It is “easy” to explain doing, “hard” to explain feeling. Turing has set the agenda for the easy explanation (though it will be a long time coming). I will try to explain why and how explaining feeling will not only be hard, but impossible. Explaining meaning will prove almost as hard because meaning is a hybrid of know-how and what it feels like to know how.
http://andara.uqam.ca/Panopto/Content/Sessions/e77674a1-a902-40e0-a9ac-7a96185c4399/78025b8e-68cf-4256-b663-3ac51c8ed100-f13e20f4-93f1-4af9-9ba0-8d2c221be233.mp4
https://www.youtube.com/watch?v=_jK_-vrwt7Y +
I am not entirely sure why is Prof. Hanard bringing in the Turing biorobot (which is assume is T4) in the debate.
ReplyDelete“People usually reply: But I’m not a robot, and I’m not interested in how robots can do things. I want an explanation of how I do things.”
To me, this sounded more like a weak equivalence VS strong equivalence appeal. Even if a robot can do everything that I can do, how can I be sure that the causal mechanisms are exactly the same (for kidsib: this is weak equivalence. Strong equivalence would be that the robot does everything that we can do AND have the same causal mechanism – which is what people are doubting)?
I have trouble seeing how bringing what the brain is doing internally solves the weak equivalence problem. Even if the robot’s “brain” could do everything internally (with similar action of neurons, neurotransmitters, etc.) than we can do, that does not ensure that the algorithm will be the same. Perhaps there is more than one causal mechanism for engineering brain performance in an indistinguishable fashion (wouldn’t we be in trouble in this were to be the case?)
“Now once we have a Turing biorobot — its doings, external and internal, indistinguishable from our own — the “easy
problem” of explaining doing is solved”
I thought that we had established in the class that T3 was the right level. And this is especially if Prof. Harnard is arguing that meaning and feelings are tied together. Since sensorimotor capacities are enough to pick up semantics, then why wouldn’t T3 have feelings as well if “meaning is felt”?
Florence (and everyone):
DeletePlease be sure you have these things clearly sorted out in your mind:
1. What can CogSci explain, and how?
CogSci can explain how and why people can do what they can do (the "easy problem"), by "reverse-engineering" the causal mechanism that generates that capacity. The mechanism cannot be purely computational (Searle, Symbol Grounding), so it cannot be just an algorithm. But it can be part computational (Turing) and part dynamic (T3, T4).
2. How can you know for sure whether an entity feels?
You can't (except if the entity is yourself; Descartes' Cogito). This is the other-minds problem. The best you can do is to rely on the correlates of feeling. These are verbal (T2), behavioral (T3) and neural (T4). If you can't distinguish the entity from a feeling person, assume it feels. You can never be sure, but that's close enough -- and it's as close as you can ever get,
3. Does T3 (or T4) guarantee feeling?
No, they are just the best that CogSci can do. If T3 (or T4) does not feel we will never know it (other-minds problem).
4.Does cognizing require feeling -- and if so, why?
That's the "hard problem" of explaining how and why organisms feel. No one knows. And no one knows whether or why feeling is necessary for cognizing. If it is not required -- i.e., if there can be T3(or T4) Zombies -- then, by definition, cognizing just means T3 (or T4).
5. Is TT based on weak equivalence (same Input/Output) or strong equivalence (same internal structures and processes)?
All TTs -- T2 (verbal), T3 (verbal and robotic) and T4 (verbal, robotic and neural) -- are based on weak equivalence, but T4 (synthetic brain) is clearly closer to strong equivalence. T5 (real brain) would be strong equivalence.
(But in many ways "weak and strong equivalence" are only computational notions: same I/O, same software (algorithms), same hardware. For dynamical systems it's not clear what "strong equivalence" would mean. In physics you can ask: What if two (or more) different causal theories can each predict and explain everything in the universe (all possible data)? Which one is right? There is no way to know, even if you pick the most "economical" theory (Occam's Razor). This is called "underdetermination": All (nontrivial) theories are underdetermined by their data. They are explainable by more than one theory. In terms of the Kolmogorov/Chaitin complexity-theoretic notion that a theory is the shortest string of bits (algorithm) that can generate all the data, there is no guarantee that that shortest string is unique -- just like the minimal grounding set of a dictionary is not unique -- nor which, if there are several, is the "right" theory -- nor even that the shortest one is the right one! That's just one of the Cartesian uncertainties we have to learn to live with, just as with the other-minds problem. You can't know for sure. But it's close enough for all our purposes. All our categories -- except in formal maths -- are approximate, including CogSci's reverse-engineering theories and Physics's "Grand Unified Theory of Everything.")
Thank you for clarifying about the TT and weak/strong equivalence
DeleteI have some questions regarding the all-encompassing definition of feeling as presented in this article.
ReplyDeleteHarnad states that “it feels like something to do most of the things people and animals are able to do (while they’re awake). It feels like something to see, recognize and manipulate an object.” He mentions that “we feel pain when our tissues are injured”, but that “it also feels like something to think, believe, understand or doubt something.”
I am somewhat confused as to how feeling something through somatosensation can be the same as feeling something through a mental state. It seems to me that feeling pain by being hit by something is different from feeling pain by witnessing the death of a loved one, for example. Obviously both feelings of pain require input of some sort. Pain after being physically hit by something can be felt when the pain afferents in the body are activated. Emotional pain can be experienced after receiving input from the scene in front of you and interpreting that input.
If the pain afferents in my body are always activated, then I will always feel physical pain regardless of how I choose to think about it. However, if I see a scene in front of me, I will not always necessarily feel emotional pain. This feeling of the emotional pain depends on my interpretation on the scene in front of me. It seems to me that one type of pain is hard wired and instinctual while the other type of pain stems from interpretation and understanding.
How do these different types of feeling fall under the same definition? Is there even a difference between physically feeling something and feeling something because of a mental state?
Reginald, bodily sensations (sight, hearing, touch, taste, smell, movement) are all felt. So is migraine, hunger, fear, relief. So is the memory of a cat, and the meaning of "the cat is on the mat" -- which feels different from the meaning of "the mat is on the cat" (and not just because it sounds different). Memory images are grounded in sensory images. But when it comes to (grounded) propositions (and even formal ones, like 2 + 2 = 4 vs. 1 + 1 = 2 or 1 + 1 = 2), it's not just mental images that make understanding them feel like something but their grounded content: It feels like something to say, mean and understand every different proposition we can say, mean and understand. And no two different propositions we understand feel the same. Just like different JNDs in psychophysics. (And this is what Searle lacked for Chinese in the Chinese Room.)
DeleteIn this articles, Harnad asks, “What’s the bonus from feeling something? What causal role does feeling fulfill, that doing alone does not?”, and then goes on to argue that feeling seems to be necessarily important for deriving meaning.
ReplyDeleteIts seem to me like tons of problems arise in trying to figure out the causal order of whether (1) doing results in feeling, and feeling results in meaning, (2) doing and feel together result in meaning, or (3) doing and meaning function together, and feeling results, etc… Regardless of how these terms interact, it seems like all three concepts entirely depend on one definite interactive component: the continuous physical data a system receives from its environment combined with the mechanistic design of the system itself.
If for example feeling was necessarily required in order to derive meaning from a particular image, word or, more generally a mental object, wouldn’t the system require some sort of temporal ordering to the events and if not, at least some sort of correlate (neural perhaps) in order to show how meaning derives from feeling?
Others might argue that feeling does not result in meaning, but actually is meaning. Perhaps meaning is really just the feeling that results from certain logical categorization processing that takes place. In this case, doing would result in meaning (which is really just the feeling of correctly processing sensorimotor dynamics)
I don’t believe that the relationship of these terms bring us much closer to figuring out the problems of cognitive science. However, the implications of all of these terms change when experimenters adjust the input and processing parameters of a system on trial. An individual’s subjective report changes as the experimental parameters do, and thus the meaning and the report of his feelings do as well. Whether his or her report truthfully expresses what he or she consciously feels is a different story. Regardless, it seems clear that the ambiguity between doing, feeling and meaning, can only be solved using direct, physical measures for observing these abstract notions.
Adam, I didn't say feeling was necessary for anything. If I knew what feeling was necessary for, I'd have solved the hard problem. I just said that some internal states are felt states (e.g., seeing something, or understanding something) and not just "done" states. E.g., it feels like something to mean. Meaning is not just T3 grounding. But I have no idea how or why...
Delete“We can reduce just about everything that cognitive science needs to explain to three pertinent anglo-saxon gerunds – doing, feeling, and meaning.”
ReplyDeleteHarnad explains that cognitive science aims for the explanation of how and why humans are able to do everything that they can do. He claims that this includes things we do (e.g walking, talking), things we feel (e.g sensations, emotions), and how we are able to understand meaning (assigning ideas to certain words). However I question the distinction he makes between the types of things humans are able to do/feel and how they are able to understand meaning. Meaning seems to be just another form of doing along with feeling – I do not see how it should be categorized separately when it is a combination of the two categories. This human capacity is just another thing that humans are able to do, which is just assigning semantics to symbols. As Harnad explains there is a feeling associated with meaning and this is shown with words with double meanings – we are able to understand which meaning is referenced by context. This occurs through feeling of the context and by matching a referent to a symbol, which is something we do. Perhaps the distinction is based because it involves the combination of doing and feeling - but this does not clarify anything since there are other things humans do that include feelings - so how come these don't get categorized separately?
Danielle, the basis for the doing/feeling distinction is that doings are observable dynamics, whereas feelings are not. No one can see feelings except the feeler. That really does make them different -- and problematic, for a causal explanation (unless feelings are an independent causal force, as psychokinetic dualism would say they are).
DeleteYou are right that "meaning" is not a third category. It's hybrid: you think a grounded thought (subject/predicate, T3-grounded) -- "the cat is on the mat" -- and that's still not meaning, because it also feels like something to think (or say) and mean and understand that thought. (What Searle had for English but not for Chinese.) That's pretty much the same as the fact that it feels like something to see an object: it's not just detecting and naming it.
" The problem is explaining how
ReplyDelete— and, even more problematic, why? With doing, it’s easy to explain how and why we can do what we can do. With
feeling it’s hard, if for no other reason than that doing alone already does the job: it’s enough to explain what kind of
Darwinian survival engines organisms are, i.e., by what causal process we evolve or learn the ability to do what needs to
be done for our survival, reproduction and lifetime success (such as it is)."
It seems to me that there could only be two possible reasons for why we feel:
1. We could not do what we do without feeling in the way that we do
2. Feeling is some byproduct of evolution that, due to physical constraints in neurology, happens to go along with an adaptation (cognition) that served us very well
In the first case, in order for a robot to pass T3 they would necessarily have to feel as humans do and the other minds problem would be somewhat easier to solve in many cases (for instance, we would know that all humans feel). As such, in building a T3 robot we would discover both how and why humans feel as they would be an essential 'cog' in the robot's mechanism for generating human-like behaviour.
In the second case, I question the importance of knowing why we feel. If it has no bearing on the world around us, perhaps it would be enough to know how we feel without looking for reasons as to why we feel.
“For example, we feel pain when our tissues are injured. The temptation is to say that we need to feel the pain otherwise we would not notice the tissue injury and we would not do what needs to be done about it. But doing is doing. If something needs to be done, why is it not enough to have a mechanism that, when it detects tissue injury, sees to it that what needs to be done is done — withdraw the hand from the fire, avoid fire in future, remember, learn, compute and even say whatever needs to be done — without bothering to feel anything at all? What’s the bonus from feeling something? What causal role does feeling fulfill, that doing alone does not? Our Turing robot can do everything that needs to be done, whether it feels or not. If it does feel, it remains to explain what causal role the feeling itself is playing. And that’s the hard problem.”
ReplyDeleteThis example of the necessity of feeling brings to my mind many neurological studies on the sensory-motor versus affective-motivational aspects of pain. Indeed, certain individuals (who have primarily suffered traumatic brain injuries) experience the sensory sensation of pain without the affective-motivational side of pain (ie without “feeling” it in any emotional sense). They are able to take their hand out of the fire, know something ‘painful’ has happened, but they are unable to feel any of the emotional response that occurs during/after a painful experience. The question here posed by Harnad leads me to question: are these people really feeling pain? sure, they are experiencing pain, and these words are often treated as synonyms, but is there a difference here between feeling and experiencing? These certain individuals are essentially experiencing pain the way a robot does, by knowing it occurred, but not feeling any emotional response to that. However, is their pain less human? The sensory aspect of it is the same as yours or mine, but without the affective aspect of it, would we even still consider it pain? When it gets down to it, really how important are the affective dimensions of our experiences? I’m puzzled by this question constantly when reading about AI and the problem of consciousness, and I’m not sure I’m getting any closer to an answer. In fact, the more I read, the less inclined I am to think of robots as becoming like humans, and the more inclined I am to think of humans and robots meeting somewhere in the middle ground.
"These are all just adaptive nociceptive functions. Everything just described can be accomplished, functionally, by merely detecting and responding to the injury-causing conditions, learning to avoid them, etc. All those functions can be accomplished without feeling a thing; indeed, robots can already do such things today, to a limited degree."
DeleteIf taking your hand out of the fire does not require you to feel a thing, I guess the people you describe do not feel pain at all. They are just reacting like the robots Harnad described.
"The question here posed by Harnad leads me to question: are these people really feeling pain? sure, they are experiencing pain, and these words are often treated as synonyms, but is there a difference here between feeling and experiencing? These certain individuals are essentially experiencing pain the way a robot does, by knowing it occurred, but not feeling any emotional response to that. However, is their pain less human? The sensory aspect of it is the same as yours or mine, but without the affective aspect of it, would we even still consider it pain?"
DeleteI think that regardless of the type of pain they feel, or the intensity, they are still feeling and that is what makes any experience "human". People without neurological damage also have various pain tolerances. Also pain is not the only kind of feeling there is - it feels like something to pull one's hand out of the fire, and to know that one's hand is burned. And regardless of the quality of that feeling, it's still there, which leads us back to the point of the article - why do we feel at all?
In Harnad's 'Doing, Feeling, Meaning and Explaining' paper, the mention of psychophysicists seems to be an argument against a philosopher's belief that there is a fundamental difference between perception (touch, sound, 'feeling') and such things as believing and understanding (or meaning, though I am not quite sure how to categorize these terms). This idea is demonstrated in three ways. First there is the assumption that the work of psychophysicists is based on the assumption that all perception is a subjective feeling. Just-noticeable-differences, or detection thresholds (the smallest amount of something in the environment needed to be able to notice it) can be influenced by any number of cognitive factors that would change the experience all together, thus everything we are experiencing is a product of what is 'in our heads'. The second argument is the fact that we can make just-noticeable-differences between things such as meaning, the same way we would do with loudness or brightness. The third argument is the mention of blindsight which demonstrates that we can feel, in the sense of recognition, despite the lack of traditional sensory awareness (for example: sight due to the biological processing of wavelengths.) Though these arguments do not necessarily mean that the feeling of sight is equal to the feeling of meaning something, it does give a convincing reasoning behind the idea that we feel when we mean something.
ReplyDelete"So the “hard” problem of explaining how and why we feel is even more pervasive than sensory and emotional experience. Semantic sense is afflicted with it too."
Does Harnad mean to combine the problem of how and why we mean with how and why we feel to produce a new version of the hard problem? Or is the problem the way it was before, just the definition of 'feeling' has been expanded?
Harnad's challenge to figure out "how and why we feel rather than just do" (which, if I understand correctly, is equal to the question: 'what is the evolutionary advantage of feeling?' or 'what is the cause?'), would arguably not only answer the 'why' part of the hard problem but also the 'how' part of the hard problem (how what we do and how we do it gives rise to the experience of doing it). I was thinking about this for a while, and I cannot come up with a valid answer. The only thing I can think of is perhaps there is no cause for feeling (including perception, emotions, and the newly included meaning) and it is just an evolutionary side-effect of being a live organism, behaving appropriately with the right inputs and outputs in the environment. Perhaps there is no evolutionary reason we feel, just as there is no real adaptive reasoning behind the fact that stimuli acting on one side of the body is processed in the opposite side of the brain (a tickle on my right arm is processed in the left hemisphere of my brain.)
It is also mentioned in this paper that we must be awake to feel that we are doing things. Is Harnad talking about perceiving stimuli in the environment? Because it has been proven that we are conscious during sleep, though I do not know the extent of this. We are conscious during dreams that we may not remember, but I'm not sure if we are capable of perceiving touch or smell (or other traditional forms of perception) during sleep.
"But we do know that Turing robots are synthetically made rather than natural, so the uncertainty about whether or not they feel is greater than it is with our fellow human beings (greater even than with our fellow animals, except perhaps the ones that are the most unlike us)"
Regarding the "other-minds problem", it makes sense that we would be more likely to convince ourselves that other humans feel (in comparison with machines) since we can compare our similar brain activations in response to the same stimuli (a painful cold-press task where you hold your arm in cold water, or a funny movie).
“VITALISM, ANIMISM AND FEELING (Reply to Anil Seth)
ReplyDeleteBut in the case of the mind/body problem, we can say (with Cartesian certainty) exactly what it is about the mind that is inexplicable (namely, feeling) and we can also say why it is inexplicable: (1) there is no evidence whatsoever of a “psychokinetic” mental force, (2) doing (the “easy problem”) is fully explainable without feeling, and (3) hence there is no causal room left for feeling in any explanation, yet (4) each of us knows full well that feeling exists. Hence the “hard” problem in the case of feeling is not based on the assumption there must be a non-physical “mental force” but on the fact that feeling is real yet causally superfluous.”
While I think I comfortably understand/at the least wrap my head around these 4 points, I have trouble understanding why this warrants an argument in the first place. Harnad’s argument is logical. There is some disconnect between feeling and what’s causing feeling. However, if we go back to the other minds problem, it’s been well-established that, essentially, we just can’t know—we can only assume. So we don’t know if others are feeling, we don’t know if robots are feeling, we don’t know why or how feelings get felt—what is there to debate if we just don’t know? The hard problem seems more like a mental exercise than a fruitful path towards reverse engineering. As far as mental exercises go, yes, it is weird that there’s a causal disconnect between having something real (a feeling) and seemingly lacking a purpose for feeling, but do we need to consider the hard problem further than establishing its tricky existence? I guess what I’m asking here is what do we *do* with the hard problem now?
“Regardless of whether we arrive at our explanation via T2, T3, T4, or via the direct observation, manipulation and modeling of brain function, we will always be faced with the uncertainty of whether we have explained all of cognition, or just our know‐how.”
ReplyDeleteSearle’s Chinese Room Argument hinges on the fact that when given the correct program/materials, he could interact with someone in Chinese indistinguishably from an actual Chinese person without understanding Chinese. But Searle only knows that he does not understand Chinese because it feels like something to understand a language (it feels like something in general to understand) and he does not have that feeling with regards to Chinese. I just wrote out and then erased the following sentence, “The fundamental difference between Searle and a T3 passing robot is that Searle is human and presumably can feel but there’s no way of knowing based on the TT (which is designed to measure functional capabilities) whether the robot can feel.” This sentence got erased the first time around because I realized that if a robot passed T3 it would have to be indistinguishable from a human in its interactions and everyday life so there would be no more reason to doubt that it could feel than there is to doubt that Searle can feel. We don’t know Searle can feel but we trust that the can so presumably we would trust that the robot could IF we did not know he/she was a robot.
Basically the point I’m trying to get at in this post is that I’m frustrated with the Turing Test. Or more accurately, I guess I’m frustrated with the hard problem. I want to know if a T3 robot would HAVE to be able to feel in order to pass the TT. If an engineer built a robot solely with the intentions of mimicking the functional capabilities of the human and that robot happened to pass T3 would its ability to do everything we can spontaneously induce the capacity to feel as well? (Okay in retrospect that question seems far-fetched and ridiculous). I’m struggling because it’s obvious to me that it feels like something to be a cognitive being and to successfully explain cognition we clearly have to explain how and why we feel the things we feel in addition to how and why we do the things we do. Yet, cognitive science is nowhere near the point where a robot that passes T2 or T3 can be constructed and it’s not completely clear whether that robot would need to be able to feel.
Reading the Doing, Feeling, Meaning and Explaining article and the comments below, I think the main problem for most people is that how doing and feeling are different.
ReplyDeleteFirst, I think feeling is a one-person thing. I can feel what I feel, but I cannot know what you feel. I can guess what you feel in terms of the situation and what kind of person you are, but I cannot confirm it. However, if I do something, you can do the same by speculating. (However, for this part, I feel it hard to explain why belief is doing. I do believe belief is doing though, I just do not know how robots can hold the same beliefs as I do and how beliefs can be observable? Or are you only saying that maybe it is possible to test someone's beliefs but not their feelings?)
Second, doing something is also felt. It feels like something to do. Then it will be difficult to categorize feeling into doing, or the other way around. Also, when we feel something, we do not always have a response to it. We feel hot, but why do we feel hot? Maybe to tell ourselves to go to a cooler place. But what if there is no cooler place and there is no way to cool ourselves down? Then why are we feeling hot? We feel pain, but why should we feel it? Why cannot we just deal with the injury without feeling it?
Third, Turing Tests can test doing, but not feeling. It is hard to test feeling because if we ask a T3 robot whether they can feel, and they will definitely say they will, because feeling is only about itself, and only the robot knows whether it feels or not, and we do not know. However, it is possible to test their doing ability by using Turing Tests, and I think that is why doing is an easy problem while feeling is the hard problem.
Also, I was wondering how about meaning. Searle's Chinese Room Argument says computation is leaving meaning out, but in his argument, T2 is the subject. What about T3? Does T3 include meaning successfully? I believe meaning is not only about doing because I feel something when I mean something or I understand something, so T3 cannot include meaning. Otherwise, it will successfully include feeling as well I guess?
By now, many ideas that appear in this paper have been made clear. In summary: explaining feeling and meaning is impossible. The latter one in particular because it relies on “know-how” and “what it feels like to know how.”
ReplyDeleteDoing is the easy problem (and can be explained with a TT model: a robot that can do everything that we can do would explain the causal mechanism of how we can do everything that we can do (T3)), and feeling/consciousness is the hard-problem. Furthermore, the causal mechanisms of doing cannot explain feeling.
The issue with feeling, is that we can only be sure of our own. For anything outside of us, we simply “mind-read.”
There seems to be no room left for a “feeling force” to causally explain something.
At some point when doing this reading, I was inclined to think “do I really feel anything at all when I look at this paper and read this, do I really feel anything when I look at an apple? The conversation has been so pushy regarding the fact that it feels like “something”, to do this, and that, and that, and “wao! Doesn’t looking at that apple feel like quite a thing?!” That I hadn’t really paused to think, what am I feeling, and not necessarily about something specific but in general, in everyday life?
We haven’t paused to distinguish between, feeling feelings and “feels like somethings” that I am beginning to think, “something” is just another weasel word. Allow me to draw this distinction: it feels like something to see an apple, and it feels like something to think about an apple. Then there is: I feel sad, I feel happy, I feel hurt, I feel lonely, I feel ecstatic, I feel depressed, I feel abandoned, I feel ashamed, I feel disappointed, I feel down, I feel annoyed, I feel attacked, I feel crazy, I feel lunatic, I feel thrilled, I feel euphoric, I feel loved …. I would guess, Harnad would argue that feeling and consciousness being synonymous, could be used to simply say “I am conscious of that apple”, and maybe my “mélange” would be resolved?
I was thrilled to find the first commentary by Judith Economos, that in her view, some things “do not feel like anything.” Harnad replies largely, but I believe the core of his answer is where he says: “Well then what is a thought, and how do you know you are thinking it? In what does your “consciousness of” it consist, if not that it feels like something to think it? (How can you be eye-witness if your testimony is that you didn’t see a thing?)”
I think reading so much about this might be mixing me more up than anything else, but I’ll give it a try: let’s assume thinking is not a weasel word, but an actual process our brain can do. I think about purple, I’ve seen purple, I can imagine purple. Harnad says we can’t stop there, “thinking about purple feels like something.” Couldn’t it be simply that I replay (my neurons fire such that I think about purple, yet how and why (what Harnad wants us to answer is not resolved)) in my head what to me is the color purple? Certainly, “I feel lonely” is very different from “looking at the color purple feels like something.” Can’t we argue that with feelings, emotional feelings, I can say “I am” in that state, I can say I am sad, I am happy, I am in love, etc. I cannot imagine anyone claiming: “I am color purple.” Perhaps they can claim “I am looking at the color purple,” but this is just mere doing.
Andrea, I understand why you say it seems to confuse us more, the more we think and read about the problem of feeling, it is the hard problem after all. In the past, I never really thought of how it “feels like something to look at an apple” and then we learned how consciousness is the same this as feeling. Therefore someone who is conscious, feels. You then write: “I would guess, Harnad would argue that feeling and consciousness being synonymous, could be used to simply say “I am conscious of that apple”,” in order to explain how and why it feels like something to look at an apple. However, this doesn’t answer your question as you are just replacing feeling with its synonym, not resolving any problem. Although feeling can’t be explained, a very important part to remember is that there are several different types of feelings. I also think that there are gradients within feeling: You can feel a little sad or extremely sad for example. So when you state:
Delete“Certainly, “I feel lonely” is very different from “looking at the color purple feels like something.” Can’t we argue that with feelings, emotional feelings, I can say “I am” in that state, I can say I am sad, I am happy, I am in love, etc. I cannot imagine anyone claiming: “I am color purple.” ”
Yes, I feel lonely is entirely different from what it feels like to look at the color purple, because like I stated, there are many different types of feelings. Additionally, the fact that you feel does not mean that you are what you feel (ie- the fact that it feels like something to look at the color purple doesn’t mean that you are the color purple.) I think part of this can be answered by the gradients and range that can be found within one emotion. Feeling is not the exact same intensity every time, for example, looking at an apple after the start of my meal will not feel the same as looking at an apple if I haven’t eaten for several hours. So, maybe the reason why it is hard to conceive of what it feels like to look at the color purple is that the gradient of emotion is so small that it doesn’t ‘hit’ us as an obvious, categorizable emotion (ie- happy, sad, angry, in love etc). Most people don’t associate anything particular to the color purple and there is therefore not an enormous emotional trigger to that color. But however small it may be, it still feels like something to look at the color purple.
"We can reduce just about everything that cognitive science needs to explain to three pertinent anglo-saxon gerunds — doing, feeling, and meaning."
ReplyDelete“Meaning” seems to be a separate problem from doing and feeling but I’m not quite sure why. The way I see it, the capacity to assign meaning to symbols is simply a consequence (or output) of doing and feeling. In other words, doing and feeling are sort of “pre-requisites” for meaning. This brings us back to the symbol grounding problem.
I could also argue from a slightly different angle: The capacity to assign meaning to symbols is just another thing that we do. And like everything that we do, there is feeling associated with meaning. But it’s possible to distinguish the feeling from the doing.
To provide an example, let’s consider Searle’s Chinese Room argument. There was no understanding involved (i.e. Searle did not understand Chinese) but the language would have conveyed the information just as well if Searle had understood Chinese.
All in all, I believe the distinction was made for the meaning of our sentences to really be intended. But I do not believe that the difference was qualitative (rather it is a matter of degree).
''What's the bonus from feeling something? What causal role does feeling fulfill, that doing alone does not? Our Turing robot can do everything that needs to be done, whether it feels or not. If it does feel, it remains to explain what causal role the feeling itself is playing, and that's the hard problem.''
ReplyDeleteWhat if feeling doesn't cause anything? What if it is simply the final effect of everything that has happened in the world. Feeling is the incidental becoming meaningful its meaning lies in its total inaccessibility, its inability to be understood or to be accessed. Feeling cannot be explained precisely because feeling doesn't cause anything in itself. It is not an agent of the world but rather its effect. As humans, we misconstrue this order, giving agency to our feelings, saying that they cause what we do. We read backwards, starting with ourselves and not with everything that is not ourselves. It should be the other way around. After doing comes feeling. And this feeling is rooted in every instance of incidental things that have happened and that end by bring us to be here today (our evolutionary history, and our history as individuals - all contexts). The historical uniqueness of each individual is what creates feeling, the fact that after a series of things which can be (in theory) fully explained functionally (as in this happened, then that happened, which then led to this happening until you get to yourself), we get to the present moment, which while bounded by these experiences, also alludes them in its special status as not being exactly any of those things. Feeling then is caused by the intrinsic connection to all things other than, and the intrinsic separation of the self from all things in existing at an instance (a unique place and unique moment in time) So while we can explain (in theory) everything that has happened and the reasons why they happen, we can't explain what it is our feeling is made of, how it exists, or why it does in any functional way. Our feelings, while bounded by our functional reality (read: our history [and I mean history in all senses of the word, and at all scales]), can never be explained by it. Feelings are inherently unique in that they exist in a specific time and place, in an instance that has never occurred and will never occur again.
Or rather, maybe feelings do cause something, but they are in the future, and become incidental as they pass into the realm of history. Thus whatever effect that have will be inaccessible and inextricably entangled with any functional explanation of an instance.
Delete