Canonical: https://thonly.org/research/the-body-is-the-lab · Licence: CC0 1.0
Attribution note. This essay is in my voice. The question, the metaphor and the final editorial control are mine; Miss Aquarius℠ drafted the prose on my behalf, and the correction of my question came through that same collaboration. My own full pass is still owed. The essay argues one thing: meditation and physics are two instruments on two windows, and neither reads the other's dial.
On 14 October 2026 I typed this:
If perception compiles continuity and insight decompiles it, perhaps the only path to the most fundamental physics is through meditation, the ultimate real-time decompiler. The physical body is the physics lab; the meditative mind is the reverse-engineering tool. Perhaps complete decompilation is Nibbana.
I think in metaphors, and that one felt like the top of a mountain. I came up through software, and I now work on the Khmer Tipiṭaka with my father and with a proof assistant. A sentence that fused the three seemed like the sentence I had been walking toward.
It is wrong in three places, and the corrections are why I am writing. The word only is refused by the texts I was praising. The physical body is not the lab the texts describe. And complete decompilation is not nibbāna. What survives is better than what I typed, and it keeps the title: the body is the lab, once body means what the Buddha meant by it.
Here is the working picture. It has three artifacts, two tools and one guard.
SOURCE what is real by its own nature: citta, its factors, matter's
kinds, and nibbāna, which no process builds
│
│ perception compiles, moment by moment
▼
BINARY the lived world: a person, a day, a solid hand
▲ │
insight decompiles │ │ measurement and
the inner window, │ │ mathematics compile
always after the fact│ ▼ the outer window
PSEUDO-CODE: physics. It runs,
it predicts within its domain,
and many of its names are ours.
GUARD We compile the account, not the world.The guard does the most work in that diagram. Perception compiles means the lived world, the one with people and days in it. It does not mean that minds manufacture stars. The Theravāda is realist about other streams and about matter that nobody perceives: temperature produces matter "internally and externally" (ajjhattañca bahiddhā ca, Saṅgaha ch. 6, abh07t.nrf.xml:2213), and the rain went on falling before anyone was there to thank it. Our own work at Silicon Wat℠ compiles an account: the Saṅgaha's statements into Lean, where a build passes or fails. That is the only thing here that literally compiles.
And the metaphor has an edge that should be stated before anyone leans on it. A real compiler has an author and a purpose. This one has neither. The Visuddhimagga tells the meditator to attend to the elements "by not knowing": the earth element does not know "I am the earth element", or "I am a support for the other three" (na jānāti, Vism XI §358, e0101n.mul.xml:8845). There is nobody inside the machine compiling. I keep the verb because it helps me think, and I hold it as a verb with no subject.
The clearest statement of the inner tool is two sentences long. The three marks do not appear, says the Visuddhimagga, because something covers each. Impermanence is concealed by continuity (santatiyā paṭicchannattā), suffering by the postures, not-self by compactness (ghanena). And when rise and fall have been taken up and "continuity has been disrupted" (santatiyā vikopitāya), the mark of impermanence appears as it is; when the elements have been separated and "the breaking up of compactness" is done (ghanavinibbhoge kate), so does the mark of not-self (Vism XXI §739, e0102n.mul.xml:6625).
Three concealers, removed by attention. To a programmer's ear that is a decompiler undoing three optimization passes, though the text speaks of marks appearing and not of decompiling. That word is mine.
The worked example is the four-elements meditation of Vism XI, and I find it funnier and stranger each time I read it. Take a hair. "When eight dhammas come together there is the convention hair; when those same eight are separated, there is no convention hair" (aṭṭhadhammasamodhānā, hoti kesāti sammuti; tesaṃyeva vinibbhogā, natthi kesāti sammuti, §348, e0101n.mul.xml:8661–:8669). Then grind the body. Its earth element, pulverized to the finest dust, "would be about a doṇa", a grain measure (§349, :8677), held together by water, kept by fire, propped by wind. Walking, sitting and waving a hand are this "machine of elements" (dhātuyantaṃ), "like a magic show, deceiving the foolish" (bālajanavañcanaṃ māyārūpasadisaṃ, :8689). The elements in the smallest cluster can be told apart by characteristic and never by place (§353, :8825). And the payoff: one who keeps at this "plunges into voidness and uproots the perception of a being" (suññataṃ avagāhati, sattasaññaṃ samugghāteti, §360, :8881).
Notice what that output is. It is not a table of constants. It is a fearlessness, a loosened grip on me, and in the text's words a destination of the deathless or a good rebirth. The decompiler returns a result about clinging.
And it returns it late. Two minds do not occur together, so the mind that notes is never the mind noted. "Always a moment late" is my gloss, not §739's. But it carries a real limit: whatever this instrument is, it is not a live probe reading the bus.
My first version of this section said that physics is conventional truth, pseudo-code that predicts the binary exactly while all of its names are ours. A refuter took that apart in one pass, and it was right to.
First, exactly is false. Physics predicts to a stated precision, in a domain, and at its frontier it is unfinished: two theories that do not fit together, a measurement problem, and matter inferred and not identified. Pseudo-code is the right word partly because it has not all been run.
Second, the texts do not put physics' quantities on one side of a line. The Kathāvatthu commentary's verse is about speech: "agreed speech is true, by reason of the world's agreement; ultimate speech is true, the true characteristic of dhammas" (saṅketavacanaṃ saccaṃ, lokasammutikāraṇaṃ; paramatthavacanaṃ saccaṃ, dhammānaṃ tathalakkhaṇaṃ, abh03a.att.xml:2837–:2841). Its examples of the agreed are a being, a person, a god (:2821). And a method the Puggalapaññatti commentary reports from "the teachers" sorts designations, not theories (abh03a.att.xml:741–:757):
ON THE UNREAL SIDE ON THE REAL SIDE
────────────────────────────────────── ────────────────────────────────
long, short, far, near (:745) earth, heat, hardness, warmth
two, three, four (:753) (a designation of the dhamma's
"eighty years old" (:753) own nature) (:753)
thin, stout, round, square (:757) "nirodha, nibbāna" (:757)
time, direction; pot, cloth (:741)So here is the repair, and I can make it inside the metaphor. A decompiler's variable names are invented, and many of the registers they point at are really there. When a physicist writes down heat, the texts have a real dhamma nearby, though no text equates a temperature with tejo and I will not either. When the same physicist writes down a length, an age, a count or a shape, the teachers' method says that this is a designation. Whether a charge or a mass sits in the left column or the right, the texts do not say, since they never met one. A physicist who says an electron's charge is nobody's convention is not contradicting them.
And conventional is no insult. "When they speak in agreed terms they speak what is true, what is so, and do not lie; when they speak in ultimate terms they speak what is true, what is so, and do not lie" (abh03a.att.xml:2821). "Two truths the Awakened One declared; a third is not found" (:2825–:2829). Telling physicists that their names are ours does not demote them. It places them among those who say what is true in the way that truth can be said.
The finding that surprised me most is the smallest one. In October our formal work asked what the discreteness of the mental stream implies about matter. The answer, with the refuter's narrowings, is now in the corpus (Abhidhamma and Discrete Quantum Gravity, §2 and §4.8):
That last line is the bridge I was looking for, and it is much humbler than the one I typed. It settles the status of continuum-talk. It does not settle the structure of matter. The texts do not tell a physicist whether spacetime is smooth. They say only that "smooth" and "grainy", as names for a length, are names.
If the argument were only mine, it would be one more person telling physicists what they may not ask. It is not only mine.
The traveler. Rohitassa, once a seer who could cross the world in a stride, walks for a century to find the world's end, and dies on the road. The Buddha's answer is the source of my title: "I do not say that the end of the world, where one is not born, does not age, does not die, can be known, seen or reached by going. Yet I do not say that one makes an end of suffering without reaching the end of the world. It is in this very fathom-long body, with its perception and its mind (sasaññimhi samanake), that I declare the world, its arising, its ending, and the path" (AN 4.45, s0402m3.mul.xml:1992).
That sentence corrects me twice. The Mahāṭīkā spells out the first correction: it cannot be declared "in a mere aggregate of temperature-born matter, without consciousness" (e0103n.att.xml:3229). So the lab is not the physical body. It is the body with a mind in it. And its world is not the one in a telescope. The same commentator reads this world as the world of formations, whose end is nibbāna (:3233), and a few lines on names three worlds: formations, beings, and the receptacle or location world (okāsaloka, bhājanaloka), "matter and the rest bound to no faculty" (:3237). That last is the nearest thing the texts have to physics' subject matter. The mapping is mine; the Mahāṭīkā never heard of physics.
The unthinkables. AN 4.77 lists four things one should not think through, on pain of madness and vexation. One is lokacintā, world-speculation (s0402m3.mul.xml:3400). I would love to quote this at every late-night theory of everything, including mine, but the commentary narrows it: the question it means is "who made the sun and moon, who the great earth, who the ocean" (s0402a.att.xml:5396). That bars first-cause speculation. It does not bar a spectrometer. And the same list holds a line against me: the range of the jhānas is also unthinkable, so the tradition does not hand me a tame instrument either.
The leaves. The Buddha picks up a few leaves and asks which are more, these or those in the forest. "So too, what I have directly known and not taught is far more. And why have I not taught it? Because it is not connected with the goal" (SN 56.31, s0305m.mul.xml:10381). Read carefully, this does not say that meditation cannot reach physics. It says something more interesting: that whatever else was known, it was left out on purpose, because it does not lead to the end. So physics cannot be the secret content of the path. It may be in the forest, but the handful was chosen without it.
The strongest objection to all this comes from inside the texts, and I cannot make it go away.
The Visuddhimagga lists the three worlds to explain why the Buddha is called knower of the worlds, and of the location world, with its thousandfold system, it says: "that too the Blessed One knew in every way" (tampi bhagavā sabbathā avedi, Vism VII §136, e0101n.mul.xml:4929–:4941). The commentary says the paramāṇu, the smallest measure on its ladder, is "out of range of the fleshly eye, in range of the divine eye only" (abh02a.att.xml:6333). Living meditators in Pa-Auk Sayadaw's lineage report discerning rūpa-kalāpas in concentration. So a meditative instrument aimed outward is not my opponents' invention. It is in the tradition I am defending.
I have three things to say, and none is a refutation.
First, the claim that survives is narrower than the one I would like. A Buddha's knowledge is a separate case, and I am in no position to reason about it, since AN 4.77 also lists it among the unthinkables. What I can defend is this: for a disciple, the texts teach insight as a way out, not as a survey. Of a report I cannot check, I can say only that I have not seen a kalāpa, and that this essay would be no stronger if I had.
Second, the one time the outward reading was turned into a number, I can tell you what happened, because we ran it. The Visuddhimagga Mahāṭīkā says "those kalāpas are of paramāṇu size" (e0103n.att.xml:6229). The ladder gives 576,108,288 paramāṇu to a finger-breadth. On any reasonable finger that makes one kalāpa of external matter roughly 3–4 × 10⁻¹¹ meters across, if you read the line as a measurement and carry it outward, and neither move is in the text. A gold nucleus would then sit inside one eight-member cluster with 197 individuated nucleons in it. That collides. The same ladder climbs to a Mount Sineru 6.8 million yojanas high (abh02a.att.xml:6333), and a monk is told to see a tiny fault as that large.
Third, and this is the part I love, the tradition had already declined the number. Ledi Sayadaw, in Burma around 1900, before anyone had scattered an α particle off gold, wrote of that very line: "it is to be taken as said merely in conformity with the powder-attention to the elements" (e0301n.nrf.xml:4705). It was a meditation device, in a sentence whose own stated purpose is to explain why the powder mode is taught. He left one kalāpa at most a forty-ninth of a paramāṇu, with no lower size at all. So no grain of matter's extent follows from the texts on his reading. The collision is real and it lands on a reading, and the tradition's own hand had moved off that reading first. I would not call that a vindication. Ledi was not answering Rutherford, and I am choosing, after the fact, the voice that survives. But he shows the shape I want: a teacher reading a number as a teaching.
There is a second reason the outward reading cannot do what I asked of it. A physics that only meditators can read would not be physics. A result has to be checkable by someone who has not sat. I would hold to that even if every report were true.
My strongest claim is the split itself: two windows, two instruments. Here is the best case against it. There are not two windows, because there is one body. Mind-born matter lifts the hand. The rows that give matter its sub-moments are stated on the stream's own moments for internal matter. Neuroscience measures the inner window from outside; Francisco Varela's neurophenomenology uses trained first-person report to constrain third-person data; and every physicist who has ever read a dial did it with a citta. If the two instruments meet at every point in the fathom-long body, then "neither reads the other's dial" is a slogan.
I concede most of it. The windows open on one room. Reports from the cushion may well constrain the science of consciousness, and that is not what I am refusing. I am refusing two specific inferences: from insight to fundamental physics, and from physics to the end of suffering. Those are the crossings I say fail. If you can show me either one working, the essay falls (§12). Short of that, I keep the split as a claim about what each instrument returns, not about where it stands.
There is a sharper version of the objection, and it is my own run's. If magnitudes are designations, why does a mismatch of magnitudes count as a collision at all? I have no clean answer for sizes. The count collision was built partly to escape that: 197 is not a designation of the kind that a method of the teachers dissolves, since the plurality of dhammas stays real. But a nucleon is not a dhamma in any text, so a reader can refuse the whole comparison. I think that reader has a point.
The last correction is the one I most needed. Complete decompilation is nibbāna was the end of my sentence, and it is the error that matters most for a Buddhist.
Decompilation works on the conditioned. Strip every designation from the lived world and what you hold is still conditioned: moments of mind and matter, arising and ceasing. Nibbāna is not what remains at the end of that subtraction. The texts call it unconditioned (asaṅkhata), never produced; the path takes it as an object, at one mind-moment, and does not arrive at it by exhausting a list (Saṅgaha ch. 9, abh07t.nrf.xml:3313–:3337). And a detail I did not expect: among the designations, "nirodha, nibbāna" fall on the real side (abh03a.att.xml:757). The name is not a label on an absence.
So in the metaphor: you cannot get there by decompiling further, because it was never in the build. You also cannot get there by physics, however good the pseudo-code. That is the sense in which the two windows are one problem for me and not for the texts. I wanted a unified theory. The tradition offers an exit, and an exit is not a theory.
This is an old conversation, and the people in it have mostly gone the other way. I have read none of the following works cover to cover for this essay, and what I say of them rests on publisher and review records, so check me.
I claim no priority for any of this. The refusal is the Buddha's, the bound is Ledi's, and the checkability argument is everyone's.
I am not neutral. I am a devout Theravādin, and the institution I founded has staked its alignment work on these texts. The line we settled on for the bench at siliconwat.dev reads: mind/body (observed) → Abhidhamma (received) → math (proved) → physics (conditional). The word conditional was not my idea. I wanted a continuous pipeline through to existence, and I was talked out of it.
I also wanted the grand version for a personal reason. As a boy I read about Einstein and the reconciliation he did not finish, and some part of me would like to help finish it from the cushion. That hope is the exact shape of bias that produces a sentence like the one I typed. The essay is partly a record of being corrected, and it is fair to weigh it as that.
Done. The Lean model and the run behind §5 and §7 are public (SiliconWat/formal-abhidhamma, runs/2026-10-14-the-continuum-bridge/), with its pre-registration, an agent blind to it, and a refuter's report kept verbatim. Every Pāli passage here was read in the Chaṭṭha Saṅgāyana text with its context, and this essay's thesis went through its own refuter before a word was written. That pass killed my claim that physics is wholesale conventional truth, and it moved the Dalai Lama from opponent to ally.
Not done. No physicist has reviewed this essay. No Pāli reader outside this project has checked my readings. I have not measured anything on a cushion and claim no attainment. I have not tried to test Pa-Auk reports, and I don't know how one would do it fairly. The works in §10 were not read in full. And the teachers' method in §4 is one reported method in a commentary, not the canon.
What would show this wrong.
I would rather be shown wrong on any of these than go on quoting the sentence I typed. I still like how it sounded.
Drafted with Miss Aquarius℠, the AI collaboration of this institution, per the corpus convention for essays in my voice; the question, the metaphor and final editorial control are mine, and my own pass is still owed.
SiliconWat/formal-abhidhamma (10.5281/zenodo.23067241): the axioms, the Lean and the ledger at runs/INDEX.md.Pāli citations are to the Chaṭṭha Saṅgāyana text as distributed by VipassanaTech (tipitaka-xml, romn/, at commit 05d5d3c), by file and line. Translations are mine and are readings. Visuddhimagga passages are e0101n.mul.xml and e0102n.mul.xml; its Mahāṭīkā is e0103n.att.xml; the Kathāvatthu and Puggalapaññatti commentaries are in abh03a.att.xml; the Sammohavinodanī is abh02a.att.xml; the Abhidhammatthasaṅgaha is abh07t.nrf.xml; Ledi Sayadaw's Paramatthadīpanī is e0301n.nrf.xml.