Psychology, Neuroscience, and the Consciousness Dilemma

Presenter: Katalin Balog, Rutgers University Newark

Commentator 1: Bénédicte Veillet, University of Michigan-Flint

Commentator 2: Elizabeth Schechter, Washington University St. Louis

Advertisements

6 Comments

  1. Response to Elizabeth Schechter’s commentary:
    Thanks to Elizabeth to raising interesting questions.
    First, a quick correction. Elizabeth is quoting me as suggesting that cognitive neuro-science is not a good source for learning about the connection between phenomenality and accessibility. Here is what she says: “Balog’s claims to the effect that “The ball is in the court of those who argue for empirical… tractability” and that “there isn’t an immense amount that cognitive neuro-science is going to teach us about phenomenality and accessibility” (p. 7) appear to assume that we can confidently draw a priori conclusions about what is and isn’t knowable a priori, and I am not sure that we can.”
    This quote is a bit out of context. The relevant passage reads: “The upshot of this is not [italics just added] that there isn’t an immense amount that cognitive neuro-science is going to teach us about phenomenality and accessibility.” (p 10)

    On a more substantial note, I disagree with Elizabeth’s interpretation of the dilemma I am setting up for investigating the relationship between phenomenality and accessibility. She says on p 4: “Balog may simply have articulated more explicitly than anyone else the bind that (on some interpretations) Block finds himself in: either phenomenality requires accessibility or it is epiphenomenal and thus in and of itself empirically invisible. “
    This is not the dilemma I wish to explore. The dilemma is this: either our conception of phenomenality is conceptually linked to accessibility or it is not; in the latter case I claim there is reason to think that connection between phenomenality and accessibility is not empirically tractable. The second horn of the dilemma is not that phenomenality is epiphenomenal. I don’t believe phenomenality is epiphenomenal; nor does the claim that there is no conceptual connection between phenomenality and accessibility implies epiphenomenalism. That the concept of phenomenality is non-functional doesn’t imply that phenomenality itself is not a functional property. And even if phenomenality is not a functional property it doesn’t follow that it has no causal powers.
    Accordingly , her explanation of my argument is a bit off: “because if phenomenal properties are purely non-functional, non-causal properties, then no explanation appealing to representations with whatever phenomenal properties could enjoy an empirical advantage over an explanation appealing to those same representations minus their phenomenal properties” (p 4-5)
    I would like to correct this reconstruction of my argument. The paper leaves room for this misunderstanding by not making clear that the argument crucially turns on a distinction between concepts and properties. Explanation is, at least in one clear sense of the word, a conceptual affair. One might not be able to explain the bravery of Clark Kent by appealing to features of Superman’s upbringing (even if one can by applealing to Clark Kent’s upbringing); at least not under that description. So the point is that an appeal to the phenomenality of cognitively unaccessible neural state e (under that description) does not increase the explanatory power of an appeal to the occurrence of e under its mere physical/functional description. For all this, it might be that e is phenomenal. My point is that, given the second horn of the dilemma, we will never find out.
    Finally, Elizabeth says that rather than relying on my argument one could rely on a simpler one: “the reason that one can’t show that phenomenality dissociates from accessibility doesn’t seem rooted in the non-functional nature of our phenomenal concepts at all. It seems, rather, rooted in the fact that our intuitions about what counts as behavioral evidence for phenomenality picks out, again, just those capacities that will be accounted for in some-or-other cognitive theory of (access) consciousness. “ (p 5)
    I don’t see that this would be a good argument. There are countless cases where we can show that two phenomena are dissociated even if one counts as an evidence for the other. The simplest example is one that Elizabeth herself cites: the relationship between the external world and perc

  2. Hi Katalin,

    Thanks so much for this thoughtful reply. I’m terribly sorry about the misquote—I didn’t mean to quote you out of context but rather simply misread (not sure which is worse!), presumably because the misread made more sense to me given my interpretation of the rest.

    It’s really helpful to receive the clarification I’d asked for re: the relationship between PHENOMENALITY being a non-functional concept and phenomenality being a non-causal property. I didn’t mean to imply that you personally advocated epiphenomenalism about phenomenality (I think I laid out correctly, on the first page of the paper, the dilemma that you take yourself to be presenting to Block); rather it seemed to me at the time that if some of what you say in your first, my second horn of the dilemma is right, then epiphenomenalism seems to follow. But perhaps I was simply confused about how you were individuating properties. In your reply, you write,

    “an appeal to the phenomenality of cognitively unaccessible neural state e (under that description) does not increase the explanatory power of an appeal to the occurrence of e under its mere if physical/functional description. For all this, it might be that e is phenomenal.”

    I suspect that the issue here (which concerns not only how to individuate properties but also the relationship between ontology and scientific explanation) relates to the debate between the phenomenal concepts strategy and qualia eliminativism. I was thinking that if an appeal to phenomenality per se adds nothing to any explanation (even if we don’t construe explanation metaphysically), then why think that the property of phenomenality adds anything to the world (and, thus, why think that it’s a real or instantiated property at all)? But are you thinking that scientific theories may well appeal to the property of phenomenality (to the property of being phenomenal)—just not, as you say, under that description? If so, is part of your idea that even if or when a scientific theory does appeal to the property of phenomenality, we may not know that it is appealing to that property? (I suppose that whether or not you can mean that, depends on what you mean when you say that there is an “immense amount that cognitive neuro-science is going to teach us about phenomenality and accessibility” (10)…?)

    Re: the issue about evidence, I agree that “There are countless cases where we can show that two phenomena are dissociated even if one counts as an evidence for the other.” I certainly didn’t mean to deny that dissociable phenomena can provide evidence for each other. I only meant to suggest that in order to show that two phenomena dissociate, the evidence for one (or for its lack) can’t be identical to the evidence for the other (or for its lack). But of course my example re: the existence of a mind-independent external reality suggests that (depending on what you mean by “show!”) even that may not be right; arguably this is a place where considerations of simplicity (which you mention on p. 7) come into play….

    Really what I was struggling with here was to reconcile your claim that the dissociability of phenomenality and *accessibility* is empirically intractable, and your claim (p. 4) that the dissociability of phenomenality and *access* is empirically tractable. These claims seemed to me to fit uncomfortably together *if* the explanation for the intractability of the former debate is rooted in the non-functional nature of PHENOMENALITY, since that concept figures into the framing of both dissociability debates. So the best way I found to reconcile the two claims was in terms of a difference between what we intuitively view as evidence of *cognitive accessibility* and what we intuitively view as evidence of actual *cognitive access.* If you have time for another reply, I’d be eager to hear your thoughts on what makes one debate empirically resolvable, while the other is not.

    But, whether you have time or not for a second response, thanks for this first helpful reply!

    Lizzie

  3. Hi Lizzy, thanks for your reply; I’ll get back to you tonight. Meanwhile here is my

    Response to Bénédicte Veillet’s commentary:

    Thanks Bénédicte for raising interesting questions.

    First, you express puzzlement as for the structure of the paper. The paper does pose a dilemma for researchers interested in the connection between phenomenality and access. The form of argument is not, however, as you suggest, to pose an exclusive dilemma of some sort and then continue by arguing that one of the horns are clearly unacceptable.

    You seem especially puzzled by why I am discussing the view that there is a conceptual connection between phenomenality and accessibility. You think that people (like me) who are interested in the Phenomenal Concept Strategy (PCS) should reject this view since the PCS only works on the assumption that analytic functionalism is false.

    As you put it: “Those interested in theorizing about the relation between phenomenal and functional concepts typically have ulterior motives. Many of them believe they can use phenomenal concepts to diffuse antiphysicalist arguments (such as the Explanatory Gap argument), and they may assume that their strategy can only work if phenomenal concepts are entirely conceptually isolated from any functional concept, including the concept ACCESSIBILITY.” (p1-2)

    Two points about this. First, as you yourself point out the claim that there is a conceptual connection between phenomenality and accessibility doesn’t imply analytic functionalism. Second, and more importantly, this claim gets the dialectic wrong .It is not that analytic functionalism spells trouble for certain defenses of physicalism. If analytic functionalism was true, the physicalist would have no need to resort to the PCS; the anti-physicalist arguments would be already satisfactorily answered. It is simply that proponents of the PCS do not think that such a way out from the anti-physicalist arguments is plausible at all.

    On p. 1, you continue: “The upshot of this part of Balog’s argument is that the existence of a conceptual connection between PHENOMENAL and ACCESSIBILITY does not in fact commit us to analytic functionalism. And if it doesn’t, then it isn’t clear that what she presents us with is actually a dilemma at all.”

    I am not sure I understand why there is no dilemma here; unless you think someone posing a dilemma needs to reject one horn of the dilemma; and you also think I am not willing to reject either.

    You are trying to clarify your problem in the following way:

    “This might suggest that Balog’s goal is not really to present a dilemma for phenomenal
    concept theorists generally, but to present a dilemma for a subset of these theorists,
    namely those who (like Block) believe that there is no connection between
    PHENOMENAL and ACCESSIBLE and yet believe that the metaphysical relation
    between phenomenality and accessibility is empirically tractable. This would make the
    scope of the paper quite a bit narrower than it first seems.” (p. 2)

    I am not sure I know what you think the scope of the paper seemed to be at first. My goal in the paper was simply to explore the dilemma („is there a conceptual connection between PHENOMENAL and ACCESSIBLE?”), provide some considerations pro and con of each horn and, most centrally, to argue for the claim that those who take the second horn have strong reasons to doubt the knowability of the exact nature of the connection between phenomenality and accessibility.

    Finally your reconstruction of this argument:

    “The idea seems to be that if our concept PHENOMENAL is wholly distinct from the concept ACCESSIBILITY, then we will committed to the view that phenomenality has no causal profile.” (p. 3)

    The idea is not that phenomenality has no causal profile – that would be odd for a physicalist which I am most of the time to say – it is that PHENOMENALITY is not a functional concept. The reason this is important for the prospects of a science of phenomenality is that formulating a hypothesis about the presence of phenomenality under a phenomenal conceptualization in a brain state e does not add to the causal/explanatory power of say, neurophysiological descriptions of e. Not at least if what is to be explained, as it is in cognitive neuroscience, are psychological/neurophysicological phenomena.

    The last point Bénédicte makes is that my final argument presupposes that there are no conceptual connections between PHENOMENAL and any other functional concept, not only that PHENOMENAL and ACCESSIBLE are conceptually distinct. My response to this is that although it is an interesting question if there are any such connections but the issue doesn’t concern my argument. The assumption that e is phenomenal doesn’t – if physicalism is true it cannot – carry with it any new independent functional/physical characterizations that were not already present in the neuro-physiological/functional/physical characterizations of e.

  4. Hi Lizzie,
    Thanks for your replies. Here are some further thoughts.
    You say: “I suspect that the issue here (which concerns not only how to individuate properties but also the relationship between ontology and scientific explanation) relates to the debate between the phenomenal concepts strategy and qualia eliminativism. I was thinking that if an appeal to phenomenality per se adds nothing to any explanation (even if we don’t construe explanation metaphysically), then why think that the property of phenomenality adds anything to the world (and, thus, why think that it’s a real or instantiated property at all)?”
    You are right to understand me – as you further discuss the above passage – as making a careful distinction between properties and concepts. The fact that you don’t get an explanatory surplus by positing your neural state e to be phenomenal (*under this description*) doesn’t mean that phenomenality is not some straightforward physical or functional property of e. And you are right, in the end what I am saying is that even when a scientific theory appeals to the property of phenomenality (not under that description; but rather under a scientific description) we won’t really know that it does for the reasons I elaborate in my main argument. I argue that whether accessibility is constitutive of phenomenality is not empirically tractable and therefore the *exact* biological/neurophysiological nature of phenomenality is not tractable. Let’s suppose neural state n is found to be correlated with phenomenal state ph in uncontroversial cases (i.e., ph is reported to be present). Now suppose you’d like to determine whether phenomenality is present in certain inaccessible representations; i.e., whether phenomenality and accessibility can dissociate. Can you just answer the question by determining if n is present in those cases of inaccessible representation? No! You could only do that if you antecedently knew that the machinery of accessibility that exists in the uncontroversial cases is constitutive of phenomenality. But that is the very question we are trying to answer. And the rest of my argument in the paper is that there is no good model as to how you can bootstrap your way out of that problem by IBE. That of course is not to say that we cannot learn a lot about the science of phenomenality in many other respects; just not with respect of the question of the separability of phenomenality and accessibility.

    Another issue you raise is how to reconcile my “claim that the dissociability of phenomenality and *accessibility* is empirically intractable, and [my].. claim (p. 4) that the dissociability of phenomenality and *access* is empirically tractable.”
    Well, the Sperling experiment I think show pretty convincingly that access (being in the global workspace) is not constitutive of phenomenality. But we are convinced about this via a certain accessibility of the entire array of, say, characters presented: it is intuitively evident to the subjects that they have had a phenomenal experience of it. All the different characters are not *accessed* (at the same time) by these subjects, but at the moment of seeing their phenomenal presence is evident. This is the kind of accessibility I chiefly have in mind when I consider the possibility of accessibility being necessary for phenomenality.

    Cheers,
    Kati

  5. Kati,

    Thanks a lot for your response!

    Let me just say, first, that when I saw that you were presenting a dilemma for those interested in thinking about the connection between ACCESSIBILITY and PHENOMENALITY, I did expect there to be two horns both of which would be problematic in important ways. I was puzzled because I couldn’t see from your discussion why assuming that there was a connection between the concepts was problematic at all…So that got me puzzled about the goal of the paper.

    But I definitely agree with you that the question of the relationship between the concepts PHENOMENAL and ACCESSIBLE is very much worth exploring. I was convinced, at some point, that the concept PHENOMENAL was isolated from every other functional concept in some way. I very much like what you say at the end of the paper when you point out that it’s hard to imagine a state that is phenomenal but not accessible. Did you have in mind something like the connection between RED and COLORED? Like I mention in my comments, it seems not just hard but impossible to imagine something being red but not colored…Now it may be that after some careful thought, we do come to realize that it is indeed impossible to have something be phenomenal but not accessible. The difference, then, may just be that the impossibility may not always be immediately obvious. I’d like to know what you think about this, if you have the time!

    Benedicte

  6. Hi Benedicte, Lizzie,

    There is much about the connection between accessibility and phenomenality that needs further thought. One of the most interesting questions are, as Benedicte pointed it out, the question of the existence or lack of a conceptual connection between them. I do not think the mere fact that access provides our only “hard” evidence for phenomenality is in itself a reason to posit a conceptual relationship. And certainly the connection is nothing like the one between RED and COLORED. However, you can’t easily peel some form of access away from phenomenality either. More needs to be said of the “form” of access since it is not simple access in the sense of being in global workspace. We are aware of much more phenomenality then what is directly accessed, as the Sperling experiments show. I think there is a form of accessibility that is not purely dispositional, the one that gives you the certainty in the Sperling experiments that you have actually seen all the characters – even without actual access to all of those characters. I think that that form of accessibility might be a minimal conceptually necessary condition on phenomenality – but I haven’t figured out how exactly to argue for that. I just have the strong sense that there is something fishy about the notion of inaccessible phenomenality.

    Finally, thanks Lizzy for broaching the question of whether the quotational account sits easily with views about a conceptual link bw accessibility and phenomenality. I’ll need to think more about that.

    Thanks again to both of you for your insightful comments.

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