Experimental Philosophy of Consciousness

One of the most exciting developments in experimental philosophy these days is the explosion of new work on intuitions about consciousness. How do people determine whether an entity is capable of having phenomenal states like feeling pain or experiencing happiness? And how do the criteria for these phenomenal states differ from those for other states like belief and desire?

It seems to me that the results from these studies are gradually converging on a view according to which people's intuitions about whether an agent has conscious states are especially influenced by facts about that agent's body. We can see this pattern coming out if we look at a number of different kinds of agents that don't seem to have the right sorts of bodies.

First, take 'group agents' (corporations, teams, governments, etc.). People are perfectly willing to describe group agents using sentences that involve mental states like 'Microsoft intends to release a new product,' but they refuse to describe group agents using sentences that involve phenomenally conscious states like 'Microsoft is feeling a little depressed.' Perhaps this is because group agents don't have bodies of the right sort. (For discussion, see Arico; Huebner, Bruno & Sarkissian; Knobe & Prinz; Sytsma & Machery.)

Turning to intuitions about robots, we find exactly the same sort of pattern: a willingness to ascribe beliefs but an unwillingness to ascribe, e.g., feelings of pain. The lack of an appropriate sort of body strikes again. (Huebner; Sytsma & Machery)

And now suppose we look to the ultimate disembodied agent: God. There, we find the very same pattern recurring, with a willingness to ascribe belief, planning, etc. but a reluctance to ascribe occurent feelings and emotions (Gray, Gray & Wegner).

Overall, it definitely is beginning to look like there is some sort of connection between people's willingness to say that an agent has phenomenal consciousness and their beliefs about that agent's body. But now a new question arises. Why exactly is all of this happening?

[For an incredibly brief review of these studies, see the paper here.]

Comments

[Florian Cova]

God is the ultimate Zombie.

[Justin Sytsma]

Hi Josh!

And happy New Year to the x-phi community!

I think that the answer to your question has to do with the type of state we are talking about. While philosophers tend to divide mental states in one way (grouping sensory perceptions, bodily feelings, and emotions together as phenomenally conscious mental states), non-philosophers typically do it differently. Philosophers group those types of mental states together on the basis of their thinking that there is “something it is like” to be in each of those types of states. They treat each of these states as being similar in that each is thought to have phenomenal qualities or qualia. But, I see little reason to think that ordinary people generally think this way. Unlike philosophers, they tend to be naïve realists and to see many of these qualities, not as qualia or qualities of mental states, but as mind-independent qualities of worldly objects. If this is correct, it suggests that there might be a different answer for your question for each of the types of state at issue.

That is, I deny that people are generally willing to say that an agent is phenomenally conscious (because they do not group mental states together in that way); instead we need to focus on their willingness to ascribe different types of states to the agent and I expect that there will be a different explanation for their willingness to do so depending on whether we are talking about sensory perceptions, bodily feelings, or emotions. For sensory perceptions (excluding imaging, dreaming) the relevant qualities are taken to be out in the world, open for everyone to see (hear, smell), and the key factor for ascribing such states to the agent will be whether the agent has the relevant sensory apparatus (eyes or their equivalent, ears or their equivalent, etc.). For bodily sensations, however, the relevant qualities are in the agent’s body; as such, to feel pain, for example, not only does the agent need the relevant sensory apparatus (nerves or their equivalent), but it also needs to have the right sort of body (body parts that can be damaged and that are appropriately innervated). Emotions are even more difficult; they are largely in the head and wrapped up with “brain chemicals” and what not. I suspect that many people will feel that to have emotions you need to be a living biological entity or at least have the right sort of biological brain (thus, in sci-fi books/TV/movies emotions are typically denied even of rather complex robots).

[Joshua Knobe]

Hi Justin,

That's a really interesting alternative hypothesis, and it would be wonderful if we could find some way to put it to the test.

Suppose we told subjects about a computer that could answer questions exactly like a human being and then asked whether it would be right to say: 'The computer can vividly imagine a purple square.'

It seems like I should predict that people will think the computer can't vividly imagine things (since that would be a phenomenal state) but that you should predict that people will think the computer *can* vividly imagine things (since that would be a state with no affective character or sensory apparatus).

Does that sound right to you?

[Edouard Machery]

Josh,

Happy new year to you and the readers of Experimental Philosophy.

It seems to me that you are putting the cart before the horses. Before trying to understand the cues people use to ascribe conscious mental states, it would seem required to know how people think about conscious mental states: What is it for a mental state to be conscious according to ordinary people?

You seem to assume that we have a rather clear understanding of how ordinary people conceive of subjective experience - roughly, they conceive of it as philosophers do. Then, the question would be: What cues do people use to ascribe conscious mental states?

But, as Justin notes, it is far from obvious that ordinary people conceive of subjective experience as philosophers do (and our findings suggest that this hypothesis might well be false).

As for God and its body, we have had a few occasions to discuss this point: It seems to me that the claim that people do not ascribe conscious mental states to god reflects badly on Gray et al.'s experiment. Just open the bible.

Edouard

[Joshua Knobe]

Edouard,

I certainly agree that there is a difficult empirical question as to whether the folk have a concept of phenomenal consciousness, and I think that your arguments against my view are very impressive (hence the links in the original post).

In my comment above, I suggest a possible experiment that might help to resolve our dispute. Subjects would be given a story about a computer that could answer questions as well as a human could. They would then be asked whether the computer was capable of having certain kinds of states that were both (a) clearly not affective in any way and (b) completely, unambiguously phenomenal in the philosopher's sense. (For example: vivid mental imagery.) Does that sound like a fair test to you?

In any case, it seems like we might be able to put aside our disagreement about this issue and work together to solve a common problem. Regardless of what is involved in people's folk conception of subjective experience, there is a question as to what cues people look for in determining whether a given entity has this sort of experience. My suggestion is that people pay special attention here to the entity's *body*. Are you disagreeing also with that claim?

[Adam Arico]

Happy New Year, x-philes!

I want to first say that I completely agree with Josh about the common ground that's emerging in this literature. We all seem to be converging on the idea that folk judgments of mental state attributions seem to take physical features as their primary desiderata. Of course, there's differences between us, I think, in which features are important and why...

I do, though, take some issue with Josh's claim that people are willing to attribute non-phenomenal states to group agents, but not phenomenal states. I think there's good evidence for thinking that the folk don't distinguish (even tacitly) between mental states types in the way that philosophers do. In that regard, I'm in agreement w/ Justin and Edouard.

As far as the proposed experiment goes, Josh, it seems like really testing Justin's proposal would require that the computer be described not only as behaving in a functionally-equivalent manner, but as being equipped with whatever sensory apparatus is appropriate for detecting the imaginative state. Obviously, it's gonna be pretty tricky to figure out what counts as an appropriate description. Given Justin's claim about the folk being (moer or less) naive realists, we'd have to give an account of a mental image of a purple square as an "out there in the world" phenomenon, which can then be detected in some way. So, in order to actually test between your and Justin's accounts, I think we'd pretty much have to solve the hard problem of consciousness.

[Justin Sytsma]

Good stuff!

I think that visual imagery is an interesting test case; in fact, how people think about imagery, as well as imagination in general, is a great topic. (My personal opinion is that even if you think that seeing a purple square is a phenomenally conscious mental state, or involves a purple quale, that imagining a purple square does not involve such a quale because there is no purpleness involved; I think that the sense that it does is an illusion. I used to think that imaging was more like seeing, and involved actual colors, and I even thought I was quite good at it. Heck, I thought I had reason to believe that I was good at it from my time working as a designer. But I’m now fairly certain that there is no visual phenomenology for my imaging. I think that what I can do is better described as being able to reflect on what something *would* look like *if* I were to see it, but without actually seeing it in any way. So, in general, I’m a bit skeptical about what this test case would tell us.)

Despite these worries, I am willing to predict that people would say that a suitably described computer (or robot) could image. I think that the system will probably have to be described somewhat differently than you suggest, however. That is, while I think that people will be wiling to say that a robot can image, I don’t think that simply being able to pass a Turing test will be sufficient for the ascription. The robot will need a bit more than that. Let me try to suggest what I think might be required.

1. While visual imagery doesn’t require sensory apparatus, I would also expect that most people would deny that somebody who was blind from birth and had never seen a purple square could vividly imagine one. I suspect that one requirement on the agent for folk ascriptions of visual imagery will be having had some visual perceptions. But, cameras could be added to the robot along with a story about it using the cameras to make visual discriminations.

2. Even with cameras, I’m not sure that the robot being able to answer questions in a Turing-test fashion is a sufficient. Note, that it doesn’t seem to be enough for humans. While I assume that pretty much any normally functioning human adult can imagine a purple square in some sense, for many this isn’t a plausible candidate for a phenomenally conscious mental state. That is, many claim to have little to no ability to create visual imagery. So, it strikes me that we need some behavioral indicator for having the ability to visually image that could be added to the description for the robot (and a human control). Unfortunately, it is not so clear what the ability to visually image is supposed to do for us, so coming up with a realistic behavioral indicator is tough. If I remember correctly, people’s reports about their ability to visually image don’t correlate well with their skill at saying whether two shapes are the same or mirror-images. Fortunately, I don’t expect that most naïve subjects would know that. So, perhaps you could describe a rotation task experiment for subjects and tell them that most people who are good at visual imagining do well at the task. Then have the robot (or human control) do the experiment, do it well, and afterwards make some (verbal?) claim about having visually imaged the shape to complete the task.

3. I am not sure what the “vividly” adds to the test question (and it might be best to just ask them whether the computer or human control imagined the relevant shape from the experimental task described in 3). My concern is that imagery isn’t itself vivid (so, for example, when we say that somebody has a “vivid imagination” we are saying that they use their imagination a lot and that what they imagine is interesting, not that there visual imagery is vivid as opposed to muted). Personally, while I can readily imagine a vivid purple square (or one that is a dull purple), I don’t know what would be involved in vividly imagining a given purple square as opposed to just imagining it.

Edit: I see that Adam posted while I was composing this. I think this fits in with his suggestion. With regard to naive realism, I expect that in a strange way many people are "naive realists" about dreams, hallucinations, and the like (that is, I suspect that they think that it is possible to have a visual hallucination of a pink elephant and that in doing so you see a pink elephant (which just so happens to be both hallucinatory and pink)); as to how that could be the case, or what such "objects" amount to, I doubt they have much of a theory at all.

[John Matthewson]

Hi Justin,

Your denial of non-sensory visual experience is intriguing. You said: "My personal opinion is that even if you think that seeing a purple square is a phenomenally conscious mental state, or involves a purple quale, that imagining a purple square does not involve such a quale because there is no purpleness involved; I think that the sense that it does is an illusion"

Have you experimented with strong hallucinogens or believed you have woken up whilst actually still being in a dream? Have you seen paintings by Impressionists/Pointillists and realised that there is no real image? Have you seen phosphenes, migraine wheels or stars after a blow on the head? Have you noticed that you dont see the retinal image during saccades of the eyes (look in the mirror and see if you can see your eyes move) - where is the image that you see during the almost continuous saccades of your eyes? Do you see cartoons as continuous action - quite a miracle of internal motion modelling when you consider that cartoon refresh rates (10Hz) are less than the flicker fusion frequency (50Hz). Is perspective a property of the world or a form of the retinal image see http://en.wikibooks.org/wiki/Consciousness_Studies/The_Neurophysiology_Of_Sensation_And_Perception#Depth_perception? Etc..

[Eric]

"Overall, it definitely is beginning to look like there is some sort of connection between people's willingness to say that an agent has phenomenal consciousness and their beliefs about that agent's body. But now a new question arises. Why exactly is all of this happening? "

Didnt Wittgenstein make just this observation in the PI and attempt to answer just this very question? (see, for example, the aphorism about attributing pain to a fly and to a stone, and the one about imagining yourself turning to stone while you are in pain.)

[Joshua Knobe]

Justin,

That sounds like a great experiment! I would definitely love to know how it would turn out.

Maybe it would be helpful to include other states, in addition to visual imagery, that are as unambiguously phenomenal as possible without being in any way affective. Then one could test to see whether there is a significant effect whereby subjects are less willing to ascribe these sorts of states to robots than they are to ascribe paradigmatically intentional states like belief, understanding and planning.

Adam,

I was thinking that your recent work might shed some light on the question Edouard raises above about the mental states of God. Your studies show that people are willing to use sentences like 'Microsoft is upset about that recent newspaper article' but that they are not willing to use sentences like 'Microsoft is upset right now.' I was thinking that the very same phenomenon might arise for God. Maybe people think it sounds okay to say 'God was angry at the Israelites' but think it sounds wrong to say 'God is very angry right now.'

If so, we would have evidence that your effect is really tapping into some deeper phenomenon that arises not only for group agents but also for other kinds of disembodied agency.

[Justin Sytsma]

John,

I recognize my views on this issue are rather controversial. I do think that I can give answers to the various cases you raise, but I don’t think that I need give the same type of answer to each one (since there is no reason to think the same mechanism is involved in each). In some cases I think people are mistaken, in others their recollection of the events just isn’t trustworthy, and in still others I think there just isn’t a problem. So, for example, while I won’t comment about whether I have experimented with strong hallucinogens, note that people on drugs tend to have lots of weird beliefs; sometimes they even believe that they are seeing things that aren’t actually there. But, we shouldn’t take their reports of this to be trustworthy. I had a friend who became hooked on meth and at one point, besides claiming to see bugs that weren’t there, claimed that his roommate was coating his walls, floors, and furniture with poison. Of course, I didn’t believe the latter claim; but, then, why give any more credence to the former claim? I imagine I want to say something similar for blows to the head. (Or for schizophrenia, etc.) Likewise for dreaming. I know that it can seem like a past dream (even immediately past) was vivid and sensory; but it can also seem like you remember seeing something that was suggested to you but that you didn’t actually see. Skepticism about dreams being colored is, I think, rather easy. For Impressionists/Pointillists and cartoons, I don’t see a problem (although maybe I am missing your point, here). Focusing on cartoons, for example, it just seems like a clever exploitation of one of the limits to our ability to see. Nonetheless, what you see is actually there (and you can convince yourself of this by going through it frame by frame).

But, I do recognize that my position is rather counterintuitive. Note, however, that there are serious problems for the intuitive view as well (such as explaining how the brain conjured up purple squares, etc.) I find the problems involved in attempting to explain away examples like the ones you raise quite a bit easier than the hard problem.

Eric,

You are probably thinking of Section 283 from the Investigations:

---------------------------------
What gives us so much as the idea that living beings, things, can feel?

Is it that my education has led me to it by drawing my attention to feelings in myself, and now I transfer the idea to objects outside myself? That I recognize that there is something there (in me) which I can call “pain” without getting into conflict with the way other people use this word?--I do not transfer my idea to stones, plants, etc.

Couldn’t I imagine having frightful pains and turning to stone while they lasted? Well, how do I know, if I shut my eyes, whether I have not turned into a stone? And if that has happened, in what sense will the stone have the pains? In what sense will they be ascribable to the stone? And why need the pain have a bearer at all here?!

And can one say of the stone that it has a soul and that is what has the pain? What has a soul, or pain, to do with a stone?

Only of what behaves like a human being can one say that it has pains. For one has to say it of a body, or, if you like of a soul which some body has. And how can a body have a soul?
---------------------------------

I take it that the point we are supposed to end up at is the first line of the last paragraph; that our very idea of pain is intimately connected to our embodiment as reflected in our animate behavior. This is perhaps a bit more clear in the first paragraph of the following section (Section 284):

---------------------------------
Look at a stone and imagine it having sensations.--One says to oneself: How could one so much as get the idea of ascribing a sensation to a thing? One might as well ascribe it to a number!--And now look at a wriggling fly and at once these difficulties vanish and pain seems able to get a foothold here, where before everything was, so to speak, too smooth for it.
---------------------------------

I think that this observation is actually quite telling about the common-sense conception of pain. Where many philosophers today think of pains as being mental states that we can then conceive of as floating free of a given body, I think the more “naïve” perspective is to see pains as instantiated in certain bodies.

Josh,

I'll put some thought into other states to try. I should have access to some students this semester and can try out the visual imagery case. I don't know that it will settle the debate, but it might help! If the results are interesting, perhaps we can write something up about it together?

[Adam Arico]

Josh,
That's an interesting suggestion. Though, I must say, I don't think the rejection of phenomenal attributions to God is a product of people failing to see God as an Agent (for those who are here but haven't read my paper: the hypothesis--put somewhat roughly--is that people attribute mental states (both phenomenal and nonphenomenal) to a thing if and only if that thing triggers their concept 'Agent').

I think the worry about God is a bit of a red herring. If we're basing the idea that people don't attribute phenomenal experiences to God on the Gray, Gray, & Wegner data, it's important to keep in mind that their term 'Experience' is a technical, post-hoc category. They use it to group together their data into categories of attributions (like hunger, fear, pain, pride, and embarrassment... all attributions that don't necessarily comport with the image of an all-powerful, supreme being) to account for the variance in responses. It seems likely to me that God rated so low on their "Experience" scale because so many of the attributions they presented (especially the ones listed above) are imperfections that people have trouble reconciling with their notion of God.

So, while your suggestion seems right (i.e., that situating the attribution of anger to God in a context would, in fact, positively influence judgments), I don't think that it's the phenomenal aspect of these states that people were denying of God in the GG&W experiment, but rather the "imperfection" aspect of these particular phenomenal states that they were denying.

I actually suspect that people would be largely willing to say that 'God was angry' sounds natural, as well as other states that are compatible with the Western notion of God. And while I also think that those ratings would only increase if we situated the attribution in a relevant context, I don't think that it would be the same "context effect" that I'm interested in. That is, I don't think the context would make people more likely to see God as an Agent; it would only make them understand _why_ God is angry, which might then make them more comfortable saying that 'God was angry [at the Israelites' worshiping a false idol]' sounds natural.

[Joshua Knobe]

Justin,

I would be more than happy to help you out with this! It seems like a really great way of making progress on this question.

Actually, I was thinking that it might be a good idea to get specialists in the philosophy of mind to rate the ascriptions on a scale from 'unambiguously phenomenal' to 'unambiguously non-phenomenal.' That way, we could be sure that we were tapping into the very same notion that is at work in traditional philosophy of mind.

Adam,

It seems like the question you raise here is easily susceptible to empirical testing. For example, one could ask people whether it sounded right to say: 'God is feeling very happy right now.' 'God is now feeling a deep sense of calm.' 'God is now feeling extremely pleased.' One could then check to see whether these sentences are regarded as less acceptable than sentences like: 'God was angry at the Israelites.'

Also, I think your paper is very interesting in its own right. Instead of modestly mentioning it in this comments section, you should be boldly unveiling it in a post of its own!

[Brandon]

Dear Joshua,
You make a very valid point and the state of consciousness is one of my favorite debates. I think the conflict arises from the subjective consciousness of man. Man is only capable of perceiving the world from his own psyche or conscious. Even when "walking in an others' shoes" we are left with a hugely gray area that the individual fills in with personal experience. Man can not relate to being a "multi-intellectual" organism such as a company. Man also can not relate to a being with out physical form, and realistically, man even has trouble relating to other physical entities that inhabit the planet.
The next area of debate is that of sentience. What makes one "alive" or "conscious?" How do we define the existence of the conscious along with what it entails? If consciousness is merely fulfilling the body's need, then plants would also have a conscious state of existence. The other question I'll pose is the consciousness' dependence on a physical body. Can a conscious be separated from a physical form? This leads into a very controversial debate of a soul and the sorts, but I think it is an important point to bring up. If the "mind" is nothing more than a chemical and neural reaction of the brain that creates the facade of a more complex design, then the individual consciousness will fade along with the body and eventually die, as the body surely will in time. This debate isn't strictly of importance in your theorem here, but relates to many medical and scientific fields as well. As we become more advanced medically, at what point do we stop being "human" and start becoming "machine?" On that same notion, can the conscious or "soul" survive inside of a machine? Which leads into another debate of what does it mean to be human?
So I feel that your short poke into this area leads into a much greater jab of debate, however, I did greatly enjoy reading your idea. Thank you for your time.

[Eddy Nahmias]

Relevant to the discussion of attributions of experience to God, notice that the Gray et al. study finds that the infant (and the child) rate a bit higher on the experience scale than the adults (including the subjects' attributions to themselves). And the monkey and dog are also slightly higher than the adults. Why would that be?

My hypothesis is that some of the experience terms, like hunger and pain and fear, are ones people may ascribe to animals and infants more than adults because adults have more control over those experiences and may experience them less often. (I don't know how they did the stats, but those three feelings are the ones that are loaded highest on the experience category and lowest on the agency category.) Given the head-to-head comparison method they used, it is likely some people respond, e.g., that the infant feels hunger more (often and severely) than the adult.

Similarly, in such head-to-head comparisons, it is understandable that people would rate God below the humans and animals. Why would an all-powerful God feel hunger or fear? But it's still surprising God came out so low, given that many religions (and certainly the Old Testament) seem to attribute anger, pride, joy, etc. to God. I wonder if the researchers tracked religiosity, since non-religious people presumably aren't going to attribute consciousness to God (but then how'd he get such high agency?).

[Josh Weisberg]

I wonder, then, what to make of the claim that "God is love"? I, too, think the God results are hard to decipher, though very interesting.

Josh--

I find that while body does seem important to my undergrads (admittedly, without an x-phi survey!), more important seems to be the goo/non-goo distinction. Feelings seem to go with slimy stuff, while robots and group entities lack slimy stuff. Many students feel that aliens have feelings, but if they learn that the aliens are made of metal or rock, this changes. If this is true, I wonder what it is about goo that fits with the phenomenal? Is it a throw back to humors?

Note also that a crucial moment in many sci-phi plots is the "falling off of the robot's face plate." This reveals what was previous thought to be a normal, phenomenally conscious entity to be a "mere automaton." Again, it's not body shape or behavior, but internal composition that matters. Likewise with animals and robot animals. A dolphin can feel pain, surely, but a robot dolphin?

Interesting stuff! Heterophenomenology at work!

[Justin Sytsma]

Hi Josh -

I think you're right that goo/non-goo is an important distinction, here. In a study I've been working on looking at attributions of pains, for example, I found that whether the afflicted body part was gooey or not was one important factor (basically, people were more likely to say that pain was involved in a person cutting an artificial but biological replacement hand than a robotic one). But, in this case, even more important was the presence of nerves of some sort (biological, robotic). The data suggests that for attributions of pains, at least, being gooey is not itself essential, but rather is a cue that the body has all of the stuff that is in fact essential.

I'm not so sure that this is heterophenomenology at work, though. In laying out heterophenomenology Dennett typically states that we should interpret people's utterances in terms of beliefs about phenomenally conscious experiences. He concedes that to his philosophical opposition, instead arguing that one shouldn't move from the existence of beliefs about phenomenally conscious experiences to the existence of phenomenally conscious experiences themselves. The new experimental work that is being done on the folk psychology of consciousness, however, involves something of the reverse. The base question, coming out of Josh & Jesse's paper, is whether the folk have the concept of phenomenal consciousness (and hence whether their utterances about states that the philosopher might think of as phenomenally conscious are actually best interpreted in terms of beliefs about phenomenal consciousness). That is, one way of looking at it is that we have been focusing on the question that Dennett concedes, while largely conceding the issue that Dennett focuses on (as nobody has been moving from folk beliefs about phenomenal consciousness to the reality of phenomenal consciousness).

[Josh Weisberg]

Hey Justin--

Your study sounds very interesting! I look forward to seeing it when it's ready. In this context, I've been wondering about the androids in the Aliens movies. They have a white gooey liquid inside them, but are explicitly artificial (I'm not sure if they have nerves). I wonder if folk hold that they have experiences, or if they are more likely to have experiences than non-gooey androids like Data.

Interesting point on heterophenomenology (HP). I'm not sure that Dennett concedes that folk have beliefs about phenomenal consciousness. Rather, he says we should make a record of what the folk say about experience (or some more general neutral mental term) and assume that this expresses their beliefs. This gives us an interpersonal folk account of the mind, which fixes the data a theory of consciousness must explain. Then (as you note) there is the question of how best to explain these beliefs. We might explain them in terms of accurate introspection and inference, of a systematically over-reaching interpreter system, of social influences unrelated to introspection, of desire to please the questioners, etc. Dennett argues for a considerable degree of error and illusion to explain the beliefs, rather than accepting their accuracy, though that is certainly not the only way to go.

My guess is that Dennett would take these results as Jesse and Josh intend them: as showing that "phenomenal consciousness" is a philosopher's theoretical construct, not one simply obvious from the folk perspective. Hence, his theory of consciousness is not at pains to explain phenomenal consciousness--it's not part of the neutrally-fixed data a theory must explain. My idea, then, is that x-phi is simply HP at work. It takes the reports of subjects and forms an "HP world," similar to interviewing religious believers about the denizens of their religious world. (Note that the high priests' view of the gods may be very different from that of the average believer.) But whether or not these denizens really exist, or are just the product of a psychological need for explanation, comfort, etc., is an open question.

[Justin Sytsma]

Josh:

I hope that you are correct that Dennett would not have a problem with the experimental work being done on this topic. But, I still don’t think it is HP, at least not as Dennett typically describes it. The dialectic that Dennett gives HP within assumes a great deal about what he calls “the folk theory of consciousness.” It is assumed that people have beliefs about phenomenal consciousness and HP interprets their utterances in that light. For example, take Chapter 2 of Sweat Dreams (2005). Dennett asserts the following about the folk psychology of consciousness:

“We – nudge, nudge – know about our consciousness because we communicate about it all the time. In our everyday dealings with each other we presuppose a vast sharing of understanding in all our public representations of consciousness, and we contribute to that common stockpile, our presupposition is apparently vindicated.

“The folk theory of human consciousness is a hugely successful mutual enterprise, but it does have its well-known puzzle-points. Can a person born blind share ‘our’ understanding of color? What about a color-blind person? What about ‘spectrum inversion,’ a thought experiment at least three hundred years old? Might it be that what I see as blue you see as yellow, but nevertheless you call that subjective color blue?” (31).

(The last bit follows rather clearly from Dennett’s assertion in Consciousness Explained that the folk are committed to Locke’s secondary quality view of colors.)

Later in this chapter he gives his standard description of HP in terms of four data stages, contrasting it with Levine’s view:

“Levine’s claim can be most clearly understood in terms of a nesting of proximal sources that are presupposed as we work our way from raw data to heterophenomenological worlds:

(a) ‘conscious experiences themselves’;
(b) beliefs about these experiences;
(c) ‘verbal judgments’ expressing those beliefs;
(d) utterances of one sort or another.

“What are the ‘primary data’? For heterophenomenologists, the primary data are the sounds recorded when the subjects’ mouths move, or (d) the utterances, the raw uninterpreted data. But before we get to theory, we can interpret these data, carrying us via (c) speech acts to (b) beliefs about experiences. These are the primary interpreted data, the pretheoretical data, the quod erat explicatum (as organized into heterophenomenological worlds), for a science of consciousness.” (44-45).

Dennett locates the motivating dispute for HP at the divide between (a) and (b), accepting (with his opponents) that the utterances are best interpreted in terms of beliefs about phenomenally conscious experiences. This, of course, makes sense given Dennett’s beliefs about the folk theory of consciousness that he opens the chapter with. But, I think that the work being done on the folk theory by experimental philosophers is showing that Dennett’s understanding of it is largely incorrect. For example, while he suggests that the folk theory is committed to the possibility of spectrum inversion, that isn’t what I found when I asked the folk (the majority denying that an inverted spectrum is possible).

[Josh Weisberg]

Justin--

Good stuff. But I think we can distinguish between HP and Dennett's application of HP. He does not go out and systematically interview folk; rather, he assumes he can access that data from his armchair, given his own knowledge of the "vast sharing of understanding" on consciousness. But your results and Jesse and Josh's, etc. (arguably) show that he got the folk wrong. Dennett misapplied his own method.

Still, what x-phi gives us is a better sample of what goes into (b) above: folk beliefs about experiences. X-phi is the data gathering step in building the proper HP world. That Dennett misjudged what should go into (b) does not mean that x-phi is providing some other sort of information. X-phi is the corrective on Dennett's still-to-cozy relationship with his armchair.

Note that Dennett nowhere says "phenomenal consciousness." He intends his method to let the folk have authority over step (b) in fixing what needs to be explained. That he mistakenly thinks there are "puzzle-points," or that the puzzle-points are more prevalent than they are among the folk, is a failure to do HP properly, using x-phi, not a failure of HP generally.

I've found that asking about blind people and color seems to generate the p-consciousness intuitions most easily among my undergrads (and my relatives). Did you try any questions in that direction? I agree that lots of students don't share the inversion intuition (though, some of course do), nor do they all get the zombie idea. But the blind/color case, and it's philosophical offspring Mary, seem to get more folk excited. Is that puzzle-point enough to make Dennett's claim? How puzzling must it be? How many folk must be puzzled?