Column: Noam Chomsky Walks into a Bar

On language: It's great for talking to yourself, but not so much for talking to others

[Editor's note: David Erik Nelson writes a monthly column for The Ann Arbor Chronicle called "In it for the Money."

David Erik Nelson, Noam Chomsky

At the end of the interview.
DEN: Do you mind if I take a picture of us?
NC: Sure. I’ll put my glasses back on so people will know it’s me.
DEN: You actually look remarkably like yourself. [Chomsky laughs.]

Instead of his regular column, this month we’re publishing a piece Nelson wrote based on an interview he conducted with Noam Chomsky a few weeks ago, when Chomsky visited Ann Arbor. The piece includes long chunks of transcript, interspersed with commentary from Nelson. It begins with Nelson, whose thoughts are presented in italics throughout.]

I’m interviewing Noam Chomsky in the bar of the Campus Inn a block from the University of Michigan campus in Ann Arbor, Michigan. The bar is dim and entirely abandoned at 10 a.m. on a Friday morning. Because I’m highly distractible, I can’t help but periodically marvel at the symmetry of this: I only ended up interviewing Noam Chomsky at all because I’d Tweeted a link to a joke about Heisenberg, Gödel, and Chomsky walking into a bar [1], and Dave Askins (editor of this fine publication) had responded by noting that Chomsky would be speaking at the University of Michigan a week or so later, and essentially dared me to interview him.

I’d agreed, on the assumption that it would be impossible to land an interview with the man almost universally regarded as America’s foremost public intellectual. I was wrong – and Chomsky chose the bar as our quiet little nook! It was almost too good to be true: Noam Chomsky walks into a bar and … and …

And lemme tell you, there is more than a little pressure inherent in being the straight man in a classic joke set-up, even if the set-up is only in your head – which is germane, since from Chomsky’s perspective, it’s the conversation in your head that is most essential to the nature of language.

David Erik Nelson: Did you primarily come for the linguistics piece [as part of the 2013 Linguistic Institute held at the University of Michigan this year], or to speak to the union?

Noam Chomsky: I was first invited by the linguists. That goes way back. I have pretty close connections with the linguistics department here.

DEN: Um, so, the substance of that talk … I was looking at the abstract online, where you mention this traditional view that the purpose of language is to construct thoughts, and the more contemporary view that its principal use is communication. Um … I was hoping that you could elucidate that a little further.

NC: By now it’s a kind of a dogma that the function of language is primarily communication. There’s no support for it, and there’s actually pretty strong evidence against it, and I ran through some of the evidence yesterday [during the lecture delivered at Rackham].

The most important evidence is actually just when you take a look at the actual nature of language, and I think what you find is that – well, I mean, you have to really look closely, but – you have to find that there are conflicts between what’s called “computational efficiency” – getting the best possible system, which is the way a language would develop, just in the mind, it seeks the best possible system – there are conflicts between that and “communicative efficiency,” what’s good for communication. And, when you investigate those conflicts, you find that invariably, as far as is known, computational efficiency wins out and communication is simply ignored.

The most striking example, and the one that I did talk about a little, had to do with just the linear order of words. That’s essential for communication, of course, but it doesn’t seem to even be part of the structure of the language. It seems to be something tacked on on the outside just to meet the conditions of the sensory-motor system. But it’s not part of the language fundamentally.

The Campus Inn bar – which I previously had no idea existed – is appointed in dark wood, ancient leather, and glowing bottles of liquor I can’t afford. It is the sort of bar where crooked politicians in Hollywood thrillers scheme. As we sit across from each other in this dim little rich-man’s haven, I’m acutely aware of the fact that I’m not a classically trained journalist, and that I’ve not shared this fact with Chomsky. I am, at best, an opinion columnist, which is basically just a blogger with a four-year degree. So, I don’t know exactly what I should be doing, although I have questions prepared – 16 of them, printed out on a sheet of 8.5×11 paper, which I totally forget I’ve even brought, and thus do not reference for the entire hour we’re together – and a legal pad.

At this point I write on the legal pad “very hard to break eye contact,” because it is. Chomsky is intense, not in that he is fiery – because he is, in fact, exactly as warm-milk grandpa-pleasant as he looks in the photo and sounds in the recording – but in that he is so present and connected and focused, a hawk’s brain and eye, while my attention generally flits about like blind goldfinches in a high wind. Just look at the full transcript: I can hardly string together two dozen cogent words. Chomsky speaks in unbroken, coherent thousand-word mini-essays. If you’ve ever been stuck in an hour-long chat with a fairly intelligent parrot, then you know what Chomsky probably felt like talking to me in the Campus Inn bar at 10 a.m. on July 12.

DEN: So you’re saying that the fact that it [language] is linear, that we speak in a stream of words –

NC: – It’s because we can’t speak in parallel. That’s the nature of our sensory-motor system. If we had two channels, and could speak in parallel – which, incidentally, dolphins do – we could have a more complex system. But whatever’s going on in the brain has to be – if it does come out – has to come out through the sensory-motor system, through touch or sight or, usually, sound. And, incidentally, very little of it comes out. If you just introspect for a minute, you’ll discover that you can’t go thirty seconds without talking to yourself. It’s almost impossible. And 99% of your use of language never even is externalized.

DEN: Um [very awkwardly long pause] That’s really wonderful and stunning.

NC: Well, just think about it. It takes a real act of will just to keep from talking to yourself.

I’m already chronically, painfully aware of my own perpetual internal dialogue, so Chomsky might just as well have said, Don’t think of a pink elephant!” For the rest of the interview it’s effectively impossible for me not to focus on the constant side-chatter of my damned monkey mind. Pro-tip: It is loud in here, guys.

DEN: Um … so, when you’re talking about … What’s the edge of language for you? ’cause, we do things that are communicative that you probably aren’t grouping into “language.” Like, we’ve made choices in the garments that we’ve worn, in part to communicate certain things about ourselves.

I actually shaved for this interview. You can see that in the picture. Why? Noam Chomsky (the linguist Noam Chomsky, not the political thinker; I wasn’t aware of that writing until sometime around 9/11) was an enormous influence on me in college. I probably should have mentioned that earlier – and I actually intended to do it first thing in the interview, but I got so rattled by the fact that Noam Chomsky walked into a bar(!!!) that I flaked and went off-script. Seriously, if you are getting any joy from reading the italicized portions of this today, it is almost certainly in part due to the fact that a linguistics professor told me Chomsky’s great truth about human language: That colorless green ideas sleep furiously. [2]

NC: Almost everything we do is some form of communication.

DEN: But when you’re talking about language you’re taking a subset –

NC: – Language is also used for communication, and for many other things. Like, for whatever reason, humans find it difficult and hard to be in the same place with somebody and not to talk to them. If you’re standing at a bus stop for half an hour the chances are you’ll talk to the person next to you –

DEN: – I have a seven year old son, so I understand.

This seems really out of the blue, because I’m basically only externalizing the least viable stub of an internal dialogue that, frankly, is a little too tangled and inchoate to encapsulate meaningfully here, on a two-dimensional page, using a running stream of letters, spaces, and punctuation. The point, more or less, is that my older kid talks constantly, and much of it is clearly him organizing and testing the soundness of his thoughts, rather than him actually trying to meaningfully exchange data with me.

NC: It’s just really hard. If you’re stuck in a room with somebody – it’s a way of torturing people. Put two people in a room, close the door, and ask them not to talk to each other. I mean, it just doesn’t work. You have to do it. And that kind of interaction is not serious communication. So, if you’re talking to someone at a bus stop, you talk about what happened to the football team yesterday or something. You don’t care, they don’t care, you’re not trying to transmit any information. It’s what anthropologists sometimes call “phatic communication,” just setting up relations among people, and most interchange with language is actually that, telling jokes –

Jokes! “Noam Chomsky and a parrot with a four-year-degree walk into a bar … “ 

– at a party, not communicating anything. This is just a way of creating and solidifying social relations. If communication means anything – we can deprive it of meaning if we like – but if it means something, then it has something to do with transmitting beliefs, information, something like that, and that’s a very marginal aspect of language – or wearing clothes, incidentally. And you’re communicating in other ways. But even that is – plus, the fact that very little of language is even externalized.

All of that isn’t really as significant as what’s more difficult to think about, but much more important: What’s the fundamental design of language? If you want to understand what the function of the visual system is, you can’t just say “I use it to watch television.” That’s not biology. You want to understand the visual system, you have to look into its nature. See what it does. Investigate it. It’s a scientific problem. And then you find things about the visual system that are surprising.

It’s the same with language. It’s not enough to say, “Yeah, I use it to talk to my friends.” If you want to understand something about it, you have to look into its nature. And you can’t do that in a casual conversation. That requires investigation, like any other hard topic. And when you investigate it, you find things like I just mentioned, that linear order just isn’t part of the way language is used. You can see it in simple cases. Take the example I used last night, take this sentence:

Eagles that fly swim.

Ok. And put an adverb in front of it:

Instinctively, eagles that fly swim.

“Instinctively” goes with a verb. Which verb? With “swim”? With “fly”? So, it’s “instinctively they swim” not “instinctively they fly.” Well, “fly” is the word that’s closest to “instinctively,” but you never use linear order, you never use closeness, to determine how things are understood. Or suppose you put “can” in front of it:

Can eagles that fly swim?

You’re asking a question about swimming, not flying, even though “swim” is farther away. And in fact there is a minimal-distance relation between “instinctively” or “can” and “swim,” but it’s a structural relation, and that requires looking into the nature of structure. Then you find that, yes, it’s a minimal distance, but a minimal structural distance, not a minimal linear distance.

Now, calculating linear distances is far simpler than calculating structural distances. If you look at the computational properties, calculating order is trivial. Calculating structure is quite difficult. Nevertheless, you just automatically, reflexively calculate structure. A child, for example, never makes a mistake about this. Children have no evidence, they have zero evidence, about what the adverb goes with, but every infant understands it, and it’s because the nature – and this is not just this example, but every example, every construction in every language works the same way: It uses computationally complex notions like structural distance, but not computationally simple notions like linear distance. And there’s only one possible explanation for that: Linear order just isn’t available for the computational system. It’s just not available, period. Therefore you can’t use it, and you use the simplest thing there is, which is structural distance.

But that just tells you that linear order comes from somewhere else. Not from the language. And it’s obvious where it comes from: You can’t speak structurally, you can’t produce structures , you can only produce linear orders of words. So, you’re forced, whatever’s going on in the mind, has to work its way through this system on the outside, the sensory-motor system.

If you look at the evolution of – we know very little about the evolution of these things, but something – the sensory-motor system was around for millennia, centuries before the language ever emerged. It was just sitting there. For example chimpanzees have approximately the same auditory system we have, without any problem detecting the sounds we detect, but it’s just noise to them, they can’t do it, they’re not designed to do it. So it takes a newborn infant – with all kind of noise going on in the environment, a mass of stimuli, what William James called “blooming, buzzing confusion,” that’s what the child is born into – but an infant reflexively extracts from that environment certain things that are language related. An ape just hears noise. Similarly, humans can’t detect what other animals pick out of the environment, were not built for it. They’re not built for language. We’re built for language.

An infant picks it out reflexively, and virtually reflexively goes through the stages of development that lead to what you and I are now doing with very little teaching. It’s called “learning,” but that doesn’t really mean anything; it’s just a system that grows, like other biological systems, it’s internally designed. When you study its nature – which takes some work – you discover that, for example, things like linear order just aren’t part of it. That’s a deep, fundamental property of language, and it indicates that any form of externalization, getting it into the outside world, is tacked on, it’s secondary, and therefore any use of that is even more remote from the nature of language. Like communication, for example.

One of the few meaningful notes on my legal pad is at this point, and reads: “11-ish [minutes into the interview]: yr brain explodes.” Because that was the point in the interview when mine did. We have about 37 minutes left. I’m still on question #1.

DEN: Lemme step back: Why is it that we know, or that we’re led to believe, that externalization is an afterthought, rather than fundamental?

NC: Because externalization requires linear order, and linear order is not part of the computational system, and therefore it’s tacked on afterwards. And the only obvious reason is because whatever’s going on in our minds has to work its way through the sensory-motor system to get outside.

DEN: So there’s –

NC: – And the sensory-motor system imposes constrains that have nothing to do with language.

DEN: I have … bifurcating questions now, predictably.

NC: Pardon?

DEN: [I make a crazy "forking paths" gesture.] I have bifurcating questions. Um … I guess, one, just about linguistics, is that … I get the sense that you kind of feel like linguistics has gone off in the weeds. There’s too much interest in communication and not enough into the fundamental process that’s generating language.

NC: Actually, there isn’t very much interest in communication. This is a dogma , but nobody really investigates it. Nobody studied much about commutation. And there’s a good reason for that. Communication – if you’re really interested in communication, you’re studying a process of extreme complexity. When you and I do try to communicate – like what we’re doing now – we’re bringing in all kinds of background assumptions about each other, which we’re not aware of, but they’re critical to the interchange that’s going on. They come from all over the place. I mean, the way we look, my guesses about your background, your guess about my background, an enormous, complex background of implicit understanding – partial understanding – and assumptions is involved in the simplest act of communication, and that’s just too hard to study. Scientists can’t study complicated things, they can only study simple things. So nobody studies it. [ ... ]

There’s something obvious about language. Every language – yours, mine, everyone’s – provides an unbounded array of potential expressions that could be thought and could be used, and it’s unbounded, it’s infinite and you can keep adding new ones without bounds, kinda like the numbers: there’s always a bigger one. That’s gotta be a finite system, because you’ve got a finite brain. So, there’s a finite system that’s yielding – “generating” is the technical term – an infinite array of potential expressions, each of them is a structured object which has some kind of meaning, a semantic interpretation connected to thought, and sometimes it’s externalized into the outside world – rarely, but sometimes. That’s the core property of language.

Now, it wasn’t really possible to study that property seriously until about 60 years ago. And the reason was that the fundamental concept of the finite system yielding an infinite array of structures was not very well understood. That’s why you can go back to, say, Darwin, he says the fundamental difference between humans and all other animals is language, and language provides, he says, an “almost infinite number of expressions.”

Well, the phrase “almost infinite” doesn’t mean anything. That’s like saying there’s an “almost infinite” number of integers. It’s not “almost infinite”; it’s infinite. You can’t be “almost infinite.” But the concept wasn’t understood, so the standard traditional term used by a lot of people was “almost infinite.” Doesn’t mean anything. What it means is infinite , but the brain is finite, so you have the problem of understanding how a finite object – like, say, your laptop computer or your brain, which is a finite object – can, given enough time and memory, go on indefinitely.

You can write a program for your laptop which will add any numbers no matter how big they are. Of course, for the laptop to do that you might have to stick in extra disks, because it doesn’t have enough memory, but it has the program, the system, which can do it. And that range of concepts was not really fully understood until roughly the 1930s. By then the fundamental nature of computational systems was clarified and discovered, and it became possible to pursue these questions in a much more serious way than before. There were stabs at them in the past, but it could only be stabs. But the basic understanding of such systems made it possible to look at the core property of language, the basic property, the one that I just mentioned. But when you do it, it’s not easy. It’s like other problems in the sciences: It sounds easy, but it isn’t.

As I mentioned yesterday, what started to happen roughly in the 1950s is a little bit like what happened in the basic sciences in the early 17th century. You go back to the history of science, and for millennia the greatest scientists thought they had answers – Aristotle, Galileo thought they had answers to very simple questions. Like, suppose I have a cup in my hand, and it has boiling water in it, and I have my hand over the cup. If I let go with my hands the cup will fall and the steam will rise. Why? Well, there was an answer: They’re each finding their natural place. The natural place of steam us up there, the natural place of the cup is down there. OK. End of problem. If I have two cups, or two objects, a big heavy one and a little light one, which will fall faster? Well, obviously, the heavy one.

And that was – how do I – if I look at that rectangle over there [He points at my iPhone, which was at that very moment totally failing to make a backup recording of this conversation.], how do I see it? Well, it’s form flits through the air and imprints itself on my brain. That solves that problem. And there were answers like that to problems, except none of them made any sense.

And when, around the time of Galileo, the early 17th Century, when scientists began to allow themselves to be puzzled about these things, and not to be satisfied with what are really meaningless answers, then you get the beginning of science. It turns out all your intuitions are wrong, all your guess are wrong, all your beliefs are senseless, and you enter into the rational spirit which became the Enlightenment and became the modern era. It’s not that there was nothing there before, of course there was a lot there before. But there was a big change then. And that’s the same kind of change that, in my opinion – and this is a minority opinion – ought to take place in the human sciences. And can, in some areas. The study of language is one of them. The study of vision is another. In fact, if you go back 60 years it was quite commonly assumed among biologists that the variety of organisms – the same phrase was used, “almost infinite.”

Well, you’re either infinite or fixed; there is no “almost infinite.” But that was a common belief; I’m actually paraphrasing a famous biologist. And it turns out to be completely wrong. The range of organisms is very narrow, so narrow that there are even proposals – which may be wrong, but are taken seriously – that there’s what’s called a “universal genome,” a basic genome for metazoa – you know, complex organisms – that’s essentially fixed and emerged at around the Cambrian Explosion, and ever since then we just get slight variations on it, like microbacterium and humans, slight variations on the basic pattern which was fixed. And that may not be right, but it doesn’t seem crazy, and it’s not kind of an “infinite” variety. And the same is true of language and other system, and now something is beginning to be understood about it.

But, to a large extent it hasn’t penetrated into the human sciences, and this ability just to be puzzled, just that, Galileo’s ability, that’s a hard barrier to overcome. The case that you brought up at first, about communication and linear order, the case in point, people ought to be puzzled by the fact that in the simple sentences I mention you interpret them one way and not another way, automatically, you don’t have any choice, it’s just the way your mind works. If you’re puzzled, then the question is “Why?” and there’s an immediate answer: The computational system just doesn’t pay attention to the externalization, hence the communication is something that’s tacked on to the outside. [ ... ]

DEN: So, then, your inclination would be to believe that language developed among our specific species because it did, and once it started popping out of our mouths and hands we had use for it?

I can’t say precisely why, but as this conversation progressed I became super-wary of inadvertently leaving sign-language speakers out. I don’t know why this is; I only know, like, two signs, one of which is “more” (you can teach it to pre-external-language babies and make your life much easier. At one year old a human’s inner life is much more nuanced than his or her ability to form words with his or her mouthparts, which is enormously frustrating for a baby) and the other is “I love you” (well, actually, it’s “You complete me,” which I learned from the movie “Jerry Maguire,” and use all the time, because “You complete me” is indeed much more accurate to my sense of the bond of marriage than the word “love.” For details please see, for example, the tattoo on my left arm.)

NC: There was actually use for it before it popped out: You could think. We don’t have any tape recordings from 100 thousand years ago, but you try to reconstruct what must have happened, what very likely happened. Proto-humans are hunter-gatherers. In fact, humans were hunter-gatherers until very recently, that’s practically the whole history of homo sapiens. And there were small groups, maybe 50 to 100 people. Evidently, in one of these groups, some very small, probably very small, wiring of the brain took place, by some small genetic change – not enough is understood about genetics to imagine what it could be – but things like that do happen. We have evidence for them.

And so some small genetic change took place, and it yielded the capacity to produce, and to meet the basic property of language, to produce this infinite array of expressions that latched on to thought systems that were kind of inchoate, limited, restricted, maybe shared with other animals. And at that point you have the ability to plan, and to interpret, and to think, and that gives you advantages. You can’t communicate, because nobody else has it, this happens in a person. You don’t have mutations in a group, they’re in individuals. It could be – it would probably die out.

Most advantageous mutations do die out; it’s hard for them to latch on. There’s statistical evidence for that now. But this one apparently latched on, and it transmitted to offspring partially. After a while, it could take over a small group. At that point there becomes a reason to externalize it. Externalizing it is a hard problem, you have to connect this thing that’s inside your mind, which developed without any selectional pressures at all, just by laws of nature, and you have to connect that to a sensory-motor system that had been around for hundreds of thousands of years, and they don’t match. They’re different systems. So, that’s a hard problem, and that’s the problem of externalization.

When you learn a second language – Suppose you learn French. French is going to have the same property – the linear order that I mention – but you don’t learn that, nobody teaches that too you. You know it, because that’s part of being human, to know that. What you have to learn is what amount to trivialities: how it’s pronounced, what’s the order of the words, what kind of affixes does it have. You know, you memorize inflections [shrugs]. What you learn when you learn a second language are a selection of trivialities that are around the periphery of the language.

But the core part of the language you don’t learn because you know it. That’s just the nature of humans. And the same with a child acquiring the language. And that’s the externalization problem, and it’s a hard problem, so it can be done in many different ways, and we get the kind of illusion that there are many different languages, but fundamentally they’re all the same. And that’s how we can learn any of them, a child can learn any of them because basically the core computational processes are probably identical, or very similar, with variations in almost entirely the externalization system.

DEN: You had mentioned the anguish people feel if they’re in proximity and denied the ability to communicate, or obviously if they’re isolated and unable to communicate –

NC: – To interact , which is quite different from communication –

DEN: – to interact with another, or with their sense of another, person or entity.

NC: To just have some relation to that other person there.

DEN: Would you – I’m asking you to guess, I suppose – that it came with that original creation of language device, or that it only arose after people started [Chomsky and I talk over each other here for a bit, and the only discernible word on the recording is Chomsky saying "animals."]

NC: There are a lot of things like that. For example, it’s very, you can’t look into somebody else’s eyes for more than a fleeting second. If you do, it’s either hostile or very intimate. It can’t just be ordinary.

I now feel super self-conscious about that first thing I wrote on my legal pad.

Like, you and I can’t look into each other’s eyes for more than a second, it would be very hostile, or else it would be that we were in love or something like that.


There’s nothing in between. That’s just part of being human. Other animals find it very hostile. Sometimes you just can’t stare at an animal, it’ll get angry, you know? Why? And it’s got nothing to do with language. It’s something deep in the history of organisms. Nobody knows why.

DEN: So that’s just an aspect, like the sensory-motor system has become very tightly linked to this, but it’s sort of a separate –

NC: – It’s got its own characteristics, which were there long before language. As I said, the auditory system is probably common to primates, very close to it. There’s even evidence that chimpanzees pick out the same sounds that humans pick out. We don’t use all the possible sounds for language, only some of them, and it seems that apes are pretty much tuned to the same sounds. Of course, they can’t do anything with them, to them it’s just noise.

DEN: So, are there questions being asked and investigated in linguists right now that you find very interesting, or is it –

NC: – Oh yeah, all of these, for example. And this is just the beginning. You can go on and find many other things. In fact, almost everything you look at is puzzling. It was just like physics in the 17th century. Once physicists allowed themselves to be puzzled, everything became a puzzle, because you understand nothing, you know? In fact Galileo, at the end of his life, kind of commiserated that – I forget exactly how he put it, but he said, I think, “We can understand nothing about the world, nothing.” And what he meant is we can’t get an explanation for it in terms that are intelligible to us.

DEN: So we’re blinded by all the things we think we know?

NC: We know it in some sense, like I know that if I let the cup go it will fall, nobody is telling me that, I know it. But why does it fall? That turns out to be not so trivial. If you look at how fast does it fall? Do two things of different weights fall at different speeds? That’s the bare beginning already. By that point your intuitions are all wrong, and then you go on to more complicated things, and they’re even more wrong, you know? And that’s rational inquiry, that’s science.

DEN: [moderately awkward pause] Man, I don’t know which question to go with next! [ ... ]

There’s a section I’m eliding here, because it’s pretty barren. First I ask a question that is so out-of-context that my actual meaning is basically unintelligible – and so Chomsky addresses the only interpretation that would make sense, as opposed to what I actually meant, which just wasn’t there in the words I said out loud – and then I bumble into asking about his word processor and … ugh. The whole thing is a wreck. You can go check out the full transcript if you’re curious. What’s most interesting in that bit – and really the only reason I’m butting in right here to point out the cut – is that Chomsky is pretty clear that he’s been giving a highly abridged layman’s description of universal grammar to this point, and that there really is a very complicated, data-driven, math-y backdrop to his claims that he’s left out because he’s sitting in a bar with a shaggy non-journalist, and not because it doesn’t exist. I felt it was important to make that clear: What Chomsky does – this part, the linguistics part – isn’t op-ed; it is science. That he has interpreted the results in a different way than many of his colleagues, because he is interested in a different set of more fundamental questions than they are, isn’t a differing of opinion; it is a differing of interpretation.

There’s also sort of a nice moment embedded in there, where Chomsky cops to the principal problem of becoming accomplished in any field, which is that the more you understand the stupider you feel.

Dogs, I doubt, feel this. Ants and bees likely don’t. Probably not monkeys, either. Maybe dolphins. This, after all, is a hell built from language, which allows us to code and store our momentary notions so that they can shame us again later. Language gives us the capacity to be nostalgic for the time that we were nearly entirely ignorant, and thus knew everything worth knowing. Learn a little, and the first thing you learn is how very, very, very far you have to go. The well is deep.

It’s just a moment, and a pretty one, but there was no way to untangle it from the mess I made incoherently asking about his computer, so it’s gonna stay on the cutting room floor. But it’s a momentary gem, and you can find it in the complete transcript, or in the recording around the 35 minute mark.

[ ... ]

DEN: When you read the things that your critics have written about you, politically, do you recognize yourself in how they characterize you?

NC: Virtually never.

DEN: I was wondering – I read some pieces David Horowitz wrote around September 11 that make you sound much more terrifying than you are.

NC: What you should do when you read these things is just check back to the original sources, and see what’s there.

I had, in fact; that’s how the question arose. 

Ask how it compares to what’s being said. You’ll notice that there rarely are references, and there’s a reason for that.

DEN: One of the accusations I see leveled against you frequently is one of “moral equivalency.”

NC: Moral equivalence is a very interesting concept. Do you know where it comes from?

DEN: Please go on.

NC: I mean, the term is kind of old, but it came into common usage with Jean Kirkpatrick – take a look at the history, do a Google search, and you’ll find that until Jean Kirkpatrick it was almost never used. With Jean Kirkpatrick it got started. What did she use it for? She was in the Reagan Administration. She used it to try to prevent criticism of the United States. So if anybody made any kind of critical comment about U.S. policy, they were accused of “moral equivalence,” mainly equating the United States with Nazi Germany, or something.

It’s just a crude, vulgar propaganda technique to prevent criticism of the Holy State and no one else – that’s the only way the term is ever used, like the way you used it. The accusation of “moral equivalence” means you’re being critical of the Holy State. You’re not allowed to be critical of the Holy State. So that’s “moral equivalence,” whatever that’s supposed to mean. But it’s just a term, a vulgar term of propaganda, and that’s where it comes form. [ ... ] A cheap, cowardly way of protecting divinity from criticism. The divinity, in this case, is State Power.

We did indeed do the Googling Chomsky urged, and found that his thumbnail sketch of the history of the phrase “moral equivalence” seems to be basically spot on – as illustrated by this following comparative Ngram

Google N-gram for "moral equivalency" and "Jean Kirkpatrick."

Google Ngram for “moral equivalency” and “Jean Kirkpatrick.”

DEN: By the same note, though, people do seem frequently upset by what you write, despite the fact that [although] what you write might be strongly worded, [it] is not especially strident. I’m wondering how that strikes you.

NC: If you were in the Soviet Union 50 years ago, people were very upset by what dissidents said. If you’re in Iran today, what critics of the government say is very, really angers the religious population. I’d say the rural population support the clerics. People tend to have doctrines that they want to hold on to. And it’s not – it’s unpleasant. I mean, we have some share of responsibility for what the government does, and it’s not easy to look into the mirror, in your personal life or anything else. There’s no human being who hasn’t done things that are wrong, but they very rarely want to see it. What they usually do is construct some justification for it, not to try to look into the mirror and see what’s true. That’s too simple.

DEN: Do you ever see this as a consequence of the language mechanism itself?

NC: Not really, it’s much deeper than that.

DEN: Deeper than language?

NC: It’s kind of separate from language. It’s just the way – I mean, take a five-year-old kid who has a three-year-old brother. And he took a toy from his brother when his mother wasn’t looking. And the little kid starts screaming and the mother comes in and asks the older kid what happened, and he says “Oh, he didn’t want it, and it wasn’t his anyway, and I just happened to have it” – you know, he doesn’t say, “I took it from him because I’m bigger and stronger than he is,” even though that was the reason. It’s just normal human behavior. It’s not particular admirable, but it’s normal human behavior. And it goes on to the vulgarity of Jean Kirkpatrick. You can’t criticize the Holy State, so we’ll just try and shut people up somehow.

DEN: Just based on what you said today I wonder about that discomfort people feel when they’re told an unpleasant truth.

NC: About themselves.

DEN: About themselves or –

NC: About anything they’re associated with.

DEN: Exactly!

NC: Like your football team or your country or whatever it may be, so you’re defensive.

DEN: And it just seems to me that there’s perhaps an argument to be made that if the primary thing that I’m using language for is to clarify and make coherent my thoughts, when I hear someone saying back to me – it’s like they’re shoehorning into my head a set of unpleasant thoughts.

NC: [ ... ]

I’m gonna cut Chomsky off here. He does have an answer, and it’s interesting if you are interested in politics, but I am not interested in politics, and his answer does not answer, or even really acknowledge, my question.

There are two positions on Chomsky: The larger camp sees him as a leftist incendiary (with those on the right – like Horowitz – often implying [or outright stating] that Chomsky basically lives on the dole, enjoying an undeserved, cushy ivory-tower gig in academia). The smaller camp sees Chomsky as an influential 20th-century linguistics researcher who has a second act speaking out-of-field on political matters.

Both are wrong. For almost his entire career – from the 1950s to today – Chomsky has both been politically active and a highly influential “big picture” linguistic thinker. There’s certainly disagreement about the import of his linguistic findings – this core, but largely unaddressed, question of the primary purpose of language – but not really about their validity, or even about the bulk of Chomsky’s interpretation. His findings have spread, becoming fundamental to fields as diverse as computer science, psychology, and immunology; within the last decade Chomskyian notions have grown to dominate our quickly evolving understanding of neuroplasticity. He’s no intellectual lightweight, even at his advanced age.

Meanwhile, he’s been politically aware and active since his boyhood. Most of us didn’t become familiar with Chomsky as a political thinker until the 1990s, but by then he’d already been an outspoken opponent of much of U.S. foreign policy for more than two decades.

Despite these two Chomskys – the Linguist Chomsky and the Political Chomsky – developing in tandem and throughout the same period (and occupying the same body), both have consistently denied there being any fundamental connection between language and politics, which is what Chomsky does here – despite my clarification. Because I’m not asking about politics and language, per se.

To reiterate my question: When people get pissed off at Chomsky, are they actually pissed off by his ideas, or just by the visceral discomfort of having two ideas in their head that are in conflict?

It’s really clear, from what he’s said throughout his career – and what he goes on to say in the portion of this interview that I’m skipping – that Political Chomsky believes it is the former: Political Chomsky is telling Truth to Power, and Power doesn’t like that, because the Truth weakens Power.

But that doesn’t seem consistent with what we’ve learned from Linguist Chomsky. That guy has totally persuaded me that language, first and foremost, is the means by which we organize our otherwise nebulous thoughts. The process of externalizing these encoded thoughts means cramming those rich, complex thought structures through a funnel, so that they can be spewed out of our mouths. And then, in order for those reduced thoughts to get into someone else’s head – and thus access the rich interpretive engine within – they have to go through yet another distortive funnel.

The whole thing is a terrible hack. I mean, for all intents and purposes, we use vibrations at the free end of our digestive tract to communicate with others of our species. It would make almost as much sense, evolutionarily, for us to communicate via farting.

That’s absurd in the extreme, but so is what we’ve got. The way we communicate creates hazards that otherwise wouldn’t exist. Think about choking: We’ve run the food intake right alongside the air intake, which might be OK, since we also have a few valves set up to keep everything going down the right tube. But then we insert a whole new practice, where-by we insist on using the food-processing parts (jaw, lips, teeth, tongue) simultaneously with the breathing parts. Toss a handful of popcorn in the mix, and sooner or later someone is going to choke.

My dog can’t talk. We’ve been together for a decade, he has a brain the size of a plum, he eats basically all the same stuff my kids do (because they drop it on him), and he’s never choked on anything. Meanwhile, both my children are smart little monkeys, and within the first 18-months of their lives I had to Heimlich each multiple times. Most parents I know have had to extract at least one obstruction from their kids’ wind-pipes. None of the pet owners have had to do likewise for their proto-children.

The only reason we accept this absurd communication arrangement as a rational and efficient way of doing things is that it just happens to be what we do, not because it actually makes any damn sense.

At any rate, Linguist Chomsky – and his findings – insist that forcing our language out through our food-holes diminishes the quality of what can be encoded and transferred. I tend to agree.

That being the case, it’s a bizarre sort of blind spot, isn’t it? To insist that after all that hashing around, the thing that happens in the other guy’s head can conceivably be very rational? Just picture it: Your interlocutor has just patiently listened to what you have to say. Now, inside his head, he’s got his own ideas – which are full and rich and attractively coordinated and structurally complex – and sitting next to them there’s this ugly, pulped melange of what he thinks you think you tried to say about the stuff you were thinking. How can he make a rational comparison of these two totally unlike things? It’s like asking someone to decide which is better: ”Casablanca” or “The Wire” – but in this hypothetical, the person you’re asking wrote a bunch of “The Wire,” and her only experience of “Casablanca” is a garbled description from a YouTube comment.

So, that’s what I wanted to ask Chomsky about: Is it possible that guys like Horowitz aren’t freaking out about what Chomsky says, but instead about what it felt like to have this ugly, clanging, discordant, conflicting idea blob thrust into his or her head?

I mean, heck, isn’t the proof in the pudding? I asked Chomsky this same question twice in a row – the second time quite directly – and still couldn’t get him to hear what I was asking because it didn’t fit in among the ideas already there in his head. There was no place for it, and so it simply could not be addressed. How can I use my imperfect, externalized word-pudding to influence the well-built, ironclad pure ideas a listener has already constructed inside his or her head?

Neither Political nor Linguistic Chomsky offer us any meaningful way to handle that problem. Throughout his career, Chomsky has been pretty direct in saying that “rational debate” is the way to go: The way to change minds is to discuss contentious topics openly and engage these ideas directly, with little or no ornamentation or rhetorical flourish. But that’s demonstrably nonsense: The linguistic machinery with which we’re equipped – and the expressive limitations saddling that machinery – pretty much guarantees that it’s impossible to meaningfully compare and evaluate a new idea against a preconceived notion. (Incidentally, this whole issue seems to be yet another puzzle linguists and those in the “human sciences” don’t seem particularly interested in investigating. If you’re legitimately interested in the mess of human communication, evaluation, and persuasion, I strongly suggest you begin reading marketing books, and meditating on how marketers implement the findings from modern psychology and linguistics.)

As I think about this now, weeks after the interview, the true problem of “moral equivalence” becomes obvious, and it has little to do with State Power, or whatever. The thing Chomsky does that gets branded as “moral equivalence” is rarely an actual call to rate parties as to their moral character. More often, he seems to be attempting to elicit a sort of Kantian compassion. It is, in fact, one of his more consistent rhetorical devices. For example, following the assassination of Osama bin Laden, he said something to the effect of: Well, how would we feel if Pakistan had sent a team of commandos to Texas to break into George W. Bush’s compound, scare his wife, shoot him in the head, and dump the body in the Gulf of Mexico?

And, I’ll level with you, I’m not a fan of George W. Bush, and don’t really have a meaningful opinion of Pakistan, but I still would not be cool with that.

Chomsky isn’t saying these are equal, interchangeable situations – because they obviously aren’t – but he’s framing a thought-experiment to allow us to make contact with how Pakistanis might feel about a glib, rah-rah bumper sticker metric for how well a sitting PotUS is doing: “Bin Laden’s Dead and General Motors is alive.” (The fact is, that’s just a creepy thing to say, both clauses.)

If we’re quick to criticize a rhetorical device as the verboten “moral equivalency,” the problem isn’t that we’re acting to support State Power, or whatever – that’s really just an accidental by-product. The problem is that by avoiding any risk of being accused of “moral equivalence” – or “being PC” or ”blaming the victim” – we constrain our own capacity to fully inhabit another’s worldview. And that is the necessary condition that must precede a rational comparison of ideas. I can’t just listen to Chomsky – or David Brooks, or Heather Mac Donald, or Christopher Hitchens, or my father-in-law – talk, then set those desiccated externalizations of their ideas next to my rich self-generated notions, and evaluate them against each other. I need to permit myself to fully swallow what they suggest and regenerate it inside myself, to fully animate those ideas, so that they are as alive and vivid as my own. Only then do I get to be rational.

But, again, that’s hard work. Very few of our decisions will be made that way. So, it’s sort of a terrible irony: One of our sharpest living thinkers has dedicated half his lifetime to justice, using methods that could never work. And he should already know that these methods won’t work, on account of the fact that he dedicated the other half of his life to sussing out precisely why just telling it like it is has almost no chance of causing broad political change. Changing politics means changing minds, and cognitive dissonance doesn’t do that. Cognitive dissonance triggers an array of defense mechanisms – the formulation of repetitions and justifications – that serve one end: To preserve that first thought we already had in our heads.


[1] “Werner Heisenberg, Kurt Gödel, and Noam Chomsky walk into a bar. Heisenberg turns to the other two and says, ‘Clearly this is a joke, but how can we figure out if it’s funny or not?’ Gödel replies, ‘We can’t know that because we’re inside the joke.’ Chomsky says, ‘Of course it’s funny. You’re just telling it wrong.’ “ (via where the joke is explained)

[2] If this makes no sense, just Google the phrase “Colorless green ideas sleep furiously,” read its Wikipedia page – or, skip to the end of the full transcript, where I gush a little about Chomsky to Chomsky, and also talk about my poetrybots, and Chomsky has some pretty interesting reflections on the life-cycle of that sentence as it’s matured.

The Chronicle relies in part on regular voluntary subscriptions to support our publication of local columnists like David Erik Nelson. Click this link for details: Subscribe to The Chronicle. So if you’re already supporting us on a regular basis, please educate your friends, neighbors and colleagues about The Chronicle and encourage them to support us, too!


  1. By Mark Koroi
    July 25, 2013 at 5:59 pm | permalink

    Noam Chomsky is one of the most authoritative figures in America.

    He pioneered the field of psycholinguistics and has taught at MIT since 1955.

    He worked on a kibbutz in Israel and later became a top scholar on the issue of Israel/Palestine relations. He visited Gaza just before the commencement of the November 2012 armed conflict with Israel in solidarity with Palestinians.

  2. By Daniel O'Neil
    July 25, 2013 at 6:42 pm | permalink

    Great interview. One thing that occurs to me is that Chomsky’s assertion about sequential versus structural grammar being a result of an artifact requiring thoughts be expressed linearly is making the assumption that sequential grammar is informationally sufficient and, in fact, computationally simpler than structural grammar. There are mathematical models that would argue against both these assertions. So, I disagree with that part of his logic.

  3. By Daniel O'Neil
    July 25, 2013 at 7:14 pm | permalink

    One additional thought: I think that your discussion at the end misinterprets Chomsky’s point about linguistic communication. He does NOT believe that the transmission method is inadequate; he is only arguing that the grammar structure we use is not the most intrinsically efficient, assuming that transmission method was a consideration during the message’s creation.

    There’s a lot of good evidence that indicates the signal to noise ratio for language is actually pretty good, if nothing else because it’s unbelievably informationally redundant. Point being that if people are able to think clearly about what they want to say, they can usually say it and have it be understood by another native speaker.

    So, it’s not a transmission problem. It is, I think, closer to the issue of cognitive dissonance you describe. The reason Chomsky argues that language and politics are separate is because he’s saying, “look, the fact that you need to have the courage to absorb what I’m saying is not because you won’t be able to because of inherent problems with linguistic transmission. It may, however, be a heavy problem for you to deal with existentially”.

    And then, as a corollary, as all visionaries do, he says, “are you up for the challenge?”. Some folks do come around to his worldview, you know. I certainly have changed my thinking on some stuff because of him.

    I love that you created a context where we can have this discussion. Lots to think about.

  4. By Maria Huffman
    July 27, 2013 at 11:08 pm | permalink

    Great article. I don’t agree with Noam Chomsky’s conclusions, but great article.

  5. By Steve Bean
    July 28, 2013 at 9:07 am | permalink

    “The way to change minds…”

    “Only then do I get to be rational.”

    You’ve done one of those things (along the lines of an assumption) that Chomsky referred to. You’ve forgotten that communication isn’t always just between two people – there can be observers, like us readers, or some other audience, who, by dint of not being as egoistically invested in the process, can often be more objective and open minded toward “rational discussion”. So it’s not “nonsense”, at least not entirely, and you (and we) have other opportunities to be rational.

    From my seat here on the couch I’d say that your case for the Political Chomsky and the LInguistic Chomsky being at odds with one another is a little weaker than you imagined when you wrote it.

  6. By Steve Bean
    July 28, 2013 at 11:05 am | permalink

    @5: And of course while that was directed to you, Dave, it’s really for everyone else. :-)

  7. By Patrick Powers
    August 11, 2013 at 6:42 pm | permalink

    Werner Heisenberg, Kurt Gödel, and Noam Chomsky walk into a bar. Heisenberg turns to the other two and says, “While this seems to be real, it could be a joke. How can we determine whether or not this is real?” Gödel replies, “If it is real it is funny, but if it is a joke it is not. Jokes are funny, but cruel reality is not. Either way we get a contradiction. Our status as a joke cannot be determined from within this system.” Chomsky says, “I never tell jokes, so I can’t help you.” Then the author of the joke says, “Of course it’s funny. You’re just telling it wrong.”

  8. August 16, 2013 at 4:45 pm | permalink

    OK, that LOLed me, as the sort of meta-DFW version of the gag.