Best of Series: On Human Nature and Human Progress With Noam Chomsky

and creativity. I'm doctor Scott Barry Kaufman, and in each episode I have a conversation with a guest who will stimulate your mind and give you a greater understanding of yourself, others, and.

Chomsky is also known for helping initiate and sustain the cognitive revolution. He's Laureate Professor of Linguistics the University of Arizona, and Institute Professor Emeritus at MIT. Thanks so much for meking time to chat with me today.

as well. And I was wondering if you could we could trace a little bit, you know, in the sixties fifties, how did your work, you know, when linguistics intersect with the other work going on during that cognitive revolution at the time, the work on decision making and Marvin Minsky's work, How did all this stuff call ask that that epic? Can you take me back to the fifties sixties right now.

what we're now doing. This core problem of the cognitive sciences. What's the nature of the capacity, how can an individual acquire it during their lifetimes, in fact, during their early childhood, and ultimately how could such a system evolve? Those are the core problems of the study of linguistics beginning around nineteen fifty. It's a pretty sharp break from structural linguistics, which had quite different goals and aspirations. I've actually written

some about these years. If you want me to send you something, that'd be great, Yeah, just I have your email, so to do that'd.

of years ago. Later became his book on language, and the drafts were very widely read, and it was influential, in part because of Coin's advocacy, and part it just did fit the tenor of the times very well. So it was kind of a bible when I got there. You look at say, George Miller's first early books. Out

of the later ones, they're pretty strict behaviors. It was even you know, some of his early experiments were considered rather shocking and sounds kind of obvious, showing that you could understand a word better if it was in a sentence, then if it was in isolation, which shouldn't happen on what should happen on rigid behavior, as Browns is that you're the first word of a sentence, then there's a certain probability for the next one, and kind of like

what's done with deep learning today. And by the time you got to the end of the sentence, you can barely guess what the word is because the probably is go down. But of course the results are exactly the opposite. As you hear the sentence, you can just the last work. None of this works, But that was some of his early work and it was considered quite surprising. By the mid fifties, George had significantly changed and became one of

the founders of cognitive science. But when I got there in fifty one, this was orthodox with a couple other things that happened. Claude Shannon had come along with information theory. The Shannon and Weaver book, with Weaver's kind of popularization and extension of the technical ideas, was another bible. Cybernetics was another. Signal detection radio engineering was another, and they all kind of converged into a four area of which Scinierian behaviorism was a central part a sense that we're

cracking the last frontiers. When Krick and Watson came along fifty three, that enhanced the idea that we're now moving to a new era. It was called unified science, and which we'd be able to we had the tools to deal with the problems that were called problems of mind and psychology. There were a couple of us who thought this was all nonsense. Three in fact, to three grad students, my friend Mars Halle, who I worked with till the end of his life, and Eric Lenneberg, who also a

grad student. He went on in later years to found modern biology of language. But the three of us just didn't believe any of it. We thought it made no sense.

We began reading European comparative pathology, Tinberg and Lawrence others looked at comparative psychology work, and I was introduced by a friend, Mar Shapiro, and art historian suggests that I should read Carl A. Ashley's Serial Order and Behavior, which was a very important article back in around nineteen fifty, which just knocked the props out of the whole behavior system.

Nobody knew it. I mean, neurologists knew it. It was in the neuroscience literature, but psychologists and I think and others were totally unaware of it when I wrote about it for the first time in a review of Skinner's or Behavior the Press nineteen fifty seven, But that, I think was the first mention of it in the general psychology cognitive science literature, so that these were things that we were Eric went on to start writing articles on He was doing a good deal of work at the

time on various a variant forms of linguistic behavior, studying pathologies, early studies on use of SNIGN, cognitive deficiencies in language, and neural deficiencies, and this went on to become his Biology of Language book, very very important book, Biological Foundations of Language. But at first it was essentially the three of us. Then George Miller interested link, linked up with Jerry Brunner, who was injured, formed the Cognitive Science Group

at Harvard, spent a year there. I was working quite a lot with George Miller in the mid fifties, books, a couple of articles together, taught together, and so on, and then it just kind of expanded, linked up to some extent with the early work and artificial intelligence Simon Simon and Newell, har Minsky and McCarthy, and by the early sixties it become kind of a fairly i can't say integrated because there are a lot of internal disputes,

but interlinked approach to many questions the euphoria, but then had pretty much anticipated you get it. You can get a good sense of it by reading theoshu of Barhelil's essays around nineteen sixty five. He's an Israeli, a logician who was a regular visitor to Pralely Research Level Electronics MIT, which was kind of the center of most of it.

And at first he was very much a partisan of the euphoric hopes, but by the mid sixties he was a close personal friend also back from about nineteen fifty, but he had pretty much come around to agree with the skeptical approach and wrote some retrospectives about it which are quite interesting and knowledgeable around the mid sixties. You can get a sense from his work that's kind of what it was like in the early days in Cambridge.

There was no linguistics. Practically the only linguist, the romaniocopsin was there. But from the European tradition, but there were no American linguists, so I was, in fact about the only person there was an American linguistics tradition factor consensus, but it was quite different in character. It was it was described its proponents described as a taxonomic science set of procedures, but you could apply to any purpose of data.

It would identify the elements and their distributional arrangement. That was linguistics. That's basically my own background. This was quite different.

couple of others got interested in it. By by the mid fifties, we felt that we were pursuing something that should displace the I think it was the consensus, and as I say, none of us agreed with the None of us means three or three of us with the behaviors dogmas that were rampant at the time, but never the question wasn't is it a revolution? But is this

the right way to pursue things. Gradually I learned that there were ant of students, I say, we were reading European ethology, which had some similarities, and I began looking into the history a little further. Learned that there were much earlier tradition that had been totally forgotten some of that. So my own view in retrospect that you might call this the second cognitive revolution. Turns out, in the seventeenth century, the time of the origin of modern science, a lot

of things of this kind were happening. They didn't have what they lacked, and what we had was the poria of computation didn't come along till mid twentieth century, so there were no tools for trying to explore the kinds of questions that they were raising. But I look back to Galileo as contemporary as they recognized something very significant wasn't discovered later until I wrote about it in the

mid sixties, and almost nobody knows it now. In fact, if you look back, the early founders of modern science like Galileo, were just amazed by what they regarded was the most astonishing, remarkable fact in the world, that with a few dozen symbols we could construct in our minds infinitely many thoughts and even convey to others who have no access to our minds the innermost workings of our minds.

The Galileo, for example, alphabet was the most stupendous of human inventions because it enabled this miracle to take place. Of course, it was understood that the alphabet was just representing some system in the mind which does the same thing, and that didn't develop. The tradition of what was called general and rational grammar through the eighteenth seventeenth eighteenth centuries, which tried to develop these ideas, but lacking the theory

of computation, there was no way to formulate it. How do you formulate a computational process that captures these capacities.

By the mid twentieth century, when I was a student, I was studying logic and recursive function theory, and you could see that that offered that modern recursive function theory and theory of computation provided the tools in which you could proceed to develop computational systems which a recursive enumeration of the expressions of the language basically expressions of thought, and you could find also provided means by which this could be translated into the map into sensor remote motor

outputs and inputs, so you could link it to perception production, learnability, How can it be acquired? All these questions on the agenda as soon as this fell together.

I was wondering how your thoughts on this have evolved over the years, especially in light of Arthur Reeber's work, for instance, on implicit learning, showing that artificial grammars can be learned that we are not sort of hardwired. I'm not a big fan of that phrase. But you know what I mean, hardwired to have that specific rule structure. But we perhaps maybe like through to statistical learning, we can learn languages, and so there's a lot less built

in than maybe we once thought. And I'm wondering how you're current thinking of that is.

Street Journal that's of zero scientific interest. If you did a ton of statistical analysis of chemical experiments that say, you know, people mixing things in the laboratory and so on, you could get a fair approximation to what they're actually doing. Would I tell you anything about chemistry. Nothing. It's a game. It's a way of selling a propaganda for IBM, you know, or for Google these days. But it tells you absolutely nothing about the nature of the system and the way

it's acquired and learned that human beings don't. A two or three year old child has mastered the basis of the language. We now know from statistical analysis the data available to each child that it's extremely empoverished. I mean, sounds like the child is hearing millions of sentences. But when you take into account such elementary facts as Zip's law, you know this the rank frequency distribution, and it turns out that almost everything you barely in hearing bigrams, let

alone trigrams. So from the what's back in this is what's there is a problem that was understood in the early fifties called poverty of stimulus. How do you go from the impoverished stimuli available to the rich knowledge attained? It was obvious in the early fifties is a major problem.

By now we know it's a much worse problem, far worse than was assumed, because because by now we have extensive studies, first of all of the data available, we don't have to guess anymore, and extensive studies of acquisition of language. None of that existed at a time. And it turns out that at about the earliest age you can start testing that kids already have very rich knowledge, which goes way beyond what they produce. This is work that's going on for fifteen sixty years. So the problem

of poverty stimulus is overwhelming. Now the deep learning approaches have no problem. That they have basically as much data as you want, vast amounts of data, huge amounts of computing capacity, and with that what they get is kind of what you get if you did what I just suggested, looked, looked at or takes an example closer to the language. Suppose that you did deep learning studies of the communication system of bees. Okay, you could get a fair prediction,

probably pretty good predictions of what bees do. You know, bees start going out of the hive, they wander around, they find a flower, they right, they have the capacity of dead reckonings, or they go straight back to the hive. They waggle their wings and the other bees start fluttering around when they go to the flower. But they'll tell you anything about bee communication. Nothing. You know, it's not what when Frish was doing. It's not what other bee

scientists do. They want to find that it works. The fact that you can kind of approximate to phenomena just tells you another thing. You can look at the phenomena without a activating.

their conscious awareness. And I found that people are some people are better at it than others, but overall, you know, most people. I debriefed them afterwards, I said, did he notice that fifteen percent of the time this sequence it followed this sequence, eighty five percent of the time it followed that sequence, and that you got faster at the reaction time for the eighty five percent. And they had no idea, you know, consciously that they learned these things.

And these were artificial you know, languages, These are not, you know, something that we evolved. And so do you think that's telling us anything about how much is build in versus how much is not building? Because I think it might. I think it does tell us something tells.

lot of biagrams, let alone trigrams. And this is not there because what the kid is hearing mostly is the function words, the end of and so on. And then you look at the rank frequency distribution, it goes very sharply down and tails often to a long tail. So I mean there's some statistical learning, but very much at the margins. So there's nothing wrong with the experiments they're studying.

They're interesting things to study of cognition, and of course everything goes on unconsciously, like you and I have absolutely no awareness of the rules that we're following in this conversation, way beyond the levels of consciousness. I mean, people are deluded to think that what we call inner speech is somehow our thinking processes. Absolutely not. What we call inner speech is a pale reflection of the externalized form of

what's going on in our minds. And if you actually think about it, it's just bits and pieces of fragments. Our construction of sentences in our minds is vastly more equipped than what we call in our speech, and you can see it very simply by just doing things like reading out loud, which is much slower than reading. You can read a page in a fraction at the time and understand it of reading it all because the externalization is just very much slowing everything down. There's all kind

of stuff going on in our minds. We can study it the way we study becommunication from the outside that can't introspect into it totally beyond the level of consciousness.

I mean, even just figuring out how you get protein folding you can barely do by Now that's an area where AI has made some contributions, but that's remember this the bare beginning of what makes an organism. It's good to study these things, but you should study simple traits, not a very complex trade life language which is going to be way out of sight for any such investigations.

a second look. Maybe there's some maybe there's some sparks here that makes him do creative work and put aside things he's not interested in. So it's something, but I don't think it's much of a criterion. Of course, if somebody scores very low in everything, you think, well, probably not qualified.

of exciting work. Others just did routine technical work, you know, perfectly competent, but like it adds a brick here and there to which you know, if you look back, it's

pretty hard to find the distinguishing characteristics. I was in the Society of Fellows at Harvard, which is a very highly selective of three or four year research graduate fellowship, no duties or responsibilities, and that they tried to have very high standards for selection and occasionally look back and try to do the same judgments and the same story. You can't tell. These are really matters of character in

many ways, more than technical more than mental ability. I mean, they're going to have somebody who's a you know, a math genius, but will never discover anything. This just doesn't have the right right characteristics.

How do you define creativity? Like, how do you even conceptualize that such a word?

theoretical systems. But creativity, we is the ability to First of all, it's the ability to be puzzled by things. You have to start with that. There are infants have that capacity. They're puzzled by everything. They're constantly asking questions that can't stop. Is annoying, you know how what's this? How does that work? You know, if you're a parent, drive you bananas. They want to the world's very puzzling, strange.

They want to understand it. That's driven out of people's heads in many ways worked by the educational system, part of another way. But there are some people who retain it. And in fact, if you look at the great moments of history of science, that's pretty much what they were. So take say, the scientific Revolution seventeenth century, the great science that really changed science practically, it basically was based on being puzzled. Galileo's contemporaries were just dissatisfied with the

what we're called explanations of things. So why do alls fall to the ground, Well, it was an explanation, that's their natural place. They're attracted to the earth. They didn't regard that as an explanation. As soon as they began looking at it, they found that we don't really understand and in fact, all that was believed turns out to

be wrong. Like it was versely taken for granted that a heavy lid ball full faster than a small lid ball until they showed by thought experiments incidentally, never carried any experiment by thought experiments, ingenious thought experiments. Should this can't be the case. And that's what happens when you're puzzled. Appillion people are willing to question orthodoxy, not as to accept it because that's what everybody says, but to be to want an argument for it, to challenge it. Sometimes

it's right, you get convinced. Sometimes you see, it's just dogma. Behaviorism was like that, it was just dogma. It's soon as it began to look at it carefully, totally fell apart. But it's hard to do that. There were people who did like lastly, for example, and who were ignored. But if enough people do it begin to interact. Uh, they start to try to work out real answers. They keep finding flaws in their own answers, address those. Then you get creative work. But how to pick it out among people?

want them to win. And if you ask, as I did, ask, if you look back at the they suppose it turns out that what the proletariat will institute is inhuman, destructive, cruel and so on, does that mean we oppose it? He said, no, it's just a matter of which side you're on, there's no right or wrong. There's no questions of justice, there's no basic human nature, just which power system it takes control. That may be a caricature, I'm not sure, but that's what I understood, is position to be.

the origins of classical liberalism just won't been forgotten. It's the reason, for example, why working people in the nineteenth century early Industrial Revolution were bitterly opposed to the wage labor regarded as a fundamental assault on human dignity and human rights. In fact, that was a very standard position at the time because, in fact, the slogan of the Republican Party that wage slavery, wage labor, what they called wage slavery, is no different from slavery, except that it's

temporary on the program Lincoln. And in fact, that's the whole classical liberal tradition. There's way back to the Greeks and the Romans. That's a modern idea that having a job is a good thing. I think that the tradition is probably right that people don't want to be subject to masters. And I don't care whether the masters are the Central Committee or the the corporate sector about the same thing. But that's a guess about human nature. I can't demonstrate it today.

are big fans of foco. You know, do you ever, do you have any criticize, criticisms of maybe some of the methods used in modern day social justice movements that have moved us away from the classical liberal sort of ideal.

dreamt of. Like Stalin couldn't tell you that at three o'clock in the afternoon, you have five minutes to go to the bathroom, and you got to wear these clothes and on some other clothes, and this is the past you have to take when you're moving from here to there. I mean people that lit alone on an assembly line, which is kind of control that no dictator could even dream of. Or a person working at a cash register just totally turned totally into an automaton under total control of

an authority. I mean, that's an extreme form of authoritarian control. We know. One of those dogmas question kind of like behaviorism in nineteen fifty, is that this is a good thing. As I say, in the early Industrial Revolution, this was regarded as totally intolerable, built on basic human rights. Now it's considered the highest thing in life. I can get a job flipping Hamburg. That's it. I think that's so

what you're saying about. No, there are a lot of people like Fucoh and many others who say there's no human nature. That's first of all, it's insanity. It's like saying we're no different from cats. I mean, it's insane. You know, of course, the human nature. What they mean is something different. They mean. What they probably mean, which makes sense, is that the particular social forms and arrangements

in which we are integrated are subject to change. That's correct, that seems reasonable, But it's not going to turn us into insects. It's not going to give us an insect visual system. It's not going to turn us into creatures that are incapable of language. I mean, we have a nature, and in fact it's very rigid and strict. Within it are variations and social and cultural and other arrangements that lead to variations. I mean, take say the visual system.

Since we just think of the classical experimental work Hubil and Weasel, for example on human on vision a male en vision, I mean, what they show is that early, very early modification of the visual system, story cortex can lead to radical changes in the phenotype and the outcome that appears. It doesn't happen with us because we all

have about the same visual experience. But if there was a mangle around who could stick electrodes into our visual cortex or control the stimuli that we see, we'd have totally different visual systems that the as adults maybe no visual system a cat. As they showed, if you deprive a kitten of structured visual stimuli for the first couple of weeks of life, it's essentially blind. The analytics systems just degenerate. Okay, that means and it's kind of the

same as speaking different languages. But the point is you can't change an a mammalian visual system into an insect visual system with compound eyes. That's a mammalian nature. Within mammalian nature, there's a lot of possible options. Within human nature, there's a lot of possible options. And it's the options

that interest us as human beings. We take for granted what's common to us, kind of like sports, Like when you go to the Olympics, you don't see a competition and walking across the room with anybody to do that. You see competitions and things that humans are no good at all, like pole holding. So if you go way to the edges of human competence, you start to find differences among people, but the overwhelming mass of competence is just shared.

biological sex binary you know, what do you?

you're interested in the differences. They don't care about the fact that we're all frogs that you just take for granted. Yeah, interested in what's a little bit different between this frog and that frog. But from the point of of us looking at frogs, it's so minor you can't even see it. It's the same with this what pronoun you use. From the point of view, the nature of humans is so marginally you need a microscope to detect it. But for

human life, that's what matters. We don't care. We don't even pay attention to the fact that we're all fundamentally humans, that we just take for granted. What's interesting to us in our lives is these very slight differences around the edges. Do you say everyone thought he was here, or do you say everyone thought they were here or something? It's worth thinking about for human life, and it's important. So for example, I wouldn't like it if people called me

a kit. Let's say, I mean, I don't think there should be a law against it, but I certainly wouldn't like it.

around nineteen thirty was thirty two, early thirties. When I was a kid that was a it was Fortune magazine had a front page cover saying the wasps are unwhopping themselves, meaning Mussolini's doing a good job. So the wops are unwhopping themselves. I don't think I can say that now, I don't think so.

or not. But just like a physicist isn't going to tell me how to adjust the books. These are parts of human life which are way beyond the sciences of their comprehension. But yes, it's a it's a topic you have to be concerned about. And I think you have to have to ask yourself seriously, should we burn down the city of Washington because George Washington was a slave owner? Okay, I'm sure you can find some people to say we should.

I don't think so. These are the judgments you have to make as a human being living in a complex society.

And I was wondering your some of your thoughts on that now five years later.

kind of white supremacy is very wide spread. But on the other hand, there is a change in consciousness in many circles. You can see it in the reaction to the George Floyd murder. Blacks have been murdered for a long time. This reaction was quite different, spontaneous, enormous in scale, way beyond anything in American history, and have widely supported at around sixty percent support, a lot of solidarity, sensible goals. It was a very striking phenomenon, led by Black Lives

Matter organizers, but many others coming in. Well, that's one sign of serious changes. I want to see another sign take a look at this morning's New York Times as a very good op ed by Eric Kohner. Fine historian calling for abolition, and what he's calling for is abolition of criminal labor, mostly black. He's a historian of abolitionism, reconstruction, Civil War, and so he points out that the Thirteenth Amendment banning slavery had a qualificatient it said forced labor

is legitimate as a punishment. That, as he discusses, was the opening that the South used to say. And he says, we still have it today in private prisons where convicts are forced to work for ridiculously low wages of profit. So we should really move on to abolition. You wouldn't have had that up it a couple of years ago. It's a sign of the increasing consciousness and awareness of the really black history that is our actual dark, cruel history,

its history that's part of our legacy. Now we've never grown out of it. Part of it the extermination of the indigenous populations and other part of it. And these are things that are gradually seeping into consciousness, not anywhere near enough. So there are holocaust museums all over the country. Try to find out how many museums there are for slavery of nation of Indians.

in human history. That's that's a pretty big that's a pretty big deal. Why more so than like Hitler?

very confident that will have passed irreversible tipping points. Practically one hundred percent of climate scientists agree on this.

what they're going to do. And I think, you know, i'd love to get your thoughts how we can get the Democrats to actually focus on reducing economic inequality and meaningful ways so they can bring on board and convince the poor and uneducated people to stop voting for you know, the next Trump, you know, or even Trump four years from now. You know what, what can what can we do as Democrats?

So what can the Democrats do? Well, that's a very pertinent question. In fact, we've just seen what happens when they don't do anything. So take the November election. There's been a lot of There are areas of the country that voted for Trump that haven't voted for a Republican for one hundred years, and there's been a good deal of discussion of it. Like South Texas, Mexican American area,

it's an oil economy, hadn't voted Republican since Harting. A lot of them voted for trum even some counties went for him. The reason what they heard from their sources, Fox News, the White House, whatever they get their information from. What they heard is Biden wants to take away your job, destroy your community, destroy your businesses, devastate your families. Don't vote for him. Vote for Trump, who says I'm going to keep your jobs, keep your families, keep the oil

production working, and so on. That's what they heard. Now, if the Democrats were a political party that had any concern for the general population instead of being committed to Wall Street and rich donors and the wealthy professional classes, if that were the case, they'd have had down in South Texas saying, look, it's a fact that we're going to have to get off the oil based economy. There's just no debating that. That's like debating the weather. It's a fact. We have to face it. Now here's the

way we can face it. We can face it with feasible, sustainable measures which will give you better jobs, better lives, better communities, happier existence. Here's the way to do it. But they didn't do that. No, okay, so there's the answer your question. You don't you don't break through the propaganda. Yeah, that's what's going to happen.

So it's possible to take important issues and to destroy them, okay, by just not handling them like what I said before, somebody came along and said, let's burn down Washington. That's you know, maybe you can make an argument for it. That's not the way to deal with the world, not a sensible way.

mean by defund the police? And her answer was, just go into any white suburb. That's what we mean by defund the police. The kids using drugs, you know, pick them up, beat them up, take them to jail, and incarcerat them. You deal with it sensibly through community services. You have the police when you need them, but they're not supposed to be involved in community service operations of family disputes, overdoses, mental health issues, because they're not these issues.

That's community service. I think the police are in favor of that kind of defunded. Bernie Sanders did the same. He said, yeah, that's what we want. We want police freed from obligations that are none of their business and are just a burden. And in fact, they're probably ninety percent of what they do. And let's fund let's have better pay for police, so it's more desirable, profession people,

better trained, and have them focused on police work. But the things that should be done like they're done in a white suburb, they should be done everywhere. I lived in a suburb of Boston. Just give you an example. We were away for the summer and some neighbors called us up and said, somebody broke windows in your house. So we came back to see what happened. It turned out somebody had broken in, so we called the Lexington police and they came over and the first thing they

said is look in your medicine cabinet. We looked in our medicine cabinet and yeah, somebody had gone through it and taken some stuff in. And they said, is this kids in the neighborhood. We know who they are. We're not going to send them to jail. It's up to their parents and the community to discipline them and put them on the right path. That's the way to deal with it. I didn't want them to go to jail. I suppose it was a black community. You know what happened.

So when Kassio Cortez says defund the police means make it look like a white server, that's a message that makes.

much when I believed as a young teenager. I mean, there are individual there are individual things that are wrong, like text day of the Vietnam War, which I devoted a lot of my life to. I started in the early sixties when it was very unpopular, but that was much too late. It started ten years earlier, but nobody even heard of it. By the early sixties. That already getting out of control. There are things like that.

it's really up to you. You have to create your own life.