No Bilingual Advantage.

Back in 2011, I posted The Bilingual Advantage, featuring a NY Times interview about “how bilingualism sharpens the mind”; now it seems that was hooey. Emily S. Nichols, Conor J. Wild, Bobby Stojanoski, Michael E. Battista, and Adrian M. Owen have a new article in Psychological Science (Volume 31, Issue 5; titled “Bilingualism Affords No General Cognitive Advantages: A Population Study of Executive Function in 11,000 People”) whose abstract says “We assessed 11,041 participants on a broad battery of 12 executive tasks whose functional and neural properties have been well described. Bilinguals showed an advantage over monolinguals on only one test (whereas monolinguals performed better on four tests), and these effects all disappeared when the groups were matched to remove potentially confounding factors.” Here are some salient passages from the article:

In this study of 11,041 participants, no reliable differences in executive function were observed between monolinguals and people who reported speaking more than one language. First, when we created matched groups to eliminate confounds that may be masking an executive function advantage in bilinguals, and to ensure that our groups met the criteria for being either monolingual or bilingual, we found no significant group differences. Second, when utilizing the entire (large, though unbalanced) data set, we found that only one task, Digit Span, showed an advantage in performance in bilinguals. Although this result is statistically significant, it is important to put it in perspective: The regression coefficient was 0.05. In real terms, this means that, statistically, speaking a second language is associated with better memory for digits, but that difference is one twentieth of 1 standard deviation. […]

These results demonstrate that, across a broad battery of cognitive tasks of executive function, no systematic differences exist between monolinguals and bilinguals. […]

We conclude by emphasizing, however, that despite the fact that no meaningful relationship was found between bilingualism and executive function, the broader social, employment, and lifestyle benefits that are available to speakers of a second language are clearly numerous.

I’m glad they added that last paragraph to ward off despair!


  1. J.W. Brewer says

    “Study’s Findings Contradict Those of Earlier Study On Same Topic” tells you, at most, that one of the rival findings is hooey, but not which one. Of course, getting published with a boring result (X makes no difference) may require higher-quality work than getting published with an exciting-sounding result (X makes a difference), so that’s at least a heuristic giving the advantage to the more recent story …

  2. J.W. Brewer says

    As to the last paragraph, surely it to some extent depends on the language(s) involved? How do the “social, employment, and lifestyle benefits” available to a hypothetical someone fully bilingual in Estonian and Latvian but with no knowledge of English (or French or German or Russian etc.) compare to those of the median monolingual Anglophone?

  3. tells you, at most, that one of the rival findings is hooey, but not which one.

    Fair point; I guess I just assume that the NYT one is going to have a higher hooey quotient.

  4. David Eddyshaw says


    But the relevant point of comparison there is surely not a monoglot English speaker but a monoglot Estonian or monoglot Latvian speaker.

    And I mean, it’s not likely to damage your broader social, employment, and lifestyle prospects to be able to speak a second language, even if the second language is Warlpiri. At worst, you could always keep quiet about knowing Warlpiri when you go for that job interview with Goldman Sachs.

  5. Dmitry Pruss says

    I didn’t read the whole new paper due to the paywall, but I assume that it was a study of adults. The earlier effects were reported in children. If so, then it’s yet another confirmation of the fact that enriching learning environment impacts the kids and adolescents, but its effects wear off later in the adult life.

  6. J.W. Brewer says

    @David E.: Why is that the relevant point of comparison (other than the plausible assumption that the paper’s authors are addressing a presumed audience composed entirely of a) monolingual Anglophones; and b) people who know English-plus-something-else)?

    Whether Warlpiri-knowledge “damages” those other prospects in part depends on the opportunity cost – what else could you have been doing with the past/present/future hours necessary to obtain and/or maintain that competence (a number of hours that may of course be dramatically higher or lower depending on individual circumstances) and what potential benefits have you foregone by not doing that something-else?

  7. David Eddyshaw says

    I assumed that these were people who had become bilingual per vias naturales (so to speak), rather than opportunity-costing labour, but the summary doesn’t really say. If they lumped all categories of “bilingual” together in this study, I would have thought that that rather limited how much you can reasonably conclude from it.

    Mind you, I would not myself be very surprised (in retrospect*) if it turned out that bilinguals-but-not-by-means-of-formal-schooling had no particular cognitive advantages, as until recently that was probably the default human condition; still is, in a lot of places. It’s just humans doing what comes naturally.

    * I find few things surprising in retrospect.

  8. Stu Clayton says

    I find few things surprising in retrospect.

    That may be a side effect of self-confidence and explanatory zeal. You are usually more flexible.

    Unless you lead a very sheltered life, I suppose you will find many things surprising in circumspect. We’ll leave prospect out of this, as being too speculative.

  9. J.W. Brewer says

    If you grow up fully fluent in Warlpuri but move (especially at an early enough age) to an overwhelmingly Anglophone environment where Warlpuri speakers are quite scarce, maintaining your active Warlpuri fluency will often require the ongoing investment of time and mental energy that is thus not being used for something else. If you live amidst the Anglophones in a household where everyone else speaks Warlpuri at home, that may suffice, but not everyone in such circumstances does. And in any event we have gotten back to the notion that the advantages to the median Warlpuri-speaker of acquiring English are not symmetric to the advantages to the median Anglophone in acquiring Warlpuri.

    It is easy to find times and places in human history where bilingualism (or more) was ubiquitous – I’m not sure I would conclude from that it’s the default, since it’s also easy to find times and places where monolingualism seems predominant (obviously even predominantly monolingual societies might have “dealing with outsiders” specialists who were otherwise). It seems to me difficult to know which pattern was more prevalent when our species’ neural hardware reached its current form.

  10. Is it normal to say “no general cognitive advantages” when in reality it is no advantage in 12 specific tasks?

  11. Stu Clayton says

    It’s a generalization, in this case from performance on 12 specific tasks. You seem to have rather spectacularly overlooked the generalization involved in your own question. On the basis of one claim about “general cognitive advantages”, in one single study, you pose a general question as to whether such a claim is “normal”.

    Normality is important to vice squads. For the rest of us, plausible argument is what counts.

  12. I think the logic of the conclusion is even a bit stronger than that, because really it’s *refuting* a generalization.

    A “general” cognitive advantage would be one that shows up in a wide variety of kinds of activities. They tried 12 different tasks, presumably (one hopes) selected to represent a variety of different areas. If there was a “general” advantage, then, one would expect it to show up in most of those areas and so in most of those tasks. If an advantage fails to show up in any of the 12, on the other hand… well, either it wasn’t “general” after all, or one was awfully unlucky in picking 12 tasks that just happen to fall into the narrow exceptions to its general applicability.

    I expect an average bilingual person, whether it’s English-Warlpiri or any other pair (but especially when they’re widely divergent like those), would have a head start on monolinguals in certain kinds of wisdom: understanding that the world is big and complicated, that people in different cultures do things in different ways and all think their own ways make sense, and so on. But it’s hard to make a test for that, and in any case it’d be fair to say that’s not a “general cognitive” advantage.

  13. Greg, one of the tasks (I used sci-hub to access the article).

    A target word (either “RED” or “BLUE”) is displayed on the screen in either the color red or the color blue. The participant must select the probe word that correctly describes the color that the target word is drawn in. The problem’s color mappings can be congruent (if every word correctly describes the color it is displayed in), incongruent (if either the target word or both probe words are displayed in the opposite color), or doubly incongruent (if the target and probes are both written in the colors opposite to what they describe). Participants have 90 s to complete as many trials as possible. A correct response increases the total score by 1 point, and an incorrect response decreases the score by 1 point.

    I guess this one measures “short-term memory”. Looks like a generally useful skill (that can save you some time) and I suspect you are wrong, they are not speaking about “most” tests.

    (1) how many independent skills with a similar level of complexity are possible?

    (2) what about more complex problems?

    Consider math olympiads (e.g. IMO, the international math olympiad). They design tasks in such a way, that a talented but sloppy and generally bad at explaining her thoughts child performs better than an very accurate but less talented child. “Talented” is the ability to come up with an idea of a solution.

    Is this cognitive? I think yes. There can be complex skills that are not reducible to “performing task 1 fast”.

  14. I recently read (somewhere) that Berlin and Kay’s generalisations about colour vocabulary are completely biased towards English and are not universal at all. Their whole case is bunkum. (I wish I were more assiduous in recording where I read these things.)

    So “A target word (either “RED” or “BLUE”) is displayed on the screen in either the color red or the color blue” is totally useless as a linguistic test.

  15. drasvi: Sure, there are lots of complex problems in the world where it’s very useful for people to be able to do them. Among them is communicating in any given language. So bilinguals, by definition, have at least one complex cognitive task that they’re able to do and monolinguals (in any given language) aren’t.

    The question is whether they also have a *general* cognitive advantage, one that applies across a wide variety of tasks. That’s what “general” means. In order to answer that, especially to answer it in the negative, there’s no need to test every possible task in the world – testing a diverse assortment of tasks is enough.

    Bathrobe: I don’t think that task has any relationship to the Berlin and Kay “color universals” idea. In any case it’s not being used as a “linguistic” test; it’s a “test of inhibition”, a test of how well you can ignore one kind of input that’s trying to distract you while paying attention to another that has the information you want.

  16. And on math olympiads: as it happens, I was an olympiad competitor some years ago, and have spent a lot of time with other olympiad competitors.

    From scanning through that list of 12 tasks in the paper (thanks for the link!), I’d expect most of the people I know who were good at solving math olympiad problems would do very, very well at most or all of those tasks. So whatever “cognitive advantage” might be found among that group (be it a “general” one or not), a study like this one would spot it very clearly.

  17. If there was a “general” advantage, then, one would expect it to show up in most of those areas

    Greg, “general” is not an obscure term, it means something in English and will be immediately interpreted by the reader. If in their professional jargon it is a defined term with a distinct meaning, I would say it is an unfortunate name.

    As it happens, it can mean a variety of things. But yes, usually “in a wide variety of kinds of activities”. There will be disagreement over how wide this variety must be and over the set of possible activities this variety must be sampled from. Some can (and will) understand it as “advantages, unrelated to language use” (thinking of the “general” set of possible advantages rather than about advantages each of which is “general”).

    But “a wide variety” does not, I think, mean “most tasks”. Many, not most.

  18. P.S. sorry, this was a comment I typed yesterday (before falling asleep). I haven’t revised it in light of your answers, just woke up and pressed “send”:)

    Olympiad: I only know about olympiads in Russia and the IMO. What I meant here is that their approach is very different.

    I’d expect most of the people I know who were good at solving math olympiad problems would do very, very well

    Perhaps. But I think we need “all” (all people generally believed to be “smart”, not IMO medalists) not “most” to speak of “no advantages”.

    Of course, being able to think fast is good! Among two math geniuses the slow one will solve fewer problems in a day (and maybe will have other issues with complex problems):) And yes, most IMO winners think fast. But the possibility we need to rule out is that bilingualism increaces your performance in a class of complex tasks unrelated to language without affecting the skills discussed in this study.

  19. PlasticPaddy says

    General is not really ambiguous here, it means “wide-ranging, non-specific”. It is possible that Greg wrote “general”, because it is difficult to demonstrate an ability which transfers to multiple diverse contexts (e.g., no one appears to have created an AI demonstrating such an ability); I think it would be a reasonable POV to deny the possibility of “general” faculties, except for very basic ones allowing the identification and processing of relevant data to solve a particular (type of) problem or to acquire a new skill or competence.

  20. Finländare says

    I think the benefits to being bilingual are more hidden and cannot necessarily be measured quantitatively. English speakers have access to information that is of outstanding quality, which is not always the case for speakers of much smaller languages, such as Estonian, or even Finnish with its surprisingly large corpus of written texts. You can’t rank languages or their speakers according to the hypothetical quality of ideas expressed in them of course (Finnish monolinguals are no dumber than English monolinguals).

    But, as a bilingual person, you do have much more at your fingertips, which eventually shapes the way you see the world, the things you come to know and digest, and, if you were especially curious coming from a place where the smaller language dominates, you’d arguably come to know more than you could have known if you’d simply only been monolingual. For example, Finnish universities make their students read textbooks in English because some of the newer information is just not available in Finnish (or if it is, it’s scattered about and a pain to hunt down).

    This is written from the point of view of a person who is bilingual in Finnish and English. Your mileage may vary as a bilingual of Finnish/French, Finnish/Russian, Finnish/Warlpiri…

    It should go without saying that knowing more won’t make you an Übermensch, and it’s pointless to judge monolinguals and bilinguals based on that (in case someone understood my comment as an attempt to bolster the notion of alleged bilingual supremacy). Trying to find out who has the cognitive upper hand is even more ridiculous since, all things considered, as people we are all equally unequal and language is merely a cognitive façade.

  21. David Eddyshaw says

    a *general* cognitive advantage, one that applies across a wide variety of tasks

    Ah, dear old g.

    The idea seems to be that being “bilingual” (definition vague) does not increase your IQ, whatever it is that that measures: it seems to be the (reified) statistical correlation between the kind of tests that are chosen for psychometric testing on the grounds that they often correlate quite well with one another and with what modern artificially homogenised nation-states think makes an ideal productive citizen/soldier.

    Once again, I do not find this very surprising.

  22. J.W. Brewer says

    It would seem plausible-to-obvious that any “general cognitive” benefits from bilingualism would typically be less for someone who was fluent in two closely-related languages and typically greater for someone who was fluent in two radically disparate languages — differing widely from each other significantly in phonology, lexicon, and morphosyntax. Just because the latter involves habituating your brain (or “mind”) and related organs to handle a wider range of inputs/outputs and decode/encode in a wider range of structures, which to my mind seems congruent with any plausible mechanism whereby multilingualism could confer such “general cognitive” benefits. But by contrast the cultural/economic benefits of being fluent in A+B rather than merely monolingual in A, have little to do with that and instead depend in large part on the salience-in-the-world of B-speakers (and/or texts written in B) given the life circumstances of the particular A-speaker, which will be a matter of contingency/fortuity having nothing to do with the genetic affiliations or typological features etc. of the languages in question.

    Maybe a radical difference in the general culture (material and otherwise) of typical A-speakers v. typical B-speakers would also be advantageous in mind-stretching? But that will correlate only loosely to other sorts of differences – to go back to my example from above, Estonian-speakers and Latvian-speakers have quite a lot in common culturally (and socio-economically etc.) even given the dramatic differences between their languages.

  23. @PP, just for example: they rely on “three discrete factors relating to overall cognitive performance” from another study. “Each one of these factors is something that no single test can assess; each represents an independent aspect of cognitive function that is best described by performance on a combination of tests. They were labeled, for convenience, as encapsulating aspects of short-term memory, reasoning, and verbal abilities, respectively.

    Within this set of three “short-term memory” is specific. But if I ask you to classify cognitive “activities” or just “kinds” of cognition, you will come up with a different list, and possibly “short-term memory” will positively contribute in each. (cf. rotating your coordinate system to find that your former unit vector is now a mere combination of new unit vectors in the new basis. Now, I don’t think it is the right anology, but I think the object we are dealing with, namely all possible sorts of “activities” is worse:)).

    Greg too thought about a “general advantage” from bilingualism which would affect one’s performance in “most” (complex?) tasks because it’s “general”. And instead it is 12 tasks designed to measure 3 elementary “aspects” of cognition thought to be independent.

    What I mean is just that to speak of “general”, “many”, “more”, “wide” etc., you need to first have a list, majority of whose members we call “majority”.

    P.S. as for the reading “advantages, unrelated to language use”, it results not from ambiguity of “general”. I suspect some readers will have in mind a different contrast: the “general” set of cognitive advantages as opposed to specifically “advantages beyond those potentially afforded to cognition. For example, being able to communicate with a larger audience“.

    Not “general cognitive advantages” as opposed to “specific cognitive advantages”.

    Their own words make me think they too have this contrast “cognition vs. knowing languages” (rather than “general vs. specific cognition”) in mind even if it is not what they mean in the title.

  24. Stu Clayton says

    How Not To Do Things With Words.

    Chapter 1: The distinction between constative and performative statements.

    Chapter 2: An example: lecturing the speakers of a language foreign to you, as to the meanings of words in that language.

  25. David Eddyshaw’s comment about “what modern artificially homogenised nation-states think makes an ideal productive citizen/soldier” brought to mind that we never find out how many languages JS/07 M 378 could read or speak.

  26. @Stu, it is SAE.

  27. David Eddyshaw says


    I did have that exemplary citizen in mind. I am fairly sure that the answer is “one.”

    Normality is important to vice squads

    Auden, once again:

    I hate the modern trick, to tell the truth,
    Of straightening out the kinks in the young mind,
    Our passion for the tender plant of youth,
    Our hatred for all weeds of any kind.
    Slogans are bad: the best that I can find
    Is this: ‘Let each child have that’s in our care
    As much neurosis as the child can bear.’

    Goddess of bossy underlings, Normality!
    What murders are committed in thy name!
    Totalitarian is thy state Reality,
    Reeking of antiseptics and the shame
    Of faces that all look and feel the same.
    Thy Muse is one unknown to classic histories,
    The topping figure of the hockey mistress.

  28. Oh well, bilingualism now appears to confer no cognitive advantage.

    At least we can still console ourselves that bilingualism can stave off dementia for a few extra years. Or can it?

    Maybe being monolingual in a dominant and widespread language affording access to a wealth of intellectual and other resources is all the linguistic ability you need….

    Incidentally, the earlier article said:

    If you have two languages and you use them regularly, the way the brain’s networks work is that every time you speak, both languages pop up and the executive control system has to sort through everything and attend to what’s relevant in the moment. Therefore the bilinguals use that system more, and it’s that regular use that makes that system more efficient. …

    So perhaps it is about not being tethered to a single way of viewing things and being able to sort through relevant and important details better… although I also suspect that being bilingual might also be a distraction if you are aware of alternatives that are not available to monolinguals. E.g., being aware that there are two words for ‘blue’ in one of your languages and not the other (sorry, Berlin and Kay keep coming to mind) might make you pause if the word BLUE popped up in the wrong shade of blue in your other language. Monolinguals are perhaps less distracted by questions like this.

  29. I’m not remotely bilingual but I can drive a car with more or less equal facility on either side of the road (intentionally, I mean, in accord with local custom and habit). I would love to know what enhanced cognitive ability comes with this remarkable talent. (And I’m talking stick shift here).

  30. Lars Mathiesen (he/him/his) says

    For various reasons I’ve driven both types of car in both kinds of traffic. Stick shift. It’s always a surprise to me how quickly my behaviour changes over; basically I’m aware of left- or right hand drive before getting in the car–if I’m not, a few near-death experiences while crossing the road on foot will correct it–and then it’s just a question of checking where the hand brake lever is. (Strangely, getting into the driver’s seat from the left or right is not what triggers the switch. It’s releasing the hand brake).

    I’m also pretty good at reading upside down or mirrored text, but I remember doing that before the time of British outfitted cars.

  31. David Marjanović says

    The way I received the wisdom, it’s that being zweisprachig – having two native languages – will give you a general smartness advantage throughout life.* For learning languages later, general smartness was supposed to be an advantage rather than an effect (beyond the obvious things like letting you look at the world in a somewhat different way).

    The dementia thing… using your brain helps you not to lose it; how you use it is not supposed to matter.

    * Counterpoint: Donald Trump Jr. speaks Czech (there’s a video out there, he sounds like any other narcissistic native speaker), and probably so does Eric.

  32. @David Eddyshaw: Yes, as soon as I thought about the question, one seemed the most natural answer. However, in the context of this discussion, it immediately also occurred to me that Auden might have considered even a mention of how many languages JS/07 M 378 knew to be too relevant to the questions of, “Was he free? Was he happy?” to even be brought up in the poem.

  33. If you have two languages and you use them regularly, the way the brain’s networks work is that every time you speak, both languages pop up and the executive control system has to sort through everything and attend to what’s relevant in the moment. Therefore the bilinguals use that system more, and it’s that regular use that makes that system more efficient. … …

    (an earlier article quoted above)

    This is not very convincing:

    We found that if you gave 5- and 6-year-olds language problems to solve, monolingual and bilingual children knew, pretty much, the same amount of language.

    But on one question, there was a difference. We asked all the children if a certain illogical sentence was grammatically correct: “Apples grow on noses.” The monolingual children couldn’t answer. They’d say, “That’s silly” and they’d stall. But the bilingual children would say, in their own words, “It’s silly, but it’s grammatically correct.” The bilinguals, we found, manifested a cognitive system with the ability to attend to important information and ignore the less important. …

    Is it dukuduku correct to say that “the Earth is flat”?

    The question implies familiarity of children with the term “grammatically correct”.
    Possibly they indeed know the phrase – many (not sure about all) Russian children are occasionaly informed that this or that utterance is “incorrect”, but usually without “grammatically” – but I would love to know the circumstances where they got acquainted with it.

    They make a conclusion that 6-year-olds have or not a general (not specific to language, sorry Stu) ability “to attend to important information and ignore the less important” rather than the ability to understand what this adults needs from them. As an adult who has no slightest idea of what is “grammatically correct” (but usually understands what answer I am supposed to give) I have a doubt.

  34. David Eddyshaw says

    I don’t think it actually does imply that the children know the term “grammatically correct”: it does say “in their own words.”

    On the other hand, I’m having some difficulty thinking of how a typical five-year-old would express the concept, whether “in their own words” or not. Even quite sophisticated adults often have some difficulty explaining this idea. The absence of any specific examples of just what these budding Chomskyites actually did say strikes me as suspicious. Handwaving …

    Come to that, how on earth do you measure “the amount of language” someone knows in any way which is not highly language-dependent? Do English speakers lose marks for not being able to manage complex polysynthesis? Do Warlpiri speakers have to show that they have mastered the pluperfect? Or are we into languages-as-bags-of-words territory, and counting lexical items (itself a meaningless exercise given how much languages vary in the semantic range covered by a typical lexeme)?

    This may all just be the researcher dumbing down her explanations for the benefit of the expected audience of unHattery types, of course.

    I must say, though, that I am not astonished when research of this kind fails to be, like, replicable.

    The bilinguals, we found, manifested a cognitive system with the ability to attend to important information and ignore the less important. …

    So … the fact that “apples grow on noses” is “grammatically correct” is “more important” than the fact that it is nonsensical? Well, it’s a point of view, I guess … Or maybe there isn’t actually supposed to be any logical connection with the preceding sentence. Just a string of observations.

  35. zweisprachig
    Wow! I automatically assumed that their “bilinguials” were people with two native languages… Instead it is “people who reported speaking more than one language”.

  36. I don’t think it actually does imply that the children know the term “grammatically correct”: it does say “in their own words.”‘ – Actually I was thinking of how adults formulated the question (in their own words?).

    I must say, though, that I am not astonished when research of this kind fails to be, like, replicable.”

    An exceptionally good formulation.

  37. > Ah, dear old g.
    > The idea seems to be that being “bilingual” (definition vague) does not increase your IQ, whatever it is that that measures

    Yeah. I agree that that’s what they seem to mean (even though, the year being 2023 2019, they have enough discretion to avoid ever mentioning that term).

    > Wow! I automatically assumed that their “bilinguials” were people with two native languages… Instead it is “people who reported speaking more than one language”.

    They are pretty unclear about this, which isn’t good. But AFAICT their definition of “bilingual” is much stronger and more reasonable than that.

    Looking at their supplemental material (there’s a link near the end of the paper, just before the references), the relevant questions look to be just:

    5. What language(s) do you primarily speak at home?

    6. How many languages do you speak?
    Select one: 1-20

    And then in the paper text there’s this (p. 5):

    > participants who indicated that they were bilingual but selected only a single language were not included in data analysis on the assumption that this was an error or that they did not consider themselves fully bilingual.

    Given that set of questions, the only definition I see that they could reasonably have described that way would mean that someone was included as “bilingual” only if they listed at least two languages that they “primarily speak at home”.

    That’s still self-reported — but it is a lot stronger of a statement than just saying they *speak* a given language. If someone names more than one language that they “primarily” speak “at home”, then unless they’re outright lying I think it’s almost certainly accurate to call them bilingual.

    It sure is awkward that they’re unclear about this, though. Who’s in their sample, and who they call bilingual, seems pretty central to thinking about what real-world meaning their results might have (the results’ “external validity”).

  38. On the subject of “who’s in their sample, anyway”, their Table 1 (p. 5) shows it has some pretty distinctive features: 29% have a graduate degree, 77% have at least a “College diploma/Bachelor’s degree” (they reword it less clearly for the table, but that’s what it said in the questionnaire), and 100.0% finished high school. So, a much more educated crowd than the general population.

    I’m not sure a priori whether I’d expect any hypothetical effect here to be stronger in that population than elsewhere, or weaker. But it’s easy to spin stories of how there could be a real effect that goes away with education — one where monolinguals catch up over the course of 12-17 years of formal schooling. This study would miss such an effect completely.

  39. Counterpoint: Donald Trump Jr. speaks Czech

    Yes, but how much stupider would he be if he didn’t speak Czech? Also, excessive drug use presumably undermines any benefits of bilingualism

    In any case people who have worked with Don Jr and Eric tell me they aren’t as stupid as their social media personas suggest (Eric in particular is supposed to be a smart businessman), whatever other flaws they have. No one is claming, I hope, that bilingualism makes you a more moral person, and there is also no real correlation between intelligence and morality.

  40. that someone was included as “bilingual” only if they listed at least two languages that they “primarily speak at home”.

    But primarily…
    I think information about bilingualism came from elsewhere (not the questionnaire).

    On the other hand, the question “how many languages do you speak” does not imply language names as an answer. And they say they excluded people who didn’t write they speak English:/ So maybe you are right…

    Well, it won’t skew their sample, they just won’t include some bilinguals who thought “well, 55% of time it is Moroccan Arabic”.

  41. January First-of-May says

    On the other hand, I’m having some difficulty thinking of how a typical five-year-old would express the concept, whether “in their own words” or not. Even quite sophisticated adults often have some difficulty explaining this idea. The absence of any specific examples of just what these budding Chomskyites actually did say strikes me as suspicious. Handwaving …

    For that matter, how exactly was the question formulated in the first place? It’s indeed possible that they just used a term that most children weren’t familiar with.

    On that particular sentence, I suspect that a typical five-year-old would say something like “no they don’t”. Which by some analyses is a statement that the sentence in question is meaningful; it has enough meaning to be able to sensibly assign a truth value to it.
    Perhaps a typical bilingual five-year-old could see it as a “you keep using that word” situation, where they would think that the examiner had mistranslated something. I’m not sure what kind of reply it would invite, but it’s likely to be a very different kind of reply from that invited by the default monolingual expectation that there was no mistranslation involved.

    Depending on the exact languages involved, there could also be a difference in the level of exposure to similar phrases; Russian children’s poetry, in particular, frequently plays with grammatically correct nonsense. To a Russian six-year-old, “apples grow on noses” would sound like something he’d read in (and/or had read to him out of) a book of funny verse. I’m not aware of any similar tradition in English children’s literature.

  42. Stu Clayton says

    I’m not aware of any similar tradition [of playing with grammatically correct nonsense] in English children’s literature.

    I immediately think of Carroll, Lear, Seuss and Anon. There are many, many other writers in that “tradition”.

    # When a fox is in the bottle where the tweetle beetles battle with their paddles in a puddle on a noodle-eating poodle, THIS is what they call…a tweetle beetle noodle poodle bottled paddled muddled duddled fuddled wuddled fox in socks, sir!”# [The Fox in Socks]

    # Did you ever have the feeling there’s a WASKET in your BASKET? # [There’s a Wocket in my Pocket]

    # “The time has come,” the Walrus said, “to talk of many things: of shoes and ships and sealing-wax, of cabbages and kings, and why the sea is boiling hot and whether pigs have wings.” # [Alice]

    # If wishes were horses, beggars would ride. If turnips were watches, I’d wear one by my side. # [Anon]

  43. @Stu Clayton: I notice that those passages actually illustrate two different kinds of (seemingly) grammatical but nonsensical usage. In English literary nonsense, writers often use made-up words, like Seuss’s “wocket’ or Carroll’s “outgrabe.” I don’t know how common this is in other literary nonsense traditions, but in English it seems it may be related to the traditional use of nonsense syllables in English (from England) folk songs. (Tom Bombadil’s love of that kind of song is one of the things that makes him particularly English.)

    Separately: We have previously established that even among linguistically sophisticated Hatters, there can be big differences in how grammaticality is perceived. Some of us feel intuitively that there is a property of grammaticality that is separate from pragmatic acceptability; it is straightforward to conceive of sentences that are totally grammatical, but simultaneously absurd, unnatural, or incomprehensible. However, some other people do not see things in the same way. It may well be true that how many languages a kindergartner knows can affect which way they feel about grammaticality, but I think the description of this particular experiment underestimates the intersubjective complexity in how formal grammaticality is perceived.

  44. “grammatical, but … absurd”

    I thought* about a possible simplistic typology based on whether the mismatch occurs between different parts of the text or between the text and something outside of it (applicable to “grammatically correct” versus “factually correct” but not necessarily to “grammatically correct” vs. “idiomatic in my idiolect”).

    (1) the text and something outside (2) within the text (e.g
    and then
    (3) in the text as in “apples растут on noses” (this type is familiar to bilingual children if they code-switch)

    But it is not so straightforward. I am not sure whether “colourless green”, “colourless ideas”, “pregnant man” are principally different. Does the information that men can’t be pregnant, that ideas lack the attribute “colour”, that “green” and “colourless” are incompatible belongs to the outside “world” or…. to the lexicon?
    E.g. the lexical entry “idea” has no attribute “colour”. Then the mismatch happened within the text composed of elements of your lexicon. And then let’s apply the same logic to the lexical entry “man”:)

    Eventually and “apples grow on noses” can be added to this row.

    * I realise that I’m not be the first one to do so.

  45. In a world where imagination, metaphor, and other semantic extensions exist (ours), all these sentences can be semantically and pragmatically fine. Pinocchio’s nose can grow apples. A colorless idea is just the opposite of a brilliant one. Transgender men can be pregnant.

  46. (unrelated to Y’s comment)

    I think I did not make clear what I’m trying to do. The starting point was an observation that when people say “factually correct” or “truth value” they imply, among other things, that the clash happens in different places (in the text, outside the text, between the text and something outside the text). “Grammatically incorrect” is text-internal.

    Then to test (and question) this idea I’m trying to construct a gradual transition from grammatically to factually incorrect. “Colourless green”, “colourless ideas”, “pregnant man” are, of course, examples of grammatical correctness.

    Nevertheless I suspect that people see their relations with facts somewhat differently: they will say that “pregnant man” contradicts language-external facts (“language-external” is slightly different from “text-external” which I speak above), and they will be less sure about whether “colourless green” and “colourless ideas” are not a conflict between (culturally dependent?) lexical entries.

    Also we can easily imagine apples on noses but maybe not colourless green ideas.

    Now, I think to the left I can attach – because maybe it is just the conflict between natural numbers? But it is commonly named a “grammatical” error (on the assumption of course, that the speaker has a specific natural number in mind).
    And to there I attach “water”, “scissors” etc. and then wrong prepositions.

    To the right I attach “this man is pregnant” then “men can get pregnant”, then apples.

    Maybe I failed to make this transition gradual and one of steps involves a principal change.

    Also, I do feel that in different examples the clash happens in different “places” in my mind. But that’s feelings, introspection etc.

  47. A colorless idea is a trivial one, the opposite of a brilliant one.
    A green idea is an unripe one, one that is not well thought out.
    A sleeping idea is one that hangs around without producing anything useful.
    A furiously sleeping idea is a sleeping idea whose proponents proclaim it with much noise and emotion.

    An example of a colorless green idea sleeping furiously is the idea that syntax and semantics are entirely separable.

  48. JFoM: I’m not aware of any similar tradition [of playing with grammatically correct nonsense] in English children’s literature.

    @StuI immediately think of Carroll, Lear, Seuss and Anon. There are many, many other writers in that “tradition”.

    Yes, my eyebrow raised at JFoM’s claim. Nursery Rhymes would be the earliest ‘literature’ containing nonsense syllables/words/scenarios — some of them maybe minced from more macabre adult ditties. (Not that I’m claiming those are peculiar to English.)

    Gargantua & Pantagruel, Don Quixote, Tristram Shandy, Gulliver’s Travels contain copious quantities of language nonsense — in the ‘drunk on words’ sense. Gulliver in particular (Brobdingnag, Glubbdubdrib, Houyhnhnms) is treated as children’s literature — for reasons I can’t fathom: it’s straight-up political parody.

  49. David Marjanović says

    I’m getting the feeling we’re about to witness a large-scale experiment on whether second-language learning provides any such advantages.

    I also remain a master strategist. Or someone does anyway.

    — for reasons I can’t fathom: it’s straight-up political parody.

    If you don’t know what it’s a parody of, it’s just another fantastical story. Alice in Wonderland is straight-up political parody, too (…and in fact there’s a very well done Brexit version of it).

  50. Speaking of political parodies, one of my middle-school English textbooks printed Holmes’ “The Deacon’s Masterpiece” (AKA “The Wonderful One-Hoss Shay,” about a carriage made so perfectly that it lasted 100 years and then fell apart all at once in the middle of the street) purely as a comic poem, with no indication that it was a satire on Jonathan Edwards’ theology. (This was long ago when schoolbooks must contain the Schoolroom Poets, a tradition whose meaning had perhaps gotten lost.)

  51. Same here. I loved it, and had no idea it was a satire until I was well into adulthood.

  52. John Cowan says

    Alice in Wonderland is straight-up political parody

    Well, parts of it are: certainly the caucus-race is. But who is the political counterpart of the man in the railway train in Through the Looking-Glass? Frankly, this claim reminds me of the notion that The Wizard of Oz is all about Free Silver.

    But speaking of Brexit, yesterday I discovered “Breexit: how the Remainers left the Leavers high and dry”, about the departure of the Breelanders from the Eriadorian Union:

    The Breexiteers of Breeland were full of resentment at what they called the ‘unelected Rangerocracy’ imposing laws on them from the Eriadorian Folkmoot in Fornost. ‘That Strider was just an ordinary bloke drinking beer in the Prancing Pony,’ they grumbled illogically, ‘and now he’s made himself king of Gondor and Arnor and drinks wine out of a great gold goblet in Minas Tirith.’ Their other complaint was at the numbers of Southerners coming north and settling in the empty lands around Bree. ‘They ain’t like us. They don’t look right and they don’t talk proper, and they’re taking away our jobs.’ Even Hob and Nob complained, despite the fact that they needed the help of the Southerner barman and groom to cope with greatly increased business at the Prancing Pony.

    I particularly like the names of the Folkmoot Foreman, Terebinthia Mayweed, and the “hard-Breexit Big Person, whose name was Giacomo Reed-Mugwort”. Terebinthia is one of the islands in the Great Eastern Ocean of Narnia, not to be confused with Terabithia, an imaginary forest kingdom in Bridge to Terabithia, which (it turns out) is #8 on the list of most often banned children’s books, probably because the heroic characters don’t believe in life after death. The author has said that the name is probably an unconscious reminiscence of Lewis’s.

  53. David Marjanović says

    Alice in Wonderland is straight-up political parody

    Uh, sorry: straight-up satire of society more generally, politics included.

  54. Alice in Wonderland (especially the first book) is a mashup of all different kinds of jokes. Some of it is political parody, but other sections focus on other sources of humor. The tea party is a nonsensical comedy of manners; the episode with the griffin and the mock turtle is full of puns; and Alice’s trial is (quite obviously, to the extent that Alice herself notices it) a parody of the legal system, particularly with the antiquated traditions Monty Python would later skewer.

  55. AiW also contains many parodies of pious verse for children, whose stuffiness Carroll publicly recognized. More’s the pity that he wrote Sylvie and Bruno.

    The Walrus and the Carpenter surely was inspired by The Rime of the Ancient Mariner? “The Sun came up upon the left / Out of the sea came he! / And he shone bright, and on the right / Went down into the sea” asks to be parodied, and “The sun was shining on the sea / Shining with all his might / He did his very best to make / The billows smooth and bright / And this was odd, because it was / The middle of the night” delivers.

  56. I loved it, and had no idea it was a satire until I was well into adulthood
    Me too, mostly. Even as a pre-teen I recognized that the story about enmity between countries caused by a dispute about how to best eat eggs made fun of religious / ideological differences, but mostly I read it as a captivating adventure.
    BTW, the edition I had only contained the Liliput and Brobdingnag stories, which I guess is a frequent selection in children’s editions.

  57. David Eddyshaw says

    A number of the nonsense verses in Lewis Carroll’s works are parodies of Wordsworth, who deserves to be parodied if ever a major poet did. (My view of Wordsworth aligns with that of Byron and Auden, so I may not be a fair witness. Still …)

  58. PlasticPaddy says

    Scorn not the Sonnet; Critic, you have frowned,
    Mindless of its just honours; with this key
    Shakespeare unlocked his heart; the melody
    Of this small lute gave ease to Petrarch’s wound;
    A thousand times this pipe did Tasso sound;
    With it Camöens soothed an exile’s grief;
    The Sonnet glittered a gay myrtle leaf
    Amid the cypress with which Dante crowned
    His visionary brow: a glow-worm lamp,
    It cheered mild Spenser, called from Faery-land
    To struggle through dark ways; and, when a damp
    Fell round the path of Milton, in his hand
    The Thing became a trumpet; whence he blew
    Soul-animating strains–alas, too few!

  59. Wordsworth (quite properly) was prominently featured in Lee and Lewis’s The Stuffed Owl: An Anthology of Bad Verse; I have never forgotten the opening line of “To the Spade of a Friend”: “Spade! With which WILKINSON hath tilled his lands.”

  60. A large research library will contain a largeish shelf of Wordsworth’s complete works, of which only the first volume is of any interest to anyone not writing a dissertation on Wordsworth.

    Hans, I don’t know how Swift got into the discussion, but you’re right about that too. What always gets me about Gulliver is the number of scholars who fail to see, or I think used to (in the heyday of academic rationalism), that the Houyhnhnms are as satirically portrayed as all the rest of the nations in the books, rather than being some idealized portrait of how people ought to be.

  61. January First-of-May says

    Yes, my eyebrow raised at JFoM’s claim. Nursery Rhymes would be the earliest ‘literature’ containing nonsense syllables/words/scenarios — some of them maybe minced from more macabre adult ditties. (Not that I’m claiming those are peculiar to English.)

    I thought of Carroll, Joyce, and Seuss (and to a lesser extent Swift) as representing a somewhat different variant: nonsensical words (as combinations of syllables) rather than nonsensical sentences (as combinations of words); the grammatical-example-sentence equivalent being “The gostak distims the doshes”.

    But I indeed hadn’t thought of Mother Goose, and Edward Lear’s works are full of both versions but especially the limericks are absolutely on the combination-of-words side. It shows up in Carroll sometimes too.
    (TIL that the Mad Gardener’s Song, an excellent example of nonsensical phrases, is from Sylvie and Bruno. I’ve always seen it presented as an independent work.)

    (Incidentally, I do kind of like Carroll’s statement on truth values:

    “The first two are, as it happens, not strictly true in our planet. But there is nothing to hinder them from being true in some other planet, say Mars or Jupiter—in which case the third would also be true in that planet.”)

  62. It doesn’t seem that this very recent publication has been discussed already?

    “Linking early-life bilingualism and cognitive advantage in older adulthood”

    Individuals reporting bilingualism (i.e., daily use of L2) in the early life stage outperformed monolinguals on learning & memory, working-memory, executive functions and language. Bilingualism in middle life stage showed a significant advantage on learning & memory, while no effect of bilingualism in old life stage was identified.

  63. Dmitry Pruss says

    PS: Reading through the methods / results sections leaves me with lots of doubts about both papers.

    The “encouraging one” (in older German residents) had big differences between bilingual vs. monolingual cohorts. Bilinguals typically had college education and monolingual did not. But they tried limiting the groups to just matching the educational level and some differences remained significant – barely. OTOH stronger education may partly be a consequence rather than a cause of abilities.
    Not controlling for health conditions / employment and types of jobs is definitely a problem when an older population is studied.
    Significance is FDR rather than Bonferroni adjusted and wouldn’t have survived Bonferroni. But the metrics of the abilities all shift in the same direction, both where they are significant and where they are not (like the older-age bilingualism group was too small to reach significance anywhere, but the trends looked the same). Since the metrics are expected to be somewhat correlated, even where these multiple trends are not reaching individual significance, they may still represent a significant shift when properly added together?

    The “discouraging” study drew its matched cohorts from the large Anglo countries (it was conducted as an online quiz for all takers from anywhere in the world rather than a lab study, and of course 3/4 of the participants didn’t finish the quiz). The assignments seem to be silly (I know that it was already discussed here). Especially silly because in the matched groups, 80% had college education (of the 11,000 ppl who completed the tests, only about 700 ended up matched). But the authors explain that the same tests have been used before, for example in Dx of psychiatric conditions (I sort of suspect that the past reports of associations between these cognitive tests and the disease may have been bogus / un-replicable, but I am too lazy to check it)
    They are able to detect an age effect in declining scores. But as far as I can tell, no correlation with the educational attainment or poverty (they don’t show it but they only mention that the age effect was visible in some tests). The matched sample may have been too small, especially in their numbers of undereducated or poor subjects, to detect it. Honestly I am a little surprised that they even saw the age effect, because all of their metrics are noisy with IQR ~~1 (Fig. 2), while respective age effects are typically less than 1 (Fig. 3)

Speak Your Mind