race, it's a totally meaningless metric that will get you absolutely nowhere when you deal with actual people and in some cases will handicap you with preconceived notions. I've conducted enough business with enough people in this economy to know that. Younger or less experienced people may have not.

> but aren't Mexican's mostly of Spanish descent

Racists never blame white Europeans for anything, especially not fundamentally flawed colonial systems like the Spaniards' that were devised by greedy people who were bent on raping native people and lands for as much wealth as possible in as little time as possible - actually, why should they blame them, it worked flawlessly, didn't it?

It's a nonsense argument posited by younger people who really haven't had much life experience, and have not met many people who aren't of their same race.

Agree with everything you said except for this. Most of the most racist people I know are 40+. Sadly, I don't think life experience tempers prejudice...often it just reinforces it.

> Personally I don't believe in average anything when it comes to race, it's a totally meaningless metric that will get you absolutely nowhere when you deal with actual people and in some cases will handicap you with preconceived notions.

I agree that you can't use race to judge individual people, but it's ridiculous to not "believe in average anything when it comes to race". Again, individuals must be judged on their own merits. But if only 20% of Hispanic students graduate college, versus 40% of white ones, you can't just automatically assume institutional racism. All human sub-populations are not equal. Check out this new book - http://the10000yearexplosion.com/

> All human sub-populations are not equal

You look down on hispanics, do you think Asian Americans look down on you because your racial "subpopulation" is behind theirs in college degrees, graduate degrees, and median income?

I don't "look down on" Hispanics - intelligence is not a moral value. And I've been around enough Chinese Americans to know that yes, a lot of them do look down on white Americans, for the aforementioned reasons as well as others.

> And I've been around enough Chinese Americans to know that yes, a lot of them do look down on white Americans, for the aforementioned reasons as well as others.

Yeah, so you know it's bullshit logic.

I feel like we're talking at cross purposes here. Here are my thoughts on the matter, as neatly as they can be summarized in a 1-minute forum posting:

1) We should allow & encourage more immigration of high-IQ, educated people. Said IQ would be determined by using IQ tests on individuals, not by racial/country of origin quotas. Exceptions would be made for people with outstanding achievement in some field - we shouldn't exclude a world-class dancer because he has an average IQ, for example, or a great programmer because he doesn't have a college degree.

2) We should discourage immigration by low-IQ people, regardless of how hard they are willing to work. Everyone would be assumed to be in the "low-IQ, uneducated" group until he has proven otherwise.

Do you disagree with either of these points?

All human sub-populations are not equal.

That depends on how big your sub-population is.

If your population is 1, then the difference between two populations, unless they happen to be maternal twins, will be HUGE. Statistically speaking.

But as your population size grows the difference between populations begin to shrink. Statistics again.

The reason for that is simple, humans are among a few species that happen to have very little genetic diversity.

Approximately 80 thousands years ago, something (bad) happened and the totality of humans on earth was reduced to a terrifyingly small number.

We don't know exactly how small but it could have been as little as a few hundred individuals.

And of those only 4 or 5 (I forget) matrilineal lines survive today.

Besides humans, cheetahs are one of the other species that is similarity genetically pure.

Not that genetic purity is a good thing, quite the opposite, it makes us susceptible to extinction by disease.

We could be wiped out by something like a more easily transmissible HIV.

Even a single tribe of chimps has more genetic diversity then all of humanity.

And that's why looking at the genetic difference of human populations is not very insightful.

Take one of the most famous reproductively isolated human sub-populations - the Ashkenazi.

DNA suggested that about 40% of the current Ashkenazi population is descended matrilineally from just four women.

As is common in cases of the founder effect: http://en.wikipedia.org/wiki/Founder_effect

A few genes that are very rare in a population, can be common in the population resulting from the few founders, if even just one of the founders had the rare genes.

And the Ashkenazi appear to have a much higher incidence of genes responsible for neural diseases but perhaps also increased intelligence.

So what happens if you pull a "Trading Places" trick, but with two babies instead of Eddie Murphy and Dan Aykroyd?

Nothing, remember humans sadly have a very shallow gene pool, so even an Inuit and an Ashkenazi could have more in common genetically with each other then the Ashkenazi with most other Ashkenazi.

And that's why talking about DNA in human populations, is like talking about the gold plating in Monster audio cables.

The people who do, aren't very smart about the subject they're ranting about.

Mitochondrial and Y-chromosomal DNA are not good ways of determining population bottlenecks, since they are inherited exclusively from the mother and father, respectively.

The "Two individuals from one race are more different than two random people" myth is a pernicious one that needs to die. The short answer is that you need to look at correlations in DNA, especially of regions under selection. Imagine a single human's DNA as a point in 4-billion dimensional space (where each dimension can take on the value A,C,G, or T). If you do a principal component analysis on the DNA of large numbers of people, you will see clusters, and these clusters correspond to geographic ancestry.

http://www.nature.com/ejhg/journal/v16/n12/abs/ejhg2008210a....

"A major East–West gradient from Russian (Moscow) samples to Spanish samples was identified as the first principal component (PC) of the genetic diversity. The second PC identified a North–South gradient from Norway and Sweden to Romania and Spain."

In picture form, from that paper in case you don't have access to Nature: http://www.gnxp.com/blog/2008/11/another-genetic-map-of-euro...

Now, please tell me again with a straight face that I don't know what I'm "ranting" about, and that DNA in humans is irrelevant.

I get paid to do PCA analysis for the biotech industry.

Except that PCA is the least I do, PCA is what you learn in a bioinformatics 101 class, there are MUCH better tools then PCA.

If two things are noticeably different you don't need PCA, it's redundant.

You use PCA, when you're plumbing a huge data set, where things are mostly indistinguishable and you're desperate for ways to separate things out.

That's why PCA analysis is so convenient for human DNA, because we're so similar.

Look at this image: http://scienceblogs.com/gnxp/genmapeuropeGermany.jpg

Are you telling me that the Germans from the blue part of Dresden bordering Munich and the Germans from the green part of Munich bordering Dresden are super duper different?

That picture sure makes it looks like Hans living in the exurbs of Munich and Frans from the exurbs of Dresden are two different species.

It looks that way if you don't have a fucking clue about what PCA is and how it works and what DNA is and how it works.

And two individuals from the same race are more different then two large groups of people from the opposite end of the earth, is not a myth.

It is a fact.

It is a statistical fact, the math doesn't care what you think and your grasp of stats and genomics is not nearly as tight as you imagine it to be.

No one imagines DNA as 4-billion dimension space, you're full shit.

And if you start looking at what ever regions you're fantasizing about, and apply a bunch more conditions, you can prove anything you want. That's not science it's charlatanry.

In conclusion, you do not know what you're ranting about, so stop embarrassing yourself in front of biotech engineers on the internet.

I intentionally simplified because I didn't know you were a biotech engineer. Also I'm impressed that you were able to divine how much I know about statistics from my few short posts here.

So you're telling me that if I do PCA on the covariance matrix for many loci, I'm not going to be able to correlate DNA with geographic ancestry?

How do you explain this: http://www.sciencemag.org/cgi/content/abstract/298/5602/2381

I will grant you one point - "Within-population differences among individuals account for 93 to 95% of genetic variation; differences among major groups constitute only 3 to 5%." However, if you read the paper you'll find that they were able to identify clustering along geographic lines, and could place individuals with a discrepancy rate of 0.14%.

Obviously you can correlate DNA with geographic location.

But in-population differences of individuals are bigger then inter-population group differences because all the populations are so similar.

And because the differences between the populations are so small you'll only be able to measure them, when you measure very, very, very narrow things. Like PCA does.

For example you could measure skin color. The amount of melanin deposited in your skin is controlled by one gene. So that's about as narrow a measure as you can have.

But can you draw conclusions about broader characteristics in phenotype?

Nope. That's why the whole race concept has fallen out of favor, not because it's a lefty conspiracy to avoid hurting people's feeling. It's not that, it's science.

The same goes for IQ and societal success.

Can you correlate education levels with a phenotype or genotype?

If you could, scientists would have already done it, and they would be famous. Parents would be screening their embryos. You could do it and be famous.

Again, it's not some lefty conspiracy, it's just not possible.

Oh we do have genes that seem to result in more scientists and engineers, the Ashkenazi have them, but they correlate with neural disease. And the Ashkenazi don't have a higher incidence of CEOs and politicians and musicians.

And who's on top of the societal ladder in western society? Politicians and CEOs or us geeks?

And who has a higher reproductive success, musicians or engineers?

That's why people talking about poverty, race and IQ, are like people talking about Monster audio cables.

Your assertions about IQ and Mexicans are not evident from the real world genomic data. But if you're prejudiced you can try and twist the facts to match your prejudices.

I think it's pretty well established that IQ is inversely correlated with reproductive success. It's also well known that IQ is not the only component of financial / societal success. Things like wisdom, creativity, diligence, or agreeableness can't be measured by an IQ test. I don't worship at the altar of IQ. That said, the cognitive abilities that an IQ test measures are becoming increasingly important for the types of jobs that make America competitive. And I'd rather live in a society where the average IQ is 100, than in one where it's 90 (relative to our current 100 blah blah blah)

Anyway, let's leave the race question off the table. Why is it that when large-scale IQ tests are conducted on people from Mexico, and on people whose parents are from Mexico, the average score is around 87?

Well it might be the weather, or the phases of the moon, or it might just be education!

If they couldn't read their IQ score would be 0.

Obviously the IQ test questions strive to be about general problem solving, not culturally specific factual knowledge, etc.

But it should be just as obvious that strive is all they can do. They just can not be divorced from a while lot of stuff that's learned.

Which is why IQ test are very highly correlated with education, they are also correlated with some natural ability, but it's impossible to tell which contributes how much.

East Asian along with us East Europeans, come from very rigorous public schools. I've attended public schools in Eastern Europe and Western Europe and the US. And I have Japanese friends, I met when we were both exchange students in a top notch US high school.

We were seniors and my Japanese friends and I were amazed how US seniors barely had a 6th grade level of science knowledge in 12th grade.

Now do you think my Japanese fiend and I had all that knowledge because of the systems of public education we came from, or because we were both genetically superior to the US kids?

So it's possible that fist and second generation Mexican immigrants score lower on IQ tests, because of genetics.

But it's much, much, much more likely it is because of a different educational background.

Even if the second generation attended the same schools, that doesn't make up for the educational deficit in their home.

And importing highly educated immigrants is a great way to reap the benefits of an education you didn't pay for.

And important any kind of humans, even illiterate ones, is a good way to grow the economy. Even the lowest of the low buy stuff, and because humanity is so closely related, they are not bringing down the national IQ in the long term.

Or rather they shouldn't be, because if your country is a well functioning state it should be able to educated people and reap the economic benefits from that in the long term.

If it can't do that, immigration is not your problem, your problem is you're barely able to come out even, and unskilled immigration is just making that problem worse.

Reply on news.ycombinator.com