Craniopagus Twins and the Possibility of Introspective Misidentification

Presenter: Peter Langland-Hassan, University of Cincinnati

Commentator 1: Annalisa Coliva, University of Modena

Commentator 2: Elizabeth Schechter, Washington University in St. Louis


  1. Thanks very much to Analisa Coliva for raising some interesting issues. Being familiar with her work on the subject (and anticipating that she would disagree with my main thesis), I’ve been eager to hear what she would say. Unfortunately many (but not all) of the criticisms seem to stem from a misunderstanding. The main point that needs clearing up is that I do indeed intend to be arguing that the One Token scenario is one where a person might wrongly self-ascribe an introspectively detected visual experience or “visual impression” (in Coliva’s terms), where it is agreed that 1) one could have this kind of mental state even if one is hallucinating, and 2) where it is agreed that this state contributes to one’s phenomenology. In many places Coliva seems to assume that I am working with a more demanding notion of perceptual experience, where one can only have a perceptual experience of certain kind if it is caused in the right way by the things that it represents. She then notes that this would leave room for the idea that self-ascriptions of introspectively detected visual impressions would still be IEM. Her idea is that, while “I see a canary” might be false because my visual impression as of a canary is just a hallucination (and so not a case where I am really seeing a canary), a more tentative judgment of the form “I have a visual impression as of a canary” will still be IEM, and that this is what matters (I will use “visual experience” to refer to this kind of mental state, where one can have a visual experience as of an x even when one is merely hallucinating an x). But when I discuss the case where Krista might misascribe to herself an introspectively detected visual experience as of a pony, I mean to be describing a case where she is misascribing to herself a visual experience (or “impression”) as of a toy pony.
    Why would self-ascriptions of this latter sort always be IEM? Coliva says, “if such a mental state were one felt or had by one of the two twins, while its physical basis took place in the other twin’s brain, it would be the former’s (too) anyway.” Now, certainly if such a perceptual experience is “had by” one of the twins, then that twin is an owner of the state. But that just raises the main question of the paper, which is whether being introspectively aware of a visual experience is sufficient to make it true that one “has” the state (in the sense of being an owner of the state).
    These points are meant to go for sensations as well: I am arguing that, in very unusual circumstances such as this one, a person could be introspectively aware of a sensation—and so feel the sensation—without thereby becoming an owner of the sensation. The idea is that it is not a logical or conceptual necessity that one can only feel a sensation that occurs in one’s own mind. I am afraid that Coliva found this view so crazy that, aiming for charity, she assumed it was not what I was suggesting! Yes, it is very strange to think that a person might feel a sensation that does not occur within her own mind; but then, this is a very strange situation—more than very strange. Moreover, the arguments I give later with respect to “Midas Touch” not being a sufficiency condition are meant to pay special tribute to the counterintuitive nature of the claim I am making (while preserving its key point). (Note: I do not take the argument against Midas Touch being a necessary condition to impugn its status as a sufficiency condition, as Coliva suggests. Rather, I tried to think of the best reason I could for why someone would think Midas Touch is a sufficiency condition, and then showed that it does not extend to perceptual states. Of course, there may be other reasons one might give for finding it to be a sufficiency condition—the key is to figure out what those reasons are).
    Now, Coliva points out that the One Token (OT) scenario still is not one where the relevant mistake occurs, since Krista (in the video, and example) does not self-ascribe the state that (I claim) belongs only to Tatiana. That is quite true, but my purpose in describing the OT scenario is just to describe a case where a person is introspectively aware of a mental state that is not her own. That is the hard part. Once that possibility is granted, I assume it is not too hard to see how Krista might, in a moment of distraction, misascribe such a state to herself (and so violate Introspective Immunity). But I should have been clearer about that in the paper.
    2) ON INTROSPECTION: A second point concerns theories of introspection. Analisa finds it “slightly inconsistent” that I first characterize introspection neutrally (as being “the way that people ordinarily become aware of their own mental states) and then go on later to say that Introspective Immunity can be violated if we assume a certain (not universally held) kind of view on introspection (namely, the view that introspection involves first and second order states or processes that are “distinct existences”). Now, the point of being neutral initially was simply to characterize the widely accepted thesis being discussed (“Introspective Immunity”). I then argue that the principle can be violated if one adopts a certain style of view on introspection. Further, if such views are at least possibly right, then Introspective Immunity is contingent. Now, one might try to argue that distinct existences views of introspection are not possibly correct, because they commit a logical or conceptual error. However, given that so many people accept and find such views comprehensible, this seems a hard claim to establish.
    Also, it is sometimes not clear whether people accept the (“constitutive”, non-distinct-existences) view on introspection they do because they accept that introspective judgments are IEM, or vice versa. An upshot of the argument is that the (putative) logical necessity of Introspective Immunity cannot be used as a reason for doubting “distinct existences” views of introspection. That is, the logical necessity of Introspective Immunity is not something that “distinct existences” views of introspection need to “get right.” So, if one does not accept distinct existences views, it cannot be for the sole reason that they would allow for violations of Introspective Immunity.
    Coliva notes that those who accept the logical necessity of Introspective Immunity “are impressed with the intuition that (at least) one notion of introspection is such that to have introspective access to pain, say, is for one to feel pain (and mutatis mutandis for other mental states and their introspective awareness).” The important point for her is that “no matter how one wants to account for self-knowledge, one should respect this phenomenological feature.”
    As noted above, I accept that a person who has introspective access to a pain feels that pain (and so I accept the “phenomenological feature” in question). At issue is whether the person must also be the owner of such states—whether a pain that is felt by a particular person must, as a matter of logical necessity, occur within that person’s mind (assuming that pains are mental states).
    Here Coliva might respond that it is just impossible to feel a sensation that does not occur within one’s own mind. Perhaps it is, but what is our reason for thinking it is? To me, this is an interesting question. The answer cannot be that holding otherwise would wrongly entail that a person could misascribe to herself an introspectively discerned (and felt) sensation; that would beg the question. Another reason one might give is that Midas Touch is a sufficiency condition with respect to mental state ownership (Coliva seems to accept that it is). That is why I argued that the best reason I could see for accepting Midas Touch as a sufficiency condition does not extend to perceptual experiences and sensations.
    Again, one may well have other reasons beyond the ones I considered for holding that one cannot, as a matter of logical necessity, feel a pain that occurs outside of one’s mind (and for accepting Midas Touch as a sufficiency condition on the ownership of all mental states). For instance, as Analisa notes, people sometimes defend “constitutive” accounts of introspection that are at odds with distinct existences views. I certainly don’t take myself to have challenged all such views. I simply note that these kinds of views are most plausible—and most commonly advanced—with respect to propositional attitudes like belief and desire. Arguments of a very different kind are needed to extend “constitutive” accounts to perception and sensation.
    3) INDIVIDUATING BRAINS. Coliva raises a nice and relevant question about how we are to think of the extension of Krista’s brain. In the OT scenario, the first order visual experience v is realized in Tatiana’s brain, and Krista becomes aware of it via the thalamic link. But can we be sure that the region of Tatiana’s visual cortex that realizes v is not also a region of Krista’s brain? After all, their brains are connected. As Coliva notes, if it is a part of Krista’s brain, then there is no clear reason to think that she could misascribe experiences realized therein (there would be no better reason than for any mental state she is introspectively aware of and that is realized in her brain).
    In responding to this, there are two issues that should be kept separate: 1) how are we to individuate brains, and 2) how are we to individuate minds? Looking at the first, why should we say that the girls have two brains and not one unusual brain? Or, why think that Krista’s brain does not overlap with all the parts of Tatiana’s brain to which it is (closely) causally connected? Well, ordinarily when two bodily organs are physically fused at some point, and where states of one are closely causally connected to states of another (e.g. the stomach being causally and physically connected to the small intestine), we do not conclude that we have just one organ. When parts of the small intestine causally interact with parts of the stomach—however closely and systematically—we do not consider parts of the small intestine to be parts of the stomach. So, I don’t think that general principles for individuating bodily organs (and their parts) weigh in favor of Analisa’s proposal.
    Now one might respond that brains are special kinds of bodily organs, and so our principles for individuating them should be different. But (recalling that the twins’ brains only obviously overlap at the thalamus) I would wager that the only reason one might be tempted to hold that a part of Tatiana’s visual cortex is a part of Krista’s brain is that Krista has a kind of mental access to states within that brain region. If we became confident that v does not occur within Krista’s mind, there would seem to be no lingering temptation to count that region as part of her brain. So everything hangs on our original question, which is whether v occurs in Krista’s mind. This is a question about individuating minds, not brains. If one responds that we cannot know how to individuate the minds here until we know how to individuate the brains, we come back to the question about brain individuation: independent of the brain’s connection to the mind, there is no biological criterion that would count Tatiana’s visual cortex as part of Krista’s brain.
    4) PERCEPTUAL EXPERIENCE AND PERSONHOOD: Here is Coliva’s conclusion, concerning the impossibility of misascribing to oneself an introspectively detected visual experience: “This conscious element of perception [viz., the visual experience] no matter how it is accounted for, seems to be constitutive of the person one is just as much as one’s beliefs, desires, memories and sensations are, precisely because it would make a difference to inner phenomenology and could lead to characteristic behavioral output.” Introspectively discerned visual experiences seem “constitutive of the person one is,” to Coliva, for two reasons: they make a difference to inner phenomenology, and they lead to “characteristic behavioral output.” With respect to the first, I have already noted that the view I propose fully accepts that the states one is introspectively aware of make a difference to one’s inner phenomenology; I’ve just questioned whether that must, as a logical necessity, make them a part of one’s own mind. I am less clear on the other reason relating to “characteristic behavioral outputs.” Earlier Coliva suggests that, on the view I propose of introspection, v would not “be taken as a basis for first-personal action.” I’m not sure how that is relevant; Krista will only act on v as though it is her own mental state (e.g., reaching out in front of her face for the pony) if she becomes confused; that does not suggest that she has not in fact become introspectively aware of it in the cases where she does not become confused. Is the idea that, if Krista mistook v for her own experience and then acted on v in a “first person way” (e.g. reaching out in front of her own face for the pony), that that would by itself qualify v as occurring in Krista’s mind? I am not sure why it would—especially if that is not the typical functional role of that kind of state. But perhaps a substantive objection could be developed along those lines.

  2. Hi Peter,

    As one who has been known to use empirical evidence to argue against seemingly obvious philosophical theses, I am sympathetic to your strategy in this paper. However, I am also very sympathetic to a hard line on IEM, so I am going to try to push back on part of your reply to Annalisa.

    In discussing the “one-token” (OT) hypothesis in your paper, you write that one thing that might “make it plausible that [the twins] might not share ownership of” a given visual experience v is that the experience occurs in only one of the twins’ brains, and “a natural presumption is that a person owns all and only the mental states that are realized in her brain” (p. 8). I take it that this is what Annalisa is pushing back on in her commentary, when she asks (p. 3): “What would prevent one from thinking that Krista’s brain is, after all, more extended than normal ones, even if the region of brain where the neurological realization takes place, by normal standards, would count as Tatiana’s only?” — which you then gloss in your reply as the suggestion that “the girls have two brains and not one unusual brain”.

    But of course Annalisa doesn’t need to say anything that extreme. Rather, her view need be only that the twins’ brains overlap *in those areas that contain sensations to which they both have introspective access*. Moreover, I take it that part of what makes this plausible is the thought that brain regions need to be identified *functionally* rather than anatomically, especially in those cases where the anatomy is abnormal. Thus what makes something a part of my visual cortex is, at least in part, that it’s directly responsible for giving me visual sensations: in normal cases, the region of my brain that does this has a certain location, shape, and structure; but in abnormal ones (such as cephalic conjoining, brain damage, radical cortical rewiring, TVSS (perhaps), etc.), my visual cortex might have a different structure.

    I’ll say more about this later, but have to run out now. In the mean time, I’d love to hear some more of your thoughts on this question.

    John Schwenkler

  3. Hi Peter

    Thanks for the paper, it’s really interesting. And thanks for reading my old paper (I knew somebody would one day!). One question (which I think has already been raised by Elizabeth), then a challenge.

    Do we have any more information about the experiences in the case of Krista and Tatiana? Just from watching the video, it’s not implausible that we have something akin to blindsight going on there. That is, whilst, say, Krista is able to correctly judge that Tatiana’s foot is being tickled, she does not have an experience as of her/a foot being tickled. This might, I suppose, be tested by seeing the extent to which she is disposed to giggle when Tatiana is tickled. If this is going on, the I don’t think that the empirical case is really doing anything for you. Since, no matter how neutral you are on the nature of introspection, is surely isn’t just accurate blindsight-style guessing. So, your case really would rest on a thought experiment. That’s fine with me though.

    Now a challenge. Suppose I say that the scope of introspection (the range of states to which one has ‘i-access’) is determined by the function of that faculty, by what it’s for. Further, I suggest that it’s very plausible to view that function is providing knowledge of one’s own psychological states. With this in mind, I would want to press a dilemma: either v is Kristas state, or she lacks i-access to it. The reason for believing this is not the stuff about personal identity/Humean bundles, but rather that introspection is merely a label for our capacity to know our own minds. If I am putatively of what is in fact another’s state, then it’s not via introspection but telepathy, or suchlike.

  4. REPLY TO JOHN SCHWENKLER — Hi again John, I’ve been weighing two kinds of response—a hard line and a soft line—to your question about whether v occurs in Krista’s brain after all. The idea you’re pushing is that brain regions (and region-ownership) are functionally individuated in a way that would make any brain region that is directly responsible for giving a person visual sensations a part of that person’s brain (and that we need to keep this clearly in mind when considering unusual cases). The question is a very good one, and you and Annalisa are right to press me on it.

    The “hard line” response I would give is essentially just that given above under the “brain individuation” part of my response to Annalisa. Let me give the soft line response here. The soft line grants that some of the principles by which brains and their parts are individuated may be of the kind John and Annalisa have in mind, and that theses would weigh in favor of v being part of both Krista and Tatiana’s brain. However, I would insist that it is very far from a logical or conceptual truth that brain regions should be individuated in that way. Consider people like Andy Clark who defend an “extended mind” thesis (where, e.g., Otto’s notepad or iPhone literally becomes a part of his mind because of its close causal connection with other parts of his mind). They do not conclude that the relevant external apparatus is also a part of the person’s brain. The whole point is that the mind is supposed to extend beyond the brain in these kinds of cases—nevertheless, these are also cases where there is close, fast, immediate (however you want to put it) cognitive access to representations. So, there are plenty of people who don’t seem to share the view that brain parts are to individuated functionally in the way being proposed. (Granted, I then have a disagreement with them about minds…it is interesting to think about the relation between the extended mind thesis and IEM). One could also add (suddenly adopting a tone of modesty) that we as philosophers are not in a very good position to say just how brain regions are to be individuated (though we *might* do better with minds).

    The key point to the soft line response is that all I have to establish is that Introspective Immunity is not a logical or conceptual necessity. Introspective Immunity in its logical/conceptual truth version is a *very* strong claim, after all. And if the principles of brain individuation to which you and Annalisa appeal are not themselves things we can know as logical or conceptual necessities, then it seems we can still conclude that the OT scenario might be as I describe it—and so Introspective Immunity is contingent.

    I admit that this reply still feels a bit unsatisfying to me; I’m still somewhat tempted by the hard line approach. But I’m not sure if I should feel unsatisfied with the soft line. If this is where the argument really lies—in a debate over principles for individuating parts of brains (not minds)—then it seems we have already shaken the Introspective Immunity principle from its a priori perch.

  5. RESPONSE TO JOEL SMITH: Hi Joel, thanks for weighing in on this! I really enjoyed your “old” paper as well—and I think the distinctions it draws are very important. You’re right that my paper (and the NYT video) don’t make clear that Krista has anything more than a kind of blind-sight version of introspection—that, for all we know, she is able to judge that Tatiana sees a doll, without really being introspectively aware (in the ordinary sense) of Tatiana’s experience. However, there are cases described in Dominus’ article that seem to make clear that the introspective link is more profound. One girl starts to drink water as fast as she can and the other grabs at her own throat and says “whoa…” One girl loves the taste of Ketchup the other hates. When the first eats it, the other tries to wipe it off her own tongue (even though there is none on it). So they seem to both feel the sensations—the question is whether they are always both owners of those kinds of states. All that said, the case as a whole is still just a thought example—mainly, as I see it, because the neural situation is poorly understood, and because other (more deflationary) explanations of their behavior have not been adequately ruled out.

    Concerning your challenge: I think it would be unusual to define the *possible* scope or uses of a faculty (or tool) in terms of its ordinary function. Hammers are for hitting nails, but one doesn’t cease to use a hammer when using it for other functions. Hands have various purposes (defined in evolutionary terms) that don’t include playing guitar…etc. So we would need a special reason to think that introspection stops being introspection when it makes one aware of something that is not one’s own mental state (which is not to say that a special reason could not be given). So, when you say that “introspection is merely a label for our capacity to know our own minds,” one can agree, while holding that in extraordinary cases it can make us aware of someone else’s mind as well. Another interesting thing to note with respect to your proposal is that it would allow for situations where Krista thinks she is using introspection but where she really is not (since she is, in fact, aware of Tatiana’s visual experience). Whether she were really introspecting or not would make no phenomenal difference to her.

  6. RESPONSE TO ELIZABETH SCHECHTER — Hi Lizzie – sorry to take so long to get my responses to you..thanks for agreeing to be a commentator. Now I see that the first lines of my paper were needlessly confusing—they (understandably) led both you and Annalisa to wonder whether I was using ‘see’ only in a success term sense. Anyway, when you ask, “Does Langland-Hassan mean to ask us to imagine that Krista enjoys a visual experience as of a bird, but that the experience is Tatiana’s alone?” – the answer is Yes.

    You suggest that I lean too much on OT, since that may still be a situation where the girls *share* ownership of the mental state. But their sharing the mental state was a possibility I explicitly allowed for. I argued against Midas Touch as a sufficiency condition as a means to showing that it is at least possible that they did not share ownership of v.

    Your point is well taken that I may overstate the importance of core psychological features to personhood—it is not as though every one of a person’s beliefs forms an essential part of that person; and it seems that a person could have beliefs that are not well integrated with her other mental states. I was trying to make the best case I could for viewing Midas Touch as a sufficiency condition, and so may have oversold it a bit. The idea was just that a defender of Introspective Immunity could try to say that, for any well-integrated belief, there will be no more relevant criterion for determining whose it is than looking at the network it is a part of, since persons just are such networks. I probably haven’t gotten that idea as sharp as it could be. I’ll keep working on it.
    I certainly agree with your closing remarks: “it is likely that there are a whole host of actually pretty heterogeneous relations that we view as constituting ‘ownership,’ and while these relations may usually stand or fall together, they almost certainly don’t in extraordinary cases such as this one.” I take this to be at least leaving to door open to the kind of possibility I describe. One could see this paper as an investigation as to how different conceptions of ownership can pull apart in perhaps unexpected ways. One point (a little buried) in my discussion is that, once we are willing to allow that a person can own mental states to which she lacks introspective access and that are not well integrated with her other mental states (as certainly seems possible), then brain-based criteria for ownership seem almost inevitable. Then those criteria can pull against others, such as those on which Introspective Immunity is based.

  7. Hi Peter

    I think that hammers, being artifacts with conventional functions, probably provide poor analogies. Consider instead sight. I suggest that the function of visual experience is to provide information about/knowledge of one’s surroundings. We can use this to answer hard cases. Do I see an object presented to me on a live long-distance video relay? I suggest that the fact that the object is not in my surroundings (I realise that’s a vague term) would be sufficient to justify a negative answer. Similarly, suppose that the function of introspection is to provide (non-inferential) knowledge of one’s own psychological states. Consider a set-up in which I am able to make non-inferential judgements about someone else’s psychological states. Could it be that I am introspecting those staets? No, since they are not mine. The suggestion is, then, that we can use considerations of the function of these modalities to answer such questions about what is or isn’t visible/introspectible. This isn’t to deny that we can use those modalities for other purposes, just as you can use your hammer for propping up a wobbly table. I can ‘use’ vision for purely aesthetic purposes, I can ‘use’ introspection to impress my kids (perhaps!). What I can’t do is ‘use’ vision to see through solid objects, or ‘use’ introspection to inspect another’s conscious states. If you like, you can think of this as one of of cashing out Evans’ claim that information links are not enough (but without his own positive view which relies on Russell’s Principle, which I doubt).

    (Hands, I take it, have a function which will be specified at a highly general level – something concerning the manipulation of appropriately sized objects – so I’m not really worried about the guitar example either.)

    Consider now the case at hand. You’ve ruled out the blindsight-like interpretation. So we should suppose that Krista is consciously aware of the visible properties of the toy, its colours, its shape, its egocentric location, etc. Right? I suggest that this is sufficient to show that she has a visual experience of the object. Reason for believing this might be the very point you make against me at the end of your last comment (which, I think, doesn’t follow from the view I outline, since I think that telepathy, or whatever, would be access to another’s mind in the blindsight-like way – you just find yourself making the spontaneous judgement that T sees something red, without having a conscious experiences of redness). Consider two cases, one in which Krista has a conscious experience of a red dot, another in which she has i-access to Tatiana’s conscious experience of a red dot. We have ruled out the blindsight-like interpretation. What is the phenomenal difference between these two cases? I don’t see how you can say that there is one. Doesn’t it follow that there are some conscious experiences for which it makes no phenomenal difference whether or not you are enjoying that experience? If so, I’d be tempted take that as a reductio.

    By the way, as you’ll remember, in Shoemaker’s original papers he offered a regress argument for the claim that at least some self-ascriptions must be identification free (so IEM) if any are to be identification-dependent (so subject to such errors). He supposed that these would include introspection-based judgements. Do you think that there is some other class of judgements that would serve the purpose that Shoemaker thought needed serving, or do you think that his regress argument is unsound?

    I hope that’s all reasonably clear – I realise that some of my earlier remarks weren’t.

  8. Hi Peter,
    No problem, I’ve been swamped as well! Thank you for your reply.

    I didn’t mean to imply that you lean completely on OT–I did see where the Midas Touch principle fits into the dialectic. I guess I just worry that OT is doing some behind-the-scenes work in making the non-ownership seem more compelling than it otherwise would–which would be natural, precisely because there ordinarily *is* so strong a relation between individuating experiences and identifying the subjects to whom they belong. I may be wrong though.

    Thanks for clarifying that you did in fact mean us to imagine “that Krista enjoys a visual experience as of a bird, but that the experience is Tatiana’s alone.” Your reply to me here (and to Analisa Coliva) leaves me confused on another point though.

    You want a case where there is an actual error (a violation of IEM), right? But then it isn’t clear to me what the error is. Of course assuming that “see” is a success verb, then if Krista thinks, “I am seeing a bird,” she has made a kind of error–but one could reasonably say that she’s making the kind of error I am making when I say, “I am seeing a bird” but in reality am only hallucinating seeing one. (So that there is an error, but not by virtue of misidentifying the subject.) If she thinks, “I am enjoying a visual experience as of a bird,” she is correct. You seem to be left with the claim that IEM is violated only if Krista thinks, “This experience as of a bird is mine.” (Since on your account she is not the owner of the experience–it is not within her mind.) But since you now concede not only that she has introspective access to it but that she in fact undergoes the experience (right?) then your requirements for genuine ownership seem implausibly high. You can of course insist upon a brain-based criterion of ownership–but (arguably!) we only have a reason for accepting such a criterion because ordinarily M’s occurring in X’s brain is necessary and sufficient for X to stand in the various relations to M that we think we each stand in to our own.

    You write: “Your point is well taken that I may overstate the importance of core psychological features to personhood—it is not as though every one of a person’s beliefs forms an essential part of that person; and it seems that a person could have beliefs that are not well integrated with her other mental states. I was trying to make the best case I could for viewing Midas Touch as a sufficiency condition, and so may have oversold it a bit.” But I don’t think you oversold it. (Actually if anything, calling it the “Midas Touch” principle sounds to my ear like a bit of a slur–as if it’s magic.) I think you described one possible basis for accepting the condition, but certainly not the only one. A moral basis stands out to me: if Krista literally feels Tatiana’s pain (a possibility that you imply in your comments above is consistent with non-ownership) then it seems to me for instance that Krista as well as Tatiana should be given nitrous oxide when Tatiana has dental surgery, etc. The degree of sharing that you *allow* between the girls, in other words, has various kinds of significance.

    That said I still do believe that in other respects the experience will act like Tatiana’s. So I think I’d be more inclined to say that this is an intermediate or indeterminate case, rather than a case in which non-ownership is determinately true. So we don’t have to resort to a brain-based criterion of individuation in this case; we can just carefully work out the respects in which the state acts like Tatiana’s alone and the respects in which the state acts like Krista’s as well.

    Thanks again for the great paper and ensuing discussion!

  9. Dear Peter, dear all,

    thanks for your detailed reply and sorry for not writing earlier on. I’m rather busy with the start of the term here and I won’t be able to participate in the discussion as much as I’d like to.
    Anyway, I’ll say very quickly that I agree with Elizabeth Schechter’s comments on the issue of visual perceptions. If you mean to talk just about visual experiences then there doesn’t seem to be any EM; if you mean to talk about seeing, it is always possible to distinguish between the visual impression and the perception. IEM would hold in the former case, while EM would arise in the second. Yet this would be significant only insofar as it would call for a qualification on the idea of logical IEM of certain mental states such as perceptions. For these reasons, I actually think that my gloss on your use of “seeing” were the most charitable way of interpreting what you wrote. For you’d come out as recommending an important qualification.
    As to the individuation of brains, I share John’s suggestion of a functional characterization and I’m not entirely sure I got the gist of your reply to it.
    Finally, regarding the individuation of persons, here again I share the intuition, if you wish, pressed by Elizabeth. If one of the two twins *feels* pain, even if it originates in a region of the other twin’s brain, wouldn’t we be inclined to regard it as hers and to treat her too (assuming it were possible to treat them separately) so as to relieve her from it? And wouldn’t this show that we think of that mental state as hers?
    That’s all for now.

    Thanks a lot for the reply and the discussion.


  10. RESPONSE TO JOEL SMITH #9: Thanks for these clarifications, Joel. The point about artifacts is fair enough. We can stick with sight. Suppose we go with your rough characterization of the purpose of vision as being to represent (or make one aware of) objects in your surroundings. It could be that various contingent facts prevent certain kinds of objects from ever entering your surroundings. For instance, things that are currently in some other galaxy are not visible, for contingent reasons. That does not mean that you could not see such a thing if it somehow made it into your surroundings—into your field of view. A reasonable (though, I grant, not inevitable) analogy to introspection would then be: the purpose of introspection is to make one aware of mental states through relatively direct channels. Usually, one’s own mental states are the only mental states in one’s introspective “field of view”; however, there is no reason in principle that someone else’s mental states could not come into your introspective field of view. And one would in such a case be introspectively aware of them in the same way one would see an object from another galaxy if it somehow showed up in one’s living room. Anyway, the issue between us seems to be whether to characterize the exclusive function of introspection as “making one aware of *one’s own* mental states through some direct channel” or “making one aware of mental states through some direct channel.” In the case of vision, the kind of things one can see is not limited to things that are one’s own. So it seems to me that we still need a special reason to think that introspection *must* be viewed differently. (I’m also not sure why you think that OT, or telepathy, would *have to* be a case where the person is aware of the experience in the blindsight way—do you find it impossible that Krista and Tatiana could share ownership of v, with the single token mental state contributing to *both* of their phenomenologies? If so, you might agree with Annalisa, but not Lizzie—John S., which way do you go on this?).

    You move on to ask whether it follows from my view that “there are some conscious experiences for which it makes no phenomenal difference whether or not you are enjoying that experience,” taking it as a possible reductio if the answer is yes. If you enjoy any experience that you feel, then my argument would be that you can enjoy an experience that is not your own. At least, that is the counterintuitive possibility I aim to be considering. That bullet bitten, I avoid the reductio. It makes a phenomenal difference whether you are enjoying an experience; it just might not make a phenomenal difference whether the experience enjoyed is one’s own. Here moral considerations of the kind Lizzie raises become relevant, however. I’ll address that below.

    And, yes, you’re right that Shoemaker’s regress argument against the idea that all self-ascriptions could involve identification is quite relevant here. While I don’t think the argument succeeds, I also don’t have any quick refutation to offer. My general feeling is that the most the argument establishes is that some self-ascriptions have to work *differently* than the kind of self-as-object ascriptions that clearly involve identification (where it is judged that some x is F and then, based on a standing belief that x is y, it is inferred that y is F). That is, there must be two different ways of making self-ascriptions that can then be checked against each other. And the introspective route may be more direct and may not involve conscious inferences of the kind just described. That still seems to leave the door open to the kind of error I’ve described. (This issue is difficult and deserves much more discussion—I’m only giving my general impression here…).

  11. RESPONSE TO ELIZABETH SCHECHTER #10: Hi Lizzie – I am still not sure I follow your worry about OT. The work it is doing is not meant be behind the scenes. It is meant to highlight the fact that the brain in which an experience occurs is a relevant factor in determining whose experience it is. You raise some relevant worries about that idea as you go on, so I’ll just address those (and hope that I haven’t missed something).

    Here is a first point you raise: “You seem to be left with the claim that IEM is violated only if Krista thinks, ‘This experience as of a bird is mine.’….But since you now concede not only that she has introspective access to it but that she in fact undergoes the experience (right?) then your requirements for genuine ownership seem implausibly high.” I grant that she “undergoes” the experience, if that just means that she is introspectively aware of it (the same goes for the words “enjoys” and “feels”). The experience contributes to what it is like to be her. And then I argue that it is logically possible for a person to “undergo” an experience without being an owner of that experience. Your charge is that this makes the requirements for ownership implausibly high because “we only have reason for accepting a brain-based criterion because ordinarily M’s occurring in X’s brain is necessary and sufficient for X to stand in the various relations to M that we think we each stand in to our own” mental states.”

    Your claim here is that the *only* reason we think that the issue of which brain realizes an experience is relevant to assigning ownership is that, normally, whenever a mental state is realized in person S’s brain, that is necessary and sufficient for S to stand in the “various relations” to that mental state that we stand in to our own. To assess this idea, we need to know what the “various relations” are. Presumably you do not think they include: “the mental state is realized in your brain.” And presumably you think they do include “you have introspective access to S.” But, at this point, there is no obvious reason not to run the argument the other way. One could say: the only reason you think that introspectively accessing a mental state suffices to make it your own is that, normally, introspectively accessing a mental state is sufficient for its occurring in one’s brain—and its occurring in one’s brain is what really matters (as evidenced by our willingness to count unconscious, and inferentially isolated mental states realized in one’s brain as one’s own). (By the way, the term ‘Midas Touch’ was meant to be needling, but not quite a slur!).

    That said, I think you raise a nice challenge concerning the relationship between being introspectively aware of a sensation and various moral considerations (Annalisa gestures toward this as well). You note, “if Krista literally feels Tatiana’s pain (a possibility that you imply in your comments above is consistent with non-ownership) then it seems to me for instance that Krista as well as Tatiana should be given nitrous oxide when Tatiana has dental surgery, etc.” Now I don’ t know exactly how nitrous oxide works (on a first order or second order state?), but the larger point seems to be this: just as considerations about coherence in a set of core psychological features are closely linked to notions of personhood, so too are considerations about pain and suffering—the capacity for pain and pleasure are arguably necessary for personhood (and for making one an object of moral concern). This might give an independent reason of the kind I’ve been asking for, for including all felt sensations within the mind of the person who feels (and is introspectively aware of) them.

    Can I say that I am just going to have to think about this some more? As I said, I grant it is a good challenge. For the time being, I note that this kind of consideration seems not to extend to visual sensations of the kind I mainly consider in the paper (experiences as of toy ponies). Visual/perceptual experiences don’t play the same kind of role in moral discourse. It does however put some pressure on my attempt to extend my argument to all sensations.

  12. Hi Peter,

    Sorry to have fallen behind in the discussion (I’m swamped), but very quickly: I don’t think the appeal to the Extended Mind Hypothesis is going to be much help in responding to Annalisa’s and my objection. For part of the point of that argument is simply to *deny* that we should restrict the physical realization to a person’s mind to his or her brain — and if this is a possibility, then we can address your argument by saying it’s possible for one of the twins’ visual experiences to be realized in the other’s brain. (Of course the experiences would then be had by *both* of them, as you suggest.)

    However, perhaps a more promising reply here would be to say that it’s only when a person *consistently* has experiences corresponding to what happens in a certain bit of brain (just as Otto uses the notebook consistently, and not just now and then), that we would say either that that bit of brain is part of her brain (assuming a functionalist criterion of brain-individuation) or that the vehicles of her conscious experiences extend beyond the confines of her brain proper (assuming an extended-mind approach that takes an anatomical view of brain-individuation). I imagine something similar could be said w/r/t some of Joel’s worries about introspection: if one twin *always* has introspective access to the other’s (or “the other’s”) visual experiences, then we might say that they are hers and that she is accessing them by introspection; but if this only happens now and then, perhaps this move could be resisted.

    Sorry if this is unclear — I am shooting from the hip. Thanks everyone for the excellent discussion so far!


  13. Hi Peter,

    Thanks again. My worry re: OT was just that it was prejudicing the matter–that if one didn’t state OT in advance, but described Krista’s situation without mentioning it, one would immediately conclude that Krista was the subject of the/a visual experience. But maybe that’s not surprising; maybe the point is just that, still, there could be enough considerations to outweigh what would otherwise be our immediate assumption, which is that Krista was the “owner” of the mental state.

    On the moral issues you write “… that this kind of consideration seems not to extend to visual sensations of the kind I mainly consider in the paper (experiences as of toy ponies). Visual/perceptual experiences don’t play the same kind of role in moral discourse. It does however put some pressure on my attempt to extend my argument to all sensations.”

    Well, I don’t know, there might be some kinds of moral considerations that were still relevant. (Suppose one twin gets a thrill out of looking at violent images or something, while the other really hates to see them–or even to have the experience as of seeing them!–they always give her nightmares afterward.) On the other hand, could you maybe say that the relevance of moral considerations show that IEM isn’t a *logical* truth…? (Depends on your view on the nature of psychological attributions I guess.)

    To me one of the most interesting aspects of the reality t.v. show starring the Hensel twins (Brittany and Abigail), by the way, is that their oneness-or-twoness is continually treated as a pragmatic issue, to be worked out on a case-by-case basis, in conjunction with other people in their social world. For instance after the twins earn their driving license (one of them controls the steering wheel and one the accelerator and brakes, I think) they and their mother have a conversation about what would happen if they were pulled over for speeding–whether they would both be given a ticket or only the one controlling the accelerator, and their mother just says something like, “Well, you’ll just have to see, and we’ll deal with whatever happens after it happens.” They don’t even appear to have a determinate view themselves about who should be held responsible; it would depend upon context, the details of the case, etc.


  14. “If you enjoy any experience that you feel, then my argument would be that you can enjoy an experience that is not your own”.

    So Krista is consciously aware of a coloured surface in just the way that Tatiana is aware of a coloured surface via her visual experience, but Krista is not having a visual experience.To my mind that’s not so much biting the bullet, but swallowing the pistol, magazine and all.

    “I’m also not sure why you think that OT, or telepathy, would *have to* be a case where the person is aware of the experience in the blindsight way—do you find it impossible that Krista and Tatiana could share ownership of v, with the single token mental state contributing to *both* of their phenomenologies?”

    No, I’m not against sharing. But, as above, I don’t think we should bite your bullet. So, I think that if Krista is introspectively aware of some state that contributes to her phenomenology in the way in which visual experiences do, then she thereby has a visual experience. So, it’s just the same point again, I think.

    I guess my question, given the way you’ve put things above, would be what ‘enjoys’ means. It certainly often gets used in the literature, but it seems to me that it’s typically used interchangeably with ‘has’. But that can’t be right on your picture.

  15. Re: Joel’s comment–
    Yes, I actually didn’t expect you (Peter) to agree that Krista does enjoy the visual experience as of seeing a bird (or whatever). I thought you would state that she forms the higher-order thought, “I am seeing (or enjoying the experience as of seeing) a bird,” by virtue of her “introspective mechanism” having the same access to Tatiana’s (first order) experiences as Tatiana’s introspective mechanism does…. But in fact only Tatiana really enjoys the experience. So, in the pain case for example, only Tatiana would really feel pain, though Krista too would (at least initially) form the higher-order thought, “I am in pain.” Then there’s a clear violation of IEM. But maybe there are other kinds of intelligibility issues with this proposal…?

    (Of course there are likely to be profound behavioral differences between someone who is in terrible pain versus someone who merely believes they’re in terrible pain. Still, even if the Krista/Tatiana case can’t be accounted for in these terms, there might be another hypothetical case that would be…)

  16. Thanks to everyone for continuing to help me think these things through. A few quick (well, we’ll see if they’re quick) replies:

    JOHN: Yes, you’re right that the Extended Mind view cuts against my view in the way you described, but I raised it only to as a means to arguing that issues of mind individuation are separate from issues of brain individuation. Then I can say something like: and furthermore, certain assumptions about mind individuation made by Extended Mind folks—namely, that their “functionalist” criterion applies equally to all kinds of mental states—may not be correct. That said, your idea about a person having intermittent introspective access to another’s experiences is I think relevant and helpful. It does seem to make my view more plausible if one twin only has a kind of intermittent, occasional introspective access to the experiences of the other. For then the experience in question does not seem to have the kind of normal functional role that we would associate with an experience owned by the twin in question. And perhaps this is in fact the case, with each twin normally suppressing her (potential) awareness of the other’s experiences in order to avoid sensorimotor confusion.

    LIZZIE: About the moral relevance of visual sensations. To me it seems that what is bothering her in the violent images case is either the images themselves, or the feelings she has upon seeing them. The visual experience “as of” the images is of course relevant as a causal intermediary between those two things. As such, it does not carry the same moral significance as a sensation of pain. Perhaps it is just because we can imagine someone having the same kind of visual experience and it not bothering them at all that the experience itself is not the morally relevant feature. I like your observation that the relevance of moral considerations *might* show that IEM is not a logical truth—that’s quite interesting to think about.

    I haven’t seen the reality show on the Hensel twins, but I agree that these issues of thought and action ownership are the kinds of things that have to be worked out on a case by case basis. Perhaps as I revise this paper I should add in more “pragmatic” details (things about the interests of the twins, the intermittency of the introspective link, etc.) in describing a possible situation of IEM violation.

    JOEL: I guess I draw the line at the word “having”—Krista enjoys an experience without having it in her mind. The clearest way to put it, I think, is that I want to accept the phenomenological claim—that the state contributes to what it is like to be her in more or less the way introspected sensory states normally do—without occurring in her mind. I should not have said that I was biting a bullet in holding to this view, as that is the main idea I’ve been meaning to defend all along (though it seems no one could believe I was saying that). So I don’t think I’m swallowing a pistol here. I’m saying that there are very unusual situations where being introspectively aware of a mental state—“enjoying it”, having it contribute to one’s phenomenology, etc.—just are not sufficient for its occurring in one’s mind. No one has so far questioned that an experience can occur in one’s mind without contributing to phenomenology or being available to introspection; so we seem to have some other grip on the notion of ownership that does not involve introspectability. So then it is not obvious why *every* case where one is introspectively aware of an experience would *have* be one where that experience occurs in one’s mind. Some reasons have been given for why it would (I considered Neo-Locken ones, Lizzie considered moral ones)—but it seems to be that the request for reasons here is not unreasonable.

    LIZZIE (and all): I think I am finally starting to see why people have been assuming I was advocating a kind of “blind sight” version of introspection (and why most have been surprised when I say that a person could enjoy an experience that does not occur within her own mind). I’ve been playing along with the assumption that being introspectively aware of a mental state suffices to have it contribute to one’s phenomenology. But, of course, one of the interesting things about the debate on introspection is that it can (arguably) be conducted without committing oneself to any specific view on phenomenal consciousness. That is, one can say, “well, I don’t understand phenomenal consciousness, but the question of how a person normally comes to know about his own mental states is something I can approach.” From that perspective, I could argue that, in the case described, there is good reason to think that Krista has introspective access to Tatiana’s experiences—and then just be neutral on the question of whether this must entail that the introspected experience contributes to Krista’s phenomenology in the way that it does for Tatiana. For answering that requires wading into the separate (and perhaps more difficult) question of phenomenal consciousness; here different theories of phenomenal consciousness will return different answers. This would perhaps at least put the defender of IEM in the position of having to defend a strong connection between introspection and phenomenal consciousness; such a connection is not usually assumed necessary for the defense of IEM…is it? I’d be interested to hear people’s thoughts on this.

    It seems I may have ended up where people thought I began. The journey has been helpful for me, at least. Thanks to you all again for participating!

  17. Pingback: The Stone Philosophy Links -

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s