Change my mind

I have long thought that there is something deeply anti-scientific about the way that certain people argue against physicalism, and in particular the way they defend panpsychism. This summer, before the premiere of Phil Halper’s video Panpsychism Debunked: Why Consciousness Isn’t Everywhere, which I’m in along with Ken Williford and Josh Weisberg (I didn’t pick the title!), I posted on Bluesky that panpsychism is antiscience and invited people to change my mind. Philip Goff answered the video with one of his own, and I answered his with mine, once I figured out OBS (my first attempt, an hour on just the first seven minutes of his video, was lost to a frame rate setting).

“Antiscience” is a strong word, so I want to be careful about what I mean. I am not saying that anti-physicalists are anti-science, or that panpsychism or dualism, the views, are. If one adopts a certain philosophy of science then one can see panpsychism as largely compatible with the findings of modern science. Nor am I saying that arguing against physicalism is antiscience. To argue against physicalism you need an argument, and Philip, who has been on Consciousness Live! more than once and with whom I have had a lot of good arguments over the years, is right that making one is not anti-scientific. What I object to is what usually comes with the arguments: the certainty that they succeed, the sense that this is just obvious, and the demand that the physicalist show, right now, how consciousness could be physical. One can reject physicalism, and argue against it, without any of that, but it is difficult!

What I mean by antiscience

It is hard to define antiscience, but as I understand it, it is typically committed to at least two claims. The first puts the burden in the wrong place: it has got to make sense to me, right now, and until you can show me how something could be true, I am entitled to say that it isn’t. The second is certainty about what science can’t do: it is obvious that the arguments succeed, so whatever science finds, it can’t settle this question.

Back in December 2024 I put both claims in terms of specialness. As I said then, “Specialness comes from requiring that we be able to understand why [physicalism] is true,” or else from claiming that we already know enough to know that it is false. Either way, our understanding gets treated as a test that nature has to pass.

The first claim is the one I run into most. Someone on Bluesky told me he wasn’t making a logical argument against physicalism; it was just that science has no account of how qualities arise from quantities, and that makes experience mysterious enough to justify thinking it is something else. But a missing account is a reason to keep looking, not a reason to think there is nothing there to find. The burden is on whoever wants to add something to what science already posits, and “I can’t see how” doesn’t discharge it.

I am not attributing the second claim to anyone uncharitably. When I posted that “physicalism is the scientific option in so far as there is no good reason to go beyond science,” Philip responded that “because consciousness is not publicly observable, the scientific data is limited to establishing correlations.” I agreed that physicalism goes beyond the data; every theory does. But this year he went further. “The success of science has no bearing on this dispute,” he wrote, and he described himself as defending “the standard anti-physicalist view that physicalism can be ruled out on a priori grounds.” That is the second claim, about as plainly as anyone could state it.

The two claims work together. Russellian panpsychism is built to agree with physics about everything physics says, so it is hard to see what physics could find that would count for or against it. That by itself would just be underdetermination, and choosing between empirically equivalent theories on other grounds is ordinary philosophy of science. But Philip doesn’t say the evidence leaves the question open; he says physicalism can be ruled out a priori. And the success of physical explanation everywhere else is evidence about which kind of theory to bet on here, so it does bear on the dispute. The case against physicalism has to come from introspection and reflection rather than from the sciences, and at bottom it is that we can’t see how experience could be physical (or, in Philip’s stronger version, that we can see that it isn’t). The first claim does the work and the second protects it from evidence.

Why call this antiscience rather than just philosophy? Because the demand that nature make sense to us, right now, has a terrible track record. Newton himself thought that one body acting on another at a distance was “so great an absurdity”that no competent thinker could fall into it, and his theory won on the evidence anyway. Feynman said that nobody understands quantum mechanics, and it is the best-confirmed theory we have. And Dennett had a name for the slide from the first claim to the second, Philosophers’ Syndrome: “mistaking a failure of the imagination for an insight into necessity” (Dennett 1991).

Revelation

The centerpiece of Philip’s case is what he calls revelation. In Consciousness and Fundamental Reality he says that when we attend to a conscious state and form a direct phenomenal concept of it, “the complete nature of the type to which it belongs is apparent to the concept user” (Goff 2017, 107). Attend to a pain and you know exactly what pain is.

What I said in the video is that there is something underhanded about the way this is defined. “Complete nature” is ambiguous. It could mean the experiential nature, what it is like for me to be in pain, and in that sense I think revelation is true. Or it could mean the metaphysical nature, what pain ultimately is. In that sense revelation is false, and Philip has to agree.

Here is why. On Philip’s view, as on mine, a pain is ultimately made up of, or grounded in, things that are not themselves pains. For the panpsychist they are micro-experiences (or, on the cosmopsychist version Philip has defended, a cosmic consciousness); for me they are neural processes. Either way, attending to my pain does not show me what grounds it. I don’t find neurons in my pain, but I don’t find micro-experiences there either. They have to be inferred, and there is very little we can say about them. Philip saw a version of this problem himself years ago, when he argued that revelation is inconsistent with our consciousness turning out to be “constituted of the experiential being of billions of micro subjects of experience” (Goff 2006, 57, quoted in Roelofs 2020). So when he says, a page after the definition, that there is “nothing in any way hidden from you about the reality of how you’re feeling,” the innocuous reading is true and the strong reading is false by his own lights. You read the definition and nod along, and then out comes the rabbi from the hat: since the complete nature of pain is revealed and nothing neural shows up in it, pain can’t be neural.

When Philip was on Consciousness Live! I put it to him that what grounds a state isn’t part of the nature that gets revealed. That is just the innocuous reading, and Philip has to agree with it, since on his view the grounds of a pain are hidden too. But then revelation is as compatible with pain’s being grounded in neural processes as it is with its being grounded in micro-experiences. To get anti-physicalism out of revelation you need a third reading, on which what is revealed is pain’s essence and that essence is not physical. That is the conclusion written into the definition, and it is the underhandedness I was complaining about.

I’m not the only one who finds this fishy, and it isn’t just physicalists. Torin Alter has spent much of his career defending the knowledge argument and the zombie argument, and when he was on Consciousness Live! recently he told me that, although he admires Philip’s work, he doesn’t see why anything in those arguments shows that experiences “don’t have backsides.” He also has no sympathy with the recent turn to an intuition of distinctness, which he called “a step backwards.”

I take the innocuous reading very seriously. A few months ago I broke my ankle skateboarding, badly, and while I waited for the ambulance the pain took all of my attention. Then they gave me fentanyl and it was just gone. No one could ever convince me that the way I felt before the fentanyl was the same as the way I felt after. I knew everything there was to know about how it felt from the inside. None of that told me what it was underneath.

Philip’s argument for revelation is an inference to the best explanation. Some introspective judgments are “super-justified,” held with something close to the confidence we have in basic arithmetic (Goff 2017, 112). I accept super-justification. But the innocuous reading explains super-justification just as well as the strong one. If I know what my two experiences are like, I can tell that they are similar without knowing anything about what underlies them. So the best explanation of super-justification doesn’t favor strong revelation, and only strong revelation makes trouble for physicalism.

This is where revelation connects to antiscience. Super-justification gives us something close to certainty about how our experiences feel. The strong reading quietly transfers that certainty to what they are, and that is where the confidence that physicalism has already been ruled out comes from. If attending to a pain really revealed its complete nature, nothing neuroscience could find would ever matter.

Selective skepticism

The correlations argument has the same shape. Because consciousness is not publicly observable, Philip says, science can only establish correlations between brain states and experiences, and so every theory of consciousness goes beyond the data. True! But so does every identity claim in science. No experiment shows that water is H2O, as opposed to being perfectly correlated with H2O. We accept the identity because it explains everything the correlations explain, and does so more simply. Ned Block and Robert Stalnaker made this case in detail a long time ago (Block and Stalnaker 1999).

When someone on Bluesky told me that the conjunction of all statements made by science is compatible with there being no subjective experience at all, I replied that by the same reasoning “The conjunction of all statements made by science is compatible with the fact that there is no water at all.” Nobody would accept that, and anyone who had insisted on it before modern chemistry would have been wrong. If correlations plus simplicity are good enough for water, why aren’t they good enough for pain? Chemists accepted that water is H2O long before they could explain why ice floats.

The answer, when it comes, is always some version of the same thing: we know what pain is in a way that we don’t know what water is. The careful version is that “water” picks out whatever plays the water role, so the facts about water follow a priori from the physics (Chalmers and Jackson 2001), while “pain” isn’t a role concept at all. Plenty of physicalists grant that last part. What does the anti-physicalist work is the further claim that our concept of pain hands us pain’s essence, and that is revelation again. (In two-dimensional terms, revelation is the claim that a phenomenal concept’s primary and secondary intensions coincide, which my Type-C view denies. Chalmers says his argument can do without it, but then it shows at most that consciousness has non-physical modes of presentation, and only if zombies are ideally conceivable, which is what is in dispute.) So the skepticism is selective, and revelation does the selecting. The same goes for the charge that a physical explanation of consciousness is a category mistake. Philip’s version is that “physical science is all about what stuff does,” and feeling isn’t a matter of what stuff does. But nobody found out that feeling isn’t something stuff does. That is the conclusion, dressed up as a premise.

Vitalists were just as sure that chemistry could never explain life. The usual reply is that life was always a matter of what stuff does, and consciousness isn’t. But how do we know that? By attending to our experiences, which is to say, by revelation.

Appearance is not nature

A lot of the popular arguments run the same way. Experiences don’t seem to have locations, or parts, or mass. Brain processes have all three. So, the argument goes, they can’t be the same thing. Someone on Bluesky put it to me in exactly these terms: for two things to be identical they have to share all of their properties, and consciousness and brain processes share none.

But not seeming to have a property is not the same as lacking it. As I said at the time, “You sound like someone insisting that superman can’t be Clark Kent because one wears glasses and the other doesn’t.” Water doesn’t appear to have parts either, but, as I put it later in the same exchange, “I think drops of water are made of parts even though it doesn’t appear that way” (here). If pain is a brain process, then it is located wherever that process is, even if it seems to be located somewhere else, or nowhere at all. That is just what it would be like if physicalism were true.

Some versions are cruder than others. Back in the early 2000s, when I was a graduate student, someone at a reception after a talk heard me say that I liked the identity theory. “Let me ask you this,” he said. “I have a pain in my toe, but I don’t have a brain state in my toe, so my pain can’t be a brain state. How does the identity theory answer that?” I was so dumbfounded that I didn’t answer right away, and he looked smug, as though he had actually made a coherent point against my view! The answer is that the toe is where the pain is felt to be, not where the experience itself is happening. People who have lost a leg can feel pain in a foot that is no longer there, and wherever that pain is, it isn’t in the foot. (It’s even funnier than he knew: there are sensory neurons running right into your toe, so there are brain states in your toe! There is no central nervous system state in your toe, but that is what the peripheral nervous system is for. I dare not think what he would have said had he learned about the gut brain, haha.)

Physicalists have been pointing this out from the beginning. U.T. Place called it the phenomenological fallacy: supposing that when we describe our experience we are describing “the literal properties of objects and events on a peculiar sort of internal cinema or television screen” (Place 1956). David Papineau has a nice diagnosis of why the appearance is so compelling. When we think about a pain with a phenomenal concept we use something like the pain itself, and when we think about it in neural terms we don’t, so the neural description seems to leave the pain out. He calls the resulting inference the antipathetic fallacy (Papineau 2002, 170–71). The appearance is real, and it even has an explanation, but it is not a window onto the metaphysics.

What would change my mind

Physicalism is a metaphysical position, and I have never claimed otherwise. As I said to someone who thought I had, “I already agreed that physicalism is a metaphysical position. The point is that it is the one best supported by the evidence.” My commitment is also more modest than people tend to think. As I put it to Torin, I’m not claiming to know that physicalism is true; what I’m committed to is that it hasn’t been ruled out a priori. It could turn out to be false. If the science of consciousness ever needed to posit something that the rest of science doesn’t, I would follow it there.

Armchair arguments also cut both ways. Years ago I argued that we can just as easily conceive of shombies, creatures who are physical duplicates of us and whose consciousness is wholly physical, and of non-physical zombies, creatures who duplicate us in every non-physical respect but lack consciousness. If conceiving of zombies refutes physicalism, conceiving of these refutes dualism (Brown 2008; Brown 2010). The lesson I drew is that the a priori arguments cancel out, and the empirical evidence should decide.

I told a version of this story years ago (Brown 2010, 49), and I put it to Torin again. Imagine someone in 1768 defending the Aristotelian view that water is a simple element. Told that it might be a compound of hydrogen and oxygen, she says she can imagine a world microphysically identical to ours, H2O and all, but with no water. It turns out she was imagining something contradictory; it only seemed conceivable to her. What we know that she didn’t is not some extra stuff. It is a theory, chemistry backed up by physics, that makes it intelligible how a clear liquid can have parts. In the end both sides are betting on what future theories will look like. I’m betting it will be more of the same, as it was with water. The anti-physicalist is betting on something that has never happened before.

Torin’s reply was that his side has an advantage, since it can accept the premises of the knowledge argument, which he thinks are probably true. Fair enough; that is exactly where we disagree. But notice how he holds the view. He told me that “there’s room for reasonable disagreement,” that he doesn’t claim the argument is certainly true or that we have anything like a decisive reason to accept it, and that he is “always open to” the possibility that he has made a mistake. That is what arguing against physicalism without the antiscience looks like.

What I won’t accept is the certainty that we already know, from the armchair, that physicalism is false, or the demand that physicalists show, right now, how consciousness could be physical before anyone is entitled to believe it. That is what I mean by antiscience. It don’t mean to impugn anyone’s character. Nor do I mean to say you can’t argue for you believe. The issue is over what the argument asks us to believe about evidence. So by all means, change my mind. Just do it with empirical evidence.

References

Alter, T., and R. Brown. 2026. “Consciousness Live! Discussion with Torin Alter.” YouTube video.

Block, N., and R. Stalnaker. 1999. “Conceptual Analysis, Dualism, and the Explanatory Gap.” Philosophical Review 108 (1): 1–46.

Brown, R. 2008. “Non-Physical Zombies.” Blog post.

Brown, R. 2010. “Deprioritizing the A Priori Arguments against Physicalism.” Journal of Consciousness Studies 17 (3–4): 47–69.

Brown, R. 2026. “The Underhandedness of Revelation Based Arguments Against Physicalism.” YouTube video.

Chalmers, D. J., and F. Jackson. 2001. “Conceptual Analysis and Reductive Explanation.” Philosophical Review 110: 315–61.

Dennett, D. C. 1991. Consciousness Explained. Boston: Little, Brown.

Goff, P. 2006. “Experiences Don’t Sum.” Journal of Consciousness Studies 13 (10–11): 53–61.

Goff, P. 2017. Consciousness and Fundamental Reality. New York: Oxford University Press.

Goff, P. 2026. “Is Panpsychism Debunked?” YouTube video.

Halper, P. 2026. “Panpsychism Debunked: Why Consciousness Isn’t Everywhere.” YouTube video.

Newton, I. 1693. Letter to Richard Bentley, 25 February 1692/3. The Newton Project.

Papineau, D. 2002. Thinking about Consciousness. Oxford: Oxford University Press.

Place, U. T. 1956. “Is Consciousness a Brain Process?” British Journal of Psychology 47 (1): 44–50.

Roelofs, L. 2020. “Consciousness, Revelation, and Confusion: Are Constitutive Panpsychists Hoist by Their Own Petard?” Dialectica 74 (1): 61–93.