Very interesting! I wasn't aware of subjective physicalism, I'll have to read more. I do wonder if this is a bit of a cheat: "the properties that underwrite conscious experiences (e.g. qualia) are physical", as I see the existence of qualia as a defining characteristic of modern dualism. The physicalist can always expand their notion of what the word "physical" means, but at some point they are just relabeling dualism as physicalism.
Thanks for reading! That's a fair point. Subjective physicalism need to have its own ontology in order to avoid possible dualist conclusions. Howell does that! (We must assess his arguments and whether he achieves the goal of making subjective physicalism different though).
I love this. I was thinking something similar recently, about how even if we imagine there are objective qualia particles/whatever in immaterial souls, we could still make Chalmer's zombie argument. Like, we could imagine angels in heaven making all the same arguments about each other and how they can conceive of zombie angels that are behaviourally identical.
Thanks for reading, Joseph! And yeah! That's the main thrust of Paul Churchland's tu quoque KA argument against dualism because if dualism presents itself as an objective theory, then Mary couldn't know without experience about this psychic, nonphysical stuff that isn't expressible by a complete objective description of the world. When dualism and physicalism are viewed as objective theories, they both encounter the same problem if the KA is sound. Dualism is not better simply because it postulates the existence of a new kind of stuff.
I should add, to my other much longer comment, that what you are calling "Subjective physicalism" is similar to what I have called "Gap Deflationism" or "Gap Compatabilism".
Among physicalists who accept that there is an Explanatory Gap, I think we need to distinguish between those who think the failed explanatory mission was ill-conceived and those who thought it was sensible. Before deciding to be puzzled abut Mary's failures, does it even make sense to expect textbook neurons to give her colour concepts? Should human scientists expect to feel like bats?
We also need to distinguish between those who consider the failure itself opaque (mysterians like McGinn) and those who think the failure is explicable or even mundane (myself and many others).
I didn't know about your Gap Deflationism! I'll check out more of your posts. And I agree with you and those who believe that the failure is explicable rather than unfathomable.
I lay out a spectrum of views on the explanatory gap in the the post called “The Spectrum”. Was my second post ever. Could not link it here. Some of the positions could be considered Gap Compatibilism, which could itself be subdivided according to how explicable we think the gap is.
I was taken to task for not fully defining “Jacksonian Derivation” in this post, but there is only so much one post can cover. I think it is clear what would be involved in succeeding at deriving a natural concept of redness from circuit diagrams alone. I think it is impossible, but we know why it is impossible, so it has no metaphysical implications — only implications about what we can feel comfortable with when we consider metaphysics.
Nice write-up. I got notification you had linked to me, but then got busy in real life and never got a chance to follow-up.
My opinions on this are pretty clear, and you explain the core point of my own distinction between epistemic irreducibility and ontological irreducibility.
You frame physicalism as coming in two flavours, one that assumes epistemic transparency from explicitly physical facts, and one that doesn't. I think the assumption of epistemic transparency is not a version of physicalism, but a bad idea that anti-physicalists (and a few physicalists) assume when they sit down to consider the plausibility of physicalism and its explanatory commitments.
The Knowledge Argument is one of those arguments that never struck me as even vaguely convincing. Following the information pathways through Mary's brain, I see absolutely no reason for textbook neurons to give her the concepts she is reading about. It makes no sense even for those with no neuroanatomical knowledge, and it is ludicrous for those who have a sense of the diverging pathways and different brain regions involved in: 1) analysing circuit diagrams vs 2) entertaining direct representations of colour.
Once you feel the utter pointlessness of trying to recreate concepts of redness from black-and-white diagrams, then qualia no longer see mysterious. There is residual cognitive frustration, but that is not the same as a sense of meta-physical bafflement. There are just some things brains can't do, for reasons that make sense. See, for instance, by post on the hypercube. I don't think physicalism has any commitment to cognitive transparency on hypercubes, qualia, or a number of other difficult concepts.
If minds come from brains, then physicalism is instead committed to cognitive limits, so our default expectation should be that we face major epistemic blindspots.
Nagel's famous paper sits awkwardly in this space. He presents an epistemic case for a sense of bafflement (which I don't share), but he also argues (or strongly implies) that physicalism fails as an ontological thesis if human brains cannot map their own cognition onto the perceptions of bats. This is a bizarre belief. We might have to pause for a few minutes to see why Mary cannot adopt a common human conceptual state, but we don't have to pause at all to know why homo sapiens brains cannot map their own perceptual states directly onto bats.
One way of looking at the concerns of Nagel, Chalmers, and the initial concerns of Jackson (before he saw the light) is to note that they all made the same mistake as Searle: judging one cognitive system according to what can be achieved in a different cognitive system that has objective, explicable and often mundane reasons for not being able to adopt some cognitive state. The two cognitive systems involved (the human scientists and the bats, or the human operator and the Chinese Room) have obviously divergent features in the objective domain, so none of these examples serve as evidence for some mysterious divergence requiring ontological speculation. The failures of Searle's human operator, Nagel's bat scientists, and pre-release Mary are all of a similar kind, and all explicable without even engaging in philosophy. Standard mechanical accounts of cognition do just fine.
Thanks for reading, Zinbiel! The proposition that if physicalism is true, then it is possible to derive or instantiate phenomenal, qualitative, and subjective facts and properties from an exhaustive objective description and knowledge of the world is suspicious if we think that there is no a priori knowledge and the supposedly a priori and speculative means of knowing are unreliable. Usually, physicalism is thought to be an important tenet of naturalism, and naturalism often rejects the existence of a priori knowledge or at least reduces the importance or value of this kind of knowledge. Thinking that our brains have epistemic, transparent access to their own cognitive structure seems foolish in the face of neuroscientific knowledge about the brain's cognitive limitations, and this is not actually a literal meta-physical problem. At best, it's a scientific problem. In any event, some philosophical problems seem to me to be produced by traditional, speculative philosophical methods, and this is especially true in the philosophy of mind. Regarding Nagel, I have no objections to him (at the moment). As I wrote in the article, a physicalist might be content with his insights while remaining a physicalist. As I understand him, some facts (like facts about consciousness considered as a point of view) are only knowable through that particular point of view. However, this fact about consciousness raises an epistemic or cognitive issue rather than an ontological one. The failure is in seeing ontological and speculative implications in mere epistemic and cognitive problems that can be handled with (in principle) by science. So yeah, as you say, none of these issues and antiphysicalist arguments require positing new, non-physical entities.
Very nice analysis. In the end, I'm not sure I can make sense of the idea of subjective knowledge as the same kind of thing as objective knowledge, though.
When we imagine what subjective knowledge consists in, we invariably use objective (or intersubjective) categories (i.e. linguistic descriptions). When we try to eliminate those by some kind of internal gesturing, we're left with something purportedly ineffable, but how could that count as knowledge? Equally, such 'facts' are supposed to be known absolutely and by acquaintance, i.e. without reasoned justification, demonstration or corroboration. Are those hallmarks of anything we want to call knowledge? It sounds rather like revelation and faith.
Well, we can stipulate that it's useful to speak of such things as 'knowledge' because it comes naturally, but this knowledge seems so different from the objective kind in origin, content, function and practice that we ought to recognise these as two quite distinct kinds of 'knowledge' talk.
That then raises the question of whether what we're calling subjective knowledge matters to discussions of physicalism - I suspect it ultimately doesn't because it's a way of talking about human behaviour rather than fundamental ontology.
Very interesting take! I think that acknowledging the existence of subjective/phenomenal/qualitative knowledge doesn’t necessarily mean that “subjective facts” (which are the objects of that knowledge) are ineffable (i.e., inexpressible by our languages, ordinary and scientific) or that they can only be known by acquaintance or subjectively. Suppose that an instance of a subjective fact is qualia. Howell (in his 2007 article) mentions that the issue involved by the KA is logically distinct from the Ineffability Thesis—qualia cannot be captured by physical description because they are ineffable. Phenomenal experiences are less amenable to communication to standard propositional facts. However, the Ineffability Thesis is distinct from the claim that acquaintance is necessary for subjective knowledge, implying that a full understanding of qualia logically requires actually undergoing the state. The issue is not whether subjective facts are ineffable but whether there are subjective facts, aspects of the world that can only be grasped through subjective experience. Now, whether those facts are partially or totally excluded by objective knowledge is another issue. Both versions of subjective physicalism respond differently to that question: the first version only says that they cannot be fully grasped by objective knowledge, but objective, physical knowledge doesn’t leave anything out, and the second one says they are totally excluded by it. This last version, however, resembles property dualism. In any event, it doesn’t seem to me that it’s so obvious* that there is subjective knowledge. I don’t totally understand those philosophers who think that Mary learns something new. What exactly does she learn? What is a subjective fact? What does “what it is like to see red" even mean?
Interesting. I'll take a look at Howell's article - I can probably buy that the ineffability thesis is a logically distinct problem in relation to qualia, my issue would be that to express something at all is necessarily to express something public/objective because of the way language works.
So suppose Mary wanted to say something like "this is what red looks like", allowing at least an expressible predicate. Well, we have to be careful how we interpret such statements, avoiding grammatical illusions - it could look like she's saying that 'red' is an abstract or platonic category, and her particular experience is an instance of it. But if one is not a platonist and acknowledges the cultural relativism of colour terms, a better interpretation might be something like: this is what people in my language call 'red'. But now you're in a tricky situation because people don't have access to each others' experiences to be calling experiences 'red', they're calling objects red (albeit mediated by experience), so Mary really ends up saying something like: this is the kind of object that people in my language call 'red'. Now you're fully third-person, and what looked like an expressible subjective fact was an illusion because, by speaking, language has brought you into the objective space.
I don't know if this is splitting the wrong hairs, but maybe instead of distinguishing objective and subjective knowledge, I'd distinguish world knowledge and conventional knowledge - both objective, but the first has to do with human-independent facts and the second has to do with human conventions (e.g. language use). With the second, the tricky thing about it is that we both learn and co-construct the facts by participating in conventions which change as we make use of them. It's subjective in the sense that individual experience is involved in making the facts what they are (e.g. by using 'red' in a certain way) but I don't think a person 'knows' something subjective so much as they're embodying what objective conventional knowledge is about.
I cannot respond to that objection fully now. Responding to it necessitates an understanding of what "subjective knowledge" is or means—which, to be honest, I still don't fully grasp. I understand that you're arguing that Mary communicates to others what it is for her to see red by means of language and using it implies that her knowledge of what it is to see red is, in the end, public rather than private—that is, it's actually objective or intersubjective knowledge, not subjective knowledge. Other people understand what it is to see red or what red looks like; this fact makes possible the use of "red" as a meaningful term. I need to read more about this issue, but this argument is, by my lights, very persuasive.
Excellent post! I have been extolling the virtues of subjective physicalism for several weeks after stumbling across articles by Robert Howell and Tim Crane. It's a terrible name though, since it suggests some mythical beast with the body of a physicalist and the head of a idealist.
Howell and Crane don't try to figure out what features of our physical world permit subjects and subjective facts. I'm guessing the answers will arrive via work on autopoiesis and organizational closure.
Thanks for reading! Yes, the name is terrible cause it suggests it's kinda property dualism. In fact, an objection I could have addressed (but I didn't want) is that the fact that Mary learns a new facts means physicalism is false; that new fact should be non-physical, and it doesn't seem that subjective physicalism can do justice to her epistemic achievement. I don't think it's a decisive objection. Crane and Howell address it. But yes, we must figure out what are subjective facts and if they really exist.
The failure of objectivism is simply an epistemic inadequacy. We should be used to this sort of difficulty by now: Godel's Theorem, independence results in set theory, inaccessible regions of spacetime, quantum uncertainty, free will paradoxes, etc.
Very interesting! I wasn't aware of subjective physicalism, I'll have to read more. I do wonder if this is a bit of a cheat: "the properties that underwrite conscious experiences (e.g. qualia) are physical", as I see the existence of qualia as a defining characteristic of modern dualism. The physicalist can always expand their notion of what the word "physical" means, but at some point they are just relabeling dualism as physicalism.
Thanks for reading! That's a fair point. Subjective physicalism need to have its own ontology in order to avoid possible dualist conclusions. Howell does that! (We must assess his arguments and whether he achieves the goal of making subjective physicalism different though).
I love this. I was thinking something similar recently, about how even if we imagine there are objective qualia particles/whatever in immaterial souls, we could still make Chalmer's zombie argument. Like, we could imagine angels in heaven making all the same arguments about each other and how they can conceive of zombie angels that are behaviourally identical.
Thanks for reading, Joseph! And yeah! That's the main thrust of Paul Churchland's tu quoque KA argument against dualism because if dualism presents itself as an objective theory, then Mary couldn't know without experience about this psychic, nonphysical stuff that isn't expressible by a complete objective description of the world. When dualism and physicalism are viewed as objective theories, they both encounter the same problem if the KA is sound. Dualism is not better simply because it postulates the existence of a new kind of stuff.
I should add, to my other much longer comment, that what you are calling "Subjective physicalism" is similar to what I have called "Gap Deflationism" or "Gap Compatabilism".
Among physicalists who accept that there is an Explanatory Gap, I think we need to distinguish between those who think the failed explanatory mission was ill-conceived and those who thought it was sensible. Before deciding to be puzzled abut Mary's failures, does it even make sense to expect textbook neurons to give her colour concepts? Should human scientists expect to feel like bats?
We also need to distinguish between those who consider the failure itself opaque (mysterians like McGinn) and those who think the failure is explicable or even mundane (myself and many others).
I didn't know about your Gap Deflationism! I'll check out more of your posts. And I agree with you and those who believe that the failure is explicable rather than unfathomable.
I lay out a spectrum of views on the explanatory gap in the the post called “The Spectrum”. Was my second post ever. Could not link it here. Some of the positions could be considered Gap Compatibilism, which could itself be subdivided according to how explicable we think the gap is.
I was taken to task for not fully defining “Jacksonian Derivation” in this post, but there is only so much one post can cover. I think it is clear what would be involved in succeeding at deriving a natural concept of redness from circuit diagrams alone. I think it is impossible, but we know why it is impossible, so it has no metaphysical implications — only implications about what we can feel comfortable with when we consider metaphysics.
Nice write-up. I got notification you had linked to me, but then got busy in real life and never got a chance to follow-up.
My opinions on this are pretty clear, and you explain the core point of my own distinction between epistemic irreducibility and ontological irreducibility.
You frame physicalism as coming in two flavours, one that assumes epistemic transparency from explicitly physical facts, and one that doesn't. I think the assumption of epistemic transparency is not a version of physicalism, but a bad idea that anti-physicalists (and a few physicalists) assume when they sit down to consider the plausibility of physicalism and its explanatory commitments.
The Knowledge Argument is one of those arguments that never struck me as even vaguely convincing. Following the information pathways through Mary's brain, I see absolutely no reason for textbook neurons to give her the concepts she is reading about. It makes no sense even for those with no neuroanatomical knowledge, and it is ludicrous for those who have a sense of the diverging pathways and different brain regions involved in: 1) analysing circuit diagrams vs 2) entertaining direct representations of colour.
Once you feel the utter pointlessness of trying to recreate concepts of redness from black-and-white diagrams, then qualia no longer see mysterious. There is residual cognitive frustration, but that is not the same as a sense of meta-physical bafflement. There are just some things brains can't do, for reasons that make sense. See, for instance, by post on the hypercube. I don't think physicalism has any commitment to cognitive transparency on hypercubes, qualia, or a number of other difficult concepts.
If minds come from brains, then physicalism is instead committed to cognitive limits, so our default expectation should be that we face major epistemic blindspots.
Nagel's famous paper sits awkwardly in this space. He presents an epistemic case for a sense of bafflement (which I don't share), but he also argues (or strongly implies) that physicalism fails as an ontological thesis if human brains cannot map their own cognition onto the perceptions of bats. This is a bizarre belief. We might have to pause for a few minutes to see why Mary cannot adopt a common human conceptual state, but we don't have to pause at all to know why homo sapiens brains cannot map their own perceptual states directly onto bats.
One way of looking at the concerns of Nagel, Chalmers, and the initial concerns of Jackson (before he saw the light) is to note that they all made the same mistake as Searle: judging one cognitive system according to what can be achieved in a different cognitive system that has objective, explicable and often mundane reasons for not being able to adopt some cognitive state. The two cognitive systems involved (the human scientists and the bats, or the human operator and the Chinese Room) have obviously divergent features in the objective domain, so none of these examples serve as evidence for some mysterious divergence requiring ontological speculation. The failures of Searle's human operator, Nagel's bat scientists, and pre-release Mary are all of a similar kind, and all explicable without even engaging in philosophy. Standard mechanical accounts of cognition do just fine.
Thanks for reading, Zinbiel! The proposition that if physicalism is true, then it is possible to derive or instantiate phenomenal, qualitative, and subjective facts and properties from an exhaustive objective description and knowledge of the world is suspicious if we think that there is no a priori knowledge and the supposedly a priori and speculative means of knowing are unreliable. Usually, physicalism is thought to be an important tenet of naturalism, and naturalism often rejects the existence of a priori knowledge or at least reduces the importance or value of this kind of knowledge. Thinking that our brains have epistemic, transparent access to their own cognitive structure seems foolish in the face of neuroscientific knowledge about the brain's cognitive limitations, and this is not actually a literal meta-physical problem. At best, it's a scientific problem. In any event, some philosophical problems seem to me to be produced by traditional, speculative philosophical methods, and this is especially true in the philosophy of mind. Regarding Nagel, I have no objections to him (at the moment). As I wrote in the article, a physicalist might be content with his insights while remaining a physicalist. As I understand him, some facts (like facts about consciousness considered as a point of view) are only knowable through that particular point of view. However, this fact about consciousness raises an epistemic or cognitive issue rather than an ontological one. The failure is in seeing ontological and speculative implications in mere epistemic and cognitive problems that can be handled with (in principle) by science. So yeah, as you say, none of these issues and antiphysicalist arguments require positing new, non-physical entities.
Very nice analysis. In the end, I'm not sure I can make sense of the idea of subjective knowledge as the same kind of thing as objective knowledge, though.
When we imagine what subjective knowledge consists in, we invariably use objective (or intersubjective) categories (i.e. linguistic descriptions). When we try to eliminate those by some kind of internal gesturing, we're left with something purportedly ineffable, but how could that count as knowledge? Equally, such 'facts' are supposed to be known absolutely and by acquaintance, i.e. without reasoned justification, demonstration or corroboration. Are those hallmarks of anything we want to call knowledge? It sounds rather like revelation and faith.
Well, we can stipulate that it's useful to speak of such things as 'knowledge' because it comes naturally, but this knowledge seems so different from the objective kind in origin, content, function and practice that we ought to recognise these as two quite distinct kinds of 'knowledge' talk.
That then raises the question of whether what we're calling subjective knowledge matters to discussions of physicalism - I suspect it ultimately doesn't because it's a way of talking about human behaviour rather than fundamental ontology.
Very interesting take! I think that acknowledging the existence of subjective/phenomenal/qualitative knowledge doesn’t necessarily mean that “subjective facts” (which are the objects of that knowledge) are ineffable (i.e., inexpressible by our languages, ordinary and scientific) or that they can only be known by acquaintance or subjectively. Suppose that an instance of a subjective fact is qualia. Howell (in his 2007 article) mentions that the issue involved by the KA is logically distinct from the Ineffability Thesis—qualia cannot be captured by physical description because they are ineffable. Phenomenal experiences are less amenable to communication to standard propositional facts. However, the Ineffability Thesis is distinct from the claim that acquaintance is necessary for subjective knowledge, implying that a full understanding of qualia logically requires actually undergoing the state. The issue is not whether subjective facts are ineffable but whether there are subjective facts, aspects of the world that can only be grasped through subjective experience. Now, whether those facts are partially or totally excluded by objective knowledge is another issue. Both versions of subjective physicalism respond differently to that question: the first version only says that they cannot be fully grasped by objective knowledge, but objective, physical knowledge doesn’t leave anything out, and the second one says they are totally excluded by it. This last version, however, resembles property dualism. In any event, it doesn’t seem to me that it’s so obvious* that there is subjective knowledge. I don’t totally understand those philosophers who think that Mary learns something new. What exactly does she learn? What is a subjective fact? What does “what it is like to see red" even mean?
Interesting. I'll take a look at Howell's article - I can probably buy that the ineffability thesis is a logically distinct problem in relation to qualia, my issue would be that to express something at all is necessarily to express something public/objective because of the way language works.
So suppose Mary wanted to say something like "this is what red looks like", allowing at least an expressible predicate. Well, we have to be careful how we interpret such statements, avoiding grammatical illusions - it could look like she's saying that 'red' is an abstract or platonic category, and her particular experience is an instance of it. But if one is not a platonist and acknowledges the cultural relativism of colour terms, a better interpretation might be something like: this is what people in my language call 'red'. But now you're in a tricky situation because people don't have access to each others' experiences to be calling experiences 'red', they're calling objects red (albeit mediated by experience), so Mary really ends up saying something like: this is the kind of object that people in my language call 'red'. Now you're fully third-person, and what looked like an expressible subjective fact was an illusion because, by speaking, language has brought you into the objective space.
I don't know if this is splitting the wrong hairs, but maybe instead of distinguishing objective and subjective knowledge, I'd distinguish world knowledge and conventional knowledge - both objective, but the first has to do with human-independent facts and the second has to do with human conventions (e.g. language use). With the second, the tricky thing about it is that we both learn and co-construct the facts by participating in conventions which change as we make use of them. It's subjective in the sense that individual experience is involved in making the facts what they are (e.g. by using 'red' in a certain way) but I don't think a person 'knows' something subjective so much as they're embodying what objective conventional knowledge is about.
I cannot respond to that objection fully now. Responding to it necessitates an understanding of what "subjective knowledge" is or means—which, to be honest, I still don't fully grasp. I understand that you're arguing that Mary communicates to others what it is for her to see red by means of language and using it implies that her knowledge of what it is to see red is, in the end, public rather than private—that is, it's actually objective or intersubjective knowledge, not subjective knowledge. Other people understand what it is to see red or what red looks like; this fact makes possible the use of "red" as a meaningful term. I need to read more about this issue, but this argument is, by my lights, very persuasive.
Excellent post! I have been extolling the virtues of subjective physicalism for several weeks after stumbling across articles by Robert Howell and Tim Crane. It's a terrible name though, since it suggests some mythical beast with the body of a physicalist and the head of a idealist.
Howell and Crane don't try to figure out what features of our physical world permit subjects and subjective facts. I'm guessing the answers will arrive via work on autopoiesis and organizational closure.
Thanks for reading! Yes, the name is terrible cause it suggests it's kinda property dualism. In fact, an objection I could have addressed (but I didn't want) is that the fact that Mary learns a new facts means physicalism is false; that new fact should be non-physical, and it doesn't seem that subjective physicalism can do justice to her epistemic achievement. I don't think it's a decisive objection. Crane and Howell address it. But yes, we must figure out what are subjective facts and if they really exist.
The failure of objectivism is simply an epistemic inadequacy. We should be used to this sort of difficulty by now: Godel's Theorem, independence results in set theory, inaccessible regions of spacetime, quantum uncertainty, free will paradoxes, etc.
I'm uninformed when it comes to phil of mind but this was informative, I'll surely make sure to read Zinbiel's post soon too!