The Official Scientific Worldview Is Incompatible With Consciousness

Philip Goff is a philosopher of mind at Durham University, a self-described panpsychist, and the author of *Galileo's Error* and the denser academic book *Consciousness and Fundamental Reality*. He also hosts the podcast *Mind Chat* with the committed illusionist Keith Frankish — the gimmick, he later jokes, is that one of them thinks consciousness is everywhere and the other thinks it is nowhere. Lex Fridman opens with a scene from his second conversation with Elon Musk, when he asked whether consciousness permeates all matter. Musk looked at him, Goff is told, as if the question were nonsense, and answered no: we could not tell if it did, so the claim sits outside the scientific method.

Goff disagrees, but not by pretending consciousness is a public measurement. You cannot look inside someone's head and see their feelings. We know about consciousness, he says, from immediate awareness of our own experience, not from experiments. Science is used to unobservables — fundamental particles, quantum wave functions, other universes — but in every other case we postulate the unobservable in order to explain what we *can* observe. Consciousness is unique: the thing to be explained is itself not publicly observable. If science is to take it on, we need a more expansive conception of the method.

The word *publicly* matters. If aliens visited Earth with utterly different sense organs, they might not understand our art or music, but if they could do mathematics they could understand our physics, inspect the data, and rerun the experiments. Pain is not like that. Goff's toothache is directly evident to him; Fridman can see pain behavior or ask about it, but he cannot access the pain itself. Fridman presses the devil's-advocate question: could we sneak closer and closer to observing someone else's experience the way they do? Goff's reply is that we are hardwired to read smiles as happiness, and we get good at taking another person's perspective, but strictly speaking all we perceive is behavior.

If you were only trying to explain publicly observable behavior, Goff says, you would never need to attribute consciousness. You could just postulate a mechanism. Daniel Dennett is, on this point, the most consistent person in the debate. Dennett's data for a science of consciousness is what he calls heterophenomenology: what can be accessed from the third-person perspective, including what people *say*. Crucially, testimony is not treated as evidence of an unobservable realm of feelings. It is just more public data to be accounted for by underlying mechanisms. Dennett's conclusion, as Goff renders it, is that his own consciousness cannot be empirically verified and therefore does not exist.

Goff thinks he is consistent too, in the opposite direction. The confused position is the one he attributes to a lot of people: they think the business of science is to account for public observation, and they also believe in consciousness, without noticing that the second claim implies another datum over and above experiments — the reality of feelings and experiences. Fridman sketches Dennett's multiple-drafts model, in which the brain weaves competing narratives and the editing *is* what we take to be consciousness. Goff grants that narrative theories are a plausible contender for a theory of the *self* over time. He does not think they are good as theories of consciousness, if consciousness just means subjective experience: pleasure, pain, seeing color, hearing sound.

A hamster, he says, has consciousness in that sense. There is something it is like to be a hamster; it feels pain if you stand on it. You do not need the capacity to tell sophisticated stories about yourself merely to have experience. What is special about humans, under panpsychism, is not a rupture with the rest of the universe. Human consciousness is a highly evolved form of what exists throughout nature. The distinctive capacities Goff is willing to claim are reflection on experience, planning for the future, and — more controversially — responsiveness to reasons: weighing money against family when deciding whether to take a job in the United States.

That last claim pulls him away from David Hume. As a graduate student he held a more Humean view, on which reason is the slave of the passions and what we have reason to do arises from our motivations. Professors at the University of Reading talked him out of it with a thought experiment: someone whose basic life goal is counting blades of grass, and who does not even enjoy it. Goff wants to say that goal is objectively pointless in a way that pursuing pleasure, helping others, or scientific inquiry is not. The moment you say that, you are no longer a follower of Hume. Fridman defends the grass-counter as someone who might find mastery in the task; Goff keeps stipulating a person who does *not* enjoy it, because otherwise the example collapses.

Fridman asks what death is doing in all this — the fear that the flame of consciousness goes out. Goff is blunt: panpsychism does not particularly help with life after death. There is nothing supernatural, nothing beyond the physical. All there ultimately is are particles and fields; the claim is only that their ultimate nature is consciousness. When the matter in his brain ceases to be ordered in the way that sustains the particular consciousness of waking life, then in some sense he will cease to be.

The last chapter of *Galileo's Error* is more experimental. The first four chapters are the cold-blooded case that panpsychism is the best solution to the hard problem; the last talks about meaning, free will, and mystical experience. Goff stresses that the view is not necessarily spiritual. David Chalmers and Luke Roelofs, among others, are secular atheists who just think mundane consciousness needs explaining and that the conventional scientific approach cannot cut it. But if you already take mystical experience seriously, panpsychism is more consonant with it. A materialist has to treat the sense of a higher consciousness at the root of things as a delusion. A panpsychist already thinks the fundamental nature of reality is constituted of consciousness, so the leap is smaller.

Experienced meditators across cultures have claimed that there is an element of consciousness that is universal: your mind and mine are not totally distinct; each is built on the same foundation. Goff has never had such an experience, but he does not see why it should be impossible. If it is true, then when the particular aspects of his experience fall away, the universal element at the core of his identity would continue — an impersonal afterlife, absorption rather than survival. Buddhists and Hindu mystics try to meditate their way there; maybe, he says, if there is no karma, everyone gets enlightened at death and simply sinks back into universal consciousness. The same picture, he adds, could underwrite altruism, because egotism assumes we are utterly distinct individuals, and on this view we overlap.

Fridman offers a smaller version of togetherness: say *pink elephant* and listeners visualize one, rotating it together. Is that one elephant or many? Goff says collective consciousness is certainly real in some sense; the question is how ontologically serious to be. Are these phenomena just constructions of our minds plus ordinary interaction, or, as Rupert Sheldrake would have it, fields beyond the scientifically understood ones? The Australian philosopher Miri Albahari defends a mystical picture rooted in Advaita Vedanta, but from inside analytic philosophy, and she argues that experienced meditators should count as expert testimony — the way most of us trust climate scientists or cosmologists. Goff's plea is not that everyone accept the view. It is that terms like *new age* can function like slurs, picking out a set of ideas and implying fluffy thinking, when the actual arguments, in Albahari or Roelofs, are rigorous even when they are hugely controversial.

Fridman is allergic to the word *expert* because, in his ear, it rhymes with ego. Put someone on that throne and humility starts to dissipate. During COVID, he says, genuinely knowledgeable colleagues dismissed what millions of people were arguing online without looking at the arguments with empathy and rigor. If you do not have time to examine the case, do not dismiss it so strongly. He cites the heat he took over Francis Collins and the lab-leak hypothesis: Collins, on Fridman's telling, showed that he had not really explored the circumstantial evidence and the tense public fight around it.

Goff wants to defend experts a little. You cannot research everything; if you want to know about biology, you need some sense of who has spent years reading the material and having their views torn apart. He is especially sharp about conflicts of interest. A UK organization called the Institute of Economic Affairs appears on the BBC as experts on economic questions and does not declare who funds it. You should not get to call yourself a think tank, he says, without being transparent about who is paying. His analogy is Churchill on democracy: the worst system apart from all the rest. He even applies the joke to his own view — panpsychism is the worst theory of consciousness, apart from all the others.

Fridman's favorite illustration is Brazilian jiu-jitsu. In the gi, a black belt changes how people treat you; they defer. In no-gi, nobody knows your rank, and the only thing that matters is the live exchange. Both have value: knowing someone is a black belt helps you decide whose advice to weigh, but the practice of science still has to throw rank away and test the ideas. Goff is frustrated that philosophers do not reach out more. Academic training pushes toward watertight arguments and jargon; a popular book requires loosening the argument, which then makes you fear colleagues will think it is crap. His solution was to write the academic book first, then *Galileo's Error*, so he can point back to the full case.

The same itch shows up in a *Scientific American* piece he had just written. There is a popular move from the cosmological fine-tuning of the laws of physics to the multiverse, associated with physicists such as Max Tegmark. Philosophy journals, Goff says, have treated a line of that reasoning as fallacious for twenty or thirty years, but the argument is buried in technical articles. Physicists argue for a multiverse, theists argue for God, and neither side knows the objection exists. That, for him, is what goes wrong when expertise stays inside the guild: the important point never reaches the people having the public fight.

The textbook menu Goff was taught had two options. Materialism says consciousness can be wholly explained in the terms of physical science, as electrochemical signaling in the brain. Dualism says consciousness is non-physical, outside the workings of body and brain though closely connected. He became disillusioned with both. Panpsychism is the middle way: there is just the physical world, particles and fields, but there is more to the physical than physical science reveals, and that more is consciousness. It is not outside the physical, as the dualist thinks. It underlies the description physics gives.

The core problem with materialism, as he puts it, is that physical science works with a purely quantitative description, while consciousness essentially involves qualities — the smell of coffee, the taste of mint, the deep red of a sunset. As long as the description of the brain is framed in the quantitative language of neuroscience, those qualities are left out, which is to say consciousness itself is left out. Dualism he has partly walked back. In the book he argued that an immaterial mind impacting the brain every second of waking life would show up in neuroscience like a poltergeist, and that not finding such traces is a growing inductive argument against dualism. Talking to neuroscientists at Durham's interdisciplinary consciousness group changed his confidence. We know basic chemistry, how neurons fire, and a fair bit about large-scale function. We are almost clueless about how those functions are realized at the cellular level.

The numbers he cites are meant to puncture premature certainty. Every pixel on a brain scan corresponds to 5.5 million neurons. Researchers are only about 70 percent of the way through a connectome for the maggot brain, which has on the order of 10,000 to 100,000 neurons; the human brain has 86 billion. Martin Picard at Columbia is experimentally exploring whether mitochondria in the brain should be treated as social networks, rather than reducing everything to underlying chemistry and physics. Dualism remains, for Goff, an inelegant theory that splits reality into two radically different kinds of property. Why believe in two kinds of thing when you can get away with one? Materialism is also simple, but he does not think it can account for the data.

The move that defines his panpsychism is inversion. The standard approach starts with matter and asks how to get consciousness out of it; he does not think that problem can be solved. Start with networks of very simple conscious entities instead. Because their experience is simple, they behave predictably; their interactions realize mathematical structures; those structures *are* the ones physics identifies. We call the entities particles, the structures fields, the properties mass, spin, and charge. Doing physics, he says, is like playing chess when you do not care what the pieces are made of. Sabine Hossenfelder's critique — that extra consciousness properties would make the Standard Model give false predictions — misses the view: there are not two kinds of property. Mass, spin, and charge *are* forms of consciousness. Bertrand Russell's phrase, which inspires much of this tradition, is that physics captures the causal skeleton of the world, not the flesh and blood.

Fridman wants to know whether, in a panpsychist framework, consciousness can be measured — whether some things are more conscious than others. Goff splits the job. There is an experimental project of finding neural correlates: which physical brain activity goes with which experience. There is also a theoretical why-question that he does not think an experiment can settle, because consciousness is not publicly observable. One theory he is attracted to is integrated information theory. Consciousness, on that account, shows up at the level with the most integrated information. A cup of tea is probably not conscious as a cup, because there is likely more integrated information in the molecules than in the liquid as a whole. The brain is distinctive because the system has more integrated information than individual neurons.

Where he gets pedantic — his word — is degrees. IIT talks about gradations of consciousness. Goff can make sense of grading complexity or information, but not of experience itself coming in amounts. Something either has experience or it does not. It may have very simple experience or very complex experience; that is not the same as more experience versus less. The unusual notion of information in IIT is about constraining past and future possibilities. The retina can be in a huge number of next states depending on incoming light; the brain's next states are much more constrained. And unlike a computer, where pulling a few transistors may not destroy much information, taking a tiny bit of brain out loses a lot, because storage depends on interconnection.

Suffering is the test case Fridman cares about, because that is how the public connects consciousness to animal cruelty and policy. A hamster, he feels, has less capacity to suffer than he does; maybe a cockroach or a bacterium still less. Goff's reply is layered. Humans have kinds of suffering unavailable to a hamster: heartbreak, fear of one's own death, anxiety about whether existence has a purpose. That is a bigger menu, not obviously a deeper instance of the shared items. On physical pain, he is not sure humans feel it more deeply than a hamster. And some experiences run the other way. Bats echolocate; Thomas Nagel's famous point is that no amount of bat neurophysiology tells you what it is like to find your way by listening to echoes.

Suffering, for Goff, is a kind of experience, so you have to be conscious to suffer. That is one reason illusionism looks, to its defenders, like a way to dissolve the whole problem: if consciousness in the philosophers' sense does not exist, there is nothing to explain. Frankish, Goff's *Mind Chat* co-host, is ambivalent even about the word *pain*. Another illusionist, François Kammerer, has written about how to think about morality if pain as we ordinarily conceive it is a trick the brain plays. The concept David Chalmers made famous is the philosophical zombie: a creature whose body and brain work just like ours, who screams and runs when you stick a knife in it, but for whom there is nothing it is like to be that creature. Nobody believes such beings are walking around, Goff notes, except perhaps the one philosopher who thinks everyone else is a zombie.

Would a silicon Commander Data, able to converse about hopes and fears but with no inner life, have moral rights? Is it murder to turn that being off? Goff is inclined to say no. Consciousness is the basis of moral value and moral concern. Conversely, as a panpsychist, he thinks the view can transform your relationship with nature: a tree as a conscious organism, however unusual, is a locus of concern in its own right, and forest fires become the burning of conscious organisms. The ethics get messy. People assume a panpsychist will be vegan, but if plants and trees are conscious too, the clean dividing line disappears. The practical rule he is working toward is not eating factory-farmed animal products — not because kinder farms involve no suffering, but because more people might actually do it, which could mean more harm minimized.

Fridman then puts the problem in the workshop. He can make a robot move, recognize the world, interact with people. He does not think the metal and electricity are conscious, but he forms a connection that makes him feel they are, and he becomes less and less okay with 'murdering' the machine. Will there be a time in twenty, thirty, fifty years when we are not morally okay turning off the power to a robot? Goff's first move is epistemological. How would we *know* it was a zombie? Ned Block's paper 'The Even Harder Problem of Consciousness' asks whether a silicon duplicate's pain, if any, is grounded in hardware or software. In humans the two always come together, so you cannot run the control. Even the search for neural correlates is a mess: some people put consciousness in the front of the brain, some in the back. Goff suspects Fridman's current robots are not conscious on any reasonably viable model, but sophisticated future systems would not be so easy to dismiss.

Fridman pushes into history. The difference between zombie and non-zombie, if you look at civilization, is often that the zombie accepts the role and marches to slaughter; you stop being a zombie when you say no. Groups declare themselves real and others disposable until the others refuse the name. Goff's reply is to split the questions. Ontologically, there is a fact of the matter as to whether something is conscious, and on his view it has rights if it is and does not if it is not. Epistemologically — how the hell do we know? — is a minefield. They will have to cross that bridge when they get to it. Sean Carroll's objection that Goff's view makes consciousness causally idle gets a software-hardware answer: physics is the software, consciousness the hardware. That Microsoft Word can run on two machines does not mean the desktop is doing nothing.

The day before this recording, Goff had been on Joe Rogan's podcast, and Fridman asks for the postmortem. Rogan put up a fight. The sticking point, as Goff reconstructs it, was whether consciousness can already be explained in conventional scientific terms. Goff's pressure was a distinction: correlation versus explanation. Everyone agrees that certain kinds of brain activity go along with certain kinds of experience. That does not answer why those activities go with those experiences.

Rogan's argument, roughly, was that we know consciousness is explained by brain activity because funny chemicals change the brain and thereby change consciousness. Goff grants the correlation. Changing the chemicals changes the experience, and that is part of the story. It still leaves open why those particular chemicals go along with that particular kind of experience rather than a different experience, or none at all. Materialists try to explain the experience in terms of the brain activity; panpsychists reverse the direction; dualists think they are separate but perhaps tied by special laws of nature.

His analogy is quantum mechanics. The equations work; they predict outcomes. There is a further question about what is going on in reality to make the equation predict correctly, and some physicists want to say shut up and calculate. Consciousness research has the same split: one project is finding physical correlates, the other is asking what undergirds those correlations. The options are as hard to evaluate as interpretations of quantum mechanics. Then, with the grandiosity he has been joking about all conversation, Goff adds that actually it is easy — panpsychism is obviously the best one — and immediately flags the joke.

Fridman wants the story Goff tells in *Galileo's Error* about a voyage funded by the Russian investor and philosopher Dmitry Volkov, who co-founded the Moscow Center of Consciousness Studies at Moscow State University. On board were Dennett, Paul Churchland, David Chalmers, Martine Nida-Rümelin — a substance dualist at the University of Fribourg — and Goff as part of the official opposition. The original plan, Goff says, leaned toward a Dennett-type camp on free will and consciousness; Chalmers insisted they needed people who disagreed.

The technical point Goff is proud of is that dualism is consistent with conservation of energy. Churchland had argued that an immaterial soul doing things in the brain would add energy to the system and violate conservation. Goff, drawing on materialist philosophers such as David Papineau, notes that naturalistic dualists like Chalmers want consciousness in a law-governed world. They already believe in psychophysical laws over and above physics. Those laws can be written to respect conservation, the way multiple physical forces already work together to do so. Why not add more laws? He raised the point in Churchland's talk and, as a Moscow graduate student later told him after asking for a translation, everyone turned on him like a pack of wolves. Patricia Churchland asked whether he believed in magic. He was not even a dualist; he was making a pedantic point.

That evening everyone went onto the island except Goff and Dennett. Dennett was on deck carving a walking stick. Goff honed in on the conservation point, Dennett called dualism nonsense, and in the end — maybe Dennett would deny this — he said maybe that is right. Goff calls it a win, a tiny one, not a conversion about the fundamental worldview. He cannot tell Fridman much more about philosophy in Russia, except that Volkov has funded the center, brought graduate students along, and written a book on Dennett titled, Goff thinks, *Boston Zombie*. The trip did not, at least, end with anyone thrown overboard in the Arctic.

Goff wants much more agnosticism about free will than about consciousness. We do not have the same certainty. It could turn out that free will is an illusion, that it feels as though we are free when we are not. The idea that nobody really feels pain is, he thinks, much harder to make sense of. What he does feel strongly is that there are no good scientific or philosophical arguments against strong, libertarian free will — some of our decisions being uncaused. He disagrees with Sam Harris's claim of an overwhelming case. Ultimately it is an empirical question, and we do not know enough about the brain to settle it.

One common argument, which he has heard from Harris, is that there is no middle way between a choice determined by prior causes and a totally random, senseless event, like the decay of a radioactive isotope. Jonathan Lowe, a hugely influential Durham philosopher who died of cancer, had a reply: what distinguishes a free action from a senseless random event is responsiveness to reasons. Deciding whether to take a job in the United States, you weigh standard of living, family, cultural difference, and edge toward a decision. That is not a senseless occurrence, because the decision essentially involved weighing considerations. That, Goff thinks, answers that particular objection.

The Libet experiments, in which people press a button at random while brains are scanned, claimed that activity leading to the action starts about half a second before the conscious decision — later estimates go to seven to ten seconds. Goff's complaint is not only the usual methodological ones. As far as he knows, the 'choices' studied are totally random, senseless actions. The free will we care about involves weighing whether to get married, and those decisions may not happen at an identifiable instant; you edge slowly. Maybe Libet shows we cannot freely choose to do something totally pointless. That does not show we cannot respond, in a strong libertarian sense, to reasons and values.

The physics objection is the one he keeps having with Sean Carroll. Carroll focuses on the 'core theory' — the Standard Model plus the weak-gravity limit of general relativity — and says that if there were libertarian free will in the brain, the core theory would be wrong. Goff is not sure. Perhaps strong emergence in the brain would mean not that the core theory is false in its own terms, but that particles and fields are not the only things running the show: systems might have fundamental causal capacities too, co-determining what happens. The branding matters less than the empirical point. If evidence showed everything reducible to underlying physics and chemistry, he would drop libertarian free will in a heartbeat. We are, he says, a million miles from that stage.

Do we live in a simulation? Goff's first word is no, but he will not laugh the argument away. It is important and serious. The issue he has with it is a crucial assumption: that consciousness is substrate-independent, software rather than hardware, dependent on organization rather than stuff. As a panpsychist he thinks consciousness *is* the stuff of the brain, the stuff of matter. Taking the organizational properties, the software, and uploading them would not get you the stuff.

That is why he worries about a future in which we start uploading minds, think granny is still there, and email her after her body has rotted in the ground. We might just be committing suicide — getting rid of our consciousness — because replicating software features would not, for him, preserve the experience. Fridman finds the stance surprising. Intuitively, if consciousness is the base layer, moving it into the cloud or a computer should be easier, not harder. Goff agrees that panpsychism allows artificial consciousness: there are no souls, no extra magical ingredients. His only point is that replicating computational or organizational features would not preserve *his* consciousness.

Opponents of materialism disagree among themselves here. Chalmers is a kind of dualist, but he thinks psychophysical laws hook onto the computational or organizational features of matter, so uploading could work. Goff tends to think not. In that sense, substrate matters to him, and he agrees with Carroll that physics is our best way of capturing what the stuff of the world does — just not the whatness, the being, the 'is-ness' of the stuff. Fridman notes that Russell Brand had reached for the same word.

In *Galileo's Error*, Goff imagines the Hitchhiker's Guide entry for our reality: a physical universe whose intrinsic nature is constituted of consciousness, worth a visit. The whole conversation has been about the first half of that sentence. Fridman wants the second. Why is this place worth a visit? Why does it have meaning or value at all? Goff starts with the same incompatibility he opened with. Consciousness is the root of everything that matters in life — deep emotions, subtle thoughts, beautiful sensory experiences — and yet the official scientific worldview, as he reads it, cannot admit that consciousness is real. People feel this on an intuitive level. It is perhaps part of what Max Weber called the disenchantment of nature: they know their feelings are not just electrochemical signaling.

That intuition, he thinks, can be rigorously supported, and its cultural effect is alienation — a missing framework for the meaning and significance of our lives. In the absence of that framework, people turn to nationalism, fundamentalist religion, consumerism. Panpsychism matters here because it brings together the quantitative facts of physical science with what he calls the human truth: the qualitative reality of our own experience. He also thinks there are objective facts about value, about what we ought to do and believe, mostly to do with kinds of conscious experience. They are mysterious, both in how they could exist and in how we could know them, but much of the human condition is the work of discovering those sources of value.

At a still bigger level, it is important to him to live in hope that there is a purpose to existence, and that what he does contributes in some small way to it. He does not know whether there is such a purpose; some things point toward it, some away. You do not need certainty, or even high probability, to have faith. Suppose a friend is seriously ill and there is a 30 percent chance they will make it. You should not *believe* they will get better, because they probably will not. You can still say you have faith they will — meaning you choose to live in hope, to orient your life toward that possibility. He does not think we know whether existence has a purpose. He thinks we can choose to live in hope of that possibility, and he finds that a worthwhile and fulfilling way to live. Fridman, playing editor, suggests amending the Guide entry from 'worth a visit' to 'hopefully worth a visit.'