Living with Uncertainty IV
(This is an original excerpt from a never-before published new work)
Chapter 4
IT’S TURTLES ALL THE WAY DOWN
A legend has it, a renowned cosmologist, who has just concluded a public lecture on the structure of the known universe, is approached by a visibly irate woman, “You are wrong,” she says, “the universe cannot and does not exist in pure space. It must have support.”
“Then what’s holding it up?” asks the cosmologist, his interest piqued.
“The back of a giant turtle”
“And what’s holding the giant turtle up?”
“Another giant turtle”
“And what’s holding…,” but here the interlocutor, anticipating his objection, waves away his question.
“Its turtles all the way down”
The first time I heard this was an NYU undergraduate majoring in philosophy. This was the sixties and it was a heady time for American philosophy. My teachers were Paul Edwards, the celebrated analytic philosopher, William Barrett who helped introduce European existentialism to America and a couple of rooms away, if I ever wanted to dip my toes into radical political philosophy, there was always the iconic Sidney Hook. And if that wasn’t enough — if I wanted to hear the latest thoughts of the new rising star of philosophy, — the then young Arthur Danto (to whom the idea of becoming a famous philosopher-critic of postmodern art had not yet occurred) — there was the hallowed Green Room.
My favorite was Paul Edwards, so much so that when I learned decades later from a mutual friend that he had just died, I took it as a personal loss. It seemed, as though yesterday, that I could remember my confidence soaring whenever he seemed especially pleased by one of my questions, or when I was once told that he had privately cited me as one of the best students he had ever had. None of which stopped him, however, on my off days, from informing me, in front of my classmates that something I had said was, “as clear as mud”.
It was Paul Edwards who introduced me to skeptical thinking, who was a living example of what it meant to dedicate yourself to the pursuit of truth, but who recognized the need — (perhaps to preserve one’s sanity) — to balance profundity with humor.
In short, like his hero Einstein, he seemed to like jokes, comic asides, ironic remarks, almost as much as deep truths. Thus, he was the author of what to my mind was the single most memorable quote I had heard in my undergraduate year: after doing his best to explicate a revered but ridiculous sounding tenet of nineteenth century idealism, he suddenly sighed, lay down his chalk, and said “You know, these are the kind of ideas that only someone in a philosophy class could ever believe.”
Inevitably, skeptical thinking would lead to point of view, to context, to nuance. Years later, as a candidate in the American Institute for Psychotherapy and Psychoanalysis, I would learn firsthand the folly of assuming you can read the mind or thoughts of the other. To sit in a class, a group, say a seminar on dreams, and to listen to the same dream interpreted six different — but equally plausible ways — is a sobering experience. Demoralizing at first, it can in time be liberating to see the same case presentation, the same person, the same psyche conceived in ways that seem unimaginable. You learn in short time that it is not so much philosophy but your philosophy that is required: a philosophy personal enough to sustain you but flexible enough to accommodate the pluralism of viewpoints you are bound to encounter. You need experience, fine-tuning, and a certain instinctive grasp of the surprising complexity of your own inner world.
The great flaw of the reductionistic mind is that it tries to get too much from too little. There is nothing wrong with trying to milk quantifications, compressibility and algorithmic thinking for all it is worth. It is another matter to try to get nuance from numbers. Even Richard Feynman, — whom Hans Bethe called the “greatest calculator he ever met” -had to admit that the fundamental equations of physics do not (yet) contain qualia: e.g. the wetness of water.
All of which is encapsulated in the apocryphal confrontation of the woman and the cosmologist. He, in the Faustian spirit of our remorselessly scientific age is searching for the most measurable, yet abstract model of our universe. A model so dynamically rich it is forever changing. She, in her need for certainty, insisting on the unequivocal if preposterous image of an infinitely regressing series of giant turtles.
To maintain such an out outlandish viewpoint, it is necessary to be immune to the embarrassing sea of inconvenient truths that rise up against it. This the woman does with such intrepid brio that she became not only comical but unassailable. There is no amount of compelling counter evidence to which she cannot respond with yet another giant turtle. So convinced is she, that it does not bother her that no one has seen such a fabulous creature; that there is absolutely no evidence that there is a series of such beings — each one supporting the universe above — stretching backwards without end through all of time.
She does not claim that she knows cosmology. That she understands mathematics. That she grasps the necessary physics. She only knows that she is right. And nothing else matters.
Now fast forward to the twenty first century, a world in which there are literally thousands of brilliant thinkers in a dizzying array of emergent disciplines. Daniel Dennett is one of them. He believes, like the others, he is on to something big. His typically brilliant new book — Intuition Pumps and Other Tools for Thinking — is a dazzling compilation of what he considers his best insights. They are the product of decades worth of profound thinking about thinking. In other words he is a cognitive psychologist who is also a computer analyst. Above all, he is a philosopher’s philosopher, someone who does not make discoveries, who neither proves nor disproves rival theories and who is himself famous for creating breathtakingly original, thought experiments.
But what I find endlessly provocative — even more than his encyclopedic, overpowering intelligence is his seemingly supreme self-confidence. Of the multitude of brilliant minds I’ve encountered, he is hands down, the most doubt free. In the half dozen books, I have read by him, over a span of twenty years he has never presented himself as less than in thorough command of his material. As a lifelong student of human nature, I know this cannot be the whole story. No one can be one thing all the way down. Somewhere along the line, complexity, conflict, difference and yes defense mechanisms must enter the picture. That said what I want here is to analyze the ideas, not the mind, of Daniel Dennett (who is a favorite writer of mine and from everything I know about him, a splendid person).
Those ideas, however, are perhaps the most counterintuitive I have ever encountered, outside of quantum mechanics. By way of contrast and in its defense, even quantum mechanics can always fall back on its equations: the most astonishingly precise and repeatedly confirmed in the history of physics. To his credit, Dennett is as candid about what he does not do as he is about what he has achieved. He does not offer new facts, new experiments and testable predictions. He cannot therefore be falsified. He cannot offer proofs. He cannot offer science. What he can do is demonstrate in example after example the philosophical tools of one of the sharpest minds around. His ability to unpack ideas — to think about thinking — is uncanny. Few concepts can withstand his sustained assault. His specialty is to turn cherished theories on their head. You are worried about what David Chalmers calls the hard problem in consciousness studies: how could a three pound lump of brain tissue ever produce something as nonphysical as thought? Dennett’s answer: because it is an illusion, a self-induced trick to think that thought is a nonphysical phenomenon. You can never be convinced that even the most sophisticated computer can ever be conscious in the sense that humans are? That is because you do not understand the revolutionary import of Alan Turing’s discovery — that thinking itself can be mechanical: which means that artificial intelligence machines can not only think, they can, in certain simulated tasks (like play chess) outthink humans.
Two of the most extraordinary books I have read in the past twenty years are by Daniel Dennett: Consciousness Explained and Darwin’s Dangerous Idea. Consciousness Explained is perhaps the best book I have read that makes the case for artificial intelligence. His ability to analyze and demonstrate human thought is simply breathtaking. His fund of original ideas seems inexhaustible. It is a book that I would not hesitate to recommend to anyone; a book, whether you agree with it or not that is endlessly provocative. It is also a book that at the end left me not only profoundly unconvinced — but mystified as to the mind that had produced it.
By contrast Darwin’s Dangerous Idea is about evolutionary biology, not artificial intelligence. Right from the start Daniel Dennett makes it clear that he considers Darwin’s idea of natural selection working on random mutations to be the greatest single idea that any scientist ever had. So far so good. Darwin happens to be one of my intellectual heroes and I was excited to see what Dennett was going to do with it.
What Dennett did — in an audacious move — was to attempt to marry evolutionary theory with his own brand of cognitive psychology. To achieve this, he introduces two novel terms: skyhooks and cranes, conceptual tools that mentally perform an analogous function — in the realm of theory building — that their physical counterparts provide in the mundane physical world. Thus cranes, by leveraging immense mechanical forces manage to lift objects too heavy to be hoisted by a single concentrated effort. They show what can be accomplished by steadfast, cumulative pressure.
It is Dennett’s contention that the natural selection of adaptive, random mutations is, in effect, an evolutionary crane. Without foresight, without intelligence, — operating silently with dumb automaticity across vast stretches of geological time — it manages to construct marvelous life forms- that have the appearance of design. Dennett likens the process of natural selection operating on random mutations, therefore to nothing less than an algorithm, an algorithm that can be trusted; an algorithm that can demystify the attempts of contemporary creationists to demonstrate that the world we inhabit could only have been made by an intelligent designer (ID). Thus Dennett, by a process he calls reverse engineering strives to reduce the profound dynamics of evolution to the workings of a dumb (because completely unconscious) but brutally effective machine; think of Deep Blue — that does not even have a child’s sense of what it means to play chess but is able to scan billions of moves per second.
My point is simply that Daniel Dennett, brilliant through he may be seems to have forgotten that the indispensable components of human intelligence are: meaning as the existentialists have time and again made clear; emotion as neurobiologists, neuroscientists, and cognitivists are increasingly recognizing; and the influence of the dynamic unconscious that for the first time is being taken seriously by experimental psychologists.
In what follows, I therefore attempt merely to complement what he is saying and I do this primarily by incorporating the psychodynamic factor.
“Love is Just a Word”
Dennett is right that this is hardly a profound cliché: it means either that love is an ambiguous word whose meaning is hard to establish or that it refers to a certain kind of relationship that primarily resists being pinned down. What is important according to Dennett, is that any mystery that may be thought to reside in the universally popular word ‘love’, has thereby been thoroughly- semantically and philosophically debunked and demystified — by such careful analysis.
A moment’s thought plus the use of the psychodynamic method, shows just how premature this conclusion is. First, there is a reason “love is just a word” is a universally popular cliché and that is because it contains or points to more than just a kernel of truth. As a therapist, who has listened to just almost every usage of love that there is, I can attest it is indeed an extraordinarily ambivalent term; one that is used in a multitude of ways many of them blatantly self-contradictory in its quest to define a kind of imaginary ideal relationship that does not seem to exist.
But by focusing primarily on the semantic confusion informing the many usages of the work ‘love’, Daniel Dennett ignores everything that underlies that confusion. By sidestepping everything but cognitive semantics, Dennett deletes the powerful dynamic of affective attraction and repulsion that can contribute to a deeper understanding of love. People do want to understand what it is that their heart desires, to whom they most want to be affiliated, and what are the things that keep getting in their way.
One wonders why it is that genuine love or passionate engagement rarely endures? The reasons are myriad but one of the biggest obstacles to a more steadfast grasp of the intricacies of love is simply the simple but profound puzzlement people feel when it comes time to distinguishing the difference between “wanting and shoulding”. In other words do I love my wife, my child, my parents, my job, my country, my religion, because I want to or because I am supposed to?
“My Daddy is a Doctor”
Meaning says Dennett is not a unitary acquisition, it is built up piece by piece. When a child refers to her “daddy” we assume too much to think we know what she means. Is she referring to her biological father, her stepfather, or simply a father figure who acts in a loving way? Does she know there can be many kinds of father? Whatever she knows, says Dennett, give her time and she will know more. Furthermore, we never believe just one thing (about anything). If I believe, for example, that you have a cat, I also (have to) believe the cat has (in order to be called a cat) four legs, two eyes, a mouth, etc.
But “No,” says the philosopher, Fodor, “I can imagine a person who believes just one thing. As a thought experiment I can imagine a world, a world space with just one atom.” What I myself cannot imagine is how such thing can come about, or how a person can become someone who can actually experience and therefore believe only one thing. Daniel Dennett, however, is doing little more than asserting his belief. Because he is so disdainful of the psychodynamic, contextual point of view he does not need to show, to present a compelling narrative of what he believes. Cleverness plus his formidable analytic skills will suffice. Once he has exposed the cognitive dissonance, the hidden contradictions of a train of thought, his work is done.
In that spirit Dennett asks us to imagine someone (Tom) who has this belief: “I have an older brother who lives in Cleveland.” Furthermore, we are told it is a belief that has been implanted (unbeknownst) to Tom by a brilliant cognitive microsurgeon. Such a person has no memory of the implantation, but continues to have the persistent thought, “I have an older brother who lives in Cleveland.” This being one of Dennett’s beloved thought experiments, we are also told that Tom, in reality, has no older brother and that the belief is false.
Nevertheless, Tom can not shake the thought and — because of the wizardry of the cognitive microsurgeon — the resulting belief that in some strange way, he really does have an older brother who lives in Cleveland. Dennett asks us to take this thought experiment seriously (as he certainly does). He asks us to imagine Tom’s dilemma when sitting in a bar and a friend asks do you have any brothers or sisters? It should be noted we are to assume that Tom’s rationality was in no way damaged when the microsurgeon inserted the false belief. So should Tom reply — as the operation spontaneously prompts him to — “Yes, I have an older brother living in Cleveland,” — he will immediately realize something is terribly wrong. To wit; he does not have an older brother living in Cleveland; he never did, he does not know any such person and does not know why he said it.
What Dennett is saying is that belief implies mental competence. It implies you understand not only what you are saying but that you have a repertoire of uses of that belief that are not contradictory and that do not flout reality. This is Dennett at his cognitive best — breaking down the mental world into bits of behavioral data, that are progressively smaller, simpler and as he likes to put it “dumber and dumber.”
Once again we see his strategy. He starts with the mental and by a process he calls reverse engineering, he works his way down to the brute physical, the dumbly mechanical, the machine-like. From such a perspective there is no need for the evolutionary psychologist’s celebrated theory of mind. The child who is learning to ride a bicycle does not, says Dennett, have or need a theory of how to ride a bicycle. Perhaps an autist like Temple Grandin will require a theory but a normal child will learn by doing, will rely on what is called folk psychology, and will acquire a theory, if he or she ever does have one afterwards.
The reader will not be surprised that this, I believe is exactly what Daniel Dennett himself does — in other words he is hoist by his own petard — when it comes to the project for which he is perhaps most famous — artificial intelligence. On the one hand he readily admits (being a passionate Neo-Darwinist) that the process of natural selection operating on random mutations, being devoid of foresight, can not engineer in advance necessary adaptations. On the other hand, he expends boundless creative energy on the problem of just how to fine tune a working design of an intelligent machine. It is no small irony that Dennett attempts this (as does the entire AI community) by eventually bypassing biology and evolution. Instead armed with only the theory of strong AI, he will settle for the laboratory tools of the computer scientist.
To do so, is to engage in what I call Dennett Denial. This is when — rather than offer unambiguous experimental proof, substantive evidence or watertight argument in support of your hypothesis — you put most of your effort into ridiculing the position of your adversary: i.e. famously in the case of Daniel Dennett the claim that no computer can ever be, in any meaningful sense, artificially intelligent.
It is worth noting, that to disprove your opponent’s thesis is part of a successful defense in science, — equivalent to showing that a certain experiment allegedly conclusively confirming their adversarial position — is upon closer analysis thoroughly bogus. As important as this is, it does not compare to the lion’s share of the work when it comes to breaking new ground in a particular area of science. It is this that Dennett does not do and to my knowledge has never done. On the contrary he assumes that artificial intelligence has already been achieved: on the celebrated PBS series The Glorious Accident, Dennett, when confronted with just this question, asserted without hesitation that he “could build one now (an artificially intelligent machine) if I had sufficient money and materials.” When then asked if he could ever envision an intelligent computer that also had emotions he was equally unimpressed: not only could he imagine such a thing, but work in that direction — i.e. teaching a machine to give feedback on what was presently happening- was already underway.”
In other words, it is possible to program human evolution. To make such a claim is to fail to see the elephant in the room: in this case that to date there has never been a remotely intelligent AI machine. As Brian Christian — author of the wonderful The Most Human Human notes, — the gold standard for determining artificial intelligence in still the Turing Test, and that, by wide consensus, is woefully inadequate. And as I first remarked twenty five years ago in my essay “A Psychoanalyst takes the Turing Test” expert impersonation is not a proof of identity. Trying to pass the Turing Test is like trying to get away with a forgery — a forgery of human intelligence. It is to deliberately ignore everything in the human mind that is not computational, algorithmic, quantitative, operational and replicable; i.e. what is commonly called emotional intelligence, free-associative thinking, creativity, reflecting, musing, wondering, fantasizing and so on. For all of which I admit I was on board, at least in principle, with the famous philosopher Thomas Nagel’s remarks, in his classic work, The View from Nowhere: “centuries from now we will look back on AI as a gigantic mistake.”
Daniel Dennett, in short, is a first class philosopher himself who- although he can offer masterful critiques of established patterns of thinking and is a virtuoso when it comes to creating new ways of looking at old questions — has made no demonstrable discovery of his own. Considering the range and breadth of his studies it is striking that he shows no discernible interest in the psychodynamic perspective. He believes that what is described subjectively, as qualitative interior feeling can ultimately be viewed in physical behavioristic terms. He attributes the divide between mental and physical in large parts to Descartes (and the problem of the “Cartesian theatre”. This to Dennett is analogous to the millenniums long inability (prior to the arrival of Bottzman and especially Einstein in 1905) to establish once and for all the physical nature of the alleged but heretofore invisible atoms.
In Consciousness Explained, Dennett admits that Cartesian dualism is so entrenched in Western thought that it will take a mighty effort to dislodge it. Yet that is what he believes he has done in the above book. From such a lofty view of his own contribution to the mind-body problem, Dennett likens his legion of critics to those Luddites, who, lacking all technical savvy, saw computers as performing same kind of calculating “magic.” They do not realize, computers are only “machines.” Analogously people who believe in consciousness (99.999 of the human race) do not see there is nothing in the world of the mental that ultimately can not be reduced (i.e. explained) to the behavioral, the computational, and the algorithmic.
But Dennett goes much further than this and that is one of the reasons, no matter how much one disagrees with him, that he continues to be fascinating. Somehow, he has managed to convince himself that the classic mind body problem is at bottom a self-induced illusion: perhaps by slipping in so as many anthropomorphic attributions to the nuts and bolts vocabulary of computers science it now becomes meaningful to ask questions such as what do neurons want? As though neurons have wants. As though it makes perfect sense to discuss how much “memory” a given computer has? How much information can a computer can access? How fast it can think? How well it can compete with an intelligent human being visa-a-vis. answering a specific set of challenging questions? To actually wonder to what extent, and how much of thought processes of an advanced artificially intelligent machine can occur consciously or unconsciously? As though a computer can have an unconcious as well as a conscious. As though it is possible to program not only emotion but emotional conflicts — and therefore meaningful feelings — into inanimate computers.
Why one wonders or at least I wonder does Daniel Dennett work as hard — and to my mind no one in the world does it to better — to convince us that there is no such thing as a mind? And at a time in history when the cumulative powers of the human intellect have reached an all-time high?
I am reminded of something that Richard Feynman, the great quantum physicist, once said: that unless he could build something — he didn’t understand it. Feynman, of course, was not talking about the human mind, he was talking about the fundamental principles of particle physics. In other words, he was talking about physical reductionism in its present, simplest and most radical form.
As of this writing, no one as far as I know, has remotely succeeded in reducing a thought to a physical thing. That is as true today as it was when I first heard it as an undergraduate in Paul Edward’s philosophy class. So impressed had I been then that I immediately tried to apply my wonderful new insight to my own mind. I would think of a thought, a recent one, the more idiosyncratic the better, and then do my best to reduce it. The thought which I chose — sports related, and connected to my father, who was a baseball fan — was this. It concerned the pitching immortal. Bob Feller, the first man reportedly who could throw a baseball (in the 1940s) over 100mph. How did they know? According to an old newsreel which my father saw, by having Feller release his fastball, at the exact moment a man on a motorcycle travelling over 100 mph flashed by his side. It didn’t matter that the stunt sounded hokey — if not fantastical. What mattered is that the picture and the thought it represented, stayed with me even more since my father told it to me.
In my mind as I thought about this, I saw a blurry motorcycle cop travelling along at 100 miles per hour. I pictured an iconic, fireballing pitcher named Bob Feller in his prime. I pictured Feller’s fastball going faster than the motorcycle, as measured by a precision camera. I saw the whip like right arm that could throw such a thunderbolt. And I considered how different this imaginary thought was from the reality it referred to. How different, for example, was the idea of a motorcycle travelling at 100mph from a real one. For the idea has no speed, it cannot be measured in the sense the firing of a synapse can be calibrated. The thought of a pitcher throwing a fastball has no dimensions, no weight, no color, no sensory qualities, no felt immediacy. There is nothing to see, to feel, to measure, to hear, to smell. There are no equations, no fundamental forces of nature to which it can be reduced. There are no atoms, molecules, and electrical charges, no underlying quarks that have been detected and need to be carefully described. The particular imagist thoughts I have presented cannot be found in physical space. They occupy a peculiar interior world — varying from person to person and cannot be replicated, — that is akin to mental space.
None of which seems to matter to the AI community. To think otherwise, to conceive of an inscrutable gap between brain and mind is to treat our biological brain as though it is wonder tissue. It is tantamount says Dennett, to “giving up, of assuming that it is a mystery that can never be solved.” (p.10) Here Dennett treats the classic mind body problem as though it is equivalent to saying there is a palpable supernatural presence in the world. As though to talk about thoughts, ideas, feelings, consciousness — the mental world — is like talking about the presence of ghosts, gremlins, angels, and spirits in the external physical world.
In short Daniel Dennett acts as though all the evidence of recorded history, properly understood. points in the direction of consciousness being made of strictly material things and to claim otherwise is to be under the spell of a self-induced illusion. He, unfortunately, has it backwards. The burden of proof is not on everyone who is not in AI to prove that a given computer can not think. To say that the mind-body problem is a great mystery — in the sense that quantum entanglement, dark matter, dark energy, the inexplicable fine tuning of the startup conditions of the Big Bang can not presently be explained — is only to note that substantial mysteries continue to exist. It does not mean that there are supernatural forces at play in the interstices of known reality. It does mean that there are telltale gaps in our current knowledge which not only resist our best efforts to explain them but may point to the need for a different explanatory mode.
From this standpoint, Dennett and his cohorts in strong AI, are simply adamantly refusing to admit that there is a mental world, a non-physical consciousness that unquestionably exists. That does not mean that consciousness is not concomitant with a physical substratum. Nor that one day it may eventually be reduced to a strictly physical monism.
It does mean that for the present, the elephant in the room for AI is that there never has been the slightest evidence for their central claim. No one has ever devised a computer software program that could remotely simulate the complexity, breath, depth and the phenomenological multi-layered richness of ordinary human thought. Not to speak of the even more perplexing question — will there ever be a software program that will give us the equivalent of computer dreams, nightmares, personality quirks, attention deficit disorder, grandiosity, narcissism, mood swings, character disorders? Will there be a day when artificial intelligence becomes so sophisticated that we will worry — if a certain personal computer to which we have become attached — is showing signs of depression?
What I find hardest to understand is how someone as remarkably intelligent as Daniel Dennett can convince himself that the difference between human intelligence and machine intelligence is, from a fundamental epistemological sense, trivial. Certainly Dennett in the half dozen books I have read by him has never taken the trouble to tell us. It is as though the central truth of what he calls strong AI is so self-evident that it would be a waste of time to dwell on it. Better to convince anyone who will listen that- belief in the existence of consciousness- is the end product of a nest of self-induced illusions.
Daniel Dennett can be masterful in demonstrating in stunning detail how a computer can perform a certain formidable computation. Afterwards, he will not hesitate to gloatingly point out, “Look, no ghost in the machine! — no wonder tissue- just some ordinary software and hardware.” He seems oblivious to the fact only a diehard believer in strong AI would ever compare the electronic circuitry of an ordinary computer to the profound workings of the human mind. Incredible though it may seem, Daniel Dennett seems to place the “discovery” by Alan Turing of what he calls “mechanical thinking” on the same plane “as the Big Bang model of the origin of own Universe”.
In a fantasy debate my proxy has this comeback after Dennett triumphantly announces, “Look, it’s just a machine… just software/hardware… no wonder tissue!”- “You’re right there’s no wonder tissue, THERE’S NO MIND EITHER…LIKE YOU SAY- IT’S JUST A MACHINE!” Here Dennett, is like the man in Wittgenstein’s famous analogy who, staring at a statue of a man in a garden — having forgotten it is just a statue — keeps waiting for it to move.
None of which impedes Dennett in his quest to find a machine model for the human mind. He notes, for example the difference between intrinsic meaning and derived meaning. The human mind supposedly has intrinsic meaning while the artifacts it endlessly fashions have only derived meaning. No says Dennett, human beings when you think about it, are just the artifacts of evolution (thereby taking a page out of Richard Dawkin’s The Selfish Gene). Here Dennett exercises the second prong of his two prong attack against the existence of a non-materialist consciousness. On the one hand he relentlessly reduces — deconstructs — demystifies the human mind until it is seen as just the final artifact of a multi-layered system of parallel computational processes. On the other hand, he likens it to the most recent representative and survivor of a primal evolutionary process that has been going on for millions and millions of years. His strategy seems to be to not only blur but destroy any hypothetical line between mind and body.
It may be that I am incapable of imaging a world that is devoid of consciousness. But let me speak on behalf of thirty years of intensively exploring the consciousness of literally thousands of diverse patients. Of immersing myself in the literature of the field of neuropsychoanalysis. Of faithfully attending over the past twelve years one hundred professional lectures by cutting edge world class neuroscientists, neurobiologists, psychiatrists, psychoanalysts, neuropsychologists, experimental psychologists. They include Nobel Laureate, Eric Kandel, Joseph Le Deux, Antonio Damasio, V.S. Ramachandran, and Jaak Panksepp. While they disagreed more than they agreed, I can honestly say not one ever mentioned strong AI as having any relevance to the mind body problem. Not one ever referred to Daniel Dennett’s 1991 magnum opus, Consciousness Explained, in which, as mentioned, he triumphantly announced he had for the first time cracked or solved or dissolved the centuries old mind-body dualism. (More than once, I would have liked to hear the opinion of acknowledged world experts on the origin of consciousness according to this contrarian but undeniably brilliant book.
Had such a conversation ever ensued, I for one would have wanted to know how a machine that is not self-aware, devoid of consciousness, a complicated circuit board of electronic processing can be in some equivalent sense as dynamic, animate and alive, as a human mind. What is even more frustrating is that the question is not even addressed. Instead a machine that passes the Turing Test — the gold standard for evidence of true artificial intelligence — is presumed to be alive in the sense that all intact human functioning brains are alive. So the question for me becomes — is a functioning AI computer dead or alive? On what basis do you claim that a calculating number crunching machine in any meaningful sense is emulating a significant aspect of human thought? How do you refute the biologist’s assertion that thinking, from everything that is known, is impossible without a viable brain? Isn’t this tantamount to building from the bottom up, in a laboratory, from purely inorganic materials — a fully functioning thinking entity that can really do — not simply mimic — everything that a human brain does while alive?
It follows that radical cognitivists such as Daniel Dennett aspire to do no less than to build a human mind. There is, it seems, to me more than a little hubris in such a project. How, for example does one program experience… the sense of being embedded in a time — bound world?
Dennett, as mentioned, belittles the importance of the so-called theory of mind. A boy, he says, rides a bicycle not because he has a theory of mind about how to ride a bike, but because he learns by trial and error how to balance himself from falling or swaying side to side while moving forward. Only after he learns to ride a bike, does he construct a theory of it. The same thing applies to what the psychologist calls the theory of mind. First we learn about our own intentionally — then we apply it to others. We are not born with a theory of mind. We build it up, by trial and error, piece by piece, from our experience in the world.
Thus Dennett, as previously mentioned, arrives at his philosophic distinction between intrinsic intentionality and derived intentionality — the products of technology, utensils, cars, machines, etc. — are artifacts specifically designed to meet the designer’s intentions. By way of contrast, intrinsic intentionality refers to the essence of the object — its basic meaning. But this, however is what Daniel Dennett, empathatically refutes. Meaning, he insists, is always derived. All we can say is that the intrinsic meaning of our most personal thoughts and feelings is somehow derived — in ways we cannot fathom — from our long evolutionary heritage.
We see that Dennett, in his unswerving loyalty to the philosophy of strong AI, wants to demystify language and meaning, separating it from any trace of mentalism. He wants it instead to be something an AI programmer can design: i.e. something that is ultimately material, concrete, something that can be put into a machine, something that far from being transcendental can be understood, explained and designed according to fundamental mechanical principles.
As noted, in Darwin’s Dangerous Idea, Daniel Dennett — in his attempt to explain the phenomenal success of evolutionary biology — introduced the terms cranes and skyhooks. Cranes were biological devices or adaptations that (like their machine counterparts) worked on the principle of maximizing small increments into palpable improvements. Darwin called this the summing up of small effects. Add time, geological time, to this equation and you begin to appreciate the power of natural selection. By way of contrast, skyhooks were the biological analogue for saltation: broadly discontinuous variations and transformations of one organic form into another, or radical mutations. Dennett’s point is not that such mutations do not occur, but that they occur rarely, improbably. To imagine that evolution occurs systematically in this way is like imagining that enormous loads are miraculously transported by suddenly appearing skyhooks: i.e. cranes from above with no apparent foundation. Not surprisingly, Daniel Dennett sees the creationist theory I.D. (intelligent design) as a bogus (skyhook) attempt to undercut the dazzling explanatory power of Darwinian natural selection (the summing of small incremental effects over vast geological stretches of time).
Daniel Dennett is not shy about adding new terms to his new ideas. Accordingly, design space is the space of all possible evolutionary designs and designoid refers to those natural selection successes — now referred to as biological adaptations — that are best viewed as nature’s designs (although mindless). Here we can see Dennett laboring to forge a synthesis of neo — Darwinian thinking with the computer scientist’s paradigm of an electronic-processing machine. The hybrid he comes up with is cranes and skyhooks. Marshalling all of his philosophical sleight of hand Dennett tries to blend biology and computer science. Here I will admit that it took me years of reading both Darwin and Dennett before I realized that not once did Darwin ever imply that the model upon which he was basing his theory of natural selection was anything if not biological: (far more based on the lifestyle of the barnacle (cirripedia) — on which he spent 8 years than on the spectacular triumphs of the burgeoning industrial revolution that was all the rage in England.
There is a sense then, in which Dennett is perhaps too brilliant: a leading cognitive philosopher — who is bewitched by the computer model of the brain — yet who considers Darwinian evolution to be the greatest idea anyone ever came up with in the history of thought. He seems certain that such a workable model of a mind-brain synthesis can be forged. No obstacle is too great. He delights in constructing and then mercilessly deconstructing seemingly formidable obstacles to his visionary view.
Thus “stotting” — the well documented tendency of some gazelles when being chased by a predator, to conspicuously leap high into the air in their attempt to escape their pursuer — is actually beneficial, from the standpoint of survival. That is, gazelles who stot seldom get caught and eaten. Why? Because the signal is being sent to their pursuers — that since it is so obvious they are the fastest and fittest of the herd — they will be much too hard to catch. Dennett calls this “a free floating rationale” and his point is that a predator such as a lion need not understand it, just as a computer need not understand a particular software — so long as it is compatible with it, — so much as instinctively follow it.
In the same vein, Dennett attempts to demystify the mechanism by which the frog’s eye relays to the frog’s brain that a fly is within reach. His point once again, is that there is no intrinsic meaning being conveyed that the frog’s brain needs to understand. There just has to be a fly-detector mechanism “appropriately wired to the hair trigger in the frog’s tongue” (p256). In his explanation, Dennett does not mention Konrad Lorenz who pioneered the concept of the IRM, (innate releasing mechanism). Nor does he mention that ethology, upon what he relies so heavily — except in his theories — has no discernable link to artificial intelligence. (To my knowledge, in the many books, I’ve read by Konrad Lorenz, the concept of artificial intelligence does not appear).
None of which deters Dennett. Citing the ubiquitous presence of the Krebs cycle in evolutionary biology, he likens it to the repetitive cycles and feedback loops of the functional computer. Impressed by the explanatory power of Richard Dawkins’ famous concept of “memes”, he calls them “a great intuition pump.” He is unfazed by the distinction between a philosophical zombie- — a fictional being who behaves exactly as you or I; but who is actually is a machine, having no inner mind, no unconscious and no idea of what a self or an unconscious would be like — and a “zimbo”. The latter is another fictional being, who is also a machine, but who has been programmed to have recursive self-representation, even unconscious ones. Amazingly, Dennett challenges the reader to derive a test that could determine, conclusively, which of the two they were.
There is an ironic sense in which AI philosophers like Daniel Dennett, are like blinkered academics who are convinced they have found an ironclad theory of mind (one that typically leans heavily on Alan Turing and his celebrated Turing Test). They therefore believe they are capable of building a machine model of the human mind. And in this sense, they are like top down engineers — the very opposite of how evolution, which is blind and has no foresight — works.
From this perspective, his spectacular intellectual fireworks notwithstanding, Daniel Dennett seems to want his cake and eat it too. When evolution serves him he believes in it. When it doesn’t, he switches to Richard Dawkins’ cultural memes and serves a dazzling array of imaginative thought experiments that he seems to delight in. Not surprisingly, he is comfortable substituting silicon and other materials for nervous tissue. He is content to ignore emotions and pretty much all of what is called affective neuroscience. He has forgotten that people become themselves analogous to the way that they learn to ride a bike — not from a theory of how to be a person, or how to think like a person, or how to build a person but via the myriad processes of developmental psychology and neurobiology.
From the standpoint of my own travels through the fields of academic philosophy, literature, and a brief stint as the recipient of a four year engineering scholarship (which I would later relinquish) and an eventual career as a psychotherapist and author I can say this: Daniel Dennett seems to operate as though logical positivism — the project to translate everything meaningfully mental into a series of utilitarian behavioral statements — that reigned in the fifties, is still viable. He acts as though the mechanics of nineteenth century physics still reigns.
How one wonders, can Dennett aspire to design a computer that can model not only consciousness but much of the functional brain if he has never, qualified as a professional neuroscientist, neurobiologist, neuropsychologist, or psychotherapist.
Undeterred Daniel Dennett sees all such attempts to limit the boundaries of all artificial intelligence as evidence of a failure of imagination. For all his inexhaustible intuition pumps, his undeniable philosophical fecundity, he seems himself unable to imagine that there might be something in the universe other than the physical. That the phenomenon of the mental, of consciousness, of subjective experience may force us, whether we want to or not, to rethink (e.g. Penrose) our existing physics.
