This was originally going to be a review of What is Intelligence?: Beyond the Flynn Effect, by James R. Flynn. However, like so many of my reviews, it quickly became something else – I realized that I couldn’t fully address what I found important about Flynn’s ideas without drawing also on his earlier work, Race, IQ, and Jensen. Also, his more recent theories are summarized very well in the article “Heritability estimates versus large environmental effects: The IQ paradox resolved”, which is much shorter than his later book. I will also reference some things Thomas Sowell has written, for reasons I’ll explain later.
(Incidentally, for purposes of this post, I will be assuming that the reader correctly understands the concept of heritability.)
James R. Flynn is one of the most important intelligence researchers in recent years; whether you think “second only to Arthur Jensen” or “even more important than Arthur Jensen” is a better end to this sentence probably depends on whose theories you think are more plausible. He is known for publicizing – but not discovering – the “Flynn effect” – the large, long-term increase in IQ scores seen in many countries throughout the world. He didn’t name the Flynn effect, incidentally; that was Charles Murray, known for writing The Bell Curve – there’s a high level of civility between researchers who strongly disagree with each other.
There are some things about intelligence that are not seriously disputed by mainstream researchers:
- IQ is highly heritable as measured in twin studies; more than 50% of the differences between individuals are associated with differences in genes.
- There are fairly large differences in measured IQ between people of different races; in particular, black Americans score about 10-15 points lower than white Americans do.
- IQ at least roughly measures some things that we think of as “intelligence”, and it is highly correlated with educational and professional outcomes such as grades and income, as well as less traditional measures of success such as health.
- Many popular arguments against IQ tests, including “cultural bias” and “stereotype threat”, have been disproven.
That leaves quite a few things that mainstream researchers don’t agree about:
- Does IQ, and in particular a refined measurement called the “g factor”, measure a single, central quality about someone’s brainpower?
- What exactly does heritability as measured by twin studies tell us about the role of the environment?
- Why do IQ scores vary between races, and are these differences “innate”?
- What is the explanation for the Flynn effect?
The most common view among experts is probably what we might call the “Jensenist” view, after Arthur Jensen:
- The g factor represents a single, central, coherent aspect of someone’s cognitive ability. And so, IQ tests are a fairly precise measure of intelligence.
- The high heritability of IQ shows that differences in IQ within broad social groups (read, “among people of the same race”) are innate and genetic.
- It is extremely difficult or impossible for twin studies to show whether racial differences in IQ scores are innate or genetic, but there is no plausible environmental explanation for differences as large as we see, so the simplest assumption is that the differences are at least partly genetic.
The last of these three beliefs is by far the most controversial, and so I think we could probably divide Jensenists into two kinds:
- Non-racial Jensenists, who believe that intelligence is mostly innate and genetic, but who believe that discrimination and other race-specific environmental factors explain most or all of the differences between races.
- Racial Jensenists who believe that intelligence is mostly innate and genetic, and that this is true both within and between races.
(Jensen himself wasn’t 100% sure what kind of Jensenist he was, though he clearly leaned toward the latter. The “non-racial Jensenist” theory is the view Richard E. Nisbett seems to defend, although it’s possible he’s simply accepting the more common positions on 1 and 2 so he can focus his arguments on 3.)
James Flynn was the first to call attention to a series of problems that rising average IQ scores over time present to both kinds of Jensenist. The problems all stem from the fact that the gains in IQ scores over time have been very large – so much so that people whose IQ scores indicate average intelligence today would be considered near-geniuses in the 1930s, and people from the 1930s whose IQ scores indicated average intelligence in the 1930s would be considered severely mentally disabled by today’s standards.
- Even if we believe that people have gotten somewhat smarter over time, we cannot possibly believe that people have gotten that much smarter – books written in the 1930s don’t seem to have been written by imbeciles. So the first innatist belief – that IQ tests measure intelligence – can’t be universally true.
- Likewise, it isn’t reasonable to believe that genes have changed enough during that period of time to account for gains in IQ. So the second innatist belief, that heritability implies that the environment doesn’t matter, can’t be universally true either.
- The third innatist belief – concerning race – is an extension of the second, so if the second innatist belief isn’t universally true, there’s no reason to believe that the third and most controversial innatist belief is true, either.
In one stroke, the Flynn effect put Jensenists on the defensive for the first time since the before they were called Jensenists. Their story had never been airtight, but it had always had the advantage of being simpler than the alternatives. Forcing the Jensenists to explain changes over time separately from differences among individuals makes their story just as complex as any other, taking away their one great advantage.