>My point is that functionalist accounts of the property of e.g. 'being in pain' are transparently not accounts of anything we can recognize as potentially being a phenomenal property.

To me this is assuming more than we have a right to. It's presupposing an ontological distinction between phenomenal appearances and functional processing. Sure, we have an intuition that says they are distinct as well as explanatory gaps that cause us to question such an identity. But that in itself isn't enough.

>but in the absence of any positive argument to this effect, why assume so?

Depends on what you consider a positive argument. There are certainly many reasons to prefer a functionalist approach, e.g. its the most parsimonious way to cash out the correlations between brain states and phenomenal states, it doesn't suffer from the combination problem, inverted spectrum/valence problems, etc.

>None of Searle's arguments against functionalism rely on the premise that physicalism is false.

It depends on what you mean by physicalism. Often there's an ambiguity between physicalism and materialism, so let me clarify. Physicalism is the idea that brain processes are logically identical to phenomenal properties. To put it another way, there's no possible world where you have physical properties identical to this world that doesn't have consciousness. But physics tells us that physical properties just are sets of physical interactions. Thus physicalism leads directly to substance independence, and so a perfect simulation of a brain would be conscious.

>There isn't any distinction between feeling pain and seeming to feel pain.

I agree, but the point of the "it seems that P" is to cast the problem in a theory-neutral manner. You take functionalism to be a non-starter as an explanation of consciousness, and so you're quick to erase the "it seems that..." from the equation. This wording is necessary to avoid our biases from infecting our language thus biasing the investigation. The point of referencing Chalmers paper was for his exploration of the problem space that was theory-neutral. I disagree with Chalmers' conclusions but he always offers good exposition.

> It's presupposing an ontological distinction between phenomenal appearances and functional processing. Sure, we have an intuition that says they are distinct as well as explanatory gaps that cause us to question such an identity. But that in itself isn't enough.

Right, we agree on where we disagree. Probably not much point in hashing out this long-standing philosophical debate in HN comments.

>Physicalism is the idea that brain processes are logically identical to phenomenal properties.

That's a rather boutique definition of physicalism. Standard definitions (e.g. "the thesis that everything is physical", according to the Stanford Encylopedia) make no reference at all to the brain.

>But physics tells us that physical properties just are sets of physical interactions. Thus physicalism leads directly to substance independence

Hmm, that seems like a total non sequitur to me. I'm sure people have constructed arguments to that effect, but it's very far from obvious that physicalism entails substance independence. Searle certainly doesn't think so.

I don't really understand your last paragraph. As far as I can tell, you don't think we should start with the assumption that people feel pain, but only with the assumption that it seems to people that they feel pain. But that just dodges the main issue. If people don't really feel pain, then most of the philosophical problems we're talking about dissolve immediately. On the other hand, if they do in fact feel pain, then these problems remain, regardless of whether or not we have an account of why it seems to them that they feel pain. As Chalmers says, to really get anywhere with this line of argument, you end up having to deny that people really do feel pain -- which is absurd.

>That's a rather boutique definition of physicalism... make no reference at all to the brain.

Perhaps my definition was too on the nose. Going by the SEP: "Physicalism is the thesis that everything is physical, or as contemporary philosophers sometimes put it, that everything supervenes on the physical". Supervenes on the physical, at least in the context of consciousness, implies substance independence of consciousness. Chalmers depends on this understanding in his zombie argument.

Just to belabor the point, we can posit some property of physical matter that entails conscious experience only in certain kinds of physical processes (say in biological brains but not in microchips). This difference-making property's actions either are or are not mediated through physical interactions. If it is mediated through physical interactions, then we can include those interactions in the simulation, thus the property obtains in microchips, contradicting the premise. If it is not mediated through physical interactions, then by definition it's mode of influence is non-physical. But this contradicts the premise that "everything [including the mind] supervenes on the physical" (i.e. we have a change in consciousness but no change in physical properties).

Chalmers cashes out such a difference-making property of physical matter as his microphenomenal properties in panpsychism. But importantly, this isn't Physicalism (his zombie argument is specifically against Physicalism). He argues for an expanded notion of the physical world that includes microphenomenal properties at the base. Just to tie all this back to Searle, from this discussion it seems clear that Searle cannot be a physicalist under this understanding.

>I don't really understand your last paragraph. As far as I can tell, you don't think we should start with the assumption that people feel pain, but only with the assumption that it seems to people that they feel pain.

I agree that people feel pain. The problem is that there is some ambiguity in common usage that can mislead and confuse the issue. For example, if I stub my toe and I shout ouch, you would say that I'm in pain. But in the context of philosophy of mind, we can't assume that from only outward appearances. So my characterization is an effort to shed all possibility of misunderstanding by zeroing in on the phenomenal character of the thing, as well as the only indisputable statement of fact that describes our relationship to the phenomenal character.

So in this case "I am in pain" is operationalized as "It seems that [pain quale]". The "it seems that..." is important because we cannot be mistaken about seemings, i.e. it is not possible that an evil demon or simulation can trick us into thinking something seems a certain way while not being the case that it seems that way. But once phenomenal consciousness is operationalized as such, it becomes clear that functional explanations cannot be ruled out simply by definition.

So far as I can see, your argument that physicalism entails substance independence works equally well for the substance independence of wetness:

“We can posit some property of physical matter that entails wetness only in certain kinds of physical processes (say in rain storms but not in microchips). This difference-making property's actions either are or are not mediated through physical interactions. If it is mediated through physical interactions, then we can include those interactions in the simulation, thus the property obtains in microchips, contradicting the premise. If it is not mediated through physical interactions, then by definition its mode of influence is non-physical. But this contradicts the premise that "everything [including rainstorms] supervenes on the physical" (i.e. we have a change in wetness but no change in physical properties).”

>So in this case "I am in pain" is operationalized as "It seems that [pain quale]"

That is not operationalization in the usual sense, since it isn’t possible to observe which quales are or aren’t seeming to someone.

>The "it seems that..." is important because we cannot be mistaken about seemings

That depends on how you cache out the formula “it seems that [quale]”. If it’s an attribution of a propositional attitude, then it certainly could be mistaken. If it’s not an attribution of a propositional attitude, then I don’t know what it means. Or at least, I don’t know how “It seems to John that [quale]” differs from “John is experiencing [quale]”, or how formulating things this way helps anything.

I don't think there is in fact any ambiguity in common usage. "John is in pain" unambiguously means that John is undergoing a particular sensation, not that he is exhibiting a particular kind of behavior. The latter interpretation of the statement is entertained only in the work of certain behaviorist/verificationist philosophers.

>So far as I can see, your argument that physicalism entails substance independence works equally well for the substance independence of wetness:

I have no problem with this if the terms are properly understood. Wetness as a relational property can obtain in a sufficiently precise simulation. Wetness as a metaphysical property can't obtain in a simulation because the term includes suppositions of ontological grounding that interactions of electrical signals don't satisfy (why my head can't get wet from simulated water). If we note this distinction then there is no issue or reductio.

Let me ask you this: if you take it that there is some difference-making physical property of matter for consciousness that doesn't manifest as a kind of physical interaction studied by physics, then what is the nature of that property? Physics tells us that physical properties are defined by their interactions. So if we rule out the standard kind of influence that physical properties have, then whatever is left is necessarily unobservable. This is plainly non-physical influence if you take physics seriously.

>That is not operationalization in the usual sense, since it isn’t possible to observe which quales are or aren’t seeming to someone.

It's not measurable in practice, but it might be in principle if physicalism or materialism is true (i.e. influence should be measurable).

>That depends on how you cache out the formula “it seems that [quale]”. If it’s an attribution of a propositional attitude, then it certainly could be mistaken.

Right, attributing such a statement externally could be mistaken. But we do know that such a self-report, if genuine, necessarily is correct. And so in the context of analyzing consciousness, we can assume genuine reports of phenomenal properties and so attribution isn't an issue.

> "John is in pain" unambiguously means that John is undergoing a particular sensation, not that he is exhibiting a particular kind of behavior.

The issue is less about ambiguity (I know I used that word), and more about bias. If we want to avoid bias in our specification of the problem, we need to use theory-neutral language. "John is experiencing pain" isn't theory-neutral, as it assumes John and pain are distinct things rather than pain being a state or property of John.

Some random thoughts on the subject (feel free to skip):

With my "it seems that [quale]" formulation and the idea that genuine self-reports of that nature cannot be mistaken, some interesting questions arise. In Chalmers' zombie argument, a zombie would give utterances of this sort exactly like we do. But on what basis can we say that their reports are not genuine without begging the question against physicalism? On the other hand, if we say their reports are not genuine, neither are ours! Actual phenomenal properties by assumption play no causal role in our behavioral reports of phenomenal properties and so the utterances are not genuine. When we think "I seems that [quale]" we are correct, but every time we create a physical artifact to that effect, we are wrong. What a strange state of affairs non-physicalism is.

Under what conditions are we justified in believing phenomenal reports are genuine? If I were to evolve or train an artificial neural network that, without any explicit training, started to make such phenomenal reports, are we justified in believing it genuine?