Color-Consciousness Conceptualism

Presenter: Pete Mandik, William Paterson University

Commentator 1: Jacob Berger & David Pereplylotchik, The Graduate Center, CUNY

Commentator 2: Charlie Pelling, Birkbeck College, University of London

https://consciousnessonline.files.wordpress.com/2010/02/comments-on-mandik.mp3>Comments%20on%20Mandik

Commentator 3: Philippe Chuard, Southern Methodist University

Pete’s Response

Advertisements

99 Comments

  1. First off, let me thank Pete for offering such a rich and thorough reply to all of the comments, including mine and Jake’s. It will take me some time to think carefully through everything he’s put on the table, so what I say below is by no means the last thought I’ll have on his replies. But it is the first.

    In response to Charlie Pelling’s comments, Pete writes:

    “there’s no obvious harm done to the conceptualist in allowing that the conceptualizations deployed in conscious experience are contradictory. Of course, I don’t mind supposing that reality has no room for contradictions. Something cannot at one at the same time be just like A and not just like A. But it’s much less problematic allowing that there are contradictory representations. There is, for example, the following sentence: “B is a color that is simultaneously just like A and not just like A.” That sentence gets on just fine being contradictory. Perhaps analogous mental representations exist while being analogously contradictory.”

    I must confess that Pete’s move here struck me, at first, as somewhat desperate, partly for the reasons he goes on to mention in the remainder of that paragraph. He writes:

    “Of course, when the representations in question are beliefs, and the believers are rational, and the contradictions are very simple and obvious, many philosophers will want to say that there’s some sort of problem here.”

    Pete attempts to defuse the worry by noting that “the conceptualism on offer is not committed to conscious experiences being beliefs. Conscious experiences need only be similar to beliefs in the following manner: they are attitudes toward contents exhausted by deployed concepts.”

    I take issue with the claim in the very last sentence. It seems to me that whatever mental attitude conscious experiences bear will have be assertoric–i.e., committal. In this crucial respect, conscious experiences resemble beliefs. And I suggest that it’s *this* that animates the worry concerning contradictions. Commitment to contradictory contents, especially *conscious* conscious contents in a *rational* agent, is just plain unpalatable–a conclusion that can be reached from a wide range of theoretical standpoints.

    Upon further reflection, however, it seems to me that cases of commitment to something like a contradictory content can be offered. Consider the well-known images that perceptual psychologists have dubbed “impossible figures”. Several such figures can be viewed at the following site.

    http://www.fink.com/papers/impossible.html

    I have not delved into the relevant literature, but it strikes me as well worth examining whether cases of geometric “contradiction” are on a par with the chromatic contradictions Pete has in mind.

  2. David,  
    Thanks again for your commenting on this material. I dig it.

    I think I can agree that conscious experience is assertoric , but    I’d like to here wonder out loud about how much trouble this really raises about contradictory contents. I presume conscious experience to not just include veridical perceptions, but also illusions, afterimages, hallucinations, and imaginings. Certain cases of these latter sort make various philosophers reluctant to regard conscious experience as a kind of belief. I can afterimage a green square on a white wall without believing that there’s a green square on a white wall. I can be subject to a powerful case of Titchner circles or Muller Lyer arrows without believing that the figures really are as the appear in consciousness.

    Whatever is assertoric about conscious experience, it looks like it can be so without being accompanied by a full blown endorsement by the rational believer. Conscious experience can make assertions that may or may not be simply echoed in belief. 

    If being assertoric is consistent with not receiving the endorsement of belief, then maybe that’s they way to see contradictory conscious contents as not posing any special problems for the theorist.

  3. David,

    Here’s an additional thought about the chromatic contradictions that might arise if there are such things as synchronic phenomenal sorites.

    First off, here are some preliminaries.

    Consider the conceptualized content
    (1) The color chip on the left is red all over and green all over.

    Is (1) contradictory? It’s not logically contradictory. In contrast, the following *is* logically contradictory:

    (2) The color chip on the left is red all over and green all over and there is no such thing as a color chip that is red all over and green all over

    While (1) is not contradictory, it says something that, for purposes of conversation, let’s assume is false, since nothing is red and green all over. What makes (2) contradictory is the logical inconsistency of the conjunction of (1) and that empirical information.

    Now let’s take a look at an alleged synchronic phenomenal sorites.

    One conceptualization of the stimulus might be like this:

    (3) A is just like B, B is just like C, and C is not just like A

    Another conceptualization might be more complicated. Consider a conjunction of (3) with an explicit conceptualization of the transitivity of being just like, so:

    (4) A is just like B, B is just like C, C is not just like A, and being just like is transitive

    Perhaps a conceptualist can pursue a case whereby (1) and (3) aren’t contradictory, they just violate empirical truths that can yield contradictions when conjoined to yield (2) and (4) respectively.

    I’m not sure I’m convinced that this is a great way to go, but I thought I’d throw it out there and see if it stuck to anything.

  4. Like David, I want to thank Pete for his thoughtful replies to our comments. I should say that I also really enjoyed Philippe’s and Charlie’s commentaries. I’ve learned a lot from this whole exchange. Lastly, I’d like to thank Richard Brown for organizing this great conference. I have to think about everything here, but I’ll say a few things now (I note that these are my remarks, and may not necessarily reflect David’s opinions).

    I have something like a quibble about Pete’s rephrasing of our statement of the two conditions for conceptualism. I regard talk about concepts much in the way that nominalists regard property talk – if it’s philosophically useful to use such language, I do, but I doubt these posits will make it into the final analysis. David and I went along with the concept talk at the onset for ease. But I’m unconvinced that thoughts are wholes composed out of compositionally combined concept pieces. I would prefer to reword Pete’s version of the conditions to remain neutral about whether thought has such structure into something such as:
    (1’’) For every color one can consciously experience, one is capable of having an intentional state about that color.
    (2’’) Every time one consciously experiences a color, one has such an intentional state.
    This restatement has the virtue of remaining neutral about the compositionality of thought, which I, unlike Pete and Fodor, do not think is nonnegotiable.

    However we restate these conditions, one of the major points of our commentary is that one can hold something like (1) without holding something like (2), and still be called a conceptualist. It doesn’t follow from the fact that we can have intentional states about all the colors we can experience consciously that we must always have an intentional state about a color when one consciously experiences that color or that all there is to a conscious experience is intentional content. Just as one doesn’t have to be a conceptualist to be a representationalist (i.e. Tye), one doesn’t need to be a representationalist to be a conceptualist.

    Our proposal is that one’s nonconceptual sensory impressions can be conscious, and yet one need not have any intentional states about the colors of which those sensory impressions provide us sensory access (though, of course, one could). Pete, however, denies that sensory impressions are ever conscious. His reasons have to do with his concerns about the HOT story of how sensory impressions become conscious. First, Pete thinks that it’s reasonable to identify the conscious state with the intentional HOT (for that it what is responsible for “what it’s like for the person” to be in that state) rather than with the first-order nonconceptual impression. I agree that what it’s like for two individuals with the same HOTs will be the same, but that does not entail that “what it’s like for the person” or the “consciousness” are properties of the HOT. HOT theory holds that the HOTs are themselves seldom conscious, though they are responsible for the first-order state’s being conscious.

    Recall one of the central motivations for HOT theory. Folk psychology tells us that sometimes the very same state can be both conscious and nonconscious. A pain in one’s leg may at one moment be conscious, at the next moment nonconscious (as one becomes no longer aware of it, say because one is distracted), and then be conscious again if one later attends to it. Throughout this process, intuitively it’s the same pain. If so, then we need a story about what distinguishes the state in its conscious form from its nonconscious form. If we identify conscious states with HOTs, as Pete suggests, then we have no such explanation. This is not to say that Pete couldn’t tell some story, but I think HOT theory provides this account.

    But Pete thinks there’s a bigger problem with HOT theory, which is the problem he targets with his Unicorn Argument. Pete claims that since we sometimes represent things that don’t exist, representation isn’t a relation. If so, then a state’s being conscious can’t consist in its being represented by a HOT. I admit that there are some delicate issues here, but in short I’ll say that this argument rests on a misreading of HOT theory. Although HOT theory is often glossed as the claim that a state is conscious only if there is a suitable HOT representing it, there is way to read this claim that defends it from the Unicorn objection.

    Another way to read the HOT claim is that state is conscious only if one represents oneself via a suitable HOT as being in that state. This reading does not commit one to the target state’s existing. At first blush, this other reading may sound paradoxical, but it only sounds that way if we assume that state consciousness is a property of states. Instead, state consciousness is a property of persons. What it means to say a person is state conscious, or to say that the person is in a conscious state, is to say that the person is aware of herself as being in that state.

    Pete might reply that what it’s like for one is the same whether or not the first-order target state is present, and I agree. And, he might continue, if one can be state conscious without there being any target state, then why should we suppose that (a) the target state is what is conscious or (b) that we should even posit the target state in the first place? But (b) doesn’t hold because we often erroneously think things are present when they aren’t, but that does not entail that those things don’t exist. I might think there is a table in front of me when there is in fact a table in front of me, or I might think that there is a table in front of me when there is no such table. The fact that we couldn’t distinguish from the first-person the veridical case from the illusory case doesn’t entail that we should not posit any tables. And (a) doesn’t follow for the reasons stated above. HOT theory is an account of what distinguishes conscious states from their nonconscious counterparts. If the same state can be both conscious or not, then we have reason to think it’s the target state that becomes conscious (and, of course, I mean by this that one becomes conscious of oneself as being in that state).

    There’s a lot more to say, but I’ll leave it here for now.

  5. Hi Jake,
    Thanks for the further thoughts. I’m really grateful for the exchange.

    Re LOT stuff, I don’t have a whole lot to say. Most of what I have to say comes from my Supervenience and Neuroscience paper. If at a single time, a subject has a conscious experience as of seeing a red circle and a blue circle and at a later time, a conscious experience of just seeing a red circle, it seems overwhelmingly plausible that at t1, there are 2 distinct phys properties instantiated, one of which is common to both t1 and t2, the other of which is present only at t1. 

    Re Unicorn stuff:
    I’m having a hard time seeing how this defense of HOT is all supposed to hang together. It seems to me that the version of the HOT theory that results from reading it to be Unicorn-immune is a version that cannot be advertised as explaining (as opposed to explaining away) the so-called central motivation (CT). If CT is to be regarded literally, then it must be literally true that one and the same impression can be conscious at one time and unconscious at another.  There are only two general ways this can be explained. One is for there to be a change in an intrinsic property of the impression. The other is for there to be a change in the relations born to the impression. It seems quite clear that the friends of HOT don’t take the first option ( and for what it’s worth, neither do I in my AEI theory). So, as long as HOT-heads want to offer an explanation of the literal truth of CT, it looks like they are committed to an impression’s changing from unconscios to conscious by virtue of some change in what relations are born to it. However, if it is being granted to me and my pet Unicorn that there is no relation of being represented, then whatever is the explanatory relation that’s born, it’s not that the relevant HOT is a representation of the impression.

    So, how is it that CT gets explained while simultaneously granting that representing is not a relation?

  6. My thanks to Jake for what strikes me as a fine defense and account of HOT theory. And my thanks also to David for the assertoric point–and to both Jake and David for what I thought was an outstanding comment on Pete’s challenging paper. I thought I’d just add a thought or two to the discussion.

    The consciousness of mental states is, from a pretheoretic, intuitive point of view, a matter of mental appearance. It’s how our mental goings on appear to us.

    That much is clear. There some that’s not. ‘Conscious’ as an adjective applied to mental states (as opposed to creatures), didn’t get much use in English (or, to my knowledge, other European languages) until relatively recently. So is it clear that there is stable way we speak about conscious mental states? I’m less sure than Pete.

    That’s why Jake’s paraphrase of what’s involved in mental states’ being conscious strikes me as especially felicitous. Since the consciousness of mental states is at bottom a matter of how our mental lives appear to us, best we put things by saying that a mental state’s being conscious consists in one’s being conscious of oneself as being in that state.

    This is, as Pete puts it, immune to the Unicorn Argument. If I’m conscious of a tree, the tree had better be there. So whatever my being conscious of something consists in, we can construe it relationally. Still, I can be conscious of the tree as being purple even if it’s green; the relation holds between me and the tree; I don’t need a relation between me and the way I’m conscious of the tree as being.

    HOT theory says that the way I’m conscious of my conscious mental states is have having suitable thoughts about them. So I anticipate Pete’s objecting that thinking in general isn’t the way just described the way we’ree conscious of our conscious mental states. That’s because we sometimes have thoughts about things that aren’t there.

    I don’t know that there’s a settled, good way to deal with this issue. Thinking pushes in two opposing directions. On the one hand, some thoughts are about things that aren’t there. Indeed, it’s worse than that. I can have a thought about something that I’m convinced that’s there but isn’t–and conversely.

    On the other hand, however, thoughts do connect somehow with extramental reality–or, to let HOTs into the picture–they do connect somehow with the things the thoughts purport to be about.

    When one regiments to within an inch of one’s life, I agree with Quine’s eventual position that there’s no way to ascribe intentional states that’s genuinely relational, as opposed to notional, using here Quine’s own terms.

    But it seems unacceptable to suppose on that account that when I think about things, I never connect with the things my thoughts are about.

    I don’t have a nice way to wrap all that up in a single pretty package. But I’m not inclined–and I suspect few people are–to draw the conclusion that some of it is just wrong. So pending something more decisive, I’m inclined to accept Pete’s outright rejection of the idea that thoughts and representation are ever relational. If I don’t connect somehow with Obama when I blame him in thought for the failure of the public option, my thoughts are a pretty meaningless matter. My not having a good, unified theory of how all that works shows, I think, not that there’s no relational aspect to thought, but that our theorizing about thought needs more work. Since nobody could be at all informed and doubt that conclusion, I’m happy with it.

  7. Let me just chime in with my $.02 and say that is precisely where the causal theory of reference steps in and become useful. When it comes to regimentation we may need to abandon names but we need not on that account bar them from our mental ontology. Clearly my thoughts about Obama are about him precisely because I am causally related to him in the right way. That does not mean that other thoughts do not work differently (I am open to the claim that thoughts about unicorns may just be thoughts about horses with horns (clearly ‘bachelor’ is like this)…though if ‘unicorn’ is a natural kind term then it is mot likely causally related to an act of the imagination)…

    Or another way to make the point. Why isn’t what you say here reason to adopt some kind of hybrid theory like that of Michael Devitt’s?

  8. Pete, David, Jake, Philippe, and Charlie,

    Thank you all for your presentations. I find myself in agreement with much of what Pete said about conceptualism, and share the same concerns that David and Jake have about Pete’s rejection of the HOT theory.

    For Pete:

    I hope you could expand a little on the mechanisms you posit in the functional analysis. I’m especially interested in how the right color concepts get to be applied to the right sensory impressions. It seems a little mysterious how concepts (and thoughts) hook up with those impressions when all there is to the model is arrows and words. One worry to keep in mind is the one that David Rosenthal (and perhaps Kant before him) brought up in his illuminating comment: we should have some story to tell about how thoughts connect up with the things they are about, or expect these thoughts to be meaningless.

    Which brings me to a question that might be just a request for clarification of the ensuing discussion of HOT theory in light of Pete’s unicorn objection:

    Jake’s exposition of HOT theory above states that a mental state’s being conscious consists in one’s being conscious of oneself as being in that state. It seems to me that this is a thoroughly relational account, as the relation holds between a thought and the person that has that very thought. What is wrong with saying that this relation is ‘being represented,’ i.e., that the person that has that the thought is being represented by that very thought (as being in some mental state)?

  9. Richard, if I understand him, suggests that the so-called causal theory of reference is what’s needed to resolve issues about how HOTs might make one aware of mental states when they conscious. I would differ.

    I don’t know that there really is any causal theory. (Kripke sensibly calls it a *picture*, reflecting as I understand it a recognition that it doesn’t really rise to the standards of theorizing–even in philosophy.)

    Richard’s formulation is that “my thoughts about Obama are about him precisely because I am causally related to him in the right way.” Leaving aside what the right way is: If I’m causally connected to Obama in the right way, but my descriptive content (my dispositions to say what’s true about him) are all the lines of his having proved incompleteness of arithmetic, been at the Institute for Advanced Study, etc., no causal connection will do any good at all to have my thought be about Obama and not, e.g., about Gödel.

    Causal theorists tend to lampoon descriptivists as relying for reference on single or at least short descriptions for each referent and then appealing to Kripke’s Gödel-Schmidt case to show that that won’t do. But arguably I could be referring to Gödel even if he didn’t prove incompleteness if enough of the descriptions that figure in my Gödel thoughts landed.

    Richard concludes by saying we need both descriptivist and causal factors. But what would the causal factors do? What we need is to enhance the usual descriptivist paradies with the resources of conceptual-role semantics, providing suitably rich descriptive dispositions for me to think about each referent.

    In any case, a mixed theory would require knowing when to appeal to causal ties and when not, and it’s hard to see how to do that in a way that’s not ad hoc.

    And without that, invoking causal ties won’t by itself get around Pete’s Unicorn.

  10. Thanks to everyone for their excellent comments and to Pete for his very thoughtful responses: lots to think about here!

    I wanted to come in on the synchronic phenomenal sorites issue, if that’s ok.

    I thought it might be helpful to clarify what I see as the first part of the synchronic indistinguishability argument (SIA) against conceptualism.

    A useful distinction to bear in mind here is that between (i) sameness of representation and (ii) representation of sameness.

    These notions come apart. Suppose my experience represents A and B as having the same colour, B and C as having the same colour, but A and C as having different colours. If so then we have representation of sameness as between A and B, but we don’t have sameness of representation: A but not B is represented as having a different colour from C.

    Now distinguish two claims about synchronic indiscriminability:

    1) Synchronic indiscriminability entails sameness of representation.

    2) Synchronic indiscriminability entails representation of sameness.

    Suppose there are synchronic phenomenal sorites series. If so, this needn’t put any serious pressure on (2). If there are such series and if we accept (2), then at worst the result is that we have to allow for necessarily false perceptual contents. But that is no problem, because we already know that perceptual contents can be necessarily false (as with experiences of Escher scenes, etc.).

    The thought underlying the first part of the SIA is rather that synchronic phenomenal sorites series, if they exist, force us to reject (1). If there are such series, then we have to reject (1) because otherwise we’ll have a contradictory account of perceptual contents in sorites cases. In particular, we’ll have to say that the way that A is represented is the same as the way that B is represented, the way that B is represented is the same as the way that C is represented, but the way that A is represented is different from the way that C is represented.

    If there are synchronic phenomenal sorites series, then, we have to reject (1): we have to allow for indiscriminable differences in the ways that objects are represented to us. On the conceptualist view, differences in representation are constituted by differences in the concepts we deploy in experience. So if there are synchronic phenomenal sorites series, then the conceptualist must allow that objects can be indiscriminable even though there is a difference in the way that those objects are conceptualized.

    The second half of the SIA then tries to show that this is problematic for the conceptualist.

  11. I’m very gratified by the continued interest in my Unicorn Argument and have found quite helpful the recent comments. I wanted here to make some responses that are hopefully brief.

    (For a whole lot more, I refer interested parties to the published article (Mandik, P. (2009). Beware of the Unicorn: Consciousness as Being Represented and Other Things that Don’t Exist Journal of Consciousness Studies. 16(1). pp. 5-36 ) available here:
    http://www.petemandik.com/philosophy/papers/unicorn.pdf )

    I very much appreciate David Rosenthal’s candid remarks. I wanted to point out that there’s perhaps more agreement between us than he registers concerning the stability of the way we pre-theoretically speak about conscious states. I think David is right to be unsure about whether there’s much stability about our folk-talk and -thought on these matters. I would like to think that I’m not presupposing very much stability. I also grant that there’s a pretheoretic way of thinking and talking about conscious states whereby they are states we are conscious of ourselves as being in (the transitivity principle).

    But now I would like to point out where it is I think we start to disagree. I think that the transitivity principle is not worth depending on very much. The first reason I have for this is that whatever entrenchment it has can be explained by the observation that every conscious state that we are conscious of is one that just seems to prove the rule. Conscious states of which we are not aware are by definition not going to pop up on our personal radar.

    My second reason for thinking that the transitivity principle is not worth depending on much is that there are additional pretheoretic principles about conscious states, not all of which sit well with the transitivity principle and the theories of consciousness that have been based on it. While it’s important to connect with a pretheoretic way of talking about consciousness, the transitivity principle is not the only game in town. I think that there are other pretheoretic ways of thinking about conscious besides the transitivity principle. To mention a few:
    (1) Conscious states are states in virtue of which we are conscious.
    (2) Conscious states are states with which we are conscious of something.
    (3) Conscious states are states in virtue of which there is something its like to have them.
    (4) Conscious states are states that, when one tries to be conscious of them, one just is conscious of what the states are experiences of.
    It’s (2) that conflicts HOT-theorists’ contention that one can have an unconscious state that allows one to be conscious of something. And (3) conflicts with HOT-theorists’ contention that untargeted higher-order thoughts aren’t conscious. (4) is a version of the so-called Transparency or Diapheneity intuition, and it can be put into direct conflict with the transitivity principle.

    Anyway, I don’t favor an approach whereby we just pick one pretheoretic principle and build a theory around it. I don’t intend here to suggest that that is what Rosenthal is doing. I just want to be clear that I myself don’t have a favorite pretheortic principle. I think we should use the whole gaggle of pretheoretic principles to mark out, in a very rough way, what the subject matter is.

    I want to turn now and attempt to address some remarks that Michal makes.

    First, regarding ‘being represented’:

    Here’s an analogy that I like when thinking about this issue. Suppose I argued that whatever triangularity consists in, it can’t consist in blueness. I suppose that one of the most important premises in such an argument would be pointing out that some triangles aren’t blue. Suppose further that you presented me with some blue triangles and asked whether these triangles, the blue ones, can have their triangularity consist in their blueness. I would reply that the blueness of these triangles is irrelevant for assessing whether their triangularity can consist in blueness. What’s relevant is the other triangles, the ones that aren’t blue. It is clear that since these other triangles can fail to be blue without being disqualified as triangles, it’s true even of the blue triangles that their triangularity cannot consist in their blueness.

    So let’s turn to the question of thinking of oneself. This is something that is hard to fail at. If there’s a thought, there’s a thinker. So self-directed thoughts generally aren’t empty. So we have two relata: the thought and the person the thought allegedly is about. Why, one might ask, can’t these two relata enter into a representing relation with each other? This question strikes me as relevantly similar to the question about the blue triangles and it may receive a relevantly similar answer. The fact that in this case, the thing thought about exists is irrelevant. What’s relevant are the cases where you’ve got representing without the represented’s existing. It’s because of those other cases we know that in this case, even though the candidate relatum exists, a reprenting relation is not born to it.

    I’d like to turn now and say something about Michal’s request for further details on what sorts of mechanism might underwrite the arrows in the target article’s key flowchart. I take it that part of what Michal wants to suggest is that whatever is supplied is also going to go some way toward answering questions about whether representing is a relation to representeds. I think, however, the issues may be kept distinct.

    I do think that I need for there to be a plausible causal mechanism whereby some determinate shade of color can reliably cause, via a mediating impression, a deployment of a concept that we’d be correct to call a concept of that shade. I suppose that in very many cases, that mechanism is something that will be installed via learning. I’ve got no serious idea of how learning works though, so I don’t have further details to supply about these mechanisms. I hope, however, that the current state of the dialectic between me and the nonconceptualists doesn’t oblige me to supply further details. If they could make a case that there couldn’t be any sort of mechanism, or any plausible one, then my conceptualism would be in trouble. But so far, I don’t feel that I’m in trouble.

    Anyway, all this stuff about learning to reliably conceptualize stimuli strikes me as orthogonal to the question of the metaphysics of intentionality. It seems to me that there could be the sort of mechanism that I describe above regardless of what intentionality turns out to be, or even if there is any intentionality. I guess I would like to hear more from Michal about what he has in mind here and if I’m even interpreting is remarks correctly.

  12. Pete,

    Many thanks for your reply to my remarks, and I am sorry you had to do interpretation–I was aiming at being brief, but that sometimes gets in the way of being clear.

    On mechanisms:

    I don’t doubt that there is a causal process that triggers the application of concepts. I also don’t think that this bears all that much on intentionality.

    So let me just throw some things out there.

    I was expecting you to say something about how the relevant concepts are related to each other. These relations would be implicated in the kind of discriminations a subject can make about stimuli. And, if locutions such as …darker than… correspond to concepts, as I take you to believe, we should expect a web of relations between concepts that is structurally similar to the web of relations that exists between comparative color locutions.

    With that said, I wonder whether you think that the web of relations between concepts is constitutive of the concepts themselves. If yes, then it seems to me that the concept that makes it into the short-term memory buffer will have enough fineness of grain to make your explanation of diachronic indistinguishability not work. Just a hunch, for whatever that is worth.

    If those relations are not constitutive of the concepts, then I wonder how the store of concepts that interfaces with impressions implements the network of relations. It seems to me that to get a relevant web of relations on the non-constitutive view, the view would have to be supplemented with a story about how concepts hook up with the world, and that takes us beyond the straightforwardly causal process of concept application.

    With that aside, do you also think that there is going to be a straightforward causal account of how concepts get to be about stuff in the world?

    Regarding ‘being represented’:

    You say:

    “Why, one might ask, can’t these two relata enter into a representing relation with each other? (…) The fact that in this case, the thing thought about exists is irrelevant. What’s relevant are the cases where you’ve got representing without the represented’s existing.”

    What would be helpful to me is if you could tell me which of the following claims are false (if any):

    1) Mental states are properties of organisms.

    2) Mental states sometimes represent organisms.

    3) Mental states sometimes represent the organisms that instantiate those very states.

    4) Mental states sometimes misrepresent.

    5) Misrepresentation involves mis-assigning a property to what is being represented.

    If you think that 1-5 are true, then I’m afraid I don’t understand why a HOT cannot represent its owner as having a mental state she doesn’t have. I would think that it is the owner that is being represented in some way, not the mental state as such. This also seems to respect your criteria that relations hold between existent relata. I thought that is the thrust of Jake’s defense of HOT theory, and David Rosenthal was himself saying something along the same lines. If I am misrepresenting (;) their point the fault is all mine, but I hope you still can respond to my concern above.

  13. I cannot see what incentive one would have in the first place to defend conceptualism- what is the use of saying that the content of conscious states is conceptual? I think all this argumentation in favour of conceptual content, irrespective of the usefulness of having the ‘concept’ ‘darker than’ (and one might wonder what kind of ‘concept’ this is- is it like ‘further away’- and is the latter even a concept?

  14. Hi Pete, interesting stuff going on in here!!

    I think that the transitivity principle is not worth depending on very much. The first reason I have for this is that whatever entrenchment it has can be explained by the observation that every conscious state that we are conscious of is one that just seems to prove the rule. Conscious states of which we are not aware are by definition not going to pop up on our personal radar.

    This actually pushes for leaning on TP. What we need is some reason to think that there are conscious states that I am not conscious of. In fact we need more than that we need a theoretical account of what it could even possibly mean for there to be some state that was conscious and which there was something that it was like for me to have but which I was in no way conscious of myself as being in. On the face of it this looks like a non-sensical, confused way of talking. No one, not you, not Ned, not anyone, has ever tried to give an account of how this would work. Care to give it a try?

    (3) Conscious states are states in virtue of which there is something its like to have them. And (3) conflicts with HOT-theorists’ contention that untargeted higher-order thoughts aren’t conscious.

    Just to hammer (:0) the point home; it is not clear that there is a conflict here since it is nonsensical to talk about there being something that it is like for me to be in a state when I am not conscious of myself as being in that state. In fact it seems to me the other way entirely. When there is something that it is like for me I am conscious of myself as being in some state and when I am not conscious of myself as being in some state then it is not like being in that state for me…

  15. Hi Richard,

    My thought about (3) is that without tacking on a lot more epicycles, the HOT theorist would have to admit that HOTs are states in virtue of which there is something it’s like to have them. Consider: there are 3 creatures, A, B, and C with the same HOT. A has a true HOT, B has a false HOT, and C has an empty HOT. But since the HOTs are all the same, what it’s like to be A is the same as what it’s like to be B and also C. So, what state of those creatures is the state in virtue of which there’s something it’s like? It’s the HOT. What else could it be. And if (3) is true, then the HOT itself is a conscious state.

    Re: nonsense. Do you hold TP to be analytic?

    Re: trying. I think lots have tried and it strikes me as weird to claim otherwise so maybe I just don’t understand. How, for example, would Dreske’s 1995 book Naturalizing the Mind count as not even trying?

  16. Hi Michal,

    Re: being represented, I hope you’ll forgive me for being very brief about this. I’ve written an article that deals with precisely these sorts of issues at length and I’m reluctant to rewrite it here. But hopefully the following helps a bit: I regard 1-5 as all being true. I don’t deny that a HOT can represent its owner. I still deny that the owner instantiates a relational property of being represented. I really like my blue triangles argument, but I guess it wasn’t very convincing for you.

    Re: webs. I think you raise an interesting problem. I don’t think, ultimately, that the (noncompositional) concepts are going to wind up counting as fine grained. I think that we can find different sets of functional roles for impressions and the concepts whereby the impressions turn out to be way more fine grained than the concepts. Think of a machine that converts analog signals to low resolution digital signals. We can functionally define the finegrained inputs and coarse grained outputs. That there are causal relations to the finegrained inputs doesn’t automatically infect the outputs with fineness of grain.

    Re: aboutness, I don’t think it’s going to be either straightforward or causal (in the sense of ‘causal theory theory of reference’).

  17. Aspasia,

    I think that there are a lot of good motivations for conceptualism. One that I’ve worked out at length and like best is spelled out in my article, “Transcending Zombies”.

    http://www.petemandik.com/philosophy/papers/tz.pdf

    The line of thought is summarized in fn 1 of “Color-Consciousness Conceptualism” like this:

    The gist of that line of thought can be conveyed briefly as follows: States of knowledge are conceptual states. And our self-knowledge of our own non-zombie-hood requires an especially intimate relation between the conscious states that we have knowledge of and the states with which we know them. I argue that this intimacy can only be accounted for if the contents of conscious states are conceptual.

    Other motives for conceptualism might go like this:

    (1) we already have reason to believe that thought and belief is conceptual. It gives us a satisfying unification of the mind to show that sensory experience is also conceptual.

    (2) We can talk about and think about what we experience. The talking and the thinking is the expression of conceptualizations. If there were nonconceptual contents of experience, it would be inexplicable how anyone could think about or talk about such contents.

    (3) Our states being conscious in the first place is due to our having certain sorts of thoughts about them, thoughts that themselves involve the deployments of concepts.

    I don’t endorse all of these. (3) is pretty much the higher-order thought theory of consciousness, which I don’t like. But conceptualism cannot be dismissed as lacking motivation.

  18. I don’t know that there really is any causal theory. (Kripke sensibly calls it a *picture*, reflecting as I understand it a recognition that it doesn’t really rise to the standards of theorizing–even in philosophy.)

    I agree with this characterization of Kripke…but one of the virtues of Michael Devitt’s work consists in taking this intuitive picture and working it into an actual theory….

    Richard, if I understand him, suggests that the so-called causal theory of reference is what’s needed to resolve issues about how HOTs might make one aware of mental states when they conscious. I would differ.

    Actually I was suggesting that this would fill the gap between my thoughts and the world. As you rightly point out there has got to be some way in which my thoughts “connect” with their objects.
    The causal theory of reference offers a theory of how that is done. It also offers an nice answer to the so-called “problem of intentionality”: if ‘unicorn’ is like ‘tiger’ then my thoughts of unicorns are about whatever objects we “christened” with that natural kind term. In that case it may ultimately refer to a deformed goat (one theory of how the unicorn myth got started) or it may ultimately trace back to some imaginative act (as fictional reference works for Kripke); either way representation is a relation. On the other hand if ‘unicorn’ doesn’t turn out to be like ‘tiger’ but instead turns out to be like ‘vixen’ that is no problem either since all that it shows is that we need a mixed causal theory and not a pure causal theory (in Devitt’s terms if I remember them correctly). In that case ‘unicorn’ will simply be (analytically) equivalent to ‘(mythical) horse with a horn’ and if so representation is still a relation since when I think about unicorns I think about horses with horns, both of which I am related to in the right way.

    In fact Devitt argues that we must adopt a descriptive-casual theory of reference-fixing even for natural kind terms. if this is right then the way that the reference of ‘unicorn’ is fixed may well be a mix of the two above options. Something like this is what I think is going on with HOTs. When I have a HOT to the effect that I am seeing red I would cash that out as (1)

    (1) I think I am seeing Dthat red

    The ‘dthat’ is not meant to evoke Kaplan but is just a handy way of symbolizing what I mean. Namely that the higher-order state picks out the state that it is causally related to and characterizes it conceptually as well.

    If I’m causally connected to Obama in the right way, but my descriptive content (my dispositions to say what’s true about him) are all the lines of his having proved incompleteness of arithmetic, been at the Institute for Advanced Study, etc., no causal connection will do any good at all to have my thought be about Obama and not, e.g., about Gödel.

    Suppose that I am at an Obama rally. Suppose that someone points at Obama and says to me “that’s Obama”. Now suppose that I said in return (2)

    (2) Yep, I know. I really like Obama, he’s such an intellectual. Did you know that before he was president he spent time at Princeton’s Institute for Advanced Study where he discovered the incompleteness of arithmetic?

    Suppose my friend responds;

    (3) Man, I didn’t know that! Obama’s a genius!

    Do you really think that ‘Obama’, as used here, refers to Gödel?

  19. About thought compositionality: I didn’t want to rest too much on my skepticism about compositionality (explaining why I’m skeptical about it is a really long story). And, I didn’t want to deny that there may be reasons, like the one Pete offered, to think that thought is compositional. I only wanted to flag that we could put the conditions for conceptualism in a way that is neutral about the issue, which might be preferable. One simple reason it’s preferable is that some might have confusions about what exactly counts as a concept (cf. Aspasia’s question below), but there are less confusions about what constitutes an intentional state like a thought.

  20. It seems to me that all of these motivations for conceptualism, save for the complex reasons Pete gives in his “Transcending Zombies” paper, are compatible with there being nonconceptual aspects to conscious states. One can grant that knowledge or belief states are conceptual, that we have beliefs or knowledge about our conscious states, and yet deny that our conscious state are themselves conceptual. As Pete notes in his paper, we can have beliefs or knowledge about rocks without rocks being conceptual. So, the only real argument for conceptualism are those presented in “TZ,” which, if I understand them correctly, rely on there being that there’s an especially intimate relationship between our conscious states and our consciousness of them. But why should we think there’s such a relationship? Why should we think that our beliefs about our conscious states are any more secure than our beliefs about rocks? I know that it seems that way, but why think it is that way? In other words, why think that appearance and reality coincide in the case of our beliefs about conscious states? Pete says in the paper that it’s to ward off zombie worries (i.e. how can we know we’re not zombies unless we have this intimate relationship to our conscious states?). But if we had independent reason to deny the zombie intuition, is there any other reason to posit such an intimate relation? One independent reason to deny the zombie intuition is that one can conceive of zombies only if one assumes that our conscious states are only knowable (in their full glory) from the first person point of view. If one denies that supposition from the onset, then zombies aren’t conceivable. So what’s the need for intimacy about consciousness?

  21. Pete Mandik :

    Hi Richard,

    My thought about (3) is that without tacking on a lot more epicycles, the HOT theorist would have to admit that HOTs are states in virtue of which there is something it’s like to have them. Consider: there are 3 creatures, A, B, and C with the same HOT. A has a true HOT, B has a false HOT, and C has an empty HOT. But since the HOTs are all the same, what it’s like to be A is the same as what it’s like to be B and also C. So, what state of those creatures is the state in virtue of which there’s something it’s like? It’s the HOT. What else could it be. And if (3) is true, then the HOT itself is a conscious state.

    Ah, I think I see what is going on here…the problem is here: ” if (3) is true, then the HOT itself is a conscious state”. In one sense the HOT is the state in virtue of which there is something that it is like for me since there being something that it is like for me is constituted by having the relevant HOT (as your example shows). But that doesn’t make the HOT itself a conscious state, since there is no HOT about it. The phenomenally conscious state is the one that I am conscious of myself as being in. This is because a state’s being conscious simply consists in me being conscious of myself as being in that state so A B and C all have exactly the same conscious mental state (and it ain’t the HOT). This may be weird for Rosenthal since he has to say that the conscious state that C has is merely a notional state (but he really thinks that this is true of all conscious states I would say)…whereas for me the HOTs have complex demonstrative content and so (3) is virtually ruled out…

    Re: nonsense. Do you hold TP to be analytic?

    I am not sure if I think it is analytic, though I do think that it is something that has a strong intuitive pull. I have trouble understanding what it could mean for there to be some state that was phenomenally conscious but which I was not conscious of unless that simply means that were I to be conscious of it then it would be like such and such for me.

    Re: trying. I think lots have tried and it strikes me as weird to claim otherwise so maybe I just don’t understand. How, for example, would Dreske’s 1995 book Naturalizing the Mind count as not even trying?

    Dretske never, to my knowledge, gives an account of what a phenomenally conscious state that I am not conscious of myself as being in would be like. What he does is to tell us that a conscious state is one in virtue of which I am conscious of things and then goes on to argue that I can be conscious of red when I am not conscious of myself as so being conscious. But this doesn’t address the fundamental issue. When I have a phenomenally conscious experience of red that I am not conscious of myself as being in how do we know that it is phenomenally conscious? What reason is there to think that there is something that it is like for me in these cases? if there is nothing that it is like for me in these cases then why say that I have a phenomenally conscious state?

  22. Hi Pete.

    I guess that the structure of the second half of the SIA will depend on the details of the conceptualist theory that it’s attacking.

    Once the conceptualist concedes that objects can be synchronically indiscriminable but conceptualized differently, then some account is needed of this difference in conceptualization.

    The conceptualist might try to explain the difference by saying that different demonstrative concepts are deployed in experience. If so, then the SIA will try to undermine that proposal by appeal to Wittgensteinian hand-on-head considerations.

    Alternatively – and I take it that this is the approach that you would prefer – the conceptualist might try to explain the difference in conceptualization by saying that different comparative concepts are deployed in experience. If so, then a lot is going to turn on whether the worries that I raised in the first part of my comments are good worries.

    I’m wondering, though, if there’s a more immediate problem for you here, given the line you take in response to the DIA.

    Start by distinguishing two claims:

    SYNIND: Synchronic indiscriminability entails sameness of representation.

    DIAIND: Diachronic indiscriminability entails sameness of representation.

    Suppose that SYNIND is false, as we agree it must be if there are phenomenal sorites series. If so, then there can be a difference between the way that A is represented and the way that B is represented, even if A and B are indiscriminable when perceived simultaneously. But if such an indiscriminable representational difference is possible when A and B are perceived simultaneously, then it seems to me that it should also be possible when A and B are perceived at different times. I can’t see what we reason we might have to deny this. So it seems to me that if SYNIND is false, then DIAIND should also be false.

    As I understand it, though, your response to the DIA depends on DIAIND’s being true? (Correct me if I’m wrong here.)

    If this is right, then it seems to me that you need to deny the existence of phenomenal sorites series.

  23. Thanks very much, Pete, for your useful thoughts in response to my own. But let me correct the record a bit.

    You say you think I’m “right to be unsure about whether there’s much stability about our folk-talk and -thought on these matters.” That’s not really what I said. I said I thought that some things in our folk ways of thinking and talking about mental states’ being conscious are clear and others not.

    I offered as something very clear the principle that no mental state is conscious if the individual in that state is not in any way aware of it. You think that this transitivity principle “is not worth depending on very much.”

    You give two reasons. One is “that whatever entrenchment it has can be explained by the observation that every conscious state that we are conscious of is one that just seems to prove the rule.” But confirming instances are seldom if ever the right place to look in testing a conditional. That’s why I’ve stressed that entry into the basic folk-psychological idea is best stated as the contrapositive: that if one is *not* in any way at all aware of a mental state one is in, that state isn’t conscious. Disconfirming instances aren’t easy to find–I think there are none, because that conditional captures an extremely stable, well-entrenched feature of our folk-psychological notion of a mental state’s being conscious.

    Your second reason for not relying on the transitivity “is that there are additional pretheoretic principles about conscious states, not all of which sit well with the transitivity principle.” But the examples you cite of pretheoretic principles are, I think, unconvincing.

    (1) Conscious states are states in virtue of which we are conscious.

    I assume that the ‘we are conscious’ part alludes to a creature’s being conscious. But we count a creature as being conscious if it walks around, takes in sensory information, makes decisions, and the like. It’s question begging to assume that these things require conscious states. Decisions and sensing sometimes occur without being conscious.

    (2) Conscious states are states with which we are conscious of something.

    I see no folk-psychological basis for that at all, though it’s been ably defended by Fred Dretske as an alternative to the transitivity principle. People generally distinguish in talking about subliminal perception between being aware of a stimulus (which could be subliminal) and being consciously aware of it (which must be supraliminal). (2) is incompatible with that–and indeed with the fact that people naturally describe the distinction that way.

    (3) Conscious states are states in virtue of which there is something its like to have them.

    You think this is compatible with the transitivity principle, but not with the idea that it’s implemented by way of higher-order thoughts (HOTs). I don’t understand that. Is that because (3) implies that no state is conscious if there’s nothing it’s like for one to be in it? If so, (3) is probably at best poorly stated, and should say ‘include’ instead of the first ‘are’.

    (4) Conscious states are states that, when one tries to be conscious of them, one just is conscious of what the states are experiences of.

    This appeal to so-called diaphanousness, about which Moore himself was equivocal, doesn’t seem to help. For one thing, it likely appeals to attention–what do we *attend* to when we try to become conscious of something–which is not the issue. For another, it’s pretty clear that we’re often aware of being in mental states–though not perceptually, which the so-called intuition about diaphanousness tacitly but illicitly has in mind.

    So I’m not convinced that any of your alternative principles are well supported by folk psychology in a way that at all undermines the transitivity principle.

  24. Regarding (3), I agree that it is in virtue of the HOT that there is something that it’s like for one. But I wrote above that “I agree that what it’s like for two individuals with the same HOTs will be the same, but that does not entail that “what it’s like for the person” or the “consciousness” are properties of the HOT. HOT theory holds that the HOTs are themselves seldom conscious, though they are responsible for the first-order state’s being conscious.” The fact that it is the HOT in virtue of which there is something that it’s like for one does not entail that it is the HOT that’s conscious. Think of a runner running home in baseball (I borrow this example from David Rosenthal). If the umpire calls the runner out at home, then it is in virtue of the umpire that the runner is out. But the property of being out isn’t a property of the umpire, but a property of the runner. Likewise, if a HOT makes a first-order state conscious, it is in virtue of the HOT, but the state consciousness isn’t a property of the HOT, but of the first-order state (or person, depending on how you want to read the HOT claim).

    So it’s perfectly fine to say it is nonconscious states in virtue of which there is something its like for one, and I think the HOT theorist might be better off denying (3). But the reason the HOT theorist should be fine with denying (3) is that the phrase “what it’s like for one” is quite ambiguous and, I think, to a large extent misleading. As has come out in the discussion above, the HOT theorist wants to bifurcate the qualitative character of mental states, which are properties of first-order states, from the consciousness of those qualitative states (which is a function of the higher-order states) in virtue of which the first-order states become conscious. Once we separate consciousness from qualitative character (and thereby grant that states with qualitative character may occur nonconsciously), then it’s not clear what work “what it’s like” is doing anymore. If what it’s like should be identified with the qualitative property, then it’s a property of first-order states. If it should instead be identified with some property of persons (i.e. it is state consciousness), then it also doesn’t need to be a property of HOTs.

  25. Thank you, David. I’m happy to let you here have the last word on most of this since we have entered territory that’s already heavily trod. I do want to focus on one small thing, though:

    You write:

    Disconfirming instances aren’t easy to find–I think there are none, because that conditional captures an extremely stable, well-entrenched feature of our folk-psychological notion of a mental state’s being conscious.

    But should this be the other way round? Isn’t something well-entrenched because of a lack of disconfirming instances? It seems weird to say that it lacks disconfirming instances because it is well-entrenched. Am I in error to detect a weirdness here?

  26. Hi Pete,

    I’d said that I think we won’t find disconfirming instances of the transitivity principle because there are none, and that I think there are none because our folk-psychological notion of a mental state’s being conscious is embodied in the transitivity principle. That’s an inference to the best explanation: We can infer that that’s how folk psychology sees what it is for a mental state to be conscious because we find no disconfirming instances.

    And I was asking for similar folk-psychological support for your alternative principles:

    (1) Conscious states are states in virtue of which we are conscious.
    (2) Conscious states are states with which we are conscious of something.
    (3) Conscious states are states in virtue of which there is something its like to have them.
    (4) Conscious states are states that, when one tries to be conscious of them, one just is conscious of what the states are experiences of.

    My suspicion is that, especially with (2-4), they’re principles that sum up contentious philosophical theorizing, and not found in folk psychology.

  27. Hey Jake,

    Sorry, I almost missed this comment. There’s been a lot to keep track of.

    Anyways…
    I do like the umpire point. However…

    Be careful about WIL. If it’s a property of FOstates, then there’s nothing it’s like to be C, which contradicts the earlier statement that C is the same wrt WIL as A. If it’s a property of persons, then let’s ask ourselves what a state is. Isn’t it the having of a property at a time? So A, B, and C are in the same WIL state. And they’re in the same HOT. What non-ad hoc basis is there for denying that the WILstate is identical to the HOT state?

  28. Hello Pete,

    Thanks again.

    About ‘webs’:

    So when blue1 and blue2 are put side by side, we would expect blue1* and blue2* to be instantiated. Why are things different, on your view, when the blues are presented one after the other? Is it just a matter of not enough learning?

    About ‘being represented’:

    I want to understand your triangle argument. The triangle argument is this:

    1) those triangles are blue, but they are still triangles
    2) those triangles are not blue, but they are still triangles
    C) being a triangle doesn’t consist in blueness

    By analogy:

    1*) those empty HOTs are about their owner, but they are still empty HOTs
    2*) those empty HOTs are not about their owner, but they are still empty HOTs
    C*) being an empty HOT doesn’t consist in being about their owner

    If I am getting the analogy right, then let me point out that on the view presented premise (2*) is false. Every HOT, even in pathological cases, is about its owner. But, as I expect, I am getting the analogy wrong.

  29. Hi Pete,

    In your interchange with Jake, you wrote: “Be careful about WIL. If it’s a property of FO states, then there’s nothing it’s like to be C, which contradicts the earlier statement that C is the same wrt WIL as A. If it’s a property of persons, then let’s ask ourselves what a state is. Isn’t it the having of a property at a time? So A, B, and C are in the same WIL state. And they’re in the same HOT. What non-ad hoc basis is there for denying that the WIL state is identical to the HOT state?”

    A few thoughts. Is there being something it’s like for one to be in a state itself a state? I wouldn’t have thought so. I do think that the (allegedly) folk-psychological notion of there being something it’s like for one to be in a state is itself a matter of what higher-order awareness there is–i.e., of how one is aware of oneself as being in a state. It’s a matter of how one’s mental life appears to one.

    Your identification of what-it’s-like states with HOTs is, I take it, meant to draw a silly or absurd conclusion from HOT theory. But I don’t think that works.

    We posit subatomic particles, and then Eddington raises alleged, silly-sounding problems about whether the table is the commonsense table or the congeries of particles. Combining folk concepts with theoretical concepts often seems to have that result.

    HOTs are theoretical posits; what it’s like for one is (allegedly) a folk notion. If the identification is silly, you need to argue that it’s silly in a way that’s other than Eddington’s two-table problem.

    On the other hand, all that may be beside the point. You may be just arguing that the what it’s like for one should be the first-order state and not the higher-order awareness. I don’t see that. What non-question-begging reason is there to think that what it’s like for one should be a feature of a first-order state?

    You’ll perhaps say the non-question-begging is that what it’s like for one simply is the property of a state’s being conscious, and that it’s the first-order state that’s conscious (if any is).

    But then you need a non-question-begging reason to hold that what it’s like for one is not, instead, a property of the individual–i.e., how that individual’s mental life appears for that individual.

    Just a few thoughts.

  30. Peter, thank you so much for your reply.

    I think the deeper point of the nonconceptualist is to deny your assumption that all knowledge is conceptual. That is why we draw the distinction between knowledge by acquaintance and knowledge by description. Also it is not clear to me what conceptual content is, on your view. If you allow contradictory contents it cannot be belief content- so the kind of conceptual content you are talking about is not ‘conceptual content’- but of the nonconceptual variety which allows for contradictions like the waterfall illusion and stuff. With respect to the view that we can sometimes withdraw judgment concerning our experiential states- I think it really misses the phenomenology of perception which has a matter-of-fact character.
    With respect to 2) I do not see why a nonconceptualist in the end has to assume a TOTAL EXPRESSIBILITY thesis – of the sort I can describe everything I have experience of. It is exactly the point of a rather strong nonconceptualist that the translatability from experiential content to lignuistic content is not 1-1. Also the nonconceptualist can press hard the conceptualist to explain the data (like Kosslyn’s) that appeal to depictive or iconic representations for all types of problem solving. The argument from ‘economy’ does not appeal to me at all- because if we insist on economy we might miss certain interesting distinctions of significance in cognitive architeccture- ending up claiming that all psychological states are like beliefs- or require knowledge or implicit knowledge of some sort.

  31. Hi David, 
    I’m not seeing a clear way out of the trouble I posed for Jake. I’m afraid I don’t have a lot to say to help diagnose what my problem might be here. 

    I don’t mind saying that WIL is a property of persons. I wonder, though, if you and I differ over how to sort properties from states. I think that there are exactly as many states of things as ways things are. I think this because they are one and the same.

    In the problem case posed to Jake, persons A, B, and C are all the same wrt WIL. I suppose you are committed to saying they all have a property in common. They also all have the same HOT. Their HOTs differ wrt truth, but are the same wrt to content. If consciousness is how one appears to oneself wrt one’s mental life, then we can throw that into the mix too: A, B, and C all appear to themselves in the same way. But this leads to saying that they are all in the same conscious state.

    All these ways that A, B, and C are the same coincide so closely that I have a hard time seeing how to avoid the temptation to just identify those ways with each other. Their common states and properties all strike me as boiling down to the same commonality.

    (The closest thing to your theory that I can envision myself agreeing with holds that phenomenal character, WIL-ness, is one and the same as the conceptual content of a certain kind of thought, thoughts that thereby have state consciousness.)

    I don’t see that the problem posed is relevantly similar to how you describe the one concerning Eddington’s tables. I think I’m avoiding the sorts of equivocations that mixing folk and technical terminology leads to in the posing of Eddington’s puzzle. The terms I use in posing the problem all have a life in your theory, and I don’t see a clear way out, in the theory’s own terms, for avoiding the problematic identifications. 
     

  32. Hi Michal,

    Re webs, I’m mostly inclined to just chalk it up to learning.

    Re triangles
    I’d draw the analogy like this
    Blue triangles = thoughts about things that exist
    Nonblue triangles = thoughts about things that don’t exist
    Triangularity does not consist in blueness = being a thought about something does not consist in bearing a relation toward what it’s about

  33. Hi Aspasia,

    You’ve made very many highly interesting remarks, none of which I find at all convincing and each of which would require a separate paper to adequately address.

    One remark I do want to make here, though, concerns content and contradiction. It seems to me that the existence of contradictory contents in perception favors positing conceptual composition. Being contradictory requires having logical form and I don’t suppose anything not conceptually articulated can have logical form.

    I’m aware of no account of nonconceptual content that can accommodate contradiction. One account of nonconceptual content oft promoted is in terms of causal-informational relations, but since nothing in reality can be a contradiction, no representation can be a causal consequence of it. Another account is that nonconceptual content is pictoral or analog and represents via isomorphism or resemblance. But a similar problem arises, there being no contradictory states of affairs for pictures to resemble.

    I may very well be ignorant of the relevant literatures here, but what account of nonconceptual content explicitly explains the possibility of contradictory representation? I know of none.

  34. petemandik :
    I’m not seeing a clear way out of the trouble I posed for Jake. I’m afraid I don’t have a lot to say to help diagnose what my problem might be here. 
    I don’t see a clear way out, in the theory’s own terms, for avoiding the problematic identifications. 
     

    I’m afraid I don’t see what the trouble is, nor what’s problematic about the identifications. Suppose that there being something it’s like for a person to be in a (qualitative) mental state is identical with that person’s have a higher-order thought that that person is in the relevant state. What’s the problem?

    My last post, which doesn’t strike you as forceful or germane, was intended to explore possibilities as to what the alleged trouble might be. But since I haven’t managed to grasp what it is, why don’t you explain what’s problematic.

  35. Hi Pete,

    I think we’re just going around in a small circle.

    I wouldn’t identify a HOT with a conscious state unless one is conscious of oneself as having that HOT. I would identify each conscious state as a state one is conscious of oneself as being in. The HOT is the awareness of oneself s being in that state.

    I think I’ve said all this earlier in this thread.

    Let me repeat: ‘Conscious state’ refers to a state one is conscious of oneself as being in, not the consciousness of oneself as being in it.

    Maybe you know of a non-question-begging reason why that’s wrong. I guess I haven’t seen it.

    Here’s a question-begging reason. ‘Conscious state’ refers to a states with the property of being conscious. I suspect that that’s what you think. But I see no serious reason to think that that’s so.

    Here’s a reason to deny it. Everybody agrees that the consciousness of mental states is a matter of mental appearance–how one’s mental life appears to one. My formulation flows from that.

    If you think that that’s not so, I’d like to hear why.

    If you agree that that’s so, then you need to explain how it is that it fits with your surface-structure belief that ‘conscious state’ refers to a states with the property of being conscious–assuming that I’m right that that’s what you believe.

    An appearance of being in a state doesn’t always imply that there’s any such state that one is in. And even if there is such a state, its appearing to one that one is in it is not in any straightforward way a property of that state.

    So if you think that ‘conscious state’ refers to a states with the property of being conscious, I think you can’t accept that a state’s being conscious is a matter of mental appearance. And I think if you deny that, you’re denying the most basic folk-psychological thing about mental states’ being conscious.

  36. Hi David, 

    I do want to avoid a circular dance and also repeating myself. I offer the following remarks for clarity in the record about what I take my argument to be:

    I am very happy to here concede “that a state’s being conscious is a matter of mental appearance”. That particular way that you there word it is quite congenial to me.

    I can also go along with this, it seeming very much in the spirit of the Unicorn and all:  “An appearance of being in a state doesn’t always imply that there’s any such state that one is in.  And even if there is such a state, its appearing to one that one is in it is not in any straightforward way a property of that state.”

    Note, however, that on your view, its appearing to one that one is in such a state does suffice for what it’s like, whatever it’s like, and what it’s like remains the same across contexts in which the HOT is (A) true, (B) false, and (C) empty.

    Were I to identify a specific point where we part company, it seems to me to be right around here:

    It strikes me as enshrined in common sense to hold that whatever conscious states are, they darn well better be such that persons A, B, and C are in the same conscious state. And if not enshrined in common sense, then such that the denial sounds pretty gosh-darn odd to anyone with enough philosophical accumen to even understand what’s being said.

    Anyway, thanks for taking the time for discussing these issues with me. I feel like it’s aided my understanding of the current state of your thinking on all this.  

  37. Hi Pete,

    petemandik :
    Were I to identify a specific point where we part company, it seems to me to be right around here:
    It strikes me as enshrined in common sense to hold that whatever conscious states are, they darn well better be such that persons A, B, and C are in the same conscious state. And if not enshrined in common sense, then such that the denial sounds pretty gosh-darn odd to anyone with enough philosophical accumen to even understand what’s being said.
    Anyway, thanks for taking the time for discussing these issues with me. I feel like it’s aided my understanding of the current state of your thinking on all this.  

    I’m still not seeing what’s problematic here. I agree that A, B, and C are all in the same conscious state. But note that anything “enshrined in common sense” isn’t going to require any “philosophical accumen to even understand what’s being said.” Nobody among the folk talks of being in conscious states. People have conscious sensations, perceptions, experiences, thoughts, and desires. Do A, B, and C have, e.g., the same conscious experience? Why not? Does that imply they have the same experience? Not if that means something beyond their all being conscious of themselves as having that experience. If you think philosophical acumen requires otherwise, I don’t see how. None of this new in my thinking; it’s all in, e.g., Rosenthal 2005.

  38. Hi Peter, I am sorry you found none of my comments at all convincing. It is fairly clear to me that one way to understand nonconceptual content is as spatial content which does not require any concepts, (and perhaps spatial content might require only iconic representations, and not descriptive/conceptual representations.) As I understand it one argument in favour of nonconceptual content is Crane’s (The Waterfall Illusion) according to which the content of perceptual experience can be contradictory (thus when U look at the rock behind the waterfall illusion it is seen as being still and moving at the same time) (Crane 1988) Since any theory of meaning would not allow that a competent speaker can endorse p and not p at the same time but the content of looking at the Waterfall illusion CAN be contradictory, the content of perception differs from the content of belief (in so far as rationality consists in not endorsing p or not p or even withdrawing judgment from p.) So I do not understand your initial point that only conceptual content can be contradictory because it has logical form.
    Secondly, I still think it is phenomenologically implausible to suppose that as we endorse a thought or reject it, or withdraw judgment, the same thing happens in perceptual experience- where unless we have mastered some zen technique it is preety hard to withdraw from it-in an analogous manner.
    So I think the point is the opposite- whereas one cannot hold at the same time two contradictory conceptual contents one can hold at the same time one nonconceptual content with contradictory content.
    So that is why I think your conceptualism (and your view that all concepts are analogous to characteristic ‘nonconceptualist predicates'(like taller than further away darker than) cannot really be characterized as conceptualism.

  39. I assume this because the most well known argument against nonconceptual content (McDowell, Brewer) is that if it existed (why does this window become so small?) it would not be able to justify belief content- because it would not be ‘inferential’ or as U are putting it it would lack logical form. My view is that saying that conceptual content is propositional or ‘intentional’ is not very informative to be honest (Also I would say that attempts like Peacocke’s scenario contentsto demaracate nonconceptual content- are misguided)

  40. I would like to return to issues that Charlie earlier raised and rehearse some solutions that strike me as having some promise.

    The issues center on two problematic theses that Charlie formulates:

    SYNIND: Synchronic indiscriminability entails sameness of representation.

    DIAIND: Diachronic indiscriminability entails sameness of representation.

    I continue to think that Charlie makes a convincing case that if synchronic phenomenal sorites series exist, then SYNIND needs to be abandoned.

    Earlier in the comment thread, Charlie tentatively describes my response to DIA as depending on accepting DIAIND. I’m thinking now that perhaps DIAIND can be rejected while preserving the essence of my conceptualist model.

    One counterexample to DIAIND that’s pretty clearly consistent with my general functional analysis involves very long delays between presentations. When delays are longer then the term of the short-term memory buffer, a failure to discriminate can quite easily be consistent with my model even when positing distinct conceptualizations for distinct presentations. Even a red chip and a blue chip can wind up not being discriminated diachronically if the conceptualization of the first chip fades from memory before the presentation of the second chip.

    Of course, the nonconceptualist can try to raise trouble by coming up with problematic situations that can’t be dealt with by memory fade, situations involving relatively short inter- presentation delays.

    One direction the nonconceptualist can turn for a source of problematic stimuli involves what I’ve been calling diachronic phenomenal sorites. (I should mention that something like the following problem was brought to my attention by Nick Saint at the CUNY Grad Center in 2007):

    Consider what we can call “a forced march” through a phenomenal sorites series wherein colors are presented one at a time. If the delays between color presentations are shorter than the term of the memory buffer, then it seems tempting to me to say that diachronic indistinguishabilty is going to need to be explicated, on my functional analysis, by sameness of representation. However, here’s where a problem arises: each member of a pair of adjacents, for all adjacents in the series, is diachronically indistinguishable, and thus what’s conceptualized as red at the start of the march is going to lead to a RED conceptualization of unique yellow at the end of the march.

    This is a highly implausible consequence. Something needs to give. My first inclination is to think that some further rejection of DIAIND is what’s needed.

    One way to envision an implementation of this can be spelled out by reference to the boxology from my target article. The conceptualist can posit an instability of the representations in the boxes or noise in the channels that input to one or more of the boxes . One representation (impression or concept) in one of the boxes can change in a brief moment without any change in external stimulus.

    Suppose the march is from red to orange, and somewhere in the middle of the series are two adjacent colors c13 and c14. If there’s no instability at the level of impressions, then c13 and c14, will have corresponding impressions c13* and c14*. If there is instability at the level of impressions, then c13 might trigger a c13* which quickly changes to a c14* without any change in the color presented. If there’s instability at the level of conceptual deployments, one way this might go is that c13 gives rise to a conceptualization of RED in conscious experience but is misremembered with an ORANGE in the short-term memory buffer. When c14 is presented, it gives rise to an ORANGE in conscious experience, which, when compared to what is now an ORANGE in the buffer, gives rise to an erroneous judgment that c13 and c14 are the same color.

    However the hypothesis is implemented, the central feature of this hypothesis is that there will be at least one color, somewhere in the series, that gets represented by two different representations at two different times within a small window of time.

    This general way of dealing with sorites is pretty much the solution to the sorites paradox offered by Raffman in her “Vagueness without Paradox” (VWP) (which I guess is kind of ironic given my main motivation of responding to her “Persistence of phenomenology”). Her VWP treatment is largely discussed by her in terms of pair-wise presentations through the forced march, but the adaptation to my one-at-a-time presentation doesn’t change the core features of the approach.

    At bottom, when two colors are indiscriminable, there’s only one general way in which this is so: when there are factors leading to an absence of a judgment that the colors are different. On some occasions this absence will be due to the upstream presence of sameness of representation. But this won’t be the case on all occasions. For presentation delays exceeding short-term memory, the absence of a judgment of difference will not be due to an upstream presence of sameness of representations of the two colors. For presentation delays not exceeding short-term memory, there may be instabilities in the representations deployed, so that at least one member in a color pair winds up being represented in two different ways during the relatively small window of time. We might describe these latter sorts of cases as involving, within a small window of time, both a sameness of representation and a difference of representation.

  41. Hey Pete. Thanks for calling my attention to that point again. I’m sorry–there’s so much going on here that I think I just missed it. First, let me say that I’m happy, as you say, to agree to disagree about how to interpret those hard cases like dental fear.

    As for unification, I see how adopting intentionalism offers a nice way to explain what’s common about all mental states. Brentano would have been right that the mark of the (conscious) mental is intentionality. And, another way of putting the point, I think, is that beauty of intentionalism is that the sensory (qualitative mentality) is supposedly really hard to explain, whereas intentional content is putatively better understood, so reducing the former to, or explaining the former in terms of, the latter seems like a good thing.

    But I worry about several things. For one thing, folk psychology does draw a distinction between the qualitative and the intentional. And, folk psychology surely draws a distinction between, say, believing, which is intentional, and perceiving, which is qualitative. The intentionalist must then provide an explanation of what distinguishes believing from perceiving solely in terms of conceptual content. I often find these sorts of explanations rather strained. Distinguishing nonconceptual qualitative character from conceptual intentional content makes this very easy.

    I have broader worries about how, if we do not posit nonconceptual sensory properties, we could explain how creatures develop intentional states to begin with, but that’s a longer story that I’ll set aside.

  42. I should say that, in sum, there are two marks of the mental: intentionality and qualitativity (for lack of a better word). I think that’s unification enough.

    I’m in the process of writing something expanding David Rosenthal’s account of qualitative properties according to which they are, like intentional properties, representational, but in a nonintentional way. So, I do think there’s a way in which these two kinds of properties fit together nicely. But I still think they are distinct and separable.

  43. Thanks Jake,

    There’s a lot there that I agree with and a lot of that is spelled out in my initial responses to the commentaries. One thing I’d like to focus on here…you write:

    For one thing, folk psychology does draw a distinction between the qualitative and the intentional. And, folk psychology surely draws a distinction between, say, believing, which is intentional, and perceiving, which is qualitative.

    I agree with the second sentence (under certain readings of “qualitative,” though not others. But I get the impression that you think this is an additional point beyond the first sentence. Is that correct? If so, what’s the evidence, independent of belief/perception distinctions, that the folk distingusih the qualitative from the intentional?

  44. Hey Pete and all. Great discussion.

    Pete—here’s a crack at reconstructing what you’re saying to David:

    1. A, B, and C are all in the same (type of) conscious state

    2. If a subject is in a conscious state, then a state of the subject has the property of being conscious

    3. In case C, the state the subject is aware of herself as being in does not exist

    4. So that state can’t have the property of being conscious (argument ad unicornum)

    5. The only other state A, B, and C share is the HOT

    6. So the HOT must be the state with the property of being conscious in A, B, and C.

    I take it David, among other things, is challenging premise 2. All agree that A, B, and C share WIL, and are all in the same conscious state. But why can’t that just be a global property of A, B, and C—that they are all aware of themselves in the same way, and so all share the property of being in the same conscious state? Why must the property of being conscious be tagged to something more fine-grained here? In particular, why must it attach to a thought-sized state? You seem to want to put consciousness “into” a state, and that’s just what puts pressure on the idea that consciousness is a matter of appearances. Putting consciousness into a state makes it a matter of the reality of that state—in conflict with the TP.

    For the record, I tend to dislike 4 more than 2, because I think theories of properties and intentional objects is pretty thin stuff. This seems just a matter of metaphysical bookkeeping, one that will be pushed around by whatever is the best empirical theory we’ve got. Further, there seems a good commonsense way of talking about inexistents having properties. Santa has the property of being fat, Sherlock is smart, the fountain of youth is desirable, etc. Whatever we end up saying about these cases, apply it to the property some nonexistent (though represented) states have: the property of being conscious. All that amounts to is that we’re aware of ourselves as being in these states, so it’s a pretty thin property on anyone’s theory. Call it a “smroperty” if that helps keep the unicorns from haunting your dreams!

    Anyway, I found your arguments against the anti-conceptualists compelling. 😉

  45. Josh, you are awesome (and not to mention, my second-favorite type-q materialist)! Thanks for the note.

    I think that 1-6 does a pretty nice job of representing my line of thought and really helps shine some light on the dialectic.

    I have some remarks about your comments on 2 and 4.

    First off, regarding your attempted resistance against 4, I don’t think that truth-in-fiction stuff is going to cut it in any way that lets you HOT guys and gals go around bragging about how you’ve got an implementation of the transitivity principle (TP). Here’s a reminder of the gist of what I say about this in my “Sherlock does coke” discussion in my Unicorn paper. ‘Sherlock does coke’ is true in a sense because Sherlock is depicted, in the story, as doing coke. However, it’s not true, in *that* sense of “true”, that Sherlock is written about by Sir Arthur Conan Doyle, since it’s not depicted, in the story, that Holmes is written about by Sir Arthur Conan Doyle.

    Let’s connect this to issues about TP and HOTs via the following analogy. If you have a HOT, the content of which is something like “I am perceiving a red square”, what’s depicted in the “story” here is you perceiving a red square. It is not depicted in the story that you are conscious of yourself as having a perception of a red square. Nor is it depicted in the story that you are having a higher-order thought that you are perceiving a red square. So, it looks to me that remarks about Santa aren’t really going to help connect TP with the properties that are notionally “instantiated” by nonexistent first-order states, what you call the “shmroperties”. If you want to say that state-consciousness is a notionally instantiated shmroperty, I don’t have a beef with that. (I realy like consciousness anti-realism.) But I think that this problem I raise about Sherlock means you can’t get shmroperty-ism from TP.

    Ok, now let’s talk about 2. I think that a lot of the CUNY Cogsci crowd–and I include myself–is reluctant to get too metaphysical. I’m perfectly happy saying that the metaphysics of states and properties, heck, the metaphysics of everything, is just people making stuff up as they go along.  And that if the check can’t be cashed by science, it ain’t worth even the paper it’s written on. But it strikes me that this reluctance to be too committal about states and properties can cut both ways in the current dialectic. The way it cuts in my favor is that, since I don’t take the surrounding metaphysics very seriously, I don’t see any real reason against supposing things like 2.

    I suspect that at this point The dialectic over 2 turns into a kind of burden-of-proof tennis. When you ask me questions like…

    Why must the property of being conscious be tagged to something more fine-grained here?  In particular, why must it attach to a thought-sized state?

    … I very much want to shift the burden back to you to say why we should take seriously a metaphysics that says that we shouldn’t.

    I guess at this point y’all think the burden is born by TP and you want to hit the ball back to my court. And then I’ll refer you back to my remarks about your remarks about 4… .

    Anyway, before I say much more beyond this, I’ll wait to see what your reactions to my reactions are so far.

  46. I’m sorry to jump in the middle of this conversation, but this is just jumping out at me and screaming loudly.

    You say:
    “It is not depicted in the story that you are conscious of yourself as having a perception of a red square.”

    Isn’t the “I” in “I am perceiving a red square” just the part of the story that is relevant to being conscious of yourself as having a perception of a red square?

    If the content of the HOT was “some perceiving of red square is going on” then you would be right in drawing the analogy with the lack of Doyle in a story about Holmes doing coke. But, since that is not the stipulated content of the HOT, then the analogy doesn’t work–or so it seems to me.

  47. I agree! Three cheers for Josh

    Hey everyone, sorry I have been absent from the discussion…I have been sitting in committee meetings dealing with the self-study we are doing for our re-accreditation…jealous? Then I got to grade a bunch of papers…On top of that last night I was out at the premiere party for Fly Girls (the new reality show that Jen and I have a small part in. So my brain is a bit frazzled but I can’t resist jumping in here.

    First I would just point out that one may just as easily challenge 3. In fact David suggests this in places when he says that every case we could describe as C we could re-describe as B. That is we can always say that the HOT is misrepresenting some existing state. This is in effect the strategy that the complex demonstrative HOT content I suggested above fits nicely with. The first-order state that is “pointed” at is the state that is misrepresented. David doesn’t like the causal aspect of this account so I am not sure how we would determine which first-order state was being misrepresented.

    As for the case against 4. I think that is pushed David does have to end saying that all intentionality works this way. Cf his earlier remark above;

    When one regiments to within an inch of one’s life, I agree with Quine’s eventual position that there’s no way to ascribe intentional states that’s genuinely relational, as opposed to notional, using here Quine’s own terms.

    Finally, though I wouldn’t endorse this move I don’t see what the argument Pete gives against the fictionalist account is supposed to be. To implement TP, according to Pete, we need to be able to say which state has the property of being represented. The answer is: a state will have that property only if the story says that it does. Pete takes this to mean that the story will have to include the fact that I am conscious of the first-order state whereas the story just says that I perceive a red square. But this is exactly what we want! Think of the “I am conscious of myself as” part of the HOT as the “author” of the story and the “perceiving a red square” part as the story itself, just as you suggest. So which state does the HOT “say” I am in? The seeing a red square one. Since TP tells us that the conscious states are the ones we are conscious of ourselves as being in we then can say that the red square state has the property (or shmoperty) of being a conscious state even though it is a notional state.

  48. Michal and Richard,

    Let me try to put the Sherlock argument a different way. Tell me what you think of this:

    i. A shmoperty is instantiated iff it is mentioned in the story.

    ii. Being conscious = being represented by a HOT

    iii. HOTs tell stories that never mention being conscious or being represented by a HOT

    C. Being conscious Is not a shmoperty.

    Comments
    Re: i. Josh sez so, and Josh is awesome.
    Re: ii. This is the higher- order thought theory of consciousness, bound in a nutshell, subscribed to by you guys
    Re: iii. Consciousness-conferring HOTs that counterexample this claim are so rare by the admissiion of you guys that we can safely pretend they don’t even exist

    Tell me why C doesn’t follow from i-iii, you guys!
    (Richard seems to think that HOTs are third-ordered, with his “I am conscious of myself…” formulation, so I guess he has an out. But I don’t see what motivates his formulation. (Besides his being a drummer who always wants to take it to the next level, of course!))

    😉

  49. Richard,

    Re 3, and what DR would say, being not a causal theorist: I’m guessing he thinks the ultimate regimentation of HOTs involves attributing to them existentially quantified content. This will have the consequence in making both the B case and the C case equally false and equally concernining states that do not exist.

  50. Pete, thanks.

    Sherlock’s being depicted by Doyle=me being represented as having a perception of a red square.

    And “me being represented as having a perception of a red square” doesn’t make it into experience, so it isn’t a part of the story HOTs tell.

    That puts pressure on the story that claims that TP is a part of our folk-psychological talk. We would seem to never have any non-theoretical evidence for it.

    But it seems to me that the evidence for TP is not to be found in experience, but in the way we talk. Take these sentences:

    (1) “It’s raining”
    (2) “I believe it’s raining”

    We can all agree, I hope, that they have different truth conditions, but their performance conditions are the same. However, “It’s raining but I don’t believe it’s raining” is self-contradictory. This is a part of folk theory if anything is.

    The only way to explain the contradiction is to say that expressing mental states and reporting on them is a differnt sort of speech act. So, (1) expresses thought, and (2) reports on it. And it is this that suggests that the mental state that is expressed by (1) is represented by another mental state, not the story that experience tells us.

    This comes back to David P’s initial comment in this thread about the mental attitude of HOTs being assertoric. We know they are such thoughts because of the illocutionary force of the speech acts that express them. We would not be able to utter sentences like (1) if they were not represented by something with the right mental attitude.

  51. So, Michal, I’m thinking that there’s something not quite right with this:

    And “me being represented as having a perception of a red square” doesn’t make it into experience, so it isn’t a part of the story HOTs tell.

    I don’t want to say that failure to get into the story is due to failure to get into experience. Failure to get into the story is simply failure to be one of the conceptual constituents of the HOT. It seems then that the stuff about Moore’s paradox etc leaves unscathed my premise iii.

  52. Pete, thanks again. This is a really cool exchange.

    And I see. I thought that the story=experience. I take it then that story=conceptual constituents of HOT. From your iii you are worried that “being conscious or being represented by a HOT” isn’t a part of the conceptual constitents of a HOT.

    But that’s the point of a relational account of consciousness. Consciousness is explained as a relation, so it doesn’t have to be a part of the relata.

    Take as example the explanation of gravity as attraction between bodies with mass. Do we need to say that gravity is in some way a constituent of either body to explain it?

    Doing so doesn’t seem to be helpful to me. Actually, it sounds a lot like Aristotle’s theory of intelligent falling, where bodies move toward the ground because of an inherent disposition to do so. Aristotle’s theory got people messed up about gravity for a long time. Thinking that consciousness is a part of conscious mental states got people messed up (since Aristotle) for a long time, too.

    It seems to me that you thinking that ‘being conscious’ or ‘being represented as’ needs to make it into the conceptual constituents of a HOT is similar in thinking that gravity needs to make it into the object that exerts gravitational pull on something.

    But, since being conscious is explained relationally, we would not expect ‘being represented as’ or ‘being conscious’ to feature as a part of the conceptual constituents of a HOT. I guess you will come back with Sherlock or A, B, C to tell me that it has to.

  53. Hi Michal,
    I’m glad you think the exchange is cool. I’m enjoying it a lot, too.

    I think you anticipate me correctly: I’m going to come back with Sherlock.

    I think that at this point in the dialectic it won’t do for you to lean on relationality. As I see it, the reason we are talking in terms of “stories” is because in Josh’s response to #4 (the argumentum ad unicornum) he wanted to back off of making a relational claim about what consciousness is and instead make it notional. All the stuff about Josh’s “shmroperties” is supposed to make being conscious notional instead of relational so as to dodge unicorn worries. My Sherlock argument (i-iii) was supposed to raise trouble for the notional move.

    To sum up what I see as the most recent segment of the dialectical trajectory:
    I take it that Josh’ll grant this: Being conscious is either relational or notional.
    then Pete says: Well, unicorn stuff makes it real weird to say it’s relational.
    Then Josh says: ok, so it’s notional
    Then Pete says: Well, Sherlock stuff makes it real weird to say it’s notional

  54. Do I have the property of being awesome or just the schmroperty? (Answer: yes.)

    Ah, what fun.

    So, about the burden-tennis on 2. I say that no one has the burden here: it’s metaphysics, so it’s just book-keeping after the real game of science has ended. The point being, you don’t get to beat up us poor, innocent HOT folks with some argument about properties–that’s not the type-Q way to deal with things. Let’s instead see how the science plays out, that is, how well the HOT theory handles the empirical evidence, fits with commonsense ways of fixing the data, makes interesting predictions, etc. If it comes out ahead on all that (say, in 1000 years or so), then we’ll figure out the right theory of properties (or of property talk). So there’s no need to fear the unicorn–the metaphysicians will corral him by and by, y’all.

    As for shmroperties, why do you have such a restrictive view of shmroperties? Being explicitly mentioned in a story is hardly the only way to acquire shmroperties. They might be implicit in a story. Or they might be posited by a theory explaining a story. Consider Ponce de Leon (constantly on like robotron). He was looking for the fountain of youth. So the fountain of youth has the property of being looked for by PdL. That’s true, I take it, even if no one ever knew that’s what PdL was up to, if PdL never said it aloud, etc. Fiction is a red herring here. Now it’s true that some representing or other must go on for there to be shmroperties, but the shmroperties thus instantiated (shimstantiated? Oy…) need not be explicitly represented. Or so goes my theory of shmroperties.

    About what the TP commits one to, etc. I don’t think the TP entails any of this. The TP is just some commonsense thingy: it picks out a reasonable everyday way of distinguishing between conscious and nonconscious states. We then try to figure out how this works in us. We thus posit HOTs to explain the TP. Now, it may turn out (as evidenced by this debate, perhaps), that we need to go back and revise or modify what we think about commonsense after we’ve got a viable theory up and running. So the TP may not commit us to anything about shmroperties, but the best theory of the realization of the TP does, so shmroperties are in. So the TP is a starting point, but it’s not the ending point. We explain how we’re aware of our mental states, so we explain the TP. That we commit ourselves to shmroperties in doing so may strain the “commonsense” of “folk metaphysics” but who cares about that? (Besides all of us, say 😉 )

  55. Heya Josh,

    A lot of this strikes me as acceptably deflationary and I can see myself going along with it. However, something is bothering me in your penultimate paragraph.

    So, Fair enough, plausibly, notionally instantiated properties don’t need explicit representation. However: For the move to implicitness to work for you, you need some kind of analog to being written about by Doyle. I don’t mind saying that it’s implicit in the story that Holmes has an intact spinal cord even though that is not explicitly stated about him. But I don’t see how it’s at all analogously implicit that Holmes is written about by Doyle.

  56. I’m glad you agree with a lot of what I said, Pete. I do think, as you’ve mentioned a few times and we say in the commentary, that we do see eye to eye on a lot of these things, and that’s a good thing.

    But I did have in mind that my above points were distinct. So here are some ways that I take it folk psychology distinguishes the qualitative and the intentional. First, FP holds the contents of intentional states to be truth evaluable (the belief content that snow is white is either true or false), whereas I take it that mental qualities aren’t truth evaluable (the bluish quality of a sensation isn’t true or false).

    Second, FP has it that intentional contents are sentence-sized (snow is white); by contrast, mental qualities are term-sized (bluishness). Lastly, FP has it that intentional states have a two-property structure of content and mental attitude (such as belief or desire) toward that content, whereas purely qualitative states (if you believe there are such things) have no such analog structure because they exhibit nothing like mental attitude.

    I guess the question to you, then, is how intentionalism can accommodate these folk-psychological platitudes.

    I’d like to reiterate that I think this debate about the plausibility of intentionalism is, though extremely interesting, I think somewhat orthogonal to the issue of conceptualism. Pete, I’m wondering, though maybe you addressed this above and I missed it, whether you think the view that David and I put forward in the commentary counts as a version of conceptualism (i.e. non-intentionalist conceptualism) or if you think you have to be an intentionalist to be a conceptualist. Perhaps this is a verbal matter, but I’d be interested in your thoughts. Thanks again for the neat conversation!

  57. Actually, I think you have the shmoperty of awesomeness and the property of shmawesomeness 😉

    I think this is what i was trying to say above (though said much better!)

    Let’s put it like this.

    0. TP=(platitude) A conscious mental state is a mental state that I am transitively conscious of myself as being in
    1. To be transitively conscious is to represent something in an appropriate way
    C1. Therefore, to implement TP is to be able to say which states have the property of being represented (this is supposed to be Pete’s buildup to the argumentum ad unicornium)
    2. Having the property of being represented=being transitively conscious of myself as being in that state=having a suitable HOT (best theory of how TP is implemented)
    3. I am transitively conscious of myself as being in notional states (arg. ad uni)
    C2. Therefore notional states have the property of being represented and so have the property of being cosncious (revision of common sense)

    It looks like Pete would have to deny 2. to avoid C2. but he hasn’t given us any reason to think that 2. is false (and the Sherlock case doesn’t do that)

  58. Hi Richard,

    Maybe I was misreading you, but you wrote:

    Think of the “I am conscious of myself as” part of the HOT as the “author” of the story and the “perceiving a red square” part as the story itself, just as you suggest.

    I figured you meant it literally that part of the HOT is “I am conscious of myself as” yielding a HOT which is the thought that “I am conscious of my self as perceiving a red square”.

    Assuming that this is what you intend the HOT to be, here’s what’s third-ordered about that.

    First, I assume you are going along with the analysis of “conscious of x” whereby it is a (certain kind of) representing x. Hereafter I’ll just drop the “certain kind of”, so conscious of x is representing x.

    Second, I assume n-ordering is determined by number of representations of representations. So, zero order is a non representation. first order is representation of a non-representation. Second order is a representation of a first order representation. Third order is a representation of a second order representation.

    Putting this altogether.

    Richard perceives a redsquare -> Richard first-order represents a redsquare

    Richard thinks of himself as perceiving a redsquare -> Richard second-order represents a first-order representation

    Richard thinks of himself as being conscious of himself as perceiving a redsquare -> richard third-order represents a second-order representation

  59. Hey Richard,

    Actually, my beef with this is 3. It seems like a misuse of the word “as”. I assume, and so do a lot of the HOTties that, if one is conscious of x as being P then one must apply to x a concept of P. If I’m conscious of Jake as being in his twenties, I have to apply to Jake the concept of being in his twenties. But it strikes me as deeply implausible that very many people besides us freaks go around applying the concept of a notional state. So it strikes me as deeply implausible that anyone is transitively conscious of themselves *as* being in notional states.

    So, C2 has been blocked. And…

    …you have be Sherlocked.

  60. Thanks Jake,

    Re the FP stuff, I appreciate you spelling that out. I agree that it’s important to keep in mind the orthogonality of intentionalism and conceptualism, so I’ll be brief about the intentionalism bit.

    First, I’m not super sure that stuff is in FP. I don’t have any arguments proving that it isn’t, so I’ll just move on to the next point. Second, insofar as that stuff is in FP *explicitly*, then it seems as such to the folk, which requires their conceiving of it as such, and their conceptualizations are adequate to account for its seeming presence. Their conceptualization’s being true is not required. Third, if that stuff is not in FP merely explicitly, but is instead implicit and need be theoretically posited to explain a bunch of stuff, I have my doubts (*cough* Unicorn *cough*) about why we need to regard that stuff as counting among the contents of consciousness But hopefully everyone is sick of talking about that by now . (I recall Rosenthal once saying something at Brendan’s that paraphrases roughly as this: No one ever abandoned a theory because there was an argument against it. Theories are abandoned in favor of superior competing explanations.)

    Ok, conceptualism: As best as I understand the view you and David P hold, the three of us can easily be equally conceptualists. So, here’s a way of formulating color conceptualism that can hopefully help make explicit the orthogonality of the Great Unicorn Wars.

    (1) For each determinate shade that a creature is able to, at time t, consciously perceive something as having, the creature must apply, at time t, a concept (compositional or not) of that determinate shade.
    (2) For each color-determinable that a creature is able to, at time t, consciously perceive something as having, the creature must apply, at time t, a concept (compositional or not) of that color-determinable.

    Maybe nonconceptualists will object that this formulation stacks the deck against them, since arguably the construction “perceive…as” implies concept application. I don’t know. But maybe we need to reformulate by replacing “consciously perceive something has having” with “consciously perceive”. I’ll leave that issue alone for now.

    I take it that the HOT story about consciously perceiving something as vermillion goes like this. I have this first-order sensing which has a sensory quality or whatever you guys say. Then I have a HOT which is a thought like “I am perceiving vermillion”. We can show the same determinate shade to a color-sighted yet uneducated numbskull, Dummy, who doesn’t know vermillion from brick-red, and Dummy with have a HOT such as “I am perceiving red”. So, there will be a determinate shade that Smarty is able to consciously perceive that Dummy cannot.

    But I have my doubts about whether I’m representing your position correctly here. I take it you want to say something along the lines of this: While Dummy, lacking a concept of vermillion, cannot think of himself as perceiving vermillion, he can apply to his firstorder reaction to vermillion a thought along the lines of “I am perceiving red”. Due to the miracle of the Transitivity Principle, this will suffice to make his firstorder response to vermillion a conscious sensing of vermillion and he’ll be consciously perceiving vermillion even though he is not conscious of himself as perceiving vermillion. I’m not totally sure I’m representing your position correctly in part because I’m not sure what you think of all the relations between (a) perceiving a vermillion square, (b) perceiving a square as vermillion, (c) consciously perceiving a vermillion square, and (d) conscious of oneself as if perceiving a vermillion square. Specifically, I’m not super duper sure which of (a)-(d) you guys think requires the application of a concept of vermillion.

    Of course, HOT people think (d) requires VERMILLION. But I’m not HOT. I’m cold. In the world according to Pete, and what’s important to the present paper, is that (c) requires the application of the concept of vermillion. If you guys are on board with saying that ( c ) requires VERMILLION, (and rereading your commentary, it looks like you are) then you can be conceptualists without being intentionalists.

  61. Hi Pete,
    I want to follow up on B&P’s question about why you don’t like the demonstrative concept reply to the indistinguishability argument. For what it’s worth, I think the demonstrative concepts reply is sufficient to defend color consciousness conceptualism. I think part of the issue depends upon how we should characterize the dialectic between non-conceptualists and conceptualists. The non-conceptualists use the indistinguishability argument to argue for the claim that in principle our color concepts could not be as fine-grained as our color experience. In Evans’ and Peacocke’s arguments, the in principle claim about color concepts is defended by the claim that our color concepts are limited to common color terms, like ‘blue’ or ‘red’. In Raffman’s argument, the in principle claim about color concepts is defended by the claim that our color concepts are limited to those concepts we can remember, namely identify and reidentify. I don’t remember what Kelly’s claim about concepts is, but it’s something that also commits to some in principle claim about concepts… While I like your response by way of comparative concepts of color, I wonder whether the response to the in principle argument by way of demonstrative concepts doesn’t diffuse the argument better by giving an in principle reply, rather than providing a reply that focuses on the details of our color concepts. (Another way to put this is that you also make some commitment to an in principle outer limit about color concepts, namely that concepts are bounded on the outside (forgive the metaphor) by comparative concepts, to which the non-conceptualist can reply that experience is beyond that limit.) So, the demonstrative concept reply is that we could have (there is no principled reason why not…) demonstrative concepts for shades of color, like THAT SHADE OF BLUE. The quasi-Kantian reply is that insofar as these demonstrative contents function as long-enough term recognitional capacities, then they 1) serve the function that is demanded by the indistinguishability argument and 2) count as concepts. In your response to B&P, you suggest that the demonstrative reply engages in “an unmotivated profligacy of posits.” But, you seem to be assuming a kind of mental particulars view of concepts that for instance, Evans wasn’t committed to. I don’t see why a deflationist response by way of appeal to concepts as recognitive abilities rather than as mental particulars couldn’t assuage that worry. Also, in your response to Pelling you say, “A natural way, then, to characterize the opposing side, which I’ll just call “nonconceptualism,” is as the view that conscious experience of color is more fine-grained than the aforementioned conceptual repertoire.” But, I think this delivers too much to the non-conceptualist, because the argument is that it is more fine-grained than a conceptual repertoire *could be*. I understand that you are not concerned with a priori conceptual arguments, as you point out in your initial paper. But, then why would you concede *what I take to be* apriori conceptual claims about concepts to the non-conceptualist in the dialectic? You might think that Evans, Peacocke, Raffman, Kelly, et al have captured some essential property of concepts, but I’m not sure why you would think that.

    Thanks,
    James

  62. Hi James,

    Thanks so much for your comments.

    Here are some initial reactions. Please let me know if you think I am missing points you wanted to raise. I’m not confident that I’m connecting with everything you raise here.

    One thing I think that motivates me, and perhaps connects with some of your remarks about what in principle may or may not exceed our conceptual grasp, is that I’m impressed with the power of a combinatorial view of concepts. (By the way, I take this to be entirely consistent with what Evans says about abiding abilities, and not necessarily assuming Fodorian mental particulars, but what I have to say about this specifically is in progress so I shall set it aside.) A second thing that motivates me is the Wittgensteinian head-on-the-hand remark and what I see to be the disappointing “blindness” of demonstratives. A third thing that motivates me is to give a satisfying account of how it is that these colors can make a proper appearance in phenomenal consciousness. I know that this is all very vague and hand-wavy, but it strikes me that the combinatorial view can do a better job than the demonstrative view in satisfying these three motivations.

    An analogy that impresses me, though I haven’t run it by anyone else yet, is the following: Suppose someone were to object to a conceptualist account of *thought* by saying that there are more states of affairs or propositions that we can think about than we have concepts for. It strikes me that it would be an exceedingly lame response for the conceptualist to lean on a demonstrative construction like, THIS_state_of_affairs. It seems to account insufficiently for the appearance of the state of affairs in my mental life. And as for accounting for an indefinitely large number of states of affairs: with an appropriate stock of recombinable elements, it’s not clear that there *will* be an in-principle limit of what can be conceptualized.

    Another line of thought that I have about what might in principle exceed our conceptual repertoire which, I confess, is still in progress and so is unlikely to be very convincing, goes like this: If the nonconceptualist is posing their challenge as concerning a state of affairs concerning perceptible colors that obtains *in principle*, then this can be restated as a state of affairs that *conceivably* obtains. But nothing conceivable outstrips our conceptual grasp.

    Anyway, I would say that, ultimately, I am not uninterested in the a priori aspects of the issue. But I deemphasized the a priori aspects in the paper because I thought that what I had to say about them was not as ready-for-prime time as what I had to say about the empirical issues.

    The history of this project is that it grew as an offshoot of a paper still in progress that develops a transcendental argument for conceptualism based our knowledge of the character of phenomenal consciousness. Along the way I got obsessed with discriminability-based objections to conceptualism, and took a break from the transcendental argument to develop the present off-shoot. This is how papers make babies.

    I’ll stop there for now and wait to see how much you thinks this helps, if at all, connect with your comments.

  63. Oh dear, it’s gonna be tricky catching up! I shouldn’t have waited so long!!

    But first things first: some responses to Pete’s responses (thanks for the amazingly prompt response, Pete! Sorry I couldn’t get a serious look at them earlier!)

    1) ON CONCEPTUALISM & CONCEPTUAL CONTENT
    I tried to steer clear of this issue in my comments, but I’d like to point out that the formulation offered by Jacob and David—and modified by Pete in terms of combination of concepts in his response—falls short.

    One way to see this is by considering the following possible view: suppose I’m a nonconceptualist and I believe the representational content of experience to be entirely non-conceptual (it’s not determined or constituted by concepts in any way) and yet I endorse the following constraint on perceptual experience: for any object, property, relation, etc, represented in the nonconceptual content of my experience, I must entertain a conscious thought, the content of which is fully conceptualized, about that object, property, relation, etc., at the time of the experience, so that I must possess a concept for that object, property, relation, etc. Call this the thinking constraint. My satisfying such a constraint entails my satisfying the two clauses in Pete’s—and Jacob and David’s—characterization: (1) I possess a concept for everything I experience, and (2) I deploy it whenever I experience what the concept picks out. And yet, the content of my experience isn’t conceptual at all, only the content of my accompanying thoughts is.

    But, I take it, conceptualists like McDowell and Brewer (pre-2006) typically make a claim about the nature of perceptual content—to the effect that it is fully determined or constituted by concepts—and they wouldn’t rest content with the weaker claim that I must think (conceptually) about everything I experience. I don’t know of anyone who endorses the view I just sketched, but it’s consistent, and it shows that clauses (1) and (2)—as well as Pete’s (1’) and (2’)—fail to entail conceptualism, as it is usually understood.

    2) ON THE ROLE OF CONTEXT IN PETE’S ARGUMENT
    Again, I’m not sure I understand Pete’s proposal and response (§3). The suggestion was that, over time, it’s possible to present two distinct (yet highly similar) shades of blue and to ensure that they are presented in the same perceptual context. If so, there’s no ground for Pete’s claim that, over time, the two different shades are in fact presented in same way, since there is no difference in context between the presentation of blue1 at t1 and of the presentation of blue2 at t2. Given this, I’m not sure I see how this (“There are many plausible mechanisms that may be appealed to by the conceptualist whereby the difference between blue1 and its white context is insufficiently different from the difference between blue2 and its white context to trigger different concepts.”) addresses the issue.

    A couple of things:
    (i) to repeat, the suggestion was that, at t1 and t2, there is no difference in context because the white background, the lighting conditions, the perceiver’s focus of attention, can remain constant from t1 to t2, even the shade of blue presented is different. Pete mentions various “plausible mechanisms” which, presumably, should induce relevant changes in the context. But presumably, those mechanisms are such that, one can design a case where they are held constant too.

    (ii) it’d be nice to hear what these mechanisms are, by the way!!

    (iii) I entirely agree with Pete that the difference between blue1 and its white background at t1, and the difference between blue2 and its (same) white background at t2, is “insufficiently different”. So what? The issue is whether there’s a difference in perceptual context between t1 and t2 and I don’t see that there is—if there is one, according to Pete, we need to hear what it is. Without that, there is still no ground for the claim that the two distinct shades of blue are presented in the same way, due to differences in perceptual contexts. And it’s not clear why, exactly, the difference between each shade of blue and its respective background is relevant.

    (iv) in fact, Pete seems to grant that the contexts are the same between t1 and t2. He shifts the issue to whether the difference between shade blue1 and its respective white background at t1 is the same as the difference between shade blue2 and its respective white background at t2, and admits that these differences between blue shade and background are different, only “insufficiently” so. I had assumed that the white backgrounds are exactly the same from t1 to t2, but the shades are different, which is presumably why the difference between blue shade and background is slightly different between t1 and t2. So, here, he seems to agree that the contexts are the same, and the shades different. Again, that was just presupposed and stipulated in my example.

    (v) Why does it matter whether the difference between the shade blue 1 and its white background at t1 and shade blue2 and its same white background at t2 is “insufficiently different”? Admittedly, the difference in colour between blue1 and blue2, when presented diachronically (and, as a result, the differences between these different shades of blue and their identical white background), may be too small to be noticed by a perceiver. Everyone can agree with that! (In particular, those of us who think there is diachronic fineness of grain can insist that there are differences in who the two shades of blue look between t1 and t2, which may go unnoticed—though that’s largely irrelevant to the cogency of the argument from the fineness of grain of experience, which presumably shouldn’t rely on this point!) The point was only that, if the perceptual contexts are the same between t1 and t2, shifts in perceptual contexts provide no ground for Pete’s claim that the two shades are perceived in the same way—since there are none!

    (vi) the point, then, still stands: there are situations where different fine-grained shades are perceived at different times, but where the perceptual contexts remain the same. In which case, differences in perceptual context do not support the claim that different fine-grained shades are perceived in the same way. Since Pete admits (in §5 of his response to me) that a proponent of diachronic fineness of grain need not assume that different fine-grained shades are always presented differently over time, but only in some (many) cases, it seems to me that Pete still hasn’t said anything that undermines the assumption that there can be diachronic fineness of grain. Now I suspect from what Pete now says in his response that what he’s really trying to do is something a bit different (of which more below). So the only point of criticism so far is that to present his case as one against diachronic fineness of grain based on context effects is a case of mis-advertising. It’s still compatible with this that Pete does in fact have some response that works against AFG.

    (vii) on a side-note, all the above apply to Pete’s example in §2 of his response to me: his example provides no reason to think the two shades of blue look the same when presented at different times, since the respective contexts can be the same. At the very least, then, the claim that they do look the same can’t be based on the suggestion that the different shades are perceived in different contexts, and so look the same. What is true is that a perceiver may not be able to notice that the two shades look different. But again, this is not something a proponent of AFG can rely upon in arguing against conceptualism. All this, I think, suggests that the action in Pete’s response to AFG lies elsewhere.

    So what does Pete’s proposal really amount to? My sense is that it’s something like the following: Pete writes that the difference between shade blue1 and its white background at t1 and shade blue2 and its white background at t2 is “insufficiently different”. Insufficiently different for what? For the subject to notice the difference, I take it, and to form a discriminatory judgment regarding the chromatic difference between blue1 and blue2. This need not be about how perceptual contexts can affect how things look (since the contexts can be the same). It’s about how what figures in the conceptual content of experience is determined by what the subject notices and identifies or recognizes. Thus, conceptualists can admit there is synchronic fineness of grain because subjects can discriminate (and notice) the differences between distinct (yet highly similar) specific/determinate colour shades. However, they can reject diachronic fineness of grain on the ground that subjects aren’t very good at noticing and, hence, forming discriminatory judgments about, fine-grained differences over time. Putative explanations for the latter fact may be varied, but it’s likely to have something to do, inter alia, with how visual memory loses its fineness of grain over time. Everyone can agree on that point! It need not have anything to do with changes in the perceptual conditions over time, since they can remain constant.

    Pete’s conceptualist, I think, is perfectly entitled to make this sort of move. The dispute between conceptualists and non-conceptualists will then be over whether how things look is determined by what perceivers notice. And that there can be such a dispute is at least unlikely to make AFG more cogent.

    The real question, however, is whether rejecting diachronic fineness of grain in this way (or, in fact, in any other way) helps to block AFG. I don’t see that it does. Here’s why.

    3) DOES AFG NEED DIACHRONIC FINENESS OF GRAIN AT ALL?
    I suggested it doesn’t, and Pete doesn’t seem to have addressed this point. Here’s a version of the Kelly/Raffman argument as I understand it:

    (1) at t, S conceptualizes the fine-grained difference between shade A and shade B thus, according to Pete: BLUE-DARKER (for A) and BLUE-LIGHTER (for B), say (allow for complex combinations of comparative concepts, as Pete suggests, but the details don’t matter here).
    (2) S possesses a fine-grained concept such as BLUE-DARKER for shade A at t, only if S can reliably re-identify shade A as being the same BLUE-DARKER shade S experienced at t, in different contexts at different times, and in different situations (situations where, in some of them, A will be brighter than the other shades presented simultaneously).
    (3) S cannot reliably re-identify shade A as blue-darker in different situations over time.
    (4) Hence, S doesn’t possess the concept blue-darker for shade A.
    (5) Hence, conceptualism is false.

    In his response to Jacob and David (§5), Pete notes that, though he thinks concept-possession has something to do with memory, he rejects the re-identification constraint. In which case, he rejects premise (2). I couldn’t agree more! But then, there’s nothing really new ABOUT Pete’s response TO AFG. Appeal to determinable-cum-comparative colour concepts (instead of demonstrative concepts) doesn’t really make a difference, as this version of the argument shows. And appeal to context-effects against diachronic fineness of grain (or to noticeability constraints on how things appear) doesn’t really serve to block AFG. If anything, the fact that subjects are unable to discriminate different fine-grained shades over time, and the fact that the same shades can appear differently at different times, are considerations which can be brought to bear in support of premise (3) in this version of AFG: they explain why reliable re-identification is not something we’re very good at. But it doesn’t show that (2) is false. So really, like everyone else, Pete wants to resist (2)—the rest is largely irrelevant insofar as diagnosis of AFG’s failure is concerned.

    (note that this diagnosis of the situation doesn’t really change if AFG is formulated with a weaker re-identification constraint, such that S must be able to identify repeatedly shade A as BLUE-DARKER in different situations over time—and not to re-identify A as the very same shade S encountered at t. In those situations where A is presented with C and C is darker than A, it seems S will have difficulties correctly applying the same concept to A in different situations, or so I presume).

    More interestingly, Pete might rightly insist that determinable-cum-comparative concepts need not be governed by such re-identification constraints. Still, one would like to know why! Perhaps, the suggestion is that, as with demonstrative concepts, determinable-cum-comparative concepts are highly context-dependent (they depend on the other shades presented at the time), and that’s why we shouldn’t expect a subject who possesses such concepts to be able to re-apply the same concept (such as BLUE-DARKER) to the very same shade in any circumstance. Perhaps, but in any case it would be nice to have some sort of non-ad-hoc reason for this way of denying (2).

    4) ON CONTROLLING FOR CONTEXT-EFFECTS & THE “NEW EXPERIMENT”
    I’ll try to be brief. Sure, if the subjects are told that they are being presented with only two colours, they will be able to tell reliably over time which of the two shades is brighter (or darker). So there are cases where subjects can successfully re-identify the same shade over time. But this is completely irrelevant for two reasons: (a) the issue, to repeat, was whether subjects can re-identify not the relatively darker (or brighter) shade (which they can do even if they’re not told that there are in fact only two shades, and even if they’re presented with lots of different shades of different hues over time), but whether they experience the very same shade, and (b) the fact that we can design some settings where subjects succeed at such diachronic discriminatory tasks is irrelevant, since we can design many experiments where subjects fail pretty badly to re-identify the very same shade over time, which is all AFG needs—especially when they are presented with many more shades, or wrongly told that they are.

    5) ON THE(SYNCHRONIC) FINENESS OF GRAIN OF EXPERIENCE
    I’m not sure I understand Pete’s responses to (a) my demand for a detailed proposal of the conceptual content of an experience where you’re synchronically presented with 6 distinct shades of green (§6), and (b) the example of diachronic fineness of grain I give (§4).

    I would have thought that the following was common-ground about synchronic fineness of grain:
    (i) you experience each of the 6 different shades and its specific colour, as distinct from the other 5.
    (ii) you experience all (or most) of the relations of comparative difference between all 6 shades—that is, for each shade, you experience its relative differences with the other 5.

    I assumed that, by conceding synchronic fineness of grain, conceptualists are typically willing to grant that the information conveyed in the content of conscious experience encompasses information about (i) and (ii). But now, if the content of experience is conceptual, and for each different specific shade, you possess a particular concept for that particular specific shade (and likewise for different properties and relations), so that the information that’s conveyed in experience can be entirely captured by concepts the perceiver deploys in her experience, I would have thought that an account of the synchronically fine-grained conceptual content of experience would at least include the following:
    (i*) how different colour concepts (presumably, 6 different concepts) are applied to each distinct shade.
    (ii*) how distinct comparative colour concepts are applied to each relation of difference/similarity between one particular shade and the other 5.
    (iii*) how all these concepts are combined together to capture (i) and (ii), but also combined with other concepts about the locations of the 6 shades, as well as their respective backgrounds and surroundings, and their spatial and chromatic properties and relations.

    Clearly, Pete’s proposal (“a row of greens, each darker than the next”) doesn’t even attempt to meet that challenge. In which case, it seems, we’re talking past one another when we say that Pete grants to the proponent of AFG there can be synchronic fineness of grain! Now, it strikes me that most conceptualists usually seem to take the challenge of synchronic fineness of grain as how to conceptually capture (i) and (ii) fairly seriously—I think McDowell and Brewer do, at least.

    Now, Pete can indeed characterize the dialectic any way he likes, though it did initially seemed as though he was prepared to grant synchronic fineness of grain, only not diachronic fineness of grain. In his response to Charlie (§1), however, it seems as though all the action now is on how much synchronic fineness of grain there is. By resisting assumptions such as (i) and (ii), as I now understand he would, the sort of conceptualism you get doesn’t strike me as very plausible at all. For one thing, the content expressed by “a row of greens, each darker than the next” strikes me as significantly poorer in information than what my experience of such 6 shades actually conveys—a lot of very different experiences can share that content. If so, I lose my grip on what exactly Pete is trying to achieve by answering AFG in such a way. After all, insofar as synchronic fineness of grain is concerned, he only suggested that he could capture everything in terms of determinable-cum-comparative concepts (again, that may be right, it’d be nice to see how exactly, though). He did not argue, or provide any reason, for the claim that experiences are in fact less fine-grained synchronically than non-conceptualists assume.

    In this respect, I also think Pete is far too quick in accusing his opponent of begging the question—again, in response to Charlie (§1). When describing the synchronic fineness of grain of experience, non-conceptualists never rely on the assumption that there’s some information that’s more fine-grained than the concepts a perceiver possesses. The description of the fineness of grain of experience is supposed to be a piece of phenomenological data, which both sides can agree upon, as most conceptualists seem to. That’s the first premise in the argument. The task of the non-conceptualist is then to provide an argument as to why normal subjects don’t have the conceptual resources to capture that fineness of grain, and that typically goes via some assumption about constraints on concept-possession. That’s premise (2). But typically, the description in premise (1) need not rely in any way on premise (2)—typically, it’s the other way around.

    Likewise, my description of what synchronic fineness of grain amounts to—claims (i) and (ii)—did not in any way rely on the assumption that perceptual content is more fine-grained than the conceptual repertoire of normal subjects. Indeed, I think AFG fails, and I think demonstrative concepts are as fine-grained (if not more) than perceptual contents. That is, insofar as fineness of grain is concerned, I’m happy to grant everything about fine-grained concepts to conceptualists. What I’d like to hear is how exactly all these fine-grained concepts combine to one another to form a whole conceptual content—the whole content of the experience. It’s not an objection (not yet, at least), just a request for more detail! In other words, I’d like to hear more about the structure of such conceptual contents, if they have any, that is. (My suggestion that, if they involved different comparative concepts, it was possible to have two distinct conceptual contents, arranged differently, for what seems to be the same perceptual content, is prompted by the same sort of worry. Pete (§7) seems happy to grant that there can be two different conceptual contents, but then doesn’t address the worry that the content of experience really seems to be the same in both cases.)

    One last note on this: nothing Pete says against Charlie there applies to my point. I can grant to Pete that, if there exist, say, 25 distinct shades of blue, our experiences need not represent any such shade as BLUE1, BLUE2, …, BLUE25, but only as DARKER-BLUE and LIGHTER-BLUE in a given context. So I’m happy to grant that experience need not be as fine-grained as all the chromatic differences that there in fact are, but only as fine-grained as all the chromatic differences that are currently represented in a given experience. Still, I’d like to hear how, when you simultaneously experience 6 or 12 different fine-grained shades of blue, the conceptual content of your experience represent these 6 or 12 different shades and their differences, assuming—with claim (i)—that such conceptual content will involve distinct chromatic concepts (whatever they are) for each distinct shade. In my comments, I presented a case where it seems you can simultaneously experience the differences between 12 different shades of green. Again, my question to Pete is: what are the 12 different colour concepts you deploy in that case exactly?

    On a related note about diachronic fineness of grain, I agree with Pete (§4): my example where 2 sets of 6 simultaneously presented shades of green are presented one after the other doesn’t raise any special problem. If Pete can account for the conceptual content of an experience of 12 simultaneously presented distinct shades of green (with 12 distinct colour concepts, one for each distinct shade), then he can just break up that content in two, and ascribe the first part to the content of the experience of the first 6 shades at t1 and the second part to the experience of the other 6 shades of green at t2.

    The point of that example, however, was to suggest that there can be, in addition to synchronic fineness of grain (6 shades simultaneously perceived as different), cases of diachronic fineness of grain (6 shades perceived simultaneously and perceived as different from 6 other shades presented earlier). I thought Pete denied such diachronic fineness of grain. But in his response to this case (§4), he now seems to grant that a subject can be aware that the 6 green shades she is now experiencing are slightly different from 6 green shades experienced previously–so the different sets of shades are not perceived in the same way after all. So I was wrong: he doesn’t deny diachronic fineness of grain after all!! (Though he did say I had characterized his view accurately on that score!)

    6) on introspection & extra layer
    No, the claim wasn’t that it’s possible to perceive without introspecting (quite possible indeed!). The claim was that when I introspect my consciousness, it’s possible to introspect lots of conceptual contents (the conceptual content of the thoughts I’m having, whether they are about what I see or not). But it seems very hard to introspect any conceptual content beyond the contents of these thoughts.
    • I can introspect different conceptual contents about the scene in front of me, but the scene as it appears to me doesn’t seem to be changing as I entertain different thoughts with conceptual content about it.
    • I can introspect various conceptual contents about the scene in front of me, but I could introspect thinking those contents even with my eyes closed, so there seems to be something in addition to those thoughts with conceptual content I can have with and without experience—by the way, premise (4) didn’t say there was nonconceptual content, only something extra which wasn’t just the conceptual content I can also entertain in a thought.
    • of course, I can introspect an experience as such, and a thought as such. So I can introspect the content that George is blue as a thought when I think it, and I can introspect the content that George is blue as the content of an experience when I have the experience (let’s grant), but there are lots of different ways in which George being blue could appear to me visually, which are introspectible, and yet aren’t captured by the content that George is blue, or the content that I experience George as being blue.
    • if Pete could tell us what the conceptual content of 12 distinct simultaneously presented shades of green was exactly, it seems as though it should be fairly easy to introspect that one is indeed having an experience with that conceptual content. So why isn’t it just obvious whenever I introspect my experience of these 12 distinct shades?

    Perhaps, Pete and I experience things differently, though I think it’s part and parcel of non-conceptualism to suspect that conceptualists, when introspecting, typically confuse the contents of their accompanying thoughts for that of their experiences. Though there’s probably not much to be gained for either side by running this line of thought. My bad!

    There’s a lot more to say, but I’ve already said too much! Maybe later. Thanks again to Pete for the prompt and detailed responses!

  64. Pete Mandik :

    Hi Richard,

    Maybe I was misreading you, but you wrote:

    Think of the “I am conscious of myself as” part of the HOT as the “author” of the story and the “perceiving a red square” part as the story itself, just as you suggest.

    I figured you meant it literally that part of the HOT is “I am conscious of myself as” yielding a HOT which is the thought that “I am conscious of my self as perceiving a red square”.

    Assuming that this is what you intend the HOT to be, here’s what’s third-ordered about that.

    First, I assume you are going along with the analysis of “conscious of x” whereby it is a (certain kind of) representing x. Hereafter I’ll just drop the “certain kind of”, so conscious of x is representing x.

    Second, I assume n-ordering is determined by number of representations of representations. So, zero order is a non representation. first order is representation of a non-representation. Second order is a representation of a first order representation. Third order is a representation of a second order representation.

    Putting this altogether.

    Richard perceives a redsquare -> Richard first-order represents a redsquare

    Richard thinks of himself as perceiving a redsquare -> Richard second-order represents a first-order representation

    Richard thinks of himself as being conscious of himself as perceiving a redsquare -> richard third-order represents a second-order representation

    Yeah, I think that we were talking past each other at this point. I agree with the way you are ordering stuff but I only meant the second-order one.

    Pete Mandik :

    Hey Richard,

    Actually, my beef with this is 3. It seems like a misuse of the word “as”. I assume, and so do a lot of the HOTties that, if one is conscious of x as being P then one must apply to x a concept of P. If I’m conscious of Jake as being in his twenties, I have to apply to Jake the concept of being in his twenties. But it strikes me as deeply implausible that very many people besides us freaks go around applying the concept of a notional state. So it strikes me as deeply implausible that anyone is transitively conscious of themselves *as* being in notional states.

    So, C2 has been blocked. And…

    …you have be Sherlocked.

    I think I see what the worry is here. But why isn’t it enough that we are transitively conscious of ourselves as being in a state & (as it happens) that state is a notional state. So we are conscious of ourselves as being in a state which as it happens is notional but we are not applying the concept of a notional state. Substitute this for (3) and the argument still goes through…Sher(un)locked!

  65. Hi Richard,

    You write:

    But why isn’t it enough that we are transitively conscious of ourselves as being in a state & (as it happens) that state is a notional state?

    There are two main challenges facing the HOT theorist that prefers the notional to the relational route, but wants to take you up on your suggestion here.

    The first is that they need, for this to not just be an ad hoc stipulation, some principled independently motivated grounds for saying when a property is and isn’t notionally instantiated. Santa has a hat because he’s explicitly represented as having a hat is one way to motivate notional instantiation, but that won’t help here. Sherlock has a spinal cord because he’s explicitly represented as walking without assistance and it’s implied by walking without assistance that that one has a spinal cord is another way to motivate notional instantiation, but that won’t help here either.

    The second challenge is that if the HOT theorist is going to follow your suggestion, they need to make it hang together with all the other ways they lean on “…as…” formulations. What it’s like is determined not by the way the first order state *is*, but by how the higher order state represents it *as*. If a first order perception of green is represented *as* a first order perception of red, then what it will be like will be precisely how it’s represented *as*. If the first order state happens to be made out of exactly one trillion molecules, this won’t figure in what it’s like unless something about one trillion molecules figures into how things are represented *as* being. When crass anti-HOT-ists object that it doesn’t seem like they have HOTs, the HOT-ist replies that this is because it is seldom that there is a HOT that represents them *as* having a HOT.

    But if the HOTist starts relaxing these *as*-based restrictions so as to allow notional states having the notionally instantiated property of being represented even though it is not represented *as* being represented, then why are some of the notionally instantiated properties constitutive of what it’s like and some are not? Why doesn’t it seem as if there are HOTs even when there is no third order HOT? etc etc.

  66. Philippe,
    Thanks for another batch of very useful and thorough remarks. I hope I can make a bit of progress here on clearing things up. Unfortunately, for many of your remarks I need to think much further about them and will be setting them aside for now.

    What’s conceptualism?
    Philippe’s formulation of conceptualism yields a conceptualism that I agree with, and highlights continuity between my account and McDowell/Brewer’s. However, there are continuities between my account and the higher-order thought theory of consciousness that seem worth highlighting and that seem just as targeted by nonconceptualist arguments as are McDowell/Brewer/Mandik conceptualisms. What position is best to apply the label “conceptualism” to? I’m basically happy following Philippe’s recommendation here. (But for highlighting both sets of continuities in the future, it might be worth bifurcating conceptualists into two camps. The first camp holds a matching thesis: there are concepts deployed for each element in experience. The second camp holds a constituency thesis: each element in experience just is the content of a deployed concept.)

    Context (and memory)?
    I think a lot of what I’ve said so far about context has perhaps not been put in a very clear way, so here’s another stab at trying to get clear in my thinking about it:

    For any two colors that one fails to discriminate diachronically, there are two general possible explanations. One is that the failure is due to memory failure. The other is that the failure is due to perceptual failure.

    Regarding memory, even if the colors are very different, like red and blue, with a really long delay (days, years) one may very well forget the first color and thus, on presentation of the second color, be in absolutely no position to discriminate it from the first. However, memory-failure explanations seem implausible for very short delays, especially when the stimuli are conceptualized. For colors that we non-controversially have concepts for, like red and blue, a short-delay diachronic discrimination failure seems implausibly due to memory failure. While I don’t want to impose a hard-core reidentifiability criterion on concept possession, it does seem that for very short delays, if one does conceptualize the stimulus, one tends to be able to remember it.

    (Are there clear counterexamples to such a claim about conceptualization and memory? )

    To instead explain the failure as due to a failure of perception, it needs to be made plausible that despite being different colors, the way the first color is perceived is the same as the way the second color is perceived. (I’m here setting aside bells and whistles that need to be added to accommodate sorites worries.) An obstacle to this being made plausible is that the colors may be sufficiently different to be synchronically discriminable. Plausibly, when synchronically discriminated, the way color1 is perceived is different from the way color2 is perceived. How can it be at all plausible, then, that c1 is perceived in a way that is different from the way c2 is perceived in the synchronic context and that c1 is perceived in the same as way as c2 in the diachronic context? This would seem to entail that the way c1 is perceived when synchronically presented with c2 is different from the way c1 is perceived when c1 is presented by itself. How can it be at all plausible that c1 be perceived differently in these two different contexts? It is in answer to this sort of question that I see an appeal to general context effects in color perception playing a role. Our familiarity with such context effects is supposed to make the claim that c1 is perceived differently when alone and when with c2 plausible. I don’t think an appeal to context effects can explain how this can be so. I think the appeal just helps to make it plausible that it can be so. I guess it shouldn’t be surprising that context effects are relatively ineffectual in explaining much. Such effects are data in need of a theory to explain them. They themselves are ill-poised to do much explaining.

    Fineness?
    It might be dialectically useful here to represent my view in a very extreme form. Consider it as an extension of the language of thought hyposthesis (LOT) to conscious experience (LOE). For sake of argument, suppose also that the conceptual repertoire is exactly mirrored by a subject’s spoken language. So, there will be exactly as many noncompositional color concepts as individual color words the speaker has mastered, and there will be as many compositional color concepts as the speaker is able to form phrases for. For a monolingual English speaker, their LOE is English. As I view the dialectic over discriminability arguments, then, the burden is on nonconceptualists to present evidence of discriminability behaviors that cannot be accounted for by the LOE hypothesis. And if there’s evidence that someone can perceive 12 different greens each as being different from the other, I would like to know what evidence there is that it’s inadequate to think the LOE content isn’t simply “12 different greens, each different from the other”. I don’t see that in the present dialectic I need to grant, for example, that there are 12 different color concepts deployed.

  67. Hi Pete, thanks again for the follow up.

    I think we must be talking past each other a bit here. I am not claiming that it is part of teh content of the HOT that I in fact have the HOT. I completely agree with the way you characterize the ‘Rosenthalian as’ (as we might call it). The argument that I put forward had a faulty premise and I suggested a way to fix it. So here is the argument again (with key differences highlighted):

    0. TP=(platitude) A conscious mental state is a mental state that I am transitively conscious of myself as being in

    1. To be transitively conscious is to represent something in an appropriate way

    C1. Therefore, to implement TP is to be able to say which states have the property of being represented (this is supposed to be Pete’s buildup to the argumentum ad unicornium)

    2. Having the property of being represented=being transitively conscious of myself as being in that state=having a suitable HOT (best theory of how TP is implemented)

    C2. Therefore to implement TP is to be able to say which states figure in the contents of HOTs

    3′ I am transitively conscious of myself as being in states, which as it happens, are notional states

    C3. Therefore notional states have the property of being represented and so have the property of being cosncious (revision of common sense)

    Doesn’t this answer both of the objections?

  68. Hi Richard,

    Thanks for sticking with this. I think I’m understanding better what’s going on. I don’t know, though whether I understand this argument well enough to say precisely how it may or may not connect with problems previously raised. But a new problem arises as I examine this.

    2 and 3′ and C3 strike me as being in tension with each other. For simplicity’s sake we can represent the identity statement made in 2 as: 

    r = t = h

    And we can say that 3′ and C3 imply that r is notional.

    But, by Leibniz’s Law, 2 entails that t and h are notional too. But this is weird, since I don’t suppose any parties to the debate take h to be notional.
     

  69. HI Richard,

    I’m glad you’re diggin’ it!

    I interpret the proposal that r can notional without always being notional as the proposal that r is notional only at some worlds. And that, by itself, isn’t problematic. But I’m thinking that r isn’t going to be notional at every world in which h is instantiated, and therein lies the problem.

    So, there’s still a Leibniz’s law type of problem associated with 2. Whatever it is that you want to say about r – that it’s notional only in some worlds, or whatever – has to be said also of t and h. More to the point: given that r, t, and h are identical, they must be at all the same worlds. However, I take it that part of what’s involved in saying that r is only notional in some worlds is that there are worlds with h & notional r and worlds with h & non-notional r. Think back to the scenario comparing persons A, B, and C – persons with true, false, and empty HOTs that are otherwise identical. We might think of A, B, and C as at different worlds all of which contain h. But unless I’m losing my grip on what this notional stuff is all about, r is notional at A’s world but not at C’s. If r is notional at A’s world but not C’s, then by the law of Leibniz, h is notional at A’s but not C’s. But I don’t see how it makes sense to say h is notional at any of those worlds.

  70. Hey Pete. About whether the stuff I talked about above is in folk psychology. I’m not sure I understand your explicit/implicit distinction. By ‘explicit,’ do you mean that if you did some, say, x-phi and asked people whether mental qualities were truth evaluable, they’d be able to tell you? I doubt they’d know what to say at all, if you could even get them to understand the question. So I guess I think all of these folk-psychological facts are implicit. I say these things are enshrined in FP because they’re consistent with the way the folk talk about mentality. Take, for instance, the TP. If you asked a person on the street whether we should understand conscious states as those states a person is conscious of herself as being in, I’m sure the person would look at you blankly. But if you ask a person X about person Y’s belief B, and told X that we had good reason to think that Y believes B but Y is not aware of B, X would likely tell you that B is unconscious. I think the same kind of thing holds for these other folk-psychological ways distinguishing the qualitative from the intentional. The game is to try to give a theoretical explanation of as many of these platitudes as possible. If we cannot give a theory that captures some of them (i.e. if we cannot vindicate the TP in any theoretical way because the Unicorn argument is a good argument), then so be it. But I’m happy to lay that debate to rest for now.

    As for whether the view we sketched can be called conceptualism, I think Philippe does an excellent job of showing why our view can’t be called conceptualism, as the term is generally understood in the literature. Conceptualism, as he describes it, is that perceptual experience is exhausted by conceptual content. I wonder if there’s room for another view that has the attractive aspects of this kind of conceptualism, without requiring such a strong thesis.

    Of course, no one wants to get into verbal debates about what falls under what heading, but I’m trying to get straight what’s really at the heart of the debate. According to our two conditions for conceptualism, we were thinking that the more important of the two clauses was (1), whereas Pete and Philippe take something like a strengthened version of (2) to be what marks conceptualism.

    Given Pete’s distinctions between the matching vs. constituency versions of conceptualism, I think Philippe would say that genuine conceptualists are committed to the constituency thesis (and Pete’s on board with this) because the matching version could meet his thinking constraint but not require that perceptual experience is exhausted by conceptual content. Even if one could hold the matching version and count as a conceptualist as Pete says, both of these theses entail that one always deploys a concept of a perceived color when one experiences it (which is what condition (2) holds).

    So what about rejecting (2), as we do, and only holding something like (1), which says that for every color one can consciously experience one possesses a concept of that color? (2) is much stronger than (1), and I’m wondering what benefits one gets from holding (2) over merely (1). At this risk of rehashing many of the comments we made in our original commentary and some well-worn details about HOT theory, let me just draw a few distinctions.

    On the HOT view, perceptions have both a conceptual content and a mental quality. The perception of red has both the content “I am seeing red” and a red* mental quality. Sensations, by contrast, have only mental qualities, and no conceptual component. A sensation of red only has a red* mental quality. Both perceptions and sensations can occur nonconsciously, or consciously when they are the targets of HOTs (Unicorn worries notwithstanding). HOTs are purely conceptual states. The content of HOTs are that one is in such and such a mental state. If I have a conscious perception of red, I have a HOT with the content “I am in a state with red* and the content “I am seeing red.”” Notice that the concepts figuring in the HOT are concepts not of the perceptible colors (i.e. RED for the perceptible color red), but concepts of the mental qualities, such as RED*. Moreover, the HOT also involves concepts of concepts, such as the concept of the concept RED.

    So, if you have a perception, conscious or otherwise, you not only have the concept of the perceptible color you are perceiving, but also will deploy that concept (not, of course, in the HOT). But, on HOT theory, it’s also possible that you might have conscious sensations, wherein you have a first-order state exhibiting only red*, and a HOT with a content in which only the concept RED* figures (and not the concept of red). This is all consistent, I maintain, with holding (1) but not (2). Just because one needn’t deploy the concept each time one has a conscious sensation doesn’t mean that one doesn’t have such a concept. And, I think for various reasons, one couldn’t have mental quality concepts without first developing the perceptible quality concepts. But that’s another story.

    Where this leaves us with Dummy and Smarty I’m not sure. I take it that Dummy, lacking the concept for vermillion, doesn’t deploy it in either the first-order perception or the HOT, and so he is not consciously perceiving vermillion even though his first-order perception may have the vermillion* quality. I agree that he’s not able to consciously perceive vermillion without having the concept for it. Moreover, he cannot consciously sense it without the concept. Were he to develop the concept, he could consciously sense or perceive it. But that does not mean that every time he does so, he must deploy it (he might not do so if he consciously sensed it). I hope this is making sense.

    As an aside, I have a terminological question, which is perhaps better directed at Philippe: what, if anything, is the difference between conceptualism and varieties of intentionalism according to which all content is conceptual? I suppose, for instance, Tye isn’t a conceptualist on this line, but what about intentionalists that reject nonconceptual content? Are these just different names for the same position?

  71. Pete Mandik :

    HI Richard,

    I’m glad you’re diggin’ it!

    I interpret the proposal that r can notional without always being notional as the proposal that r is notional only at some worlds. And that, by itself, isn’t problematic. But I’m thinking that r isn’t going to be notional at every world in which h is instantiated, and therein lies the problem.

    So, there’s still a Leibniz’s law type of problem associated with 2. Whatever it is that you want to say about r – that it’s notional only in some worlds, or whatever – has to be said also of t and h. More to the point: given that r, t, and h are identical, they must be at all the same worlds. However, I take it that part of what’s involved in saying that r is only notional in some worlds is that there are worlds with h & notional r and worlds with h & non-notional r. Think back to the scenario comparing persons A, B, and C – persons with true, false, and empty HOTs that are otherwise identical. We might think of A, B, and C as at different worlds all of which contain h. But unless I’m losing my grip on what this notional stuff is all about, r is notional at A’s world but not at C’s. If r is notional at A’s world but not C’s, then by the law of Leibniz, h is notional at A’s but not C’s. But I don’t see how it makes sense to say h is notional at any of those worlds.

    I am not really following what the worry is supposed to be. So, if ‘r’ is ‘property of being represented’, ‘t’ is ‘being transitively conscious of’ and ‘h’ is ‘have a suitable higher-order thought then I don’t see what the problem is supposed to be. None of those things are notional…what is notional is the target of the HOT. so, r=t=h tells us that whatever the targets of HOTs are those things have the property of being represented. Sometimes the targets are notional, sometimes they aren’t. So sometimes existing states have the property of being represented and sometimes notional states do.

  72. Richard,

    I apologize if I’ve been sloppy with “notional” and related terminology. Below I try again to make my point, this time with more care:

    When something notional has a property, the property is thereby notionally instantiated. It’s a “shmoperty” in Weisberglish. r is supposed to be a property. Your formulation of 2 that I compressed to “r=t=h” is:

    Having the property of being represented=being transitively conscious of myself as being in that state=having a suitable HOT

    I take it that your formulation is itself a kind of shorthand. Longhandishly, it would be this:

    A state’s having the property of being represented = my being transitively conscious of myself as being in that state = my having a suitable HOT that I’m in that state

    The three items linked by the two “=”s are properties. The first is a property of a state. The second and third are properties of me. When the state in question doesn’t exist, it’s notional and any properties of it are notionally instantiated properties.

    But the state in question can fail to exist even though I exist and I have a suitable HOT about it and transitive consciousness of it. If I exist and I have a suitable HOT, then my having the sutible HOT is not a notionally instantiated property. I exist and so does my HOT.

    And here’s where the trouble arises: in the false/empty HOT cases, the property of being represented is a notionally instantiated property, but the property of my having a suitable HOT is not a notionally instantiated property. So, contra 2, the property of being represented can’t be identical to my having a suitable HOT.

    To sum up in a very condensed form: r and h have different modal properties and therefore cannot be identical.

  73. Reluctant as I am to contribute further to this discussion, I impelled to register my sense of oddness about this notion of a notionally instantiated property. Is my representing Santa Claus a notionally instantiated property of Santa? Or a regularly (nonnotionally) instantiated of me?

  74. Hi David,
    I’d say that your representing Santa is a regularly instantiated property of you. Josh, and then Richard on Josh’s behalf, seems to want to say something along the lines of this: that Santa’s being represented by you is a notionally instantiated property of Santa. I take it, also, that Richard is (on behalf of Josh, I don’t think Richard really goes for this himself) floating the idea that is something along the lines of the following: a notionally instantiated property of Santa can be identical to a regularly instantiated property of you. But anyway, that’s just Pete on behalf of Richard on behalf of Josh. Who knows what might have got lost in translation?

  75. Hey Pete, fwiw, what I was saying (on behalf of Josh; you are right that I don’t myself slum in these areas :)) was that the property of being represented is identical to the property of being the target of a higher-order state and then I was suggesting that in A an existing state is the target and in C a notional state is the target. so it would be the case that in A you have the property of state consciousness whereas in C you have the shmoperty.

  76. Richard, I don’t see how that suffices to defend 2 from the Leibniz’s Law based objection, so the argument you were trying to fix looks to have little hope of being shown to be sound.

  77. Hi Jake,

    Thanks for your remarks. They are quite clear and helpful. I think you get my meaning just fine wrt explicitl/implicit. To answer the question you pose to Philippe, I’d say that, yes, those would just be different names for the same position.

    Further thoughts re: Dummy and Smarty.

    I take it that the unconvinced-by-the-unicorn-argument HOT theorist is quite happy with first-order states being conscious despite certain kinds of, let’s say, mismatch between predicate deployed in the HOT and property of the targeted state. So, one might, in actuality, be in a first-order state of perceiving red but have a HOT that one is perceiving green. In such cases, the HOT theorist is comfortable saying that one is conscious of oneself as perceiving green. And perhaps also, that that perception of red is conscious (since, perhaps, that’s the state of which one is conscious, though in an inaccurate way (see Rosenthal’s 2005, p. 354, fn. 34)).

    We might see a similar tolerance for mismatch concerning transitive consciousness of external objects. So, a green external object that’s first-order represented as red is an object of which one is conscious, according to you guys, though in an inaccurate way.

    One might wonder about less extreme cases of mismatch. One sort that I’d like to focus on here concerns determinate/determinable “mismatch”. So, consider when dummy has a vermillion sensing but the HOT says nothing as determinate as vermillion, but is just a determinable like red. I take it that you are comfortable saying that the vermillion sensing is conscious even though Dummy is only conscious of it as a red sensing.

    Let’s look now at the possibility of determinate/determinable “mismatch” for first order representation AKA transitive consciousness of external objects. A vermillion object is represented by Dummy simply as a red object. Wouldn’t he still, by the lights of your theory, be conscious of the vermillion object, even though he wasn’t conscious of it *as* vermillion? I’m thinking that you’d want to say “yes” here.

    Anyway, this line of thought is supposed to raise trouble concerning this thing you said:

    I agree that he’s not able to consciously perceive vermillion without having the concept for it.

    I wonder if it would be better for you, by the lights of your own theory, to grant that he *can* consciously perceive vermillion while lacking the concept for vermillion; he just can’t consciously perceive vermillion *as such*.

    Does this sound ok to you?

  78. Hey Pete. Thanks for these thoughts.

    I have just a few thoughts about your example of the possibility of a first-order determinate/determinable mismatch. It’s not clear to me that we want to say that, in the case of Dummy first-order perceiving a vermillion object as a red object, Dummy perceives a vermillion object. I think it’d be more natural to say that he perceives a red object. Since vermillion is a breed of red, I think we’d also like to say that he veridically perceives a red object. If, however, Dummy were to deploy GREEN in the perception, he would perceive the vermillion object as a green object – and it’d be natural to say he misperceived a green object (he mistakenly perceived a vermilion object as a green one).

    If we add consciousness into the mix, I’m still not sure that we want to say that, if Dummy has a conscious perception of the vermillion object as a red object, he consciously perceives a vermillion object. Yes, the perception is conscious. But since I think it’s better to say that he first-order perceives a red object, I think he consciously perceives a red object (not a vermillion one, despite it being vermillion). So, it seems to me that Dummy can’t consciously perceive vermillion without having VERMILLION.

    This will likely be my last post as the conference is ending tomorrow. So let me say that I’ve really enjoyed this exchange. I’m sure there are many outstanding issues that haven’t been adequately hashed out, so I hope we can continue this discussion in the future (perhaps offline :). Many thanks to everyone who has contributed to the conference, and Richard Brown again for organizing it.

  79. Concluding Un-diachronic Postscript to Phenomenal Pigments

    Here I try to make some wrap-it-up-y remarks highlighting stuff that strikes me as important with an emphasis on stuff that hasn’t been said already.

    I take it that the distinct theories of consciousness of the first-order conceptualists and higher-order thought theorists can agree on this much: the appearing to consciousness of a color just is it’s being conceptualized by a suitable mental state. Details about what suitability of mental state amounts to can be here set aside, but such details concern what distinguishes unconscious states from conscious ones and what distinguishes perceptual states from non-perceptual ones.

    I’m concerned to counter arguments alleging empirical evidence that certain profiles of discrimination successes and failures challenge the above-described consciousness conceptualism. I’d like to say a bit here about how the Diachronic Indistinguishability Argument (DIA) may still seem threatening even when a reidentifiability criterion for concept possession has been abandoned.

    Considered from the point of view of analysis, reidentifiability seems not to impose a criterion on concept possession. Consider SwampSlice, my time-slice doppelganger who exists only long enough to conceptualize some item just once. SwampSlice possesses a concept of that item without being able to deploy it twice, and thus without being able to re-id the item. I concede that. But I’m not super-concerned with the conceptual analysis of concepts (or anything).

    Considered from the point of view of an empirical science of the relations between concepts, perception, and memory, the following seems an unchallenged generalization:

    (M): Relative to short time periods, if a stimulus is conceptualized then it is remembered.

    Of course, an analogous generalization doesn’t hold for very long stretches of time. I can conceptualize a passing stimulus as a man with a mustache, but it’s highly likely that this will be forgotten eventually in a few minutes, hours, or days.

    But when we look to shorter time periods, periods spanning just a few seconds, (M) seems pretty good. Consider diachronic discrimination tasks comparing performance for words or written characters in known and unknown languages. It is plausible to predict support for (M) in such tasks. And we can see this as consistent with the sorts of effects that psychologists chalk up to “depth of processing”.

    We are in a position now to see DIA as depending not on a re-id-based conceptual analysis of the concept of “concept,” but instead on the reasonable empirical generalization, (M). Colors c1 and c2 discriminable synchronically but not diachronically even across very short delays seem to give conceptualists a problem. Because of (M), we’d expect a conceptualized c1 to be remembered long enough to support an appropriate comparison to c2. And because of the synchronic distinguishability of c1 and c2, there seems to be no reason to think c1 isn’t making an appearance to consciousness.

    The most appealing conceptualist move to make at this point is to say that c1 and c2 appear to consciousness in variable and variably coarse-grained ways. This allows for saying things like:

    * Sometimes c1 appears to consciousness differently in the synchronic and diachronic situations, allowing that the appearance of c1 is the same as the appearance of c2 in the diachronic situations.

    * In “New Experiment” kinds of set ups, when c1 and c2 are synchronically co-present even for diachronic discrimination tasks, the coarse grained conceptualization is insufficiently fine-grained to enable knowing that the darker of the two colors this time is the same color as the darker of the two colors from a moment ago.

    The objections to my conceptualism that I feel like I have developed pretty good responses to are:
    (1) Introspectively, conscious perception seems more determinate than conceptualism allows
    (2) Phenomenal sorites involve patterns of discrimination failures that can’t be explained by the general sort of model appealed to in responding to the DIA

    The objections to my conceptualism that I feel require a lot more time at the drawing board to respond to adequately are:
    (3) Conceptualism seems to allow for excessive fineness of grain, that is, the implausible difference in appearance of c1’s being darker than c2 and c2’s being lighter than c1
    (4) It’s not clear what’s supposed to be wrong with demonstrative-based approaches that the present form of conceptualism improves upon

    As I make further progress on all this, I’ll be putting bits up on my Brain Hammer blog (http://petemandik.blogspot.com).

    * * * *

    I am enormously grateful to the highly detailed and thoughtful commentaries and subsequent discussion from Jake Berger, Philippe Chuard, Charlie Pelling, and David Pereplyotchik. I am grateful, too, for helpful and interesting thread comments (this thread had the most, baby!) from Richard Brown, James Dow, Aspasia Kanellou, Michal Klincewicz, David Rosenthal, Josh Weisberg, and an anonymous unicorn. And ultra-super-big thanks go to Richard Brown for all his work in organizing Consciousness Online. I think the right thing to do at this point is to just declare physicalism true in Richard’s honor.

    Cheers,

    Pete Mandik, William Paterson University

  80. Hi Pete,

    Thanks for the insightful presentation and subsequent comments! As a student of neuroscience, it occurred to me that an empirical complication might help diffuse the DIA: there has been some debate over the nature of “neutral” visual stimuli. In this case, the “blank” screen which replaces the coloured stimulus is usually a white stimuli, which is not a neutral stimulus at all and may serve as a kind of mask over the previous stimulus. Therefore, the memory of the colour will have been altered. Just my two cents; sorry for submitting this post-conclusion.

Leave a Reply

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

WordPress.com Logo

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

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s