One of the things that gives me the greatest confidence in the diametric model of cognition is the way that it makes sense of what have so far seemed like insoluble paradoxes relating to intelligence in general and measures of IQ in particular. I have already posted comments about how distinguishing between mentalistic and mechanistic intelligence can readily explain the paradoxes of both the Flynn effect and of racial differences in IQ. Now I can reveal that the same solution resolves yet another: the paradox of the high fecundity of seemingly less intelligent women.
Throughout the first half of the twentieth century, authorities as diverse as the philosopher Bertrand Russell, the statistician, R. A. Fisher, and psychologists too numerous to mention were predicting dire consequences for society as a result of the fact that more children were born to mothers of lower intelligence than to those of higher IQ. According to Russell:
There can be no doubt that … civilization … has this singular characteristic, that in proportion as men and women absorb it, they become sterile. The most civilized are the most sterile, the least civilized are the most fertile, and between the two there is a continuous gradation. At the present, the most intelligent sections of the Western nations are dying out.
Even today, parents with lower IQ do indeed have larger families, with the average IQ of mothers of 5-child families being about 20 per cent less than that of mothers who have only one child.* Furthermore, about 20 per cent of women remain childless in the US and UK—and this figure is more like 50 per cent in the highest-paid occupations which also tend to go with the highest IQs.* Nevertheless, the measured IQ of Western nations has risen inexorably. Indeed, these figures suggest that the Flynn Effect would have been even greater had it not been for the countervailing effect of this inverse relation between female fecundity and IQ. What on earth is going on?
As with so many other paradoxes relating to the fraught issue of IQ, the diametric model of intelligence can make sense of it. According to the model, there are two parallel modes of cognition—mentalistic and mechanistic—each with its own implied measure of intelligence. Raven’s matrices and suchlike tests are quintessentially mechanistic, and as the model would predict, autistics often do surprisingly well on these, even if they do abysmally on other tests.
A case in point is a test of mentalistic intelligence not currently included in proprietary IQ tests but used in the diagnosis of autism: Reading the Mind in the Eyes (RME). Given a picture of someone’s eyes only, the test asks which of the four words offered best fits the expression (below).
Women averagely do better than men on the RME, and recent research by Jennifer Bremser and Gordon Gallup reveals a strong correlation between high RME scores and low waist-to-hip ratio (WHR, below).
WHR is defined as the ratio of the narrowest measurement around the waist to the largest measurement at the greatest protrusion of the buttocks, and correlates strongly with fat content of the body. Differences in body fat distribution between men and women are at a maximum during early reproductive life. After puberty estrogens stimulate women to deposit adipose fat (which is critical for fertility) on the buttocks, thighs and breasts. Testosterone in men, by contrast, causes them to lose fat from buttocks and thighs after puberty, and begin depositing it on the stomach, shoulder and nape of the neck. WHR is similar for the sexes before puberty, but afterwards women’s range from 0.67-0.8, whereas men’s range from 0.85-0.95.
Studies in which subjects were asked to judge the attractiveness of female body profiles with differing WHRs showed that, although women tended to prefer thinner profiles to men, both sexes preferred profiles with low WHR (0.7). A comparison of older and younger men’s preferences showed that, although younger men had more of a preference for the thinnest profiles, neither age group inferred the reproductive capability of a woman from weight or fatness alone. No subject found female profiles with WHRs typical of men attractive. Subjects rated heavier profiles as older, independent of WHR, perhaps explaining why normal weight profiles were generally preferred to lower or higher weight ones.*
The same is true of cultures where heavier profiles are preferred. The prehistoric figurine illustrated here has a WHR estimated to be 0.7, suggesting that even in what probably were ice-age conditions where extra body fat may have been positively valued and perhaps associated with youth rather than age, ideal WHR fell within the same range that it does today.
WHR is an accurate measure of fertility. Women with higher WHR and lower body weight are less fertile than those with the contrary indications. WHR is also a good indicator of health overall: higher WHR indicates increased risk of death in women independently of weight.
Data from the Third National Health and Nutrition Examination Survey reveal that, controlling for other correlates of cognitive ability, women with lower WHRs and their children had significantly higher cognitive test scores. The data also suggest that teenage mothers with lower WHRs and their children were protected from the cognitive decrements normally associated with teen births. In the words of the researchers, “these findings support the idea that WHR reflects the availability of neurodevelopmental resources and thus offer a new explanation for men's preference for low WHR.”* Indeed, the findings offer additional evidence for the heritability of intelligence from the mother as argued in previous posts.
The finding that WHR correlates both with fertility and with higher IQ as conventionally measured goes part of the way towards explaining why intelligence may not have been falling in Western societies in the way in which luminaries like Russell and Fisher feared it would, but it does not go the whole way. On the contrary, only the diametric model of cognition can fully resolve it. According to this way of looking at cognition, the two modes of intelligence normally vary inversely: in other words, higher mechanistic IQ, lower mentalistic IQ, and vice versa (genius is the exception). As I have pointed out before, our great-grandparents and sub-Saharan Africans only seem a standard deviation below us in measured IQ because of their mentalistic intelligence probably being at least a standard deviation above ours.
Exactly the same insight explains the paradox of the high fecundity of seemingly lower IQ women. Their IQs may indeed be lower as measured by current tests, but the reason is that such tests are heavily biased towards mechanistic measures. If tests like the RME were included, the results would underscore the link between female fertility and mentalistic intelligence revealed by WHR. But the neurodevelopmental resources of the mother as evidenced by her WHR could be just as easily realized as mechanistic, rather than mentalistic, IQ in her children. This means that just because a woman’s measured IQ appears to be low, that of her children need not—especially if cultural factors like those underlying the Flynn effect impinge on them more than they do on her. In other words, the diametric model allows for trade-offs, not just between mentalistic and mechanistic intelligence in individuals, ethnic groups, and history, but also for differences between parents and their children, despite the very high heritability of intelligence.
(With thanks and acknowledgement to Jennifer Bremser.)