His argument is based on the assumption that any physical fact can be expressed purely by language. That's supposed to be the justification for Premise 1 I guess. But then it also assumes that there are facts about qualia that can't be expressed linguistically, but only experientially, in particular, facts about what it's like to see red. These assumptions are where the flaw in the argument lies, because I don't see why physicalists would be forced to believe both of them. Physicalists who believe that there's no proposition expressing what it's like to see red have no reason to believe the second premise. Those who believe that there is, but that it can be expressed linguistically also have no reason to accept it. And those who believe that there is, but that it can't be expressed linguistically have no reason to believe the assumption that all physical facts can be expressed linguistically.
Yes, I tend to think that proponents of these arguments don't make necessary distinctions like that between theoretical reduction and ontological reduction. It's possible for the thesis of ontological reduction to be true and for theoretical reduction to be false.
If a physicalist believes that the experience of color is a physical fact/process/experience, then the first premise couldn't possibly go through. There would be no way to know all the physical facts without having all the physical experiences. Right?
It's always weird to me that the first premise is seen as a gimme.
Just to give some reaction to the points in your article, apart from the Ability Hypothesis stuff.
I'd be inclined to be as charitable as possible here. Yes, there's a lot that isn't specified, there are a lot of questions that could be asked, but I think there are charitable ways to answer them.
It's not important that Mary only learns from textbooks. She can learn by interactive demonstrations, by having conversations with teachers, etc. She learns in whatever ways you might like as long as she never actually experiences colour. So this means she doesn't see anything colourful, and also that she doesn't do something like take drugs or build a robot neurosurgeon that can rewire her brain to give her colour experience or knowledge of what red is like.
Mary is taken to be arbitrarily smart. Presumably smarter than any actual human. Not necessarily omniscient. The point is that the barrier to her knowledge of what red is like is not supposed to be that there is some important fact she hasn't yet learned or some understanding she hasn't figured out. So she's as smart as the physicalist would like, in other words. If the physicalist thinks that she just has to be smart enough or to know enough facts (via textbooks etc) to know what red is like, then that's how smart she is. The antiphysicalist intuition is that it's just not a question of being smart enough or knowing enough facts, so how smart she is or what she knows doesn't matter.
Yes, it is just a hunch that Mary wouldn't know what red looks like. But I think it's a broadly shared hunch, and one I share. I agree that we shouldn't be too confident about these hunches, but it's worth taking them seriously if it's all we have to go on. And if this hunch is right, it's worth thinking about what that might mean for physicalism. At the very least, the argument does a good job of explaining the intuition that causes people to reject physicalism. If that hunch is right, and if there are (propositional) facts about what red is like (which, again, seems obvious to many people), then physicalism is in trouble.
That's fair enough. What about people who do have the hunch though, as I do? Ultimately everything to do with the HP seems to come down to hunches. So physicalists who have the hunch need to take the KA seriously. Perhaps you don't.
Mary's Bike does not have the same punch because many of us remember how we acquired bike riding skills. We do not remember how we acquired color vision and qualia. It certainly wasn't by reading neuroscience.
There is a deficit known as achromatopsia. Somewhere I read about an adult recovering from it and having this 'alien vision'. Color was splotched onto the objects and ran out of the lines like watercolor. Color vision is comparative for one, and requires some brain wiring to keep the color in the lines.
Scenario: Mary is a medical doctor and sports expert who knows everything about the human body and the mechanics involved in riding a bicycle. She understands the physics, the muscle movements, the balance required—every conceivable detail on a theoretical level. However, she has never actually ridden a bike herself because she has been confined to a room without one.
Outcome: When Mary finally gets the opportunity to ride a bike, she finds that she cannot do it immediately. Despite her exhaustive theoretical knowledge, she lacks the practical skill that comes only with physical practice.
There is a nuanced move in my argument that opens up some different kinds of knowledge. This scenario does not seem to offer the same intuitive punchas MR. But why?
I have so much to say about this, for so many years. This argument is where I started making fun of philosophers. Not that I am that mean, but rather that I was so disappointed that MR reached the printed page.
Also read this: ramachandran2008-I See, But I Don’t Know
BTW, this general kind of response is known as The Ability Hypothesis, and I think it's the right one.
It posits that to know what red is like is not to know a propositional fact as is supposed in the KA. Instead, it's a a suite of abilities, e.g. the ability to imagine red (if not aphantasic), the ability to recognise it and so on. These abilities have a physical, neural basis. And not one that there is any reason to think can be achieved just from learning facts. Like riding a bike.
What I like about the Ability Hypothesis, apart from seeming pretty plausible, is that it grants all the dubious premises the anti-physicalist might like about Mary's ability to learn from books, and that Mary would learn what red is like only when she leaves the room, but it shows how this doesn't disprove physicalism.
Despite agreeing with Dennett on almost everything else, I dislike Dennett's response to the KA, where he suggests that if physicalism is true, then Mary would presumably know what red is like if she knows all the physical facts, and that our problem is just that the case is so unrealistic. I dislike it because there's no reason to think she would know what red is like, and assuming that she would is unnecessary to defeat the argument.
I'm not sure I really see this distinction between defeating an argument easily or accurately/usefully. I think the Ability Hypothesis is accurate and useful.
It's useful because it saves us from futile endless arguments about what Mary would or would not know. Intuitions differ. It doesn't get us anywhere (although, it seems to me that Dennett's intuitions on this are perhaps motivated by the conclusion he wants to get -- the antiphysicalist intuitions here really do seem more plausible to me).
It's accurate because I think it gives the right account of what sort of knowledge Mary acquires when she experiences red. The idea that what red is like is a fact in the sense of propositional knowledge seems quite weird given that it's ineffable and so cannot be expressed as a proposition. But people who know what red is like clearly do have abilities which are lacking in people who don't. And there's no problem for a physicalist who thinks these abilities have a physical basis.
I agree that learning all physical facts is a crazy thing to try to intuit. You can learn a list of facts but how can you possibly internalise all the consequences of those facts? There are plenty of physical systems which we can describe using some simple rules, but have no real idea how to see what the systems actually do without simulating them (e.g. rule 30 for Wolfram’s cellular automata).
However, maybe she doesn't need all the physical facts in order to know what red looks like? I think it’s perfectly physically possible that Mary in her room could read a finite textbook which would give her enough info to work out what red looks like before going outside.
There are people who claim to be aphantasic (they have no mind’s eye). There are also people who claim to have ‘cured’ their aphantasia by reading a particular book and practicing a set of techniques. If Mary’s brain has the ability to see red, it’s easy to imagine that her imagination can see red even before seeing it physically for the first time. Reading this textbook and fully internalising what it says would mean she had perfected these techniques and so she does already know what red looks like.
Maybe this argument isn't in the spirit of the thought experiment though.
His argument is based on the assumption that any physical fact can be expressed purely by language. That's supposed to be the justification for Premise 1 I guess. But then it also assumes that there are facts about qualia that can't be expressed linguistically, but only experientially, in particular, facts about what it's like to see red. These assumptions are where the flaw in the argument lies, because I don't see why physicalists would be forced to believe both of them. Physicalists who believe that there's no proposition expressing what it's like to see red have no reason to believe the second premise. Those who believe that there is, but that it can be expressed linguistically also have no reason to accept it. And those who believe that there is, but that it can't be expressed linguistically have no reason to believe the assumption that all physical facts can be expressed linguistically.
Yes, I tend to think that proponents of these arguments don't make necessary distinctions like that between theoretical reduction and ontological reduction. It's possible for the thesis of ontological reduction to be true and for theoretical reduction to be false.
If a physicalist believes that the experience of color is a physical fact/process/experience, then the first premise couldn't possibly go through. There would be no way to know all the physical facts without having all the physical experiences. Right?
It's always weird to me that the first premise is seen as a gimme.
Hi Nathan,
Just to give some reaction to the points in your article, apart from the Ability Hypothesis stuff.
I'd be inclined to be as charitable as possible here. Yes, there's a lot that isn't specified, there are a lot of questions that could be asked, but I think there are charitable ways to answer them.
It's not important that Mary only learns from textbooks. She can learn by interactive demonstrations, by having conversations with teachers, etc. She learns in whatever ways you might like as long as she never actually experiences colour. So this means she doesn't see anything colourful, and also that she doesn't do something like take drugs or build a robot neurosurgeon that can rewire her brain to give her colour experience or knowledge of what red is like.
Mary is taken to be arbitrarily smart. Presumably smarter than any actual human. Not necessarily omniscient. The point is that the barrier to her knowledge of what red is like is not supposed to be that there is some important fact she hasn't yet learned or some understanding she hasn't figured out. So she's as smart as the physicalist would like, in other words. If the physicalist thinks that she just has to be smart enough or to know enough facts (via textbooks etc) to know what red is like, then that's how smart she is. The antiphysicalist intuition is that it's just not a question of being smart enough or knowing enough facts, so how smart she is or what she knows doesn't matter.
Yes, it is just a hunch that Mary wouldn't know what red looks like. But I think it's a broadly shared hunch, and one I share. I agree that we shouldn't be too confident about these hunches, but it's worth taking them seriously if it's all we have to go on. And if this hunch is right, it's worth thinking about what that might mean for physicalism. At the very least, the argument does a good job of explaining the intuition that causes people to reject physicalism. If that hunch is right, and if there are (propositional) facts about what red is like (which, again, seems obvious to many people), then physicalism is in trouble.
If we specify all this stuff, I have no idea what Im imagining and place no confidence in inferences I make about that imagined case
That's fair enough. What about people who do have the hunch though, as I do? Ultimately everything to do with the HP seems to come down to hunches. So physicalists who have the hunch need to take the KA seriously. Perhaps you don't.
I don't know -- they can have their hunches and neuroscience/linguistics/psychology will continue to progress regardless!
I don't know what 'taking it seriously' means? Quit your neuroscience job?
Well, it might mean, for example, rethinking what is meant by "knowing what red is like".
Im definitely open to doing that
Mary's Bike does not have the same punch because many of us remember how we acquired bike riding skills. We do not remember how we acquired color vision and qualia. It certainly wasn't by reading neuroscience.
There is a deficit known as achromatopsia. Somewhere I read about an adult recovering from it and having this 'alien vision'. Color was splotched onto the objects and ran out of the lines like watercolor. Color vision is comparative for one, and requires some brain wiring to keep the color in the lines.
Mary's Bike
Scenario: Mary is a medical doctor and sports expert who knows everything about the human body and the mechanics involved in riding a bicycle. She understands the physics, the muscle movements, the balance required—every conceivable detail on a theoretical level. However, she has never actually ridden a bike herself because she has been confined to a room without one.
Outcome: When Mary finally gets the opportunity to ride a bike, she finds that she cannot do it immediately. Despite her exhaustive theoretical knowledge, she lacks the practical skill that comes only with physical practice.
There is a nuanced move in my argument that opens up some different kinds of knowledge. This scenario does not seem to offer the same intuitive punchas MR. But why?
I have so much to say about this, for so many years. This argument is where I started making fun of philosophers. Not that I am that mean, but rather that I was so disappointed that MR reached the printed page.
Also read this: ramachandran2008-I See, But I Don’t Know
BTW, this general kind of response is known as The Ability Hypothesis, and I think it's the right one.
It posits that to know what red is like is not to know a propositional fact as is supposed in the KA. Instead, it's a a suite of abilities, e.g. the ability to imagine red (if not aphantasic), the ability to recognise it and so on. These abilities have a physical, neural basis. And not one that there is any reason to think can be achieved just from learning facts. Like riding a bike.
What I like about the Ability Hypothesis, apart from seeming pretty plausible, is that it grants all the dubious premises the anti-physicalist might like about Mary's ability to learn from books, and that Mary would learn what red is like only when she leaves the room, but it shows how this doesn't disprove physicalism.
Despite agreeing with Dennett on almost everything else, I dislike Dennett's response to the KA, where he suggests that if physicalism is true, then Mary would presumably know what red is like if she knows all the physical facts, and that our problem is just that the case is so unrealistic. I dislike it because there's no reason to think she would know what red is like, and assuming that she would is unnecessary to defeat the argument.
Why should our goal be to defeat the Mary's Room argument as easily as possible, rather than to reject in in the most accurate / useful way we can?
I'm not sure I really see this distinction between defeating an argument easily or accurately/usefully. I think the Ability Hypothesis is accurate and useful.
It's useful because it saves us from futile endless arguments about what Mary would or would not know. Intuitions differ. It doesn't get us anywhere (although, it seems to me that Dennett's intuitions on this are perhaps motivated by the conclusion he wants to get -- the antiphysicalist intuitions here really do seem more plausible to me).
It's accurate because I think it gives the right account of what sort of knowledge Mary acquires when she experiences red. The idea that what red is like is a fact in the sense of propositional knowledge seems quite weird given that it's ineffable and so cannot be expressed as a proposition. But people who know what red is like clearly do have abilities which are lacking in people who don't. And there's no problem for a physicalist who thinks these abilities have a physical basis.
Thanks for the recommendation, I will check out that paper.
I agree that learning all physical facts is a crazy thing to try to intuit. You can learn a list of facts but how can you possibly internalise all the consequences of those facts? There are plenty of physical systems which we can describe using some simple rules, but have no real idea how to see what the systems actually do without simulating them (e.g. rule 30 for Wolfram’s cellular automata).
However, maybe she doesn't need all the physical facts in order to know what red looks like? I think it’s perfectly physically possible that Mary in her room could read a finite textbook which would give her enough info to work out what red looks like before going outside.
There are people who claim to be aphantasic (they have no mind’s eye). There are also people who claim to have ‘cured’ their aphantasia by reading a particular book and practicing a set of techniques. If Mary’s brain has the ability to see red, it’s easy to imagine that her imagination can see red even before seeing it physically for the first time. Reading this textbook and fully internalising what it says would mean she had perfected these techniques and so she does already know what red looks like.
Maybe this argument isn't in the spirit of the thought experiment though.