Face recognition not correlated with IQ

Heritability of the Specific Cognitive Ability of Face Perception:

What makes one person socially insightful but mathematically challenged, and another musically gifted yet devoid of a sense of direction? Individual differences in general cognitive ability are thought to be mediated by "generalist genes" that affect many cognitive abilities similarly without specific genetic influences on particular cognitive abilities. In contrast, we present here evidence for cognitive "specialist genes": monozygotic twins are more similar than dizygotic twins in the specific cognitive ability of face perception. Each of three measures of face-specific processing was heritable, i.e., more correlated in monozygotic than dizygotic twins: face-specific recognition ability, the face-inversion effect, and the composite-face effect. Crucially, this effect is due to the heritability of face processing in particular, not to a more general aspect of cognition such as IQ or global attention. Thus, individual differences in at least one specific mental talent are independently heritable. This finding raises the question of what other specific cognitive abilities are independently heritable and may elucidate the mechanisms by which heritable disorders like dyslexia and autism can have highly uneven cognitive profiles in which some mental processes can be selectively impaired while others remain unaffected or even selectively enhanced.

Here's some more from ScienceDaily:

For the study, Liu and his colleagues recruited 102 pairs of identical twins and 71 pairs of fraternal twins aged 7 to 19 from Beijing schools. Because identical twins have 100 percent of their genes in common while fraternal twins have just 50 percent, traits that are strongly hereditary are more similar between identical twins than between fraternal twins. (Identical twins still show variability because of the influence of environmental factors.)

Participants were shown black-and-white images of 20 different faces on a computer screen for one second per image. They were then shown 10 of the original faces mixed with 20 new faces and asked which ones they had seen before. The scores were more closely matched between identical twins than fraternal twins, and Liu attributed 39 percent of the variance between individuals to genetic effects. Further tests confirmed that these differences were specific to face recognition, and did not reflect differences in sharpness of vision, general object recognition abilities, memory or other cognitive processes.

In an independent sample of 321 students, the researchers found that face recognition ability was not correlated with IQ, indicating that the genes that affect face recognition ability are distinct from those that affect IQ. Liu and Kanwisher are now investigating whether other cognitive abilities, such as language processing, understanding numbers, or navigation, are also heritable and independent from general intelligence and other cognitive abilities.

I've blogged about prosopagnosia before, which is a much more extreme manifestation of lack of face recognition than they're talking about here. These "face blind" individuals are shockingly common within the general population, 1 out of 50 humans. I haven't heard it reported that these individuals are particularly unintelligent, so it stands to reason that there'd be no relationship between intelligence and the ability to recognize faces. I happen to know some individuals with this issue, and they've developed "tricks" to compensate, so I know firsthand that they're not necessarily a dull lot.

So how to reconcile the heritability of IQ with domain specific competencies that exhibit modularity? I think the model in terms of genetic architecture is pretty simple. General intelligence is a quantitative trait characterized by a normal distribution; the bell curve. This is because variation on the trait is due to many genes of small effect; i.e., each genetic variant which controls variation of IQ has only a very small effect on the trait. By contrast, imagine that face recognition is controlled by a few genes of large effect, for example, five genes controlling most of the variance of the trait. Even if these five genes which control face recognition ability also affect intelligence, their contribution to the variance on the latter is going to be trivial because any specific gene has only a small effect, at least in the normal range. I assume that this sort of dynamic could characterize many domain specific cognitive traits.

Citation: Heritability of the Specific Cognitive Ability of Face Perception, doi:10.1016/j.cub.2009.11.067

More like this

Human face recognition ability is specific and highly heritable: Compared with notable successes in the genetics of basic sensory transduction, progress on the genetics of higher level perception and cognition has been limited. We propose that investigating specific cognitive abilities with well-…
THE perception and recognition of faces is crucial for the social situations we encounter every day. From the moment we are born, we prefer looking at faces than at inanimate objects, because the brain is geared to perceive them, and has specialized mechanisms for doing so. Such is the importance…
I just read an interesting new paper, Genetic Variation in Political Participation: The decision to vote has puzzled scholars for decades...The results show that a significant proportion of the variation in voting turnout can be accounted for by genes. We also replicate these results with data…
As promised I have a response to this article in the New York Times (I had to spend a couple days marshalling my evidence). I thought I would summarize some evidence about what we know from behavioral genetics so you could understand why I think this article was so wrong. I have tried to classify…

Not surprising. I've trained artificial neural networks to recognize faces better than most humans. Facial recognition is just an algorithm, related to cognition in the same way as being able to multiply large integers in one's head.

By Chris Irwin Davis (not verified) on 19 Jan 2010 #permalink

A person with poor cognitive ability of face recognition is highly likely to find the Visual-Spatial Processing Part of an IQ-Test a hard or even insurmountable task. The lacking points in this field will lower his IQ-overall-score, generally taken as the measure of intelligence.

Apart from that I can confirm from personal experience that far below average face recognition capability can go along with pretty good results in the other parts of an IQ-test.

The study confirms my point of view, that Visual-Spatial cognitive intelligence should be separated from general intelligence.


A person with poor cognitive ability of face recognition is highly likely to find the Visual-Spatial Processing Part of an IQ-Test a hard or even insurmountable task.

Was that in the study or is it a prediction on your part? I don't have access to the journal right now, but they mention nothing about it in the abstract. Since it seems contrary to their stated conclusion, it seems like it would be pertinent to mention there if so.

Our brains have a lot of resources dedicated to facial recognition. It's not at all obvious (well, not to me) that an allele causing a defect in that system need do so at the level of visual-spatial processing.

I have pretty high IQ on the tests, but I suck at facial recognition.

In anything but good light, or even just out of context, I can completely fail to recognize people I don't know really well. I've been sworn at (!) more than once for failing to speak to people I know but simply failed to recognize. It really seems to upset people.