The Idols of Inner-Awareness: Towards Disjunctive Self-Representationalism Presenter: Chad Kidd, Auburn University Read Chad’s paper Commenter 1: Glenn Carruthers Get Glenn’s comments Commenter 2: Stephen Pearce, University of Western Ontario Get Steve’s comments Advertisements Share this:FacebookTwitterEmailRedditPrintPinterestGoogleTumblrLinkedInLike this:Like Loading... Related 28 Comments I would first like to thank the organizers of this conference for this great opportunity and privilege. And I would like to thank my commentators for working their way through my paper in such a short period of time (roughly a week’s time: a task of heroic proportions for busy academics! Thanks so much for your time). I want to start us off by addressing some of the concerns expressed in Glenn’s comments. He begins by pressing an issue that, I believe, does not reveal a true flaw in the argument of the paper. But I am happy that Glenn raises it because it gives opportunity to discuss a crucial interpretative mistake that is easy to fall victim to. Glenn opens by going to the crux of my argument against the explanatory adequacy of HO: that it does not have the resources to provide a positive explanation of the “non-grippingness” of global skepticism about introspection of phenomenal states, i.e., for the fact that there is a striking asymmetry between (i) the (relative) ease with which one can be taken by Cartesian-style evil demon thought experiments to a skepticism about knowledge of the external world and (ii) the (relative) ease with which analogous considerations can bring us to skepticism about the subjective qualities of our own first-order mental states. Standard Cartesian thought experiments get a grip on us psychologically (which is, of course, not the same as saying that they are sound or convincing); but such considerations directed towards the quality and structure of our own experience do not. The mistake I see, however, is evinced in Glenn’s gloss on what it is that underlies this non-grippingness. Glenn writes, Such a feeling must arise, he supposes, given an SR account because it is in fact the case that being in a mental state is partially constitutive of it seeming to be in that state. We cannot seem to be in a mental state without actually being in that state. So it seeming that one is in a mental state justifies the belief that one is actually in that state. As such a kind of mental state scepticism doesn’t appear plausible; one is always justified in believing one is in a mental state if one represents oneself as such. (p. 1). Now, it is true that SR posits, alongside the intentional relation between the higher-order state (what I call a state of “inner-awareness” in the paper and represent with the symbol M*) and the first-order mental state (aka: M), a constitutive relation of “one-sided dependence” of M* on M. So that, aside from M*’s representing M, it is necessary that if M* is realized in the mind of a subject, its intentional object M is also realized. And so, in such cases where M* is realized in the unity of an experience, the subject is justified in believing that she is in M. M* is, as I put it, a “factive” representational state. None of this, however, as I construe it, issues in the claim that Glenn suggests follows from it: that we cannot seem to be in a mental state (i.e., have a state of inner-awareness representing M) without actually being in that state (i.e., M). On the contrary, I argue that this is a possibility that SR *must* accommodate, if it is to be an adequate theory of phenomenal consciousness. And it is precisely this is point that I use to motivate (however incomplete it may be) the disjunctive refashioning of SR (in section 3 page 7). In short, what this move allows is the SR theorist to say *both* that states of inner-awareness are one-sidedly dependent on their intentional objects (first-order states) *and* that, in certain cases, the inner-awareness in the unity of experience is not so dependent. The latter, however, are not states of inner-awareness ‘proper’, but states that are (to use an Aristotelian term) privations of proper states of inner-awareness; they are, as it were, ‘faux’ states of inner-awareness. As such, these “non-standard” states of inner-awareness (as I call them in the paper) are such that they fulfill many of the same functions as “standard” state (i.e., where M* is one-sidedly dependent on M): they still serve to ground a manifestation of experience for the subject, and they seem to play many of the same roles as genuine states of inner-awareness in motivating behavior and judgments. But what they lack – and, crucially, what makes non-standard states a different most specific type of state of inner-awareness from (even subjectively indistinguishable) standard states – is the essential capacity of standard states of inner-awareness to ground the truth of introspective judgments about our own first-order mental lives. So, on the SR view I am defending, we cannot be assured of the truth of our introspective judgments simply by turning our regards inwards and noting, e.g., that “it seems to me that I am now visually experiencing black letters on a white background.” For this experience might be the product of a non-standard state of inner-awareness that, because it is not one-sidedly dependent on its intentional object (i.e., because it is non-factive), cannot issue in the relevant epistemic guarantee. In short, the idea is not that (as I said it is easy to read into my paper) the relevant epistemic guarantee can be attained only if inner-awareness in experience across the board is factive, but only that it is possibly factive. This also helps us to begin to see the answer to the first criticism Glenn issues in his comments: that there is no significant difference between HO and SR. The difference is that SR stipulates *more* than HO stipulates: that, for certain cases of inner-awareness in experience, a constitutive relation of one-sided dependence is essential to it. And it is this that allows us to make sense of the non-grippingness of internal-world skepticism. Now, on to what I take to be the real worry Glenn is expressing: that, as he puts it, [since] this relationship of partial constitution is not […] represented by the subject […] the justification for me believing that I [am having the experience of] see[ing] a screen is inaccessible to me and so cannot explain the experiential difference [i.e., the asymmetry] between external world scepticism and mental state scepticism. This, I think, is a point deserving of deeper reflection. I attempt to address it in the closing section of my paper. There I argue that the skeptical challenge indicated in Glenn’s worry need not worry the disjunctive SR theorist. For the epistemic guarantee of the veridicality of her introspective belief based on cases of standard inner-awareness comes not in anything that the subject can introspectively discern in her experience, but rather in the very metaphysical structure of the experiential state itself. The idea, roughly, is as follows: just because it is possible that my experience involves a non-standard state of inner-awareness, it does not mean that I am in the same (deficient) epistemic situation when I am in a standard state that is dependent on its intentional object. When I am fooled by a non-standard state into thinking that my first-order mental life is a certain way, I am (to be sure) not in a position to know (from within the experience) that I am fooled. But it does not follow from this that, when I am not fooled, I do not know that I am not. For the guarantee comes from the structure of my experience itself, not from something that my experience represents. (This way of responding should be reminiscent of the way disjunctivists about perceptual knowledge of the external world respond to the argument from illusion. In fact, I here model this way of putting the point on the discussion of the argument from illusion by Sebastian Rödl in his book *Self-Consciousness* (Harvard 2007, p. 158).) Since HO abstains from any such postulations, it also keeps itself from this way of responding to the worries of internal skepticism and accommodating a positive explanation of the asymmetry of external and internal skepticism. This also, I think, is all I need to give in response to Glenn’s challenge to differentiate HO and SR. But, of course, if he or anyone else sees something I am missing here, please point it out. Finally, I want to address the main issue that Glenn raises in the second section of his comments on Epistemology and Ontology of CS. First, a small correction that draws on the foregoing clarifications (a correction that Glenn suspects is coming; see p. 3 of his comments). SR does not say that the “kind of *self-knowledge* needed to *make something an experience* must be i) knowledge that one-self (qua subject) is in a mental state, ii) that the mental state must be partially constitutive of the knowledge.” The knowledge does not make the experience. Rather, the experience, as it were, makes the knowledge, i.e., it makes it possible. And it makes it in the way, as indicated above, that does not need to appeal to subjectively discernible criteria to fend of skepticism. Yet, lurking nearby is an interesting issue, which I suspect many will be concerned with in the comments that follow: Why prefer a *positive* explanation of the asymmetry of the grippingness of external and internal world skepticism over a “negative” or “debunking” explanation? In other words, why think that this intuitive asymmetry is not just (as I put it on p. 8 of my paper) a persistent cognitive illusion? My response in the paper is that, so long as there is a consistent and plausible positive explanation available, the burden of proof lies on the side of the debunking theorist. I here end my response and look forward for further development and investigation of these themes over the coming weeks. All, I plan to post my response to Stephen’s comments tonight. They came a bit late. Hopefully they will prompt more discussion. If the paper is too long, I encourage you to watch the video. In it I do not present all that is in the paper, but I present and elaborate the key points. Also, my response to Glenn replays the primary epistemological thrust of a “disjunctive” conception of Self-Representationalism. More anon, Chad Hi everyone, really interesting stuff going on in here! Before I jump into the fray and defend higher-order theories I have something of a clarificatory worry. On page 2 Chad says, Thus, both HO and the versions of SR with which I am concerned accept the claim that a subject’s mental state is made phenomenally conscious – i.e., made into an experience – by virtue of the special intentional connections that the state has with other mental states of the subject When things are put this way Chad seems to be targeting a version of higher-order theory that no one endorses (at least no one I know). It is not the case that, according to higher-order theories, some first-order state is made conscious or made into an experience in Chad’s terms. How could an intentional connection possibly do something like that? That seems to make ‘intentional relations’ into some kind of magic. Rather what higher-order theories aim to do is to explain mental appearances. That is, they aim to account for why one’s mental life appears to one as it does, and they do so by claiming that mental appearances just are states of higher-order awareness. This issue also comes up when Chad is discussing the division of phenomenal labor. He says that this division amounts to positing two components, viz, (i) the aspect that determines what is phenomenally experienced by the subject (e.g., the subjectively experienced blue in visually experiencing a clear daytime sky); (ii) the aspect that determines whether a subjective has a phenomenal experience at all (i.e., whether the subject is ‘phenomenally conscious’ or not). and that, “On both HO and SR, certain intrinsic properties of a first-order mental state (e.g., the sensation of blue, of pain, or of anxiety) fulfill function (i); and higher-order intentional states with a certain kind of intentional content fulfill function (ii).” He then says in a footnote that this view is endorsed by David Rosenthal, but this is not right as far as I understand Rosenthal’s views. But I don’t want to get bogged down in Rosenthal exegesis so I will just say that this is not right according to the version of higher-order theory that seems the most likely to be right to me (or that I am interested in defending). The first-order state is in no way involved in determining what is phenomenally experienced by the subject (except in so far as it is required in order to acquire the concept and in order for normal functioning (i.e. behavior) to occur). Rather it is the intentional content of the higher-order awareness that determines what is phenomenally experienced by the subject. So, given this, I am not sure if the argument is even supposed to be targeted at this kind of higher-order view (which I take to be the standard one, by the way). Now, assuming that the argument is supposed to be targeted at this version of the higher-order theory, there are a few things to say. First, in responding to Glenn Chad says, Why prefer a *positive* explanation of the asymmetry of the grippingness of external and internal world skepticism over a “negative” or “debunking” explanation? In other words, why think that this intuitive asymmetry is not just (as I put it on p. 8 of my paper) a persistent cognitive illusion? My response in the paper is that, so long as there is a consistent and plausible positive explanation available, the burden of proof lies on the side of the debunking theorist. Here is one possible line of argument. People are not gripped by internal world skepticism because they are implicitly harboring a naive theory of consciousness on which it is wrong. I’ll bet that you would *not* be able to get Aristotle to feel the grip of external world skepticism (in fact many people do not feel the grip of it). So whether one is gripped by this (or any other) version of skepticism depends on what theories one accepts and so this cannot be used as an argument. I think that is enough to deal with Chad’s argument, but it is worth pointing out that the argument actually has no purchase on (what i take to be) the standard form of higher-order theory. So, Chad says (p. 12) For now it is possible for a state of inner-awareness M* that provides justification for phenomenal belief on some occasions – i.e., on those occasions where a corresponding first-order state is present – to be present in and constitutive of the experience of a subject while not justifying her introspective belief that her phenomenally conscious experience actually is the way it seems, since the corresponding first-order state is actually absent. And so in any case the subject formulates her introspective beliefs about current and conscious experience on an evidential basis that does not guarantee truth. So, by definition, these beliefs are subject to conservative defeat. According to what I take to be the standard version of higher-order theory one’s belief is justified since one’s phenomenal experience is in fact the way that the higher-order state says that it is. The first-order state need not be there (it is not transformed into an experience, rather the experience, (if we use it the way Chad does) is the higher-order state itself). So I may be mistaken that I am in the relevant first-order state but that has nothing to do with my phenomenal experience. Now, of course, we might be able to run the argument ‘one level up’ and suggest that one can have the introspective belief that one is consciously seeing red (that is that one is in a suitable higher-order state) when one has no higher-order state to the effect that one is seeing red. But even so we don’t get the relevant kind of skepticism. In such a case it will seem to me that it seems to me that I see red, and so what it is like for me subjectively is like seeing red introspectively and so I am right about the way my mental life appears to me and so I am right about what my conscious experience is like at that moment. I just am not right about the way it actually is non0consciously. Finally, Chad says “The HO theorist, then, must specify what else is to be added to the nature of subjectively manifest data that can overcome the obstacle to knowledge the theory entails.” The answer to this seems easy: one must add third person data. If one is consciously experiencing red but is not behaviorally disposed to act in various way (like pressing a button if one sees red, etc) then one can infer that one is in fact not in the relevant first-order state. This is in fact exactly what we find in the kinds of dental fear cases that Chad himself appeals to. It takes the dentist telling us that we cannot be having a pain sensation to get us to see the mismatch. Anyway, that is a lot and so I’ll leave it there for now. Richard (first comment): I now realize that my language is suggestive of this claim that it is the first-order state that is made phenomenally conscious. But I did not intend to endorse this claim. (BTW, this is also a view that Stephen attributes to me.) What I intended to say was this: there is a division of explanatory labor that seems most natural for all extrinsic views of phenomenal consciousness that construe the relevant “relation” as “intentional.” This division of labor is that the presence of a higher-order representational state determines whether the subject has phenomenally conscious experience at all; and the represented first-order state determines the quality of the experience (i.e., “what” is experienced, like whether it an experience the subject could correctly as being as of a red ball or as of a clear blue sky). Given that the relevant relation is intentional two possibilities follow: First, that it is possible for the relevant first-order state M (let’s stipulate that it represents there being a red ball in front of the subject) to exist in the mind of a subject without the subject having the experience as of seeing a red ball in front of her; second, it is possible for the subject to have a relevant higher-order representational state (M*) (let’s stipulate that it represents the subject as having mental state M) without the existence of M in the subject’s mind. Thus, it is not the first-order mental state M that is ‘made’ conscious. Because M need not even exist in order to have a conscious experience. In fact, I want to abstract from any such claims about which state in the unity of a phenomenally conscious experience is the “bearer” of the phenomenal properties in the experience. For I am not sure how to go about answering such a question (I am not sure it even makes sense to ask – as Richard suggests, it moves into mystical-sounding territory very quickly). And nothing in the main argument of my paper depends on it. This does, however, suggest that I need to modify this second section of the paper. I will have more to say about what I am committed to concerning the role of the higher-order and first-order states in the constitution of phenomenally conscious experience in my response to Stephen’s comments, coming soon. But does that help with the first worry you expressed, Richard? Hi Chad thanks for this response. It does help but you still have to be careful when you say “the represented first-order state determines the quality of the experience” as ‘represented state’ is ambiguous in just the way that matters. It is not that the actual first-order state, which as it happens in the good cases, has the property of being represented that determines what one’s conscious experience is like, rather it is that the HO state is representing things to be a certain way does the work. So if by ‘represented state’ you just mean ‘the intentional content of the HO state’ (note, not the intentional object of the state) then we are on the same page. This is the first of a two-part response to the comments by Stephen Pearce. Thanks, Stephen, for you careful commentary. You seem to have penetrated quite deeply into the argument of the paper and come up with a formidable response in a short period of time. I really appreciate your careful and creative work on this! In this part of the response, I address a confusion that is engendered by the language of section 2 of the paper (see also my first response to Richard Brown above). The confusion is the idea that the extrinsic view of phenomenal consciousness attempts to explain how it is that first-order states are *made* phenomenally conscious, which implies that it is first-order states that are phenomenally conscious. Such a view of “state consciousness” has been attributed to David Rosenthal (see, e.g., the discussion under section 4 of http://plato.stanford.edu/entries/consciousness-higher/ ). Whatever the case is with Rosenthal, I don’t adhere to such a view in this paper. Rather, what I emphasize is a difference in explanatory role for the features of a first-order state and the features of a higher-order state in the production of a conscious experience. The difference I employ (that I now am fully aware need not be present in the paper, as it serves no role in the argument of the paper) is that: A. the first-order state and its subjective features determine what a subject’s experience is like for her and B. the higher-order state and its intentional content determine whether there is a phenomenally conscious experience at all. Note: B does not say that the HO state determines whether the first-order state (the state a HO state veridically represents) is made conscious. And it is important that it does not say this. For otherwise it would not be possible to have phenomenally conscious experiences that are due to empty higher-order representations, i.e., higher-order states whose intentional content is not satisfied by any first-order mental state in the mind of the subject. At another point, Stephen suggests that this way of construing the common extrinsic core of HO and SR must not be right. For, he writes, we might hold that it is higher-order states that have phenomenal character. On this view, a state has phenomenal character when it has the right sort of higher-order content: namely, when it represents that one is in another state with a certain qualitative character. But this is to abandon an extrinsic view of phenomenal consciousness. (p. 4) I learned the true significance of an extrinsic view of phenomenal consciousness by meditating on Josh Weisberg’s “Misrepresenting Consciousness” (which is, I think, still only published online: see http://philpapers.org/rec/WEIMC ) As Weisberg argues, one need not maintain that the intentional relation that obtains between a higher-order state and its first-order object requires the existence of both. Rather, we should think of the higher-order state as inten*s*ional state, which can represent a state that does not actually exist. But how does this not reduce to a kind of strange intrinsic view, wherein the object is (to use Brentano’s phrase) an intentional inexistent in the higher-order state? Here theories of intentional content must work their magic. I don’t want to get caught up in the niceties here. But, I take it, a story like this might do the trick for many of us: The higher-order state M*, which has an intentional content , would not have this content if it weren’t for the fact that M* is an instance of a state type that has a causal history involving actually existent states of the type M belongs to. So, states of the type M play a role in determining “what” the higher-order state represents, i.e., its intentional content, in a way that does not require the actual existence of a state of type M in every case where M* is present in the subject’s mind. Now, having said that, I don’t really think that I need to go into any of this for the purposes of my paper. For nothing I argue for in it depends on it. Moreover, as I said, I resist specifying a state to be the subject of “state consciousness.” [PART 2 of 3 of my Response to Stephen] Now, as I understand it, the core of Stephen Pearce’s criticism of my argument for the superiority of SR over HO proceeds by two interlocking steps: 1. To suggest a necessary modification to my argument against the adequacy of HO vis-à-vis the conservative indefeasibility of introspective belief about current conscious experience; 2. To show how this modification also reveals a flaw in my overall strategy. My response, in short, is that, while Stephen is right to point out that the modification in 1 would undercut the effectiveness of my argument, I do not think that the modification is required. Moreover, the modification it suggests would basically reduce it to a non-disjunctive view of inner-awareness. Stephen argues that my argument against HO (stated on p. 12) is not valid without the addition of a further premise. I argue at p. 12 that the HO view, because it abstains from positing a constitutive relation of one-sided dependence, cannot make sense of the conservative indefeasibility of introspective belief. That is, it cannot make sense of the idea of conservative indefeasibility (which I argue follows from the proper analysis of first-person privileged access to experience – see the second half of section 2, starting on p. 9) because it leaves it open for the same most specific kind of higher-order state to exist in both normal experiences and in experiences that have no first-order state corresponding to the higher-order state. I must admit that I do not fully comprehend Stephen’s argument for the invalidity of this argument. For I get the impression that Stephen thinks that the need for the modification depends crucially on his interpretation of my SR view as attempting to explain how first-order states are made conscious (an interpretation which I repudiate in the first part of my response to him). But I do not see why even the adoption of this would undermine the argument or require the modification he suggests, which he formulates as: “(HOJ) When one forms a phenomenal introspective belief, in the usual way, to the effect that one is having an experience with phenomenal property P, what justifies this belief is a higher-order state, which represents one as being in a lower-state with qualitative property P.” The reason that this does not help is because the issue of whether an empty higher-order state in the unity of a conscious experience “justifies” an introspective belief or not is simply not the issue here. It is, rather, whether such a state can provide justification *of the right sort* for introspective belief, i.e., a kind that maintains the indefeasibility of introspective beliefs about current conscious experiences that are “standard” in the sense explicated above. [PART 3 of 3 of my response to Stephen] The final point in my response (2 of 3) to Stephen leads to a key point (and I think that this also partially addresses a concern that Richard voices in comment #4 above): I am not arguing that states of inner-awareness on the HO view cannot justify an introspective belief. It seems to be unproblematic, even given the truth of the SR view, that empty cases of inner-awareness do provide *prima facie* justification to introspective beliefs, i.e., justification that is, in principle, subject to defeat in light of further evidence. But since this is the extent of the justification that states of inner-awareness can provide for introspective belief on the HO view, it cannot accommodate the asymmetry of our intuitions about internal- and external-world skepticism. Let me try to clarify this point by starting from passages in Stephen’s commentary. He writes: “Consider the following two cases. In case A, you see a dog down the road. In case B you see what is in fact a cardboard cutout of a dog down the road. The cutout in case B is so lifelike that it is perceptually indistinguishable (from where you are standing) from the real dog. (p. 6)” Stephen then urges that, in both case A and B, the perceiver is justified in believing that she is seeing a dog down the road. After all, your perception as of a dog down the road does not need to be veridical in order to justify your belief that there is a dog down the road. This belief, then, is not evidentially grounded on your perception being veridical. (p. 6) Certainly the ability for a perception to justify a belief does not require the truth of the perception. For it can justify belief even when it is false. But this is just what I was claiming is the crucial distinction between beliefs that are “defeasible simpliciter” and those that are “conservatively indefeasible” when based on standard cases of experience, i.e., experiences whose higher-order states are instances of a type that are one-sidedly dependent on their intentional objects. I would like to end my response by considering a kind of skepticism that Stephen alleges is present to both standard HO and disjunctive SR views. He writes: “the SR theorist who adopts disjunctivism about inner- awareness is committed to there being a gap for internal-world skeptical worries to take root. The gap is not between states of inner-awareness and corresponding first-order states, but between one’s experiences and the data of introspection: it seeming to one that one is having certain sorts of experiences.” Here again, I think that the disjunctive SR theorist need not be concerned with this challenge. For the “data of introspection,” which I take to mean “What is discernible by the subject in introspection,” are not what renders an introspective belief indefeasible. Rather, it is very nature of the occurrent experience that the introspective belief is about: i.e., that it has the ontological structure that guarantees that the way things (pre-reflectively) seem to the subject in the state is the way things are with her own mental life. I’d like to respond directly to the issues that Richard raises: Richard writes: “According to what I take to be the standard version of higher-order theory one’s belief is justified since one’s phenomenal experience is in fact the way that the higher-order state says that it is. The first-order state need not be there (it is not transformed into an experience, rather the experience, (if we use it the way Chad does) is the higher-order state itself). So I may be mistaken that I am in the relevant first-order state but that has nothing to do with my phenomenal experience.” Some of these issues have already been raised and addressed. But I belabor these points a bit: 1. We know now that I am not committed to the claim that the HO state is the experience or anything like that. I think of experience as the explanandum of the HO and SR theories. And the higher-order and first-order states are crucial parts of the explanans. I never meant to say or imply that either the first-order component or the higher-order component is identical to this explanandum. Rather, as I envisioned the matter in writing this paper, I wanted to be neutral on how to individuate the “experiential elements” of the mind-brain. 2. And again, the argument against HO is not that empty HOTs don’t justify introspective belief. I think its plausible that they do. The question is, rather, whether the general structure of experience, i.e., the structure that belongs to the relation of higher-order and first-order representations, can provide the basis for a positive explanation of the assymetry between the grippingness of internal- and external-world skepticism. Now on to the two new issues Richard raises: 3. He suggests that we can explain the non-grippingness of internal-world skepticism in this way: “People are not gripped by internal world skepticism because they are implicitly harboring a naive theory of consciousness on which it is wrong.” I think that this is not enough. For it is compatible with what I call (following Horgan, Tienson, and Graham) a “debunking explanation” of non-grippingness: that is, an explanation that acknowledges the fact that it doesn’t grip us, but also asserts that this rationally ought not be the case. What I am looking a theory of consciousness to provide is an explanation that also vindicates this intuitive asymmetry. As I put it in the paper, I am looking for an answer to the question, “What is it about consciousness (i.e., as I could also put it: what is it about the constitution of HO and FO states in the unity of conscious experience) that makes our knowledge of it so special?” And I am looking for an answer that actually purports to tell us what is special about it, not just tell us why it seems special, but really isn’t. 4. Richard also suggests that there might not be any assymetry here. For “I’ll bet that you would *not* be able to get Aristotle to feel the grip of external world skepticism (in fact many people do not feel the grip of it).” If by “feel the grip” Richard means, “Are actually global skeptics” then count me in this number as well. I am not inclined to skepticism at all. But if by “feel the grip” Richard means, “can’t, by means of artful cajoling, get someone to see it as a live possibility (even though they are not convinced that it is),” then I disagree. I think that the possibility of external world skepticism is easily brought forth on the basis of the kinds of considerations that Descartes raises in the first Meditation: by what criteria do you discern that you are not dreaming? Or subject to the powers of an evil deceiver. Now, whether these are ultimately rationally compelling is another issue. But we do not find them incoherent, as we do their application to introspective beliefs about current conscious experience. I like and admire this paper a lot for its care, thoughtfulness, and attention to theoretical details. I have a few comments and thoughts. First about so-called dental pain (often called dental fear, because of dentists’ conviction that it’s anxiety that causes one to be aware of oneself as being in pain). This doesn’t seem to be quite the same sort of thing as blindsight. In blindsight, one fails to be aware–inner aware, as Chad says–of one’s being in visual states; in dental fear, one is aware of something that isn’t pain as though it is. A plea for putting my own extrinsic higher-order-thought view (HO) in a slightly better light (echoing a comment or two of Richard’s): Chad correctly characterizes the difference between my view and the self-representation (SR) and HO on p. 3, last full paragraph. SR wants to explicate a constitutive relation between higher-order content and the mental properties that higher-order content is about, whereas HO “sees no need for any constitutive relation.” It’s worth mentioning the reason the HO gives. Both views agree on the transitivity principle–that a state is conscious only if one is aware of oneself as being in that state. HO’s positive claim is that this is all that’s needed. We don’t need anything more constitutive than that, since being aware of oneself in the right way as being in that state in that state suffices for the state to be conscious. Anything more is gratuitous. Richard stresses one side of this, that (extrinsic) higher-order theorists dont think that there’s some change in the first-order state itself when it comes to be conscious. This is of a piece with the point I’ve just stressed. All there is to a state’s being conscious is that one is aware of oneself as being in that state. And we would need a positive reason to suppose that merely being aware of a state would result in a change in the state itself. I think Chad’s argument begs some questions by taking infallibility, following the interesting appeal he make to its etymology, to involve a cognitive capacity. Chad usefully stresses the difference between states’ being conscious in the ordinary, nonintrospective way (what he calls inner awareness) and their being conscious introspectively. I agree that that’s important, and I think he makes very good use of the distinction. Still, most people would say that inner awareness or introspective awareness are fallible if they sometimes get it wrong about what’s going on in one’s mental life; infallibility would simply be never getting it wrong. Never getting it wrong is just the higher-order content’s always being correct. No need here to appeal to the process or actualization of a capacity–just accuracy. And without that appeal, it’s doubting that the case for a disjunctive theory of these things can be sustained. Put differently, the considerations Chad invokes in the relevant part of §3 (and which are of course also essential to disjunctivism) are essentially epistemological. E.g., there’s the talk of legitimizing toward the end of the third full par of p. 6. And there’s a serious question, I think, about whether epistemological considerations should enter at all into one’s theorizing about consciousness. One might think that any higher-order theory must invoke or at least allow some epistemological considerations, since one’s being aware of something is, intuitively, partially factive: If I’m aware of a table as being brown, arguably the table had better be there. But I can be aware of the table as being brown with the table’s actually being brown; I can, e.g., be aware of a green table as being brown. Things get complicated with this very fast. Can I be aware of that object as a brown table even though it’s a brown couch? I don’t see why not. In any case, my version of the transitivity principle, mentioned in passing above, prescinds altogether from this complication about objects vs. properties. A state is conscious only if one is aware of *oneself* as being in that state, and there’s of course no worry about one’s not being there to be aware of aware as being in the state. This, by the way, is the *theoretical* reason that I don’t care about so-called empty HOTs: My being aware of a state as being a pain might imply that *some* state or other be there for me to be aware of as being a pain, but my being aware of myself as being in pain doesn’t require any state that the awareness is about. Or so I argue. But in addition to that theoretical reason, there are actual cases that are best seen as the higher-order content’s being inaccurate, sometimes radically. These include esoterica, such as so-called Rare Charles Bonnet syndrome and Grimes’s saccade-based change blindness, but also quotidien occurrences., such as simply seeing things consciously but inattentively. Back to the quasi-factivity of being aware of things and the question whr theories of consciousness should have truck with epistemological considerations. I would argue that that quasi-factivity is so weak that it doesn’t underwrite bringing in epistemological considerations. And independent of that, I think it’s reasonable to hold that they’re extraneous to asking what it is for a state to be cconscious. A state’s being conscious is a mental property of that state (on my view an extrinsic mental property). Epistemology again adds something extra and unnecessary. This is worth stressing. Many today argue that since mental states of one sort or another do bear epistemological relations to other mental states, we must understand the mental natures of those states at least partially by appeal to those epistemological considerations. Echoing Kant, the mental properties a state has must be whatever is needed to make the epistemology come out right. This neo-Kantian thinking puts the cart before the horse. If particular types of mental state do bear epistemological relations to other types of mental state, such as justifying them or whatever, then they bear those epistemological relations *in virtue of* the mental properties they have, whatever they may be. Intentional content, mental attitude, qualitative character, and nonconceptual content if there is such a thing are *responsible* for which epistemological relations obtain. So we must first settle what those nonepistemological mental properties are, and then see whether they sustain whatever epistemological relations we want to ascribe to the states. It’s not in serious question that we are in states that exhibit mental properties; epistemological relations are far more in dispute, and not a vert sturdy foundation for drawing conclusions about mental properties. Or so I would argue. And whether I’m right about that is less important than highlighting the issue and, if one wants to appeal to epistemological considerations, giving some reason why that’s a good way to procede. If one doesn’t invoke the kind of epistemological considerations Chad brings by insisting that infallibility attaches to capacities, it doesn’t seem that the argument goes through that being receptive drives out infallibility. (Indeed, being receptive, as Chad describes it, itself strikes me as epistemological.) Hi everybody, A very interesting paper and a great discussion! I would like to challenge the commitment of SR to disjuntivism by rejecting that SR is committed to accept cases of empty inner awareness (apologize if the topic has been already touched) SR and HR explain inner awareness by appealing to certain kind of intentional content. They disagree in what is required for a mental state to have such a content. At least in some SR views, empty inner awareness can be ruled out for metasemantical reasons (what is required for a mental state to have the content that accounts for inner awareness –call this content C). As in Millikan’s teleosemantics, to take a well know example, a frog cannot be in a mental state with the content that there is fly around unless (extremely roughly) flies have caused mental states of this kind in the frog’s ancestors; for some SR theories one cannot have a state with content C unless the first-order state is present. Consider Kriegel’s theory, in this case, in order to have content C, indirect representation is required, what in turn requires the presence of the first-order state. The same is true in my proposal (this conference last year). In order to have content C (de se content in my view), the first-order state has to affect certain structures; again here if there is no first-order state then there is no content C. If there is no content C, then there is no inner awareness, and therefore empty inner awareness is not possible. Hi Chad I hesitate to step back in as I suspect many of my concerns are orthogonal to how you like to approach these issues (and don’t we all hate the ‘you didn’t write the paper I would have written’ comments…), but [here comes the but- don’t worry it’s a little one] I think David hit the nail on the head suggesting that setting up an account of consciousness to lead to particular epistemic outcomes is putting the cart before the horse. Although I think from your reply I now understand you better that the epistemology is explained in terms of consciousness (not the other way around as I initially thought), I would still agree with David that designing a theory of consciousness to explain a certain fact about epistemology is designing it to explain a contentious phenomenon. Unlike say the saccade based blindness that David suggests which is a clear and well estblished phenomenon. Now another option that you could take- I I thought for a while you might be trying to hint at- is to say that your account doesn’t explain the epistemology so much as it explains an epistemic feeling, i.e. your account explains why internal world skepticism doesn’t feel grippy. Now as in the comments from Richard there is room to debate whether or not everyone has this feeling, and I think it would be better for you if everyone does, but I don’t see any reason why you couldn’t test this. It would be easier than establishing what the actual epistemology is; just give people the evil demon story and ask them about their feelings. If you do this then you tie your account to something ore directly related to consciousness (the feeling of grippyness). But, I’d worry that if you did take this approach though this is where my concern- in the commentry- about getting the justification into the subject comes in. So it seems to me that your best option for differentiating your account is to focus on how it might explain some well defined phenomena, like those that David suggests. (This also is why I hesitated to step back in because now it sounds like i’m saying “you should care about what I care about, not what you care about” that’s not my intention- I really think you can get a better argument for SR by adapting- or better building- it to explain things like change blindness etc). I hope that doesn’t come off as too abnoxious or territorial! Hello Chad, A very interesting paper! After reading your paper, I have been toying with some speculative idea (which might already exist in the literature, but I do not know) that might be able to account for the grippingness assymetry. I would like to know what you make of this. We (and the literature) are familiar with the phenomenon of “transparency” (from Moore and Harman). One way this phenomenon is described is by saying that when I introspect my experience of a tree, I somehow “pass (or see) through it” to the tree itself. Now, in order to pass or see through X, it prima facie seems, X must exist. This might generate the feeling that a Cartesian demon could not get between my introspective awareness of X and X, for my introspective awareness of X appears to go *through* it. I do not know, however, how to test this hypothesis (i.e., the hypthesis that transparency accounts for the feeling of the absurdity of internal world skepticism). If this is correct (and as I said, I am merely suggesting that it might be), the question now becomes: what is the correct account of transparency? If transparency requires a constitutive connection between M* and M, then the epistemic asymmetry you discuss is real, and transparency explains it. If it does not require this constitutive connection, then the epistemic asymmetry is not real (this will be a debunking explanation, in your terms). Thanks again, Assaf Hi again everyone, I am very interested to hear Chad’s response to David’s worry (with which I whole-heartedly agree) but a couple of quick thoughts in response to Chad’s response to my comment in the meantime. Chad suggests that the argument in the paper does not depend on the confusion I point out but this isn’t true. What is ‘internal world skepticism’? Is it skepticism about the way ‘my mental life really is’ or is it skepticism about the way ‘my conscious mental life is’ (that is about the way my mental life *seems to be to me*)? Chad seemingly adopts the latter but in response to my worry he seems to retreat to the former. However, the point I was trying to make was that the higher-order view doesn’t let the latter kind skepticism off the ground. The standard higher-order approach does in fact license conservative indefeasibility *about phenomenal consciousness* even if it doesn’t do so about the existence of the first-order mental state (I realize now this is what I should have said in my original comment instead of talking about justification simpliciter, since that obscures my real point, as Chad rightly points out). Chad’s argument crucially depends on claiming that the first-order state is somehow a part of my conscious experience (or is phenomenally conscious) but that simply does not follow as long as one avoids the confusion I was pointing out. Now it is true that this kind of account cannot provide any reason for denying skepticism about the way my mental life really is; that is to say that the argument on p. 12 does suggest that my belief that I am actually in the first-order mental state is not conservatively indefeasible but that kind of skepticism is easy to motivate. It is in fact done so by appeal to the kinds of cases that Chad himself appeals to (dental fear cases, etc). However, as I suggested, this doesn’t get a grip on us *unless* we have third-personal kinds of evidence that suggest that the first-order state isn’t really there. In Cartesian kinds of cases we don’t have this kind of evidence and so we find it implausible that the first-order states aren’t there (our behavior seems in every way the same in the Cartesian skeptical situation) but as soon as you introduce third-personal differences (you say you are experiencing red but you don’t behave in the way we would associate with that) then skepticism about our mental reality does get a grip. Unless, that is, one is assuming that the first-order states must be phenomenally conscious. There is a lot more to be said about the other issues that Chad raised in response to me, but I’ve got to run (committee meetings and then off to Rosenthal’s excellent class I am sitting in on), I hope to be able to get back to this tomorrow though. Hello all. Excellent questions so far! I intend to post responses to as many as I can tomorrow afternoon. And I will, of course, return to post more over the weekend, when I do not have teaching duties. But one quick general response: while reworking the paper and thinking through the issues again in light of some of these comments I notice that I make more in the first few sections of the paper rely on the strange idea that it is first-order states that are “made conscious” by the higher-order representation than I need. I think that there is a way of pursuing the line of argument in the paper that is neutral on this point. Thanks for pointing this flaw out to me. Great help. More anon! Hi Chad, I think you’re right that much of what you say can be recast so that it’s neutral about whether a HOA results in any intrinsic change in the first-order target mental state. That’s why I went on about the epistemological dimension: My hunch was that if you recast your argument so that it was neutral in respect of the foregoing issue, you might have to prescind from the issues I urged were epistemological. And that might weaken–or even, I suspect–do away with the apparent advantage of SR over HO. Anyway, that was my thought. Well, I’m jumping in late, but what the hey.,.. Chad, I really liked the presentation (I haven’t had a chance to look over the paper itself in detail–gotta love youtube!). I’ll present some thoughts/worries here. My apologies if some are echoes of earlier comments. 1. I guess I just don’t see why a debunking explanation of non-grippiness is a problem. Here’s my crack at an explanation (I think I got it from Eric Schwitzgebel): in the external world case, we have examples of perceptual error where we get feedback from one sense to contradict what we get from another (think the “bent” stick in water or the Mueller-Lyre, etc.) We can, as it were, lay out side by side contrary perceptions of the external world. We can recognize cases of error for this reason. But in the inner case, there is no additional inner line of evidence to contradict what we’re apparently aware of. This is true even if HO is correct. So we’d expect people to have the commonsense intuition that inner skepticism isn’t (as) possible. And we’d have an explanation of how things seem epistemically “from the inside. What more do we need? This point holds for any good debunking claim, I would have thought. Anyway, I don’t get how burden is being determined here and what the requirements are. If there’s no privilege, we don’t need to explain it. 2. Why can’t the extrinsic HO view go disjunctive as well? Your dependency way of cashing out “constitutive” seems to leave the following possibility open: We can define standard and nonstandard HOTs. Standard HOTs can only occur when the FO state they are about exists (and perhaps exists pretty much as the HO state represents). Nonstandard cases can occur and such cases are subjectively indistinguishable from standard cases, but that’s ok vis-a-vis privledge, as you argue. Given that both SR and HO accept the transitivity principle, I don’t see why the sort of dependency you posit can’t be causal rather than constitutive in the stronger, Aristotlean sense. 3. If HO misrepresentation occurs very rarely, due to the awesome feedback loops in the brain (or whatever), wouldn’t that explain nongrippiness in a positive way? So, given the super-reliable brain processes (or anyway, FAR more reliable than external perpcetual processes), we’d understand why we aren’t gripped by this skeptical worry. No debunking needed. Anyway, sorry if you go into some of these in the paper, or if I missed them in the talk. I wanted to get this in before too long, sloppiness be damned. Nice paper and nice discussion all. Response to Miguel Sebastian: I have argued for a thesis like what you suggest as well. (See my paper in Phil Studies: http://philpapers.org/asearch.pl?strict=1&searchStr=Kidd,%20Chad&filterMode=authors ). As a model of inner-awareness, I think there is a lot to be said in favor of it. For instance, it is a way of explicating the ‘content’ of inner-awareness in standard cases of experience that I would like. But I think, posed as an argument against HO, it simply begs the question. For it attempt to hoist HO on the petard of a constitutive relation that HO simply denies as a gratuitous addition to the theory. That said, one way to read my Idols paper is as an attempt to show why HO cannot blithely deny the constitutive relation, even granting the possibility of ’empty’ experience. Response to Rosenthal. Thanks, David, for your comments. Very incisive. I feel that these comments won’t be persuasive to you. And I am not altogether happy with them myself, since, as I said, your comments are very incisive (as are the comments from Richard and Josh) and I feel like I am rushing my thought on them. Nevertheless, here are my thoughts: First, concerning infallibility you say: “most people would say that inner awareness or introspective awareness are fallible if they sometimes get it wrong about what’s going on in one’s mental life; infallibility would simply be never getting it wrong. Never getting it wrong is just the higher-order content’s always being correct. No need here to appeal to the process or actualization of a capacity–just accuracy.” While this might be true of what most people would say, I don’t believe that it gets at the core of the notion. For it does not sufficiently distinguish between a representational state’s necessarily being accurate and its ‘just so happening’ to be accurate. In other words, it seems to allow the identification of being infallible and being lucky. And this is crucial to the argument I am making against the infallibility of inner-awareness. So that, if we construe infallibility as you say most would construe it, my argument against the notion of infallible inner-awareness would not go through. For I see no tension in the notion of a really lucky mental representation as I do in the notion of a receptive capacity whose success requires nothing more than the actualization of the capacity itself. I must admit, however, that I am somewhat surprised that you would want to resist grounding representational mental states on capacities of the subject to produce these states. This seems the most natural way to combine a philosophical representational theory of mind with results from empirical cognitive science. For what else (from the perspective of a representational theory of mind) could the hypothesized cognitive functions and neural functions be aside from capacities that the brain has to produce brain states on which representational states supervene (and that may be nothing over and above the brain states)? In any case, assuming here a notion of infallibility that abstracts from grounding representations in cognitive capacities, and that construes it as merely “never being inaccurate,” one might construe the necessity of infallibility (as I did in the paper I mentioned in response to Miguel) in terms of nomological necessity, instead of metaphysical necessity (which I here assume to be just as strong as logical necessity, viz., true in every possible world). So that the “infallibility” that attends inner-awareness is a sort of “contingent infallibility,” i.e., an infallibility that pertains to mental states only in possible worlds with such-and-such nomological structure. Here too, it seems, the HO theorist would resist this theory, pointing to the well-established empirical cases you (and Richard) discuss in your (respective) paper(s) with Hakwan Lau, arguing that the actual world does not have the appropriate nomological structure for even contingent infallibility. And it is for this reason that I do not try to wed my disjunctive SR theory even to this weakened analysis of the infallibility of states of inner-awareness. Concerning you claim that the notion of receptivity seems epistemological: I attempt to construe the notion of “receptivity” in a way that does not rely on any epistemological theory or involve any particular epistemological notion. I do understand the notion to be one applicable only to capacities. But I do not see this notion as specifically epistemological either. For the fact that my capacity to shoot a freethrow is prone to failure is not an epistemic failing on my part. As I analyze it, a capacity is “receptive” just when the conditions for its successful execution involve more than is provided by the mere exercise of the capacity itself. And, here too, there seems nothing epistemological about this. For it applies to the capacity to see or hear (both “cognitive”) just as readily as it does to the capacity to catch a baseball or shoot a freethrow (both “non-cognitive). Now for what I take to be the most incisive and challenging comment you made to the effect that the epistemological considerations are theoretically gratuitous. As I understand your view, the *central* explanandum of a theory of phenomenal consciousness is the transitivity principle: that the core of phenomenal consciousness is a subject’s being (appropriately) aware of herself as being in a certain mental state. And so the adequacy of a theory of phenomenal consciousness should be measured by how well (i.e., how elegantly and consistently with findings from cognitive science) it explains the transitivity principle. The theory of subjective qualities of experience need not be brought to bear here, since these are (basically) special representational properties of first-order mental states (the pre-reflective awareness of which in a higher-order state accounts for the “phenomenal redness” of experience). So, even though the HOT thesis by itself explains nothing directly about the nature of properties like “phenomenal redness,” its consistency with an independent representational theory of sensory qualities of first-order states is all that is needed. So, from the perspective of a theory of phenomenal consciousness, the theory of subjective qualities of experience is gratuitous. For it need not prompt any changes to the theory of phenomenal consciousness itself. And, I think, you are taking a similar stance with regard to the non-grippingness explanandum. If my construal of your position is correct, then the way I respond is that the stakes are different with regard to the non-grippingness of internal-world skepticism. As I understand it, unlike the theory of phenomenal qualities of experience, the choice between the SR and HO views turns on saving or sacrificing a crucial part of our commonsense way of construing our access to our own mental lives: that the mind is known in a way that is more intimate and with significantly more limited possibility for error than our knowledge of the external world. In the paper, I rely on (i.e., I merely cite) the arguments in Horgan, Tienson, and Graham for support at this point. (And I would follow these lines of argument as well for further support in response to Weisberg’s challenges against a non-debunking view, inspired by Schwitzgebel). But, at the very least, my point here is that the epistemological explanada are not gratuitous as “phenomenal redness” might be. For the choice seems to be: save the appearance of non-grippingness by postulating a constitutive relation in standard cases or abandon the appearance of non-grippingness. This is not a choice that is far from common sense, as is the choice between seeing “phenomenal red” as an intrinsic property of consciousness (as the intrinsic conception of consciousness would have it) or not. It is, rather, a choice on which common sense has already issued a verdict in support. Now, here, I suspect, an empirical study of such intuitions (like Glenn suggests would be interesting) would support this claim. And this is part of the reason I think that the burden of proof lies on the side of the debunking theorist, especially given a non-debunking theory that also accommodates the possibilities of inner-misrepresentation as well as its best competition (which I take HO to be). Finally, David writes: “This (i.e., my) neo-Kantian thinking puts the cart before the horse. If particular types of mental state do bear epistemological relations to other types of mental state, such as justifying them or whatever, then they bear those epistemological relations *in virtue of* the mental properties they have, whatever they may be. Intentional content, mental attitude, qualitative character, and nonconceptual content if there is such a thing are *responsible* for which epistemological relations obtain. So we must first settle what those nonepistemological mental properties are, and then see whether they sustain whatever epistemological relations we want to ascribe to the states.” It might seem strange that I agree with everything in the first paragraph, but disagree with the second paragraph. I agree that epistemological features of mental life are features mental states have in virtue of their non-epistemological features. This is why I argue that a theory of consciousness that is to explain the non-grippingness of internal-world skepticism must appeal to a metaphysics of phenomenally conscious states. But I do not think that we can approach the metaphysics of mind without some expectations as to what to look for in it. Of course, this must also be balanced by what we will find in cognitive neuroscience. But, as I tried to argue in the paper, I see nothing in it as of yet that would give us reason to accept a debunking view over a non-debunking view. I do not yet have any strong arguments against a debunking view. I have ideas here. For instance, I think that there are grounds for the claim that only a non-debunking view of our knowledge of those states that are “spontaneous” in the Kantian sense (e.g., states of thinking, willing) and certain spontaneous activities (e.g., all cases of intentional action and certain kinds of “second-person” activities, like getting married) require a non-debunking conception of privileged access to our own mental states. The case is obviously harder to make with regard to “non-spontaneous” aspects of mind, such as “sensations.” That is probably not really clear. But it at least gives some indication of how my thought on these issues goes. Perhaps I will issue a post later with more discussion of some of these ideas. Thanks, again, David, for the great comments! BTW, just in case you are wondering. The foregoing response to David R does not address another challenge raised by Richard Brown, concerning the nature of the skeptical worry in the paper. The response to this will come in my next installment. Hi Chad, Thanks for your very useful comments. With regrets, I think I did not understand everything you say quite well enough, and your reactions to my thoughts are very helpful in clarifying places I went astray. Still, I’m not comfortable with your argument. So let me take another shot, with any luck without missing the target this time. First something about infallibility. I said I thought infallibility was just getting it right all the time. You said that didn’t preclude amazing good luck. I agree. I’m not sure how credible it is that we would get things right all the time just by luck; I was assuming, as you say you were, that some naturalist explanation would be operative–if any necessity, then natural necessity. You ground the infallibility is a natural capacity. Here I’m not sure how to understand things. I assume your saying we have a capacity is more than just a way of saying that the thing we have a capacity for happens. It’s something like the way we’re constituted mentally or psychologically. (I would say iI passing that I have less faith than you express in your reply to me in the alleged findings of cognitive science, which seem to me a very mixed bag, and less faith in cognitive neuroscience unless it’s interpreted correctly in the light of psychological findings. But I don’t think these methl matters figure in anything we disagree about here.) So it’s not exactly that I “want to resist grounding representational mental states on capacities of the subject to produce these states.” I just don’t think that that says what needs to be said. And it did seem that the epistemological considerations that enter more full force in §4 were taking root here. But I’m happy to take your word for it that this was my misreading you. I don’t think we are right about all our inner mental goings on, and I don’t think we’re *exactly* right about very many of them. Is this the “global skepticism about the reliability of inner-awareness [that you say] hardly gets any firm intuitive grip”? I don’t think so. I glance at something red and I’m aware of myself as seeing something generically red. I look back and see it’s scarlet–a rather special type of red. I hadn’t been aware of myself before as seeing something scarlet–just as seeing something generically red. Now I’m aware of myself as seeing something scarlet; there’s been a change. There are two ways to go. On one we dig in our heels and say that the visual state was generically red the first time and I was wholly accurate in being aware of myself as being in such a state. I see no reason for that. We correct ourselves about all sorts of things by looking twice; why not about what exact visual states we were in? Better I think to say that I was in a visual state in both cases of scarlet, and aware of myself as being in a scarlet state only the second time, while aware of myself as being only in a generically reed states the first time. You might argue that this is global skepticism. I don’t think it’s any type of skepticism; I think it’s being open to a real possibility, that’s eventually going to be fully testable neurally when we can identify the neural counterparts to the two kinds of first-order visual states. You might argue that it’s not only skepticism, but global; it’s not global. It’s that we are sometime more accurate about our inner states than on other occasions. You might demand an explanation of why it is that we’re as accurate as we are. That’s not difficult. The more inaccurate we are and more more often we’re inaccurate in ways that matter, the less efficient and effective our psychological functioning will be. So as we develop, we come to be very accurate–though not fully and not necessarily always–because inefficiency in psychological functioning is something the system drives out. No special capacities; just efficient functioning. People do think that even getting *anything* wrong *ever* about our own mental goings on is unintuitive. Two points. Intuitions are not one-liners we have to accommodate; that would be weird. Theory typically overturns such commonsense one-liners, about heavy things’ falling, about time and space, and the earth’s surface; people in philosophy seem especially resistant, however, to allowing theory to go up against so-called intuitions when it comes to mental goings on. That seems strange–and mistaken. We need to save the appearances, of course; but saving the appearances doesn’t ever mean taking them to be true. It means explaining why we have them. (And their being true couldn’t help at all to explain why we have them except on an assumption, question begging in this context, of infallibility or transparency of the mind to itself.) And we can explain the appearance that consciousness always gets our mental lives right: There is no *subjective* check on whether consciousness is right in any particular case. (That’s why there seem to be two ways to go in the red-scarlet case.) Since there is no *subjective* check on consciousness, consciousness seems always to get it right. That’s the explain of that appearance. But though there is no subjective check, there are objective checks. There is good reason to think that the first-order state in perceiving is often at variance with the state we’re aware of ourselves as being in. (I urge looking at Grimes’s striking results for especially dramatic cases: https://wfs.gc.cuny.edu/DRosenthal/www/Grimes.pdf) About epistemology: You write: “As I understand it, unlike the theory of phenomenal qualities of experience, the choice between the SR and HO views turns on saving or sacrificing a crucial part of our commonsense way of construing our access to our own mental lives: that the mind is known in a way that is more intimate and with significantly more limited possibility for error than our knowledge of the external world.” I don’t have any clue what intimacy means here; I suppose it must mean if anything that one is unaware of any mediation. If that’s what it means, we can certainly understand why common sense ensorses that view. But then we have no reason to give any special credence to common sense. If it means anything else, why think it’s true? As to the “significantly more limited possibility for error” you say common sense ascribes to our access to (I assume that means awareness of) the hpl qualities of experience, I have already said something between that above. Hence my worry about the epistemological dimension of your argument–which consists, as I understand you to see it, in giving credence to what I see as an unfounded doctrine of pretheoretic common sense, though a doctrine that we can readily explain without giving it credence. So I would take issue when you say in your reply to me that “hat the epistemological explanada are not gratuitous” A small remark about my own view. You describe me as holding that first-order states constitute a prereflective awareness; I wouldn’t say that–partly from not knowing what it means. I think that first-order states unaccompanied by suitable higher-order awarenesses simply aren’t conscious states, though they make us aware of stimuli–as subliminal seeing does. A final point: I see no advantage to SR when it comes to saving the commonsense dicta–something’s representing itself can be every bit as inaccurate as one thing’s representing another. This is especially so with Kriegel’s version of SR, in which it’s questionable whether the state with the higher-order content really is literally the same as the state whose (other) mental properties are represented by that higher-order content–whatever “constitutive” relation Kriegel may posit. (On that constitutive relation, see Ben Phillips, http://www.academia.edu/2344320/Indirect_Representation_and_the_Self-Representational_Theory_of_Consciousness, forthcoming in Phil. Studies. But forget questions about whether the state with the higher-order content is the same as the state that content represents, and let them be a single state. Why should the higher-order content be acurate about that state’s other mental properties? What magic ensures that? If it were so, wouldn’t be need a nontrivial explanation? Nothing about immediacy or the like would do, since higher-order content can be as unmediated as you like as yet inaccurate. I do hope that I have at least done better this time in understanding your position. Response to Richard (and a little more for DR): Richard has a great way of putting a question that should trouble anyone working on inner-misrepresentation. He writes about my internal-world skepticism: “Is it skepticism about the way ‘my mental life really is’ or is it skepticism about the way ‘my conscious mental life is’ (that is about the way my mental life *seems to be to me*)? ” In other words, is the person in the grip of internal-world skepticism worried about whether her introspective beliefs about current, conscious experience accurately represent the way her first-order mental life actually is? Or is she just worried about whether her introspective beliefs accurately represent the way things pre-reflectively *seem* to her. Obviously, these two are importantly different in many ways. And, I think, Richard is arguing that HO does not fall into internal-world skepticism when we take the worry to be about how things seem pre-reflectively. For “how things seem (i.e., what it is like) for the subject is determined *solely* by the HOT (or whatever other higher-order representational state you prefer) in the overall unity of the conscious mental state. And so, the accuracy or inaccuracy of the state makes no difference. The only way this would make a difference is if the accuracy of the HOT vis-à-vis the first-order state is at issue in the explanation of how ”what it’s like for the subject” is determined. Now, as I said in response to Rosenthal above, one of the main tasks of the paper is to show that there is more at stake in a theory of phenomenal consciousness than merely explaining how “what it’s like for the subject” is determined. We must also take into account the intuitions concerning the privileged accessibility of phenomenal consciousness (which is brought out in the intuition of the asymmetry of external- and internal-world skepticism. As I said there, I think that a decision on this issue is not gratuitous. Rather, the decision depends on whether we specify that the dependence of the HO representation on its first-order object is one that renders the HO representation factive (at least in standard cases) or not. Now, the question becomes: what is this factive state about? Is it: (a) The way things “seem to me” in the experience (i.e., the phenomenal character determined by the content of the higher-order representation) or (b) the way things actually *are* with regard to my first-order mental life? Obviously, the answer, as I construe it, must be (b). For there is no problem for HO in capturing the sense in which our beliefs about (a) are “factive”. (This, I think, is a really important insight, often missed by those on the other side of the “empty” HOT debate.) But I do not miss it. In fact, I want to bring it out as exactly the kind of “factivity” that does not help us overcome the problems of internal-world skepticism. For what we are worried about is whether the phenomenal character of experience itself actually, in some way, *reveals* to us our own first-order, world-directed mental life. That is, more concretely, we are concerned about whether, when it seems to me (from within the experience) that I am seeing a red ball, that I am actually in the state that I seem to myself to be in. But what HOT reverts to, when we follow the straight Rosenthal line, it seems, is an attitude that would be unmoved by the possibility that the phenomenal character of experience never actually indicates what is going on in our first-order mental life. For since the veridicality of HOTs makes no difference to phenomenal consciousness, we would not take this to be an epistemic loss for pure introspective thought. This tactic, if it is indeed what the HO theorist has in mind, reminds me of O.K. Bouwsma’s excellent little paper diagnosing and trying to disabuse us of the skeptical pull of Descartes’s Evil Deceiver scenario. (Here’s a link: http://philpapers.org/rec/BOUDEG ) There he argues that if the radical form of deception – the kind that we are subject to in global illusion – makes no difference to our cognitive behavior, given how things seem, then there is no epistemically relevant difference at issue. I, on the other hand, think that there is good sense of what is epistemically lacking in the case of radical external-world deception: access to the actual structure of the external world. Mutatis mutandis, I think that there is good sense of what is epistemically lacking in the case of radical internal-world deception: access to the actual structure of our own first-order mental lives. This also speaks to Weisberg’s question about why HOT could not go disjunctive. Now, I think that HOT could go disjunctive. But to reduce the constitutive relation to a merely causal correlation is not enough, i.e., if we take (as I do in the paper) the causal relatedness of M* and M to be compatible with the mutual independence of M* and M (i.e., in Hume’s terms, compatible with taking them to be ‘distinct existences’). Given mutual independence, the doubts giving rise to radical internal-world skepticism still loom. (Moreover, even if we take the constitutive relation to be nomologically-based, as I did in my Phil Studies paper on Infallibility of Self-Awareness, there is still the question as to whether this world actually has the appropriate nomological structure. If the demon can make a piece of soap look like a lemon, there seems no reason why he couldn’t make a world whose nomological structure is not factivity-saving ‘seem like’ a world that is.) Now, to the other point worrying Richard: I take all of this to be (at least I think) importantly different and independent of the claim that states of ‘HO awareness “make first-order states phenomenally conscious.” There is no need to take inner-awareness (my terms) to be like shining a high-powered light on an object: where phenomenal consciousness is a intrinsic property, developed in the first-order state by virtue of its causal relation to a state of inner-awareness, like how french fries stay hot (here taking “hot” – perhaps wrongly – as an example of an intrinsic property) by virtue of their causal relatedness to a heat lamp. In fact, I think (as I am attempting to do in the paper at present) that I could simply take Rosenthal’s apparatus of phenomenal consciousness – constructed as an explanation of the Transitivity Principle – on board as the explanation of what it is to “make states conscious,” but motivate an additional explanans – the privileged access of first-order mental life – that requires the constitutive relation between higher-order and first-order awareness. Thanks for your reply Chad, just for clarification. I was not and have not presented any model as objecting to HO, just as an alternative way of implementing the idea behind the transitivity principle. I am not, on the basis of the empty case objection demanding that the FO be constitutive of inner-awareness, but providing a a metasemantic theory of the content of experience, the result of which is that the relation between FO and inner-awareness is not merely intentional. Think of it simply as a mere alternative to HO (one that shares with HO theories many features). This position is SR, is not disjunctive, is not ad-hoc and doesn’t seem to suffer from your objection. Where have I gone wrong? I am looking forward to read your paper, thanks for the recommendation. Now, just in the light of your reply to David in #21(I am sure I am missing many important points, I apologize for that): In your paper you construct the necessity of infallibility as nomological necessity and this is not what I was pointing out to. I maintained that any metaphysically possible world which is a physical duplicate of the actual one is a world in which there is no empty inner awareness, because the FO is required for a mental state to have content of the C kind –and a subject has a conscious experience iff the subject has a mental state with content of the C kind. Hi Chad, thanks for your response, it helps to clear up a lot. I now wonder, though, how hard it actually would be to convince these intelligent undergrads that this kind of skepticism might hold (especially if one gave them the kinds of empirical cases that Hakwan, David and I like to appeal to)? I doubt very much that they are clear on the distinction that our discussion has brought into focus and it is all too easy to slide into talking about skepticism about my conscious mental life (as you do in your paper and video) rather than about my first-order mental life. Descartes clearly didn’t have it in mind since he thought all mental states were conscious and so he would have only been reacting to the kind of inner-world skepticism that you allow the higher-order views have no problem with. But now with the argument in full view David’s ‘final’ worry from above becomes all the more pressing. At the end of your paper you address the worry of how we know that any of our states of inner-awareness are standard. Your answer, it seems to me, is to argue that since we do have cases of standard states of inner awareness where our first-order mental life is in fact revealed to us there is no room for skepticism. The higher-order views on the other hand will always have this worry because for all they say there may never be any cases of factive higher-order awareness. But it is not clear why you get to say that we know that there are standard cases of inner-awareness. As David has been pressing it is presumably an ubiquitous fact about inner-awareness that it represents our first-order states in a less fine-grained way than they actually are. So if my first-order visual state represents Red48, but my inner-awareness represents me as merely seeing Red then what it will be like for me is like seeing red, it will not be like seeing Red48 (except in so far as it is a shade of red). I take it that on your account this is *not* an example of a standard state of inner-awareness, is that right? If so then there is room enough for a global internal skepticism that puts self-representational views on all fours with the higher-order views. One last question occurred to me while reading your response to Josh. You say ” If the demon can make a piece of soap look like a lemon, there seems no reason why he couldn’t make a world whose nomological structure is not factivity-saving ‘seem like’ a world that is” And I agree, but so too there is no reason why she couldn’t make a world of non-standard inner-awarenesses look like a world where there are standard ones, right? I mean to say that there is a possible world out there where all there are are cases on non-standard inner-awareness and so at that possible world global internal-world skepticism is true. But even though you agree that this is possible you argue that we have reasons to think the actual world is not like that. Why does it matter if those arguments appeal to constitutive relations or causal ones? The point is the same: there is a logical possibility that we cannot completely dispel but even so we have good reason to rule it out as actual. Thanks for the great discussion. I’d like to jump in on Richard’s worry about the different kinds of internal-world skepticism, because I think everything turns on this. According to Richard, the version of HO theory that we are considering holds that: ..”it is the intentional content of the higher-order awareness that determines what is phenomenally experienced by the subject.” Now I’m not quite clear on whether Chad has accepted this claim. He says: “I think of experience as the explanandum of the HO and SR theories. And the higher-order and first-order states are crucial parts of the explanans. I never meant to say or imply that either the first-order component or the higher-order component is identical to this explanandum.” In cases where the higher-order state misrepresents the lower-order state (say by representing that you are having a blue experience, when the lower-order state is in fact representing red), HO theorists need to take a stand on whether the phenomenal character in such a case is blue or red. And as far as I can see, there are only two options: either the content of the lower-order state or the content of the higher-order state determines the phenomenal character of such an experience. So lets follow Richard and take the latter route. To have a phenomenally blue experience, then, is to be in a state that represents that you are in a state that represents blueness. Now, as Richard notes, there are two ways of understanding “inner-world skepticism”. The first is skepticism about “the way my mental life really is”. The second is skepticism about “the way my conscious mental life is.” Since “the way my mental life really is” includes (among other things) the content of my first-order states, let’s call the first first-order inner-world skepticism. Skepticism about “the way my conscious mental life is”, however, is specifically about how things seem to me: in particular, it is about how things phenomenally seem to me. It is skepticism about ones occurrent phenomenal experiences. You are gripped by this form of inner-world skepticism when, upon noticing that you are having a pain experience, you wonder: “am I really having a pain experience?” Let’s call this phenomenal inner-world skepticism. Since the relevant difference between SR and HO theories is, as I understand it, the factivity of the higher-order content, and since the higher-order content alone determines the phenomenal character of your experiences, then SR has no obvious advantages over HO in terms of phenomenal inner-world skepticism. But arguably SR does have an advantage over HO in terms of first-order inner-world skepticism. But Chad argues that this is enough, because first-order inner-world skepticism is important: “For what we are worried about is whether the phenomenal character of experience itself actually, in some way, *reveals* to us our own first-order, world-directed mental life. That is, more concretely, we are concerned about whether, when it seems to me (from within the experience) that I am seeing a red ball, that I am actually in the state that I seem to myself to be in.” It is not clear to me why we should be troubled by this form of inner-world skepticism, however. Our first-order, world-directed mental life in general does not seem to be the usual domain that it is intuitive to think that we have privileged access to. When I have a perceptual experience as of a red ball, I am not gripped by the possibility that I am not really having an experience as of a red ball. Nor am I gripped by the possibility that my perceptual experience is really about something other than a red ball (which I take to be the meat of Horgan, Tienson and Graham’s argument for phenomenal intentionality.) This is because I am not gripped by skepticism about the phenomenal character of my experiences, and my experience being about a red ball is part of its phenomenal character. Experiences do have a world-directed component, but this is just part of their phenomenal character, which is determined solely by our higher-order states. What is open to doubt is whether, in addition to this perceptual experience as of a red ball, I am in any other mental states that represent that there is a red ball. I may be skeptical about whether the phenomenal character of my experiences reveals anything about the nature of my non-phenomenal mental life: first-order inner-world skepticism. SR theories seem to entail that we cannot (do not? should not?) doubt this, but this doesn’t appear to constitute an advantage over HO theories. Hello all. Thanks again for the great comments and questions. Instead of posting a ‘final’ reply, I will provide a link to the revised paper. I have benefitted greatly from your interaction. http://www.academia.edu/2459817/The_Idols_of_Inner-Awareness_Towards_Disjunctive_Self-Representationalism Comments are closed.