May 26, 2004
A Thought Experiment At Close Range
Close Range: A Thought Experiment
Go read the set up. I'll wait.
Marc says that Option 1 (Kurt knows that p at t1, fails to know it at t2, knows it at t3, fails to know it at t4, … , knows it at tn.) "requires that knowledge be able to flicker in and out of existence over very short intervals of time. This seems to me like an immensely unattractive position."
Why is that such an unattractive position? A simpler case might be you tell me your phone number. After finding a paper and pencil, I write it down. I immediately forget it. When I need to call you, I look it up, but then forget the last digit as I'm dialing. I look again at what I've written and complete dialing. Later that day I forget it again. Eventually I've called you enough times that I no longer have to look it up.
Wouldn't it be a pretty ordinary way of looking at things to say that I knew your phone number when I wrote it down, I didn't when I had to look it up again, I did when I dialed it correctly, and so on? At each of those moments I have Justified True Belief--and yet it certainly looks like my knowledge flickers in and out of existence for cases much simpler than proofs that can only be apprehended at the barest edge of my mentally acuity.
Maybe I'm missing something , but it does seem to me that we know things even if they are only in our iconic memory, despite our knowing that they'll vanish immediately unless reinforced. If that weren't the case we'd have to deny knowing the appearance of something even when we're looking right at it--and to me that's also an immensely unattractive position.
May 07, 2004
Meanwhile Back on The Farm
In Fake Barn Country: Color in Dreams and Fiction, they're arguing about whether you dream in color, and can you be mistaken about that. (A bit of a coincidence, since my previous post on color perception was prompted by a thread on Sequential Tart )
The first thing that springs to mind is that nowadays it's probably possible to settle the question of dreaming in color empirically, by doing a PET scan while someone is dreaming.
The second thing is that what Jonathan is proposing (or agreeing with Ernie on) is really quite weird in light of, for instance, what the Damasios have written about color perception, memory, and language use. Normally, seeing, remembering seeing, and imagining seeing all activate the same parts of the brain. Destroy certain of those parts and you not only become incapable of seeing color, you become incapable of even imagining it any more. On the other hand, destroying the parts that let you assign names to colors, or speak words (including but not limited to those for colors) and you can still perceive, remember and compare colors non-verbally. Physiologically, then, dreaming without perception would seem to require that we dream using only the linguistic centers of the brain--i.e. that we dream in sentences, and confabulate the sensations that we seem to recall having in the process of remembering or recounting the dream. Another possibility would be that the visual centers would light up, but only those responsible for black-and-white imagery. That would be strange, but probably testable, although I'm not sure whether our current undertanding is that color vision is seperable from black-and-white vision as far as what the brain does (obviously there are people who lack color receptors in the retina, so they see in black-and-white, but I don't know if PET scans of their brains show any systematic differences from people with normal vision). Either that, or being in a dream state would somehow turn us into Zombies, where the visual parts of the brain would be doing all their usual stuff which when awake leads to perception and memory and sensation, but while asleep is simply neural activity unattached to qualia but generating memories that will have qualia attached if and when we reactivate them. This latter hypothesis does seem to me to be untestable, but also wildly implausible.
So why were people widely convinced prior to the 60's that dreams were never in color? I don't know. Did they really, or was that one of those stupid things like Behaviorism that only an intellectual could believe?
Talking About Damasio
Hanna and Antonio Damasio have written about language and the perception of color, and three different kinds aphasia relating to color: achromatopsia, damage to the occipital region that prevents percieving or even imagining colors (this relates back to what I mentioned about the same areas being used both to see and remember seeing); lesions to left posterior temporal and inferior parietal cortex, which prevent proper formation of words ("buh" instead of "blue") but which affect all speech, not just color speech; and color anomia, damage to the temporal segment of the left lingual gyrus, which prevents you from being able to name any colors even though you still perceive and can match and discriminate between colors non-verbally. So, there seems to be a very specific locus where color perception gets coded as color language, but the perception can still go on normally even when the ability to encode the perception as language is destroyed.
May 06, 2004
Gorilla My Dreams
Crooked Timber: What we don't notice... recounts an interesting experiment in perceptual framing and focus, where people who are shown a film and told to focus on a particular task (counting the number of times a team passes the basketball) don't notice something really obvious if unrelated to the task (a woman in a gorilla suit walking slowly through the players, stopping, thumping her chest, and moving on).
More evidence of just how much processing of our perception goes on beneath the level of consciousness, and the great extent to which conscious intention directs and shapes that processing.
April 08, 2004
April 06, 2004
Brains...Brains...
Richard, of Philosophy, et cetera, suggests that it's easy to imagine a simple robot that falls victim to the Motion Induced Blindness illusion I mentioned before. Quoting from one of his comments:Anyway, the central point to note is that any simple robot (as could be built today), could easily have 'hidden' information (e.g. raw visual data) which is not made available to its other decision-procedures. That is, its camera would capture the photons (same as our eyes)... but the interpreting algorithm could be fooled into not noticing any yellow dots (same as our brain).
The thing about Zombies, though, is that they do considerably more than this robot; by posit, they are indistinguishable from human beings in their behavior. This means that Zombies not only have to be fooled by the illusion, but to notice that it is an illusion, and to say "Neat!" So more is going on than just the information getting hidden by a sub-system (which is presumably what it happening even in humans); there also has to be an ability to note and act on the fact that not only that certain information is present to the decision procedures but that other information ought to be present but is not (which requires not only a model of the world, but a model of your "mental states"), to infer that the reason this is so is due to an optical illusion and not some other kind of failure (Zombies don't look at it and say, "Augh! I've gone partially blind!") and to decide, despite the absence of anything subjective going on, to express a faux-emotional reaction: "that's cool."
Perhaps that's still easy to picture (I don't really find it so, but I may be deficient in this regard).
Suppose we flip it around, though. According to the logic underlying the Zombie hypothesis, if it's conceivable, then it's possible. Isn't it conceivable that the simple robot has qualia? After all, in the Zombie hypothesis qualia don't attach to a mental state by virtue of the behavior that state induces, not even the behavior of being able to note and act on that mental state, nor are qualia the cause of any mental state. (If they were, then there would in principle be some way of distinguishing Zombies from Realies.) Is there a reason that it's inconceivable that the robot that falls for the Motion Induced Blindness illusion has qualia as it does so? If not, then there is a possible world where such robots do have qualia...but if a robot, why not a rock? Can we imagine that there's something that it's like to be a rock?
I am a rock,
I am an island,
and a rock feels no pain,
and an island never cries
It might be that the Zombie proponents would say, "Sure. That proves that consciousness has nothing to do with that lump of meat in your head." I'll stick with my brain, though. I'm very fond of it, you know. We grew up together.
You could slice it another way, as well. Imagine that there are some beings who have qualia, just as we do, but they don't experience emotions the way we do. Call them Vulcans. Unlike people who actually don't experience emotions (such people really exist), and unlike Star Trek Vulcans, these Vulcans are indistinguishable in their behavior from normal people (Feelies). They laugh, they cry, when asked why they laugh or cry they give perfectly consistent and lucid explanations of their behavior in terms of emotions, their physiology changes exactly as humans do in the grip of emotions (bloodpressure, galvanic skin response, etc.)...they have all the normal sensory qualia, but they don't actually have any emotional qualia.
Is this conceivable? If Zombies are conceivable, then Vulcans certainly are.
Have we just proved that the soul, or whatever metaphysical thing provides the qualia, is bipartite, and the question of having sensory qualia is completely distinct from having emotional qualia? Seems so. Whatever "the answer" is to how people can have qualia, it's possible (in the Zombie-possibility sense) to have an entity in that same situation who's a Vulcan.
Why stop there? Is it conceivable that the qualia of thinking are severable from the qualia of perceiving and emoting? What about specific sensory qualia? Can you imagine an entity who has qualia when perceiving visually, but is a Zombie when it comes to auditory perception? What about one who is a Feelie with respect to all emotions except, say, lust, where he's a Vulcan? (Note that he would still act as if he experienced lust, he just would have no lust qualia; it's not the same as having no sex-drive.) I'm reminded of Daffy Duck, who once explained, "I'm different from other people. Pain hurts me." What about someone who is only a Zombie part-time...say every other day. Does that prove the detachability of souls (or whatever)? Or does it prove a new type of metaphysical entity: the every-other-day-qualia-inducer?
It seems like you can multiply the necessary metaphysical entities to explain qualia endlessly. The conceivable => possible argument is an anti-Occam's Razor.
What I'm really arguing is that what's imaginable or conceivable is a really loose constraint if it's wide enough to admit Zombies, and so by imagining Zombies the Zombie-hypothesizers haven't really proved anything possible--let alone proved that there must be something metaphysically or spiritually other going on in our brains or we wouldn't have qualia.
Crawling up the Ecosystem
I'm now A Wiggly Worm. Woohoo!
There's some interesting back and forth at The Panda's Thumb about the TTLB ecosystem rankings, and whether the "progress" that higher rankings in TTLB represent is antithetical to the proper understanding of the evolutionary relationship of organisms. I think that some of the people, in order to deny teleology in evolution, are bending over backward too far in denying that there's anything interesting going on at all in terms of complexity and specialization over time. There's a pretty strong sense in which organism A is "more evolved" than organism B if organism A could be the decendant of organism B, but not vice-versa, and I think that sense carries over to organisms C and D where C is a descendant of A that shares the features that make A later than B and D is a descendant of B that lacks those features, even if C and D are both equally removed in time from A and B. Evolution is a tree, not a field of grass, and time's arrow is stamped all over organisms if you know how to look. That doesn't mean that more evolved is somehow better in some moral sense, or even in terms of fitness, any more than having a lot of inbound links (the measure in TLBB rankings) makes a blog better or even more "fit" in the blogosphere.