Ep. 222: Debating Functionalism (Block, Chalmers) (Part One)

0:00 Mark: You're listening to the Partially Examined Life, a podcast by some guys who at one point set on doing philosophy for a living, but then thought better of it. Our question for episode 222 is still what is the mind? And we're considering arguments attacking functionalism through Ned Block's troubles with functionalism from 1978 and defenses against some of these attacks in David Chalmers essay absent Qualia, fading Qualia, dancing Qualia from For more information please visit partiallyexaminedlife.com this is Mark Linsenmayer with Qualia dancing the Macarena in Medicine, Wisconsin.

0:34 Seth: This is Seth Paskin, paradigmatically embodied in Austin, Texas.

0:39 Wes: This is Wes Alwan in Cambridge, but supervening on the economy of Bolivia.

0:46 Mark: All right, we are back to Ned Block, who in preparation for our interview with him next episode. Very exciting. This is basically the second half of episode 221. Our four part episode, which is itself was the second part of our 218. 219. We've done a lot of these damn things.

1:06 Wes: Well, I'm sure the fans love it.

1:10 Mark: We had another Chalmers paper this time which was shorter yet equally lucid and pleasant is the last one we talked about. And we had another Block paper which also very lucid but very technical and quite long and there's some good parts in it, but parts made me cry.

1:29 Seth: Agreed.

1:30 Wes: What made you cry?

1:31 Mark: Ramsey sentences, which are not actually that difficult when I slow down. And there is things about multiple instantiation, about how you have to define functionalism with reference to a specific psychological theory. This is one of the things I'm hoping you guys can clear up for me. There was some dense stuff on that. So the thing that we're actually going to be reading for next time with Block related to this is an essay called Michael Tye's Homunculi Heads and Silicon the Importance of History to Phenomenology. And then Block's reply fading Qualia.

That's kind of the most important thing we need to get out of today is to understand Chalmers, fading Qualia and Dancing Qualias objections to Ned Block's famous absent Qualia famous thought experiment, which is in turn supposed to refute functionalism that we talked about last time. So I felt like maybe we should instead of just trying to stick just to the block for the first half, let's spend the first half dealing with that issue that we for sure have to hit. And then in the time remaining, if any, we can touch on those perhaps more difficult issues in Ned Block's rather lengthy essay.

2:38 Wes: Do we want to talk any more about his characterization of functionalism, or are we going to jump straight to absent Qualia and the homunculi headed systems that he's talking about?

2:49 Mark: I think Block had some interesting ways of putting it. I tried to incorporate those a little into last discussion, but yeah, was there a particular quote you want to start with about that to at least remind folks what functionalism is and what block in particular takes functionalism to be, since the Putnam and Armstrong that we presented last time were quite different for each other.

3:07 Wes: So this is page 262 or PDF page 2 so one characterization of functionalism that is probably vague enough to be accepted by most functionalists is each type of mental state is a state consisting of a disposition to act in certain ways and to have certain mental states given certain sensory inputs and certain mental states. So put, functionalism can be seen as a new incarnation of behaviorism. Just skipping down a little bit. Functionalism replaces behaviorism, sensory inputs with sensory inputs and mental states, and functionalism replaces behaviorisms, dispositions to act with disposition to act and have certain mental states.

We got into this very sort of evolution of functionalism out of behaviorism last time. Where behaviorism you describe a mental state like pain for instance, in terms of dispositions to behave in response to certain stimuli. Pain is the thing that's apt to make you go ow, to put it very simply, or to engage in aversive behavior, so on and so forth. Where functionalism allows you to refer not just to inputs and outputs, stimuli and behavior, but to internal states of the system, in this case the person of the mind itself. And that could be certain beliefs or psychological states or anything you like.

Functionalism allows you to take that into account. That's why you modify disposition to act with disposition to act and have certain mental states, because the outputs may not just be behaviors, but they may themselves be mental states that then lead to further outputs, and so on and so forth. And the end output is behavior. But it's not the only one we're concerned with. And sensory inputs become sensory inputs and mental states because in a way the intervening mental states or mental states that are causally related to the mental state you want to explain are also, well, let's just say they're causally relevant to the description.

So hopefully that's not complete gibberish. We have been saying this over and over again in one way or another for many episodes now, and it's always

5:25 Mark: weird to talk about pain in terms of that, because pain doesn't Seem like something that you have to use other mental states to explain. It really does seem like it has certain typical causes and it has certain typical behavioral effects. But it's not like you feel pain if you think you should feel. But it's not like there's a belief involved. There's a desire. I don't want to feel pain, so I won't feel pain. Like it's.

5:49 Wes: Well, it certainly could cause a belief. So it could cause the belief that, for instance, that I should avoid such and such a behavior.

5:56 Mark: Sure.

5:56 Wes: And so that's one of the causally relevant beliefs that a functionalist is thinking about in this case.

6:01 Mark: And in that causal thing, it also sort of makes reference to a desire to avoid pain. So there's the feeling of pain, there's the desire to avoid pain. Those clearly have some sort of internal logical connection between them. And then there's the belief that some particular thing will cause me pain or is causing me pain, might have the

6:20 Wes: belief that more pain, more gain in some circumstance. So pain doesn't just mean aversive behavior. Depending on kind of the beliefs that are causally related to pain, it might involve me persisting through that. And that's actually one of the important things that behaviorism can't deal with, that functionalism, in a way, is designed to deal with that kind of complexity.

6:45 Seth: Right.

6:45 Mark: And I think throughout these discussions, Chalmers is going to be defending functionalism. So the issue can't be that. Like in Armstrong from last time, it seemed like he was trying to reduce mental talk to functional talk and say that there is nothing more to a mental state than sort of a functional flowchart or a causal connection to behavior and to other mental states. So Chalmers is going to explicitly call what he's arguing for non reductive functionalism. It's just the idea that functionalism is the mark of the mental. And the way that Chalmers puts that is by using the term supervenience, a term that I've used enough, but I still constantly forget exactly what it means.

And I think it really just has to do with logical dependencies or physical dependencies. The functionalist claim there are various sorts. Of course, there are reductive ones and non reductive ones, but they all have in common the fact that whenever you have a particular functional state, if functionalism is true, then you're going to have that mental state, and whenever you have that mental state, you're going to have that particular functional state. And we're not saying anything else about which causes which, or are they the same thing ultimately metaphysically, and that's given

7:52 Wes: what Chalmers calls natural necessity. So these are empirical facts about the word. It's as we know it's important, Chalmers, that it's conceivable or it's metaphysically possible to have certain functional states without certain qualia or certain mental states. But on his view, it's not empirically possible. So let me just give you a quote because I think this is important to point out from the very beginning, as Mark has done, because they may, Chalmers and Block may be talking past each other to a certain extent. They might be talking about two different things.

Because Block says up front in the beginning that when he talks about functionalism, he's saying, I mean, an identity statement analogous to a physicalist identity statement. Functional organization just is constitutive of mental states. And this is what Chalmers says towards the end he said, the arguments therefore fail to establish a strong form of functionalism upon which functional organization is constitutive of conscious experience, but they succeed in establishing a weaker form on which functional organization suffices for conscious experience with natural necessity.

We can call this view non reductive functionalism as it holds a conscious experience as determined by functional organization without necessarily being reducible to functional organization. As things stand, the view is just as compatible with certain forms of property dualism about experience as with certain forms of physicalism.

9:17 Seth: There's actually an amusing and interesting interplay here, and I'm not sure whose side I'm on, but when Marx said the kind of functionalism that Chalmers is arguing for, I think I'm sympathetic to Block saying nobody actually argues for functionalism. They argue against non functionalist theories. And Chalmers essay is really kind of an extended critique of criticisms of functionalism. And that's why he ends up saying, well, the best I can get to is this weak form of non reductive functionalism is that he hasn't actually argued in favor of functionalism.

He's just essentially said trying to argue for something other than functionalism doesn't make any sense. So this is kind of what we're left with. It's a strategy, but it's not a great strategy if you're trying to establish something. I think that's what Block says. Well, I don't think it's what he says explicitly in the essay. I think the thread of the interplay between the two, I'm not sure they're talking past each other. I do think that Chalmers is responding directly to Block, but I don't Know that I buy it.

10:23 Wes: What Block says is that most philosophers just assume functionalism and they don't bother to argue for it. Once you accept that physicalism has certain fatal problems, you know, why is it that just that only this particular stuff can be conscious? Then you say, okay, this is the best we can do, and you work on that. But I think Block gives a set of arguments that I think are pretty convincing, pretty persuasive, that actually functionalism doesn't really work if you think of it as constitutive. I think what Chalmers is doing is something a little bit different in a way. I think ultimately it's compatible with Block's critique of here, of constitutive functionalism or reductive functionalism. Because Chalmers is just saying, I'm not embracing reductive functionalism.

11:10 Mark: I just think that Block has so many lines of attack on functionalism that it's trying to catch for all forms of functionalism. The whole reason that you would take functionalism is because there was something wrong with materialism. And I think we saw that in both the Putnam and Armstrong. They both wanted to be scientific, which sounds like you should just be a materialist. Sounds like you should say that mental states are the same as brain states at a type level.

11:32 Wes: So I think we should use the word physicalist just because functionalism can be materialist. The physicalism, though, is the alternative that functionalism opposes.

11:42 Mark: Well, so they're both. Again, the starting point is wanting to be materialism and assuming that physicalism and materialism are the same thing. Right. Okay. And so it was an advance when guys like put him in Armstrong in the late 60s, early 70s are arguing that. No, that would be, for instance, chauvinistic. If you want to say an octopus feels pain, but octopus obviously doesn't have the same. It has a brain, but it has a brain that's very differently evolutionarily involved. And the parts that go off when there's pain are, I think, not the parts that go off in us.

So even just forgetting about androids and potential aliens with really weird biologies, like just different creatures on our Earth, that there's something chauvinistic about. Yeah, there's something chauvinistic about physicalism. And so wouldn't it make sense that we could say that there's human pain and there's octopus pain? And what makes them both pain? Well, it's their functional role. Functionalism kind of became a default position when you object to physicalism. And Block is kind of giving a Rejoinder to that that says actually the advantages that you thought functionalism had over physicalism are something very parallel plagues.

Functionalism as well. It itself is chauvinistic, or depending on how you define it, might be in fact too liberal. It might end up ascribing mentality to collections of rocks and things like that that you don't want, or to economies. Right. One of the points he's making is about how you specify inputs and outputs.

13:12 Seth: Yeah. Which is Wes's little tag when he came in. Yeah.

13:15 Wes: I mean, I think if you were to talk about his objection, in the grand scheme of things, the big picture for his objection is what he calls homunculi headed systems. And this is sort of an outgrowth of Searle's Chinese room argument in which a system is made to pass a Turing test to simulate consciousness. Just using a guy in the room is translating Chinese to English or whatever. The other example is just you could make the whole nation of China, as long as its population operates in the right way and according to the set of rules, you could make it functionally equivalent to a brain, let's say.

13:54 Seth: Yeah.

13:55 Wes: So the idea is that this is where the homunculus part comes in. You have a consciousness which consists of smaller conscious units. And some of their operations are like, predicated on their consciousness. Right. So in a brain, everything just happens naturalistically according to chemistry and physics and whatever. But in a homunculi headed system, you have units, they operate consciously, going, okay, I got to fulfill this rule now, or I got to look something up in this table or do such and such. And it's extremely counterintuitive that these sorts of elements could add up to a conscious. It's really counterintuitive and objectionable to think that China, you know, as long as people are doing the right thing. Right. If they just behaved in certain ways, they would constitute a consciousness. We dislike that idea intuitively.

14:47 Seth: You mentioned that Block has a lot of, you know, brings a lot of arguments to bear. I would also say that Block, he's as much a fiction writer as he is a philosopher, because what he does is bring an insane number of analogies and metaphors and all that to bear. That's the currency that he plays in because his argumentative strategy is to say, okay, there is something it's like to be you. And I'm talking to all the listeners. You have a qualitative experience of what it's like to be you. And then there's that bodily system that generates that qualitative experience.

And what he wants to do is say, can you conceive of something which functionally works like your body, your brain, what have you, but doesn't have that same qualitative experience that there's nothing it's like to be so the homunculi headed robot. The argument rests on the idea, first of all, that you can characterize brain functions. Forget about reading the paper and Turing test and all this, that essentially brain functions in themselves, a neuron firing and not firing is kind of a chemical, mechanical type process. And that if you could simulate the process of brain neurons firing, you could in theory replicate all of the brain activity that you're currently having right now.

But you could do so in such a way where you wouldn't say that the thing in which you're instantiating all of this activity, this neuronal activity or firing, has a conscious experience of being the way that you do. That's the argument. And so the homunculi headed and the Chinese things are all intended to show you, like if you could get everybody in China to function according to some set of rules to duplicate the activity

16:39 Wes: in your brain, we should amend that, by the way, just to say we're not sure that you would have to duplicate neuronal activity. Because the whole point of getting beyond physicalism is to say it could be instantiated in many different the physical realizations. And it's not clear that what we have to replicate precisely is neuronal activity. It could be a much higher level of functional stuff that the brain happens to accomplish with neurons, but could be accomplished.

17:08 Seth: Fair enough. The point is that there's a translation mechanism where you have to be able to characterize neurons firing in such a way. And that's input. So he talks about input and output. Right. It's the behavioralist thing, which is there's a set of inputs, red, heat, whatever. And the behavioralist says there's an output which is run away. Right? Fear, run away. So we say, oh well, that person's afraid of fire because there's heat and red. And we know that because they ran away. The functionalist says, well, we know they're afraid of fire not just because they ran away, but because there are a series of functional states of the brain which essentially cause that response.

Maybe I'm not saying that exactly right, but the idea is that you can characterize mental activity in a functional way that could be duplicated in non gray matter, such as a bunch of homunculi, or in the entire nation of the Chinese all functioning as if they were a single entity with the appropriate transmitters and all that sort of thing and you know, functionally acting in the same way as your brain. And you wouldn't want to say that there's something it's like to be the entire Chinese nation or there's something it's like to be that thing that's filled with homunculi in the same way that there's, it's something there's like to be you.

For Block, all he wants to do is introduce the possibility, the idea that it's logically possible, or it's at least intuitively possible that such a thing, we could do such a thing and we would not say that. Yes, that has exactly the same qualia is the exact same experience of itself that you have of yourself.

18:48 Wes: Yeah, and he's going to try to give arguments for this beyond the non intuitiveness of it. Right. He thinks there are actually good reasons

18:54 Mark: to reject what do these intuitions rest on. And that's what makes his articles so difficult because it gets rather technical when he's trying to figure out like where exactly is this intuition coming from? Drilling to a deeper layer than we saw just reading about Searle's Chinese room argument, for instance, I think these nation of China and these couple of examples are important enough that we should read a little bit of them. So the beginning of the homunculi headed system example is in 278. Imagine a body externally like a human body, say yours, but internally quite different.

The neurons from sensory organs are connected to a bank of lights in a hollow cavity in the head. A set of buttons connects to the motor output neurons. Inside the cavity resides a group of little men. Each has a very simple task to implement. A square of a reasonably adequate machine table that describes you and if you remember from last time we talked about machine tables, you can go back and listen to that. On one wall is a bulletin board on which is posted a state card. That is a card that bears a symbol designated one of the states specified in the machine table.

Here's what the little men do. Suppose the posted card has a G on it. This alerts the little men who implement G squares. G men, they call themselves. Suppose the light representing input i17 goes on. One of the G men has the following as his sole task. When the card reads g and the i17 light goes on, he presses output button O191 and changes the state card to M. This G man is called upon to exercise his task only rarely. This is actually A better description of like what a Turing table is than talking about in general terms. Last time.

20:32 Wes: He does give a very simple example of a table in this as a vending machine. But finish this first.

20:37 Mark: Yeah. In spite of the low level of intelligence required of each little man, the system as a whole manages to simulate you because the functional organization they have been trained to realize is yours. A Turing machine can be represented as a finite set of quadruples or quintuples if the output is divided into two parts. Here they are current state, current input, next state, next output. Each little man has the task corresponding to a single quadruple. Through the efforts of the little men in the system, realize the same reasonably adequate machine table as you do is thus functionally equivalent to you. So a few details in there that make it a little confusing, but just the idea that each part has its current state, current input, next state, next output. And that's kind of what each part of a computer. That's what it comes down to.

21:17 Wes: How many homunculi are required? Perhaps a billion are enough, since there are only a billion neurons in the brain.

21:24 Mark: He spent quite a few pages before this in this article to say a functional realization could occur at many different levels. So it could be. You could give a functional account of the brain just having left hemisphere be one square and right hemisphere be another square. And the states that you would have to use to describe them and how they react to each other would be kind of complex. But clearly, to talk about functional analogs, we want to get at a finer grain than that. It has to at least account for behavior. But going all the way to the neural level is probably too far, as you were saying, Wes, to Seth earlier.

But that is the way he starts it here. Suppose we convert the government of China to functionalism. We convince its officials. It would enormously enhance their international prestige to realize a human mind. For an hour, we provide each of the billion people in China. I chose China because it has a billion inhabitants with a specially designed two way radio that connects them in the appropriate way to other persons and to the artificial body mentioned in the previous example. We replaced little men with a radio transmitter and receiver connected to the input and output neurons.

Instead of a bulletin board, we arrange to have letters displayed in a series of satellites placed so that they can be seen from anywhere in China. Surely such a system is not physically impossible. It could be functionally equivalent to you for a short time, say an hour. There's the example I just wanted to lay out. Does it help to have the details.

22:39 Wes: Yeah, it does. No, I think it's good. I think a lot of listeners are going to find this baffling. And of course, some listeners are going to dislike thought experiments like this. But I don't want to spend any time defending them. You know, I think they're perfectly fine. The point of talking about these things is because they're in principle possible. Thinking about what would happen if we did them is a good way of analyzing this sort of situation. Analyzing our intuitions, let's say. I'm not sure that's right.

23:07 Mark: Well, I like it makes clear. So later on this page279. Here you have to define the functional states of something. With respect to specific types of inputs and outputs. So he says, of course, there are signals the system would respond to that you would not respond to. For instance, massive radio interference or flood of the Yangtze River. Such events might cause a malfunction, scotching the simulation. Just as a bomb on a computer can make it fail to realize the machine table it was built to realize. But just as a computer without the bomb can realize the machine table, the system consisting of the people and artificial body can realize the machine table.

So long as there's no catastrophic interferences, for example, floods, et cetera. In other words, a flood is technically an input to the system. But it's not the right kind of input. The right kind of input. We're defining what a functional analog is here by saying the inputs are the letters that show on the satellites that are viewable from anywhere in China. Anything else that affects the operation of the system is kind of coming in through at a lower level. So if you think like, again, the brain is kind of running the software. And the software is the functional organ.

Things that might hit somebody in the head with a brick or they drink alcohol or something, that affects the system. But it's not actually an input to the system considered as a functional organization. Would that be the same with behavior? In other words, like the results of being hit in the head with a brick is not an output of the machine table system? Or the results of drinking a lot of alcohol. But that it sure seems like there would have to be some interaction when you're in pain. I'm giving a functional analysis of pain. You normally back away from the thing or you consult your beliefs.

Is this a good kind of pain or is this a bad kind of pain? Am I trying to train for the marathon, so I should work through this or not? Well, if you're drunk, that's obviously going to affect Those decisions, it's going to affect that matrix. But yet we still consider the drinking is not a proper input for the functional system. He doesn't actually raise in here about the outputs. His talking with slurred speech. He wouldn't have to give a functional explanation for that. Right, because it's reacting to something non functional.

25:16 Wes: No, it's part of the inputs and outputs. I think alcohol is an input and slurred speech is an output. I'm not sure. Why would you say otherwise? I mean, if you have your brain, if something damages your brain and it can no longer perform the functions of a typical brain, then you have something analogous to the flood in the case of China or something. If you want to say alcohol, obviously there's a problem involved in defining inputs and outputs. Where does the system begin and where does the world begin? And what constitutes an input as opposed to something that actually interferes with functioning? What distinguishes an input that gets processed functionally as opposed to something that actually interferes with typical functioning? Those are difficult questions.

26:00 Seth: Well, yeah, and I think they're difficult. And I think that's part of the reason why Bloch tries to just sort of say, you know, I'm talking about this, whatever, trying to rule out the flood case and for his analogy for the Chinese. But it does seem as though paradigmatically what they're talking about when they're talking about inputs is perception. Nobody in any of these essays we've read does the hard work of actually talking about, at least I don't think so, of what inputs might be. And then it's assumed perception. It's assumed the paradigmatic case is like seeing a color or reacting in a certain way to, oh, pain, Right.

Everybody talks about pain for exactly the reasons that Mark mentioned at the beginning of it. They want to use pain as the standard case because it theoretically gets you out of some of the complexities of behavioral response that involve memory history, experience, trauma, all these different things. It's just basically, oh, well, we all feel pain. It's a response to a particular stimulus. So I think that's what's at the back of everybody's mind. It's just very few of them seem to do a good job of like, articulating that and being clear about it from the beginning.

27:10 Wes: I think Block will say whoever says what the system is, says what the inputs and outputs are, says what counts as inputs and outputs. And for him it's only a neural activity. So we think of the frontier of the brain, basically, and instead of thinking about sounds and Sights, color, or something like that. I go a little further in and I'm just thinking about the electrical stimuli that are entering the brain or exiting the brain to my nerves, and those are the inputs and outputs. I don't think this is something we want to get bogged down in.

27:43 Seth: Block wants to introduce the idea. He wants to say I can provide a thought experiment which will be plausible enough to suggest that there can be a functional system which is equivalent to you, but which does not have your qualia about which it can't be said that there is something that's like to be it. That's the same way that it's something that's like to be you. And then the rest of the essay in many respects is metaphor after metaphor, analogy after analogy. Trying to tease that out on page

28:16 Wes: 281, he's going to say a similar thing throughout for various forms of functionalism. But what makes the homunculi headed system count the two systems as variants of a single system just described? A prima facie counterexample to machine functionalism is that there's a prima facie doubt whether it has any mental states at all, especially whether it has what philosophers have variously called qualitative states, raw feels or immediate phenomenological qualities. So we're doing a reductio ad absurdum here. Oh, you say mental states are just these functional states. Well, let me give you an example where that seems absurd.

28:54 Mark: I'm trying to find the place where he talks about the subatomic homunculi. So you might say what's wrong with the nation of China example? What's wrong with the homunculi headed system example? Exactly the same thing that was wrong with Searle's Chinese room example in is that you have parts of the system that are themselves little conscious beings. It seems obvious, like if those little creatures have mentality, then the whole system that they make up can't itself be a superorganism that has mentality.

29:26 Wes: And what they do depends on them having mentality. Right. The little man, when he sees a G on the board, I forget the exact example, but he himself is doing the work of a conscious being in order to fulfill his function in the larger consciousness. There's something inherently objectionable about that, and

29:42 Mark: that specification anticipates this related example he gives. So if you remember, the Putnam article just ruled this out by stipulation. He just said for two systems to be functionally the same, one of them can't have little bits that are themselves functional. Entities. Block just thinks that that is bullshit. You could give multiple functional explanations for even the brain. And if you can give a functional account of the brain where you're only looking at the two hemispheres, and then a different functional account of the brain where you're looking at smaller bits, that is just not a legitimate thing to rule out that you can only work at certain levels.

So he gives this example of if you imagine like the Whos in Dr. Seuss, their whole world lives on a speck of dust. And imagine if every molecule had something like that going on. It's way too small for us to see. And importantly, so this is contra what you were just saying, how the larger units, which these little worlds make up, the behavior of those larger units is not determined by the intelligence of the individuals involved. So it's not like the individuals are like steering the stacks of dust or determining the ultimate physical properties that a quark exhibits.

It's because there are little dudes in there looking at a rule book and saying, using their mentality to make the quark move in the way that quarks are supposed to move. No, they're just living their little lives. And even though then every single macro organism, every single mind that we look at, according to this far fetched but potentially metaphysically possible setup, it's not an issue that if the parts are conscious, then the whole can't be conscious. Because there's an example right there where it seems at least coherent that the whole you and me and Seth and Dylan are all conscious, yet the parts are also conscious.

It's just the relationship between those consciousnesses that it can't be, as you were saying, Wes, that the behavior of them acting in the functional organization uses some form of intelligence itself. Did you buy that distinction between the two types of homunculi?

31:52 Seth: Who cares? Why are we talking about two different kinds of homunculi? I mean, this is the part I don't understand.

32:00 Mark: This is a source of the intuition here that, oh, you can't have a macro organism that's made up of little homunculi that is itself conscious. Whereas if you believe in like the Gaia hypothesis, if you're Spinoza, yeah, that's kind of his metaphysics, is that you put all the creatures in the world together and you get an uber mind and let's call it God, and it even has intelligence. There's nothing incoherent. We'd have to prove that somehow. But this is ultimately what we're trying to determine here, whether it matters whether some alternate kind of structure could be conscious, what it's made up of.

32:39 Seth: Why are we arguing about what kind of homunculi there are? The bottom line here is I think this is a rat hole that we don't need to get into. Block proposes the homunculi thought experiment to basically say, is it intuitively possible that there's something that's functionally equivalent to us but doesn't have conscious states? Okay, here's a homunculi thing, here's a nation of Chinese thing. And then it's like, well, somebody might object to that thought experiment and try to say, well, I have a problem because I don't believe that something that is constituted by conscious entities could itself have a mega consciousness on top of it.

33:19 Wes: Well, what Block is trying to do, he's trying to bring up a counterargument.

33:22 Seth: He's bringing up a counterargument to his example to try to defeat that counter argument. So let's put that aside for a second and just say, instead of getting wrapped around the axle of his thought experiment, is the intuition that we could create something that was functionally equivalent and yet did not have mental states or did not have what we would describe as the consciousness states. Is that something that's.

33:47 Wes: Well, what he's worried about right now is whether we could be functionally composed of homunculi in particular.

33:55 Seth: So let's not argue about homunculi. Let's argue about whether or not you can agree with the intuition that we could conceive of a system that is functionally equivalent to you, that would not have the experience of what it's like to be you.

34:13 Mark: It would not have qualia.

34:14 Wes: I think homunculi are actually essential to this because, yes, in a broadly metaphysical sense, in the Chalmers sense of, oh yeah, we could conceive of zombies. Like we could even conceive of someone who has exactly my brain activity, who isn't conscious. That's just a metaphysical possibility that restates the mind body problem because I don't understand the causal mechanisms or whatever the identity relation is, however you want to put it, that makes my brain states mental states. So, yeah, in that broad sense, it's entirely conceivable that not just something that's functionally equivalent, but something that is materially, physically equivalent to me doesn't have mental states.

But we're trying to go beyond that conceivability to the idea that there's something wrong with functionalism. And that's what this, the Homunculus example does. So in other words, it's not just that I can imagine a functional equivalent that doesn't have mental states, it's that intuitively we think that there would be certain functionally equivalent systems that just we can't bite the bullet on them having mental states, we think they don't. And that becomes a real objection to functionalism. Whereas I don't think the conceivability functionally equivalent systems that don't have mental states is in itself an objection.

35:41 Mark: If you think that any of these examples, the whole notion of functional equivalence, doesn't actually make any sense, then that would be a major knock against functionalism, since the whole thing is a reductio ad absurdum. It's not necessarily that the piece that has to fall here is the necessary supervenience of the mental on the functional. The piece that fell could just be the notional functional equivalence itself, which seemed to be what you were just asking intuitively, could we imagine that there is some kind of thing that's homuncula headed, but yet that is functionally equivalent to us.

And I think that's hard for us to just be intuitive about because we don't really understand functionalism all that well. I mean, we just had a whole episode on it. And you can think about it in very simple ways and okay, what are the functional components of a mousetrap? And they could be realized in different things. But actually imagining something that is functionally equivalent to a whole human being, a whole, seems astounding. And it seems like it might not just be a matter of increased complexity. It's a mousetrap with more pieces. It's a factor of complexity greater that you really just couldn't have a functionally equivalent other thing.

36:55 Seth: Right. That's because human beings are historical too. I was trying to get back to this point. And functional equivalency rests on the idea that you can characterize mental states ironically, functionally. So in other words, in order to talk about functional equivalency, you need to characterize mental states as something that's capable of being characterized functionally. And in his case, he's talking about the Turing square input, state, change, state output. It's the idea that there is a way to characterize the mental state in some kind of, I hesitate to use the word mechanistic, but essentially that it's ordered.

37:35 Mark: Let's say algorithmic instead of mechanistic.

37:37 Seth: Algorithmic, sure. That's a good one. Because we know that as complex as physical mechanisms of the brain are they can be characterized algorithmically. I guess it's natural to say, well then mental states must also be able to be characterized. Or at least if we're going to explain mental states, we should piggyback on the algorithmic structure of the physical state of the brain, which gives us something we call functional states. Functionalists want to talk about the causality of mental states, and so there has to be some kind of structure in order to them that permits that.

If I deny that that first step is even possible, that you can algorithmically characterize mental states, then the whole functionalist enterprise, I think is doubtful. But again, we haven't positively characterized functionalism here other than to have Block define it in order to provide all these counterexamples. So I thought we were going to get the Chalmers early on and we weren't going to spend a lot of time on block. So that's why I'm like, we can talk about all of the 22 different examples that I've cataloged of Block bringing in analogies and different metaphors to try to get at that intuition.

And we could spend all of our time just arguing about which ones are more satisfying or less satisfying, or which ones have flaws or which ones don't. But there's some other things that we would be more interesting to spend time on.

39:00 Wes: Yeah, I think there are more things in Block though, that I don't think that's all there is is him reiterating the same point.

39:06 Seth: It's more complex than that. But ultimately that's the strategy of argument that he's employing.

39:11 Mark: This is one of the reasons I was just bringing in the whole alcohol thing as a potential either a challenge to clarify what functionalism is, or perhaps an objection to functionalism that I think, contra Wes's thought that, oh, alcohol, you know, its deadening effects count as an input. No, if you're thinking about the brain algorithmically, well, the taste, as we were saying before, it's the sensations. You define what counts as an input. So the sensations, the taste of the alcohol and the feeling that you get as a drunk person, those qualia would count as inputs, but the liquor itself slowing the neurons.

I'm more putting it as a question. I don't really know how to deal with this, but maybe we should just move on. I think we've defined pretty well what absent qualia is imagining that there's this homunculi headed thing, it's got the same functional organization of us. Imagine that you understand that. And yet it seems obviously not to have qualia. So that's the absent qualia argument. So do we want to say briefly what Chalmers response to that? He's got two counter reductio ad absurdum arguments to say if you believe in absent qualia, you must also believe in fading qualia.

Well, it's actually inverted qualia. In other words, my green is your blue, etc. Then you must believe in dancing qualia. And dancing qualia also work against absent qualia. So between those two counterexamples, those two reductios, he thinks he has just knocked down Block's absent qualia argument altogether.

40:45 Wes: We can do that. Since Block also talks about inverted qualia, we can backtrack and talk about. So we can do these in two separate steps.

40:52 Mark: All right, so assuming this is page three of his 12 page article here, assuming that absent qualia are empirically possible, that's we're just assuming that to test it. It follows there can be a system with the same functional organization as a conscious system such as me, but would lack conscious experience entirely due to some difference in non organizational properties. The organizational ones are exactly the functional ones. Without loss of generality, suppose this is because the system is made of silicon chips rather than neurons. Let's call this functional isomorph.

Same functional setup as me robot. The causal patterns in robot's processing system are the same as mine, but there is nothing it is like to be robot. Given the scenario, we can construct a series of cases intermediate between me and robot. Such as there is only a very small change at each step. The functional organization is preserved throughout. If we're talking about the neural level, just to simplify things, we're saying to really give a complete map of the functional organization, you actually have to say what each neuron is doing, which we've made very clear.

That's probably not the case. Well, you could just swap one neuron out with a silicon chip. It says its replacement is a silicon chip that performs exactly the same local function as the neuron. We can imagine it is equipped with tiny transducers that take in electrical signals and chemical ions and transforms these into a digital signal upon which the chip computes, etc.

42:08 Wes: This bit of detail is actually good, with the result converted into the appropriate electrical and chemical outputs. Because I think for many listeners this will be the first taste of the specifics of how one different kind of material might functionally replace what the brain does.

42:25 Mark: You might say like no, you can't even do that with one, because there's no way it would actually do the Same thing. But when you think of how simple a neuron actually is.

42:33 Wes: Yeah. And as he puts it, as long as the chip has the right input output function, the replacement will make no difference to the functional organization of the system. And we have to think, even if it's not a neuron that could be replicated in this way, there has to be some level. Right. Where we could model input and output, because that's what's going on. Presumably we'd have to deny that there's input output in order to deny that we could do that with a silicone chip.

42:59 Mark: Yeah. So an artificial heart, it's not like you go in each cell, you figure out how many cells were in the original heart. No, it had just something that plays the same.

43:07 Wes: Yeah. So I think it's entirely plausible. And then if you're going to say that if we replaced all the neurons with silicone chips, if you're going to say, well, that being doesn't have any qualia, isn't really conscious, then you have to account for this gradual replacement thought experiment in which I go from being a conscious being to one without consciousness, and then where do I stop being conscious? Obviously, if you replace just one neuron in me, or even a lot of them, I'm not going to display any behavioral problems or any other sorts of abnormalities.

I'm going to function just the same as a person. But the antifunctionalist is going to have to say, well, at some point as you're undergoing those replacements, you may look and act just like you normally would, but somehow your subjectivity is being drained away. And that, I think, is a very powerful thought experiment that Chalmers is giving

44:07 Mark: us the gist of. It is, what is it like to be one of the in between states? Before we even get to that, Seth, what did you think of this example in terms of just the idea? I don't know if you guys know the origin story of the Tin Woodman, but it's the same as he's describing here. I was out chopping my wood one day and my axe slipped and it hit my leg. And so I went to a guy and he replaced my leg with tin. And then I did the same thing the next day, and he replaced my other leg with tin. And then he replaced my. It's ridiculous that that's how in the wizard of Oz, how the Tin Woodman got that way.

He used to be a normal dude and he chopped his own face off eventually, and they just replaced it with tin. And somehow he retains functional isomorphism. He Acts virtually the same. Obviously, he has things like rusting is not a matter of the changing algorithm, that's like a matter of the changing material. But he has the same personality and the same behavior.

45:02 Wes: Yeah, so maybe the functional units are quite gross, quite, you know, quite macro. Yeah, just cut my head off and replace it with a. With a tin head.

45:12 Mark: That example is from, like 1899, that this kind of stuff was out there already.

45:17 Seth: What Chalmers is doing, he's starting with consciousness. He's starting with the idea of consciousness, and he's essentially doing a replacement. So in the case of the absent Qualia, what he's saying is, okay, let's take Block at his word here. I am on one end of the spectrum, I'm functionally organized in the manner of A, and I have the quality, the experience of being Seth. At the other end of the spectrum is something which is also functionally organized in the manner of A. But it is not Seth. It does not have the experience of being Seth that represents Block's homunculi, or Chinese nation unit, and it is absent qualia.

So what Chalmers is saying is in that spectrum between the two, if I start replacing pieces of me with pieces of the simulacra, functionally still the same, but just not the same material basis or physical basis, we start working our way from one end of the spectrum, and then likewise, that you could start replacing piece by piece, something in the simulacra to get it to work more towards the Seth brain stuff and what have you. How do you describe what's in the middle? Like, how do you get from A to B? He says, well, if you believe that it's a continuum, and on the one hand you have absent, and on the other hand you have completely present qualia, then either qualia has gradations, so you have the fading Qualia, meaning the further away you get from me, the less qualia you have.

And the closer you get to the simulacra, the less qualia you have. Or there's a switch that flips somewhere in the middle, there's a tipping point. And when you're 51% towards Seth, you're essentially full on Seth, or at least that's when Sethness turns on. And then it's not until you get to the end that it's fully realized. Or something like that.

47:19 Wes: Yeah, it's either gradual like that, or there's some point where just switching out a single neuron would completely turn out the lights.