The mind/body problem is the feeling/function problem (Harnad 2001). The only way to "solve" it is to provide a causal/functional explanation of how and why we feel...
Click here to view --> HARNAD VIDEO
But... "Claustrum Nostrum"
Saturday, January 6, 2018
10b. Harnad, S. (unpublished) On Dennett on Consciousness: The Mind/Body Problem is the Feeling/Function Problem
Harnad, S. (unpublished) On Dennett on Consciousness: The Mind/Body Problem is the Feeling/Function Problem.
The mind/body problem is the feeling/function problem (Harnad 2001). The only way to "solve" it is to provide a causal/functional explanation of how and why we feel...
The mind/body problem is the feeling/function problem (Harnad 2001). The only way to "solve" it is to provide a causal/functional explanation of how and why we feel...
Subscribe to:
Post Comments (Atom)
-
Harnad, S. (2003b) Categorical Perception . Encyclopedia of Cognitive Science . Nature Publishing Group. Macmillan. http://www.youtube.com...
-
Harnad, S. (2003) The Symbol Grounding Problem . Encylopedia of Cognitive Science. Nature Publishing Group. Macmillan. or: Harnad, S...
-
Pinker, S. & Bloom, P. (1990). Natural language and natural selection . Behavioral and Brain Sciences 13(4): 707-784. Many people hav...
Dennett: “People undoubtedly do believe that they have mental images, pains, perceptual experiences, and all the rest, and these facts the facts about what people believe, and report when they express their beliefs are phenomena any scientific theory of the mind must account for.”
ReplyDelete“Dear Dan, you keep giving examples of successful prediction of functions from functions, and then an overall causal/functional explanation of the correlation. But when feeling rather than function is what is being predicted, all progress stops with the prediction. There are no further steps to be taken; only regression back to the functional explanation of the functional correlates.”
“Only felt feelings count. If I didn't feel it at the time, I didn't feel it.”
Dennett’s argument is that the Hard Problem doesn’t exist because we don’t truly have feelings, but “believe” that we have feelings (last quote by Dennett). This in turn justifies his standpoint of ‘heterophenomenology’, in which we can learn from a 3rd person perspective the reasons behind people’s beliefs about their feelings. This is primarily very convenient, and secondly ignores the very fact that we actually feel. I have to agree with Harnad’s entire argument, namely because it doesn’t pretend to be able to know things that are clearly out of our reach (for the moment). His emphasis on feeling our feelings seems very plausible because there is a reason why feelings are called feelings--they are felt, not simply thought about. I agree with the fact that Dennett’s argument falls apart the moment it attempts to predict feelings from functions, as only functions can be plausibly predicted from functions (if we want to get to the root of the problem, at least–I guess we can always explain why we feel a certain way with respect to our hormones, but this doesn’t answer the Hard Problem.) In my opinion, therefore, Harnad’s response is very reasonable.
Besides, it feels like something to believe something. Believing that 2+2=4 feels different from believing 2+3=5 -- and not just because it looks and sounds different.
DeleteI enjoyed this paper, as Harnard addressed the problems of Dennett’s claim to have solved the hard problem through heterophenomenology. Most importantly, heterophenomenology focuses on functions, and it does not explain on the causal mechanisms of feelings. Also, heterophenomenology focuses on the verbal reports of human beliefs. Like Harnad states, learning about beliefs will not tell us about human consciousness, but learning about the causal mechanism of feeling will.
ReplyDeleteScientists discover the on-off switch for human consciousness deep within the brain
“When the claustrum was stimulated, the woman just stopped whatever she was doing (speaking, reading, moving) and stared blankly into space; when stimulation was removed, she continued as normal with no recollection of what had just happened.”
This was interesting to read that the patient was still awake when she stopped doing these functions, but I wonder if she is still conscious or not (if she still feels or not). We wouldn’t be able to know since all we know is that there is an absence of behaviour. I’m not sure if we can conclude that there is an absence of consciousness as well.
But this correlation, too, fails to explain, causally, how and why organisms feel.
DeleteI really enjoyed how Harnad broke down all the arguments for Team A from Dennett’s paper and explained his issues with them in a clear and concise way. I noticed that a main theme of his responses was in relation to the debate between feeling and thoughts. I wonder then, if the questions posed by Kant, Turing, and Descartes were re-worded to encapsulate “thoughts” as input/output rather than “feelings” if the arguments would make more sense, and if so, would Dennett’s article answer the question and solidify the A team argument.
ReplyDeleteIt's not thoughts vs feelings, it's doings vs feelings. (It feels like something to think, as is already implicit in Descartes' Cogito.)
DeleteOn (10a) I commented a question asking if we need to be able to feel to do things, and made a comment that the zombie thought experiment doesn’t seem realistic to me.
ReplyDeleteThis reading made me realize that the question doesn’t really matter. Regardless if the answer is yes or no, it leads back to the same question: how/why do we feel.
If we do need to be able to feel to do things (which I assume is the case) - great - how/why do we feel?
If we don’t need to be able to feel to do things (which the zombie thought experiment implicitly suggests) - also great - but to be able to explain how it is we don’t feel, we have to first explain how and why we do feel.
Explaining how and why sentient organisms feel is exactly the same (hard) problem as explaining how and why sentient organisms (or T3/T4/T5) cannot be zombies.
Delete“But I can tell you that until you explain why and how a pinch hurts, the game's not won. (That is does hurt, and that that hurting correlates perfectly with some functional story, is not the how/why explanation we were seeking...)”
ReplyDeleteI think this precisely hits the nail on the mark. It’s why the thought experiment of that lady living in a black and white room works so well (I forgot who came up with it and specifics). So I think the idea is that this lady is an expert on vision; for example, she knows everything about the retinal cone coding of the colour blue, the path “blue” information takes to the relevant regions, etc etc… She knows all of it. But will any of that let her know what it’ll feel like when she sees “blue” for the first time? No. Because the felt experience of blue is just that: felt. And behavioural correlates don’t matter that much, either. There’s no description of her experience of the colour blue (her beliefs) that we can provide to another person which will automatically allow them to have that same felt experience of blue.
But I do think Dennett’s position is important. Because any good theory of what feeling actually is will have to meet Dennett's objection that what's being put forward isn't actually a set of beliefs that can be examined by heterophenomenologuy (because some people confuse the two....) But that doesn't change the point that what's really being put on trial is feeling in the sense of feeling, not some other cognitive feature of people like their beliefs and behaviours.
I agree with you completely. The thought experiment by Jackson (dubbed the Mary experiment or the knowledge argument) perfectly encompasses the point that Dennett seems to miss. How and why we feel is not learnable from any third person data (as of yet and maybe never). My only question for you is this: when you say that there is no description of her experience we could impart to make someone else understand, if the brain is responsible (somehow or other) for feelings, would it be feasible with a completed neuroscience (i.e. we know everything there is to know about the brain and how it works) to snapshot her neuronal firing pattern the instant she sees blue and then replicate that pattern in someone else's brain? In other words, could we feasibly use her brain's experience of something to impart that experience to someone else and would that count as "feeling" in the technical sense?
DeleteHugo, Blind neuroscientist Mary is as much an allegory about symbol grounding and a picture always being worth more than 1000 words as it is about the difference between direct experience and verbal description. (BTW, I don't think congenitally blind people's descriptions of the visual world are completely ungrounded; they are just vague, like our sense of what it feels like to be a bat.)
DeleteDan's really just talking about T3/T4/T5, i.e., the easy problem, not the hard one.
Zachary, T4/T5 would not solve the hard problem, "just" the easy problem. Nor would it solve the other-minds problem, even if it predicted feelings perfectly -- and even if it could induce them in someone else by telemetry or something... But having got that far, who cares?
The major comment from this paper is that the paper by Dennett for this week isn’t telling us anything about the how/why of our feeling. It skirts around the point by trying to explain that it can be determined through correlational measures of brain functioning and other behavioral measures which cannot be used to close the gap between what they actually tell us about (functionality) and the questions we really care about (the how/why of feelings). The measures reveal nothing more than the functional systems which underlie our feelings, interesting in their own right but irrelevant to the hard problem.
ReplyDeleteHarnad’s critique of subliminal perception where something could be considered to be felt without conscious awareness and later impact behavior is in essence the idea of homuncularity. If we were to posit that there is some sort of self within oneself which felt the feeling that we didn’t consciously feel we must therefore need to posit even more “inner selves” which feel what another self does not. I quite agree with this criticism, reminiscent of that for the pitfalls of introspection as it does little to reveal the how/why of feeling that we are really interested in. Since the hard problem is unable to be solved because of the above limitations, I wonder if anything could bring us closer to the answer, perhaps not solve it but give us a better approximation?
The solution to the easy problem would be as close an approximation to a solution to the hard problem that we could ever hope for ("Stevan says...").
DeleteIn response to Dennett’s heterophenomenology, Harnad declares that there is no possible way to solve the hard problem of consciousness. Once doing is explained (T3 and T4 AI beings included), you have no degrees of freedom to explain feeling. Causal mechanisms are limited to doing and therefore subjective phenomena will remain unexplained. A feeling cannot readily be reproduced as accurately in scientific terms as they can freely in an organism’s mind and thus only correlational conclusions can be drawn. The problem of other minds does not allow for subjective consciousness to be accessed by others in the first place so there is no use in looking for a way to do so.
ReplyDeleteIt's not "accessing" feelings that's the problem, it's explaining their causation and causal role.
DeleteAs Professor Harnad makes clear, heterophenomenology doesn't get us much closer to solving the hard problem. Although Dennett conveniently dismisses feeling and experience as something too hard to pin down scientifically (in fact, he almost seems to come to the conclusion that feelings don’t exist at all), he fails to understand that this is precisely what the hard problem explores. He offers a few cases of “feeling” which appear too transient or complex to describe accurately, and concludes from this that it’s all too difficult to study. I have several points about this:
ReplyDeleteHowever, we don’t need to look for special cases where the “feeling” is ambiguous in descriptive terms. As Professor Harnad points out, once can just use the simpler case of not-warm vs warm-now. The other special cases are all just variations on the same theme. Complexity doesn’t rule feeling out. We can pin down feeling in our own experience. I know what it feels like to be warm or not.
Pointing to the other minds problem as a barrier to investigating feelings shows that Dennett was asking the wrong question to begin with. The feelings are there; we all know it because we feel them. We know that the detailed accounts of them we can produce in heterophenomenological terms don’t amount to feeling them. The question is not to prove their existence or to capture their every quality in what amounts to simulation. The question is to explain them. Why are they there? What gives rise to them? Heterophenomenology (and arguably, whatever method Chalmers proposes as well) can only get us to a solution to the easy problem at best.
Spot on, again...
DeleteAfter reading Harnad's comments on Dennett's article, I still feel that important questions regarding the hard problem have not been answered. While I do feel that heterophenomenology will not get us closer to answering the hard problem of consciousness, I am at a loss for any solution that could give us these answers. Is it then true that until we can access the mental state of another individual, it is an unanswerable question? If verbal communication and "telepathy" don't currently cut it, what is the next step? Is there even a plausible next step?
ReplyDeleteYou (and many others) keep mixing up the HP and the OMP. The HP's not about accessing others' feelings, but explaining how and why organisms have them.
DeleteBesides, if the OMP's what's bothering you (with humans) you just have to ask!
DeleteThank you for the correction; I'll make sure to word my questions more carefully, as I did mean to ask about the hard problem, not the other-minds problem!
DeleteI have an additional question, however: if we could simply solve the OMP by asking other humans about their feeling states (as we do already), wouldn't that mean that the OMP is in fact not a problem that has to be solved? I feel like there's something missing here; we did discuss that feelings cannot always fully be conveyed through verbal communication, and it remains unclear to me if being in the same mental state is just shared feelings (i.e. we both feel sad about something) or if it is something more exact, akin to being in the same physical state (except with feelings).
Even if we did ask other humans about their sentient states, we cannot actually know if they are feeling that state. We only have an objective "3rd-person" account of their feelings. What they say they feel is still a behaviour that we observe, and so we cannot conclude that they are feeling. The only "mind" that we know are feeling are our own. This is why the OMP is hard to solve.
DeleteAll that matters is that you felt something at all, be it ever so skittish, malleable, ambiguous, or whatever other fallibility you want to ascribe to it. It is a feeling, and that's a fatal liability for a functionalist...
ReplyDeleteI think that this point is important to keep in mind when we are looking at reverse engineering our own experiences. Even if we are able to figure out what kind of area maps generally to a certain feeling, how would we go about implementing this at a deeper level? For example even if according to our biology we should be experiencing a certain phenomenon if we don’t report it according to the info we are wrong. We can’t simply take neurological correlates and expect them to tell us what the results are. As well since humans have variability in thresholds for what registers as a stimulus how would we control for this variability when creating AI? The best way here would be through randomization which seems overly complicated and unnecessary. Therefore I tend to agree with this point that functionalism cannot simply be the answer because it doesn’t understand humans and can’t account for the subtleties that play into our lives.
I think you are talking about the OMP not the HP...
DeleteDennett, in his paper, tries to prove the ephemerality of scientific measure of feelings, therefore demonstrates the futility to focus on it at all. However, the reason why the hard problem is so hard is not because of the difficulty to quantify it, but more about the fact that why and how do feelings even exist? What function do they serve? Why we can feel them, in different magnitude and dimensions and categories at all, since we can build machines to do certain things that we are capable of, leaving out the feeling part? There is hardly no any doubt on the fact that all organisms feel something, which can be as easy as the feeling of pain and no pain. If the author cannot contradict this simple fact, then whatever attempts on proving the non existence of the hard problem that he is trying to make is not valid.
ReplyDeleteOne question that intrigues me is that, is zombie really possible? Can we really make something that can do everything we are able to but do not have feelings, or certain things require feelings to be conducted? As the technology on AI keeps enhancing, one day we might encounter a point where feelings are necessary for robots to achieve cognition, and it would be intriguing to figure out what they are and why they need feelings to be enabled. Will the hard problem be solved by then?
Feelings, for the most part, are an evolutionary advantage. If you look at where we process the most basic senses, like smell and sound, they are first processed in the old parts of cortex (the limbic system) which helps us to know that they evolved very early on in evolution. Because our evolutionary ancestors had an advantage over others when they were able to hear or smell a predator or dangerous situation before it attacked, they survived over those who lacked these sensory traits. So, the reason we have these senses is not a question, they have helped us evolutionarily. The real question is how these senses arise from the neural circuitry we possess.
Delete"That they correlate (feelings and function) is an interesting fact. Explaining how and why is another matter."
ReplyDeleteDennett did a decent job explaining heterophenomenology, which at the surface seemed to be an optimistic candidate to solve the hard problem. By translating feeling from a first-person point of view to the more objective and observable third-person, it gave us hope that we could get somewhere with the hard problem. But then we have to think about what the hard problem is exactly: it is about answering "how" and "why" someone feels. Even if we can objectify the feelings of someone using heterophenomenology, perhaps it could explain causality for the easy problem. But when we turn to the hard problem, this method completely fails to answer "why" the cognizer feels. Thus, we're back at square one - the hard problem remains unsolvable.
I agree that the hard problem remains unsolvable, although we could answer the “why” question of cognition on a surface level- it is because it is evolutionarily advantageous to us. We are still left with the how problem which we can’t currently answer and I doubt we will be able to in the near future.
DeleteThe article ‘Scientists discover the on-off switch for human consciousness deep within the brain’ contends that researchers have identified a brain region (the claustrum) that, when electrically stimulated, turns consciousness off. Hence, they conclude to have found the neural center for human consciousness.
ReplyDeleteHarnad refutes this conclusion by providing a distinction between awakeness and consciousness. The stimulation of the claustrum shuts off not just consciousness, but awakeness. Subjects cease not just to FEEL, but to speak/read/move etc. Consciousness, Harnad argues, is far more expansive than just awakeness and in order to locate a center causally responsible for consciousness, it would have to, when stimulated, shut off ONLY our feeling capacity, and maintain our doing capacity. (And if we are able to achieve this, we will have generated human zombies who do and not feel…).
Moreover, even if we are to localize a claustrum-like neural correlate responsible for ONLY our feeling capacity (consciousness), would this even answer the hard problem? We still require a causal mechanism for HOW and explanation for WHY such a brain region can generate feeling.
“Dear Dan. All bodily and molecular motions are there and available. And obviously 100% coupled with the feelings. The question (I hate to be repetitious) is how and why?”
ReplyDeleteHarnad's article explicitly addresses the ‘hard problem’ of consciousness, which asks how and why we (or any organism) feels. The article explains that correlations between our brain and feeling (generated by heterophenomenology) are not sufficient to answer the question; rather we need a causal mechanism that explains how and why we have these correlated feelings.
“In other words, the Hard Problem is simply explaining how and why we are not Zombies! (This does not require us to believe that Zombies are possible. Maybe they are, maybe they are not. Nolo contendere. Explaining how/why they are impossible would be even harder than the hard problem. I'll settle for just an explanation of how/why they are not actual in our own case.)”
ReplyDeleteI don’t understand how explaining how/why zombies are impossible would be harder than the hard problem. Wouldn’t showing how and why we are not zombies be enough? If we prove we cannot be zombies and how and why we have feelings, then there cannot be zombies in the first place unless we again doubt that others are feeling the same as we are, once again bringing the other minds problem. Once we know why we feel, we should know why it would be impossible for us not to feel and so zombies should be out of the question entirely.
“How/why does it feel like something to have (or to be!) certain functional powers? Although that sounds superficially like asking "How/why does gravity pull?" it isn't, because pulling is gravity, but feeling is not doing.”
ReplyDeleteThis statement got me thinking… Why exactly are we saying that feeling isn’t doing? Feeling is closely related to doing as we have discussed it in this course. Feelings are generated from our doing, and whenever we don’t do (ex. in delta sleep) then we don’t feel either. But that’s pretty much all we know about “feeling”, and since we haven’t resolve the how and why around it, why then are we excluding the possibility that these two things may actually be the same…
Emma I think that this is a very interesting hypothesis! However I'd like to add that saying feeling is a byproduct of doing depends on how we define doing. In certain cases that is true, the classic example being the ability to have felt beliefs, those are the result of being exposed to and processing things and then "feeling" something about that perceived construct. That being said there are certain cases where feeling seems to precede doing, for example feeling hungry makes you go eat. Then again the feeling of hunger is triggered by numerous biological factors, so we need to make the distinction between conscious and unconscious doing, and figure out which one we would consider when discussing the roots of feeling.
DeleteThe big issue with Dennett is that he is trying to prove an argument while using the wrong reasoning and explanations to prove that argument. He is arguing for computationalism but not addressing the issue of feelings or even the hard problem, instead he is trying to essentially discuss how zombies would not have feelings and heterphenmenology (3rd person science) when the real issue at hand is why and how we feel. I think prof. Harnad makes this very clear throughout his commentary of Dennetts paper and tries to shift the issue back to how and why we feel.
ReplyDelete“But who cares about ‘beliefs’ (which, as we know, can be ascribed to books and computers if we like)? What we wanted to know about was feelings. Not beliefs about feelings, but feelings. What is their functional role in a heterophenomenological theory?”
ReplyDeleteAfter reading Dennet’s article (10a), I found myself caught up in Dennet’s notion of beliefs (and more specifically, false positive/negative beliefs). Harnad now highlights a significant limitation in Dennet’s article, as he is transfixed on beliefs not feelings – but what we are interested in is the role of FEELINGS in heterophenomenology. Beliefs it appears are rather a component of feelings, as it FEELS like something to believe (or not believe) something… Therefore, was Dennet’s scope too narrow?
“Why not pick an open-and-shut case like feeling or not feeling warm-now? Not thermosensitivity or thermoregulation or thermolucution. Not being warm now (that's function); not being ready to say an instant later 'well, maybe I didn't feel warm then after all, just a little tense' etc.”
ReplyDelete“We are not interested in whether your toothache was real or psychosomatic, or even if your tooth was hallucinated, nor in the conditions under which these various things may or may not happen or be predicted. We are interested in how/why they feel like anything at all.”
In a comment on Dennet’s article (10a), I highlighted my struggle to understand how heterophenomenology would remain neutral when the first- (subjective/felt) and third-person (objective/unfelt) data don’t always align. At times, the subject’s personal cognitive state and feelings are uncorrelated to the context, and in these situations, even when the observable and the context points the other way, one cannot doubt that entity still has the feeling. In the present paper, Harnad provides several examples that better illuminate this point. For example, feeling warm – even if there is no increase in body temperature, or it is -20 outside etc., one cannot deny that the individual still ‘feels’ warm. Similarly, with respect to the feeling of a toothache, it doesn’t matter the function (i.e., whether the pain is from a rotten tooth or rather from referred pain or simply psychosomatic). These examples reminded me of phantom limb patients, where the context surely doesn’t correlate with the feeling – no physical limb is present and yet pain/sensation is still felt. The bottom line here appears to be that regardless of the correlation (or lack thereof) of mental states to objective and observable contexts and causes, we cannot deny that the individual has this feeling. Now, the problem is to determine HOW and WHY they feel this.
Just as Dennett focuses on the correlational/behavioural data of the feelings and not the causal/functional explanation, finding the claustrum as the "seat of consciousness" gives us a location in the brain where feelings might originate or a source of integration of all areas of the brain, but tells us nothing about how or why feelings exist, or how or why it feels like something to have this integration, to think, to sense, to do, etc. It was emphasized many times in the Harnad paper that the focus of science should be on the hard problem and the hard problem alone - how and why do organisms feel? This question is impossible to answer through objective data, since the subjective experience is the only one that matters or that can be "right" when it comes to feeling. From what I gather, Dennett is more concerned with people's perception of their feelings (arguing that the responses are bracketed by neutrality), rather than their feelings alone. But it feels like something to feel something, or to perceive what you are feeling. This brings us back to the ultimate, end all, and insoluble question of finding a causal mechanism or explanation for how and why organisms feel.
ReplyDeleteDr. Harnad makes a number of crucial points about Dennett’s paper on heterophenomenology. He articulates, over and over again, the absence of how/why explanations for feeling in Dennett’s plan to empirically address the hard problem. This certainly echoed my own thoughts, and truly made me question the value of the research program that Dennett is suggesting in terms of scientific merit. I found especially valuable the point about how at some point we may be able to reliably predict what someone is thinking (feeling?) from fMRI. But this does not get us any closer to why we have these feelings (e.g. their Darwininian function), or how this particular brain activation is causally related to feeling. I also shared Dr. Harnad’s skepticism of the zombie example. How is a feasible that there could exist a being with the exact molecular make up as us that does not feel? Also, should we truly believe that a being like this exists, even at the t3 level, would we feel ok kicking it? After all, it does not attach meaning to pain, it only reacts as we would to pain. But how could you really be sure? I believe that Dr. Harnad’s answer is that you can’t. You simply cannot answer how or why we are any different from this Zombie, from Isaure. I certainly wouldn’t kick Isaure.
ReplyDeleteI thought this article was interesting to read as a response to Dennett. I like how Dr. Harnad clarified that he is neither team A or B but actually his own team C because he believes that the hard problem cannot be solved and that the Turing test is more of an explanation than the science of consciousness. However, the main focus of the paper is how and why. These two questions are asked repeatedly throughout this paper. I think this repetition is necessary to emphasize its importance and role within the hard problem and why its so insolvable.
ReplyDeleteAdditionally, I feel like it was important to understand the following:
"You are not authoritative about what is happening in you, but only about what seems to be happening in you,"
"Let me put it another way: seems = feels-like... What matters is that anything is felt at all! (How/why?)"
From this its clear that the hard problem is not why and how we feel a certain way after something happens but how and why we feel at all.
This article helped me really understand what heterophenomenology is. To my understanding, heterophenomenology attempts to combine all the measurable aspects of feeling, aka the physiology when one is feeling something, the brain activity, the facial expressions, the verbal reports etc, the idea being one can explain the hard problem if we can get a global understanding of what happens when we feel. Yet, the problem is in the method itself, which is, all you end up answering is the what happens when we feel, so all you end up getting is correlations and predications, but never getting to the direct question (how/why is anything felt at all), which penetrates the behavioral and neural processes that will have their functional and adaptive causal explanations but never getting an explanation for their felt correlates!
ReplyDeleteNote, I loved the breakdown of the term emotional reactions: You have reactions and you have correlated feelings. That was such a great way of framing the feeling/doing problem. Likewise the distinction between cerebral and mental.
Dr. Harnad you bring up a point where you say the Freudian type of unconscious mind is superfluous among other adjectives, I don't have an opinion on this but would like from your perspective a distinction between the unconscious and the subconscious.
Also when you talk about beliefs, you say they can be ascribed to computers and books, what did you mean by that, because you go as far as to say one can make the argument that walls have belief, I'm not quite sure how this is possible (is this the other minds problem)? Were you referring to the concept of an as-if belief?
Overall I found the counter arguments to Dennett, very strong, clear, and with an ever present thesis (none of this answers the how/why, just a bunch of side-stepping jargon). The main difference for me between Dennett and Dr. Harnad lies in that the latter argues the hard problem is insoluble, whilst Dennett obviously doesn't as he attempts to apply normal science methods to answer the "what" in hopes that it leads to the how/why.
"You are describing an empirical psychophysical data-gathering paradigm for getting all the measurable ("3rd person") correlates of feelings."
ReplyDeleteThis paper successfully addressed misconceptions about the illustrative 'zombie', as well as underlining over and over again what the real issue at hand is.
I enjoyed this response paper in particular because I felt that it explicated the intuitions we have about consciousness and our mental lives, while still insisting on what information we need to answer the hard problem: why people feel and what the causal mechanism of feeling is.
In particular, it shows step by step how Dennett's article on heterophenomenology sets out a goal that is not meaningless, but nonetheless simply describes the feelings and experiences we have without explaining how we got there.
I am not sure that I can agree 100% that that hard problem is insoluble, nor that heterophenomenology and using objective scientific measures of correlation won't provide any insight or direction.
This comment has been removed by the author.
ReplyDeleteI agree with the argument made in this paper in that there needs to be a distinction made between thoughts and felt thoughts. Felt thoughts (understanding, etc.) should be equivocal to feelings; but, then this brings up the question: can you have thoughts without feeling? And can you have feelings without thoughts? And how does this relate to the hard problem and understanding the how and why of feeling? If we understand how thoughts come about, can we then link that as some sort of explanation of how feelings come about? I would be interested in investigating just how much these two phenomena are related and if we can learn more about one when learning about the other.
ReplyDelete"The (hypothetical) Zombie does not "fervently" anything, because he does not feel! He only behaves in a way that is interpretable (by us) as if he felt. If there can indeed be such a Zombie, the how/why difference under discussion would be that difference between actually feeling and merely functioning-as-if-feeling. If there cannot be such a Zombie, then you need to explain, causally/functionally, exactly how/why there cannot."
ReplyDeleteProfessor Harnad does a great job of putting into words many of the things that I felt (irony intended) while reading Dennett's article. By breaking it down piece by piece, it becomes evident exactly what Dennett was arguing and why he misses the point -- answering the hard question -- completely. I think that using the explanation above about the Zombie question helps to hone in on the argument being made by Harnad. All details aside, the arguments keep coming back to the how/why question. If there CAN be such a zombie, then we need to understand the difference between feeling and "functioning-as-if-feeling". If the cannot be, then we have to answer how/why not! Up until this point, there seems to have been a fair amount of confusion with regards to the Zombie problem, and Harnad essentially clears up the confusion in this sentence. Furthermore, I found this article to be particularly refreshing in that Professor Harnad doesn't to pretend to know the answers to things that cannot be answered. Much unlike Dennett who manages to beat around the bush with a plethora of examples, though coming no closer to the core of the question.
I enjoyed how this paper combed through each of Dennet's claims and broke them down into digestible chunks, while identifying and rebutting the flaws. In a similar fashion, I have my own questions
ReplyDelete1) Relating back to the symbol grounding problem, how do we define the word "thought" and who is to say what the right definition was? Are all thoughts "felt thoughts" or can they be mutually exclusive? Conversely, is it possible that there is spectrum of "thoughtiness" ranging from thoughts-felt thoughts-feelings? It seems to me that not every thought consciously addresses feeling, like if you were thinking your way through a math problem you wouldn't be feeling an emotion as much as computing an equation.
2) What are the parallels between heterophenomenology and the Turing test? It seems to me that the Turing Test is a more elevated, sophisticated concrete example of heterophenomenology.
3) How can we ever prove if something is truly causal beyond the slightest hint of doubt that something else may be confounding the conclusion? It's obvious that we need to go beyond answering just the what and when, but is it even possible to create a mechanism precise enough to determine one answer to how and why?
Cognition is also possible below JND's. Specifically, in backward masking paradigms, stimuli are presented to participants too quickly to actually be consciously recognized (below the level of their awareness) and replaced with another image. Though participants don’t report seeing the first image, it has been shown that it can affect their behavior (speed up word priming, etc). Because of these findings, how does that complicate the importance of JNDs as it relates to the subjective experience of feeling for cognition?
The hard problem asks how and why we can feel and is most likely unanswerable. It is unanswerable because to answer how and why we feel, would mean that “feeling” would have to act as some kind of category, with “not feeling” as its complement. This seems kind of like the Lelek (?) category from class because we can’t really separate feeling from our daily life, making it impossible to know what “not feeling” would be like, which limits us significantly.
ReplyDeleteIn 10a, I asked a question about T5s and psychopaths. It seems to me answering the how and why we could feel, would be similar to examining why T5 robots couldn’t be zombies. I was wondering if it is possible to learn more about this question by examining people that are essentially devoid of feelings or is this essentially just the same thing that we experience except the converse? It seems curious to me because although they don’t necessarily have emotions the ways in which others do, they do interact (touch, eat, etc.) with the world, which gives them a unique perspective.
Harnad’s comments to passages of Dennett’s text helped me better understand Dennett’s ideas and the fact that they avoid feeling. Harnad repeatedly expresses that the question Dennett and other cognitive scientists should be focusing on is the hard problem: how and why we are not zombies or, in other words, how and why we feel.
ReplyDeleteDennett’s heterophenomenology is not a method that can get us closer to solving this problem (if it is even soluble at all – Harnad claims it is not), because all it can do is “inferring and describing feelings” rather than explaining how and why they arise. Therefore, heterophenomenology can only contribute to solving the easy problem.
Hi Myriam!
DeleteEnjoyed your response. It touched on a lot of questions and hesitations I had while reading the Dennet piece. Would you not agree that since all heterophenomenology does is "infer and describe feelings", and since it does so second-hand to the feeler themselves, that it merely continues reiterate what we already know about thinking and feelings. It doesn't even take into consideration the other-minds problem, as far as I can see.
I would go as far as to deem heterophenomenology as useless after reading Harnad's piece. Like the discovery of the mirror neurons, it put a name to something we knew was already present, and claims that this is somehow the answer to cognitive science research on the feeling brain. Just like the discovery of the mirror neurons, however, it fails to establish cause, and leaves a highly incomplete picture about what it is the "feel" or "experience".
I engaged one of my more argumentative friends in a conversation about the hard problem a few days ago. She comes from an anatomy/cell biology background and is skeptical of philosophical approaches to biology, but I felt up for the challenge. I tried to do justice to explaining the main arguments of the course in the three hours we spent talking, but either I didn’t summarize properly (which is fair considering the material is presented over the course of four months not three hours, and I was arguing with the lay person who hadn’t read any of the articles/arguments we were given) OR I just blindly drank the cool-aid and didn’t actually understand the course at all enough to defend it. Anyway, she didn’t agree that the hard problem was really that hard. Her argument was that feeling is central to our success as mammals/other feeling organisms, and that their purpose is to compress vast amounts of data from sensory input etc. and distribute it quickly instead of us having to process each separate thing independently. I said that the compression could theoretically be achieved without being felt, and that I certainly don’t disagree that feelings are important, but the problem is explaining them. Talking about zombies honestly didn’t get us anywhere because she doesn’t believe it’s possible to have functing humans void of feelings. Her argument was that a human would not be able to do anything without feelings (like assigning meaning) because the reason we do most things (like thinking) is thanks to feelings.
ReplyDeleteThis is the argument we kept snagging on, which to me seems like the most acceptable one:
Delete“Don't say "The function of the feeling is to draw his attention to..." -- because the reply will always be: Why/how does any of that "drawing" have to be a felt drawing, rather than just a drawing? Why/how does any of that attention have to be a felt attention, not just a "selective processing" etc. This is where the ineluctable difficulty lies, not in a hybrid data-base or in authenticating correlations/predictions between functional data and feelings.) “
The point when the tide of the debate started turning was I brought up that trees/plants are capable of doing extraordinary things without feeling at all (we hope). So why aren’t we like plants, just reacting to environmental stimuli/biological imperatives, or like the quote above says, “drawing” unfelt attention? Why does it feel like anything when we do things? Once we got on the same page about what I meant by feeling (sensing/functing could be done without feeling and saying that you need feelings to think isn’t an explanation for feelings) we started to converge more on understanding why explaining the how/why of feelings is so difficult.
The reason I’m recounting this debate is because I feel like at this point in the course we’ve entered an echo chamber, and there’s really no interesting way for me to discuss this article without just saying “yeah, the hard problem’s quite a doozy, eh?” I’ve spent days mulling over this response and I finally came to the conclusion that I should stretch my kitten legs and see if I can defend what I believe. I was moderately successful in that by the end she didn’t outright disagree with what I had to say, and I’d like to believe that she wasn’t just humoring me in the end.
This comment has been removed by the author.
Delete
DeleteThanks for the share, I like your personal reflection and I found it interesting as it is always difficult to make arguments when people argue from radically different perspectives. I agree with this " the compression could theoretically be achieved without being felt, and that I certainly don’t disagree that feelings are important, but the problem is explaining them." As we have said in the course, what, inter alia, makes the hard problem so hard is because we could theoretically take out feeling and still do everything that we can do, and yet, conscious states are felt states so cognition is in part and to some degree, feeling.
Could you remind me what is the basis of the argument that plants/trees do not feel ? Of course, there is the Problem of Other People's Minds, but more ?
I found Harnad’s distinction between thoughts as “internal goings-on that generate certain outputs in response to certain inputs” and thoughts as “feelings” helpful in deepening my understanding of the hard problem. Indeed, it feels like something to think, thus Turing’s question regarding how to build a robot with thoughts becomes very complex since it involves feelings, which we do not yet understand causally.
ReplyDeleteIn order to build a robot with thoughts, therefore, we would first need to solve the hard problem (why and how we feel), and this is one of the main ideas that Harnad tries to communicate to Dennett, who does not seem to grasp what the hard problem is.
Although Harnad believes we should attempt to solve the hard problem, he believes that it is ultimately insoluble. Therefore, even if we managed to build robots with thoughts as “feelings” as opposed to thoughts as simply “internal goings-on that generate certain outputs in response to certain inputs”, we would not understand how and why feelings arise in them just like we do not understand how and why feelings arise in us.
I agree with what Professor Harnard has to say in this paper. When I was reading the original one I got really hung up on these "unfelt" experiences because Dennet uses them as such an integral part in his explanation but to me they don't mean anything or help at all in explaining how and why we feel.
ReplyDeleteIn another class of mine we were discussing if we are experts over our own experiences or not and someone argued that we are not because there is evidence for these unfelt feelings or the capability of hypnosis which suggests someone else has control over your experiences. They proposed one study that says if you are high up on a bridge and encounter someone you find attractive, you will find them more attractive on this bridge because of the adrenaline from the height and therefore you are fooled by your experiences and not an expert. This argument is flawed in a similar way as the "unfelt" feelings argument Dennet proposes because it does not matter if you are misunderstanding your feelings or experiences, the important part is that you are feeling and experiencing.
What most people are missing, and which you have made a point of repeating, is that we still don't know why or how we feel. Is there actually a purpose to us feeling? In some sense, every type of species feels. So it must be a matter of organisms. But that still doesn't answer anything. Great! all animals feel, but why? and how? Why is it that my pencil can't feel anything but my cat can? Is it a matter of cellular formation? Genetics? Simple luck? In the same way, what would happen if we created something that could not feel? Because of the OMP, it would be impossible for us to know if they did not feel because if we pinched them and they did not react, we could only infer it based on our expected results. Personally I think the HP is such an abstract concept that we don't have the capacities to answer or understand this question at this time, but it would be nice to see that someone else figured it out. But we'd probably have gone extinct by then...
ReplyDeleteDon't say "The function of the feeling is to draw his attention to..." -- because the reply will always be: Why/how does any of that "drawing" have to be a felt drawing, rather than just a drawing? Why/how does any of that attention have to be a felt attention, not just a "selective processing" etc. This is where the ineluctable difficulty lies, not in a hybrid data-base or in authenticating correlations/predictions between functional data and feelings.
ReplyDeleteI want to highlight this passage because it captures a couple important points:
(1) I initially have the thought that this is what feelings were for. I thought it was obvious and couldn’t understand why the hard problem was so hard. For me, this is a reformulation of the hard problem. Instead of why/how do we feel anything at all, this poses the question as a sort of reply. Sometimes it’s helpful to understand the problem when it’s formulated in different ways.
(2) It reinforces why the hard problem is so hard. Once we move past heterophenomenology just providing us with correlates, we understand that the hard problem is, in fact, really hard. We can solve the easy problem, even to T5, and still not be any closer to solving the hard problem. Not only does heterophenomenology only provides us with correlates, but it’s attempting to predict whatever word Dennet calls “feelings”. But this is not the hard problem at all. We know that we feel, and we have a good idea what other people feel, too (ask them!), but knowing and predicting feelings is not a problem for us. The hard problem is why and how we feel what we feel at all.
Additionally, Harnad’s response illuminates how the other-minds problem and the “zombie hunch” are connected to the hard problem. It seems that Dennett was confusing the hard problem with the other-minds problem. And the zombie hunch only exists because we can’t explain the hard problem.
"We are not interested in whether your toothache was real or psychosomatic, or even if your tooth was hallucinated, nor in the conditions under which these various things may or may not happen or be predicted.” Stevan thinks that the distinction between real or hallucinated feelings is irrelevant because we should be concerned with how/why we feel in the first place. Originally found this point of view cumbersome, because feelings, real, hallucinated, or otherwise, must have some source. This source (I thought) certainly wasn’t the same for real and hallucinated feelings because real feelings must have originated from some sensory experience (the uncovering of the source attempts at answering the where/when). However, my opinion has changed. I realized that regardless of the realness of the feeling, the source of any feeling must be something physical. To insist that the origin of hallucinated feelings is something different from the origins of real feelings would be to assume a dualist perspective. However, refusing to accept a dualist perspective further complicates the attempt to understand feeling because there are certainly physical ways to experience feeling that can be either sensory or non-sensory in origin.
ReplyDelete"But until further notice, the only one who can actually feel the feelings themselves is the party of the first apart, the feeler (1st person)"
ReplyDeleteIt seems like heterophenomenology is more of a reiteration of what we already know to be true that an endeavour that will yield significant new results. We already know what it is to think, and feel. We already know that human sensory systems have limitations not accessible by conscious thought. The reports of heterophenomenology only give us second-hand access to experiences that we have, as humans, already experienced first-hand. It seems almost circular to use heterophenomenology for research introspection is more direct and has also been rejected as a cognitive science practice.