Developing Intelligence

Several high-profile studies have shown that bilingual children outperform their monolingual peers in terms of several cognitive abilities – including tests of verbal and nonverbal problem-solving, selective attention, flexibility (e.g., task-switching) and others. These studies have captured the public imagination and probably guided many moms to expose their kids to a second language.

But a new article in Developmental Science suggests that these impressive results may be somewhat overblown: bilingual children may be likely to come from wealthier families than monolingual children (EDIT: this may not be true in general, but is plausibly true for the samples used to show a bilingual advantage, which predominantly come from Canada, where wealthier families may be more likely to speak both French and English). Since wealth has a strong and well-established relationship with cognitive function, these effects may be due not to a cognitive advantage but a financial one. None of the often-cited “bilingual advantage” papers seem to have measured socio-economic status independently.

To address this shortcoming, authors Morton & Harper asked 17 monolingual and 17 English-French bilingual 6-7 year-old children to complete a series of tasks to measure vocabulary (a picture naming test), selective attention (the Simon test, where subjects have to press one key for a red square and one key for a green square, but red squares may appear closer to the green key and vice versa, providing a measure of selective attention just to the color), and non-verbal intelligence (the Matrix Analogies test, which seems loosely similar to Raven’s Matrices). Every parent completed a questionairre on their socio-economic status (SES).

Their results:

Not only did the results fail to show an effect of bilingualism on the Simon task (in contrast to some previous work by others) but they showed a strong effect of SES: children from higher-income families showed better selective attention on the Simon task (i.e., they were less slowed when the stimuli appeared near the incongruent response button).

Furthermore, this SES difference remained even after controlling for differences in nonverbal IQ, suggesting that the high SES advantage for selective attention is not merely due to differences in IQ between SES groups.

Although the current sample size was small, the significant SES results suggest that the sample size was not too small to detect group differences in general. Morton & Harper suggest that there are other reasons to suspect that bilingualism does not benefit cognitive function in the way many of the previous papers have described (in particular, those researchers had argued that bilinguals must constantly suppress or “inhibit” the language which they are not using at any particular time, leading to a benefit in the degree of control they can exert on their behavior).

Nonetheless, the study does raise another issue: why is there a relationship between SES and cognitive function, beyond differences in IQ? Important work by Martha Farah and colleagues has investigated this question in detail, and demonstrated that the differences seem less related to visual/spatial abilities as to the elusive “executive functions” and a fronto-temporal network of brain regions. The exact causes of these differences remain unknown, as do possible interventions. In the presence of these larger-scale differences driven by SES, the influence of multilingualism is likely to be small – a conjecture confirmed by Morton & Harper’s work described above.


  1. #1 Chris Waigl
    October 31, 2007

    bilingual children are likely to come from wealthier families than monolingual children

    I wonder what the basis for this assertion is. A whole lot of multilingual children — indeed, the archetypal multilingual child in Western-type societies — come from migrant families, many of which are below average income.

    Indeed, multilingualism is wider-spread in poor countries and independent of educational and literacy status.

  2. #2 David Harmon
    October 31, 2007

    Hmm. I’d only heard of a specific advantage for early bilingualism, namely that such kids were supposed to (later) be able to learn new languages more easily. Did they examine that more-specific claim?

  3. #3 DR
    November 1, 2007

    But I thought executive functions were 99.9% genetic?!?

  4. #4 CHCH
    November 1, 2007

    DR – I had to remove that post due to complaints from the authors, who were concerned that it might affect JEP:G decision over whether to do a press release (with the concomitant media embargo until then). Anyway, you must have missed the paragraph where I describe why those results *don’t* mean that environment doesn’t matter. Individual differences around a mean are genetically determined, but group means can be elevated based on SES, education, etc.

    Chris: I should have been more specific; the location in which this study was run had an SES difference between bilinguals and monolinguals; in Toronto, where most of the studies about the bilingual advantage come from, this may also be the case. Regardless, you see why there could be a “bilingual advantage” for a research study run in a wealthy suburb of Toronto, where kids might speak both french and english.

    David: Check the link to google news for “bilingual advantage” (the first link in the post) and you’ll see what you’ve been missing.

  5. #5 CHCH
    November 1, 2007

    Chris – I’ve clarified the post to address your comment. Thanks!

  6. #6 GL
    November 1, 2007

    I see there is some concern over SES of the bilingual samples in Toronto. I am a graduate student from the Toronto lab that conducts the bilingual studies. If you followed the studies that show bilingual advantage, you would realize that the samples often consist of children speaking many different languages as their non-English languages (i.e., not only French). Also, in Toronto suburbs, the language population is far more diverse than just French and English. Toronto has a very diverse language profile compared to other cities. I usually work with young adults. In a sample of 25 bilingual subjects, the number of non-English languages are typically between 12 to 15.

    I agree with Chris Waigl that bilingual families, because of their immigration history, usually have lower income. Therefore, I am not sure if bilingual families = wealthier families.

    The monolingual and bilingual samples in the Toronto studies are obtained from the same neighbourhoods to control for potential SES difference. In fact, we are currently conducting a study that will systematically examine SES and bilingualism in the Canadian context.

  7. #7 CHCH
    November 1, 2007

    Hi GL, thanks for the input. I’m glad to hear that your lab is now controlling for SES differences. I am curious if that will prove to be an explanation for why some other labs have had difficulty replicating the “bilingual advantage” in other cities.

  8. #8 GL
    November 1, 2007

    We have always controlled for SES by choosing monolinguals and bilinguals from the same neighbourhood (middle class). The current study would be able to answer the question whether bilingualism affects other social class with the similar magnitude.

    About the failure to replicate the bilingual advantage in other cities, other than SES, it could also be due to the specific profiles of the bilingual subjects, i.e., levels of bilingualism and the specific kind of bilingual experience. This is what I am looking at in my dissertation. The results are interesting and hopefully I can wrap it up in a few months 🙂

  9. #9 Al Fin
    November 1, 2007

    Thanks Chris, for clarifying why you removed the “executive function” post. Since I had commented on your post at my blog, a few people asked where your article had gone.

    I agree that the whole concept of “executive function”, being somewhat independent of IQ, but still significantly heritable, is fascinating. No doubt there are several other variables that impact on life success which while independent of IQ are still significantly heritable.

    That is a problem for many psychologists who battle so hard to deny the significance of the entire concept of “intelligence” and “IQ”. All of those other pesky variables related to life success that may also be found to vary with population groups.

  10. #10 Brian Mingus
    November 13, 2007

    Why don’t bilinguals show less general intelligence than their peers?

    “Brain potentials reveal unconscious translation during foreign-language comprehension”

    If you have to reallocate neural soup to the translation process, you are presumably doing this to the exclusion of some other cognitive faculties. On the other hand, you could also theorize that this translation “hidden layer” is developing higher order representations of meaning, akin to an “interlingua.” Man, that’s gotta’ be useful for something 🙂

    I’m not sure that this article has much to say about the language representations of children who simultaneously learned the languages at a very young age. That’s where the interesting results are to be found, imo.

    More on interlingua

  11. #11 dr. priti singh
    November 14, 2007

    i wonder what the basis for thisfinding is. A whole lot of multilingual children in India are product of a a three language formulae adapted in schools(not in private schools)as there are as many as 16 recognised regional languages with different scripts.
    this put’s indians at advantage to learn Europeans and asian languages.So an early expose to a second language is the dominating than SES.

  12. #12 Lee35Diaz
    June 9, 2011

    Some time before, I really needed to buy a building for my corporation but I didn’t have enough cash and couldn’t purchase something. Thank heaven my sister adviced to get the mortgage loans from reliable bank. Thus, I acted so and used to be satisfied with my financial loan.

New comments have been disabled.