58Why I Am Not a Buddhist

Evan Thompson, Why I Am Not a Buddhist, Yale University Press, 2020

Linked to 20 chapters: supports 11 · contradicts 13 · mechanism 4 · same claim 2 · pairing 1

Thompson's target is not Buddhism but the dominant modern form of it, which historians call Buddhist modernism: the transnational Buddhism that plays down metaphysics and ritual, plays up meditative experience and scientific rationality, and presents itself as Buddhism's original core although it dates from the nineteenth century. At its centre sits what he names Buddhist exceptionalism, the belief that Buddhism is superior to other religions in being inherently rational and empirical, or that it is not a religion at all but a mind science based on meditation. That belief is false, the assumptions about religion and science holding it up are false, and its scientistic descendant — neural Buddhism, which holds that neuroscience has confirmed there is no self, that meditation rewires the brain, and that enlightenment has neural correlates — is not merely wrong but confused. The book's critical half takes exceptionalism apart chapter by chapter; its positive half argues for cosmopolitanism, the idea that all human beings belong to one community that can hold different ways of life. The title echoes Russell's 1927 lecture "Why I Am Not a Christian", with one difference Thompson insists on: he is not arguing against religion, and his feelings toward Buddhism are not hostile.

The book opens with how he got there. In 1974, aged eleven, he came home from a family trip to Scotland to find the commune he lived in taken over by Zen. His parents had founded the Lindisfarne Association in Southampton, New York, and his father had asked the abbot of the San Francisco Zen Center to send a teacher; Reb Anderson had arrived while they were away. Shoes came off at the door of the lodge, meals turned silent, and the children ran about imitating the pious expressions of the adults. Thompson had nothing against Buddhism — his father gave him a book about the Buddha at eight and a small Buddha statue at twelve, which he still keeps — and he followed the translator Robert Thurman to Amherst in 1979, at sixteen, to read Nāgārjuna and Tsong Khapa beside Kant. Then came Francisco Varela, whom he met at Lindisfarne, and with whom, along with Eleanor Rosch, he wrote The Embodied Mind in 1991, the first academic work to bring Buddhist philosophy to cognitive science.

What never came was a practice he could join. He visited meditation centres for years and connected with none, and in those same decades one prominent male Buddhist teacher after another was found to have abused students. After Varela's death in 2001 he worked with the Mind and Life Institute and watched a research community he had helped build harden into an in-group where sceptical questions were pushed aside. On a retreat at Barre, Massachusetts, designed for scientists, he sat ten hours a day being told he was learning to see things as they are, while thinking that everyone present was being handed a collectively reinforced conceptual system for making sense of whatever happened in the silence. Since he would not enter a monastery, the only way for him to be a Buddhist was to be a Buddhist modernist; and since Buddhist modernism is philosophically unsound, that would be bad faith.

The myth of Buddhist exceptionalism

The chapter opens with a test. Imagine books called Christian Biology, Islamic Biology or Why Hinduism Is True. No such books exist, but Buddhist Biology, by the evolutionary biologist David Barash, and Why Buddhism Is True, by the journalist Robert Wright, both do, and the second was a New York Times bestseller in 2017. Even the sharpest scientific critics of religion make the exception: Richard Dawkins, endorsing Barash, writes that Buddhism is surely "religion's best shot".

Thompson names the belief by analogy with American exceptionalism, then dates it. In the nineteenth century, Christian missionaries in Asia and European colonisers argued that their superiority was shown by their possession of science. Buddhist reformers, especially in Burma and Ceylon, turned the argument around: Buddhism is the truly scientific religion, since it has no creator God, relies on reason rather than faith, and treats its founder as a man. Each claim requires ignoring something — an array of celestial deities, many objects of devotion, a Buddha traditionally held to be supramundane — but the move worked, and the hybrid forms were exported west and imported back into Asia.

When Yamada Ryōun, abbot of the lay organisation Sanbō-Zen, says that religion requires faith in a transcendent being whereas Zen is experientially finding one's true self, the argument works only on a modern conception of religion built from the outlying case of Protestant Christianity. Faith in a transcendent being is not a universal feature of religion; Zen has ritual, scripture, liturgy and priests; and "one's true self" is a soteriological notion. Take religions instead as social forms of meaning-making and every form of Buddhism qualifies, secular Buddhism included.

The deeper claim is that Buddhism is exceptional in being a mind science, and Thompson traces its method to one man. Nyanaponika Thera, a German-born monk from Ceylon, identified the Buddha's mind-doctrine with a science of the mind in 1954, and coined "bare attention" to gloss sati, the word Thomas William Rhys Davids had chosen to translate as "mindfulness", whose root sense is memory. Bare attention, Nyanaponika wrote, tallies with the procedure and attitude of the true scientist. He also wrote that it slows the transition from thought to action and increases the plasticity of the mind. But to disclose something is not to change it, and to shape the mind is to change it. A microscope does not alter the cell, because it is separate from what it reveals; bare attention is a mental process that changes other mental processes. Is it a light or a mould?

The answer cannot come from the experience of bare attention. It must come from Buddhist philosophy, which is normative, making value judgements about how a mind should be cultivated, and soteriological, concerned with liberation; the doctrine gives the meditation its meaning rather than the other way round. Nirvāṇa and bodhi are not constructs whose validity measurement can settle, any more than beauty is. Nor is the historical Buddha available as an authority: S. N. Goenka called him a spiritual super-scientist who never founded a religion, but liberal Christians write the same sentences about Jesus, and the Buddha the appeal invokes was built by nineteenth-century Orientalists — a figure, in one historian's phrase, forged in a European philological workshop. B. Alan Wallace holds that critiques from anyone but professional contemplatives deserve no weight; imagine a psychoanalyst saying it. Meditation is a skill and is testable the way skills are. What it does not do is test hypotheses.

Is Buddhism true?

Wright argues that science corroborates the core ideas of naturalistic Buddhism. Natural selection built us to anticipate pleasure strongly and feel it briefly, because that keeps us reproducing; craving is the result; there is no CEO self, only competing mental modules that seize control through feelings; and mindfulness weakens craving's grip, which makes the Buddhist path a rebellion against natural selection's values. Wright gives the clearest statement of a widespread position, and, unlike the exceptionalists, he grants that even secular Buddhism is a religion. Thompson asks four questions and answers all four no: whether evolutionary psychology is the right science of mind, whether it is the right way to relate science and Buddhism, whether naturalistic Buddhism is compelling, and whether the title question is the right one.

He means evolutionary psychology in its narrow sense, a programme many biologists and philosophers of science regard as faulty. It treats problems as existing independently of organisms, where niche construction theory shows that organisms modify their environments and so bias the pressures they generate. It treats the Pleistocene as the source of what matters, where cultural transmission is at the centre: there is no innate reading module, yet a literate brain looks different from an illiterate one, not only while reading but while hearing speech. And it requires the brain to be nearly decomposable, where the evidence points the other way — Broca's area, long tied to language, turns up in movement preparation, action sequencing and imitation.

The better science is the embodied and enactive kind, and the difference from Wright is one of direction. Wright takes scientific naturalism for granted and uses it to justify modern mindfulness practice; he never lets Buddhism question science. Thompson's image is circulation, each tradition flowing into the other and back. Vasubandhu's Abhidharma analyses everything into relations among elementary processes, so any easy identification of mind with brain fails at the outset. Nāgārjuna's Madhyamaka adds emptiness — not that things do not exist, but that nothing has a nature from its own side, since what something is depends on a conceptualising mind, a basis for applying the concept, and a term in a language.

Wright uses emptiness only on the perceiving side: the essences we seem to perceive are an illusion engineered by natural selection, while genes and organisms are objective and science tells us how the world really is. Emptiness, Thompson replies, is a claim about everything, science included. If any nature we can point to is concept-dependent, science shows us reality in relation to our concepts and methods, not apart from us — which changes the standard from correspondence with a mind-independent world to empirical adequacy for observers like us.

Naturalising also costs the tradition its hardest idea. Nirvāṇa is the unconditioned, and the unconditioned is by definition uncaused, so it cannot be the effect of following the path — and yet the Buddha is held to have attained it. Robert Sharf's formulation is the one Thompson keeps: "Liberation is impossible, yet it is achieved." In Wright's hands the unconditioned becomes the deconditioned, feelings that no longer trigger craving, and the question of whether there can be a transformation so radical that from outside it looks transcendent disappears. Nor is mindfulness a rebellion against natural selection's values: our cravings probably owe more to consumerism than to the Pleistocene, and natural selection has no values, being either a force changing gene frequencies or a statistical trend. Only people have values.

No self? Not so fast

Here Thompson changes a position he once held in print. The Embodied Mind argued that cognitive science supports the Buddhist no-self view, and that Buddhist practice supplies a way of not clinging to something one knows does not exist. He now thinks that only part of the story, on two counts. It is facile to hold the Buddhist view superior to the Brahminical views it argued against for centuries, since those opponents raised objections that anticipate what cognitive science calls the binding problem. And cognitive science does not show that the self is an illusion. It indicates that the self is a construction, which is a different thing.

The self Buddhism targets is the object of self-grasping: a personal essence, that which is essentially what I am, such that if it is present I am present and if it is absent I am absent. As a boy Thompson played street hockey imagining he was Bobby Orr and played chess imagining he had Bobby Fischer's mind — not imagining being identical with either man, but imagining himself having their body and their abilities. The experiment asks you to imagine a different body or mind while still being you, and shows how readily we take the self to transcend both.

The Pali sources deny exactly that. In the "Discourse on the Characteristic of Nonself" the Buddha says of each of the five aggregates that make up a person — body, feeling, perception, volition, consciousness — "this is not mine, this I am not, this is not my self", for two reasons: the aggregates are not subject to our control, and they are impermanent. The criteria implied, permanence and control, are the Vedic-Brahminical criteria of the Upaniṣads, so the arguments establish that no aggregate is a self on those criteria rather than that there is no self. It was understood early as meaning there are only the aggregates, which we call a being out of convenience: the nun Vajirā answers Māra with the chariot, a word used of an assemblage of parts, and "Questions of Milinda" has Nāgasena press the same analogy on the Bactrian-Greek king Menander.

Here the Nyāya philosophers Vātsyāyana and Uddyotakara apply pressure, and Thompson thinks they found something. A chariot is inanimate; a person has an inner life and a unified perspective. If all we have is impersonal elements related as cause and effect, what makes some of those relations mine and others yours? When I say hello I cause you to hear the words and I cause myself to hear them: one cause, two effects, one in the series we call me and one in the series we call you. Nothing in the causal relations alone marks the difference. A principle of identity is needed, and the same gap opens with perceptual binding: the yellow must bind to the mango and not the bowl.

Thompson's verdict is split. With the Buddhists he agrees there is no substantial self. With the Nyāya philosophers he agrees some principle of unity is required, and denies it must be a substance. The Brahminical self-theorists were no less rational and empirical than their opponents, so singling out the Buddhists as the scientific ones is partisan.

Then the modern versions. Thomas Metzinger, who once described his own position to Thompson as analytical neuro-Buddhism, argues that a self would have to be a substance, single at a time and over time, distinct from any other; science finds no such thing; therefore nobody ever was or had a self. What exists is the content of the brain's self-model, taken for a real self because we cannot see the model as a model. Miri Albahari argues similarly that repeated identification with transitory experiences creates the appearance of an "I" purporting to be unconstructed, and so an illusion. Thompson's objection concerns a premise, not a conclusion. Each defines the self as a personal essence in an independent thing, unconstructed by the experiences it owns. Since William James's Principles of Psychology of 1890 and George Herbert Mead's Mind, Self, and Society of 1934, the self has instead been treated as a developmental and social construction. The measured conclusion is that a certain kind of self does not exist, not that there is no self.

The positive account refuses the split between the self and the sense of self, and defines the self in terms of self-awareness, the structure whereby one experiences oneself as oneself. Prereflective self-awareness is implicit: watching a sunset you are aware not only of the colours but of your seeing, which you live through rather than look at. Reflective self-awareness, which psychologists call meta-awareness, is explicit, as when you notice your mind has wandered; bodily and social self-awareness are the other two. Ulrich Neisser's five kinds of self-knowledge map onto these closely.

None of these entails a personal essence, and this is where the distinction does its work. Illusions are mental constructions, but not every mental construction is an illusion, so saying the sense of self is under constant construction does not imply there is no self. Buddhist philosophy marks the point by reserving ātman for the personal essence and pudgala, person, for the construction; neural Buddhists ignore the distinction, so their announcement that the brain generates an illusion of self reads as a denial of the person. A self, on Thompson's account, is an ongoing process that enacts an "I", in which the "I" is no different from the process, as dancing is a process that enacts a dance in which the dance is no different from the dancing. Nobody looks for a dance inside a dancer's muscles. Being dependent and relational does not make it unreal.

Mindfulness mania

Mindful living, mindful parenting, mindful leadership, mindful colouring books; Wired announcing that in Silicon Valley meditation could make your career. Many Buddhists object that mindfulness is not an ethically neutral stress-reduction technique but the seventh limb of the Eightfold Path, and that selling it as a corporate commodity is McMindfulness. Thompson thinks the objection right and shallow. Appeals to authenticity are a dead end, since Buddhism has always been entangled with the economies of its hosts, funded by bankers and rulers in return for merit. So are appeals to a single traditional conception of mindfulness, because there is none: classical styles judge what is before the mind by an explicit set of values, nondual styles reject evaluative judgement for mere non-distraction, and the contemporary style draws mostly from the second.

The scientists share the blame with the journalists. Anything you do changes your brain, so the claim that mindfulness literally rewires it says nothing, the evidence of lasting benefit is tentative, and the idea of an isolable active ingredient inside one brain is likely a mistake, since much of the reported benefit is inseparable from the communal settings of practice. Two internalist ideas prop each other up: that mindfulness is inward awareness of one's own private mind, and that its effects are best seen inside the head. Together they produce the mindful brain, and people say things like: to be a mindful parent you need to down-regulate your amygdala. People are mindful, not brains. A bird needs wings, but its flight is not inside the wings; it is a relation between the whole animal and the air.

His target is a figure from a 2014 Scientific American article tying each phase of focused-attention meditation to a network: mind wandering to the default-mode network, noticing the distraction to the salience network, sustaining focus to the dorsolateral prefrontal cortex. The mapping is empirically unwarranted. On Christopher Mole's account, attention is not a distinct process but the mode in which many processes run together in the service of a task; there is no place in an orchestra where unison resides. And Michael Anderson and Luiz Pessoa show across large imaging databases that any region is activated across many tasks and any task recruits many regions, so moving from regions to networks leaves the mapping many-to-many.

It is also a conceptual mistake, because mindfulness consists in the integrated exercise of cognitive, affective and bodily skills in situated action. The analogy is parenting: being a good parent is a set of skills exercised in the world, which depend on the brain and change it but are not private mental states, and what counts as good parenting depends on the society. Locating it in the brain would explain nothing. Edwin Hutchins's cognitive ecology draws the general consequence: two brains with similar prefrontal activation can be doing quite different things depending on how each person is being culturally orchestrated. Mindfulness meditation is therefore not the inner perception of a private mental theatre; it is the metacognition and internalised social cognition of socially constituted experience, and even hermits are socially supported.

The rhetoric of enlightenment

In 1857 an anonymous article in the Times of London called the Buddha "the Enlightened"; Max Müller, the Oxford philologist who wrote it, republished it under his name the same year. Bodhi means awakened, from the root budh, to wake, and Müller was among the first to render it as enlightenment. The English word carries two inheritances, one Christian and one from the Age of Reason, where Kant in 1784 defined enlightenment as humanity's emergence from its self-incurred immaturity. Müller's Buddha, striking out alone and rejecting Vedic ritual and priestly caste, fits both. The discrepancy is about the self. Using your own understanding, for Kant, requires taking yourself to be a rational agent and asserting your moral autonomy, while Buddhist modernists describe enlightenment as the realisation that there is no autonomous self. Where the European Enlightenment presses for freedom of the self, modernist enlightenment offers freedom from it.

The larger problem is that there is no agreed content to the awakening. In one early discourse it is a cognitive insight: the Buddha comprehends the Four Noble Truths and knows the task is done. In another it follows the four absorptions, entered one by one, after which three knowledges arrive over the three watches of the night — recollection of past lives, knowledge of how beings pass away according to their actions, and knowledge of the destruction of the taints. The fourth absorption has no thought or feeling, which seems incompatible with actively discerning that this is suffering and this is its origin; yet the text places the insights in meditative concentration and mentions no other state. Psychology cannot decide, and the textual history offers no rescue: the teachings were carried orally for roughly four centuries before being written down in Sri Lanka in 25 BCE.

Buddhist modernists want both halves of an impossible bargain. One prong demythologises awakening into a comprehensible psychological state; the other romanticises it into a nonconceptual epiphany, which David McMahan compares to the literary modernist's moment of being; neural Buddhists add a third step and look for correlates. But either you accept faith in an awakening that transcends conceptual thought, in which case science can neither legitimise nor delegitimise it, or you believe only what can be made scientifically comprehensible, in which case the fullness of being has to go.

Thompson's own position goes further than the critique. Not sharing the faith, he doubts that "enlightenment" has a referent outside the sphere of concepts. Nothing is a game apart from the concept of a game: bits of wood are not chess pieces unless conceptualised as pawns and knights. Money, in Dennett's example, would vanish if everyone forgot what it was, leaving engraved paper but no exchange rates and no monetary value. Love is the closer analogy: the concept is multivalent, forms such as courtly love are no longer available to us, and the biology of attachment is necessary without being sufficient. None of this makes love unreal; its reality depends partly on our concepts of it.

Any experience called an enlightenment experience is concept-dependent in the same way, and three consequences follow. Enlightenment cannot be an essentially nonconceptual epiphany, since nonconceptual elements get to be part of it only under a concept. No brain state is intrinsically a state of enlightenment, any more than one is intrinsically a state of love — and if we found neural patterns reliably correlated with Yo-Yo Ma playing Bach's first cello suite, the finding would tell us about the effects of training, not about music. And modern Buddhists must ask not what enlightenment is but what it could be. That is not idle, given the record of misconduct by teachers extolled for their realisation, the jingoism of a good number of Japanese Zen masters during the Second World War, and the persecution of Muslims by monks in Myanmar.

Cosmopolitanism and conversation

The last chapter opens with what Thompson calls his favourite moment in the discourses. In "The Noble Search", the newly awakened Buddha meets a seeker named Upaka, who asks who his teacher is. The Buddha answers in verse that he has transcended everything, has no teacher, and that no one like him exists even among the gods. Then you must be a universal conqueror, says Upaka. Yes, says the Buddha. "May it be so, friend," Upaka replies, and shaking his head, takes a side road. Upaka was an Ājīvika, of a movement that rejected Vedic religion as the Buddhists and Jains did and did not survive; all its texts are lost and we know it only through its rivals. The Buddha was one seeker among many in a plural and competitive world, whose teaching had to be argued for.

Sheldon Pollock calls the sphere that later ran from today's Pakistan to Indonesia, for roughly a millennium, the Sanskrit cosmopolis, and contrasts it with the Latin countercosmopolis of the Roman Empire. Latin described its own universality and travelled as the language of a conquest state, obliterating what it found; Sanskrit never conceptualised its universality, and those who took part in its culture chose to. Two kinds of cosmopolitanism, then, one coercive and one voluntary. In its normative sense the idea must square allegiance to those near and dear with the thought of humanity as one community, which Kwame Anthony Appiah's partial cosmopolitanism does by holding both strands. His model is conversation, which needs only enough overlap in vocabularies of value to begin and need not end in agreement: it is enough, he says, that it helps people get used to one another.

Thompson tests the model on a conversation he has taken part in. Its origin was Varela, whom he met in 1977 at a conference organised by his father and Gregory Bateson at Lindisfarne; Varela was known for the theory of autopoiesis and had recently fled Chile after the 1973 coup. A talk he gave in 1978 on science and Buddhism contains the seed of what the Mind and Life Dialogues later attempted. Varela asks what happens when you belong to two communities and bring them into conversation, and answers that what you do is factually transformed. In meditation one notices that experience is gappy, a series of discrete noticings rather than a continuum; Varela asked whether that could be tested, and published an experiment supporting the hypothesis that perception is discrete, inspired by Vasubandhu and now recognised as an early source of current work. Thompson was one of the subjects.

Varela's second example is harder: doing biology means killing animals, which is central to science rather than incidental, and which he carried out himself. Standing before an animal he would ask whether it was worth it, and had no answer beyond the observation that he could not drop the whole thing. That is what conversation across traditions does, Thompson says: it destabilises your background commitments. Varela also named the two ways such conversations fail. The embellishment attitude stays inside science and decorates it with Eastern metaphors, as The Tao of Physics had done in 1975; the justification attitude stays inside a spiritual tradition and uses science to validate it, as research on Transcendental Meditation did; the hype around mindfulness today resembles the hype around TM then.

The first Mind and Life Dialogue took place in Dharamsala in 1987 and thirty-three have followed. They work when the participants let their own positions become unsettled; they fail when participants slide into embellishment and justification, or when Buddhists reframe their theories as a science. They are also limited by design, since the strategy is to bracket the deeper commitments on both sides — karma, rebirth and liberation on one, physicalism and reductionism on the other. Thompson agrees with Donald Lopez that those are precisely the topics that must be unbracketed.

The closing argument returns to values. Appiah observes that the natural sciences have not produced the progress in our understanding of values that they have produced in our grasp of facts, so their success gives no reason to think our understanding of values better than other people's. A cosmopolitanism that is not Eurocentric will draw on many vocabularies, from the Akan proverb that in a single polis there is no wisdom to the Ubuntu principle that a person is a person through other persons. Thompson's summary is a list of tenets he takes to be indefensible: that Buddhism is a mind science; that there is no self; that mindfulness is inward awareness of a private mental theatre; that neuroscience establishes the value of mindfulness practice; and that enlightenment is a nonconceptual experience outside language and tradition, or a brain state. What Buddhism offers instead is a critique of our preoccupation with the self and of our overconfidence that science tells us how the world is apart from how we measure it — a critique that Buddhist modernism, by reinforcing both, has helped to blunt. He is not a Buddhist, he ends, but he wishes to be a good friend to Buddhism, and a viable cosmopolitanism would be Buddhism's greatest ally.

Since 2020

The historical half of the argument has held and been extended. David McMahan, whose account of Buddhist modernism Thompson leans on, followed it with Rethinking Meditation in 2023, arguing that modern meditation is a genuinely new cultural formation rather than the recovery of an ancient one. The claim that Buddhism is exceptionally immune to the political pathologies of religion has aged worst: the Myanmar coup of February 2021 and the violence that followed put monastic nationalism back on the front pages.

The empirical half has moved his way on the science of mindfulness and against him on its clinical uses. The MYRIAD trial, reported in 2022, taught mindfulness in English schools to some twenty-eight thousand adolescents and found no benefit over ordinary lessons, and work on meditation-related adverse effects has grown, vindicating the questions he says were pushed aside. At the same time a 2022 randomised trial found mindfulness-based stress reduction no worse than escitalopram for anxiety disorders, a stronger clinical result than the tentative verdict he records — though it is a clinical claim rather than a neural one, and so leaves his actual target, the inference from brain imaging to what mindfulness is, untouched.

On the self, the debate he wanted has arrived. Jay Garfield's Losing Ourselves of 2022 is the sharpest reply from the other side, defending the no-self view as both true and good for us and arguing that it is the person, not the self, that survives the analysis — close enough to Thompson's own terminology to make the remaining disagreement precise rather than verbal. Anil Seth's Being You of 2021 argues that the self is a controlled perceptual construction built from predictions about the body, which is the construction thesis Thompson wants, though Seth keeps the word illusion that Thompson says does the damage. Nothing since has settled the question, which is the outcome the book's argument predicts: these are philosophical matters, and no imaging result is going to decide them.