Ep. 223: Guest Ned Block on Consciousness (Part One)

0:00 Mark: You are listening to the Partially Examined Life, a podcast by some guys who are at one point set on doing philosophy for a living, but then thought better of it. Our question for episode 223 is still what is the mind? And we are joined by Ned Block, professor at New York University, to discuss two sections of the new book about his work, Blockheads, edited by Adam Pautz and Daniel Stoljar. We'll be looking at Brian McLachlan's essay in that collection, could an Android Be Sentient? And Ned's Functional Role Superficialism and Commander Data replied to Brian McLachlan. And then Michael Tighe's essay Homunculi Heads and Silicon the Importance of History to Phenomenology. And Ned's reply fading a response to Michael Tighe. For more information, please visit partiallyexaminedlife.com this is Mark Clinton Meyer, quasi Sentient in Madison, Wisconsin.

0:43 Seth: This is Seth Paskin, full of Qualia in Austin, Texas.

0:47 Wes: This is Wes Alwan in Cambridge, Massachusetts.

0:49 Dylan: This is Dylan Casey, consumed with elementary Homunculi in Madison, Wisconsin.

0:55 Ned: This is Ned Block on Martha's Vineyard.

0:57 Mark: So we've been building up to this for a while. The past few episodes we've been covering some of your other essays, some of the things that those essays were responding to in the first place, trying to give our listeners some grounding here. And Blockheads is just full of so much cool stuff. I really. I skimmed through several of the other essays. There are just a lot of big names, a lot of interesting people in there. Do you want to say a little bit about how this book came about, how this relates to your actual work? I know several years ago I emailed you to be on the podcast and said, we're one of the only places to promote your book. And you said, well, I'm not going to have a book anytime soon. That was several years ago. But now you have one.

1:34 Ned: Yeah, and actually I'm working on another book. I'm almost done with another book, but this book was put together by Adam Pouts and Daniel Stoljar, both of whom were students where I was their thesis advisor. I've been teaching since 1972. So they thought it'd be nice to get some former students and some colleagues to write articles and for me to reply to them. And that's what Blockheads is.

2:00 Mark: Well, and it's great that. Yeah. That I wouldn't say your replies to all of them are really substantial. Some of them are two pages long. Like, that's totally irrelevant to what I was talking about, but at least the one for the tie. Here is one you recommended as being a real step in that ongoing discussion between you and.

2:18 Ned: And.

2:18 Mark: Well, really, David Chalmers.

2:20 Ned: It was very helpful to me to write that reply to Ty and to talk to Chalmers about it. We had a lot of back and forth about it, so I feel like that that reply really does advance the issues.

2:31 Seth: Ned, I wanted to thank you for coming on and appreciate it. This is not a topic with which I'm terribly familiar. My background is in existentialism and phenomenology.

2:43 Ned: My original introduction to philosophy was via existentialism and phenomenology. My first teacher was Bert Dreyfuss.

2:50 Seth: Really?

2:51 Ned: Oh, yeah. In the fall of 1960, I started as a freshman at MIT. The first day of classes, he was my freshman humanities instructor.

3:02 Seth: Wow. Now I'm officially 2 degrees removed from Birch Dreyfus. So there we go. I guess my question to you, framing the conversation, is it feels like this is a conversation that you've been having for 30 or 40 years with a variety of people. And as we kind of dip into it, we've read a few papers here and there trying to set the stage for the hard problem, the easy problem, and functionalism. And I'm wondering how this conversation's gonna go, since we're not as steeped in it as you are, and how we can touch on the critical topics, but not having been part of the conversation for 30 years. So hopefully you'll help us navigate that.

3:36 Ned: Sure, I'll try. I wouldn't put it in terms of a conversation that the people involved have all been having for 30 or 40 years. You know, there is this issue about what the nature of the mind is. And I think there's been an ongoing discussion of many, many, many people, but a lot of people have come into it pretty recently.

3:55 Seth: Okay, great.

3:56 Mark: Yeah, A lot of fresh names in blockheads here that I was very impressed by. Some of them.

4:00 Dylan: My background isn't even. Except for a little bit of undergraduate work in political philosophy. My background's in physics.

4:06 Ned: That's what I was an undergraduate major in, actually, at mit, as it turns out. Oh, yeah, Okay. I actually did a double degree of physics and philosophy.

4:16 Dylan: I got a degree in political philosophy and a degree in physics, but I continued on in physics. So when I'm reading this and when I was going through it again, is I just got myself really tied up with trying to understand what we mean by having an experience and how we talk about something. Having had an experience, I say that I want to Be very tentative about it. Because I think that that problem ends up basically being in the background of a lot of these papers. Because the argument ends up being about saying, well, that is obviously an experience.

And a lot of the reductio ad absurdums end up going to say, well, it's clearly not having an experience, or it is having an experience. And so that question is rooting around in the background. But I found myself deeply confused about what the signs of having an experience were. That I would say that some other entity was having an experience. And to me, I had a hard time disengaging myself from thinking about that problem when I was reading any of these papers.

5:16 Ned: So, look, I think the key thing is to make a distinction between a number of different mental states. What is at issue in all these papers about phenomenology and experience is what I call phenomenal consciousness. And that is what Nagel calls what it's like to have an experience. You know, sometimes people say, try to evoke it with the redness of red or the smell of roses or the taste of something. But that's the key issue involved, the issue of, you know, what the nature of phenomenal consciousness is. Importantly, that has to be distinguished from a number of causal notions that capture some of the causal relations between phenomenal consciousness and other aspects of the mind.

For example, there is what I call access consciousness, which is the ability to use phenomenally conscious information in one's thought and reasoning and reporting and behavior. And sometimes it can be pretty difficult to distinguish between phenomenal consciousness and access consciousness. But access consciousness is an information processing notion. And then another kind of consciousness that has to be distinguished from those is what one might call self consciousness, which is awareness of oneself as an acting being and awareness of one's own mental states.

So that is the notion that is at the heart of the higher order thought approach to consciousness. It's important to distinguish all these things. And of course, whether there is really a distinction between these things is itself a major topic of discussion. So a lot of people think phenomenal conscious just is access conscious. Or other people think that phenomenal consciousness just is a matter of self consciousness of some kind. So the definitions and the disagreements are all woven together.

7:07 Mark: I can kind of tell that you maybe started being very much aware of phenomenology. We had an episode on Sartre where we argued about this very thing that, you know, is Sartre claiming that his big advance over Kant was that actually the self is not present in consciousness. That more like Hume says, the self is something that is built up afterwards, it is a concept. But then at the same time, if you read Sartre literally, he says, we always have a non positional consciousness of self whenever we're consciousness of anything. Whereas that seems something you directly deny. In other words, there could be creatures, probably animals, certainly us, in a lot of our normal day to day life, that when we're focusing on the things around us, there is no sense of self at all.

7:49 Ned: And you know what? Those animals are us. The Israeli neuroscientist Rafi Malek has shown that in intense flow experiences, self related circuits in the brain are suppressed. The circuits that are active when one is doing what intuitively are self related activities are suppressed during, for example, watching a movie where one is totally engrossed in it and loses one sense of self. So he does his experiments where he plays a Clint Eastwood movie and the subjects get lost in the movie. And when that happens, they're, you know, at least judging from the neuroscience, their sense of self is completely drifted away and they're just having vivid phenomenal experience.

And then he also does some really interesting experiments where he tries to experimentally create a situation where he can compare two cases. So one of them is he shows people a series of pictures and asks them their reactions, their emotional reactions to the pictures. And then that activates these self circuits. Then he shows them the same pictures a much more rapid pace. And they have to do a very quick identification of classification of them in one of I think three classes. And in that the self circuits are suppressed.

9:12 Mark: And then to distinguish between phenomenal consciousness and access consciousness. I know you've written a lot about the neuroscience of this as well. Usually blindsight is brought up as with certain cognitive defects, you could report in a much greater than chance rate when something is shown in say your right visual field, when you've had a certain lesion in your brain so that you're not conscious, you don't report consciousness, let's say, of anything in that visual field. But yet if I ask you how many fingers am I holding up in that field? You'll almost always get it right. So that shows that you are in fact conscious of it, but it's not the sort of consciousness that allows it to be used in reasoning. And so it's not phenomenally conscious.

9:49 Ned: What you describe is definitely not my take on blindsight. Just a first, just a factual point. If you hold up a number of fingers in the blindfield of a blindsight patient, they cannot guess the number of fingers, okay? What they can do is something Way more minimal than that. If you give them an alternative between certain options like horizontal line, vertical line, they can do that at a very high degree of reliability and then 90%. But they do not have, in my view, they do not have what I call access consciousness. They do not have free use of information and reporting and reasoning.

They can make certain kinds of guesses that are done on the basis of subcortical circuits. So I don't think they have either kind. I don't think they have either phenomenal or access consciousness. I don't take that as a case of phenomenal without access consciousness. I think the cases of phenomenal consciousness without access consciousness are very special experimental situations. One of the most interesting experiments has been experiments in what are called the no report paradigm, where people have been using brain scanning, been able to show that if you don't ask people to give any report, they've been able to show that conscious experience occurs before the global availability involved in access consciousness.

Too early for access consciousness. So it's that kind of very technical experiment that I think provides the best evidence for phenomenal without excess consciousness.

11:17 Mark: So that's the libet experience. That's what you're talking.

11:20 Ned: No, the people who've done this work. One person is a guy named Michael Pitts at Reed College. He's done these really cool experiments where he uses a form of brain imaging that's very sensitive to timing. His method is to give people what's called a dual task experimental paradigm, where their official task is to tell whether little disks on the periphery of their visual field get slightly dimmer, but what's in the center of the screen. There are lines that are constantly changing and occasionally form a square or a triangle. And what he finds is that he can set this thing up so that many people do have transitory awareness of the triangle or the square, but those are suppressed from their reporting mechanism because they're concentrating on this other task.

And then he finds that awareness of those figures occurs too early in the processing stream to be accessed at the time of the experience. So it's before what's called global broadcasting. So that's one item of evidence for phenomenal consciousness without access consciousness.

12:38 Mark: So are we positing when we're considering whether data the Android is conscious, to segue to the McLaughlin paper and back to sort of continuing our harder problem of consciousness discussion? The idea is that we're really concerned with whether data has phenomenal consciousness. Are we assuming that because he is passing the Turing test, because he is displaying language skills, because he's reasoning, because he's acting, he seems to have moral responsibility, that he does, in fact, have access consciousness.

13:05 Ned: Yeah, that's an assumption.

13:07 Mark: So McLaughlin's claim in this paper is that you had claimed that determining whether Data, that is an Android, that is a functional isomorph of a human being, of Brent Spiner, say, the actor that plays him is the way McLachlan puts it. In other words, he's behaviorably indistinguishable. Was there more to it than that, really, besides behaviorally indistinguishable?

13:27 Ned: Well, yeah.

13:28 Wes: Let's let Ned give his spiel on what a functional isomorph is. Superficial functional isomorph.

13:34 Ned: It's a lot more than behavioral. And distinguishability, for every mental state we have, there's a corresponding state. Whether it's conscious or not is up for grabs. But there's a point by point correspondence and isomorphism between us and Commander Data. And then the question is, if that isomorphism is produced by extremely different mechanisms, like maybe silicon mechanisms, how can we know whether Commander Data is conscious?

14:01 Mark: Now, are we just positing that he is a superficial isomorph of this sort? In other words, that he has these states? Okay, yeah.

14:07 Ned: That may not, in fact, be possible.

14:08 Mark: Right. So in the context of the show, for instance, you know, again, discounting the fact that he's super intelligent and all the other obvious differences that he has with human beings, it is behaviorable indistinguishability that would make us think that the best explanation for this is that he's a functional isomorph of us. That it could be that there is a very complex algorithm that is producing the same sorts of behaviors, intelligent responses, the appearance of moral responsibility, et cetera, but yet there's no corresponding state. His functional structure is very different. But that's just part of the setup of the thought experiment.

14:41 Ned: So what I argue is that functional isomorphism does, in fact, give us a reason to think he is having phenomenal states, and McLaughlin thinks not. I take functionalism more seriously than McLaughlin does.

14:53 Dylan: Am I wrong to say that we're using superficial functional isomorphism in this case? You're using it as. The reason why we would say that Data has consciousness is we're on the one hand saying that we don't have any sign that Data is having a phenomenal experience because he's doing everything behaviorally the same way as we do.

15:14 Ned: But we do have a sign. If he's doing everything behaviorally, we do have a sign that he has phenomenal Experience.

15:19 Dylan: And I was just going to say other than that. But we don't believe the behavior sign. So the behavior sign is insufficient for us to say he's having a phenomenal experience. So we throw that one out and now we're absent any other sign for it. And so now we're going to try to come up with an account for the structural account of him to come up with a way to say, well, yeah, maybe he actually is having a phenomenal experience. Is that basically right?

15:43 Ned: Yeah, that's right.

15:44 Dylan: So then that goes a little ways to answering the question of saying that we don't have a sign for what it means to have an experience.

15:53 Ned: Well, I think the functional isomorphism is a sign. It's not a totally reliable sign, but it is a sign. So I think it does give us a reason, but it's a defeasible reason. Now, McLaughlin thinks it isn't even a reason, and he gives a lot of arguments against functionalism by way of arguing that it isn't even a reason. So he thinks we just have to pay attention to the physiology.

16:16 Mark: So I thought that in reading the harder problem, yes, you were saying that there's a prima facie reason to believe that he has phenomenal consciousness because of the behavior, because of the functional isomorphism, but that this is defeasible given what we know. I mean, this, I think, is what McLaughlin's argument really comes down to, is that the other consideration that you consider in that paper, which is that he has very different physiology than we do, overrides that.

16:40 Seth: Right.

16:40 Mark: That in our case, we know from first person experience that we are conscious as individuals. Well, why is the problem of other minds not a really serious skeptical problem? Because other people are similar to us physiologically. You know, I know that in my own case, I'm conscious. I can look at the science. This obviously is not the case for ancient Greeks, but we at least now could be pretty confident that other people are also conscious, not only for the normal behavioral reasons, but because we can chart correlations between what seem to be phenomenal states and brain states.

What exactly the relationship between those are? Well, that's the hard problem. We don't maybe have to decide that. But just knowing that that's enough of a reason for us to say other things that have similar brain states are also going to have phenomenal states. That seems a reasonable leap. It's not 100%. You can't argue definitively against the philosophical skeptic about that, but it's enough for any purpose that we might need but yet if some other being like Data has silicon instead of brain, then we don't have that near guarantee of sentience.

And so that disanalogy from us is enough to knock down this prima facie reason for attributing sentience to him.

17:53 Ned: I pretty much agree with that. The difference between me and McLaughlin is McLaughlin thinks that the functional organization gives us no reason at all to think the data is conscious. I think it's some reason, but it's a defeasible reason. The main thing I was arguing for is we at the moment have no conception at all of how to find out whether data is conscious. Even if we understood, even if we solved the hard problem for us and knew what it is about our physical makeup that made us conscious, that might not give us any way of extrapolating to completely different creatures. That was really the issue that I was trying to bring out. And that's why I call it the harder problem. Because even if the hard problem were solved, it does not follow that the harder problem would be solved. I think the problem of other minds is this really difficult problem.

18:43 Mark: One thing, when we were talking about the harder problem paper, we didn't really say what you meant by meta inaccessible, that you were saying that not only is knowing whether data is conscious inaccessible to us, in other words, we're not in the right epistemic position. We are not him to see his phenomenal states, but we have no way it's meta inaccessible. In other words, we don't even have the conceptual apparatus to determine how we would figure out whether he's conscious or not.

19:07 Ned: You know, there are many things that we don't know where we have a pretty good idea of how to find out. But this is different. We don't know whether data is conscious, and we have no idea how we would go about finding out whether data is conscious.

19:20 Seth: So one question I have about that is, isn't that statement true of other human beings? Or is it because do we have some kind of privileged access to our own experience which gives us the meta framework to solve that? When we got to that resolution, I felt like you could apply that same criticism to the problem of other minds, just as you mentioned.

19:38 Ned: So my take on that is that we have extremely good reason to think that the neural basis of our mental states gives us the information needed to understand whether they're conscious or not and what the conscious qualities of them are. And that to the extent that we think other people are similar to us in that respect, we have extremely good reason to think that they're conscious too. So other people who are physiologically like us, I think we can be almost completely sure that they have consciousness much like ours. In fact, we do know of some differences between people.

You know, different brains are different in the cortical thickness of different areas. We know enough about how the brain works to know that that actually underlies differences in perception, but it also underlies similarities in perception. So we have a pretty good way of telling when other people are perceptually similar to us.

20:32 Seth: You're saying that in the context of some awareness of, let's just call it neurophysiology or some sort of scientific neuropsychology. It feels like what McLaughlin's pointing to page 338. He's talking about how we're tuned to faces and voices and we're disposed to attribute consciousness based on conscious states like and desires and intentions, with very simple triggers. So to the extent that I have little to no experience of the literature around the scientific inquiry into all of these various mental states, from my position as just a regular schmo, isn't the Commander Data problem and the Ned Block problem for me essentially identical or do I need to go research all of that neurophysiology?

21:22 Ned: So I guess what I think is that even prior to understanding as much as we now do about the neural basis of consciousness, we had a pretty good idea that it was something about the body, including the brain, that was underlying our anomalous experience. So I think we had a pretty good sense of the same point without a lot of the detail. And I think your ordinary Shmuel on the street has read something about how the brain underlies experience. So he has some idea about what the epistemology of the attribution of their minds is. But, you know, in the last 20 years we really have learned a huge amount of, about the neural basis of experience.

But of course, you know, even, I don't know, 40 or 50 years ago, you know, Wilder Penfield was doing these experiments where he was sticking electrodes in people's brains and they were saying what their experiences were. It's really quite striking that certain areas, when stimulated, will give rise to certain experiences. You know, now we have, as part of the lead up to operations to control epilepsy. We have a lot of patients who have these microelectrode arrays on their surface of their cortex and we can stimulate certain areas like we stimulate the fusiform face area, for example.

People start seeing faces everywhere. There's a video on the web actually of A Japanese patient who is saying what his experiences are when they're stimulating his face area. So we have a very good idea of at least some part of the neural basis of what certain experiences are.

22:53 Mark: So one of the points raised in the harder problem paper, and we actually read that in conjunction with Pap's, David Papenow's Could There Be a Science of Consciousness? Is Papenow seems to take very similar view to yours in some ways, but seems to take seriously the idea that the reason the harder problem seems so hard is because it's sort of a Wittgenstinian response that it's a problem of language, that consciousness, phenomenal consciousness, is a vague concept and Data is perhaps a borderline case. And so saying yes, he is or no, he isn't would be entirely inappropriate linguistically.

Right. The reason that we came up with the concept of phenomenal consciousness was from our own experience. We have meat brains. And so why would we even think, even if Data has something like phenomenal consciousness, that it's going to be a single term, that disjunctivism is going to make sense, that we would say, yes, we are phenomenally conscious and he is phenomenally conscious. Maybe he just has something else.

23:47 Ned: I think that kind of reasoning might be reasonable when applied to a creature like, say, a fish, a creature that really may be a borderline case. In some ways, its processing is like our conscious processing, in some ways like our unconscious processing. But that doesn't apply to Commander Data. The point about Commander Data is he may be phenomenally exactly like us. Forget about the issue of whether he might have a different form of consciousness. Let's ask the question of how we would know whether he's exactly the same phenomenologically as us. And that is not a borderline case issue, and it certainly isn't a verbal issue. It is a genuine question which we have no idea how to answer.

24:32 Seth: So in the Commander Data case, just to tie this to what you were just saying, Ned, could we stipulate that stimulating all of the relevant parts of his cranial areas, sticking probes in the right places and stimulating would cause him to have the same states that you were just talking about where the Japanese guy, where he sees faces. Is that entailed in the isomorphism?

24:55 Ned: No, that's not. So that would be a different kind of case that moves in the direction of giving us real neural evidence that he has the same consciousness as us. So, no, no, I was assuming that all his psychological mechanisms work differently from ours. But if you start stipulating that there's a correspondence even at the neural level, then it begins to look much more like you have a real case that he is a conscious being.

25:21 Seth: Okay, that clarification helps.

25:23 Mark: Well, this is a question that came up in our previous discussions about. So it seems like going into the brain and stimulating something directly is a different kind of input than one would normally receive. And so when you're talking about how do we describe a functional model of a person and so describe the functional isomorph of a person, we're just talking about inputs and outputs at the sensory level. We're not talking about materials. In other words, if you pump alcohol into the brain, that's going to change the way the sensory inputs and outputs operate, but that does not count as part of the functional model. Is that correct?

26:00 Ned: Yeah.

26:00 Mark: Okay. So the fact that we have that thing with being able to see the faces everywhere if you put an electrode in the right place, and Data doesn't, doesn't mean he's not a functional isomorph. It just means that there are architectural differences, which, which was the whole point of us defining what a functional isomorph is. We're saying, despite the architectural differences, there are functional things in common.

26:20 Ned: That's right.

26:21 Dylan: I'm wondering why the case of Data stimulating different areas of his brain would mean that we were making a tighter connection. Because even if structurally his brain was completely different and the map between the sensory inputs and the phenomenological experience was completely different than ours, even topologically different, we're still supposing a kind of functional arrangement in his case that's causal in some respect, so that there would be a mapping. And what we're leaving open is the question is that his mapping is completely different than ours, but the functional relationship is isomorphic.

27:00 Ned: There are different levels of functional relationships. So the one that I used in describing Data is the superficial level. It's the common sense level. So if we have a pain, he has some corresponding state. If my pain causes me to say, then there's a causal process leading from his pain to emitting the sound. Ouch. But then there are functional relationships at the brain level too, and those can be described functionally. And even there are functional relationships at the elementary particle level that can be described functionally. So there is a hierarchy of functional levels. And the lower you go, that is the closer to physics that you go in this hierarchy and you say, you, you stipulate that Data is like us in that respect, the better a case you have that he is A conscious being.

27:48 Seth: Would it be fair to say then that in the circumstance where we're talking about an isomorphism related to those abnormal inputs, that we're talking more about the case of the Chalmers example where you talk about fading qualia versus that. It's really more where you're substituting one piece of silicone for one neuron firing or whatever the. That's. It's more that case as opposed to the Commander Data case.

28:13 Ned: Well, that would be a dramatic case of varying at a physiological level while keeping the superficial stuff the same.

28:22 Mark: So are we capturing the same distinction? McLachlan spells out what he takes to be your view on a superficial functional isomorph versus a deep functional isomorph. That a deep functional isomorph of a normal adult human being is a being that shares the functional states that the as yet unknown scientific psychology that is in fact true of us will associate with the mental states that the human being has?

28:43 Ned: Okay, so that's a more fine grained functional isomorphism where the fine structure, the behavior on psychological tests is similar between us and data. So I was thinking of a different kind of hierarchy where we go down into the brain, but that example is where we make. The timing of people's response is the same too. And of course that suggests a physiological similarity, since why would the timing be the same?

29:11 Mark: We had made that distinction between regular functionalism and psycho functionalism. Regular functionalism was more at the level of the superficial. So we read the classic Putnam paper and the classic Armstrong paper on this. So Armstrong was kind of giving the a priori superficial functionalism. Just look at our everyday psychological conceptions. You can see that psychological states in fact correspond to typical causes and effects in internal structure. So that gives rise to the idea of a superficial functional isomorph. Whereas you had included Putnam as among those the psycho functional functionalists.

Like if you just look up Wikipedia functionalism, it actually says, oh no, there's machine table functionalism, that's one type. Psycho functionalism, that's a different type. Are we really saying that to do an accurate machine table is to know the best possible psychological theories and get the timing right and everything like that?

30:02 Ned: Well, of course, machine tables can be at different levels of grain too. One thing I think I said about Commander Data is that where we have a memory intensive process, he has the corresponding complexity and procedure. And where he has a memory intensive process, we have a corresponding complexity in procedure that has the same effect without the memory intensive process. So I was thinking about differences in processing at the psychological level. Now if you stipulate that there's a psycho functional correspondence that increases the plausibility of inference to phenomenal states like ours.

30:38 Mark: And we're not assuming that you could do a machine table functional analysis of him, are we? Just the fact that we're talking about him being silicon, that you could have a neural net model or some other model that doesn't yield itself to a machine table flowchart, sort of diagram of his mental processes, that's not essential for defining what a functional isomorph is, right?

30:57 Ned: No, no, I don't think I even mentioned the machine table correspondence in that paper.

31:02 Mark: Guys, were there other things in the McLaughlin, do we want to pull out any particular quotes, any particular issues?

31:07 Seth: Yeah, there's one thing for me which I found curious and I'd love to hear what Ned has to say about it. So Starting on page 358 he goes on and he starts talking about all of these mental states that you can have without phenomenal consciousness or all these things that we could conceive. So you can think of perception without phenomenal consciousness, language and intelligence, a current thought, self consciousness, moral personhood. And on page 363 he says it might be wondered why phenomenal consciousness matters if it isn't required for any of the things mentioned above.

But we've just been talking about what's conceptually required. Although data is coherently conceivable, he may well be nomologically impossible. Side note, at some point in the past, nomological became popular in the literature, and that was past my time. But he says further, it's a mistake to think that if phenomenal consciousness has value, that value must be instrumental. So in other words, if phenomenal consciousness isn't causal, we ask ourselves why do we even care about it? And he says subjective experience subjectivity is intrinsically valuable.

So the best short answer concerning the value of phenomenal consciousness is the Louis Armstrong like answer. If you have to ask, you'll never know.

32:20 Ned: Yeah, I think I agree with all of that. So it has been a persistent problem to find a function that can be done consciously but can't be done unconsciously. And actually in Stan DeHaan's recent book he has a whole chapter on this and there are lots of cases where people have not been able to show something to be done unconsciously. So a pretty simple example is you can do one stage reasoning unconsciously, but not two stage reasoning. He has a number of experimental paradigms which involve one stage reasoning or two stage reasoning. And it looks like from those experimental paradigms that two stage reasoning cannot be done unconsciously.

So it does look like consciousness does have a real function. No, basically it's a kind of mental blackboard. It allows you to keep some things in mind at the same time and do conscious manipulations on them. And at least so far, nobody has been able to find humans doing those things unconsciously.

33:22 Seth: Wouldn't that fall under the category of what you described at the beginning of this as access consciousness?

33:27 Ned: Yes, yes. So that's all in the domain of access consciousness, but the two correspond fairly well. The question arises, if it takes consciousness to do two stage reasoning, maybe phenomenal consciousness is part of the explanation of that, but nobody knows whether it is for sure.

33:44 Dylan: I haven't heard this term access consciousness before, but it seemed to me that phenomenal consciousness would be acting as a special case of access consciousness, that access consciousness would be a generalized version of it that could operate on things that weren't phenomenal, that you could be operating on things that were conceptual, but you'd have that kind of interaction. But that it's a special case of phenomenal consciousness. I mean, the other way around.

34:10 Ned: So you're saying that phenomenal consciousness is a special case, but there might be access consciousness without phenomenal consciousness?

34:16 Dylan: Yes.

34:17 Ned: Yeah. But there might be phenomenal consciousness without access consciousness too.

34:20 Dylan: Either you take one being the special case of the other, or you say that they're instances of the same kind of general case, but they're information processing.

34:28 Ned: I guess what I think is that we don't know that phenomenal consciousness is an information processing process, and that it might be more dependent on molecular mechanisms or electrochemical mechanisms. So I think these are separate realms that are probably in humans anyway. Phenomenal consciousness is needed to implement the information processing mechanisms of access consciousness. I think I would take that as a lesson of what DeHaan foundation is, that we need phenomenal consciousness to do certain jobs, which makes sense from an evolutionary point of view too. Why do we even have it? I suspect that the reason we have it is because it does certain things that at least in our evolutionary history, could not be done without it.

35:13 Seth: Now that's fascinating. And that also, I feel like circles back to where we started when we were talking about Sartre, where this idea that phenomenal consciousness could in a certain sense, play the same role as a limit or horizon. It's the field in which the field in which conscious activity takes place in some very interesting way. I like that as an explanation, or at least as a hypothesis.

35:37 Ned: So you might be interested in some work by the philosopher of biology, Peter Godfrey Smith. He gave a talk at NYU the year before last and has written some articles. And some of it is in his book on octopi, suggesting that there are examples in the animal kingdom of very good candidates for being zombies. His leading example is the solitary wasp. So what he suggests is that the solitary wasp first of all is an example of a creature that has no wound tending. That means if you cut off a leg or a wing, it does not favor the damaged part. It just continues to do everything it can do without really paying any attention at all to bodily damage.

In addition, it has extremely stereotyped behavior patterns where, you know, there's these famous examples of if you intervene in its routine, it does the same thing over and over and over again. And this is all coupled with an excellent visual system. It is able to track and ambush its prey. It has great constancies of vision. He notes that there might have been a precursor being that maybe had some more sophistication, but that the wasp has great evolutionary pressure to reduce its brain size for flight. So mechanisms of thinking and mechanisms of phenomenal consciousness, you know, maybe they just took a path where it didn't use any of those things.

And so it doesn't have the kind of whatever sophistication phenomenal consciousness buys us in terms of our access consciousness. It doesn't have. So it's a really good candidate for a non thinking zombie. Non thinking phenomenal zombie.

37:24 Seth: I definitely have to go look that up. You got me excited because I'm a big Octopus fan. I don't eat them because I think they're smart, they're intelligent. So I have to go see what he says about octopi.

37:33 Ned: Oh, well, his book on octopus is terrific.

37:35 Mark: So it sounds like you might have had some evolution in your thinking. You'd also read this 1978 troubles with functionalism paper where I'm not sure if you were just claiming or toying with the claim that qualia are really not in the domain of psychology at all. In other words, phenomenal consciousness, if that can't be used to explain any of the processing that's going on in thinking, it cannot be used to explain intelligence, then it's an undeniable phenomenon and it seems like it should be mental. But ironically, psychology, the domain of the mental, if it's not usable as an explanatory tool in psychology, which seemed to be the reason that, for instance, Dennett was just rejecting quality altogether as just being useless as an explanatory tool, then they're something else.

38:22 Ned: I don't have the view of Dennett who doesn't believe in consciousness to begin with. So I was focusing on the idea that the domain for thinking about consciousness might be at the level of brain science. Of course that would have psychological ramifications. So there's certain things at the psychological level that a conscious brain activation could do. Certain things that advise us like solving two step problems instead of one step problems, but that the real explanation of consciousness was to be found at the level of brain science. That's what I was arguing for.

38:53 Mark: Sure. So that's just a view of psychology that is taking functionalism seriously, is that right? So that what functionalism should be doing is at least psycho functionalism is. Psychology is all about mapping correlates and things, you know, within human behavior to mental states to. So if you're coming up with the whether it's machine table or whatever model it is you're coming up of the psyche, you're really giving a functional explanation. And Qualia didn't seem to be playing an element in that.

39:22 Ned: I forget what I said in that paper. I haven't looked at it in many, many years. But what I would now say is that qualia do play a psychological role, but that the explanation of that role would have to be found at the level of the brain.

39:35 Dylan: I was also drawn to the section that Seth referred to in the McLaughlin partly because of it giving me some entree into again what the signs of different kinds of consciousness are. And you replied, Ned, that you agreed with this notion formulated by McLaughlin that the best answer concerning the value of phenomenal consciousness is that if you have to ask, you'll never know. So what that meant to me was that there are lots of different forms of consciousness, that is ways in which an entity is responding, processing inputs and generating outputs.

And there are lots of different ways that's happening. One of those ways is the one that we frequently identify with ourselves, which is phenomenal consciousness, which is the way in which we talk about what it's like to have an experience, what it's like to be ourselves. And if I have that last comment that if you have to ask, you'll never know, it makes it sound to me like, well, it might be the case that phenomenal consciousness is kind of the super added part of consciousness that we have experience as human beings, that is the character of our experience of our lives, but really doesn't matter that much in terms of the ins and outs of us as entities.

And so it's sort of like the color of the car, but it doesn't matter for the car operating as a car. That's what it sounded like to me.

41:01 Ned: McLaughlin made an important distinction between instrumental value and non instrumental value.

41:09 Dylan: Yep.

41:10 Ned: So I think phenomenal consciousness has both instrumental value and non instrumental value. So the instrumental value is not like the color of the car, it's like what's going on under the hood.

41:22 Dylan: Absolutely.

41:24 Ned: So it's because we have phenomenal consciousness that we can solve two step reasoning problems.

41:28 Dylan: Okay, so we're brought back to your response earlier. Okay.

41:31 Ned: Yeah. So when he invoked the Louis Armstrong quote, he was talking about the non instrumental value, the value of, you know, like good conscious states, bad conscious states, apart from their role in enabling us to do things. That's where I endorse that idea. The idea that its non instrumental value is something that you have to appreciate from the inside.

41:53 Mark: So I still feel like when I raised Papenow's vagueness charge, there's still something in the neighborhood here. As McLachlan is saying, maybe the data can do all these other things. It doesn't make sense for him to be held on trial to say whether he has phenomenal consciousness or not. If he acts as a moral agent, if he seems to have plans, projects, if all these things that potentially could happen only with access consciousness, not with phenomenal consciousness. There's another essay in Blockheads that's specifically about this. This Jeffrey Lee's alien subjectivity and the importance of consciousness.

I don't want to get in the details of that, of course, since we didn't all read that, I guess I want to just push you a little more on what you thought about the idea that there could be things, you know, even if Data does not have phenomenal consciousness, maybe he has something that is like phenomenal consciousness, that for all moral purposes we might want to say, fine, maybe we shouldn't sweat about the harder problem too much because he meets all these other criteria.

42:45 Ned: I agree with Lee that there might be other kinds of states that have the same moral epistemological upshots that aren't consciousness, or at least not our kinds of consciousness. But I don't take that to devalue the issue of consciousness. I think the issue of what the nature of it is or the importance of it, I think that it's possible that there are other things that have that same nature and importance that aren't it. I would agree with him that far.

43:13 Mark: It also just seems to be another kind of harder problem. Like if you're going to say, okay, phenomenal consciousness for us is the what it is like to have an experience. Well, there's something that creatures that are very much unlike us, they don't have phenomenal consciousness, but they have something like what it is like. Like I don't even understand that.

43:32 Ned: It's hard to get your mind around that, that there might be something that doesn't exactly have I what it's like, but something close to it. So I think that's another. It's in the realm of the borderline case thing that we talked about before where like fish might be a borderline case. Maybe there would be a verbal issue there about whether to call it consciousness. That's possible. But I don't think that the issue of data is a verbal issue because I think that we can ask the question how would we find out whether data's consciousness is exactly the same as ours? And that is not a verbal issue, that is a genuine puzzle about how it would ever be possible for us to do that.

44:12 Mark: To me, the elephant in the room is still Kripke Putnam on rigid designation. The whale oil example filled such a large part of the hard to problem paper and we actually didn't end up bringing that up at all. Maybe we can just revisit that briefly to say what the relevance of this is. The idea is that maybe what to us is consciousness and we think that if I'm picking out my own phenomenal consciousness, then I'm picking it out rigidly, as Kripke would say. So in other words, I don't know if I'm not a brain scientist, what brain state is going on?

What brain state is perhaps identical to who knows? But certainly correspondent with supervenient under the phenomenal state, I'm pointing specifically at that. Then if we want to say if data has something that's like phenomenal consciousness but doesn't have that brain state, then there's a disconnect there. This is, I think, the thing that Papena was pointing to that we don't know that whether the natural kind concept is in fact phenomenal consciousness itself. And maybe in fact that's a superficial concept. We might say it's a phenomenal concept, which I think this is actually Kripke's solution out of this, right?

That whatever we have and what the aliens that don't have brains or data or whatever haves, those are both going to be consciousness. But according to the strict materialist. No, no. Once we've discovered that in us that phenomenal consciousness is identical to this brain state, then if there are creatures in some other possible world who seem to have something like phenomenal consciousness but don't have brains, then that must not really be phenomenal consciousness. Do you want to talk about the general relevance this sort of example has and what we're talking about?

45:49 Ned: So look, I think consciousness does, the term does have a natural kind semantics and not a superficial functional kind semantics. So Gareth Evans example was that in some cases, whether a term has functional kind semantics or natural kind semantics can just be a matter of decision. And he considered an Eskimo word that they use to designate the oil that they use that they get from Wales and seals. He said, well, look, maybe a petroleum company could introduce a mineral substance that had the same use and call it by the same name, in which case it would be a superficial kind term, just something with a certain use and overall appearance or something, but not a natural kind notion.

So what I was saying was the term consciousness is not like that. The term consciousness really does have a natural kind semantics. But then there is the point that there could be different natural kinds involved here. So there could be a brain natural kind and a phenomenal natural kind. And so I think that there may be another kind of genuine indeterminacy there.

46:55 Mark: The question is with whale oil in particular, is that if the Eskimos continue to use this stuff, which is no longer from whales, but they're still calling it by the same term whale oil,

47:05 Ned: and most importantly, it may not even be an animal product at all, it might be a mineral product.

47:12 Mark: So that just tells us something that if they don't actually care, you'd think that they point at that stuff and they say, you know, whatever the word is for whale oil, that they're rigidly designating it. In other words, they're picking out the actual chemical property, but then you swap it with something else, they keep using it in the same way, they keep using the same term for it. That just shows that actually it wasn't after all a rigid designated natural kind term.

47:34 Dylan: This gets us back to the data and human consciousness problem, right? Is that does consciousness refer to whatever it is that is manifest both in us and in data in the same way, but is manifest in different forms, that is one is mineral product and the other is whale flesh? Or are they different natural kinds that actually are different in some important way, such that they're not referring to the same thing? And I take it Ned was point out in the first paper we read that we can't tell we don't know enough about what we mean by consciousness to be able to say what the answer is. Whether or not one is, to use the analogy whale oil as from whales versus whale oil as from minerals, we don't know enough.

48:20 Ned: I think that consciousness does have a natural kind. Semantics. And when we ask the question of whether data is conscious, we're asking whether it has that same natural kind of. And of course, since we know that the brain is different, it's the phenomenal natural kind that's at issue. If it was really like the whale case, then there would be an indeterminacy about whether we were just asking whether he was the same functionally as us, which, by stipulation, he is.

48:48 Dylan: I thought that it was like the whale oil case because functionally we're saying that they're the same, that data is behaving in a deep, isomorphic functionality way with us. And in this case, the whale oil is. I'll use the term whale oil from being from whales and the term mineral oil being from minerals, just to make the distinction, but they're functionally performing the same thing. The distinction is that there's some characteristic about ourselves that we don't know if data is manifesting it. So that there's something about consciousness which is a natural kind term, that we're not sure that data actually is manifesting consciousness as a natural kind. And we furthermore don't have a way to know.

49:32 Ned: Yeah, exactly.

49:33 Mark: So does that seem to capture what McLachlan's argument is, that he thinks that consciousness is a rigid designator in us, and so it does drill down to the brain state. And so we've kind of ruled out, and it's okay for us to rule out in advance. You know, that's a sufficient reason, though not certain reason, for us to say that whatever's going on with the functional isomorph with silicon, there's not phenomenal consciousness there. And you just don't think that we have the epistemic warrant to make that leap.