The existence of a ~1S.D. B/W gap in IQ is beyond dispute.
It is probably the single most studied and replicated finding in psychometrics. The graph of all the different studies of the B/W gap themselves form a bell curve, with Black IQ centered around 85.
Here is Emil Kirkegaard’s plot of John Fuerst’s massive (although unfinished) metastudy:
The only debate now, and since the past few decades, is whether the source of this difference is primarily environmental, cultural, or genetic.
The preponderance of the evidence now indicates it is mostly the latter (at least within socio-economically homogenous countries, such as the US), and in the next 5-10 years, further GWAS will very likely confirm it.
And that is a good outcome.
If the environment was to blame, it would imply that the already huge interventions to raise Black performance to White (and Asian!) levels had barely scratched the surface, so the only option would have been to abandon any further dreams of social justice or to embark on social engineering projects on cardinally bigger, possibly dystopian, scales.
If culture was to blame, it would have validated conservative critiques of Black and Hispanic lifeways and actually constituted an argument for a intensive imposition of White (or Asian) social standards and expectations on under-performing groups.
If, however, the ultimate cause is genetics, then nobody is to blame and we can all go about in peace rationally discussing the best way we can adopt to biological realities to everyone’s mutual benefit.
Here is Charles Murray himself, the conservative/libertarian, on this topic in The Bell Curve:
If intelligence plays an important role in determining how well one does in life, and intelligence is conferred on a person through a combination of genetic and environmental factors over which that person has no control (as we argue in the book), the most obvious political implication is that we need a Rawlsian egalitarian state, compensating the less advantaged for the unfair allocation of intellectual gifts.
The liberal Steven Pinker, in The Blank Slate:
Can one really reconcile biological differences with a concept of social justice? Absolutely. In his famous theory of justice, the philosopher John Rawls asks us to imagine a social contract drawn up by self-interested agents negotiating under a veil of ignorance, unaware of the talents or status they will inherit at birth — ghosts ignorant of the machines they will haunt. He argues that a just society is one that these disembodied souls would agree to be born into, knowing that they might be dealt a lousy social or genetic hand. If you agree that this is a reasonable conception of justice, and that the agents would insist on a broad social safety net and redistributive taxation (short of eliminating incentives that make everyone better off), then you can justify compensatory social policies even if you think differences in social status are 100 percent genetic.
Robert Lindsay, that rarest of breeds, an HBD-realist Leftist:
Here is the conundrum for Left-liberalism:
Just supposing that there are differences between the races that are not caused by oppression, racism, etc. This is painfully obvious to anyone who will look. The Left refuses to look, because the reality of the whole mess is bad for the Left. So we say it doesn’t exist, unscientifically. We wish the reality away. …
Suppose Blacks had the same abilities as Whites, genetically.
All of the problems, including low IQ, were simply due the fact that they are fucking up, often on purpose. If this were true, and strangely enough, this sort of follows from liberal beliefs about genes and environment, I would argue for a harsh response to Blacks. Not necessarily cutting them off altogether, but I would certainly be a bit less likely to help them.
But there’s no evidence that that is true.
If Blacks do have low IQ due to things they cannot control, then, as a socialist, I would argue that there is no reason that the higher IQ group ought to obtain dramatically higher income, wealth, housing, living spaces and health than the lower one.
As much as possible, socialists should try to attempt to more equalize incomes, housing, living spaces and health care access for both groups, the higher IQ and the lower. …
Why should Whites be allowed to become dramatically richer, healthier, better housed, and live in better places than Blacks, simply because of how the genetic dice got rolled?
Answer: They have no such right. If both groups were equal, and Whites got that way by simply trying harder, then we could make the argument that the White position is just.
Why should Blacks be forced to become dramatically poorer, less healthy, worse housed, and live in worse places than Whites, simply because of how they were born, a variable that they had no control over whatsoever?
Answer: This is not right. It is not just. They should not be forced into these outcomes, and that they are is an outrageous injustice.
There seems to be a broad agreement across the thinking parts of the ideological spectrum that “social justice” (sanely defined) is both perfectly compatible and possibly even more defensible under an HBD-realistic lens.
However, the Pink Guards are utterly uninterested in looking at things from such perspectives, and it just so happens that neither will the Black Shirts they might eventually conjure up into being.
Welfare payments will always be unpopular with taxpayers. Make work is more palatable. The streets could be cleaner. Phones could be answered by more people and fewer machines.
And I don’t know if the futurist stuff that you’re interested in will come true. The people who were young in 1950 expected to see Mars colonies and flying cars within their lifetimes. If it does come true, this problem will be replaced by new, completely unrelated ones.
Are blacks less intelligent than whites on the average? It sounds likely, though I would be very happy if these studies were to be proved wrong. In that case, there will always be an income gap, yet I totally disagree with Murray’s recommendation. It’s not my problem. Forced redistribution of income will always be wrong.
You’re confusing “genetics” and “heredity”. For example: coal mining is highly hereditary, but obviously there’s no such thing as a ‘coal mining gene’.
This confusion makes your subsequent reasoning kinda pointless.
As I understand it, the rule of thumb is that measurable outcomes (material and psychological) are explainable by both heredity and socialization, in more or less equal parts. Of course there must be threshold effects, so that if you are particularly (un)fortunate genetically or socially, this has preponderance. With this in mind, I agree with the leftist HBD position that it is reasonable for society to intervene on behalf of genetically disadvantaged subpopulations, but I would add that such interventions need to be carefully thought through so that they have a positive effect on the social environment of those subpopulations. Although an intervention is not going to affect genetics, it may very well affect the social environment, and one should not forget that this also is important.
This is a pretty asinine complaint. Obviously, when the subject is biology, the word ‘heredity’ refers to genetics–not your last name, or your family’s property, or your social class.
You’re making a circular argument.
Intelligence is obviously hereditary, but the degree of genetic influence is unknown. Countries like Korea that went from Africa-tier to Europe-tier IQ’s in one generation is an obvious counterexample.
Many things that are hereditary are not genetic: nutrition or child rearing habits, for example.
I’d imagine that for IQ the critical period is from conception to 3 years, raise your kid wrong and it’s probably already too late to make him smart.
It’s quite probably genetic. I’m sure you’re aware of the many interventions tried, which included adoption, daycare with constantly two caregivers per child, and similar such things. It’s basically impossible for it not to be largely genetic.
You never did respond to my comment regarding a likely and mismeasured environmental source of at least part of the IQ discrepancy. In particular, a sporadic dosing (interdosing time much greater than biological half-life) will yield a blood level that is poorly correlated to integrated dose. I await your reply, here or there.
It is with some hesitancy that I am commenting here, as I failed to keep a promise to Astabada, and will not have time to fulfill that promise in the coming year, writing my doctoral thesis and all. I’ll make some observations here with regards to Karlin’s writings, regarding HBD and regarding a recent post by him regarding the Malaysian plane shot down in the Donetsk area. I shall then shrink away again, until I get to providing previously promised items.
@Karlin
1. I understand that you have a non-superficial background in mathematics. I was curious for a long time about HBD and the like, so I read all I could, including much of your, Steve Sailer and others’ writings. I also engaged a number of relatively anonymous HBD proponents with my initial understanding, based on alternative models. Their criticisms showed that the alternative models regarding crime and IQ, as stated (lead poisoning, et cetera, with their associated dosing/epidemiological models, about which more below) cannot account for the data. I therefor regard the genetic causation hypothesis as a working null hypothesis, and revisited an old objection of mine to the lead epidemiology, namely that the biological half-life is on the order of a month.
One may approximate the blood lead level (BLL in what follows) as (LaTeX notation for all math)
$$\frac{\mathrm{d}\,BLL}{\mathrm{d}t} = R_\mathrm{intake} – \frac{BLL}{\tau}$$
where $\tau$ is the biological half-life over ln2. Strictly speaking, one needs four coupled ODEs to account for blood, soft tissue excluding brain, brain and bone, but blood+soft tissue dominates the lead account as enters the brain. Lead poisoning occurs discretely in time, i.e. at time $t_n$, dose $D_n$ is administered. Let dose $D_n$ be described in terms of additional short-term BLL. The intake fuction is
$$R_\mathrm{intake} (t) = \sum_n D_n \delta (t-t_n)$$
where $\delta(t)$ is the Dirac delta function, and the $D_n$ corresponding to the earliest $t_n (= 0)$ is the maternal BLL at conception. This allows one to solve for the BLL
$$BLL (t) = \sum_n D_n u(t-t_n) e^\frac{t_n-t}{\tau}$$
where $u(t)$ is the Heaviside function.
In the automobile fuel dominated cased, the time between subsequent poisonings is on the order of hours, i.e. much less than the biological half-life. Therefor, irrespective of the appropriate dose integration, if only automobile fuel fallout lead poisoning occurs, blood lead level (BLL) is highly correlated with delayed versions of itself (due to consistent habits of exposure, e.g. driving; I refer to the autocorrelation function of the BLL) and with total expected dose; note that IQ loss due to poisoning is usually during infancy. If the expected time between subsequent poisoning is much greater than the biological half-life, the BLL may drop several powers of 2 relative to peak, and the correlation between BLL at the time of measurement and total dose may easily fall well below a tenth; ditto the BLL autocorrelation time function with substantial delays. A consequence of this is that very substantial poisoning may be obscured by time, e.g. an infant suffering one severe poisoning of 102.4 $\mu$g/dL (microgram per decilitre) will have a BLL of 0.1 $\mu$g/dL$ a year later, without intervention; presumably the brain damage will be due to the 102.4 $\mu$g/dL right after poisoning, rather than due to the 0.1 $\mu$g/dL at the time of measurement. Note also the need for very large $n$ in recent US lead poisoning / IQ studies, as well as correction for the usual HBD variables (SES and parental IQ as proxies for race), as well as the race data for the highest BLLs in NHANES III.
In the sporadic lead poisoning case, one may wish to calculate the probability density function (PDF) of BLL for a given total dose. I will only give an outline, and a very approximate (but revealing) result. This result is of interest, as it is quite comparable to NHANES III blood lead level data. NHANES III contains two data sets, namely prior to 1990 (but after the initiation of the phaseout of leaded petrol/gasoline in USA), and after 1997 iirc. The NHANES BLL data is single sampling of e.g. infants within a given year-bound age bracket, and may thus be modeled as a randomly timed sample with a uniform distribution within a year.
To estimate the PDF of BLL for a consistent but sporadic level of poisoning, assume that the poisoning is a stationary Poisson process, but with $D_n$ consistent. My simplification for this post is that one assumes one poisoning, at the beginning of a year, with the time of measurement being some time later, uniformly distributed between immediately after and a year after the poisoning. The cumulative distribution function is logarithmic, so the PDF is inversely proportional to the BLL. Compare this to the NHANES III data for most states, especially after 1998, and contrast this with the near-Gaussian NHANES II data, during the use period of leaded petrol. The consistency of nearly inverse BLL histograms is a phenomenon pointing to sporadic poisoning of a large subset of the population, rather than a population consisting of a tiny minority of highly poisoned individuals and largely unpoisoned supermajority.
All of this points to substantial remaining gains in black IQ (and lower end of distribution white IQ) in USA, as cleanup largely started in 1992. Note also that cleanup with dry sanding (thus causing a dust hazard in a neighbourhood) is still a problem, federal LEADSAFE regulations notwithstanding. I had predicted that 2015 would see the start of a long exponential decline in US murders, but instead 2016 saw a rise—probably due to additional poisoning due to liberation of lead bases and salts in paint, due to dry sanding. The Chicago data suggests that the peak has passed, so it should be smoother sailing from here, as youth crime is substantially down, with murders by 11 year old perpetrators at the time of perpetration down more than an order of magnitude in USA, 2013 vs 1993 iirc wrt dates, but the decades for sure.
I have a second issue, with regards to attractiveness and HBD. I notice that on an individual level, Steve Sailer is more objective than you—he at least admits of individual attractiveness (and associated higher expected IQ—see a comment by him regarding ahem a “Nubian princess” re a certain BLM activist). But the issue is more general. If attractiveness is deficient to members of another ethnic group due to genetic difference, then that should not affect intra-ethnic attraction too much, and I believe the HBD argument is recent/incomplete speciation to explain black-white pairings. If this argument is valid, then we can do a simple test.
If differences in attractiveness is due to speciation difference, one may expect on the basis of koinophilia that the aligned visual mean of several faces of the unattractive ethnic group should remain unattractive, as the difference in phenotype would be the problem, rather than individual homeliness. A simple sociological test could be done, using e.g. Mexican and say Southern African mean images, to rate attractiveness of the mean within another ethnic group, e.g. a white European ethnic group. Do I need to spell out the application of the modus tollens, and the likely null hypothesis?
I have a third issue, with regards to the application of random variables by HBD enthusiasts, to IQ distributions. Take for example the careless talk of the central limit theorem. If a given cause of IQ (or any random variable) fluctuation is distributed in a binary fashion, with the cause (genetic or environmental) having a deterministic and large effect relative to the unaffected population’s standard deviation, the resulting distribution of the population, affected and unaffected taken together, will be bimodal, even if the separate distributions are mean shifted identical perfectly Gaussian distributions. A Gaussian added to a binary may yield something grossly non-Gaussian.
Another example that suggests poor application is smaller effect causes—if a given cause of binary fluctuation yields a small difference relative to the unaffected standard deviation, then it should not be visible in the resulting probability density function—adding statistically independent random variables produces a random variable whose probability density function is the convolution of the individual probability density functions. A concrete example will illustrate. Allow that the genetic standard deviation of intelligence corresponds to 80% of observed standard deviation in a given population, or 12 of the 15 points; 9 points in this case will be environmental variation (12^2+9^2=15^2). Any binary distribution that causes a differential effect smaller than 12 points will not make a visibly non-Gaussian total PDF. Compare this to the NLSY black IQ data, and the varying slopes of the PDF, and remind yourself of the PDF of a function of a random variable; higher lead poisoning has a lower differential IQ loss per unit additional poisoning—what does this suggest about the width of the intra-black genetic IQ distribution, or do I need to spell it out?
I have a fourth issue, with regards to a recent article by you claiming that the JIT explanation is likely. I’ll leave aside that you merely stated that ex cathedra and provided no argument against the arguments made against the JIT, but rather allude to the lack of competency of the JIT’s opponents in operating the equipment in question—does one need to know how to operate the equipment in order to understand the specifications? If so, elementary Newtonian physics must be beyond the learning abilities of the vast majority of the above 120 US2000 normalised IQ population, at least until they learn to operate special machinery, and engineers don;t qualify as humans
excellent point imo
also – once the truth is publicly accepted, women in the current underclass will change their breeding habits and the average IQ of future generations will slowly increase
James Thompson (an academic working in the field) addressed just this issue a few days ago, here:
“7. Most measures of the “environment” show significant genetic influence
…..
8. Most associations between environmental measures and psychological traits are significantly mediated genetically
He also, in the same piece, pointed out that this is far less the case than we might have expected:
“5. The heritability of intelligence increases throughout development
This is a strange and counter-intuitive finding: one would expect the effects of learning to accumulate with experience, increasing the strength of the environmental factor, but the opposite is true.
Shared environmental effects, as from family life and school, decrease with age. Good family lives and good schools are not the essential start in life that many people have always imagined, or at least not crucial in societies where family life and schools are reasonably good.“
Let me add that if shared environment’s effects are the larges at an early age, than this simply means that how good parent you are will have little effect on how well your child will turn out later in life. However, it does mean that you have a great influence in the here and now – how your child behaves right now. This means that parenting should be about making your life (and, of course, that of your child) as good as possible. By playing with your child, perhaps you won’t make him into a more successful or happier person later in life, but you can make him happier right now. Interestingly, you will probably also become happier – playing with your child is fun and can make you happy.
(Of course, occasionally, which means quite often, you will need some time out – take your child to daycare or kindergarten or organize him programs where he can leave you alone, letting you reading books, doing sports, watching movies, commenting online etc. Then you’ll have more energy and inclination to play with the child when next time he’ll be at home with you. But that’s a different story.)
Also, forcing him to be polite to – but also, for safety’s sake, suspicious of – strangers will make both your and his life easier, just as forcing (or convincing in any way) him to be well-behaved in general.
Coerced redistribution relies on a separate line of argument. You could have people who are naturally disadvantaged, whether by IQ or something else, and accept that there is a natural social hierarchy but also have a culture of noblesse oblige whereby the higher ups look after the lower downs. The arguments against coercing these obligations would still apply. For example, even if we grant that intelligence helps a lot in getting ahead, we still want to encourage those people to work hard and get ahead as best they can, while guaranteed welfare will undermine whatever work ethic they still have after realizing that they’re probably too dumb to get truly wealthy.
Also, looking at it as a social engineer, how much do you really want to subsidize what are essentially the unfit? Wouldn’t it be more rational to reduce the population of the low IQ folks, if you’re looking out for the long-term survival of the whole society? I can see a humanitarian case for keeping the stupid on welfare, as long as they are sterilized or otherwise forbidden to propagate.
Oxford Dictionary of Biochemisty and Molecular Biology: “Hereditary: of, pertaining to, or caused by heredity”. “Heredity: the transmission from one generation to another of genetic factors that determine the characteristics of an organism”.
Not saying they should be subsidized, that is up to people to decide however they want to decide it.
I was wondering about the definition of sanely defined social justice. It seems to me that any attempt to institute social justice creates a whole slew of moral hazards. And, while there are sane ways to address these hazards, like
few sane people believe them to be practically implementable.