Sally Adee is an author and freelance science writer living in London. Her book, “We Are Electric: Inside the 200-Year Hunt for Our Body’s Bioelectric Code, and What the Future Holds,” was published in 2023.
Photographs courtesy of the Eden Project.
From a snarl of roots that grip dry, shallow soil, the knobbly trunk of an ancient olive tree twisted into a surprisingly lush crown of dense, silvery-green leaves. Far above, the retrofuturistic pattern of a geodesic dome framed the blue sky outside. Dan Ryan considered the tree: “It’s probably close to 1,800 years old.” When it was still a shoot, the Roman Empire was at the height of its influence. Ptolemy was drawing epicycles in a doomed effort to model the paths of the planets and the sun as they revolved around the Earth. For nearly two millennia, this tree managed to evade death by drought or predation or pestilence, forging alliances with alien species in the soil below and the air above.
It was humans who almost did it in. When the tree stopped producing olives, the people who owned the land where it lived made brisk plans to demolish it. But in stepped the Eden Project, a nature preserve on the Cornish coast of the United Kingdom, offering the tree a kind of retirement home, as Eden does for many other plants. Here, people like Ryan make sure all its needs are attended to.
Ryan has deep knowledge of the plants in his care and a consequent respect for them. “The timescales of these plants puts any notion of intelligence into some pretty sharp focus,” he said. “Is that olive tree cleverer than us?” Given its lifespan, I was willing to think: Maybe?
Nonhuman intelligence has been the subject of a long-running and contentious war in science whose sides have periodically skirmished over the past 150 years. It was Charles Darwin who first popularized in the West the abilities in plants that in any human we would be comfortable describing as a display of intelligence.
But we don’t. Intelligence is still, for the most part, tightly defined as a human quality. The strict rules have relaxed somewhat in the past few decades, thanks to animal behavior scientists from primatologists to insectologists agitating to admit the objects of their study into the smart club. Crows can use tools, dolphins and bees use language, whales appear to communicate across hundreds of miles, octopuses are extraordinary escape artists — the list is getting longer every year.
“It would seem from recent research that the term ‘intelligence’ has too many shades and variations and synonyms, an undifferentiated sludge with little scientific utility.”
Plants, however, are still excluded by most scientists. The idea that they could be intelligent threatens to stretch the definition of some concepts that are fundamental to humans’ perception of themselves and the cosmos — cognition, understanding, reasoning, sentience — into meaninglessness. Plants, after all, don’t speak or move or react in ways most people would recognize as the actions of thinking beings with independent agency. And yet …
We seem to be entering a new era of cries du coeur to gather more life, including plants, under the umbrella of intelligence. Bookstores these days are heaving with volumes with titles like “The Revolutionary Genius of Plants,” “Planta Sapiens” and “The Light Eaters.”
Their authors are not even at the vanguard anymore. Some boldly go even further, finding behavior they label intelligent in fungi, bacteria, slime molds and paramecia. Even the cells that constitute our bodies are now standing at the velvet ropes, backed by frontier scientists waving evidence of behavior that might qualify as the hallmarks of intelligence if it were observed in an animal.
What on Earth is going on? Should we consider everything to be intelligent now?
There’s some evidence that the question is exactly backward. A small but growing number of philosophers, physicists and developmental biologists say that, instead of continually admitting new creatures into the category of intelligence, the new findings are evidence that there is something catastrophically wrong with the way we understand intelligence itself. And they believe that if we can bring ourselves to dramatically reconsider what we think we know about it, we will end up with a much better concept of how to restabilize the balance between human and nonhuman life amid an ecological omnicrisis that threatens to permanently alter the trajectory of every living thing on Earth.
A Potted History Of Intelligent Plants
It took a long time for anyone to notice what plants were getting up to. For millennia, Western intellectuals had dismissed plants as inert scenery at the bottom of what Aristotle called “scala naturae,” the “ladder of life” or “great chain of being.” But after observing the animal-like hunting movements of Venus flytraps and other carnivorous plants, Darwin began to doubt the dogma. He speculated that perhaps plants might be thought of as upside-down humans whose “brains” in their roots controlled the motion and activity of the limbs up above. Quite unlike the plaudits and societal disruption that had attended “On the Origin of Species,” however, the book in which he asserted this new idea was met mostly with embarrassed silence.
Sporadic reawakenings of scientific interest in the question were torpedoed in the 1970s, when the former CIA agent Cleve Backster published findings that he had hooked a house plant up to a lie detector, thought pointedly about setting it on fire, and then took the squiggles on the machine to mean it had telepathically read his mind. This nonsense muffled many intriguing findings over the following 30 years.
By the turn of the 21st century, however, a renegade group of plant physiologists had had enough. They argued that it was past time to bring existing theories of plant behavior into line with the avalanche of new observations enabled by late 20th-century advances in molecular biology, genomics, ecology and neurophysiology. Perhaps they weren’t reading anyone’s mind, but it sure started to look like plants had (some version of) their own.
Among many findings that precipitated the revolt and have proliferated since: Plants can sense — and with a bigger sensory suite than the one humans have. More importantly, they can integrate the information those senses carry and use it to make decisions. For example, the molecular biologist Edward Farmer and his colleagues at the University of Lausanne in Switzerland found in a 2000 study that Arabidopsis (the main model organism in plant physiology studies) markedly alters its hormone response depending on the size of a caterpillar munching on its leaves. When the attacker is small, the strategy is to keep them that way. “It’s better to be eaten by something small than by something big,” Farmer told me. And so, when attacked, “the leaf makes itself harder to eat” by producing toxic chemicals and proteins that interfere with digestion. This strategy slows the caterpillar’s growth and can delay pupation. One interpretation of this, according to Farmer, is that the plant can make a decision about how much energy to expend to repel pests based on the severity of the threat to its vital anatomy. Other plants have similar responses. Phaseolus lunatus (lima bean) mounts a particularly Machiavellian response: When a caterpillar starts snacking on it, it emits a chemical tailored to entice parasitic wasps, which swoop in like cavalry to pick off the predators. Ten years ago, researchers discovered that Boquila trifoliolata, a vine native to southern Chile, is somehow able to pass itself off as whichever species of plant is nearby, imitating its characteristic shape, color and pattern, possibly to entice pollinators or put off herbivores by assuming the guise of a less tasty snack. In one experiment, it even seemed to imitate a plastic houseplant.
Researchers also found that some plants show evidence of memory. A Venus flytrap counts the number of times an insect has triggered its sensory hairs — two in a row, not one, are usually required for it to clamp down on an interloper. To avoid a false alarm, it then seems to count to three (the trapped insect’s struggle?) before deploying costly digestive juices. Plants that track the sun seem to be able to retain their knowledge of when it will rise, even after a few days in the dark. Others appear to learn lessons from droughts, shrinking or completely closing the evaporation pores on their leaves, perhaps to prepare them to cope with continued dry conditions.
“In every way, a ‘human’ seems to be a collaboration between many organisms that have cooperated to form a superintelligence.”
It’s not just plants. Fungi, the objects of numerous investigations over the past several decades, exhibit goal-directed behavior — even, in the case of some parasitic species, manipulating the goals of their hosts. Like plants, they seem to sense: Their microscopic thread-like tendrils (hyphae) feel their way around the world, and they use sensory input to inform their next moves, such as whether to grow toward, around or away from whatever stimulus they touch. Writing in the magazine Psyche, the fungal biologist Nicholas Money saw in these behaviors the hallmarks of “spatial recognition, memory and intelligence.” In 2022, another researcher, Andrew Adamatzky, found evidence that enoki, split gill, ghost and caterpillar fungi may incorporate environmental information into stable internal concepts, like a human in a forest processing their surroundings into the words tree, rock and squirrel. The fungi may even recognize shapes.
And then came the slime molds — one group, Physarum polycephalum, in particular. Long misclassified as fungi, they have a gelatinous aspect due to the fact that they are constituted only of a single cell: a really, really big one. If two or more hungry molds meet, instead of clustering into a group of discrete individuals, they merge into a single genetically undifferentiated ooze. (In this way, one variant colloquially known as the tapioca slime mold attained the weight of a small child.) Investigators have been cataloging their surprising abilities for years now. The molds have a remarkably efficient ability to maximize reward for effort. In experiments where snacks were hidden in mazes, they identified not only the shortest routes to food but also solved complex mathematical problems like the famous Traveling Salesman Problem, which involved determining the quickest way to reach multiple destinations only once before returning to the starting point. They make judgments: Confronted by harmful or noxious stimuli, they accurately balance safety with efficiency, sometimes in offbeat ways reminiscent of our own human irrationalities and preferences. There is even evidence of memory and pattern recognition in their apparent ability to anticipate the administration of regular electric shocks. One observing researcher could only conclude that he was observing an “intelligent system.”
Sophisticated and seemingly intelligent actions can be found in the behavior of other single-celled organisms. Amoebas like Dictyostelium discoideum can perceive and act on positive stimuli like light, food and mates, and they can steer well clear of toxins and predators. They can even hunt in packs. So do some bacteria, which make complex decisions about when to fly solo and when to team up with others to form societies that reach billions of individuals, thereby becoming a biofilm, a good hedge against danger and starvation and even humans’ most sophisticated antibiotics. These societies may even be capable of a kind of associative learning, a Pavlovian response akin to dogs drooling at the sound of a dinner bell. E. coli colonies were found to be able to associate higher temperatures with a lack of oxygen, consequently altering their metabolism as temperatures dropped, apparently predicting an associated drop in oxygen levels.
Individuals in such a biofilm can also choose to cleave off from it, Albert Siryaporn told me. Siryaporn’s group at the University of California, Irvine, studies the physics of bacterial interactions. When threatened by viruses called bacteriophages spreading through their biofilm, bacteria can engage in strategic “social distancing,” as Siryaporn put it. When the researchers infected one part of the population, the uninfected part figured out how to avoid contact.
The developmental biologist Michael Levin, among others, is going deeper still, arguing that even the individual cells that comprise our bodies are capable of making decisions similar to bacteria. Levin has shown how they coordinate action during development to cleave off into organs and tissues. The same cells can also decide individually to stop cooperating with their collectives — if they turn cancerous, for example — prioritizing their own growth and food over the needs of their team. Such work is not even the radical departure it may appear to be: In her acceptance speech for the 1983 Nobel Prize, the plant biologist and cytogeneticist Barbara McClintock mused about the thoughtful action of cells. In fact, these ideas have been around for well over a hundred years. They are now popping up again everywhere you look, from work on Pavlovian learning in single cells to a 2024 study in which New York University researchers found evidence that one of “the canonical features of memory” — the fact that periodic repetition leads to better information retention than cramming does — is enlisted by kidney cells. The hits just keep coming.
Objection, Your Honor
Under this cascade of research suggesting that some of the smallest components of larger systems display intelligence, some scientists are growing weary. Lincoln Taiz, an emeritus plant physiology researcher at the University of California, Santa Cruz, and the author of one of the foundational textbooks of the field, has been on the diss track beat for at least a decade. Years ago, he told The New Yorker that plant neurobiology researchers suffered from “brain envy.” In the same story, Yale University microbiologist Clifford Slayman went further, deriding “plant intelligence” as “a foolish distraction” that was only establishing the line between “the scientific community and the nuthouse.”
In 2023, enlisted by Vox to comment on yet another apparently intelligent plant, Taiz reiterated that “the vast majority of mainstream plant scientists do not give the work of ‘plant neurobiologists’ much credence.” A year later, he and five co-authors categorically rejected the idea that fungus might have language as “too vague to be evaluated scientifically.” The “sentient cell” idea fared no better under his withering gaze: It was, he and other dissenters argued, “based on an elaborate series of speculations for which empirical evidence is lacking.” Instead, they wrote, the observed behavior was “evolutionarily genetically hardwired and has nothing to do with learning by individual organisms and cannot be taken as evidence for conscious or even cognitive behavior.”
The behaviors that seem so intelligent are in fact “tropisms,” more conservative scientists argue. Popularized by the German-American physiologist Jacques Loeb, the term describes the opposite of cognition or agency: an automatic, mechanical reaction to external or internal physical or chemical factors. Mechanists like Loeb declared that tropisms govern all plant behaviors, such as turning to grow in the direction of sunlight, and perhaps even some animal behaviors, such as a moth’s compulsion to fly toward light. A tropism precludes any internal will or individual agency. It is the opposite of intelligence and cognition.
For today’s critics, perception, communication, learning and memory in seemingly simple organisms are actually just innate, reflective responses performed by automata genetically programmed by hundreds of millions of years of evolution. If this happens to look intelligent to an outsider, that does not mean it is intelligent.
“If we can’t settle on a discrete and meaningful definition of cognition, what is a new prefix going to do? What good is the ‘minimal’ version of an undefined concept?”
Taiz offers a simple heuristic for the easily confused: True intelligence, not tropist imitations of it, requires a brain. This is a widely held view. “Intelligence requires mental representations of the external world that can be manipulated and that can be used to predict, explain and control the world, and I’m pretty sure those representations would not exist in a non-neural organism,” said Michael Anderson, who studies intelligence and cognition at the University of Cambridge. So: no brain, no mental representation, no intelligence.
But the argument creates a certain circularity. Later in his life, Loeb in fact thought that all animal behavior could be chalked up to tropisms — and if the controlling physical or chemical forces could be identified, he even envisioned a “mathematical theory of human conduct.” More than a century later, some researchers have argued that organisms with brains may engage in behavior that is simultaneously intelligent and a tropism. As Enrico Sandro Colizzi, a theoretical biologist who studies multicellularity at Cambridge, put it: “The process of mutation of genetic material actually works very well as a kind of machine learning tool. Evolution can learn.”
Adamatzky, who does not consider the behavior of the non-neural organisms he studies to be tropisms, nonetheless still said he was “against using phrases such as ‘slime mold intelligence’ or ‘fungal intelligence.’”
His quarrel was with the word itself: “We do not have a good definition of intelligence,” he said.
The Definition Of Intelligence
In a 2007 paper, Shane Legg, who later went on to co-found DeepMind, cataloged at least 70 definitions of the word intelligence. It would seem from such research that the term has too many shades and variations and synonyms, an undifferentiated sludge with little scientific utility. This has been an open secret in psychology and cognitive sciences for well over a century.
Controversy has long attended the scoring system Western countries use to measure intelligence. The IQ test assesses an individual’s ability to use memory and reason to manipulate mathematical and linguistic symbols in their head. The ability to do so has been correlated with good life outcomes. The resulting score — said to assess general intelligence, or g — is how we “determine” the “intelligence” of any individual.
No plant, fungus or bacterium can sit an IQ test. But to be honest, neither could you if the test was administered in a culture radically different from your own. “I would probably soundly fail an intelligence test devised by an 18th-century Sioux,” the social scientist Richard Nisbett once told me. IQ tests are culturally bound, meaning that they test the ability to represent the particular world an individual inhabits and manipulate that representation in a way that maximizes the ability to thrive in it.
What would we find if we could design a test appropriate for the culture plants inhabit? “Over half their biomass is underground in the soil, and we miss 100% of what they do there,” Paco Calvo, who directs the Minimal Intelligence Lab at the University of Murcia, in Spain, told me. “Their roots need to grow away from danger while scanning for pockets of nutrients and water — and they have to execute these goals all while solving the particular problems imposed on them by the totally different timescales they live in.” In other words, the test that measures a human’s ability to thrive in a specific environment is irrelevant to the question of whether plants or other organisms are intelligent in theirs.
Some have even started to question the utility of the IQ test as a measure of the entirety of the human mind. In an attempt to broaden the scope of the test in the late 1970s and early 80s, the cognitive psychologist Howard Gardner described a theory of what he called multiple intelligences. His original formulation included novel constructs like musical intelligence, bodily-kinesthetic and spatial intelligence, and inter- and intrapersonal intelligence. Over the years, he added more, including existential intelligence. This quickly became a new schism. You either believed in general intelligence or you believed in multiple intelligences.
Later, evidence of intelligence in non-mammalian animals like ants led to the study of so-called “swarm intelligence” to explain how many distributed organisms acting on tropisms could collectively transcend the brain power of any constituent individuals. “Collective intelligence” also began to appear in the literature to make the same point.
“Intelligence, according to some, is a biological function that evolved not with humans or brains but way back in some form to the earliest organisms, a fundamental biological function like respiration.”
For fungi and cells, scientists tried to propose new appendages like “minimal cognition,” “basal cognition” and, most recently, “proto-cognition.” These terms would account for instances of cognition in creatures that didn’t meet the most basic requirements of “intelligence” — a central nervous system, more than one constituent cell — while acknowledging that some of their behavior looked awfully smart.
Pamela Lyon has had enough of this. Rather than offering any illumination, she told me, slapping a bunch of modifiers and subcategories on the definitions of intelligence and cognition has only added more ingredients to a semantic word salad. None have shed any light on the core meaning of the word they all purport to modify. After all, she pointed out, if we can’t settle on a discrete and meaningful definition of cognition, what is a new prefix going to do? What good is the “minimal” version of an undefined concept?
Lyon, a researcher at the University of Adelaide in Australia, argued that these increasing complications are the equivalent of Ptolemy frantically refining epicycles in his hopeless bid to explain away the increasingly evident fact that the Earth was not at the center of the cosmos. Another paradigm shift, she said, is just around the corner. The very foundations of our ideas of intelligence are rotten, constructed with the same set of priors that placed Earth at the center of creation. The idea that intelligence revolves around the human has no scientific grounding.
Lyon told me that cognitivism — the current theoretical lens through which we seek to understand the mind — is nearing time for an exit interview. In Aeon, she enumerated the reasons why. “We still don’t have a good grip on the fundamentals of cognition: how the senses work together to construct a world; how and where memories are stored long term, whether and how they remain stable, and how retrieval changes them; how decisions are made, and bodily action marshaled; and how valence is assessed,” she wrote. These are just some of the problems that have been left on the table by Noam Chomsky, John B. Watson, Sigmund Freud, René Descartes and the many other thinkers who, over the years, fronted the philosophical inquiries into how mind emerges from matter.
So if cognitivism is out, what’s going to replace it? Lyon is petitioning for a return to Darwin: Intelligence, she has argued, is a biological function that evolved not with humans or brains but way back in some form to the earliest organisms, a fundamental biological function like respiration.
Biological Cognition
As Lyon sees it, cognition is like being pregnant. Nothing can be “minimally pregnant.” Creatures reproduce; reproduction is a function shared by all creatures and carried out at appropriate scales to each. But its purpose is unchanged across phyla — everything seeks to make more of itself. Different species’ approaches to reproduction are all, still, reproduction. And so it is, Lyon said, with cognition. “We’re going to have to think biologically about the way systems evolve rather than thinking in terms of our own categories,” she told me. In other words, as Anthony Trewavas, a botanist and molecular biologist who has also worked to shift the Overton window around how we evaluate intelligence across kingdoms, told me recently, our narrow, “scholastic” variety of intelligence needs to be seen as a specific manifestation of a much broader biological intelligence.
Studying cognition across all the species that enlist it could help us answer some of the questions previous approaches could not. Among the most interesting, for Lyon, is: What is a mind for?
Instead of arguing endlessly about how to use contested words to interpret various creatures’ behaviors, studying intelligence as a biological property is a radical new proposal that opens up a world of commonalities across species, phyla and, potentially, planets. According to Lyon and others on this frontier, such research could break us out of a centuries-old conceptual impasse and lead us to new insights about fundamental characteristics shared by all life.
Electrophysiological readings, for example, have for a long time revealed striking similarities in the activity of humans, plants, fungi, bacteria and other organisms. It’s uncontroversially accepted that electrical signals coordinate the physical and mental activities of brain cells. We have operationalized this knowledge. When we want to peer into the mental states produced by a human brain’s 86 billion or so neurons, we eavesdrop on their cell-to-cell electrical communication (called action potentials). We have been measuring electrical activity in the brain since the electroencephalogram was invented in 1924. Analyzing the synchronized waves produced by billions of electrical firings has allowed us to deduce whether a person is asleep, dreaming or, when awake, concentrating or unfocused.
Motor and sensory information is also encoded in the coordinated firing of specific groups of neurons. The match between specific sets of neurons firing and the effect — a leg kicking or an emotion triggered — is known as the neural code.
“Humans’ narrow variety of intelligence needs to be seen as a specific manifestation of a much broader biological form of intelligence.”
In the past few decades, it has become increasingly clear that similar electric signals mediate the actions and senses of all kinds of creatures without nervous systems. “Non-neural cells can be wired up too,” said Alison Hanson, a neuroscientist at the University of Iowa. “They’re found in bacteria, they’re found in plants, they’re found in fungi, they’re found anywhere. You put epithelial cells together, you get an electric network, just on a slower timescale. They’re not unique to human brains. They’re everywhere.”
Take plants. They may not have nervous systems but they do process and transmit information using methods that range, like ours, from hormonal to chemical — and electrical. A leaf has on the order of 30 million cells, each of which is studded with thousands of tiny electrical conduits called ion channels. This turns every plant cell into an electrical conductor. To instantiate their defenses, plants may employ fast electric signals whose rhythms look a lot like human action potentials. For example, tomato plants send them when their fruits are being eaten and release an antimicrobial chemical, possibly to guard against infection. But plants also make use of other types of electric signal — the variation potential, which may tell it about non-biotic attacks like fire, and the slower, more localized “system potential,” whose meaning is contested.
Fungi may also use electric signals to process the valence of stimuli from their environment. Researchers have measured oscillations in the electrical voltage of their constituent hyphae when they colonize their food. The function of these oscillations has not yet been as extensively probed or characterized as a plant’s. However, as Adamatzky told me, “Fungi respond to different stimuli with consistently different patterns of electrical spiking.” His recordings of electrical activity in four species of fungi suggested that these differences could be encoding representations of their external world. The “language” he had identified was not like our social chatter; it was akin to how analog electrical signals encode our own brains’ experience of the world around us — a kind of neural code. For example, the electrical signals may be a signal from a scouting tendril that alerts the rest of the body to the discovery of a good source of nutrition. Other studies of wood-decaying fungi, including oyster and honey mushrooms, have also found action potential-like responses to light, fire, salt and alcohol, among other stimuli. And some mushroom caps change their electrical activity after rainfall, possibly propagating their “knowledge” down to underground hyphae.
Slime molds, whose “bodies” are single cells without central command structures, still manage to pass electrical signals. Environmental stimuli cause synchronized rhythmic oscillations that appear to encode a memory of the original stimulus. Researchers theorize this is important to their ability to learn.
Scientists are divided on the implications of all this — some, like Siryaporn, think that the electrical signals in these creatures are just stimulus-response tropisms. Others believe they might be a reflection of the sensory information the creatures are picking up, potentially analogous to some kind of “concept” they are able to form of the world around them. Which brings us to the most striking idea — that some types of electrical oscillations could mediate an experience of self.
All Intelligence Is Collective
Arthur Prindle, a researcher at Northwestern University, has spent years unpicking the electrical signals of bacteria to understand how one becomes many. His favorite species is a friendly bacteria called Bacillus subtilis. Like slime molds, in times of stress — say it got stuck somewhere or a noxious chemical is nearby — Bacillus subtilis forms a biofilm. It’s protective, but it’s also a bit of a Ponzi scheme.
As the initial progenitors populate the society, they stay at the center of a slowly expanding mass. This provides insulation from outside dangers, which is one reason why, like all biofilms, it can resist a course of antibiotics (or other attacks); the medicine may kill many layers, but if it doesn’t reach all the way to the center, that inner circle will start producing more colonists to replace the dead.
There’s a downside to the cushy, protected life of a center-dweller. They may be far from danger, but they’re also far from food. The outer circle, normally preoccupied with the task of continually creating a new frontline barrier between themselves and danger, can be convinced to occasionally stop and divert their energies to pumping nutrients inward.
For a while, no one was able to understand how they coordinated these strange metabolic rest stops. But when Prindle and his colleagues examined the electrical patterns that accompanied them, they found rhythmic oscillations that looked similar to human brain waves. These preceded the work stoppage every time. It was as though the oscillations were a representation of the concept of “hungry.” Prindle told me he thinks disrupting the electric signals is a promising route to getting around antibiotic resistance. When he used a blocking drug to switch them off, the entire system broke down and an administered antibiotic was able to wipe out the whole colony.
In 2021, Hanson found that similar electrical activity — spontaneous low-frequency oscillations — is evident across many different organisms, from E. coli to humans. She concluded that across a diverse range of creatures, the oscillations may have a shared function: constructing a single organismal whole from many parts. In humans, such a pattern of oscillations is associated with what’s known as the default mode network. This “at-rest brain state,” whose function is still under heavy debate, is linked to a person’s subjective mental awareness. It’s active whenever your mind isn’t engaged in some specific task — when you’re daydreaming, recalling an autobiographical memory or just resting quietly. Previously, most research had focused on the electrical activity that was elicited by sensory, perceptual and cognitive activity. But even without stimulus or planned motion or any other input, there is a “background” signal that some consider the signature of a baseline awareness of being a self.
When Hanson looked through the literature on electrical activity in non-neural organisms, she found similarities in oscillatory electrical patterns across many multicellular collectives and organisms: not just bacterial colonies but honey fungus, oyster mushrooms, and some protists and plants. Hanson concluded that electrical signaling was allowing many parts to integrate information from the outside environment into the whole — a function she dubbed an “electrical organism organizer” — and that it was also a way to draw the group’s boundary between collective “self” and “not self.”
“Cognition is a relational property in between the organism and its environment. It’s not something that is sitting in your head or in your heart. It doesn’t reside within the organism.”
Prindle was skeptical of the notion that his bacteria had self-awareness, but he had an unbridled enthusiasm for the idea that synchronized electrical oscillations can function as an organism organizer. “I totally buy that. I love that,” he said. He has seen for himself that a biofilm responds as a whole when an antibiotic arrives, rather than as discrete individuals reacting as themselves.
Other researchers have noted similarities between the complex behavior emerging from, say, the networked connections between billions of neurons in a human brain and the synchronized behavior of swarms. Swarms of animals can encode information within the collective that is not necessarily legible or actionable to the individuals within it: Large groups of fish, for example, are able to detect light gradients across an area that no individual fish could perceive. Though a fish might prefer dark areas where it can better hide from predators, it often can’t locate the dark regions by itself. In groups, however, this ability — an emergent behavior — increases with size, and the group hightails it to the dark. Something similar seems to be at work in the formation and retrieval of memories by groups of neurons firing in synchrony, none of which would individually activate such patterns of action potentials themselves.
“The reality is that all intelligence is collective intelligence,” Levin told me. “It’s just a matter of scale.” Human intelligence, animal swarms, bacterial biofilms — even the cells that work in concert to compose the human anatomy. “Each of us consists of a huge number of cells working together to generate a coherent cognitive being with goals, preferences and memories that belong to the whole and not to its parts.”
He and Lyon both see cognition baked into the biological foundations of life. The “cognitive glue” that connects individual neurons into a brain or binds a bunch of cells into a coordinated human body, Levin said, is the same bioelectric pattern that allows plant cells to organize a joint defense against caterpillar attack. The plant’s activity is simply not cognition troubled by the philosophical baggage humans have attached to the term.
Electrical signaling may also help dissimilar organisms link up into a larger superorganism. The research is evolving, but there is compelling evidence that underground mycorrhizal connections link most plant roots into their symbiotic clench with fungi. Electrical currents appear to play a part in the formation of these interactions. And after they form, a hurt plant, for example, can transmit information related to its experience to a neighbor (whether of its own species or not) via their common underground fungal network or even aboveground by leaf-to-leaf touch.
Other intriguing experiments suggest that when two larger societies of bacterial biofilms encounter each other in the wild, their electrical oscillations begin to synchronize to alternate their feeding times. Cooperating in this way can be more beneficial for the biofilms than competing. As the evolutionary biologist Lynn Margulis wrote, life “did not take over the globe by combat, but by networking.” Networks themselves may be crucial to intelligence in ways we are just beginning to realize.
Maximal Cognition
The discovery of markers of intelligence in creatures without nervous systems is not the only thing shaking the foundations of the conviction that the brain is the seat of intelligence. Other findings from biology have also troubled the previously clean divide between human intelligence and the world from which it purports to stand discrete. Take the fact that humans are not made of only human cells. “We are not even individuals at all,” wrote the technologist and artist James Bridle in “Ways of Being,” a 2022 study of multiple intelligences. “Rather we are walking assemblages, riotous communities, multi-species multi-bodied beings inside and outside of our very cells.”
Bridle was referring to (among other things) the literal pounds of every human body that consists not of human cells but bacteria and fungi and other organisms, all of which play a profound role in shaping our so-called “human” intelligence. “The health or otherwise of our microbiome affects both brain development and our ability to cope with stress and trauma,” Bridle wrote. This influential microbiome was thought to populate mainly our gut, but new findings show that even areas previously thought sterile teem with nonhuman others: The brain microbiome is implicated in Alzheimer’s and Parkinson’s diseases, and there are now hunches that these diseases may represent the dysregulation of an otherwise commensal population. In every way, a “human” seems to be a collaboration between many organisms that have cooperated to form a superintelligence.
If we can let go of the idea that the only locus of intelligence is the human brain, then we can start to conceive of ways intelligence manifests elsewhere in biology. Call it biological cognition or biological intelligence — it seems to manifest in the relationships between individuals more than in individuals themselves. “Cognition is a relational property in between the organism and its environment,” Calvo told me. “It’s not something that is sitting in your head or in your heart. It doesn’t reside within the organism. Organisms don’t exist in a void — they are always in an environment and acting with each other.”
Thinkers like Bridle and Michael Muthukrishna, a professor at the London School of Economics, would agree. In humans, Muthukrishna wrote in “A Theory of Everyone,” the complexity of the ideas any individual can have is a function of the ideas they are connected to and the larger cultural “software” imprinted into that network.
As an example, he explained how mean IQ scores rose steadily through the 20th century: As more people entered more standardized classrooms, they became acculturated to the symbol-manipulation ability that was taught in those schools, which was captured by the IQ evaluations. “IQ tests are useful as a measure of cultural competence,” he wrote. “The broad structures of the collective brain affect the smarts of its constituent cultural brains.”
“What is the reason a group of individuals can exponentially multiply their abilities into something beyond what would be expected from the linear addition of their strengths? In some ways, it’s the most critical question of our time.”
In some circumstances, a particular group becomes so much smarter than the sum of its parts — more effective at achieving complex goals that no individual could reach — that only mathematics can explain it. This phenomenon is called “synergy” and the networked informational flows that underpin collective intelligence are its focus: ant colonies, schools of fish, bacteria. Groups of humans can experience it, but we seem to be especially bad at harnessing cooperation except by accident or in very specific contexts, such as team sports.
The individual members of a team behave in some fundamental ways like a swarm, and the mathematical principles underlying the relationships can be identified in the study of complex systems. “How do teammates coordinate their movement to increase the likelihood that a behind-the-back pass is caught by a teammate rather than stolen?” wrote Jessica Flack, a professor at the Santa Fe Institute. What is the reason a group of individuals can exponentially multiply their abilities into something beyond what would be expected from the linear addition of their strengths? In some ways, it’s the most critical question of our time.
Everywhere across nature, the message is the same: Species are smarter when they team up.
The next question is: How far can this scale? Is it possible to form a collective macro-organism with other kinds of intelligence? This is part of the vision Levin has begun to articulate in his work. “Given the ability of human subunits to merge into even larger (social) structures, how do we construct higher-order Selves that promote flourishing for all?” he wrote in the journal Frontiers. “The goal of this research program beyond biology is the search for optimal binding policies … [to enable] a scaled-up Self at the level of groups and entire societies.”
The Planetary
One aspect of human intelligence — all intelligence, actually — is its fluidity, its plasticity. As Muthukrishna has noted, the collective brain can marshall a lot of information to change the way individuals in that collective do things.
Expanding our understanding of intelligence beyond our self-referential framework would awaken us to ways of having relationships with other minds. The benefits of doing so are a lot more practical than they may appear.
For example, it could give us a surprising way to improve agriculture. Mildew is a major reason we dump so much toxic pesticide on our crops — by the time human eyes sense the first signs of infestation, the crop is already lost. The only solution is frequent prophylactic spraying, which is bad for the soil and the water the toxins leach into.
For the past few years, a Swiss company called Vivent has been working on a different strategy: find patterns in plants’ variation, action and system potentials that reveal how mildew infestation manifests in their internal state. Just as we use EEG on human brains to diagnose health and states of mind, Vivent’s electrodes can detect signals that reveal that the plants are feeling cold, say, or irritated by a pathogen. Nigel Wallbridge, the company’s chief scientist, said the plants’ electrical patterns convey clear signatures of mildew infestation far earlier than other detection methods. Identifying the problem so early opens the door to new treatments that eschew toxic sprays. Several big agriculture companies are now trialing the sensors, including Bayer.
Fungi that infest humans may also be amenable to electrical manipulation. “I believe we can understand fungal minds better by decoding their electrical activity,” Adamatzky said. He suspects the distribution of electromagnetic fields on the skin influences whether and where fungal colonies form. Wearable devices that alter those electrical relationships could lead to better ways to prevent pathogenic fungi from getting an unpleasant foothold on human biology.
“Everywhere across nature, the message is the same: Species are smarter when they team up.”
A similar managerial role may be relevant for bacterial control: the skin-dwelling opportunistic bacterium Staphylococcus epidermidis — a culprit in many common infections — changes its electrical excitability as it makes itself at home on healthy skin. Using mild electrical stimulation, researchers were recently able to manipulate the bacteria’s electric signaling apparatus, which suppressed the cells’ ability to grow and form a collective biofilm.
Machine learning and large language models — another form of intelligence — are critical to identifying relevant patterns in the din of biological electrical signals and the quest to bridge human and nonhuman intelligences. The development of AI could also help cause a conceptual awakening of humans to different manifestations of “more than human” intelligences, from plants to protists. Interestingly, much of the heated rhetoric around AI is echoed in the discourse around plant and other nonhuman intelligences.
“The boundaries between humans and nature and humans and machines are at the very least in suspense,” wrote the philosopher Tobias Rees. Moving away from human exceptionalism, he argued, would help “to transform politics from something that is only concerned with human affairs to something that is truly planetary,” ushering in a shift from the age of the human to ‘the age of planetary reason.’”
Interviewing Levin on his podcast, the computer scientist Lex Fridman mused: “We really don’t want to see human civilization or Earth itself as one living organism; that’s very uncomfortable to us.”
“We have to grow up past that,” Levin replied.
During the time the olive tree at the Eden Project has been alive, two fundamental scientific paradigms were laid to waste — Ptolemy’s Earth-centric view of the universe was swept aside by the revelation that our planet orbits the sun, and human exceptionalism among other animals was devastated by Darwin’s discovery of the mechanism of natural selection that would eventually link us back to the first bacteria.
Exiting the olive tree’s dome, Ryan and I were joined by the botanist Jo Elworthy and Tim Pettitt, a plant pathologist and microbiologist. Both have been with Eden from the start. In 1998, this lush green bowl was a bleak, lunar landscape, ecologically devastated over years as a china clay pit. Nothing grew because there was no soil to grow in. Elworthy described standing at the bottom of the pit with the man who had just purchased what seemed likely to be an eternal wasteland. “He told me, ‘I’m going to turn this into a global garden. I am going to change this from a scarred landscape.’ We wanted to create this circular system of people working with nature and looking at all the different interconnected parts to demonstrate that you can leave the world better than you found it.”
Nearly three decades on, every inch of the valley is covered in a riot of greenery. Under the vast protective domes, two gardens flourish: a wet rainforest and a dry Mediterranean biome. Pettitt told me about the soil mixture he invented to nurture the plants that flourish here, a homebrew that contains, among more standard ingredients, a blend of fungi, wormery leachate, bacteria and protists. Once, when the protists got out of hand and began to assume a parasitic stance, instead of turning to pesticides, Pettitt treated the plants to a drench of lactobacillus, which gobbled them up until the system came back into balance.
Late last year, the Eden Project got funding to expand to a second site, a former British fairground that became a lifeless concrete block when the glory of Victorian seaside resorts faded into memory with cheap flights to Spain. Here, project staff will work with the coastal ecology to restore the site. Step by small step, the goal is to put back into balance what’s been lost.
The goal is also to show in concrete terms that, even at this late hour, it is still possible to pull ourselves back from the brink of ecological crisis. The argument about what does and does not have intelligence isn’t a detached, academic debate between philosophers concerned with defining abstract concepts. Ecological restoration is, in many ways, an exercise in building symbiotic interdependence between species, which is non-negotiable for human survival over the long term. Just as intelligence seems to exist in relationships between cognitive selves, survival is impossible if humans think they can do it alone.
Elworthy told us that she likes to ask the people who come to visit where they’re from. They’ll usually answer with their neighborhood in London or foreign country of origin, to which she responds: “No. You live on this ball — this spaceship — that’s hurtling around the sun at 67,000 miles an hour. It gives you all your air and food and cleans your water.” The point she was trying to make was that we can be more cooperative, more embedded in the biogeochemical systems that sustain all living things — to act like crew, not passengers, on Spaceship Earth.
“And somebody said to me, ‘Oh, that’s such a good metaphor!’” Elworthy said. “For god’s sake!” She smacked her palm to her face. “It’s not a metaphor!”
