The Science of Language

by Noam Chomsky

  NC: Well, peripheral systems are systems that are input systems and output systems. So, the visual system receives data from the outside and transmits some information to the inside. And the articulatory system takes some information from the inside and does some things, and has an effect on the outside world. That's what input and output systems are. Language makes use of those systems, obviously; I'm hearing what you say and I'm producing something. But that's just something being done with language. There's some internal system that you and I pretty much share that enables the noises that I make to get into your auditory system and the internal system that you have is doing something with those noises and understanding them pretty much the way my own internal system is creating them. And those are systems of knowledge; those are fixed capacities. If that's not an internal system, I don't know what the word means.

  JM: OK; there are other systems, such as facial recognition. That also is not a peripheral system. It gets information from the visual system.

  NC: Well, the facial recognition system is an input system, but of course it makes use of internal knowledge that you have about how to interpret faces. People interpret faces very differently from other objects. Show a person a face upside down; he or she can't recognize it.

  JM: So it's not sufficient (to be a central system) that it get information from some other system.

  NC: All kinds of internal processing is going on – thinking, whatever that is. And most of it is totally unconscious and beyond the reach of consciousness. But there's plenty of evidence that it's going on. The evidence, of course, is always from the outside. Even if you're doing brain imaging, you're getting evidence of the effects of the internal object; but that evidence pretty conclusively shows that plenty of mental internal operations are going on, using systems of knowledge, interpretation, planning, action, and so on. And language just seems to be one of them. It may well be the one that binds them all together. In fact, that's been proposed in a clear form by people like Elizabeth Spelke (2003, 2004, 2007) at Harvard. She's a major cognitive developmental psychologist, who argues that in her work (with children, mainly) that as language capacities appear to be developing they're always there, but they don't manifest themselves until certain ages. As they begin to manifest themselves, you get interactions among different other kinds of cognitive activity. There are primitive forms of recognition of, say, where you are and where you ought to go that are shared right through the mammalian kingdom and they are used by children. There are more sophisticated ones that are used by adults. So a young child can learn about the difference between a blue wall and a red wall, as can a rat. But if you want to learn about the difference between, say, to my left and to my right, or some more complicated thing that is much more complex – well, she argues that these more complex capacities seem to be integrated at about the time that the language capacities are manifested, and therefore she suggests, plausibly, that the language capacities facilitate this kind of interaction – which certainly makes sense.

  The archeological-anthropological record suggests similar conclusions. As mentioned before, there is what is sometimes called a “great leap forward” in human evolution in a period roughly 50,000–100,000 years ago, when the archeological record suddenly changes radically. There are all sorts of indications of a creative imagination, planning, sophisticated use of tools, art, symbolic representation, and taking careful note of external events, such as phases of the moon, things like that. And it looks as though that is about the time that whatever put language together emerged. So it's not an unreasonable speculation that they're connected. If some hominid has a language capacity, it can plan, it can think, it can interpret, it can imagine other situations – alternative situations that aren't around – and make choices among them, or have attitudes toward them. At some stage it can convey some of that to others. All of that could well be the source of whatever brings together the other various capacities, which probably existed – at least, in some rudimentary, unsophisticated form – brought them together and gave rise to this evidence of sudden complex, creative activity.

  JM: It's almost certain that language has the place of constituting or providing much more complicated, much more complex, highly structured forms of conceptual ‘objects’ – offering a conceptual range that is just not available to other types of creatures. But it sounds as if language, having that kind of capacity – and being given the capacity of coordinating, and perhaps even integrating information from other systems – that . . .

  NC: And also innovating. Language doesn't just bring together information; it's no recording device. We know this by introspection – you can think about what the world would be like if you chose this course and not that course. In fact, you can imagine things that can't even be physically realized. There's a range of possible creative activities available to you – to some extent, you have that available from other internal systems too; with vision, you can imagine pictures, that sort of thing – but it's far richer when you can actually formulate internal expressions. You have propositional attitudes; descriptions of possible organizations and interactions between people; possible physical events, and so on. And that's available to you if you have internal language; we all know it, just by introspection. We all have that capacity. And presumably our hominid ancestors at that point had the same capacity.

  One thing we're pretty certain about is that existing humans are virtually identical in this respect. Which means that whatever it was that gave us this capacity – it couldn't have happened later than about fifty thousand years ago, which is about when the trek from Africa starts. In fact, it's likely that whatever happened led to the trek from Africa. Hominids physically very much like us were there for hundreds of thousands of years – as far as the archeological record shows. And it began with a small group, one of many small breeding groups. And that one suddenly exploded all over the world. It's hard to imagine that that's not related to the same developments that led to the human capacity to innovate.

  JM: But if language has this role in the mind, is there any need for what Jerry Fodor used to think of as a ‘central processor’? Isn't it sort of taking over the role of the homunculus – being that, whatever it is, that coordinates and brings together all kinds of information, makes decisions, and so on . . .

  NC: Is it the central processor? How could you know? Probably not, I would guess. There are ways of integrating, say, sound, and smell, and sight, that you wouldn't even know how to describe in language. And they're certainly part of life. I imagine that there have to be central processors for that.

  One suggestion that Jerry [Fodor] proposes which seems to me to require more evidence is that there is a language of thought. And the question is whether the language of thought is any different from whatever our universal, internal language is. As far as I can see, we can't tell anything about the language of thought other than it's a reflection of whatever our language is. And if it's true – as it is likely – that the existing and, indeed, the attainable languages are only superficially different, then the core that they share has a good claim to be the language of thought – so far as I can see.

  JM: What, then, about the sciences? They seem to have a syntax that is far different from the syntax of natural languages.

  NC: Not only that, but they seem to operate in different ways. There is a debate about this. There are people like Sue Carey, for example, who tried to show that our scientific capacities are just the natural development of our ordinary commonsense capacities for reasoning, exploration, and so on. I'm not convinced. I've got a feeling that science involves quite different capacities of mind.

  There's really no serious evidence on these things. But if you just look at the history of the sciences, it looks also as if there was also a great leap forward, but much later, and – unlike the one that came about through language – presumably requiring no biological change. It's not that there weren't precursors, but around the seventeenth century, the attitude toward investigating
and understanding the world just changed radically. Right around the time of Newton, there was a dramatic change – so dramatic, that Newton – who basically helped create it – couldn't accept it. The previous assumption – without anybody actually saying it exactly – was that the world was intelligible. God made it perfect, and if we're smart enough, we can see how he did it, and it will be intelligible to us. All we have to do is work hard. The core psychological effect of Newton's discoveries, I think, is that that's not true. It involves what are to us, intuitively, mysterious forces. That's why Newton resisted his own conclusions, which effectively undermined what was called the “mechanical philosophy” – the idea that the world works like a machine, with gears, levers, and things pushing each other, kind of like a medieval clock. It should be something like that. But what he showed is that it just isn't true.

  JM: Is this the end of folk physics and folk psychology?

  NC: This is the end of folk physics. Folk psychology is not refuted. It just doesn't work for physics. The Cartesian mechanics was a sketch of something that was a kind of folk physics. It's our commonsense understanding of how the world works. If I hit that over there, it will move. But I can't move that thing just by moving my arm. It's intuitively obvious that I can't move it by moving my arm; but it just happens to be false. The idea that there are principles of action and reaction, interaction, growth, development, and so on, that are just not of the mechanical type – over time, it caused a real shift in the standards of intelligibility for science. It's not the world that's going to be intelligible; we give that up. But the theories have to be intelligible. So we want intelligible theories of the world that we can work with and that meet our epistemological criteria, which are just other aspects of our cognitive system. And then science goes off on quite a different course. It's not that people give up the commonsense models; talk to a mathematician studying some abstract topic in topology, and in his study, he's probably drawing pictures and thinking about them, and so on. But you know that there's a gap, and that your intuitive, commonsense understanding of the world is simply not a guide to what the world is. That's an important change, and it leads in other directions. It's only happened in small areas of science. Nevertheless, it's quite clear that the other picture of the world – the scientific one – is quite different from the commonsense.

  It goes until very recent times. Until the 1920s, chemistry wasn't considered part of the core sciences, but a means of calculation, because it just had these pictures and diagrams and so on, from which you could get very good approximations to the results of experiments. But it was regarded by many leading scientists as a mode of calculation only; it can't be real. It wasn't reducible to physics as then understood. That was because, for reasons that were later discovered, physics just didn't have the conceptual tools needed to incorporate it. Physics underwent another radical change and became even more unintelligible from the commonsense point of view – although the theory, of course, is intelligible. And then chemistry becomes part of the natural sciences. And so it continues.

  If you take a look today at the debates that are going on about ‘psychological reality’: can the mind be following the rules, and so on and so forth – you'll see that they're very similar to the debates going on in the 1920s as to whether chemistry can be ‘real.’ What do you mean, chemistry is following laws? Show me how to explain what you observe in terms of Newtonian mechanics, or something else I understand.

  JM: That particular dispute – whether the mind is following rules or laws, or whether they're ‘psychologically real’ – has been around at least since Helmholtz, with his idea that the mind somehow carries out inferences, and does so in the case of vision, audition, and faculties like that.

  NC: Yes, these ideas keep coming back; it was a Cartesian issue too, in a way.1 The modern debates about psychological reality in linguistics and cognitive processing – computer metaphors, and so on – they're very similar to the arguments about chemistry and the Bohr atom back in the 1920s. Now it's “How can you explain it in neurophysiological terms?” You can't. But you couldn't explain chemistry in physical terms, as physics was understood then. The conclusion at the time was that there's some problem with chemistry. Later, it was understood that there was some problem with physics. Do we know more about neurophysiology today than people did about physics back in the 1920s? Far from it; quite the opposite. You have to know that you're looking at the right things. And there's nothing in the way of any depth of theory. There's a slogan – that the mind is neurophysiology at a more abstract level. But chemistry wasn't physics at a more abstract level, as it turned out. Rather, a new physics came along, which was chemistry at a different level. And we don't know that that won't happen in the study of mind.

  The point is that if you really accept the achievements of the modern sciences since Newton, what you're trying to do is construct the best explanatory theory you can. And you would like to unify it with other guesses about what the best theories of other phenomena are. But if one of them can't be reduced to one of the others, it doesn't say anything. It just shows you that something is wrong.[C]

  JM: Returning to your remark about language providing innovation, in addition to coordination, integration, and the like. Science – unlike language – seems to offer a different kind of innovation – although again, unique in the animal kingdom. It provides new conceptual materials, new concepts. Language is innovative, but its innovation is compositional; it takes what is available and puts the pieces together.

  NC: The language faculty itself uses the conceptual resources that are available . . . It's a little hard to say what language is ‘itself.’ Does the English language include the word gravitation? We're somewhere in an area now where our linguistic capacities and our science-creating capacities are interacting. We don't understand either of these systems enough to know where to go on.

  JM: I'm thinking of language as primarily the core system, the computational system that operates so that one form of Merge yields argument structure and another provides various forms of edge effect, such as focus. That [system] seems best suited to dealing with concepts that are – as it were – available – and that will serve our interests as people who use common sense to try to deal with the world and think about what their position is and how they can change it, and the like. Science, on the other hand, really does seem to offer the opportunity to introduce new conceptual materials in a different kind of way.

  NC: The sciences provide completely different kinds of ways of looking at the world, which are completely counter to common sense. In fact, common sense – at least, in the advanced sciences – has been completely abandoned; it's not even a criterion any more. But that's a very modern, very special development, which holds [only in] certain areas.

  1 Descartes offered what amounts to a computational theory of vision, one that indicates that the visual system ‘solves problems’ such as determining visual depth by performing a geometric calculation of sorts entirely in the mind, given ‘input’ concerning the degree of convergence of the eyeballs. It might seem puzzling that he introduced the rudiments of a science of mind that shows how the mind offers humans sensations of depth while insisting that a science of mind that offers sensations of depth and other mental phenomena is out of reach. The puzzle disappears when it becomes clear that what he had in mind by science is a contact mechanics, which – if the discussion above is correct – is heavily tied to common sense and folk physics. For some discussion, see my introduction to the 2009 edition of Chomsky (1966/2002/2009).

  12 Chomsky's intellectual contributions

  JM: Noam, let me ask about what you take to be your most important contributions. Do you want to say anything about that?

  NC: Well, I think that the idea of studying language in all its variety as a biological object ought to become a part of future science – and the recognition that something very similar has to be true of every other aspect of human capacity. The idea that – there was talk of
this in Aspects, but I didn't really spell it out – the belief . . .

  [Wait; I'll start over. B. F.] Skinner's observation is correct that the logic of behaviorism and the logic of evolution are very similar – that observation is correct. But I think his conclusion – and the conclusion of others – is wrong. Namely, that that shows that they're both correct. Rather, it shows that they're both incorrect, because the logic of behaviorism doesn't work for growth and development, and for the same reason, the notion of natural selection is only going to work in a limited way for evolution.1 So there are other factors. As I said in Aspects, there's certainly no possibility of thinking that what a child knows is based on a general procedure applied to experience, and there's also no reason to assume that the genetic endowment is just the result of various different things that happen to have happened in evolutionary history. There must be further factors involved – the kind that Turing [in his work on morphogenesis] was looking for, and others were and are looking for. And the idea that maybe you can do something with that notion is potentially important. It's now more or less agreed that you can do something with that notion for, say, bacteria. If you can also do something with it for the most recent – and by some dimension most complex – outcomes of evolutionary history like language, that would suggest that maybe it holds all the way through.

  JM: Well, it would be pretty radical progress if we're actually at the stage now where we can begin to ask for language the old question, “Why are things the way they are?” I guess you think we're at that stage.

  NC: To some extent. I think that there are even some answers . . . In recent work, I've been trying to compare what now seems plausible with what seemed plausible ten years ago. And a good deal of machinery that was thought to be needed has in fact been cut away. How far you can go with that – who can tell? That's like asking what really is specific to language. These questions were coming up all along; that's why I brought up that 1974 biolinguistics conference [I mentioned before]. When you read through the transcript, the questions kept coming up – what could it be that is specific to language? How could it be so remote from everything else in the biological world? It didn't make biological sense. But you were stuck with it. Well, by now you're less stuck with it, and you can begin to ask more seriously the basic questions of biology of language – some of them, answer even. There are still huge gaps. Take the first point you mentioned, about the nature of the concepts. We have nothing to say about how they evolved.