Damn you, David Chalmers!
This is a paper I have written for philosophy of mind. There's some use of technical terms, but I think it should be readable for a non-philosophical audience.
I was particularly proud of the review comments I got on this one: First time I’ve seem a my little pony reference in a philosophy essay, but the example is good none the less ... well-argued and entertaining to read which is always a good thing
The topic is David Chalmers "zombie" argument, and the challenge it poses to physicalism (the idea that the mind is purely physical, i.e. nothing more than the nervous system).
I've left out all the citations, to make it easier to read.
Damn you, David Chalmers!
I’m going to have to start this essay with a bit of a confession: I have no idea.
This is generally considered pretty bad form in philosophy essays, where one is expected to take a stance on the issue at hand, and make as solid an argument for that stance as one can. But when it comes to Chalmers’ Zombie Argument, I can only throw my hands in the air and admit that, I just don’t know.
What’s especially galling is that when I’ve previously covered this topic in an introductory metaphysics course, I’ve been a pretty hardcore physicalist. It seems so obvious, after all, given the correlation of brain states with conscious states, such problems for dualism as causal closure, and the powerful principle of Ockham’s Razor, that “we have no choice but to accept materialism as the only suitable theory of mind.
But philosophy and science are often bedevilled with what Robert Anton Wilson dubbed “Damned Things”: those niggling problems that defy easy resolution. Just when the Damned Thing we called consciousness seemed fairly comprehensively cornered in the physicalist camp, along comes Chalmers with his Zombie Argument that lurches along like one of those Hollywood creatures that refuses to lay down and die.
But of course, Chalmers’ Zombie is nothing like a Hollywood zombie, nor even a traditional Haitian voodoo zombie. Rather, Chalmers’ zombie is an influential thought-experiment first proposed in his 1996 book, “The Conscious Mind”, as a challenge to the reductionist theory of physicalism.
Physicalism is the idea that all mental states, up to and including consciousness are simply physical states and functions of the central nervous system; specifically the brain. Consciousness is said to necessarily supervene on the physical: given the requisite physical states of a person, consciousness is logically entailed. In this way, physicalists argue (see, for instance, Jeffrey Smart) that mental states (including consciousness) are identical to brain states.
Chalmers’ Zombie Argument is a direct challenge to this idea. Chalmers proposes the possible existence of a “Philosophical Zombie”: a perfect physical replica of a person, atom for atom, but which lacks consciousness. The Philosophical Zombie acts just like an human ordinary human being: it walks, talks, senses and interacts with the world around it, but its experiences lack qualia, the “what-it-is-like” inner sensations that, for us conscious beings, accompany experiences: the thrill of seeing the sun break through the clouds, the exquisite taste of chocolate, the heartbreaking sweetness of a minor key. For the Philosophical Zombie, to use Chalmers’ own phrase, “it’s all blank on the inside”.
The challenge to physicalism from Zombies lies in the fact that the Zombie shows that it is possible to conceive of brain states existing without consciousness. Physical states do not logically entail, a priori, conscious states, in the same way that the state of being a horse and female logically entails being a mare. If physical states do not similarly logically entail consciousness, then consciousness is not identical to brain states. Not only this but physicalism must be false because physicalism has to deny the conceivability of beings that are physically identical to us but not conscious.
At first glance, this seems a bizarre argument. Zombies (of the philosophical kind) just don’t seem to exist, and even Chalmers concedes that it seems to be a fact that consciousness is always correlated with brain states. But this is an a postierori observation of the “horses do not fly” kind. But as the popular cartoon character Rainbow Dash shows, a flying horse is at least conceivable, if not probable. But Rainbow Dash the male mare is not.
So it is conceivable that there could be identical brain states that are not correlated with consciousness; thus a Zombie could exist. And, as Philip Goff observes, “what’s possible has implications for what’s real”. The implication of the Philosophical Zombie for physicalism is that, if consciousness does not, a priori, accompany brain states, then physicalism is false.
Indeed, in a very limited way, it could be argued that Philosophical Zombies of an unsophisticated kind do exist. Chalmers uses the example of the Google Car, or Watson, the computer that beat the human champion at Jeopardy! Both of these machines perform at least the rudimentary functions that creatures with brains do, albeit in the most primitive way (in Chalmers’ words, ”maybe we’re up to earwig level, or earthworm level”), yet I doubt anyone would claim that they have any kind of consciousness. For instance, that the Google Car has any qualitative “feel” about its environments as it negotiates it, or that Watson had any feeling of triumph at winning.
James May demonstrates another kind of (primitive) Philosophical Zombie, when he compares a tune composed by a computer, programmed to emulate Beethoven, with an actual Beethoven composition. One, he says, is plainly the work of Beethoven, the other “merely the vomit of a digital contraption”. Like Searle’s Chinese Box, the computer has analysed the data and produced a logically coherent result, but there’s no “something extra”. Beethoven wrote the “Moonlight Sonata” in the throes of unrequited love and, as May notes, “you can tell. There’s something going on there that defies purely logical analysis”.
That “something” would be qualia: “what-it-is-like” to be a musical genius hopelessly in love. The Beethoven-emulating computer is, in a very primitive way, a Philosophical Zombie.
Needless to say, the Philosophical Zombie argument has attracted trenchant criticism from defenders of physicalism.
Patricia Churchland dismisses the Zombie argument as “a demonstration of the feebleness of thought experiments”. We can imagine anything is possible, but that doesn’t mean that it is possible, or particularly enlightening. But Churchland’s objection here seems to be missing the point, as her example of Crick’s “Thermodynamic” thought experiment illustrates. The Thermodynamic thought experiment imagines a world were heat is not identical to mean molecular kinetic energy. It is true that that is “feebleness itself”, but that is because mean molecular kinetic energy is necessarily heat (Cf Kripke). If heat is strictly identical with mean molecular kinetic energy, they must be identical in all conceivable worlds.
On the other hand, as argued by Kripke, it does not seem to be the case that consciousness is strictly identical with brain states.
Daniel C. Dennett, on the other hand, seems to object to the Zombie argument on the grounds that “the purported ‘subjective qualities’ or ‘qualia’” simply don’t exist. All there are, are the performance of functions of the brain, or the dispositions to do so. (This seems a strangely behaviourist answer although perhaps not so strange, considering that, as Chalmers points out, Dennett was a student of Gilbert Ryle. Philip Goff also notices the behaviourist “flavour” of arguments like Dennett’s). Explain the physical functions and dispositions, Dennett argues, and you’ve ipso facto explained consciousness (Facing Backwards). There is nothing more to be explained.
Chalmers responds that the Zombie argument shows that there is something further than the physical to be explained: what it is like. One might contrast Dennett’s argument with the process of digestion. To explain all the physical facts of digestion: food chemistry, the human anatomy of taste receptors, the processing of information in the brain, and so on, still does not explain the ineffable experience of eating a really good tiramisu.
Dennett and Churchland both also liken the Zombie argument to the Vitalist argument, which held that the physical processes in living things were insufficient to explain how life itself could arise out of dead matter: there had to be something missing, the so-called elan vital. Ultimately, however, biology was able to show that there was no need of such a “life spirit”.
Chalmers responds that, yes, given the physical facts of reproduction, metabolism, response, adaptation, etc., life necessarily entails. All those functions are part of the a priori definition of “a living thing”. The Zombie argument shows, though, that “any physical description of a brain seems at least to be logically constistent with the absence of consciousness. So somehow you need to put an extra ingredient in there”. Brain states and functions do not necessarily entail consciousness. One can conceive of all the physical processes that normally correlate with consciousness, but without consciousness. One cannot do the same with life.
Churchland also objects that Chalmers is committing a fundamental fallacy of informal logic - an argument from ignorance. As she sees it, Chalmers argues that from the premise that because we scientifically understand little about the nature of consciousness, we can never explain consciousness, no scientific discovery would deepen our understanding of it, nor can it be explained it terms of physical properties. But I’m not sure that Churchland’s accusation holds. Indeed, it seems that she is committing, at least in part, the informal fallacy of the straw-man.
The first conclusion of the alleged argument from ignorance, that we can never explain consciousness, is clearly false. Chalmers obviously thinks that an explanation exists, but that (if it is ever found) it will be a novel one involving new physical laws. Of course, an explanation may never be found, but that we may never find an explanation of something is not in itself controversial, as even Dennett admits.
The second alleged conclusion, that no scientific discovery could ever deepen our understanding of consciousness, besides seeming to be a partial restating of the first, is also clearly false. Chalmers frequently alludes to discoveries in neuroscience which are daily, it seems, advancing our understanding of the mind, and explicitly states that “neuroscience is not irrelevant to the study of consciousness”, even if it is not enough.
The third conclusion, that consciousness can never be explained in terms of physical properties, is not an unfair characterisation of Chalmers’ stance. But Churchland gives only one premise - that of ignorance; on that structure, the argument is indeed invalid. It seems to me, though, that Churchland has excluded a second premise of Chalmers’ argument; one which is the very point of the Zombie argument: while we may be largely ignorant of the nature of consciousness, it seems that physical brain states do not logically entail consciousness.
Thus it seems that the accusation of argumentum ad ignorantam fails. But both Churchland and Dennett make a much stronger claim against the Zombie argument, one that seems to me to be the most damaging: the feebleness of Chalmers’ explanatory hypothesis.
Chalmers proposes that consciousness will be explained by introducing new fundamental “psychophysical” laws to link conscious states to brain processes. Chalmers agrees that, unlike the Zombie, it does seem to be “a law of nature … that when you get a brain of a certain character you get consciousness along with it”, but that, “you actually need these laws of nature, special laws of nature, to connect the brain to consciousness”. Chalmers suggests that conscious experience be included in the canon of irreducible, fundamental features of the universe, alongside mass, charge, spin, space-time, etc.
Dennett objects that Chalmers, unlike physicists, cites no objective evidence that justifies his hypothesis. The only data Chalmers offers is a subjective “belief in a fundamental phenomenon of ‘experience’” (Facing Backwards). Dennett complains that this is too drastic a move to even contemplate without independent evidence. Like Smart, Dennett objects to such seeming “nomological danglers”. Greg P. Hodes likewise cautions that “to accept the first hypothesis that comes to mind without any attempt at verification is to be intellectually irresponsible”.
But Chalmers is in some august company. Consider, for example, String Theory. The theory is logically consistent, but it does not make any predictions that can conceivably be tested. True, String Theory rests on a foundation of profound and elegant mathematics, while Chalmers is by his own admission left to fall back largely on subjective experience and (allegedly feeble) thought experiments. The lack of testability for String theory similarly leads some in the physics community, like Sheldon Glashow, to dismiss the theory out of hand (”not even wrong”, in the words of Wolfgang Pauli). But its logical coherence and mathematical elegance nonetheless means that it, like the Zombie, refuses to die.
At the same time, Chalmers has some heavy-duty supporters in the science community. Physicist John A. Wheeler has suggested that information is a fundamental physical property. Chalmers argues that information has correlated physical and consciously experiential aspects that might lawfully (in the sense of his psychophysical laws) connect the brain to consciousness. Francis Crick and Christof Koch as also accept, however tentatively, that information may indeed be the key for a neurological basis of consciousness. Indeed, John A. Wheeler and Roger Penrose have suggested even a central role for conscious observers in the universe.
I confess, Chalmers’ hypothesis, however tentative, makes me slightly uncomfortable. It seems a little to like the “quantum flapdoodle”, to use Murray Gell-Mann’s memorable phrase, of many New Age cranks. Chalmers’ hypothesis, though, unlike the musings of the white-robe-wearers and crystal-wavers, has a logic that is difficult to deny, maddening though it may be.
Chalmers acknowledges that, at heart, all he has is “a collection of intuitions”, although he defends these intuitions “natural and plain”. But this seems a pretty weak claim: many of our intuitions are natural and plain - and wrong. The earth is not flat, the sun does not move around the earth, and time does not tick by second for second equally for all observers, everywhere.
And yet, Damned Thing that it is, Chalmers’ core intuition just won’t go away: there exists something which demands explanation. If Chalmers’ intuition seems weak, a denial (like Dennett’s) that there is anything to be explained seems even weaker. As Descartes first observed, the fact that we are conscious cannot be denied.
So the Damned Thing lurks, leering and sniggering, out of our explanatory reach, thanks to Chalmers’ Zombie, another Damned Thing. The Zombie doesn’t lurk in the explanatory shadows, though: it lumbers directly for the throat of our physicalist assumptions and refuses to die, no matter how many argumentative rounds we pump into it.
All in all, it seems that, as Philip Goff puts it, there is indeed good reason to be “more than a little afraid of zombies”.
I was particularly proud of the review comments I got on this one: First time I’ve seem a my little pony reference in a philosophy essay, but the example is good none the less ... well-argued and entertaining to read which is always a good thing
The topic is David Chalmers "zombie" argument, and the challenge it poses to physicalism (the idea that the mind is purely physical, i.e. nothing more than the nervous system).
I've left out all the citations, to make it easier to read.
Damn you, David Chalmers!
I’m going to have to start this essay with a bit of a confession: I have no idea.
This is generally considered pretty bad form in philosophy essays, where one is expected to take a stance on the issue at hand, and make as solid an argument for that stance as one can. But when it comes to Chalmers’ Zombie Argument, I can only throw my hands in the air and admit that, I just don’t know.
What’s especially galling is that when I’ve previously covered this topic in an introductory metaphysics course, I’ve been a pretty hardcore physicalist. It seems so obvious, after all, given the correlation of brain states with conscious states, such problems for dualism as causal closure, and the powerful principle of Ockham’s Razor, that “we have no choice but to accept materialism as the only suitable theory of mind.
But philosophy and science are often bedevilled with what Robert Anton Wilson dubbed “Damned Things”: those niggling problems that defy easy resolution. Just when the Damned Thing we called consciousness seemed fairly comprehensively cornered in the physicalist camp, along comes Chalmers with his Zombie Argument that lurches along like one of those Hollywood creatures that refuses to lay down and die.
But of course, Chalmers’ Zombie is nothing like a Hollywood zombie, nor even a traditional Haitian voodoo zombie. Rather, Chalmers’ zombie is an influential thought-experiment first proposed in his 1996 book, “The Conscious Mind”, as a challenge to the reductionist theory of physicalism.
Physicalism is the idea that all mental states, up to and including consciousness are simply physical states and functions of the central nervous system; specifically the brain. Consciousness is said to necessarily supervene on the physical: given the requisite physical states of a person, consciousness is logically entailed. In this way, physicalists argue (see, for instance, Jeffrey Smart) that mental states (including consciousness) are identical to brain states.
Chalmers’ Zombie Argument is a direct challenge to this idea. Chalmers proposes the possible existence of a “Philosophical Zombie”: a perfect physical replica of a person, atom for atom, but which lacks consciousness. The Philosophical Zombie acts just like an human ordinary human being: it walks, talks, senses and interacts with the world around it, but its experiences lack qualia, the “what-it-is-like” inner sensations that, for us conscious beings, accompany experiences: the thrill of seeing the sun break through the clouds, the exquisite taste of chocolate, the heartbreaking sweetness of a minor key. For the Philosophical Zombie, to use Chalmers’ own phrase, “it’s all blank on the inside”.
The challenge to physicalism from Zombies lies in the fact that the Zombie shows that it is possible to conceive of brain states existing without consciousness. Physical states do not logically entail, a priori, conscious states, in the same way that the state of being a horse and female logically entails being a mare. If physical states do not similarly logically entail consciousness, then consciousness is not identical to brain states. Not only this but physicalism must be false because physicalism has to deny the conceivability of beings that are physically identical to us but not conscious.
At first glance, this seems a bizarre argument. Zombies (of the philosophical kind) just don’t seem to exist, and even Chalmers concedes that it seems to be a fact that consciousness is always correlated with brain states. But this is an a postierori observation of the “horses do not fly” kind. But as the popular cartoon character Rainbow Dash shows, a flying horse is at least conceivable, if not probable. But Rainbow Dash the male mare is not.
So it is conceivable that there could be identical brain states that are not correlated with consciousness; thus a Zombie could exist. And, as Philip Goff observes, “what’s possible has implications for what’s real”. The implication of the Philosophical Zombie for physicalism is that, if consciousness does not, a priori, accompany brain states, then physicalism is false.
Indeed, in a very limited way, it could be argued that Philosophical Zombies of an unsophisticated kind do exist. Chalmers uses the example of the Google Car, or Watson, the computer that beat the human champion at Jeopardy! Both of these machines perform at least the rudimentary functions that creatures with brains do, albeit in the most primitive way (in Chalmers’ words, ”maybe we’re up to earwig level, or earthworm level”), yet I doubt anyone would claim that they have any kind of consciousness. For instance, that the Google Car has any qualitative “feel” about its environments as it negotiates it, or that Watson had any feeling of triumph at winning.
James May demonstrates another kind of (primitive) Philosophical Zombie, when he compares a tune composed by a computer, programmed to emulate Beethoven, with an actual Beethoven composition. One, he says, is plainly the work of Beethoven, the other “merely the vomit of a digital contraption”. Like Searle’s Chinese Box, the computer has analysed the data and produced a logically coherent result, but there’s no “something extra”. Beethoven wrote the “Moonlight Sonata” in the throes of unrequited love and, as May notes, “you can tell. There’s something going on there that defies purely logical analysis”.
That “something” would be qualia: “what-it-is-like” to be a musical genius hopelessly in love. The Beethoven-emulating computer is, in a very primitive way, a Philosophical Zombie.
Needless to say, the Philosophical Zombie argument has attracted trenchant criticism from defenders of physicalism.
Patricia Churchland dismisses the Zombie argument as “a demonstration of the feebleness of thought experiments”. We can imagine anything is possible, but that doesn’t mean that it is possible, or particularly enlightening. But Churchland’s objection here seems to be missing the point, as her example of Crick’s “Thermodynamic” thought experiment illustrates. The Thermodynamic thought experiment imagines a world were heat is not identical to mean molecular kinetic energy. It is true that that is “feebleness itself”, but that is because mean molecular kinetic energy is necessarily heat (Cf Kripke). If heat is strictly identical with mean molecular kinetic energy, they must be identical in all conceivable worlds.
On the other hand, as argued by Kripke, it does not seem to be the case that consciousness is strictly identical with brain states.
Daniel C. Dennett, on the other hand, seems to object to the Zombie argument on the grounds that “the purported ‘subjective qualities’ or ‘qualia’” simply don’t exist. All there are, are the performance of functions of the brain, or the dispositions to do so. (This seems a strangely behaviourist answer although perhaps not so strange, considering that, as Chalmers points out, Dennett was a student of Gilbert Ryle. Philip Goff also notices the behaviourist “flavour” of arguments like Dennett’s). Explain the physical functions and dispositions, Dennett argues, and you’ve ipso facto explained consciousness (Facing Backwards). There is nothing more to be explained.
Chalmers responds that the Zombie argument shows that there is something further than the physical to be explained: what it is like. One might contrast Dennett’s argument with the process of digestion. To explain all the physical facts of digestion: food chemistry, the human anatomy of taste receptors, the processing of information in the brain, and so on, still does not explain the ineffable experience of eating a really good tiramisu.
Dennett and Churchland both also liken the Zombie argument to the Vitalist argument, which held that the physical processes in living things were insufficient to explain how life itself could arise out of dead matter: there had to be something missing, the so-called elan vital. Ultimately, however, biology was able to show that there was no need of such a “life spirit”.
Chalmers responds that, yes, given the physical facts of reproduction, metabolism, response, adaptation, etc., life necessarily entails. All those functions are part of the a priori definition of “a living thing”. The Zombie argument shows, though, that “any physical description of a brain seems at least to be logically constistent with the absence of consciousness. So somehow you need to put an extra ingredient in there”. Brain states and functions do not necessarily entail consciousness. One can conceive of all the physical processes that normally correlate with consciousness, but without consciousness. One cannot do the same with life.
Churchland also objects that Chalmers is committing a fundamental fallacy of informal logic - an argument from ignorance. As she sees it, Chalmers argues that from the premise that because we scientifically understand little about the nature of consciousness, we can never explain consciousness, no scientific discovery would deepen our understanding of it, nor can it be explained it terms of physical properties. But I’m not sure that Churchland’s accusation holds. Indeed, it seems that she is committing, at least in part, the informal fallacy of the straw-man.
The first conclusion of the alleged argument from ignorance, that we can never explain consciousness, is clearly false. Chalmers obviously thinks that an explanation exists, but that (if it is ever found) it will be a novel one involving new physical laws. Of course, an explanation may never be found, but that we may never find an explanation of something is not in itself controversial, as even Dennett admits.
The second alleged conclusion, that no scientific discovery could ever deepen our understanding of consciousness, besides seeming to be a partial restating of the first, is also clearly false. Chalmers frequently alludes to discoveries in neuroscience which are daily, it seems, advancing our understanding of the mind, and explicitly states that “neuroscience is not irrelevant to the study of consciousness”, even if it is not enough.
The third conclusion, that consciousness can never be explained in terms of physical properties, is not an unfair characterisation of Chalmers’ stance. But Churchland gives only one premise - that of ignorance; on that structure, the argument is indeed invalid. It seems to me, though, that Churchland has excluded a second premise of Chalmers’ argument; one which is the very point of the Zombie argument: while we may be largely ignorant of the nature of consciousness, it seems that physical brain states do not logically entail consciousness.
Thus it seems that the accusation of argumentum ad ignorantam fails. But both Churchland and Dennett make a much stronger claim against the Zombie argument, one that seems to me to be the most damaging: the feebleness of Chalmers’ explanatory hypothesis.
Chalmers proposes that consciousness will be explained by introducing new fundamental “psychophysical” laws to link conscious states to brain processes. Chalmers agrees that, unlike the Zombie, it does seem to be “a law of nature … that when you get a brain of a certain character you get consciousness along with it”, but that, “you actually need these laws of nature, special laws of nature, to connect the brain to consciousness”. Chalmers suggests that conscious experience be included in the canon of irreducible, fundamental features of the universe, alongside mass, charge, spin, space-time, etc.
Dennett objects that Chalmers, unlike physicists, cites no objective evidence that justifies his hypothesis. The only data Chalmers offers is a subjective “belief in a fundamental phenomenon of ‘experience’” (Facing Backwards). Dennett complains that this is too drastic a move to even contemplate without independent evidence. Like Smart, Dennett objects to such seeming “nomological danglers”. Greg P. Hodes likewise cautions that “to accept the first hypothesis that comes to mind without any attempt at verification is to be intellectually irresponsible”.
But Chalmers is in some august company. Consider, for example, String Theory. The theory is logically consistent, but it does not make any predictions that can conceivably be tested. True, String Theory rests on a foundation of profound and elegant mathematics, while Chalmers is by his own admission left to fall back largely on subjective experience and (allegedly feeble) thought experiments. The lack of testability for String theory similarly leads some in the physics community, like Sheldon Glashow, to dismiss the theory out of hand (”not even wrong”, in the words of Wolfgang Pauli). But its logical coherence and mathematical elegance nonetheless means that it, like the Zombie, refuses to die.
At the same time, Chalmers has some heavy-duty supporters in the science community. Physicist John A. Wheeler has suggested that information is a fundamental physical property. Chalmers argues that information has correlated physical and consciously experiential aspects that might lawfully (in the sense of his psychophysical laws) connect the brain to consciousness. Francis Crick and Christof Koch as also accept, however tentatively, that information may indeed be the key for a neurological basis of consciousness. Indeed, John A. Wheeler and Roger Penrose have suggested even a central role for conscious observers in the universe.
I confess, Chalmers’ hypothesis, however tentative, makes me slightly uncomfortable. It seems a little to like the “quantum flapdoodle”, to use Murray Gell-Mann’s memorable phrase, of many New Age cranks. Chalmers’ hypothesis, though, unlike the musings of the white-robe-wearers and crystal-wavers, has a logic that is difficult to deny, maddening though it may be.
Chalmers acknowledges that, at heart, all he has is “a collection of intuitions”, although he defends these intuitions “natural and plain”. But this seems a pretty weak claim: many of our intuitions are natural and plain - and wrong. The earth is not flat, the sun does not move around the earth, and time does not tick by second for second equally for all observers, everywhere.
And yet, Damned Thing that it is, Chalmers’ core intuition just won’t go away: there exists something which demands explanation. If Chalmers’ intuition seems weak, a denial (like Dennett’s) that there is anything to be explained seems even weaker. As Descartes first observed, the fact that we are conscious cannot be denied.
So the Damned Thing lurks, leering and sniggering, out of our explanatory reach, thanks to Chalmers’ Zombie, another Damned Thing. The Zombie doesn’t lurk in the explanatory shadows, though: it lumbers directly for the throat of our physicalist assumptions and refuses to die, no matter how many argumentative rounds we pump into it.
All in all, it seems that, as Philip Goff puts it, there is indeed good reason to be “more than a little afraid of zombies”.