Bilingualism

In fact, our ability to learn lots of things changes as our brains are remodeled over time. Teenagers famously go through a ferocious remodeling, but my understanding is that remodeling is fairly constant throughout childhood and slows down after the teen years, but never really stops.

I only learned English as a child, however a lot of people are surprised when I tell them my second language is English. I myself am often amused by the fact that my written English is much better then my written French (to the point that I wasn’t sure I was going to get access to higher education).

Because of this I have always had a practice based opinion to language learning we learn the second language much more slowly and poorly because we dont need to, we already have another language to use. This does not explain deaf people however or anyone who hasn’t learned a language by their fourth year.

If we are wondering why super language skills shut off after a while we should ask ourselves if said super language learning skills bear a certain cost to the person in some way, shape, or form. I guess we could hunt down some poly linguistic person and find out if they are more limited in some other field. (assuming they didnt get that way by just taking lots and lots of language courses).

“understanding speech in noisy environments”

There is a silver lining to this: being able to tune out the outside environment when desired! Returning back to the States after a time abroad, my first complaint is usually about the inane conversations I suddenly understand.

When learning a first language, you don’t have any choice but to do so. That’s what’s presented. You have few others ways to communicate, and everyone big and important around you is requiring and encouraging you to learn, often to the point of withholding goods and services until you cough up the correct word.

I daresay there are few, if any, similar situations an adult, or an older child, might experience. Even if immersed in a language as an adult, you already have another one (ie your first) to think in.

Really, when you think about it, it takes a small child quite a long time to get fluent. But, again, they don’t have any choice.

I don’t think that, at a “wiring” level, the distinction between learning one language and learning two is meaningful. Within any given language, there are lots of different modes (what vocab you use, lots of details of syntax) depending on who is speaking to whom and so on – this is highly formalized in some cases like Japanese, less formal but equally critical in English. So, if a child learns to speak to Mum in one way and Dad in another, it’s only much later that they will learn these are different languages.

Coming to the evolution question, I’ve long thought that this is the one clear exception to the “blank slate” idea. To quote myself from a review of Pinker’s book of the same name

Pinker is a linguist and takes the acquisition of language, more precisely, the acquisition by children of their native language, as the paradigm example of learning. It’s hard to disagree with the conclusion that children’s brains are hardwired for the learning of language, based on the simple observation that two-year olds perform with ease a feat which most adults find exceptionally difficult.

But the exceptional nature of this feat should alert us to the dangers in using it as a paradigm. Langugage is the only characteristically human cognitive feat for which we are obviously hardwired (like most other complex animals, we are also hardwired for vision and other senses). For nearly everything else, the Blank Slate metaphor seems appropriate. Thanks to the environment in which I grew up, I can solve functional equations, swim the Australian crawl and perform many other tasks unknown to my putative hunter-gatherer ancestors. On the other hand, I can’t make or throw a spear or distinguish edible from deadly forms of bush tucker.

A striking instance of the absence of hard-wired functionality relates to kinship systems. Pinker lays much stress on the cultural universality of kinship. Yet even a relatively simple kinship system such as that prevailing in modern Western societies presents a formidable learning task for most children, and puzzles of the form ‘brothers and sisters I have none, but that man’s father is my father’s son’ baffle many adults. There is little to suggest that the capacity to learn kinship systems is any more hardwired than the capacity to learn trigonometry.

Still, this is not at all an area I’m an expert in, and I’m sure many CT readers will be able to point to more informed speculation than mine.

Though linguistics is one of the stops I’ve made on my dilettantish path through life, I’m not an expert in anything. But it’s fun to speculate, so here goes.

I realised a few days after the symposium that there was a huge question I wish I’d asked. Why are we so good at learning a first language, and so poor at learning a second language. What cognitive system would have such a feature(/bug), and what evolutionary advantage could there be to having such a system?

What if we ask this question the other way around, i.e. what evolutionary advantage would there be to maintaining, throughout life or at least a lot longer than we seem to, the brain system/structures that facilitate learning first language(s)?

If I try to imagine measuring the advantage of knowing the language of the people you’re born amongst as compared to the advantage of knowing a language that you might encounter later, there seems to be no comparison. I.e., the former seems immeasurably greater than the latter, all the more so if the vast majority of early humans rarely if ever encountered a second language.

Everyone needs a first language; once language exists, not to be able to speak the language of your people is a major handicap. But not everyone, in fact hardly anyone, proportionally, needs a second language.

That framing doesn’t fit with this suggestion (hypothesis? speculation?): Our ability to learn language isn’t switched off, it is inhibited by some other abilities we acquire.

If there’s dedicated brain machinery that facilitates early language learning, and it ties up a lot of brain resources and is rarely ever used again, the advantage gained by maintaining it might be much smaller than the advantage gained by repurposing the machinery for some other use. Dismantling the structures and reusing the parts doesn’t seem quite the same thing as “inhibiting” language learning, it seems more like an obvious response to the fact that for most people, later language learning wasn’t/isn’t that important.

Really, when you think about it, it takes a small child quite a long time to get fluent. But, again, they don’t have any choice.

It probably reveals a certain bias on my part, and perhaps the shortcomings (and/or out-of-date-ness) of my dabbling in linguistics, but I am under the impression that their lack of choice is like their lack of choice in whether to get taller, not like their lack of choice in whether to eat their peas.

Just my layman’s understanding- can’t find a good cite- but I believe 42 mo is just about the age of brain lateralization/ major neural pruning, which I think is also consistent with what others have said about filtering. This pruning means that a lot of developmental decisions are made, and alternative possibilities excluded.

Here’s another link, though it seems to refer to the same study:

http://www.washington.edu/research/pathbreakers/1992b.html

The point of relevance to the OP is that the efficiency mechanisms seem to start very early. (If I remember correctly, some of them apparently start to operate within days of birth.) Newborns can distinguish any and all language sounds (makes sense; human infants are not pre-programmed to learn only the language of their biological parents), but they lose that ability very early in order to concentrate on the important sounds: the ones used in their native language.

But what benefit could it be that inhibits language learning?

It could be the benefit of only distinguishing the phonemes, sequences of phonemes, and tones that make up your mother tongue(s) and making it easier to understand despite background noise. In order to learn a language, a baby has to first recognize the sounds that convey meaning and discard the ones that do not. Once these sounds are blocked out as meaningless, it can prove difficult to regain the ability to hear them again, hence the difficulty of learning a language at an older age.

Just out of curiosity, why is ‘language’ the unit? Just because adults distinguish them, doesn’t mean the infant’s brain does.

Also, I don’t think it can just be assumed that there is an evolutionary benefit to learning many languages — why? I know married couples who have children but don’t share a language.

JanieM and purple raise very interesting questions, seems to me.
Another question (I hope interesting), among all those one could raise around the definition of “bilingual”: how about distinguishing reading/writing a language from speaking it? I suspect OP privileges speaking over reading/writing. Is that justified? (You don’t have to be Derridean to wonder.) This question arose for me, immediately, as I investigated Burmese this evening (thanks to obvious, recent news). Apparently a language (another one) with notably distinct written and read vs. spoken forms. Caught my attention I guess because I can read a number of languages very well and can even write in some of them, but speak only English, really.

I think the evidence about how good young children are at learning two languages, either simultaneously or sequentially, raises some doubts for what Jason and Katherine say.

Once again, the confusion between first/seond language and early/late language arises. What I meant was that a young child has no choice but to learn what “language” (meaning words) is placed before her at a very young age (1-3, when language is truly forming, rather than later) whether that is one or two “languages” (meaning body of particular words). Later on in life, there is no such extreme compulsion, even in a state of immersion as an adult.

As purple notes, there’s no reason to assume a priori that this inability is an evolutionary feature rather than a bug with no direct or indirect selective advantage. There may or may not be biological reasons for this difficulty, but there is what strikes me as an obvious social reason: children generally get to spend all day and night in the close proximity of native speakers, don’t lose any esteem in their own or others’ eyes when they say things wrong, and automatically get treated as insiders rather than strangers. It’s very rare for an adult second-language learner to experience anything like that situation. There are also structural reasons, of course: a person who’s already learned that [p] and [pÊ°] count as the same sound (as in English, but not Hindi), and that eating crunchy food, eating soft food, and eating soup count as the same activity (as in English, but not Songhay) tends to have some difficulty unlearning these, whereas a first-language learner does not have preconceptions of this kind to correct.

Hypothesis: it’s all about incentive (this is to follow on from JanieM at 18). There is a very strong incentive to go from no language to one language; but once a person has one language there is only a relatively small incentive to add languages. This would explain why we see the difference in linguistic aptitude in immigrant families — Mum and Dad will learn the new language, but they’ll speak it with an accent, unidiomaticallty and so on, because although they are incentivized to undertake this onerous, complex task (they need to be able to communicate, go shopping, get by at work and so on) they’re not strongly incentivised . At the same time peer-group pressure will result in daughter and son speaking accentless, fluent and idiomatic New Language, because they have a strong incentive not to stand out amongst their schoolfriends and so on.

How is a 3 year old German speaker placed in an immersive French environment being compelled in a way that a 4 year old German speaker placed in an immersive French environment not being compelled?

Because by the time they get to 4, or 8, or whatever, they already know language and how to communicate more. The “speak this word or you won’t be understood and have any chance of getting what you want/need” element doesn’t exist in the same way. A four year old or an eight year old can communicate, by vociferous gesticulation, or climbing up and getting the damn thing themselves, for the sake of argument, in a way a 2 or 3 year old can’t.

I daresay a 4 or 8 year old will be quicker at learning a language than an adult, perhaps for the same reasons moved further along. I’ve found myself, for example, in a place where there was literally no shared spoken language. I made myself understood by use of rudimentary sign language, and, when it came to it, drawing what I needed. A 4 or 8 year old wouldn’t be able to do that in the same way, let alone a 2 or 3 year old.

As I say, I’m not an expert, even slightly, it’s just a practical consideration that occurred to me.

I have absolutely no expertise in linguistics, so the following is solely vague speculation on the basis of personal experience and can as such be ignored, but I’m struck by the tendency in many comments here to reify individual languages-as-defined-by-modern-linguists (presumably subsuming dialects, even those that bridge two different modern languages like Plattdeutsch?), and also to conflate different kinds of linguistic abilities into a single fluent/not-fluent dichotomy. Assuming that the capacity for language is innate, does that imply that every individual tribe on the veldt would have developed a completely different language, or might we expect (a) grades of differentiation, perhaps increasing with distance and (b) hence a tendency to be more comfortable with closely related languages and better able to cope with the linguistic consequences of inter-marriage with neighbouring rather than distant tribes?

I didn’t start learning German until I was in my twenties; I certainly don’t speak it like a native, but I’m fluent enough – and after a week or so of immersion, I do find myself thinking in German at least some of the time, and dreaming in it. I do think that, at least in part, this is because of the shared roots with English, so that in a sense they’re different forms of the same language. It might be interesting to compare acquisition of related languages (an Italian learning French) with acquisition of unrelated ones (a Chinese person learning English).

Maybe I’m not normal, but I have to take objection to the comments in the original post that people think in their first language. I find that I have to translate a lot of my thinking into English from non-verbal forms to explain what I’m thinking about. Sometimes I have to say that I can’t say what I’m thinking. This is especially the case when I’m thinking of something with spacial characteristics, such as the design of something, or a mathematical pattern.

There is a lot of limiting structure around the words we use and this structure may be laid down with the first language and then optimised for speed of processing around 42 months, which then makes language learning harder.

There is a theory that creoles are the creation of children brought up in a (typically colonial) environment in which there are at least two mutually incomprehensible languages bridged at most by pidgins. The supposition is that language is not only absorbed but even perhaps invented by children, whose brains are not completely wired when they start to master speaking.

It might also be reasonable to consider bilinguality in cases where neither language is English, which is itself a sort of Germanic-Romance creole with an unmanageably large vocabulary.

Oh, and what Frank Ashe and JanieM just said. I’m another programmer who finds it hard to describe some sorts of thinking which are quite abstract, certainly neither verbal nor visual. I’d go further and say that my stream of consciousness is by no means primarily linguistic; when I’m hunting an idea the words only come into play like the talons of a hawk, at the last moment when the prize comes within reach.

One thing that’s always puzzled me is how French people ever get fully competent in their literary tenses, since exposure to those will post-date the learning window and isn’t picked up from regular speech. Maybe most of them don’t and it too is a class marker (only graduates of the grandes ecoles can do it properly or something).

You don’t mean the _passé composé_ Phil. I’m talking _passé simple_, _passé antérieur_ and so on.

Z. The childhood development thing you mention is not uncommon in English, but I think that what typically happens is that the child first “gets” the irregular form and then, discovering the rule, starts to overgeneralise it. So “He’s flied to New York” might well be something a small child in the overgeneralization phase would say. But never an adult, I think. (Comment based on personal experience and anecdote rather than research, I hasten to add).

I don’t think the passé simple is any great mystery – it is a mark of education, just one that most French speakers attain (in first-world Francophone countries at least). Just because people’s language-learning abilities are strongest when very young, doesn’t mean that almost everything is learned by the age of 4. But it’s clear that people are at a disadvantage when it comes to grammatical structure learned later in life, even in their first language. For instance, most people have no trouble distinguishing between ‘I’m’ and ‘my’, because it’s obvious in speech, but for many people the difference between ‘you’re’ and ‘your’ is difficult and has to be deliberately learnt, even though it’s an entirely analogous distinction, because they are only exposed to it after learning to read. ‘He’s’ and ‘his’ might be interesting, because I’m sure there are some accents that pronounce these identically. Maybe children who have grown up only hearing such accents find this distinction as hard as ‘you’re’/’your’?

bad Jim @ 50
There is a theory that creoles are the creation of children brought up in a (typically colonial) environment in which there are at least two mutually incomprehensible languages bridged at most by pidgins. The supposition is that language is not only absorbed but even perhaps invented by children, whose brains are not completely wired when they start to master speaking.

A fascinating recent example of this is Nicaraguan Sign Language, which appears to have been created by deaf children at special schools for the deaf in Managua in the late 1970s and 1980s. In this case, there were no pre-existing languages involved; instead, children elaborated a mixture of simple home-signing gestures into first a crude pidgen and then later (starting with the younger children) into a fully fledged language. More detail here.

As the mother of two “natural” bilinguals (now adults), there was a period in my life when I was thinking alot about this subject. My conclusions are the same as Z’s above @55, i.e. the reason (very) young children can acquire full native speaker competence in more than one language (and older children/adults cannot) is connected with neural development/ apoptosis. Eric Lenneberg’s “critical age hypothesis” initially posed the problem, and I found Gerald Edelman’s Neural Darwinism helpful too.

It’s been a long time since I have read about this subject, but iirc there was a very interesting study of bilingual aphasics sometime in the nineties (Univ of Oregon?). Basically, the study demonstrated that true bilinguals – the subject of this post – may entirely lose one language through traumatic brain incident, and retain the other in full. This would suggest that such individuals possess two separate language systems extending from the limbic system all the way up to the highest neocortical layer and back again. SL users, otoh, would seem to have a single language acquisition and production system, with SL “grafted onto” the NL system. This (theory) helps us account for the persistent phenomenon of interference (from L1) in all but natural bilinguals, where there is no interference. Interestingly, such interference is most marked in the “lower level” functions (e.g. pronunciation), and least marked in the “upper level” functions (e.g. reading comprehension), which also would tend to support the neural darwinist theory for language, viz. that neurons hard-wired to recognize all human phonemes at birth die off in the absence of exposure. Stronger neural connections are accordingly formed (during “pruning”) for the native language, including extremely fine gradations in sounds (accounting for the fact that by and large, native speakers of a language are more sensitive to differences in pronunciation and intonation in the native language than are non-native speakers – a distinct “advantage” btw for native speakers). Brain plasticity appears to be highest for higher-level cognitive functions, and lowest for functions hard-wired at birth.

I recently watched two excellent lectures by Dr. Allan Schore (UCLA) discussing pretty much the same issue, though regarding a different area of brain development (emotional attachment). What is rather new in our understanding of brain development is the massive amount of genetically-programmed brain growth that continues throughout the first year or two of life, i.e. until initial neural pruning occurs. The continued genetically programmed growth of neurons/synaptic connections takes place in direct contact/interaction with the “facilitating” (or non-) environment, implying that nature (genetic development) actually responds to nurture (learned development) and program further post-birth brain connections accordingly. It’s pretty exciting stuff!

@Chris Betram: Confusingly, the phrase ” m’a tuer” (sic) is now a well-known phrase in France, after an infamous murder case in 1991 in which “Omar m’a tuer” was found written in blood near the victim’s body. For instance there is now a book out called “Sarko m’a tuer”.

@Peter Erwin: If you want an example that can be heard in speech, some people say “should of” as an emphatic form, as in “I should of, but I didn’t”, where standard English would use “should have”. (I say ’emphatic’ because “should have” is usually rendered as “shoulda” or “should’ve” in speech when it’s not being emphasised, and “should of” appears to be descended from “should’ve”.)

bad Jim (and Colin Reid):

I think Colin is suggesting situations where someone is speaking slowly and deliberately — situations where you might normally avoid contractions and the like, saying “I will not do that” instead of “I won’t do that” — and they are still saying “‘ve” (“of”) instead of “have”. I found an interesting if extreme written example in this web page of a British law firm: “Container door which injured client should never of been used”. (I.e., the headline writer evidently believes that the normal form is “‘ve”, even when it’s not directly attached to a modal auxiliary like “should”. This is combined with the moderately common orthographic error of writing “‘ve” as “of”.)

You’re right that describing people as saying “should of” instead of “should’ve” is a form of “eye dialect” almost always meant to stigmatize the speaker as uneducated, even though the pronunciations are the same. In almost all cases, “should of” is not descended from “should’ve”, as Colin suggested — they’re the same thing, just written differently.

Nonetheless… there’s some evidence for a phenomenon (which I’m guessing might really horrify Colin) where people have taken this reduced form of “have” used with modal auxiliaries like “should” and essentially turned it into a new word, even to the point using it in circumstances where you wouldn’t use “have”. There are examples in this Language Log post, and in the comments. As Mark Liberman says in one comment: “Barbara’s post made the point that this ‘of’, after a long career as a reduced form of ‘have’, has become a different morpheme in its own right, and therefore often appears in contexts where ‘have’ couldn’t”.

However, I don’t think this counts, quite, as “failing to learn the language” in the sense of language acquisition — instead, it’s an example of language change, where some dialects have transformed the reduced form of “have” into a new word. If we wait a few hundred years, this might become a normal dialectical variant (like “ain’t” or “y’all”) or even a standard form. Or it might disappear, of course.

If you want a concrete example of language learning that continues into adulthood, then I’d pick vocabulary: people do continue to learn new words in their own native language long after childhood, after all.

Hi, I’d like to offer a comment, and make a suggestion. First, some background:

I learned French as an adult… didn’t really start learning it until my thirties, when I moved to Montreal for the better part of a decade. I got a job in a French firm writing English for them (they needed their software translated from French to English and to have it documented; eventually they translated my manual back into French because it was much better than their existing documentation). One of the things they did was to send me to a night course to help formalize my French, so as to better understand the French docs as well as to aid in intra-company comm.

Since then, I’ve moved into programming and system administration, mostly on Solaris, though the programming I do is more about using tools like shell, awk, sed, grep, etc to be able to tie software together to do a particular job… though I do get pulled in pretty often during the design phase because I’ve got a good grip on complexity and general algorithmic approaches to solving particular problems.

My other job is as a musician. I’m a professional guitarist, playing anywhere from seventy to a hundred and fifty gigs a year. At that time I wasn’t nearly as active as I am now, but I had been doing it for some ten years or so at that point.

My brother had been living in French in Montreal for many years at that point (his wife and his daughters speak French as a first language). And yet, I soon exceeded him in French, esp. wrt to the actual sounds that came out of my mouth. He still kicks my ass in vocabulary, but I generally end up doing better in the area of gender agreement/verb tense (until it gets to some of the less common tenses), and when we speak I sound a LOT more Quebecois than he does.

A lot of the work I do as a player is improvisational in nature. No, that doesn’t mean I play jazz. What it does mean is that as a sideman I often find myself in the position of standing in front of an audience accompanying someone I’ve never met, let along jammed with, playing a song I’ve never heard. There’s a particular process that goes into this, generally starting with the condensed version of the song (what key is it in? what chords are you using in it? what’s the general feel of the song (i.e. – what’s the basic rhythmic pattern? train? shuffle? swing? funk? reggae? straight eighths? sixteenths?))and then having them count me in. The safe way is to start by hitting the one of the key you’re in, and then listening like hell to figure out what’s going on and getting there in a bar or so (or less if you can manage it).

Now, I’m not an academic, but I suspect it might be interesting to find out if musicians have an easier time of picking up second languages as an adult. If part of the issue is about cognitive filtering, it may be that music as a practice helps keep that from happening. In my general experience, I end up having a MUCH easier time than most of the people I know picking up new sounds that exist in other languages and applying them to successfully pronounce new words. My experience learning French (in reality my first real French lessons were in a taverne on Rue St-Jacques which had good live music and cheap booze, not to mention going to francophone parties to mostly hang out in silence, because, you know, I couldn’t speak the language ;) is that the hard part is picking up the music of a language, its rhythm… because once you get that it stops being a stream of undifferentiatable sounds and becomes a stream of words. At that point, it becomes about figuring out what the words mean, rather than figuring out what the words are, and that’s just straight up grunt work of pounding your way through the reference(s) for that language.

For the folks here that are actually studying this stuff, I’d suggest finding some musicians to study, and perhaps to study what’s actually happening in their brains while they’re talking, while they’re learning, and while they’re actually working on music (aside from learning natural language… just playing) to contrast and compare what’s going on inside the heads of both other adults as well as small children who are engaging in some of those activities.