In Our Time - Language and the Mind

Episode Date: February 11, 1999

Melvyn Bragg and guests discuss the history of our ideas about the formation of language. The psychologist George Miller worked out that in English there are potentially a hundred million trillion sentences of twenty words in length - that’s a hundred times the number of seconds since the birth of the universe. “Language”, as Chomsky put it, “makes infinite use of finite media”. “Language”, as Steven Pinker puts it, “comes so naturally to us that it’s easy to forget what a strange and miraculous gift it is”. “All over the world”, he writes, “members of our species spend a good part of their lives fashioning their breath into hisses and hums and squeaks and pops and are listening to others do the same”. Jean Jacques Rousseau once said that we differ from the animal kingdom in two main ways - the use of language and the prohibition of incest. Language and our ability to learn it has been held up traditionally as our species’ most remarkable achievement, marking us apart from the animals. But in the 20th century, our ideas about how language is formed are being radically challenged and altered. With Dr Jonathan Miller, medical doctor, performer, broadcaster, author and film and opera director; Steven Pinker, cognitive scientist, Professor of Psychology and Director of the Centre for Neuroscience, Massachusetts Institute of Technology, California.

This BBC podcast is supported by ads outside the UK. Thanks for downloading the In Our Time podcast. For more details about In Our Time and for our terms of use, please go to BBC.co.com.uk, forward slash radio four. I hope you enjoy the programme. Hello, Jean-Jacques Rousseau said that we differ from the animal kingdom in two main ways, the use of language and the prohibition of incest. Language and our ability to learn it

has been held up traditionally as our species' most remarkable achievement, marking us apart from the animals. But in the 20th century, our ideas about how language is formed are being radically challenged and altered. To discuss how and why, I'm joined by Dr Jonathan Miller, Britain's most celebrated polymath. He started a career in medicine because of his interest in language. Since then, he's been a performer, a broadcaster, an author,

which is still is, of course, a film and opera director. He's just curated his first art exhibition mirror image at London's National Gallery. And Professor Stephen Pinker, one of the world's leading cognitive scientists, and also one of its most controversial, who's radically rewritten how we view language in the 20th century, currently professor of psychology and director of the Centre for Neuroscience

at Massachusetts Institute for Technology. His books, most recently, How the Mind Works and the Language Instinct, promote his vision of a computational theory of the mind and the innateness of language. Seem Finger, you claim that language is innate. Let's start with a basic building block, like your colleague at MIT, Namski.

How do you prove that it is innate? You have to establish it with many different kinds of evidence because there isn't any single discovery that establishes it completely. For starters, one has to ask how particular languages are learned. Obviously, no individual language can be innate. English isn't an innate. Japanese isn't innate. Nonetheless, in order for the child to acquire English or Japanese, they can't just repeat back sentences like a parrot, they have to analyze it into units that they can then recombine to express brand new ideas.

and I think any attempt to simulate language acquisition, say as a computer program or as a mathematical model, always has to build in at least some assumptions about what are the units in language worth paying attention to, and that's how the child gets off the ground, by listening to speech and looking for things like nouns and verbs and subjects and objects as the patterns worth paying attention to. You're saying language is part of what we have, just as we have something in the brain which makes the muscles work, we have something in the brain which makes language work. That's right. I think there's a, the brain develops circuitry that's optimized for extracting words and grammatical rules from parental speech. I think you also see evidence for it in the universal design of language that the 6,000 languages spoken across the planet aren't just any old computational systems,

but they're built according to a common plan. What's a common plan briefly? That there are words which are arbitrary pairings between a sound and a meaning, which are composed of things like vowels and consonants, which themselves don't have meaning, but which can be ordered in different ways to multiply out the set of words that you can have in a language. That there are rules for combining the words, and the words are categorized into grammatical part of speech categories, such as noun and verb. They have roles that express who did what to whom by being labeled as subjects and objects. There are mechanisms for inflecting words that is modifying them according to their role in the sentence. So this is over 6,000 language. Jonathan Miller, what's your view of the innateness of language?

I can't help quoting someone I met in the United States some years ago, Lila Gleitman, who said that if a visitor were to come to the earth, It or he or she would be struck by the fact that the various languages which we think of as are so separate, so different and so difficult to learn once you're trying to acquire a second language would simply appear to him like dialects of the same language. But as Steve was saying, they have a common design structure, which I think are not so much the result of our having language inside us, but that language is built onto certain fundamental inherited assumptions about the,

the way the world is. The world contains things which act, things which are acted upon, and in a sense language is structured upon certain structural assumptions that we have and we better have about how the world is, otherwise we wouldn't do very well in it. Is there any way of calculating, perhaps this is a silly question that you make this? Is there any way of calculating when it became innate? I mean, at what stage in our evolution it became innate? You can put a late boundary on it, and that's because all of the cultures and ethnic groups in the world have, as far as we know, identical language abilities. They began to split apart, say, 50,000 years ago at the latest, and that's when they stopped interbreeding until just the last couple of hundred years. And so unless you had massive parallel evolution in the different branches of humanity, it's a good bet that the language ability was in place before our species fragmented, before the New Guineaans went to New Guinea and so. on. So that would put 50,000 years as the late boundary. In terms of the early boundary, that's obviously much harder. There have been attempts to reconstruct the evolution of language from the anatomical signs that it left behind in skeletons, things like the diameter of the canal that carries

the nerve to the tongue, the diameter of the canal in the bone that carries the nerve that controls the breathing apparatus, each of which has been modified in the course of evolution, presumably in the service of the control over the breath and the tongue that's necessary for language? I've never been convinced by that particular, by the anatomical argument, because it goes against the idea of the arbitrariness of the sign and because, in fact, when it's constantly struck by the fact that when people are disabled, that they can make themselves understood,

for example, laryngectomy patients can do speech by burping and they can articulate it, and people can speak with severe injuries to the jaw and to the tongue. So it makes one feel that the linguistic ability must be some sort of abstract structure, which can make use of whatever there is, and you can see the extraordinary eloquence of the deaf when using sign, which makes it look as if it's something upstream of the anatomical capabilities of the larynx or the tongue or the jaw.

And I think when people go searching for whether the larynx is high enough up the neck to allow these flows of sound and flows of air. I think they're looking at the wrong thing. So what should they be looking at? Where do you think it does be getting? I don't think you can look in where it's implemented because we don't have the soft tissues of pre-pliceyne brains to look at, nor do we know enough about how brains are organized now

to be able to say how it happens. I mean, we talk about these areas which are not exactly responsible for language, but certainly are vulnerable linguistically, where when, you know, a broker's area, so-called, when that goes, people lose the capacity to utter speech in a certain way, certain areas the temporal lobe go, people lose the capacity to understand language.

Now, that's really all we know. People can sort of cut sections to their heart's content, and we still can't see from the circuitry how it does it. If that's so, Stephen Binker, what argument, what hard argument can you bring to bear against a hypothesis that language began by people reacting to things, grunts of fear turned into words which said run, as it were, I don't remember it's ridiculous, but that sort of thing.

Is there any, given what Jonathan said, which seems to me to be quite convincing, is there any reason why that could not have been the start of it, rather than it being this word innate, which is a beguiling, almost magical word, almost sort of divine word? Well, those two ideas aren't incompatible, because even if it's innate in modern Homo sapiens, it must have had some kind of origin in our evolutionary ancestors, and that origin could have been a conversion of calls,

such as for fear or warning and affiliation and so on. We certainly know now that the calls and equivalence of animal calls that we still have, moaning, sighing, laughing, shouting and pain and so on, are controlled by different parts of the brain than the ones that control, voluntary articulate speech and that attempts to get primates to communicate

have been quite unsuccessful of getting them to control the vocal apparatus so it's reason to at least be puzzled as to what the origin is the obvious thing would be that language would come out of these calls and grunts but they seem to be to have nothing to do with one another at least in the modern human brain

what about the brain Jonathan is there found in the mechanism of the brain something in which this is located which you feel has always been there? No, no. I mean, all that we know is something about the vulnerability of the brain. And I think it was the great pioneer neurologist Jackson, who said that we mustn't confuse vulnerability with the responsibility.

The fact that when you knock something out that certain linguistic capabilities are injured doesn't mean that the part that you've knocked out is responsible for broadcasting those capabilities. It's obviously a massively interconnected network which is vulnerable at certain points. And at the moment we really have the very, very vaguest idea of the actual implementation of it. But that doesn't mean that, therefore, we have to go looking for something else.

It can't be implemented by anything other than brain unless you start to invoke all sorts of weird sort of psychic substances, which they did, of course, at the time of Descartes, you know, thought there was out of spirits. I think that's absolutely. Although it's changing now that we have new technologies for functional neuroimaging. And I think in the next five years there are going to be some interesting discoveries about parts of the brain that we may not have looked at so far based on the literature from brain damage,

but which may turn out to play an important role in language. Well, I think that's certainly true that neural imaging shows that there are more parts involved than we thought before. I mean, the progress is often equated with technical discoveries. But the funny thing is that Chomsky's revolution was achieved without looking at the brain at all. achieved by having a new approach to the subject, in other words, to language. But in effect, a comparable approach to language was already fairly firmly in place by the 17th century, as a result of these strange Jansenists in outside Paris, the Port Royal linguistic philosophers, who were actually proposing that there was, in fact, some sort of native structure to language.

Now, in a sense, what Chomsky did in the 1950s and 60s could have been done by the Port Royal people in the 1670s because both of them had the same material to look at, which was simply how do we talk and what is the structure of language. And I think that although it's interesting to see what's going to happen in the brain, I think that the big advance has been the result of listening more carefully to what we do when we talk to one another. But when we talk, is there a clash, do you see, Jonathan, between culture and biology? I mean, just to chuck in a couple of little things. We know that the young boy in Russia brought up with wolves and the young girl brought up with chickens. They only were able to communicate with the sounds and reflexes they'd heard from these animals.

Does that tell you anything about the culture of passing on language? Well, it tells us what Steve said at the beginning, that people don't inherit Japanese, or English or French, in order to be competent performers of one or other of those, they must be exposed to it. But they wouldn't be competent performers of it unless they had this innate capacity to acquire language very rapidly, much more rapidly than it could possibly be explained by sort of trial and error learning. But that's not, I think, where the controversy arises between culture and biology.

I mean, we know linguistically that this is something. So why do you say the center of that controversy between? culture remote. Well, I think that, and this is where I think, Stephen, I differ to some extent. I think that there has been a sort of vast trade of evolutionary psychology, which I think has overstressed the extent to which we are, in fact, creatures of our heredity, rather than creations of our culture as well. Stephen Finker. I would see our discussion of language as being a model for how to look at the interaction

of biology and culture elsewhere, namely that what biology gives us is an ability to learn in certain ways to pay attention to certain aspects of the environment and ignore others, to have certain goals to analyze the world in certain categories. And just as we don't inherit Japanese, but we inherit an ability to acquire a range of languages that includes Japanese, we might inherit a way of parsing the cultural world to pay attention to certain aspects of behavior, to be able to learn certain things easily and certain things, not at all. So I don't even like thinking of biology and culture as two complementary ingredients because they're really different categories.

Culture is one of the things that happens when you have an innate learning ability in a species and then you let the members of the species interact and bump into each other and culture is what emerges as a result of that interaction. No, but I think that there has been a frightful disparagement on the part of people who have perhaps drunk too deeply at the wells of neurobiology.

A disparagement of what, for example, for one couple of authors, Tubby and Cosmedes, have attacked as the standard social science research model, indicating that social studies and sociology are sort of derelict enterprises, and that if only sociologists were to direct their attention towards, for example, game theory and to genetics and notions of inclusive fitness and so forth,

we'd have the human sciences down to a T, and I think it's a Philistine approach to the extraordinary varieties of culture, Of course they co-evolve in some way, but I don't think that culture is on this sort of short leash of hereditary compulsion, which some people imply. It leaves the study of history in a state of derelict shame, really. I just don't think it's right. Stephen Binker. Yeah, I think it differently.

The leash metaphor isn't particularly appealing. I think it more as the biology. provides the elements and the rules of combination. Perhaps it's the rules of chess and culture as a particular game of chess. And that it's not so much that social sciences are derelict enterprises, but just that they shouldn't float free of the rest of knowledge, in particular knowledge of science, that ultimately they should be connected to the natural sciences via an understanding of human nature and the innate learning abilities that give rise to culture when you watch people interact over large spans of time.

I've had a sneak preview of your forthcoming books. You make a words and rules, the ingredients of language, and you talk about the combinatorial system of language. That was first put forward in the Enlightenment. Can you tell us why you think that is so extraordinary? The extraordinary thing about language is the vast range of thoughts that we can communicate.

It's not just that we have a list of a dozen messages that we can bark at each other, but we can talk to each other about theories of the origin of the universe or the latest twists and turns in the Monica Lewinsky scandal or the football scores or soap opera plots

or cosmological systems. What's the secret behind our trick to convey so many different kinds of ideas? And I think the secret is combinatorics that what language allows us to do is take a fixed stock of ideas, the ones that we have words for, but then to combine them in

phrases and sentences in which the meaning of the sentence can be computed from the meanings of the words and the way that they're arranged. Because the number of messages grows geometrically or exponentially with the length of a sentence, it means that you can get truly enormous numbers of thoughts that can be expressed, even with a finite set of tools, as long as the tools allow for combining the nouns and verbs into bigger strings. There are lunatic statistics, either, that in 20-word sentences, you can have 100 million trillion variations of a 20-year-old.

word sentence, and this is more than the double the second since the universe began, something idiotic like that. Yes, yes, many orders of magnitude more. And it's just simply because of the mathematics of combinatorics, you can get mind-boggling numbers. I mean, you can't get your mind around it,

but where does it lead us? It's such a... Why is that fastness, that infinity? How do you not account for that? What does it say to you? It reflects back on the thoughts that we put into language. reason that we have this mechanism for converting thoughts into so many different noises is that we

have so many different thoughts to begin with. And I think that the, perhaps even the essence of human intelligence is to be able to imagine brand new scenarios, to put things together that other members of the species haven't put together before. Language is just the medium by which we share those combinatorial thoughts. So it's a way of taking combinatorial thoughts and converting them into combinatorial noises so that we can get them from one head into another. But it also creates the possibility of having thoughts which were not actually conceivable before,

not just simply expressible. It may well be that, I think Forster once said, I don't quite know what I think until I hear myself say it. Now, there is a sense in which new thoughts get created as a result of utterances of an unprecedented form, which goes back to really what I was arguing earlier, that if in fact there is this

vast combinatorial possibility, I think von Humboldt pointed it, out the possibility of using limited resources to express an infinite series of thoughts. That was already in place at the beginning of the 19th century. In a way, that reinforces what I was saying about the autonomy of culture, that if, in fact, using this biologically inherited device of a vastly combinatorial system, which allows us to express infinitely more than was previously thought to be possible, then it does start to float free of its biological basis.

It still is, it's anchored in it, but nevertheless it becomes possible to create a cultural artifact which is entirely within the medium of language, which actually transcends what was previously thought to be possible on the basis of its biology. Is this actually bringing in the imagination? No, the imagination is a sort of loose and rather unfocused concept, which perhaps includes some of these things. but I think that it indicates the autonomy of what in fact begins to appear in language. There are different ways of I guess looking at the same phenomenon because we agree with what the phenomenon is.

It is autonomous in the sense that any particular idea can be thought for the first time and communicated for the first time, although the elements and rules of combination that allow us to have those thoughts are in a sense specified by the mental apparatus that we inherit. And indeed, even an infinite combinatorial apparatus doesn't cover everything,

just as the integers are infinite, but they don't include the fractions. An infinite combinatorial scheme for having thoughts also leaves possible there's certain kinds of thoughts that we can't think or have great difficulty thinking, even though the ones that we can think are, as you point out, open-ended and infinitely creative. But it is interesting how thoughts which perhaps at the time seemed unthinkable. I mean, one knows, for example, that Newton virtually bust his... head trying to think of the mathematical concepts which enabled him then to think to the point where he could create the Principia.

Now, a pretty smart first year of physics students finds the Principia, which was almost unreadable by more than two or three of Newton's contemporaries, he finds it perfectly easy to understand it. So this creates this vast, autonomous mathematical culture. And I think that language also creates this enormous autonomous, autonomous. culture in which we speak to one another. Although, of course, it's rooted in biology. How else can it be rooted?

What part does philosophy play in this? A lot of philosophers believe that philosophy is just the advancing edge of science, and that the problems that the scientists think aren't worth thinking about yet are the ones that the philosophers worry about. As soon as they start to get answered, the philosophers get bored with them, and they slough them off to the sciences. There are also, though, some problems

that traditional philosophical problems that don't seem to be yielding to science and these might even be examples of the unthinkable thoughts that our combinatorial apparatus for cognition leaves out. And in how the mind works, I reiterate an argument from Colin McGinn

that some of the mysteries that haven't become less mysterious over time, the ones where you and I could have a debate with Aristotle and we would be speaking in the same terms, maybe cases where the very nature of thought doesn't allow us to see a problem from the right perspective. It's thought that perhaps the problem of sentient, subjective experience,

why does neural firings give rise to something that I actually experience as redness and pain and saltiness? Could be a kind of problem where limitations of our own brain never allow us to be completely satisfied with any answer. We know that it must be the physiology of the brain that gives rise to the sensation of saltiness. But what bridges that gap between the physical and the subjective may be something that we don't have the mental categories to appreciate. And yet you see there are these sort of radical eliminativeists, both east and west coast in the United States, who think that they're really the consciousness and redness and saltiness and fear and so forth

are epiphenomena. And what's really going on is brainwriting. And once we get brainwriting down, we'll have got it. and then you wonder what they do when they're off duty, these people, when they're getting redness and saltiness and fear and lust and so forth. But I agree with Colin McGinn and, in fact, and others, that there may be some sort of fundamental mystery about how it is that grey salty porridge by firing electrically

can make the owner of the grey salty porridge see red. What role do you think the unconscious plays in language, Stephen Binker? Well, most of language processing is unconscious, None of us has any idea what the rules are that allow us to string words together in well-formed sentences to make ourselves understood. So I think with most cognitive processes, what we experience subjectively is a tip of the iceberg. And it's an interesting scientific question why certain aspects of the process should be things that we can talk about and ruminate about and reflect back on while the rest goes on beneath the surface. It needn't be Freud's answer that what goes on beneath the surface is socially unacceptable or too painful to witness. There may be good engineering reasons why you want to have some subset of the information processing in the brain

mutually available so the language system can talk about some aspects of the perceptual system, and the decision-making system has access to the results of those computations, where the rest you farm out to autonomous processes that can just spin away in their black box, and not bother these other parts of the brain. Yeah, I think that the popularization of the Freudian unconscious has actually done almost irreparable harm to the more interesting unconscious with which people like Chomsky deal and others, you see.

And the funny thing is that whereas Freud's notion of the unconscious is a custodial, repressive one, in which you keep things under lock and key because you don't want to hear them talked about and because they're dangerous and subverses and disruptive to society. But actually, long before the Freudian unconscious was developed, a much more interesting what I would describe as an enabling unconscious, the unconsciousness that allows us to do things like speaking

or sleeping on a problem and coming up with it next morning, remembering things surprisingly after a journey, suddenly remember me. Oh, yes, I had seen that. I had seen that. There's a vast amount of stuff. This notion of the enabling unconscious was already in place by the 1870s. There were two scientists in England, William Benjamin Carpenter at University College and Thomas Laycock, who had actually already formulated a notion of what was called then, either the reflex functions of the brain, or a wonderful clumsy term called unconscious

cerebration. So that the idea of an enabling unconscious was fully in place until the Freudian unconscious, with all its sort of naughty connotations of Freudian slips and naughty thoughts, replaced it in the public imagination. And actually, if it hadn't, I think, been for behaviourism in the 1920s, the enabling unconscious, which was already formulated by the end of the 19th century, would have actually brought the cognitive revolution into existence long before it did. As it was, we had to endure this ghastly drought of behaviourism. We're watching rats running mazes and giving monkeys rewards and punishments

and stop people thinking cognitively. Finally, can I ask you, Stephen Pinker, with street language and street slang and computer language and different scientific languages growing, do you see more and more languages? We talk about the disappearance of languages as the disappearance of certain tribes and cultures. Do you see more and more languages evolving, more complex and more different? Certainly there's within a language such as English, we mourn the loss of certain words, the distinction between lie and lay, the distinction between, infer and apply and disinterested and uninterested, but new words are being bred all the time to maintain the richness and the diversity.

Unfortunately, it is something that will be preserved in English, which won't replace the vast number of languages in New Guinea and Russia and Alaska and Africa and so on that will be lost. So there is an increase in diversity within a language, but a loss of different languages. When we lose those languages, finally, finally, in those places. Are we losing something valuable, or are we losing, as Jonathan said earlier in the program, a dialect?

No, I think we're losing something very precious. It's like losing a species that each language is an astonishingly rich and beautiful system that captures something about the culture of the people that speaks them, and it's a great tragedy for these species. I did say there were dialects only to Martians. Yes, and I slapped myself on the knuckles for saying only a dialect as I spent. quite a bit of time trying to preserve the dialect of the north-west of England, the extreme north-west of England, which is a great loss of civilisation. Thank you very much Stephen Pinker and Jonathan Miller, and thank you for listening. We hope you've enjoyed this Radio 4 podcast. You can find hundreds of other programmes about history, science and philosophy at BBC.com.com.uk forward slash radio 4.