Naturalism and Dualism in the Study of Language and Mind

[00:00:00] STEPHEN NEALE:
Welcome to the Howison Lecture in Philosophy for 1994. I’m Stephen Neale from the Department of Philosophy, and it’s my great pleasure and privilege to introduce this year’s lecture. The Howison Lectures were established in 1919 as a memorial to George Holmes Howison, who came to the university in 1884 as Mills Professor of Intellectual and Moral Philosophy and Civil Polity, a position he held until his death in 1916.

In the letter to the University Regents, the Howison donors expressed their debt as follows: “Professor Howison held the reasoned conviction that this world to its very depth is kindred to the human spirit; that it is a community of free persons, finite and infinite, sustained by the vision of the Perfect; and all his great powers were directed to awaken in others a loyalty to these ideas.” We are honored to have as this year’s Howison Lecturer, Noam Chomsky, world-renowned linguist, philosopher, and political activist, Institute Professor at MIT, where George Howison himself was once Professor of Logic and Methodology of Science. Chomsky spent his undergraduate and graduate years at the University of Pennsylvania, receiving his PhD in linguistics in nineteen fifty-five.

From nineteen fifty-one until nineteen fifty-five, he was a junior fellow of the Society of Fellows at Harvard. In nineteen fifty-five, he joined the faculty at MIT, where he still teaches today in the Department of Linguistics and Philosophy. Chomsky’s work in the nineteen fifties and nineteen sixties revolutionized the study of language.

Fusing ideas from traditional grammar with ideas from parts of mathematics known as recursive function theory, he showed how to provide rigorous theories of grammatical structure. He argued that the capacity for language is a biological system, perhaps analogous to vision, its structure largely the product of a shared genetic endowment. Detailed empirical work and powerful arguments for non-conscious and innately determined grammatical knowledge combined to topple the dominant strains of structural linguistics, behaviorist psychology, and empiricist philosophy.

Generative grammar flourished, and linguistics was transformed from a taxonomic into an empirical discipline. The representation of knowledge, linguistic or otherwise, soon became the very subject matter of much work on the mind, undertaken by linguistics, psychologists, and philosophers. In the ’70s, ’80s, and ’90s, Chomsky has been no less active, refining his earliest arguments and positions, deploying new arguments, and exhausting colleagues and successive generations of graduate students with major theoretical advances in generative grammar.

Chomsky is also well known for his political writings and activism. He’s been a tireless critic of US foreign policy, of relations between government and multinational corporations, and of the timidity, complacency, and complicity of the mainstream press. He’s lectured widely on linguistics, philosophy, intellectual history, and international affairs.

His many books include Aspects of the Theory of Syntax, Cartesian Linguistics, Language and Mind, American Power and the New Mandarins, For Reasons of State, Rules of Rep- Rules and Representations, Lectures on Government and Binding, a linguistics book, not a politics book, Towards a New Cold War, Fateful Triangle: The U.S., Israel and the Palestinians, Pirates and Emperors, Language and Problems of Knowledge, The Culture of Terrorism, Manufacturing Consent with Edward Herman, Deterring Democracy, and most recently, Year 501 and Language and Thought. Professor Chomsky has received honors and awards from around the world. If I were to list them all, there’d be no time to hear him speak, so I’ll mention just a few.

He earned the Distinguished Scientific Contribution Award of the American Psychological Association and the Kyoto Prize for Basic Sciences. He was in residence at the Institute for Advanced Studies in Princeton in nineteen fifty-eight, fifty-nine, and Visiting Beckman Professor of English here at the University of California in nineteen sixty-six, sixty-seven. He delivered the John Locke Lectures at Oxford in ‘sixty-nine, the Bertrand Russell Memorial Lectures in Cambridge in nineteen seventy, and the Nehru Memorial Lecture in New Delhi in nineteen seventy-two.

He’s a fellow of the American Academy of Arts and Sciences, a fellow of the National Academy of Science, a corresponding fellow of the British Academy, and a recipient of many honorary degrees. Tonight, he’ll deliver the Howison Lecture in Philosophy entitled Naturalism and Dualism in the Study of Language and Mind. He’ll enter to– entertain a few questions afterwards.

Please welcome Noam Chomsky.

(applause and cheering)

[00:05:04] NOAM CHOMSKY:
I was impressed to see how many people are interested in arcane questions of linguistics and philosophy. It shows that the, uh, complaints about the intellectual level of American students are totally misplaced. Uh, well, let me begin by trying to clarify what I mean by the terms of the title, uh, which I’m gonna use in perhaps somewhat unconventional ways.

Uh, in fact, I want to use them in ways that I think are innocent of, uh, far-reaching implications. And under that interpretation, I want to suggest several theses to indicate why they’re plausible and what directions they suggest for future inquiry. So let’s begin with, uh, the last word, mind, or to begin with mental.

I want to use the term roughly as we use such terms as chemical or optical or electrical. Uh, certain phenomena, events, processes, states are informally called chemical, uh, et cetera. Uh, we, uh, there’s no metaphysical divide suggested by that.

We merely select certain aspects of the world as a focus of inquiry. Uh, we don’t seek to determine the true criterion of the chemical or the mark of the chemical of the optical or the bounds of the electrical or anything of that kind. Uh, and, uh, I want to use mental, uh, very much the same way with something like the ordinary coverage, but no deeper implications.

And by mind, I just mean the mental aspects of the world with no more interest in sharpening the boundaries or finding a criterion than in the case of the chemical or any of the others. Second, take language and linguistics. Well, again, we can focus attention on certain aspects of the world that fall under this informal rubric, and we can try to understand them better.

Uh, in the course of doing so, we may, uh, develop and apparently do develop a concept that is more or less re-resemble it– more or less resembles the informal concept of language and, uh, postulate that such objects, uh, are among the things in the world alongside of chemical bonds and electrical fields and photons. And the kind of inquiry I have in mind is ordinary scientific inquiry, what I’ll call naturalistic inquiry, which is the third term in the title. Uh, so a naturalistic approach to language and mind investigates certain aspects of the world, the linguistic and mental aspects of the world, as we do any others, uh, seeking to construct, uh, intelligible explanatory theories, taking as real whatever we are led to posit in this quest and hoping for eventual unification with what are sometimes called the core natural sciences.

Notice that I said unification and not reduction. Reduction is rather rare in the history of sciences. Commonly, the more fundamental science has had to undergo radical revision for unification to proceed.

The case of chemistry and physics is a recent example. Dogmatism aside, we have no idea how eventual unification of the study of cells and the study of language and mind might proceed, uh, nor do we know if these are the right categories, uh, to seek to unify, uh, nor do we even know whether the question lies within the, uh, our cognitive reach. Uh, there are, uh, plenty of dogmas about the matter.

Uh, for example, there’s the dogma that mental properties should be reduced to neural network properties. I’m quoting from the presidential address of the American Philosophical Association Pacific Division last week, Patricia Churchland. Uh, but we should recognize these as what they are, just dogmatic beliefs of the kind that have repeatedly proven false in the course of scientific inquiry, and they’re without any particular scientific merit in the present case.

So for example, comparable hypotheses have in fact been rejected as fallacious in some of the work studying. And nematodes, these are little worms with 300 neurons, uh, the wiring diagram and the developmental pattern is completely known, but there are serious questions as to whether the abstraction to neural networks is at all the right course. Uh, if the thesis about neural networks is in– understood as a research proposal, some proposal about neural networks and what they might conceivably do, well, that’s another matter.

Uh, we wait and see. If more is intended, it is simply irrational. Uh, in any event, a naturalistic approach accepts no such dogmas and proceeds as we do elsewhere in the sciences, which brings me to my last term, namely dualism, by which I simply mean here, non-naturalism with regard to mental and linguistic aspects of the world.

That is the do- the doctrine that we must abandon ordinary scientific rationality when we study humans above the neck, metaphorically speaking, uh, imposing arbitrary stipulations and a priori demands of a sort that would never be contemplated in the natural sciences or in other ways departing from normal canons of inquiry. That’s what I’m going to mean here by dualism. Side comment, there are serious questions that can be raised about naturalistic inquiry and how it should proceed, but we can put these aside in the present context unless it’s shown that they have some unique relevance to the study of the mental and linguistic aspects of the world.

For example, we can put aside skeptical questions if– unless some special relevance is shown here. The place to look for answers to questions is where they’re likely to be found. In this case, the place where answers are likely to be found is in the hard sciences where richness of understanding may provide some insight and some guidelines.

Uh, to raise such questions with regard to inquiries that are less advanced, uh, is just a form of harassment of emerging disciplines, not a serious pursuit in my opinion. Uh, so if certain kinds of questions can’t be answered for, say, chemistry or physics, uh, we can put them aside here unless some argument is given to show that they have a special relevance in this connection. And I don’t know that such arguments can be given.

Uh, the questions usually are raised in connection with psychology and linguistics and so on, but as I say, I think that should be regarded just as a form of harassment, not as serious inquiry. Uh, I’m also going to put aside questions of factual accuracy, which are obviously crucial, but in some different context. Well, with those terms, uh, I hope more or less understood, let me turn to the theses.

Uh, I had three in mind. One thesis is that naturalistic inquiry into language and mind is uncontroversial, and furthermore, it’s been pursued with a certain amount of success. Second thesis is that dualistic approaches are highly controversial.

And a third thesis is that a good deal of the most serious and thoughtful and influential work in philosophy of language and mind and the sort of reflective aspects of the cognitive sciences is dualist in the sense that I was using the term, including a good deal of what prides itself as, uh, on its hard-headed, uh, naturalist stance. Well, those are the three theses. Uh, to clarify further what I mean, uh, let me pursue the terminology a little bit and try to compare this deflated usage of naturalism that I want to employ with more conventional ones.

So for concreteness, take a recent paper by Cambridge University philosopher T.R. Baldwin called Two Types of Naturalism. Uh, his two types are both standard and both different from what I mean here. Uh, again, what I mean here is a kind of methodological naturalism, inquiry in the manner, in the manner of ordinary science, so that naturalistic inquiry into language and mind stands alongside of the study of chemical or electrical aspects of the world, of insects and mammalian vision and so on and so forth.

Uh, Baldwin’s two types of naturalism are the standard ones, and they’re different. Uh, one is epistemic natur– uh, and the other is metaphysical naturalism. So let’s have a look at those.

Uh, epistemic naturalism, he traces plausibly to Quine. Uh, um, metaphysical naturalism is also traced to Quine in a recent and useful review of Philosophy of Mind by Tyler Burge, which I’ll use as another convenient reference point in, in the discussion here. That’s in the centennial issue of, uh, Philosophical Review.

Very readable as very well as very expert. Uh, so let’s look at both of those notions, epistemic and metaphysical naturalism. Uh, epistemic naturalism to begin with.

Well, let’s start with Quine’s version, to which the modern, uh, notion of epistemic naturalism is traced, that opened the way to what’s called naturalization of epistemology. Quine stipulates that the study of language and mind must be incorporated within a branch of behaviorist psychology of no known scientific interest. Uh, this is a paradigm example of what I mean by dualism.

That is, nothing of the sort would be tolerated in the natural sciences for a moment unless some argument is given, and in this case, there’s no argument. It’s just stimulated, stipulated, as far as I know. Well, there are other, uh, versions of dualistic versions of epistemic naturalism, again, notably among those who want to make philosophy tough-minded, uh, the presidential address I mentioned is a case in point, but the issue is a bit to the side, so I won’t, uh, have some comments on it, but I’ll skip them.

Uh, Baldwin, uh, also discusses the epistemic naturalism of Hume, Thomas Reid, uh, others of the Scottish Enlightenment, which seem to me, uh, more reasonable in their conception. Although I think that even this work is a regression from earlier versions of epistemic naturalism that I think are better grounded, uh, which have been in fact, partially resurrected in recent work. However, whatever one’s evaluation of this may be, the Scottish Enlightenment, the Cartesians, the English Platonists, and others were proposing theories of mind that are readily construed as naturalistic in my deflated sense of the term.

So I’m suggesting no quantum– ah, qualms about that earlier enterprise of epistemic naturalism apart from questions of fact. The modern version, however, seems to be tainted by a deep-seated dualism, and hence I am suggesting qualms about that. Uh, well, turn to metaphysical naturalism.

That seems to me a much more obscure doctrine. Uh, Baldwin quotes Dan Dennett, uh, as saying that this… I’m quoting him, Dennett now, “This naturalization of philosophy is one of the happiest trends in philosophy since the 1960s,” uh, referring to the idea, again quoting Dennett, “That philosophical accounts of our minds, our knowledge, and our language must in the end be continuous with or harm– and harmonious with the natural sciences.”

Similarly, uh, Tyler Burge in his review describes metaphysical naturalism, sometimes called materialism or physicalism, as one of the few orthodoxies in American philosophy in the same period, roughly since the sixties. Uh, this is in Burge’s formulation, the view that there are no mental states or properties and so on over and above ordinary physical entities as identified in the physical sciences or by common sense. And one major strand of this effort, as he puts it, to make philosophy scientific is eliminationism or eliminative materialism, various other terms.

And this is again quoting Burge, “the view that mentalistic talk and mental entities will eventually lose their place in our attempts to describe and explain the word. There are other– world, sorry. There are other variants, uh, which are more or less the same.

Uh, this includes critics of the doctrine like Tom Nagel, who present the doctrine much in the same way. Uh, now, whether you propose it or criticize it, it seems to me a– there seems to me a problem. Namely, the doctrine, as far as I can see, is unintelligible, uh, and remote from the sciences.

In fact, the whole doctrine or the willingness to discuss it seems to me a form of methodological dualism or, under another interpretation, true and uninteresting. So let’s begin with eliminationism. Uh, this is, remember, the view that mentalistic talk and mental entities will eventually lose their place in our attempts to describe and explain the world.

Well, replace mental there by physical. So we now have the thesis that, uh, the view that physicalistic talk and physical entities will eventually lose their place in our attempts to describe the world. Well, if by physical here we mean what the sciences develop as they proceed, then the thesis is trivially and uninterestingly false, uh, and the same would be true of the study of language and mind.

If we mean, on the other hand, common sense discourse, the thesis is true but uninteresting for physicalistic entities. That is, uncontroversially, ordinary physicalistic talk has long ago lost its place in the sciences, and there’s no reason to expect mentalistic talk to fare any differently. So it is true, as Donald Davidson puts it, that we shouldn’t expect to find psychophysical laws relating ordinary talk about what people do and why, on the one hand, with principles like the law of falling bodies on the other.

But it’s also true that no one expects to find physico-physical bridge laws or whatever, uh, which involve informal descriptions of things like, say, a rock falling from the sky and rolling down a hill and ultimately hitting the ground below. uh, and that’s not a matter of imprecision. So for example, the statement that the sun set this evening at, you know, seven forty-three PM or whatever is precise enough and either true or false.

I presume false since I just made it up. But, uh, naturalistic inquiry has just led to very different ways of thinking about the world, which are not continuous with or harmonious with folk physics, again, a truism, and there’s no bridge laws, so eliminationism in this sense at least seems true of physicalistic talk as well. Well, there are traditional views, uh, claiming that we have some special access to the workings of our minds that we do not have to the external world.

That goes back hundreds of years. And recently, there’s a variant that holds that so-called folk psychology is different from folk mechanics or folk chemistry or intuitive views about things, uh, because of its a priori character and its intimate relation to notions of rationality, reasons, intentions, first-person perspective, and so on. There’s no doubt that the domains are different, but it’s not clear to me that they differ in what’s sometimes called anomalism in the sense of these discussions.

That is, insofar as scientific inquiry might undermine one’s conviction that the sun is setting or that objects are impenetrable while incidentally leaving such convictions in place in other parts of life, uh, it seems to me that it might, in principle, have similar effects on one’s convictions about the nature of beliefs, say, with regard to the role of rationality. Uh, much of what people believe about beliefs is a posteriori, so consider the debates about holism and innateness, uh, and we surely have a priori beliefs about rocks rolling down hills and storms brewing in the west, and so on. Uh, in any event, it’s not clear what the thesis of eliminationism is trying to assert, uh, and I’m not convinced, or I’d like to see how a coherent thesis could be presented.

Uh, matter has only become more obscure as we turn to other parts of the picture. So consider again Dennett’s notion, quoting him again, that “philosophical accounts of our minds, our knowledge, and our language must in the end be continuous with and harmonious with the natural sciences and other variants “of this,” as Burge calls it, “orthodox of, orthodoxy of” American philosophy.” Uh, first question to ask is, what are philosophical accounts as distinct from naturalistic accounts, particularly if they’re supposed to be continuous and harmonious?

More seriously, what are the natural sciences? Well, surely not what’s understood today. That may not be continuous and harmonious with tomorrow’s physics, at least if the past is any guide.

So perhaps some Peircean ideal. Well, that doesn’t seem very promising. Perhaps what the human mind can attain in the limit.

I think that leaves us even in worse shape. Uh, suppose that a 19th-century philosopher had demanded that chemistry in the 19th century, since, you know, Kekulé’s, uh, organic molecules and so on, that, uh, uh, that chemistry must be continuous with and harmonious with physics. Uh, if that was any more than a hope for eventual unification, it would have been a serious error.

As we know, there was no continuity and no harmony, uh, until physics radically changed. If Dennett’s thesis is that we should hope for unification of the science of the mental and the s– and other parts of science, no one could disagree, but it’s a thesis of very little interest and not a happy trend in philosophy. It’s just saying, you know, we hope that we’ll understand things better someday.

Uh, it seems to me that a deep-seated dualism, in my sense, underlies all of these discussions. And some belief, perhaps, that we have a grasp of the bounds of the physical, uh, and this grasp, uh, leaves the mental and linguistic aspects of the world somehow outside, in a way different from ordinary physicalistic talk, which is also outside for uninteresting reasons, and also different from the chemical, the electrical, and the optical, and so, so on, which were indeed outside of what turned out to be inadequate and erroneous physics very recently, in fact. Well, the point’s worth exploring a bit, I think.

There was once a fairly clear concept of the physical and material, namely what was called the mechanical philosophy in the seventeenth century. Uh, as is well known, Descartes argued that certain aspects of the world, in particular the normal use of language, lie beyond the bounds of the mentalist– of the mechanical philosophy. Uh, his arguments were actually rather reasonable, and his conclusion was correct, uh, though for reasons that he did not know.

The reasons were given by Newton a generation later. Uh, Newton showed that ordinary terrestrial and planetary motion lies beyond the bounds of the mechanical philosophy. Now, that is beyond what was understood to be body or matter.

Uh, recall that, uh, Newton did not exorcize the ghost, rather he exorcized the machine. Uh, Descartes’ theory of mind was left untouched, and what remained from the Newtonian revolution was, uh, quote, intellectual historian Margaret Jacob, it was a mechanical philosophy that relied heavily on immaterial and spiritual forces. Uh, now, Newt- this, uh, Newton’s invocation of the, what was called occult quality of gravity, was sharply condemned by leading scientists of the day.

So Leibniz and Huygens in their correspondence saw Newton’s work as a regression to the Dark Ages that they had somehow escaped from. Uh, and Newton himself seems to have been dismayed at these ghostly properties that he was forced to invoke. He even gave his reasons.

Uh, as he put it, quoting Newton, “The idea of an act of action at a distance through a vacuum is so great an absurdity that I believe no man who has in philosophical matters a competent faculty of thinking can ever fall into it.” Recall, philosophical means scientific here in our sense. Uh, so no one who has the slightest grasp of science could ever believe this nonsense, although the conclusion seemed unmistakable.

Uh, well, Newton reluctantly concluded that universal gravity exists, but the notion of the physical disappears. As Hume wrote later, while Newton seemed to draw off the veil from some of the mysteries of nature, he showed at the same time the imperfections of the mechanical philosophy and thereby restored nature’s ultimate secrets to that obscurity in which they ever did and ever will remain, meaning we no longer have an intuitive understanding of the notion physical. It’s gone, uh, and gone forever.

With the disappearance of the physical, the mind-body problem becomes unformulable, and the several versions of metaphysical naturalism and eliminative materialism and the like, uh, uh, also seem to lose any clear content, at least until some new notion of physical body replaces the lost mechanical philosophy, which hardly seems like a reasonable enterprise. I’m not aware that anybody’s undertaken it. Uh, so let’s take, say, a modern version, say, say, Quine’s version, uh, in its nineteen ninety-two formulation, so recent, uh, Uh, he presents the naturalistic thesis.

Remember, this is the source of the happy tendency in philosophy that we’re celebrating. Uh, the naturalistic thesis is, uh, that the world is as natural science says it is insofar as natural science is right. Right.

Okay. Next question. What is natural science?

Well, he gives an answer to that. Natural– Here’s the total answer.

Natural science is theories of quarks and the like. So the question is

(laughter)

what’s enough to be alike, uh, uh, uh, to be natural science? So for example, what about mathematical objects like fields of force, which are taken to be real objects because they push each other along through empty space? I’m quoting physicist Roger Penrose.

Or what about curved space that takes all definite structure away from anything we can call solidity? Quoting John Wheeler, another major physicist. Or what about infinite one-dimensional strings in multidimensional space?

Uh, twenty-seven dimensions last time I tuned in. Uh, all of this is apparently like enough, uh, by Quine’s standards to count as physics. Uh, well, suppose it’s argued further that on quantum theoretic grounds, uh, at a very– I’m quoting, \”at a very deep bottom, there is nothing but bits of information, answers to the questions that we pose to nature.\” That’s John Wheeler again in a book that just came out.

Uh, by Quine’s standards, that should be like enough, uh, hence physical. Recall that physics is first philosophy. We don’t second-guess the physicists.

Uh, suppose that a substantial explanatory theory is constructed involving something like Lord Herbert’s common notions, uh, Humean ideas, Fregean thoughts, computational principles and states, and so on. Is that like enough? Well, Quine’s practice indicates that it’s not.

Uh, but if so, that just seems to be irrational dualism, at least unless until some argument is given. A naturalistic approach will simply adopt the post-Newtonian course, recognizing that we can do no more than seek the best theoretical account of the phenomena of experience and experiment, seeking no ultimate explanation. It’s quoting I. Bernard Cohen, a leading Newton scholar on the, uh, legacy of the Newtonian revolution.

Uh, that– what I just said was indeed the conclusion of some leading eighteenth-century scientists. So, for example, the eminent chemist Joseph Priestley concluded that human thought and action are properties of organized matter, like powers of attraction and repulsion, electrical charge, and so on. In this case, the organized matter in question is the brain, not the foot or the kidney.

Uh, that conclusion, in fact, had been reached by La Mettrie a generation earlier, but his work was suppressed. In fact, it was suppressed almost to the present day. Uh, and it’s pretty hard to think, at least I can’t think of a coherent alternative to that conclusion.

With the collapse of any independent notion of the material or the physical or a body after Newton, what else can we say about various aspects of the world than essentially that? Uh, accordingly, it’s hard for me to see why it’s now resurrected more than two centuries later as a bold and innovative, innovative hypothesis. the hypothesis that capacities of the human mind are in fact capacities of the human brain.

Again, I’m quoting the same presidential address. That sounds to me as about as close to a truism as one can imagine, and as uninformative as truisms typically are, since the brain sciences hardly even know where to look, contrary to Churchland’s rather enthusiastic report. Uh, the current situation is that we have good and improving scientific theories of aspects of mind, including language, but only the most rudimentary ideas about the relation of any of this to cells or whatever turn out to be the relevant aspects of the brain.

It could be something quite different than we think about, at least if the history of science is any guide. Well, let me take a concrete example. Uh, a good deal is known about computational principles of the language faculty, the brain.

For example, we have pretty good evidence for fairly subtle distinctions among different kinds of what’s called deviance, that is departure from one or another principle of language, for those of you who know the game, uh, subjacency violations, ECP violations, uh, violations of the specificity condition, semantic deviance. Doesn’t matter if you understand what I’m talking about. These are just different kinds of departures from certain principles.

Uh, and they, they, they have a meaning inside the explanatory theory of language and evidence. Well, recently, there’s been quite intriguing work at the Salk Institute in, uh, Helen Neville’s lab on electrical activity of the brain. And researchers there have been able, pretty surprisingly, to find correlates in event-related brain potentials, ERPs so-called, to categories of deviance, namely the categories that are discovered in the computational theories.

There’s more recent work at the Max Planck Institute in Nijmegen, which has found a distinctive kind of electrophysiological response to syntactic violations on the one hand and semantic violations on the other that improves these results. Uh, still, these results are interesting, but they’re kind of a curiosity. Uh, the reason is that there’s no theory of electrical activity of the brain and its causes and character, no reasons why one should find these results, not all sorts of others.

Uh, the computational theories, in contrast, are much more solidly based, at least from the point of view of scientific naturalism. The situation is somewhat reminiscent of chemistry and physics, uh, a century ago, though the analogy naturally should not be pressed too far. Well, a naturalistic approach to language and mind will simply proceed to improve, try to improve each of these approaches, the ERP approach and the computational approaches, hoping for more meaningful unification.

Uh, a dualist, on the other hand, will see something deeply problematic in the scientifically more solid theory, namely the mental one, and will worry about problems of eliminationism and physicalism that have yet to be formulated in a coherent way. Uh, the dualist tendency dominates discussion and debate, uh, in fact, is virtually presupposed, which seems to me a curious phenomenon and one that merits explanation. Uh, well, I wanna keep then to methodological naturalism, uh, putting aside metaphysical naturalism and its variants as apparently meaningless, and epistemic naturalism as reasonable in its traditional, if not modern version.

Uh, next question: Where does the naturalistic approach take us in the study of language? Well, putting aside questions of the factual accuracy, which I’m not going to try to justify, it seems to me to lead to something roughly like this. Uh, we take a naturalistic approach.

We therefore begin with objects that biology presents to us as natural objects, say, Tom Jones, which biology tells us is a natural object. Uh, and we’re interested in particular aspects of Tom Jones, namely the linguistic aspects. We discover pretty quickly that these involve Jones’s brain, and indeed perhaps more, like his articulatory organs or perhaps his limbs and other components of the body that may well indeed have specific language-related design.

Uh, however, let’s keep to the brain, which is clearly fundamental. Some aspect of Jones’s brain is dedicated to language, it appears. Let’s call that the language faculty.

Uh, we furthermore have very good evidence that the language faculty has at least two different components. One of them, a cognitive system that stores information in some manner, the other various performance systems that access and make use of this information for articulation, interpretation, speech acts like talking about the world or asking questions or telling jokes, and so on. Uh, the cognitive system and the performance systems have initial states that are fixed by genetic endowment.

Uh, it’s generally assumed that the performance systems are language independent, not meaning– I don’t mean not specific to language. They do seem specific to language, but language independent in the sense that any state changes that take place are either internally directed or the result of language independent events like injury, for example, not the result to exposure to one or another language. Uh, the reason for that assumption is essentially ignorance.

It’s the simplest assumption, and we don’t know that it’s wrong, though it probably is. So therefore, it’s just assumed to be true because we don’t know anything else. Uh, well, on that simplifying assumption, uh, we then look just at the cognitive system of the, at the language faculty, its initial state, and its later states.

Uh, it’s plain in this case that state changes reflect linguistic experience. So for example, English is not Swahili, at least not quite. Uh, a rational Martian scientist who was looking at us the way we look at, say, frogs would probably find the variation rather superficial, concluding that there’s just one human language with rather minor variants.

It looks like a plausible conclusion, in fact. Uh, in any event, the cognitive system of Jones’s language faculty does change in response to external events, changing state until it pretty much stabilizes, perhaps as early as six to eight years of age, some recent work indicates, uh, which would mean, if true, that later changes, which indeed have been found up to about puberty, uh, are probably inner-directed, not the results of specific experience. Uh, well, let’s call a state of the cognitive system of Jones, of the language faculty, an I-language.

Uh, I here is to suggest internal and individual, since this is a strictly internalist, individualist approach to language. Uh, if the cognitive state of Jones’s language faculty is in state L, uh, we can say that Jones has or knows or speaks the I-language L. Uh, so far I’ve kept to Jones, Uh, his, uh, uh, uh, brain, its language faculty, its components, uh, all of which are natural objects. So let’s now turn to Smith.

Uh, we discover that the initial state of Smith’s language faculty is virtually identical to that of Jones. Meaning we find that given Jones’s experience, Smith would have attained Jones’s I-language. In fact, that seems to be true very broadly across the species, meaning that the initial state is a species property, and to a very good first approximation, we may therefore speak of the human language faculty and the various I-languages that are its manifestations as natural objects.

We’re still within the domain of ordinary scientific naturalism. Uh, if Jones has the I-language L, he knows many things. So I suppose Jones is one of us.

Uh, he knows that the word house rhymes with mouse. He knows that the phrase brown house consists of two words. Uh, the two words are in the formal relation of assonance, meaning same vowel but different consonants.

uh, that the phrase is used, uh, to refer to a structure that’s designed and used for certain purposes and that has a brown exterior, not interior. So if you paint the house brown, you’ve painted its exterior brown, and a lot of other things of that sort. Uh, we would like to know how Jones knows things of that kind, and the answer seems to work something like this.

The I-language seems to consist itself of two components, one a computational system, the other a lexicon. The lexicon is a collection of lexical items, each of which is some complex of properties, usually called features, like the property begins with a bilabial stop or is an artifact. Uh, the computational system possibly is uniform virtually.

It doesn’t vary across languages. This can’t be quite true at the sort of phonetic level. There’s some variation there, but it seems to– It’s very plausible.

It’s a plausible idea that it’s true elsewhere. Uh, if that’s the case, language variation is in the lexicon, and indeed, it seems to be in narrow subparts of the lexicon, rather form, formal aspects of it. Uh, the I-language has lots of properties.

One of its clearest and best confirmed properties is that it’s what’s called a generative procedure, a cogn– co– a computational system which characterizes an infinite class of expressions. Uh, each expressh– expression is a complex of properties, uh, drawn from the lexicon, except again for the phonetic properties. Uh, well, notice that these are proposals, true or false, of ordinary natural science, proposals about the states of certain natural objects and their properties.

In this case, states and properties of a kind that are unusually clear and well understood. So we’re still in the domain of non-controversial, except for questions of fact. Uh, the I-language, again, a state of the cognitive system of the language faculty, is embedded in performance systems.

Uh, these have access only to the expressions generated by the I-language. Recall the expression is a complex of properties which we can think of it as providing instructions to the performance systems. Uh, furthermore, some of these properties seem to provide instructions only to the articulatory and perceptual aspects of the performance systems.

So one element of the expression then seems to be a certain symbolic object, sometimes called phonetic form, let’s call it PF, which consists of instructions to the articulatory and perceptual systems, and the elements that constitute it are called phonetic features. Uh, the expression also provides instructions to other systems, conceptual and intentional systems, and maybe the same set to both, to all, because there may be many of these. Uh, one element of the expression then would be another symbolic object, sometimes called LF, which is to suggest logical form, but without connotations, uh, consisting of instructions to the conceptual intentional systems.

Uh, its elements are sometimes called semantic features on a par with phonetic features. So if that’s correct, the language generates expressions, each of which is a pair, uh, of a PF and an LF representation. The PF is a complex of phonetic features and the LF a complex of semantic features.

Each of them provides information and instructions to certain performance systems. Notice as we proceed, we are becoming more controversial, but empirically, because things don’t have to work like this. It could easily work in a thousand other ways.

And all of this requires evidence, and there is evidence of varying kinds. Well, this picture induces a partitioning on the features of the lexicon itself. So the feature begins with the bilabial stop would be a phonetic phonetic feature which survives only to PF.

the instructions for the articulatory perceptual system, uh, whereas the feature artifact in the lexicon is a semantic feature surviving only at LF, the instructions to the conceptual and intentional systems. And then there are others that enter into the computational system but don’t survive at the interface to other systems, formal features, if you like. Uh, certain formal properties of PF and LF are interpreted by the performance systems in particular ways, for example, as rhyme or as assonance or as entailment.

Uh, if the I-language were embedded in different performance systems, it would be interpreted in different ways, perhaps as instructions for locomotion, let’s say. Uh, the I-language, in other words, is a real object in the natural world with its specific properties, and the performance systems are as well. Well, how does that interpretation work?

As we proceed, we get more controversial. So let’s take the phonetics side, the, uh, the articulatory and perceptual systems and its elements. Let’s take the element ba, let’s say, which we idealize away from individual speakers.

Well, here’s one theory. Ah, one theory of how the percept– the performance system works on that side is that Ba picks out an object in the world, an external object, let’s say some set of noises or motions of molecules, and we could now spin out some kind of theory of perception and of communication and so on in these terms. Uh, it’s conceivable, though it’s total nonsense.

That’s not the way it works. No one has ever proposed such a crazy theory. No one defends it or even considers it.

Although with really heroic dedication, uh, you could defend it. There’s no doubt of that. Uh, in fact, what happens is apparently is that Ba provides instructions to articulatory and perceptual mechanisms which we seek to discover, and we find that they’re pretty complex, like it might involve a motor theory of perception, for example.

Well, turning to LF, the other side, the relation, the other interface of the language faculty with other systems, uh, conceptual, intentional. Uh, so let’s take things like, say, brown house or London or other elements of the– now looked at from their semantic side, instructions to the computational intentional systems. Here’s a theory which is parallel to the absurd theory of phonology that I j– of performance that I, that I just described.

Um, the theory is that there’s an entity in the world or an entity in the world, as it’s believed by Jones to be, that stands in some fixed relation to these things, like in some fixed relation to London or brown house or whatever. Uh, we can call that relation reference or denotation. Notice that that’s a technical term.

That is to say, it means what we say it means. We have no intuitions about it any more than we have about, say, angular velocity. It’s a technical term.

You have some intuitions, you better throw them away because they’re not– You’re not supposed to. It’s invented.

There is no such notion in any natural language, as far as I’m aware. Uh, that’s why Frege, uh, had to invent terms. Bedeutung for Frege didn’t mean what it means in German.

Uh, uh, now, this means that there can’t be any what’s called intuition pumping, you know, sort of pumping your intuitions about these things, uh, about reference, because it’s an invented technical notion, which means exactly what you say it is. Uh, uh, it’s, uh, in the s– in the externalist literature, for those of you who are familiar with this, there’s plenty of efforts to find out what your intuitions are about these things, as like in all the Twin Earth thought experiments, you can’t have any, you know? It’s like having intuitions about angular velocity, so it doesn’t m-matter.

Or like, say, undecidability. And if you’re studying metamathematics, and you use your intuitions about undecidability, you didn’t understand the topic. You’re not supposed to have it.

You know, it means what that guy up there says it means, period. No intuitions. Uh, and, uh, that’s true, I think, of the entire externalist literature.

As far as I can see, it’s mostly meaningless because it’s seeking to explore your intuitions about something you can’t have intuitions about. When you try to rephrase these discussions in terms of what you do have intuitions about, like talking about something or referring to something, as far as I can see, it all collapses. Nothing– none of the arguments go through.

Uh, you can try that. Well, anyhow, if you pursue this idea, you know, that there are external objects out there connected one-to-one with expressions like London or Brown House, you could spin off a theory of use and interpretation and communication. But as in the perceptual-articulatory side, it just seems to be the wrong way to proceed.

Although again, with sufficient dedication, you could doubtless defend it. Uh, the big problem is that the objects that you’re in– have to invent to stand in a relation to terms would be extremely queer. Uh, for example, they would often have contradictory properties.

Well, let’s say take London. Uh, if there– if one tries to establ– inven– you know, to postulate some entity standing in this one-to-one relation to the phrase London, uh, it’s a very strange object. Uh, on the one hand, it’s obviously concrete, so you could…

London could be, say, destroyed with a bomb, so it’s plainly concrete. On the other hand, it’s also completely abstract because London could also be destroyed by a stroke of the pen, let’s say a decision of some bureaucrats in Brussels, you know, to draw boundaries differently. In fact, London could even be destroyed and rebuilt somewhere else, and under certain conditions it would still be London.

You know, just decided it’s too polluted, let’s move it up the Thames ten miles or something. Uh, these, uh, so it’s– And if you really pursue it, it has extremely weird properties.

Furthermore, that generalizes to everything, as far as I know. I don’t know of any examples for which it’s not true. Uh, I’m not going to pursue this for lack of time.

Uh, the basic observation, in fact, is quite familiar from eighteenth century critiques of the theory of ideas, uh, which were revived by Thomas Reid and people like that. Uh, they were revived in modern ordinary language philosophy. For some reason, which I don’t quite understand, they’ve been largely forgotten in recent work but apparently they’re valid, and it seems to me they offer quite rich possibilities for inquiry which to some extent are now being pursued.

Uh, in any event, it seems that implausible, as far as I can see, that the– on the performance– that if you’re looking at the performance systems, that there is this invented relation, you know. It doesn’t seem to exist as far as I can see, and by pursuing the belief that it exists, you are led to very strange conclusions with no explanatory force. They don’t seem to do anything except allowing some kind of computations to run through which don’t explain anything.

Uh, what it looks like is that the semantic side, so-called, is rather similar to the phonetic side. That is, these semantic features provide certain perspectives for looking at the world in ways that focus on particular regions in quite complex space of human interests and concerns, uh, directing attention to one or another aspect of these, in a manner that’s surely innately determined and completely inaccessible to awareness by humans, rather like the phonetic properties which are quite the same. Incidentally, it’s a conceptual possibility that some creature might exist.

It is probably a biological impossibility, but it’s a conceptual possibility that there might be some other creature, uh, which has complete awareness of all these operations, uh, and the way its visual system works, and so on. Uh, that would be an extremely fa– interesting factual discovery, but from a naturalistic point of view, it doesn’t seem to me to have any further consequences. Uh, now, side comment again, it could well be the case that in one particular human enterprise, namely the kind we call science, uh, it’s possible that we do try to construct as a kind of goal symbolic systems that have the property, uh, uh, that terms pick out things in the world, in fact, that have the sort of Fregean perfect language properties generally.

So, say, take the term quark or I-language. Those, I think, are invented with this goal in mind that they actually pick out something in the world. Here we aim for a kind of a Fregean ideal.

And we may even speak of symbolic systems that are contrived in this quest as scientific languages, but that’s just a metaphor as when we say that airplanes fly, they’re not at all like languages. They differ radically from natural languages in, uh, all their aspects, syntactic, semantic, and everything else. Uh, natural languages again are just natural objects in the world, states of some system of the brain, which are what they are, and they don’t look like Fregean languages, so it seems.

Well, pursuing this course, we carry out a naturalistic study of linguistic aspects of the world, which is part of mental capacity, uh, to, uh, to a large extent, at least. Its mental aspects may involve others. Uh, naturally, we try to find con-connections with other naturalistic inquiries.

Uh, so the, uh, the scope of this internalist inquiry is quite broad. It includes a large part of the study of language and its use. It includes much of what’s called sociolinguistics and psycholinguistics, or at least they presuppose it.

Uh, it doesn’t include all kind of other topics for it doesn’t, doesn’t include the topic whether, uh, some standard language should be selected for school instruction or how language is used for coercion and control, and so on. In fact, it has nothing to say about these topics, despite what many linguists claim. Uh, you, you can drop all the linguistics and study those topics quite readily.

Uh, the, uh, notion of I-language seems fairly close to the ordinary notion of language, but that’s not a very important fact any more than it’s important to determine how close energy in physics is to the standard notion. But just as a matter of fact, it seems to be sort of close in some respects, where we understand the traditional– the common sense notion of language in one traditional way. In fact, as a way of speaking, if that’s what a language is, well, okay, an eye language is sort of like that.

Uh, it also differs. So for the, for example, the, the ordinary notion of language has quite intricate normative and teleological elements. So for example, suppose we say that Johnny, who is a three-year-old, uh, has a partial knowledge of some language that he will approach as time goes on, though what he’s speaking now isn’t any language at all.

That’s in fact the way we use the term. You talk about a three-year-old, you don’t say he has a language which he knows perfectly. What you say, he has partial knowledge of some goal that he’s approaching.

Well, these informal notions, like all informal notions, are perfectly okay, but we don’t expect to give any coherent account of them, uh, within an inquiry that seeks to understand and explain what’s going on in the world, what Johnny is up to, for example. Similarly, the theory of vision, uh, is not interested in, say, Clinton’s vision of a free market future or something. That’s not part of the theory of vision, and these normative teleological aspects of the ordinary notion of language, whatever they may be, uh, are not part of the naturalistic inquiry into understanding what Johnny is doing.

So the pu– the I-language, again, is purely internalist, individual. The inquiry into it is uncontroversial, as far as I can see, rather extensive in scope, including just about anything we can investigate within the linguistic aspects of the world, and indeed seems to be presupposed by other inquiries. Well, let’s turn to another question.

Are there different externalist notions of language? For example, take the notion common language to which appeal is constantly made in philosophy of language and mind and the cognitive sciences, and there certainly is such a notion in ordinary discourse. So we say, for example, that Hans and Maria speak, uh, a common language, German, even though they can’t understand– neither of them can understand one word that the other says, while Hans and Jan speak different languages, namely German versus Dutch, although they understand each other perfectly living on opposite sides of the border.

We do say things like that. In the empirical study of language, there’s no such notion as German or Dutch in this sense, except as shorthand conveniences. Uh, so there is then no way, as far as I can see, to understand somebody like Michael Dummett when he says that a person has a partial and partially erroneous knowledge of his or her language, namely the common language.

Similarly, as far as I can see, there is no way to understand the thought experiments that are suggested by Hilary Putnam, Tyler Burge, Saul Kripke, a lot of other peoples, all of which crucially rely on this completely unexplained notion. Now, here I mean the term understand rather narrowly. Of course, you sort of get the sense of what they’re up to.

Uh, what I mean is give a coherent account of what they’re saying. Uh, that will be possible only when they give us some indication, let’s say even a hint as to what a common language might be, or for that matter, what a norm is, or a convention, or a set of dispositions to respond, or a cultural practice, or any of these other terms that are thrown around as if they’re supposed to mean something. Uh, that has not been done as far as I’m aware, and in the empirical study of language, it’s always been assumed to be misguided, uh, kind of a misinterpretation of ordinary usage.

Uh, the unargued beliefs to the contrary seemed to me a form of dualism in my sense. That is an insistence that in the mental realm we must depart from, uh, canons of rational inquiry over canons that are adhered to in a spectrum ranging from physics to anthropology, uh, but somehow we have to depart from them here for unexplained reasons. That’s dualism.

Uh, well, returning to the theses Thesis one is that naturalistic inquiry into language and mind is uncontroversial, except in the normal sense of empirical inquiry. Second, dualistic approaches are highly controversial. They require argument, which I think is hard to find.

And thirdly, a large part of the most serious and influential work about these topics is radically dualist, including, uh, much of what calls itself very hard-headed and talks about how scientific it is and so on. I’ve given a couple of examples. If there was a lot of time, I would go on to a lot more.

Uh, but let me just tell you the kinds of topics that I think could fall under this discussion. One is the whole debate about what’s been called innatism. Another is the whole discussion of philosophical behaviorism, say Quine’s latest version in his book, Pursuit of Truth.

Uh, another is the highly influential radical translation paradigm, uh, including the use of that paradigm in an approach to the theory of interpreters. Uh, all of this seems to me radically non-naturalistic, ah, dualist in ways which, uh, I don’t think are intelligible. At least I can’t figure out how to make them intelligible.

Uh, or, uh, I’m sorry just to throw these comments around without discussion, but time’s short. Uh, so I’m saved, in other words. Uh, let me just take one, uh, example to comment on to illustrate, which is much broader.

Uh, that’s the belief that the gaps in unification, which doubtless exist, are a crisis for cognitive science, psychology, and linguistics, maybe worse, uh, and, uh, the efforts to overcome them and, uh, various comments on that, a few things. But I’ll just say a few words about that, uh, and about the attempts to overcome the so-called crisis. Uh, remember again, naturalistic inquiry into the mind yields theories about the brain, its states and its properties, for example, theories about I-language.

And again, the unification problems are very far from resolution, so no one knows how to relate these theories to properties of atoms or cells or neurons or whatever may be relevant. We probably don’t even know what’s relevant. Uh, this disparity wi– this is a situation very familiar in the history of science.

For example, it was true of chemistry up until Pauling’s, uh, use of quantum theory in the 1930s to explain the chemical bond. Up until then, it was a complete gap because physics was just wrong. The disparity between theories of the mind and what’s been learned about neurophysiology, I’m now quoting, “Creates a crisis for those who believe that the nervous system is precise and hardwired like a computer.”

That’s, uh, Nobel Prize-winning biologist Gerald Edelman in a recent book. Uh, he also points out that there’s the same crisis for connectios- connectionist and neural nets theories. Uh, his argument is that the varied individual histories of the nervous system and the enormous individual structural variation of brains, I’m quoting, “Provide the coup de grâce, uh, to, uh, multiple coups,” he says, “to attempts to construct computational or neural net theories of the mind.”

And apparently Edelman takes this to be true no matter how successful such studies might be now or ever by the standards of the sciences: explanation, insight, opening new inquiries, and so on. Uh, well, by similar logic, one could have argued– Maybe he’s right, I don’t know.

But what– But, but the logic is very flawed, you know. The logic– By similar logic, one could have argued not long ago that there’s a terrible crisis for the study of matter and organisms in terms of colors, valence, the solid state, a multitude of other properties.

And earlier, one could have argued that there’s a terrible crisis for the investigation of electricity and magnetism, uh, planetary and celestial motion, and so on. In fact, virtually the whole of science was in a terrible crisis because of the huge gap between what had been learned about these topics and the principles of the mechanical philosophy, or in fact even much more recent physics, uh, that would– by the same logic, you could argue that. Uh, the problem that Edelman perceives is quite real, but seriously misconstrued, I believe, in a common but misguided way.

What about the enormous variation in structure of brains and experience? No doubt that’s true, but it doesn’t tell us anything. Uh, not many years ago, languages appeared to differ one another– from one another as radically as neural structures do to many a trained eye today.

Uh, and they were considered, in fact, merely reflections of infinitely variable experience. In fact, any complex system will appear to be a hopeless array of confusion before it comes to be understood and before the principles of its organization and function are discovered. Uh, Edelman argues that introducing considerations of meaning will somehow, uh, overcome these alleged problems.

Uh, but, uh, that’s just a very mistaken view of semantics. Simple semantic properties pose all the problems that Edelman perceives in syntactic theories and constructions. That is, they appear to be rule-governed, sharply delineated, fixed in relative independence of experience and known aspects of neural structure.

Hence, they too induce the crisis caused by the gap between the apparent algorithmic digital character of language and the observed variability and continuous flux of individual experience and neural structures. So we face a typical problem of unification in the sciences, which may, as often in the past, require that what someone might consider the more basic science be fundamentally recast if it’s to be integrated with successful explanatory theory at other levels. Recast because it just isn’t understood.

You don’t know what to look for there. Uh, so it looks like a mess, like anything that you don’t understand. Uh, well, various remedies have been proposed to deal with this alleged crisis.

One is the idea that the mental is just some way of talking about the neurophysiological. That could turn out to be true, but at the moment, it’s a hypothesis about the neurophysiological and not a characterization of the mental. In fact, the shoe is on the wrong foot in the light of what is at all understood now, at least.

Uh, another is the version of, uh, eliminative materialism, so-called, which holds that we should concentrate on neurophysiology. Well, that has all the merit of a proposal some time ago that chemistry should be abandoned in favor of the study of solid particles in motion. Uh, there’s a, uh, no merit that is.

There’s a substantial literature asking what it would imply if neural net and connectionist models could account for the phenomena that have been explained in computational terms. Such discussion may appear to be naturalistic in temper, but that’s hardly clear. A few biologists would be much impressed by the suggestion that unstructured systems with unknown properties might someday make it possible to account for development of organisms without appeal to complex theories about concentration of chemicals, the cell’s internal program, production of proteins, and so on.

Maybe, but that’s not a very interesting proposal. In some domains, language in particular, successful theories are commonly of the computational type. That’s a fact that has caused considerable uneasiness.

And to relieve it, uh, computer models are often invoked to show that we have robust, hard-headed instances of the kind, and then psychology is supposed to study software problems, and now we’re supposed to feel happy. But that’s a very dubious move, as John Searle, for one, has pointed out. Artifacts pose questions of a very serious kind.

In fact, those questions don’t arise for natural objects. So whether some object is a key or a table or a computer, that depends on the designer’s intent, on standard use, on mode of interpretation, and so on. And the same considerations arise when you ask whether the device is malfunctioning or following a rule and so on.

Point is there’s no natural kind. There’s no normal case. It’s just whatever you say it is.

Uh, um, now, these questions don’t arise in the study of, say, organic molecules or the wings of chickens or the language faculty or other natural objects which just are what they are. Uh, the belief that there was a problem to resolve beyond the normal empirical ones reflects an unwarranted dualism, and the proposed cure is a lot worse than the disease. Uh, related to these matters is the much-debated question of whether machines can think and all of its variants.

A question that was opened in the modern period, of course, it’s classic, uh, by Turing’s nineteen fifty paper, famous nineteen fifty paper, in which he proposed what’s called the Turing test for machine intelligence. Without going into details, it’s basically, can some machine fool an observer into thinking it’s a person? Actually, there isn’t any such thing as the Turing test.

There’s ten zillion different tests depending on how you set this up, all of which are totally arbitrary and pointless. Uh, but, uh, th-there are these things called the Turing test. Uh, now, that question has aroused quite lively discussion and controversy, which incidentally is quite contrary to Turing’s intentions.

Turing, I’m quoting him in his 1950 paper, uh, he regarded the question as too meaningless to deserve discussion. Uh, although he said that in half a century, meaning year 2000, conditions might have changed enough for us to alter this usage, just as languages, uh, have commonly, though not always, uh, changed enough to use the metaphor flying for airplanes, although there’s no empirical question whether airplanes fly like eagles. If we say they do, okay, if not, not.

Uh, Turing, in other words, seems to have agreed with Wittgenstein about the pointlessness of the discussion and the debate that’s ensued until today over whether machines can in principle think or play chess or understand language or do long division and so on, and about how we could empirically establish that they do or don’t, or whether robots can reach for objects and pick them up, can commit murder and so on, or for that matter, whether brains can think and play chess and so on. The answer is no, uninterestingly, if we’re using natural language, and yes, equally uninterestingly, if we choose to adopt certain metaphors, perhaps as part of a change in the way of thinking about the world, which could be interesting from some other point of view. Well, I think myself that the Turing stand was correct.

Uh, the questions are too meaningless to deserve discussion. They’re questions of decision about sharpening and altering usage, not questions of fact, just as there never was a question about whether airplanes can fly to London like eagles or if they can fly that far, or whether submarines really set sail but do not swim. Yeah, if that’s what you want to say.

No, if not. There’s no empirical question. And that conclusion remains, and as far as I can see, if you add further ease– further sensory conditions or criteria, as has been per– like performance criteria, has been proposed, say, by Stevan Harnad, for one example.

Now, there’s a completely separate issue which shouldn’t be confused with this, and that’s whether simulation might teach us something about the process that’s simulated. So for example, could a chess-playing program teach you something about human thought? In that case, the topic seems to be very badly chosen, at least for naturalistic inquiry.

The answer is almost certainly not. But in principle, simulation could provide insight. And in fact, that much was understood centuries ago, though the classical discussion didn’t fall into the errors of the modern revival, post-Turing revival.

So when Jacques de Vaucanson, famous eighteenth-century, you know, skilled craftsman. When he amazed observers with his remarkable contrivances, he and his audience were concerned to understand the animate systems that he was modeling. He had a famous clockwork duck, for example, uh, which was supposed to simulate the digestion of a duck.

But the idea was not to see if you could fool the audience into thinking it was a duck. In fact, he made it transparent to make sure he couldn’t fool the audience. Uh, the idea was to see if he could learn anything about the digestion of a duck.

Well, that’s the pur-purpose of simulation generally in the natural sciences. There’s very little, if any, role here for operational tests of one or another sort, which never come up in the sciences as far as I know. Uh, and there’s surely no point in a debate over whether Vaucanson’s duck really digests.

Uh, in this regard, there’s been considerable regression, in my opinion, in the modern cognitive revolution. Although Turing himself, interestingly, was quite clear about the matter, and I think we can regard the whole discussion as another reflection of the methodological dualism that seems to me to underlie a good deal of contemporary thinking about these topics. Well, let me s-mention that this methodological dualism differs from traditional pre-Newtonian metaphysical dualism.

That was not– That was meth-metaphysical dualism, but it was not methodological dualism in my sense. Rather, it was methodological naturalism.

It was a plausible, reasonable theory about the world based on the mechanical philosophy Soon shown to be false. Uh, it led to factual theses shown to be incorrect with Newton’s demolition of the me– mechanical philosophy. In contrast, contemporary methodological dualism seems to me to have no redeeming features, or so I’ve been trying to suggest.

Well, to finish, uh, these remarks barely skim the surface of what seem to me dualist elements and much of the most sophisticated and interesting and influential thinking about language and mind. Uh, if correct, these, uh, elements should either be justified or abandoned. The critique of naturalistic approaches that’s been given also seems to me to be flawed.

In any event, it seems to me there’s good reason to examine more closely doctrines that have, have been assumed much too casually, and if they do not withstand such analysis, to ask a further question, namely why they seem so compelling. Thanks.

(applause and cheering)

We have about, uh, ten minutes- I was told that there’s a couple of microphones around here, if anybody wants to say something, but I don’t see them. We have about ten minutes for questions. If people- Could you- -can’t make themselves heard, they can use the microphone. Where you are?

[01:12:25] AUDIENCE MEMBER:
Hello? Right here.

[01:12:26] NOAM CHOMSKY:
Over where?

[01:12:27] AUDIENCE MEMBER:
Professor.

[01:12:28] NOAM CHOMSKY:
Where am I looking?

[01:12:29] AUDIENCE MEMBER:
Here.

[01:12:29] NOAM CHOMSKY:
Oh, you’re there.

[01:12:30] AUDIENCE MEMBER:
Professor, yeah.

[01:12:30] NOAM CHOMSKY:
Yeah, okay.

[01:12:31] AUDIENCE MEMBER:
Sorry.

[01:12:32] NOAM CHOMSKY:
Okay. Voice is coming from the stratosphere, you know.

[01:12:35] AUDIENCE MEMBER:
Yeah, apparently. Um, this is just a, a preliminary question to establish a dialogue. Um, so I don’t wanna pretend that it’s exactly precise, but, um, if a dialogue is what we’re after, I’ll, I’ll phrase it like this.

Um, if wherever we look, or if there’s the possibility of interest wherever we look, um, does it seems to me that there is, there’s hope for me meaningful activity wherever we look. Um, if this is so, then presumably if ethics– presumably ethics, communication, and the and the self also hold that same possibility of interest. Um, if, if activity is solicited by these interests and are being pursued, then they’re being pursued everywhere at all times by everybody here, um, just by living.

Because it’s– I mean, we’re being solicited to live and I– And to communicate and to, uh, well, you know, we’re, we’re solicited first primarily to look at the soul, to look at the self, and then from there move on to to, you know, communication with the person next to us and- And yeah, and if I, if it’s making sense.

[01:14:10] NOAM CHOMSKY:
Well, look, I, I kind of get your point, but I’m, I’m not saying anything about those problems

(laughter)

I mean, you know, they’re, the-

[01:14:16] AUDIENCE MEMBER:
But there is something to be said about that.

[01:14:17] NOAM CHOMSKY:
Look, there’s all kinds of things in the world. You know, naturalistic approaches, sort of scientific approaches are extremely narrow. You know, they can do some things very well.

Most things they can’t do at all. So if you look at the questions that were, say, raised by the Greeks, for example, and with regard to most of them, we’re about in the same state of ignorance that they were. Uh, there’s a little narrow point of light that broke through around the seventeenth and eighteenth centuries, uh, which is called the natural sciences, and it led to enormous understanding using particular capacities of the human mind in very specific areas.

Most of the things that work concern us and interest us in life, that has nothing to say about. You know, people may pretend that it does, but it basically doesn’t. Uh, those problems you sort of deal with the way you, the way you live, you know.

[01:15:03] AUDIENCE MEMBER:
Well, I, I, I’d have to disagree there.

[01:15:04] NOAM CHOMSKY:
Okay.

[01:15:05] AUDIENCE MEMBER:
Because- Yeah. I’d have to disagree because-

[01:15:08] NOAM CHOMSKY:
Yeah, could be wrong. That’s the way it looks to me. Yeah. Uh- If anybody… I think the idea would be to go to the microphones. I can’t really see too well because the lights- Professor Chaffee? Is somebody else at a microphone somewhere?

[01:15:27] AUDIENCE MEMBER:
Where? I have a m- There. Do you have any advice for, uh, people like myself who are interested in teaching English as a foreign language in other countries?

[01:15:38] NOAM CHOMSKY:
Yeah. My advice would be to pay some attention to what’s… I mean, it would be the same kind of advice that I think a, an honest physiologist would give to somebody who’s teaching swimming.

Uh, probably it’s worth paying attention to what people have learned about physiology, but probably it’s not going to do you much good. Uh, most

(laughter)

of the problems of teaching are human problems. They’re problems of motivation and making things be interesting and getting people to try using– be willing to use their natural intelligence to explore the thing, and so on. That’s probably about ninety-eight percent of the problem.

And what’s understood from a sort of scientific point of view may shed some light on this, and it’s certainly worth being aware of what people know. But if you think you’re going to learn much from it, I think that’s a mistake. Uh, I disagree with a lot of my colleagues on this, I should say.

Uh, for example, if you read The New York Times this morning, uh, there’s an op-ed by a close colleague and friend, Steven Pinker, uh, in which he says some things which I think are perfectly sensible. But what I don’t like about it is that he says these things come from what’s known in the cognitive sciences, and I don’t think they do. I think they’re just common sense.

You know, I don’t think the cognitive sciences contributed anything to these things. And it’s not that the cognitive sciences are sort of weak or something, it’s the same elsewhere. I mean, it really wasn’t until the n– I mean, take, say, physics, you know, way in advance of anything we’re talking about It really wasn’t until the nineteenth century, so I’m told, so it seems from the literature, that, uh, physics had anything much to contribute to engineering.

Uh, the crafts were so far ahead of what people understood that they just didn’t have a lot to say, you know. In fact, take somebody like, say, Thomas Edison, seemed to know nothing about the sciences, but he was inventing things all over the place. Now today it’s different.

You know, today you wanna do– you know, be an engineer, you better learn science. But that’s pretty recent. And in other areas, you know, crafts are often way ahead of– uh, cr– and just crafts and intuitive understanding and things like that are often way ahead of, of any theoretical understanding.

Uh, which does what it does, but people shouldn’t claim about it more than is in fact accurate.

(clears throat)

Uh, and it seems to me in these domains, there’s, some interesting things. I, I think there’s very interesting results, but I don’t think they bear much on questions of human life. You know, things like, say, teaching a second language or, or others like them.

And I think one should be rather wary about professional intellectuals who claim otherwise. Uh, there could– It’s, they might be right, but you have to be very skeptical.

There could be careerist reasons. That’s always worth asking, you know?

[01:18:31] AUDIENCE MEMBER:
Yeah, I wanted to ask about, um, uh, some comments you made earlier about, uh, when you look at concepts and you use the simple concept of a house, and you, and you think about it, it has these odd properties, and this seems to be true about, um, a variety of concepts and, um, when we– uh, in particular, the way we approach it, it seems is through language. We look at particular words and how they’re used and we, we try to flesh out what’s behind those concepts. And I was wondering how that inquiry is pursued

[01:19:04] NOAM CHOMSKY:
to gain insights, um,

[01:19:06] AUDIENCE MEMBER:
from a conceptual analysis as opposed to

[01:19:10] NOAM CHOMSKY:
Linguistic analysis. Okay. Yeah.

Yeah. It’s true that we often look at it from the point of view of language, that’s limitations of our methods of inquiry. In fact, if, if you go to the study of, say, conceptual development, people have done very interesting work approaching concepts without any language.

So there’s very interesting work on, uh, on infants, pre-linguistic infants. Now, when you say they’re pre-linguistic, you have to be a little cautious because other lines of inquiry are tending to show that as you improve your experimentation, kids who seem to not perform at all know an awful lot. In fact, uh, four-day-old infants can distinguish, uh, their mother’s language from another language as spoken by a third, a second woman, you know, a bilingual woman who speaks the two languages.

And in fact, ch-children– infants seem to know an awful lot– you know, the, the, we don’t know how early knowledge is there. We only know we can tap it at certain points. In any event, there is an interesting study of concepts, uh, at a time when children manifest no language, like a couple of months old, say.

Uh, so that’s non-linguistic inquiry, And it’s quite interesting. I mean, take, say, the current or maybe the last issue of Cognition, one of the journals that deals with this, has extremely interesting article by, uh, several people– several articles. In fact, one of them– uh, about things like, uh, notions like, uh, infant.

This is a couple of months– infants a few months old, studying things like how they interpret motion, objects in motion, and it looks at thr– There’s inquiry into three properties which turn out to be different in the way they work. One of them is that motion is smooth, meaning things don’t suddenly– you know, if something isn’t moving along here and suddenly it starts moving up there. So it’s gotta be con-continuous somehow.

For– sorry, I misnamed it. That’s continuity. It has to be continuous.

The second property is it has to be smooth, meaning no sharp inflections, like motion doesn’t go like this, you know, instead, unless there’s an external cause. And a third is the notion of inertia in the modern sense. Uh, well, it turns out these concepts apparently they do develop, but they develop in very different ways.

Uh, so very young infants, as early as you can test, a few months old, seem to un– to, to, to take it for granted that motion is continuous. If they’re presented with a situation which appears to be discontinuous, you get a surprise reaction. You can’t test a lot at those ages, but some things.

On the other hand, the, the idea that motion is smooth, you know, no sharp inflections, that seems to come along a little bit later, but still pre any– before any manifestation of language, so it seems to be non-linguistic in that sense. Like, I forget, maybe ten months old or something. What about inertia?

Well, it turns out people never learn it. Adults can’t figure it out. So if you give, uh, adults this– you know, if you give adults a, a sort of a picture where something is falling like this, and suddenly it drops as it’s on a, it’s on, on top…

So like say you have a ball on top of a little cart or something, and the ball falls off at this point, so it’s going like this. A lot of people think it goes straight down, you know? Uh, and in fact, more…

If, if you take a, something moving through a curved tube, let’s say, uh, how does this work? If it’s wa– I think the way it works is if it’s water that’s moving through the tube, lots of people think it goes straight.

Uh, but if it’s a ball, they think it keeps going, continuing the curve, you know. Uh, and there’s all sorts of funny ideas. In short, people just don’t have the idea of inertia, which is why it was so hard to discover.

You know, it was really hard to discover because it’s just not part of our conception of the world. It’s not part of our folk physics. Now, continuity and smoothness seem to be part of our folk physics, but they develop differently.

Apparently, if this work is correct, at different times. And by now, there’s quite a lot known about infants’ conceptions of objects, about solidity, you know, the fact their impenetrability, I mean, all sorts of other things. All right, this is an inquiry into non– apparently into non-linguistic concepts.

It’d be hard to argue that a six-month-old kid is using knowledge of language to do these things. Uh, and to deal with such problems in the case in– when you deal with an adult, it’s harder to dissociate because of course the adult has a ton of linguistic material, and you’ve got to figure out a way to dissociate that from non-linguistic material. But here too, there are studies, uh, like the work that’s come out of, say, David Marr’s school on vision as an example.

Uh, surprisingly to me, this work is now being given an externalist interpretation in a lot of the philosophical literature, and I don’t understand how that’s possible. Uh, I mean, what’s going on is a kind of a textual reading of Marr of the kind you do with, you know, like Baker and Hacker do with Wittgenstein or somebody, And it’s beside the point. Marr couldn’t have meant the things that are being attributed to him on the basis of this textual analysis because he was a scientist.

And furthermore, and I, I knew him personally. I know perfectly well he didn’t mean those things. But, uh, the, but the more interesting fact, which is independent of textual analysis, is if you look at the experimental materials, it’s clear that this can’t be right because here we get real analysis of…

I mean, the, the exper- the experiments are mostly things like lights on p-p-tachistoscopes. So a person sees a bunch of, you know, a couple of lights shining on tachistoscopic presentations, and they say, “I see a cube rotating in space,” you know, or, “I see a line going around a barrier,” or some such thing.

Well, they’re not seeing a cube rotating in space because you set it up. In the– if you use the philosophical literature as analogies, you’re setting it up as a brain in the vat, basically. You know, here’s this brain in the vat, you’re sticking stimuli into it.

They happen to be dots of light. There’s no cube anywhere. The world doesn’t have to exist for these experiments.

Uh, and on– But the point is, with regard to the conceptual analysis, that people do put very rich interpretations on things in very specific ways, And it appears, for example, that they somehow– that the visual system is constructed to, um, in terms of what’s sometimes called a rigidity principle. That is, it tries to see things as rigid objects in motion.

Well, you know, uh, now they may be dots on t-t-tachistoscopes, and it’s been shown that very few presentations suffice to give an even unique interpretation of something as a rigid object in motion, even though the evidence could be like a bunch of dots showing up on, on tachistoscopic presentations. Well, okay, that’s the way, that’s the basis for our concepts. We somehow interpret the world as a bunch of rigid objects interacting with one another.

Now, that’s not experience, surely, because most of our experience is with non-rigid objects. Rigid objects only exist in technological societies, you know. Uh, most of–

And it can’t have anything to do with natural selection ’cause most, you know, human evolution was finished before people had anything like geometrical objects. There’s none– There’s almost none in nature.

You know, you take a walk in the woods or something, you never find anything that’s, uh, like a cube. You’re not gonna find a cube, say. Uh, but the, the, uh, uh, so somehow the system developed for whatever reason, uh, in order to yield particular kinds of concepts, not others.

Uh, and that’s an inquiry into, uh, conceptual structure that is independent of language, but they’re hard to do. The trouble is language is just so interpenetrated with everything you do that It’s hard to sort it out.

It becomes experimentally difficult. But your point is right. It’s a separate inquiry, and it ought to be distinguished.

Uh, there, there are also, you know, for years, I mean, decades, there have been attempts to discover what– How language influences thought or whether it does. And it’s a very tricky inquiry.

It’s, it’s kind of interesting that if you take the standard theories of reference in language, you know, the theories that assume that if you have a word, you have a thing, let’s say a piece of four-dimensional space or something that it refers to, if you think about those approaches, they’re really taking a kind of a super Whorfian approach, going back to Benjamin Lee Whorf, the guy who argued that the way you talk determines the way you think. They’re really assuming that if you have a word, you’ve got a thing, you know, that you postulate associated with it. I mean, that’s super Whorf, you know, Whorfian extremism.

Uh, and, uh, anthropologists who work on this topic don’t find it that easy to decide whether these– this conception is true. They don’t find it as easy as people who are sitting around, uh, you know, writing philosophical theories. And in fact, there’s been a lot of work to try to find out whether it’s true, and there’s, you know, the beginnings of evidence where you can find little effects, uh, of, you know, you find some Mexican language that, uh, deals with plurals and animacy somewhat different than English, and you can show slight differences in sorting and memory experiments, which seem to indicate some effect of, uh, language structure on thought, but they’re hard topics, you know?

It’s not easy, and it’s been like forty years that people have been working on this one with very limited results and we just don’t know to what extent thinking and conceptual development and so on is independent of language. I mean, you know a lot about language. It’s harder to find things out about language-independent conceptual development, but not impossible.

Uh, but the way you, whoever it was who said about it, I forget where, uh, was putting the point quite accurately. You really wanna study these things independently and see how they’re related to one another.

[01:28:41] AUDIENCE MEMBER:
Yeah. Could you go into some more detail on what you find incoherent in certain uses of twin Earth examples?

[01:28:45] NOAM CHOMSKY:
In which?

[01:28:47] AUDIENCE MEMBER:
Can you go into more detail on what you think is incoherent in some uses of Twin Earth examples and also radical translation?

[01:28:54] NOAM CHOMSKY:
Well, let’s take Twin Earth. The Twin Earth experiments ask– We, we set up this guy who’s molecule by molecule identical with Oscar, uh, except he happens to be in a world in which, uh, what he calls water is XYZ, not H2O. And for the purposes of the experiment, we put aside the fact that if that were true, he couldn’t be molecule by molecule identical.

So we put that aside. Uh, now– And then we ask ourselves, well, does the word water in Oscar’s language refer to XYZ?

And does the word water in Twin Oscar’s language refer to H2O? And the answer is to that question is yes, if Hilary Putnam says yes and no if Hilary Putnam says no, because he’s in- he’s inventing the term.

He’s saying there’s a term which I invent, I call it refer, like, as often you use terms from natural language, like when you talk about energy or work in physics, uh, or angular velocity for that matter. He’s saying, “Here, I invented this term,” and I tell you what it refers to, what it means. You know, that’s the way you do it when you invent a term.

It means what you say it does. But you can’t ask whether you have any intuition about the question. Now, when you try to run through the same experiment a-and asking, “Is Oscar referring to XYZ?

Is he talking about XYZ?” I think you get no answers. You know, it depends on the conditions.

Like if twin, twin Oscar is coming to Earth, and there’s a glass of water in front of him and a glass of Sprite, and he can’t tell the difference, the water is H2O, and twin Oscar says, “I’m thirsty. I want a glass of water.” If he points to the Sprite, he’s making a kind of mistake which he’s not making when he refers to the water, at least by my intuition, which would mean that, you know, somehow he was referring to H2O all along, you know, even though he never had any.

Uh, and in fact, uh, if you really look at– if you push it further, I think it becomes even worse, because it turns out that what you and I are referring to, uh, when we use the word water is not H2O at all. It’s some mixture of H2O and deuterium oxide in the proportion of, you know, six thousand to one or some other thing. Uh, and when this was discovered, actually nuclear engineers and people like that just made up two new words.

They talked about heavy water and light water. Uh, they didn’t say they were both water. Well, they could have done that, or they could have done some other thing.

But the point is all of these things seem to be decisions. There aren’t so many facts at issue. On the radical translation matter, it seems to be even more radical.

You know, the problem seems to be even deeper. Uh, the radical translation paradigm stipulates, it, it’s supposed to describe the situation of two people in a communication situation of a child acquiring a language and of a linguist studying a language. It’s supposed to describe those three things.

First of all, those three things are illegitimately mixed. They’re completely different in character. Uh, two people in a communication situation already have some kind of language in their heads.

The linguist, the child acquiring a language doesn’t have a language in his head. He has in the innate properties of the language faculty in his head. The linguist is approaching the situation from totally a third point of view.

What the linguist has is whatever’s been learned about natural language and, you know, some ideas about how to construct the scientific theory. They’re all doing totally different things, and assimilating these already doesn’t make any sense. But let’s take any one of the three.

For any one of the three, the radical translation paradigm stipulates what you’re allowed to do. So let’s take the linguist. What the linguist is allowed to do, according to these orders from on high, is the linguist is supposed to go to the jungle community.

You know, I find some guys running around. And there’s, you’re only allowed to look at their behavior in that community. You’re not allowed to look at, say, what you learned about Japanese.

Well, nobody would ever approach a problem in the natural sciences this way. If you go to this jungle community, the first thing you would ask is, Are those human beings? Now, that’s an empirical question.

And there’s an answer. the way to answer it, the answer is, if they– you ask about a counterfactual. You say, “If they– if I had grown up in their community, would I be doing what they do?

And if they had grown up in my community, would they be doing what I do?” That’s an empirical question. Like other empirical questions, not trivial, but you know how to investigate it.

All right, once you– suppose you’ve pursued that question, and you’ve sort of got some evidence that they’re human beings. Okay, well, now you don’t restrict yourself to what they do. You look at whatever you know about what human beings do.

I mean, that’s the way any scientist would approach a similar question. If you, if you’re looking at, you know, a bunch of fruit flies and you’ve convinced yourself that those are fruit flies, you don’t say, “I’m not allowed to use any evidence about what’s found out about fruit flies. I just have to look at them.”

No, that would be crazy. You know, you don’t do that. Uh, and in fact, nobody does do it.

So everyone who’s studying language empirically, once you’ve sort of convinced you, you don’t even bother asking it usually, but once you’ve convinced yourself that those are human beings, meaning same initial state, uh, you, you s- you s- you s-

Now suppose you have two theories about them. Well, you could, I suppose it’s Japanese, let’s say. You have two different theories about how Japanese works.

Now, from the point of view of the radical translation paradigm, if these two theories both describe what you’re allowed to look at about Japanese, you’re finished. There’s no further question about truth or falsity. Nobody would do that.

I mean, what you would do is, for example, look at English and ask, uh, what do you know about English? Well, it might turn out that these two theories about Japanese require different assumptions about English on the assumption that they all come from the same base, innate basis, and one of those might be refuted for English. Well, in that case, the corresponding ones are refuted for Japanese, even if you had no evidence on the, the argument is perfectly straightforward.

You know, it is, well, look, we believe we have evidence, reason to believe that they come from the same initial state. Therefore, what we found somewhere about that initial state is usable when you’re looking at the next, uh, manifestation of that initial state. That’s just ordinary garden variety science, completely normal, but it’s excluded by the radical translation paradigm.

And it gets worse, because not only are you forced to rest– you’re, you’re also forced to restrict all, to eliminate all sorts of other evidence. Like suppose something comes along from event-related, you know, from, uh, uh, from ERPs, from, uh, evoked reaction potentials. So I suppose something’s really learned someday about this stuff, not just, you know, giving numbers to things you knew, but something new, which could happen.

Well, according to the radical translation paradigm, you’re not allowed to use it. You know? Uh, it’s not really–

You’re only allowed to look at the behavior of the people, uh, in their jungle community. It’s kind of ironic that this is being proposed by the very same people who insist that science is holistic, that everything is relevant all the time. Uh, that I leave for someone else.

But in any event, this is saying only certain kinds of evidence are relevant in this particular case. And furthermore, when you pursue it further, that evidence is very narrowly circumscribed. Uh, you’re allowed to look at their answers to certain questions which are stipulated by the philosopher.

Uh, so Quine says you’re allowed to look at– he designs a kind of an experiment, uh, an assent-dissent experiment. You give somebody something, and you say, “Gavagai,” and he’s supposed to say, “Yeah, it is a Gavagai,” or, “It isn’t,” and that’s it. Uh, then you’re allowed to do simple induction, whatever that is.

Uh, something apparently which isn’t done in language or in anything maybe. Uh, and the, the, a-and beyond that come various other stipulations about association and so on. Well, you know, where does this come from?

Why not twenty thousand other experiments you might think of? Uh, why not other kinds of evidence? Why not looking at, at other organisms?

You know, why not looking at– look at what you might learn about cells or anything, you know? You know, if it’s science, no such constraints are even remotely thinkable. And it’s in– so, so this is really, in my view, radically dualistic.

And any conclusions that follow from it about thinking or rationality or beliefs and so on are highly suspect. In fact, they seem to me to make no sense at all. It’s in that respect that I think it’s a, it’s really radical dualism.

And there are some ironies here because it’s being proposed by people who want to make philosophy scientific. You know? They wanna say, “Look, it’s really part of the sciences, all this kind of old metaphysics and stuff we throw out.”

And it seems to me quite the opposite. You know, they’re, they’re going into a form of dualism which is much more pernicious than traditional metaphysical dualism, which as I said, I don’t think was pernicious at all. I think it was quite a plausible scientific hypothesis proven wrong when the theory of body was demolished.

But other than that, a plausible hypothesis, you know, given with reasons. This kind of stuff just seems to me, you know, some other kind of dualism which has, again, no redeeming features. And I, I think if this is correct, you know, maybe I’m wrong, but if this is correct, it seems to me an interesting question arises, why is it so compelling?

You know, why are people so driven to it? My own– I have a suspicion of it.

That’s an unfair question to ask unless you’ve shown that it’s really wrong, you know. You first have to show that. But if you can show that, then the question arises, why is it so compelling?

And it seems to me there’s a possible answer to that second question. It could be so compelling because we are just designed to be dualists. Uh, like we’re designed to see the sun setting, you know, no matter how much relativity theory you learn, you know, when you’re out there on the beach at night

and you see what’s going on, you see the sun setting, and you can’t help it. You can be the greatest physicist in the world, you still see the sun setting, ’cause that’s the way you’re designed, you know? And it could well be that we are just designed to see people as, you know, minds and bodies or something like that.

I don’t know. Maybe that’s the way people are. You know, we just see things that way.

Doesn’t mean that that’s the way they are, just doesn’t mean it’s… I mean, it, it’s not false any more than that the sun setting is false, but it’s just a, from a naturalistic point of view, it would just be the wrong way of looking at things. That’s possible.

Um, that’s a, that’s a hypothesis about something which has to be first shown, namely that these really are radically dualist approaches. But that’s, uh, wherever the questioner was, why I think that these two things in particular are indeed radically dualist.

[01:39:12] STEPHEN NEALE:
Well, unfortunately, we don’t have any time for any further discussion. Oh. So I’d like to thank Professor Chomsky on behalf of us all.

(applause)