Zombie Spouses

Would you be worse-off with respect to well-being if your spouse didn’t really love you, but only seemed to? Lots of people think you would be and are therefore persuaded that well-being/welfare must consist of something more than pleasure, happiness, or other mental states.

For example, Shelly Kagan’s “Deceived Businessman” believes he has a faithful, loving spouse but has an adulterous spouse who secretly despises him. Kagan argues that since the businessman is blissfully unaware of his spouse’s deceptions, mental state theories of well-being cannot distinguish between the deceived businessman and his doppelganger whose identical beliefs about his circumstances are true. Kagan concludes, “In thinking about this man’s life, it is difficult to believe that it is all a life could be, that this life has gone about as well as a life could go. Yet this seems to be the very conclusion mental state theories must reach! … So mental state theories must be wrong” (see Kagan’s Normative Ethics, p. 35).

OK, how about this: In 25%-Zombie world, every fourth person is a zombie, and every fourth spouse is a zombie. The non-zombie denizens of 25%-Zombie world know that every fourth person/spouse is a zombie who behaves exactly like (and is indistinguishable from) a normal person but feels nothing and doesn’t love anyone. (Never mind how they know this. It’s true, they believe it, et cetera.) I think the non-zombies would find it disconcerting that there were so many zombies around, but most would eventually get over it, marry someone, and assume that some other sucker had the zombie spouse. Many of the non-zombies would be wrong about this, but neither they (the non-zombies with zombie spouses) nor anyone else would ever know which spouses were zombies.

I don’t think the non-zombies with zombie spouses would, on average, be any worse off with respect to well-being than non-zombies with non-zombie spouses. (Of course, they’d all be worse off than average spouses in our world, since we don’t have to worry about whether our spouses might be zombies, but this worse-off-ness would apply to all married non-zombies in 25%-Zombie world, whether or not they were married to zombies.) What do you think? If you agree with me, do you think this constitutes a counter-example to thought experiments like Kagan’s Deceived Businessman? I sure do, but you might have guessed that.

I think that in our world, disloyal or unloving spouses usually are distinguishable from faithful, loving spouses because the former don’t feel and display the same respect and love for their spouses as do the latter. So deceived spouses typically suffer tangible, discernible harms to well-being that undeceived spouses typically do not suffer. However, if unloving spouses were completely indistinguishable from loving ones (e.g., because they were zombies or superb deceivers who never came out of character), I don’t think the unloved spouses would suffer any harms to well-being. Their situation might “look worse” to epistemically privileged readers of thought experiments since we know, e.g., that Donna’s husband is unfaithful or that Jack’s wife is a zombie, but that strikes me as a purely aesthetic matter — something that doesn’t look quite right from “the view from nowhere” — rather than anything to do with well-being.


66 Responses to Zombie Spouses

  1. Yikes, I think the zombie’s spouse is much worse off.

    Though it’s worth noting that one could deny this without also denying Kagan’s point. For one might think that being loved by a genuinely conscious person is not one of the objective goods in life, but that being intentionally deceived or betrayed by such a person is an objective bad. (I’m not sure why anyone would be drawn to this combination of views, but it’s at least possible.)

  2. On the “aesthetic” point: does the badness really seem so impersonal to you? I would have thought that, e.g., it was precisely for Donna’s sake that we have reason to prefer that her husband not cheat.

  3. Shen-yi Liao says:

    I don’t have strong or clear intuitions either way. I guess that means, at least, I don’t have the intuitions necessary for Kagan to make his case. But it also does not clearly seem that the zombie spouses are no worse in your scenario than non-zombie spouses.

    I have a side question, though: why must mental state theories of well-being be committed to the relevant mental states being internal? As our colleague Sven Nyholm shows, external factors affects judgments of happiness. One explanation (probably not the one that he endorses) for this is that happiness is an external mental state, like what Williamson thinks about knowledge. Similar things might be said for other relevant mental states.

  4. Jonathan Livengood says:

    I don’t think spouses of zombies are any worse off at all. But then, I don’t think zombies are physically realizable. One problem here is the stipulation that you cannot (I suppose this means no possibility — logical, metaphysical, physical?) distinguish zombies from non-zombies. (Incidentally, I think it is unfair to simply slide under the rug how the people in the possible world you imagine come to know that a quarter of their population are zombies, which is just to point out that under the indistinguishability condition they cannot legitimately acquire this knowledge.)

    I think that our sympathy for the spouses of undiscovered cheaters comes along with a deep conviction that cheaters are eventually uncovered, and when they are, their partners are hurt (probably something that cannot be really shut off — sort of like taboos). Also, with the cheating spouse, we can readily understand a counterfactual of the form, had the spouse discovered the cheating, he/she would have been deeply hurt. Whereas, I for one have no grip on the counterfactual, had the spouse discovered the zombification, he/she would have been deeply hurt (since by stipulation such discovery is impossible).

    A similar story is probably available from the perspective of harm. We can ask, “Was agent x harmed by action p of agent y?” I think it makes sense to talk about undiscovered harms. But in those cases, it makes sense — at least in part — because the harms are discoverable. Whether an always undiscovered harm is an actual harm is, I think, a very difficult question. Moreover, it’s not one with any practical teeth, as far as I can see.

  5. Steve Nayak-Young says:

    Richard, thanks for your comments.

    Re the “Yikes,” I should mention that these are the “no phenomenological experiences” zombies, not the brain-eating variety. :-)

    I agree that we might think that since zombies don’t deceive anyone (at least not deliberately), they don’t harm us in the way that a deceptive spouse does. But what is that way, exactly? The mere fact of deception isn’t sufficient to constitute a harm, so I suspect something else must be necessary, e.g., something that one can notice. Since we can’t notice any difference between loving spouses and perfectly deceptive spouses or zombies, I’m inclined to think they’re all equal with respect to well-being.

    This take on the matter trades heavily on the intuition that well-being tracks harms and benefits, which undergirds most of what I think and say about well-being.

    Re Donna, we have no reason to worry about her, because it’s already been stipulated that she’ll never find out that her husband is cheating, he’ll never behave any differently from a loving spouse, etc. We can’t help but dislike the idea of cheating, but I like what Jonathan Livengood has to say about this in his comment above. Cheating usually causes harms, so we always think it’s bad, but mental state theories can track the usually-caused harms.

    The only way Kagan and others can say that mental state theories can’t identify the harm is by making it so that the *subject* can’t identify the harm. My theory-soaked intuition is that if we stipulate that a subject can’t and won’t ever notice any difference as a result of some putative harm, there’s no harm at all.

  6. Steve Nayak-Young says:

    Thanks, Sam. It’s not so much that zombie spouses are no worse than non-zombies — for one thing, they’re zombies! — but that having a zombie spouse is no worse for one’s well-being. Again, I think that well-being can’t be worsened without some sort of harm, and I don’t think we should bother straining to come up with a concept of “unnoticeable harms.”

    Of course, if one holds an “objective list” theory of well-being, one will disagree with my focus on harms, but I’m tempted to think that “objective list” types just associate a different concept with “well-being.”

    Re your side question, I think Williamson’s view is crazy. I’d much rather say that knowledge isn’t a mental state than that mental states can somehow be “external.” So I don’t know whether all mental state theories are committed to mental states’ being internal, but I don’t have the time of day for “external” mental states. I think the more usual strategy (I believe Kagan takes this line re happiness) is to deny that happiness is a mental state and hold that it is one or more mental states plus various external facts. That’s not how I think about happiness, but I know lots of people who do.

  7. Steve Nayak-Young says:

    Thanks, Jonathan. I agree with and like most of what you wrote.

    Re zombies, I’ve never quite understood the ins and outs of zombie biology, psychology (if any), etc., so I’m on shaky ground here. I certainly didn’t mean to imply that there’s no way for anyone to distinguish between zombies and non-zombies. It’s only important, for my purposes, that the non-zombies can’t do it. Let’s say that the reigning deity of 25%-Zombie World can pick out zombies no problem, and she/he/it is the one who tipped off the non-zombies to the zombies among them in the first place, perhaps via some stone tablets or reliable prophets.

    As for the rest of your post, I couldn’t have said it better myself, unless I were smarter and less lazy.

  8. Hi Steve, I agree that wellbeing concerns “harms and benefits”. More precisely, we can neutrally analyze ‘wellbeing’ as concerning what is desirable for the sake of the individual in question. So the substantive question is then what sorts of things we may have reason to prefer for the sakes of individuals.

    Most people, faced with cases like Nozick’s experience machine, find that they prefer a life of genuine struggles, relationships, and achievements, to a (subjectively) happier life of delusion. We care about many such objective matters in our own lives (for our own sakes), and we also care about such objective matters in the lives of our loved ones (for their own sakes).

    The hedonist seems committed to the claim that these ordinary preferences are irrational. (Perhaps they could hold that we have other kinds of reasons — perhaps aesthetic reasons, as you proposed — to prefer these states of affairs. But they must deny, what strikes me as overwhelming plausible, that we have reason to hope for Jack’s sake that his life contains meaningful human connections rather than the mere appearance thereof.)

    I’d be very surprised if your intuition in the zombie case was shared by anyone who wasn’t already a committed hedonist. Most people, I expect, think that it matters a lot to genuinely share themselves with another person. They want a genuine relationship, and not just the (even reliable) appearance of such. The thought that their spouse might really (though undetectably) be a robot or zombie would seem horrific. Yet it seems you must disagree with this judgment. This state of affairs that Jack so worries about would not, according to hedonism, be bad for him at all. Instead, hedonism implies, it is only the state of affairs of Jack’s discovering that his wife is a zombie that is really horrific (or undesirable for Jack’s sake).

    Given these incredible implications, I’ve always been rather baffled as to why anyone would be drawn to hedonism. Can you explain the attraction?

  9. Shen-yi Liao says:

    Richard– You seem to rest a lot on intuitions of these cases. But it seems to me that intuitions of these cases are quite fragile. As I said above, I don’t have clear or strong intuitions both in the Donna case or Steve’s zombie case. As Steve will probably tell you, nor is empirical evidence clear that “Most people … think that it matters a lot to genuinely share themselves with another person.” (my emphasis) — as the inverted experience machine cases suggest http://peasoup.typepad.com/peasoup/2009/02/the-inverted-experience-machine.html

  10. Jonathan Livengood says:

    Well, then, be surprised Richard. I am no hedonist, but aside from qualms about the existence of zombies, I share Steve’s intuitions. That is, I completely agree with Steve that “if unloving spouses were completely indistinguishable from loving ones (e.g., because they were zombies or superb deceivers who never came out of character), I don’t think the unloved spouses would suffer any harms to well-being.”

    I also agree with him that in most typical cases, cheating is not only in-principle detectable but also often actually detected, even if not correctly attributed to the cheating (e.g. less time spent with the spouse, less money spent on behalf of the spouse by the cheater, the cheater also hogs all the covers at night, etc.).

    Anyway, I suspect that the aesthetic point–what makes one feel sorry for a duped and cheated-on spouse, for example–is the prediction (a pretty reliable one) that the actual relationship will end badly.

    I once had a friend who married a woman he was not suited to at all. It seemed that everyone but my friend knew this to be the case. At his wedding, when he was presumably very happy, I told some of my other friends that I pitied him, and they agreed. Three years later, he was divorced. Given that my assessment of my friend’s relationship was sound, pity would have been the correct reaction *even if their marriage had been a long and happy one*. Why? Because the risk of failure (at least at the time) was high.

    What most people see that is “off” in cases of undetected cheating is some sort of instability in the relationship–something that is likely to give way at the slightest touch. The increased risk of actual harm is real enough, even if in the particular case at hand, the harm does not occur. That is what makes the duped person worse off, and that is what makes a counterfactual analysis here conceptually important.

    Maybe I’m missing something, but I don’t see how any of what I’ve said here commits me to hedonism. Nor do I see why Steve needs to be committed to it, though he might well be. (Sorry if I’m stealing some of your thunder, Steve!)

  11. I am surprised! Do you mean the same thing as I do by ‘harm’, namely, that which we have reason to dis-prefer for the individual’s sake? I just have trouble wrapping my head around the mindset of people who don’t think that relationships, etc., actually matter.

  12. Steve Nayak-Young says:

    Hi Richard,

    I’m about as fond of desire-based theories as I am of “objective list” theories, but that’s no reason not to let them in on the zombie fun. However, I’m not prepared (as a matter of will or ability) to unpack “what is desirable” and “what we have reason to prefer,” not to mention “interests” and “sakes,” on a sleepy Sunday morning.

    However, I have just had what I hope are some interesting thoughts about preferences. Namely, there is a sense in which we’d think it odd to say that you prefer something you haven’t experienced. For example, if you had never tried any ice cream at all, it would be odd for you to say that you preferred chocolate to vanilla. Instead, you might say that you *would* prefer chocolate to vanilla, but this would be a prediction about how your preferences would/will look if/when you’ve had a chance to try both kinds of ice cream. You might turn out to be wrong, and find that even though you prefer chocolate in all its other guises, you prefer vanilla ice cream to chocolate.

    Now, imagine that you agree to engage in a taste test of Soda A versus Soda B. You try both sodas, and you do your best to discern a difference and state a preference, but you simply can’t. Soda A tastes exactly identical to Soda B, and you couldn’t truthfully say that you prefer either. Then the generic covers are removed, and the testers reveal that Soda A is actually Real Soda, while Soda B is actually Zombie Soda.

    “Yikes!” you declare. “Don’t give me any more of that nasty Zombie Soda! I strongly prefer Real Soda!” But you don’t. You don’t prefer either, because you’ve tried both, and you couldn’t discern any difference between them. It’s certainly true that you *want* Real Soda, now that you know it’s called “Real Soda” and the alternative is called “Zombie Soda,” but to say that you prefer or would prefer one over the other rings false, at least in the sense of “prefer” that I’ve been exploiting.

    Instead, you just like the sound of one name better than the other, and once someone tips you off to the names, you want the prettier one. But that strikes me as aesthetic or perhaps evaluative about appearances, appearances which (by stipulation) aren’t available to the thought-experiment denizens who are “taste-testing” zombie spouses versus real spouses. The more relevant evaluation — in my view — is the one you attempted to perform in the blind taste test, and I’m inclined to say that if you can’t discern any difference there, we should forget about the names and treat the two sodas as well-being neutral.

    I presume that’s all wonderfully clear, but please let me know if any of the subtleties of my soda-spouse analogy are eluding you. :-)

    Re “hedonism,” I guess I’m not one, because I don’t think there is any such thing or property as value, so I’m not a hedonist about value, and I think there are various things or properties we could call “well-being” without linguistic or conceptual perversity, but none of them is reducible to pleasure or happiness. I’m probably a mental-state theorist of some sort or other, but I’m not willing to tie myself down to just one category of mental states. As for what might draw me to such a view, the attraction lies in the fact that there clearly are such things as mental states, and these, unlike anything else, come in unmistakable flavors: good, bad, and neutral.

  13. Hi Steve, I’m using ‘hedonism’ as shorthand for [internal] ‘mental state theories’. So if you think that people can’t be harmed by what they don’t (and never will) know, then you’re a ‘hedonist’ in my sense.

    Now, note that there also clearly exist such things as desires, which also come in flavours: fulfilled (good) and thwarted (bad), depending on how the external world matches up to them. So I’m not seeing any reason here to favour hedonism. Furthermore, it’s clear that people can (and often do) care about more than just their own internal states, even for their own sakes. So that seems a powerful reason to favour a broader view of wellbeing. (Unless we’re talking past each other. If you don’t mean your talk of ‘wellbeing’ to have any implications for what people have reason to desire, then I don’t know what you’re talking about, or why it would matter.)

    Your Soda/Spouse analogy seems dodgy. First, you seem to be conflating preferences (in my broad sense) with hedonic likings, which in practice is what people tend to be reporting when they say stuff like “I prefer chocolate to vanilla”. But once we distinguish these, it’s perfectly conceivable that someone might desire not to have one of two equally tasty drinks. (Suppose one of them is derived from processed horse piss, and they find the thought of this repulsive.) If you trick them, say by swapping the labels, then the person hasn’t gotten (all of) what they wanted; they’ve merely been fooled into thinking that they have.

    Secondly, you’re at risk of conflating desires about counterfactuals with the desires that would be held in the counterfactual scenario. If Zombie Soda is derived from horse piss, and I have a standing desire to not drink anything so derived, then it’s already the case (though I don’t realize it) that drinking the zombie soda will thwart my actual desires. The resulting state of affairs is not, to my mind, a desirable one. (As a general rule, you can’t always identify an agent’s all-things-considered preference by asking them which of two outcomes, only partially described, they would prefer. After all, their preference might turn on some feature of the situation that was left out of the description. But that’s no reason to deny that they may already have a standing desire that counts one way or the other.)

  14. Steve Nayak-Young says:

    Hi Richard,

    “there also clearly exist such things as desires, which also come in flavours: fulfilled (good) and thwarted (bad), depending on how the external world matches up to them.”

    It’s not the same. I certainly desire that my wife is not a zombie or perfect deceiver, but if that desire is unfulfilled, I’m not getting the flavor. As I’m fond of saying, desires aren’t fulfilled or frustrated, *people* are. In other words, there’s no evaluative flavor without an evaluator, and evaluators can only evaluate what they’re aware of. If I think my desire is fulfilled, it’s gonna taste fulfilled, whether it is or not. Accordingly, I might guess that I wouldn’t like the zombie flavor, but I’d be wrong, because it’s just as good as the real flavor.

    “Furthermore, it’s clear that people can (and often do) care about more than just their own internal states, even for their own sakes.”

    Yes, they do, but I don’t think that constitutes “a powerful reason to favour a broader view of wellbeing.” I think that caring about things is part of being evaluative creatures with affective responses to our environments. We assume that we are pretty good at knowing a fair amount about what our environments actually comprise, but if one is systematically, incorrigibly wrong about something — e.g., if I’m a brain in a vat, if my wife is a zombie, or if I believe that Odin loves me and there isn’t any Odin — one’s well-being seems unaffected. Caring about things feels wonderful, whether or not they’re real.

    Re horse piss, I’m going to go out on a limb and assume that it’s bad for you. It might not be much worse for you than some soda, but I’d guess it might make one feel sick at some point. That’s a hedonic reason not to drink horse piss. Even if it didn’t make one feel sick or detract from one’s health in any way, one would be very angry to learn that one had been fed horse piss. So that’s a hedonic reason not to drink it if it’s possible to learn about the horse piss. As for your standing desires, so long as they are fulfilled as far as you know, and you suffer no ill effects, I don’t see how they have the thwarted flavor in any way that will affect your well-being. So perhaps we are talking past one another.

  15. It’s not the same… If I think my desire is fulfilled, it’s gonna taste fulfilled, whether it is or not.

    That’s true enough, but I don’t see the relevance. If I desire that P, then what I want to be the case is that P, not that I believe that P. (One might also desire the latter, but that would be a very different thing to want.)

    Of course, you may just insist that subjective experiences are all that matter, but that isn’t any kind of argument for why anyone else should find that view appealing. What I’ve pointed out is that there’s an equally naturalistic, non-mysterious alternative to hedonism, since whether or not the world is how we want it to be is another candidate for evaluation. The desire theorist may agree that there’s “no value without valuers”. But they may then point out that people value (desire) a whole lot more than is recognized by the hedonist.

    Indeed, as I recently pointed out to John Bronsteen, hedonism is really a kind of ‘objective’ or ‘paternalist’ view, insofar as it overrides evaluator’s own judgments of what is best for them. It insists that all that matters is having experiences that we like, even if we would strongly prefer to achieve some goal in actual fact than to merely believe ourselves to have done so.

    perhaps we are talking past one another.

    Perhaps. It’s hard to tell since you haven’t said anything about the implications of your claims. You might just stipulate that ‘wellbeing’, as you use the term, concerns the quality of one’s felt experiences, in which case you haven’t said anything substantive at all with which I might disagree. I certainly don’t mean to argue that secretly thwarted desires have a subjectively discernible experiential “flavour” to them, if that’s all that you mean to be ruling out!

    Alternatively, you might share my understanding of ‘wellbeing’ as concerned with what’s desirable for an individual’s sake (and hence what decision would be rational in certain forced choice situations), in which case I strongly disagree with you, and have argued that your position has crazy — radically counterintuitive — consequences.

  16. Steve Nayak-Young says:

    Talking past or not, I’m enjoying it!

    I’m certainly not insisting that subjective experiences are all that matter, nor that hedonic likings matter. To be clear, I don’t think that anything matters or is valuable or is desirable sans phrase. Things matter to us (i.e., we care about things), we value things, and we desire things. Of course, some or all of the “things” we value might be imaginary, but that wouldn’t make us stop valuing them or make them matter to us any less.

    For example, it matters very much to Eric that Odin loves him; Eric deeply values Odin’s love; and Eric wants Odin to love him forever. Eric wants it to be the case that Odin loves him forever, not just that Eric believes it. Since (I’ll venture) there is no Odin, Eric’s desire is frustrated. Nonetheless, Eric feels fine, and he is happy, content as he basks (he thinks) in Odin’s love, and fulfilled by his contemplation of the majesty of Odin. If we caused Odin to pop into existence and love Eric in the standard undetectable way that deities love people, I don’t think Eric’s well-being would change one whit. Similarly, if Odin does exist today but will cease to exist tomorrow, I’ll be just as ready to say that Eric is sitting pretty tomorrow as I am today.

    I use the Odin example because belief in deities is common, but there might be no deities, or only a few, so lots of people’s fervently held desires might be frustrated. Do you think that affects their welfare? Would they be better off *with respect to well-being* if their gods existed? (Leave aside the benefits that their gods might heap upon them if they existed; assume that these are the hands-off, no-intervention deities with which we’re familiar from our field observation of deity behavior.)

    If you do, you’re adopting a broader view of well-being than mine. Although I agree that it could be naturalistically respectable to talk about what S desires and whether the world conforms to it, I think that’s a big mess of a natural property to track, and it’s not clear that it’s identical with well-being. Of course, you probably have various refinements that permit you to ignore some desires and demand others (whether the subject actually entertains them or not), but then it becomes even messier. Once some putative natural property gets messy enough, I think it looks more like we’re just talking about our theories rather than talking about the world.

    An organism’s affective responses to stimuli, on the other hand, constitute a comparatively neat natural property. Every evaluative organism trundles along encountering, interpreting, and evaluating stimuli, and if most of the stimuli ring up as “good,” I’ll say that organism is faring well. If aspects of the organism’s current or future circumstances threaten to introduce lots of “bad,” I can cite that as reason to worry about the organism’s well-being. It’s not fancy, but it’s something we can nail down and talk about.

    As for your suggestion that hedonism “overrides evaluators’ own judgments of what is best for them,” I suppose it might if it were a hedonistic theory of value, but I don’t hold any such thing. A hedonistic theory of value, I take it, holds that only certain sorts of mental states are valuable. I don’t think anything is valuable, but I value things. If, when I value things, I think that I am noticing that they contain value, I am misguided (i.e., there’s no value with or without valuers). Instead, I am noticing that I have positive affective responses to certain stimuli. I can prettify the way I talk about the affective responses, but I don’t think that changes what I’m talking about.

  17. Could you clarify what the meaning or significance of ‘wellbeing’ is, on your view? Again, does it have any implications for what people have reason to choose, or to desire? I gather that you want to associate the word ‘wellbeing’ with affective responses, but what non-linguistic upshot are you trying to establish here? If your claims about ‘wellbeing’ have no implications for what we have reason to choose or to desire for our own sakes and for the sakes of our loved ones, what do you take the point of this line of inquiry to be?

  18. Jonathan Livengood says:


    Actually, I’m not sure I understand your definition of “harm” well enough to say whether I agree with it or not. For example, I’m not sure what you make of cases where agent x desires p, but I (or we) judge p to be physically injurious to x. In some cases, I would judge x’s desire (or acquiring the object of that desire) to be harmful — e.g. a desire for self-death. But in other cases, I don’t think I would, e.g. a desire for a stiff drink. I’m not sure if having a reason is enough. Or do you mean for the reason to be a good one, or a prevailing one? (After all, I can have reasons both for and against an action and still engage in that action.)

    In any event, in order to get the fully zombified thought experiment, you need to consider whether or not a person is harmed by what they don’t, never will, and never *could* know.

  19. Steve Nayak-Young says:


    My “mental states” definition of “well-being” is closely related to health and happiness, though I don’t think it’s identical to either one. It might be close to the answer to the question, “How you doin?”

    It’s not just physical and mental health, because some things we might call ill health are compatible with great well-being, specifically when, e.g., being in a wheelchair, being blind, or having mild OCD doesn’t bother one. It’s also not quite the same as happiness, because “well-being” takes into account those lurking threats to mostly-positive affect that isn’t currently busting through the Prozac or otherwise penetrating one’s consciousness.

    My definition includes anything the subject could notice and evaluate, and excludes anything the subject couldn’t notice. That’s a lot of stuff. I’d say we have reason to choose and desire things that we can notice and value. And if subjects have false factual beliefs, they can “notice” that they are, e.g., pleasing Odin by praying to him. That’s not strictly true, but it’s harmless fun. And if the subjects in question happen to be brains in vats or married to zombies, but they can’t notice the difference, they probably have thwarted desires, and we thought-experiment evaluators won’t like the sound of it, etc., but I just don’t think it would harm their well-being.

  20. Jonathan – we can talk of both ‘pro tanto’ and ‘all things considered’ harms, corresponding to whether we have merely pro tanto or all things considered normative reasons to dis-prefer them for the individual’s sake. So a stiff drink may be harmful, insofar as your liver is concerned, but beneficial (desirable) all things considered if it contributes to an enjoyable evening, or whatever.

    So yeah, the question is whether we could have reason to prefer for a person’s sake that they avoid some outcome which they could never learn of in any case. We often do have such preferences (again, Jack obviously hopes that his wife isn’t a zombie, and his motivations here aren’t purely altruistic), so the question is whether those preferences are rationally warranted, or irrational, or perhaps ‘a-rational’ in the sense that there’s neither reason for or against it. Another way to get at the question is in terms of rational choice: suppose that (years earlier) Jack’s mother dies and goes to heaven, where God tells her that Jack’s future wife (Sally) is a zombie. He then offers to “dezombify” Sally, making her a genuinely conscious person, if Jack’s mum requests this. Given that she loves her son, does she have a reason to ask God to dezombify Sally and thereby ensure that her son will get to share his life with another real person? It seems so to me. Surely any loving mother would want this for her son. Genuinely caring about a person surely requires wanting them to have a meaningful life, with meaningful connections to other people, and not the mere appearance of such (however reliable the deception may be).

    Steve – I take it you disagree with the normative claims made in my previous paragraph? Jack can’t tell whether his wife is a zombie, so (you apparently think) he has no reason to hope that she isn’t? His preference, which I’m sure most people would share and consider perfectly reasonable, is instead irrational (or at least a-rational)? His mum would have no reason (at least for Jack’s sake) to ask God to de-zombify Sally?

    These seem mighty large bullets to bite…

  21. Steve Nayak-Young says:


    Jack has reason to hope that his wife isn’t a zombie, because he doesn’t like the idea of being married to a zombie. He wants her to be a non-zombie, like him, and it makes him happy to think that she is. His mum would have similar reason to ask God to de-zombify Sally. She cares about Jack, she thinks zombie wives are no good, etc., etc.

    As you rightly insist, we care about lots of things other than our mental states, and I wouldn’t change that for all the zombie wives in the world (and some of them are real lookers, I imagine). As for what people desire or prefer, I believe there’s a longstanding Humean view holding that *all* desires or preferences are a-rational, and I’m somewhat tempted in that direction. Jack doesn’t want to be married to a zombie, you don’t like Zombie Soda, and I hope that Cody Hodgson, the Vancouver Canucks’ 2008 first-round draft pick, will be the 2009-10 NHL rookie of the year. I really do.

  22. Ok, I’ll guess I’ll bow out at this point, since it seems that we must have been talking past each other. If you agree that people who care about Jack thereby have reason to (want to) dezombify his wife, then that is just to say that having a zombie spouse was (as I use the terms) bad for Jack. When you use the terms differently, it does not appear that you are actually saying anything of substance, so it isn’t clear that you’re really a “mental state theorist” in any sense that matters.

    P.S. Resist temptation! I’d agree that some desires are a-rational. But surely not all. For example, it would be either crazy or malicious for someone (Jack’s mum, say) to actively want Jack’s spouse to be a zombie. And even a hedonist must think that we have reasons to prefer pleasure to pain. The alternative is normative nihilism, the view that nothing is really good or bad for anyone (let alone overall).

  23. Jonathan Livengood says:


    I think I’m having trouble getting my brain wrapped around zombies here. I still have a very strong intuition that spouses of zombies are neither better off nor worse off than spouses of non-zombies. I doubt that even God could make a philosophical zombie, and were I told on top of Sinai that my wife is a zombie, I would want to know what that could possibly mean. What are the consequences of her zombie-ness? And were God to say, “I can unzombify her if you want,” I think my reaction would be something like, “Uh … What would that mean?”

    I can’t see any reasons to prefer a non-zombie spouse, child, relative, or self to a zombie version of the same. But maybe I’m missing something, so let me turn it around. Why would you want your spouse, child, whatever, to be a non-zombie? What do they gain by being a non-zombie? What is lost by being a zombie? (This is not a joke or a taunt, I honestly don’t know.)

  24. Steve Nayak-Young says:


    Thanks for seriously engaging with my half-baked zombie talk.

    I find a lot of discussions about well-being end up in one or more parties sticking hard to the intuition that only subjectively discernible stuff can affect well-being, and others sticking equally hard to the opposite intuition. My fellow grad student Steve Campbell has raised the idea of a distinct concept of “phenomenal well-being,” which tracks only those harms and benefits of which a subject is or could be aware. My inclination (as I’ve made abundantly clear) is to say, “How can X be a harm or benefit to the subject if the subject isn’t and couldn’t be aware of X?”

    But splintering a “phenomenal” bit out of a larger, more inclusive concept of well-being isn’t the only way to go. I’ve heard Tom Hurka say on several occasions that he has no use for the concept of well-being, as it’s not nearly so interesting as the concept of a good life. I doubt that Hurka would endorse anything like my take on well-being (if he bothered to develop a view on the matter), but I prefer this way of pulling the two things apart: well-being only concerns those harms and benefits of which a subject could be aware, while “the good life” is more accommodating to Aristotelian fun and can incorporate anything we want to say about what’s necessary, sufficient, etc., for a subject to have a good life.

    I’m aware that many people use “the good life” as somehow synonymous with welfare and well-being, and I don’t have any real issue with the labels we attach. Nonetheless, I think there should be some way of isolating and referring to the harms and benefits of which a person is or could be aware. Whether that’s Campbell’s “phenomenal well-being” and “well-being” or my “well-being” and “the good life,” there oughtta be some way of dividing the two concepts (or perhaps technical terms). Otherwise, we just end up with deadlocked intuitions, which can be fun for a while (especially if experience machines, deceived businessmen, and zombies are involved), but is inevitably dissatisfying.

    As for the charge that I’m not saying anything of substance, one good parting shot deserves another: I’ll admit that my view isn’t nearly so “substantial” as yours, but as I explained above (at 4:28 pm), I think that’s one of its virtues! :-)

    By the way, how’s Eric doin’?

  25. jmsytsma says:

    Late to the party and only a cursory read of all of these comments… but let me jump in anyway!

    Question: You write that the zombie “behaves exactly like (and is indistinguishable from) a normal person but feels nothing and doesn’t love anyone.” Why no love for the zombies? What does love feel like and do you think that it is no longer love if you lack that what-it-is-like? It seems to me that love is far messier than that and that even if you believe that love (typically) has love-qualia (whatever those are) associated with it, love still must involves much, much more than those qualia… and the zombie, of course, has all the rest of the stuff that makes up love.

    I’m assuming that the intuition that the zombie’s spouse is worse off has to do with the seeming betrayal involved in loving someone who doesn’t *really* love you back (and that a betrayal is a betrayal whether you could ever know about it or not). Does that intuition still pump if it is just love-qualia that are missing? It seems to me that the zombie loves his wife, even if he lacks some love-qualia (whatever those are); couple that with the facts that the zombie doesn’t know he is a zombie and that the wife doesn’t know her spouse is a zombie, and I just don’t see how the wife could be any worse off: Nothing of substance is different.

    Let me try to pump that intuition. Imagine that I have strange phenomenology: Whenever I feel a particularly intense bit of love for my wife, I feel a tingle one foot above my head… and that exhausts my love-qualia. (I am assuming that love-qualia are only involved in such occurent episodes of loving and not, for example, when I am not even thinking about my wife, even though I love at all times.) The tingle is rather forgettable; not especially pleasant nor unpleasant… and of course, this tingle has nothing to do with my love-behavior (if having no love-qualia makes no behavioral difference, then having love-qualia makes no behavioral difference). Does it really seem like my wife would be worse off if I didn’t have that occasional do-nothing tingle one foot above my head?

  26. Steve Nayak-Young says:

    To paraphrase Bart Simpson, you could fill up a barn with what I don’t know.

    I thought that zombies were like robots running on some kind of Searle-Chinese-Room program, such that they produce all the expected responses for humans, but they’re all dark inside.

    I think there’s “something it’s like” to see what’s in front of my face or think, “I am here now,” or just about anything, so I guess I thought zombies didn’t have much, if anything, going on upstairs. Shows what I know.

    So here comes Captain Ad Hoc: These zombies are magic. They act like humans, but they’re as conscious, sentient, etc., as blocks of wood. So having a relationship with one is (phenomenally) as fun as having a relationship with a normal human, but it’s (objectively) a terrible sham and a horrible blight on one’s well-being. Weep for the spouses of zombies (unless the zombies’ spouses are also zombies).

  27. GNZ says:

    are you not removing all the possible “good” of love by swaping the qualia for a foot tingle?

    Ie even in the case where love has a greater value in terms of having a network of loving relationships that would still rely on the fact that love had some sort of intrinsic value.

  28. Jonathan Livengood says:


    You’re correctly understanding philosophical zombies for the most part, I think (Justin will correct me if I’m wrong, I hope). But the point that Justin (and I) have been trying to make is that they are conceptually incoherent. You just can’t have something that does all the things a normal human does and yet is only as sentient as a block of wood. (The part that you may not have right is that zombies are usually understood to be physically identical to their non-zombie counterparts, right down to the structure and firing of their neurons. I think this is really quite important to the debate about zombies, since if one could get a Searle-style lookup table to mimic human behavior — something that every reputable computer scientist will tell you is physically impossible — then maybe his thought experiment would be troubling.)

    Part of the trouble, then, is why one should think that agency, person-hood, the self, or anything else important to having a loving relationship requires qualia.

    GNZ and others,

    Why think that an agent with no qualia derives no “good” from loving relationships? Or is incapable of living “the good life”? Or is incapable of suffering harms? I’m fairly confident that the oak tree living in my parents front yard has no qualia, yet I think it makes sense to say that I have reasons to prefer some things over others for the sake of that tree, to borrow Richard’s language.

    Now, suppose that some tree or other has tree-qualia, and suppose that you are forced to choose between killing the tree and killing a philosophical zombie. Which is the morally correct course and why?

  29. GNZ says:

    Well it is kind of “by definition” that things with no “qualia” derive no good.

    I imagine the issue here is that you are arguing that the zombie is to you an incoherent idea and so intrinsically the thought experiments about zombies are not coherent. that isnt the sort of issue one can just skip over.

    The standard thing to do here is to just say – for your purposes the qualia are just “what it is like” including “what it is like to suffer a harm”

    BTW that thought experiment is indeed quite a intuitively big bullet bite. But if the zombie by definition cannot be harmed or helped then you have to favour the tree.

  30. jmsytsma says:

    I hardly think that that can be by definition. Look, if you go the standard zombie route (microphysical duplicates that are nonetheless lacking in phenomenal qualities), then phenomenal qualities must be epiphenomenal; but, it is dubious that some epiphenomenal quality is a necessary condition for being open to deriving good or whatever. If qualia are understood in a way that makes them irrelevant, and that is the upshot of taking the zombie thought experiment seriously, then it is hardly a radical position to deny that they are what “makes life worth living” or some such.

  31. Steve Nayak-Young says:

    I think I still don’t quite get zombies. I assume that consciousness is caused and/or constituted by biochemical stuff, so a microphysical duplicate of someone with phenomenal consciousness will have phenomenal consciousness, too. So I don’t think they’re possible, either. (I’m prepared to be flippantly dogmatic about that assumption, if necessary, so please don’t try to talk me out of it or ask me to defend it.)

    If I ever try to use this crazy example again, I’ll have to distinguish my zombies from the philosophical kind by making them slightly microphysically distinguishable when carefully examined by hypothetical super-duper neuropsychologists who know exactly how neurons cause consciousness in us (and therefore can tell that zombie neurons don’t cause consciousness in zombies), but indistinguishable to primitive slobs like us and the denizens of 25%-Zombie World.

  32. For the record, property dualists agree that (given the natural laws governing our world) brain states give rise to conscious mental states. So we don’t think zombies are nomologically possible. A world where physical stuff fails to give rise to phenomenal stuff would be importantly different from our own.

    But in any case, even those who take it on faith that zombies are metaphysically impossible can usually still make sense of the idea (they just don’t think reality would ever co-operate with it), which is all you need for this sort of thought experiment. Supposing that zombies were really possible, wouldn’t it be awful to end up married to one?

    To address Justin’s optimism: remember that we’re not just talking about a missing “tingle” here. Zombies never feel anything. They’re not happy to see you. They don’t feel comforted when you give them a hug. They don’t care about your wellbeing, or share your hopes and dreams. They don’t have any genuine desires or other mental states at all. (Phenomenology seems partly constitutive of all genuine mental life. Granted, we can talk of “subconscious desires”, but they’re really an altogether different kettle of fish — we certainly don’t care about these purely dispositional/behaviouristic forces in the same way.)

    In short, zombies are mere automata; effectively stimulus-response machines, albeit with advanced information-processing capabilities. There’s a lot more missing than just Justin’s “tingle”. And again, that’s not to say that they’re really possible. But for the sake of the thought experiment, this is what you’re supposed to be imagining: creatures that look and behave like us without really possessing minds.

    Even if you think the separation strictly impossible, you must agree that it is in respect of their having minds that we care so much about other people (rather than just, say, in respect of their behaviour).

  33. Jonathan Livengood says:

    Sure, I can say the sentence, “Zombies are creatures just like us but without minds.” I have at least a vague sense of each of the words on its own, just like other examples of syntactically well-formed sentences composed of actual English words. Sometimes, I can even fool myself into thinking that I can give sense to the sentence, but I can’t.

    And zombies aren’t “mere” automata. They’re fantastically, mind-bogglingly complex automata. All that complexity is important. Having that complexity (or at least some features of that complexity) just is to have a mind.

    Insofar as one (namely me) thinks that the separation of physical and phenomenal is strictly impossible, one thinks that having a mind and having all the right behaviors (including internal behaviors, like neuron firings) are exactly the same thing.

    I agree that it would be ashame (and maybe even illegal) if one were married to something without a mind, but in that case, one might be married to a statue or a tree but not to a zombie.

  34. GNZ says:

    So are you saying that love matters even if it is between two things that don’t feel anything when they love? If so what is the root cause of the value of that?

  35. jmsytsma says:

    Hey Richard,

    “Zombies never feel anything. They’re not happy to see you. They don’t feel comforted when you give them a hug. They don’t care about your wellbeing, or share your hopes and dreams. They don’t have any genuine desires or other mental states at all.”

    This is rather contentious. That zombies aren’t happy to see you, don’t care about your wellbeing, don’t hope or dream… none of that is obvious. The zombie thought experiment on its own doesn’t establish that; it asks us to imagine micro-physical-duplicates that lack qualia. Happy to oblige (to the best of my ability to do so). As the qualia don’t have any impact on the physical world, don’t alter one particle or change one behavior, what I’m being asked to imagine are epiphenomenal qualia. Again, happy to oblige (in so far as I can). Now, explain to me why having these qualia is necessary for being happy, for caring, for hoping and dreaming… and of course for loving. I neither see why they should be nor how they could be. The qualia at issue (the one’s we are imagining in attempting to entertain the zombie thought experiment) are too impotent for that. They do not motivate, do not alter behavior, aren’t noticed, aren’t known. What is required at this point, I suppose, is an account of love (or happiness, caring, etc.) that applies to normal humans but not zombies; I don’t see how that would be done except by stipulating that certain qualia are necessary. I’m not sure how you stipulate things about certain epiphenomenal qualia, but regardless I would still want some motivation for thinking that epiphenomenal qualia are necessary for love.

    I take it that this is beside the point, however; Steve wasn’t really interested in *these* zombies, but in elaborate (magical) fakes such that we would all agree that they aren’t really happy to see you, don’t really care about your wellbeing, and so on. Once that is made clear, I think I share your intuition: your spouse being such a “zombie” is a betrayal, betrayals are harms.


    Wow, off the top of my head I have no idea. I’m saying that my zombie wife and I still love each other, because the zombie versions are exactly the same as the normal versions of us minus some do-nothing qualia. As such, it seems to me that everything that matters, still matters, and its mattering still has the same root causes. (This has to be the case as nothing is causally different between the normal versions and the zombie versions.) What the root cause of love mattering is, however, I don’t think I’m prepared to speculate on.

  36. Jonathan Livengood says:


    I take it that the power of the thought experiment comes from something peculiar to zombies — the fact that they can’t be distinguished from non-zombies. Do you still have the intuition that the non-zombie spouse is betrayed under the condition that no one could tell that the zombie spouse isn’t really happy to see his/her partner, doesn’t love his/her partner, doesn’t really care about his/her partner’s well-being?

    I still don’t think the spouse of such a zombie is harmed. And I still don’t think that the case to be imagined makes any sense.

    However — and here I think I was either making a mistake earlier or at least not clearly avoiding one — I don’t think that the zombie intuition helps make Steve’s case. He asks in the original post, “If you agree with me [that spouses of zombies are no worse off on average than spouses of non-zombies], do you think this constitutes a counter-example to thought experiments like Kagan’s Deceived Businessman?” I should have answered, “No, the zombie-spouse case is not a counter-example.” The fact that the zombie spouse is modally indistinguishable from a non-zombie spouse makes the case importantly different from the Deceived Businessman. Maybe the story can be patched up in some way, but I don’t know how.

    Anyway, I think that we’ve come to the point where we need to argue about Leibniz’s principle of the identity of indiscernibles. (I’m for it, if you couldn’t tell.) However, I don’t know the current literature on that question … does anyone care to suggest some reading?

  37. Jonathan – I think you’re confusing the identity of indiscernibles with logical positivism! The former principle concerns ‘discernibility’ in the technical sense of having different [qualitative] properties. Zombies are discernible from humans in this sense (they lack phenomenal properties that we have).

    Justin – maybe you’re speaking a different language from me, but in my language, the idea that one could be “happy” even though one lacks a first-personal point of view (or “anything that it is like”) is a straightforward contradiction in terms. Similarly with other mental predicates.

    what I’m being asked to imagine are epiphenomenal qualia

    Not necessarily; Chalmers has noted a couple of other possibilities, most notably “type F” monism.

    But yeah, this is all a bit besides the point.

  38. Jonathan Livengood says:


    I don’t think so, although I will cop to having sympathy for the logical positivists. (Maybe I’m wrong and I just am a variety of positivist. In that case, I’d prefer to argue on the merits of my case, as opposed to losing an argument by bad association. You *did* mean to insult me, right?) As I understand it, I admit more ways of knowing things than the positivists (e.g. the limits of infinite processes count as knowable for me but not for Carnap), and I don’t think metaphysics can be replaced by linguistic analysis.

    That said, I don’t think the identity of indiscernibles is so empty that the mere *assertion* that two things have different properties is enough to give them different properties. I don’t think Leibniz would have liked this sort of move any better than I do, and it doesn’t jive with the little bit of the recent literature that I am aware of, e.g. Black’s paper. (I’m still happy to read something new if you want to suggest where to look!)

    Let me illustrate my concern here. Were I to say, “There are two perfect spheres in an otherwise empty space,” Leibniz would reply, “You think there are two, but owing to the symmetries of the case, you are wrong. There is one sphere.” If I replied, “But God sees a difference,” I haven’t really improved my position. In a sense, I’ve added a property, but not really. Now, it looks to me like we’re in a similar position. You say that zombies are like regular people but without phenomenal properties. I say that this is an illusion — zombies have phenomenal properties iff their non-zombie counterparts do. Then you (actually, Steve) say, “Well, I can’t really give you a *story* here, but just stipulate that God can tell the difference between zombies and non-zombies.” I’m not satisfied with that reply, but I don’t have any patch to suggest. Do you?

  39. GNZ says:

    So if push came to shove would you be saying desire for love is a-rational? (ie it is neither rational nor irrational to have the desire)? If so then that may be the root of the issue, if not then I’m not sure why you think we are in such a poor position to determine why things matter since we have so many apparent data points.

  40. Steve Nayak-Young says:


    Here’s a story: The deity of 25%-Zombie World decided to make the place live up to its name by making every fourth human-like organism a zombie. These zombies aren’t microphysical duplicates of any humans with phenomenal consciousness. Instead, they are very similar to non-zombies, but their “neurons” are ever-so-slightly scrambled, such that they lack any semblance of phenomenal consciousness, while still having enough information-processing complexity in the wetworks to behave exactly like non-zombies. Thus, the deity can distinguish zombies from non-zombies, and so could any other omniscient person or 30th-Century (or perhaps 40th or 50th) Earth neuroscientist, but nobody in present-day (or near-future) Earth or 25%-Zombie World has a hope of telling the difference.

    Richard: I’m relieved to see that you share a similar view to mine re the import of “qualia,” as you seem much better acquainted with phil of mind stuff (e.g., you recently wrote the phrase “‘type F’ monism”). I have long thought that “qualia” refers to phenomenal consciousness and what-it’s-likeness, such that there’s nothing it’s like — i.e., no first-person point of view — to be a creature without qualia. I gather that this view is not universally held, as evidenced by the dissenting comments from jmsytsma and GNZ, but at least it’s not just me.

    GNZ: We’re in a poor position to determine why things matter for much the same reason we’re in a poor position to determine why unicorns became extinct. However, if we want to determine why things matter *to us* or why people think about unicorns, we’re in much better shape.

  41. jmsytsma says:


    I’m not sure that I am following. I think that the “cannot be distinguished from” causes all the problems. On the standard zombie thought experiment, this means that you can take it one of two ways (and the two of us might be generating confusion by taking it in different ways): You can either start with a conception of phenomenal consciousness and qualia on which they are robust and important and deny that zombies are conceivable because you cannot strip-off what is robust and important without otherwise changing something; or, you can let the thought experiment define what is being stripped away, in which case whatever it is that the zombie lacks is not robust or important and is in fact epiphenomenal. I think that you have been taking the former route, I’ve been taking the latter. Of course, they amount to the same thing.

    (Steve, I think this accounts for the confusion that you are noting, as well.)

    The only thing to be done to save the thought experiment for Steve’s purposes is to drop the cannot and make the pseudo-zombie a mere behavioral duplicate that is actually quite different from the real thing. I’m now thinking of something like the spouse being filled with completely homogenous jelly and controlled by an evil demon. Now there really is something importantly different between the jelly-spouse and the real deal, even if you couldn’t tell the difference because the evil demon would zap your memories or some such whenever you stumble onto the secret, for example. And, I am comfortable calling this a harm. I’m guessing that this isn’t importantly different from the deceived businessman case anymore, however?


    I think you are coming at it from the wrong end: You are brining your conception of phenomenal consciousness to the zombie thought experiment and assuming that zombies lack phenomenal consciousness, but what you need to do is to let the zombie thought experiment do the work; you need to let the zombie thought experiment tell you what zombies lack. If I take the zombie thought experiment seriously, calling whatever it is that I have and your micro-physical-duplicate zombie-twin lacks “qualia,” then I see no reason to think my zombie twin lacks a first-personal point of view. Again, what the zombie lacks must be epiphenomenal according to the zombie thought experiment, because lacking (or having) these qualia makes no difference; they can’t make a difference because the thought experiments stipulates that *nothing else* changes except the presence or absence of the qualia. And, I don’t see how appeal to Type F Monism could change that, as this is something that follows from the set-up of the thought experiment.

    Anyway, I don’t really see why Type F Monism would mean that what the zombie lacks aren’t epiphenomenal. Chalmers writes (“Consciousness and its Place in Nature”): “Type-F monism is the view that consciousness is constituted by the intrinsic properties of fundamental physical entities: that is, by the categorical bases of fundamental physical dispositions. On this view, phenomenal or protophenomenal properties are located at the fundamental level of physical reality, and in a certain sense, underlie physical reality itself.” But, if a micro-physical-duplicate duplicates the physical entity down to the fundamental level of physical reality (and why shouldn’t they?) then they should either deny the conceivability of zombies or should hold that whatever the zombie lacks are epiphenomenal. Chalmers seems to suggest as much himself: “A type-F monist may have one of a number of attitudes to the zombie argument against materialism. Some type-F monists may hold that a complete physical description must be expanded to include an intrinsic description, and may consequently deny that zombies are conceivable. (We only think we are conceiving of a physically identical system because we overlook intrinsic properties.) Others could maintain that existing physical concepts refer via dispositions to those intrinsic properties that ground the dispositions. If so, these concepts have different primary and secondary intensions, and a type-F monist could correspondingly accept conceivability but deny possibility: we misdescribe the conceived world as physically identical to ours, when in fact it is just structurally identical.[*] Finally, a type-F monist might hold that physical concepts refer to dispositional properties, so that zombies are both conceivable and possible, and the intrinsic properties are not physical properties. The differences between these three attitudes seem to be ultimately terminological rather than substantive.”


    I don’t think I want to say that the desire for love is a-rational. I just wanted to say that I don’t know what the root cause of the value of love. I don’t suppose that the root cause is rational… I imagine it is something like evolution? Anyway, the point I wanted to make is that zombie love matters just as much as non-zombie love, however it comes to be that matters!

  42. Justin – right, on the Type-F route, we do best to understand the zombie world as merely structurally identical to ours. It is alike in all respects that scientific observation can detect. But the underlying intrinsic natures of the entities differ in that one, but not the other, is capable of giving rise to (fully-fledged) phenomenal consciousness as we know it.

  43. jmsytsma says:


    Maybe I am missing your point, but I can’t for the life of me see how adding these “underlying intrinsic natures” would help: The qualia are still epiphenomenal. As far as I can see, you’ve just added another extra bit to our world that doesn’t matter (magical natures that do nothing but give rise to stuff that does nothing). What is the point?

    Regardless, how could the stuff that is given rise to on our world but not the zombie world be “phenomenal consciousness as we know it”? That is, how do we manage to *know* something without being structurally and behaviorally different than if we didn’t know it?

  44. Steve C. says:

    Hey Steve,
    Fun case! So, is this just a variation on Deceived Businessman, or do you think that the zombie case has some greater intuitive pull in favor of mental statism? (I guess I’m inclined to think that it’s only a variation. I’d be surprised if anyone had asymmetric intuitions about these two scenarios.) In any case, it’s worth discussing.

  45. Steve Nayak-Young says:

    Good question, Steve!

    How is a zombie spouse better than a cheating one?

    First, no slip-ups! Zombies are guaranteed never to come out of character, get “caught,” or let on that they don’t love you (unless they appear to stop loving you the way a non-zombie would — they might have a 50% divorce rate in 25%-Zombie World, too).

    Second, nobody will find out and treat you like a chump! Often, if your spouse is a cheater, you’re the last one to know, but you’ll get to enjoy heaping servings of scorn and pity from others who do know, until (and after) you finally find out. With a zombie spouse, nobody will ever find out.

    This is an improvement on Deceived Businessman because even though we carefully stipulate in DB that he’ll never find out, we probably can’t get the threat of all the usual misery associated with deception out of our heads. With zombies, there is no deception and zero threat of ever finding out that one’s marriage is a loveless sham, having one’s world come crashing down, etc.

    This is all silly, in any case — I hate arguing about what words mean. But I thought the zombie bit might be amusing.

  46. GNZ says:


    > GNZ: We’re in a poor position to determine why things matter for much the same reason we’re in a poor position to determine why unicorns became extinct.

    I have sympathy for that position. But jmsytsma doesn’t seem to want to go quite that far leaving me uncertain of what his view rests upon. Maybe that is just a definitional issue…

  47. GNZ says:

    well there is the epiphenominal qualia difference. The definition of knowledge just has to be extended to include those epiphenominal qualia.

    Of course that cannot have a causal effect on knowledge’ in as far as knowledge’ is defined as just the patterns in your physical brain.

  48. David Manley says:

    when I was about five I had the horrible thought that despite all her apparent love for me, my mom might be a robot inside. (I assumed, roughly, that this would involve lacking a first-person perspective.) It was, palpably, one of the worst possibilities I had ever considered. Not only did I think that I would be worse off, but it was clear to me that I much of what I thought was valuable in my life would turn out not to be. That was a visceral certainty, which spurred the fear. Of course, it was a skeptical scenario and I realized I could never be sure.

    the curse of a philosophical temperament, I suppose. Anyway I still think I was pretty much right about the value judgment, though not about the skeptical concern.

  49. Steve Nayak-Young says:

    Hi David,

    Thanks for posting. You could be right about the value judgment while still accepting that there could be a sense of “well-being” that wouldn’t rate a robot mom as a harm. As I suggested in an earlier comment, I like to think of someone’s well-being as tracking what that person might ever say in response to the question, “How are you doing?”

    So, for example, if Bob has a tiny cancerous tumor that isn’t bothering him now but will bother him later, we can still think of the tumor as a threat to his well-being, a harm, etc.

    However, if Bob would be devastated if he found out that the deity he prays to doesn’t exist, and the deity doesn’t exist, but Bob will never, ever learn this, I think that the fact that Bob unwittingly prays to a non-existent deity is no harm to Bob. Indeed, it’s probably good for Bob — it makes him happy.

    Steve Campbell has suggested carving “phenomenal well-being” out of “well-being,” while I’m in favor of keeping well-being simple (i.e., how you’re doing, as far as you’ll ever know) and using “the good life” to refer to all the splendiferous Aristotelian benefits and harms that aren’t noticed by their beneficiaries/victims but make all the difference to omniscient evaluators of thought experiments. ;-)

  50. Lhea says:

    I disagree that in this world, disloyal spouses are distinguishable from loyal spouses. I completely disagree.

    There may be a distinction between loving and unloving. But I don’t think that loyalty and love are synonyms. I think it is possible to love someone and be unfaithful. But perhaps that would be another article/post?

    I reckon I would agree, how is a zombie spouse better than an cheating/unfaithful one?

    Fundamentally, I disagree that a person who unknowingly has a zombie spouse is worth off than someone who has genuine love in their life.

    The person who has the zombie will feel loved, and have all of their other intimate needs met.

    Now the question is really whether the Zombie is worse off for having to act that way. And one would have to question whether the Zombie has feelings at all. If the Zombie does, then the Zombie is much worse off, while the Zombie’s partner is completely fine.

    Now there is another scenario that must be addressed.

    If the Zombie’s lover does not know that she is with a Zombie, and the Zombie does not let on. Fine. All is well for her, she will feel like she is genuinely in a healthy relationship.

    However, if the Zombie’s lover is the only one in this universe who does not know that her lover is actually a Zombie — that is the case when she, the Zombie’s lover, will be worse off.

    Surely her parents, friends, coworkers and neighbors will all judge her and wonder why she is dating a Zombie without realizing it.

    But if the world doesn’t know, and if she doesn’t know. Then she is no worse off.

  51. Steve Nayak-Young says:


    Thanks for your comments. I’m afraid my use of “zombies” probably wasn’t as great an idea as I initially thought, as there are many different takes on what philosophical zombies are. For my purposes, I wanted them to be indistinguishable from non-zombies, but to have no phenomenal experiences at all. Accordingly, the zombie spouses (in my scenario) can’t be worse or better off no matter what happens.

    Similarly, the zombie doesn’t know that it is a zombie, so it can’t let on to its spouse that it is one, and nobody else can tell who is a zombie, so the non-zombie won’t be “the last to know.” Nobody will ever know. I think this sort of harm is part of why we feel sorry for people who don’t know their spouses are unfaithful, unloving, etc. I think the “cheating spouse” case can’t help but tug at such intuitions, which is why I introduced the zombie business. There’s no threat that anyone will find out that someone is married to a zombie, because nobody (except deities, etc.) can ever find out.

  52. dtlocke says:

    @David: my thoughts exactly, both now and when I was about 5.

  53. dtlocke says:

    @Steve: does the fact that zombies are physically identical to humans play a key role in motivating your intuition? Or do you have the intuition that the spouses of puppets are just as well off as the spouses of humans in 25% (undetectable) puppet world?

  54. Lhea says:

    Well Steve,

    Now that you have made that distinction clear. Then I definitely don’t think the Zombie’s spouse is worse off.

    Intuitively, I would want to assume that it would be better to be actually loved genuinely by a human, than to only think one is being loved by what is actually a Zombie.

    But, honestly, I can’t think of a rational explanation for why the Zombie’s spouse is worse off (If the Zombie doesn’t know that it’s a Zombie, and no one in the world can distinguish between Zombies and Humans).

  55. Steve Nayak-Young says:


    Wouldn’t undetectable puppets be physically identical to (or at least indistinguishable from) humans? You’d think that the puppets would be “detectable” if they looked and talked like the cast of Avenue Q.

    In any case, the non-conscious spouses couldn’t be identical to humans because humans are conscious because of their physical structure (I assume), and beings that were physically identical to them would also be conscious. Instead, I’d like these thingies to be indistinguishable from humans but with ever-so-slightly differing neurons that don’t cause/constitute consciousness but somehow magically cause them to behave like humans. I don’t mind whether we call these thingies “puppets,” “zombies,” or “accountants,” so long as they have the features I need.

    Once we have these thingies in place, I have the following intuition: “I sure hope that my wife isn’t one of those thingies, but if she is, I’ll never know the difference, so I don’t see how it could affect my well-being.”

  56. dtlocke says:

    “I’d like these thingies to be indistinguishable from humans but with ever-so-slightly differing neurons that don’t cause/constitute consciousness but somehow magically cause them to behave like humans.”

    So when you say ‘zombies’, you don’t mean what other philosophers mean by ‘zombies’–i.e., creatures that are physically identical to ourselves but lack consciousness.

    “Wouldn’t undetectable puppets be physically identical to (or at least indistinguishable from) humans?”

    No, just as your ‘undetectable zombies’ are not physically identical to humans. The difference between the zombies and the humans is that they have ‘ever-so-slightly differing neurons’. In particular, the puppets have neurons that are remotely controlled by an unknown and undetectable alien species (use your imagination on this last part). And of course they don’t ‘look and talk like the cast of Avenue Q’. They look and talk just like humans. So, what say you: puppet wife or human?

    “I sure hope that my wife isn’t one of those thingies, but if she is, I’ll never know the difference, so I don’t see how it could affect my well-being.”

    I thought your intuition was that it would not affect your well-being, not that you ‘could not see *how* it could affect your well-being’. On the assumption that what you really meant is the former, why do you hope for something that you think would not affect your well-being?

  57. dtlocke says:

    “The difference between the zombies and the humans is that they have ‘ever-so-slightly differing neurons’.”

    Should say:

    “The difference between the *puppets* and the humans is that they have ‘ever-so-slightly differing neurons’.”

  58. Steve Nayak-Young says:

    “So when you say ‘zombies’, you don’t mean what other philosophers mean by ‘zombies’–i.e., creatures that are physically identical to ourselves but lack consciousness.”

    I’ve learned that I wasn’t particularly clear on what most philosophers mean by “zombies.” As I mentioned, I don’t think it is possible for a creature to be physically identical — as in, a particle-for-particle doppelganger — to a conscious human without also being conscious.

    Now I see what you mean by “puppets,” and I think they’re also well-being neutral, unless there’s some chance the puppet-masters could slip up and cause the puppets’ spouses to catch on.

    “[W]hy do you hope for something that you think would not affect your well-being?”

    I don’t like the idea of it, so I hope it isn’t true. But when I hope that something isn’t true, if I’ll never find out whether it’s true or not, then I don’t think it matters, with respect to my well-being, whether it is true or not. This is because I take “well-being” to concern all and only those considerations that could be responsive to the question, “How are you doing?”

    If I were to answer that question with, “I’m a bit nervous today because I’m hoping that my wife isn’t a zombie, but I don’t know how I could ever find out,” that’s all there is to know about my wife’s potential zombie status and its effect on my well-being. If she is a zombie, “a bit nervous” is nonetheless the worst well-being hit in store for me, and if she isn’t a zombie, I’ll still be “a bit nervous” because I don’t know any better than to worry about such absurd things.

  59. dtlocke says:

    You: How are you doing?

    Me: Pretty bad, my wife is a zombie.

    That’s responsive, no? Of course, I could never be in a position to know/justifiably believe what I said, but that wasn’t your criterion. Your criterion was that a consideration x counts towards well-being iff x is responsive to the question ‘how are you doing?’. To rule out my response, you need an epistemicl version of your criterion. Unfortunately, such a criterion would be false, and would be so precisely because it was epistemic. FYI: you might be interested in chapter three of my dissertation, if only because you would probably think that it was all wrong.

    Sent from my iPhone.

  60. Steve Nayak-Young says:

    Dustin: I guess I’ll have to put it more carefully when I write my masterpiece. I meant the “How are you doing?” bit to be a colloquial, intuitive way of expressing my theory-bloated, unintuitive notions.

    My notions about well-being (and normativity in general) are a tough sell because they involve thinking of humans as organisms with various responses to stimuli. Some of those responses (we might call them the affective and/or evaluative) can be flavored positive, neutral, or negative — end of evaluative/normative story. We can think of all sorts of fanciful ways of talking about the evaluative responses, and that’s largely harmless fun, but I’d recommend that we be clear about what we’re *really* doing, at least every once in a while.

    Getting back to well-being, if someone says something is crucial to my well-being, I’d like to hear about how it will cause some of those positive affective responses or help me avoid negative ones. Having true beliefs or matching desire-world pairs isn’t necessary. In other words, we’re already living in experience machines. We think they’re made of neurons, other wetworks, and an external world, but it wouldn’t make any difference to our well-being if we were brains in vats, Matrix pods, or strapped to Nozick-style experience machines.

  61. dtlocke says:

    I think I understand what your view is. I was just trying to figure out what your argument for that view is. As it is, I take Nozick’s experience machine and your 25% zombie world to be counter-examples to your view. Do you have an argument as to why I shouldn’t take them that way?

  62. Richard says:

    Dustin – if Steve is just stipulating a definition for how he uses the term ‘wellbeing’, what is there to disagree with? He hasn’t said anything substantive at all, and so a fortiori hasn’t said anything false.

    (Rather than disputing terminological claims, it would seem more useful to just “change the subject”, to be more explicit about what ethicists really care about in these sorts of disputes, viz., the substantive question of what we have reason to want for a person’s sake.)

  63. Steve Nayak-Young says:

    Richard, thanks for having my back. You see, Dustin? You can’t “understand” my view, and I don’t need an “argument” for it, because I don’t have a view. I’m merely stipulating.

    But hang on, Richard, aren’t you and the ethicists doing the same thing (assuming that you’ve got your finger on the pulse of what “the ethicists” are up to) when you stipulate that “well-being” refers to “what we have reason to want for a person’s sake”? And even if I cave and join your more popular style of stipulating, I don’t think we’ll be able to talk about well-being until we figure out what reasons and sakes are, perhaps via further stipulation.

    Changing the subject slightly, one of my hobbies over the past year has been to write blog posts that Dustin might find amusing. For example, I might write that one of the things one can do with words is to refer (or purport to refer) to stuff. In the same general neighborhood, one can use words to describe (or purport to describe) states of affairs (i.e., lots of stuffs and relations between them). If I were to say, e.g., “Bob has reason to want for Mary’s sake that Mary isn’t hit by a bus,” and if I intend by those words to refer to stuff and/or describe some state of affairs, it would be nice if I could at least gesture in the direction of the sorts of stuff(s) to which I’m referring.

    In that case, I’d narrow down the sorts of stuffs to include Bob, Mary, perhaps some other people, and one or more buses, as I’m not aware of any other relevant stuffs. Bob and Mary are organisms made of carbon, water, etc., and buses can smush them. If we wanted to find reasons, sakes, and/or well-being in there, it would take some fancy describin’, but I think we could cobble something together by focusing on the evaluative capacities of the humans and the smushing capacity of the bus.

    The resulting cobbled-together referents would have at least two noteworthy features: (1) They would be real, and (2) They wouldn’t be what anybody currently means by “reasons,” “sakes,” or “well-being.” Despite the latter, if you like talking about real stuff, and you think reasons have something to do with what you like, then the foregoing considerations might lead you to conclude that you have reason to adopt revised definitions of normative terms along the lines I ever-so-sketchily propose.

    Of course, you could also bail out of the whole referring, describing, and asserting business and go expressivist. But that doesn’t sound like any fun to me.

    I’ll miss these attempts to amuse Dustin after he takes his show on the road to sunny California. And I appreciate his efforts to amuse in return, via the guessing game he offers, in which I try to guess both what he thinks my view is and why he thinks the experience machine and 25%-Zombie World are counter-examples to it. But that game is too hard … I like my fun to be spoon-fed to me, like Youtube clips … ;-)

  64. dtlocke says:

    Your view: Well-being is purely a matter of positive/negative ‘affective responses’.

    Why 25% zombie world is a counter-example to your view: If your view were correct, then the spouses of zombies would be as well off as the spouses of humans. However, the spouses of zombies are not as well off as the spouses of humans. Hence, your view is false.

    In any case, when I start being characterized as someone who doubts that we can refer to stuff, it’s time for be to bow out. You amuse me indeed, but not for the reasons that you think you amuse me!

  65. dtlocke says:

    @Richard: nice post over at your blog.

  66. Steve Nayak-Young says:

    Dustin: I certainly didn’t mean to characterize you as someone who doubts that we can refer to stuff. I was just trying to be clear about what *I* intend to do — as opposed to expressing, endorsing, etc. — when I use normative terms in apparently descriptive statements.

    I thought I amused you because I make ridiculous claims and try to write about language without really knowing how. Perhaps you’re *also* laughing with me …

    As for your account of the counter-example, it looks a tad question-begging to me, but no more so than if I claim (as I have) that the spouses of zombies are as well off as the spouses of humans. Of course, if you’re just making a linguistic/terminological point, I’ll happily concede that an accurate dictionary of current English usage would not define “well-being” as “purely a matter of positive/negative affective responses.”

%d bloggers like this: