Special of the week; Issue 3-70 "Learning to Think in Our Native Language"

And we are the national educational radio network presents special of the week from Yale University from its series called Yale reports throughout the world when a baby starts to babble. There are six sounds he makes and international baby vocabulary from these six sounds develop the baby's ability to speak in his native tongue. They lead to all the different languages we have. Do babies learn their language from imitating the adults around them. Or is there something in the brain which helps them pick and choose among the various sounds they make. Discussing these biological and cultural constraints are William caisson professor of psychology Sydney and professor of linguistics and Alvin Liberman professor adjunct of linguistics is to Keston a couple of problems in trying to understand the child's acquisition of language one is a purely empirical one of finding out what the facts are and I think that's a place where lots of interesting things are going on but there's another contrast which is a theoretical one about whether we can understand

the learning of language the way we understand the learning of simple motor habits the kinds of skills the child learns when he learns to pick up a ball or something of the sort. Whether we have to call on more elaborate. Propositions having to do with biological constraints that is what does the child bring to this to this language learning task which limits what he can learn. Though you seem to propose these two points of view as alternatives that are in some kind of conflict with each other and I'm not sure I can see what the conflict is that is good for example. When you talk about learning simple motor skills isn't there also some equipment that the job brings to that situation. Yeah there is I think that probably if we look at this contrast very closely turns out to be more a question of the move ability of the system. And is that the notion that if you reinforce or reward particular motor habit you can you can make it and anything you can shape it up into anything. And the first proposition I think we must call into question is that one about language that isn't there the biological Givens and

language learning are such that not everything is possible. Isn't it really the case that there are also limitations on what you can teach. A goose or something to do in the area of simple motor habits and these limitations already find exactly boy his nervous system and physiology and so on. And isn't this really the case also for language. The most interesting word that you use there I think Sydney is teach languages really don't have to be taught. They've got to be acquired. But they don't have to be taught in the same way that one has to teach a child to play the piano or two. Teaching mathematics for example. Yes I quite agree with that and I think that's enduring I would still question whether there is really something different going on inside the nervous system under these two circumstances. That is teaching as we think of it is a process in which some other individual is consciously interacting with the person who is learning. In that sense teaching is different from the situation in which that is not present. What if we look at what's going on just

inside the brain of the one who is learning. I wonder whether these two processes are really that different. I think probably that our life from the traditions of learning psychology are speaking of a dogmatic position which is now under serious doubt even outside the language area that is that all learning not only. A good program it's kind of shaping but the shaping takes place by a social agent. But is that the child makes a proposal of some sort he says something and then they they the teacher is standing there by some adroit procedures of rewarding the right kinds of approximations gets him to speak. And I think this is the this is the aspect of the model which simply doesn't doesn't hold for kids learning language. Speech is a deeply biological kind of activity and that's it's the same thing as saying it doesn't have to be taught it's much like walking requires experience it requires activity requires a great deal of experience and we know that speech is something that just comes about without any conscious

teaching on the part of others and yet of course there is a considerable amount of diversity among different languages doing upon the particular environment that the child finds himself in. Now let us consider also in the area of various kinds of motor activities things that one can do with one's arms and legs. Here too there is a great deal of constraint that is just built in to the physiology of the brain and the particular things that the child learns to do with his arms and legs are going to be to a large extent culturally determined. In other words we can't make these two situations appear to be similar. The I think though that if you take a somewhat different attitude one that has popularized lately that what you've got are two systems in interaction namely a set of givens that the child brings a kind of set of restraints and constraints on his behavior and then ascend a set of pressures from the environment. What. What James Marc Bolan called in insisting self in the resisting world and that any behavioral outcome

particular in the case of language is the reason the result of this of this intersection. You begin to think in somewhat different ways about what constitutes early language learning. How is it for example that the child comes to to adopt a logical patterns of his own particular natural language the one that's in it's in his environment. Certainly they simplified learning model that somebody teaches him that this is an appropriate sound in English and not an appropriate sounding which is wrong and right. Oh yes I agree entirely abandoning that model permits us to ask a set of more interesting questions about how this kind of effect of experience takes place and I guess maybe the simplest way of saying it is that all effects of experience cannot be reduced to the simple kinds of learning models that have been available in psychology over the last 40 years you know here or perhaps the question of learning languages not so different from learning anything else.

What you were just saying about the process of learning language that is that people don't really teach children to speak. It seems to me that's pretty largely true of just about everything that you do. The child learns until he gets into the first grade and then when he gets into the first grade Well really perhaps what he learns from the conscious efforts of his teacher isn't really all that important in comparison to the other things that he's learning outside of the classroom or even in the classroom apart from what the teacher is doing. In other words I still don't see that there's much of a contrast between language and non linguistic activity that the child is learning. But I think there is a difference. Within the within language in the way the child learns his first language and the way he learns some language in college. A very great difference. Yes when you are in a school situation then it's different here I think we have a real contrast. Well there might even be a difference in the way a 25 year old adult learned a second language not in his office.

Yes yes indeed there is. And in fact I was not nearly so I'm sure that you're aware of this Al but I think we won't mention it that the brain seems to work a little differently. After about age 11 and before age 11 which means that especially with regard to language Yes and very likely with regard to many other things to all the languages the particular type of system that has been most perhaps looked at in this connection. I would agree with you in any case that there is really not so much a conflict between these two points of views as there is a question of which aspect of the problem do you find more interest. Are you going to learn more that is somehow interesting and useful by looking at these biological constraints. Or are you alternatively going to learn a greater number of interesting and useful things by looking at the conditions under which the the acquisition occurs conditions of reinforcement itself. Which I happen to think are relatively trivial. I think that at the moment what we need to know is a great deal more about the biological constraints.

It might be interesting that's that's good to talk just a bit about what what we have to explain here. But you mentioned one of these things namely that it seems to be increasingly difficult to learn a language and a very nearly impossible to learn it with a with the clarity of a native speaker after about adolescence. But of course with my own interest in infancy I'm particularly struck by these patterns of regularities that appear in the first two years. I don't think it's too exaggerated to say that that we have what we have to explain as a sequence that goes something like this. The child early on has these vocalizations which are almost exclusively of vowels. And then somewhere in the first half of the first year B and the first magic moment you know where he started saying. And to repeat some babbling thing which by the way is an interesting constraint and its own right it has been asserted by some people that there is an international big baby vocabulary that has the the six words Mom and Dad up above I've recently seen fascinating data on that very point.

Is it confirming it all yes youthful when children babble the stops. But I've got one can ask about that babbling. What how they time two parts of the gesture two parts of the gesture or the pulling together of the chords for Foundation and the opening of the vocal track for this stuff. And I've seen records of 8 and 10 month old children some growing up in American homes and some growing up in Lebanese homes. They babble this in exactly the same way with the mode at 20 millisecond lag between the two gestures and right smack at 20 milliseconds of like. It turns out that when these children are 3 4 and 5 years old and are asked to mimic a set of synthetically produced sounds that sample the whole range that there is only one of these synthetic sounds that is mimicked accurately by both the Lebanese children and the American children that's the one with the 20 mil a second life. And that's in spite of the fact that the

particular phonological or phonetic distinction that we're talking about here is used very differently in Lebanese Arabic and in English. The but still there is this this pivot this this clear universal that all these children do this one thing in exactly the same way. And then the cultural influences begin to pull them in opposite directions and this is also apparent. And one can see in this one very interesting study the interplay of these two factors the biologically very interesting biologically given constraint and the cultural influences. And one can also see why it is that this particular phonetic distinction this is a voicing distinction tends to be used in all the languages of the world or is used in very nearly all the languages of the world. There is obviously of a biologically given notch in the system that makes this highly distinctive a very easy thing for human beings to do. And I suspect an almost impossible thing for any other primate. That's all I can say by the way that this work is done by Malcolm Preston racing to compete in Johns Hopkins.

That's splendid to hear all because that of course indicates in one way why language studies are so fascinating here you have a system which obviously has some sort of constraint which is put under pressure by a bicultural system and one ought to be able to start making predictions about how that how that conforming ought to go. One would expect that a child will show a hierarchy of development that as he has he can't move from his initial system to the final system and one in one leap and the next level the next generation after the mound I got taught to be describable as a function of one how far they are from the initial set and what the definition of the ultimate natural form logical structure is so that one ought to be able to say for example that the child can say monarch Insania up before you can say. And also that in a particular cultural system he will move more easily in one set of lines in another. As far as I know we have no information what the secondary sources is called the Irwin and chin stuff about how all children make all sounds of all languages and then they're

selected from which is surely wrong. I mean that that that I think we can we can get that. Well perhaps it's somewhat exaggerated statement but I don't think it's completely wrong. That is we do know that a child at a certain age makes a great variety of zones. More variety than here in the U.S actually use later on but that is a rather interesting observation that we also know from this same study that I referred to that children do not babble. What you Sydney know is the aspirated stops in English. Yes I would like to say in this connection that in the case I didn't say it before that. There are interesting biologically biologically given constraints and uninteresting constraints and I think an uninteresting constraint an obvious one is that all speech has to occur within a certain frequency range and certain levels of intensity just so that people can hear it. But I think the kind of phonetic universal that I referred to a few minutes ago is quite different

from that and is a good deal more interesting as it is I think the fact that all languages are syllabic and all children babble in syllables because the syllable represents a particular way of organizing the phonetic structure of the language is a very complicated way. And we have every reason to believe that that children can do this only if they have a very special kind of apparatus which enables them to package the phonetic the phonetic segments into syllables and they also need a very special aberrated if they are to recover their eye segments from the spoken syllable. And this apparatus is almost certainly part of their there. Biologically given equipment which is by the way not to say that it doesn't require some kind of experience in order to you know bring this to its full flowering. Yes the critical observation here is that deaf kids will begin this process but very very quickly drop it off if they aren't permitted to hear it. It's interesting also in that in

that we've been talking for the number of minutes and have got the kid up to about about eight months in and his development. We've been struck in some observations we're making a very casual kind mothers report as a matter of fact of early vocabulary with this problem of the first word which seems to me to have such great psycholinguistics and psychological implication. And at least preliminary observation suggests that this very very small vocabulary that kids have between 8 months and 15 months which for most of them is two or three or four words fall fall into two groups one group can be called Action words or sensorimotor words if you really want to make the connection to pages I do such as hire by buyer up and then the other group are particular names mama that the name of the dog the name of the security blanket and there are there are no categorical names no class names no indication of sets that contain more than one member until until 15 months.

And these action words and particular names you see suggest that in fact these the first word is a verbal act like like other acts and just as the kid can can grab the ball or suck the ball or look at the ball. He can he can name it that these first these first words may not be different at all from the other kinds of things that he's that he's doing and that the next magic moment is at 15 to 18 months and I guess from little burgs data this is a quite regular dating not only in our culture but in others. Three things happen. One he begins to make these class words. He begins to to say dog to refer to more than that and his second almost immediately starts to whir grammatical utterances. We don't have any case and I guess our data are terribly limited but there are no irregularities in them. You don't have any two word utterances before you have evidence of a category noun. And the third thing that takes place of course is this fantastic explosion in vocabulary it's almost as though the

kid has said to himself Gee I can't get along with particular words. But once I've got this concept once I've got this category thing I can really increase my my list tremendously. So at the same time there's a sharp rise in the book. There's probably not just this category thing as you call it Bill but also the fact that his phonetic equipment is getting a little better. You just can't have a very large vocabulary unless you have the facility to produce systematically a certain amount of phonological diversity. And once he acquires the ability to make further the logical distinctions and he has the ability to acquire many new vocabulary items it's useful to state they a apparent empirical complication that children do not say merely what adults say. That is the simple imitative model is not going to I is not going to do for us here all the way to the difficulty with the simple imitative model

is really just the simplicity is as I see it now many people have I think gone too far and they have wanted to reject the idea of imitation altogether. And if one looks not just at the surface of the linguistic data but it underlying patterns that are in the linguistic structure then you can see a lot of evidence for actual imitation that would not have been apparent if you look only at the surface data. It may be that we or you know swing of the pendulum that is going too far against those imitative wishes. But I think the question is simply whether imitation and trial and error behavior are sufficient conditions for the child to have to learn a language. And I think it's quite clear that these are not sufficient that the grammar of any language is sufficiently complicated that it's almost impossible to conceive that a child or a machine could possibly acquire this grammar acquire the system mastered this system by by simply trying out all kinds of hypotheses at random. Hi Graham inclined to agree with

those people including in particular Chomsky who argue that the child must come into the situation with certain prejudices certain innately given tendencies to entertain certain kinds of hypotheses. Yeah that's what makes it such a fascinating problem I don't think any of us would assert that the child by Russia as a nation would it that the age of 3 and with no experience begin to. Again begin to talk. But there he does. Along the way by golly invent some sense invent the language that is on the basis of the data. Come up with proposals or hypotheses or ideas about what the what the structure of the system is and you might see on the basis of very bad very bad data Yeah and and the environment simply doesn't doesn't give him that much help. I think our model our teaching model of what happens between parent child which may in fact happen and in some like middle class families will not do as a general model for the way that the kids are now. It does not seem to be so hard to understand how he arrives at these strict patterns. That is it seems to me that the very kind of equipment that we

have to. Hypothesized to account for how the brain acquires information at all is if it is allowed to function is automatically going to go up to a higher and higher levels of abstraction. Let's take an actual example children of a certain age when they hear a passive sentence like the wolf was killed by the tiger. The child will interpret this at a certain age as meaning that the wolf kill the tiger rather than the other way around. No this is because he hasn't yet learned the purpose of construction. But. Since he here isn't learned it what he does do in decoding that utterance is to use the equipment that he has equipment that he who knows will allow a certain amount of that utterance to get through to his understanding and the rest is simply going to be filtered out. Although what gets through is will kill Tiger and he interprets that with the structure that he has the child uses whatever linguistic machinery has at a moment interpret into the environment and for interpret in a linguistic environment so he can only understand

his his utterances or comprehension by understanding his theory as well as the input and you can do this on on from a logical level as as well watching three little kids simultaneously exposed to the same language environment is a revealing experience and I recall in the case of the of the word spool and where they had all three groups of triplets heard the same sounds coming into the environment one of them at a particular moment didn't have any words visible in one of them called it move many of them called it barren. It is that with the same inputs they are clearly developing somehow their own their own full logical classifications within the classifications of the language and will take that input of spoon and give it back in the theory of the sounds of the language that they happen to have that you know. Let me go on to suggest that what you were just talking about is really a very very general process that applies not only to language would to non language and not only to children with two adults.

The general idea is that whatever input is presented to a person's brain through one of the sense one of more of the sense organs is in perceived and interpreted on the basis of the equipment that the person happens to have in his brain. Give it time. And so this is why it is that different people representing different professions when they go to see a museum they will look at the same object and they will have entirely different reactions to it. This is just a case of processing that input with whatever equipment they have and it's exactly the same thing that's going on when the child has an imperfect knowledge of his language and he hears these areas inputs. He doesn't act the same way that the adult does because his equipment for decoding the linguistic input isn't the same as the adults. And it isn't the same as other animals. I think perhaps somebody should make explicit what has so far only been implicit in this discussion which is that the language is what the biologist calls species specific behavior all men speak all men have language no none man no non-human animal has language. It's true of course that many

animals communicate in one way or another but they don't communicate with language and this is to say not only that they don't speak but that whatever communication system they use is not organized in the way that language is organized Indeed so far as we know non-human animal communication doesn't have much in the way of organization. Each utterance each signal is uniquely in whole distinctly different from every other. I think it's very clear on this basis for this reason and others that man has very special devices in his head. Which process language it at all levels and the children have these devices in their head heads. I think for example that we have in our heads something like what neurophysiologists have been referring to as feature detectors devices that in the case of the human being making particularly sensitive to the speech signal and to those parts of this very complex signal which carry the important linguistic information I happen to think there's a great deal of evidence for this. And this is a very special kind of device at a deeper level of course. It's not all that special. Other animals also

have special devices which may be constructed on similar principles but which are turned to other purposes it's well known for example that the frog has a device that makes him particularly sensitive to objects moving across his visual field at a certain rate. This is come to be known as a bug detector and this is you know just as it should be the frog makes his living catching flying bugs and we make our living talking and listening. It's interesting how far one can push they the animal speech. There's one point where the where the animal the human being may not be that distinct and that is on the use of not linguistic communication signals we've been talking perhaps because we're dominated by the concept of the radio as we are by the concept of the telephone that all the information we we. Give out has to be interpreted interminable purely in linguistic signals and I'm sure if the people who are listening to us could see us we'd be communicating in many many different ways that don't involve language and I think it's useful particularly in regard to the kid to think about how he learns that one of the characteristics of children speech is that they are insensitive to this need to to

state the domain of what's going on that is a telephone conversation with a child is very interesting because he doesn't confide in all of his all of his utterances to those which are coded linguistically and is likely to say look at this or what's this Deady of over the telephone. Clearly not having recognized this distinction. The other question said I know you've thought about this is how we think how we signal non linguistically certain parts of our communication in the in the child. I wonder whether it isn't absolutely necessary for him to learn a particular language for the mother to be present coating what she sees him do as well as what he does when he says for example baby milk. Unless she's there too to get all the rest of the communication. She can't reflect that sentence back to him proper. Well in general among actually not only babies but also adults language is just one part of

a total communication system that we are using all the time. It happens to be of course a very fascinating and highly developed part. But in the child especially it's really not differentiated from just communication in general. We probably have been particularly focused on the verbal kind of communication since we make so much use of writing and the telephone. But writing pushes us in this direction even more than the telephone because we get no tone of voice or volume or speed and so on we have only the words in the punctuation and it gives people in our society perhaps a rather distorted view of what communication is all about in the usual human situation people are face to face and they're using gestures and body distance and body motion and all kinds of things looking at each other. There is some studies been going on recently on the nature of the stair for example the stairs a threatening gesture and the luminary observations indicate that but one study

let me tell you about very informal a man would drive in his car up to a stop light and someone would come next to him and he would simply turn and stare at the driver in the other car. And then his response measure was how long it would take the guy when the light went green to go from where he was to the other side of the street and had the proper control groups non staring groups in this clearly. If you stare at people they will they will leap off their off their starting point very much faster. There clearly is some fairly regular signal contained in the speechless stare which must be only part of a very very much more complicated communication system and this system to point out a great many unique characteristics. Certainly not human. You probably in us to yes there are here we can speak about universals not only for the whole human species but properties communicative devices and so on that are shared by human beings with other mammals. And it certainly is worth saying having covered so much so many tasks so that

child is just as well he comes in with some part of this problem already solved. The idea of his having to learn all of the things that we've talked about starting from scratch would be would be overwhelming for any system and we'd be more than a lifetime learning disabled. How children learn their native tongues discussed by William caisson professor of psychology Sidney m professor of linguistics and Alvin M. Liberman professor adjunct of linguistics scripts for these programs are available without charge by writing to Yale reports 1773 station in New Haven Connecticut 0 6 5 2 0. This is Charles Dillingham for Yale reports which originates in the audio visual center of Yale University. And Norman Holmes peers and professor of English and American studies scripts for these programs are available without charge by writing to Yale reports 1773 station in New Haven Connecticut 0 6 5 2

0. This is Charles Dillingham for Yale reports which originates in the audio visual center of Yale University. Any ARS special of the week. Thanks Yale University for this edition of Yale reports. This is an E.R. of the national educational radio network.