Lewontin’s Deception

by David Sims

THE IDEA THAT race is a social construct began in 1973 with a hypothesis proposed by a Harvard geneticist named Richard Lewontin (a Jew). He asserted that the differences between races were so small that nobody, working with genetic information alone, could tell the difference between a White, a Black, an Asian, an Arab, an Aboriginal, and so on.

Leftist activist groups of that time hastily invented a number of slogans for chanting in public, including “race is a social construct” and “there’s only one race, the human race” based on Lewontin’s claims.

By the late 1990s, genetic forensic science had advanced to the point where it was clear that Richard Lewontin had been wrong. By 2000, police were routinely identifying criminals by DNA samples found in blood, hair, and flakes of skin. In 2005, one of the definitive scientific studies on the demonstrable physical reality of race was published. Here it is:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1196372

In that study, researchers, using genetic information alone, accurately predicted how 3,631 out of 3,636 test subjects would racially identify themselves. That’s an accuracy of 99.86%, even if you assume that the test subjects were always right, rather than the genetic analysis. It is possible that the opposite is true: that five of the test subjects were in error about their own ethnicity, and the results of genetic analysis were correct in the five instances of conflict.

Leftists are fond of declaring that there aren’t any (or, anyway, there aren’t many) race-specific alleles: variants of a gene that only one race has. That’s true. All of the races of mankind have all, or nearly all, of the variants of every gene in their DNA. However, the proportions of those alleles differ from race to race.

Some alleles are better than others, in terms of determining a physical or mental character when expressed in an organism. Some races have a higher frequency of the premium quality alleles for any particular character than other races do. The Black race, for example, can boast of alleles that configure the legs and pelvis for quick bursts of running speed over short distances; e.g. the 100-meter sprint events at the Olympics.

Intelligence is another genetically determined character, and it is a character that is very important for creating advanced technical cultures and in adapting to living within them. In this case, the races that shine are Whites and Asians. Most of the brown races don’t do nearly as well, and the Blacks trail behind all of the other races in this respect.

Although someone’s race doesn’t determine his individual IQ, it does determine the probability for a randomly chosen member of a race having an average IQ being at, or above, a specified value.

The normal distribution that most closely matches the IQ distribution of White male US citizens is 103.08±14.54 (Jensen & Reynolds, “Sex Differences on the WISC-R,” Personality and Individual Differences, volume 4, number 2, pp. 223-226, 1983).

The normal distribution that most closely matches the IQ distribution of US-resident mulattoes (usually called “Blacks” or “African-Americans”) is 85.0±13.0 (a typical finding of studies since 1950).

A good approximation of the fraction, f, of a race having an average IQ of x̄ and a standard deviation in IQ of σ, which is above the minimum IQ of μ, can be found as follows:

f(μ) = [σ√(2π)]⁻¹ ∫(μ,∞) exp{ −[(x−x̄)/σ]²/2 } dx

Taking advantage of the normal distribution’s symmetry, we make it more easily integrable.

f(μ) = ½ − [σ√(2π)]⁻¹ ∫(x̄,μ) exp{ −[(x−x̄)/σ]²/2 } dx

You can avoid integrating the probability density function if you have a handy error function to call.

f(μ) = 1 − ½ { 1 + erf [(μ−x̄)/(σ√2)] }

Let us suppose that an employer wants to hire workers for a job that, in his opinion, requires a minimum IQ of 130 for satisfactory performance. He lives in an area that is demographically typical for the United States, where Whites outnumber Blacks by a ratio of five.

The fraction of Whites who are qualified for the job on the basis of IQ is

f( μ=130.0, x̄=103.08, σ=14.54 ) = 0.0320528311

The fraction of Blacks who are qualified for the job on the basis of IQ is

f( μ=130.0, x̄=85, σ=13 ) = 0.0002685491

If the population of Whites and of Blacks were equal in size, then the ratio of mentally qualified Whites to mentally qualified Blacks would be 119.355755.

Since Whites outnumber Blacks in the area where the employer’s business is, by a ratio of five, the actual ratio of mentally qualified Whites to mentally qualified blacks is 596.778775.

If the employer needs fewer than 100 new employees, it could very easily turn out that he will hire no Blacks at all, even if he uses no “racism” whatsoever in selecting them. In fact, of the occasions in which this scenario plays out, and exactly 100 new workers are hired, the employer will have hired…

100 Whites and 0 Blacks on 84.5% of occasions
99 Whites and 1 Black on 14.2% of occasions
98 Whites and 2 Blacks on 1.2% of occasions
97 or fewer Whites and 3 or more Blacks on 0.1% of occasions

Recognizing that that is the case, however, is Politically Incorrect to the extent that the law will punish employers if they fail to put racial bias (in favor of Blacks) into their hiring practices.

Let me indulge in a bit of math to show how a normal distribution can arise from biological inheritance.

Suppose that there are ten locations on the human DNA that pertain to intelligence. I don’t know how many such locations actually exist, but for the purpose of illustration, let’s assume that there are ten.

Suppose, further, that there are 11 different alleles (variants of a gene) that would be candidates to fit on each of those ten locations on a strand of human DNA. I don’t know how many such alleles actually exist, but we’ll assume for the sake of argument that there are eleven of them for each location.

We’ll imagine that some of these alleles are more potent in creating an intelligent brain than are some of the others. Indeed, we’ll assume that one of the alleles that can occur in a location will contribute no IQ points at all. This allele is a dud.

The second of the alleles that can occur at the same location provides 2 IQ points; the third provides 4 points; the fourth contributes 6 points; the fifth contributes 8 points; the sixth contributes 10 points; the seventh provides 12 points; the eighth grants 14 points; the ninth applies 16 points, the tenth bestows 18 points, and the eleventh and strongest allele kicks in 20 points.

If, by an amazing coincidence, a person inherited at conception the most potent allele at all of the intelligence spots on his DNA, he’d have an IQ of 200 and be a super-genius. But, of course, that isn’t very likely.

Just for example, suppose that, at conception, a member of Race A has a 9% chance of getting any of alleles #1-#5 or #7-#11 and a 10% chance of getting allele #6 at each of the ten locations. A random sample of one million individuals from Race A might have this spread of IQ:

Race A.
n=1000000
average IQ = 100.0
IQ range, number individuals in range:
0 to 10, 0
10 to 20, 7
20 to 30, 55
30 to 40, 671
40 to 50, 3724
50 to 60, 14606
60 to 70, 40997
70 to 80, 87945
80 to 90, 114389
90 to 100, 187760
100 to 110, 191867 ← biggest bin
110 to 120, 155869
120 to 130, 98898
130 to 140, 48651
140 to 150, 18460
150 to 160, 5012
160 to 170, 969
170 to 180, 113
180 to 190, 7
190 to 200, 0

That was from an actual computer run using a random number generator. The random numbers were without weighting, but the way they add up produces a normal distribution.

Suppose again that a member of Race B has an 11% chance of getting any of alleles #1-#6, a 10% chance of getting allele #7, and a 6% chance of getting any of alleles #8-#11 at each of the ten locations. A random sample of one million individuals from Race B might have this spread of IQ:

Race B.
n = 1000000
average IQ = 85.8
IQ range, number individuals in range:
0 to 10, 1
10 to 20, 21
20 to 30, 456
30 to 40, 3709
40 to 50, 17788
50 to 60, 53662
60 to 70, 113931
70 to 80, 176219
80 to 90, 205926 ← biggest bin
90 to 100, 186466
100 to 110, 129347
110 to 120, 70351
120 to 130, 29718
130 to 140, 9648
140 to 150, 2313
150 to 160, 394
160 to 170, 47
170 to 180, 3
180 to 190, 0
190 to 200, 0

Now, of course, this is a very simplistic representation of genetic inheritance, using numbers instead of atoms and molecules. But the principle is about the same.

When leftists declare that all races share all alleles, they’re right. (Or almost right.) But then the lefties try to make that fact appear to mean more than it really does. The races share alleles for intelligence, but they don’t share them all equally. Some races have more of the premium quality alleles than other races do. And that is why some races are generally smarter than other races are.

* * *

Source: Author