Re-Introducing The Concept of Mind

Daniel C. Dennett

Department of Philosophy and Center for Cognitive Studies,

Tufts University

( )

[1] Imagine somebody whose enthusiasm for metaphysical musings is so matched by ineptitude that when his bank informs him that his account is overdrawn, he manages to convince himself that the modern banking system had created a new kind of Cosmic Substance: anti-cash , convertible into minus-pounds-sterling, nego-dollars, vac-euros, and so forth. Being a staunch 'realist' about all things, he decides that his bankers have just informed him that they are holding, somewhere within their walls, in a container with his name on it, a particular quantity of anti-cash. It's like matter and anti-matter, he thinks, and he supposes that the annihilation that happens when his overdraft of -£200.43 meets his deposit of £300.46 is– shazam!– the explosive generation of £100.03 of ordinary cash (minus a small quantity extracted by the bank) plus, perhaps, a few stray photons or quarks or gravity waves. He wonders: What kind of containers does the bank use to hold the anti-cash till the regular cash arrives? How are they insulated? Can you store cash and anti-cash in the same box and somehow prevent them from getting in contact? Might there be zombanks that only seemed to store cash and anti-cash? How could we tell? This is a hard problem indeed!

[2] What this poor chap needs is a good dose of Gilbert Ryle's 'logical behaviorism,' which is the plain truth about bank balances, however controversial it might be as a theory of the mind. Your bank statement does not report a set of facts about containers and viaducts and machines within its walls; rather it is an expression–one among many possible expressions–of what Ryle would call a multi-track behavioral disposition: roughly, the bank is disposed to honor your monetary commitments up to a certain amount, disposed to charge you for the current condition of your account at the following rate, disposed to expel currency from its automatic machines at your command in such-and-such denominations, and so forth, an indefinitely large system of interlocked if-thens. The bank needs to keep track of all these dispositions, and how it manages to do this is a 'wires and pulleys' question of interest to certain sorts of technicians in the banking world, but your bank statement–and indeed all your communicative interactions with the bank–are not about these details of implementation at all. You can know everything worth knowing about bank balances–you can be a financial genius–and be clueless about the actual mechanisms by which banks maintain their breathtakingly elaborate dispositional states, the states that govern all their financial behavior . You don't need to be a mechanist, and you don't need to be a 'para-mechanist' (inventing anti-cash and the para-machinery to deal with it).

[3] If only the case of the mind were as straightforward! If it were, however, there would have been no need for Ryle to write The Concept of Mind , one of the most original and influential–if still hugely underestimated–works of philosophy of the century. The goal of the book was to quell just such sorts of confusions about mental events and entities, the confusions that had generated the centuries-old pendulum swing between Descartes's dualism ('para-mechanical' hypotheses) and Hobbes's materialism (mechanical hypotheses), both sides correctly discerning the main flaws in the other, but doomed to reproducing them in mirror image. Since minds are so much more complex and confusing than banking systems, and since the tempting confusions about minds have centuries of tradition giving them spurious authority, Ryle's task of re-educating our imaginations had to be correspondingly subtle and difficult, so much so that my parallel with banking, if taken dead literally, would be just the sort of caricature that so often leads to premature dismissal of an iconoclastic voice. The multi-track dispositions of a bank vis-a-vis a depositor can be–indeed legally must be–spelled out definitively, without significant ambiguity or loss, but Ryle knew better than to accede to requests that he define the disposition of vanity, or wittiness, or any other mental treasure in terms of 'input and output' or behavioral responses to stimulation. That was not the sort of contribution he was setting out to make. He had something more modest–certainly more realistic–in mind, not a formal or scientific theory of the mind, but still something in its own way highly ambitious: he hoped to break some of the most deep-seated habits of thought we have about our own mental lives.

[4] But isn't it just obvious that minds are not at all like banking systems? Isn't it obvious that we know our own minds 'from the inside' in a way that nothing knows or needs to know banking, which is all outside and no inside? Perhaps it is just obvious–until you read The Concept of Mind . You may then discover that even if he fails to convince you, you can at least harbour the hunch that maybe, just maybe, the giant step we need to take to solve the mysteries of the mind is some version of Ryle's sideways step off the pendulum. But it certainly is a radical step.

[5] How did Ryle hope to dispel the confusions he saw in the tradition? 'The Concept of Mind ,' he tells us, 'was a philosophical book written with a meta-philosophical purpose.'

I wanted to apply, and be seen to be applying to some large-scale philosophical crux the answer to the question that had preoccupied us in the 1920s, and especially in the 1930s, the question namely 'What constitutes a philosophical problem; and what is the way to solve it?' . . . by the late 1940s it was time, I thought, to exhibit a sustained piece of analytical hatchet-work being directed upon some notorious and large-sized Gordian Knot. . . . . For a time I thought of the problem of the Freedom of the Will as the most suitable Gordian Knot; but in the end I opted for the Concept of Mind–though the book's actual title did not occur to me until the printers were hankering to begin printing the first proofs. (1970, p12)

[6] Ryle set out to demonstrate the absurdity of what he calls 'the Official Doctrine' and warns at the outset: 'I shall often speak of it, with deliberate abusiveness, as 'the dogma of the Ghost in the Machine.'' Who are these benighted champions of the Official Doctrine? Are there–were there–actual 'Cartesians' (or 'Hobbists') whose susceptibility to 'category mistakes' blinds them to the truth? Is this an affliction only of philosophers or do scientists or others also commit these errors of thought? One of the idiosyncrasies of the book is that there are no footnotes and no references. No thinker living in 1949 is mentioned or quoted anywhere in its pages, in spite of the fact–perhaps because of the fact?–that those rollicking pages often purport to be demolitions of contemporary confusions. The only person from the twentieth century who is mentioned even in passing is Freud, and Ryle has nothing controversial to say about Freudian ideas.

[7] Was he tilting at windmills, then? No, I think not; I think Ryle knew just what he was doing when he left his targets anonymous and timeless, for he was going after mistakes that lie just beneath the surface of reflective thought, errors that, when pointed out, everybody can scoff at but few can avoid being poisoned by. His quest was quixotic not in the usual sense, but in Jorge Luis Borges's sense. In his famous fiction, 'Pierre Menard, Author of the Quixote, ' (1962) Borges tells the tale of a literary theorist who set out to compose (not copy, not write from memory) Cervantes' great work anew in the twentieth century. He succeeds, and Borges tells us 'Cervantes' text and Menard's are verbally identical, but the second is almost infinitely richer.' (p42) How could this be? It could be because of the context in which the two texts were written–and then, of course, read. We don't need the fantastical exercise of the fictional Menard to give us a real example of this phenomenon. At the end of the twentieth century, The Concept of Mind is a much richer text than it was when Ryle wrote it in mid-century. It certainly has much more in it now for me than when I first read it as an undergraduate in 1960. In fact I have recently been struck by a pattern: many of the themes that are emerging as hot new directions in up-to-the-minute cognitive science bear a striking resemblance to long-disregarded Rylean themes: embodied and 'situated' cognition; your mind is not in your brain; skill is not represented; intelligence without representation--to name only the most obvious. Ryle himself certainly did not understand his ideas in the way we are tempted to understand these returning versions of them. Today's problems–the theoretical problems to which his ideas might be part of the solution–were largely unimagined by Ryle. How did he arrive at his ideas, then? I think the answer lies in his method, which more than most methods welds its strengths and weaknesses into an indivisible lump, take it or leave it.

[8] Ryle's method is exasperatingly informal, not just a-systematic but positively anti-systematic, the brilliant piling on of analogies and examples and rhetorical flourishes, cunningly designed to cajole the reader out of those bad habits of thought, a sort of philosophical guerrilla warfare that never settles into or commits to a positive 'theory' for long enough to permit a well-aimed attack. The reason his method is so informal gradually becomes clear: when people set out to do serious theorizing about the mind, the first thing they do is to ransack 'common sense' for a few hints about which direction to march; if they then set off on the wrong foot, they soon create problems for themselves that no amount of theory-repair or refutation-of-the-opposition will solve. The mistakes are earlier, pre-theoretical presumptions that are unlikely to come up for re-examination in the course of formal theory-development and criticism. Ryle suspects that some of the standard goods delivered by 'common sense' don't deserve their high standing, but to show this he must fight fire with fire: he must charm us into pausing and reflecting, so that we may pit better common sense against worse before running off to theory-land. But is Ryle right ? Are all these traditional ways of thinking mistakes? Fifty years later, we can see that many of them are still tenaciously defended by deeply thoughtful and adept theorists, but this hardly shows he was wrong. A. J. Ayer, writing in 1970, candidly assessed the state of play at that time:

In short, what Ryle has succeeded in doing is to reduce the empire of the mind over a considerable area. This is an important achievement, and one that is brilliantly effected, but it does not fulfill Ryle's professed intention of entirely exorcizing the ghost in the machine. The movements of the ghost have been curtailed but it still walks, and some of us are still haunted by it. (1970, p73)

The tide is still changing, as I just noted, and the defenders of the ghost and its kin today are ever more on the defensive (though their sallies, from their ever more precarious toehold in common intuition, have become desperately extravagant). I am inclined to think that Ryle just underestimated the strength of the philosophical therapy required to accomplish his aim.

[9] That is not the only thing he underestimated. Ryle was no scientist, and he sometimes betrays an almost comical optimism about the compatibility of what Wilfrid Sellars called the scientific image–the world of sub-atomic particles and forces–and the manifest image–our everyday world of people and their activities, houses and trees and other 'middle-sized dry goods,' as Ryle's colleague, John Austin, once put it. It seems to have been a point of unexamined faith for Ryle that whatever the scientists might learn about mechanisms of the brain, however necessary these were in grounding our behavioral dispositions, they would shed scant light on the questions that interested him. This might have been true, had brains not been so much more complicated than banks. Ryle's questions are about what people do , questions at what I call the personal level of explanation (Dennett, 1969), not about how brains make it possible for people to do what they do; those sub-personal level questions were completely outside his purview. Cognitive scientists have often promulgated similar differences in level or perspective, such as David Marr's famous (1982) trio of computational, algorithmic and physical levels, or the 'ecological' perspective advocated by J.J.Gibson (1975), and it has been widely recognized that many of the false starts in cognitive science have been due to failures to find the proper level of analysis for the topic. The bold claim might be defended that all the really tough problems in cognitive science reside in the murky and embattled zones where the relations between these levels must be clarified, and on these issues Ryle is imperturbably silent, content to protect the personal level from misguided incursions of mechanical hypotheses and para-mechanical hypotheses (an inspired coinage of Ryle's that hastened the extinction of its referent, though a few endangered species of dualism still cling to dubious life). Whether Ryle's silence was due to complacency or just prudence, it leaves some genuine philosophical puzzles unaddressed. The strains of Ryle's wishful thinking show through at times in the book. Many of Ryle's dismissive analogies are, in a word, glib, shots in the dark that cannot persuade us today. But even when Ryle is wrong, he's usually right about something , or as Austin astutely noted in a masterful review in 1950: 'Not only is the book stimulating, enjoyable and original, but a quite unusually high percentage of it is true, the remainder at least false.' But which portions are which? The informality of Ryle's presentation leaves that up in the air.

[10] The Concept of Mind is one of those books that is often cited by people who haven't read it but read about it, and think they know what it is in it. They have read that it epitomizes two woefully regressive schools of thought that flourished unaccountably in mid-century but are now utterly discredited: Ordinary Language Philosophy and Behaviorism. Yes, and imbibing alcohol will lead you inexorably to the madhouse and masturbation will make you go blind. Don't believe it. The dismal excesses of both these schools of thought (like the dismal excesses of sex, drugs and rock 'n roll) are terrible to contemplate, but a few works of genius defy the labels and brilliantly sidestep all the standard 'refutations.' This is one, but even those who have read it often come away with curious misconceptions. Non-philosophers, in particular, not being acquainted with the folkways of academic philosophy, often generously assume that philosophers must occasionally achieve results the way workers in other disciplines do. Having read so often about Ryle's famous doctrine of 'category mistakes,' they jump to the conclusion that Ryle must have exposed, definitively identified, and proved the fallaciousness of, something called a category mistake. If only. Ryle is a tireless alluder to the 'logical geography' of various concepts, and the errors that accrue to those who lose their way in this terrain (or is it a multi-dimensional space?), and this must spawn fantasies in many readers about some technical volumes somewhere in which one can learn this logical geography, laid out like the Periodic Table, something every philosophy graduate student is drilled in. But that is not how philosophy at its best proceeds. Ryle tells us in a brief (and glorious) 'Autobiographical' (1970) that when he was at school, one of his masters said: ''Ryle, you are very good on theories, but you are very bad on facts.'' Ryle went on: 'My attempts to repair this latter weakness were short-lived and unsuccessful.' During his undergraduate days at Oxford, he says, he 'took greedily to the subject of Logic. It felt to me like a grown-up subject, in which there were still unsolved problems.' But logic 'was in the doldrums' in Oxford at that time. Russell and Whitehead were 'still only subjects of Oxonian pleasantries' twenty-five years after the publication of their monumental Principia Mathematica . Nevertheless, 'It was Russell and not Moore whom I studied, and it was Russell the logician and not Russell the epistemologist.'

[11] So is Ryle, then, like Wittgenstein or Quine, a serious contributor to mathematical logic or logical theory? Not at all. 'Having no mathematical ability, equipment or interest, I did not make myself even competent in the algebra of logic; nor did the problem of the foundations of mathematics become a question that burned in my belly.' (1970, p7) Ryle's interests were indeed the interests of a logician, of somebody deeply curious about the abstract relationships between premise and conclusion, arguments and concepts and propositions and inferences, but his vision of 'Logic' was old-fashioned, and not clearly the worse for it. He thought, correctly, that all too often the shiny new tools of logical formalisms tempt their adepts into substituting formal derivation for . . . thinking. There are times when the bracing task of translating one's ideas into a canonical idiom for which proof procedures are defined is a path to philosophical discovery, but even there, almost all of the heavy lifting is done in coming up with the translation. The formal proofs in philosophy that have ever made a significant contribution are vanishingly few. Ryle actually knew a good deal about logic, but when he holds forth in The Concept of Mind about the logic of concepts, he is shooting from the hip most of the time, and trusting his good, peculiar brand of common sense. So would you, if you had such a fine faculty. His distinction between knowing how and knowing that , the topic of chapter two, has stood the test of time (and been reinvented by others) across a host of disciplines, and his informal observations on the logic of dispositional statements, the 'systematic elusiveness of 'I'', and other idiosyncracies of ordinary language have grounded or reformed more than a few philosophical projects.

[12] Ryle was no logician, and no scientist, but he was also no ivory tower humanist, in spite of his purely classical education in Greek and Latin (with self-taught Italian, German and French). Aside from his work in intelligence during World War II, his entire adult life was spent at Oxford, but within that insular world, he was, as his lifelong friend Geoffrey Warnock (1979) has said, 'an outstandingly friendly, sociable, and (a word that particularly fits him) clubbable man.' (p xiv ). Ryle himself thought this was what protected him from the ego-fevers that afflict so many philosophers. Comparing Anglo-American philosophers to their counterparts on the Continent, he once opined:

I guess that our thinkers have been immunized against the idea of philosophy as the Mistress Science by the fact that their daily lives in Cambridge and Oxford Colleges have kept them in personal contact with real scientists. Claims to Fuehrership vanish when postprandial joking begins. Husserl wrote as if he had never met a scientist--or a joke. (1962, p181)

[13] Few Anglophone philosophers, by the way, have matched Ryle's deep knowledge of Husserl and the Phenomenologists. 'I even offered an unwanted course of lectures, entitled Logical Objectivism: Bolzano, Brentano, Husserl and Meinong.' These characters were soon known in Oxford as 'Ryle's three Austrian railway stations and one Chinese game of chance.'' (1970, p8). Most of Husserl's topics can be found in The Concept of Mind by anybody who knows what they are, but in these pages you will find no talk of intentionality , no noemata –and no talk of qualia either, I am happy to report. His distrust of philosophical jargon approached a shibboleth, and his love and mastery of his own 'South Country English' idiom served him well, providing a palette for one of the most recognizable styles in English letters.

[14] P.G. Wodehouse was one of Ryle's favourite authors (along with Jane Austen, whose novels he read and reread) and like Wodehouse's world of Bertie Wooster and Jeeves off on their sun-kissed round of country house weekends, Ryle's particular Oxbridge is an intensified English world too good to be true, one would think, but strange to say, it is a portion of the real world that Ryle actually inhabited: a hearty world of gardening and cricket and tea and bridge parties, rowing and swimming and imagining Helvellyn in one's mind's eye, humming Lillibullero, and, of course, dealing with students in tutorials and dons at high table.

[15] Ryle was under no illusions about the shortcomings of this book. His own characterization of its aim left little or no room for half-measures, and invited incredulity by its sheer sweep.

On the view for which I am arguing consciousness and introspection cannot be what they are officially described as being, since their supposed objects are myths; but champions of the dogma of the ghost in the machine tend to argue that the imputed objects of consciousness and introspection cannot be myths, since we are conscious of them and can introspectively observe them. . . . . I try to show that the official theories of consciousness and introspection are logical muddles. (p155)

As Austin had noted in his laudatory review–and Ryle never disagreed–there were clear overshootings in his campaign:

Undoubtedly he does persuade himself that what he has to show is that 'occult' episodes 'in the mind,' which are 'private' to one person, simply do not occur at all–not merely that they are never mysterious causes, themselves mysteriously caused, of our physical movements, nor merely that their numbers and varieties have been exaggerated. (p47).

[16] Given Ryle's insistence in The Concept of Mind that thinking was not in any important sense a private phenomenon, a question that quite properly dogged him for the rest of his life was vividly put in terms of Rodin's famous sculpture of the huddled, frowning thinker: what is le Penseur doing? He is not, to appearances, behaving, or if he is, his behaviour is consistent with too many different accounts of his ongoing thinking (his 'inner story,' we are tempted to say, but Ryle fights hard to keep us from saying it). It is obvious enough that the Thinker is probably talking to himself, at least part of the time, and Ryle happily allows that we are all capable of such 'silent soliloquy.' But are we to adopt talking to ourselves as the model for all thinking–is all thought conducted one way or another in a 'language of thought'? Ryle sees that even if some thought is in language, not all thought is in language–and sometimes talking to yourself is not even an instance of thinking, but rather a substitute for thinking. (The confusion between talking to yourself and thinking is often encountered in philosophical books, especially those that maintain that thinking is a sort of talking to yourself!) So what is the Thinker doing, and what is different about what he's doing when he's doing it well? Ryle wrote a series of papers, none entirely satisfactory by his own lights, attempting to answer this question (some of them collected posthumously in Ryle, On Thinking , 1979). As he comments in his Collected Papers :

. . . like plenty of other people, I deplored the perfunctoriness with which The Concept of Mind had dealt with the Mind qua pensive. But I have latterly been concentrating heavily on this particular theme for the simple reason that it has turned out to be at once a still intractable and a progressively ramifying maze. Only a short confrontation with the theme suffices to make it clear that and why no account of Thinking of a Behaviorist coloration will do, and also why no account of a Cartesian coloration will do either.' (pviii of Vol II, 1971)

[17] Where does that leave us? With a book of breathtaking ambition in one dimension and refreshing modesty in another, a book whose hints and asides have sometimes proven more influential than its major declarations, a book that may in another fifty years prove to have an even higher proportion of truth than we find in it today. In any event, it has already fulfilled Ryle's 'meta-philosophical purpose' of showing us a good way of doing philosophy. And just as one would expect, one cannot learn this good way by memorizing a few rules or doctrines, but only by immersing oneself in the practice and letting the method do its work. When I was writing my dissertation under Ryle's supervision, I didn't appreciate this subtle fact, and told myself (and my fellow graduate students, I am sad to say) that I had actually learned almost nothing from the great man; he was a wonderful booster of my often flagging spirits, a charming example-spinner and conversationalist, but almost useless as an argument-critic, doctrine-refiner, debater. We never argued ; he never attempt to refute my propositions. But then, on the eve of my viva-voce examination in the spring of 1965, I compared the submitted draft of my dissertation with a version I had written roughly a year earlier, and was amazed to discover Ryle's voice, perspective, method, and vision on almost every page of the later version. You, too, may read The Concept of Mind , and walk away thinking you haven't learned very much. Don't be so sure. In due course you may discover that you have become a Rylean like me.

[18] In one of his hilarious novels, Peter de Vries has a character say 'Oh, superficially he's deep, but deep down, he's shallow!' How could it be otherwise, come to think of it? Philosophy is above all supposed to be profound, though, and a student asked me the other day if, in the end, I thought Ryle's book was deep. No, I decided; it is shallow–wonderfully, importantly shallow. There are those who love to tread water, the deeper the better, and who think that philosophy without depth is guaranteed to be . . . superficial! Ryle unhesitatingly defied this fashion and taught us how some of the deepest waters in philosophy could be made to evaporate. Those who still find themselves over their heads on the topic of 'consciousness and introspection' would do well to follow Ryle onto the shore of common sense, where the remaining problems are much more interesting than treading water. [1]


Austin, J. L., 'Intelligent Behaviour: A Critical Review of The Concept of Mind ' in Times Literary Supplement, April 7, 1950. Reprinted in Wood and Pitcher.

Ayer, A. J., 'An Honest Ghost?' in Wood and Pitcher.

Dennett, Daniel, 1969, Content and Consciousness , London: Routledge & Kegan Paul.

Gibson, J. J., 1975, Cognition and Reality , San Francisco: Freeman

Gibson, J. J. 1979, The Ecological Approach to Visual Perception , Boston: Houghton Mifflin.

Marr, David, 1982, Vision , Cambridge, MA: MIT Press

Ryle, Gilbert, 'Phenomenology Versus 'The Concept of Mind ',' 1962, [translated from the original French] in Collected Essays , vols. I, 1971.

Ryle, Gilbert, 1970, 'Autobiographical,' in Wood and Pitcher.

Ryle, Gilbert, 1971, Collected Essays, Vol I, II, London: Hutchinson.

Warnock, Geoffrey, 1979, 'Preface' to Ryle, On Thinking (ed. Konstantin Kolenda, a posthumous collection), Totowa, NJ: Roman and Littlefield. ppvii-xv.

Wood, Oscar P., and Pitcher, George, eds., 1970, Ryle, A Collection of Critical Essays , Garden City, NY: Doubleday.


© D. Dennett

EJAP is a non-profit entity. EJAP may not be printed, forwarded, or otherwise distributed for any reasons other than personal use. Do not post EJAP on public bulletin boards of any kind; use pointers and links instead. EJAP retains the right to continuously print all the articles it accepts, but all other rights return to the author upon publication.

EJAP is hosted at The University of Louisiana at Lafayette . Please send comments and questions to .


[Note 1] I am grateful to Victoria McGeer, Anthony Chemero, and Richard Rorty for comments on an earlier draft.