Hmm… Is the line between physicalism and illusionism really as strong as you suggest? I’m very much not well-read, so take this with a grain of salt. But in the video you linked, around 15:50 David Papenau says:
> if you look at what Dave Chalmers says when he originally introduces the supposed hard problem, he says the problem is to understand why the brain gives rise to conscious feelings. Why is it that certain brain processes generate this extra stuff? If you put the problem like that, I think physicalists should just deny that there's any such problem. After all, if you think about the way that problem is posed, it presupposes the falsity of physicalism. It says the brain is giving rise to this extra stuff.
To me that sounds very similar to your characterization of illusionism as saying “the whole idea of ‘phenomenal consciousness’ is confused bogus to begin with”.
So I’m not too familiar with the details of Papineau’s views, but I certainly don’t think he’s an illusionist. He’s a type B physicalist, meaning he thinks there’s an epistemic gap between the physical and consciousness (i.e. he thinks zombies are conceivable). I think what he’s trying to say here is that the hard problem can’t be that consciousness is something extra over the physical—it’s conceptual, not metaphysical.
He is the type-B physicalist par excellence. To see what he'd say just imagine the Greatest Analogy Ever (mind is just like heat!). When you identify temperature with mean molecular energy, you don't deny that there is mean molecular energy. You also don't deny that there is temperature. You deny that there are two different things. So it would be silly to say mean molecular energy 'gives rise to' temperature. It *is* temperature. They are one and the same.
I think I may have noticed some sarcasm with the phrase “Greatest Analogy Ever”, but I’m still going to point out that the analogy fails for the non-illusionist. Suppose we want to understand how blueness could be neurons firing, and someone suggests “it’s just like temperature being mean molecular energy”. This begs the question: what is temperature?
If we define temperature as mean molecular energy, then the analogy fails. Blueness isn’t defined to be neurons firing, it is defined to be the experience you have looking at the sky on a clear day.
Is temperature mercury rising in my thermometer? It can be proven, a priori, that mean molecular energy will increase the volume of mercury in a thermometer. Where is the a priori proof that neuron activity will cause the experience of blueness?
Is temperature the sensation of heat? But sensation is what we are trying to explain, so the analogy is too close. It would be like someone saying “You don’t understand how blueness is neurons firing? Well, it’s just like how redness is neurons firing!”. Okay, thanks I guess.
Then only other options is that temperature is some nebulous undefined thing that we eventually came to define as mean molecular energy due to empirical investigation. But in this case temperature was just some unknown combination of the three options above, which still doesn’t help.
While I don't think the Great Analogy works, this seems too quick. If you think that mental concepts are opaque in the way that natural kind concepts (or similar "kripkean" concepts), then this becomes perfectly explicable.
Temperature is mean molecular energy a posteriori, because we rigidly designate the thing that causes heat sensation, but it's opaque what this is (it could also have turned out to be electromagnetic radiation). Nevertheless, it's impossible for there to be heat with electromagnetic radiation.
Similarly, if you think that "pain" is opaque, referring to the thing that realizes the feeling, rather than "painy stuff", then it turns out that it can be a posteriori necessary that pain is C-fiber firing or whatever.
So it'll have to turn on whether pain is opaque in this way or not. Now it seems pretty clear to me that we can just stipulate "pain" to be "painy stuff", and proceed with the argument, thereby showing that "painy stuff" is non-physical--whether or not there is some other sense of "pain" that is physical. In any case, I think the argument you give here jumps the gun.
Yes, you are correct. There is a fourth option I neglected where temperature is "whatever causes the sensation of heat". In this case then it seems like analogously blueness is caused by light of a certain wavelength which still doesn't help me, but perhaps a direct realist with regards to perception (or others) would think this makes it work.
Well, we have to be careful here. "Blue" would be a certain wavelength, since that's plausibly talking about a property of objects. But "the sensation of blueness" would be a brain state (or functional state or whatever--again, assuming you think that this sort of mental concept is opaque.
I really think we should stop using the term "Type B physicalist". Chalmers poisoned it in his original description. Need to write a post on this.
On denial: we can deny that the difference between two things exists, which is a roundabout way of saying that they are equivalent. if Papineau thinks that there are two different concepts of the same thing, he is also saying (implicitly) that the delta is an illusion. To some extent, all physicalists are illusionists with respect to whatever distinguishes their metaphysics from alternatives. That doesn't mean they are denying consciousness, in all its undefined confusion.
The hardists equate the delta between two differently viewed entities as an entity in its own right, and then conflate that spurious difference with one view of the original thing, so they tie a semantic knot in the centre of the discussion. We need to highlight that knot and refuse to get trapped in the confusion.
That seems a bizarre way to describe it. I mean, I believe in the kinetic molecular theory of heat; I think that the temperature just IS the mean molecular kinetic energy. If I put my hand into hot water, it encounters a lot of water molecules that are rotating and perhaps vibrating at a high speed. And I can do a mathematical analysis (in principle) of how fast the molecules are moving, or I can feel the sensation of hotness through the firing of certain sensory neurons. But I wouldn’t say that the idea that there’s a distinctive sensation of heat means that I have the “illusion” that heat is different from random molecular motion. The sensation of heat is just the way that I perceive random molecular motion.
There’s a passage by Ayn Rand that’s relevant (from Introduction to Objectivist Epistemology, chapter 8): “All knowledge is processed knowledge . . . An ‘unprocessed’ knowledge would be a knowledge acquired without means of cognition.” We have to have some specific means of being aware of random molecular motion. The sensation of warmth isn’t an “illusion”; it’s just the way in which we sense random molecular motion. In fact, it’s the primary access we have to random molecular motion; the mathematics of molecular motion is a sophisticated theoretical explanation that only became possible sometime after the emergence of Newtonian physics.
You wrote: "With the phenomenal and the physical, however, we’re still waiting for even the smallest example of a reduction. I have never seen a good explanation of what the first step might even look like. There’s just not anything there—and I strongly suspect nothing could be."
Do you have reasons for this suspicion? I used to have the same suspicion but I'm not sure it was well justified, so I'd be interested to hear whether you have some reasoning behind it
There’s of course the fact that I don’t think we have any clear idea of what the first step might even look like. But more fundamentally, there is also the point that phenomenal properties are supposed to be “intrinsic” and “qualitative”. But physical properties (at least on certain views) are structural or relational. If that’s right then that might be a pretty good reason to think that we’ll never get there.
Maybe I'm misunderstanding what you mean by 'reduction' but haven't we found that c-fibres always fire when pain occurs, so that's perhaps the beginning of a reduction? We at least have a plausible target of something to which pain could be reduced. I suppose that isn't very promising to you given your anti-illusionist view of qualia.
Sorry if I'm butchering the dialectic here - just trying to understand!
So I guess that it's helpful to compare to other sorts of reduction. With the case of heat, we have a theory that predicts that such and such microphysical states will give rise to such and such macrophysical states (more or less as a logical consequence, given the theory), and I take it that we can do experiments that confirm this theory. So we have good evidence that the theory is true, and given the theory, we can show how the (logical) reduction would go for any system.
With the pain-C-fiber case, we can get experimental confirmation for some theory about which neural correlates of consciousness there are, but I don't see that the theory itself gives a meaningful reduction. You can't tell a story about there being such and such microphysical states (or even more higher order functional states) where, having told that story, it's just a logical consequence that there will also be such and such phenomenology.
In other words, the reduction happens at the theorizing stage, not the experimental stage, as far as I can tell. The experimental stage is just for telling us which of the proposed reductions is most likely to be true. Any candidate theory will give us a story where once you've fixed the level you're reducing to, there can be no question as to whether the phenomenon you're reducing is there or not. But for phenomenal consciousness, it seems to me that you can tell whatever story you want about brain states or whatever, but there will always be a further question as to whether or not such and such phenomenal state occurs. What I'd be looking for is a description of a physical system that leaves no further question about whether or not there is experience, which doesn't have to be filled in by hand (or by induction from our own experience, and others' behavior).
That got a little rambly haha, I hope it's reasonably clear what I'm getting at.
[Sidenote: You might think that scientific reductions also face this problem. Once I've described all the microphysics about my kettle boiling, that leaves no question as to the macrophysical properties of the system. But it *does* leave a question as to whether it's the hot transparent moving liquid in front of me--in other words, whether the macrophysical system in question will appear to me as the thing in front of me. But in that case, the problem is that you're implicitly talking about experiences again, and the physical part of the story is entirely reduced in the sense I described.]
I think it's really good, but it's also like, *really hard* sci-fi, so there's nothing even remotely approaching things like lightsaber-fights between space-wizards. The reactions to this book seem to mostly fall into either "I can't read this" or "I love it".
>> What we’re trying to capture is whether or not something is epistemically possible—that is, whether there is any a priori inconsistency with the thing conceived
What then is the difference between conceivable and possible?
Something might be epistemically possible, without being metaphysically possible. For instance if you think God is necessary if he exists, then it can't both be possible for him to exist and possible for him to not exist. But both of those are presumably epistemically possible.
Or take the water and H2O example. Epistemically, water could have turned out to be a basic element like the Greeks thought, or turned out to be XYZ or what have you. But if you buy Kripkes she Putnam's stuff, then you'll think those are not metaphysically possible.
What does "if you think God is necessary if he exists" mean? It seems to me that you should either think he's necessary or not. I'm not seeing the dividing line between Him and the sixth digit of pi.
The idea is that God is often conceived as a necessary being. That means that if he exists, he necessarily exists, and if he doesn't exist, he necessarily doesn't exist (since if he could possibly exist, he would be possibly necessary, and so actual (given certain assumptions about modal logic)).
Basically, God either exists in all possible worlds or in no possible worlds. If atheism is correct, he's not in the actual world, so he'd have to be in no possible worlds, and vice versa for theism.
However, it's not at all obvious that it's contradictory for God to exist, nor that it's contradictory for him not to exist (unless you buy ontological arguments or something like that). If that's right, then both his existence is conceivable, and his nonexistence is conceivable (ideally conceivable, that is). But one of those is impossible! So we can ideally conceive of something impossible (given the assumptions I made).
The difference with pi, is that with pi, there is only one logically consistent answer, if you actually work it out. So there's only one ideally conceivable option.
But water and H2O is probably the more straightforward example.
* Water being made up of, say, 2 nitrogens and 1 oxygen is conceivable but not possible
* Pi having the first digit of 7 is neither conceivable nor possible
But I don't understand the distinction. It seems to me:
* There is a truth about water's composition which is unknown to us until we discover it.
* There is a truth about pi's digits which is unknown to us until we derive it.
Water doesn't have to necessarily exist in the universe (maybe its combination of elements just never happens to materialize), but I don't see why the recipe of water wouldn't be just as necessary as mathematical equations, or how theists view God.
--
Looking back at, "If zombies are conceivable, then they’re possible", there's the part of me that finds this intuitive, actually. It reminds me of my own argument that an infinite multiverse can never be disproven ( https://ramblingafter.substack.com/p/infinity-kills-the-fine-tuning-argument ) because "there is something more beyond what we can see" is always conceivable, and therefore always possible.
Is there a difference between those two cases?
With the multiverse example, there's no way you could ever disprove extra-existence (or maybe "realityfluid" as Yudkowsky might say?). With p-zombies, could we disprove them if we understood consciousness well enough? And say it always necessarily arises from cognitive software of sufficient complexity?
Or would it remain a problem of the form, "Well sure in THIS universe consciousness must arise, but p-zombies are still conceivable and therefore plausible".
--
So I'm left with conflicting intuitions here. I'm leaning towards "p-zombies are possible" but also that "p-zombies are conceivable, therefore possible" is not a well-founded enough argument that we should accept.
Firstly, I think that a posteriori necessity is a giant red herring, so no comment.
Secondly, I think that your replacement version of premise 2 relies on some unstated assumptions. Why, given the generally weak link between conceivability and possibility, are we making some exception for this particular poorly informed imaginative exercise? We would need a ton of argument to show that this is not like the mathematical examples, which simply demonstrate the fallibility of the human mind, which is now a straw-manning response, because that fallibility is the same reason this whole argument is embarrassing.
Chalmers' treatment of the mathematical counter-examples is an excruciating exercise in special pleading. The onus of proof is on those who think we can just invoke ideal cognitive skills as though we had a magic lamp to rub. We don't, so conceivability leads nowhere.
So I'm not perfectly clear on which part you're disagreeing with in particular. Is it the mitigated ideal conceivability-possibility principle, or is it the idea that we have any reason to think that zombies are ideally conceivable? It sounds like the latter.
That's what the third section is supposed to address: In other cases where there are a priori entailments between two domains, we can get prototypes or examples for how this goes. But this is not the case for the phenomenal and the physical, and so, is right, this means that zombies are ideally conceivable.
I don't think Hume is your friend in this. Although I'm not at all sure I've understood Kripke's argument.
So when you perform a reduction of say heat to molecular motion, either you know what the necessary connection is between motion (the necessary connection that means collisions between molecules must be as they are, dynamics space and time function as they do and so on and could only possibly behave that way) and you show that that leads to heat phenomenon, or you are just reducing the constant conjunction of heat phenomenon (I mean per Kripke there are heat sensations and then heat empirical evidences I guess I'm not clear what heat sensation is doing in his account, but I'm betting on a Humean picture its all just more constant conjunction ultimately) to the constant conjunctions of molecular motions, you've just listed more conjunctions and these are all actual conjunctions (you don't have to worry about possible conjunctions only actual) and that is enough to prove identity.
If the former I think the anti-reductionist can say that like Hume they can't conceive of what that necessary connection you claim exists between cause and effect is and without that you haven't even taken the 1st step towards a reduction. So even rather standard physical reduction just seems inconceivable exactly like physical-mind in that the 1st step is impossible.
If you take the latter as genuine examples of reduction then it seems like finding that constant conjunction of c-fibre firing and pain just is an example of reduction in that vein. If you take sort of positing/expecting constant conjunction of things as a reduction of things to other things then I can easily see how you get the first step in a mind to physical reduction. Since actually identity relations are apparently just cases where you find more constant conjunctions. Just show we have reason to believe there is a constant conjunction of c-fibre firing to pain for example and that would be a reduction of pain to c-fibre firing. Now you could say its possible that they would come apart in some non-actual case, maybe it is but that's true of heat and molecular collisions if causation is nothing more than constant conjunction also (if different interactions were shown to be impossible that would just be the necessary connection Hume says he can't find). So if you invoke that we are back to the first horn of the dilemma (all anti-reductionists can just invoke their inability to specify a necessary connection as failure at the 1st step in reduction).
The failure of identity one way is easy to imagine and I can't imagine Kripke-Putnam argument you are gesturing at depends on it. It could easily be the case in the actual that there are two liquids with identical physical (boiling point, viscosity etc.) and chemical properties and so that are equally water, but one is made of H2O the other of XYZ. So likewise there might always turn out to be heat that is not molecular motion.
So it's the other way. Could there be molecular motion that does not become heat (given identical microphysical circumstances)? Could there be sets of water molecules going through the same properties qua water molecules (the same constant conjunctions of molecular collisions) that do not form into water droplets and so on. I'd say yes. The water molecules could go through the same motions, but interact with non-water differently, fail to dissolve salt say and so not act like water droplets as we know them. The water is the same even though the non-water is different, is easily conceivable if we can conceive of the physically identical zombie's mental properties being different from the normal humans. The salt not dissolving is a different set of constant conjunctions, but the water is going through the same set of constant conjunctions with itself and so qua water is clearly the same water in both scenarios (the watery properties of the water are identical even if their non-water properties differ).
They might also do this while going through all the same types of constant conjunctions, but different instances. After all to go through the same type is just to obey the same mathematically dictates series of outcomes. But any finite series can be generated by an infinity of possible rules. On a Hume constant conjunction definition of causality, two different series (rules) that generate the same set of constant conjunctions for the subset of possible instances that are actual are the same cause (the same constant conjunction). So their divergence in counterfactual cases would be an instance of the same causality leading to different outcomes.
Consider two trajectories for a ball that have the identical positions at times, but one is generated by a random process, the other by obeying a rule, on Hume they are just the same cases (trajectories) it was an error to think we knew more of the cause and effect (what generated the trajectory) than their constant conjunction.
I certainly wouldn't want to say that reduction consists just in constant conjunction. I'd be closer to the necessary connection side--but NOT in Hume's sense of "necessary connection." If we're using Humean terminology, then it'd be more like relations between ideas.
When I reduce boiling to molecular motion, give a story where, given that that story holds, it's entailed that the macro-behavior of boiling occurs. A very simple example would be reducing squareness to things arranged square-wise. Once I've told you how things are arranged, there's just no further question as to whether they make a square or not. The case of boiling is just a much more complicated version of that.
I think it's important to note an asymmetry in reduction: For any macro-phenomenon, there are many possible candidates for what it reduces to. Sticking with the square example, having a square leaves open whether it's instantiated by, say, gold atoms, or by pyrite atoms. But the indeterminacy doesn't go the other way! Having some microphysical arrangement of gold atoms automatically settles the question of whether or not there's a square there. The role of empirical experiment is to decide which reduction is actually the case.
So yes you're right, watery stuff could both turn out to be H2O and XYZ. But once you have H2O, it couldn't turn out not to be watery stuff.
You give a counterexample to this idea with the case where you have the water fixed, but it's still open whether it'll act like water, since the non-water stuff might be different. That's fine, but it's not a counterexample: In that case water *simpliciter* is reducible to H2O, but the whole world isn't reducible to H2O, since other stuff might be different.
But note: If you then *also* fix what salt is. If you tell a story not just about what water reduces to, but what everything that interacts with water reduces to, then there is no longer any question here. Given such a story, it is totally fixed that water will behave the way it does. So it's not a counterexample to the picture of reduction I'm giving here.
The point about indeterminacy of rules you raise at the end also doesn't stick to this version. Given any reductive story there is *no possible* indeterminacy, since the microphysics entails the macrophysics. Where there is indeterminacy is in the empirical work to tell which reduction story is correct. That's fine.
[Now, to be clear, I'm not endorsing Hume more generally. The quote was just to make a joke because Hume often makes arguments of the form "I don't really have a proof, but I can't see any good counterexample being forthcoming".]
I think the square is reduced to atoms arranged square-wise if it works allows a principle that allows one to reduce physical to mental and if it doesn't work it doesn't work. (All this is in principle in practice some reductions fail for various more gross reasons like failing to show functional equivalence in practical cases).
So we have a series of H2O molecules arranged square-wise and you claim there we've reduced the moat around the castle to a collection of H2O molecules.
The moat anti-reductionist goes, no because I see that the square around my castle is continuous, whereas the collection of H2O molecules you describe are discontinuous and so completely different. Also the H2O molecules are in constant motion whereas the moat is completely still and free of motion. I could go on forever listing differences. You haven't even done the first step to reducing the moat water to H2O molecules.
Now you give a story about how light and eyes work and show that discontinuous H2O molecules will appear as a continuous plenum to the human eye (and will also lack any visible motion and so on).
The anti-reductionist counters, all that shows is that there is empirical evidence of a constant conjunction between H2O molecules arranged a certain way and the tendency of certain images to be formed on the retina. Likewise a constant conjunction of those images formed on the retina and the ones formed in my mind by visual sensation. You've not thereby shown that the continuous and discontinuous are the same by my lights. Your story is just that we can and are justified in ignoring the differences (the approximation is good enough that we can just say there is no difference), I agree if we take that account to hold then that would dissolve the difference but only by a kind of circular reasoning that would also apply in the case say of c-fibre firing and pain (if you ignore the differences they are the same).
Between the image on the retina and the image in the mind are certain constant conjunctions that can be shown to hold and they can be detailed more and more such that there is no property of the image in the mind that does not have a covarying property in the image of the retina perhaps (especially if we extend beyond just the image of the retina to the physical firing of certain neurons and so forth). Yet if I understand your position this would not prove that the image on the retina and the image in the mind are the SAME image (identity/reduction). Such a mere perfect correlation of properties that can be described as different would not prove they are the same properties (or even similar to be similar they'd have to be similar in themselves not in the sense of having similar/isomorphic covarying relations).
So likewise we can't prove that the image on the retina and the "appearance" of the moat are the same image all we can prove is they have a set of covarying properties. In order to give the first indication that the arrangement of H2O molecules and the continuous body of water in the moat are the same you'd have to show that the appearance of each (their own image) is the same image as the image formed in the retina, then by the principles that things that are equal to the same thing are equal to each other you'd have proven that a moat and an arrangement of H2O molecules have the same appearance/image. However if you could do that you could prove that the physical image on the retina is the same image as the mental image in the mind. In which case you could prove that pain and c-bundle firing are the same thing (assuming they have a complete set of covarying properties a la a psycho-physical parallelism say).
As long as you allow reduction by saying that a difference can be ignored so long as one shows a close enough similarity in relation between things (and not in the thing itself), that allows one to reduce pain to c-bundle firing. If you don't allow that identity-reduction move you can't actually achieve any reduction because the anti-reductionist can always believe that the two things are just different in themselves even if they have a set of identical covarying relations (are functionally equivalent).
I think there are many interesting and maybe even somewhat compelling arguments for all kinds of anti-reductionism, even if I am mostly unconvinced by them. Some practical and some in principle (some about empirical questions, some about what reduction could ever be), my vague and broad comments here are no doubt very unfair to that rich intellectual vein. If someone is an anti-reductionist about the chemical bond (you don't think Pauling et al. actually reduced it to Quantum Mechanics), there are good reasons for that and those principles will almost certainly allow one to find good reasons to be an anti-reductionist about mind to brain also. I don't see how it is remotely plausible that there could be a situation where anti-reduction arguments fail in every other case, but work so easily in the mind-body case.
> If a statement is such that its truth is fully fixed by how things are presented in a conceivable situation (so there is no further hidden fact that could overturn it), then ideal conceivability of that statement entails its possibility.
Does Kripke somehow prove that Zombies are like that? Or is it just an appeal to conceivability once again - that he can't conceive how it can be any other way? This citation seems to be pointing to the latter:
> Now can something be said analogously to explain away the feeling that the identity of pain and the stimulation of C-fibers, if it is a scientific discovery, could have turned out otherwise? I do not see that such an analogy is possible.
But surely Kripke's argument isn't *that* weak?
I'm also confused by what exactly is being conceived here. According to Kripke, it's enough that P-Zombies are a priori conceivable for them to be logically coherent- i.e. one doesn't need to know all the physical facts of the universe to correctly infer from conceivability of the zombies to their logical coherency. But aren't all the physical facts of the universe already *priced in* the concept of the p-zombie?
P-zombie is an entity from the universe with *the same physics* but without consciousness. A statement about P-zombies is a statement about compatibility between all the physical facts of our universe and the negation of all the facts about consciousness. If this is true, then physicalism which claims that facts about consciousness are physical facts, is false. But evaluating logical coherency of this statement *requires* knowing all the physical facts of our universe, otherwise what are you even conceiving? Just the negation of facts about consciousness? That will not disprove physicalism.
Hmm… Is the line between physicalism and illusionism really as strong as you suggest? I’m very much not well-read, so take this with a grain of salt. But in the video you linked, around 15:50 David Papenau says:
> if you look at what Dave Chalmers says when he originally introduces the supposed hard problem, he says the problem is to understand why the brain gives rise to conscious feelings. Why is it that certain brain processes generate this extra stuff? If you put the problem like that, I think physicalists should just deny that there's any such problem. After all, if you think about the way that problem is posed, it presupposes the falsity of physicalism. It says the brain is giving rise to this extra stuff.
To me that sounds very similar to your characterization of illusionism as saying “the whole idea of ‘phenomenal consciousness’ is confused bogus to begin with”.
So I’m not too familiar with the details of Papineau’s views, but I certainly don’t think he’s an illusionist. He’s a type B physicalist, meaning he thinks there’s an epistemic gap between the physical and consciousness (i.e. he thinks zombies are conceivable). I think what he’s trying to say here is that the hard problem can’t be that consciousness is something extra over the physical—it’s conceptual, not metaphysical.
He is the type-B physicalist par excellence. To see what he'd say just imagine the Greatest Analogy Ever (mind is just like heat!). When you identify temperature with mean molecular energy, you don't deny that there is mean molecular energy. You also don't deny that there is temperature. You deny that there are two different things. So it would be silly to say mean molecular energy 'gives rise to' temperature. It *is* temperature. They are one and the same.
I think I may have noticed some sarcasm with the phrase “Greatest Analogy Ever”, but I’m still going to point out that the analogy fails for the non-illusionist. Suppose we want to understand how blueness could be neurons firing, and someone suggests “it’s just like temperature being mean molecular energy”. This begs the question: what is temperature?
If we define temperature as mean molecular energy, then the analogy fails. Blueness isn’t defined to be neurons firing, it is defined to be the experience you have looking at the sky on a clear day.
Is temperature mercury rising in my thermometer? It can be proven, a priori, that mean molecular energy will increase the volume of mercury in a thermometer. Where is the a priori proof that neuron activity will cause the experience of blueness?
Is temperature the sensation of heat? But sensation is what we are trying to explain, so the analogy is too close. It would be like someone saying “You don’t understand how blueness is neurons firing? Well, it’s just like how redness is neurons firing!”. Okay, thanks I guess.
Then only other options is that temperature is some nebulous undefined thing that we eventually came to define as mean molecular energy due to empirical investigation. But in this case temperature was just some unknown combination of the three options above, which still doesn’t help.
While I don't think the Great Analogy works, this seems too quick. If you think that mental concepts are opaque in the way that natural kind concepts (or similar "kripkean" concepts), then this becomes perfectly explicable.
Temperature is mean molecular energy a posteriori, because we rigidly designate the thing that causes heat sensation, but it's opaque what this is (it could also have turned out to be electromagnetic radiation). Nevertheless, it's impossible for there to be heat with electromagnetic radiation.
Similarly, if you think that "pain" is opaque, referring to the thing that realizes the feeling, rather than "painy stuff", then it turns out that it can be a posteriori necessary that pain is C-fiber firing or whatever.
So it'll have to turn on whether pain is opaque in this way or not. Now it seems pretty clear to me that we can just stipulate "pain" to be "painy stuff", and proceed with the argument, thereby showing that "painy stuff" is non-physical--whether or not there is some other sense of "pain" that is physical. In any case, I think the argument you give here jumps the gun.
Yes, you are correct. There is a fourth option I neglected where temperature is "whatever causes the sensation of heat". In this case then it seems like analogously blueness is caused by light of a certain wavelength which still doesn't help me, but perhaps a direct realist with regards to perception (or others) would think this makes it work.
Well, we have to be careful here. "Blue" would be a certain wavelength, since that's plausibly talking about a property of objects. But "the sensation of blueness" would be a brain state (or functional state or whatever--again, assuming you think that this sort of mental concept is opaque.
I really think we should stop using the term "Type B physicalist". Chalmers poisoned it in his original description. Need to write a post on this.
On denial: we can deny that the difference between two things exists, which is a roundabout way of saying that they are equivalent. if Papineau thinks that there are two different concepts of the same thing, he is also saying (implicitly) that the delta is an illusion. To some extent, all physicalists are illusionists with respect to whatever distinguishes their metaphysics from alternatives. That doesn't mean they are denying consciousness, in all its undefined confusion.
The hardists equate the delta between two differently viewed entities as an entity in its own right, and then conflate that spurious difference with one view of the original thing, so they tie a semantic knot in the centre of the discussion. We need to highlight that knot and refuse to get trapped in the confusion.
That seems a bizarre way to describe it. I mean, I believe in the kinetic molecular theory of heat; I think that the temperature just IS the mean molecular kinetic energy. If I put my hand into hot water, it encounters a lot of water molecules that are rotating and perhaps vibrating at a high speed. And I can do a mathematical analysis (in principle) of how fast the molecules are moving, or I can feel the sensation of hotness through the firing of certain sensory neurons. But I wouldn’t say that the idea that there’s a distinctive sensation of heat means that I have the “illusion” that heat is different from random molecular motion. The sensation of heat is just the way that I perceive random molecular motion.
There’s a passage by Ayn Rand that’s relevant (from Introduction to Objectivist Epistemology, chapter 8): “All knowledge is processed knowledge . . . An ‘unprocessed’ knowledge would be a knowledge acquired without means of cognition.” We have to have some specific means of being aware of random molecular motion. The sensation of warmth isn’t an “illusion”; it’s just the way in which we sense random molecular motion. In fact, it’s the primary access we have to random molecular motion; the mathematics of molecular motion is a sophisticated theoretical explanation that only became possible sometime after the emergence of Newtonian physics.
I don’t see a point in all of this.
I didn’t call heat an illusion. I think sensation occurs.
You wrote: "With the phenomenal and the physical, however, we’re still waiting for even the smallest example of a reduction. I have never seen a good explanation of what the first step might even look like. There’s just not anything there—and I strongly suspect nothing could be."
Do you have reasons for this suspicion? I used to have the same suspicion but I'm not sure it was well justified, so I'd be interested to hear whether you have some reasoning behind it
There’s of course the fact that I don’t think we have any clear idea of what the first step might even look like. But more fundamentally, there is also the point that phenomenal properties are supposed to be “intrinsic” and “qualitative”. But physical properties (at least on certain views) are structural or relational. If that’s right then that might be a pretty good reason to think that we’ll never get there.
Maybe I'm misunderstanding what you mean by 'reduction' but haven't we found that c-fibres always fire when pain occurs, so that's perhaps the beginning of a reduction? We at least have a plausible target of something to which pain could be reduced. I suppose that isn't very promising to you given your anti-illusionist view of qualia.
Sorry if I'm butchering the dialectic here - just trying to understand!
So I guess that it's helpful to compare to other sorts of reduction. With the case of heat, we have a theory that predicts that such and such microphysical states will give rise to such and such macrophysical states (more or less as a logical consequence, given the theory), and I take it that we can do experiments that confirm this theory. So we have good evidence that the theory is true, and given the theory, we can show how the (logical) reduction would go for any system.
With the pain-C-fiber case, we can get experimental confirmation for some theory about which neural correlates of consciousness there are, but I don't see that the theory itself gives a meaningful reduction. You can't tell a story about there being such and such microphysical states (or even more higher order functional states) where, having told that story, it's just a logical consequence that there will also be such and such phenomenology.
In other words, the reduction happens at the theorizing stage, not the experimental stage, as far as I can tell. The experimental stage is just for telling us which of the proposed reductions is most likely to be true. Any candidate theory will give us a story where once you've fixed the level you're reducing to, there can be no question as to whether the phenomenon you're reducing is there or not. But for phenomenal consciousness, it seems to me that you can tell whatever story you want about brain states or whatever, but there will always be a further question as to whether or not such and such phenomenal state occurs. What I'd be looking for is a description of a physical system that leaves no further question about whether or not there is experience, which doesn't have to be filled in by hand (or by induction from our own experience, and others' behavior).
That got a little rambly haha, I hope it's reasonably clear what I'm getting at.
[Sidenote: You might think that scientific reductions also face this problem. Once I've described all the microphysics about my kettle boiling, that leaves no question as to the macrophysical properties of the system. But it *does* leave a question as to whether it's the hot transparent moving liquid in front of me--in other words, whether the macrophysical system in question will appear to me as the thing in front of me. But in that case, the problem is that you're implicitly talking about experiences again, and the physical part of the story is entirely reduced in the sense I described.]
Thanks for the help, I'll think this through🤔
If you like hard sci-fi, or having a severe sense of existential dread, I would highly recommend the book "blind sight" to you.
By Peter Watts?
Yes, he put it for free online, if you're okay with ebooks: https://www.rifters.com/real/Blindsight.htm
I think it's really good, but it's also like, *really hard* sci-fi, so there's nothing even remotely approaching things like lightsaber-fights between space-wizards. The reactions to this book seem to mostly fall into either "I can't read this" or "I love it".
Sounds very interesting!
I’ll have to check that out then!
>> What we’re trying to capture is whether or not something is epistemically possible—that is, whether there is any a priori inconsistency with the thing conceived
What then is the difference between conceivable and possible?
Something might be epistemically possible, without being metaphysically possible. For instance if you think God is necessary if he exists, then it can't both be possible for him to exist and possible for him to not exist. But both of those are presumably epistemically possible.
Or take the water and H2O example. Epistemically, water could have turned out to be a basic element like the Greeks thought, or turned out to be XYZ or what have you. But if you buy Kripkes she Putnam's stuff, then you'll think those are not metaphysically possible.
Sorry I'm still not following.
What does "if you think God is necessary if he exists" mean? It seems to me that you should either think he's necessary or not. I'm not seeing the dividing line between Him and the sixth digit of pi.
Whoops, completely forgot to respond here!
The idea is that God is often conceived as a necessary being. That means that if he exists, he necessarily exists, and if he doesn't exist, he necessarily doesn't exist (since if he could possibly exist, he would be possibly necessary, and so actual (given certain assumptions about modal logic)).
Basically, God either exists in all possible worlds or in no possible worlds. If atheism is correct, he's not in the actual world, so he'd have to be in no possible worlds, and vice versa for theism.
However, it's not at all obvious that it's contradictory for God to exist, nor that it's contradictory for him not to exist (unless you buy ontological arguments or something like that). If that's right, then both his existence is conceivable, and his nonexistence is conceivable (ideally conceivable, that is). But one of those is impossible! So we can ideally conceive of something impossible (given the assumptions I made).
The difference with pi, is that with pi, there is only one logically consistent answer, if you actually work it out. So there's only one ideally conceivable option.
But water and H2O is probably the more straightforward example.
So I think you're saying:
* Water being made up of, say, 2 nitrogens and 1 oxygen is conceivable but not possible
* Pi having the first digit of 7 is neither conceivable nor possible
But I don't understand the distinction. It seems to me:
* There is a truth about water's composition which is unknown to us until we discover it.
* There is a truth about pi's digits which is unknown to us until we derive it.
Water doesn't have to necessarily exist in the universe (maybe its combination of elements just never happens to materialize), but I don't see why the recipe of water wouldn't be just as necessary as mathematical equations, or how theists view God.
--
Looking back at, "If zombies are conceivable, then they’re possible", there's the part of me that finds this intuitive, actually. It reminds me of my own argument that an infinite multiverse can never be disproven ( https://ramblingafter.substack.com/p/infinity-kills-the-fine-tuning-argument ) because "there is something more beyond what we can see" is always conceivable, and therefore always possible.
Is there a difference between those two cases?
With the multiverse example, there's no way you could ever disprove extra-existence (or maybe "realityfluid" as Yudkowsky might say?). With p-zombies, could we disprove them if we understood consciousness well enough? And say it always necessarily arises from cognitive software of sufficient complexity?
Or would it remain a problem of the form, "Well sure in THIS universe consciousness must arise, but p-zombies are still conceivable and therefore plausible".
--
So I'm left with conflicting intuitions here. I'm leaning towards "p-zombies are possible" but also that "p-zombies are conceivable, therefore possible" is not a well-founded enough argument that we should accept.
Firstly, I think that a posteriori necessity is a giant red herring, so no comment.
Secondly, I think that your replacement version of premise 2 relies on some unstated assumptions. Why, given the generally weak link between conceivability and possibility, are we making some exception for this particular poorly informed imaginative exercise? We would need a ton of argument to show that this is not like the mathematical examples, which simply demonstrate the fallibility of the human mind, which is now a straw-manning response, because that fallibility is the same reason this whole argument is embarrassing.
Chalmers' treatment of the mathematical counter-examples is an excruciating exercise in special pleading. The onus of proof is on those who think we can just invoke ideal cognitive skills as though we had a magic lamp to rub. We don't, so conceivability leads nowhere.
So I'm not perfectly clear on which part you're disagreeing with in particular. Is it the mitigated ideal conceivability-possibility principle, or is it the idea that we have any reason to think that zombies are ideally conceivable? It sounds like the latter.
That's what the third section is supposed to address: In other cases where there are a priori entailments between two domains, we can get prototypes or examples for how this goes. But this is not the case for the phenomenal and the physical, and so, is right, this means that zombies are ideally conceivable.
This is a duplicate comment. Some Substack glitch.
I don't think Hume is your friend in this. Although I'm not at all sure I've understood Kripke's argument.
So when you perform a reduction of say heat to molecular motion, either you know what the necessary connection is between motion (the necessary connection that means collisions between molecules must be as they are, dynamics space and time function as they do and so on and could only possibly behave that way) and you show that that leads to heat phenomenon, or you are just reducing the constant conjunction of heat phenomenon (I mean per Kripke there are heat sensations and then heat empirical evidences I guess I'm not clear what heat sensation is doing in his account, but I'm betting on a Humean picture its all just more constant conjunction ultimately) to the constant conjunctions of molecular motions, you've just listed more conjunctions and these are all actual conjunctions (you don't have to worry about possible conjunctions only actual) and that is enough to prove identity.
If the former I think the anti-reductionist can say that like Hume they can't conceive of what that necessary connection you claim exists between cause and effect is and without that you haven't even taken the 1st step towards a reduction. So even rather standard physical reduction just seems inconceivable exactly like physical-mind in that the 1st step is impossible.
If you take the latter as genuine examples of reduction then it seems like finding that constant conjunction of c-fibre firing and pain just is an example of reduction in that vein. If you take sort of positing/expecting constant conjunction of things as a reduction of things to other things then I can easily see how you get the first step in a mind to physical reduction. Since actually identity relations are apparently just cases where you find more constant conjunctions. Just show we have reason to believe there is a constant conjunction of c-fibre firing to pain for example and that would be a reduction of pain to c-fibre firing. Now you could say its possible that they would come apart in some non-actual case, maybe it is but that's true of heat and molecular collisions if causation is nothing more than constant conjunction also (if different interactions were shown to be impossible that would just be the necessary connection Hume says he can't find). So if you invoke that we are back to the first horn of the dilemma (all anti-reductionists can just invoke their inability to specify a necessary connection as failure at the 1st step in reduction).
The failure of identity one way is easy to imagine and I can't imagine Kripke-Putnam argument you are gesturing at depends on it. It could easily be the case in the actual that there are two liquids with identical physical (boiling point, viscosity etc.) and chemical properties and so that are equally water, but one is made of H2O the other of XYZ. So likewise there might always turn out to be heat that is not molecular motion.
So it's the other way. Could there be molecular motion that does not become heat (given identical microphysical circumstances)? Could there be sets of water molecules going through the same properties qua water molecules (the same constant conjunctions of molecular collisions) that do not form into water droplets and so on. I'd say yes. The water molecules could go through the same motions, but interact with non-water differently, fail to dissolve salt say and so not act like water droplets as we know them. The water is the same even though the non-water is different, is easily conceivable if we can conceive of the physically identical zombie's mental properties being different from the normal humans. The salt not dissolving is a different set of constant conjunctions, but the water is going through the same set of constant conjunctions with itself and so qua water is clearly the same water in both scenarios (the watery properties of the water are identical even if their non-water properties differ).
They might also do this while going through all the same types of constant conjunctions, but different instances. After all to go through the same type is just to obey the same mathematically dictates series of outcomes. But any finite series can be generated by an infinity of possible rules. On a Hume constant conjunction definition of causality, two different series (rules) that generate the same set of constant conjunctions for the subset of possible instances that are actual are the same cause (the same constant conjunction). So their divergence in counterfactual cases would be an instance of the same causality leading to different outcomes.
Consider two trajectories for a ball that have the identical positions at times, but one is generated by a random process, the other by obeying a rule, on Hume they are just the same cases (trajectories) it was an error to think we knew more of the cause and effect (what generated the trajectory) than their constant conjunction.
I certainly wouldn't want to say that reduction consists just in constant conjunction. I'd be closer to the necessary connection side--but NOT in Hume's sense of "necessary connection." If we're using Humean terminology, then it'd be more like relations between ideas.
When I reduce boiling to molecular motion, give a story where, given that that story holds, it's entailed that the macro-behavior of boiling occurs. A very simple example would be reducing squareness to things arranged square-wise. Once I've told you how things are arranged, there's just no further question as to whether they make a square or not. The case of boiling is just a much more complicated version of that.
I think it's important to note an asymmetry in reduction: For any macro-phenomenon, there are many possible candidates for what it reduces to. Sticking with the square example, having a square leaves open whether it's instantiated by, say, gold atoms, or by pyrite atoms. But the indeterminacy doesn't go the other way! Having some microphysical arrangement of gold atoms automatically settles the question of whether or not there's a square there. The role of empirical experiment is to decide which reduction is actually the case.
So yes you're right, watery stuff could both turn out to be H2O and XYZ. But once you have H2O, it couldn't turn out not to be watery stuff.
You give a counterexample to this idea with the case where you have the water fixed, but it's still open whether it'll act like water, since the non-water stuff might be different. That's fine, but it's not a counterexample: In that case water *simpliciter* is reducible to H2O, but the whole world isn't reducible to H2O, since other stuff might be different.
But note: If you then *also* fix what salt is. If you tell a story not just about what water reduces to, but what everything that interacts with water reduces to, then there is no longer any question here. Given such a story, it is totally fixed that water will behave the way it does. So it's not a counterexample to the picture of reduction I'm giving here.
The point about indeterminacy of rules you raise at the end also doesn't stick to this version. Given any reductive story there is *no possible* indeterminacy, since the microphysics entails the macrophysics. Where there is indeterminacy is in the empirical work to tell which reduction story is correct. That's fine.
[Now, to be clear, I'm not endorsing Hume more generally. The quote was just to make a joke because Hume often makes arguments of the form "I don't really have a proof, but I can't see any good counterexample being forthcoming".]
I think the square is reduced to atoms arranged square-wise if it works allows a principle that allows one to reduce physical to mental and if it doesn't work it doesn't work. (All this is in principle in practice some reductions fail for various more gross reasons like failing to show functional equivalence in practical cases).
So we have a series of H2O molecules arranged square-wise and you claim there we've reduced the moat around the castle to a collection of H2O molecules.
The moat anti-reductionist goes, no because I see that the square around my castle is continuous, whereas the collection of H2O molecules you describe are discontinuous and so completely different. Also the H2O molecules are in constant motion whereas the moat is completely still and free of motion. I could go on forever listing differences. You haven't even done the first step to reducing the moat water to H2O molecules.
Now you give a story about how light and eyes work and show that discontinuous H2O molecules will appear as a continuous plenum to the human eye (and will also lack any visible motion and so on).
The anti-reductionist counters, all that shows is that there is empirical evidence of a constant conjunction between H2O molecules arranged a certain way and the tendency of certain images to be formed on the retina. Likewise a constant conjunction of those images formed on the retina and the ones formed in my mind by visual sensation. You've not thereby shown that the continuous and discontinuous are the same by my lights. Your story is just that we can and are justified in ignoring the differences (the approximation is good enough that we can just say there is no difference), I agree if we take that account to hold then that would dissolve the difference but only by a kind of circular reasoning that would also apply in the case say of c-fibre firing and pain (if you ignore the differences they are the same).
Between the image on the retina and the image in the mind are certain constant conjunctions that can be shown to hold and they can be detailed more and more such that there is no property of the image in the mind that does not have a covarying property in the image of the retina perhaps (especially if we extend beyond just the image of the retina to the physical firing of certain neurons and so forth). Yet if I understand your position this would not prove that the image on the retina and the image in the mind are the SAME image (identity/reduction). Such a mere perfect correlation of properties that can be described as different would not prove they are the same properties (or even similar to be similar they'd have to be similar in themselves not in the sense of having similar/isomorphic covarying relations).
So likewise we can't prove that the image on the retina and the "appearance" of the moat are the same image all we can prove is they have a set of covarying properties. In order to give the first indication that the arrangement of H2O molecules and the continuous body of water in the moat are the same you'd have to show that the appearance of each (their own image) is the same image as the image formed in the retina, then by the principles that things that are equal to the same thing are equal to each other you'd have proven that a moat and an arrangement of H2O molecules have the same appearance/image. However if you could do that you could prove that the physical image on the retina is the same image as the mental image in the mind. In which case you could prove that pain and c-bundle firing are the same thing (assuming they have a complete set of covarying properties a la a psycho-physical parallelism say).
As long as you allow reduction by saying that a difference can be ignored so long as one shows a close enough similarity in relation between things (and not in the thing itself), that allows one to reduce pain to c-bundle firing. If you don't allow that identity-reduction move you can't actually achieve any reduction because the anti-reductionist can always believe that the two things are just different in themselves even if they have a set of identical covarying relations (are functionally equivalent).
I think there are many interesting and maybe even somewhat compelling arguments for all kinds of anti-reductionism, even if I am mostly unconvinced by them. Some practical and some in principle (some about empirical questions, some about what reduction could ever be), my vague and broad comments here are no doubt very unfair to that rich intellectual vein. If someone is an anti-reductionist about the chemical bond (you don't think Pauling et al. actually reduced it to Quantum Mechanics), there are good reasons for that and those principles will almost certainly allow one to find good reasons to be an anti-reductionist about mind to brain also. I don't see how it is remotely plausible that there could be a situation where anti-reduction arguments fail in every other case, but work so easily in the mind-body case.
> If a statement is such that its truth is fully fixed by how things are presented in a conceivable situation (so there is no further hidden fact that could overturn it), then ideal conceivability of that statement entails its possibility.
Does Kripke somehow prove that Zombies are like that? Or is it just an appeal to conceivability once again - that he can't conceive how it can be any other way? This citation seems to be pointing to the latter:
> Now can something be said analogously to explain away the feeling that the identity of pain and the stimulation of C-fibers, if it is a scientific discovery, could have turned out otherwise? I do not see that such an analogy is possible.
But surely Kripke's argument isn't *that* weak?
I'm also confused by what exactly is being conceived here. According to Kripke, it's enough that P-Zombies are a priori conceivable for them to be logically coherent- i.e. one doesn't need to know all the physical facts of the universe to correctly infer from conceivability of the zombies to their logical coherency. But aren't all the physical facts of the universe already *priced in* the concept of the p-zombie?
P-zombie is an entity from the universe with *the same physics* but without consciousness. A statement about P-zombies is a statement about compatibility between all the physical facts of our universe and the negation of all the facts about consciousness. If this is true, then physicalism which claims that facts about consciousness are physical facts, is false. But evaluating logical coherency of this statement *requires* knowing all the physical facts of our universe, otherwise what are you even conceiving? Just the negation of facts about consciousness? That will not disprove physicalism.