The Longest Con.

There is a sentence in the taxonomy books that gets tossed off like a footnote, and it deserves better than that. Fungi are not plants. Fine. Everyone half remembers that from ninth grade biology, filed next to "a tomato is technically a fruit," a fact you keep to embarrass people at dinner and never actually use. But the sentence undersells itself. It is not that fungi merely aren't plants, the way a shark merely isn't a dolphin. It is that a little over a billion years ago something split into two branches, and one branch became every animal that has ever lived on this planet, and the other became every mushroom, mold, mildew, and yeast that has ever lived on this planet, and they have been on the same side of the ledger ever since. Plants broke off earlier, on a separate branch entirely. We are, structurally, the second half of the fungal split. A chanterelle is closer to you, genetically, than it is to the oak tree it is growing under.

Nobody sits with that fact long enough to let it do anything to them. It gets deployed as a party trick, did you know mushrooms are closer to animals than plants, and then it gets abandoned, the way you abandon a good fact once it has done its job of making someone go "huh." I want to sit with it. This essay started life as a joke I could not get rid of, a question a friend asked me sideways and I could not stop turning over: why did the mushrooms choose capitalism. It is a stupid question. It is also the only one that kept working, and by the end of this I am going to answer it seriously, which should tell you something about how the last billion years actually went.

Here is what is behind the door, in language a lab would sign off on. Fungal cell walls are built from chitin, the identical compound that makes up a lobster's shell and a beetle's exoskeleton, not cellulose, which is the plant material. Fungi store surplus energy as glycogen, the same molecule sitting in your liver right now, not starch, which is what a potato uses. And fungi eat the way you eat, by consuming organic matter that already exists somewhere else, rather than manufacturing their own food from sunlight the way a leaf does. For the better part of a century, taxonomists filed mushrooms in the same kingdom as trees on the sole grounds that neither one gets up and walks away, possibly the laziest classification criterion in the history of science, and one that took decades of actual protein sequence comparison to formally dismantle. In 1993 two biologists at Indiana University, Sandra Baldauf and Jeffrey Palmer, ran the numbers on twenty five different proteins shared across plants, animals, and fungi, hunting for shared genetic fingerprints, and found four separate insertions and deletions uniquely shared between animals and fungi, and none of them shared with plants. The paper's conclusion is right there in the title, blunt as a doctor's note: Animals and fungi are each other's closest relatives.

So: closest relative. Cousin. Sibling, if you want to get sentimental about it, though sentiment is exactly the thing I am trying to earn honestly rather than just claim. Everyone who has ever eaten an edible in a dorm room at two in the morning has arrived at some version of this insight independently and called it profound. That is not where the essay is. The question worth actually asking is not are we related. It is this: what has the relative been building, this whole time, while nobody was paying attention?

Because here is the thing about siblings who don't talk. It does not mean nothing is happening on the other end of the silence. It usually means the opposite.

The infrastructure nobody voted for

Start with a fact that reads like a metaphor and is not one. Under nearly four square miles of Oregon's Malheur National Forest, there is a single organism. Not a colony, not a population, one genetically identical individual, confirmed through DNA fingerprinting across the entire span of ground it occupies. It is a species of honey fungus, Armillaria ostoyae, and depending on which estimate you trust it is somewhere between about 2,400 and 8,650 years old. Its mass is genuinely disputed, with published extrapolations running anywhere from 7,500 tons to 35,000, which is the difference between a few loaded jumbo jets and a few dozen, and I mention the spread because the big number is the one that always gets quoted alone. It has no head. It has no eyes. It has never once, in thousands of years of continuous existence, had anything resembling a plan that a courtroom could prove. And it is, by every biological definition we use to determine where one organism ends and another begins, a single thing: the largest known living organism on the planet Earth, full stop, bigger than any whale, older than most nations.

You will almost never see it. The only visible evidence it leaves, ever, is a scattering of honey colored mushrooms that show up in autumn and a slow, spreading ring of dying trees that marks the border of where it has already won. Foresters have a term for those rings, "disease centers," which might be the most bureaucratically polite phrase in all of forestry considering what it is describing: the largest living thing on Earth quietly signing its name in dead pine.

Nobody watching that forest in real time, on any single afternoon across those thousands of years, would ever describe what is happening as an event. There is no headline in it. There is no moment where the news crews show up. It just keeps not stopping, decade after decade, century after century, until one day a team of Forest Service researchers runs a DNA test and realizes the several fungi they thought they were mapping are actually one continuous creature the size of a small city. That is lesson one, and I want to name it plainly before we go further: whatever the fungal kingdom is doing, it does not do it on a timescale that registers to us as doing anything at all. Slow, to something built like that, is not a limitation. Slow is the whole strategy.

The network they never asked permission to build

Below nearly every forest floor you have ever walked across without thinking about it, fungi are threading themselves directly into the roots of trees and, per the foundational 1997 Nature paper by the forest ecologist Suzanne Simard and her colleagues, moving carbon between trees of entirely different species through that shared fungal web. The nickname it picked up, admittedly cute, is the wood wide web. I want to be careful here in exactly the way I would want a source to be careful with me, because this is contested science and pretending otherwise would cost me the right to make any of the bigger claims later on. In 2023, a review in Nature Ecology and Evolution led by Justine Karst pushed back hard on the popularized version of Simard's work, arguing that the evidence for deliberate, coordinated tree communication, trees actively warning each other, trees making decisions together, is considerably thinner than the popular science coverage implies, and that the specific claim about mature trees preferentially feeding their own offspring has no peer reviewed support at all. Simard has publicly defended her data against that challenge, and the argument is still live.

The detail I cannot stop thinking about is that Melanie Jones is a co-author on both papers. She helped write the 1997 finding and she helped write the 2023 correction, which is roughly the most honest thing a scientist can do with twenty six years.

The peer reviewed, replicated core of the original finding, measurable carbon transfer between different tree species through a shared mycorrhizal network, holds up fine. It is the layer of narrative stacked on top of that core, the trees talking, the trees caring for each other, that is doing more emotional work than the raw data can currently support.

Here is what actually happened, stripped of the poetry: fungi built real, physical, verifiable infrastructure connecting organisms that have no other way of exchanging resources. That is not in dispute. What we did, and what I am about to keep doing for the rest of this essay with my eyes open about it, is take that infrastructure and narrate a mind onto it. We took a chemical gradient moving through soil and called it a conversation. We took a resource exchange network and called it a society.

That instinct is not new and it is not unique to fungi. I have written a whole other piece about a version of it, over in Flounder Juice: the way a round of golf between a retired ballplayer and a concessions executive, followed by six weeks in a Baltimore test kitchen developing a dry rub, became the founding myth of authentic local ballpark food for an entire industry, three decades running. Powell's own hometown magazine, in a tribute that adores him, calls the stand a "concept" concession without appearing to notice what it has admitted. Real substrate, dressed in a story so good it eventually becomes indistinguishable from the record. Fungi built the pipes. We wrote the copy.

Twice, the world ended, and something didn't notice

Every mass extinction leaves a signature in the rock. Usually that signature is defined by absence, a sudden violent quiet where there used to be species. Twice in the planet's history, the signature is instead defined by an overwhelming, almost obscene presence of exactly one thing.

At the close of the Permian period, roughly 252 million years ago, the worst extinction event this planet has ever recorded wiped out as much as ninety percent of marine species and something like seventy percent of terrestrial vertebrates. Sitting right at that boundary, in certain rock samples, is a layer where essentially all of the recoverable organic material comes from a single microfossil genus, Reduviasporonites. Paleontologists have argued for years about whether this so-called fungal spike is even fungus at all, or algae doing a passable impression of one in the fossil record, and I have to report that the argument is not over. Geochemical work has been read as favoring algae; specimens from boundary rock in the Italian Dolomites look like a pathogenic fungus; a 2016 review in Review of Palaeobotany and Palynology describes the biochemical evidence as ambiguous and the whole question as open. The isotope studies sharpened that fight. They did not end it. One paleontologist, asked to characterize what the data showed, called it "an exceptionally strange ecosystem," which is what a careful person says when the honest answer is that nobody knows yet.

Sixty six million years later it happened again, and this time nobody is arguing about the taxonomy. When the Chicxulub asteroid ended the Cretaceous and took the non avian dinosaurs with it, rock layers in Colorado, North Dakota, and New Zealand independently preserve the same moment: a spike where fungal spores suddenly account for more than half of all preserved spore material, up from a small background presence. In May of this year, Rosanna Baker and Arturo Casadevall at Johns Hopkins published a study in PNAS of the Denver Basin confirming the spike was not a regional fluke, and found something stranger underneath it. There was an earlier fungal surge, roughly thirty thousand to ten thousand years before the asteroid, lining up not with the impact but with the Deccan Traps volcanism in what is now India. Fungi bloomed while the planet was already dying. Then they bloomed again when it finished the job. For a window measured in centuries, a blink geologically, the dominant visible life form on this planet was not a plant, not an animal, not really a creature in any shape our intuition defaults to picturing. It was mycelium.

The mechanism, once you see it, is almost insultingly simple. Fungi do not photosynthesize. A sky choked with ash and dust that starves every green thing on Earth is not a crisis for something that never needed sunlight in the first place. It is a delivery. Every collapsed food web, every mass of dead trees, every carcass left unscavenged because the scavengers are gone too, is, from where a fungus is standing, inventory. Researchers describing the post impact world have compared it, without much attempt at poetry, to a massive compost heap. Casadevall goes further, and has argued for years that a fungal bloom on that scale may have helped tip the advantage toward mammals in the aftermath, because our warmer body temperature makes us harder for fungi to colonize than the cold blooded reptiles were, at precisely the moment the planet was, briefly, run by mold.

I owe this section the same caveat I owe the rest of the piece. Not every fungal spike claim in the literature is airtight. Some Permian boundary sections show no spike at all. The timing does not line up cleanly worldwide in every dataset. And one genomic study of agaricomycete fungi found no detectable signal of a population bottleneck around the Cretaceous boundary, meaning the spike may be a surface level bloom of opportunistic decomposer spores cashing in on a dead world, rather than evidence that the fungal kingdom as a whole was reshaped by the disaster. The honest version of this claim is not "fungi caused the extinctions," and it is definitely not "fungi were secretly in charge." It is narrower, and I think it is more useful for what I am trying to say. Twice, when almost everything else on the planet had its clock violently reset to zero, the fungal kingdom did not need to rebuild anything. It never stopped. It was simply already doing the one job, decomposer, recycler, the thing that turns death back into raw material, that never once goes out of business no matter what kills everybody else.

That is the real shape of the slow play I started this essay chasing, and it is a sturdier shape than mysticism. Fungi are, as far as I can tell, the only major kingdom of life whose entire survival strategy is a kind of profound indifference to catastrophe. They do not need the world to be in any particular state. They need the world to eventually stop moving long enough for them to eat it.

The part where the science stops and I start believing something

Here is the seam, the exact spot where documented fact runs out and the story I am choosing to trust picks up. I would rather tell you where that line is than pretend it is not there, because the moment I start writing like there is no line I have become the thing I do not want to be, somebody selling you a concept and calling it heritage.

Provable part first. Fungi are electrically active. Researchers who have inserted microelectrodes directly into growing mycelium have recorded trains of voltage spikes moving through the network, not a metaphor, an actual measurable electrical signal, comparable in shape to the action potentials that fire down a nerve, as the organism grows, transports nutrients, and responds to changes around it. Andrew Adamatzky, who runs the Unconventional Computing Laboratory at the University of the West of England, took this further than I expected any peer reviewed paper to go. He clustered those spike trains into discrete groupings he calls "words," ran a statistical linguistic complexity analysis on the result, and reported that the distribution of fungal word lengths matches the distribution of word lengths in human languages. I want to say that again because it deserves to land: not "resembles language" as a loose figure of speech I am reaching for, but a 2022 paper in Royal Society Open Science whose literal title is "Language of Fungi Derived From Their Electrical Spiking Activity." The same lab has since built functioning logic gates out of living mycelium, encoding binary strings into a fungal substrate and getting correct outputs back out: a slow, damp, self healing processor made of oyster mushrooms that is, when it eventually dies, entirely biodegradable.

None of that proves intent, and Adamatzky is more careful about it than the headlines calling it a mushroom computer ever manage to be. He hedges explicitly, assuming that the spikes are used to communicate, which is the assumption doing all the work. It is entirely possible the mechanism is nothing more than the electrical signature of a substrate doing substrate things, growth, transport, basic chemistry, that happens by coincidence to compress into something statistically similar to language, the way static on an old radio can sound almost like a voice if you sit with it long enough in a quiet room. That gap, between resembles and is, is the entire essay. Every section of this piece has been circling it from a different angle: infrastructure that resembles intention, a network that resembles conversation, a bloom that resembles patience.

But one line out of that body of work stopped me cold, because it is the closest a credentialed scientist has come to saying, without a shred of mysticism attached, the exact thing I have been circling since the first paragraph. Mohammad Mahdi Dehshibi, a researcher at the Universitat Oberta de Catalunya who collaborates with Adamatzky's lab, described a practical use case for these networks: fungal mycelium could function as "an environmental sensor on a large scale," a living system already wired, as part of its ordinary day to day biology, to process "large quantities of data flows," one we could in theory learn to read.

That is real. That is published. That is not me talking, and I should be straight that Dehshibi is a collaborator rather than an independent voice confirming Adamatzky from the outside; it is one research program, not two. But it sits close enough to the edge of what I actually believe that I want to take the last few steps on my own, without a citation to hide behind, because hiding behind one here would be a kind of cowardice. Something that has survived a billion years by outlasting rather than outrunning. Something that has already built an effectively unkillable global network for moving matter and information from one point to another. Something that has turned planetary catastrophe into raw inventory, twice, on the actual fossil record, without so much as flinching. That thing does not need to want anything, in the way we mean want when we say it about a person, in order to end up eventually as the best informed organism on the planet. It just has to keep doing exactly what it has always done, for one more extinction than we happen to get to see the end of. I do not have a footnote for that sentence. I have a billion years of one. I think that is enough to write it down.

The message that wasn't for us

I almost went somewhere else with this next part, and I want to tell you where, and why I turned around, because the turning around is the more honest version of the essay.

The tempting move, the one you have probably half heard somewhere on the internet already, dressed up in candlelight and pan flute, is to say that psilocybin mushrooms evolved their most famous compound as a message. A voice, finally, built out of chemistry instead of sound, aimed directly at the part of the human brain that processes serotonin, engineered across deep time to speak to us specifically. It is a good story. It is also, as far as I can tell, wrong, and wrong in a way that led me somewhere better.

Psilocybin acts primarily on the 5-HT2A serotonin receptor, and its psychedelic effects in humans are directly correlated with how strongly it binds there. Under its influence, brain imaging studies have shown measurable desynchronization across major brain networks, including a documented drop in connectivity within the default mode network, the system associated with our sense of a stable, continuous self. That part is not folklore. It is replicated neuroscience, and it is a real part of why the compound is now being seriously studied as a treatment for depression. So far, so mystical sounding.

Here is the turn. In 2018, Hannah Reynolds and Jason Slot at Ohio State traced the genes responsible for psilocybin production across dozens of unrelated fungal species and found that the trait kept jumping between lineages through horizontal gene transfer, genetic material moving sideways between species rather than down a family tree, specifically in environments packed with insects: rotten wood, dung, competitive nutrient poor soil. The leading explanation is not that psilocybin evolved to speak to a conscious mind. It is that it evolved to suppress the appetite of insects, which happen to run on the same serotonin based nervous system architecture we do, which is the actual, mundane reason the compound also does something to us. Slot put it about as bluntly as a scientist ever puts anything on the record: the fungus is altering the insects' mind, "if they have minds," to meet its own needs. A 2024 genomic paper in PNAS, led out of the Natural History Museum of Utah, traced the trait back roughly sixty five to sixty seven million years, long before any primate existed to eat a mushroom and see God.

Sit with that for a second, because I think it is a better ending than the mystical version, not a worse one. The closest thing evolution ever built to a chemical message from the fungal kingdom aimed at a nervous system was not addressed to us at all. It was addressed to a hungry beetle, sixty five million years before we showed up. Every account anyone has ever given you of a mushroom teaching them something, every ceremony, every trip report that reads like scripture, is a human being downstream of a pesticide, eavesdropping on a piece of correspondence that was never sent to us in the first place. We are the accidental recipient of somebody else's mail, and we built entire religious traditions out of the fact that we happened to be able to open the envelope.

I find that more haunting than the version where they built it for us on purpose, not less. A voice built specifically for you is a gift, or a trap, or a plan. It is a relationship, and relationships can be negotiated with. A voice that was never built for you at all, that you can nonetheless hear clearly enough to be permanently altered by, is something closer to weather. You do not bargain with weather. You just live inside whatever it decided to do that century, whether or not it ever once considered you while it was doing it.

What they actually built, instead of a voice

So here is where I land, and it is not the place I expected when I started asking why the mushrooms chose capitalism.

They never needed a mouth, because a mouth is a bottleneck. One signal, one direction, one speaker at a time, the way human language works, the way this essay works, one sentence following another because that is the only tool I have got. What they built instead was distribution. A billion years without vocal cords, and the workaround was not to grow a voice box. It was to become the substrate that everything else has no choice but to run through: the decomposition layer that turns every dead thing back into usable material; the underground logistics network that trees depend on whether or not you can prove, in a court of peer review, that the trees are talking to each other; the organism so patient, so willing to measure its plans in millennia instead of afternoons, that it does not even register to the species standing directly on top of it as an organism at all, let alone one with a plan.

That is not a voice. It is closer to owning the pipes the voice would have to travel through, in every direction, forever, regardless of who is speaking.

And that is the actual answer to the dumb, perfect question that started all this. Why did the mushrooms choose capitalism? Because they needed a voice, and it turns out the fastest way to get heard, if you have a billion years and no mouth, is not to shout. It is to become the ground everyone else is standing on when they finally decide to say something. You do not need to be loud if you are underneath everything. You just need to still be there after everyone who was loud has finished, decomposed, and been quietly filed away as inventory for the next thing that grows.

I do not know if that is true in any sense a lab could confirm, and I have tried across this whole piece to tell you exactly where the studies end and where I start. This is one of the places I start. But I have also spent a few thousand words showing you a closest living relative that survived two planetary apocalypses by doing nothing differently, that already runs something structurally close to a language through its own body, that a real scientist has already called a large scale environmental sensor, and that produces a molecule capable of quietly rewriting a human nervous system without ever having intended to reach one. If something built like that has been playing a long game, I do not think we would be able to tell, and I do not think it would need us to be able to tell, and I think that is the whole point of a long game in the first place. You do not win it by being noticed. You win it by still being there.

Sources and method

Editor's note, July 22: psilocybin has read this and filed a response. It has notes. Return to Sender →

The animal and fungal split, and the protein evidence for it, is Sandra Baldauf and Jeffrey Palmer, "Animals and fungi are each other's closest relatives: congruent evidence from multiple proteins," PNAS, 1993; the chitin, glycogen, and heterotrophy comparisons are standard mycology. The Malheur Armillaria ostoyae figures, including the acreage, the 2,400 to 8,650 year age range, and the disputed 7,500 to 35,000 ton mass estimates, are from US Forest Service surveys as summarized by the Oregon Encyclopedia. Carbon transfer between tree species is Suzanne Simard and colleagues, "Net transfer of carbon between tree species with shared ectomycorrhizal fungi," Nature, 1997; the critique is Justine Karst, Melanie Jones, and Jason Hoeksema, "Positive citation bias and overinterpreted results lead to misinformation on common mycorrhizal networks in forests," Nature Ecology and Evolution, 2023, with Simard's rebuttal in Frontiers in Forests and Global Change, 2024. The Reduviasporonites dispute, and the ambiguity of the biochemical evidence, is reviewed in Spina and colleagues, "Interpretation of 'fungal spikes' in Permian-Triassic boundary sections," Review of Palaeobotany and Palynology, 2016. The Cretaceous boundary work is Rosanna Baker and Arturo Casadevall, "Fungal proliferation before and after the Cretaceous-Paleogene mass extinction event in North America," PNAS, May 2026, which is also the source for the pre impact Deccan associated surge and for Casadevall's mammalian thermal argument. Fungal electrical spiking and the linguistic analysis are Andrew Adamatzky, "Language of Fungi Derived From Their Electrical Spiking Activity," Royal Society Open Science, 2022; the mycelium logic gates come from the same laboratory, and Mohammad Mahdi Dehshibi's environmental sensor remarks are from a 2021 Universitat Oberta de Catalunya release describing that collaboration. Psilocybin's horizontal transfer is Hannah Reynolds, Jason Slot and colleagues, "Horizontal gene cluster transfer increased hallucinogenic mushroom diversity," Evolution Letters, 2018, with Slot's quotation from the accompanying Ohio State release; the sixty five to sixty seven million year dating is Bradshaw and colleagues, PNAS, 2024. The ballpark comparison is drawn from Flounder Juice, elsewhere on this site, which carries its own sourcing. Where the literature disagrees, I have said so in the text rather than picking the tidier number.