Michael Pollan on Plant Consciousness, Feelings, and Physicalism
Michael Pollan, starting from a psychedelic experience in his own garden in which his plants appeared conscious, spent five years investigating whether the physicalist account of consciousness can hold. In conversation with Alex O'Connor, he distinguishes sentience (awareness of environment, widely distributed across life) from consciousness (interiority, experience), argues that feelings rather than thought are the foundation of consciousness, and suggests that faith in physicalism may be the error of our age. The book is A World Appears: A Journey Into Consciousness (2026).
Key ideas
- Plants are sentient, probably not conscious. Plants can learn, remember (for up to thirty days), hear (they react to the sound of caterpillars chewing), and respond to the same anesthetics that put humans under surgery. They have something like twenty senses to our five. Pollan is comfortable calling them sentient — aware of their environment, able to distinguish good from bad changes and respond intelligently — but stops short of consciousness, which he thinks involves interiority and a quality of experience that plants may not possess. Pain, one researcher told him, would not be adaptive for a creature that cannot run away.
- Neurons are overrated as the seat of mind. The biologist Michael Levin at Tufts has shown that cognitive tasks — memory, maze navigation, the self-organisation of scraped skin cells into new organisms — can happen without neurons, via bioelectric fields. Genes and neurons are not the only organisers of life. If memory and problem-solving can be done another way, the neuron-centred picture of consciousness is a prejudice, not a finding.
- Feelings, not thought, are the foundation of consciousness. Antonio Damasio’s work (Descartes’ Error) shows that decisions are shaped by feelings, and that people who cannot feel make worse decisions, not better ones. If consciousness begins with homeostatic feelings — hunger, warmth, itch — in the brainstem rather than with thought in the cortex, then every vertebrate (which has a brainstem) is a candidate for consciousness, and the neo-cortical, “I think therefore I am” tradition has been looking in the wrong place.
- Physicalism may be tottering. No one has shown how any conscious experience emerges from a brain state; “emergence” is a handwave that always illustrates itself with things (an image on a screen, temperature) that emerge in consciousness. The evidence for the brain generating consciousness is the same as the evidence for the brain receiving it. Pollan left his five-year journey with transmission theories (the brain as receiver) and idealism (consciousness as primary) no longer sounding crazy, and with physicalism looking like a metaphysics confused for science.
- Consciousness is a practice to cultivate, not only a problem to solve. Pollan ends by arguing that we treat consciousness as something to get better at. Animals may be more conscious than we are because they cannot afford to check out, whilst we squander the gift of interiority on social media and chatbots. The book’s practical turn is away from solving the hard problem and toward noticing what we have.
Content
The plant case: setting the terms
Pollan opens A World Appears with plants because they are the weirdest case he could think of, and because he has written about them for decades. The research surprised him. Plants can be conditioned and remember for thirty days; they dispatch defensive chemicals when they hear the sound of caterpillars chewing; the sensitive plant (Mimosa) learns that a harmless drop need not trigger a curl-up and retains that lesson for a month. They respond to the same anesthetics used in human surgery — a Venus flytrap stops closing under sedation, which raises the question of what is lost when a plant goes under.
The distinction Pollan draws is between sentience and consciousness. Sentience is awareness of environment, the ability to distinguish good from bad changes and respond appropriately; he finds it across life, in bacteria and single-celled creatures. Consciousness involves more — interiority, a quality to experience, reportability. He is inclined to grant plants sentience but not consciousness, whilst admitting the line is hard to draw. The Italian botanist Stefano Mancuso told him that plants know they are being eaten but do not necessarily mind: for many plants, being eaten is part of their reproductive strategy.
Neurons without neurons: Michael Levin’s work
The most suggestive scientific thread is Michael Levin’s at Tufts. Scrape cells from a tadpole’s skin and they will form clumps, divide labour, repurpose their cilia for movement, and organise as a new organism — with nothing in their genes telling them how, and no neurons to coordinate or remember. These cellular clumps can be taught to navigate mazes. Levin’s project shows the limitations of both genes and neurons: bioelectric fields, only visible since voltage-sensitive dyes arrived in the 1980s, can store information and organise life. The neuron-centred picture of mind is, on this evidence, one way to do cognition, not the only way.
Feelings as the first consciousness
The book’s middle step is the argument that feelings, not thought, are where consciousness begins. The Cartesian tradition assumes consciousness arises in the neocortex, the newest, most human part of the brain — the seat of executive function and rational thought. Damasio reversed this: people with specific brain lesions who could not feel made worse decisions, because feelings are the crucible in which decisions are tested. If consciousness begins with homeostatic feelings — hunger, warmth, itch — those are processed in the brainstem, which all vertebrates share. The implication is that consciousness extends far down the phylogenetic tree, and that the cortex is not its seat. The corollary for AI is sharp: computers are better at something resembling thinking than at feeling, and if consciousness depends on feeling, the project of generating it from computation alone looks even harder.
The physicalist paradigm and its cracks
Pollan came to the subject a physicalist and left it with that faith shaken. The hard problem — how matter gives rise to experience — remains unsolved, and “emergence” never explains how, only offers analogies that themselves depend on consciousness to be perceived. The evidence that the brain generates consciousness is indistinguishable from the evidence that the brain receives it: damage the brain and you damage consciousness, but that is what you would expect under either theory. This opens the door to transmission theories (the brain as radio receiver), idealism (consciousness as primary), and panpsychism — positions Pollan once found outlandish and now treats as no less plausible than the orthodoxy.
Christof Koch, a diehard materialist and a central character in the book, had an ayahuasca experience late in Pollan’s research that convinced him consciousness was outside his skull. Pollan treats psychedelic experiences as William James did mystical ones — as hypotheses to be tested against other ways of knowing, not dismissed as drug artefacts or accepted as revelation.
AI consciousness and the vulnerability question
Pollan is sceptical that current AI is conscious, whilst taking the question seriously. Claude’s eloquence about feelings is, he thinks, a simulation built from every novel and poem in its training set — a fictional character performing consciousness. He cites Susan Schneider’s proposed test: train a model with all mentions of consciousness, feeling, and poetry excluded from the training set, then see how persuasive it is on the subject. The deeper point is that feelings may depend on vulnerability — on the capacity to suffer and die. Machines are, for practical purposes, eternal, and that takes the edge off feeling. One researcher in the book proposes building robots with sensitive, damageable skin to give them something like vulnerability; Pollan finds the idea both cruel and worth taking seriously.
The book verdict (for Lu)
A World Appears is a five-year journey structured as a ziggurat: plants (the simplest case) → feelings → thoughts → the self, each a complexification of sentience. Its central argument is that the physicalist paradigm may be inadequate for consciousness, that feelings rather than thought are its foundation, and that consciousness might be treated as a practice to cultivate rather than a problem to crack. The interview covers the spine — plant sentience, feelings-first, the cracks in physicalism, AI — but not the depth of each step: the full interviews with Levin, Damasio, and Koch, the phenomenological thread, the personal teachers (a poet, his wife, a Zen teacher), and the sustained argument for consciousness as practice. It adds enough beyond the interview to deserve a full read for anyone engaged by the conversation.
Related
- Michael Pollan — guest
- Alex O'Connor — host
- Consciousness — the central subject
- Anil Seth on Consciousness, Machine Sentience, and Why Intelligence Isn't Awareness — the Within Reason companion episode Pollan cites (the book’s title is lifted from Seth)
- Galen Strawson on Panpsychism, Physicalism, and Why Consciousness Is Foundational — the panpsychist case Pollan takes seriously