Originally posted by Keith Frankish on the Online Consciousness Conference site.
Presenter: Keith Frankish, The Open University
Commentator 1: Richard Brown, LaGuardia Community College, CUNY
Commentator 2: Amy Kind, Claremont McKenna College
Hi Keith and Richard,
I very much enjoyed Keith’s paper and Richard’s response. (For what it’s worth, I find Keith’s arguments here to be very convincing.)
I’d be curious to hear what the two of you think of the following:
One way of thinking about Keith’s case is an instance of a general phenomenon that we find in the history of science as viewed by a Quinean/Kuhnian tradition: There just never are theory-neutral grounds by which to adjudicate competing theories.
Here’s Quine from Word and Object (p. 16):
“One tends to imagine that when someone propounds a theory concerning some sort of objects, our understanding of what he is saying will have two phases: first we must understand what the objects are, and second we must understand what the theory says about them. ….[Sometimes] there is virtually no significant separation; our coming to understand what the objects are is for the most part just our mastery of what the theory says about them. We do not learn first what to talk about and then what to say about it.”
[…]
“Picture two physicists discussing whether neutrinos have mass. Are they discussing the same objects? They agree that the physical theory which they initially share, the pre-neutrino theory, needs emendation in the light of an experimental result now confronting them….To discern two phases here, the first an agreement as to what the objects are (viz. Neurtrinos) and the second a disagreement as to how they are (massless or massive), is absurd.”
Thinking about Quinean grounds for skepticism about theory-neutrality inspires thoughts of anti-Quinean strategies for resisting the skepticism. For instance, perhaps qualia dieters may want to lean on non-Quinean views of reference to try to make a plausible case that there’s some neutrally-characterizable thing that Classicists and the Zeros can succeed in referring to.
Anyway, to try to formulate some specific questions for you guys:
To what degree is Keith’s case specific to the qualia wars as opposed to just a species of a general failure of theory neutrality?
To what degree is Keith’s case hostage to specific semantic commitments about theoretical terms (commitments like holism, description theory of reference, etc.) and thus vulnerable to semantic-based attacks?
Thanks to Keith for his paper and to Amy and Richard for their comments: I found all of them really interesting.
Just a quick follow-up on Pete’s comment.
I liked your Quine quote and I’m inclined to agree with the point that he’s making in that particular passage. At the same time, though, I think that the objection which Amy raises in the first part of her comments still stands.
Quine’s point in the passage you quote is that in some cases, e.g. the neutrino case, there is no clear separation between an understanding-the-explanandum phase and an understanding-the-theory phase.
But although I’m inclined to agree with Quine about that, I don’t think it undermines what I see as Amy’s key point. For even if we agree that in some cases, there is no clear separation between the two phases, we can still say that in other cases, there is more of a separation. So for example in Amy’s moral rightness case, it does looks as though opposing parties can reach a high level of agreement on the identity of the explanandum before they start their theorizing about it. And that remains so even though there is nothing much to be said about moral rightness which remains neutral between the opposing theories.
(I guess it would also be open to Amy to allow that there’s never a *complete* separation between the two phases – perhaps there’s always some degree of overlap. She can still make the point that in the moral rightness and qualia cases, there seems to be a significantly lower degree of overlap than there is in the neutrino case.)
So although I agree with the Quine point, it still seems to me that Keith needs to find some disanalogy between the qualia case and the moral rightness case in order to answer Amy’s objection.
Hi all — interesting paper and discussion. I see I’m pegged in the paper as a proponent of the Classic Qualia conception. I don’t think that’s right. I’m a Diet Qualia guy all the way. At least, I think the way to pin down the concept of phenomenal consciousness is the Diet way — in terms of what it is like to have an experience, and so on — and that’s how I always do things in my work.
Furthermore, I think that this is the absolutely standard way to do things — so I think these putative Diet Qualia really deserve the name Classic Qualia, or at least Standard Qualia. By contrast, Keith’s putative Classic Qualia make their main appearance in a paper by Dennett that is largely opposed to qualia — which is akin to letting Pepsi define Classic Coke! I can’t recall ever seeing a proponent characterizing their conception of qualia in this way, and if that’s ever happened, it’s much, much rarer than the “Diet” characterization.
Now, it’s an interesting further question whether phenomenal consciousness has the four Dennett-style “Classic” features, and if so whether it does so necessarily, and whether it’s doing so is a sort of conceptual truth about phenomenal consciousness. Re ineffability: prima facie there are all sorts of things I can say about my phenomenal states, so if they’re to be “ineffable” it must be in some highly specific sense that needs to be spelled out. Re immediate accessibility: I’m inclined to think that some of my phenomenal states are immediately accessible, some (e.g. those outside attention) are not. Re intrinsicness — I’m inclined to think that actual phenomenal states are intrinsic to their possessors in a certain sense, but they’re also relational in another sense (as they are relations to represented properties); and furthermore, there may be possible worlds (Edenic worlds) where phenomenal states are not intrinsic at all. Re privacy: I suppose I think that I have a sort of access to my states that others don’t, so actual phenomenal states are private in a sense — but this is perhaps contingent (Amy’s hypothetical consciousness-sharers) and certainly doesn’t seem to be a conceptual truth.
Overall, none of the four claims strike me as being good candidates for conceptual truths. Maybe there are precisified versions of some or all of them so they all end up being truths and perhaps even necessary truths, but I’m doubtful that they’ll end up being conceptual truths unless they are massaged so heavily that they in effect import something like the what-it-is-like conception.
Now, Keith says that he just doesn’t have a clear grip on the what-it-is-like conception that doesn’t rely on the four features, but that strikes me as odd. Again. it’s absolutely standard to generate anti-reductive intuitions by pointing out to people, Mary could know such and such about physical or functional states, but she wouldn’t know what it is like to see red. One doesn’t need to presuppose anything about privacy etc to make that claim — even when I suspend judgment about those four claims, along the lines of the previous paragraph, the what-it-is-like claim comes through loud and clear.
I think that what Keith’s point comes down to is that there isn’t any definition or cashing-out of the notion of phenomenology in independent terms. But of course this is a point that any number of proponents of the “Diet” conception accept — phenomenal concepts are conceptually primitive and resist reductive analysis. And when one finds a primitive concept, one shouldn’t expect this sort of independent handle on it. The same goes for Amy’s example of moral rightness (I’m very sympathetic with most of what Amy says here), and other plausibly primitive concepts such as those of existence, negation, necessity, perhaps space, time, meaning, causation — it’s familiar in all of these cases that at best we get characterizations in a tight little circle. And it’s by now a familiar point that the absence of an independent reductive grip on a concept does not suffice to call that concept into question.
I think a problem with rejecting the nonconceptual content conception of phenomenal consciousness- is that it is tied with the sense- data strategy (by which the Classic conception is defined). For to argue for privacy, it is only safe to assume that qualia are not phenomenal concepts- they are not concepts at all.So basically U lose some of the arguments for the Classic Conception.
Thanks to Richard and Amy for their presentations, and to Pete, Charlie, Dave, and Aspasia for their comments, all of which raise interesting and important points. I don’t think any of them are fatal to my argument, but they do require me to make some clarifications and qualifications. This will take several posts. I’ll take the comments in order, starting with Richard’s.
Richard’s comment is particularly useful, since it represents the position that is my target in the paper –- that is, physicalist qualia realism. I am suspicious of this position, since I don’t think there is a coherent notion of qualia that is both robust and amenable to physicalist treatment. One can’t have one’s phenomenal cake and eat it. Anti-physicalists will agree, of course, but I’m a physicalist and I’m advocating irrealism not anti-physicalism. So the quarrel between Richard and me is family one, but these are often the most bitter.
Richard begins by noting that some anti-physicalist arguments, including one by me, depend on for their effectiveness on the existence of a neutral explanandum. I agree. My anti-zombie argument (2007) did tacitly assume a diet notion of qualia. (It may, however, be possible to restate the argument without that assumption, I’ll return to this later if I have time.) My response is simply that it’s so much the worse for those arguments — though of course they will still have tactical value against opponents who share their assumptions.
I take it that what’s bothering Richard here is that if we, as physicalists, can’t defeat anti-physicalists on their home ground and account for the very same thing they are concerned with, then our own position will be undermined. I used to share this worry, but I’ve come to think it’s misplaced. Physicalist theories of subjective feel always end up doing a diet qualia/zero qualia shuffle, and constructing them diverts us from what we should really be doing — namely developing satisfying zero-qualia accounts, which explain away our belief in subjective feels altogether. Richard says this is giving up the ghost; I reply that it is so only in a therapeutic Rylean sense.
In the second part of his comment, Richard explains his conception of diet qualia. Subjective feels, he says, are things we apprehend directly in introspection and which would intuitively be missing in the zombie case. Richard identifies this view with my Suggestion #1 and goes on to deny that diet qualia, so understood, inflate into classic qualia. I agree with him. For in fact his proposal corresponds to my Suggestion #2, and my claim is that it is deflationary, not inflationary. Richard fleshes out his proposal by appealing to Rosenthal’s HOT theory, according to which there is a system of introspectable mental properties (qualitative characters) which are homomorphic to perceptible properties. Suppose this is right. Then it may explain why we are disposed to regard our experiences as having a subjective dimension that is intrinsic, ineffable, and so on. But to that extent it’s just a theory of zero qualia. The introspectable properties themselves — qualitative characters –- are properties that dispose us to think we have subjective feels, not subjective feel themselves, and our awareness of them is a cognitive process. (Richard himself does the diet qualia/zero qualia shuffle, remarking that “If you’re conscious of yourself as seeing red, then it will be like seeing red for you”.)
So what happened to the diet qualia? If they have been explained away, then we have a zero-qualia theory. If they are something over and above qualitative characters, then what are they? Well, perhaps they are ontologically basic properties that somehow emerge when qualitative characters are targeted by HOTs. (I’ll return to the suggestion that subjective feels are ontologically basic later, in discussing Dave’s comment.) But is that an option Richard is prepared to endorse? (If I understand his remark that the dispute between physicalists and anti-physicalists may be merely terminological, then perhaps he may.) But if so, his position is indistinguishable from that of people who describe themselves as anti-physicalists.
Hi everyone, thanks for the comments. This is a really interesting topic and I am really having fun thinking through the issues!
I find myself generally in agreement with Dave’s comments here but here are some thoughts.
Re the terminological point: let’s look briefly at two theories.
Let’s start with Block’s biological theory. On his view the mind is causally reducible to the mind but not conceptually reducible. He takes this to roughly correspond to the the a priori/a posteriori distinction. On his view there is no a priori way to move from phenomenal concepts to physical concepts but we do scientifically discover that pain is identical to certain brain activation.
On Dave’s view if physicalism is true then we would need to have the a priori deduction. Since he sees no way in which we could make this move a priori he thinks that physicalism is false.
Are there any real differences between these two views? That there isn’t –at least about the mind– comes out when you realize that on Block’s view you can’t even deduce that water is H2O a priori whereas on Dave’s (and mine) you can. What is really at issue here is just how you use the terms ‘physical’ and ‘non-physical’. Given the way that Dave uses them ‘water’ isn’t a physical term on Block’s view (though it is on Dave’s account), and given the way Block uses the terms pain is (though On Dave’s account it isn’t); even though both agree that there is no way to a priori deduce physical concepts from phenomenal ones or vice versa.
Now as for me, I agree that phenomenal concepts seem to resist reduction in the a priori sense but I deny that they actually do so. Ultimately something like the above may be going on here as well…I am still mulling this over. kati (Balog) has also suggested something similar (except she leans towards the view that there may be no fact of the matter as between physicalism and duaism….this too may be a terminological matter!)
Re the Diet Qualia/Zero Qualia Shuffle:
I can’t help but think that there is some kind of confusion between the explanadum and the explanans. The explanandum is simply whatever it is that we pick out in introspection and which would be missing from the zombie world (were it conceivable). I think you may be right that the explanans will either be inflationary or deflationary but that doesn’t seem to me to be a problem with picking put the thing to be explained in a neutral way. (btw notice that the reason people like Dave don’t do the shuffle is because they don’t try to explain phenomenal consciousness).
Keith says, “(Richard himself does the diet qualia/zero qualia shuffle, remarking that “If you’re conscious of yourself as seeing red, then it will be like seeing red for you”.)” Am I really doing that here? The shuffle consists in introducing diet qualia as the explanandum and then giving an explanation of zero qualia instead (reminiscent, it now occurs to me, of the “target reasoning” in Block’s “confusion about a function” paper). But what I am offering here is supposed to be an account of what it is like for you not of your judging that there is something that it is like for you. So I don’t think I am doing the shuffle or deflating the notion of diet qualia; nor am I inflating it.
Reply to Amy Kind
I’d like to thank Amy for comments, which I found very helpful. I’ll respond by trying to sharpen up the claim I’m making, in order to head off Amy’s worries.
Amy notes that the claim that there is no neutral way of specifying the explanandum for a theory of consciousness threatens to become either trivial or implausible. If the claim is simply that different parties have different ways of describing subjective feel, reflecting their interests and values, then it’s trivial; the same is true of most contentious debates. Amy assumes, correctly, that I’m not making this banal claim. My claim isn’t that the different parties employ partisan descriptions of the explanandum, but that they (or some of them at least – see below) haven’t identified a coherent explanandum at all.
On the other hand, Amy argues, if my claim is that there’s no way of getting a handle on subjective feel independent of a theory of its nature, then it’s implausible. We can have an intuitive, pre-theoretical grasp of what we are talking about, perhaps relying on demonstrative identification, and we can consult our intuitions to adjudicate between different theories of the nature of the thing in question. These theories themselves may be inflationary or deflationary, but that doesn’t mean that we had no neutral ‘diet’ notion to start with. This is true in many areas, Amy notes, so why not for subjective feel? This is clearly an important challenge for me, and Dave and Richard have raised similar worries.
Now I agree that in many cases diet notions of the sort Amy describes are available. But there are other cases where it’s not clear that we can get a theory-neutral grip on the explanandum — most obviously with theoretical posits or notions that are heavily theory-laden. (Pete spells out this point in his comment.) And it is arguable that the notion of subjective feel falls into this category. I suspect that an argument against a diet notion of qualia could be constructed along these lines. However, I don’t make that case in my paper, and don’t propose to make it now. My worry is different and more specific.
First, a point of clarification. I grant that there is a general neutral explanandum for theories of phenomenal consciousness. We can pick it out as “what is going on when we have conscious experiences” or something similar. This, of course, is a much broader notion than that of diet qualia, and it is one that all sides, including qualia irrealists, can sign up to. The notion of diet qualia is a more specific one: roughly, it’s the notion of whatever it is that qualia realists are realists about. More specifically still, it is a version of that notion that is designed to be acceptable to physicalists. As Dave points out, anti-physicalists may also employ a diet notion of qualia (I’ll say something about this in a separate reply to him), but the notion I have in mind is subtly different. It is one that is introduced by physicalists to serve as a starting point for reductive theories that aim to engage directly with the hard problem. This notion needs to be rich enough to satisfy anti-physicalists but weak enough to be open to physicalist strategies. There is inflationary pressure from one side and deflationary pressure from the other. Note that these pressures arise, not in the process of explanation (as Richard points out, anti-physicalists don’t need to explain qualia at all), but in the process of establishing a target that’s *amenable* to explanation (for physicalists). My claim is that there is no notion that meets these constraints. Moreover, I suggest that the practice of physicalist qualia realists reveals this: while professing to explain subjective feel, they actually explain *beliefs* about subjective (the diet qualia/zero qualia shuffle).
Now the important thing to note here is that, so conceived, the notion of diet qualia is a theoretical one, which is introduced to do specific work – namely to permit physicalist engagement with the hard problem. So in one sense it’s not a neutral notion at all. And this, I suggest, is enough to set it apart from the unobjectionable neutral notions Amy discusses. The problem with diet qualia are specific to it, not common to diet notions generally.
Of course, this isn’t to say that I’m right about the problem with the notion, and Amy’s comment raises other issues I need to address. Why can’t physicalists use demonstrative identification to set up the notion of diet qualia they need? And why must there be inflationary pressure from the anti-physicalist side? These issues are raised by Dave too, and I’ll address them in separate reply to him (later today, I hope).
Finally Amy asks why classic qualia, defined by the four features I mention, is the only alternative for realists. Couldn’t there be other notions, stronger than the diet one but weaker than the classic one? I’ve no wish to deny this. I am sure other coherent inflationary notions of qualia might be constructed, varying in the strength and detail of their commitment to the four features, and perhaps including others as well. I don’t think anything in my case against diet qualia turns on this, since these notions will be equally unacceptable to physicalists. (Certainly, any that build in intrinsically will be.)
Thanks again to Amy for a stimulating comment.
Replies to Pete and Charlie
Thanks to Pete and Charlie for their comments and apologies for the delay in responding. Here are my responses.
Pete — To answer your first question, although I’d locate myself firmly in the Quinean/Kuhnian tradition, in this case I don’t mean to rely on general Quinean skepticism about theoretical neutrality. As I explain my reply to Amy, I’m trying to highlight specific problems with the notion of diet qualia as used by physicalist qualia realists. (By the way, note that the notion of diet qualia I’m discussing is not meant to be neutral between classicists and zeroes, but between anti-physicalist and physicalist realists.) Thus — to answer your second question — I don’t see my case as hostage to the semantic commitments such skepticism entails (though if it did, I’d probably tough it out).
Having said that, however, Quinean/Kuhnian considerations are relevant here, and I think they offer an another way of characterizing my worry about diet qualia. Classicist and zero-ist approaches to phenomenal consciousness involve such different theoretical assumptions that the overused term “paradigm shift” really is appropriate here. And one way of expressing my quarrel with physicalist qualia realists is to say that they are trying, in vain, to straddle the divide between the two paradigms. They are taking an explanandum identified within one paradigm and trying to neuter it so that it can be explained using the theoretical tools of another paradigm. Diet qualia is a notion that inhabits the no man’s land between warring paradigms.
Charlie — You are surely right that there are degrees of theoretical embeddedness, and that in milder cases it is possible to adopt useful neutral notions. However, as I explain in my replies to Amy and Pete, my case doesn’t rest on general Quinean skepticism about theory neutrality, but on more specific worries about the notion of diet qualia. Moreover, I do accept that there’s a very general neutral explanandum for a theory of consciousness (roughly “what is going on when we have experiences of the kind we call ‘conscious’”). I suggest that this is the appropriate parallel for the moral rightness case.
I’m not clear what’s wrong wrong with starting off with qualia as ‘the referents of phenomenal concepts’. If this is a concept with a priori content, then this will constrain what phenomenal properties can turn out to be (just as the traditional theisic concept of God puts constraints on what God can turn out to be: if something ain’t omnipotent, then it ain’t God). Also, you may think that there is a certain class of concepts (e.g. the semantically stable ones) such that, with reference to those concepts, conceivability implies possibility, so we can get on with arguing over whether it is conceivable that those concepts are satisified in the absence of physical properties (or vice versa), and whether phenomenal concepts are in that special class of concepts which imply a link between conceivability and possibility.
Assuming we have a shared concept here, the concept we get at my ‘what it’s like talk’, it seems to me that we can start with the referents of that concept, and just get on with it.
Reply to Dave
Thanks to Dave for his comments. I’ve already made some relevant points in my reply to Amy, which I won’t repeat here, but I’ll try to address Dave’s main concerns. This will enable me to flesh out the argument and, I hope, make it stronger.
First, I take Dave’s point that he doesn’t subscribe to the classic qualia conception, and, indeed, that few people do so explicitly. (It’s true that I implied he did, though the main point of my “Chalmers or Dennett” line was to underscore my conclusion that the only coherent options are anti-physicalism or irrealism.) I concede, too, that “classic” may not be the best term for Dennett’s conception. At the risk of extending an already tired metaphor, we might dub it “qualia max”.
Dave goes on to defend a what-it-is-like conception of qualia. He argues that we can get a grip on the notion independently of the four features of classic qualia, and he points out that the fact that the concept doesn’t yield to reductive analysis doesn’t call its coherence into question.
I take these points, but I don’t think they seriously affect my argument. For I think there is an important distinction to be made between the what-it-is-like conception that Dave, as an anti-physicalist, employs (let’s call it “regular qualia”), and the what-it-is-like conception that a physicalist needs (”diet qualia”). The former notion, I’ll argue, is no more acceptable to a physicalist than that of classic qualia, and the argument in the paper can be re-run substituting the notion of regular qualia for that of classic qualia. Briefly, I want to argue that the notion of regular qualia, while not explicitly inflationary, is tacitly so.
First, we have strong anti-physicalist intuitions about regular qualia — that physical duplicates might lack them or have inverted ones, that one might know all the physical facts without knowing what it’s like, and so on. If sound, these intuitions entail that regular qualia are not physical properties. (They probably also entail precisified versions of at least some of the four classic claims, but I won’t pursue this.) Now I assume Dave will say that these claims are not conceptual truths: our grip on the concept of regular qualia is independent of our intuitions about the nature of regular qualia. I’m not convinced. It seems to me that thought experiments about zombies, inverted spectra, what Mary wouldn’t know, and so on, played an important role establishing my grip on the notion of regular qualia in the first place. The quale of an experience is just what is missing in zombies, different in inverts, or hidden from Mary. I’m suggesting, then, that the concept of regular qualia is tacitly theoretical or theory-laden.
Or consider the hard problem. The hard problem is the problem of giving a scientific explanation of regular qualia. But, by definition, it is not the problem of explaining informational states and processes (discrimination, categorization, integration, access, etc.). Hence, by definition regular qualia are not information-processing states. Again, there are theoretical commitments built into the notion of regular qualia, and they are ones most physicalists will want to reject.
Now, I guess it will be objected here that we can identify regular qualia demonstratively in introspection (Amy suggests this), without involving a theory-laden qualia concept. Again, I’m not convinced. This is a big topic, but, briefly, I’d want to argue that introspection is itself heavily theory laden. An obvious point to make is that people differ even on the very basic issue of whether introspection acquaints us with non-representational properties of experience. Another worry is that it’s conceivable we could introspectively demonstrate properties of experience other than qualia, such as representational and neurological properties. Indeed, HOT and HOP theorists hold that we do just this. And given this, it isn’t clear that demonstrative identification alone could give us a grip on the notion of qualia.
I suggest, then, that the concept of regular qualia involves tacit commitment to inflationary claims which physicalists cannot endorse. The notion of diet qualia, by contrast, is one that is supposed to have been stripped of any such commitments and that is fully compatible with physicalism. Diet qualia are qualia a physicalist can be realist about. And now I can repeat my former complaint about the diet notion. When we try to strip out the tacit commitments from the notion of regular qualia we are left with an eviscerated notion that doesn’t pick out anything at all.
Of course, physicalists can and do agree that we have an inflationary concept of qualia, which embodies anti-physicalist intuitions and represents qualia as intrinsic, etc. But they won’t accept that this concept correctly characterizes the relevant mental phenomenon, and they will disavow and explain away the intuitions it embodies.
What about Dave’s suggestion that the notion of what-it-is-likeness is a primitive one that resists reductive analysis and is not theoretically laden? Dave is right, of course, that this wouldn’t mean that the concept was incoherent or vacuous. It would, however, mean that it didn’t pick out an explanandum acceptable to physicalists. Physicalists can accept that we have a primitive concept of qualia that doesn’t yield to reductive analysis, but they won’t accept that what needs explaining is similarly primitive — an ontologically basic property that resists reductive explanation.
A few final remarks. First, in the paper, I noted that some physicalists explicitly reject classic qualia and adopt diet qualia as their explanandum. More often, however, they simply endorse the what-it-is-like conception without remark. I’m suggesting, however, that in doing this they are tacitly conflating the anti-physicalist notion of regular qualia and the gerrymandered physicalist notion of diet qualia — a conflation made all the easier by the fact that the inflationary commitments in former notion are tacit. This gives the impression that they are tackling the hard problem head on. In fact, as I note in the paper, what they end up doing is simply explaining zero qualia. Thus, two shuffles are typically involved: from regular to diet and from diet to zero.
Second, let me stress that none of this is meant to impugn the notion of regular qualia. I don’t think regular qualia exist, but I don’t think the notion is vacuous or incoherent. My target is the dummy notion of diet qualia.
Third, what I’m doing here is not simply endorsing anti-physicalist arguments, with the rider that irrealism is preferable to anti-physicalism. I’m claiming that a commitment to anti-physicalism is built into any conception of qualia worth the name, and thus that physicalist realism is incoherent. This is a conclusion that anti-physicalists can endorse, though I expect they will not like the reasons I offer for it. If the concept of regular qualia is theory laden, then it becomes more susceptible to elimination, which is indeed what I think should happen.
Hi Keith — I find the talk about “physicalist” and “nonphysicalist” what-it-is-like conceptions confusing. I am pretty sure that I have just the same concepts of qualia and what-it-is-like as many type-B physicalists. We just draw different ontological conclusions using those concepts. Maybe conceptions are supposed to go beyond concepts to include theses, but then I think that a lot of the other moves you need to make won’t go through.
But perhaps what you want to say can be put as follows. Roughly, you’re suggesting a rewritten version of the paper in which Dennett’s four feaures are replaced by physicalism-threatening features such as: (i) zombies might have lacked them (ii) inverts might have switched them (iii) Mary couldn’t have known about them (iv) they’re not functionally analyzable and (v) they generate an epistemic gap. Then we can say Qualia Max = (by definition) what it is like properties that have these five features. While Diet Qualia = what it is like properties (without a commitment to these five features). Then your claim is something like: some wishy-washy physicalists want to appeal to Diet Qualia without Qualia Max, but you don’t think we have any grip on Diet Qualia except via Qualia Max. So really the choice is just Qualia Max or Zero Qualia (some watered-down functional notion).
Put this way I’m at least somewhat more sympathetic, unsurprisingly. Still, I think the issues are subtle. First, most or all of five features in question are all ambiguous between epistemic and ontological readings — e.g. what *conceivably* might be missing in zombies (epistemic), or what *metaphysically* might be missing in zombies (ontological); Mary gains new knowledge (epistemic) vs Mary gains knowledge of a new fact (ontological); the concept is not functionally analyzable (epistemic) vs the property is not a functional property (ontological). So we really have two different concepts here: Ontological Qualia Max and Epistemic Qualia Max.
Now, if we go with Ontological Qualia Max, then I think it’s obviously right that to accept that there are Ontological Qualia Max is ipso facto to reject physicalism (options such as Russellian monism aside). So it would be fair to say that Ontological Qualia Max have nonphysicalism built into them.
On the other hand, it’s far less obvious that to go with Epistemic Qualia Max is ipso facto to reject physicalism. Your common-or-garden type-B physicalist will accept that there are Epistemic Qualia Max while denying that there are Ontological Qualia Max and correlatively denying the move to nonphysicalism. So I don’t think Epistemic Qualia Max obviously have nonphysicalism built into them. Of course my own view is that there are good inferences from Epistemic Qualia Max to nonphysicalism — but these inferences require substantive philosophy and aren’t simple conceptual truths.
As for the relation between Diet Qualia (which I continue to think is the core notion) and the two Max notions: do we have any grip on what-it-is-like independently of either of the Max notions? I’m inclined to think that we do, and that it’s possible to understand the notion of what-it-is-like without going via zombie and Mary intuitions. On the other hand, the intuitions in question are usually pretty helpful ways of getting the concept across to people who don’t have it. And my own view is that the Epistemic Max features follow fairly directly from the Diet Qualia concept — which isn’t quite to say that they are presupposed, but the distance is fairly small. But importantly, all this applies to the Epistemic Max features and much less obvious to the Ontological Max features. So even if you were right that one can’t be committed to Diet Qualia without being committed to Epistemic Qualia Max, it doesn’t follow that one can’t be committed to Diet Qualia without being committed to Ontological Qualia Max. And it’s the latter claim that you need for the criticism of middle-ground physicalism.
So it strikes me that given what you’ve said, there remains a coherent middle-ground physicalism: one that is committed to Diet Qualia and to Epistemic Qualia Max but not Ontological Qualia Max. There *may* also be a coherent position that is committed to Diet Qualia and not to Epistemic Qualia Max, although that is less clear to me. Now, of course I think that there are further considerations that rule both of these positions out — e.g. considerations about links between conceivability and possibility and between epistemology and ontology. But these are further considerations. Incidentally, the fact that I rely on these further considerations bring out quite clearly why I’m not just presupposing the existence of Ontological Qualia Max in my arguments — if I did, I wouldn’t need the further considerations. But in any case, without moving to considerations like these, then it seems to me that the wishy-washy physicalist who accepts Diet Qualia and perhaps Epistemic Qualia Max but not Ontological Qualia Max still has some ground to stand on.
Hi Dave, thanks again for these very helpful remarks. I hope no one minds if I jump in before Keith has had a chance to respond.
…I guess I am a wishy-washy physicalist (but I prefer to think of it as a stern optimism 🙂 This intermediate view endorses diet qualia as the core notion and agrees that the move from there to epistemic qualia max is a short one given what we know now about the brain; but it is one that can be resisted. This is best illustrated with shombies and zoombies. In the shombie world we have diet qualia without epistemic qualia max. Call the qualia in the shombie world New Qualia (to really overextend the metaphor!). I can then employ the same ‘further considerations’ to get to the falsity of dualism. Thus if one has Epistemic New Qualia in mind then the epistemic stuff doesn’t have any impact on physicalism whereas if one has Epistemic Qualia Max in mind it does. So the problem isn’t that you assume Ontological Qualia Max but rather lies with assuming Epistemic Qualia Max.
Thanks for your reply, Dave. The line you sketch in your second paragraph is roughly the one I’m proposing. (Though I might want to argue that the max commitments you mention can be simplified, perhaps by seeing all of them as manifestations of a general commitment to precisified forms of intrinsicality and subjectivity.)
But, as you say, we need to take account of the distinction between ontological and epistemic readings of the max features. I think I do respect the distinction in the paper, but it’s useful have it made explicit, using your terminology.
I take it the notion of EQM is this: EQM are properties of experience that we conceptualize as having the features of OQM (metaphysically capable of varying independently of the physics, not knowable by Mary, not functional, etc)? Or more loosely, EQM are properties that seem to have the features of OQM.
I argue that a commitment to DQ entails a commitment to EQM. You’re not persuaded, but you accept that the distance is short, so let’s grant it for the sake of argument. For even if I’m right, you note, one can be committed to EQM without being committed to OQM, so the physicalist still has some ground to stand on.
I agree that one can endorse EQM without OQM. My worry is that if one does, one ceases to be a qualia realist. Here’s how I think it goes.
On the view we’re considering, the neutral notion is that of EQM — properties of experience that we represent as having the max features — and the dispute between you and the physicalists is whether these properties really do possess the features we represent them as having.
Now this position corresponds closely to my Suggestion 2 (diet qualia are properties that seem to be intrinsic, ineffable, and private, but may not really be so), and my reply is roughly the same. EQM are equivalent to zero qualia — properties that dispose us to judge that the experiences that possess them have OQM. Note that this is not incompatible with their actually being OQM. The properties that dispose us to judge that the experiences that possess them have OQM might *be* OQM. (As I say in the paper, token zero qualia could be classic qualia.) That’s the position we’ll end up with if the anti-physicalist arguments are sound, and I think it’s perfectly coherent.
But what if we reject these arguments? Physicalists will deny that EQM are OQM, and will identify EQM with some physical properties that we are disposed to misrepresent as having the max features. Fine. Let’s suppose they do this convincingly. The epistemic max features have all been explained away. But if there’s no coherent notion of qualia without those features (as we’re accepting for the sake of argument), then how is this different from elimination?
Bottom line: EQM are either OQM or illusory.
Richard
Two quick belated replies to your comment from Saturday.
Re: terminological point. This is a big issue, but at first sight it seems odd to describe the difference between Dave and Block as merely terminological. After all, Block thinks that qualia supervene metaphysically on the microphysics whereas Dave doesn’t. No?
Re: DQ/ZQ shuffle. You suggest that I’m confusing explanandum and explanans. I’ve addressed this in my reply to Amy, so I won’t repeat the points here. With regard to my (somewhat cheeky) suggestion that you yourself were doing the shuffle, here’s the thought. The shuffle consists in passing off an explanation of dispositions to judge that we have subjective feels as an explanation of subjective feels themselves. Now you say “If you’re conscious of yourself as seeing red, then it will be like seeing red for you”. I take it this means that if one introspects mental red (an ordinary physical property), then one will experience the what-it-is-likeness of seeing red. And this seems to me false — at least as a conceptual claim. The discrimination of mental red could take place without subjective feel just as the discrimination of any other physical property could. What does seem true is that if you’re conscious of yourself as seeing red, then you will be disposed to *think* that it is like something for you to see red. If you can make discriminations among your experiences as such, you will be able to frame thoughts about what this experience is like as opposed to what that experience is like, and so on. So I’m suggesting that you’re passing off an explanation of our dispositions to make judgements about feels as an explanation of feels themselves — the shuffle.
With regard to your most recent comment on shombies and new qualia: This sounds interesting, but I’m afraid I’m having trouble getting to grips with the proposal. If you have time before the curtain falls, would you spell it out a little more please?
Thanks, Keith. A few quick notes.
(i) I don’t see that EQM must be understood as seeming-OQM. Some people understanding conceivability as seeming-possibility, but many don’t, and even more so for the other features. Prima facie the EQM features can be understood in wholly epistemic terms without bringing in ontology.
(ii) Even if you were right that EQM = seeming-OQM and that qualia are defined by EQM, I don’t see how the eliminativist conclusion follows. If there are no ontological extras, then OQM don’t exist, but EQM still might. That is, although there are no OQM properties there may still be seeming-OQM properties. So we can be realists about EQM, and correspondingly realists about qualia.
(iii) Even if there was something here that got one to eliminativism about EQM, I’d be inclined to deny that qualia = EQM is a conceptual truth. Although I think that it’s a priori that qualia are EQM, I think that what-it-is-like gives us a grip on qualia that doesn’t presuppose EQM. So if this is right, there’s at least conceptual daylight left open for a view in philosophical space like Richard’s, on which there are qualia but no EQM (though my own view is that this daylight is very small, like Alaskan daylight in winter!).
Thanks for those points, Dave. I’ll comment on (ii). Of course you’re right that EQM can still exist even if OQM don’t, and it was sloppy on my part to suggest otherwise. (I blame the fact that I was posting late at night.) What I said was a sort of shorthand for this: EQM are zero qualia — properties that dispose us to judge that the experiences that possess them have OQM. But then being a realist about EQM and an irrealist about OQM is equivalent to being an eliminativist about qualia in the everyday sense. For it’s the view that all that needs explaining are our dispositions to judge that we have experiences with the features of OQM. So the bottom line would be that EQM are either OQM or do not deserve the title “qualia” in any robust sense.
I’ll need to think more about your point (i), but am encouraged the slimness of the space open for denying that qualia = EQM.
Hi Keith — I guess I don’t agree that EQM are Zero Qualia. First, EQM as defined still have what-it-is-like built into the definition, and again I think we have a grip on that that doesn’t go through the other five features (here we recap the Dennett-style discussion). Second, EQM have various features built in that go well beyond Zero Qualia (where the latter simply involving dispositions to judge, etc). For example, EQM build in the claim that Mary gains new knowledge on leaving the room (whether or not this is knowledge of an ontologically distinct fact), whereas I take it that Zero Qualia are compatible with a view on which Mary gains no new knowledge (she just thinks she does).
Roughly, it seems to me that EQM goes along with the commitments of type-B materialism, while Zero Qualia go along with the commitments of type-A materialism. I guess you’re in effect trying to show that type-B materialists should ultimately be type-A materialists (or dualists) — a worthy project! But it does seem to me that type-B materialists can resist the collapse of EQM into Zero Qualia.
All posts [ -> ]
Over the years Consciousness Live! has piled up more than a hundred conversations, close to 200 hours of people arguing… Read more.
François Kammerer joins me once again to help me finally understand Illusionism as we discuss his new book, The House… Read more.
Change my mind I have long thought that there is something deeply anti-scientific about the way that certain people argue… Read more.
Leave a Reply