Repost from the old site.
Commenters recently have been bringing up the Flynn Effect (FE) of rising IQ’s in the past 80 years. There are many paradoxes and controversies over this effect that the commenters have been sharp enough to notice and point out. One argument is that the FE is not a real intelligence rise at all, since it is not on real g intelligence. In other words, the FE means nothing, and people are no smarter than before. The FE gains are not gains in intelligence at all, they are zero, nothing, null.
This is not correct, but the argument is very interesting. First we need to understand what the FE is and what it means. Then we need to understand what g is and what it means. The post concludes that there has indeed been an FE rise on g, but only on one component of g. Further, the post critiques the whole notion of seeing intelligence purely through the lens of g as senseless and meaningless, not to mention flat out wrong.
The subject matter is highly complex, but I tried to make it as simple as possible. My readers are a very intelligent bunch, and I think most of you should be able to follow this argument. You really need to read this slowly and take your time to try to understand what is going on here. It took me months of studying the FE before I finally started to get a handle on it.
The Flynn Effect (FE) is a secular rise in IQ over time that has been occurring throughout the West for 80 years now. All ages and ethnic groups are effected. Preliminary evidence indicates that it is also occurring in the Caribbean (Dominica), South America (Brazil) and Africa (Kenya).
An overview of the FE itself goes beyond the scope of this post.
The FE is quite complex, and many people do not seem to understand the concept properly, hence are not able to discuss it, much less debate it. However, most people of reasonable intelligence, if interested, seem to be able to grasp the basic implications of the FE.
Hereditarians, most of whom are White racists, are very upset by the FE (Talk about being opposed to human progress!) because they have a strong emotional investment in White intellectual superiority and the intellectual inferiority of Blacks, Hispanics and other groups. We know that the major hereditarian researchers on intelligence are racists because almost all of them support getting rid of all anti-discrimination laws.
The agenda is clear for both lab coat racists and White nationalists: if we can prove that Blacks, Hispanics and others are intellectually inferior to Whites, we can legalize discrimination, especially job discrimination, against them.
A particularly frightening lab coat racist endeavor is attempting to prove that Blacks are inferior employees to Whites on average. If they prove this scientifically, then they will have a logical reason to support discriminating against Blacks in employment.
Almost all of these folks are White, and most of them call themselves race realists. They spend a good deal of time screaming and yelling about why Blacks and Browns will not accept that they are intellectually inferior (Steve Sailer specializes in this). Why would anyone want to accept, or actually accept, such a thing?
Furthermore, given the nefarious agenda behind those promoting these theories that seeks to legalize discrimination against Blacks and Browns, any Black or Hispanic person who gets behind this would have to be out of their minds.
I have ritually added “Hispanic” after Black above, but in general, Hispanics are being left out of this debate. The real effort here is directed by racist Whites against Blacks, not Hispanics. It is against Blacks that these Whites seek to legalize the right to discriminate.
The Flynn Effect has been hard to argue against, but the hereditarians have tried hard. They have shown that the FE is not on g. G is a hypothesized common correlational factor that supposedly measures pure intelligence. Everything outside of g is “not real intelligence”.
However, according to one paper, the very concept of g is tendentious to say the least, and possibly nonsensical. The paper is titled G, A Statistical Myth, by an admittedly brilliant mathematician named Cosma Shalizi. I read through the whole thing but I couldn’t really make sense of it. Perhaps someone who knows math better than I do can have a go at it.
Therefore, intelligence has not risen in a general, across-the-board kind of way. However, certain aspects of intelligence have definitely risen, and those aspects have quantifiable benefits in modern society, occupationally, academically and in other ways.
The argument of whether or not the FE gains are on g or not is very complex, about as complex as the FE itself. First of all, the FE gains have not been across the board. In general, they have focused on verbal analysis, visual analysis, visual intelligence and problem-solving. Gains have been few to none in basic things like general knowledge, mathematics, mathematical analysis, spelling and reading comprehension.
G is a hypothesized and problematic construct that is a correlational factor all of the subtests on an intelligence test. It is thought to be highly heritable and physiologically based, and this is why the hereditarians have gone nuts over it.
It measures how someone with a somewhat more neurologically efficient brain will tend to score better across an entire range of subtests than someone who with a less efficient brain.
The reason the FE is not on g is because it is limited to a subset of intelligence subtests, and gains have been small to none across another subset. Therefore, there is no g gain.
However, Raven’s Progressive Matrices has shown larger gains than any other test. Raven’s was designed to measure pure g and nothing else. Raven’s scores are not thought to be effected by environmental factors outside of pregnancy and the first few months of life, and are thought to be purely neurologically, physiologically or genetically based.
Given the pure g basis for Raven’s, the wild secular gains on it on the FE are most puzzling.
However, recently James Flynn has shown that the FE does show a .5 correlation on a factor analyzing fluid g.
There is fluid g and crystallized g.
Fluid g is thought to peak early in life. This is why things that require raw brain processing power tend to peak in young people: creativity – artists, musicians, poets, novelists, filmmakers, mathematicians, physicists – often do their best work as young people (usually as young men). Fluid g is really a measure of how well, fast or efficient your brain works.
Crystallized g is another matter altogether. Crystallized g may be seen as “what you know” as opposed to “how fast of a brain you have.” While fluid g peaks early, crystallized g often goes up throughout life, and people can still score high on crystallized g in their 50’s, 60’s, 70’s and even 80’s.
This is what we might call “accumulated knowledge” or “wisdom”. The old person’s brain does not work as fast, but the accumulated knowledge makes up for that in that they can see connections between things easier.
The young person’s brain works very fast, but with the lack of accumulated knowledge and life experience, they are not able to put things together as well to arrive at the correct conclusion. This is why no society has ever put the 18-23 year old’s in power, no matter how zippy their brains are.
Instead, the old men have always been put in power. The accumulation of a life of learning is thought to lead to a wisdom that will manifest itself as the ability to make “wise”, correct and proper decisions.
The concepts of crystallized and fluid g are complicated, but hopefully that explanation helped you understand it better.
The FE is on fluid g, not crystallized g. Visual intelligence and analysis, problem solving, verbal analysis analytical thinking in general, on intelligence tests, is in the realm of fluid g. Those are those little puzzles that ask you to decide which figure goes next in the series.
On intelligence tests, crystallized g measures accumulated knowledge and the degree to which one has learned basic tasks of modern life. The FE gains on mathematics, math analysis, reading comprehension, vocabulary and general knowledge are small to nil, and all of these tests measure crystallized g. To sum, these are the sorts of things you learn in school.
This is why, despite skyrocketing IQ’s, we cannot read a book, add and subtract, or do calculus any faster than our grandparents. We also do not know any more than they did, and we know no more words than they did.
This is obvious in the many reports on “idiocratic” state of high school seniors, college students or college grads. And this is how a puzzle is solved – how IQ’s are surely rising at the same time as idiocracy is.
From Flynn’s chapter summary:
IQ gains over time were calculated for each WISC (Wechsler Intelligence Scale for Children) subtest and the subtests ranked by size of gain. Verbal Similarities led at 20 points per generation – larger than gains on Raven’s Progressive Matrices.
Similarities measures on-the-spot problem-solving (something akin to fluid g); verbal subtests that do not measure this show low rate of gain. WISC subtests were also ranked by their correlations with Raven’s, the latter being used as a marker for fluid g. The r between the two hierarchies was calculated to approximate a correlation between IQ gains and fluid g .
The result of 0.50 contrasts with the negative correlation between IQ gains and the g generated by factor analyzing the WISC battery itself, which is generally viewed as predominately a crystallized g.
In sum, it appears that human groups can make massive fluid g gains in a period too short to accommodate radical change in the speed and efficiency of neural processes. Moreover, once gains in intelligent behavior over historical time are seen to be independent of brain physiology, does g really provide a criterion for assessing their significance?
Finally, not only a measure of fluid g (which is highly heritable) but also inbreeding depression are shown to be correlated with IQ gains – gains overwhelmingly environmental in origin. Therefore, correlations between such genetically influenced factors and the size of the black/white IQ gap do not show that the gap has a genetic component.
The Similarities test has risen faster than any other test. It measures analytical thinking and is thought to be a good measure of raw fluid g. The final paragraph is interesting. The Black-White IQ gap is correlated with something called inbreeding depression score, a purely heritable measure. Hereditarians use this to say that the B-W IQ gap is genetic.
But Flynn shows here that the FE (a purely environmental gain) also correlates with inbreeding depression, a purely hereditarian score. Flynn uses this to say that the B-W IQ gap is not necessarily completely genetic.
Flynn notes above that the FE cannot possibly be caused by brains that actually work better physiologically than the brains of our grandparents. Genetics doesn’t work that fast.
Therefore, what does the FE measure? Flynn says it measures “intelligent behavior.” So our brains don’t work any better than our grandparents’ brains, but we show improved “intelligent behavior” over them.
Therefore, another mystery is solved, how massive IQ gains can occur without concomitant improvement in the physiology of our brains.
Since hereditarians use g as a measure of physiological efficiency of our brains, Flynn calls this into question by noting that g gains can occur too fast to be accommodated for by physiologically improved brains. Therefore, Flynn suggests chucking g as a measure of pure brain physiological efficiency.
Therefore, the White nationalist and hereditarian argument that the FE is not on g has been proven wrong.
- Flynn, James R. 2000. IQ Gains, WISC Subtests and Fluid g: g Theory and the Relevance of Spearman’s Hypothesis to Race, Chapter 12 in Bock, Gregory R., Goode, Jamie A., Webb, Kate. Novartis Foundation Symposium 233 – The Nature of Intelligence, pp. 202-227. Novartis Foundation. Published online.