Alienation and Self-Presentation

Presenter: Joel Smith, The University of Manchester.

Commentator 1: Cheryl Chen, Harvard University

Commentator 2: Timothy Lane, Institute of Humanities in Medicine, Taipei Medical University / Institute of European and American Studies, Academia Sinica


  1. Thanks Cheryl and Timothy for the perceptive comments and questions. I’ll kick off the discussion with some very brief, initial responses.


    In fact, I don’t deny that the self is the body. Rather, I’m neutral on that question. The ‘self qua self’ is not a special object, it’s a special way of being aware of oneself. That is, it’s being aware of it in such a way as to ground essentially indexical judgements. Of course, in a sense I deny that there is such a thing.

    On the mirror cases, these are supposed to be non-controversial examples of misidentification. Since I am claiming that there is no possibility of (a certain type of) misidentification regarding awareness ‘from the inside’, obviously I need to draw on outer awareness. Of course, the fact that a mirror is involved is inessential to the examples – Wittgenstein’s initial examples concerning getting tangled-up would do just as well.

    It’s a fair point to say that I haven’t defined ‘from the inside’, and I don’t have one to offer now. However, I would say that in the mean-time we have a pretty good grasp on the notion, by example. I may judge that I am jealous in a spontaneous way, based upon my jealous feelings. That’s ‘from the inside’. Or I may judge it as my wife notes my behaviour and tells me that I’m jealous (and I trust her opinion). That’s ‘from the outside’.

    On the defence of IEM in the face of alienation cases, again, it’s true that I didn’t defend the claim that, say, the hand that I experience as alien is in fact (metaphysically) one’s own. However, I take it that it’s the overwhelmingly plausible view. The most likely way to denying it is to claim that a body part is mine only if I experience it as mine. I’m not going to argue against that here, but I must say that I don’t think it has a great deal going for it (consider those suffering from body-blindness).


    Concerning visual experience, yes I do deny that it is self-presentational, although I didn’t really argue for this in the paper. In a previous paper of mine, I offered a broadly apriori argument for the view that bodily-awareness is not self-presentational (I didn’t use that terminology then), and I think that that argument is just as good (or bad!) for visual experience. In addition, though, a recent alienation case (Zahn, Talazko & Ebert in Psychopathology 41, 2008) suggests that an empirical argument might also be made. On the view that I favour, the contents of visual experiences are not ‘bird ahead of me’, but simply ‘bird ahead’, where the position from which the bird is perceived is not explicitly represented in the state’s content. If one insists that it must be, I would suggest that ‘bird ahead of here’ does the work that self-presentationalists think needs to be done (i.e. makes the contents ‘egocentric’).

    I think that a similar line should be taken on bodily-awareness. That is, the contents of bodily-awareness are not ‘My legs are crossed’, nor are they ‘someone’s leg is crossed’ (as that would, as you rightly suggest, need an identification of myself as the someone). Rather, the content is simply ‘legs crossed’.

    How does this subjectless content get articulated into a judgement including the first-person? That’s a good question, and I didn’t really say much about it. However, the rough idea is that when such subjectless contents are experientially presented, the system relies on the self-tracking nature of the sense-modality in question and, since the subject of the state must be oneself, issues in the judgement that my legs are crossed, or that a bird is ahead of me. So, the content of the judgement is richer than the content(s) of any experience upon which it is based. There would be no deductive route from the one to the other.

    I do indeed think it likely that bodily-awareness, action-awareness and memory are all self-tracking although, again, I didn’t really argue for this. I think that there is some reason to accept this if we reflect on what such forms of experience are for. Bodily-awareness has the function of providing information about the state of one’s bodily, episodic memory of providing information about one’s past, etc. When one has a q-memory that is not derived from one’s own past experience, in a sense one suffers an illusion since it will seem (here I guess I have a cognitive ‘seeming’ in mind) to be one’s own past experience of which one is aware. Ditto for quasi-proprioception. If my proprioceptive system receives information about the state of your body, that does not mean that I proprioceive your body. Rather, it’s an interesting way of causing a proprioceptive illusion in me. I would suggest that the same is true of introspection and the grounds for thinking of introspection as self-tracking are also grounded in its function (and this, incidentally, is one of the places at which I would challenge Peter Langland-Hassan in his fascinating paper on craniopagus twins – I’ll try to post something in his comments box to this effect).

    I think the toughest question for me concerns why we should think that reflection on pathological cases should allow us to infer anything about ordinary experience. The first thing to say would be that, I my view, the very least the alienation cases do is put the onus on the defender of self-presentation, since there are at least some non-self-presenting experiences in the modalities that are claimed by some to be self-presenting. If cogent, the argument would also show that a state can be phenomenally conscious without being self-presenting (contra Zahavi et. al.). There is, however, the possibility that, roughly, states are contingently self-presenting. I need to have a think about how to sink that view!

  2. Joel,
    I apologize for having misrepresented your position concerning the relationship between self and the body, and many thanks for your reply.

    (1) Perhaps I should begin by saying I’m now even more uncertain what to make of “self qua self,” since you unpack it with respect to the grounding of “essentially indexical judgements,” but then proceed to say that “in a sense” you deny that there is such a thing. Hmmmm….I’m not entirely certain what it is that you are denying, or what the implications of this denial might be.

    (2) As for the mirror cases, fair enough. since your position is that there is no possibility of (a certain type of) misidentification regarding awareness “from the inside,” you need to make your point with respect to outer awareness. But then this is one of the reasons why I’m concerned that “from the inside” is left to do too much work, especially since it is left undefined. When the notion is insufficiently characterized, it’s unclear how much work it can do. Your example of the difference is familiar and easy to grasp, as I think is mine–“seeing” fear in the eyes of someone standing in front of me and feeling fear, without recourse to perception of the external world (e.g., something I could feel while resting in the fMRI, presumably while the default mode network is active). I’m happy to allow that these two ways of being aware of the existence of a mental state, fear, are distinct. But it seems that your argument requires “from the inside” to be special in such a way that it can assist with your defense of IEM, and since you don’t adopt the anti-perceptual position, I don’t see what is so special about it.

    (3) After you acknowledge that “it’s true that I didn’t defend the claim that, say, the hand that I experience as alien is in fact (metaphysically) one’s own,” you say that “the overwhelmingly plausible view” is that, despite what is experienced, it is in fact one’s own. I wonder whether we are just on different pages here: For the only somatoparaphrenia patient that I have ever interacted with, I would not want to deny that the hand attached to his body is, in an important sense, his. But “important sense” here is dead obvious and non-contentious. My main concern, however, is with explanation, and how to construct the explanans such that it adequately explains the explanandum. To cite a familiar example, on the assumption that schizophrenics can tickle themselves whereas healthy persons cannot, explanation of this difference seems to benefit by drawing a distinction between self and non-self, despite the fact that all of the action must be happening within distinct bodies. It is in this context, worries about explanation, that I cited the example of recovery of tactile sensation when a patient suffering form somatoparaphenia was cued to expect that someone else’s hand was about to be touched–clearly it is that patient who is reporting the recovered tactile experience, but I don’t understand why (if this is your position) her ability to report the sensation entails that it belongs to her. To repeat my previous example: I can report on the existence of mental states, and nothing about my ability to report those states entails that they are mine. I fear that “overwhelmingly plausible view” is merely the report of a strong intuition, and I believe that intuition is part of the problem, an obstacle, if you will, to explaining rare but revealing phenomena. This is why I often claim that confusion in this vicinity is similar to a complaint raised by defenders of the Extended Mind Hypothesis: just as it is arbitrary to restrict vehicles of mind to boundaries of skin and skull, so too is it arbitrary to say It “is in fact (metaphysically) one’s own” just because it is somehow connected to one body rather than another.

    If I’ve in any way misrepresented your positions, my apologies. I’ll try to remain engaged in this discussion, though the next two weeks will be extremely busy here. All good wishes for the successful development of your ideas!


  3. Hi Joel,

    Thanks so much for a fascinating paper.

    You define the self-presentation thesis as the view that certain experiences – e.g. experiences of bodily awareness – have first-person contents. You then point out that there are cases where people who have the relevant experiences, but who are also suffering from certain kinds of delusion – e.g. alien limb delusion – are not prepared to make first-person judgments on the basis of their experiences. You argue that this observation undermines the self-presentation thesis.

    One way of blocking this argument that Cheryl suggests is to deny that the victims of the delusions have experiences that are relevantly similar to the experiences of those not suffering from delusions. A different way of blocking the argument would be to deny the victims of the experiences are taking their experiences at face value. The fact that a particular experience does not lead me to judge “That is a dagger” does not imply that “That is a dagger” is the not the content of my experience. I might, for reasons good or bad, be refusing to take my experience at face value. Similarly, the fact that a particular experience does not lead a victim of alien limb delusion to judge “My hand is being stroked” does not imply that “My hand is being stroked” is not the content of their experience. The effect on them of their delusion might be precisely to get them to refuse to take their experience at face value.

    I was wondering if you had a way of adjudicating claims about when someone counts as taking their experience at face value that rules out this different, and self-presentation compatible, treatment of the cases you discuss.


  4. Tim:

    1. The significance of the ‘qua’ locution is easiest to see at the level of linguistic representation. If I utter, “JS is in Manchester”, I have said something true of myself. But it need not manifest to me that it is myself of whom I am speaking (I may have forgotten that I am JS). If, on the other hand, I utter, “I am in Manchester”, I have said something with just the same truth conditions (since I am JS). However, in this case it is manifest to me that it is myself of whom I am speaking. We can summarise this by saying that whilst “JS”, in my mouth, refers to me, it does not refer to me qua self. On the other hand, “I” refers to me qua self.

    Now, the proponent of self-presentation, as I articulate that view, claims that a similar distinction holds at the level of experience. That is, that some experiences have content that is first-personal and that they serve to ground qua self representations at the level of thought and language. This is what I deny. Thus, in this sense, I deny that there is any experience of myself qua self. Of course, I don’t deny that some experiences serve as the grounds for first-person judgements. However, I deny that such judgements are grounded by way of their possessing the same content as the grounding experience.

    2. I have to say that I don’t really understand quite what you are getting at with the example of seeing fear in someone’s eyes. Here’s what you wrote,

    “The person who is in a position to introspect on a state and report it is not necessarily the self who experiences the state. If I occupy a unique position in space-time such that I can “see” fear in your eyes and report on it, it does not follow that I am the self who experiences the fear. Indeed, I might even be aware of the fear before you become aware of it, or even if you never become aware of it (in the latter case, allowing for the possibility that the fear was fleeting). How cases of introspecting on tactile sensations or seeing fear in the eyes of another should be understood are largely empirical matters, not matters to be decided by the imposition of an IEM tautological interpretation.”

    You are quite right that when I see the fear in someone’s eyes, I am not the one experiencing fear (unless I am seeing a reflection of myself). However, I find your comments puzzling, since this is clearly not a case of introspection but of visual perception. I don’t know what is supposed to be being settled empirically here. If anything, the example gives a nice illustration of my point (rather than being inconsistent with what I say). Errors of misidentification are possible with ‘seeing the fear in the eyes’ case, in a way that is intuitively implausible for introspective cases (but see Peter Langland-Hassan’s paper for an argument against this).

    3. You accept that there is a “dead-obvious and non-contentious” sense in which the alienated states/body parts are one’s own. It’s that sense with which I am concerned, no other. As it happens, I am attracted to the view that I am identical to my body (with the view that I am constituted by my body being a close runner up). Given this, how a body part is represented in bodily experience is irrelevant to whether it is mine. Thus can we explain why one and only one, say, gallbladder is mine. The cases of memory, action, etc. are less simple but, again, I would want to claim that the question of their metaphysical ownership is independent of their experienced ownership. In the alienated memory case, the memories do, in fact, belong to R.B. despite the fact that things do not appear that way to R.B.

    You say, “I don’t understand why (if this is your position) her ability to report the sensation entails that it belongs to her. To repeat my previous example: I can report on the existence of mental states, and nothing about my ability to report those states entails that they are mine.” Of course, I can report on other people’s sensations, but this is neither here nor there. What we are interested in is reporting on sensations that one feels. You suggestion, then, seems to be that I can feel a sensation without it being mine. This, I think, is wrong. Suppose the state is one of pain. If I feel a pain, this is tantamount to saying that I am in pain. Perhaps it is possible to be in pain without feeling it (although I have my doubts). However, it should not controversial that feeling a pain entails having a pain. I say this knowing that there’s a discussion over at Peter’s paper, in which he denies this. I’m going back over there later to join in properly.

  5. Daniel: Thanks for the useful comment. I’ll get to it tonight, and try comment on your excellent paper early next week (this online conference thing is hard – I feel like Napoleon fighting on too many fronts!).

  6. Daniel:

    Whether to attribute a refusal to take the content of one’s experience at face value is, I think, to be decided on a case-by-case basis. The only cases that clearly involve delusion are thought-insertion and those cases of alien limb in which the subject claims that the limb in question belongs to some other person. These, I take it, are delusory beliefs. Of course, there’s a lively debate about how to read delusions – whether to attribute an unreasonable belief about an ordinary experience or a reasonable belief about an unordinary experience. I don’t really want to take a stand on this other than to suggest that an entirely general answer, one that abstracts from the particularities of the case, is a rather unlikely prospect.

    Although I didn’t have the space to present any detail on this, my own reading of these cases is that the alienation cases involve an ordinary experience which is being taken at face value. That is, ordinarily, the content of one’s judgements outstrips that of one’s experience. So, whilst the content of one’s visual experience, say, is ‘bird ahead’, the content of one’s judgement is, ‘bird ahead of me’. We allow that this transition is rational, despite not being legitimated by the grounding content, for the reason that if I have a visual experience with that content, it will always be correct to self-attribute (in something like the way that we reasonably move from the belief that P to the belief that I believe that P). The alienation cases, I think it reasonable to suggest, involve experience with the same content as ordinary experience, but where the standard, unthinking transition to the self-inscriptive judgement does not occur.

    Two alternatives present themselves: first, Cheryl’s suggestion that the cases involve a reasonable report of an unordinary selfless experience, second, your suggestion that they involve an unreasonable report of an ordinary self-presenting experience. I think that there is some reason to doubt your reading. Consider the description given by R.B. of his memories: “I can picture the scene perfectly clearly…studying with my friends in our study lounge. I can ‘relive’ it in the sense of re-running the experience of being there. But it has the feeling of imagining, [as if] re-running an experience that my parents described from their college days. It did not feel like it was something that really had been a part of my life. Intellectually I suppose I never doubted that it was a part of my life. Perhaps because there was such continuity of memories that fit a pattern that lead up to the present time. But that in itself did not help change the feeling of ownership. (Klein and Nichols 2012, 483)”

    Which reading is best supported here? I would suggest that yours is the least plausible. This does not read as though a delusional subject is refusing to describe their experience at face value. Rather, this is an intelligent and rational adult critically reflecting on their experience and attempting to describe it in an accurate way. Consider another example (that of anonymous vision, not mentioned in my paper but that I alluded to in my response to Cheryl): “D.P. did not see doubled objects in the literal sense – instead he described his sensations as follows: when looking at or concentrating on a new visual object, he is able to see the object as a single object, but that the way he perceived things had markedly changed in a way which he had never experienced before. It appeared to him that he was able to see everything normally, but that he did not immediately recognize that he was the one who perceives and that he needed a second step to become aware that he himself was the one who perceives the object.” (Zahn et. al. 2008, 398). Once more, this does not lend itself to the delusion reading that you suggest.

    In fact, I would say that these details of the two cases also give some reason to doubt Cheryl’s suggested reading. For it is notable that each subject describes their experiences as, in a certain respect, normal. It is the immediacy of their judgements of ownership that is compromised. Now, I don’t pretend that my own reading is mandated by such details (no such description is ever like that). However, I do think that the onus is on proponents of other accounts to justify those alternative readings.

  7. Is the content of first person experience selfless or is it first-personal? I think it is possible to answer ‘both’. Take perception. If I see John kiss Mary, what I see involves John, Mary and the kissing relation. That the relevant event (John kissing Mary) is happening here and now is arguably not something that’s represented in the content of my perception, but rather something that follows from its being a perception. We perceive what is taking place here and now — that’s a feature of the perceptual ‘mode’. What’s conveyed by the mode of the experience arguably does not need to be explicitly represented in the content. So we can distinguish the ‘explicit’ content, and the overall or complete content of the experience. The latter, but not the former, factors in the contribution of the mode. Now if we accept that distinction between two notions of content, we can apply it to the relevant cases. When I know from inside that my legs are crossed, the content of my proprioceptive experience arguably is something like ‘legs crossed’. That’s the explicit content, and it is selfless. (See my *Perspectival Thought*). But the mode of the experience is first-personal: proprioceptive information is bound to concern the experiencing subject, so the overall or complete content of the experience is first-personal, despite the explicit content being selfless. On this view, it is not true that ‘the content of one’s judgements outstrips that of one’s experience.’ The content of one’s first person judgment outstrips the EXPLICIT content of one’s experience, but the overall content of the experience is first-personal, like that of the judgement which it grounds. The difference between the judgement and the experience is that (typically) the self is represented explicitly in the judgment. The judgement makes explicit what is conveyed implicitly through the mode of the experience. (On this transition from implicit to explicit, see my paper on immunity to error through misidentification, in Prosser and Recanati 2012.)

  8. Joel,

    Let me try again:

    Suppose we have two persons, TL and JS. TL might become aware of fear in JS and might also become aware of fear in TL. Although the mechanisms in virtue of which TL becomes aware of these distinct instantiations of fear differ, there is nothing inherent to those mechanisms which conveys anything interesting about the metaphysics of belonging, nothing other than just that the mechanisms must be such that they do not easily misattribute TL’s fear to JS, or vice versa. Had the mechanisms not been so designed, at the species level, creatures like us would not have survived—fear needs to be attributed correctly in most instances in order to trigger fight or flight.

    Note that when I here say TL’s or JS’s fear, I mean that in what I take to be the wholly uninteresting sense that the fear is realized in virtue of activity in the brain, the endocrine system, etc. of two distinct organisms, one named TL and one named JS. If this is all there is to metaphysical belonging, then there is nothing to disagree about. But this is not informative about IEM.

    The somatoparaphenia case that I wrote about and that you grouped with other “alienation” cases posed a unique explanatory problem: the patient, FB, suffered from somatoparaphrenia coupled with the loss of tactile sensation. She believed her niece owned that hand. When, with appropriate controls in place, she was told that she, FB, was to be touched, she felt nothing. When she was told that her niece was to be touched, she recovered tactile sensation.

    Here’s the point about seeing fear in someone’s eyes: I claim that FB is reporting a tactile sensation that doesn’t necessarily belong to her. It is at least an open question as to whether FB’s report is more like TL reporting JS’s fear or whether it is more like TL reporting TL’s fear.

    Back to the explanatory problem: when cued to expect that “FB’s” hand would be touched, FB reported nothing. When cued to expect that “FB’s niece’s” hand would be touched, FB reported tactile sensation. The problem concerns how to explain these distinct phenomena concerning which there is only one difference—the cue—in the causal chain leading up to the sensation of touch or its absence.

    My claim is that the best way to explain these two phenomena is to say that FB only reports what is represented as belonging to someone else, and that this is not significantly different than TL reporting upon what he sees in JS’s eyes.

    My hunch is that something similar is happening in Stanley Klein’s case of episodic memory. All the conditions for satisfying what it is to be episodic memory are satisfied, but it doesn’t feel like it belongs to me. The patient’s report of the events are, so I am inclined to say, importantly analogous to TL reporting fear in JS.

    It seems that one of several matters concerning which we differ is that I peg the interesting metaphysical questions—if there are any in this vicinity—to problems pertaining to explanation, what phenomena are in need of explanation and what need we posit in order to explain those phenomena.

    There are other important empirical issues here that should be flagged—e.g. when TL reports on JS’s fear he feels something, though just what he feels is not always easy to characterize. And when I’m deep in NREM sleep I adjust my posture, unlike those who suffer from congenital insensitivity to pain. Were I to again employ contrastive explanation, as a first pass, I would say that I adjust my posture because I feel pain, unlike those who suffer from congenital insensitivity to pain. I claim to be making an inference to the best explanation, even though deep in NREM sleep it is by no means clear that I could be “feeling” pain in any conventional sense.

    Hope this at least helps to clarify.


  9. Hi François

    Thanks for the comment. I must admit to having put your view to one side, in part in order to present a cleaner contrast between the two views I pit against each other. In fact, I find it slightly difficult to pigeon-hole your position, which may just say something about the inadequacy of my pigeon-holes. Still, my central puzzle concerns what it means to say that something is part of the overall content, but not the explicit content. I think I have a firmer grip on this at the level of language, where we can make a distinction between what is said (whatever that amounts do) and the words uttered. Still, it may well be that your position is something to which I could happily agree. One of the issues would be what, on your view, it is to take the content of one’s experience at face value. If it is to make a judgement that mirrors just the explicit content, then one can consistently claim that the alienation cases involve subjects taking their experience at face value (which I do want to say). If, on the other hand, taking one’s experience at face value involves judging the overall content (complete with first-person, place and time), then I think that the alienation cases pose a challenge to it.

  10. Hi Tim

    I think we disagree about a number of things here.

    When you say that there is nothing in certain mechanisms that conveys anything interesting about the metaphysics of belonging I think that this is in a sense right and in a sense wrong. I take it that you are talking about vision and introspection. I agree entirely concerning vision. With introspection, I agree that we should not define what it is for a mental state to belong to me in terms of introspection (if that was your point). However, I would claim that the converse is not true. That is, it makes no sense to suppose that I could introspect someone else’s mental state (pace Peter L-H).

    When you say that there is a boring sense of ‘belonging’ which is not ‘informative about IEM’, I want to register an agreement and a disagreement. No, it’s not informative about IEM, but then nobody has ever said it is (as far as I know, and understand the claim). However, it certainly IS relevant to the question of IEM. IEM involves the impossibility of taking something other than oneself to be oneself. That a psychological property is mine (in the boring sense) is exactly relevant here. Errors of misidentification are those in which a property that (boringly) belongs to one object is taken to (boringly) belong to another. At least, that was how Wittgenstein, Shoemaker, Evans, etc. all understood it. So whilst this sense of belonging perhaps does not illuminate IEM, it is important to an adequate understanding of what IEM is supposed to be.

    You say that, “It is at least an open question as to whether FB’s report is more like TL reporting JS’s fear or whether it is more like TL reporting TL’s fear.” But, given what I say above, I obviously must disagree with this. The grounds of her judgement are introspection. She FEELS the sensation. This is very different from the visual case, and (I say) it makes no sense to suppose that she could be introspecting someone else’s sensation. I don’t see that anything about this (admittedly very puzzling case) suggests anything to the contrary.

    I think that we agree on a number of things though. For example, I’m happy to accept (in leiu of a better explanation), that “the best way to explain these two phenomena is to say that FB only reports what is represented as belonging to someone else, and that this is not significantly different than TL reporting upon what he sees in JS’s eyes.” But that’s consistent with my line on IEM etc. The similarity with the visual case is that the state is represented as belonging to another, but there the similarity ends. In particular, it is not true (as it is in the visual case) that the state does belong to another.

    Perhaps you are right that your views on explanation and metaphysics (if I understand them) show a difference between us – perhaps you want to draw a metaphysical conclusion concerning the ownership of mental states on the pathological cases. I would be highly reluctant to do this. As far as I can tell, the cases require us to say something about how certain states are represented (as belonging to another, or as belonging to nobody), but do not mandate any claim about whether the states in question are in fact the subjects’ own.

    I’m not sure as to the connection with the sleep/pain case. However, I’m happy to agree with you for the sake of argument. I think that if I feel a pain, then the pain is mine. But I’m not committed to the converse, perhaps I have pains that I do not (in the usual sense) feel. Perhaps.

  11. Joel,

    I’m reluctant to get into exegetical matters here but concerning the Wittgenstein-Shoemaker version, Shoemaker wrote: “there is no question of recognizing a person when I say I have (a) toothache. To ask ‘are you sure it is you who have the pains would be nonsensical.”

    FB introspects and reports on the instantiation of a tactile sensation. But it would not be nonsensical to ask of her, “are you sure it is you who have the tactile sensations?”

    Shoemaker wrote: “it cannot happen that I am mistaken in saying ‘I feel pain’ because, although I do know of someone that feels pain, I am mistaken in thinking that person to be myself.”

    My claim is that cases like that of FB (actually, other alienation cases more clearly so) evince the possibility of being mistaken, especially since Shoemaker repeatedly emphasizes that his concern is with the self-as-subject, not the self-as-object (the boring sense of belonging?)

    Shoemaker wrote: “…in being aware that one feels pain one is, tautologically, aware, not simply that the attribute feel(s) pain is instantiated, but that it is instantiated in oneself.”

    To the extent that he is writing of the self-as-subject, I claim that cases like that of FB (and other cases of alienation) show that “feels x” and “in oneself” are not necessarily linked.

    And Shoemaker is not bound to introspection, because he applies the same idea to claims about visual perception and action.

    I think too that you overplay the “she feels the sensation card.” Our awareness of instantiated mental states tends to be bound to feeling, even if we don’t attribute those states to self: my awareness of your pain is, at least in part, made possible by feelings whose neural correlates overlap, to a considerable degree, with the neural correlates of pain that are experienced as mine.

    Of course the feelings differ somewhat, but even the degree of difference varies greatly depending, let us say, upon where one falls on an empathy scale (or some other measure). Although this might sound like caricature of politicians, there are times when one person can feel another person’s pain.

    Although I have not carefully followed the craniopagus twins discussion here, let me try to relate this point to them: if one of the two twins who are linked by the thalamus eats a piece of chocolate, the other twin feels something, seemingly a taste sensation. But she then wants a piece of chocolate for herself. One way of characterizing this is to say that the latter twin introspected and experienced a taste sensation, but the experience was somehow different than what she would have experienced had the chocolate been placed into her mouth. Arguably, the taste that she was aware of via introspection didn’t belong to her (self-as-subject). Or, so I claim, this is at least one viable, even testable, hypothesis.

    My point is straightforward: there is nothing about feeling or introspection, per se, that informs in a non-boring way about belonging. Also, although I am loath to engage in exegesis, for I believe that in most instances it is more of an obstacle to than an aid to fruitful inquiry, I think the boring sense of belonging is not so relevant to the Wittgenstein-Shoemaker version of IEM as you think.

    All the best,

  12. Hi Joel,

    You wonder what it means to say that something is part of the overall content, but not the explicit content, and say you have a firmer grip on this at the level of language. But take (for example) episodic memory. Suppose I have a visual memory of John kissing Mary on the beach. What I visually remember is that event: John kissing Mary on the beach. The fact that I visually witnessed that event in the past is not part of the explicit content of my memory, but something that follows from its being an episodic memory. I can imagine an alienated subject who has ‘visual memories’, but without the feeling of having witnessed the remembered event in the past. Such an alienated subject would be exactly like someone who ‘perceives’ events around him or her, but without the ‘feeling of presence’ which normally characterizes perception. The feeling of presence, like the feeling of ‘mineness’ in first-person cases, is part and parcel of the experience and belongs to its overall content even though it’s contributed by the experiential mode.

    You ask what, on my view, it is to take the content of one’s experience at face value. Is it is to make a judgement that mirrors just the explicit content, or does taking one’s experience at face value involves judging the overall content? The latter, I think. On my view the alienated subject (e.g. the subject who ‘perceives’ but without the relevant feeling of presence) has an experience with a *different* (overall) content than ‘normal’ subjects, even though the explicit content is the same. I therefore reject the presupposition that normal subjects and alienated subjects have experiences with the same content.

  13. Hi all,

    It is going to be a while (if ever) before I catch up on all these comments, esp. the exchange between Joel and Tim. But here are two quick questions:

    JOEL: You argue against the self-presentation thesis by noting several pathological cases where experience seems to be subjectless. But how does this show that the corresponding forms of experience aren’t self-presenting in the normal case?

    FRANCOIS (and also Joel, I suppose): You say that the explicit content of my memory of (say) John kissing Mary is not that “I saw John kissing Mary” or even “I remember John kissing Mary”, but just that “John kissed Mary”. What, though, about the case where I remember *myself* kissing Mary? Would you say that the explicit content of that memory experience is just “kissing Mary”? That seems like a strange position — of course I might *say* “I remember kissing Mary”, but it seems like the memory itself must have a propositional content.

    I’ll have more to say shortly depending on how these questions are answered.


  14. John: My initial answer to this is in #1 and #6 above. I realise that it’s not entirely satisfactory. At the most general, the claim is that some strong version of self-presentation is false (the view outlined by Zahavi and others), and that there is reason to think that weaker versions of self-presentation have some explaining to do.

    John again: On memory, it’s important to distinguish between episodic and declarative memory. I don’t think there’s any need to see the content of episodic memory as a proposition, rather it’s a retained experience.

    François: Lots of things follow from the fact that I visually remember John kissing Mary. For example, it follows that I exist, it follows that I have eyes, it follows that 5+7=13. But I wouldn’t want to count any of those things as part of the ‘overall content’ of my memory. So we need a principled way of getting the right things into the contents. In the case of linguistic representation, we perhaps can rely on linguistic convention, but I’m still not sure what the analogue is in the cases of memory, perception, etc.

  15. Hi Joel,

    Okay, this is helpful. I certainly agree with you that self-presentation isn’t *always* a characteristic of experience, visual or otherwise. As you note, though, that isn’t a reason to think that experience is never self-presenting at all — though I agree with you and Shoemaker that the self-presentingness of experience (if there is such a thing) couldn’t be part of what explains our capacity for first-person thought.

    Regarding the content of memory, I suppose you’re right that it needn’t be propositional. But one thing I have always found puzzling about views like Francois’s (and now yours) that deny that perceptual experience is ever self-presenting is the way it suggests such an enormous asymmetry between the presentation of one’s own properties in perceptual experience and the presentation of the properties of other people. That is, you would say e.g. that when I kiss Mary (or remember kissing Mary), my experience — and here I mean: not just my experience “from within”, but also what things are like for me e.g. visually — is not of a relation between two people, as it is when I see someone else kissing Mary, but instead has a sparser content. Or again, to modify one of Francois’s examples above, when I cross my legs and look down at them, the content of my experience is not that *my* legs are crossed, in contrast to the case when I look at you and see that Joel’s legs are crossed. These questions are really the subject-matter for a much longer paper (and as you know, I have written at length elsewhere in defense of the idea that visual experiences can present one’s own motion in an explicitly first-personal way), but I wonder: Isn’t it counterintuitive that there should be *such* a stark difference between the way these properties and relations are experienced first-personally and the way they are experienced when they involve others?

    Sorry again if I am being unclear — it wouldn’t be the first time!


  16. Hi All,

    This is an extremely fascinating discussion, and it will take me a while to work through everything. I have some (half-baked) thoughts on how a certain reading of Zahavi and others could potentially be reconciled with a view like François’, and perhaps Joel’s (which is very similar to my own, by the way) that I would like to hear your thoughts on, but I will have to get to that later.

    For the moment, I just wanted to briefly reply to Joel’s and John’s last comments:

    Joel: With regard to your last reply to François, I suppose the mode is supposed to somehow contribute to the phenomenology of the experience. The same is not true for all the other facts that follow from the explicit content of my experience. I take it that this is why, on François’ view, the former enters into the overall content, while the latter don’t. That said, I am not clear on how exactly we are to understand the way in which the mode does contribute to the phenomenology (this is related to my point about Zahavi, which I will have to come back to).

    John: I find it interesting that you find the asymmetry you describe counterintuitive. After all, aren’t we often at pains to stress the asymmetry between the way we experience properties and relations first-personally and the way they are experienced when we ascribe them to others? (Tugendhat speaks of the “epistemic asymmetry” here, for example, though one needn’t put it that way). So I wouldn’t hold it against the view defended by Joel and François that it implies such an asymmetry; quite the contrary. I think if anything it is a positive feature of such a view that it can potentially account for this asymmetry. Perhaps we have just wildly diverging intuitions here – it would be interesting to hear more about why you think this asymmetry is counterintuitive, and I’d also be interested to hear what others think about this.

  17. Dear John, Joel, and Kristina,

    John: The content of experience is hardly ever propositional, I would think. When I see, or remember, John kissing Mary, the time of the event is not part of the explicit content. The content of the perception, or memory, is a certain scene (possibly the *same* scene if the memory inherits its content from the perception) but when does that scene happen or did it happen? That, I take it, is determined by whether the experience is a perception or a memory. If it’s a perception, the scene is happening now before me. If it’s a memory, it happened some time in the past, and I was there to witness it. (I would say roughly the same about the place of the event as I’ve just said about the time). We know, of course, when an experience of ours is a memory and when it’s a perception. And we use that piece of knowledge to ‘complete’ the content by injecting the contribution of the mode.

    Of course, as Joel points out, not everything that follows from the fact that we’re having the experience we’re having goes into the overall content of the experience. But, as Kristina points out, I take the role (or one of the roles) of the experiential mode to be that of providing what’s needed to make the (explicit) content of experience truth-evaluable: we need a time, a place, and (when the explicit content of the experience is a property we self-ascribe, as in proprioception or agentive memory) a subject. When I see that it’s raining, the content of my visual experience is not propositional, it’s a property of places (the property a place has iff it’s raining at that place) or possibly a relation between times and places (a relation that holds between a place and a time iff it’s raining at that place at that time). For the experience to be truth-evaluable a place (and a time) must be provided. On my account, the experiential mode provides them, and it is its role to provide them. This, as Kristina said, takes care of Joel’s worry: “In the case of linguistic representation, we perhaps can rely on linguistic convention, but I’m still not sure what the analogue is in the cases of memory, perception, etc.” The analogue in the case of memory, perception etc. is the role (the function) of the experiential mode.

    Now what about the asymmetry deplored by John? John writes, reporting my view (and Joel’s): “when I cross my legs and look down at them, the content of my experience is not that *my* legs are crossed, in contrast to the case when I look at you and see that Joel’s legs are crossed. … Isn’t it counterintuitive that there should be *such* a stark difference between the way these properties and relations are experienced first-personally and the way they are experienced when they involve others?” Kristina rightly responds that we need an asymmetry if we are to capture the contrast between the two perspectives: ‘from inside’ and ‘from outside’. We need to distinguish two ways in which the subject can be involved in the overall content of his or her experience: the implicit, subjective way and the explicit, objective way. Suppose I see myself from outside (say, in the mirror) kissing Mary. In this case I am both the viewer (not explicitly represented in the content of the experience, but implicitly involved qua subject of the visual experience) and one of the characters explicitly represented in the scene (the kisser). There are also cases in which I am not a character in the perceived scene: I see John kissing Mary, and am only involved implicitly as the viewer. There is indeed a deep asymmetry : only the subject of experience (not the other people) can be involved implicitly via the mode; but the other asymmetry -— that which John deplores — is superficial: it’s not only the other people which can figure explicitly in the content of my experiences, *I* can also figure explicitly, as a character, in the content of my experiences.

  18. Hi Joel, thanks for your response to my comments. Your clarifications were very helpful!

    One thing I’m still wondering about: how is your claim that the content of a bodily experience is something like “legs crossed” supposed to square with your claim that quasi-perceptual (or quasi-proprioceptual) experiences are illusory? In this scenario, there are some legs that are crossed in exactly the way the experience represents. Why wouldn’t that make the experience veridical?

    If we want to say that the quasi-perceptual experience is illusory, we seem to have three choices:

    (1) We can hold on to the self-presentation thesis and say that the content of the experience is “my legs are crossed.”

    (2) We can deny that the content of the experience should be understood in terms of accuracy conditions.

    (3) We can adopt what I take to be Francois’s position and distinguish between two kinds of content: the explicit content of the experience (“legs crossed”) and the overall content (“my legs are crossed”). The overall content of the experience determines the accuracy conditions, so in the quasi-perception/quasi-proprioception case, the experience is inaccurate.

    I take it that you are inclined to go with (2) and say that the functional role of the type of experience (whether it’s perception, proprioception, introspection, etc.) contributes to the accuracy conditions. Is that right?

  19. A few very quick thoughts:

    Tim: As I pointed out in the paper, Shoemaker’s formulations are not equivalent. The ‘makes sense’ formulation is indeed challenged by the alienation cases, the knowledge formulation isn’t. Incidentally, note that Shoemaker’s claim that, “…in being aware that one feels pain one is, tautologically, aware, not simply that the attribute feel(s) pain is instantiated, but that it is instantiated in oneself” is clearly true. Notice what he says, in being aware THAT ONE FEELS PAIN one is aware that it is instantiated in oneself. This is a tautology. Of course, Shoemaker probably also thinks that feeling a pain necessarily involves feeling it as one’s own. Here, I agree with you that he is wrong. None of the cases under discussion show, on the other hand, that it is possible to feel another’s pain, since there is no relevant other on the scene.

    John: Like Kristina and François, I agree that the asymmetry is well motivated. I would add that a similar case, that of sensory imagination, seems to be more smoothly understood on this picture. For we can imaging being someone else, Buster Keaton say, kissing Mary. If the sensory content were self-presentational, then one would have to additionally imagine that I am Buster Keaton which is, arguably, not a possible state of affairs. The subjectless view simply requires that I imagine that the the subject of the kissing experience is Buster Keaton.

    François: You say, “as Kristina points out, I take the role (or one of the roles) of the experiential mode to be that of providing what’s needed to make the (explicit) content of experience truth-evaluable: we need a time, a place, and (when the explicit content of the experience is a property we self-ascribe, as in proprioception or agentive memory) a subject.” But I don’t think these all have the same status. Yes, we need time and place to get a truth-evaluable content but, at least in the case of proprioception, we don’t need the first-person. Completing “_ legs are crossed” with “this body’s” rather than “My” gets us a true/false proposition. So it’s still not obvious to me why we should suppose the first-person to get into the experiential content (this concern is less concerning for different modalities).

    Cheryl: Thanks, tough question. As it happens I am tempted by 2. However, I’m not sure that I’m forced that way by my rejection of both 1 and 3. I’m shooting from the hip a bit here — I can say (perhaps!) that my experience is accurate if the legs of which I am aware are crossed. But I need to accept that, in the quasi-proprioceptive case, I am aware of any legs. Compare: a system is set up such that a red square in the neighbouring room causes me to suffer a visual experience as of a red square. I deny that I am visually aware of the red square in the neighbouring room, and this denial is based on considerations regarding the function of visual awareness. Thus, despite the fact that my experience represents (I’m assuming for the sake of argument) that ‘that square is red’ and it is, in fact, caused by a red square, my experience is not veridical. For the ‘that square’ fails to pick out any square. Similarly, in proprioception, the content is not the open ‘some legs are crossed’, but ‘legs crossed’. This only manages to attribute crossedness to any legs if those legs are my legs. If this is right, then there is some sense to be made of the idea that the content of my experience is both exhausted by and exhausts its accuracy conditions, despite not being a full proposition. This thought rests heavily, perhaps, on a distinction between enabling conditions and accuracy conditions, but it seems to me that we need that already — for ‘my legs are crossed’ to be true, it has to be the case that I exist, for example, but we shouldn’t think of that as part of the content/accuracy conditions of the experience. Rather, it’s something that is presupposed by the ‘my’ having a referent. Something similar goes for legs, or at least that’s my story tonight.

    Whew! This is hard work.

  20. I hope to have more to say soon about what I think is wrong with this particular way of understanding the distinctiveness of the first-person perspective, but very quickly: Joel, I don’t think the appeal to imagination helps very much to show that sensory content isn’t self-presentational. First, it could be self-presentational sometimes but not always. Second, is there any reason to think we can’t imagine impossible states of affairs? E.g. I can imagine disproving Goldbach’s conjecture (or proving it, as the case may be), or having been born in the Middle Ages, or … — so why can’t I also imagine being Buster Keaton?

  21. Joel,

    At some point, perhaps now, I’ll be bowing out of the discussion. Simply, as is true for all of us, much too busy. But in addition to that I am certain we all realize that leveling criticisms at the work of others is much easier than drafting new manuscripts. For this reason, and because I think it is just good manners, I prefer that you have the final word. Just in case I do disappear from the discussion, please allow me to say that I’ve enjoyed the give-and-take and have benefited greatly from it. Joel, fight the good fight.

    Possibly my final responses:

    1. As for Shoemaker’s formulations, agreed, they are not equivalent. This is one among several reasons that I’m reluctant to get into exegesis here. I’m certain we share a concern with achieving clarity, and maybe even saying something true, new, non-trivial, constructive, and relevant to human psychology.

    2. As for Shoemaker’s tautology claim, to the extent that it is treated as nothing more than a claim about q relationship among terms and that it is not taken as being informative about human psychology, I’m comfortable with it.

    3. As for alternative Shoemaker formulations and cases that involve an “other on the scene,” consider (a) The formulation: “for a statement “a is Φ” to be erroneous through misidentification relative to the term “a” is to allow for the following possibility: “the speaker knows some particular thing to be Φ, but makes the mistake of asserting ‘a is Φ’ because, and only because, he mistakenly thinks that the thing he knows to be Φ is what ‘a’ refers to.” (b) The “other”-involving case: In the paper that you cite, I described the “Body Swap Illusion”—briefly, cameras, head-mounted displays, and hand-squeezing can be used, when a participant is interacting with the experimenter, to invoke the feeling that participants shake hands with themselves. Of course, from the third-person point of view, participants are actually shaking hands with the experimenter. (c) How does this “other”-involving case relate to this (3a) formulation of IEM? One of Shoemaker’s prototypical examples of self-as-subject is “I am waving my arm.” And, according to IEM, it cannot happen that I am mistaken in saying “I am waving my arm” because although I do know of someone that is waving his arm, I am mistaken in thinking that person to be myself. I am necessarily aware that I am, myself, waving my arm. The same should, so I claim, apply to hand shaking. But the body swap case shows that having an experience of hand-shaking does not guarantee that the agentive experience cannot be misrepresented. Here, I contend, the best way to explain the illusory state is to say that participants misattribute agency. Given the above formulation of IEM, however, mistakes of this type are not possible: whenever I say “I am shaking hands” it cannot be the case that I am mistaken in thinking that the person who is shaking my hand is me. But that is what happens here.

    4. How does this (3) relate to your concern with self-misidentification? By way of analogy you cite the famous Duck Soup example wherein the character mistakenly thinks an apparent mirror-reflection is self. You then say that “it is natural to think that when it comes to judgments based upon bodily-awareness, action-awareness, memory and intellectual experience (self-misidentification) errors are not possible,” and “it should be reasonably clear that alienation cases are not errors of self-misidentification. They are, rather, errors of other-misidentification.” Let us allow that alienation cases might not be errors of self-misidentification. But because, like Shoemaker, you include action-awareness, it seems you must be able to address Body Swap Illusions (and their like). My claim is that Body Swap Illusions are instances of self-misidentification: just as Groucho thinks Chico is self, so too does the participant think the experimenter is self.

    5. More generally, I claim that Problems of Belonging include two types: (a) instances of belonging wherein (from the 3rd-person point of view) the mental state, body part, or action could not belong, and (b) instances of estrangement (what you call “alienation”) wherein (from the 3rd-person point of view) the mental state, body part, or action does in fact belong. Body illusions, as with the example cited above and as with the famous rubber hand illusion, are instances of (a). Somatoparaphenia, the case discussed in previous comments, is an instance of (b) or, more generally, estrangement (alienation). Putting aside, for the nonce, estrangement/alienation, I think Type-A cases of belonging pose problems for your claim that “it is natural to think that when it comes to judgments based upon bodily-awareness, action-awareness, memory and intellectual experience (self-misidentification) errors are not possible.” Indeed they are possible, and it is that naturalness of thought that I contend impedes understanding of these phenomena. The intuitions that are extremely reliable in quotidian circumstances, and that serve as the basis for various philosophical formalizations, are sometimes confounded in pathological, aberrant, and illusory cases.

    6. My comment about “feeling another’s pain” was much too glib. Let me try to state my position more clearly: it is not contentious that TL’s awareness of JS’s states is in part due to TL’s having feelings that to some degree correspond to JS’s states. What distinguishes my states from yours in situations wherein I am aware of your states is “belonging”. How do I unpack this notion? There is a lot to be said here, but I think much of the action is occurring at the sub-personal level. For example, some of Northoff’s work on distinguishing self from other, with respect to patterns of activity in the cortical midline structures, is highly relevant.

    All the best,

  22. Okay, so here in brief is why I think that we shouldn’t explain the distinctiveness of the first-person perspective by denying that sensory experiences are ever self-presentational.

    1. First, I am moved by something similar to what Cheryl says in her #18 above. That is: if we deny that the proprioceptive experience of my crossed legs has a self-presenting content, then how are we to explain the difference between the ordinary, non-alienated experience of my crossed legs and the corresponding alienated experience that is had by someone suffering from alien limb syndrome? (Joel, perhaps you say something about this in your paper or in some of your comments above — if so, my apologies.) Certainly *a* natural explanation of the difference is that in the non-alienated experience, my legs are represented *as mine*, whereas in the alienated experience they are just some legs that I happen to experience from within. What alternative explanation would be given by those of you who deny (even a weak version of) the self-presentation thesis?

    2. Second, and returning to my suggestion above that a view like the one in question posits *too much* of an asymmetry between self-experience and the experience of others, here are three ways to explain what I am after.

    a) First, it certainly doesn’t seem that denial that sensory experiences are ever self-presenting is *required* to explain how the first-person perspective is distinctive. For example, I think that the distinctiveness of this perspective comes out in the fact that it makes it possible for there to be judgments about oneself that are immune to error through misidentification: thus when I have an experience from within of my legs being crossed, or a visual experience of my motion or of the location of something with respect to me, there is no question of my being wrong about who it is that has these properties (or stands in these relations). By contrast, my experience of others does not underwrite judgments that have this property, except ones using demonstratives like “that person”.

    b) Second, and given the above, I submit that the burden of proof in this debate is on those who wish to say that we can *never* have sensory experiences of properties as belonging to ourselves, except in strange cases where e.g. we recognize ourselves in mirrors. All else being equal, it seems best to preserve the idea that there are *some* real commonalities between self-experience and the experience of others, if only to put us in a position to resist various bad Cartesian arguments. So what is there to motivate this radical proposal? I would understand it if this were the only way to explain what is distinctive about self-experience, but as I’ve just said this seems not to be true. I would also understand if it were the only way to explain why certain first-person self-ascriptions were IEM (Francois, wasn’t this once your view? That a self-ascription can be IEM only if it is based on a state whose content is not explicitly first-personal? But you’ve dropped this commitment due to some of Daniel’s arguments, correct? — please someone let me know if I am wrong here), but once again this does not seem to be the case. And then there is the issue I raise in #1 above.

    c) Third and finally, I worry sometimes that those who deny that sensory experiences are ever self-presentational are not in a position to explain how we can experience *the same properties* when they are properties of ourselves as when they are properties of others. Take e.g. the case of self-motion. According to the view in question, when I experience my motion “from within” (either visually or in some other way), my experience is not an experience of *my* motion, but rather of — well, what? I can see two possibilities:

    – Either I experience the motion of “something”, but that this something is myself is not part of the content of the experience. I am not sure how to make sense of this idea, and indeed I have argued elsewhere that it is probably incoherent.

    – Or what I experience is not the motion *of anything*, but rather just *motion*, presented in a distinctively first-personal way. But then I am not experiencing the *property* of motion, since plausibly it’s possible to experience a property only as instantiated by some object. Which then leads to the conclusion that while I can experience the motion of other people, I simply cannot experience my own: not because I can’t experience motion “as mine”, but rather because I can’t experience motion *at all* when it’s my motion I experience, as the experience I have when (as I would put it) I experience myself moving doesn’t even include the same property as when I experience the motion of someone else. And this seems … wild.

    (Note that the same holds for relations: e.g. when I see that the cup is to your left, what I experience is a two-place relation involving you and the cup, but when the cup is to *my* left, this view commits us to saying that I don’t experience a relation at all, unless it’s a relation between the cup and my point of view — but it’s not clear why we should think that a PoV can figure as an element in my experience whereas *I* cannot.)

    Alright, I’ve gone on long enough. Standard apologies apply w/r/t my lack of clarity. This has been a great discussion!

    [N.B. I’ve made a couple of minor corrections to this comment, filling in words that were missing in the third-to-last paragraph.]

  23. Thanks John, those are some great challenges. Let me attack your 1 first, then come back to the others.

    Here’s what I said above in response to Daniel: “The alienation cases, I think it reasonable to suggest, involve experience with the same content as ordinary experience, but where the standard, unthinking transition to the self-ascriptive judgement does not occur.” (typo corrected).

    The difference then is in what transitions the subject naturally makes/is disposed to make from experience to judgement (I don’t assume, though, that such transitions are the only relevant ones – I suppose also that relevant transitions may occur between contentful states that bypass judgement but “presuppose” that the experienced states are one’s own). I’ve not made any claims about why such a transition does not occur. This is, in large part, as I take it to be an empirical question (concerning the localisation of the relevant deficit) that I’m not really qualified to answer (all accounts that I’ve seen seem somewhat controversial).

    I take it that your own interpretation does make such a claim. It’s that the alienated experiences have a different content. My reasons for assuming this to be the less plausible interpretation were, primarily, that a number subjects’ descriptions of alienation cases seem to report that the experience is just like an ordinary memory (or whatever), just that it’s not their own. I admit that this evidence is both sketchy and disputable. I think I was also probably swayed by the fact that self-presentational content is often put forward as being either required to explain the subjectivity of conscious states or required for the explanation of some other phenomenal feature (e.g. that visual perception is egocentric). I think that it is arguable that alienation cases show such claim to be false. The question is then whether self-presentational content is required to explain some other feature of experience. The obvious answer, and I take it yours (but also there’s your comments under 2, which I’ll come back to), is that it explains the sense of ownership. But this, so suggest at least, can be explained by the disposition to move to the relevant judgement (I’ve putting to one side, just for simplicity, issues about the sense of agency. In fact, I suspect that many of the alienation cases may be partially explained through a lack of this. However, I think/hope that this can be interpreted in a way consistent with the overall position that I’ve been sketching).

    Finally, something I didn’t say in the paper, but which I think carries some weight for me. Self-conscious self-referential content is a pretty sophisticated phenomenon. Any view that attributes it to states that, after all, we share with many much less sophisticated creatures carries, for that reason, some burden.

  24. Hi Joel,

    That does help to clarify things. One thing that I think gives reason for thinking that the alienated experience may have a different content than the non-alienated one is that there’s a common — though perhaps not universal? I am no expert on the literature — tendency among those who work on delusions to treat them as the product of experiences that have non-standard features. (I am thinking in particular of a paper by Nora Breen, Max Coltheart, and others that was published in 2000 in Mind & Language.) What motivates that view is the observation that many delusional patients don’t seem to have any trouble making reasonable inferences in everyday life; indeed, if anything their problem is that they are *too* rational, as they start from a certain premise — e.g., “That is not myself that I see in the mirror” — and move from there to a host of increasingly bizarre beliefs that follow from it. Thus the suggestion is that the root of these delusions must be something strange that is happening at the experiential level, which then justifies (to a degree, anyway) the delusive beliefs that follow. Am I right that this account of delusion is incompatible with your view? And if so, can you say more to motivate an alternative one?

    Re: your observation that subjects report a similarity between alienated experiences and non-alienated ones, I would note that it’s compatible with the self-presentation thesis that the contents of these experiences will be similar in lots of respects. E.g., when I experience the movement of my hand, I also experience the movement of *a* hand — and the alienated subject will share this aspect of my experience, though without the further aspect wherein the moving hand is experienced as my own. (Alternatively, if Tim is right then my ordinary experience will *lack* a feature that is present in that of the alienated subject — I am inclined to think that this analysis of belonging could actually be squared with the self-presentation thesis, but have to think more about it.) It would be surprising to me if this weren’t enough to account for these reports of phenomenological sameness.

    Your concern about attributing self-consciousness to non-human animals is more serious, though. I don’t have a worked-out view to offer here, but have had some things to say on the subject in Katja’s thread!

  25. A quickie (as I’m in the middle of cooking for the kids!): Are my suggestions inconsistent with Breen & Coltheart’s account of delusions? I hope not. See my comments in #6. Many alienation cases involve no obvious delusion (although some do). The failure to make the automatic transition to self-ascriptive judgements isn’t, I would say, itself a delusion. Rather, if you look at what RB says about his memories, the problem seems to be that the automatic self-ascriptions have to be arrived at via reasoning/inference. Hence the alienation.

  26. John, you ask:

    “Francois, wasn’t this once your view? That a self-ascription can be IEM only if it is based on a state whose content is not explicitly first-personal? But you’ve dropped this commitment due to some of Daniel’s arguments, correct? — please someone let me know if I am wrong here.”

    Actually I maintain the view: that self-ascriptions that are IEM are IEM because they’re based on an experience whose first-personal character is a matter of ‘mode’, not content. (Given the mode of the experience, it cannot fail to be about the subject, so a self-ascription of the property that is the explicit content of the experience is IEM.) What I retracted was not that claim, but the claim that a first-person thought can’t be IEM if the subject is explicitly represented in the content of the thought. I now hold that a thought that’s explicitly about oneself *can* be IEM because it can be based on an experience whose explicit content is selfless. The transition from “legs crossed” (explicit content of the proprioceptive experience) to “my legs are crossed” (explicit self-ascription) involves a process I call ‘Reflection’, whereby the contribution of the mode to the overall content of the experience is made explicit in the content of the judgment based on the experience.

  27. Thank you, Francois — that is helpful. I am very sympathetic to your account of IEM in terms of distinctively first-personal modes of awareness, and hope it’s compatible with my own position. (That is, my own position is that even if certain forms of sensory awareness have contents that are explicitly self-presenting, it’s not this that explains the IEM of self-ascriptions based on them — rather, that’s explained in terms of their being forms of awareness whose character is first-personal *also* in its mode.)

  28. Hi Tim

    Thanks very much for all you comments and discussion, it’s been really thought provoking. I’ll be sure to go back through all this stuff when I come to re-work the paper (which I obviously need to do!).

    Best, Joel

  29. John:

    A few thoughts.

    “First, it certainly doesn’t seem that denial that sensory experiences are ever self-presenting is *required* to explain how the first-person perspective is distinctive.”

    I agree, but I think that it is an explanation.

    “I submit that the burden of proof in this debate is on those who wish to say that we can *never* have sensory experiences of properties as belonging to ourselves”

    I’m less sure that this is right, given the worry about animals. But I’ll accept it for the sake of argument. So, what matters is how to interpret the alienation cases. I’ve suggested (and I realise that you don’t agree), there seems to be at least some reason to think that the contents of (at least some of) the alienated experiences are just as usual. If that’s right, then I say that we’ve got reason to think that ordinary every-day experience is not self-presenting. Another way to put this is to say that the ‘sense’ of ownership is not determined by the contents of the experiences themselves (some of my formulations in the paper were not helpful in this regard, but I that’s the view that I want to offer). In this regard, I have a thought about the memory case that I’m going to mull over – I’ll get back to you.

    “According to the view in question, when I experience my motion “from within” (either visually or in some other way), my experience is not an experience of *my* motion, but rather of — well, what?”

    I’m a little puzzled by this, so perhaps haven’t understood the example. I think my response is to say that what I experience to be moving is my body (or limb). I don’t deny that through proprioception one experiences one’s own body. Indeed, I rely on the claim. My body is not, however, represented as ‘mine’ in the experience but rather the sense of ownership over it is accounted for in the manner suggested above. Have I missed something?

  30. Hi Joel,

    I am going to put off saying more about the alienation cases until I’ve had a chance to revisit your paper and some of your earlier comments, and also (hopefully) look again at the paper by Breen et al. But let me say something quick in response to the last part of your #29, which I hope will clarify what I meant in that part of my earlier comment.

    In the paper where I make the case that the content of visual experience is (at least sometimes) explicitly self-presenting, I base a lot of my argument on what it is like to experience one’s own (real or apparent) motion from a first-person point of view. In brief, my argument there relies on three premises:

    1) When I move, I have a visual experience of motion.

    2) This motion is experienced visually as the motion of *something*.

    3) The best way to explain what is experientially distinctive about the “something” whose motion I experience is to say that I experience that motion as *mine*.

    Now, I don’t think you deny premise (1) of this argument, though in principle one can: this would require arguing that the one’s own motion isn’t represented visually in a way different than a corresponding motion of the surrounding world. But this is a hard road to take. Nor, I think, do you deny (2), though once again it’s possible to do this: it would require arguing either that it’s possible to experience the property of motion without experiencing a moving object, or that in a case like this the motion you experience isn’t represented as a property at all. Based on what you’ve just said, I take it that your strategy would be to deny (3), saying that while it is possible visually to experience the motion of my body, such an experience doesn’t represent my body *as mine*. (Of course your comment concerned proprioception rather than vision, and this may be significant — a word on this below.) But how exactly are we to understand this suggestion? Does my experience just represent the motion of *a body*, albeit one that happens to occupy the location from which I am looking? This seems not to be faithful to the phenomenology, not least because it seems to leave us with nothing to say about how this experience would be different from one in which I experienced my motion “as other”. (Of course those experiences would be similar in *some* regards — but then, the same is true of the experience of self-motion and the corresponding one of world-motion.) I see, though, that I’ve now come back to the point about the phenomenology of alienation, so I’ll hold off on developing this response until later. Am I right, though, that this is the premise of my argument that you’d want to reject.

    HOWEVER, a possibly significant concession before I go. The argument I’ve just given seems to work, if indeed it does, only because vision is a modality in which it is possible to experience *both* one’s own motion *and* the motion of other people and things. The same isn’t true, though, for a modality like proprioception, so perhaps it would take a different strategy to show that experience in purely self-specific modalities is explicitly self-presenting, too. (Or perhaps the appeal to phenomena of alienation would come in at the same place? I need to think more about this.) If that is right, then there may be room for a compromise position, wherein sensory modalities like vision can be explicitly self-presenting due to the richness of their possible contents, whereas ones like proprioception present the self only implicitly, since they can’t also present us with anyone else.

  31. OK, thanks John, so the focus is visually presented motion. [As an aside, notice that in the paper I didn’t give the alienation case for vision, and don’t think I denied that vision was self-presenting. Nevertheless, I think that it isn’t, so I’ll defend it as best I can here].

    First of all, whilst I am tempted to accept 1, I would insist that in the usual case, the visual representation of motion is very closely tied up with one’s kinaesthetic sense. To get a pure case, one has to presuppose body-blindness. It then becomes an empirical question whether the sorts of visual illusion to which you appeal in your paper still occur. I don’t know the answer to that question (perhaps you do?). Anyway, lets suppose that they do. That is, there is a purely visual difference between the experience of the movement of one’s surroundings and “one’s own” movement.

    As you say, in such a I’d want to deny 3. I take it that your claim is that what is needed to account for this distinction is the representation of something that occupies the point of origin of the frame of reference at play in the visual experience. Furthermore, that thing needs to be represented in such a way that it is the sort of thing that can move, spin, or whatever. The best candidate, then, for what is so represented is me. The first response to this, of course, is that we could rather have a visual representation of the body play this role. I take it that you reject this on the grounds that (in the body blindness case) one has no visual experience of one’s body. Of course, this itself involves further manipulations of the experience since I always have a visual representations of those parts of my face that bound the visual field. So perhaps we suppose that my body is invisible. I don’t think it’s trivial to decide what our resulting visual experiences would be like. Still, lets press on with this thought.

    Incidentally, I would also want to question the claim that, even in this situation there is no visual representation of my body. I think it reasonably plausible that there are, in many cases, visual representations of the occluded sides of objects, of the surroundings behind my back, etc. As such, I think it reasonable to suppose that, even in the case where I don’t see my body, my body can be (in some sense that I agree needs to be fleshed out) nevertheless visually present.

    Now, as it happens I think that the fact that visual perception is egocentric means that one has to visually represent the origin as itself spatially extended. This, I think, is needed if we are to give an illuminating account of what it is for something to be presented as to the left or to the right (in the monadic sense that Campbell employs – as you know, I have doubts about his early account in terms of action possibilities). As a first thought one might suppose that “I” is badly suited to play this role since, without further content, it’s not obvious that we should think of the representation of myself as involving the representation of something spatially extended. Remember we’ve taken bodily-awareness and the visual experience of my body out of the picture.

    Further, one might point out that the fact, if it is one, that I can’t see my body shows that my body is not visually represented might well also be taken to show that I am not visually represented. What am I? Perhaps I’m identical to a body, to a human animal, to a brain, to a Cartesian ego. Cartesian egos can’t be seen (I suppose) and the other candidates seem to require that to see myself I must see some body, animal or brain. But we’ve ruled those out.

    But haven’t I now painted myself into a corner? What’s left? One suggestion might be that the origin is located as ‘here’, with ‘here’ understood not as an extensionless point, but as an occupied region of extended space. I take it that that’s an intuitive thought in any case. But ‘here’ can’t move, can it? That is, all that remains of visually movement is the presentation of an external object O as located a certain distance and direction from ‘here’ and the subsequent presentation of O as at a different distance/location. But this doesn’t allow us to distinguish between the movement of O and the self-movement. I’m not so sure that this proposal cannot be made to work, however, since the suggestion is that ‘here’ is represeted as occupied. Thus, in the case of O’s movement, O is first represented as first to the right, then to the left, say, of ‘here’. In the case of self-movement, on the other hand, one can say that the represented occupant of ‘the left portion of here’ (if I may be allowed that ugly phrase) is now represented as occupying the region of space previously occupied by that which occupied the ‘right portion of here’. It will follow that O is first represented as right, then left, but this will not be the representation of O as moving, but rather the consequence of the representation of the occupant of here as moving.

    Is this cheating? Haven’t I just smuggled in a visual representation of the body despite the fact that we’ve agreed that no parts of the body can be seen? No, I don’t think so. In ordinary experience, the visual representation of the ‘occupant’ of the space at the origin of the frame of reference will be represented as one’s body. This will, in large part, be due to the integration of vision with the various elements of bodily awareness. In the non-ordinary scenario that we are considering, the relevant portion of space will be represented as occupied, but without any further specification of the properties of the occupant. This is significantly pared down from anything we might ordinarily think of as a representation of the body.

    That was long-winded, so I’ll stop. I’ll end with a quick thought – it would be very interesting to see whether the patient suffering from anonymous vision (Zahn et. al.) suffers from the vection illusion. If so, I would think that evidence against the self-presentation view that you think the case motivates. I have no idea which way it would go, but at least it’s empirically testable. As it happens, Zahn is based in Manchester, perhaps I’ll ask him!

  32. Dear Joel,
    You have covered a lot of ground and I hope this is relevant but I would like to raise hypnopompia. First a little context. As a biologist I am surprised philosophers still want to treat the whole human being as the subject. To me it is metaphysically pretty uninteresting. The role played by the lenses in my eyes in reading, is now done by my glasses – so what? Is the sugar in my breakfast now part of me or having an effect on me? Etc. Subjectivity is up in the cranium so let’s say so. I personally see three relevant self domains here. (1) The whole person. (2) The narrative or tableau that the brain generates about ‘me in a world’. (3) The structures in the brain that, under the causal influence of the tableau, mediate the initiation of a behavioural response. Those (3) I suspect are closest to ‘subjects’ in the traditional sense of instantiating a sense of ‘what the tableau is like’.

    I have been sceptical of IEM but I like your take on rescuing it. If ‘I’ is defined either as (1) or (3) we could be partly or very wrong if asked to justify the veridicality of ‘I felt pain’ by defining the referent of ‘I’ in other terms. You seem to be suggesting IEM is safe if we use something like (2). It must be me that feels pain because there is only one of this sort of tableau of a ‘me in a world’ and the pain is in the tableau in the right way. The ‘owner’ of ‘my’ narrative/tableau is implicit.

    If that is the right interpretation of your approach then it seems a good way to rescue IEM. However, what about the situation where the tableau seems to break up in hypnopompia. I almost sense I heard the door bell (who is ringing at 2a.m.?) yet I also sense that it was the dreaming me that heard the doorbell. I did not *hear* the doorbell, I just have a sense that I have a sense of hearing it (still making my ?dreaming ears feel raw) but am sceptical. The place of the experience in the tableau becomes uncertain. Does IEM fail here? I think it may be rescuable but am interested in your thought.

  33. Hi Jonathan

    Thanks very much for the comment. I’m afraid that my first response is that I don’t really know what “tableau” means in this context. On the face of it, though, it sounds like a property rather than an object. If so, I’d be very reluctant to think that I am identical to a tableau (the same holds for a brain structure). Obviously when it comes to generating subjective experience the action is, as you say, up in the cranium. However, I don’t think that’s an especially compelling reason to suppose that I am a brain. For one thing, if I’m a brain then very many of of the things I believe about myself is false (that I am currently typing, that I am 5’9″, etc.). For my money, there are two serious contenders. One is that I am identical to a human animal, the other than I bear the constitution relation to that human animal. Of course, certain parts of me do more work in certain areas of my life (digestion happens mostly in one spot, thinking in another), but that’s no different to many other objects.

    Now, for my defence of the claim that some self-ascriptions are IEM all we need, I claim, is to think that certain ways we have of gaining knowledge are necessarily ways of gaining knowledge of ourselves. Thus, proprioception and kinaesthesia are necessarily ways of gaining knowledge of one’s own bodily properties. Introspection is necessarily a way of gaining knowledge of one’s own mind. Episodic memory is necessarily a way of gaining knowledge of one’s own past experience. And so on.

    I agree that the example you offer, of a hypnopompic state, gives a good reason to think that many of our judgements can be mistaken. In the case you mention, I can be mistaken in my judgement that I heard a doorbell. It’s less clear, just from this example, that I can be mistaken that I seem to have heard a doorbell. But, in any case, I don’t think that this example really challenges any IEM claim. To do that, we need an example in which I attribute some state to myself, hearing a doorbell, but am mistaken for the reason that although I know that someone heard a doorbell, I am wrong to think that it is myself. This isn’t an example of that sort.

    Now, perhaps you will respond by saying that it was “my dreaming self” not myself that heard the doorbell. And, furthermore, perhaps the reason for thinking that I am not identical to my dreaming self is that I am a tableau and my dreaming self is another tableau, thus a distinct person/self. However, now I want to insist that I am identical to the person who dreamt the doorbell. After all, only in this way can sentences such as “I had a dream last night” be true. And surely they sometimes are! At least, I’d need a pretty compelling argument to be persuaded otherwise.

    So, in short, I think that such cases are cases of error, but not error through misidentification. As such, I’m not too concerned about them.

    I hope that all makes sense. Best, Joel

  34. Dear Joel,
    Thanks for the interesting reply. I sense we may be in complete agreement bar metaphysical pigeonholes. As I see it, ‘I feel pain’ has IEM because ‘I’ is defined in terms not of an object but a current event, a bit like the ‘It’ in ‘It is raining’. (My formal philosophical training is brief; I am thinking Wittgenstein, Kaplan, Travis.) The immunity comes from the pragmatics, which tightly determine the nature of the event and the role of ‘I’ in it. The relevant event is an instance of what you call ‘certain ways we have of gaining knowledge’, and I call a narrative (may be too diachronic) or tableau realized in the brain. It is that event of presentation that lacks self-presentation.

    What surprised me about the contemporary philosophy community is an adherence to a metaphysics of ‘physical objects’ and ‘properties’ when, as realized by Whitehead, physical science has now abandoned objects. We have realized we are dealing with dynamic relations, or events. What philosophers call natural kinds, like animals and species, are just useful but negotiable groups of events – maybe nexuus. It might be argued that physical science is a narrow way of looking at the world but to my mind it is just the usual way, refined to reduce inconsistency. If we want to understand consciousness I think we are going to need that refinement.

    Even in physics the scope of technical terms varies widely with context so I see nothing surprising about the fact that we sometimes use ‘I’ for one nexus and sometimes for another. But if we want to define the most necessary event I would go (I think) with Locke and (I think) Parfit in making it one of a series of instances of a ‘certain way of gaining knowledge’ or tableaux, or Humean presentations in which the self is defined by the event, not the content. The other stuff is one or other nexus we associate with this but rapidly loses necessity. Even proprioceptive signals in dorsal columns are removed, if relevant, because they respond to things that might or might not be ‘I’. Moreover, they might not even be interpreted as ‘sensed by me’. Again, the hypnopompic state of knowing that one is emerging from REM sleep and feeling that something that is nearly ‘I’ is sensing its running legs are not moving is an example. The real I no longer senses that, as a detached observer of the dreaming I. Where I agree is that ultimately, there is an event that is necessarily the I ‘tableau’ or way of knowing. Where I think I disagree is that this has to be the I of ‘I had a dream last night’; it doesn’t seem to be the dreaming I in the hypnopompic state.

    At the moment I think we have a big problem in neuroscience, which one might hope philosophers could help on, with pinning down an individual ‘I’ event. Ladyman is good on absence of things but not forthcoming on the first person side. Nobody is prepared to nail Whitehead’s occasions as circumscribed events. If such events are not circumscribed we have a regress in space and time back to Big Bang. To my mind, most current efforts to circumscribe mental events bundle things up in a non-local way that would have made Descartes, Newton and Leibniz tear their hair out – and should make those familiar with axioms of modern physics tear theirs out too.

    What I liked about your analysis is that it works so well within this event framework and suggests that we could make a link between Wittgenstein and some specific maybe prefrontal biology – a circumscribable event that is uniquely necessarily ‘mine’. I think we desperately need that.

  35. Hi Jonathan

    Thanks for these interesting reflections. I won’t comment on much, as I fear it would take me significantly beyond any area of competence I may have. However, I have some doubt about the thought that physics does without the object/property distinction. As far as I can see (and I’m no expert) one reasonable interpretation is that fundamental particles are objects with various properties (e.g. spin). The question is then when, if ever, collections or aggregates of these objects form a larger object (tables, chairs, human bodies, etc.). I very much doubt that physics alone answers that question.

    Sounds as though your ‘tableau’ is a token experience, whilst my ‘ways of gaining knowledge’ are experiential modalities (e.g. proprioception). My line is that experiences have content, but that content is not first-personal. To find the first-person, one needs to look to the relations between experience and thought. None of this, I think, really tells us a great deal about he nature of the self. I do think that there are some tight relations between phenomenology and metaphysics in this area. However, the view that I’m arguing for here tends to support the claim that, so far as the content of experience goes, the nature of the self is left wide open.

    Best, Joel

  36. Dear All

    Just a quick note to thank everybody for taking the time to read my paper and provide me with such penetrating and useful comments. I’ll do my best to answer some of the objections in a revised version. Thanks especially to John Schwenkler for inviting me to participate and for being so helpful throughout.

    All the best


Comments are closed.