Materialism Cannot Be True

Why this agnostic monist with no taste for the supernatural thinks the materialist picture of reality defeats itself, and why the evidence has started to say so.

Tell people that materialism cannot be true and they usually hear one of two things. One group hears a confession of faith wearing philosophical clothes, an argument whose last move will be a soul, an afterlife or a God who listens to prayer. The other hears someone who has failed to notice how much has been explained by physics, chemistry and neuroscience. I mean neither, and the gap between what I mean and what is heard is wide enough that I want the argument laid out whole, if only to see for myself which parts bear weight and which do not.

Monism is my starting point. Reality, on my view, is one, and consciousness, whatever it turns out to be, is neither a ghost in a machine nor some substance imported from elsewhere. I also reject the supernatural, if the word names a second order of being that intrudes on nature and overrides its regularities, and I refuse the suggestion that giving up materialism obliges me to accept anything of the kind. That suggestion is among the most effective bits of framing in modern thought. Offer only two choices, matter or magic, and any case against materialism turns into a case for magic, so that the materialist wins without having to argue. Elsewhere I have described the binary as a stage on which both sides perform, and explained why even a nontheistic metaphysics owes nothing to physicalism. Two choices was always one too few. A third exists, and it has a long history.

Einstein’s view was near to it. In 1929 Rabbi Herbert Goldstein asked him whether he believed in God, and his cabled answer was that he believed “in Spinoza’s God”, the God revealed in the orderly harmony of what exists, and not one occupied with the fates and doings of human beings. Spinoza’s God has no personality and does not interfere; it is the single substance of which all things are modes. On mind and body, his core claim is that we have one thing grasped two ways, not two things acting on each other. Part II of the Ethics compresses this into one sentence: the order and connection of ideas is the same as the order and connection of things. Extension is not caused by thought, nor thought by extension. Their parallel course follows from their being, at root, the same.

What the cable licenses is narrower than it may appear. We learn from it that the twentieth century’s most celebrated physicist saw nothing unscientific in a non-materialist monism. We do not learn that he would have endorsed my objections to materialism, and I have no intention of enlisting him in arguments he never made. The cable does show that nobody who thought seriously about physics was ever compelled to choose between matter and magic. Spinoza’s view, along with the dual-aspect monism I regard as its modern heir, treats the physical and the experiential as two aspects of one underlying order, neither of which reduces to the other and neither of which is supernatural.

The question Goldstein asked Einstein is one monism has made me answer differently too. These days I am readier than I once was to allow that a God exists. Monism raises God’s plausibility for me and puts me squarely among the agnostics. Rather than disproving God, materialism barred the question from the outset, because if the physical aspect is everything, reality has nothing mind-like at its foundation. Monism removes the bar while leaving the question open. Should experience be part of the fundamental order instead of a latecomer in complicated matter, we may reasonably ask whether the whole has an inside, and that I am not yet able to answer. Nor would a Spinozist God count as supernatural in the sense I reject, since it would simply be nature taken whole. The argument below is thus neither apologetics nor an atheist’s case.

God, moreover, names two separate questions, and they point in different directions. Nothing in monist metaphysics excludes a God who made the cosmos we live in. Plato’s Timaeus offers one in its divine craftsman, who forms our world imposing order on chaotic materials instead of producing it out of nothing. Such a maker would be one subject inside the single order, bound by the conditions it imposes, and so a God with limits. A dual-aspect monist can accommodate that God as a further standpoint alongside the others, whatever its power. Treating God as the ground of reality is another matter entirely. Were a mind the very reason there is an order at all, the experiential aspect would take priority over the physical, which is idealism, not dual-aspect monism. Neither claim is mine. Still, if I came to believe in a God of the second kind, I would shift toward idealism, and the monism defended here would have to yield.

For now, dual-aspect monism is where I argue from. According to materialism, the physical aspect is everything: what physics describes, or would describe once complete, exhausts reality, and consciousness either is some physical process or is an illusion one generates. My case against this comes in two parts. First, materialism contradicts itself, since it adopts as its criterion of truth a standpoint nobody has occupied or ever could. Second, the program erected on that standpoint is not converging. Its own assumptions predict that open questions should be dwindling, and instead they are proliferating.

⁂

A distinction I have relied on for years comes before the contradiction: method is not metaphysics. Science disciplines belief against error. Its technique is to isolate whatever can be measured, repeated and shared, and to set aside whatever differs between one observer and another. Its power comes from that bracketing. But bracketing is a choice, and choosing what an inquiry will ignore does not amount to discovering that the ignored thing is absent.

We know when the choice was made and who made it. Galileo, in The Assayer of 1623, drew a line that physics has observed ever since. We must think of a material body, he writes, as possessing shape, size, place and motion, yet the mind is under no compulsion to think of it as white or red, bitter or sweet. Tastes, odors and colors, he decides, are “no more than mere names” as far as the object goes, and they exist only in consciousness. Take away the living creature and those qualities vanish, leaving shapes, numbers and motions behind.

Color, taste and sound went unexplained by Galileo. He moved them instead, from the object into the mind, and then constructed a science of objects that was excused from accounting for them. The move was brilliant, since it allowed mathematics to take hold of nature without being held back by the qualitative grain of experience. Yet it left a bill for later. Physics set the qualities of experience aside exactly so that it could get on without them, and so it is in no position to declare afterwards that it has explained them, or that they amount to nothing beyond the shapes, numbers and motions it retained. Exclusion was the founding act of the method. What a method excluded at its founding, it cannot later prove to be nothing but the remainder it kept.

Usually the materialist replies that scientific success entitles us to extrapolate. In The Moon Does Not Prove Physicalism I dismantled that reply, and the gist is this: when a rocket reaches the Moon, what is confirmed is the physics used to build and guide the rocket. Nothing further follows about whether matter produces consciousness. The success belongs to the job the method was built for. Extrapolating from it quietly assumes that what the method set aside is the same sort of thing as what it captured, and that assumption is the entire dispute.

---

---

So to the contradiction. Third-person description is not, for materialism, merely a tool. It is the yardstick of the real. On this view a fact counts as objective when it would hold for any observer, or for none, so that it describes the world from no particular vantage. Anything that cannot be caught from that vantage gets demoted to appearance, to how things seem to beings such as ourselves. Thomas Nagel gave the ideal its name in The View from Nowhere, and the materialist hopes to describe the world from that nowhere and afterwards slot the observers in as objects inside the description.

Nowhere, though, is not a place one can stand. Each observation in the scientific record was made by somebody, at some place and time. A dial shows a reading only once someone sees it, and a measurement turns into data only when it is noticed, written down and understood. Arthur Eddington grasped this in The Nature of the Physical World in 1928 when he remarked that exact science has as its entire subject matter “pointer readings and similar indications.” Such readings are delivered to nobody from nowhere. They come in experience, to a subject, and all objective knowledge is assembled out of agreement between first-person reports. Wherever we have actually achieved objectivity, it has been a convergence of perspectives, never their absence.

This is the sense in which I call the third-person standpoint phenomenologically impossible. It is not just that we have so far failed to get there. Observation is itself a standpoint, so the idea of observing from none makes no sense. Every datum the materialist uses is borrowed from first-person experience, and the picture built from those data is then used to pronounce first-person experience derivative, illusory, or nothing beyond the picture. Away goes the ladder, though it was the ladder keeping the building up.

Few people did more to construct the quantum mechanics that materialists now invoke than Erwin Schrödinger, and he saw the problem as one of method. In Mind and Matter he set out what he named the principle of objectivation: “we exclude the Subject-of-cognizance from the domain of nature that we endeavour to understand,” withdrawing into the role of a spectator who is not part of the world, which thereby becomes an objective one. This, he said, was a simplification we adopt in order to master a problem of impossible intricacy, and he went on to name its cost. Once we have taken ourselves out of the picture, we reinsert ourselves as just another object, and what follows, as he tells it, is a series of unbearable antinomies, above all our futile hunt for the point at which mind acts on matter. Naturally that point eludes us. The material world was built by removing the mind, and a picture drawn on the condition that the mind be absent cannot contain it.

Here, then, is the contradiction in the words of one of the physics’ founders. Nobody is being called dishonest, and objectivity is not being called a sham. As a procedure, objectivation is legitimate and immensely fruitful. Contradiction enters only once the procedure is elevated into an ontology, and the spectator who left the world for reasons of method is informed that he was never in it.

From the opposite side, Nagel reached the same point in “What Is It Like to Be a Bat?”, published in 1974. His argument was that an organism is conscious if and only if there is something it is like to be it, and that this subjective character slips through every reductive analysis available, since each is logically compatible with its absence. The diagnosis concerned structure. Each subjective phenomenon is bound to one point of view, and an objective, physical theory seems bound to abandon that point of view. A method that operates by giving up the point of view has no way to recover what essentially belongs to one. Gathering more data changes nothing, because the trouble is the form of the instrument, not a lack of evidence.

Physics supplies a further twist. Bertrand Russell and Eddington each observed that physics characterizes the world through structure and dynamics, that is, through how things relate and how those relations alter. In the account given by the Stanford Encyclopedia of Philosophy, this is a structuralist picture in which physics says nothing about what underlies the structure and dynamics it describes. Mass is fixed by how something resists acceleration and draws other masses toward it, and charge by how it couples to fields. Every property is given by what it does relative to the rest, and not one by what it is intrinsically.

My own name for this elsewhere is the principle that description is not exhaustion: to say what a thing does is not thereby to say what it is. That matters a great deal to the materialist, whose claim is precisely that physics says what everything is. If what physics actually supplies is how things behave toward each other, then saying consciousness is “nothing but” physical stuff lacks definite content. Nothing but which stuff? Of matter we know the relational profile in exquisite detail, and its intrinsic nature in a single case, our own experience, in which we meet a portion of the world from within. The dual-aspect monist’s conclusion, drawn with care, is that experience is the inside of what physics describes from outside, and that neither perspective is complete.

Dual-aspect monism is not proved by any of this. What is shown is that the materialist has taken for granted a concept of the physical that physics itself does not provide. The third-person picture is the view from outside, and a view from outside has, by its very construction, no inside. Inferring from this that no inside exists mistakes the boundaries of a method for the boundaries of the world.

⁂

Neuroscience’s habits of speech display the contradiction on a small scale. Pick up a popular book about the brain, or quite a few technical ones, and the brain turns out to see, choose, believe, predict, desire and deceive us. In 2003 the neuroscientist Max Bennett and the philosopher Peter Hacker brought out Philosophical Foundations of Neuroscience, an extended charge sheet against that habit. Psychological predicates that hold of human beings as wholes, they argue, cannot intelligibly be ascribed to parts of them such as the brain. They named the mistake the mereological fallacy, after the Greek word for part. Seeing and deciding are done by people. People need brains in order to see and decide, yet a brain does not see any more than a leg walks.

They drew their examples from the field’s most eminent figures, among them Francis Crick, Antonio Damasio and David Marr. Writing in Notre Dame Philosophical Reviews, the reviewer summed up the accusation: the Cartesian mind had not been eliminated but moved, by “replacing the ethereal by grey glutinous matter.” A neuroscientist who says the brain builds a model of the world and then consults it has installed a little person in the skull, handed it a model to inspect, and labelled the little person a brain.

Replies came from Daniel Dennett and John Searle in the 2007 book Neuroscience and Philosophy. According to Dennett, neuroscientists employ these predicates in an “attenuated” sense everyone grasps, so that Bennett and Hacker were teaching their grandmother to suck eggs. Searle’s charge was that the two authors were confusing the rules for using the words with the ontology.

Dennett gives away more in that reply than he intended. Attenuated predicates make for attenuated explanations, and we are then entitled to ask what the full-strength sense is doing whenever such explanations are offered to the public as accounts of what seeing or believing really consists in. Shorthand is defensible only so long as somebody can redeem it, and redeeming it is exactly the step nobody has taken. Searle, for his part, was defending his own biological naturalism, according to which consciousness is a genuine first-person feature that cannot be eliminated. An eliminativist will find little comfort there.

Neuroscientists are not the only ones with the habit. In a recent passage, the philosopher Dan Williams writes of “a powerful but largely unconscious ‘social diffusion bias’” by which “our brains evaluate information.” The rest of his work is about the ways reputation, status and group loyalty shape what we believe, an account that is social rather than neural, and “our brains” is quite probably a figure of speech. Which is exactly my point. The figure arrives so easily that it goes unnoticed. Information is evaluated by people, in social contexts, for reasons. Hand the evaluating over to an organ and the person vanishes from the explanation, while the organ acquires the very mind it was meant to account for. In the mereological fallacy, the third-person contradiction becomes grammar: our whole vocabulary for consciousness was learned from within and applied to whole persons, and moving it onto their parts renders it incoherent while it goes on sounding scientific.

⁂

My case’s second part is empirical, and it calls for more caution, since science always has open problems and a mere list of them proves nothing.

Stated exactly, the claim runs as follows. Were materialism true, and the method grounded in it tracking reality’s structure, the frontier ought to close as time passes. It need not close steadily, nor without surprises, but overall the territory of fundamental mystery should contract as instruments get better. What we find instead is that better data brings deeper questions, a number of them questions the framework never anticipated.

Take the brain first. How do features handled in separate regions, with color processed in one area and motion and shape in others, get joined into one unified experience? That is the binding problem, and decades of research later, “no complete model exists,” while one review chapter calls it “still unresolved.” Nowhere does the outside description meet the inside fact more directly. From within we know that experience is unified, more surely than we know any fact about neurons, whereas the outside description offers distributed processes. The unity is the thing to be explained, and no one has explained it.

A similar story comes from the most generously funded effort to test theories of consciousness head to head. Nature published in 2025 the results of an adversarial collaboration setting integrated information theory against global neuronal workspace theory, in which 256 participants were studied with functional imaging, magnetoencephalography and intracranial recordings. Some predictions of each theory were borne out, while the findings ended up substantially challenging key tenets of both. It is a candid and creditable result, and it portrays a field in which both of the leading theories were badly hurt the first time they met a carefully designed test.

Particle physics next. No theory in the history of science has been confirmed more precisely than the Standard Model, and with neutrinos treated as massless its dynamics rest on 19 parameters whose values are established by experiment, a figure that climbs to 26 when neutrino masses and mixing are added. Why those numbers have the values they do, the theory does not say. It accepts them as given. A program that aims to explain everything from the ground up, and whose ground consists of two dozen numbers it cannot explain, has less a foundation than a ledger of measurements it cannot account for.

Best known among those puzzles is the hierarchy problem, which asks why the Higgs boson is “a hundred million billion times less massive” than nature’s highest energy scales. Physicists expected for decades that new particles, supersymmetric ones above all, would show up at the Large Hadron Collider and restore what they call naturalness. None showed up. According to Quanta Magazine, the discipline had by 2016 entered a textbook Kuhnian crisis, and physicists have been reconsidering ever since the long-held assumption that big stuff consists of smaller stuff. Far from peripheral, that assumption simply is reductionism. Because the costliest experiment ever constructed failed to find what reductionist reasoning foretold, some of those who operate it now ask whether the world is assembled from parts in the manner the materialist picture supposes.

Gravity after that. Each of general relativity and quantum theory is magnificently confirmed within its own domain, yet the two will not fit together. Quantum gravity, in the description offered by the Stanford Encyclopedia of Philosophy, remains under construction more than a century on, with “no single, generally agreed-upon body of theory” and with neither principal research program yet shown to recover the world we observe at low energies. My own work argues that the obstacle is in part metaphysical, because the theories disagree on what time is, and one places the observer outside the system in a way the other cannot absorb. The Birmingham experiment discussed in The Eternal Now is a Laboratory Result, which reconstructed time from correlations within a system instead of importing it from outside, offers a small empirical clue to how that dispute may resolve. I treat it as a clue and nothing more.

Then cosmology, where anomalies keep coming in faster than the theory can digest them. Many expected the Hubble tension, the mismatch between the expansion rate measured nearby and the rate inferred from the early universe, to disappear once the James Webb Space Telescope tested whether crowded starfields had skewed the nearby measurements. Instead, by one recent account, Webb ruled out the crowding explanation and confirmed the higher number. The Dark Energy Spectroscopic Instrument collaboration then reported in 2025, on the basis of nearly 15 million galaxies and quasars, mounting evidence that dark energy has weakened over time. Should that stand, cosmology’s standard model, in which dark energy is a constant, will need revision at its base. Having written already about the state of cosmology and about the habit of deferring inconvenient observations to a scale not yet surveyed, I will observe only that the field now speaks of crisis where it once spoke of refinement.

Biology, finally, where reductionists expected form to follow once the genome had been read. Work in Michael Levin’s laboratory has shown that electrical potentials in cells outside the nervous system serve as instructive prepatterns for gene expression and morphogenesis, and that a planarian flatworm’s target body shape can be rewritten permanently without genomic editing or transgenes. Levin’s work contains nothing non-physical. Still, it is one more instance in which the explanatory level that counts proved to lie above the lowest, and in which the organism acts as a whole in ways its parts failed to predict.

No single result on this list refutes materialism. Materialist research programs are attached to every one of these problems, and some of them may succeed. The list reveals a trend instead. In foundational physics, in cosmology, in consciousness science and in developmental biology, ten years of improved instruments have generated more fundamental questions, not fewer, and a number of the costliest tests of reductionist expectation have returned negative. A framework that tracked reality’s deep structure should not, taken as a whole, generate that pattern. One would expect it from a framework excellent at one aspect of reality that is being driven into ground where that aspect has ceased to be the whole story.

Nor does evidence against materialism count, on its own, as evidence for the alternative I hold. Idealism is compatible with the same anomalies, as are various forms of neutral monism, some future physicalism quite unlike the present one, and simple ignorance. Non-convergence undermines the materialist’s confidence that the frontier will close on his terms, but it does not close the frontier on mine. Dual-aspect monism would be specifically favored by a prediction that its rivals do not make, and that work is under way elsewhere.

⁂

Gödel came last on my list, and he gets a section to himself because no item is more easily misused.

Kurt Gödel showed in 1931 that every consistent formal system able to express basic arithmetic contains truths it cannot prove. J. R. Lucas, a philosopher, argued in 1961 in “Minds, Machines and Gödel” that the result shows the mind is no machine. The Internet Encyclopedia of Philosophy sets out the steps: a machine generating theorems of arithmetic can be represented by a formal system; a consistent system has a Gödel sentence, true yet unprovable inside it; anyone who understands the construction can see the sentence is true; hence that person is not that machine. Adding the sentence to the machine yields a new machine with its own Gödel sentence, and the person can always take one more step.

In Shadows of the Mind, published in 1994, Roger Penrose took up the argument and concluded that human mathematical understanding is non-computable, out of reach of any algorithm. Rather than inferring that the mind is immaterial, he infers that our physics is incomplete, and that we will need physics beyond computational simulation to understand consciousness. Together with the anesthesiologist Stuart Hameroff he built the Orch OR proposal, tying consciousness to a gravity-related collapse of quantum states inside neural microtubules. The pair’s 2014 review concludes that “consciousness plays an intrinsic role in the universe,” and Hameroff locates the precursors of experience in the fine-scale geometry of spacetime.

Penrose is hardly a figure on the margins. As a doctoral student under Dennis Sciama, Stephen Hawking built on Penrose’s singularity theorem, and later the two jointly proved the general theorem of 1970. Penrose shared the 2020 Nobel Prize in Physics for showing that general relativity robustly predicts black holes. So when he claims the mind is not an algorithm and that physics must change to make room for it, nobody can wave the claim away as a misunderstanding of physics or mathematics.

Contesting it is another matter, and since the conclusion suits me, the objections merit closer scrutiny, not looser. Consistency is the gravest issue: a Gödel sentence is assured of truth only when the system is consistent, and by Gödel’s second theorem no system can prove its own consistency. Whoever runs the argument must therefore know his own reasoning to be sound, the assumption David Chalmers singled out as its “greatest vulnerability.” Mathematicians err; a faulty proof of the four-color theorem survived for eleven years. Paul Benacerraf further suggested that a mind which is a formal system might be too complex for that mind to grasp, so that no one could ever write down its Gödel sentence. As for the physics, Max Tegmark computed in 2000 that microtubule coherence would decay in around 10⁻¹³ seconds, much too quickly to matter, to which Hameroff and his colleagues answered that he had modeled something other than their proposal. Chalmers also wondered why quantum processes should yield consciousness when classical ones do not, and proposed that Penrose may have “simply replaced one mystery with another.”

Describing the argument as “very controversial”, the encyclopedia stops short of calling it refuted. As I read the objections, they show that the argument does not prove the mind non-computable, and what remains after them is humbler yet adequate for what I need. Each of the logical objections suggests that our hold on mathematical truth may not be what it appears to be from within: perhaps we are inconsistent unawares, or running an algorithm too intricate to examine. Every one of them saves materialism by marking down first-person evidence. It seems to us that the Gödel sentence is true, and we are told the seeming may mislead. That is the same move the illusionist makes about consciousness generally, at the same price. The third-person picture is protected by casting doubt on the first-person evidence out of which it was constructed.

The point I make about Gödel is one Penrose defends, and he defends it as a physicist convinced that consciousness is physical. The argument’s most formidable living champion therefore sides with a physics not yet written, one in which mind belongs to reality instead of being something computation outputs. That is not my position, but it lives on the same street.

⁂

From an unexpected quarter comes the second wing of the argument from authority. Panpsychism, the view that consciousness or something like it is a basic and pervasive feature of reality, is currently receiving some of its most serious attention from inside neuroscience.

Panpsychism is not my view. In its usual forms it gives the basic constituents of the physical world experiential properties of their own, and then runs into the combination problem, which asks how micro-experiences merge into a unified mind. That is binding again, moved over into metaphysics, and moving it does not make it easier. As I see it, experience is an aspect of one underlying order, its depth of inside-ness scaling with how recursively a system is organized, rather than a property scattered across the smallest pieces. Still, panpsychism sits close to monism. It shares my view that consciousness is no late by-product of complicated matter, and that the physical and the experiential have belonged to one reality from the beginning.

The plainest example is integrated information theory. Its developer was the neuroscientist and psychiatrist Giulio Tononi, and its champion Christof Koch, who was once president and chief scientist of the Allen Institute for Brain Science. On Tononi and Koch’s account, IIT regards consciousness as “an intrinsic, fundamental property of reality,” and Koch has described it as “a scientific, constructive, predictive and mathematically precise form of panpsychism” fit for the twenty-first century. Consciousness, he adds, “depends on a physical substrate but is not reducible to it,” which comes about as near to a dual-aspect formulation as a working neuroscientist is likely to get.

Of the two theories examined in the Templeton-funded adversarial collaboration mentioned earlier, a program that cost more than $6 million, IIT was one. The fiercest reaction against it likewise arose inside the field. During 2023, 124 researchers signed an open letter branding IIT pseudoscience, partly on account of its “panpsychist commitments,” and among the signatories were Adrian Owen, Chris Frith and Joseph LeDoux. Anil Seth, hardly a panpsychist sympathizer, defended the worth of wild ideas against the letter’s tone.

Neuroscience, in other words, is not so much warming to panpsychism as quarrelling over it, and the quarrel counts as evidence. Were the materialist answer settled in the field’s eyes, a rival would simply be ignored, not kept out by 124 signatures. Letters like that get written when a serious portion of a field no longer finds the default answer sufficient, and the remainder senses the ground shifting.

⁂

Illusionism is the most formidable materialist answer to all of this. Keith Frankish, and Dennett before him, maintain that phenomenal consciousness, the felt quality of experience, does not exist, although it appears to. In the words of one recent defense, the position joins an “eliminative claim: phenomenal consciousness does not exist” to an appearance claim, that it nonetheless seems to. Dennett’s phrase for consciousness was a “user-illusion,” comparable to the desktop icons a naive user takes for the machine’s inner workings.

Consistency is illusionism’s strength. It takes the third-person standpoint as reality’s measure and follows that measure to the bottom: since the outside view contains no inside, there is none, and the impression of one is a representation the system generates about itself. I answer that an illusion is the way something seems to somebody, and that without a subject for whom the seeming happens the concept does not apply. Illusionists reply that the seeming may be construed functionally, as a tendency to form certain judgments, with nothing it is like to have that tendency. In that case, though, the theory has redefined the appearance in terms which, as Nagel pointed out, are logically compatible with its absence. Its starting datum is that there is something it is like to see red, its finishing theory says there is nothing it is like to see red, and the result is presented as an explanation of the datum.

History provides a second reply. Each era has had its open problems, the materialist observes, and each era’s critics have taken them for symptoms of failure. Lord Kelvin noticed two small clouds above nineteenth-century physics, which grew into relativity and quantum mechanics, both triumphs of physical explanation. What makes today’s clouds any different? One part of my reply is that Kelvin’s clouds really did overthrow the framework. Classical mechanics failed to survive them; what lived on was the method, and the physics that replaced the old was far odder, and far less hospitable to naive materialism, than Kelvin could have pictured. Quantum theory wrote measurement into the foundations of physics, and nobody has managed to erase it since. The analogy demonstrates that open problems resolve toward better physics, and better physics has again and again drifted away from the image of little objects in a container, watched from nowhere.

The third reply accuses me of borrowing the intuitive pull of the hard problem and labelling it a contradiction. A materialist may concede that data originate at the first-person standpoint and yet insist that what occupies the standpoint is a physical system, no more contradictory than a thermostat taking a measurement. That sentence, I grant, holds no formal contradiction. The contradiction sits in the completeness claim about third-person description. If a physical system observes, then there is a fact about how matters stand from its standpoint, and completeness holds that the third-person description of the system captures that fact. Here Nagel, Schrödinger and the structuralists meet: a description built by removing standpoints cannot be complete about standpoints. The materialist is free to abandon completeness, but at that point he is no longer a materialist of the kind I oppose, having become something nearer a dual-aspect monist who has not yet realized it.

⁂

A further consequence follows, and its target is organized religion, not materialism. Whatever God these terms allow must be either a limited creator inside the single order or a ground with no personhood. The God taught by most religious institutions, a personal being who is simultaneously the ground of all things, all-powerful and all-knowing, is precisely what the argument rules out. Were the institutions backed by their own scriptures and teachers, this would be merely a philosopher’s dispute with theology. Frequently they are not backed. Judged by religious teaching itself, the picture of God that organized religion offers deserves suspicion, and the evidence lies in the texts it venerates and the mystics it condemned.

Begin with the earliest stratum of Jewish scripture. Early Jewish scripture suggests a very limited God, one who takes human form and must make the physical journey to Sodom to check what he has heard. Three men come to Abraham’s tent in Genesis 18 and eat the food he offers, and before they depart YHWH declares “I will go down now, and see whether they have done altogether according to the cry of it”. The Babel story has the same shape, with the Lord who descended to see the city and the tower. In The Bodies of God and the World of Ancient Israel, Benjamin Sommer treats passages like these as remnants of a God possessing more than one body and fluid selves, an idea that biblical religion’s dominant currents later abandoned. Notice how the tradition handled them. Rashi, discussing Babel, explains that God “did not need to do this, except to teach judges not to condemn a defendant until they see” what is before them. According to the text, God went to inspect; according to the commentary, he never had to, and the trip becomes instruction for human judges. When doctrine and text conflicted, it was the reading that changed while the doctrine stayed put. The commentators deserve this much: the text is mixed, and the very chapter that dispatches God to Sodom has the visitor knowing that Sarah laughed inwardly. Even so, these stories give God a location: he is in a place, journeys, arrives and learns. Omniscience came later, once theology stripped God of every standpoint, in much the way physics after Galileo stripped out the observer, and theology did this in defiance of its own earliest scripture.

Hinduism, too, gives God many faces, and you are among them. In the Rig Veda, Indra, Mitra, Varuna and Agni appear as faces of one reality, in a verse commonly translated “Truth is one, sages call it by various names.” The schools part company over what this means for you, and they part along the same fork. For Advaita Vedanta the key text is Tat Tvam Asi in the Chandogya Upanishad, “That thou art,” whose “that” is the Existent, “the self of this whole world.” This time the critique is internal to the tradition, and stated in the open. Shankara treats the personal God of devotion, the one most worshippers pray to, as Brahman as it appears from the standpoint of ignorance, while your innermost self simply is Brahman, understood as pure consciousness. A towering teacher of the tradition places the God of ordinary worship below the ground. In Ramanuja’s Vishishtadvaita the faces stay real, since the world and its souls are modes of Brahman and form its body, separate from it yet entirely dependent, the way a body relies on its self. Dual-aspect monism lies closer to Ramanuja, minus a personal God at the center. Every one of us is an indexed standpoint, an inside of the order, genuinely distinct and never the totality. Where Genesis grants God a standpoint, Hinduism makes each of us one of God’s.

What an institution does when its own teacher takes that continuity seriously is on display in Christianity. Meister Eckhart, a Dominican and master of theology, ranked among his century’s most admired preachers. One of the claims he stood accused of preaching held that “There is something in the soul that is uncreated and not capable of creation”, a near relative of Tat Tvam Asi. Pope John XXII’s bull In agro dominico, dated March 27, 1329, condemned 28 propositions taken from or ascribed to his teaching, and by then Eckhart had most likely died. No one ever pronounced him personally a heretic; the bull noted that he had submitted and condemned what he had said. The same pattern runs through all three traditions. Their scriptures and teachers present a God who occupies a standpoint, a personal God placed below the ground, a soul continuous with the divine. The God of the institutions endures by revising the reading, as happened with Genesis, or by condemning the teacher, as happened with Eckhart, while in Hinduism a master of the tradition states outright that the God of ordinary worship is not final.

⁂

The supernatural is not what survives. Nowhere does the argument call for a soul outliving the body, a God who overrides nature’s regularities, or a domain of spirit outside it. Neither does it exclude God, which is what monism has shifted in me. My case is for an enlarged nature, not an escape from it: the view Spinoza held and Einstein admired, of a single reality known to us under two aspects, each irreducible to the other. One aspect is the business of physics, which handles it magnificently. The other we know by acquaintance, as experience, and it is the ground of every observation, those of physics included. Materialism mistakes the first aspect for the whole and then searches it in vain for the second, just as Schrödinger predicted.

A useful fiction is all the third-person standpoint ever was. Stepping back from the world far enough to measure it served us well, and the measurements stand. Yet we never actually departed. Someone read every pointer reading, and someone understood every theory. What gets called the view from nowhere has always been a view from many somewheres, pooled and agreed upon, and that agreement is a fact about subjects. A metaphysics that opens by subtracting the subject and closes by deciding the subject was never present has lost track of what it took away.

Evidence is catching up with that, I think. Materialism has not been refuted by any single experiment, nor could it be. Even so, the boldest programs founded on reductionist expectation keep coming back with more questions than answers, and some of the people who run them have started to doubt the expectations. Particle physicists wonder whether big things are made of small ones. Neuroscientists argue over whether consciousness is fundamental. Cosmologists are learning that the constant at the foundation of their model may be shifting. A Nobel laureate in physics holds that the mind is no algorithm. Proof is not what this adds up to, though it is just what we would anticipate if the view from outside was never complete.

Trouble for the materialist picture does not vindicate mine. Dual-aspect monism must win its place as any theory does, by claiming what its rivals do not and submitting to tests of it, and that work remains unfinished. The question of God falls under the same discipline. Monism has opened it, and made God likelier in my eyes than before, yet likelihood settles nothing, and an honest monist stays agnostic until something can distinguish a God from an impersonal order that has an inside, and after that separate a limited creator inside the order from a mind that grounds it. What I can assert with confidence is narrower and, I believe, weightier. Materialism is a metaphysics, not a neutral default that science has earned, taken up for methodological reasons four hundred years ago and then forgotten as a choice, and like every metaphysics it may be mistaken. I believe it is mistaken, and for the single reason it was built never to examine: the one who is looking.

---

Go Deeper into the Circus

---