The problem with the knowledge argument

100 thoughts on “The problem with the knowledge argument”

1. I have never been impressed with the knowledge argument.

   I suspect it is connected with the characterization of knowledge as “justified true belief”, and that has always seemed wrong to me. It attempts to equate knowledge and information, but those are not at all the same.

   LikeLiked by 1 person

   Reply

   1. Sounds like we agree on the knowledge argument.

      What would you say distinguishes knowledge, belief, and information? I tend to think of belief as information in a mind, and knowledge as higher certitude belief.

      LikeLiked by 1 person

      Reply

      1. Information is what is used in a communication channel. Knowledge is what connects the information to reality. I tend to think of belief as involving an emotional commitment.

         LikeLiked by 1 person

         Reply

         1. So what would you call my mental image of a stop sign knocked down at a particular corner in my neighborhood, a proposition with which I hold no particular emotion? I tend to think I believe the stop sign is knocked down, although there’s a slim possibility the city has come out and fixed it already, so I can’t claim to know it’s still knocked down.

            LikeLiked by 1 person

            Reply

            1. I’m doubtful of propositions. I think they lead to confusion.

               There are degrees of belief. Some we might believe quite strongly, while for others the belief might be shallow. And the degree of belief has to do with that emotional commitment. I would guess that your belief about that stop sign is a rather shallow one.

               LikeLiked by 2 people

2. Let’s break down the argument a few levels. To experience something is to directly acquire it via one or more of the 5 senses (sight, smell, hearing, taste, and/or touch) of our physical body. To know something is deduce, observe, read, hear, believe, hope, fear, think, guess, dream, imagine, etc. something directly or indirectly from any internal or external source. Both knowledge and experience, after being perceived, are shared via physical neurons with various other parts of the brain, which may or may not activate associated thoughts, actions, and/or behaviors throughout our physical body, which may or may not cause physical or mental consequences outside our bodies. What is non-physical here? How are processes deemed non-physical supposed to function? Redness seems to be the favorite argument of the non-physicalists but pain, fear, or sex may be substituted for redness. Imagine that Mary can’t or has never experienced pain or she has never experienced the fear of battle or facing execution, or she has never had sex. These things can’t be completely known until they are experienced but experience as well as knowledge are physical processes for animals like us. Concepts are abstract but we process them physically.

   LikeLiked by 1 person

   Reply

   1. Thanks Mike. That’s a good analysis of the practical realities involved.

      The fear of battle or execution and sex are good examples. She could, in principle, maybe be in virtual reality simulation of these things. But of course, there would always be an element missing. She’d know her life or health wasn’t really in danger in the battle and execution ones, or that the virtual person she was having virtual sex with wasn’t a real life human being. It might get close if we could arrange for her to think these are real events, but we’re getting into that non-practical territory most of these thought experiments live in.

      LikeLiked by 1 person

      Reply

      1. When we’re talking about “redness”, I think we’re talking about the real reality of redness, not the virtual reality of redness or the sweet-sour taste of an apple. Simulating a airplane crash will never give you the same experience of really crashing a plane. Experience is real and therefore physical IMHO.

         LikeLiked by 1 person

         Reply

         1. I agree that it’s real and completely physical. Of course, non-physicalists will say we’re only talking about the functional correlates, not the actual experience. Not sure, other than Occam, that there’s any way to ever resolve it.

            LikeLiked by 1 person

            Reply

            1. Why is it that when we say something is physical we have to prove it to someone’s satisfaction but when someone says it’s not physical, he/she is not compelled to prove it or explain how the non-physical process is supposed to work? I’d be willing to bet that a physicalist will eventually come up with a testable hypothesis to prove consciousness is physical before a non-physicalist can come up with a testable hypothesis to prove it’s not.

               LikeLiked by 1 person

            2. Unfortunately I don’t think a testable hypothesis to rule out the non-physical is possible. Someone can always come up with a theory with more assumptions than the more minimal one that fits the data. Our only heuristic for rejecting the additional assumptions is parsimony, Occam’s razor, and it’s often pointed out to me that it’s just a heuristic, not a guarantee, although historically it’s usually right. Sometimes another testable theory that contradicts the more complex one will decisively rule it out. But in most cases, sustained lack of evidence for the more complex theory might make it fade away, eventually.

               LikeLiked by 1 person

3. It’s another poorly stated problem. This is clear when you look at any review of the argument and see all the ‘perhaps’es that appear as attempts are made to better define Mary’s condition before and after she leaves the room. Puttings aside what colour Mary’s skin might be, or her blood, and assuming she only sees black and white (explain that one Mary) we try to keep it as simple if unrealistic as possible. But, by some miracle akin to the argument’s claim for something additional to the physical world, she gains colour sight on leaving the room. Even then we have to ignore the fact that anyone discovering a new sense for the first time will not have the biological background and experience to interpret it. So, all the nonsense aside … well, just a bit more …

   What does this mean for Mary in the room? She sees “black and white (and presumably scales of gray)” … But hold on, “black and white”? White light contains other fequencies. Does refraction and defraction splitting of colours not occur in her room? OK, let’s assume her room world is just the intensity of single frequency. But what frequency of light in this monochrome world? If it’s red light intensity, she has experienced red! She just couldn’t put it into the context of more colours. …

   OK, really, all that aside too …

   It’s about the difference between information and the exprience of information in the brain, aas ‘knowledge’.

   1 – When in the room Mary clearly isn’t acquiring all knowledge about colour, because her brain has not been triggered by channels her eyes would otherwise process.

   2 – When Mary leaves the room she is reciving information she hasn’t received before – more colour radiation that contains more information about the world.

   Both in and out of the room Mary’s brain processes the information she receives, and she receives more information outside the room that in: An item that to Mary in the room looking out looks like a single colour, but outside she receives different signals and her ‘expereince’ is the processing of that new information.

   The slight of hand in this thoughtless experiment is using ‘information’ in different ways. When they say she has all possible information about colour and colour perception they are talking about descriptive information – words, graphs, even images, but the images would be monochrome … as those images are transmitted into the room they lose information, by definition.

   But all this is missed by the dualists. They think data, as information, can be transmitted whole. But in never can be. The premise of complete information is nonsense.

   And further, we know from many illusions (genuine optical illusions, where the information received by the eyes is the same for two different scenarios, or more common mental illusions, where the brain interprest different stimuli as the same, or the same stimuli as different) – we know that the brain invents stuff; we know that when we look out at the world, most of what we ‘see’ is reconstructed by the brain from experience. Everything about the brain experience of sensory inputs is just what we’d expect from the type of biological components we’re made of.

   The “what it’s like to be” aspect is irrelevant. Without other evidence that there is something else going on that is non-physical, all we have is the physical. The fact that a physical system processes data internally, as ‘experience’, adds nothing new to our understanding.

   If the dualists want to make up stuff about ‘something else’, why do they never imagine invisible Numskulls working in the brain? What it’s like to be a human perceiving and experienceing red might as well be a brain Numskull.

   https://ukcomics.fandom.com/wiki/The_Numskulls

   LikeLiked by 1 person

   Reply

   1. For what it’s worth, it’s perfectly feasible to do the experiment largely as you describe. Just raise Mary in rooms lit with sodium lamps. These lamps emit a very restricted spectrum, mostly orange. There was a place in San Francisco, might still be there, and I think it was called something like the Exploratorium. They had such a room, with one wall a giant color photograph of SanFran, and a machine that dispensed jelly beans. Everything appeared to be shades of orange. But they also provided flashlights, and it was amazing to see the effect when you shine the light on the wall or jelly beans.

      *

      LikeLiked by 1 person

      Reply

      1. Thanks, hadn’t heard of that.

         LikeLike

         Reply

   2. Right, the room is obviously not a serious description. Even if her skin is coated in white paint or something, Mary could have the experience of red merely by closing her eyes while facing a white light, or seeing phosphenes from rubbing her eyes.

      Whenever I think about this thought experiment, I actually do a mental swap, imagining that Mary has some condition, like achromatopsia caused by a lesion in her V4 regions, and that what enables her to experience color is some kind of medical procedure.

      Of course, that’s still ignoring issues like critical development periods. She wouldn’t be able to perceive colors in the way we can. She could probably acquire some ability in the months after having her color circuitry fixed, but it might never be as much as someone who developed it as a child.

      In relation to these issues, in the old discussion on this, someone shared this study of red-green color blind monkeys having their ability to discriminate red restored by a genetic therapy: https://www.livescience.com/21275-color-red-blue-scientists.html

      I’m onboard with much of the rest of your remarks about the problems with non-physical concepts.

      LikeLiked by 1 person

      Reply

4. Mary learned something when she went outside. But someone would have to tell her the color of the shirt, or flower she saw was called blue. (As a side note, I don’t believe everyone sees blue or any color exactly the same. People have preferences, likes and dislikes depending on how colors appear to them.) Whether you say what she learned was “information” is a question of language and meaning . The accepted use of the word information would say she gained information. Changing an accepted definition of a word in common use is a tough task. But definitions of words change over time. That us why people discuss the meanings of words and why it is interesting.

   LikeLiked by 1 person

   Reply

   1. I covered in the post why I think she might be able to name the color of a flower, but admittedly in any practical sense, it would depend on how much context she had. If she’s outside and can see a clear sky, she’s should see the similarities between that hue and the flower’s. It’s why I noted the toughest case is an isolated color patch.

      People do have different numbers of S, M, and L light cones in the retina, which does change the exact mix of signaling their brain receives for the same stimuli. People even have different numbers as they age. But my memory of purple things I saw as a boy is a reconstruction of how I perceive purple today.

      LikeLike

      Reply

5. My impression is that the dispute reduces to a mere gainsay over whether there can be “non-physical” information. The defenders of the “Mary’s room” thought-experiment assert that there can be; critics assert that there can’t be; and in your two-line summary, the real argument is over whether one term can be substituted for the other. Sides have already been taken, and the thought-experiment itself provides no traction either way.

   It might be more useful to unpack the idea of “physical information,” and to understand it what sense it might (or might not) be legitimate to speak of “non-physical information.” What’s really at issue here is the assumption that “information” must always be understood as “physical information.”

   We might ask whether a train timetable is always “physical.” Certainly, it always takes a physical manifestation. But the timetable itself, considered as information, is an abstraction built upon abstractions, which can take various physical expressions and is not functionally the same as any of them. You can burn it, but you can’t burn it.

   When we lock the timetable down to a physical expression, we forfeit certain useful ways of talking about it. “I burned the timetable and the link is broken, but I can still give you the information” becomes a philosophical puzzle, for those who like to go there. I prefer to back off and allow ways of talking that do not confine us so rigorously to “the physical.”

   LikeLiked by 1 person

   Reply

   1. Good point about the physical and non-physical sides. I probably should have clarified that while the thought experiment doesn’t really challenge physicalism, it doesn’t affirm it either. Nothing about it necessarily challenges a non-physicalist position, at least other than removing one particular argument for it.

      On discussing abstract information, we do have a way of talking about that being non-physical. I’m sure I do it myself in many places. In my job, we often talk about whether something is software or hardware, with the understanding that software isn’t physical. Although if you challenge someone, they’ll likely stipulate that of course software is physical, it’s just not any particular physicality.

      One way of thinking about it is as a category of physical things. So if I think about the book “War and Peace”, which can exist in a variety of forms, hardback, paperback, ebook, audible, etc, I can either be talking about one particular instantiation, or the category of physical patterns that are instantiations of “War and Peace.”

      But this ties into the question of platonism we discussed. It certainly is a convenient way of talking to say abstract objects exist in their own right. If we’re talking about the modern version of platonism, we end up saying they do exist, but in a particular sense of “existing”. Which is why I mostly think this is a matter of how we want to talk about it, without any particular fact of the matter. But I recognize that’s a judgment I reach for a lot in philosophical discussions, maybe too much.

      LikeLiked by 1 person

      Reply

   2. Re the time table – When you read it, physical radiation reflected from the page is received and processed by you brain via your eyes. The information is copied. You might be able to recall the times, or you might be able to recall the image of the paper, with the times on it – these are details of personal memory recording, recall, and reconstruction.

      The important point herre is that living brain-bodies consume energy – the copying is not free of physical energy consumption, or the alteration of physical states in the brain. So, this copy exists even when you burn the original table.

      So, the information might be thought of in the abstract, but even thinking of the abstraction is a physical act, involving physical patterns in the conceiving of the abstration.

      You could say that whenever we recall a memory we are creating a new temporary copy in consciousness (whatever physical instantiation that takes in the neurons).

      Copying information is never free of phsyical effect.

      Pick up a pebble and drop it into wet sand. It leaves a phsycial imprint. Do it again and again – the copy imprints are not free of physical change, in the sand. Eventually, even the pebble will wear away.

      We used to know this from analogue recordings that need a master for consistent reproduction. We have been fooled by digital information transfer and copying. It seems we can copy digital data accurately over and over. But it takes error correction and reconstruction of deteriorating digital signals. There’s always a physical cost, a physical effect.

      Back the the Mary problem – as in the OP, it presupposes that there’s something other than the phsyical, then declares these thought examples show it, when they don’t. There’s also never any explaination of how the non-phsyical and the phsyical interact, as they must do – how does a physical sense affect a non-physical experience … or how do non-physical mental events affect the brian to stimulate motor actions.

      “critics assert that there can’t be [non-physical]” – I disagree. The assertion is that all we know about is the physical (ignoring solipsism, which tosses all knowledge into the air), and to claim there’s something else requires evidence to be taken seriously.

      LikeLike

      Reply

      1. The timetable is of course instantiated in some way in a human memory. But its existence as “information” requires an interpretive act of some kind. In these mind-brain discussions it’s common to note that you can’t read the timetable by examining the corresponding structures in the brain. One can imagine an amazing apparatus that makes this possible, but even the output of the apparatus is only so many pixels or marks on paper. Up to this point it is information only in the syntactical sense of some components in an arrangement. There’s another sense of information at play in these discussions, the semantic; and the connection between them is exactly as mysterious as the connection between the physical and the non-physical—if we want to make a mystery of it.

         Enthusiasts for the physical tend to make a mystery out of the semantic by insisting it must be equated with the syntactic, leaving us all struggling to make some connection between them. The assertion that “all we know about is the physical” in fact assumes that the physical has only a syntactic aspect, or that only its syntactical aspect is “real.” Can the physical have a semantic aspect? I don’t see why not—in fact, I don’t see how to avoid it, given the facts present to us—but here the traditional physicalist balks, forcing us to relegate the semantic to some mysterious realm of “mind” that runs in parallel with the syntax of “matter.”

         LikeLike

         Reply

         1. I see ‘semantic’ as only a contextual perspective of some system – the semantic meaning means something to the system only because of the context.

            A computer kernel and OS manage data at various levels. An Excel v Word program have a context for a byte/word of data, as value in a spreadsheet cell or a character in a word in Word. But as a byte with the same value is passed down to the OS, kernel, CPU, that semantic context is lost and it’s just data.

            Computer systems can monitor themselves, for errors, for operating temperature, and so on, to try to maintain something like a stasis. This is how I see the basis of consiousness: a brain process that provides a context for other brain content in the context of the sensory stimulus, driving brain-body action. And this is the context in which ‘meaning’, semantics evolves. We see this developed to a great extent in humans, compared to other mammals, inlcuding bats, and to anaimals, with far fewer neurons.

            It seems mysterious because we can’t easly delve into our own braisn to see, or rather feel or sense, the relationship between the data we conjure up in consciousness and the neurons that are doing the processing of that process – no more than a spreadheet program has information about CPU and RAM chips.

            To me this seems the source of the illusion of dualism – which in turn is the source of why philosophers and the religious are looking for ‘something else’ when Mary experiences colour. What we feel is a disconnected mind is a process that monitors its own container, the brain, but only in general higher level terms. It feels like a free floating ‘mind’ behind the eyes*.

            And this is the context in which humans give semantic meaning to syntactical information.

            (* Not sure what it feels like to someone born blind, but it may not be so different, since we are evolved to see, even if we can’t actually see – but then it may depend on how early on in the early development process sight is prevented from developing).

            LikeLike

            Reply

            1. I agree that semantics requires context. But there are also contexts for the activity of the CPU, and for the gate changes on the motherboard.

               Computer systems are our creations. Their behaviour embodies our purpose. Our purpose provides the context, and the semantics, for their behaviour. It’s an interesting question whether a computer can understand itself as having a purpose, a goal towards which it is working. Could this be how a computer experiences its programming? Many would say that computers don’t experience anything at all. I don’t know where you stand on this question. To me it seems reasonable that if anyone experiences the programming context of a flip-flop toggle, it’s those who want to accomplish something with the change.

               We could say of the human brain that the changes at the neuronal level have a meaning at the neuronal level, appropriate to the context of neurons: a context of electrical signals, chemical receptors, strengthened synapses. Whatever significance their activity is seen to have, requires a context, which represents a purpose. Examining the neuronal changes and their patterns is like looking at the gates and registers of a motherboard. We should expect to find consciousness of the programming in the physics or its relationships, because these are built for another level of purpose. They express a different paradigm, and a different ontology.

               It would take many translations, through many levels, to understand consciousness through the observation of neurons. By the time we reached the end of these translations, we would be occupying a completely different paradigm, with a context appropriate to consciousness. But this context was there to be found all along; we simply diverted ourselves with other contexts first, and then a long,complex, and pointless chain of translation.

               You’re right that some philosophers, and the religious, hope for ‘something’ else when Mary experiences colour. They have a feeling that by searching for consciousness in brain activity, we are looking in the wrong place, and asking the wrong questions. The “Mary’s room” thought-experiment cannot help one way or the other, because it can be understood using more than one semantic aspect; that is, with respect to more than one context. In fact it confuses the contexts of brain activity and human experience, and the real dispute is over which context is appropriate. This is why turning over its details, as so many are doing here, is inconclusive. We need to stand back from the ins and outs of the experiment, and ask what assumptions it reflects about the role of context; indeed, about the possibility of alternative contexts.

               LikeLiked by 1 person

            2. Sorry, for “We should expect to find consciousness of the programming. . .” please read “We should not expect to find consciousness of the programming. . .”

               LikeLike

6. Mary is in a new state she hasn’t been in before, and that’s true of Mary described as a physical entity and described as a mind. What makes that state feel like something is that it comes along with associations of how good or bad the current state is for her, what mental and physical actions she can take from this state, and what she can pay attention to and discriminate in support of those actions….together with whether the next states those actions will lead to are more or less favourable for her.

   LikeLiked by 2 people

   Reply

   1. Agreed. The trick, which it looks like you’ve caught on to, is to consider what the causes and effects of the experience might be. Once we do that, the biggest mysteries seem to dissolve. At least for those of us who are comfortable with that move.

      LikeLike

      Reply

7. I think the Mary’s Room discussion is a conflation of (at least) two significant issues. One is the distinction between knowledge of types and knowledge of specific instances of the type. You could know everything there is to know about peas, but if I put a pea under one of three cups, you don’t know everything about that specific pea. You learn that missing info when you turn over the correct cup. Likewise, the knowledge Mary gains about “red” relates to a specific instance of a “red” recognition apparatus, specifically her own.

   The second issue relates to the perspective of knowledge, 1st person vs. 3rd person. A system can be able to recognize and distinguish patterns, but be unable to explain how it does this. Internally it can have pointers to specific recognition apparati, which it can use to determine if and when a pattern is recognized, without knowing anything about the patterns themselves besides whether they’re on or off. And some systems, like us, will be able to create a memory of recognition events and refer back to the specific apparati when necessary. So when Mary sees red, she gains knowledge of the existence of a pattern recognition apparatus and the capability of re-identifying the activation of that apparatus.

   *

   LikeLiked by 2 people

   Reply

   1. There’s a lot in your analysis I agree with. But I wonder, what if Mary’s knowledge includes her own brain? Maybe she’s been able to scan her brain while experiencing black and white sights, and is aware of just which pathways are being activated, and which aren’t, and knows what usually happens at both a physiological and psychological level when the inactive circuits are stimulated. How much would she be able to anticipate her own subjective state upon the first color stimulation?

      But it is true that her inability to put her own system into that state may be a blocker. One thing I discussed in our old discussion was the possibility that she could figure out a way to induce the state without colored light hitting her retina, such as maybe building a machine that could trigger it with a series of targeted TMS pulses, or maybe even implants. Most people thought that would be cheating, although nothing in Jackson’s original description precludes it.

      LikeLike

      Reply

8. I think, we should avoid using the term “knowledge” as applicable only to humans.

   For example, in constructor theory, knowledge is defined as information that can act as a constructor. Constructor theory is a theory of information that aims to express all scientific theories regarding possible and impossible physical transformations. It also seeks to unify the laws of information and computation with the laws of physics.

   If we do not associate knowledge only with humans (or just biological beings), then a whole problem should be re-thought.

   LikeLiked by 1 person

   Reply

   1. I’ve made multiple passes through constructor theory trying to understand it, but haven’t been able to grasp it. But “knowledge” does seem to have pragmatic utility, although possibly not to the extent it may be getting stretched in this argument.

      If I know the way to grocery store, it’s something, to your point, I can put to use. What does the “knowledge” of intrinsic redness enable? Apparently I can remember and recognize it, but not communicate it to anyone else, or do anything meaningful with it. At best, it seems like a different kind of knowledge.

      LikeLiked by 1 person

      Reply

      1. One example of knowledge in constructor theory is this. In the context of a car-building factory, constructor theory suggests that the recipe for building a car is an abstract constructor that can be considered knowledge. This knowledge is information that can act as a constructor and remain instantiated in physical substrates. I thnik that knowledgs has a practical utility.

         LikeLiked by 1 person

         Reply

         1. Thanks. That description on constructors helps!

            LikeLiked by 1 person

            Reply

9. Wonderful piece! You are getting better with age.

   This recalls an argument I remember that deaf people couldn’t have any language skills without instruction. Some inventive scientists found some deaf people who had been living on the streets their whole lives. They communicated with facial expressions and hand gestures mostly. They were then, if I remember rightly, taught American sign language, and low and behold they had sophisticated worldviews and opinions. They couldn’t read or hear English, but their minds were fully capable of communicating somehow.

   I think we know a lot and I think there is a lot more we need to know. I wonder if there would be so many challenges against scientism, determinism, and naturalism is there were a large cadre of people who desperately wanted to believe in magic.

   LikeLiked by 1 person

   Reply

   1. Thanks Steve!

      That’s interesting about the deaf people. It’s a counter to the stories (mostly unverified) about children being raised without language and being little more than animals, and unable to learn language in later life.

      Yeah, my experience with the word “scientism” is it too often it just means, “Get your science away from me!” I do think there are real cases of it, but I’d say the vast majority of people use it as a pejorative against science that doesn’t validate their views.

      LikeLike

      Reply

      1. Yeah, the story you relate about deaf people was the standard story people told before anyone seriously investigated it. I wish I had a reference for that study.

         And your take on scientism is exactly the same as mine, sort of a “you smarty-pants scientists don’t know everything” not knowing their isn’t a scientist alive who even things his entire field involved everything to be know in his field.

         Back when I was still in the classroom I used to draw a line across the board from one side of the room to the other. One end was labeled 0%, the other 100%. The lne represented the sum total of chemical knowledge. I put a check mark right next to zero and wrote ‘You are here!” next to it. I put another check mark right next to the first one, “This is me.” And the sun total of chemical knowledge has tripled since I started doing that.

         That anyone would thing that any scientist knew “everything” about any topic, qualifies them for “Total Idiot” status.

         LikeLiked by 1 person

         Reply

10. One more angle.

    The implicit assumption in this puzzle is that Mary knows and uses the same language as the surrounding society. If Mary does not know the language used in the surrounding culture, she could not label anything red or blue.

    Suppose Mary did not know the language used in the surrounding culture. In that case, the answer to the puzzle would depend on when somebody teaches her the proper language – before or after Mary perceived some light patterns.

    LikeLiked by 1 person

    Reply

    1. Right, but in that case, where is Mary’s knowledge about the science of color perception coming from? Is she maybe a prisoner in a foreign country, which is piping the info from her native country? If so, then certainly she’s not answering any questions about color.

       Although I still think her knowledge would give her insights into what to expect from her first exposure to color, particularly if its in a full context of actually being outside in the world.

       On the other hand, if Mary knew nothing in particular about color perception, then that first exposure would likely be everything people intuitively think it would. Well, assuming her nervous system doesn’t have to spend months learning how to discriminate colors.

       LikeLiked by 1 person

       Reply

11. This confirms that (in this discussion) human physical resemblence and perceptions and one person relations with society play a big role in the understanding the problem. If we ditch all that then the problem will look very different.

    LikeLiked by 1 person

    Reply

12. It occurs to me now that Mary’s efforts to understand the color red are similar to something I’ve been trying to work through in my Sci-Fi writing. If I’m writing a story set on the Moon or Mars, I would really like to describe the experience of walking around in low gravity. I’d like to do that in a way that feels authentic–viscerally real–to my reader. So I’ve read papers about it, and I’ve watched videos from the Apollo missions, and I’ve listened to astronauts describe the experience. I have an intellectual understanding of what low gravity is like and how the human body tries to deal with it, but I still don’t have any first hand experience to draw from in my writing.

    The real problem, in my opinion, is that language is an imprecise tool for communicating information. There’s only so much language can do to explain what low gravity feels like to someone who’s never walked in low gravity. There’s only so much language can do to explain what the color red looks like to someone who’s never seen the color red. If bats could talk, that would be amazing, but there would still be only so much language could do to allow bats to explain what being a bat is like.

    So I can agree that some things can only be known through experience, but I wouldn’t say that that kind of knowledge is non-physical. Just not explainable.

    LikeLiked by 1 person

    Reply

    1. I know what you mean. I’ve tried to get at the same thing in my research. What is it like at a personal level to be in space? One interesting thing many astronauts seem to agree on, is that space smells like burnt metal or gunpowder. So I imagine it’s something like the smell after firing a pistol or rifle, but suspect the reality has a lot of nuance that’s missing from that association.

       And of course there are things like the way your sinuses feel full in zero gravity, or the deadening effects on taste buds, or the Coriolis sickness people might feel in spinning habitats. All stuff classic sci-fi never had, and really couldn’t have foreseen.

       I think that last point is important to remember. No matter how much we research it, we won’t be able to capture the reality of living on Mars in a way that the first colonists won’t immediately make obsolete in their first reports. The good news is people still read Jules Verne and H.G. Wells, even though they were wrong about many things.

       LikeLiked by 1 person

       Reply

       1. Fair enough. If I’m lucky enough that future Mars colonists are reading my work, I just hope they know I tried my best. What really motivates me in this research is that I’ve read a few Sci-Fi novels where low gravity and microgravity are described really poorly. You read it and it’s way too obvious the author doesn’t know what they’re talking about. I just don’t want to fall into that same trap.

          The burning smell, I’ve read, has a lot to do with oxidation. If you’re smelling dust or gas particles from space, then you’re smelling something that also hasn’t been exposed to oxygen for billions of years. The redox reaction happens pretty rapidly. Which is another example of me feeling like Mary researching the color red. I’ve read about that smell, and I’ve read about the chemistry that probably causes it. But all that reading has only taught me so much, because language can’t communicate 100% of an experience like that.

          LikeLiked by 1 person

          Reply

          1. Interesting. So they’re not so much smelling space, but our environment’s interaction with it. That’s a detail I’ve have to remember. Thanks!

             LikeLiked by 1 person

             Reply

13. I think you nailed it, Mike. The only things I would add are additional ways Mary could name the colors even in superficially challenging cases. Take the example where someone shoves an image of the spectrum under the door, with no labels and no wavelength numbers. The human eye’s sensitivity is skewed toward the shorter wavelengths, as you move away from peak sensitivity at a slightly yellowish green: https://www.pixelsham.com/2016/03/10/sensitivity-of-human-eye/ So Mary would know this, and thus know that blue was on the left (in your diagram).

    Even if a single colored patch is shoved under the door, Mary could build a spectrometer out of black-and-white materials, and deduce its reflectivity profile.

    LikeLiked by 2 people

    Reply

    1. Thanks Paul! And good points.

       Although there might be a question of cheating with the spectrometer. But it may also highlight how much this can be seen as a matter of ability rather than propositional knowledge. So David Lewis and the others who proposed that answer might have a point. (I’m increasingly finding Lewis to be someone whose views I often end up coming around to, at least for the philosophy of mind.)

       LikeLike

       Reply

       1. For what it’s worth, I don’t think the distinction between abilities (to identify, discriminate differences, etc.) and propositional knowledge (“red looks like (calls up memory) *this*”) needs to be mutually exclusive.

          LikeLiked by 1 person

          Reply

14. Mary is a lousy scientist, or has somehow been prohibited from experimentation. Her revelation feels much less dramatic if you take knowledge to be what’s reliable. Everything she can access, in principle, in the colorless (not just red free, to set the bar as high as possible ) room encompasses every thing about red that can be ‘functionalized’. What is left is the ‘what it’s like’ aspect of the phenomenon, and that little leftover is epiphenomenal.
    Mary says,”Meh, kind of suspected it would be like that.”

    LikeLiked by 1 person

    Reply

    1. Right. I think when people imagine this scenario, they think about what it might be like for them to suddenly be let into a world of color, sort of like Dorothy emerging into the land of Oz (the film switches from black and white to color at that point). Mary probably would get a dopamine hit from having her theoretical knowledge confirmed, but she’d be unlikely to be overwhelmed in the way we imagine we might be. It’s very hard for us to put ourselves in her shoes with everything she’d know.

       LikeLiked by 1 person

       Reply

       1. Although the Mary-scenario is more romantic, I think the better thought experiment is a version of the inverted spectrum. I’ll admit to a little bit of personal affront when these arguments are presented as negative cases for some sort of dualism.

          As I think you’ve noted before, the language surrounding basic issues is extremely slippery. A lot of it implies the notion of mind as distinct from everything else – good, hearty, if maybe by this time slightly overcooked Cartesian dualism versus ideas about knowledge, truth, and causation that are frozen in their development around the time of Democritus. For those who strategically employ the language, I’m sure it seems like things are just more fair that way. I don’t dispute that it makes their job easier. It is much easier, as well as more satisfying to imagine Mary’s life-changing revelation, backed by a crescendo of stringed instruments, in opposition to a mass of clacking billiard balls that somehow become conscious when they reach a threshold number.

          Those folks have ‘splaining to do in my opinion. I’m still looking for a reasonable account of what constitutes a separate category of properties, or a separate substance. I feel like I’m holding the princesses place in line on that one, so little has been offered over time. This is the only reason that I have any interest in what the panpsychistis have to say.. They at least give it a go and have had to face up to the inherent difficulties.

          The more detailed account of qualitative experience would be nice as well. The “what it’s like” formulation is a little sparse. When asked to ante up on this, the usual reply amounts to, paraphrasing Simon Blackburn, an offered to rebut an incredulous stare. It’s intuitively intuitive you lout. As far as I can tell, the otherwise unaccountable portion, and a very small portion it is, of qualitative experience is an element of that particular experience’s aspectual shape. What does it mean if that last little dangler is epiphenomenal?

          LikeLiked by 1 person

          Reply

          1. There is actually another little known thought experiment that Jackson discusses in the same paper right before the Mary one, the case of Fred, who can see a color the rest of us can’t. He discusses how we figure out an operation so we can see the new color, and argues that afterward we learn something new about Fred we didn’t know, even if we knew every physical detail. Mary gets all the attention, but Fred’s case is just as interesting. Although all the same answers work for both.

             I’m actually liking the idea that “what it’s like” refers to an ability, or abilities, rather than knowledge of facts. It’s the one answer Chalmers couldn’t respond to, accept to assert it’s implausible. (Not that Chalmers’ ability to respond is my standard for such things.)

             I do sometimes wonder if acausal qualia are worth arguing over. If someone wants to believe in them, is there harm? My concern though is believing in them tends to affect how someone assesses scientific theories. You’re a lot more likely to buy into exotic solutions than if you think capabilities are all that needs explaining.

             LikeLike

             Reply

15. How is ‘knowledge’ possible if physicalism is true? What exactly is meant by ‘knowing’? We assume there is something called ‘mind’ enabling humans to generate objective knowledge about the universe. But there isn’t any mind-stuff inside the brain. What is the mechanism linking physical processes in the brain to generation of ‘objective knowledge’?

    LikeLiked by 1 person

    Reply

    1. I think of knowledge as accurate belief, and belief as information in a mind, a representation or cluster of predictions.

       Opinions vary, but I’m a functionalist. I think it’s a mistake to look for mind-stuff. I see the mind as a process, or maybe more accurately, a collection of ongoing processes carried out by the brain. We’ll never find a mind in a static scan of a brain, just as we won’t find software in such a scan of a computer. In both cases, we have to look at what the system does. The mind is as the mind does.

       I would say objective knowledge requires the collaboration of many different perspectives, which collectively are able to overcome the blind spots of the individuals. (Overcoming the blind spots of an entire culture or our whole species is more difficult.)

       LikeLiked by 1 person

       Reply

16. I think you are dismissing the argument far too easily. Your modification of the premises misses the point of the thought experiment. Mary has acquired all the physical facts that there are to know about colour sensations., but she does not have all the knowledge concerning colour vision, since she does not know how a person with normal vision experiences colours. On her release, when she experiences red for the first time, she learns a new fact of which she was previously unaware: she learns what it is like to see red.
    There is no contradiction in the story and the premises do not presuppose the conclusion.
    The issue is whether what Mary does not know are physical facts, or whether she lacks some kind of knowledge about physical facts. In the first case the conclusion is that there are non physical facts, in the second case that there is non physical knowledge.

    LikeLiked by 1 person

    Reply

    1. I’m open to the possibility I’m overlooking something, but your remarks make me wonder if you caught my point. 

       The assumption is that experience, and any new information Mary might get via the experience, is non-physical.  The contradiction only arises if we remove that assumption, because then we’re saying she has all the information, yet learns new information with the experience.  The other side of the assumption is that Mary is able to have all the physical information without the experience.  If she can’t, then learning new information says nothing about physicalism.

       If you see a flaw in that reasoning, I’d be grateful to know it.

       Of course, as I cover later in the post, pragmatics aside, I’m skeptical she gets new actual information from the experience.  All I see gained are opportunities to exercise functionality.

       LikeLike

       Reply

       1. Admittedly, your modifications of the argument lead to contradictions, but they miss the point and do not demonstrate that the original train of thought is flawed. The experiment assumes that Mary has acquired all the physical facts about colour perception. To assume instead that she has all the information about colour perception before she experiences colour, is just to deny the point of the experiment from the outset. So there is no objection to accepting the first premise as an appropriate description of a reasonable thought experiment.
          The second premise of the train of thought states that Mary has a new experience when she leaves her room, such as what seeing red feels like, which she was previously unable to deduce. Neither premise anticipates the conclusion.

          There are really just two strategies for the physicalist. Some argue that Mary gains no new knowledge, but is merely gaining abilities (like the ability to imagine, remember and recognize colors or color experiences).
          Others acknowledge that Mary acquires new factual knowledge when she sees red for the first time, but deny that it concerns any new facts, arguing that it only concerns the same physical facts that she already knew.

          LikeLike

          Reply

          1. The only reason to make modifications is to remove the assumption of non-physical experience. If we do that without modifying anything else, we get the contradiction. We can avoid it by modifying the statement about Mary acquiring all the information to Mary acquiring all the third person information, but then her acquiring any new information in the experience has no metaphysical consequences.

             I actually see the new-ability and acquiring-old-knowledge-a-new-way as more or less the same take. I think they’re right, but I don’t see them as necessary for the physicalism point, only to the one about whether there are facts that can only be acquired through direct experience.

             The thought experiment brings out our intuitions in interesting ways, and leads to fascinating discussions like this one, but like all thought experiments, it doesn’t tell us anything about reality.

             LikeLike

             Reply

17. Bravo Mike! Very well argued! I was reluctant to weight in. My comments essentially would be a regurgitation of comments in two previous posts of yours, “Illusion and Functionalism” and “Illusionism and types of physicalism.” In the end, I could not help myself. You say “Mary has ‘all’ the information on color perception prior to experiencing color. [And then] when Mary experiences color for the first time, she learns ‘new’ information. [Therefore] the premises [challenging physicalism are] contradictory.”

    I’m not so sure you have uncovered a true contradiction. It may only be a fallacy of ambiguity, i.e., when a word or statement has more than one meaning. Off the top of my head “information” can mean facts, data, signs, pictures, or a communication or reception thereof, or a quantitative or qualitative expression by some medium of any of these concepts. I think I would doubt that I wasn’t getting into a fallacy of ambiguity with the multiple meanings of “information” in order to construct a contradiction. But that is not my main concern.

    In the end you conclude; “What I think we can say is that Mary’s nervous system is allowed to discriminate colors for the first time. Her neural circuitry is exercised in a new way. And she is subsequently able to remember the discrimination.” I totally agree. Beyond that I find it difficult to pinpoint exactly where you fall in this debate. You have a nuanced view of functionalism that is a bit difficult for me to grasp.

    But, nevertheless, for one to say that Mary can discriminate colors for the first time fails to persuade me that Mary has no new information or has not learned something new. I think such a claim is a distinction without a real difference. The same goes for the argument that it’s not any new knowledge or information, just a new so-called ability. Some philosophers seem to be playing games with definitions here. As I have said before in my comments to your previous postings, I fail to see the necessity of resolving our understanding of reality as having to be reduced to a foundational property. In other words, I have no problem with a greater evolving complexity producing a plurality of irreducible entities among all the parts of our universe—no problem with the complexity of a brain that causes a mind. Nor with a plurality of irreducible parts of reality that are understood only by explaining and understanding their complex systems as a whole.

    Getting back to basic starting assumptions, I don’t think one has to accept the assumption that the physical and the mental are concepts for the only two possible categories of reality which also happen to be mutually exclusive. That is, nothing requires us to find a way to accept one or the other or some form of dualism. There is nothing that tells me we are locked into that sort of a conundrum. In fact, I would submit that adopting physicalism as one’s ontological scheme is perhaps giving in to a poor solution to this old Cartesian-Lockean conundrum. At the end of the day it’s essentially claiming that mental experience does not exist. It seems to me that physicalism ignores in large part our first-person ontology. I am convinced that my subjective experiences are real phenomena in a real world which can be understood as causally the product of an evolving physical complexity. But it simply cannot be explained or reduced to those terms. Try as we might I haven’t as yet seen a persuasive argument that one can explain this subjective phenomena from an objective perspective. I accept the basic physicalist notion that we live in one reality. And I totally agree with you when you claim the mind is a “process.” But I also accept the presence of irreducible properties at various levels of complexity. That is a more complex understanding of reality, consistent, I submit, with a broad interpretation of evolution. Again, it all goes back to that Cartesian-Lockean starting assumption. It needs to be challenged.

    LikeLiked by 1 person

    Reply

    1. Thanks Matti!  No worries on repeating points you’ve made recently.  Part of what we do here is rehearse our arguments with each other, always testing them, seeing if others can develop new counters to them.  As long as we keep it friendly and stay patient with each other’s different views, I think it’s all in good fun.  

       Just to be clear, I’m not saying there’s a contradiction in the thought experiment, but that it assumes its conclusion.  The contradiction only arises if we try to remove the assumption without altering other parts of the scenario.  The assumption is that experience and any new info gained from it is non-physical.  We can remove that assumption, but to avoid contradiction, we can no longer assert that Mary has all the physical information.  We could assert, as I noted in the post, that she has all the third person information, but then any new info she might gain has no necessary implications for physicalism.

       On pinpointing where I fall in the debate, I’m an analytic functionalist, an a priori physicalist.  I didn’t press hard on that in the post because I didn’t think the arguments required it. But I think we can have a full logical understanding of how conscious experience relates to physical processes.  If Mary has that understanding, then, in principle, I don’t see her learning any new information.  (In practice is a different matter.)  I think what she gains are opportunities to use certain functionality in her nervous system, and it’s easy to misconceptualize that as new information.  But as I noted in the post, it would be information no one can describe or do anything with, so not information as we normally understand it.

       On playing games with definitions, I’m increasingly aware that’s what philosophers do.  I think it’s fine as long as they’re clear about it.  (The ones who play those games and refuse to clarify are, I think, the problem.)  And since I think a big part of the problem we perceive about consciousness comes down to definitions, it seems like it can be a productive endeavor.  

       We agree that the Cartesian-Lockian paradigm is problematic.  But I come at it from the other side.  I’m a mechanist all the way down.  I think the mistake they made was in blanching at that, and attempting to exempt the mind from it.  In that sense, I think Hobbes was closer to the truth than they were.  As we discussed before, I’m fine with emergence as a concept where it’s productive for us to switch models at different scales.  No one seriously thinks we can use quantum mechanics for economic forecasts.  But I still want to understand the logical relationships between those scales.  Again, it’s a difference in what we count as success, although I suspect we’d agree on intermediate successes.

       “It seems to me that physicalism ignores in large part our first-person ontology.”

       What part of that ontology do you see physicalists ignoring?  My take is that there’s nothing in our actual experience which can’t be accounted for functionally.  It’s only the assumptions we bring in that have to be questioned, and even then, in most cases, it amounts to a more nuanced understanding of what’s being assumed.  So experience can be very difficult to describe, and practically private, without being absolutely ineffable and fundamentally private.  What do we lose with those minor adjustments?

       LikeLiked by 1 person

       Reply

       1. Mike, as you know, I’m a bit new to this area of philosophical inquiry. I agree, part of the purpose here (certainly for me) is to rehearse our arguments and give them a good test run. And I’m trying to understand your position. So I’ll make a few more stabs at it. At times I think your functionalist position is nuanced which demands serious pondering and then you go and claim you’re a “mechanist all the way down” and pay respects to—of all philosophers—Thomas Hobbes. OMG! What should be clear is that functionalism is difficult for me to swallow—Hobbes’ materialism worst of all. But I remain open minded. I flirted with Hilary Putnam’s original thesis of functionality—from which he retreated as have I. And I think that is understandable since Putnam was a mathematician and computer scientist as well as a philosopher. The computer is a powerfully attractive model for the mind. But if your understanding of functionality is represented by Thomas Hobbes then you’ve befuddled me. I certainly hope you have not embraced Hobbes’ social contract political theories as well.

          I understand that one of your arguments against Mary gaining new knowledge when released from her room is that her conscious experience amounts to gaining opportunities to use certain functionalities in her nervous system. Again, this sounds like gaining new abilities. And again I say it’s a distinction without a difference. You claim it would be “information no one can describe or do anything with so not information as we normally understand it.” Not information as you understand it, that is. If one assumes a functionalist account of the brain. I.e., that it is a kind of behavioristic machine. If one assumes a functionalist account of mental states, then mental states have to be understood functionally. Thus, mental states or conscious experience that cannot be described functionally do not exist? Are we not assuming or presupposing the conclusion in the argument? Or, maybe I misread you.

          LikeLiked by 1 person

          Reply

          1. Relax Matti. I just said Hobbes was closer to the truth than Descartes and Locke, and only in the context of whether the mind is separate from the rest of world. But “closer” doesn’t mean he’s got the right answers for today. And I certainly don’t buy into all of his philosophies. For the materialism aspect in particular, there’s just too much learned since his time, too many paradigm shifts.

             But I am a mechanist, in the sense that I think reality is ultimately deterministic. We may not always have access to that determinism, but it’s what I expect a full understanding to entail. Of course, as a pragmatic matter, we may have to accept provisional theories that aren’t fully deterministic as a temporary intermediate step to make progress. But I expect a deterministic solution to be the eventual answer. Maybe I’ll be wrong.

             I’m not sure what Putnam’s stance was on functionalism later in life. He seemed to equivocate on his blog. I am a computationalist, but not really a Turing Machine computationalist. (Although it’s not clear that even the early computationalists were themselves, or they were using a liberal nuanced version of “Turing Machine”.) I think the connectionist paradigm is closer, but my commitment to any abstract model is very provisional, pending what we learn from computational neuroscience.

             But there’s also another tradition of functionalism going through David Lewis and D. M. Armstrong. It’s compatible with broad computationalism but its emphasis is on causal roles. You can think of it as an outgrowth of logical behaviorism, but recognizing internal mental states as a necessary concept.

             Right, the danger of taking experience to be functional is I may omit some aspect of it. That’s a danger with any conception of experience that doesn’t take it to be fundamental. I try to be alert to the danger. Which is why I often ask people for examples of what they think is being left out. Most of the clear examples I get seem like functionality to me. Can you think of another precaution I should be taking?

             As I noted in the previous post, I do think experience evolved to enable more sophisticated behavior. Natural selection can only operate on behavior and other relational properties (like energy usage). It’s why I think the idea of epiphenomenal experience is just broken. Again, evidence or solid logic can change my mind.

             LikeLiked by 1 person

             Reply

             1. Thanks Mike. I appreciate the clarity of where you stand. And thanks for clarifying your reference to Hobbes as merely for his adherence to a thoroughgoing materialism.

                I’ve been sitting back trying to conjure up a way to better to highlight the point I have been trying to make—a point I’ve been trying to make now in three of your recent essays. I think I’ve beaten to death the argument that much of the heat (and confusion) generated by this issue is the result of accepting the old Cartesian-Lockean assumption—an assumption that should be challenged.

                That ontological problem is accepted as expressed in the very statement of the issue—the mind-body problem. Why is that the proper statement of the issue? Does that bias our thinking? I think it does. We start there and immediately start stumbling around. We are locked into trying to find a way to accept one or the other as the foundational property in our ontology. Although we have made some progress. It does seem that most folks can’t swallow some form of dualism and few it seems contend for some form of idealism. But, as I’ve said, there is nothing that tells me we are locked into that sort of a duality conundrum. One does not have to accept the assumption that the physical and the mental are concepts for two mutually exclusive categories of reality.

                I accept that we live in one physical world. But I don’t accept that is the end of the story. One can, in the alternative, accept the presence of irreducible properties at various levels of complexity—an ontologically plural view of reality. In fact, I would contend that that mind and body are in no way a complete inventory of our reality. I can’t make the full argument here but so much of our reality is understood only by explanation of complex systems as a whole and cannot be explained or understood by reduction. Some things just don’t lend themselves to that sort of explanation. The new science of complexity adds persuasive value of this plural view of ontology. Moreover, this view seems to be consistent with our present understanding of evolution. This recent blog entry of yours states the problem as a knowledge problem—a problem in epistemology. And many folks approach it that way. What if, as I’m trying to say, we look at it as an ontological problem. No matter how complete, a third-person description of reality leaves out a lot. And what it leaves out many folks refuse to accept as not real. That’s what the Mary Problem highlights. In other words first-person experience is one part of our (complex) reality that cannot be reduced to a third-person physical ontology. I’m OK with that. I don’t see that as a problem. And at this point I think I should stop plowing that field. I apologize for my continual verbosity.

                LikeLiked by 1 person

             2. Thanks Matti.

                The other day I listened to a podcast interview of someone who said that we are all stuck in the shadow of Descartes. He felt even Daniel Dennett, of all people, was trapped in it. Although I didn’t understand what he was trying to get at. I get a similar feel from you, that even the most hardened materialist is somehow bound in this mind-body conundrum. If that is what you think, maybe elaborating on it might help.

                From my point of view, I think the notion that there are aspects of the first person viewpoint that are fundamentally inaccessible from third person observation actually is a remnant Cartesian intuition. Although it’s probably not fair to put all of that on Descartes, since he was channeling older intuitions that seem to exist in almost every human society. Just about every culture has a ghost concept, an essence of a person separable from their body. Even many hard nosed materialists show remaining hang-ups from it, while insisting they’ve utterly purged it from their thinking. Although I tend to think Dennett, Blackmore, and a few others are actually able to shake it.

                On being at peace with irreducible concepts, I guess the question I have is, are you okay with others trying to do the reductions anyway? Or do you see the attempt as more than merely unproductive?

                LikeLiked by 1 person

             3. Mike, I think that podcast may have been saying some of things I’ve tried to point out. And yes, I submit that even the most hardened materialists are bound in this mind-body conundrum—as I’ve tried to explain. I had a favorite philosophy professor who always used to say if you are having difficulty answering the question then maybe you are asking the wrong question or you are asking it in the wrong way. We all learned to tackle the so-called mind-body problem (from phil 101 and on) with certain assumptions (which I have repeated ad nauseam) that we can trace all the way back to Descartes and Locke. Maybe the way the problem is stated is the problem. I admit that looking at the so-called mind-body issue this way is currently radical. I should admit that I have been reluctant to use the word “emergence.” But, I do think that the British Emergentists of the early 20th Century were onto something. Recently some contemporary philosophers have revived that way of thinking. I refer to “The Re-emergence of Emergence” by Philip Clayton and Paul Davies and a book by William Wimsatt (a Philosopher of Science at Chicago) called “Re-engineering Philosophy for limited Beings”. I think I mentioned Wimsatt once before. I think I’ll leave it there.

                Am I ok with others trying to do reductions? Absolutely! That’s a legitimate method of inquiry and many times provides the best explanation of phenomena. There is no reason to give that up? Science and the science of complexity tell us that sometimes we can reduce complex features of our reality to its simple components. But complex systems sometimes demonstrate a complex hierarchy can only be explained in terms of the whole—a whole which cannot be simplified and explained in terms of its components because the whole exhibits unique properties. Only if you adopt the Cartesian-Lockean background assumptions as immutable facts about reality does that seem unthinkable. No one has convinced me that such a notion is required. I simply submit that we may be able to cut the Gordian knot of the so-called mind-body conundrum by taking a step back from those old background assumptions.

                LikeLiked by 1 person

             4. Matti,
                I found the podcast I was thinking of. The video is long, but the relevant parts are early, between the 1.5 and 11 minute marks. Listening to that portion again, it’s interesting, but I don’t think he understands Dennett’s view. That said, his confusion is typical, and I’m getting tired of defending the illusionists. They need to fix their pitch.

                I’m onboard with questioning our assumptions, although it leaves me to different places. But all ground we’ve covered.

                That book looks interesting. It’s a bit pricey (academic pricing), which discourages me from picking it up just out of curiosity. I see Chalmers has a chapter on strong vs weak emergence, and has a copy on his site. Lightly poking through it, it looks typical Chalmers. Strong emergence for consciousness, but weak everywhere else.

                Similar to your stance, I’m fine if people want to hold to strong emergence, as long as they don’t stand in the way of reductionist efforts. Even if the reduction is impossible, it strikes me as a productive endeavor. Persistent failure might be as instructive as success.

                LikeLiked by 1 person

             5. Mike, Thank you for hunting down that video! I’ve watch the first 30 minutes so far. I really like this Terrence Deacon. He seems to be in my camp! He also seems to get Aristotle which is refreshing. I trace my own intellectual lineage from Aristotle and tend to see reality in terms of Aristotle’s four causes; matter, form, efficient cause, and teleological cause—vastly different from Descartes. I also think he’s more or less saying the same things I’m trying to say. I think he calls what I label the Cartesian-Lockean conundrum Descartes’ box. Yeah, right! Thanks again for pointing me in his direction. I have someone new to follow!

                LikeLiked by 1 person

             6. Mike, again thank you for pointing out the work of Terrence Deacon of UC Berkeley. I should note something I missed earlier. Deacon has an excellent essay in Clayton and Davies’ book “The Re-emergence of Emergence.” I’m quite excited about Deacons’ book, “Incomplete Nature, How Mind emerged from matter.” The more I dig into it the more I find that he has a clearer way of expressing some of what William Winsatt of U of Chicago says.

                Just a very short note to make my point. You note that you are a “functionalist, an a priori physicalist.” As Terrence Deacon explains, it is difficult to fit functionality (among other things) into a physicalist world as we now understand it. As Deacon puts it, “The function of a shovel isn’t the shovel and isn’t a hole in the ground, but rather the potential it affords for making holes easier to create.” As Deacon further explains, “it is difficult to ascribe energy, materiality or even physical extension to [functionality].” So saying you are a functionalist and a physicalist still leaves out something important as I’m quite sure you sense and as it appears Deacon succinctly explains. I’ve only just dipped into Deacon’s book and I wanted to thank you again for pointing him out. I read his essay sometime ago in the “Re-emergence” book but didn’t follow up. I’m sure his work will inspire me to converse, hopefully more intelligently, with you. Have a good day!

                LikeLiked by 1 person

             7. Thanks Matti. Glad you’re finding him interesting.

                I have to admit I don’t understand his concern about functionalism. Certainly functionalism isn’t necessarily physicalist. You could in principle be an interactionist dualist and be a functionalist. But its focus on causal roles seems to fit most easily with a physicalist framework.

                Maybe it comes from being immersed in this stuff for several years, but I actually don’t sense that anything is being left out of the functionalist account (aside from cases where we don’t yet understand all the causes and effects). I do understand that many people have an intense intuition there is. Which is why I often try to open things up so someone can make the argument, just in case I’m badly missing something.

                LikeLiked by 1 person

             8. Terrence Deacon’s concern is not only with functionalism. His argument is much broader—encompassing function, meaning and value as well. I find Deacon’s analysis intriguing. If I understand him correctly. And it’s different from other emergent theories. I think he is saying that to give a functionalist account of conscious experience (within a physicalist framework) is to rely on a phenomenon which has no material, i.e., physical, presence. He’s saying that something is missing in such an account. I think he may have a point. And I suspect this argument will lead us back to that Cartesian-Lockean background assumption as the source of the problem.

                LikeLiked by 1 person

             9. Ah, ok, thanks. Now I think I see where he’s coming from. He’s concerned functionalists are being unwitting dualists.

                But I think a physicalist has to be careful not to hold theories about mental states to a different standard than other physical accounts. So it seems like the person concerned about functional accounts involving anything non-physical, to be consistent, has to also be worried that our understanding of Microsoft Windows, car transmissions, mouse traps, or any other causal mechanism, is involving something non-physical. If not, then I think there needs to be explicit logic for why we should be concerned in one case but not the others.

                We do have to be on guard against falling into dualist thinking, something we’re all innately prone to do. But I don’t think functionalism, in and of itself, falls into that trap. I actually think the reason functionalism strikes so many people as wrong comes from remnant Cartesian intuitions.

                LikeLike

18. What you’re describing at the end is merely knowledge about the structural properties of qualia, meaning which colors are more similar to each other, which are less so, and the like. Indeed, with such a vast amount of third-person knowledge, Mary would have significant knowledge about the relationships between qualia, but she still wouldn’t know what the colors look like. This is knowledge that can only be gained through experience, and to someone blind from birth, you could at best explain how certain colors relate to others.

    LikeLike

    Reply

    1. I’d say that the structures and relations of color are what they look like. It’s hard for us to hold all the structures and relations we know about them in mind at the same time, while it’s easy to just have the experience. And of course we don’t know all the structures and relations yet.

       A lot of problems in the philosophy of mind historically have just come down to people not being able to imagine how there could be a physical explanation. For early modern philosophers, that was the intellect, memory, reasoning ,etc. It’s easier to see how that can happen physically today, so people have moved on to feelings and sensations, but to me it’s just the same issue, a difficulty in being able to imagine the solution. But you get a lot of benefit from just making the effort.

       LikeLike

       Reply

       1. Mary can provide exact data for colors, for example, for red: [255, 0, 0], she may know which colors are lighter than others, and know that orange is a mix of yellow and red, but this is still just formal knowledge that doesn’t tell us what these qualia are. In other words, Mary can only know the formal properties of qualia expressible mathematically; she won’t even understand what it means for colors to be similar, she will only know which colors are called similar and which ones people react to in a similar way.

          LikeLike

          Reply

          1. Mary knows more than that. She understands how the blue / yellow and red / green ganglion opponent cells receive signals from the cone cells in their receptive fields, their portion of the overall visual field, in tandem with all the other ganglion cells which care about various things (like motion and other changes), and how these get sent up to the LGN in the thalamus and V1 in the occipital lobe. She understands how the processing takes place in the various layers, V1, V4, etc., and the automatic reactions that are triggered, as well as the process for triggering learned associations (and she knows she wouldn’t have any of that last category in her first experience).

             In fact, she would understand how her first experience would not be like ours as lifelong color perceivers. There would probably be a period where her nervous system has to be trained to be able to make many of the discriminations we make without thought. She’d understand the process far better than we do today.

             To understand how Mary could know what to expect, we have to be willing to learn about color perception as the profoundly complex biological and adaptive process it is. It only seems simple to us because we don’t have introspective access to all the underlying processing, since it was never adaptive for our ancestors during evolution to have it.

             LikeLike

             Reply

             1. Indeed, Mary would have an impressive knowledge of how the brain distinguishes colors and how it triggers various behaviors related to them. However, she still wouldn’t know what colors look like, because this is something ineffable, accessible only to someone who has color-related qualia. You might be falling for a certain illusion. You’re looking at what consciousness is from the outside and it seems to you that it’s simply some brain processes that cause certain behaviors. But pay attention to what you can learn only by being yourself. This is phenomenal consciousness, which consists of qualia. Moreover, it’s this subjective aspect that makes it matter how we are treated. Why should we care about suffering if it were only a state that causes someone to scream and shed tears, and not about how it feels from the inside?

                LikeLiked by 1 person

             2. Of course, the illusionists will say that it’s the intrinsic, ineffable, fundamentally private qualities that are the illusion.

                I prefer to say that what you’re thinking of is functionality, functionality that includes the ability to make discriminations, and to remember, imagine, and recognize those discriminations at a later time. There’s no information there beyond what can be obtained with other means. But it’s easy to misconceptualize those abilities as information. To see the misconceptualization, try to think about what you could ever do with that “information”. We’ve already established that it can’t be communicated. And non-physicalists argue that it makes no difference to behavior. In what sense is it informative?

                LikeLike

             3. It seems to me that you’re still missing something very important, which is what it’s like to be conscious from the subjective side—the “movie” that no one else “sees,” only you.

                Why is it so important that you are not subjected to pain? If there were a robot programmed to shout “ouch” in response to a certain state, would it matter to avoid triggering that state in the robot? Or would it matter if the robot were programmed in such a way that it sought to avoid that state? I don’t think so either, unless avoiding that state in the robot brought some benefit to us. It’s not about having some functional state that triggers avoidance reactions. What matters is that pain has a certain unpleasant quality that no one can observe from the outside, but only your behaviors at most.

                I understand that the concept of “information” presents some difficulty, but there is an intuitive sense in which only the experience of pain can give us knowledge of what the quale of pain is. Perhaps the knowledge of pain differs from the pain itself only in that the being experiencing the pain is in the appropriate functional state; however, the quale of pain is a necessary condition for this knowledge.

                That said, I do share anti-epiphenomenalist intuitions. On one hand, it seems to me that consciousness cannot be reduced to how the brain works because functioning can, at most, explain behavior. On the other hand, I think pain does cause us to reach for painkillers. However, I agree that explaining our behaviors is the easy problem compared to explaining qualia (phenomenal consciousness).

                How do I reconcile this? Well, I believe that qualia do cause some of our behaviors, but if our only task were to explain our behaviors, we would have no reason to accept qualia. We would simply believe that one day neurobiology will explain our behavior. However, by being myself, I notice that I have qualia, and a neurobiological explanation alone doesn’t satisfy me.

                LikeLike

             4. I actually see pain as functionality, much more sophisticated functionality than the simple robot examples you’re describing. In a recent post, I discussed the functionality in terms of when we might regard a machine as sentient:

                But what exactly is a feeling in this sense? I think it’s a draft evaluation of a situation, a learned or innate automatic reaction, which prepares the system for certain actions or inaction, a priming which might involve ramping up internal mechanisms to higher levels of alertness and readiness (consuming energy), or dampening it down, all of which reverberate back as interoceptive signals. Which is why the same word in English is typically used for both body perceptions and the visceral experience of an emotion.

                Animals experience the feeling to learn from it, and because they may need to override it based on knowledge and foresight (prediction), although doing so takes additional energy. But while they can override or ignore it, they can’t shut it off. The automatic reaction continues to consume energy, and require more energy to override or ignore, until habituation over time causes it to fade, or (sometimes) if it’s released in some display of frustration.

                What I just described is a pretty specific architecture. It exists in animals due to evolutionary history, where predator / prey relations set up an adversarial dynamic where whoever could move faster usually won, and was more likely to pass on their genes.

                AI intelligence, consciousness, and sentience

                The problem with the “being myself” strategy, is that’s just introspection, self reflection, which evolved for a number of adaptive reasons, none of which involve giving us an accurate view of how our mind works. For a lot of what you’re talking about, it only has access to things at a certain level, not to much of the underlying functionality. It makes it easy to assume that what we do have access to is somehow simple, primal, and disconnected from all that underlying complexity, when the reality is we just don’t have access to the underpinnings.

                LikeLike

19. Roy G. Biv – what happened to indigo?

    Doesn’t all knowledge require experience? The physical facts she knows are facts in experience. She learned them by observation.

    LikeLiked by 1 person

    Reply

    1. Indigo and Teal ran away to get married. They named their first baby Chartreuse. Report are they are doing well.

       LikeLike

       Reply

    2. The spectrum is continuous, so you can always squeeze in additional shades and give them a name.

       Right, her knowledge comes from experience, but not direct experience of color, only from third person observations, the descriptions of other people’s experience, etc.

       LikeLiked by 1 person

       Reply

       1. Her observations are not experience?

          My point is her knowledge of the physical facts was acquired through experience like the knowledge she will acquire when seeing color. The two are on par. Both can be described with words that have meaning.

          LikeLike

          Reply

          1. Strike that first sentence above.

             Her knowledge of physical facts is just as ineffable (or not) as her experience of color. Nobody has access to her knowledge in either case exactly as only she has it.

             LikeLiked by 1 person

             Reply

             1. I’d say aspects of her knowledge are practically ineffable and practically private, but neither in any absolute sense. Someone with all the physical facts of her knowledge would be in just as good a position as she is with all the physical facts of color perception.

                LikeLiked by 1 person

             2. We describe our knowledge with words, but we learn the meanings of words by associations with other words. So my “meaning” of any word might not match your “meaning,” because we “color” the associations slightly differently, even if they may be similar enough to communicate. That is probably the same as with color.

                BTW, has it occurred to anyone that Mary may not even see color when exposed to it? She might simply see in shades of gray until her visual cortex has learned to see “color.” I think the squirrel monkeys who had some of their green cones modified to see red couldn’t actually see the red for about a month after the modification. People blind from birth have difficulty learning to see if they have their sight restored after an early age.

                LikeLiked by 1 person

             3. I think you’re right about the variances. We all have different ratios of S, M, and L cones in our retina, which means different concentrations of B/Y and R/G opponent cells. We’re not talking inverted spectrum territory here, but our ability to distinguish various shades is affected. Even worse, it declines as we age., so we don’t discriminate colors as well as we did as a teenager.

                In the literature, I think Paul Churchland noted that Mary probably wouldn’t be able to perceive colors as we do. This is particularly true if we imagine something more plausible, like her having a kind of achromatopsia that is corrected by a medical procedure, sort of like the genetic therapy for those monkeys. In that case, it might be months before her nervous system could distinguish colors, and she likely would never be able to do it as well as someone who grew up seeing color. So it likely wouldn’t be Dorothy opening her door and walking into the land of Oz.

                LikeLiked by 1 person

20. Great post! Sorry it has taken me so long to respond. I have a great deal to say about the issues this thought experiment brings up, and hopefully I’ll find time to blog about it.

    “But remember, this is all based on a little bit of what we know today. Mary’s knowledge is supposed to be far more complete. In our revised understanding of the premises, she has all the third person information that will ever be obtainable.”

    I’ve never understood Dennett on this point (assuming he’s the first to introduce this?) To say Mary must have complete scientific knowledge from some infinite point in the future requires Mary to have god-like intellectual capacity, but that doesn’t seem fair, especially since it’s generally understood, even by scientists, that scientific knowledge is never complete. This requirement seems to me overly strict and I don’t think it was intended from the quote. Unless, of course, this requirement was introduced in some academic paper later on?

    I haven’t heard anyone bring up primary and secondary qualities, which is strange, given that this thought experiment seems as though it had been constructed with that distinction in mind. I think if we take that distinction into consideration and consider its implications for the scope of science, I think Mary’s room makes a far stronger case than p-zombies.

    If we take colors as such to be secondary qualities, then they simply are subjective. From what I understand, this is the scientific view of the matter: colors are NOT objects or in objects. So if we take the scientific view, colors as such are not objective, not mind-independent—which means colors as such don’t fall in the purview of scientific investigation. I don’t think the stipulation that science will someday at some hypothetically perfect completion in the future makes sense. To grant that stipulation I would want to hear the argument for how science might come to understand colors as such as mind-independent—I see no scientific color revolution on the horizon. Keep in mind I’m not saying it’s impossible for science to change in that direction, but I doubt Dennett would’ve loved the idea.

    To lay my cards out on the table, I don’t love the primary-secondary qualities distinction and I was surprised to find Nagel taking it for granted in The View from Nowhere. I won’t get into why I take issue with it here. That would be a very long discussion. Hopefully I’ll write about it someday.

    I’m just trying to wrap my mind around the arguments coming from the other side. Mary’s room takes the distinction for granted, and I see this as a powerful stance (even though I don’t agree with it) precisely because those who oppose the idea that experiential knowledge lies outside the purview of science would also take the distinction for granted. If you want to argue that experienced color can be known objectively, you have to deny what science currently takes color to be, which is subjective!

    By the way, here’s a quote from an actual color-blind scientist:

    “Although I have acquired a thorough theoretical knowledge of the physics of colors and the physiology of the color receptor mechanisms, nothing of this can help me to understand the true nature of colors. From the history of art I have also learned about the meanings often attributed to colors and how colors have been used at different times, but this too does not give me an understanding of the essential character or quality of colors.”

    http://www.achromatopsia.info/knut-nordbys-story/

    It’s a fascinating real life example of Mary’s room. I think Chalmers mentions another case in his book, Reality+, and in that case the guy ends up getting corrective surgery (or something like that) which allows him to see color. I might be hallucinating this detail, however, since I can’t find this in the book. I only have a vague memory of coming across it.

    LikeLiked by 1 person

    Reply

    1. EDIT: “From what I understand, this is the scientific view of the matter: colors are NOT objects or in objects.”

       LikeLiked by 1 person

       Reply

    2. Thanks Tina!

       Dennett cites Paul Churchland as also resisting the premise of the thought experiment in a way similar to how he does.  His point is, if we’re going to take the implied metaphysical implications from the scenario, then we have to fully buy into the premise, in a way that not even Frank Jackson likely did.  The problem is it won’t do for Mary to have just current scientific knowledge.  That would be far too incomplete for her then learning anything to have the implications it’s supposed to.  

       That’s the problem with a lot of these types of thought experiments.  People very rarely come to grips with the actual premise.  The reason is that they’re usually hopelessly implausible.  Yet it’s what’s required for the purported conclusion.  We can of course weaken the premise to something more realistic, but only at the cost of giving up Jackson’s conclusion.

       There is the interesting question of what Mary might learn even with just a plausible level of scientific knowledge, which is what I discussed toward the end of the post.  My point is that even there, Mary would be able to anticipate far more of the experience than people commonly assume.  But there’s no doubt that with current scientific knowledge, she learns new things.  It’s just not anything that implies we need to go non-physical.

       On secondary qualities, I think philosophers make too much of Galileo’s stance.  His take was likely much more methodological, mostly centered around what could be measured at the time.  It seems like later thinkers turned it into some kind of doctrine.  But it’s far from universal among scientists that colors only exist in the mind.  Some do try to locate them in the outside world.  None of the ones I’ve read take color to be outside of science.  The idea that they are in some axiomatic sense seems popular among non-physicalist philosophers, but scientists don’t think that way.

       Myself, I think colors are about what reflective properties of objects mean for us, about their evolutionary affordance.  We can think of blue / yellow opponent cells as older baseline mammalian salience detectors, with red / green opponent cells as additional primate salience detection, giving our ancestors an enhanced ability to identify the ripe fruit they needed to survive.  The variations in salience also enhance our object discrimination abilities.  At least, that’s my current thinking.  It might be different later.

       Searching in my Kindle edition, I see where Chalmers also mentions Nordby, including attempts to scan and stimulate his brain to see if color experience could be induced.  Unfortunately it wasn’t successful. I’m wondering if that’s what you remember?

       Later in the book Chalmers uses Mary’s Room to argue that consciousness is the one thing that can’t be explained via structure and relations.  He argues that Mary’s knowledge would only be structural, and so incomplete.  I find the lack of even an attempt to do a structural analysis, as I did a little bit above with the salience point, annoying.  Granted, we can’t do it all yet.  We’re still learning about all the causal roles of color.  But it certainly won’t be completed by people who’ve rushed to surrender any chance of its possibility.

       Looking forward to your blog entries!  I think the knowledge argument is one of the most provocative ones out there.  In the end, I think it amounts to us struggling to conceive of how it could work physically, but that’s been the case with too many things historically (intellect, memory, life, etc) for me to find it compelling.  But maybe I’m wrong.

       LikeLiked by 1 person

       Reply

21. “The problem is it won’t do for Mary to have just current scientific knowledge.  That would be far too incomplete for her then learning anything to have the implications it’s supposed to.”

    I don’t get it. I mean, no one can have perfect scientific knowledge since it’s not even possible in principle. To then say we have to have perfect knowledge and build in the idea that science will someday cover what it doesn’t seem to be in the business of covering strikes me as a superficial demand. Would it be enough to build in some clause saying “science as its paradigmatically understood today” but to allow for future development within its current framework? What I mean is, it’s entirely possible that science as we understand it today could become a much broader discipline many many many years from now. It could morph into something we wouldn’t recognize. Not implausible when we consider it once included other disciplines, and physics was one of “the sciences”.

    I hadn’t realized there were any scientists who think of color as mind-independently objective. Do you mean they think the redness of a strawberry is an inherent property of a strawberry?

    “I find the lack of even an attempt to do a structural analysis, as I did a little bit above with the salience point, annoying.”

    I get your frustration, but what I’m noticing is that color as such isn’t doing much work in the structural framework you bring up. There’s no reason to think color itself, or the experience of color, needs to be included in an explanation of salience. There could be many other ways an object could be salient, so the explanation of color strikes me as ad hoc. “We have experiences of color so they must be doing something to aid in our survival.” In other words, there’s no overarching framework other than the usual scientific theories of evolution. What’s new, is I guess what I’m asking.

    LikeLiked by 1 person

    Reply

    1. “I don’t get it. I mean, no one can have perfect scientific knowledge since it’s not even possible in principle. “

       If you think it’s impossible, even in principle, then in what sense do you think the thought experiment works? Or do you think it works? I’ve given my reasons why I don’t think it does. But I was willing to entertain the possibility in principle of Mary knowing all the physical facts. If you’re not willing to do that, you’re actually not that far from Dennett’s view. Unless you think even a plausible amount of scientific knowledge has the metaphysical implications Jackson takes. If so, how do you see it working?

       On the scientists, I don’t know that many would say that. Their views are complex, and I often don’t understand them myself. Here’s one paper on it that was cited in something else I was reading today. I’ve barely skimmed it so far (warning: it’s dense), but it seems they’re grounding colors in biological details (which count as “the environment” relative to the brain): https://vision.psychol.cam.ac.uk/jdmollon/papers/MollonJordan1997UniqueHues.pdf

       Color may not seem like it’s doing much, but we have to be careful about assuming seeming=reality. Consider pain. Whatever else it is, it’s an active state of the body. It usually comes with increased breathing, heart rate, and other physiological ramping up. Your body wants to do something. Of course, most of that is driven by lower level circuitry in the brain. From a functionalist perspective, we perceive all those aversive reactions as the feeling of pain so we can use it in our deliberations.

       I think colors are similar. They’re not just passive, but active. Consider some things about colors that seem so banal we tend to skate past them. They’re distinctive. And some jump out more than others. Red is the most attention grabbing, orange a little less, then yellow, lime, etc. By the time we get down into the greens and blues, it’s not that striking.

       A possibility I’ve been thinking through is that the color spectrum is a salience scale. If so, then our perception of a color is a computation of that scale for a part of the visual field. The whole visual field of color are those computations happening in a massively parallel fashion. We just perceive the color, but it’s the effects on our cognition that show what the color is. A color perception also comes bundled with a galaxy of unconsciously triggered associations, including a range of other affective reactions. All of this together, I think, is the rich experience of the color.

       I’m sure this isn’t the whole story. But the point is color exists for adaptive reasons. It has effects. It’s an active and utilized analysis of our perceptual systems, a micro-disposition. It’s not just passive mental paint, although it’s easy for us to think of it that way since we can’t introspect the details.

       LikeLike

       Reply

22. I think the thought experiment works to illustrate the point that the scientific investigation into color—color science—is not about color as it’s experienced, but instead about some correlate to that experience, such as wavelengths. (Assuming that is the case, which is now up in the air given what you said about some scientists taking color to be mind-independent. And if this is the case, philosophers across the board need to be put right!)

    I do think our experiences of color in general make possible a scientific understanding of color. It’s hard to imagine how a science of color could get off the ground if no one experienced color at all.

    “They’re not just passive, but active. Consider some things about colors that seem so banal we tend to skate past them. They’re distinctive. And some jump out more than others. Red is the most attention grabbing, orange a little less, then yellow, lime, etc. By the time we get down into the greens and blues, it’s not that striking.”

    Now you’re talking! And note: you’re talking about the experience of color. This is what I feel is missing from what I’ve heard of functionalist accounts. In the philosophical discussion there’s an admission of something called ‘mental’ states, a mental state that is what it does and has some impact on other mental states. But I haven’t been seeing mental states doing anything when push comes to shove. If “mental states” are actually meaningful, they shouldn’t immediately dissipate into a brain state or behavioral state.

    “A color perception also comes bundled with a galaxy of unconsciously triggered associations, including a range of other affective reactions.”

    I would add that sometimes color associations aren’t unconscious, as when we’re considering a work of art or admiring the landscape, or deciding to make stop signs red.

    No, definitely not passive mental paint.

    LikeLiked by 2 people

    Reply

    1. “to illustrate the point that the scientific investigation into color—color science—is not about color as it’s experienced”

       Science actually studies both the experience of color (through psychology) and the physics of it (through neuroscience). The trick is closing the gap between those two. But while it remains, it’s narrower than it once was. I alluded to a document I was reading yesterday, which discussed color opponency from both sides. It gets technical in short order, but if you read just the opening dialog, it’s clear science is studying it from both ends, and that we can learn a lot just by systematically studying experience.

       https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6022826/

       “And if this is the case, philosophers across the board need to be put right!”

       I think more philosophers need to do a better job of tracking the science, particularly if they’re going to make blanket statements about what is or isn’t the purview of science. Which isn’t to say that scientists are always right about that either. But if they can extend science into an area, at least some of them will do it. For that matter, some will push into areas where it’s a poor fit. I think philosophers have a role to play in both situations, but they can only play it if they’re tracking the actual developments.

       “And note: you’re talking about the experience of color.”

       Absolutely. The functionalist shift in perspective isn’t that we should only study functionality and not experience, but that experience is functionality. They are one and the same.

       Of course, many say that identification is wrong, that it’s problematic, that we risk losing something in the conflation. I noted to Mattie that I’m aware of that danger. It’s one of the reasons I always ask people to elaborate on what about experience they feel isn’t functional. I always find the concrete examples to be functional. But I remain open to learning about clearly identifiable aspects of experience that aren’t.

       “I would add that sometimes color associations aren’t unconscious”

       I agree. I think a practiced art critic, for example, is better at bringing those out. But I know I am often moved by experiences: artistic, natural, even technological, but can’t always identify exactly what about them moves me.

       LikeLike

       Reply

       1. Thanks for the article. I like the dialogue format, and it was nice to hear the psychologist complaining about confusing color percepts with nerve signals. To me the psychologist was pointing out something interesting about the characterization of what scientists are doing when they perform such experiments, and the tendency on the part of some scientists to downplay the need for reports from a subject to access that phenomenal experience in the first place. I’ve had this on my mind lately as I’m writing a fictional blog post for my novel on this issue.

          On another note: green? How is that primary? Or if green is allowed, why not a six color split primary palette? Maybe I’m making too much of what I learned from painting (and should have learned in AP art in high school) which is that the traditional three primary colors plus white and black are insufficient for pretty much everything you’d want to do.

          I think the idea that experience is excluded from functionality might come from not hearing about it in functionalist’s accounts. Add that to hearing that experience is functionality, and that just sounds like experience is such and such component of the brain or such and such computation. That identification might need to be explained in greater detail, in other words.

          LikeLiked by 1 person

          Reply

          1. Glad you enjoyed the dialog! I found it interesting for the reminder of how much Helmholtz and Hering were able to figure out at a time when the neuroscience didn’t really exist yet. And it’s interesting that the physiological details are close, but not a complete matchup. For example, it’s safer to say S, M, and L cones rather than blue, green, and red ones. Or +L/-M opponent cells rather than red/green ones. Not that scientists don’t use the color references anyway. I’ve actually seen some papers characterizing blue/yellow opponent cells as blue/green, which seems just as defensible.

             I don’t know enough to help with the green thing. I can see the judgment that it seems less primary than blue, yellow, and red.

             On the identification of experience and functionality, my standard is that the functional term needs to be something that could be swapped for the experiential one without loss of meaning. But that can quickly get dicey because a common mental term, like pain, may refer to a wide variety and/or combination of functional states. Still, for any specific instance of pain, there should be one or more functional states that could be swapped in, even if awkwardly.

             Above I tried it with color, substituting salience. Of course, that’s not sufficient, because color is also categorical. To perceive that something is orange is to categorize it in the category of orange things. It’s also an impression that can be remembered and later recognized. And it contrasts to varying degrees with other colors. Along with many other functions we don’t understand yet.

             LikeLike

             Reply

23. Hi Mike,

    Agreed with others here: great post.

    I think I have a harder time accepting your argument for the notion that there are some things that can only be known through experience than I do with your argument against non-physical information. So for the sake of this comment I’m going to explore the possibility that firsthand experience and thirdhand information about experience are both fully and only physically instantiated “dynamics” and wonder aloud about why they might be different.

    In the case when Mary is in the room knowing everything about how the experience of red works, and also in the case when Mary is out of the room viewing the color red, the only interesting thing here is what it’s like to be Mary. When in the room, Mary enjoys a physical state in which her nervous system and brain and vascular system and physical body are in some state resulting from her possession of a “comprehensive description” of how color is experienced. And I would argue, her conscious faculty is presumed, and so she has a certain experience at this stage.

    Later, she leaves the room, and has another experience. This time based on all of the previous conditions plus the sensory information from the environment. Are these two states identical? My intuition is that physical states of the “Mary system” are not the same.

    To supplement this, there’s some discussion here in other comments about all this knowledge we don’t have yet, and that if we did, would conceivably make the states of the “Mary system” inside the room and outside the room identical. But if Mary is a normal human being, I don’t know if it’s physically possible for her system to instantiate that much knowledge in the way imagined. In other words, it’s not clear to me the human organism is sufficiently complex to simulate the amount of information required–if it’s correct that more information / knowledge would bridge the gap. It may be that she would have to possess and process a HUGE amount of information.

    Related, when Mary leaves the room, the state her organism assumes upon receipt of new sensory information is an utterly unique one: it is the state resulting from the state of Mary’s system at the time, which is based on the entirety of Mary’s event history and responses–in other words, the sum total of who Mary “is” in that moment–interacting with new sensory information in a particular moment and context and place and time. I would argue this is not a repeatable experiment. Nor is it repeatable inside the room necessarily. And nor is our experience of the color red one thing and only thing only, or ever independent of the complex of sensations and knowledge being processed.

    So all that said: it seems obvious that we simply DO have to give some credence to the notion that the state of the human organism, (or of any other sufficiently complex system capable of “experiencing” what a human “experiences”), is in a different state when having a direct sensory experience of the environment than when being in possession of all possible information about that environment including how its own physical system would process certain parts of that information. How can these possibly be identical physical states? And, for starters, isn’t the quantity of information involved radically different? How can we remotely be sure that processing different types and quantities of information “about” a thing on the same physical system will produce the same output state?

    Even assuming / agreeing that firsthand experience is an illusion, and that there is only information processing, I don’t think the same information is remotely being processed…? It seems to me this would require an organism (or any equally complex system) to have the capability of simulating its entire organism in a subset of its systems, and this seems absurd.

    As a thought experiment: Wouldn’t the organism itself be the simplest possible description of its physical state? Or could the organism be fully described with less information than it contains?

    Michael

    LikeLiked by 1 person

    Reply

    1. Thanks Michael.  And I’m glad you focused on the knowledge through experience question.  To be honest, it’s the one I’m less certain of, and so it’s more interesting.

       I like your analysis, particularly the part about how implausible it would be for Mary to be able to have all the physical information, or even all the third person information.  One of the things I always have an issue with with these thought experiments is that very few people really engage with the premises.  Doing so often changes the intuitions involved, or at least balances them.  It makes it easier to accept a counterintuitive answer if we understand how counterintuitive the premises are.

       But one thing I want to call attention to in your analysis is you switch between discussing states and information.  To me, that risks conflating two questions.  Is Mary in a different state after experiencing color?  And does Mary learn something new with the experience?

       There’s no doubt she’s in a different state.  Every experience changes our state.  In my mind, it’s a crucial part of what it means to have an experience.  So, assuming she doesn’t figure out a way to induce the experience of color while in the room (usually considered cheating by most people, but not ruled out by Jackson’s original description), then yes, she enters a state she’s never been in up to that point.

       But it’s not clear to me that every change in state necessarily involves learning new facts.  So the question is whether she learns new information.  And if so, what is it?  Of course, if it’s something ineffable and epiphenomenal, then we can’t answer that last question.

       One answer that some philosophers argue for is she gains a new ability, a new path for acquiring already known information.  Seeing it this way does allow us to articulate what she gains, the ability to remember, imagine, compare, and later recognize discrimination of colors.  I’m not sure this really evades learning something new, but if not, it’s a different kind of knowledge involving different regions of the brain, maybe how-to knowledge rather than knowledge of.  This is particularly true if, as multiple people noted in this thread, there’s a good chance Mary’s ability to discriminate colors would take time to develop.  

       On the question of whether the organism itself is the simplest description of its state, I’m not sure, and it might depend on what we mean by “simple”.  I don’t think it would be the smallest.  A functioning system is going to have a lot of redundancy.  It seems like there would be compression opportunities.  It’s been pointed out before that a big part of science can be seen as finding the right compression algorithms for reality.  Of course, it seems inevitable that would involve abstraction, meaning some loss of information, so maybe it comes down to how particular we want to be.  But it also raises questions about whether Jackson’s “all the physical information” is even a coherent idea.

       LikeLiked by 1 person

       Reply

       1. Hi Mike,

          Reading your response was very interesting, because it suggested I do not understand your thoughts on things as well as I might have hoped. And that is intriguing! It’s intriguing largely because I took some pains to try and form an analysis with the starting point that physicalism is correct, and based on your feedback, I may have misunderstood some things.

          First and foremost, is the distinction between physical states of a system and the notion of learning. You wrote [collapsing a few things here] There’s no doubt she’s in a different state.  Every experience changes our state.  In my mind, it’s a crucial part of what it means to have an experience. . . . But it’s not clear to me that every change in state necessarily involves learning new facts. 

          Okay, I would tend to agree with you, but, this begs a question about your concepts of “information” vs “learning” vs “facts” in the context of our recent discussions on entropy. Let me see if I can highlight my confusion for you. I’m beginning with the starting point that all aspects of conscious experience, including the comprehension of facts, are states (or dynamics) of a physical system first of all. In physicalism they cannot be otherwise. A fact is not meaningful in terms we would intuitively deem as valid, such as what it “means” to the “illusory” consciousness I experience, but in the sense that it corresponds to a physical configuration of energy and matter that, in the context of a sophisticated system like an organism, is causal because of how the physical conditions of the system correspond to causal energy potentials that can create changes, and thus, subsequent states. (I’m not trying to be ironic with my use of quotation marks here, just highlighting key words in this logical train of thought.)

          Confusion No. 1: On the basis of the previous, it seemed irrelevant to me that Mary might imagine she has “learned” something. Any learning would / must be reflected in her physical state. And as we agree her state has changed, I’m suggesting “something happened.” If something happened that was different than what happens when she’s in the room, but she doesn’t learn anything, then I don’t understand what learning is in terms of physicalism. It would be state changes with particular properties. What are those?

          Let me keep going. The next question is, as you asked, if we assume she learns something by leaving the room and beholding the color red, then what is it?

          My answer to this one is that she learns what it is like for the organism named Mary to behold the color red in the particular conditions under which she observed it. She cannot learn what it is like for someone else, like Ted, to actually see the color red, but she can learn what it is like for Mary to behold the color red, because the state of her system IS Mary beholding the color red. She may not learn anything new about the color red that could be listed in a physics book, but she may have / probably has learned certain things about how exposure to the color red affects her physical state, as reported by sensations she has access to. This was probably not the point of the thought experiment, but I think it highlights something important: first person and third person observations of the same phenomena are (likely) not identical physical states. If there is the objection that she knew everything about how her physical organism would respond the color red as well, this is where I would disagree based on the notion the organism can never simulate the entirety of itself.

          So, we have generally agreed to the notion that her state changed. That said, you noted that Mary could have entered a new—or let’s just say “different” state—without learning new information. This gets to confusion number two.

          Confusion No. 2: “Understanding” in physicalism is identical to the realization of certain states. But if there was a state change without learning, in this thought experiment in which learning is directly caused by sight of the color red, then you are implying that only certain state changes can be considered learning. I take it from your note that learning is equated to something like “comprehension” of facts, and that a state change in my own body induced by my taking a flight of stairs, for instance, has probably not changed my general knowledge “about” things. I would agree that I’ve not learned anything new assuming I’ve taken a flight of stairs before. I’m being inefficient with language now but “comprehension of facts” is not what I would think physical information is. I interpret the physicalist interpretation of information to be something like the “set of relationships embodied in a physical state.”

          Confusion No. 2 restated more simply: I would think in physicalism that different states inherently possess, or embody, different information. They are not identical; therefore, the information that exists as the state, which is the only way information could exist, is different. This is why I was confident in using the terms as I did.

          So my own thought writing my previous note is that Mary’s state changed, and that to the extent her new state was the result of physical dynamics it was the result of physical information processing. How could she change states without there being a corresponding change in physical information? I think I’m being true to physicalism by suggesting the physical information MUST have changed.

          If all this happened and she didn’t “learn” anything new then you’re suggesting it is the case that in the future it will be possible to distinguish states that reflect an increase in “factual” information comprehensible to the physical system of the organism from states that are different but do not reflect an increase in factual information comprehensible to the physical system of the organism. This may be true. But it would make Mary’s experience of herself as something “real” because it would necessarily correspond to physical states with certain properties. In other words, the “self” wouldn’t be illusory, but would be specifically the set of states with certain common properties.

          The only thing interesting about this is that it makes it really hard for me to know what you mean when you say information. If you mean facts about the world that emerge as comprehensible notions in certain physical states, but not the actual composition of the organism itself, then a new term is required to distinguish these two. Because I would suggest the state of an organism is one of profound embodied information in terms of the relationships that are embodied by the physical state, and that the relationships that comprise the physical state must encompass any subset of information we deem as factual in nature.

          This gets to the last point: I don’t think a compressed “file” containing a set of facts about the position and velocity of every particle of my body is the same as my body. (You didn’t say they were identical either; I just took your leap of speaking about compression to be your suggesting that the information in a compressed file is the same as the information of an actual body and I’m not sure I can get on board with that.) While I don’t know if the compressed file would actually be less bits of information than my body itself, my hypothesis was that a) no system equivalent to the complexity of the human organism can simulate itself completely on a subset of its constituent mass; and b) if that is correct, learning 100% of all possible “comprehensible facts” about a phenomenon can never produce the same state as the system as a whole responding to the phenomenon itself.

          This means Mary could never predict perfectly what her state would be after seeing the color red, and thus it would be necessary that she at least entered a new state. New to her. A comparison of what she “thought” her state would be to what it actually became might be a form of learning. This is a comparison she could potentially perform because the “before” state would be whatever was achievable given the limits of her simulating capabilities, and the “after” state would be whatever her state actually was upon receiving the stimulus. If some of those differences were comprehensible differences, then she may be able to perceive them, though perhaps not 100% of the time.

          If you survived this lengthy note, thank you for taking the time!

          Michael

          LikeLiked by 1 person

          Reply

          1. Hi Michael,

             First, I appreciate the effort in trying to identify what might be a contradiction in my views.  I’m definitely human, and don’t doubt there are contradictions.  But I don’t think so in this case.  I can’t remember if we discussed the different types of information.  Physical information is the most basic, and what I think is the relevant kind for talking about physics concepts like entropy.  

             But another type is what’s often called “semantic information”, although I think a better name might have been “agential information”, because it’s information that has meaning to an agent (not necessarily a conscious one).  In other words, it’s relative to an agent.  All semantic information is physical information, but not all physical information is semantic information for any particular agent.

             If a large star billions of light years away collapses into a black hole and emits a gamma ray burst that just happens to be aimed at our galaxy and solar system, and the electromagnetic effects arrive and pass through Mary’s body, her state, including her brain state, her physical information, has been altered in some (hopefully) minor fashion, but it would be perverse to say she learned anything, because none of it would be semantic, that is, meaningful for her.

             More relevantly, if Mary gets up and goes to the frig to get a (black and white) drink, she has the experience of doing so, her state changes, but nothing about her world model really changes.  I suppose we could say she “learned” the memory of getting the drink and that it’s now part of her autobiographical memory, at least temporarily.  But if that’s the case, then it seems trivially true that she learns something when she’s let out of the room just by living the event.  

             No, I think for us to say she learns something in the way Jackson means, there has to be some change in her understanding of the world, or herself in it.  (Jackson actually says as much toward the end of the quote in the post.)  For that, the information has to be meaningful for her.  

             Saying Mary learns what it’s like to experience color can be interpreted in many different ways.  David Lewis actually interprets it to mean the new abilities.  If we say that she learns something in the stronger sense, then what about her world or self model changes?  What can she do with the information other than those new abilities?

             Point taken on Mary not being able to have all the physical information on herself.  Although I don’t think she’d need that level of knowledge to avoid learning in the substantial sense described above.  It might be enough to simply understand the effects of the experience at a coarser grained level.  But that’s coming from a functionalist.  

             BTW, just a point of clarification.  It’s not my position that consciousness is illusory.  I do think it’s functionality, so for some people that makes me an eliminativist toward their version.  But then they deny my version of it as “real consciousness”.  We’re all eliminativist toward versions we disagree with.

             Ha!  No worries on long comments.  Hope I didn’t miss anything.

             LikeLiked by 1 person

             Reply

             1. Having slept on it, I realized part of the issue might be Jackson’s use of the phrase “physical information”, which of course is the same term I use for the more fundamental type of information. However I don’t think he means the fundamental version, but is using it to refer to semantic information about the physics of color perception.

                That said, who knows what Jackson might say if pressed on this. I suspect he might answer differently today than when he wrote it since his views have changed. Language ambiguity is the bane of philosophical discussions!

                LikeLiked by 1 person

             2. Thanks for clarifying, Mike. I wasn’t trying to catch you out or anything–just trying to understand. The distinction between physical information and semantic / agential information makes perfect sense to me.

                It does seem to me, however, that Mary cannot learn any new semantic information, because semantic information is by definition inherently that information that can be conveyed meaningfully to a third party or some other agent (not necessarily conscious) that could take an action upon it by virtue of that agent’s own physical configurations. In other words, it is inherently descriptive (or perhaps instructional).

                My thought is that it is at least possible there can be physically instantiated states that arise due to direct sensory engagement with the environment that may not be identical to physically instantiated states that arise due to processing a description of that environment. Brooking this distinction does not need to undermine physicalism in my mind. It just admits of the possibility that firsthand experience may produce different outcomes in a complex system than secondhand ones.

                My question would be: why do physicalism and the uniqueness of firsthand experience need to be mutually exclusive? Or are you not suggesting they do?

                Michael

                LikeLiked by 1 person

             3. Hi Michael,
                I don’t think they need to be. In fact, the point of my argument against Jackson’s non-physical reasoning is exactly that. Maybe there is information that can only be acquired through a particular mechanism. If so, assuming that it must be non-physical is circular.

                That said, I think we should carefully scrutinize why we think it could be the case. Is there any information in any other system that could only be obtained through one pathway? Consider for example the sky constellations as seen from our place in the galaxy. It’s a lot easier to see what they look like from here than it would be on a planet around, say, Arcturus. But we could painstakingly reconstruct what constellations someone sees from another location with the knowledge we have. (There is actually software that does it.)

                My take is that first hand experience is more likely to be like that than some absolutely exclusive manner of obtaining certain information. But I’ll admit we don’t have the equivalent psychophysical knowledge yet for reconstructing another conscious entity’s color perception. Although I’ve been meaning to read Ed Yong’s book on animal perception, which covers what we can currently reconstruct.

                LikeLiked by 2 people