Emergence Is Not Magic

ai blog
August 5, 2026

Joel Kowalewski, PhD

The princess's objection

May 6th, 1643, The Hague. She writes in French, puts ink to paper. Plainly, for all who watch, her arms move like clever machines. But something hides in the mechanism. Motion is transferred by contact, pen to paper — one body pushes, and another body pushes back. She lifts her arm, pausing: What is the cause bringing the ink to the paper; what is this that decides in me? It escapes her perception, a seemingly formless and immaterial flow of being acting on the physical world. Of this world and yet somehow beyond it. She is twenty-four, writing to the mathematician and philosopher René Descartes.

Descartes’ reply would arrive in the hands of the young woman weeks later, though his three “primitive notions” — the soul and body and their union — compound her mystery. The union between soul and body, he writes, is not disclosed by pure intellect; it is known through the senses and the conduct of ordinary life. It is, in short, the vital essence of being alive.

Elisabeth of Bohemia asked the question and got the same question back. Descartes, the father of analytical geometry and inventor of the familiar X-Y coordinate system, put the ghost into the machine. A material, fluid-like substance comprised of small corpuscles — what he referred to as “animal spirits” — flowed through biological channels, rising to the brain and interacting with the immaterial soul at the pineal gland, a brain structure detectable in most vertebrate species that is now known to play a role in regulating biological cycles (e.g. circadian rhythm in mammals). At that time, strict materialism simply could not provide a reasonable account of human experience and volition. Descartes and many of history’s great thinkers failed to reconcile the apparent differences between and within species. Though these differences remain real, the difficulty has originated in part from the belief that human experience is not only distinct but superior. Without a physical, external counterpart, Descartes, as well as others throughout history, was left to describe human experience in fully subjective terms — terms that would only reinforce the apparent distinctiveness and superiority of humans, paradoxically driving human nature away from nature. We find ourselves like Descartes pondering our immaterial ghost-like existence in a physical and material world and wondering how, if at all, the interaction is even feasible. Perhaps, only mind exists. Maybe mind is just a clever illusion. That day in May of 1643 it could have been anyone writing the letters. It is the eternal dialogue. Only the language changes. In the 21st century, it is a public lecture series with a neuroscientist. Someone asks how eighty-six billion cells produce subjective experience. With the authority of Descartes in the 17th-century dialogue, the neuroscientist replies quickly and without providing further clarification, “emergence.”

Emergence, properly understood, is a rigorous and quantifiable consequence of physical law: it describes how individual parts interact and are progressively constrained by those interactions, giving rise to an apparently new level of description (the group or population). For example, imagine individual cells in the brain firing randomly. Because neurons are in networks, each neuron is constrained by the actions of its neighbors. Over time, “individual” becomes poorly defined as there is no longer a fully isolated neuron, doing as it pleases — firing randomly. Individual neurons, importantly, do not disclose what can or will happen in the network, giving rise to emergence. That means while the network is mathematical, physically instantiated, and governed by physical laws, it evolves unpredictably. We know the math and that emergence manifests itself as the boundary conditions on the math as the parts interact over time. We do not know what interactions will take place. This is nevertheless a wholly mathematical and physical description, though it does seem magical. A popular interpretation has us imagine emergence as a threshold across which new properties are said to appear, and it is difficult to not find something immaterial in emergence from this vantage point, hovering above the material substrate from which everything must, in truth, inevitably derive. But as the components or parts become too numerous, emergence presents as a paradox — arguably, the same apparent immaterial-material divide that led to the correspondence between Elisabeth and Descartes back in 1643.

Our ability to discuss human experience cogently will depend on our commitment to external, physically grounded terms and analogies. Once forbidden from the historical dialogue as it might have challenged the long-standing belief in human superiority, a physical and universal description — one incorporating other species and non-living systems — is the goal. The language of mathematics is what connects the seemingly separate threads of mind and body. We will find ourselves in emergent, elaborate and evolving patterns of chemical reactions in petri dishes, slime molds that navigate mazes, hurricanes overhead fueled by energy gradients that send powerful wind and rain toward the solid ground below, where there are plants that behave like the brains they lack — and we will find emergence.

The steam whistle

Elisabeth and Descartes represent just one half of a long standing dialogue on mind and mechanism. Throughout history, others have seen only the mechanism, untroubled by the interpretation of the human an elaborate machine. But classic and even contemporary mechanical interpretations have done little to fully address the apparent mysteriousness of our perceptual experiences. In moving the dialogue forward in time to the late 19th century, mechanical inventions have become more impressive. We find a world in part untroubled by mechanism because of its reach. Everything touches mechanism. But this very reach, something familiar and obvious, also extends further than perception can grasp. Mechanism vanishes on the horizon, regularly leaving us with just as much, if not, more mystery.

Addressing the British Association at Belfast in 1874, Thomas Henry Huxley proposed that consciousness in animals was “related to the mechanism of their body simply as a collateral product of its working, and to be as completely without any power of modifying that working as the steam-whistle which accompanies the work of a locomotive engine is without influence upon its machinery.” The whistle is real. It is audible. It is caused by the engine. It pulls nothing. Huxley meant this as the tough-minded conclusion of a consistent physicalism. For a century, it has been the shape that physicalist accounts of mind rely on. When a modern textbook says that consciousness is an emergent property of neural activity and stops there, it is in the tradition of Huxley: something else appears above or beyond the machinery, maybe it is correlated with it, but it is secondary, insubstantial or seemingly incidental to machine’s operation. There is no causal power. That is the business of the machine. Clearly, a great deal of mystery remains. Perceptual experience is not as trivial as the whistle of the steam engine. To address this physicalist accounts steadily inserted their own ghosts, unexplained physically untethered terms and entities such as “emergence.”

That word first appears one year after Huxley’s statement to the British Association. In Problems of Life and Mind (1875), George Henry Lewes distinguished a resultant from an emergent. When effects combine in a way we can trace step by step — forces adding, displacements summing — the outcome is a resultant. When they combine such that “we cannot always trace the steps of the process, so as to see in the product the mode of operation of each factor,” Lewes proposed to call the effect an emergent: something that “arises out of the combined agencies, but in a form which does not display the agents in action.” He was building on John Stuart Mill, who in A System of Logic (1843) had already separated the ordinary composition of causes from what he called heteropathic laws, where the joint effect is not the sum of the separate effects. Neither was initially reaching for a ghost. Both were making observations that the world was governed as much by non-linearity as by linearity.

In the 1920s the British emergentists — Samuel Alexander in Space, Time, and Deity (1920), C. Lloyd Morgan in Emergent Evolution (1923), C. D. Broad in The Mind and Its Place in Nature (1925) — turned emergence into a metaphysics of levels, each with its own irreducible laws, each arriving in the world as what Alexander called a brute fact to be accepted “with natural piety.” Focusing on chemistry, they suggested the properties of water could not in principle be derived from hydrogen and oxygen. Brian McLaughlin’s account of the movement’s collapse is that it did not lose an argument, it lost an experiment — quantum mechanics explained chemical bonding, and the doctrine went with it. Historians of chemistry nevertheless contest this exact interpretation story, but the point survives the dispute. Emergence began as a remark about non-additivity or non-linearity, became a claim about the strata of physical reality that would supposedly leave physics with little explanatory power, until it was quietly abandoned because of the explanatory power of quantum physics. When emergence appears in neuroscience, many initially take the 1875 claim about non-linearity seriously, but then find themselves using emergence to describe the implications of that non-linearity in seemingly metaphysical terms.

The whole is the interactions

To better understand emergence and its relevance to “mind,” it is important to define linear and non-linear. A system is defined as linear when its responses or outputs are the sum of its inputs: knowing that A is .5 and B is .5 does tell us what A and B do together (for example, Input A + Input B = Output and .5 + .5 = 1.0).

Complex systems are, however, governed by non-linear equations. The Navier–Stokes equations for a fluid are the standing example: each parcel of fluid changes the velocity field that determines what every other parcel does next, continuously. That relationality among the parts means we cannot define the system’s behavior as a sum of independent contributions such as A and B, as A and B begin to lose meaning. The imagery of anything individual has been lost, and the fundamental unit of physical reality has become the relationship. In a non-linear dynamical system, the mathematical description morphs to include an interaction term.

To illustrate with an example from biology, two proteins A and B combine to form an olfactory receptor. It is nonsensical to state that A now contributes 50% of the channel’s function, because A was formerly 0% of the channel. A and B are no longer discrete and independent parts. Because of the physical docking, they have undergone a conformational shift. The geometric shape of the A inside the complex is different from the A that exists in isolation. The functional unit of this biological system is the channel pore formed by A combining with B. That pore is made entirely of the relationship between A and B. Accordingly, the causal power of the system (what it can and will do) depends on A in relation to B.

Similarly, a single water molecule has no temperature. It is not wet, it has no pressure, it does not undergo a phase transition. These are not properties the molecule is hiding; they are meaningless at that scale, in the way that a single voter does not convey an election result. Temperature is a statistical property — the average kinetic energy of an ensemble — and when you assemble on the order of a mole of interacting molecules, the law of large numbers sharpens the distributions into quantities so stable that we can build engines on them. Macroscopic properties of this kind are, importantly, not less physical than the molecules. They are not floating above the molecules.

Philip Anderson’s 1972 article in Science, “More Is Different,” formalized the modern statement on emergence. Anderson was not attacking reductionism; he granted it. His target was constructionism — the assumption that the same fundamental laws operate at all scales, and that the universe can be rebuilt from those laws. At each scale, he argued, broken symmetry produces genuinely new organization requiring new concepts, so that each level is as fundamental as any other and “psychology is not applied biology, nor is biology applied chemistry.” This, however, was perhaps wrongly interpreted as certifying the independence of different scientific disciplines and legitimizing them, suggesting each field discloses its own laws. We find shades of the early 20th-century British emergentists, only now physics was to become a particular vantage point, not a universal science.

The whimsical interpretation of emergence was given new life. How the world appeared through subjective experience, from the vantage point of the psychologist or neuroscientist developing their specialized terminology, could very well be physical, law-like and legitimate. Should our subjective experience be misaligned with notions of “physical,” then the issue may not even be perceptual. There is something, as the argument goes, the physical cannot contain. Hence, reality could be legitimately metaphysical and the human being could once again persist above or beyond, defensible now because that persistence is operating at a “different scale.” There are nevertheless serious scientific points about the distinction between scales, prompting what might be seen as a natural extension to the classic mind-body problem first introduced in the dialogue between Descartes and Elisabeth. Dropping the obviously subjective terms — “mind” and “free will” and “consciousness” — a new generation of thinkers began to debate the causal power of the microscopic versus the macroscopic. Which is more real?

Why the microscopic details wash out

If macroscopic laws are nothing but a great many microscopic interactions, why should there be macroscopic laws at all — clean ones, with few variables, that work? This is the question that decides whether emergence is spooky, and it has an answer with a Nobel Prize attached to it. Kenneth Wilson received the 1982 Physics prize “for his theory for critical phenomena in connection with phase transitions.” The technique at its center, the renormalization group, is a mathematical procedure for coarse-graining: you average over the smallest scale, rewrite the system in terms of the resulting blocks, and repeat, watching how the description transforms as you zoom out. What the procedure shows is that most of the microscopic detail flows away. The parameters that distinguish one substance from another shrink toward irrelevance under repeated coarse-graining, and a small number of features — the dimensionality of the space, the symmetry of the order parameter, the range of the interaction — remain, governing the macroscopic behavior.

The consequence is, somewhat oddly, universality. A simple fluid at its liquid–gas transition and a uniaxial magnet losing its magnetization obey the same power laws with the same exponents. One is molecules of carbon dioxide, the other is electron spins in iron; they have essentially nothing in common at the level of their observable parts. Macroscopic laws are therefore not weaker approximations to the microscopic truth.

That verifies macro-laws, though it does not yet handle causation (the relationship) between micro- and macro-states. One objection grants that macro-descriptions might be practically useful (e.g. mathematical convention), but the micro-states still apparently do the actual work, and this introduces the classic mind-body paradox yet again. The question of what made Elisabeth of Bohemia’s pen move resurfaces in the question of what the macro-states do if they are apparently fully described by micro-states. Ghosts?

I have argued elsewhere that what we label “physical” is not a thing. It is relational — that the universe performs its relationships over time rather than storing them in things; the micro-macro divide is therefore simply a non-existent divide because “micro” and “macro” are not discrete and essential things. We are observing relationality. Our perceptual experience constantly deceives us. There would be something overwhelming, if not pathological, in encountering nature in its true form.

Consider a whirlpool. It is made of nothing but water molecules, and it is fully constituted by them at every instant. But once it exists, the whirlpool is a boundary condition on the molecules that enter it: their subsequent trajectories are set by the large-scale structure, and the individual molecule cannot be described without the macro-state. We wrongly approach them as two seemingly distinct things when they are a single, continuous process.

This is also comparable to an artist’s use of negative space; here, rather than explicitly draw or paint the foreground, the artist adds marks or strokes to the background and the foreground emerges. The human visual system assembles the composition giving rise to a complex perceptual experience. Is this composition really two distinct things or just a depiction of pure, physical relationality?

If the physical is relational, as I have argued here and elsewhere, our perceptual experience, while not altogether illusory, is very much incomplete, and we must look beyond ourselves to challenge appearances. On that view, the question of whether it is the whirlpool or its molecules that “really” do the causing has no answer — the framing is itself nonsensical, by which I mean, it lacks physical grounding.

We can see why, despite centuries of asking, the human experience is everything and nothing. Looking for a vital essence to explain differences that do not exist, we have found ourselves beyond nature: us-them and inside-outside. Our nature — mind, consciousness, and the like — is whatever we want it to be. Clearly, however, human nature — our nature — cannot be so arbitrary. And the hard questions that remain must drive, and are driving, us toward unfamiliar spaces.

There are only relationships

Figure. Negative space illustration. The vase or chalice (white) is not introduced directly (for example, that would be a drawing of a white vase on a black background). It is inferred or emerges from the silhouettes human faces in black on a white background. The brain processes relationality (this web of relationships) by breaking it up. We get two alternating percepts (faces and a vase). But is there actually two things or just one? Source: https://en.wikipedia.org/wiki/Negative_space

The hurricane has no essence

There is a tendency to imagine the world is comprised of objects with essences. The essence or properties are quite literally in the object. But a hurricane is not an object, and it uncontroversially lacks an essence. It is a process — a self-sustaining pattern of convection maintained by a flow of energy from warm ocean into cold upper atmosphere, which persists exactly as long as the flow does and dissipates when it makes landfall and the energy supply is cut off. Heat a shallow layer of fluid from below and you get the same lesson in miniature: past a threshold, the disordered jostle organizes itself into a lattice of convection rolls, the Rayleigh–Bénard cells that became the standard example of what Ilya Prigogine named dissipative structures, work for which he took the 1977 Nobel Prize in Chemistry. Order appears. Nothing was added but a temperature gradient.

And there is no saying precisely when the hurricane became one. The Saffir–Simpson categories are arbitrary — a Category 3 becomes a Category 4 at a wind speed chosen by us, for insurance and evacuation purposes, and the atmosphere knows nothing about the boundary. We cannot carve the storm into its parts and find a smaller storm inside; the eye is not a component, it is a feature of an ongoing relation, similar to the biological channel or receptor example and the negative space artwork example I discussed in the previous section. Suppose a meteorologist announced that hurricane-ness is an emergent property that supervenes on the air. There is no hurricane over and above the organized flow.

We do not paradoxically extend this logic to ourselves. We carve mind into attention, consciousness, volition, working memory, executive function — categories inherited from introspection and from a century of laboratory work — and then, having treated each as a thing, we go looking for the place in the brain where each thing is kept, and when we cannot find it we say it emerges. Alfred North Whitehead named this phenomenon in 1925, calling it the fallacy of misplaced concreteness: mistaking an abstraction for a concrete particular. The hurricane case shows the error because we do not demand anything beyond its physical nature. We accept and commit to the physical terms. Deriving the language of mind from the language of physical systems does not mean claiming that a person is a storm. It means giving up the expectation that the phenomena will come pre-divided into kinds with sharp edges and interior essences, and accepting instead that we are describing sustained organizations of matter and energy whose boundaries we draw for our own purposes and should be prepared to redraw. I have discussed this process in greater detail in “Myth is the First Draft,” where I describe the project of science as telling physically grounded myths, challenging these myths, and telling new ones. Although there is something desirable in our belief in objects with essences, as these beliefs lead us to organize the world accordingly, why these myths? Why should we be organized by something that, while useful, is also separating us from nature? Maybe it is time to ground the ghosts.

Grounding ghosts

Attention has historically been viewed as a spotlight sweeping over the sensory field, a phantom that demands explanation, unsure as we are how the spotlight is affixed. To what is the spotlight connected? We find a virtually identical challenge to the one Elisabeth had raised in her 17th-century correspondence with Descartes when she could not reconcile mental causes with mechanical consequences — two seemingly distinct, incompatible things or modes must interact.

The physical version replaces the spotlight with synchronization. A population of coupled oscillators, described in the model Yoshiki Kuramoto introduced in 1975, will spontaneously fall into phase alignment once coupling exceeds a threshold, with no conductor and no central command; the coherence is a consequence of the coupling itself. In cortex, Pascal Fries and colleagues reported in Science in 2001 that neurons representing an attended stimulus increase their gamma-band synchrony, and that this coherence, rather than firing rate alone, tracks what the animal was attending to. Synchronized input arriving in a narrow window has a disproportionate downstream effect, so alignment functions as amplification. On this account attention is not a beam falling on the data; it is the physical state of the data (neural activity) itself falling into rhythmic pattern.

Volition converts along similar lines. The ungrounded version has a ghost intervening in deterministic physics, which relates to Elisabeth’s question to Descartes. The grounded or mechanistic version relies on degrees of freedom and internal models. A falling rock has effectively none; a bacterium swimming up a sugar gradient has a few; a human brain is a non-linear dynamical system with an enormous number, and — on the predictive-coding account developed by Rajesh Rao and Dana Ballard in 1999 and generalized by Karl Friston — it does not merely react to input but continuously generates predictions and acts to reduce their error. Agency, in that frame, names a system whose behavior is driven principally by its capacity to update an internal model of the world. That model is far simpler for the bacterium, as it apparently has fewer physiological states (for example, a running mode for directed movement toward resources versus a tumbling mode or chaotic random movement to explore an environment for resources). For this reason, and I would argue this thinking derives from an arbitrary human vantage point, it seems like the bacterium does not choose, lacks volition, or has no free will. However, this explanation broadly assigns volition to all living, physical systems meeting the demands of shared thermodynamic challenges. All life shares very real macroscopic constraint on action. The often superficial impression that we are either witnessing genuine instances of free will, volition, or choice is attributable to the observable size and diversity of the state space (how life can respond and act in the world).

The freedom to choose is being alive and acting in the world

Not all mechanistic interpretations are free from ghosts. Prominent mechanistic accounts of volition have been oddly dualistic. Benjamin Libet’s 1983 finding that a readiness potential precedes the reported moment of intention in EEG experiments was interpreted for decades as evidence that the brain decides independent from conscious awareness. Aaron Schurger, Jacobo Sitt, and Stanislas Dehaene showed in 2012 that the EEG activity labeled a readiness potential was largely an artifact of averaging spontaneous fluctuations time-locked to action (movement) onset. That ghost was a statistical artifact. If our perceptual experience was indeed likened to a hapless, physically untethered observer — that goes wherever a separate mechanical engine commands us — that account demands a magical interpretation of experience, often hidden in vague and misused terms like “emergence” — implying an incidental pattern lacking causal power.

Consciousness, on the other hand, has two serious physical or mechanical proposals, but even these are less grounded than we might imagine them to be. The first, global workspace theory, in Bernard Baars’s formulation and in Dehaene and Jean-Pierre Changeux’s neuronal version, treats conscious access as the broadcast of locally computed information across a wide network so that memory, language, and motor systems can all address it at once. Whereas the second, integrated information theory, developed by Giulio Tononi, proposes consciousness (particularly the degree of consciousness) can be captured by a mathematical quantity, Φ, measuring how much a system specifies its own state above and beyond its parts, and identifies conscious experience with the structure of that integration. Both are attempts to say what kind of physical organization consciousness is, in terms that could in principle be measured. That is the right ambition, but head-to-head comparisons have been inconclusive. And it is no longer clear if any amount of data will provide sufficient evidence for either. We might return to Whitehead’s fallacy of misplaced concreteness to explain the futility. Recall this statement about our intellectual pursuits implies physical grounding alone does not make non-existent terms more real. It is instead our search for their physical basis that makes the terms seem real. Consciousness, as we understand and have labeled it today, could very well be the next phantom.

When the mechanism falls short

Even if we begin by replacing “emergence” with a physical mechanism, it is progress only if the mechanism answers to something outside the vocabulary that named the problem. We saw this with consciousness. Otherwise it becomes a cultural artifact; it is the eternal, essential, and mythical, no different from the mythical creatures of Ancient Greek and Roman mythology that still capture our imaginations to this day. Science should be different (refer to “Myth is the First Draft”).

Take the attention case. In one mechanistic interpretation, it is gamma synchrony: the idea that neural activity leads to particular rhythmic patterns (oscillations in particular frequency ranges) and these patterns can selectively synchronize, coordinate, or control neural activity across the brain. Imagine that attention is the synchronization of populations of neurons (attended) at the expense of others (unattended). However, even to the non-specialist, it is difficult to parse how attention is not some facet of consciousness. And the gamma mechanism has been contested: Supratim Ray and John Maunsell have argued that gamma frequency varies with stimulus properties in ways that make it a poor carrier for computation, and Björn Merker has argued that cortical gamma indexes activation rather than cognition.

Alan Allport asked in 1993 whether twenty-five years of attention research had been asking the wrong question; Britt Anderson published a paper in 2011 titled “There is no such thing as attention”; Bernhard Hommel and colleagues published “No one knows what attention is” in 2019. The field answered — there is a 2023 rejoinder titled “We know what attention is!” — and the dispute is live rather than settled. If “attention” is a folk category that does not exist, then finding a beautiful physical mechanism and attaching that label to it does not ground the term. The synchronization is real. But is attention?

The same argument applies to theories of consciousness. Scott Aaronson showed in 2014 that simple, obviously unconscious systems — expander graphs of XOR gates — can be assigned enormous Φ, which looks like a reductio until you read Tononi’s reply, which accepts the conclusion and holds that such a grid is indeed conscious, more so than a person. Adrien Doerig and colleagues have argued that any theory identifying consciousness with causal structure is empirically unfalsifiable in principle, since the same input–output behavior can be produced by a system with different causal structure. In 2023 more than a hundred researchers signed a preprint calling IIT pseudoscience, which drew a substantial counter-response — Anil Seth and Alex Gomez-Marin’s Nature Neuroscience reply among them — and a survey finding that only a small minority of consciousness researchers actually endorsed the charge. Most tellingly, the COGITATE adversarial collaboration, in which proponents of both theories preregistered their predictions before a large multi-site study, reported in Nature in 2025 that the results “substantially challenge key tenets of both theories.” Neither ghost was caught. Similar unease attends the free-energy principle, whose generality — the property that makes it attractive — has led philosophers of science such as Matteo Colombo and Cory Wright to question what could count as evidence against it.

These are the most serious attempts at grounding “emergent” in the neurosciences. But there is a specific pathology: taking a construct we inherited from folk psychology, finding some genuine mathematics in its vicinity, and treating the fit as a demonstration that the construct was a natural kind. Eric Jonas and Konrad Kording made the point in 2017 by applying standard neuroscience analyses to a microprocessor — a system whose ground truth is completely known — and finding that the methods yielded plausible, publishable, and entirely wrong accounts of how it worked. A mechanism that reproduces the phenomenon is not the same as a mechanism that explains the term. We arrive at an unsettling conclusion. To learn about and understand ourselves does not necessarily involve an inward search. As I have argued throughout, what actually is the “inside” of our nature? A history of neuroscience moves to reject dualism, the supernatural, and the metaphysical, only to adopt a seemingly more technical and physical language that gives modern scientists the freedom to introduce what was supposedly already rejected. Just as we are inclined to ask if attention exists, does neuroscience even exist as a science that addresses physically grounded and tractable problems, or does it make the ghosts seem more real?

The same mathematics, without a brain

What are ultimately real are the rules of interaction or the rules of relationality. I previously argued about the importance that substrates (material composition) have in altering the trajectories of physical systems in “Intelligence is a Relationship.” What can emerge, in other words, differs (call this difference “consciousness” or “intelligence” if desired). But the actual scientific work is largely in disclosing the rules that govern how physical systems interact. Where we have gone wrong, then, is our incessant focus on the appearance and features of the patterns that do emerge. That is not surprising. Although it was an abbreviated history, we did begin in the 17th century with the mind-body conundrum that Elisabeth wrote to Descartes about and later learned of its extension as “emergence” in the microscopic-macroscopic conundrum of the 20th century and finally found ourselves looking for spotlights in the brain and correlating gamma synchronization with “attention” and throughout this journey there were two distinct patterns: a human observer and an external world. Two discrete and essential things that perpetually escape us.

Physarum polycephalum is a plasmodial slime mold that has no nervous system and, being a syncytium, not even a conventional multicellular body; it finds the shortest path through a maze between two food sources: it explores, then retracts the dead ends. Toshiyuki Nakagaki, Hiroyasu Yamada, and Ágota Tóth published this in Nature in 2000. In 2010 Atsushi Tero and colleagues put food at the positions of the cities around Tokyo and let the mold grow; the resulting network matched the real rail system closely on cost and transport efficiency — though it was notably worse on fault tolerance. The mold demonstrates degrees of intelligence. It is strictly speaking a flow network in which tubes carrying more nutrient thicken and tubes carrying less wither, and this local rule leads to global optimization. In terms of the network-like structure and local, simple rule-following in these examples, the organism as as whole may be seen as transiently simulating the neural networks that would later evolve in multicellular organisms like ourselves. The networks became enclosed, which shielded them from environmental perturbations, enabling them to persist over time and grow in size and complexity.

Just as ancient human history anticipates the structure of modern day civilizations, both living as well as non-living systems reveal our evolutionary history. Even the simplest and unfamiliar cases anticipate the human brain.

Figure: Slime mold “choosing” the optimal path to discover food in a maze. Source: Marine Biological Laboratory / How Can a Slime Mold Solve a Maze?

Mix the right reagents in a dish and you get the Belousov–Zhabotinsky reaction: a chemical soup that refuses to settle into equilibrium and instead generates pulsating, rotating spiral waves, a macroscopic clock arising from local reaction and diffusion rates alone. Boris Belousov discovered it around 1950 and could not get it published — the story, told by Arthur Winfree, is of an editor who considered the result impossible and demanded he first disprove existing theory. Winfree’s own 1972 paper in Science on the spiral waves brought it into the mainstream. There are no biological components in the dish, and there is timekeeping in it.

Alan Turing, two years before his death, published “The Chemical Basis of Morphogenesis” in 1952, showing that two substances diffusing at different rates through a homogeneous medium can spontaneously break symmetry and generate stable spatial patterns — spots, stripes, and bands out of a uniform starting condition, from nothing but reaction and diffusion.

Figure: Belousov–Zhabotinsky reaction. Source: ResearchGate / Patterns in the Belousov – Zhabotinsky reaction

The activator–inhibitor vocabulary usually attached to this came later, from Alfred Gierer and Hans Meinhardt in 1972. It took another half-century to confirm the mechanism in living tissue: the ridges on the roof of a mouse’s mouth, where Andrew Economou and colleagues identified an FGF–Shh activator–inhibitor pair in 2012, and the stripes of the zebrafish, worked out by Shigeru Kondo’s group. The textbook line about zebra stripes remains an extrapolation — no one has identified the molecules, and the qualified version of this argument is stronger than the popular one. Where the mechanism has been established, what it establishes is remarkable enough: the animal does not carry a blueprint specifying where the stripes go. It carries parameters, and the pattern is generated.

Before neurons evolved and began communicating in networks through action potentials, there were cells signaling with waves of calcium ions, behaving like brains before brains. Scientists have shown that when an insect bites a leaf of Arabidopsis, glutamate release triggers a calcium wave that propagates through the plant and activates defenses in leaves far from the wound — Masatsugu Toyota and colleagues filmed this in Science in 2018.

Astrocytes, the non-neuronal cells that outnumber neurons in our brains, do not spike or communicate through action potentials like neurons; in fact, they communicate through slow calcium waves, which resemble the plant’s response patterns. How is this possible?

Michele Ballerini and colleagues provide some insight in a 2008 study of starlings. They showed that each bird only tracks around six or seven of its nearest neighbors, regardless of how far away they are. That simple rule is sufficient to produce the flock’s astonishing coherence or relationality.

Figure: Arabidopsis uses calcium signaling to communicate local damage through a network of plant structures. There are no neurons or astrocytes (glial cells) in the plant. We observe the actions of these brain cells, suggesting the observable, physical world manifests relationally. Not in things. Not as essences. Source: MDPI / Cell Type-Specific Imaging of Calcium Signaling in Arabidopsis

What we are observing in these examples is nature-as-relationship. The world, filled as it is with emergent patterns, is the history of parts becoming relational through countless interactions ultimately governed by physical laws and describable by mathematics. If these laws were not universal and scale invariant, convergence — the repeated, independent discovery of similar emergent patterns — would be impossible. Convergence, here, strongly suggests that science is not about the emergent patterns themselves — the history of countless “autobiographical” interactions, which I discuss in “Myth is the First Draft” — but the laws that determine what is and is not physically possible.

This is not rehashing the microscopic-macroscopic divide or a claim that individuals wholly disappear when describing the world relationally. The starling in the flock visually must detect other starlings to track them, preserving something uniquely starling-like in that case. The receptor or channel example — comprised of subunits A and B — was similarly not about other proteins, call them D and C. A was related to B because A uniquely fit with B and not with D or C. The issue — the individual account surfaces — is that we gain very little in focusing on A or B or the starling independently; the functional and meaningful description became what A and B or the starlings are doing and can do relationally. At first glance, this shift makes the scope extremely broad. What is the limit or boundary of this relationality? Everything is, if we ponder this further, related to everything else. So, we might ask, do you want to be a neuroscientist? Well, study everything. Fortunately, convergence and the universality of physical laws do provide us with bread crumbs. And we must follow them to plants and slime molds, chemicals in petri dishes. Only when we are fully embedded in this relational web, depriving ourselves of familiarity and comfort, will we be capable of finding what it means to be human — to perceive the world before us and truly describe it.

The last vital force

None of this necessarily refutes the thinking in Elisabeth’s correspondence with Descartes. There is perhaps something mysterious underpinning the observable, mechanical, physical, and mathematical world. We cannot say for sure. But we can also not expect science to reject or refute claims. It does not. It is a vehicle or process for proposing models of physical reality, “placing” those models of reality side-by-side — and this dissolves disputes; the victor does not emerge by clever, logical assault but by a community losing interest in the dispute over time because it had become surreal (dreamlike). The rival models, with their unicorns and mythical beasts, are exposed as stories inspired by true events but not quite actual enough to be real and useful anymore, as they are now frozen snapshots and the world they once captured has since drifted.

There are long-standing dialogues in the annals of medical research documenting the existence of four substances or humors that could mix in differing ratios and cause health and disease, or miasmas — malodorous emanations or foul smells — that could cause anyone perceiving them to become ill. They disappeared without argument as useless emergent patterns. They, admittedly, provided the best possible physical description or model at that time. But interactions occur. Relationality consumes. We just stop talking about those “ghosts” because the seemingly ghostly descriptions were genuine attempts at physical descriptions (refer to “Myth is the First Draft” for further detail).

Subsequently, the relational description I have outlined dissolves any questions based on perceived dualities (for example, mind-body). That there is something and then something else. Descartes had handed her a mind that was a thing — a substance requiring a point of contact. But the organizations that non-linear dynamics describes are not things needing to touch other things. The entire language is malformed. A whirlpool does not push water; it is a shape the pushing takes. Synchrony does not act on neurons; it is a relation among their firing. The physical world performs its relationships rather than storing them in objects, and once that is granted, the demand for a contact point between mind and matter dissolves as just another previous best attempt at a description.

There is, however, one extremely persistent form of mind-body dualism, carrying the tradition on. Even granting every mechanism, the objection goes, a complete physical account of a brain leaves something out: that there is something it is like to be. Thomas Nagel gave it its modern form in 1974, Joseph Levine later referred to it as the explanatory gap in 1983, and David Chalmers rebranded it the hard problem in 1995. But “what it is like” carries no physical commitment other than what these authors claim, and these conditions are not owed an explanation. They are owed a decision about whether they provide anything useful. Because these are all attempts at new physics or quite literally metaphysics (something beyond our current physical descriptions) and are therefore ultimately proposals about the physical grounding of dualism. At present, there is this physical description we know, but there is this something else. Clearly, this “something else” will become the new science, suggesting to many proponents that these are not dualistic claims after all but critiques of the physical description as it stands today.

From starlings to plants to human brains, however, we have found convergences resulting from physical laws. To determine if these claims are insightful — driving science toward something new — we need only compare human behavior over time. Protein A is to B as C is to D. The closest example to this moment in history — this something else — can be found in early questions about life.

Life was physical and mechanical, and yet there was this something else, then labeled élan vital. Philosopher Henri Bergson formalized the vital impetus in L’Évolution créatrice in 1907. In 2026, few recall that respectable intellectuals once considered life to be imbued with something mysterious.

While Friedrich Wöhler synthesized urea in 1828 from ammonium cyanate, demonstrating the connection between inorganic and organic chemistry — the chemistry of non-living and living systems — the vital force or essence of life remained a respectable position. Eduard Buchner’s cell-free fermentation in 1897 did not refute it either; Robert Kohler’s history shows the reception dissipated within particular professions. No one could ever measure the vital impetus and so no one could report it to be zero. It yielded no determinate testable results. Hans Driesch would eventually reformulate the vitalist claim as entelechy, which was an early attempt to describe self-organization during embryological development. This something else would be subsumed by complexity science, rendering “entelechy” a placeholder.

“What it is like” is in Bergson’s position. Every constraint on it is private. There is no measurement whose outcome can render it a placeholder, because it is specified from the outset so that none could, as beyond the current physics. Chalmers did see this objection and answered that experience is disanalogous to the vital spirit — that the élan vital was an explanatory position, discardable once the functions it was invoked for got explained, whereas experience is an explanandum in its own right and therefore not a candidate for elimination.

An explanandum earns its standing by being specified in terms the world can push back on. A phenomenon that would look identical under every possible finding has not been described; it is declared exempt from all possible obligation to make the description answerable. It is the eternal belief that there is something beyond history or beyond the current physical description.

This is not the claim that people lack experiences, or that the felt character of a smell or of grief is unimportant. It is a claim about what science is obliged to deliver, and to whom. The vocabulary that can never convert or change has no utility. What I refuse is the middle position — the one that uses the word “emergence” to mean “arising lawfully from interactions” when challenged, and to mean “appearing from nowhere above the physics” when convenient. That equivocation is where the ghost lives now in our prestige science. Moving from metaphysics to mathematics does not cost us the wonder. It is the only route by which the wonder can be lived, experienced, and shared.