I read about qualia from time to time in books or on web sites dealing with consciousness for the most part. The general consensus (as far as I can tell) is that, if one even recognizes qualia as a legitimate phenomenon, it’s somewhat difficult to even pin down a consistent definition of qualia, let alone make a detailed scientific study of it. There’s a decent amount of literature on the subject, but I’m primarily interested in a scientific (as opposed to purely philosophical) analysis and what, if any, applications could be applied to attempts at creating artificial intelligence/consciousness. Any insight that Dopers have on the subject is appreciated, and any recommendations for further reading would also be welcome. Thanks!
The topic may be better suited for Great Debates. Certainly, Great Debates have occurred on the topic, which strongly hints that there aren’t any useful “factual answers” available. The Wikipedia article, for instance, refers to several arguments regarding the concept, but doesn’t have a whole lot of real concrete answers.
“Qualia” (singular: “quale”, pronounced qhah-lay) is a philosophical term of art, and this means that (as you say) people, the experts, disagree about what it means or even if it refers to anything real at all. You are not going to find a “scientific” account or study of qualia that is not also philosophical. In the current state of the study of consciousness, science and philosophy are inextricably intertwined, and both people trained as scientists and people trained as philosophers bring very valuable and relevant skills to the table. Scientists bring new data and observations, but philosophers bring a much more informed awareness of the (long and convoluted) history of the relevant problems and their ramifications, and of the sorts of conceptual pitfalls that surround it (and that unwary scientists all too often fall into).
There are certainly people who are scientists, in the sense that they have laboratories and do experiments, who have things to say about qualia, and may even claim that they have experimental data that bear upon the matter, but that does not mean that they are not actually doing philosophy (often, though not always, rather badly) and thereby making highly questionable and controversial assumptions (which they may or may not be aware that they are making) in the interpretation of any actual empirical results. If they say anything positive about qualia, they inevitably are making questionable and controversial assumptions. If they do not show a sensitive awareness of that, they are incompetents who are probably doing more to muddy the waters than anything else.
The best scientists working on consciousness, these days, are very philosophically sophisticated, and listen to what the philosophers have to say. (That does not mean that they always get the philosophical issues right, of course, because it is of the nature of philosophical issues that nobody knows, there is no general agreement, about what is right.) By the same token, the best philosophers working in the area keep themselves very well informed about the latest scientific findings and theories. It is a truly interdisciplinary enterprise. - Actually, as things are today, even most of the second-rate philosophers in the field make an effort to follow the relevant scientific literature closely. Unfortunately, the reverse cannot always be said of the more second-rate scientists, whose understanding of philosophical issues (and what philosophy itself is) often seems to be quite superficial. (Forty years or so ago, shortly before I got in to the field, this situation was somewhat reversed. Many philosophers of mind, back then, were woefully ignorant and even contemptuous of neurological and psychological science, but things have definitely changed in that regard.)
To come back to the concept of qualia specifically, though, they are one aspect of the study of consciousness that are currently still much more the concern of philosophers than of scientists. The concept is closely bound up with the so called “hard problem” of consciousness, which some people (some scientists and some philosophers alike) think marks a sort of limit on the extent of possible scientific understanding of the mind. That is to say, some (by no means all) people in the field (both some scientists, with labs, and some philosophers, without labs) think that empirical results, and our (slowly) increasing understanding of the brain and of cognition, will never be able to cast much light on this aspect of the problem of consciousness, and that it either cannot be solved, or can only be solved by some sort of drastic revision of some of our fundamental concepts. This means that much of scientific work on consciousness that is done these days, steers away from the issues surrounding qualia to concentrate on other aspects of consciousness that seem more likely to be empirically tractable.
– Well, I see I have said quite a lot about the relationship between science and philosophy in the contemporary study of consciousness without having said very much about the concept of qualia itself. However, the point is that, as things stand, you are really not going to get a good understanding of the concept of the concept of qualia from the scientific literature. You won’t find very much by scientists on the topic, and what you do find is likely to be both superficial and biased (taking into account just one point of view, or a very small subset of the wide range of points of view that there are on the matter).If you want to understand this one, you have got to go to the philosophers and try to understand the multi-sided and very complex debate that is going on amongst them. At least they will be aware of its multi-sidedness. Although there may be exceptions, what scientist-generated literature there is on the topic tends to come from scientists who are either too dumb or too inexperienced to understand just what a minefield it is. The smarter scientists mostly leave this one to the philosophers. Maybe, if the philosophers are eventually able to reach a bit more of a consensus than they have right now on whether qualia exist, and what they are like if they do, then the topic will be ripe for empirical investigation.
Anyway, the moral is that if you want to understand the problem (or problems) of qualia, and maybe, one day, be able to contribute to a scientific solution of it, you must look to the philosophical literature, not the scientific (although there is some overlap these days, with lab-less philosophers occasionally publishing in scientific journals, and people with labs occasionally showing up in philosophical ones). A good place to start would be here. I do not say, and do not think, that that article is unbiased – it is impossible to be truly unbiased about such poorly understood issues, and, in particular, the article gives rather short-shrift to those who think that qualia just do not exist – it it does make the attempt to provide a true overview of the debate, and is likely to be a lot more aware of, and careful about, its own biases than anything you are likely to find written on the subject by a scientist, even a really smart one (and it will certainly be more reliable than Wikipedia).
That is because it is a philosophical issue, and it is of the nature of philosophical issues that people disagree in fundamental ways about them. It does not follow, however, that there are no facts about that philosophical debate and the positions within it that have proven to be sustainable. If the OP is prepared to work at it (it will be hard work) he/she can learn a lot more about the issue from the professional philosophical literature than he/she can learn from the amateur philosophical debate that that will develop if the thread goes to Great Debates.
The question seems to be one about the literature on qualia, not about what we Dopers might think about the issues. There is a real, professional and scholarly literature, and I have given what should be an excellent entry point to it. The Stanford Encyclopedia of Philosophy, unlike Wikipedia, is highly regarded as a scholarly resource, with articles written by top experts.
Not at the “What’s the Straight Dope” level. The OP might as well have asked “What’s the Straight Dope about God” or Free Will or The Soul or the like. I mentioned Wikipedia; you mentioned the Stanford Encyclopedia of Philosophy. Both are decent starting points. But the SDMB General Questions forum, I suggest (placidly) isn’t, exactly. The best we can do (and have done) is point elsewhere.
But this prompts me to a serious question: there is a Scientific Method. Is there a Philosophic Method?
The most scientific theory of qualia currently available, at least in so far as I am aware, is Giulio Tononi’s work on integrated information. The basic idea is that consciousness is, or is related to, the amount of information generated by a system that cannot be accounted for by the information generated by its parts, i.e. the information the system generates over and above the information all its parts considered separately generate.
Tononi’s classic example is that of a photodiode: it can distinguish between light and dark, and thus, generates one bit of information. You, on the other hand, can distinguish between light, dark, and any possible configuration of visual stimuli. So by making the judgement ‘it is light’, if that is indeed all the visual stimulation you receive, you are not distinguishing the state of your visual input from ‘it is dark’, but, say, from any possible frame of any possible film you could tell apart from one another. That this is related to consciousness, Tononi says, is due to the way you integrate information: consider a lattice of one million photodiodes. This could distinguish between a great number of states (2[sup]1000000[/sup]), but it would function just as efficiently if you cut it apart, into one million individual photocells: it does not generate any information over and above that generated by its parts. But if there are certain interrelationships between the parts, i.e. if the state of one part depends in a certain way on the (possibly prior) state of another, then the ‘cutting apart’ evidently results in a set of separate system whose combined information production falls short of that of the whole system.
Tononi now stipulates that this extra, ‘integrated’ information, which he calls phi, is, or at least is a measure of, conscious, subjective experience. This has one immediate consequence which most other proposals (to my knowledge) lack: by studying the functional interrelations of some system, we could, from the outside, judge whether or not it has conscious experience, just by evaluating how much information it integrates. In this sense, it makes subjective experience amenable to scientific study.
As for qualia, Tononi defines what he calls ‘Q-space’, which has an axis for every possible state of the system (and is thus extremely high-dimensional even for comparatively simple systems). A point in this space is then a probability distribution over the states of the system, i.e. it gives the probabilities that the system is in a certain state. A quale is now a shape in this space: roughly, starting from the point of maximum uncertainty about the state of the system, it tracks the effect of sharpening the knowledge about the system’s state in specific ways.
The resulting theory is a bit mathematical, as it is essentially concerned with relative entropies of different probability distributions; if you’re willing to clear that hurdle, however, the resulting framework is not entirely uncompelling (you can judge for yourself by reading Tononi’s account of the theory as it stands so far here).
Of course, whether the concept of integrated information does the philosophical heavy lifting it set out to do is a matter for further debate. Essentially, the ‘hard problem’, to the extent it is considered at all, is solved by stipulation: phenomenology, the ‘what-it’s-like’-ness of experience, is just something that is brutely attributed to integrated information, which I think makes the theory essentially a Chalmers-like dual aspect property-dualistic solution. Whether or not one is happy with that kind of a solution is, I suppose, a matter of taste.
Each animal builds and stores in its brain a model of the world outside. The world impinges on the animal, and the animal updates the model with its interpretation of the meaning of each “touch”.
Each animal’s model, brain, body, etc. is different so we can be sure that the model is full of qualia. Important note: these qualia do not exist in the world outside. They only exist in feeble models us animals construct to try to make the most of the world.
But there is also the question whether any independent reality exists. If it does, then the qualia described above are the only kind. But if no independent reality exists, then there are also the qualia caused by differing observations of the (supposedly) same phenomena. They are different.
I don’t see how that falls out of (what you have said about) the theory, or at least how the theory is committed to such a view. It seems to me one could say “qualia just is (i.e. is identical to) such and so functional process, and that functional process is integration of information” and be consistent with the theory.
Just consider the following question: Is my experience of the color red the “same” as yours? Does the question even have a meaning? It certainly doesn’t have an answer. We can say that, in some sense, red stands out more than the other colors, that blue or green is more restful, etc. but this just nibbles around the question. Only by entering in your mind could we answer it and, even then, I am not certain it would.
The problem with equating qualia with functional processes is the causal closure of the physical world: generally, for all physical effects, you can find a physical cause, in such a way that in the causal story, there is no need to refer to subjective mental states, or qualia, at all. Tononi’s theory merely asserts that ‘integrated information’ has a phenomenal aspect, without arguing why that should be the case; it seems entirely conceivable to have a system generating a high amount of phi, i.e. having a high relative entropy between the probability distribution over its possible states when considered as a unified system versus the probability distribution over its possible states when considered as a collection of parts, without that necessarily being accompanied by there being something it is like to be that system (at least to me). So it certainly seems possible to integrate information without this being accompanied by subjective feels. Thus having a phenomenal feel to it must be a nonreducible aspect of information integration.
Does Tooni, or anyone else in the literature, make this argument?
What argument exactly? That qualia cannot be explained functionally? That’s standard issue, and should be in every text on the subject (e.g. here on the SEP).
Applied specifically to Tononi’s theory? I’m not sure; I got his book for Christmas, so if I find anything there, I’ll let you know. But it doesn’t seem essentially different from other accounts to which the argument has been applied.
Applied to Tooni’s theory. In terms of the-things-you-don’t-want-to-have-qualia, Tooni’s theory is going to say something specific about such things; possibly multiplying the problem (for functionalism). Conversely, if one takes “stipulation” seriously, biting the bullet and accepting the multiplicity of qualia may end up being unproblematic, if philosophically unsatisfying.
I’m not sure what you’re saying, but again, I don’t see why the usual arguments against functionalist explanations of qualia should cease to apply if the specific function to be performed is the integration of information. Consider an inverted spectrum argument: if some quale (a shape in Q-space) is associated with a red experience, and another with a green experience, it seems entirely conceivable that these shapes in Q-space could be differently associated. But then, the nature of these qualia does not rest in their function, and there must be additional facts that fix the precise content of subjective experience.
I’m not sure what “cease to apply” means here. I’ve not said the arguments cease to apply, just that how they apply might differ (e.g. be more or less persuasive), or how one responds to them might differ.
Take two competing physical theories on qualia; they may say different things about which things have qualia. And so the “you have too many things having qualia” argument may be more or less persuasive, because what falls under the scope of the theories will differ.
Or take different definitions of “qualia”. The “you have too many things” argument may be more or less persuasive depending on the definition of qualia it is directed against. And here, qualia was said to be defined by stipulation; if I take “stipulation” seriously, I can’t see how the “too many things” argument could be effective at all. (So, does Tooni say what qualia is besides the mathematical definition? Does Tooni say “qualia” is the “what-it-is-to-be-likeness”, for example?)
For some arguments it may make no difference.
Thank you all for the input. It does seem that this subject might be better suited for GD after all, if the Mods would like it moved there, it seems to make sense to do so.
(A quick disclaimer: my own knowledge of the subject is very limited. My professional training is in visual art, and any education I have beyond that which is provided by standard requirement courses from college came from independent study undertaken to indulge my curiosity about things. I have also had the privilege of talking to a few neurologists, psychologists, and physicists about various subjects related to some of these things, though I suspect that most of them err on the side of being good scientists who dabble in philosophy. I come from a position of an artist with an interest in science and an armchair philosopher’s understanding of some basic ideas. So you will probably take whatever I have to say about the subject with the whole shaker of salt!)
Perhaps if I give more detail about the questions I have, and the current perspective I have, it will be easier for me to receive correction in the areas that I have wrong: currently I probably fall into the camp that believes that there is likely no mind/body split. The mind may be an emergent phenomenon, but I can’t find any evidence that minds exist independent of brains and bodies, so my approach to understanding qualia is to look at what is known about physiological responses associated with experiencing qualia, and worry about emergent phenomenae later.
For example: we’re fairly certain that specific hormonal and chemical processes in the body and brain occur in response to stimuli, externally and internally. Thoughts can produce arousal states as much as tactile stimuli. As I understand it, when we experience an event that triggers the release of oxytocin, for example, there appears to be an evolutionary function in that the organism experiencing this response feels a pleasurable sensation, a positive reinforcement to continue or repeat the triggering behavior. Part of this sensation seems to include physical responses that promote good health, as well. Pursuit of the pleasurable sensations in turn may drive creativity in humans, as problem-solving is also beneficial from an evolutionary perspective, and creativity can often boil down to problem solving. Problem-solving requires the asking of questions and the pursuit of the means to answer them, thus increasing intelligence and knowledge. Nature would almost certainly select for the ability to experience qualia if that is the case, wouldn’t it? Conversely, if we associate “redness” with “heat” and primitive man made the same association, causing him to be cautious around burning embers, that is also a potential benefit of qualia.
This made me think that a key to discovering more about qualia would be to get a better understanding of the total physiological experience of qualia, including hormonal and chemical responses in conjunction with mental processes. I further thought that if a better understanding of those things were gained, it might better inform efforts to construct a more human-like model of artificial intelligence or even artificial consciousness. A computer that had artificial “hormonal” responses that would, say, encourage it to find more efficient ways of working or discourage it from indulging computer viruses. A computer that would not only be able to learn, but be driven to invent and ask questions on its own. An insatiably curious computer, if you will. Does this seem like a reasonable approach to anyone?
(My self-education about I.A. is even less developed than the other areas, so while the ones seriously studying such things may be like the proverbial blind men and the elephant, I’m more like the guy who got lost on the way to find the elephant!)
Of course, none of this even begins to address the problem of the conscious self and what it really means to experience something, so I know none of this is even scratching the surface. It has more to do with trying to include as much in the equation as seems relevant to the problem, by involving overall physiological response as well as purely mental processes, and trying to figure out the evolutionary advantage of the experience of qualia.
One of the points that it seems physics and neuropsychology converge on is the concept that reality, or at least the perception of reality, is a creation of the observer. This apparent convergence is especially interesting to me, and one of the things that I keep hoping to read more about from authors more learned about those subjects, but so far I haven’t had much satisfaction. Maybe I’m reading the wrong authors, or too many authors at once, or maybe I’m just trying to come up with a half-assed piecemeal theory of mind that encompasses too many fields. In any case, I really appreciate the input and look forward to reading more. Thanks.
Well, Cuckoorex, as you see this is a pretty big topic and you won’t get a neat GQ answer.
But it’s also a minefield for people new to the concept. IME most people when first thinking about subjective experience struggle to even appreciate the problem until there’s a “penny drop” moment*.
And some of those on the web who argue most vociferously that qualia are nonsense are pre-penny drop IMO. Not all, but some.
- e.g. I was at a meeting, and we were talking about a particular species of animal, and I happened to mention that this species has colour receptors covering the same EM range as our eyes, plus part of the UV spectrum. However, they took exception when I then referred to the species “seeing” UV; “this species might see RGB light but it must merely detect UV, since UV light is invisible”. I had to explain that colour is something that happens in minds as a result of a stimulus (or sometimes is generated without external stimuli: imagining, dreaming, hallucinating), it is not the same thing as the stimulus.
But assuming that, say, red, is actually “out there” is of course very natural to us.
I think the best gauge (in so far as there is any) for having grokked qualia or not is when you realize that you can, in principle, explain every action, thought (as refers to its content, not to the experience of having the thought), speech act, and every other form of behaviour in a completely gapless manner, but that none of these explanations refers to subjective feels, to the what-it’s-like-ness, to qualia at all. This is of course the heart of the zombie argument: a world is conceivable which, regarding the material, physical facts, is identical to ours, but in which there is no conscious experience.
That, and the (to most) utterly undeniable reality of our own conscious experience – the fact that obviously there is something it is like for me to be myself, or for me to have a red sensation – combines to truly drive home the formidable nature of the hard problem.
It’s my experience that qualia are a product of attention, not perception. For example, I usually listen to my iPod when I take my dog on long walks. I also often spend time thinking hard about phenomenology on these walks. I’ve noticed that when my attention is fully concentrated on my thoughts, the qualia associated with the music I’m listening to go away. I’m clearly still hearing the music in a mechanical sense – I would certainly be surprised if my iPod stopped playing – but I don’t *experience *hearing music. If I’m really into my thoughts I may even lose the qualia associated with the dog. I’m clearly going about the business of keeping him under control on his leash, but I have no experience of doing so. So its possible, for some of us at least, to dip in and out of “zombie mode” by directing our attention away from our sensory flow.
Inasmuch as there is really a scientific method, yes there is, and they are the same. Science and philosophy are really the same enterprise and rely on the same tools, reason and observation.