10: In Search of Dharma

Throughout In Search of Dharma I have insisted that dharmas are made, plural, and unprivileged: no single The Dharma, no view from nowhere, no tradition seated above the others in advance. And then, having said all that, I have gone on using the word ‘better’. Better dharmas, worse ones, failed ones. A reader is entitled to stop me here and ask the obvious question: better by what standard? If you refuse to privilege any dharma, what exactly are you measuring them against?

Part 8 answers this, but it answers it six hundred words from the end of a longish book, after eight parts of preparation. A standard that matters should also be stated where it can be examined on its own, attacked on its own, and cited on its own. So here it is, whole.

A dharma is to be judged by two tests.

The first is substantive: a dharma is better the more fully it holds without crushing. The better ones gather human lives into shape and meaning without breaking the people inside them or the strangers outside them.

The second is procedural, and it is the keystone: dharmas are better to the extent they remain revisable by the people bound by them.

The first test asks what a dharma does. The second asks who holds the pen. The rest of this essay defends each, explains why the second is not merely a fourth clause of the first, and then meets the strongest objections I know of, including the ones I cannot fully answer.

Plural is not equal

Begin with the accusation the whole apparatus exists to answer: that a project which privileges no dharma has thereby forfeited the right to judge any, and that ‘unprivileged’ is just relativism under another name.

It is not, and the distinction is not a technicality. Isaiah Berlin spent a career on it: value pluralism is an alternative to both relativism and absolutism. The many values that human communities build their lives around are real, grounded in a shared human nature; there are simply more of them than any single way of life can hold at once, and they genuinely conflict. That is a different claim, and a far stronger one, than “anything goes”. The lazier doctrine collapses at its first two tests. It cannot condemn an atrocity, since the atrocity is somebody’s custom. And it cannot make sense of reform: the abolitionist and the suffragist both said that their own society’s settled practice was wrong, which requires a standard standing above settled practice, which is precisely what relativism denies them. Any account of dharmas that makes reformers impossible has disqualified itself. Every tradition in this book has had its reformers.

Note also what disagreement does and does not prove. That cultures disagree about a norm is a fact about acceptance, not about bindingness; “many people permit recreational cruelty” and “recreational cruelty is permissible” are claims of different kinds, and the first has never entailed the second. Descriptive plurality leaves the question of standards exactly where it found it.

So: plural, yes; equal, no. The question is what the standard is.

The first test: holds without crushing

The shape the evidence keeps converging on, at every scale this project has examined, is a thin floor with thick plurality above it.

The floor is a small set of near-universal constraints: limits on cruelty and violence, and the protection of basic human capabilities. This is not wishful cosmopolitanism. When Oliver Scott Curry’s team tested seven rules of cooperation against the ethnographic record of sixty societies, every rule was held as morally good wherever it appeared; human morality, on that evidence, is variation on a theme, which is exactly the architecture the floor predicts. Martha Nussbaum gave the floor teeth with her list of central human capabilities, the goods any defensible way of life must let the people inside it actually achieve: life, bodily integrity, practical reason, affiliation, play, and the rest. Amartya Sen, who built the capabilities approach alongside her, famously refused to fix any such list. Hold that refusal in mind; it returns below as more than a footnote.

Above the floor, diversity stops being a problem and becomes a richness. Below it lies not a different-but-valid dharma but a failed one. And a dharma can fail the first test in two opposite directions. It can hold and crush: the tyranny-and-purity spiral of Part 6, the order that gathers its people magnificently and grinds them. Or it can refuse to crush and fail to hold: the thin gruel of Part 7, the curated patchwork that never binds a community or carries a life through grief. The good ones do both jobs at once. They hold firmly enough to give a life its spine, and loosely enough that other lives, and other dharmas, have room to do the same.

Two cautions come with this test, both from Berlin. The floor identifies failures; it does not rank successes. Above the line the goods genuinely conflict, and where values are incommensurable there will be ties and tragic trade-offs, not a league table. And the judging power is not the imposing power: knowing that a dharma has crossed the floor tells us where argument, pressure and refusal are justified. It tells us nothing about who may reach for the boot. Those are different powers, and the conditions established in Part 6 bind the second, not the first.

The problem the first test cannot solve

Now the honest difficulty. The floor needs placing, and every placement is made by someone, from somewhere.

Sen’s refusal to fix the list is usually read as a technical dispute with Nussbaum inside one school. It is much more than that. It is the recognition that a fixed floor, specified once and for all, is somebody’s thick dharma at the moment of fixing, and the fixer is never a view from nowhere. One tradition’s obvious floor (gender equality, say, or individual rights of exit) is another’s thick imposition; the hard cases sit exactly on the line, and this project’s own research notes flag, repeatedly, the trap of mistaking one’s local dharma for the universal threshold. The charge lands on this book too. Western interpreters have reliably found in other people’s dharmas whatever they carried in: Kantians finding Kant, positivists finding Russell. It must be asked whether we found ‘holds without crushing’ in the traditions or brought it along, and Part 8 asks it.

The usual escape routes fail. Hand the placement to experts, and you have built a priesthood; the floor is then whatever the credentialed say it is, which is capture with footnotes. Hand it to an overlapping consensus of the traditions, and you do better, but you freeze the floor at the moment of overlap; consensus is a snapshot, and snapshots yellow over time.

The way out is not a cleverer placement. It is to change what kind of thing the standard demands: not the correct position of the floor, but a floor whose position remains permanently contestable by the people standing on it. Which is to say, the first test cannot stand on its own. It needs the second.

The second test: who holds the pen

State it again, because this sentence is the centre of the whole standard: dharmas are better to the extent they remain revisable by the people bound by them.

Every word is doing work. Revisable: the capacity must actually exist, not as a myth of origin (“our founder once changed things”) but as a living organ. By the people: not by guardians alone. A dharma whose revision is monopolised by its priesthood, its politburo or its board is not revisable in the relevant sense; it is revisable the way a locked room is ventilated, from the outside, at the discretion of whoever holds the key. Bound by them: the test is indexed to the held, not the onlookers. Reform imposed wholesale from outside fails it too, however enlightened; that is the coloniser’s shortcut, and Part 6 has already priced it.

The clearest statement of the underlying principle in this project’s corpus is a century and a half old. Bakunin, asked whether rejecting imposed authority meant rejecting all authority, answered: far from it. “In the matter of boots, I refer to the authority of the bootmaker.” He consults the architect and the savant, listens freely and with respect, “reserving always my incontestable right of criticism and censure.” What he refuses is not authority; it is authority he cannot revoke. There is, he concludes, “no fixed and constant authority, but a continual exchange of mutual, temporary, and, above all, voluntary authority and subordination.” That is the second test in one sentence, written in 1871. Transmission always requires authority; every dharma is taught before it is examined, and there is no other way to teach one. The bootmaker’s principle does not abolish teachers. It denies them tenure over the taught.

Why should this be the keystone, rather than one virtue among several? Because it is the only part of the standard that follows from the project’s own central finding. Dharmas are made, not found: that claim has been the spine of the book since Part 1. And if made, then made by people who err, in circumstances that pass. An unrevisable dharma is therefore a made thing pretending to be found. To forbid revision by the bound, a dharma must claim that its shape was fixed elsewhere: by heaven, by blood, by history’s verdict, by the founder’s unrepeatable genius. It must, in short, lie about its own origins, and institutionalise the lie. The concealment this book exists to dismantle (building the things that hold us and then kneeling before them as if we had received them) is not merely error. Enforced, it is the mechanism of every catastrophe in Part 6. Revisability is what “made, and known to be made” looks like when it becomes an institution. It is honesty about origins, given a constitutional form.

I did not arrive at this by deduction, and I should say so. The criterion converged from three directions before it was ever stated as one sentence. Part 8’s verdict already carried it as a clause (“stay honest enough to correct themselves”, since sharpened in the text to “remain revisable by the people they bind”). The research note on relativism (3.4) had, independently, listed “self-correcting and non-oppressive” among its criteria. And the note on engineered dharmas (7.7), after walking through the Jacobin cults, the New Soviet Man and Year Zero, ended its licence to build with the condition that matters: the power to unbuild kept in the hands of the built-for. Three separate investigations, three routes to the same clause. This essay’s only real contribution is to promote it from clause to keystone, and to notice what it solves: the floor-placement problem dissolves once the floor’s position is understood as permanently revisable by the placed. Sen’s refusal to fix the list stops being the approach’s embarrassment and becomes its deepest insight.

The tests at work

Run the pair across the cases this project has already put in evidence, and notice that the second test repeatedly sees what the first one squints at.

The wong sikep of Part 3, the book’s own hero and its hardest case. On the first test they pass with two pauses, honestly recorded: the children long kept from letters, the wife defined by her duty. It is the second test they pass outright. The schooling refusal softened after independence; some communities built their own curriculum rather than accept the state’s or abandon the principle; the hardest edge was adjusted from inside, without the dharma losing itself. The pen stayed in the village. That is what revisability looks like when it is lived rather than proclaimed, and it is why this tradition could pause at the floor and still stand.

The engineered dharmas of note 7.7 fail the second test before the first can even be scored. The Cult of the Supreme Being, the New Soviet Man, Year Zero: in each, an ethic was authored at the centre and enforced downward, and the one power never distributed was the power to write back. The atrocities differ; the pen-monopoly is constant.

China’s ‘ecological civilization’ doctrine is the instructive modern case precisely because it embarrasses the first test. Written into the constitution in 2018, it holds, it coordinates, and it delivers real results, including one of history’s largest reforestations, on the measures it chooses to count. The scholars who traced its history call it authoritarian environmentalism: an ethic driven by technocratic elites, enforced on people who do not get a say. The first test hesitates, because the thing plainly functions. The second does not hesitate at all.

The corporate culture-by-decree, Part 8’s minor key: values authored in a workshop, laminated, and cascaded to people who must live them but may not amend them. Revisable, certainly: by the board, at the next rebrand. Revisable by the bound, no. The comfort of the regime does not change its grammar.

And the machines, since they are where this project is ultimately headed. Delphi, the 2021 model distilled from 1.7 million moral judgements, was in effect a dharma with no revision organ at all: a frozen verdict machine, wrong the way an oracle is wrong, with no path by which the judged could talk back. Whereas a constitution for a machine, an explicit, readable, rewritable document stating the values it is trained toward, is at least the right shape: the pen exists, the text is public, revision is procedurally possible. The second test then asks its question anyway: revisable by whom? By the lab that wrote it, or by the people increasingly bound by it? The honest answer today is the former, with experiments (a thousand citizens drafting a public constitution, in one trial) probing toward the latter. The criterion does not care that the dharma’s substrate is silicon.

The objections

“You brought it along.” The sharpest objection says revisability is just liberalism smuggled in as a meta-standard: one more Western thick dharma flattering itself that it is a neutral referee. Half of this must simply be conceded. The criterion is a commitment, not a discovery; there is no view from nowhere, this project has one point of view like everything else, and an avowed commitment is the only honest kind. But the other half should be resisted, because the criterion is thinner than the liberalism it is accused of being. It demands no elections, no individualism, no rights-talk, no separation of powers. It demands only that the capacity to revise exist and sit with the bound, and traditions have built that organ in forms liberalism would not recognise as its own. Musyawarah, deliberation toward consensus, is a revision organ. Ijtihad, the labour of fresh interpretation in Islamic law, is a revision organ, and the centuries-long argument over whether its ‘gate’ ever closed is itself an argument about exactly this criterion, conducted entirely inside the tradition. The wong sikep revised a refusal most of a century old without a ballot in sight. What the criterion does refuse is narrower and starker: any dharma whose self-understanding requires that the bound never touch it. A genuinely closed orthodoxy will experience this standard as an attack. It is one. The difference between this project and the neutral-referee posture it declines is that I am saying so.

Nietzsche’s challenge. The corpus’s own fiercest counter-voice must be given his turn, because he is the most serious attempt in the tradition to do what I have said cannot be done: rank dharmas with no floor at all. Every people hangs its table of goods over itself, he says, and the tables are created, not found. So far, this book’s own thesis. But he then ranks the tables as symptoms, by the kind of life, ascending or declining, that produces each, and expressly refuses the floor, which he diagnoses as the herd’s own thick morality universalising itself. The challenge is real and I will not pretend to dissolve it in a paragraph. I will state its price, which the scholarship states too: the standard “what kind of life it produces” smuggles in a substantive aristocratic ideal while claiming to stand outside all moralities; by its own symptomatology it is itself a symptom; and its reception history required decades of scholarly repair. An option with a stated price is not a refutation of the floor. It is the standing reminder that the floor must be argued for, never assumed. That is, again, why the pen matters more than the placement.

“Revisability just measures churn.” No. The criterion names a capacity, not a rate. A dharma may stand substantially unchanged for a century and pass, because the bound held the pen throughout and chose, year after year, not to use it; consent renewed is not stagnation. Another may change every quarter and fail, because every change was the centre’s. What is measured is where the capacity lives, not how often it is exercised. A grip you cannot open is not a hold; it is a trap – and the test for a trap is not whether it is currently moving.

“Can a dharma revise itself shut?” Yes, and the paradox is more apparent than real. A community may use its living revision organ to weld the hatch: entrench a canon, close a gate, criminalise the question. The criterion does not forbid this in advance, and it does not need to. It is a test applied to the dharma as it now stands, not a licence granted in perpetuity to whatever the dharma once was. The community that froze itself by a free act is now a frozen community; the freedom of the welding does not unweld the hatch. Popper saw the same structure in tolerance, and the resolution is the same: the standard judges states, not pedigrees.

“Who are the bound, when the bound cannot speak?” The criterion indexes revision to the people a dharma binds, and my worked cases quietly assume those people are adults with voices. Most of the bound are not. Every dharma binds children first and hardest, in the years before examining it is possible; that is what the installation machinery of Part 5 is for. Many bind the dead, whose vows and bequests persist. Some bind the unborn explicitly, as the Seven Generations teaching does, and the ecological dharmas bind rivers and forests that will never hold any pen. For each of these, the test must convert or confess. For the children it converts into a demand about equipment and exit: a dharma passes only insofar as it gives the child the tools to examine it and the realistic possibility of leaving it, which is exactly what the Sikep schooling case was about, and why its revision mattered so much. For the dead it converts into a denial of tenure: their pen passed to the living at the grave, and a bequest is honoured as an inheritance, not obeyed as a command. For the unborn and the more-than-human it can only be held in trust, and there the criterion must confess its limit: every trustee is a proxy with interests of its own, “speaking for the river” is a role that capture loves, and the most the second test can demand is that the trusteeship itself remain revisable by the living who share the binding. That is thinner than I would like, and I will not dress it up. It is the honest edge of the standard: the second test is strongest exactly where the bound can answer back, and weakest where they never can – which is a reason to keep widening who can answer back, not a licence to hand the pen to whoever claims to speak for the silent.

The is/ought wall. Finally, the objection this project’s own note 7.4 built into the foundations. Nothing above is derived from the ethnography. Curry’s sixty societies show that the floor’s shape recurs; recurrence is not obligation, and no count of cultures ever crossed the gap from is to ought. The wong sikep’s revision shows that the criterion can be lived; that it can be does not prove that it must be. Both tests are wagers: disclosed, argued, and revisable, but wagers. Provenance is not truth. The most this essay claims is that these are the best-argued wagers the evidence allows, staked in the open, by an author who can be wrong.

Held, not trapped

The word this whole project grew from means that which holds. The standard, in the end, only unfolds what was folded into the root. The first test asks whether the holding is real and whether it has become a crushing. The second asks the question that tells traps apart from holds: whether the held could loosen the grip, and stayed.

So, in one breath: the better dharmas hold firmly, crush no one – and keep the pen in the hands of the held. Everything else about them is still to argue about. That the argument never ends is not the flaw in this standard. It is the standard, applied to itself.

Sources & further reading

A companion essay, not a research note: it argues from cases already presented in In Search of Dharma and its research notes, and carries no inline citations. The notes closest to it: 3.4 Relativism vs. Universalism (Berlin, Curry, Nussbaum and Sen, the relativism taxonomy, the Nietzsche challenge); 7.7 Dharma by Decree (the engineered cases and the “power to unbuild” condition); 7.4 Measurement & Falsifiability (the is/ought wall and the project’s evidentiary bar). The substantive test is argued in full, with the wong sikep worked case, at the end of Part 8 of In Search of Dharma. The criterion sentence itself (dharmas are better to the extent they remain revisable by the people bound by them) is the author’s, and is offered under the same terms as everything else here: revisable.

Key works

• Isaiah Berlin, “Two Concepts of Liberty” (1958) and The Crooked Timber of Humanity (1990) – value pluralism as the alternative to both relativism and absolutism; incommensurable goods above the floor.
• Oliver Scott Curry, Daniel Austin Mullins & Harvey Whitehouse, “Is It Good to Cooperate? Testing the Theory of Morality-as-Cooperation in 60 Societies” (Current Anthropology, 2019) – the sixty-society test behind the floor’s shape.
• Martha Nussbaum, Women and Human Development (2000) – the ten central human capabilities; and Amartya Sen, “Capabilities, Lists, and Public Reason” (Feminist Economics, 2004) – the refusal to fix the list.
• Mikhail Bakunin, God and the State (written 1871, published posthumously 1882) – the bootmaker passage; authority without tenure over the taught.
• Friedrich Nietzsche, Thus Spoke Zarathustra, “The Thousand and One Goals” (1883), and Beyond Good and Evil (1886) – ranking without a floor; the challenge met here at its stated price. The reception history was repaired by Walter Kaufmann, Nietzsche: Philosopher, Psychologist, Antichrist (1950), and successors.
• Karl Popper, The Open Society and Its Enemies, vol. 1 (1945) – the paradox of tolerance; standards that judge states, not pedigrees.
• Wael B. Hallaq, “Was the Gate of Ijtihad Closed?” (International Journal of Middle East Studies, 1984) – the centuries-long tradition-internal argument over a revision organ.
• Mette Halskov Hansen, Hongtao Li & Rune Svarverud, “Ecological civilization: Interpreting the Chinese past, projecting the global future” (Global Environmental Change, 2018) – authoritarian environmentalism and the 2018 constitutional amendment.
• L. Jiang et al., “Can Machines Learn Morality? The Delphi Experiment” (arXiv:2110.07574, 2021) – the frozen verdict machine.
• Anthropic & the Collective Intelligence Project, “Collective Constitutional AI: Aligning a Language Model with Public Input” (2023) – the thousand-citizen constitution trial.