From “Quining Qualia”, qualia are supposed to be (says Dennett):
But I don’t find any of these to be necessary features of what I’m talking about when I refer to qualia.
Ineffable? But by saying “what things seem like when things seem red,” I’ve effed it.
Intrinsic? But I’m okay with the possibility that were someone other than me having an experience I’m having, the qualia of the experience might be different.
Private? But surely there’s a way to hook up brains together such that one person experiences the other’s visual field. It seems possible that there might be a way to tune the hookup such that the secondarily experienced visually fields really does have the same qualia as the primarily experienced one.
Directly accessible? Maybe not–maybe I have access to the qualia of experience only through awareness of the thing experienced, for example. In other words, perhaps I have experience, and then qualia come in as a kind of “overlay” placed over that experience–implying that I don’t get to access the qualia directly but instead access them via my access to whatever (real or illusory) objects I’m experiencing.
So if I’m not necessarily intending any of the above when I mention qualia, then what am I talking about? As I said when I effed it–how things seem when things seem red. It appears as though no number of facts about particle positions and momenta could allow me to derive the fact “this is how things seem when things seem red.” That fact seems to be logically independent of the physical facts. This suggests that there are non-physical facts, namely, facts about qualia. Of course, this could simply be a failure of imagination on my part–maybe there’s a way to derive the qualitative facts from the physical facts, and I just haven’t figured it out yet. I’d love to see some such derivation developed. But for the moment, it appears not just to be an unaccomplished goal but an impossible one.
What reasoning from physical facts should lead us to conclude that red things typiocally look the way things seem when things seem red, rather than the way things seem when things seem green?
That they are supposed to be non-physical, and yet have physical consequences. (If Mary learns something new, she might say, ‘Well hot damn, so that’s what red looks like!’, for example.) I realize that not everybody would view this as an inconsistency, but to me, a thing’s physical nature is embodied in its physical consequences: a table is physical because it scatters light in a certain way, offers resistance to touch, holds my laptop up against gravity, and so on. We only know the table through these physical consequences, and these physical consequences entail that the table itself is physical. Thus, qualia, at least to the extent that they have physical consequences, must be physical; but the things everybody talks about when they talk about qualia can’t be physical. Trying to have something be observable, i.e. have physical consequences, yet non-physical, is trying to have your cake and eat it, too.
I think you’ve gone wrong here–the idea of a “zombot” (or more generally, a philosophical zombie) is that it is physically, functionally indistinguishable from an entity with experiences with qualitative properties. A p-zombie will insist it experiences qualia just as you and I do.. but it would be wrong, while we would be right.
Got it. I think most people have said in the past that qualia don’t have physical consequences. But you’re right that in the scenario, it sure seems as though the qualia have physical consequences in that they cause Mary to say “wow”. (I think Frankford himself later came to this conclusion.) But that just means it turns out qualia are physical after all, and in fact are physically fundamental in that their states are (apparently) logically independent of the states of all other fundamental physical fundamentals. In my view, if this is the right way to think about qualia, it represents progress in the understanding of qualia, rather than representing some kind of reductio of them. It may turn out we can fit them into a physical understanding of the world after all.
I’m not sure the qualia are what cause Mary to say “wow,” though, since even if Mary were a p-zombie, she would presumably (since functionally identical to a human being) still say “wow” on coming into the presence of something red. Something would cause her to say “wow” in just the same way Mary herself says “wow,” but it wouldn’t be the qualia since pz-mary doesn’t have qualia.
Again, not what I’m saying. We can’t conclude that we’re not zombies.
Then you disagree with the premise, that it’s possible to construct a machine that acts in all respects conscious without being conscious. I’m fine with that, as such a machine hasn’t yet been created.
That’s why I tried to be a bit more precise. An illusion could, in my terms, be defined as a real-seeming that’s different from the corresponding real-being.
It seems to me you’ve just given it a rather complicated name, and saying ‘it’s a quale’ would have the same content. If you could describe to me what it is like to you when you see red, then you would have effed it (all to hell you would have effed it, since if you could do that, the concept of qualia would be naught).
But if you believe you’ve communicated anything meaningful by referring to ‘what things seem like when things seem red’, you must believe in some common denominator behind the experience – otherwise, you couldn’t expect that I have any idea what the hell you’re talking about.
If that were possible in this way, qualia would just be reducible to boring old physical signals – those flitting around between the two brains --, and there’d be no mystery about them.
Wouldn’t that mean there’s an outside source for qualia? Otherwise, anything having reality only within the mind certainly seems like it ought to be directly accessible to the mind.
I think it’s not that hard in a model of mental content as generated by answers to questions posed to yourself by yourself. Take the things I said about how visualizations are generated, in a sense, without visualizing anything. The folk model of a visualization is that in the mind’s eye, an image is generated that corresponds to the object being visualized. This image is then observed in the mind, and things are learned from this observation. But this idea is actually circular – and viciously so, as, in order to generate the visualization in the first place, the mind would have had to know all the things it later learned from observation beforehand; if however, such visualization is the way to learn things about objects, we are trapped in an unending regress. This is just the problem of the homunculus observing the play of the mind in the Cartesian theatre.
So, what’s real about the whole thing? Well, the learning effect. When we visualize something in order to, say, solve some spatial problem, in the best case, afterwards we will know the solution. But this isn’t a case of real learning, in a way, as we had all the information about the thing to be visualized from the beginning (of course, we might have deduced new propositions from this knowledge, but we couldn’t have learned anything that doesn’t follow from what we knew before). So it’s more like an actualization of this knowledge. In fact, this is close to what I think a visualization of an object is: an actualization of our (visual) knowledge about that object. The answers to a certain set of questions, posed to yourself by yourself: is it round, does it have edges, does it have straight lines etc. Some suitable set of such questions is sufficient to uniquely characterise an object – that’s nothing else but saying that the object could be coded into a string of 0s and 1s, yesses and nos.
Now, the key is that the zombot – able to fool any however clever observer into believing it is conscious – can answer all these questions to itself just as well as you can answer them to yourself. In your case, knowing all the answers to these questions means that you are conscious of your visualization – that you visualize something in the proper sense. However, the zombot doesn’t have that luxury – it’s not a conscious entity. But, being conscious of something is just a kind of more abstract version of a visualization (both are kinds of representations), so the zombot can play the same game once over, and answer questions it poses to itself about how it is to be conscious of a visualization – just as well as you could answer them, if you were asked.
This is how the zombot is able to ‘gain’ subjective states, or at least, the illusion of having them. At first, there’s a set of answers to questions about properties of an object, asked by the zombot of itself in a bit to find out whether or not it is conscious. This is ‘raw data’, needed to construct a representation of something – a visualization, say. The zombot can’t decide the question of his consciousness on this level, as, by the hypothesis of its ability to fool any questioner into believing it to be conscious, it can answer these questions as well as a conscious being can. The zombot thus “concludes” that it possesses all the necessary raw data to create a visualization. Thus, it turns to the next level, asking questions about the properties of being aware of the properties of the object. Again, those questions it can answer as well as any conscious being would. And knowledge of the properties of being aware of something is the same thing as being aware of something.
This still sounds, perhaps, like I am making some big unjustified leap here, but it’s nothing deep – these properties of being aware are the ‘what it is like’ of awareness, and to know these properties is to know what it’s like to be aware. This knowledge, to the zombot, is still unactualized, it doesn’t know that it knows the what it is like; but it knows the what it is like. Else, it could ask a question of itself it could not answer in the same way as a conscious being would, and hence, detect its own non-conscious nature.
Think about a thought. This thought has properties: first and foremost, its content; then, its expression – in what language did you think it, what words did you use, perhaps some more ephemeral things like how it ‘sounded’ in your head, whether it was a sad or joyful thought, etc. You could answer questions asking about these properties, and so could the zombot. But then, you can think about the thought. This thinking about again has properties, thought they appear slightly more difficult to capture. You might think about what an odd thought that was, or where it came from, or what its consequences are etc. Again, if somebody asked you about this, you could answer appropriate questions, and similarly the zombot could answer the questions it poses to itself. But that’s where the magic (which isn’t that magical after all) happens: in answering the questions, the zombot generates a representation of the thought, through the answers about the properties of thinking about the thought. This representation is, to the zombot, indistinguishable from a representation a conscious being might have of the thought. Answering questions about mental content generates mental content, in the form of a representation of mental content through the answers to questions about mental content.
This is circular, but non-viciously so: it is like a river’s flow being determined by its bed, while the bed is shaped by its flow; or like stress-energy being the source of the gravitational field, which itself carries stress-energy.
Anyway, this has turned into another damn essay, which I’m certain has nothing to do with me procrastinating on any number of things I should be doing right now; in conclusion I still think that if a zombot can’t conclude that it is not conscious, we can’t conclude that we are not zombots.
But if it’s not the qualia, then what is it? If she’s had all the physical knowledge beforehand, there would not be any change within her, physically, which would ultimately lead to her utilizing her vocal apparatus to produce a speech act; that ‘wow’ would then come from nowhere, and we’re left with the whole ‘synchronized clocks’ lunacy.
Ah yes. I’d forgotten the definition of a p-zombie.
I can see how one could tie oneself in knots trying to determine whether qualia should be considered physical.
I’ve been making a weaker claim, but I guess a similar sort of claim to the one associated with Mary’s Room: Namely that an objective / physical description of qualia appears impossible in principle. This is not quite the same thing as saying qualia are not physical, let alone that they have no physical effects.
What’s the ef in “ineffable” supposed to mean exactly anyway? Was there ever actually an englis word eff or ef?
So you’re saying the ineffability of qualia consists in their being nameable but not describable? But while I can see a difference between these two concepts, I’m not sure why you think I can’t describe qualia. This qualia is more similar to that one than the other, this one is typically triggered when I hear a flute, this one is how things seem when things seem red, etc. Why do these not count as descriptions?
I was saying qualia might not be intrinsic. What I mean by this is, the association between a particular quale and a particular experience need not be necessary. For all I know, someone else could have had the same experience, with different qualia. I tend to think that people with similar experiences are having similar qualia, but I don’t take this to be a necessary truth. It could have been–or it could be really, for all I know–that the very same experience I’m having right now could have had different qualia associated with it. For example, it might be that experiences must be individuated externally–i.e. with reference to the objects they appear to be experiences of–while we need not refer to externals when individuating qualia. If that’s the case, then since qualia and experience can be differentiated in different ways, there is no intrinsic relation between them.
What I said doesn’t imply that qualia are just physical signals. It does assume that qualia can be had as a result of the transmission of physical signals, but that is not a reduction of one to the other, any more than my having a glass of water in my hand can be reduced to the transmission of water through pipes.
I don’t think it’s true that things with reality only in the mind ought to be directly accessible to the mind–the notion of a subconcious thought is at least coherent, for example. Nor do I think it’d be a problem if there’s an outside source for qualia–perhaps qualia actually inhere in some way in the objects we’re experiencing, or perhaps they inhere neither in the mind nor the experienced object but in the system encompassing them both.
Perhaps ‘communicable’ would be a better word. If you could communicate qualia, eff 'em, so to speak, you could, for instance, compare my and your qualia of redness, or you could explain to a blind person what seeing colour is like. Mary’s problem would be trivial, as you could just go and tell her what it’s like to see red. I think this impossibility, be it apparent or real, is what Dennett means when he refers to the ineffability of qualia.
Granted. But could you have the same experience with different qualia? In what way would it then be the same experience? (I’m not sure that the intrinsicness of qualia is supposed to mean that everybody experiences the same qualia, just that, to you, the experience and its quale are inseparable.)
If it doesn’t, then how can you be sure the other person experiences the same qualia? Does it depend on where you wire the brains together? Say, after one brain has ‘processed’ the input, this is sent to the other – i.e. one person’s Cartesian theatre is filmed, and projected on the other’s screen. But if the qualia are not in the data that is transmitted, there must, in the other brain, occur some processing of its own, causing qualia over there – which may be completely different from those of the first brain. More obviously, if you divert the data-stream pre-processing, i.e. just wire the two sets of optical nerves together, say, things are barely different from just showing both patients the same image. If both don’t share the same qualia from the start, they won’t after rewiring, either.
OK, by mind, I really mean something one is conscious of; I think it’s debatable to what extent the subconscious can be called part of the mind, at least the way I typically use the word. The mind, to me, is what it produces: the sum of all thoughts, experiences, feelings, perceptions, etc. And in a sense, the mind is what ‘we’ are. But what could be more directly accessible to us than our being? So if the mind produces qualia, I think there’s a good case to be made for their direct accessibility.
But then, there ought to be some mode of perception by which we become aware of those qualia. Their presence, either in some object, or in the larger system, ought to change something in us in order to become known to us. Is this a measurable change? A change that occurs only in humans, or could one build a ‘qualia detector’? Could one, perhaps, cut the part that transmits the qualia out of an object, leaving behind something that is physically the same object, yet when we look at it, though we would see that it is red, we would not experience this redness, or when we touch it, though we would feel that it is rough, we would not experience this roughness? And we are again faced with all sorts of quandaries about how something non-physical can have any impact at all on the physical.
The ghost of a queue can’t strike a billiard ball!
Was there ever actually an englis word eff or ef?