Monday 22 July 2013

Competition in the womb


I am in favour of competition. Not all school essays are of equal standard. One of my great pleasures in early life was to be asked to read out my essay to the class. Normally the accolade went to Keith Yorston or Michael Turner, but I relished my occasional turn, accepted the competition, and felt spurred on to greater literary feats. We were happy and well-fed children living in a little country where there was no television. Even a short essay constituted entertainment.

Competition in the womb is another matter. In the essay reading competition no-one was hurt. Egos, perhaps, but egos improve once they have been knocked back a bit. The untrammeled ego is a social menace.

When two twins must battle for access to scarce nutrients in their mother’s womb then competition takes on a harsh significance. For most of the 20th Century the results were very clear: twins in general were 5 IQ points behind singletons or, to put it another way, they were a third of a standard deviation behind in terms of intelligence. On average, singletons would expect to be at the 5oth percentile and twins at the 37th percentile. A significant difference in the pecking order.

This savage difference was maintained despite controlling for social class, but not dependably maintained when controlling for lower birth-weight and shorter gestation. So, during the last century twins got short commons, poor rations, thin gruel and a bad deal. Evolution is unfair, and it doesn’t even know it is unfair. Stuff happens and organisms evolve.

In more recent times the Deary Edinburgh gang have taken another look at this issue, as well they might. They have collared the largest ever sample of UK schoolchildren, and compared their intelligence at age 11 against their scholastic achievements at age 16. Correlation is not always causation, but it would be bloody unlikely if every correlation was a coincidence. As a clue, when one measure precedes another, it ups its likelihood of being the cause of the other. Anyway, those scholastic results deserve another later post, if only to discomfort those educationalists who ignore human intelligence.

In this particular paper (see reference below) the cognitive ability scores of 178,599 schoolchildren in England (average age about 11 years) were looked at in terms of twin or singleton status. There were no significant discrepancies between twins or singletons. Not only was there no difference in intelligence, which is the cause of most variation in scholastic achievement, but there were no differences in later scholastic attainments.

This is an interesting paper to look back upon when one expectantly savours the prospect of the special issue of Intelligence on the Flynn effect, due out in December. It certainly boosts the argument that the apparent secular rise in intelligence must be linked to better nutrition and health care and general living circumstances. The fact that we can now raise twins without them being behind singletons in terms of intelligence strongly suggests that we are boosting outcomes by making pregnancy much less costly for both mother and children, to the benefit of their intellects.

There is no longer a cognitive cost to being a twin. We have assured twins a well-nourished start in life.  They may never get to read out their essays to their class mates. But they can stand in line with other children, and do their very best to get chosen for the life tasks ahead.



Is there still a cognitive cost of being a twin in the UK? Calvin, Fernandes, Smith, Visscher and Deary Intelligence 37 (2009) 243–248


  1. Hold hard, there's an inconsistency there. I take the original result "This savage difference was maintained despite controlling for social class", add it to the correlation of social class with income available for food for the pregnant woman (and probably with care in nutrition too) and conclude that the problem can't be nutrition. Then you say it is. Which?

  2. The phrase I used was "but not dependably maintained when controlling for lower birth-weight and shorter gestation". The original phrase in the paper was "This observation remained after
    controlling for socioeconomic factors,while adjusting for lower birth weight and shorter gestation in twins produced variable results". Of course "adjusting for" always begs the question whether you are correcting or distorting. It remains my assumption, and not the author's, that this might be related to the causes of the Flynn effect. Deary merely says that the effect is of unknown origin.