Property Dualists Should Be Panpsychists

Presenter: Philip Goff, University of Hertfordshire

Commentator 1: William Robinson, Iowa State University

Commentator 2: Jonathan Simon, New York University



  1. I appreciate Dr. Goff for skillfully presenting us with a new and creative variant of the paradoxes of Zeno(1). For his argument depends upon reasoning that, since the property dualist cannot be sure on a microscopic scale of where consciousness ends, this lack of precise boundary in the small means there can be no macroscopic boundary for consciousness asserted by the property dualist at all.

    But whether or not we hold to property dualism, the same difficulty may occur for any non-divine intervention which would stop consciousness– say, killing Lot’s wife, or just rendering her unconscious. Let’s explore a more mundane example in more detail.

    Let us say instead that Lot’s wife is undergoing plastic surgery under general anesthesia. We know that when the concentration of the anesthetic propofol in her brain, given through an intravenous drip and transported via the blood, goes above a certain concentration, say 10 micrograms per gram, she will become unconscious, due to a high percentage of her brain cells decreasing their electrochemical activity. Empirically this would be imprecise, but let us say she becomes unconscious from the anesthetic at the point that the overwhelming majority of her brain cells decrease their firing rates. Such a scenario would correlate with Lot’s wife losing (temporarily we hope) her body’s (nonphysical in the case of property dualism) property of consciousness. Let us say that each of her brain cells drop their activity beneath that threshold one by one, over fractional nanoseconds of time, during a span of 10 minutes.

    But we cannot say exactly how many cells need to drop their activity before she is unconscious! Therefore, by Philip Goff’s argument, with or without the perspective of a property dualist, Lot’s unfortunate wife cannot ever be said to become truly anesthetized. (Let’s hope she does not have too unpleasant an experience.)

    Importantly, we should see now that this epistemic boundary deficit in knowing when Lot’s wife is properly anesthetized is not due to a vagueness in our concept of the property of consciousness (which Goff’s argument forbids the property dualist), but rather to a lack of precision in our ability to locate a precise boundary in our measurement of changes at the scale of the tiny intervals which Dr. Goff’s argument requires.

    The resolution to the paradox is to recognize that since, in the example Dr. Goff gives(2), God turns Lot’s wife entirely to salt atom by atom, she is certainly not conscious when God is done(3). The exact spatiotemporal location of the boundary for her loss of consciousness is relatively unimportant because, even if we cannot measure or count exactly when the boundary was reached, that boundary certainly is passed by God during the salting process(4). This is true whether the consciousness property is transparent or opaque, and regardless of how sharply we define that property. Just as Achilles would certainly pass that tortoise. And just as pillars of salt are not conscious, but we are; yet the exact distribution of any potentially dualistic property of consciousness among organisms is, I agree, far from clear.


    1. Black, Max, “Achilles and the Tortoise.” In Salmon, Wesley, ed, Zeno’s Paradoxes, 1970. Black argues against Zeno’s use of infinitely small divisions as imposing an unnecessary constraint on Achilles, but his argument applies to any inappropriately small unit of measure. The Sorites paradoxes would not apply since Goff asserts that consciousness cannot be a vague concept to the dualist (the commentators have looked at that point well enough I think).

    2. Goff, Philip, property-dualists-should-be-panpsychists.pdf, Section III.

    3. The story of Lot is a Biblical reference, but we ignore afterlife issues as irrelevant here.

    4. This may have relevance to Chalmer’s fading qualia argument for the potential of machine consciousness. If we assume in advance that the end result of electronic replacement of the brain is not conscious, the point consciousness was lost might be an exact point in time on each occasion, but would vary in a microscopically stochastic and thus unpredictable way between trials.


    I would like to comment on the introductory paragraph to Philip Goff’s paper which reads:

    “By ‘consciousness’ I mean the property of being a thing such that there’s something that it’s like to be that thing. The meaning of this rather cumbersome phrase can be illustrated with reference to our common sense beliefs about what things have the property it denotes. According to common sense, there’s something that it’s like for a rabbit to be cold, or to be kicked, or to have a knife stuck in it. In contrast, there’s nothing that it’s like for a table to be cold, or to be kicked, or to have a knife stuck in it. There’s nothing that it’s like from the inside, as it were, to be a table (according to common sense).Consciousness, as I will understand it, is the property of having an inner life of some kind or other; a property ordinary opinion supposes to be confined to the biological realm.“

    This line of thinking, stemming in large measure from Nagel, is very common in the contemporary philosophy of consciousness and I’m commenting on Goff’s statement partly as representative of this wider tendency. (Hacker has recently provided a much more detailed criticism of the Nagel proposition than I can here.)

    Clearly, a lot of weight is being thrown on the notion of ‘(being) like’. So what could be the meaning of ‘like’ in the context?

    There seem to be only two possibilities. It’s either a ‘like’ of ‘feeling like’ (‘I feel like going to the movies’) or of comparison (‘This one is like that one’).

    Let’s look at the first possibility in the case of the rabbit. The meaning would then be: ‘There is something it feels like for a rabbit to be (feel) cold’ – or, straightening out the convoluted syntax, ‘A rabbit feels like (doing) something when cold’. The proposition is obviously quite vacuous. Presumably, it’s true in some general, uninformative way (ignoring the anthropomorphism of the word ‘feel’) but it tells us nothing at all about how the rabbit feels when cold. (Does it feel like seeing a movie? Does it feel nostalgic? Homicidal? Like eating an orange?) The contrasting case, ‘There’s nothing that it’s like for a table to be cold’ is equally vacuous. This simply tells us that when a table feels cold there is nothing it feels like (doing). It amounts to saying that there’s nothing we can say to describe the table’s state of feeling or consciousness when cold (unless it’s a roundabout way of saying that a cold table feels lethargic and apathetic!)

    What about the second possibility? What is it ‘like’ (comparison) for a rabbit to be (feel) cold? Presumably, it is like a rabbit being (feeling) cold. But we can’t compare something to itself, so is it like a cow or a person being cold, for example? We have no way of knowing since we can’t know how a rabbit (or a cow) feels. So this possibility also leads nowhere.

    Why is the Nagel thing persisted with so often? My guess is that it’s precisely because the word ‘like’ is being used ambiguously as I’ve indicated. People, I suspect, have allowed themselves to be bamboozled by that (and by the convoluted syntax) and have assumed that something profound is going on. But as we can see, it only takes a little analysis to show that nothing remotely profound is going on and that, under either interpretation of ‘like’ (and is there any other?), the proposition is at best an empty truism.

    There’s one other element in the quote above. Near the end, the emphasis shifts from the notion of being like to being like ‘from the inside’ or ‘having an inner life’.

    Note, first, that there’s no necessary connection between ‘being like’ and being like ‘from the inside’. That is, the one doesn’t automatically generate the other. More importantly, note that we have suddenly moved from a prosaic – if ambiguous – statement (‘something it is like’) to the realm of metaphor. What justifies this move? Certainly not the ‘like’ argument, which doesn’t lead anywhere as we’ve seen. Moreover, the metaphor of ‘inner life’ for feelings and consciousness has been around for a long time. It’s part of our everyday language in statements like ‘He has a heart of stone’, or more colloquially, ‘He’s a nincompoop. There’s nothing going on inside’. But handy though it is, the metaphor doesn’t tell us anything philosophically enlightening about the concepts of feeling or consciousness. In effect it’s just another name for the idea of consciousness.

    As I say, these comments are not directed at Philip Goff specifically. They are an attempt to point out obvious weaknesses in a line of argument that is used over and over again in the contemporary philosophy of consciousness – almost like an incantation – without any real attempt to subject it to careful analysis.

  3. I am very, very grateful for these challenging responses and comments. Below is a response to William Robinson and Jonathan Simon (In the response I address the former as ‘Professor’ to return to courtesy…actually I’m just a lowly post doc…). Responses to the comments drbillh and Derek Allan to follow.

    First, just one small point that occurred to me reading the responses and may be of interest to both:

    Even if the property dualist accepts the kind of panpsychist picture I try to press on her, I think she can still accept that a given person can lose consciousness. Suppose we hold some kind of Lockean theory of personal identity, and suppose that there is a period of time, say during the dreamless sleep of David’s body, when there are no conscious experiences bearing the relevant psychological relations to each other (perhaps not involving any cognition at all). In that case we can quite happily say that David is not conscious during this time. There are of course conscious states instantiated by parts of his body; by his brain, but also by his liver and toe nails, but they are not his consciousness. This may help account for what Jonathan takes to be evidence that consciousness disappears (Note also that there is no direct route from empirical data to fact. We must build a theory to account for the data, and in doing this we are obliged to make one that is theoretically virtuous. I take it that this obligation rule out any hypothesis involving the implausible consequence).

    Response to William Robinson

    Professor Robinson accepts much of force of my argument, and hence feels that the property dualist must find some way of avoiding my implausible consequence. I like the way he puts the challenge as a demand that nature be continuous.

    Therefore, Professor Robinson sets himself the challenge of making sense of a continuous transition from consciousness to non-conscious. He thinks that the way to do this is to hold that episodes of consciousness, rather than creature consciousness, are basic, and that such episodes are episodes of consciousness tout court, rather than essentially involving attention. If we start with creature consciousness, we find ourselves stuck with something that seems like a binary property: it’s either on or it’s off. However, episodes of consciousness tout court are more plausibly taken to be the kind of things that can fade: imagine for example a single sine tone slowly fading to nothing.

    Where Professor Robinson talks of ‘creature consciousness’ and ‘episodes of consciousness’, I prefer to talk of the determinable property of consciousness – the property of being a thing such that there’s something that it is like to be that thing – and its determinates – various ways in which a thing can be conscious. Are we talking about different things here? The existence of conscious experience surely requires a subject to have, that is, to experience, the experience. And I take it that an ‘episode’ is an event, which I take to be the having a property by an object at a time. So it seems to me that strictly speaking Professor Robinson is wrong that episodes of consciousness are basic: they can be broken down into subjects, times and the determinate ways of being conscious.

    But the determinate properties involved in episodes of consciousness are surely determinates of a certain determinable: that’s what I call ‘consciousness’ (what Professor Robinson seems to be calling ‘creature consciousness’). What I take to be theoretically unpalatable is the sudden loss of the determinable of these determinate ways of being conscious: the sudden loss of this way an object can be modified. Wouldn’t it be strange – as radical discontinuity in nature as Professor Robinson puts it – if an object suddenly lost not just a certain way of being shaped, but the entire determinable property of being shaped in some way of other. It would be strange in the same way if an object suddenly lost, not just a certain way of being conscious, but the entire determinable property of being conscious in some way or other.

    Perhaps we can still capture Professor Robinson’s central point like this: unlike shape, the determinable property of being conscious can ‘fade out’. We are not surprised that an object can completely lose the property of having velocity, as, unlike shape, velocity comes in degrees and so can disappear in degrees. Its disappearance is, therefore, continuous rather than sudden. If it can be shown that consciousness comes in degrees, Professor Robinson can show that the disappearance of consciousness is continuous rather than sudden. It has to be shown that consciousness is more like velocity than shape.

    Unfortunately, I think that Professor Robinson has misdescribed the cases he takes to show that consciousness comes in degrees. When my auditory experience of a single tone ‘fades out’, I am not becoming less conscious: it is not becoming the case that I have less of a degree of being a thing such that there is something that it is like to be it (I don’t think we can make sense of the notion of this property coming in degrees). Rather, our determinate state of consciousness is changing, from one which represents a loud noise to one which represents a quiet noise. Similar to Descartes, I think we have two determinables here, neither of which come in degrees: extension and consciousness. The hypothesis that an object can suddenly lose either is strange. The hypothesis that a macroscopic object suddenly loses either one as the result of a tiny alteration to a fundamental particle is unacceptable.

    Another point: Professor Robinson talks of changes in neurons, rather than changes in fundamental particles, bringing about the disappearance of consciousness. But the example I give involves a woman being turned into a pillar of salt via a series of slight adjustments to fundamental particles. I suspect that Professor Robinson’s picture would look much less plausible if he set it up in terms of sub-atomic particles rather than neurons, but I’m not clear how he can avoid the obligation to do so.

    One final point. Professor Robinson suggests that the conceptual dualist physicalist might make use of his solution. As I say in the paper, I disagree that the conceptual dualist physicalist is subject to the challenge I raise for the property dualist, as I think they can hold that the concept of consciousness is vague (given that they are semantic externalists, they need not hold that the spectrum of sharpenings are a priori associated with the concept).

    Response to Jonathan Simon

    Jonathan spends much of his response objecting to my move from property dualism to transparency. He claims that the arguments of Chalmers, Jackson and Kripke do not involve a dependency on transparency. However, he doesn’t examine the reliance of Chalmers’ arguments upon the 2D framework (he only gives the very basic version of the argument Chalmers starts with in ‘The two-dimensional argument against materialism,’ and doesn’t explore the ‘refined conceivability argument’ he ultimately defends). Once the reliance on the 2D framework is taken into account, it is clear that Chalmers’ argument relies on a commitment to phenomenal transparency. It is an assumption of the framework that the primary intension of a concept is a priori evaluable, and it is a premise of Chalmers’ anti-physicalist argument that the primary intension of consciousness is identical to its secondary intension (this is pretty much Kripke’s claim in his anti-physicalist argument). I believe this assumption and this premise entail phenomenal transparency. I explain this in footnote 4; perhaps I should have included this in the main text to avoid confusion. Chalmers has explicitly accepted this online somewhere in his discussion of Nida-Rumelin, who is an advocate of phenomenal transparency, but I can’t seem to find it now.

    Furthermore, if one wants to attack my move from property dualism to phenomenal transparency, I think one ought to respond to my argument that the standard anti-physicalist arguments are incredibly weak without it, which Jonathan doesn’t seem to do. I try to show in the paper that there is a clear and damning physicalist response to these arguments if they are not backed up by phenomenal transparency. I take it that this is why the Mary argument and the Kripke argument are very weak if they are do not involve the 2D framework, or a straightforward commitment to phenomenal transparency.

    I am intrigued by Jonathan’s suggestion that the link between conceivability and possibility in anti-physicalist arguments might be justified a posteriori. I can’t begin to think what kind of empirical considerations could justify this kind of metaphysical principle. I’d like to hear more, but in the absence of further details I can only express my emotional reaction.

    At an number of points Jonathan seems sympathetic to the idea that the concept of consciousness is translucent rather than transparent – to use my terminology – for example on p. 4 he says, ‘Perhaps consciousness has a noumenal essence as well as a phenomenal presentation, and our concepts of consciousness disclose its phenomenal presentation, leaving its true noumenal essence opaque.’ But he doesn’t consider my concerns with this proposal, in the ‘Aside’ on p. 7-8. I think he ought to have a response to these concerns if he wants to hold this view.

    So my move from property dualism to phenomenal transparency goes like this: (i) without a denial of the opacity of the concept of consciousness, the standard anti-physicalist arguments are rubbish (and in any case those anti-physicalist arguments which are reliant on the 2D framework implicitly rely on phenomenal transparency), (ii) it is hard to make sense of the claim that the concept of consciousness is translucent, (iii) therefore, if we have to deny that the concept is opaque, and we can’t make sense of its being translucent, then we are forced to conclude that it is transparent.

    Jonathan has concerns about how a vague concept could be transparent. It’s a good point that if the concept is vague then there won’t be a single sparse property which it picks out. Although given this, we should either take the word ‘property’ in the definition of ‘transparency’ in some metaphysically loose way, or we should turn to the other definition of ‘transparency’: what is ascribed by the concept can be known a priori (Jonathan notes these two definitions, but then only discusses the former). Once we are working with the latter definition, I really don’t get the problem. If the concept is vague, then what is ascribed is indeterminate. If the concept is transparent, then one can know a priori the indeterminate content that is ascribed in an application of the concept. I think I need to hear more details of this worry with reference to specific examples. Why can’t the concept of tallness be vague but transparent?

    Jonathan sometimes talks as though, if the concept of consciousness is transparent, then we must be able to work out every single fact about consciousness. This is not the case: I cannot, for example, know how many instances of consciousness there are in the world. What I know is merely what it is for something to be consciousness, what I ascribe to something when I ascribe consciousness to it. For this reason, phenomenal transparency does not immediately entail the falsity of physicalism. Galen Strawson accepts phenomenal transparency but denies physical transparency – we know what it is for something to feel pain but not what it is for c-fibres to fire – in order to make sense of physicalism, and this seems fine to me.

    Despite his concerns about my move from property dualism to phenomenal transparency to the non-vagueness of consciousness, Jonathan likes the Michael Anthony argument to the conclusion that consciousness is not vague. The worry I have with Anthony’s argument is that it seems to assume the falsity of semantic externalism. If the semantic workings of one of my concepts is determined wholly outside the head, then I don’t think I can have any a priori insight into whether or not that concept is vague. This is why I don’t think there is a good argument to press the a posteriori physicalist to believe in the non-vagueness of consciousness.

    I think Jonathan has slightly misunderstood my concerns about what I call the ‘implausible consequence’. The worry is not just that we have to believe in brute facts: I take it that everyone has to do that. But some brute facts are theoretically more palatable than others. Newton’s law of gravity, even if we suppose it to be brute, was theoretically virtuous: simple and with universal application. The laws that the implausible consequence would commit us to would be unacceptably theoretically vicious: of arbitrarily limited application, and involving arbitrarily precise values.

  4. Thanks, Philip, for the interesting talk. I’m really no fan of property-dualism, but I don’t see why property dualists have to deny that being conscious is a property that does not come in degrees. You say on p. 15 that

    “Careful a priori reflection on consciousness yields a strong modal intuition that […] Consciousness seems to be an essentially all or nothing, on/off property.”

    Is this really the case? How will you convince me if I say that I have the opposite ‘modal intuition’? And if I tell you that the Lot argument just gave me pretty damn strong intuiitive reason to deny that consciousness is an on/off property?


    I think you may be underestimating the granularity of cognitive (or brain) function: Yes consciousness (or, as I prefer: whether or not you feel anything at all) is all-or-none, not a matter of degree. But to apply Sorites to the transition from a feeling system to an unfeeling system you have to do it by the equivalent of grains of sand — infinitesimal gradualism. But surely the normal-brain and salt poles of your continuum are too far away. The action is more likely to be in the configuration of something much more macroscopic — something somewhere between a kidney and a cerebellum. And the jumps that occur with a change in functional organization from one to the other could be quite abrupt (as in the chemical periodic table).

    So much for panpsychism being forced upon us by the gradual transition from a feeling to a nonfeeling system (or vice versa). But I do have a query about panpsychism. (If you are not a panpsychic maybe another reader is.) What is it that one is supposing, when one is supposing panpsychism? That every “entity” feels — meaning every part of anything, anything of which something is a part, as well as every possible combination of things and parts. Does that mean that if I give a chair a kick, the whole universe and every permutation of its parts feels a big bang? Or is it that whole mereological mess is just feeling a generic “ho-hum” feeling until except if it is in the right configuration (say, a biological orgnanism with an intact brain)?

    If panpsychism is coherent at all, I wonder what is the cardinality of its promiscuity?

  6. Drbillh: I think you have slightly misunderstood the argument. It is not an epistemic point that we don’t know where the boundary between consciousness and non-consciousness lies. The issue is the theoretical implausibility of the hypothesis that the laws of nature are such that a slight change in one particle of a macroscopic object can make the difference between the whole object having and lacking consciousness. Newton’s hypothesis concerning the law of gravity was extremely theoretically virtuous: simple and with very general application. The hypothesis concerning laws that non-pansychist property dualist would commit us to would be extremely theoretically vicious: of arbitrarily limited application and involved arbitrarily precise values.

    Hence I don’t think my argument is a version of Zeno’s paradox.

    Derek Allan

    In answer to you challenge, I deny that the phrase ‘what it’s like’ should be analysed in either of the ways you suggest; you’re quite right that neither of them work. Rather the phrase is used to direct listeners to attend to the qualities of their conscious experience. Perhaps you hold a strong thesis of transparency, according to which there are no intrinsic qualities of experience which can be attended to; that when we try to attend to the qualities of experience we find ourselves attending to the qualities our experience represents as present in the world? Do you have an argument for this claim?

    It is evident to me that I can attend to the redness of the surface, and I can also attend to how the red appears to me, i.e. to the quality intrinsic to the red perceptual experience. This is one of the benefits of being a human rather than a non human animal: not only can you perceive the world, you can also attend to the perceptual experiences in virtue of which you perceive the world. What do you think the difference between attending to the world via perceptual experience is, and attending to one’s perceptual experience of the world? Or do you think humans can’t attend to their experiences?

    I agree that ‘inner life’ talk is metaphorical. It is to be cashed out in terms of talk of the qualities of experience.

    Bence Nanay

    If I try to imagine a borderline case of consciousness, I just end up thinking of something with very soft, faint feelings. But of course, this is just one specific way of determinately having an inner life, rather than a matter of being in a state whereby it is indeterminate whether or not one has an inner life. I can imagine someone who is in a state in which it’s indeterminate whether or not she is tall, but I can’t imagine someone who is a in a state in which it’s indeterminate whether or not she has an inner life.

    I think someone wanting to defend fuzzy consciousness owes us a few words to help us find this situation intelligible. No problem for the analytic functionalist: if being conscious is a matter of satisfying some not very fine grained functional predicate, then something that instantiates a functional state that falls into the borderline cases of that predicate will thereby fall into the borderline cases of being conscious. But if your reject any kind of physicalist or functionalist reduction of consciousness, I find it difficult to see what one might say to help make fuzzy consciousness intelligible.

    Maybe my thought experiment gives us reason to want to make sense of a scenario where it’s indeterminate whether or not there is consciousness, but I can’t see how it helps us actually make sense of this.

    Steven Harnad

    I see the temptation to want to set things up at a more macroscopic level, but I think my opponent is obliged to say what happens in my case given how I have set things up. Is it the case that any one of those slight changes at the fundamental level makes the difference between the whole macroscopic object having or lacking consciousness? If not, then it seems the whole object must still be conscious when it’s a pillar of salt. I don’t see how one can avoid this question by shifting to the macroscopic level.

    The kind of panpsychism I am pushing on the property dualist is as follows: from the macroscopic level up all things and all combinations of things have an inner life. Presumably, though, only organisms have the kind of inner life that grounds perception and cognition. Most panpsychists, though, tend to hold that only fundamental particles and organisms have consciousness; the former having only some very primitive (non perceptual, non-cognitive) kind of consciousness.

  7. Hi all — three things.

    (i) Jon is right that I’m not committed to phenomenal transparency. The 2D argument dosn’t need the premise that the primary intension and the secondary intension of ‘conscious’ are the same (see “The 2D argument against materialism” for why). Still, I’m inclined to accept that the intensions are the same, and I’m probably inclined to accept something in the vicinity of phenomenal transparency, though maybe not exactly as Philip has stated it. So I wouldn’t put too much weight on these points in responding to Philip’s arguments.

    (ii) I’m inclined to think that phenomenal transparency is compatible with the vagueness of ‘conscious’. I do have a sort-of intuition that ‘conscious’ isn’t vague, but that intuition isn’t so strong that it plus transparency entails that consciousness isn’t vague. I think it’s quite compatible with my grasp of the concept of consciousness that it could turn out to have a priori accessible sharpenings. I can’t right now imagine such sharpenings, but I take lack of current imaginability to be very weak evidence for impossibility here, as e.g. I may lack the concepts required to imagine the sharpenings. As far as I can tell, it could easily turn out e.g. that there are protophenomenal properties (which I currently lack concepts of) and that the transition from protophenomenal to phenomenal is vague.

    (iii) I’m fairly sympathetic with panpsychism. But I certainly think that non-panpsychist forms of property dualism are coherent and viable. And even if consciousness isn’t vague, I see a challenge to non-panpsychist property dualism but not a fatal problem.

    In particular, I don’t see why there couldn’t be arbitrary thresholds. E.g. maybe consciousness in a complex system depends on some quantitative property of that system such as Tononi’s phi. And maybe it’s a law of nature that there is some constant k such that systems are only conscious when phi > k. Or more likely some more complex and elegant law involving a constant that has something like this as a consequence. Of course this requires an arbitrary constant in a law of nature — but we know already that there are numerous such constants.

    Presumably then the property dualist can happily say that pillars of salt have phi < k, and that somewhere in the transition from brains to pillars of salt, as phi approaches k from above, consciousness gradually fades out, and at phi = k winks out altogether. Maybe something like this even happens when we fall asleep (Incidentally I think Philip distorts the situation by comparing to a blue balloon suddenly going pink. A better comparison would be some physical quantity such as mass going from positive to zero.) Of course this is a simplistic picture and reality is probably far more wonderful, but I don't really see a principled objection to a property dualist view somewhere in the vicinity of this.

  8. Hi Dave, sorry to thread jack here, but I had a quick question about (i) above. You say here that you don’t need the PI=SI premise for the 2D argument against physicalism to go through, but in the online version of the paper you cited you say,

    A materialist might reasonably question (3) [If P&~Q is 1-possible, materialism is false] by holding that even if there is a world w verifying P&~Q, w might be a world with quite different ingredients from our own. For example, it might be that W does not instantiate true microphysical properties (those instantiated in our world), such as mass and charge, but instead instantiates quite different properties: say, pseudo-mass and pseudo-charge, which stand to mass and charge roughly as XYZ stands to H2O. Likewise, it might be that w does not lack true phenomenal properties, but instead lacks quite different properties: say, pseudophenomenal properties. If so, then the possibility of w has no bearing on whether true microphysical properties necessitate true phenomenal properties. And it is the latter that is relevant for materialism.

    Still, it may be that the gap between 1-possibility and 2-possibility could be closed. In particular, when a statement S has the same primary intension and secondary intension, then a world will verify S iff it satisfies S, so S will be 1-possible iff it is 2-possible. If P and Q both have primary intensions that coincide with their secondary intensions, then so will P&~Q, and we could run the following argument:

    (1) P&~Q is conceivable
    (2) If P&~Q is conceivable, P&~Q is 1-possible

    (3) If P&~Q is 1-possible, P&~Q is 2-possible.

    (4) If P&~Q is 2-possible, materialism is false.

    (5) Materialism is false.

    Here, the truth of (3) requires that both P and Q have primary and secondary intensions that coincide. In the case of Q, this claim is quite plausible. As Kripke noted, there does not seem to be the same strong dissociation between appearance and reality in the case of consciousness as in the cases of water and heat: while it is not the case that anything that looks like water is water, or that anything that feels like heat is heat, it is plausibly the case that anything that feels like consciousness is consciousness. So it is not clear that the notion of “pseudoconsciousness”, something that satisfies the primary intension of ‘consciousness’ without being consciousness, is coherent. Likewise for other more specific phenomenal properties. So there is a strong case that the primary and secondary intensions of phenomenal terms coincide.

    This makes it sound like you do recognize that you need the PI=SI premise for the argument to go through. Then the argument hinges on the “strong case” that they do in fact coincide. Maybe you think that the case is so strong that it is virtually unnecessary to make the case; but there are those like me who, while we agree that the case is prima facie strong, think that there are stronger reasons for thinking that the do come apart. These reasons come from both a priori reflection (2d framework applied to ‘consciousness’ in exact same way as ‘water’) as well as empirical considerations (pain asymbolia, dental fear, etc). I know you don’t agree, but the point is that the argument does seem to rely on this move and there is some contention of the “strong case”.

    Just as a comment. The ‘psuedoconsciousness’ that you find unlikely is no more psuedo than xyz is psuedowater. It would of course feel just like our consciousness, just as xyz looks just like our water. The fact that they have a different microstructure is irrelevant. If that world had been real then the psuedowater would have been real water, and the psuedoconsciousness would have been real consciousness….but those worlds aren’t the actual world, they are mere possibilities. But neither possibility threatens the thesis that water, or consciousness, is physical. It only sounds strange when you say things like “something feels like consciousness but isn’t consciousness,” but there is really a disguised equivocation here. What it really says is that ‘something might feel like consciousness and not be this particular brain state’ and that is no different than saying ‘something might be watery and not be H2O.

  9. KIDNEYS AND COMBINATORICS (Reply to Philip Goff)

    If I can get from salt to an unfeeling kidney, and from an unfeeling kidney to salt, then all I have to worry about is getting from an unfeeling kidney to a feeling cerebellum and back, and that would not be molecule by molecule but by reorganization.

    Since (I agree) feeling is an all-or-none matter (hence the “primitive” vs “refined” distinction doesn’t much matter), panpsychism seems to me to imagine a distribution of feeling in the universe that is either NP-complete, indeterminate or (most likely) incoherent.

  10. Hi Richard,

    I think you missed this passage a few paragraphs down from the bit you quoted:

    “Second, it is worth noting that (contrary to a common supposition), the assumption that Q has the same primary and secondary intensions is not necessary for the argument for (5) to go through. To see this, we can consider the version of the argument where we adjoin a “that’s-all” clause to P. From (1) and (2), we can derive the conclusion that there is a minimal world verifying P in which the primary intension of Q is false. If P has the same primary and secondary intensions, then this world will be a minimal P-world in which the primary intension of Q is false. This world must differ from our world, because the primary intension of Q is true in our world. (There is a small loophole here arising from the possibility that this world differs merely in the location of the center of the relevant centered world. I discuss this loophole in section 5.) It follows that there is a minimal P-world that is not a duplicate of our world, so that physicalism is false of our world. It could be that strictly speaking physicalism will be true of consciousness, because P necessitates Q, but physicalism will be false of properties closely associated with consciousness, namely those associated with the primary intension of Q. We might think of this sort of view as one on which phenomenal properties are physical properties that have non-physical properties as modes of presentation.”

  11. == quote: ==
    It is not an epistemic point that we don’t know where the boundary between consciousness and non-consciousness lies. The issue is the theoretical implausibility of the hypothesis that the laws of nature are such that a slight change in one particle of a macroscopic object can make the difference between the whole object having and lacking consciousness.

    == end quote ==

    I agree that such is the main issue with the epistemic solutions to the sorites puzzles– the idea of an unknowable transition boundary is weird to us. It also goes against Leibniz’s principle of contunuity, though I’d think that quantum physics also violates that principle. I think you are assuming Leibniz’s continuity holds more or less implicitly in your argument, but I don’t think it has to. There may be a real but microscopically unpredictable phase transition somewhere.

    Please forgive me if I’ve misunderstood.

  12. Hi Philip

    Thank you for your reply. Since this is a conference, I assume I may respond, and that a conversation can ensue.

    Some of your reply speculates about views you suggest I might hold, and then gives your responses. I don’t want to pursue those matters because they will distract from the important issue I raised in my comment.

    My comment was directed specifically at how you understand the proposition you make in your opening statement: ‘By “consciousness” I mean the property of being a thing such that there’s something that it’s like to be that thing.’

    You have agreed with me that neither of the interpretations of ‘like’ that I examined will work. But you still say in your reply: ’Rather, the phrase is used to direct listeners to attend to the qualities of their conscious experience.’

    I certainly agree that the phrase is used that way – very often. But my question was: how do we interpret it – and in particular the word ‘like’ – to arrive at that result? If neither of the meanings of ‘like’ that I examined is satisfactory, what could the word, and the proposition, possibly mean?

    This is a serious question. As I said, part of my reason for asking it is that the proposition is used over and over again in the philosophy of consciousness (I could easily find other examples) as if it were, as you say, a good way to ‘direct listeners to attend to the qualities of their conscious experience.’ My comment argued that the proposition is close to nonsense – that, under any reasonable interpretation, it leads nowhere and can only confuse people.

    If I am wrong, I imagine it should be easy enough to produce an argument showing why. If I am right, shouldn’t the proposition be dropped altogether from the philosophy of consciousness as a pointless obfuscation?

  13. I did miss that! Thanks for pointing it out. Maybe we shouldn’t go on about this in here, but it does relate to Philip’s argument tangentially…

    You say, “It follows that there is a minimal P-world that is not a duplicate of our world, so that physicalism is false of our world,” (emphasis added) whereas I would there is a contradiction and hence we reject the assumption that led to it. This is exactly analogous to the reasoning that leads us to conclude that we can’t conceive of worlds physically just like ours without water. If Q really has the PI that it does actually and the world is really physically just like ours then it will pick out a brain state, so just as was the suggestion, Q must have a different PI and what is lacking in this world is some kind of psuedo-Q. On the other hand, it may be the case that the PI assigned to P currently needs to be augmented in some way so that when you conceive P & ~Q you conceive of a world where P has a PI that is different from the the PI assigned by an ideal completed physics. Naturally one wouldn’t notice since we don’t have a completed physics. This is exactly what we should say to some Aristotlean who argues for the falsity of water=H2O on conceivability grounds. Since the PI and SI of water diverge we cannot tell from examining the PI what the SI is. Is this is the same for Q we should expect the same results, unless the PI and SI for Q coincide.

    Part of the problem, I think, is that there are two ways that a world could verify ‘pain is not a brain state’. It could do so in the innocuous way where the PI of pain picks out something other than it does here, but in a way that does not make it 2-possible. The other way to do so is via the contentious way of stipulating that P&~Q is conceivable (P&~WATER is just as bad).

  14. Let me offer Phillip the following argument:

    1) If ‘consciousness’ is vague, then consciousness must be continuous with non-consciousness.
    2) If dualism is true, then consciousness is not continuous with non-consciousness.

    I think the extremely intuitive kernel of what Phillip is getting at, comes out here. 1) is an extremely plausible principle. As far as 2) is concerned, well, if dualism is true then at least we have a special reason to think that consciousness is not continuous with non-consciousness (viz. because its not simply a transition from some physical states to other physical states).

    Incidentally, you don’t need anything like Transparency, or Property Dualism, to defend 1). With ‘continuity’ construed in metaphysical terms (similarity of properties) rather than in semantic-epistemic terms (similarity of concepts, or a priori similarity of properties), this seems to be a fairly modest platitude about vagueness: if ‘P’ is vague, then ways of being P must somehow or other bleed into ways of being ~P.

    The problem is that, given this more modest metaphysical reading, the case for 2) really is not very strong at all. For one thing, we’ll have to venture precise accounts of continuity even to rule out that there could be continuity between non-physical consciousness and physical non-consciousness. But the larger point is that, if we rely only on the metaphysical reading, we have very little ground to rule against the possibility of some sort of protophenomenology. I am undecided as to how much credence we should allot to Dave’s pet possibility that we will one day discover concepts that make a priori the reducibility of phenomenology to protophenomenology. But in the context of the above argument we don’t even have to go here: there may be continuity without any sort of a priori reducibility.

    My own view is that we CAN strengthen 1) from a metaphysical to an epistemic-semantic reading, but only as a general principle about all vagueness. This will mean that the ensuing argument tells against materialists as well as dualists.

    My view is very closely related to Antony’s. Phillip, you say you do not like Antony’s approach because it runs afoul of semantic externalism. You then offer a very extreme characterization of externalism: “the semantic workings of my concepts are determined wholly outside my head”. But what about the more moderate “there are some external constraints on the semantic workings of my concepts”? This seems closer to what most externalists actually defend, and Antony’s argument (or anyway my reworking of it) needn’t conflict with this thought. The basic idea is that even externalists can and should allow for the (perhaps contingent) a priority of truths like ‘Julius invented the zipper’.

    A quick question to Phillip: you say your sort of panpsychism is compatible with people losing consciousness. Given a Lockean theory, you say, this would be a matter of there being no conscious experiences (of my parts) bearing the right relations to each other. I take it you have in mind that this sort of thing could happen gradually, hence no conflict with the spirit of your argument for panpsychism in the first place. But I am not sure that this works unless you take there to be some sort of reduction of the consciousness of the whole to the consciousness of the parts – and my impression was that this wasn’t your bag. Que pasa?

    One last remark re: Implausible Consequences. I’d like to just report something that Ted Sider suggested (on Phillip’s facebook wall, actually) – one option for the non-panpsychist dualist is to hold that the Laws linking consciousness and the physical are Indeterminate – as many take the laws of quantum mechanics to be. If the ultimate correlations turn out not to be a clean k constant, we might just take them to be indeterministic. If that is also true of the travelling habits of photons, really hard to see whats so specially problematic about it.

  15. ps: I should comment on the relation between the two-step argument I just sketched, and Professor Robinson’s discussion: I take him to be sketching ways for non-physical consciousness to be continuous with physical non-consciousness.

    I think his discussion highlights some of the deeper issues about what we might be asking for when we ask for continuity. In Platonic heaven, Plato takes an acute angle and widens it until it is an obtuse angle. The transformation is continuous, in the mathematical sense. Yet ‘acute’ angle is not vague. This shows that the continuity condition, though necessary, is not sufficient for vagueness. But it also casts doubt on Phillip’s case that if dualism is true, consciousness is not continuous with the non-conscious. Say we have some series of increasingly dimly conscious creatures, culminating with a first creature who is not conscious. [Whether or not experience is diaphonous, it seems plausible to say that a dim experience is in at least some intuitive sense “more like” the absence of experience, than a less dim experience is.]
    Well, obviously there is a metaphysically significant inflection point: there is either a last member of the series with non-physical properties, or a first member without them. (I am discouting the protophenomenal options here for the sake of argument). But that does not mean the series is not continuous. And there may be plenty of inflection points of one sort or another, so we cannot just amend 1) by adding that there be “no inflection points”, whatever that might mean.

    Thats to say, I think Professor Robinson has a point.

  16. pps: A friendly suggestion for Phillip:

    instead of talking about turning someone into salt (gradually or not), think in terms of the Problem of the Many. Here I am, a material being that is also an experiencer with phenomenal properties. Now what about the material being that differs from me only by not containing Eli the Electron as a part…

    the trouble with setting up the argument in terms of gradually turning someone into salt (or in general, setting up the argument in terms of any sort of diachronic transition), is that there are too many places for quantum dynamical thresholds (or neuronal dynamical thresholds, or thresholds at whatever level of granularity you have in mind) to figure in. If you do things synchronically, at least you constrain the range of possible thresholds people can appeal to. (Cf Peter Unger, “The Mental Problem of the Many”)

  17. Johnathan,

    Practically speaking, the issue with this being a diachonic transion is that if one requires a real amount of time to test the system for dynamic signs of consciousness, say 10 milliseconds or even 1 msec, we need more time than the age of the universe to do 1 atomic particle at a time and then check if she is still conscious after each replacement.

    Of course this is a thought experiment so the practical bounds of time don’t limit us. But it does underline our total ignorance of changes at that scale in macroscopic systems, including persons.

  18. David Chalmers

    I was hoping I could ignore Chalmers’ alternative route to the falsity of materialism which he alleges to go through without relying on the premise that the primary and secondary intensions of phenomenal concepts are identical. The trouble is that his move here depends on assuming the non-existence of what I call radically opaque concepts – very roughly a concept is radically opaque iff it reveals neither essential nor accidental properties of the referent. Me and Dave have been talking about this issue for the past 8 years or so, but unfortunately he still hasn’t come up with a a good argument for his conviction that there aren’t any (wink emoticon). There seems something in the blood of Australian philosophers (Jackson, Chalmers, Lewis – not Australian, but an Australian philosopher nonetheless) that makes them unable to tolerate opaque concepts.

    Here’s why the move depends on there being no such concepts (in the following big letters refer to concepts and little letters to properties). Suppose Q is a radically opaque concept which picks out p, then P & ~ Q will be conceivable; because Q doesn’t reveal that its referent is p, there is no way of ruling P & ~ Q out a priori. But because Q refers to p, there is no genuine possible world where p exists without q; nor is there a genuine world where p exists without the accidental properties a priori associated with Q, because there are no accidental properties a priori associated with Q as it’s radically opaque. (I’ve changed the letters to refer to properties rather than propositions, but you get the idea).

    Chalmers claims that his 2D framework allows him to avoid postulating two kinds of modality. But I too believe in a single kind of modality which is linked to rationality. So long as we are transparently conceiving of states of affairs, conceivability entails possibility. But because there are radically opaque concepts, there are some propositions which we can’t rule out a priori but which aren’t associated with genuine possible world (you don’t need a new modal primitive to make sense of this, you just need radically opaque concepts). So we should stop talking about modality, and start having an argument about semantics: are there any radically opaque concepts?

    My response to Chalmers second concern is on p. 14 of the paper:

    There is a sense in which our inability to find sharpenings of consciousness is not entirely conclusive evidence that there are no a priori knowable sharpenings of consciousness. Certain facts which are rendered a priori knowable by concept C may be out of the cognitive reach of a given individual possessing C, due to that individual’s cognitive limitations.1 Perhaps ones might suppose that if we were better reasoners we would be able to see how to sharpen consciousness. This seems to me an implausible leap of faith. We are not dealing with some difficult mathematics which is beyond our cognitive capacities, but which greater beings than ourselves could deal with. We are dealing with the basic semantic structure of a single concept. If our best efforts to find sharpenings of consciousness do not yield them, then we must suppose that there are no such things, at least not accessible a priori.

    On Chalmers’ third point: I can’t remember the details of Tononi’s proposal, but I take it phi is a higher level property? Things seem much more palatable when the thought experiment is altered such that we are considering God directly manipulating higher-level properties, e.g. turning down that phi till consciousness disappears. But I don’t think my opponent can avoid the obligation to consider the thought experiment as I set it up. If consciousness is not vague, then to avoid panpsychism she must hold that one of those slight alterations to a fundamental particle makes the difference between the whole object having and lacking consciousness. It is this implication which is so implausible.

    Steve Harnad,

    I’m not sure I understand your point. I don’t know what you mean by ‘primitive’ and ‘refined’ and ‘NP-complete’, and I’m not sure how you get to the conclusions you reach regarding panpsychism. The concept of a ‘kidney’ or a ‘cerebellum’ is presumably vague, and hence the difficulties don’t arise that do with a non-vague concept of ‘consciousness’.


    I don’t have a problem with non-panpsychist laws of nature being unknowable, or being indeterministic. The problem is that they would be extremely theoretically vicious. In science, there are often numerous hypotheses which are consistent with the data (various interpretations of quantum mechanics, for example), and we must choose between them on the basis of the internal theoretical virtues of the various proposals, e.g. unity, simplicity, elegance, etc. My claim is that non-panpsychist property dualist hypotheses concerning the laws of nature would be extremely theoretically un-virtuous.

    Derek Allen

    Your question is what the meaning of ‘what it’s like talk’ is, and my answer is that such talk concerns the qualities of experience apparent to introspection. Nagel’s phrase has proved itself to be very useful for directing the listener’s attention to these qualities, which is why it has caught on.

    So the real question is whether there are or are not qualities of experience apparent to introspection This is what Hacker ends up talking about at the end of ‘Is there anything that it’s like to be a bat?’, and this seems to me where the real issue lies, rather than refuting implausible analyses of ‘what it’s like’ talk in terms of superficially resembling bits of language, which is what goes on earlier in the paper.

    Jonathan Simon

    In footnote 13 I emphasise that my argument is not aimed at the protophenomenal guys, but rather property dualists who take phenomenal properties to be fundamental, although I do think it has force against the protophenomenal guys:

    This argument is aimed at property dualists, whom I have stipulated to hold that consciousness is a fundamental feature of reality arising in accordance with basic psycho-physical laws of nature. But not all anti-physicalists take consciousness to be a fundamental feature of reality. Many Russellian monists (Russell 1926, Feigl 1958/1967, Maxwell 1979, Lockwood 1989, Chalmers 1996, Griffin 1998, Stoljar 2001) take phenomenal properties be realised in proto-phenomenal properties, certain qualities of physical objects which are not themselves phenomenal properties, but are somehow intrinsically suited to constitute phenomenal properties (clearly, our grasp of such qualities is frustratingly meagre). Perhaps the Russellian monist could hold that Implausible Consequence obtains in our world, but is explained in terms of some more fundamental laws involving protophenomenal properties, laws which do not involve such arbitrarily precise specifications. However, even on the supposition that consciousness is not fundamental, it still seems pretty implausible to suppose that a slight adjustment to a single fundamental particle – one of countless billions – in the brain could make the difference between the whole brain having or lacking the determinable property of consciousness. So I am inclined to think that the considerations outlined here have some force against the Russellian monist, even though the argument is primarily aimed at, and has more force against, the property dualist.

    You’re quite right that Anthony’s argument is only inconsistent with thoroughgoing semantic externalism, rather than the semi-externalist of Chalmers and Jackson. The trouble is that there are a hell of a lot of thoroughgoing externalists (outside of Oz), e.g. Millikan, Papineau, Fodor, all these theories cash out reference fixing without involving the a priori. Worse still, I think a posteriori physicalists pretty much have to be thoroughgoing semantic externalists (If phenomenal concepts revealed to us anything about the nature of phenomenal qualities, and if phenomenal qualities are physical, then phenomenal concepts would reveal to us something of the physical nature of phenomenal qualities (which is inconsistent with a poteriori physicalism). This difficulty could be avoided if phenomenal concepts revealed the accidental qualities of phenomenal qualities in virtue of which we pick them out, but Kripke has persuaded most that it is implausible to think we pick out phenomenal qualities in virtue of accidental qualities), so Anthony’s argument really begs the question against a posteriori physicalism.

    I’m not sure I get your worry with my claims about personal identity. I hold that it is a vague matter whether or not the relevant psychological relationships between mental states required for personal persistence hold or not, so I don’t think the problems outlined in my paper apply here. Also, there doesn’t need to be a ‘reduction of consciousness of the whole to consciousness of the parts’. All brain states are conscious, but only some of those conscious experiences count as ‘Philip’s conscious experiences’, namely those which bear the right kind of psychological relationships.

    I was curious as to why Sider didn’t take a parallel to this proposal to be capable of refuting his own vagueness argument unrestricted composition. In any case, it’s not clear to me how adding indeterminism to the laws would avoid the implausible consequence. I think it needs spelling out a bit more.

    Regarding the support of Professor Robinson, even if we suppose the conscious experiences get weaker in some sense, there is still a huge chasm of a leap from a whole brain/organism having a certain way of being modified, to its lacking that way of being modified.

    I don’t quite see your worries with focusing on the diachronic rather than the synchronic. Can you spell out a bit more the concerns in the next to last post? The concerns in the last post seem to me of a practical, and hence not really a worry given the speculative nature of the thought experiment.

  19. TURNING OFF THE LAST LIGHT (Reply to Philip Goff)

    P.GOFF: “I don’t know what you mean by ‘primitive’ and ‘refined’ and ‘NP-complete’ and I’m not sure how you get to the conclusions you reach regarding panpsychism.”

    I was quoting you, referring to “some very primitive (non perceptual, non-cognitive) kind of consciousness.” So “refined” would just be the complement of your own category.

    “NP-complete” means that individuating “panpsychic” states (as entities, subparts of entitities, supraparts of which entities are subparts, and all permutations and combinations and arbitrary disjunctions of entities, subparts, supraparts, permutations, combinations and arbitrary disjunctions — leads to combinatory explosion. (More likely, I think: incoherence: an uncountable infinity of feeling entities. A monstrously huge hammer to hit one tiny terrestrial nail…)

    P.GOFF: “The concept of a ‘kidney’ or a ‘cerebellum’ is presumably vague, and hence the difficulties don’t arise that do with a non-vague concept of ‘consciousness’”

    Kidneys and cerebella are not concepts but objects (like brains and salt-pillars). If (modulo ordinary physical Sorites) there is nothing untoward (or panpsychically impelling) about the atom-by-atom transition from salt to kidneys and back, then we can forget about everything between kidneys and salt and simply ask whether there is anything in the atom-by-atom transition from unfeeling kidneys to feeling cerebellum that entails panpsychism (assuming, correctly, that feeling is all-or-none rather than gradual).

    The difference between a kidney and a cerebellum is largely a difference in physiological organization — even if the reorganization is done atom-by-atom, Sorites style. Hence there is nothing anomalous or untoward (or panpsychically impelling) about the possibility — indeed the likelihood — that the difference between a physiological organization that is unfelt and a physiological organization that is felt, even if done molecule by molecule, Sorites-style, is an all-or-none one, just as the difference between a dark and a light building is an all-or-none one, depending on who is the last to turn off the last light.

  20. Philip,
    I liked this paper a lot. I just wanted to pose a question about the scope of this argument. As I’m sure you know, historically many ’emergentist’ philosophers thought that not only consciousness but chemical, biological, etc. properties appeared on the basis of certain complex compositions of matter, in a ‘strongly emergent’, i.e. unpredictable in principle, way. Largely this has become less popular because of empirical advances in e.g. explaining chemical properties physically.

    If your argument works for mentality, will it also provide a principled reason for rejecting such ‘compositional laws’, on the basis that 1) the possession of irreducible chemical etc. properties, if they really are fundamental, can’t be vague, and 2) we should recoil from arbitrarily precise laws of nature?

  21. re: personal identity. I understand the possibility that it be vague which of many (precise) conscious beings you are. My worry is that if your version of panpsychism is true, each of these beings will be having conscious experiences. I see how you can make room for coming into and going out of being, but not how you can make room for temporary unconsciousness – unless you say, as some distinguished thinkers certainly have, that temporary unconsciousness is actually temporary cessation of existence. Is that what you are thinking?

    re: Sider’s suggestion. Very interesting question why this wouldn’t block his own vagueness argument for 4Dism (or Lewis’ for Universalism). I think the major crux for Ted at this stage in his argument is whether we have to take the persistence facts to be fundamental (rather than grounded in local matters of particular qualitative facts). He offers us an alternative position – plenitudinous four dimensionalism – from which (he seems to think) we do not, and from this perspective the question of how smoothly we can write the laws of this new fundamental stuff, is like fussing over deck chair arrangement on the Titanic. In contrast, the debate between the Panpsychist and the Property Dualist is a debate where both sides agree we need fundamental laws of some form here.

  22. Quick reply to Philip’s three points:

    (i) Good, so Philip concedes that my argument doesn’t require the phenomenal transparency premise. He rejects one of the premises that I use instead (1-conceivability entails 1-possibility). There’s a lot to say about that, but in any case it’s obviously a different premise, having nothing specifically to do with phenomenal concepts.

    (ii) Our lack of protophenomenal concepts gives a potential explanation of why there may be a priori accessible sharpenings that we haven’t yet found.

    (iii) I don’t see what’s implausible about the relevant object’s consciousness going from miniscule to zero as phi goes from k+epsilon to k as some particle moves. Arguably something like this happens to me every night when I fall asleep!


    Hi Philip

    Thank you for your reply, brief though it was.

    You write: ‘Nagel’s phrase has proved itself to be very useful for directing the listener’s attention to [the qualities of experience apparent to introspection], which is why it has caught on.’

    I’m afraid I don’t see this as a philosophical argument. At best, it’s a description of an empirical state of affairs – and a very questionable one at that. ‘Proved’ itself? Where is the proof? And useful for whom? (Certainly not for me – and not for Hacker either, apparently). And useful for what? You say: for revealing ‘qualities of experience apparent to introspection’ – which I assume implies that it reveals significant aspects of human consciousness. But does it really do this, or does it just encourage people to think it does, when in fact it doesn’t? (As we know, people can believe all sorts of things that are not the case. And many of those things ‘catch on’.)

    The philosophical question I raised was in any case not about use but about meaning. You yourself say in your opening statement: ‘By “consciousness” I mean the property of being a thing such that there’s something that it’s like to be that thing.’ You are – or so it certainly appears – explaining what you mean by consciousness. Moreover, as I pointed out, attempts to explain the meaning of consciousness in this way are very common in the contemporary philosophy of consciousness. Here are just two examples from other people in the field (both of whom happen to be participants in this conference):

    ‘A mental state is conscious when there is something it is like to be in that state.’

    ‘For an experience to be phenomenally conscious is for there to be something it’s like; our talk about phenomenal consciousness accordingly is permeated by the expression “there’s something it’s like”.’

    (True enough. Even on this very page, I notice, the idea is invoked more than once.)

    We can hardly accept these claims unthinkingly, can we? Especially since they seem to have such central importance. As philosophers (and this should go double, I imagine, for analytic philosophers), we need to ask what they mean – we need to analyse them.

    And that was precisely what I did in my initial comment – and came up with the conclusion that, under any reasonable interpretation of ‘like’, the proposition is, as I put it, ‘at best an empty truism’ – camouflaged by the ambiguous use of ‘like’. This was not, to quote you, an ‘implausible [analysis] of “what it’s like” talk in terms of superficially resembling bits of language’.* It was a close analysis of the proposition in terms of the only meanings of ‘like’ that seemed remotely probable (and I invited you to suggest others if you could think of any). If my analysis is implausible, I ask you again to show where.

    So my question, again, is very clear: given that the ‘something it is like’ idea holds such an important place in the philosophy of consciousness, and given that my analysis appears to show that it is worthless, is my analysis wrong – and if so where? – or is the idea in fact worthless? I realise it may be difficult to let go of an idea that has become so firmly embedded in the philosophy of consciousness, but if that idea is a dud, surely it’s better to recognise that and look elsewhere?

    (You also raise the question of ‘introspection’ and you’ll notice I haven’t pursued that. I have my own views on the viability of ‘introspection’ as a path to an understanding human consciousness but, once again, I don’t want to distract attention from the central issue at stake here – the meaning and viability of the ‘something it is like’ claim.)


    * I think you are talking about Hacker here but perhaps the implication goes further. I should add that I arrived at my conclusions quite independently of Hacker – though it was nice to find someone else thinking along the same lines. As yet I haven’t even read his book – just a resumé of some of his arguments.

  24. Richard: The premise that 1-conceivability entails 1-possibility is general, having nothing to do specifically with the phenomenal. So the case for the premise can’t presuppose phenomenal transparency any more than it presupposes chemical transparency. And it obviously doesn’t presuppose the latter, as the water/H2O case brings out. Maybe the premise requires phenomenal and chemical “non-radical-opacity”, in Philip’s terms, but as the name suggests, non-radical-opacity is far weaker than transparency (it’s perfectly compatible with ordinary opacity of the water-H2O sort).

  25. Postscript to ‘THE SOMETHING IT IS LIKE ARGUMENT (cont’d)’

    I forgot to add in my last comment that, of course, anyone else is cordially invited to respond to the argument I have put – which some may well want to do, given that the proposition I am criticising, and describing as worthless, seems to have lots of supporters and seems generally to be regarded as a very important element in the contemporary philosophy consciousness.


  26. Hi everybody!
    Great paper and great discussion. Just one comment:
    I do not see unpalatable that a slight alteration to a fundamental particule make a difference in a higher-level property. Imagine an electrical circuit made out of transistors that has the property of being an adder of two one digit numbers. If God start removing fundamental particles from one of the transistors I see no reason for thinking that at a precise removement the circuit loses this property.
    If higher-level properties are non-physical, property dualism can embrace some form of proto-panpsychism , without committing herself to the claim that the salt statue is conscious.
    If qualia are identical to functional properties, as God replaces Lot’s fundamental particles, qualia start dissapearing (suddenly there is no headache, then no redness, then no sadness, etc) because their states cannot perform the corresponding function.

  27. Phillip, I don’t get your remark that a posteriori physicalists must embrace radical externalism (which I’ll understand as the claim that there is absolutely no a priori knowledge from concept mastery, not even defeasible knowledge of contingent claims, like ‘Julius invented the zip’, or ‘Water tends to be watery’), rather than semi-externalism. Not sure what you take that latter position to be: plenty of logical space between the negation of the claim in parentheses above, and the Chalmers-Jackson Plan. A Posteriori Physicalists have to reject versions of the Plan – they must reject that everything is a priori entailed by the metaphysically fundamental facts, for example – but that doesn’t mean they cannot take there to be any such thing as a priori justification.

    Re: the actual argument you give. As with your other uses of the transparent/translucent/opaque distinction, I sort of grasp what you are getting at, but I want to see a non-metaphorical, exact account out of what this distinction strictly amounts to. And, as I outline in my comment paper, using the ‘revelation’ formulation as a case in point, I worry that once we have it, the distinctions will not be able to do all of the work you want to put them to.

  28. Response to Philip Goff’s post on 2/22 at6:41

    Philip, thanks for the response. “Bill” will do fine.

    The most important part of your response, in my view, is your remarks about conscious experience requiring a subject to have the experience. I think that’s treacherous commitment. But it doesn’t seem appropriate to turn our discussion into a rather different debate; and in any case, it would take far too long for me to give an explanation of my views on this issue that would not raise more questions than I could adequately respond to. So, on this matter, I’m just going to point to the references at the end of this response and make a rather formal point here.

    To wit: The upshot of making the assumptions about subjects of experience (and determinables/determinates, of which a bit more anon) is that your thesis is, in effect, more complex than what’s reflected in your title. I.e., even apart from any other points that might be raised, we wouldn’t know that property dualists should be panpsychists, but at best that property dualists who hold certain non-trivial views about subjects of experience should be panpsychists.

    Now, some less formal points that may offer grounds for resisting even this more complex thesis. One of them concerns the determinables/determinates point. My model would be that consciousness is to episodes of consciousness (I’ve called them “qualitative events” elsewhere) as sound stands to a particular tone. Of course (as I think my original comments allowed for) we can get rid of a particular tone in one way by, say, tightening a string so that we get another tone; analogously, we can get rid of one kind of qualitative event by changing it into some other kind, i.e., we can have consciousness, but of a different kind. But we can also reduce the intensity of the tone until there is none left; and then (assuming there are no other sources of sound) we will have got rid of sound, and will have done so by getting rid of the particular tone. Analogously, there is a way of getting rid of a certain kind of consciousness (say, visual consciousness) by gradual steps that end in no visual consciousness. And then, if we do the same for all forms of consciousness, we will arrive at absence of all consciousness.

    You remark that “It has to be shown that consciousness is more like velocity than shape.” Yes, that has to a defensible part of a theory that’s a property dualism but not panpsychistic. But if you are assuming the contrary, that amounts to a way of complexifying your thesis, which might now be something like this: Property dualists who hold a certain assumption about subjects of experiences and a certain assumption about the determinate/determinable structure of consciousness should be panpsychists.

    You also remark that “When my auditory experience of a single tone ‘fades out’, I am not becoming less conscious . . . .” I certainly agree; I thought it was clear in my original comments that consciousness would not be absent unless all its forms ‘faded out’. I just meant to talk about one example to illustrate the kind of possibility I had in mind.

    Regarding the property of being something such that it is like something to be it, I think makes good sense to hold that if each kind of way it is like for X to be something were to fade out, then there would no longer be anything it would be like to be X. And I think the analogous point holds for the *ability* to get into a state where there would be like something to be X.

    But maybe this last point is one that can be resisted if we stick to your Lot’s wife example? I thought it would be clear how I could go back to particle talk, but I see I was wrong about that. So I’ll try to meet my obligation here. On my view, qualitative events depend on patterns of neural activation. So, let’s imagine that a god replaces Lot’s wife’s molecules one by one with NaCl. Some of the replaced molecules will be proteins in her neural membranes, some will be neurotransmitters, etc. The first few of these replacements might have negligible effect on her neural activation patterns. As more molecules are replaced, however, the activation patterns that her neurons can get into will be sufficiently different that she’ll notice that things don’t look/taste/sound, etc. the way they used to. With more replacements, vision will dim, deafness will set in, etc – all because the damaged neurons can no longer get into the activation patterns on which consciousness depends. Memory and reflection abilities will degrade too. Eventually, there will not only not be neural patterns on which consciousness depends, but also no ability for what’s left of Lot’s wife’s neurons to get into such patterns.


    W. S. Robinson (2004) _Understanding Phenomenal Consciousness_ (Cambridge: Cambridge University Press).

    W. S. Robinson (2010) _Your Brain and You_ (New York: Goshawk Books). Description and link to Amazon are available on my web site,

  29. Steve Harnad

    Where does the quotation from me come from? I’m afraid I still don’t understand your explanation of ‘NP-complete’. What does ‘NP’ stand for?

    Kidneys are cerebella are objects, but there are concepts of those objects. I take it to be reasonably uncontroversial that those concepts are vague, whereas if I am right the property dualist must accept that the concept of consciousness is vague. This is the crucial difference (for a property dualist) between Sorites series involving kidneys and cerebella turning into different things, and Sorites series involving conscious things turning into non-conscious things.

    Great question. I’ll have to think about that. I would have thought it’s more likely to entail the falsity of biological emergentism: implausible to suppose that a slight alteration at the fundamental level can make the difference between the having and lacking of a certain sparse biological property. It’s not so clear to be that it is similarly implausible to hold that a slight alteration at the physical level can result in the loss of a chemical property, given the proximity of the physical and chemical levels.
    Jonathan Simon
    re. personal identity: Hmmm…I’m not sure what to say now because of my lack of expertise re. personal identity. What do Lockeans say about an organism in dreamless sleep such that there are no occurrent mental states psychologically continuous with past and future mental states? Perhaps they say that there are dispositional mental states still present. In that case, I would say that I am still present in dreamless sleep for the same reason, but I am not conscious as there are no conscious states bearing the right psychological relations to past/future mental states.
    re. Sider: Not clear on your claim here. Both Sider and his opponents think there is a mereological principle governing composition: he holds a principle of unrestricted composition, but others hold a principle of restricted composition. Things seem exactly parallel in the debate I am concerned with: the panpsychist property dualists hold a principle of unrestricted phenomenal composition, whilst non-panpsychist property dualists hold a principle of restricted phenomenal composition.
    Re. a posteriori physicalism: Once one accepts the Kripke claim that we don’t pick out phenomenal qualities in terms of accidental properties, then either we refer to them in virtue of an implicit understanding of (part of all of) their essential nature, or reference is fixed wholly outside the head. I take it that the a posteriori physicalist can’t say the former, as this would imply that we have a priori understanding of at least some of the physical or functional nature of conscious states (conscious states have a wholly physical or functional nature for physicalists, therefore if we understand some of that nature a priori, then we would know a priori that at least part of their nature is physical functional), therefore they must hold the latter.
    How’s about this for a definition of the terms in my framework:
    Transparent concept: A concept c of a property p is transparent iff there is a possible world where someone, in virtue of possessing c and without relying on any empirical knowledge other than what it is required to possess c, comes to a complete understanding of what it is for an object to have p.
    Translucent: A concept c of a property p is translucent iff there is a possible world where someone, in virtue of possessing c and without relying on any empirical knowledge other than what it is required to possess c, comes to a partial understanding of what it is for an object to have p.
    Opaque: A concept c of a property p is opaque iff there is no possible world where someone, in virtue of possessing c and without relying on any empirical knowledge other than what it is required to possess c, comes to either a partial or complete understanding of what it is for an object to have p.

    David Chalmers
    (i) In the paper I talk about the anti-physicalist argument which tries to show the falsity of physicalist by demonstrating the possibility of zombies worlds. I think this argument doesn’t go through without a commitment to phenomenal transparency. However, I now see that Dave is right that there is a version of his anti-physicalist argument which requires only that phenomenal concepts are not radically opaque. Having said that, once we accept Kripke’s claim that we don’t pick out conscious states in terms of their accidental properties (which in general both sides do), it follows that phenomenal concepts are either transparent/translucent or radically opaque, so the category of mildly opaque drops out of the picture. Still, I concede that strictly speaking all that is required is the denial of radical opacity. I will adjust footnotes accordingly to explain this.
    (ii) I think that the thesis that concepts outside of those required to possess the concept of consciousness are required to grasp sharpenings of the concept of consciousness, is inconsistent with phenomenal transparency. If what is ascribed in an application of the consciousness concept is to be understood in terms of its indeterminacy over a certain spectrum of sharpenings, then one cannot understand what is ascribed without understanding that spectrum of sharpenings.
    (iii)I think I need to look at the details of this phi stuff. Do you have a reference?

    Derek Allan
    I think your analyses are implausible because, as you yourself have shown, they obviously aren’t up to the job. If we were obliged to define what it’s like talk in terms of other bits of English which have ‘like’ in, then you would indeed have demonstrated that ‘what it’s likeness’ talk is misguided. But you haven’t given an argument as to why we are obliged to do this.
    I do take what I have said to explain the meaning of ‘what it’s like talk’.
    Question: ‘What does ‘what it’s like to feel pain’ mean?
    My answer: ‘When you’re in pain, and you attend to your pain, there is a quality of the pain, how the pain feels, which is apparent to you. It is exactly that quality I intend to refer to by talk of ‘what it’s like to feel pain’.
    Why does this answer not give a satisfactory explanation of the meaning of that phrase?
    Miguel Sebastian
    I don’t say that for every single higher-level property, it is implausible to suppose that a slight alteration could make the difference between the presence and the absence of that property. Of course we could artificially invent a property that would, e.g. the property a tiger has of being composed of n (utterly precise) number of particles.
    I accept that the functionalist is not subject to this argument, as she can plausibly make sense of the concept of consciousness being vague. The argument is aimed at property dualists.
    So the argument is not primarily aimed at protophenomenal guys, although I do think it has some force against them, see footnote 13 quoted in full above.

    Bill Robinson
    I think there’s something to your claim that my thesis is more complex that at first appears, and I’m thinking I need to add the claim that consciousness does not admit of degree and defend it a bit at some point (at which point you will of course be suitably credited). Nonetheless, I think it’s a pretty plausible thesis. Suppose all my auditory experiences are representing progressively fainter noises, all my visual experiences fainter colours, and so on for all my other senses. It think it’s wrong to describe this as my having the property of having an inner life to a lesser degree. My inner life changes as the process goes on, but I have an inner life just as definitely towards the end of of the process as I did at the start. You either have an inner life or you don’t.
    I think the temptation to say what you want to say is given by the fact that you have chosen cases where consciousness represents properties that admit of degree. So what you describe as an auditory experiences fading out, is really consciousness representing sound getting quieter.
    I’m not clear how this is supposed to work visually. I can imagine a specific colour fading out: red fading, say, to reveal white behind. But I can’t make sense of my entire visual experience ‘fading out’.
    I think there is a serious difficulty in explaining the conditions for consciousness in terms of neuronal activity, rather than in terms of activity at a more fundamental level: for any neurophysiological state, it is surely a vague matter whether that state is instantiated. If phenomenology is determined by the neuronal level, then it would be vague which phenomenal states are instantiated. But this is not a not vague matter (on your account, although consciousness admits of degree it is not vague, right?).

  30. I’ve been pre-occupied in discussing Stevan Harnad’s Minds, Brains and Turing paper up to this point (as it deals with an issue I find especially interesting) but I have periodically poked in on some of these other discussions. I won’t pretend to be doing any heavy lifting here because of my low level of attention, however I did notice the following quote from Philip Goff (and have seen him make the same kind of comments earlier):

    “Suppose all my auditory experiences are representing progressively fainter noises, all my visual experiences fainter colours, and so on for all my other senses. It think it’s wrong to describe this as my having the property of having an inner life to a lesser degree. My inner life changes as the process goes on, but I have an inner life just as definitely towards the end of of the process as I did at the start. You either have an inner life or you don’t.”

    I think this intuition is more likely an artifact of how our minds work than a true picture of how things are. After all, levels of consciousness are surely observable in the animal kingdom where higher mammals seem to be more engaged in a wider sphere of phenomena in their world than lower level mammals and various other species.

    My late cat, for instance, lacked all the intellectual discriminatory capabilities I generally ascribe to myself and yet she was able to differentiate herself from other things around her quite easily and to differentiate other things with one another by type. She knew that a mouse was for chasing and eating but I was for petting and feeding. She didn’t expect much more from the sitting furniture than a comfortable place to curl up.

    Years ago I used to keep turtles and they had a much less extensive command of their environment. Isn’t this evidence at least of the gradualness of consciousness, that it is not an either “on or off” phenomenon? I recognize some may want to say that the cat or the turtles simply were differently equipped than humans and so lived in a different environment. But then one of the things about higher mammals is that they become increasingly able to interact with their environment as we are the higher up the evolutionary ladder they go (and the more complex their brains).

    And there are other cases wholly within human experience which argue against the all or nothing feature being ascribed here to consciousness. An alzheimer’s patient gradually loses awareness of what’s going on around him and of the wider world beyond his immediate surroundings. Most sadly, perhaps, the patient’s personal history fades away. With the progress of the condition, it becomes increasingly difficult to keep track of his personal history even over very short timespans. Did he just eat ten minutes ago and forget?

    Yes, there’s a point where the consciousness of the patient will blink out entirely, as there is for all of us, but is it unreasonable to recognize in the slowing down of the thinking mechanisms of an alzheimer’s patient’s mind the gradual shutting down of mental processes? And this surely implies that consciousness seems to occur on a continuum of sorts.

    In the end, such patients, if they live long enough, end up as vegetables and here they have less conscious capacity than my late cat who, whatever her limits for not having a human brain, still had the capacity to recognize, distinguish and respond to elements in her environment.

    The same problem occurs with a wide host of brain impairments or when we use various chemicals as drugs. Now when do we decide consciousness has been turned off? The truth is, we just don’t know. But why conclude from the fact that such a point is reached in all living things, that there’s a radical difference on either side of the critical threshold?

    What if it is just a matter of gradual accretion or removal of various functions? What if it just becomes harder for observers to discern differences along a continuum of capacities the closer they are to one another? And does not the everyday occurrence of falling off to sleep suggest that there is a fading quality to consciousness?

    My point here is not to take on the entire thesis offered in the initial paper but just to remark that the assumption that consciousness is an “on/off” phenomenon is not a given — nor is it necessarily supported by what we know about how things work.

  31. Pingback: The Zombie Argument Depends on Phenomenal Transparency « Philosophy Sucks!

  32. SORITING THINGS OUT (Reply to Philip Goff)

    P.GOFF: “Where does the quotation from me come from?

    Here: P.GOFF: “Most panpsychists, though, tend to hold that only fundamental particles and organisms have consciousness; the former having only some very primitive (non perceptual, non-cognitive) kind of consciousness.” February 23, 2011 at 05:35

    P.GOFF: “I’m afraid I still don’t understand your explanation of ‘NP-complete’. What does ‘NP’ stand for?

    P.GOFF: “Kidneys are cerebella are objects, but there are concepts of those objects. I take it to be reasonably uncontroversial that those concepts are vague”

    Are you morphing brains into pillars of salt or concepts of brains into concepts of pillars of thought? If the former, there is nothing unproblematic or vague about putting a brain on a slab here and a pillar of salt there. Then you start your soritean morphing.

    I suggested nothing obviously important would change it you started by putting your brain on one slab and a cerebellum on another, and morphed the brain into the cerebellum.

    Similarly, salt on one slab and a kidney on the other. Morph the salt into the kidney.

    So far, the cerebellum is still conscious and the kidney is not. Now you morph the kidney into the cerebellum, and there’s nothing wrong, or even implausible, about the prospect that somewhere alone the way there is an all/none change, like turning on the lights, probably mainly a change in organization rather than material. (But certainly nothing that makes it necessary to smear conscious over every entity, part and disjunction of entities and parts in the universe.)

    P.GOFF: “whereas if I am right the property dualist must accept that the concept of consciousness is vague. This is the crucial difference (for a property dualist) between Sorites series involving kidneys and cerebella turning into different things, and Sorites series involving conscious things turning into non-conscious things.”

    Kidneys are unconscious, cerebella are conscious, and consciousness (feeling) is all-or-none, not vague (otherwise the Cogito would be a matter of degree!)


    Hi Phil

    I’ll take you reply bit by bit.


    I’m sorry. What ‘job’ are we talking about? I’ve argued that your statement ‘“By ‘consciousness” I mean the property of being a thing such that there’s something that it’s like to be that thing’ is vacuous. If my ‘job’ was making that statement *not* vacuous then, true, I certainly didn’t succeed. Could you explain what ‘job’ you have in mind?


    I don’t follow the logic here at all. Unless you want to say that the word ‘like’ in your statement means nothing at all – and I assume you don’t want that – the only way we can consider its possible meanings is to look at meanings it assumes in other contexts (what you call ‘other bits of English’). This is what I did. Do you see some other way of proceeding?

    (Of course, there is another meaning of ‘like’ I did not consider – the one it assumes in ‘I Iike ice cream’ but that’s obviously not relevant so I didn’t discuss it. What I did was to show that under either of the two probable meanings, your statement – and of course it is not only yours – is vacuous.)


    No, I’m sorry it didn’t. You made a comment about the alleged *usefulness* of the proposition – which I have challenged, pointing out that it makes unwarranted assumptions. But you said nothing to address the issues of *meaning* I raised.


    First, let’s recognise that your answer doesn’t tell us anything about the feeling of being in pain specifically (e.g. it doesn’t tell us we’re going to feel very uncomfortable etc). It’s a comment applicable to any experience – joy, sadness, fear, homicidal anger, suicidal depression, and so on. But it’s an uninformative comment. One necessarily ‘attends to’ an experience – that is, it commands some degree of our attention. (Could one be ‘in pain’ and not be aware of it? One wouldn’t *be* in pain in that case.) Similarly, any experience ‘feels’ a certain way (it couldn’t not feel in any way), it must have a certain ‘quality’ (it couldn’t have no quality) and is ‘apparent’ to us in some way (it couldn’t be non-apparent).

    So this gets us nowhere. It’s tantamount to saying that all experiences are experienced, or something equally unhelpful.


    I don’t know what ‘that quality’ could mean. It can’t be the quality of being in pain specifically since you’ve said nothing about that. And, as indicated, there is no such thing as a ‘quality-less’ experience, so while I’m perfectly happy to agree that, in some general sense, all experiences have a ‘quality’ (just as all objects do) this obvious point gets us nowhere. Your reference to the ‘something it is like’ idea is apparently intended to throw more light on the subject, but all that does is land us back with the problem I explained carefully in my first post: what do we mean by ‘like’?

    If it’s a ‘like’ of comparison, then quite obviously pain is (somewhat) ‘like’ certain other experiences – suffering, agony etc. But again that doesn’t help. All sorts of experiences, things, etc can be compared – whether they’re similar or not. That’s hardly news.

    If it’s a ‘like’ of ‘feeling like’ (as in ‘I feel like going to the movies’) then, yes, presumably, people in pain can feel like doing all sorts of things – saying ‘ouch!’, bellowing, reaching for pain-killers, even asking for more if they happen to be masochists. There are a hundred and one possibilities. But again this is hardly an informative observation.


    Well, what I’ve said here, plus what I said originally, should answer that. Let’s remember, too, that although I’ve adapted my comments here to the experience of pain, because that’s the example you chose, the ‘something it is like’ idea is usually employed, as you originally employed it, as an explanation of the meaning of *consciousness* – i.e. it is given a much broader and apparently quite central significance in the philosophy of consciousness.

    I am not an analytic philosopher – there are too many aspects of this school thought that I find unattractive – but one of the central claims of analytic philosophy, as I understand it, is that it always pays careful attention to the meaning of words. This is central to its oft-repeated claim that it is ‘rigorous’. Now my original post did just that. It carefully analysed the possible meanings of ‘like’ in the ‘something it is like’ claim and concluded that under any interpretation the claim was vacuous (a vacuity camouflaged, as I pointed out, by the ambiguous way ‘like’ is used). Hacker, I gather, has advanced similar arguments. Now I assume that most of those involved in this conference are of the ‘analytic’ persuasion and it strikes me as odd that no one has as yet engaged with the issues I raised. Surely analytic philosophy has not abandoned its claims to ‘rigour’?


  34. Is it useful to speak of consciousness as “something that it is like to be”? It does, indeed, seem a very strange locution as Derek Allen points out.

    Yes, being conscious (having consciousness) implies awareness and THAT implies a way of relating to the world that is not available to the non-conscious (e.g., tables, chairs, rocks and similar stuff). But what more is gained than that implication? And does something more need to be gained?

    Is there a need to delineate consciousness from what clearly isn’t? Or, perhaps, the issue is more to do with delineating levels of consciousness among entities entitled to be at least considered as being possibly conscious such as humans and bats and even, say, computers, given certain information about some of them?

    Is anything really informative finally gained by speaking of consciousness as the condition of being something that it is like to be?

  35. S. Mirsky and D. Allen:

    I think that you are attacking the “something it is like to be” phrase as strange and meaningless, and you are (sort of) right, but the attack nevertheless is misguided here.

    Here’s why: you appear to not care or know that this phrase (as used by Nagel and others) is a philosophical convention, and is supposed to be a mere pointer to aspects of our subjective experiences which are private to our consciousness and which cannot be expressed in language alone.

    The “something it is like” is thus supposed to refer to something that is not understandable objectively, only subjectively. And I think you are criticizing that phrase to some degree because it does not describe an objective thing. You are correct, but beside the point, perhaps?

  36. When It is Like Something to Be Like Something

    I’m sure we are both aware that the use is a philosophic convention. For myself, I can only say that I always found the locution odd. When I first heard it there did seem to be something there, however strangely expressed, but over the years it has come to seem to me a pointless point.

    Well, we know that being a subject means having experience, means being aware of our world, means, as Stevan Harnad (and Antonio Damasio) puts it, “feeling” what is happenning, etc. Perhaps as you say, there is no better way to say this (language does seem to fail us here) but just because something is a convention shouldn’t mean that it should not be reconsidered and, perhaps, found wanting. In this case I think that the strangeness of the formulation works against it, no less than its seeming portentousness.

    I don’t want to butt in too much to this discussion because I delayed getting involved and am too lazy to read through all the material that’s come before. But I did just want to raise a couple of issues and this was one (the question about assuming an all or nothing nature for consciousness being another).

    I think Derek Allan does yeoman’s service here in finally calling our attention to a locution which seems to promise more than it delivers.

  37. I wanted to ask more about consciousness depending “on some quantitative property” of systems and the possible laws of nature that they “are only conscious when phi > k” or involve “a constant that has something like this as a consequence.” Chalmers notes that there are numerous arbitrary constants in theories of nature but it seems to me like this may be a different, perhaps more universal fact (but maybe this is close to it being fundamental, and so we aren’t so far apart). It seems like it is doing some work here setting the threshold of complexity needed for consciousness. If it has some explanatory part of a theory then it seems non-arbitrary.
    Miguel asks about the “precise removement” where a circuit loses this property but I am not sure on the connection to function. The issue seems less about a function stopping and more about building a critical velocity or mass. I like Whitehead’s notion of society. I find it helpful to think that the really real things are the processes that do not endure but from which arise through combination larger and more complex beings. There’s not much of a role for substances but rather events, so maybe I shouldn’t say beings: the idea I want to stress is that the continuing thing, be it a person or a stone, is made up of fundamental events, and that these processes are the real building blocks. This makes the transition to consciousness easier to imagine, at least to me, and I think that there should be (discoverable) theoretical claims about why the subjective feel starts at a point. If so, this would make a physical constant not arbitrary.
    Lastly, the idea of a quantitative property is maybe not as helpful as a theory that would explain why language or concepts enrich our experience: how the expert wine taster or opera lover has a different experience and how this is explained by more sophisticated higher order thoughts. I will say more about why I think Rosenthal’s theory sheds some interesting light on these issues but I mention some of them in my post on here last year.

  38. Hi drbillh

    Thank you for your comment.

    But I’m not at all sure what you mean by a ‘convention’ in this context.

    It’s a convention, for example, that in Morse code three dots means ‘S’ but, of course, there’s nothing intrinsic to three dots that should make it mean S rather than, for example, X or wheelbarrow. Three dots is a ‘mere pointer’ to S in your phrase – by convention.

    Are you saying that, for users of the ‘something it is like’ locution, there is nothing intrinsic to the locution that is seen as identifying what consciousness is in some way? For example, if the locution were changed to ‘there’s a bird in that tree’ or ‘swongle’ (a word I just invented) do you think those alternatives would satisfy them just as well?

    It’s worth bearing in mind comments like the following (quite recent ones by two prominent members of the philosophy of consciousness field):

    ‘A mental state is conscious when there is something it is like to be in that state.’

    ‘For an experience to be phenomenally conscious is for there to be something it’s like; our talk about phenomenal consciousness accordingly is permeated by the expression “there’s something it’s like”.’

    And then of course there’s Philip’s own similar statement which begins ‘By ‘consciousness’ I mean…’

    Those statements strike me as more than mere appeals to a convention, at least in the sense of ‘convention’ described above. To me they even sound rather like definitions…


    PS. My thanks to Stuart Mirsky for his supportive comments.

  39. Hi drbilh

    Thank you for the link but I’m not sure why you gave it to me. The posts on the page you refer me to seem to be mainly about the (silly) idea of a zombie. I was hoping you might reply to my query about what you meant by saying that the ‘something it is like’ phrase is just ‘a convention’.


  40. Hi drbillh

    I really don’t think we can say the issue is ‘off topic’ given that it is there, front and centre, in the opening paragraph of Phil’s paper.

    As for what the ‘something it is like’ proposition ‘points to’, I provided a careful analysis of that matter in my first comment here arguing that it pointed to nothing at all – that it is, as I put it, ‘at best an empty truism’.

    I realize it might be very difficult for philosophers working in this area to accept that a proposition they have relied on so long and so often simply falls apart under analysis (especially since this is ‘analytic’ philosophy…) but an obvious solution to that dilemma might simply be to show that my analysis is wrong. That would of course involve addressing my arguments which so far no one has tried to do. (I gather Peter Hacker has put forward arguments similar to some of mine if that provides a little more incentive.)

    Am I correct in assuming you don’t want to pursue your claim that the ‘something it is like’ phrase is merely a ‘convention’?


  41. Hi Derek,

    I was a little confused about the linked to article as well but I think the overall point he is making is apt (it is essentially the same one Philip made in response to your question). When philosophers say that by consciousness they mean the property of there being something that it is like to be the creature they are trying to call your attention to something that should be obvious to you. There is something that it is like for me to be cold and what it is like for me is different than what it is like for me to be warm. Both of these are different from what it is like for me to see yellow or hear an oboe. There is nothing profound about this. The phrase has persisted because it has proved to be very useful in getting people to focus on one of the properties that philosophers working on consciousness are interested in. Once I have picked this property out it is intuitively the case that tables and chairs lack it while dogs and rabbits (even bats) have it (it is also intuitive that zombies would lack it and that MAry would learn about it when released from her room, etc). It also turns out that explaining how this property is/can be physical is the problem in consciousness studies but that has nothing to do with the phrase that we use to pick it out. So, sure, the phrase that is used is not completely arbitrary and so perhaps not completely conventional but the convention involves using the phrase as denoting or picking out the relevant property and so analyzing the phrase isn’t going to be of much use. One doesn’t learn about death by analyzing ‘kick the bucket’ even though we can use ‘kick the bucket’ as a way of picking out the relevant phenomena (though I am not suggesting that we do this!).

  42. In response to Derek Allan:

    I feel I should offer my account of “what-it’s-likeness” in response to your comments, because it is an attempt to analyse the term rather than use it as an empty truism, although I recognise the points others have made about the way it is used and am not opposed to its continued use in that way. The account appeals to something I call “what-it’s-not-likeness”, and incorporates both senses of “like” that you mention, and go on to reject individually.

    The “like” of comparison:
    We identify external stimuli by placing their various properties (colour, size, rate of movement, texture, smell, etc.) along a range of scales corresponding to each type of information available to our senses, within which each differentiation relies on comparison with other points on the scale, and between scales. What things are not like (“what-it’s-not-likeness”) is part of what we think they are like. Without our having seen more vivid shades of colour, pastel shades could not be identified as pastel and have the associated qualities or “feel”, and so on. The qualitative character of sense data depends on comparison – each property’s position along a scale, relative to everything else in the same category (with categories dependent on what types of properties we are able to track) and across categories. This might seem to deny the existence of qualia, but my claim is that data only take on qualitative character once in the abstract realm – once they are compared.

    The “like” of feeling:
    Concurrently, we have physiological states that are responsive to bodily conditions (damage, disease, lack of food, sexual needs, etc.), also corresponding to points along a series of scales. These scales introduce valence – where things lie on the scale depends on whether they are conducive to wellbeing or a threat to it. Comparison, on its own, has no valence and no “feel”- this is an obvious objection to the comparison-only account given above. But the “feel” of the redness of red, for example, that is over and above our ability to pick it out as distinct from other sense experiences and recall it in the absence of red objects, arises from interaction between incidences of perception of redness and associated affective responses, grounded in physiological states, together with affective responses associated with other properties on other scales and their relation to the stimulus, including the scales for physiological states. I think there is always an affective component to conscious experience, and that this is what distinguishes it from mere computation.

    This theory needs more support than I can provide in a brief response, but there it is, with apologies to Philip Goff for not commenting on his paper in this post.

  43. Stuart W. Mirksy

    One common problem in philosophy of mind is that the word ‘consciousness’ is ambiguous, used in lots of different ways. You seem to be using ‘consciousness’ to mean awareness of the world, or some kind of cognitive ability. Of course consciousness in this sense admits of degree. I am using the word consciousness simply to mean the property of having an inner life of some kind or other (in response to your concerns about my definition of consciousness, see below). I find it difficult to make sense of consciousness in this sense being something that admits of degree.

    Steven Harnad

    Sorry for continued requests for clarification!

    Ok, so we agree on the vagueness issue. But you seem to be just denying my claim that – assuming the truth of property dualism – it is implausible to suppose that consciousness can disappear as we morph the brain into a kidney, without responding to my argument for that claim. My argument involves the intuition that the hypothesis that the fundamental psycho-physical laws of nature (remember the arguments is aimed at property dualists) specify down to minute detail how each fundamental particle in a brain containing trillions and trillions of them must be for consciousness to result is extremely theoretically vicious, and hence should be rejected if there is an alternative (which there is: panpsychism).

    Derek Allan (and Stuart W. Mirsky in this connection)

    When I say your analyses are not ‘up to the job’, I mean that they clearly do not result in the phrase having the meaning it is supposed to have. Now if there were no alternatives to analysing the term in this way, then I agree you would have shown that its use was in some way misguided. But clearly we don’t have to analyse every use of the word ‘like’ in terms of some distinct use of the word ‘like’, e.g. we can’t analyse the comparative use of the word ‘like’ in terms of the “I like chocolate” use of the word ‘like’, or any other use of ‘like’. So I think you need to offer me an argument as to why I am obliged to analyse this use of the word ‘like’ in terms of other uses of the word ‘like’.

    Actually, there is a more general use of ‘like’ which the Nagalian talk fits into: the use of ‘like’ we employ when we wish to talk of a thing’s qualities, of how a thing is. The Nagalian phrase is simply a way of signalling that we are talking about the qualities of experience which are apparent to introspection. You yourself admit that experience has qualities. But you then suggest that the fact that all experience has qualities makes ‘what it’s like talk’ vacuous. I don’t see how that follows. We are often interested in thinking about the qualities of experience which are introspectively apparent, for example, when we want to assess the thesis that those qualities are identical to physical qualities of the brain. When we are interested in the introspectively apparent qualities of experience, we use the Nagalian phrase to indicate that this is what we are interested in. We could in fact drop the Nagalian phrase, and just talk about ‘the introspectively apparent qualities of experience,’ but in my experience, using the Nagalian phrase just gets us to the same point quicker.

    Fair question: how does this allow us to focus in on some general conception of consciousness? Answer: I believe that the qualities we find when we attend to experience, how pain feels, or how red looks, are determinables of a single determinate (as square and round are determinates of the single determinate of shape). I call that determinable ‘consciousness’. Now of course there are ways in which my conception of consciousness might be challenged. Perhaps you think there are no qualities of experience apparent to introspection, or perhaps you think those qualities are not determinates of a single determinable. But I think we should cut straight to these questions, rather than worrying about what ‘what it’s like’ means.

    Andy Snyder

    Perhaps the difficulties I am raising would not arise if we go for some kind of Whiteheadian picture, however, the argument is aimed at the property dualist who has a fairly standard conception of the physical world.

    I think everything else supports me, right?

  44. Response to Philip Goff:

    I agree that mental words are generally (maybe always) vague and difficult to pin down as to referent and meaning. A term like “consciousness” is notoriously like that. Yes, I tend to use “consciousness” in the sense of referring to the condition of having a mental life as roughly equivalent to awareness. And yes, I’m very familiar with the fact that there are other uses for both terms (awareness and consciousness). By “awareness” I generally mean what Stevan appears to mean by “feeling” though I find that word more problematic in this application than he evidently does. Perhaps it’s a matter of taste and perspective? My problem with the “what it is like to be something” formulation is that, while I know what it means in the context in which you deploy it, I don’t think it adds anything to the story.

    Is that an argument not to use it thus? Perhaps not. But it is at least an argument that it is worth reconsidering it as a definition of “consciousness”. Marvin Minsky has suggested that “consciousness” is a “suitcase word” and by this he means that it doesn’t have a fixed meaning but rather many different referents which are not always easily distinguishable from one another but which often are distinguishable enough to create an ambiguity hazard.

    I don’t know that it’s possible to give a really definitive definition of “consciousness” when you come down to it. It’s a “we know it when we see it” phenomenon. But since it is a word that operates mainly in the realm of mental features and mental features are inherently private in at least one relevant sense here (I generally side with Wittgenstein that our mental words have very public criteria but that’s a different issue), it follows that fixing it with a satisfactory definition is going to be hard if not impossible. To what extent is talking about being something that it’s like to be helps is at least highly questionable.

    How this bears on the question of whether consciousness is an on/off phenomenon as you and some others seem to think or a graduated phenomenon with no clear line demarcating where the lights go out is another question. Insofar as I think of consciousness as awareness (in the sense offered above) I see awareness as a complex phenomenon and it waxes or wanes with the increase or diminution of that complexity.

    In replying to Stevan’s paper I’ve tried to make the case that we can explain what he calls feeling (which I think he sometimes mixes up by shifting between different referents, e.g., the feeling of migraines and the feeling that we understand a phrase we’ve read) as a subsystem or subsystems operating interactively within a larger computational system containing the others. Given that view, it’s not hard to see why an Alzheimer’s patient’s consciousness (or our own as we drift off to sleep or fall under the influence) might fade and eventually cease to be anything we would call conscious at all.

    If consciousness is just the operations of so many processes in an overarching process-based system then as the processes turn off the consciousness become progressively diminished. And vice versa.

    Obviously this is controversial because many prefer a different picture of what it means to be conscious. But I would suggest that this is the one that most accords with our actual experience of brains and minds in the world.


    P.GOFF: “[I argued that] assuming the truth of property dualism it is implausible to suppose that consciousness can disappear as we morph the brain into a kidney… [you have not] to my argument for that claim”

    I have to confess that I don’t know really know what it means to believe in property dualism. I just believe that feeling is inexplicable causally in the way all other properties are explicable.

    (Richard Brown had been suggesting in the thread on my paper that this meant I’m a “property dualist.” If it doesn’t then please just take my comment to be that an all/none notion of feeling — which I do believe is correct — is compatible with a sudden switch in the gradual process of morphing an unfeeling kidney into a feeling cerebellum, and that the change might well be a macroscopic organizational one rather than a molecular one. If that conclusion is irrelevant to your paper, I apologize for intervening.)

    P.GOFF: “My argument involves the intuition that the hypothesis that the fundamental psycho-physical laws of nature (remember the arguments is aimed at property dualists) specify down to minute detail how each fundamental particle in a brain containing trillions and trillions of them must be for consciousness to result is extremely theoretically vicious, and hence should be rejected if there is an alternative (which there is: panpsychism).”

    But are only molecular properties properties? Aren’t organizational properties properties too? A reverberating neural circuit is an organizational property, as well as an all-or-none-property. It’s even plausible that whether a circuit is inert or reverberating could be determined by a change in one molecule…

  46. Response to Philip Goff’s post on 2/28 at 11:48

    In the second paragraph of your response to me, you imagine the several forms of sensory consciousness getting progressively fainter, and remark that “I think it’s wrong to describe this as my having the property of having an inner life to a lesser degree. My inner life changes as the process goes on, but I have an inner life just as definitely towards the end of the process as I did at the start”.

    As I hope will soon be clear, there’s much I agree with here, but I think you’re not giving sufficient due to the remarks I put under (3) in the original, i.e., the part where I say some things about what I called “self-consciousness”. Those remarks recognize that there is more to being conscious than having what’s ordinarily included under sensory consciousness – most especially inner speech, but also feelings of ownership, feelings of strangeness or familiarity, etc., etc. When we imagine all forms of *sensory* consciousness going, we do not thereby imagine these other aspects (that I lumped under “self-consciousness”) also going.

    A good part of the motivation for the material under (3) was to use exactly this point to explain the *plausibility* of the binary character of consciousness, despite the actual continuities I was arguing for. And the motivation for most of the rest of (3) was to argue that these further aspects of consciousness too could come and go, consistently with a property dualist account.

    So, I agree that when you lose sensory consciousness you do not thereby cease to have an inner life, but not with the view that a property dualist can’t allow for daily cessations and re-gainings of an inner life (or its loss through becoming a pillar of salt).

    In your last paragraph, you say “for any neurophysiological state, it is surely a vague matter whether that state is instantiated”. That doesn’t seem “sure” at all to me; it doesn’t even seem true. Can you explain a bit how you’re thinking that leads you to say this?


    Response to Richard Brown

    Hi Richard

    Thank you for your comment.

    Here are my thoughts on some aspects of what you say.

    I’m afraid I’m wary of comments like ‘THEY ARE TRYING TO CALL YOUR ATTENTION TO SOMETHING THAT SHOULD BE OBVIOUS TO YOU’. The ‘something’ in question is not at all obvious to me as my initial analysis pointed out. On the contrary, it strikes me as confused and ultimately vacuous for the reasons I gave. So, if someone wants to change my mind, they need to give me a solid argument, and just telling me it ‘should be obvious’ sounds like an admonition not an argument.


    This is much the same claim Phil made at one point. My reply to him was along these lines:

    ‘Proved’? Where is the proof? And useful for whom? (Certainly not for me – and not for Hacker either, I gather). And useful for what, exactly? You say: for getting people (I assume you mean students largely?) ‘to focus on one of the properties that philosophers working on consciousness are interested in’. What property precisely? And does the proposition really focus attention on something important about consciousness, or just encourage students to think it does? As we know, people can be encouraged to believe all sorts of things that are not the case. That’s one of the reasons we have philosophy – to help them look carefully at what they’re told.


    I think I dealt earlier with the suggestion that the proposition is just a ‘convention’. More importantly here, it’s surely a worrying state of affairs when philosophers start saying that analysing something is not going to be much use. Surely, philosophy – ‘analytic’ or not – must always be about analysing. If at some point we start saying, ‘Well, let’s not look into this’, we’ve thrown in the towel haven’t we? We’re no longer even doing philosophy.

    As for the comparison with ‘kick the bucket’, that’s a metaphorical phrase (for dying, by the way, rather than ‘death’) and I assume you’re not suggesting that the ‘something it is like’ proposition is a metaphor. (And if one thought it worthwhile, there’s no reason why one should not analyse the phrase ‘kick the bucket’, like any other, as long as one kept in mind that it is a metaphor.)

    Final point: I’m now beginning to get the impression that you and perhaps others want to back away from the ‘something it is like’ proposition – to downplay it, and push it into the background. Drbillh tells me it’s just a ‘convention’; you say, in effect: ‘let’s not look at it too closely’; and Phil, if my memory serves me, also implied something similar.

    Let’s remember, however, that the proposition plays a major role in much of what’s written in the philosophy of consciousness. It’s invoked again and again as if it’s saying something important. Not infrequently, it’s used as an opening gambit, as in Phil’s case (and the two similar instances I quoted earlier were opening gambits by Chalmers and Helle.) Now, either the proposition has something important and substantial to say, or it hasn’t. If it has, then it should presumably stand up to robust analysis and we should be able to say what its important and substantial meaning is. If it hasn’t – if the proposition is a red herring as I’ve argued – just nudging it into the background hardly seems the right way to go. Shouldn’t we admit clearly and forthrightly that it is, and has always been, a philosophical dud?



    Response to Phil Goff

    Hi Phil

    Thanks for your further comment.

    If we don’t analyse a words in terms of one of its uses, how do we understand it at all? If someone asks me to clarify the meaning of a statement and I can’t illustrate the meanings of the word contained therein by their use in other contexts, how do I reply to the inquiry? Clearly, as I pointed out myself, the word ‘like’ in the proposition doesn’t mean what it means in ‘I like ice cream’ (that in itself is the beginnings of an analysis); but to say we can’t assign it a meaning in terms of ‘any other use of “like”’, to quote you, seems a very odd thing to say. Do you mean that it has no meaning at all? Philosophy would surely come to a standstill if we couldn’t explicate the meanings of words. Especially analytic philosophy, I would have thought.


    You’re taking my remark out of its context. I was making the obvious point that we can’t imagine any experience (the same goes for things) as not having qualities of any kind – qualities in the sense of attributes. An ‘attribute-free’ experience or thing is unimaginable. But that simple point has nothing to do with the Nagel argument or the ‘something it is like’ proposition. Neither – I assume! – is merely trying to establish that everything has attributes of some kind.


    The notion of ‘the introspectively apparent qualities of experience’ raises questions of its own, not least about the value of ‘introspection’ in learning anything important about human consciousness. But all that is beside the point as far as my argument is concerned. My argument is that the ‘something it is like’ proposition tells us nothing important – that it is merely an empty truism, as I put it. So, long before specific questions arise about ‘the introspectively apparent qualities of experience’, or whatever else, the proposition, I am saying, is just a philosophical red herring.


  49. Derek, I am not sure what to make of your comments. Surely you are conscious and you know that you are conscious. You also know that consciously seeing blue is different from consciously seeing green, both of which are different from consciously hearing a concerto and tasting mole sauce. Those experiences, at least when conscious, feel a certain way for me when I have them. Put differently there is something that it is like to be me having those experiences rather than nothing that it is like. When you or I look at a blue patch (say the background of this website) we usually have a conscious experience of blue. If I were to point my digital camera at this screen it would not have a conscious experience of blue. Put differently there would be nothing that it is like for the digital camera, and it would not be like seeing blue for the camera. Seeing this point is all that the “Nagel phrase” is marshaled for. It is just a shorthand way of saying “you know that conscious experience you have when…”

    And I do not think that it is a metaphor, I think that it is an idiom.

    But if what you are upset about is philosophers who try to draw substantive conclusions about consciousness from the phrase then I agree. That is a bad practice. Its usefulness does not lie in something profound that it reveals about consciousness but rather as a common sense starting point for identifying the phenomena that is to be discussed/analyzed.

  50. Hey Philip I like the collection you are included in called Consciousness and Its Place in Nature. I’m curious to hear what you thought of some of the replies to Galen Strawson’s target article. Let me start with this comment of yours: that general metaphysics may not help, that the unintelligibility of emergence of macroexperience has more to do with the “specific nature of experiential phenomena, rather than the general ontological categories they fall into.” This relates to why I brought up Whitehead’s metaphysics and Strawson does the “object/process/property/state/event cluster of distinctions.” Whitehead clearly picks processes and events as the fundamentals and notes that at any state/instant there is actually nothing (of any causal import or interest). Objects and substances would be enduring collections (societies) of these more fundamental drops of experience/process. And so, at least to me, he makes intelligible an answer to your worry of “how little subjects of experience can sum together to form big subjects of experience.” And even though he builds on work of James, Whitehead would not have liked the passage that you quote: “each remains the same feeling that it always was, shut in its own skin, windowless.” He insists that there is no reality at an abstract instant, separate from others but that the real things are ingressions, coming-togethers (but he was fond of the phrasing of the many coming together + 1 which is close to this idea of James here of the 101st feeling as a totally new fact). I think Rosenthal, though, would question this point based on this comment: “But the extended does arguably emerge from the unextended. Geometrical points, though unextended, nonetheless constitute lines, planes and solids; and that might be enough for us to at least keep a provisional open mind about the emergence of the experiential from the non-experiential” (121). Indeed, even though I appreciate Whitehead’s metaphysics where events and not points are the nature of reality, I do still wonder about the need to put consciousness or experience into the base actual entities. James clearly thinks it is needed (note 48): “consciousness, however small, is an illegitimate birth in any philosophy that starts without it.” But I’m not sure why it matters for evolution “to work smoothly” (in fact, it is not even so smooth is it?) and why a story about applying concepts (which seem to get more enriching and complex)to a 1st order experience wouldn’t be just as good as stories which insist on all being there at the beginning. One last inquiry is Strawson’s footnote 37 on Chalmers: “it is fabulously implausible to suppose that an atom-for-atom duplicate of an experiencing human being could be produced and not have experience.” I think he could have chosen words differently but I’m glad for this phrasing because it shows part of the problem. It is experience itself which is in question here — so it can’t be an atom-for-atom experiencer but functionally-like an experiencer which is the concern. It is still in question whether you can separate off experience from function. I’m curious whether Chalmers and Rosenthal would agree on the issue of consciousness and function. The latter seems close to Nietzsche at points actually, at least when it comes to how much we could do without being conscious. I wonder both if there is a function of consciousness but also whether functions result in consciousness. These two sometimes get combined but that’s why I am curious about the role of consciousness in a function like communicating or creating: how much feedback, awareness, reflection, etc. are needed in these aspects of the functions and how much of these aspects themselves are actually needed to execute these functions. More soon on fool’s consciousness and essence but thoughts welcome on consciousness, function, and essence. Here’s a quote from Lycan to get the ball rolling (against the Whitehead scheme though): “Kripke’s essentialism is hopeless regarding events (since particular events have no clearly essential features).”

  51. Comment on the “What it’s like” thread (mostly Derek Allen, but several others too):

    I know I’m coming to this late, but even so I think there’s something that should be in this discussion that hasn’t appeared yet. I’ll start with some examples, then make some claims.


    (A) Turow is a fine writer. If you read his _Presumed Innocent_ you’ll get a good sense of what it would be like to be a defendant in a murder trial.

    (B) “I cannot tell how it happened that death entered into me, but I can say what it was like to feel him there”. [Quotation from a novel, referenced below.]

    (C) It is difficult to imagine what it would be like to have reactive dissociation, or to be a unilateral neglect patient.


    (1) These examples are non-philosophical claims that are perfectly intelligible to native speakers of English. (Well, maybe (B) presents some problems, but not because of the ‘what it was like’ construction.)

    (2) There are plenty of unproblematic phrases we all understand but would be hard put to explain. E.g. “If it were true that P, then Q would also hold” has a singular subject (‘it’) followed by a plural verb (‘were’). Why? How did that come into English? I don’t know, but I don’t need to to understand what’s said, which is perfectly clear as it stands.

    (3) In view of (1) and (2), the “like” that occurs in these examples needs no analysis; in particular it does not need to mean the same as “like” in other contexts and does not need to be explainable in terms of what ‘like’ in other contexts means. Lexicographers just have to have at least three entries (i.e., one for the meaning in the examples, in addition to ‘feeling like doing some action’ and the ‘like’ of similarity).

    (4) The examples can be paraphrased by using phrasing constructed from the idea of describing one’s experience. (“Mr. Sabich, can you describe the experience of being tried for murder?” “The experience of having death enter into one can be described as follows . . . .” “I’ve listened to people who’ve experienced reactive dissociation describe their experience and still find it impossible to imagine myself being in that condition.”

    (5) When Nagel wondered “What is it like to be a bat?” his wording seemed appropriate because it used the same idiom that occurs in examples (A) – (C).

    (6) The same resonance with the examples (A) – (C) explains why Nagel’s phrase was found appealing and was taken up by so many.

    (7) “A subject is phenomenally conscious when there is something it is like to be that subject” is informative because the examples (A) – (C) direct our attention to the qualities of our experiences, and thus the proposition ties the phrase “phenomenal consciousness” to qualities of our experiences.

    (8) The statement in (7) is, however, only minimally informative. (You have to already understand something tantamount to “description of experience”, and thus something tantamount ot “experience”.) That’s why so many philosophers (including me) have relied much more heavily on use of examples (e.g., pain, rage, afterimages) to (at least attempt to) give a kind of ostensive definition of what they mean by “phenomenal consciousness”.

    (9) Since they’re trained in logic, it comes naturally to philosophers to make generalizations, e.g., “Being a subject is being something such that there is something it is like to be it” or “There being something it is like to be you is the most important fact about you”. Statements like these are at some distance from the examples given. Possibly, this partially accounts for why examples like the ones I’ve given have not been featured in this discussion. Another part of the explanation may be excessive use. Philosophers who see this phrase all the time have overlearned what it means. They do not have to bring to mind the illustrations that gave the phrase its resonance to begin with. They can come to regard it as an empty term of art for the character of experience. I’m saying that’s not right – the phrase still does carry the resonance with examples like (A) – (C) – but it’s understandable how that could be overlooked, as it seems to have been in this exchange so far.

    [ The quotation in (B) is from p. 8 of Hans Keilson, _The Death of the Adversary_, Ivo Jarosy, tr. (New York: Farrar, Strauss & Giroux, 2010). Originally published in German in 1959.) It’s sheer coincidence that I happened to start reading this book this week. ]


    Response to Richard Brown

    Hi Richard

    Thank you for your further comments. I think, however, you gloss over some key issues.


    As an everyday statement this is fine. But for purposes of the philosophy of consciousness it surely won’t do. I know I am six feet tall, but I can produce a tape measure to show that, and I can explain quite adequately what ‘feet’ means if asked. But what can I do if someone asks me (the philosophical question): what do you mean by ‘conscious’? No matter how many times I repeat phrases of the kind you mention (‘seeing blue is different from consciously seeing green’ etc) they will tell me – quite rightly – that I am not explaining, I am just giving instances, and begging the question.

    The same goes for your following comment:

    My interlocutor might legitimately say (for example): ‘Oh, so “conscious” means the image of blue is not recorded digitally’, or even, ‘Oh, so “conscious” means something you can have and a camera can’t’. In other words, these are only instances again. They are all perfectly correct usages of the term ‘conscious’ as we know, but there can be a big gap between a correct use of a word and an explanation of its meaning – and the term ‘conscious’ is a conspicuous example of that.

    You also say that one is not meant to ‘DRAW SUBSTANTIVE CONCLUSIONS ABOUT CONSCIOUSNESS FROM THE PHRASE’ and that it is only ‘A COMMON SENSE STARTING POINT’ for a discussion of consciousness. But if the phrase is not intended to say something substantive about consciousness, why do I keep reading statements like:



    And then of course Philip’s own similar statement, which begins ‘BY ‘CONSCIOUSNESS’ I MEAN…’

    (There are many more examples of this kind. I only give these because I happened to be able to locate them easily.)

    To me these certainly sound as if they are trying to say something substantive. Moreover, even if they were, as you say, just a ‘common sense starting point’, they must, presumably, be understood as telling us something about consciousness. They can’t be telling us nothing at all – or, for example, about growing carrots or the internal combustion engine. You yourself say the ‘“Nagel phrase” is … just a shorthand way of saying “you know that conscious experience you have when…”’ If the listener is to make any sense of this claim, s/he must surely take it as a description – even if only a rough one – of the state of consciousness. If this is not the case, we can presumably only read the statement as a kind of ‘convention’, as Bill Robinson (drbillh?) sought to argue. And I pointed out the problems with that.

    The fact that the ‘Nagel’ phrase has ‘caught on’, as Phil put it, is worth pondering, I think. As my initial analysis showed, the term ‘like’ is being used ambiguously. Is it a like of comparison or a like of ‘feeling like’? This ambiguity is only apparent once one sits down and examines the proposition (as Hacker has also I think) and a hidden ambiguity like that can be a real trap in philosophy. If one doesn’t detect it, the mind, I think, tries to assign a sense to the statement anyway – i.e. it tries its best to make it clear and meaningful. The word ‘consciousness’ comes to the rescue here. Being itself a word whose meaning is hard to pin down, as we know, we are happy enough to connect it to a phrase whose meaning is hard to pin down (due to the hidden ambiguity). I think this is part of the reason why the phrase has met with so little resistance – why it has caught on.

    This is getting too long. Just a word on an unrelated matter – the feedback you requested. I think the conference has been great, and the discussion stimulating. Only one tiny thing I would like to see – the facility to use various things like italics, bold etc in one’s contributions. Capitals alone tend to be a bit limiting.


    Derek Allan


    Response to William Robinson

    Hi Bill

    I confess I haven’t followed your reasoning in detail (my apologies for that) but I just wanted to make the point that none of the three example sentences you give raises the problem raised by the ‘Nagel’ proposition.

    Each of your sentences is a perfectly comprehensible ‘like’ of comparison. In A the comparison is between an experience as described in a book and the experience of being a defendant in a murder trial. In B the comparison is between the experience of ‘death entering me’ and what ‘I can say’ to describe that experience. In C the comparison is between having ‘a reactive dissociation’ and a difficult-to-provide description of same (a hyperbolic way of speaking but a comparison nonetheless).

    Now the ‘like’ in the ‘Nagel’ proposition (to use that shorthand) is not the same as that at all. Firstly, it is ambiguous: one can’t be certain whether it’s a ‘like’ of comparison or one of ‘feeling like’ (doing). Second, neither interpretation works anyway – we end up with vacuous results in both cases. (I won’t run through my analysis again; it’s in my first comment – though in a rather more convoluted way than I would have preferred because I had to adapt my remarks to Phil’s specific (rabbit) example.)

    As I pointed out in my comment, the very ambiguity of the thing has been a trap. Here is what I said:

    ‘Why is the Nagel thing persisted with so often? My guess is that it’s precisely because the word ‘like’ is being used ambiguously … People, I suspect, have allowed themselves to be bamboozled by that (and by the convoluted syntax) and have assumed that something profound is going on. But as we can see, it only takes a little analysis to show that nothing remotely profound is going on and that, under either interpretation of ‘like’ (and is there any other?), the proposition is at best an empty truism.’

    Or as I put it later, perhaps bit more clearly:

    ‘… a hidden ambiguity like that can be a real trap in philosophy. If one doesn’t detect it, the mind, I think, tries to assign a sense to the statement anyway – i.e. it tries its best to make it clear and meaningful. The word ‘consciousness’ comes to the rescue here. Being itself a word whose meaning is hard to pin down, as we know, we are happy enough to connect it to a phrase whose meaning is hard to pin down (due to the hidden ambiguity). I think this is part of the reason why the phrase has met with so little resistance – why it has caught on.’

    My strong recommendation to the philosophy of consciousness is that it drop the ‘Nagel’ proposition completely.


  54. Reply to Bryony Pierce

    Hi Bryony

    My apologies for not replying to you sooner.

    There’s not much I can really say, however, because you seem to be approaching the problem from a different angle from the one I’ve been adopting.

    Essentially, I’ve been doing a bit of traditional philosophical analysis (long overdue, I would argue) of the ‘something it is like’ proposition just as it stands – i.e. trying to see what it means, if anything. You have probably noticed that my prognosis is negative.

    You seem to be addressing issues more connected with the way the brain/mind operates. That’s a bit outside my remit at the moment, so I’m happy to leave those matters to others.



    My only objection to the usage is that it’s more cumbersome than other, simpler ways of making the same point, while adding no more to that point than the other terms do. Because no term or terms seem to be entirely adequate it seems like we need to find something definitive enough to fill THAT bill. But “what it is like to be X” locution isn’t anymore definitive, while the complexity of the formulation seems to imply something profound has occurred with its introduction.

    There’s no reason people should not go on using it but I think, with Derek Allan, that we ought to see it for what it is and not be prompted by its appearance in a discussion to think something new and profound about consciousness has been offered. It certainly doesn’t help much to give it definitional status either.

    Can “consciousness” be defined in any precise and relatively finite way? Probably not. But that’s why it’s such a fecund field for philosophy, isn’t it?

  56. Thinking more on essence and events, Lycan and Whitehead would agree that there is no essence to a momentary pulse or instant (although perhaps Lycan would want events to last longer or have endurance, like a festival). Societies, for Whitehead, are what continue, what change; they have a history and they endure. Actual occasions just come and go, become and perish. The essential shared character or defining characteristic or point (no pun:) is the reason for the likeness. A society is its own reason, it is self-sustaining, it imposes on the other members the conditions which lead to the likeness. I bring this up because I wonder about phrases like a “pain event.” Is pain something that is more like a pulse or actual occasion or does it have an essence and history expressed by changing reactions to changing circumstances? I take it the argument for why the actual occasions have no essence is that they are the fundamentals; they have no internal structure to examine. But part of the issue is whether there are kinds involved here. Is pain or consciousness a natural kind with an essential function or make-up? Is the feel of pain essential to its being called so or the firing of c-fibers?

  57. Really enjoyed the debate on ‘what it’s likeness’ talk prompted by Derek Allan. Derek is of course right that much of analytic philosophy of consciousness is premised on the intelligibility of this phrase; in general, it’s the agreed starting point for both physicalists and anti-physicalists. However, in my view what has emerged from the debate is that Derek and Stuart W. Mirsky seem to think we want more out of the phrase than in fact we do. It’s not supposed to ‘tell us anything’ about consciousness. It merely indicates the sense of ‘consciousness’ we’re interested in, namely the qualities of experience which are accessible to introspection, as opposed to, say, some explicitly cognitive characterisation of consciousness. It involves the use of ‘like’ which indicates concern with the qualities of something. Now you might doubt that there are qualities of experience apparent to introspection, you might even doubt that there is such a thing as introspection, but we should be having these arguments rather than worrying about the meaning of the Nagalian phrase.

    Several people, including Steve Harnad most recently, want to blunt my argument by associating the disappearance of consciousness with changes at a higher than fundamental level. I agree that the difficulty isn’t so apparent when we focus on suitably high levels, but I think my opponents are obliged to confront the thought experiment as I set it up. Suppose you’re sitting next to God as he’s changing Lot’s wife into a pillar of salt minute adjustment of a particle at a time. Suppose each change you ask God: ‘Is there still consciousness?’. Two possibilities. Firstly, after God makes a certain one of these tiny changes, he declares ‘No, consciousness is gone now’. This first possibility is implausible. It’s theoretically implausible that the laws which generate consciousness from the physical facts specify the conditions necessary and sufficient for consciousness down to the last detail at the fundamental level. Second possibility: After each change God says, ‘There’s still consciousness’. In this case, the pillar of salt is consciousness. Panpsychism is the only theoretically plausible form of realism about consciousness.

    I’m quite serious about this; I don’t take this to be just a fun philosophical paradox. Let me give a few general methodological remarks to try to persuade you to take what I’m saying seriously. A good working principle in scientific enquiry is: Side with theoretically elegance, not with common sense. Two cases in point: Special relativity versus the Lorenzian theory which preceded it. The Bohm interpretation of quantum mechanics versus the Copenhagan interpretation. In both cases we have two empirically equivalent theories, the former of which saves common sense at the expense of theoretical elegance. I suggest the rational thing to do is to go with theoretical elegance. Why on earth take what the average Joe thinks as a guide to the nature of reality as it is in and of itself?

    Now you may think that there is a crucial difference between the two scientific examples I give and my argument for panpsychism: the scientific cases involve data as well as theoretical virtues in the theory selection process. But my argument for panpsychism (the broader argument of my book which involves the conditional claim of this paper plus a massive argument against physicalism) does involve data. Each of us knows with utter certainty that there is consciousness. There is data which is more epistemologically significant that empirical data, given that it’s not subject to doubt. So we know that we have to incorporate consciousness into the world, and we are rationally obliged to do it in a theoretically elegant manner. The only way to do this is to be a panpsychist. I am quite seriously convinced that there is overwhelming motivation to embrace panpsychism, at least once one rejects standard forms of physicalism.

    I feel like there is much more to be said between Bill and I, and it would be be good to discuss this more at some point. I take it that he needs to make sense of consciousness ‘fading out’ in order to smooth over the potential discontinuity (I thank him for making me think in terms of continuity in nature). But it seems to me that the conscious representation of sound getting quieter, or of colours getting fainter, does not constitute a gradual lessening in degree of the property of having an inner life. In his last reply to me he points out that consciousness has cognitive as well as sensory aspects. But I don’t see how this helps; it seems even more difficult to make sense of these cognitive aspects fading out as it does the sensory aspects. So at this point I am still not getting a grip on how consciousness can fade out, and hence still not seeing a theoretically plausible way of making sense of the disappearance of consciousness.

    (This debate reminds me a bit of Kant’s response to the argument for the immortality of the soul – can’t remember the name of the guy whose argument it was – which goes roughly: the soul is indivisible and so can’t be destroyed. Kant’s response if I remember rightly was something like: the soul can’t divide, but it might fade away. This comparison may count against me).
    (To back up my claim that neurological states are vague: I take it that our characterisation of such states involves functional elements that will not be precise down to the level of physics. This will introduce some borderline cases; it will not be precise down to the last particle movement when you have and don’t have a certain neurological state).

    Andy Snyder:
    Very interesting stuff that I need to think about more. I may have to rethink my earlier contention that the combination problem had nothing to do with issues of general metaphysics. I’m afraid I’m no expert on Whitehead. Been meaning to look into him for a long time.

    I’m very grateful to both Bill and Jonathan and to all the people who have taken the time to give me comments on this. It’s been very helpful for developing the paper.

Comments are closed.