In the preface to his book Knowledge of Language (1986), Noam Chomsky writes,
“For many years I have been intrigued by two problems concerning human knowledge. The first is the problem of explaining how we can know so much given that we have such limited evidence. The second is the problem of explaining how we can know so little, given that we have so much evidence. The first problem we might call “Plato’s problem,” the second “Orwell’s problem,” an analogue in the domain of social and political life of what might be called “Freud’s problem”…
…Plato’s problem is deep and intellectually exciting; Orwell’s problem, in contrast, seems to me much less so. But unless we can come to understand Orwell’s problem and to recognize its significance in our own social and cultural life, and to overcome it, the chances are slim that the human species will survive long enough to discover the answer to Plato’s problem or others that challenge the intellect and the imagination” (Emphasis mine).
To contextualize this statement just a bit, it’s asking a profound question about human language and how we acquire it as a first language (with implications for second language learning).
It’s an illustration of what Chomsky has named ‘the poverty of the stimulus.’ That is, how do children figure out the grammatical or structural constraints of a particular language, so as to acquire a linguistic system (let’s say by the age of four) where they are creating new sentences they’ve never heard before, but which are ‘grammatically’ correct (in the linguist’s sense)?
Another way to think about it is How do children acquire a grammar so completely, so fast? Why do adults take longer, or if at all? And where do we draw the line between the two processes of child ‘acuqisition’ and adult ‘learning’? Is there a ‘fundamental difference between adults and children (Bley-Vroman)? Does our attention work differently at different stages of development (Schmidt)? Or is less more (Newport)? That is, because children have less cognitive resources available to them, they tend to start with the simple bits of language whereas adults dig right into all the complexity, and thus this helps children learn the combinatorial patterns of language much faster.
So how does universal grammar (UG) help to explain it all?
Personally, at least right now, I feel it’s fairly pointless to argue against such a thing as the existence of UG as (genertive) linguists argue for it, simply because it has become such a powerful explanation for how language is acquired (that we have an innate ability to learn language incredibly rapidly and completely as a child), so powerful in fact that it explains so much, it ends up not really explaining much at all. Or at least raises more questions than it answers.
Let me put it this way. It is very convincing to me that there is something special about the child’s brain that allows them to acquire a specific language so uniformly across all languages on earth, and it seems that there is something different about adult brains because there is such variation in their ability to learn another language. (I’m using fairly idealistic notions of ‘child’ and ‘adult’ here for the sake of argument, I know).
This raises questions like, “Do adults lose access to UG” after some period of time after childhood? (the critical period hypothesis). This would explain in part why it would be so difficult to learn another language when your UG ‘goes away.’
Or do I think that language acquisition is just a general cognitive learning process, the same kind of learning we do for everything else in our human brains, that allows children to acquire a language? Not so much.
Or does it comes mostly from social interaction (outside)? No, or at least not as a full explanation.
So do I think that children are born with an innate set of ‘constraints’ in their brain that allow them to figure out how the rules and structure of a new language work? Not really either.
Then is it a miracle?
No:) that wouldn’t be any more helpful either I think. My own view is that trying to tease apart what must be innate or ‘universal’ among all human beings and what must be the myriad variations in our languages, in the data ‘out-there’, or what has been called the classic ‘nature-nurture’ debate, is missing the point a bit.
Aliens and UG?
The perspectives I find most convincing are ones that attempt to take a broader and more integrated approach to understanding what language is, including what we understand from ‘core’ linguistics, neurolinguistics, anthropology, and evolution. Thomas Givón and Terry Deacon‘s approach (PDF) are two examples of this more integrative approach, where UG is neither about nature nor nurture but rather in the ‘semiotic constraints’ of language itself.
Deacon writes,
Semiotic constraints delimit the outside limits of the space of possibilities in which languages have evolved within our species, because they are the outside limits of the evolution of any symbolic form of communication. So perhaps the most astonishing implication of this hypothesis, is that we should expect that many of the core universals expressed in human languages will of necessity be embodied in any symbolic communication system, even one used by an alien race on some distant planet!
Now that is a pretty wild claim, but his argument is fairly convincing I think. Am I an expert in generative linguistics? Nope. Do linguists I’ve discussed this with think I’m a little nuts? Yup, sometimes. Linguists, both students and professors, I’ve had the pleasure of talking to are surprisingly more than just a little passionate about the positions that they take on the issue of UG. And I’ve actually witnessed some fairly intense, but brief, shouting matches between grown-ups on the issue.
So what about Orwell and Plato’s problems?
Plato’s problem is derived from the conversation between Meno and Socrates about ‘inborn knowledge.’
Meno says,
“And how will you inquire into a thing when you are wholly ignorant of what it is? Even if you happen to bump right into it, how will you know it is the thing you didn’t know?”
Orwell’s problem is derived from Orwell’s argument against totalitarian regimes that try to shove propaganda down our throats, ‘and often plainly at variance with obvious facts about the world around us.’
The two problems are crucially different from each other, but it does show how Chomsky dips into seemingly unrelated fields to explain different aspects of both. Or to make fairly convincing arguments of why we should be weary of ignoring certain questions, in trying to answer others.
Okay…
…this post is starting to go off the rails a bit, but what can you take away from this as a language learner?
I would suggest to you to be highly skeptical of methods and approaches that equate second language learning to first language learning, the ‘learn just like a child’ approaches. If we’ve learned anything from second language acquisition research, it’s that first and second language learning, although still extremely complex and mysterious phenomena in many respects, are quite different from each other.
However, I would also suggest that you be highly skeptical of people who take this information to persuade you to think that adults cannot master another language fluently later on in life. This is patently false, people do it all the time, and in a variety of circumstances and with various methods.
I’ve left out quite a few other theories and approaches to understanding what language is,but I wasn’t meaning to be comprehensive here. Rather I just wanted to express the point that complex questions require either complex answers, or more difficult questions with more nuanced answers. The last thing SLA research needs to do is follow in the foot steps of the self-help, pop-psychology literature spawned from ‘surprising findings’ in the sciences.
As Ta-Nahesi Coates, writer for the Atlantic once put it,
“…a big part of all of us likes getting answers. But we now live in a world where counter-intuitive bullshitting is valorized, where the pose of argument is more important than the actual pursuit of truth, where clever answers take precedence over profound questions. We have no patience for mystery. We want the deciphering of gods. We want oracles. And we want them right now’
So when it comes to the questions of language, What is it? Where did it came from? How do we use it? and Where is it going? adopting a slower science approach, where we make room for a little patience in the face of a deeply mysterious aspect of life, shouldn’t be seen as such a bad thing. I just think all this is incredibly interesting, and for now, I have no problem just cruising on the shoulders of these wonderfully fascinating theories out there to see if I can get a bit better perspective on what this whole language thing is all about:)
I think you forgot an, “Oh my” in the title
Really thought provoking post here. My daughter is just about to turn two. We are a bilingual family, and the explosion of language going on at the moment has left me with a lot of questions (and quite a bit of envy!).
Anyway, thanks for sharing all the interesting perspectives here. I really enjoyed reading it.
Haha, yes I did forget that didn’t I:) Thanks for your comment and glad you enjoyed the post, it was a bit rambling but I hope that it gives readers some different perspectives to think about. That’s exciting that you’re a bilingual family with the opportunity to see all of this first(s?) language acquisition stuff in action. Here is a book(PDF) I’ve enjoyed reading that might provide you with some food for thought in exploring bilingualism a bit more:
Aloha!
Great! Thanks for the book. It will definitely help to read something on the subject while witnessing the linguistic phenomenon going on at our house. One of my friends is in a house with three languages: English, Polish, and Japanese. Now that is a trip! I think those kids can actually move things with their minds…
You wrote: “Let me put it this way. It is very convincing to me that there is something special about the child’s brain that allows them to acquire a specific language so uniformly across all languages on earth, and it seems that there is something different about adult brains because there is such variation in their ability to learn another language.”
Let’s assume, for the sake of argument, that there really is a critical period for mastering accent and intonation. I think that both the scientific and the anecdotal evidence are pretty convincing here—everyone knows adults who’ve kept their accents after decades of professional work in another country.
Once we get beyond that, though, I think that far too many SLA researchers have drawn sweeping conclusions based on scientifically-inadequate experiments. If a researcher wants to claim that children have unique language-learning abilities, when compared to adults, then they need to perform an experiment where the only variable is age. And that is very hard, and so researchers rarely try to approximate it.
But let’s try to imagine a fair experiment. We want an adult to receive the same quantity and type of language input as a small child, and evaluate them by small child standards. We could approximate this by placing an adult in complete immersion for 2 years, without access to their native language. Furthermore, we want the adult to have access to the same simplified, repetitive language that children get. We can approximate this by having the adult work as a secondary care-giver to kids being raised in the foreign language (and the adult’s native tongue is still off limits!). This will give the adult two years of highly-repetitive, full-time input, and a desperate need to communicate.
Once the experiment is over, we also need to compare the adult to an appropriate standard. I propose we compare their language skills to those of a young child about to enter kindergarten. Kids that age actually speak fairly slowly (compared to adults in a fast-moving conversation) and they have limited vocabularies. And kids are notorious for using weird turns of phrase, as anybody who reads newspaper comics about kids could attest. Dennis the Menace is *not* a competent adult language user; that’s the whole joke.
Now, if we want to hold our adult to higher standard, they’ll have to go to school, read hundreds of books, interact with peers, watch thousands of hours of TV, and do all those things we expect children to do. Seriously, how many college freshmen can actually write competently after 13 years of school? I wrote better than many, and I suspect that my term papers were still pretty awful.
What’s your intuition about this experiment? I’ve actually lived through a very watered-down and imperfect version of it, and I’m pretty certain that I can do simultaneous translation for typical 4-year-olds, and do so in an acceptably idiomatic and grammatical fashion. I won’t say that I’ve never read a grammar book, but I certainly haven’t made a regular habit of doing so.
We could attack this question from another direction. Do children ever fail to acquire a language to an acceptable level? It turns out that this happens all the time, as a quick Google search for “heritage language learner” will reveal. Children actually need pretty specific circumstances to acquire a language, and they will fail in various ways if those conditions are not met. For an even more depressing body of evidence, search for things like “I have no native language”. This is typically a consequence of using one language with peers and another at school, or switching schooling languages several times.
And finally, a few theoretical comments:
- The argument about “poverty of the stimulus” seems to based on a misapplication of Gold’s Theorem. Kids get *enormous* amounts of input, and that it’s really not that hard to acquire, for example, probabilistic context-free grammars from input. See Norvig’s essay on Chomsky and statistics for a pointer to the relevant papers. Chomsky may still be convincing to his partisans, but it’s certainly not an uncontroversial claim.
- I don’t actually think adults should try to learn languages like children. I personally subscribe to a vague form of the Noticing Hypothesis. If adults want to acquire, say, an unfamiliar gender system, I think that it helps enormously to pay attention to the relevant input. In practical terms, try to cultivate the ability to be startled if somebody ever says “une sandwich” (which can happen—it’s masculine in Europe and feminine in Quebec). There’s some evidence that adult brains try to tune out grammatical details, and paying attention certainly doesn’t hurt.
- Even adult language learning can seem almost miraculous at times. Given enormous motivation and no other way to communicate, adults routinely blow through the equivalent of 10 years of language classes in 6 or 12 months, and reach 10,000+ word vocabularies in a couple of years.
Hi Eric, thank you for your comment. I can see where you are coming from and I enjoyed reading your hypothetical experiment. Alas, there are many experiments that SLA researchers would need to do to ‘really’ get to the bottom of the issue, but for both practical and ethical reasons, (experimental) researchers have to often settle for less than ideal conditions.
If I understand the point you are making about the experiment, it seems more of a question of ultimate attainment that you are arguing for, whereas I was more interested in the cognitive and interactional process of language learning. I have no doubt that you are correct, there is evidence, at least from what I have read, that individuals past the ‘critical period’ (whatever that may be) are fully capable of acquiring aspects (morphosyntax for example) of a language equivalent to ‘native speakers.’ (I’m putting contested terms in scare quotes, but I’m using them for the sake of clarity)…
As you mentioned, you, and I’m sure many many people have lived through ‘watered down versions’ of the experiment you mentioned. The general idea floating around in the SLA research is that adults are better (faster) for the first two or so years, but ‘young learners’ surpass them later on (with more native-like morphosyntax). This is of course for second language acquisition, not first. SLA by definition is looking at people acquiring a language with one (or more) already in place to various degrees, whereas children are coming to the language learning table without a first language. My point with the quote you highlighted was to show the uniformity of first language acquisition vs. the variability of second language acquisition.
This means that I won’t quite be able to agree with how you’ve characterize heritage language learning. I’m a little unclear on what you meant by the following statement: “Do children ever fail to acquire a language to an acceptable level? It turns out that this happens all the time, as a quick Google search for “heritage language learner” will reveal.” And the rest of your comment after this statement seems to imply that multilingual children (children using different languages in different domains of use) are linguistically defficient, but I may be interpreting this wrong so I’d be interested to hear a bit more about what you mean there.
As for Chomsky, my own perspective is quite different, but I tend to use him as a baseline for arguing for different positions on language acquisition since so many people tend to be familiar with his work. If you’re curious about my own perspective on UG, the closest to my own view is Terry Deacon’s position
Let me know if this helps clarify my points a bit, and thanks again for the comment, I always enjoy it when people engage with my posts in more depth.
Aloha!
Thank you for link to the Deacon essay! There’s lots of interesting stuff going on in that argument, and I’m going to have to think about it for a while.
I apologize for having elided my argument about heritage learners from my earlier post. In fact, “heritage learner” is a pretty broad category, so, for the sake of this argument, I’m going to narrow it down to kids raised in homes where one of the dominant home languages isn’t used by the larger community. Typical examples would be the children of immigrants, or children who are raised to be bilingual using approaches such as “one parent, one language” or “minority language at home.”
Heritage learners are a pretty unremarkable fact of life when you live next to a big research university. In our immediate circles of friends, there’s a whole lot of bilingual families and bilingual kids. One local high school student observed that in a class of 20 students, they could count 12 different home languages, or something crazy like that. And like any other parents, we compare notes.
So, take a bunch of kids, all of whom receive full-time input in a minority language from at least one parent. Look at them at 4, at 7 and at 10, and compare them to their monolingual counterparts. The picture actually looks pretty good at 4 for a lot of kids, although their minority language is often behind where it would be for monolingual kids of the same age. By 7, many of them are actually losing ground, especially in active skills. By 10, the minority language may be almost entirely receptive, even if it’s still used heavily by the parents at home. I know somebody, for example, who can’t put together a coherent sentence in their first language, even though their mother still uses it almost exclusively with them. (This isn’t the only possible outcome, of course.)
Some parents, of course, will send their kids to live abroad, or pay for various sorts bilingual immersion schools. And those bilingual immersion schools often come with sober warnings: “Your child will quickly develop conversational proficiency, often on the order of 6 months. But developing age-appropriate academic language will typically take 3 to 7 years.”
When you get right down to it, kids have to work pretty hard to acquire languages (even their first one!), and they’ll mostly wind up speaking the community language(s) at “native” levels with just enough minority language(s) to get by.
Meanwhile, look at the foreign graduate students and postdocs at a research university. Many of the arrive with English that’s somewhere around B2 or C1 on the CEFRL scale used in Europe: Enough to function independently, or even professionally, but often a long way off from native levels. But if those students socialize in their second language, marry a native speaker, and spend a decade or two working in their second language, they’ll generally be highly fluent, with a faint accent and a few traces of grammatical errors. And most of these people don’t have time for English classes or grammar books, because they’ve got to publish or perish, raise families, and all that.
So kids generally won’t learn languages if they don’t have to, and adults will often learn if they have no choice. There may be some differences in learning mechanisms, and there are usually some differences in ultimate attainment.[1] But my intuition says that if you could somehow hold everything else constant—with blowing the research budget or rightfully alarming the ethics committee—age would be far less of a factor than folk wisdom would have it.
[1] Though I’d argue that any limitations on adult attainment may not affect the adult much in practice. On the other hand, survivorship bias undoubtedly plays a huge role: If there were adults who were incapable of leaning a foreign language, they would avoid the kind of long-term language use that’s necessary to reach high proficiency. Anyway I look at this, it’s difficult to get rigorous evidence!