icon bookmark-bicon bookmarkicon cameraicon checkicon chevron downicon chevron lefticon chevron righticon chevron upicon closeicon v-compressicon downloadicon editicon v-expandicon fbicon fileicon filtericon flag ruicon full chevron downicon full chevron lefticon full chevron righticon full chevron upicon gpicon insicon mailicon moveicon-musicicon mutedicon nomutedicon okicon v-pauseicon v-playicon searchicon shareicon sign inicon sign upicon stepbackicon stepforicon swipe downicon tagicon tagsicon tgicon trashicon twicon vkicon yticon wticon fm
13 Nov, 2020 06:59

Babies learn to speak in the womb — linguist

Language is what makes us humans unique. So, what's behind one of our species’ most incredible talents? And how is it changing with technology? We ask Robert Berwick, professor of computational linguistics, computer science and engineering at MIT.

Follow @SophieCo_RT  

Instagram Sophieco.visionaries

Podcast https://soundcloud.com/rttv/sets/sophieco-visionaries

Sophie Shevardnadze: Robert Berwick, professor of computational linguistics, computer science and engineering at MIT. It's really great to have you with us today, professor, welcome. 

Robert Berwick: Hi. Great to be here. 

SS: Hi. So I mean, look at us, we're like 7 billion people and we speak 7000 languages, we need language for communication, but is language in that sense kind of preventing communication rather than facilitating it, I mean, since every language constitutes some sort of a private club, that's more or less closed to outsiders, right?

RB: Right. That's always been one conception of language, it’s that collectively the people that speak a language, even it varies a little bit from person to person, you can think of as a kind of cultural artifact that's been created by the people that all speak that way. Of course, the way we want to study language might differ from that because language can be used for lots of things. It can be used for communication, but it can also be used for miscommunication, as we've seen in the current election, not only just here in the United States, but lots of other places, too. So there's this aspect of language about what's knowledge of language that's inside your head that every human being shares. But then there's also a separate notion about language, about to what use you put language. And those are two different things, two different ways of looking at language.

SS: So if linguists can now think about the evolution of language, where is language evolving? I mean, is it further being divided into many different ones? Or will we just all speak in common tongue one day, sooner rather than later?

RB: Well, I don't think we'll arrive at some kind of lingua franca where there's just essentially one language that's used by everyone. I think there you have to distinguish between, again, two fundamental ways of looking at the evolution of language. So one way is to think of - because you can think of the evolution of what was the initial origin of language, you know, hundreds of thousands of years ago. But then there's the business about how languages change over generational time, from one generation to the next. And at that finer level, we see that languages do actually change but it depends upon the number of people that are speaking them. If there's enough of a group of people that speak a particular dialect or language, then by and large that survives, and that's subject to historical forces and other factors. So in fact, there are some places where people tried to revive things like the Scottish dialect, or we've actually revived some of the Native American languages, to an extent where even though they were lost, there was actually a project here, near where I am, to revive one of them in the southern part of Massachusetts. And that was at least a little bit successful.

SS: Scientists say that at least one language is dying out per week. And that means that nearly half the languages we speak will be gone within the next century. Why is our linguistic diversity reducing so dramatically?

RB: Well, I think that there's been speculation about that, I think a lot of that is due to the Internet, and the fact that, you know, people use languages like English or other common languages in the EU, and they tend to wash out these smaller dialects. And if you get down to a small enough size, then that language will just drop out. And if you don't have someone who speaks it to your kids, then you can lose it. And that's a concern. That's a concern that a lot of linguists have, and there are, in fact, efforts by the societies of linguists, like the Linguistics Society of America has one, and in Canada and other places to actually try to preserve some of these more minor languages where there's not a lot of people, not a lot of speakers for them. But you're right, that is a something to be concerned about. Because once you lose that language, then unless there's some concerted effort to try to revive it, just like animal species, or plant species, you might lose them forever. So that's a concern.

SS: Neural engineering is already making it possible to send signals from the brain to external devices like robotic limbs. And many scientists are actually also saying that one day, we might be able to communicate directly brain-to-brain without actually using speech or writing. Will we still need language for that?

RB: Well, that's a terrific question because actually, what that brings into focus is this difference between the external aspect of language, the fact that we can either speak it or if you're a sign language user, you can use manual gestures, for example. So there's that aspect of language. But there's also an aspect of language that has to do with its use inside your brain as a kind of inner thought. We all have this sense that we're talking to ourselves, that might be more prevalent in some people than others. But I think when you begin to reflect on it, a lot of your own internal communication is in fact inner thought in that sense, and that's not speech. It seems to us like speech, but you can actually show that it's not. And it looks like in that sense that language is a kind of inner mental tool that we use for pulling together many of our thoughts and ideas, and making them coherent. People have done experiments about this to actually see that language, internal language serves as a kind of universal, you know, again, like a lingua franca, system, for pulling together all the other things that are going on with your, your sensory and brain activity. So you're constantly getting information about, -  you know, when you're looking at an object, you get, you know, what that colour of the object is and where it is. So we know there are particular areas in the brain that tell us what, and where things are, like the colour, where it is, there's this kind of geometric ability that we have, it's built into us. And it turns out, you can actually show that it's language that pulls those things together so those different components of your brain can talk to each other. So language, in that sense, is actually important internally. And I don't think that's going to be sort of put aside by these kinds of abilities, these kinds of neural abilities that say, “oh, by sending a little electrical signal I can that I make inside my brain, I can, you know, turn on the lights, turn off the lights”. I don't think it's going to be so easily replaced.

SS: But I still wonder, can we think without language? I mean, toddlers or people who only speak sign language, what language do they think in?

RB: Well, that's, again, a really wonderful question because as far as we know, a language like sign language, you know, the people that grow up, kids who grow up only using sign language, think just the way that ordinary people think. I mean, the way they might express themselves can vary. So a very famous experiment that was done years ago by Lila R. Gleitman and Barbara Landau, they did the following thing. If you blindfold a normal child and you tell the child to look up, they'll go like this, makes sense, right? If you take a blind child and you blindfold them, and you say “look up”, they go up, they raise their hand and start to prob with their fingers, which is what a blind child will do. And I see that's very interesting. They have a different modality, a different way of expressing the same thing, they just do it in a different way. But inside their heads, as far as Gleitman and Landau could make out, their thinking is just exactly the same. So we think what that means is that inside us, what we have, our inner thoughts are in fact very language-like. Although it's certainly true that there are other kinds of thinking, like visual thinking and the things that, you know, a painter or a mathematician might do that, you know, lie often outside of what language strictly is. But it looks like those might all be connected up with language very tightly.

SS: Does the language I speak influence my thinking? I mean, for instance, the English language likes a more rigid word order than Russian -  

RB: Right. 

SS: Does that shape the way Anglophones or Russophones think?

RB: Again, that’s great! As far as we know, while the external part might vary, so, Russian might have a more flexible word order or not. There are languages like this, the famous Australian language Warlpiri, where the word order is completely free. So it’s even freer than Latin, so I can juggle the words in almost any order. And yet, it turns out the internal thought processes as far as we can gauge that looks invariant, that looks like it's the same from person to person. Now, of course, there's some variation, there's certainly personal individual variation. And we can't exactly tell that, but so far as we can make out the internal part, it’s all fixed, which is sort of what you want and that's what you want. So it's almost like, you know, when in the US, when they used to make all sorts of cars of a certain type, the engine and the inside will all be the same, and then you would just stick different fenders and different things on the outsides so it would look a lot different. But sort of the inside machinery would all be the same. Now, not everyone who studies language shares that view, but that's certainly a view that I hold and other people who have looked at this sort of inner structure of the way that language interfaces to other aspects of our thinking.

SS: Professor, Goethe said that the more languages a person speaks, the more languages a person knows, so many times, he's a man. Is there a limit -

RB: That’s a nice quote.

SS: Yeah, I love that quote of him. Is there a limit in the human brain to how many languages one can speak?

RB: You know, that's, again, a fun question to think about. We have some sort of natural experiments that we were lucky enough to actually look at people who know many different languages. And it doesn't seem like there's any real bound. There probably is because there's some limit on the number of words you can actually sort of keep separated and compartmentalise. But there's a very famous example of a linguist who passed away about 15, well, I guess about 20 years ago now, who was at MIT, his name was Ken Hale. And he was a person who never lost the ability to learn language. What we know is that around puberty and to some extent after that, there's something that goes on in the brain. You probably remember puberty, that's a time when the brain either matures or immatures, depending on your point of view as to whether you're a parent or not. But we know that there are certain things that happen, so the brain actually becomes less plastic and we lose this ability to learn language as if it's your native language. But there are a few people, there are very rare cases where that doesn't happen and one of these people was Ken Hale. He realized this first when (he grew up in Arizona, in the United States, actually New Mexico) his parents sent them to a cowboy camp for the summer, and his roommate was a Navajo kid and Ken Hale learned Navajo within a few weeks - boom! It’s a tremendously difficult language to learn, you may be familiar with the fact that the Navajo language was used during World War II. 

SS: Yeah, the Japanese couldn’t decrypt it. 

RB: They couldn’t break it. But it was easy for Ken Hale. And he gets back to school, he was about 12 or 13 at the time, he learned Polish, Russian, German, French, boom, boom, boom, boom, boom, boom, boom. And his French teacher said, “Doesn't that just make you more confused?” And he said, “No, it just keeps getting easier and easier”. And I was actually was lucky enough to see this in person, because I actually, this was about 25 years ago, just walked down the stairs, there was a flight of stairs, about seven flights of stairs at MIT. He just met a Miskito Native American Indian, from Nicaragua, just started talking to him and by the time he got down to the bottom of the stairs, he essentially was fluent in that language, in the sense that he could speak it perfectly.

SS: Can someone explain how that happens?

RB: We don't understand how he was able to do this so well, but he was if he was actually just, you know, a child who could just absorb this.

SS: Yeah, but I mean, we're talking about prodigies, or like, out of the ordinary things.

RB: Yes, he was a linguistic savant.

SS:  But for the most part of us - is there like a limit to how many languages one can really speak?

RB: I don't think so. In the sense that we know that even from people who aren't savants like Ken Hale was, that if you grew up in a bilingual or trilingual household, the kids have no problem with that. In fact, it seems that for all we can tell, it actually improves their cognitive ability and their vocabulary and their general intellectual functioning. So I don't think there really is that kind of limit, we're very plastic. But it's never actually been systematically tested in that regard, I don’t think so. I think we don't understand all the prodigies. There are lots of other prodigies, but sometimes they have other kinds of mental impairments. So we don't know what's going on there. So I think it remains an open question. But I think that, by and large, it seems that you can learn a fair number of languages when you're a kid, and we just don't know what that bound is for ordinary people. 

SS: Do you think learning languages can be like an antidote to brain maladies of age like Alzheimer's? I mean, does learning a language or speaking actually multiple ones make human brains more resilient, elastic, stronger?

RB: Yeah, there are actually some results on that. And it does seem that if you actually even later in life start learning a second language or a third one that actually does help with plasticity and resilience against memory loss. But again, the studies with Alzheimer's and memory loss and language is a fairly new one. We have a couple of people at MIT. One of them is Professor Suzanne Flynn, who's working on that. And you can actually look at memory loss and language as a kind of guide to what's going on with Alzheimer’s. And she's found that bilingual people actually do better. So that's at least some evidence suggests that it is of help, which I find very heartening, actually.

SS: So some languages like Italian or French have a melodic reputation, others like German are viewed as more harsh, to which extent the sound of the language influences cultural stereotypes about those who speak it?

RB: There's a funny story about that, which is that you can actually pick that up as early as newborns. So one of my acquaintances, Angela Friederici, did these experiments at the Max Planck [Institute for Human Cognitive and Brain Sciences] where what she did was tape-record the cries of newborn babies at birth. And German babies would cry in a pattern like saying, WOah, and French babies would say, woAH. So that actually followed the rhythmic pattern of “mamA” versus “mAma”. So you know. So where did they get that? And it turns out, she looked into it and of course, what was happening was that the babies could actually hear their mother's speech in the womb because there were the lower frequencies actually transmitted enough through into the uterus, that the newborns actually have picked up on that already. So that actually, in a sense, prepares them for acquiring language that they're going to be born into, which is a kind of funny thing. But then there's this funny part that it actually followed the rhythmic patterns of French versus German. 

SS: That's funny. So biological evolution facilitated the forming of language and culture. Do language and culture impact biological evolution as well? I mean, I've heard this idea that the French have some special throat architecture that allows them to pronounce this raspy ‘r’. Does this idea hold any water?

RB: It’s not clear whether there's - See, when you actually look at the cases with French, it's not clear that there's been enough time for that biologically to have influenced what the structure of your throat is, and whether that's had an effect. So let me turn to another example. That's actually, I think, a little less well-known because there was work published on this about 1.5-2 years ago, where they looked at the people - essentially, it's the Khoisan in Africa that are able to make these so-called ‘click’ noises, which are actually very rare in the world's languages. These are the noises you make like that [clicking with his tongue] something like that. Now, that's very interesting for the following reason. Again, a Dutch linguist, another friend of mine, Riny Huybregts, who's at the University of Utrecht, looked into this in some detail. And there's two interesting facts about the Khoisan. One of them is that they emerged very early in Sub-Saharan Africa, almost about 140,000 years ago, as a kind of genetically isolated group. You can show they're genetically isolated, we can do this from the genome studies now. So that's one thing. And the second thing is that this ‘click’ sound, in fact, does require a certain structure to the roof of your mouth, that makes it easier to make those sounds and others. And these ‘click’ languages, you can show aside from some, you know, there's some few cases where the people borrowed the clicks, some adjacent groups and others, but Riny was able to show that there's a really sort of noise, so you can ignore. Those click languages are just associated with the Khoisan. So his hypothesis is, in that case, there is enough time for this kind of effect, to build up on the roof of your mouth, to make it adapted to this making the ‘click’ sounds. So there is some evidence at least in this one case where that can happen. So maybe not the French because it's too late. And there's too many cheeses in France anyway. So you know, but in the case of this sub-Saharan language group, I think there actually is at least some suggestive evidence for this.

SS: Yeah. There's lots of evidence that people tried to communicate by means of pictures before writing was invented, right? I mean, look at cave art. Now we communicate with emoji and pictures more than words. I mean, it saves you time, it combines meanings and emotions. So in the written language, at least, are we going to make words redundant and go back to basics communicating with images rather than syntactic structures?

RB: Well, I think that language still has a big edge because with language, one of the things we can do that makes it so special, is we can talk about things that are not just in the here and now, so that ‘I'm happy’ and so on, but I can make the sentences as complicated as I want with their structure. So I can say something like, “I believe that Mary thinks that Bill said, that John left, you know, yesterday”, that's hard to imagine- as you see that has the structure to it that looks like branches down. And you can't do that with just a linear sequence of emojis. Because it doesn't have that same ability to, you know, encapsulate a little world, you know, the world where John left yesterday. I can't, you know, wrap that up and put it inside another emoji and take that bigger emoji and wrap that inside another one. It doesn't let you do that, the emojis don't let you do that. So that's what the emojis are, you know, in a way, have stripped language of some of its enormous expressive power that I think you still need to talk about a whole world that you can create with sentences.

SS: Professor, it's been such a pleasure talking to you, I would carry on and on and on, but it’s all the time we have for this particular programme. I hope we'll get to do this again. 

RB: Okay. 

SS: So good luck with everything. 

RB: It was great, fun. Thank you so much. 

SS: Same here. Stay safe. So we'll see you hopefully sometime soon. Take care. 

RB: I hope so too.

SS: Bye.

RB: Bye now.

Podcasts
0:00
28:29
0:00
26:22