https://harmonism.io/world/blueprint/communities-that-held
Civilizational article — Kinship pillar of the Architecture of Harmony. See also: Sacred Architecture, The Network State as Harmonist Order, Community, The New Acre, Voluntary Association and the Self-Liquidating Bond.
Eighty-three American communes founded in the nineteenth century have been counted, dated and compared closely enough to settle an argument. Thirty of them were religious and fifty-three were secular. The religious communes lasted a median of twenty-five years and the secular ones five. Across 1,556 commune-years, a secular commune was three times more likely to dissolve in any given year than a religious one.
That gap is the part everybody quotes, and it is the less interesting half of the finding.
Richard Sosis and Eric Bressler also counted what each commune demanded of its members: the surrendered property, the dress rules, the dietary restrictions, the constraints on marriage and on contact with outsiders. Religious communes averaged seven such requirements and secular communes three. Then they crossed the two measurements, and the result is the one worth sitting with. Among religious communes the number of costly requirements predicts survival strongly and cleanly. Among secular communes it predicts nothing at all.
The same demand, imposed on the same kind of person, in the same century, on the same continent, produces a durable community in one case and no measurable effect in the other. What differs is not the cost. It is what the cost is understood to answer to.
Take the finding at its narrowest, because the narrow version is already decisive.
A secular commune’s rules are the members’ own work. They were drafted at a founding meeting, they can be amended at the next one, and every member knows this. Asking someone to hand over their savings under a rule they helped write is asking them to make a large sacrifice to a document they could revise on a Tuesday. That sacrifice is real and it buys nothing structural, which is why the data show it buying nothing.
A religious commune’s rules answer to something the members did not author and cannot amend. The demand arrives from outside the room. Refusing it is not a disagreement with the neighbours; it is a refusal of the order the neighbours are also under. Nobody in the commune is in a position to grant a dispensation, because nobody in the commune issued the rule.
Rosabeth Moss Kanter reached the same territory from a different direction in 1972, working the same nineteenth-century record. She set the threshold for a successful commune at twenty-five years and identified six mechanisms that separated the survivors: sacrifice, investment, renunciation, communion, mortification, transcendence. Five of them are costs. Her sixth names what the costs are paid to. A body that has the five without the sixth is running an expensive membership scheme. This is the same structure the non-territorial polities were held by, arrived at through a different record. Five arrangements with no contact between most of them — Benedictine monasticism, the Buddhist monastic code, the Ottoman communal courts, the Sikh assembly at Amritsar, the Hanseatic League — ran on a written rule nobody present had authored, a sanction that materially hurt, periodic public re-assent, a binding vow, and property held in trust. Two literatures, two centuries apart, two continents, one finding.
Cost without a creditor is only expense.
Mark the epistemic seam before going further. Sosis and Bressler report a statistical interaction and explain it through costly signalling: expensive requirements screen out free riders, and a supernatural sanction is what makes the requirement credible. That explanation is theirs, offered on their own ground. Their measurement stands; what it shows is read here from Logos rather than from signal theory.
The nineteenth-century data are clean because the subjects are dead. The living cases are messier and they say the same thing.
The Bruderhof is the strongest working instance on earth. Twenty-four communities, roughly 2,900 members, seven countries, a century old and still growing. Members hold no private property, draw no salary, and keep no personal bank accounts. The whole thing is funded by three manufacturing businesses: Community Playthings, which makes wooden furniture for classrooms; Rifton Equipment, which makes mobility devices for disabled people; and Danthonia Designs, which carves signs in Australia. Every one of the five features is present in its strongest form, and the rule is Anabaptist Christianity, which no member conference is empowered to rewrite.
The Amish crossed four hundred thousand in the United States in 2025, growing at two and a half percent a year and up a hundred and thirty percent since 2000.
Hutterite colonies split when they approach a hundred and fifty people and start again. No ecovillage anywhere is within an order of magnitude of these retention rates.
Now the failures, and each one fails at a nameable point.
Auroville is the most credentialled intentional community ever built — founded in 1968, some 3,300 residents from sixty nationalities, a spiritual charter of real depth. Its governing rule was a statute: the Auroville Foundation Act of 1988, passed by a parliament in which no resident sat. In July 2021 a Secretary was appointed to the Foundation’s Governing Board, and within four years the Residents’ Assembly had been reduced to a body that meets. It voted ninety-eight percent against a set of land transactions in June 2024 and the transactions continued. Around two hundred residents have had visas withheld or monthly allowances discontinued. A rule authored by a sovereign is a rule that sovereign can operate.
Findhorn failed at the opposite pole. Founded in 1962, it had no rule at all in the relevant sense: no vow, no sanction that hurt, no cost to leaving. It ran for six decades on shared sensibility and on the income from teaching that sensibility to visitors. In 2023 the Foundation closed its educational programmes, declared its model no longer financially viable, and put around fifty posts at risk. The houses are still there and people still live in them. What dissolved was the transmission, which was the only thing the Foundation was.
Damanhur in the Piedmont is the most ambitious attempt of the three and it isolates the variable most precisely. Founded in 1975, it has its own constitution, its own currency called the Credito, a tiered citizenship running from full residents down to distant affiliates, and the Temples of Humankind excavated by hand into a mountainside. Every costly requirement the literature counts is present. The rule, however, was written by Oberto Airaudi, who died in 2013, and a rule one living man composed is a rule that has to survive that man. The succession problem is the shape of what Damanhur built.
One case in the record refuses the reading above, and it has to be taken in its own terms before it is answered.
Twin Oaks in Virginia was founded in 1967 and is still running. Around a hundred adults and seventeen children share all income. They make tofu, index books, pack seed for a neighbouring seed company, and until a 2024 fire they wove hammocks, all of it turning over something near six hundred thousand dollars a year. There is no creed. There is no vow. Anyone may leave on any morning and some do. Fifty-nine years, none of the five features, still there. The kibbutz movement is the same shape at national scale: roughly a century of collective agriculture built on a secular ideology of real force, followed after 2000 by a wave of privatisation that introduced differential wages across most of the movement. A hundred years is not nothing.
So the honest statement of the finding has to narrow. Communities without a received rule can persist. What they have not done is transmit.
Seventeen children among a hundred adults is the number to hold against a population doubling every twenty years. Twin Oaks continues by intake: a stream of arrivals replaces a stream of departures, and the institution stays roughly the same size while almost none of the same people are in it.
The kibbutzim reached their second and third generations and then rewrote the terms those generations had been raised under.
Both survived as institutions. Neither reproduced as a people.
Neither literature can see this distinction, because both of them measure years. Kanter’s twenty-five-year threshold and Sosis’s commune-years both count how long an address stays occupied.
A boarding house passes that test.
The strongest apparent refutation comes from the group that did everything right.
The Shakers had all five features at maximum. A rule received through revelation and not open to member amendment. Confession of sins and consecration of all property at entry. Communal worship on a calendar. A covenant with real weight. Property held wholly by the society. At their height around 1840 they numbered some five thousand members across eighteen villages, built to a standard of craftsmanship that still sets prices at auction two centuries later.
In August 2025 the community at Sabbathday Lake in Maine consisted of three people: Brother Arnold Hadd, Sister June Carpenter, and April Baxter, who had arrived that month and was not yet a full member.
The missing element is not obscure. The Shakers were celibate, and their numbers came from adult conversion, from adopted orphans, and from indentured children who could choose at twenty-one whether to sign the covenant or walk.
When the revivals stopped and the states took over the placement of orphans, the intake stopped, and a body with no births has no second source.
That makes the Shakers the cleanest confirmation available rather than a refutation. Costly requirements under a received rule bought them a hundred and fifty years, which is six times the religious median and thirty times the secular one. They bought duration. Continuity is a separate purchase, and it is made in children.
Reproduction is therefore the sixth feature, and it is not derivable from the other five.
A community can hold all five and end.
The serious objection comes from the historian who has spent longest inside this record.
Donald Pitzer’s developmental communalism holds that communal organisation is a stage in a movement’s life rather than its permanent form, and that treating longevity as the measure of success mistakes one phase for the whole. The Oneida Community dissolved its communal system in 1881 and reorganised as a joint-stock company that became a silverware manufacturer of some standing. The Amana Colonies voted the Great Change on the first of June 1932, keeping the church as one body and creating Amana Society, Inc. for the businesses as another. On the longevity metric both are failures. On Pitzer’s reading both are graduations, and the communal phase did exactly what it was for.
Grant the correction, because it is a real one, and then look at what each case actually kept.
Oneida kept the factory. Amana kept the factory and, in a separate legal person, the church — which is to say that the Great Change consisted precisely of severing the rule from the property. What graduated in both instances was the enterprise. What ended in both cases was the community, and calling its ending a developmental stage does not put anybody back at the table. A community that becomes a firm has kept its assets and lost its rule. That is a real outcome, sometimes a prosperous one, and it is what the movement was avoiding.
The finding generalises past communes, and it has already been made once in another domain.
Sacred Architecture argues that traditional building produced beauty as a by-product of following patterns that carried accumulated correction, and that modernism’s decisive move was not a change of taste but the replacement of pattern with proposition — the individual designer’s concept, built at scale, with the correction loop removed. The pattern is slow and cannot explain itself, and those are its strengths, because it cannot get far ahead of what has been tested. The proposition can be brilliant and can be wrong in ways nobody discovers until the buildings are full of people.
A founding charter is a proposition. A Rule is a pattern.
This is why the interaction in the commune data is not a sociological curiosity. A demand you agreed to is a demand you can un-agree to, and every member knows the address of the person who wrote it.
A demand received is not yours to amend, which means the cost it imposes is a cost against something, and the something is what the community is organised to stay in contact with.
Logos is not selected by whoever noticed it and does not become different by majority agreement. Dharma is alignment with that order, and a community under Dharma is not a group that agreed on a value.
Harmonism already names what the secular communes were building. Voluntary Association and the Self-Liquidating Bond articulates three forms of bond: the perpetual, the continuous, and the voluntary self-liquidating — freely entered for the duration of a specific work, dissolving without harm when the work completes, its integrity residing in its readiness to dissolve. That form is correct for a working circle and it is what Twin Oaks has been running at settlement scale for fifty-nine years, which is why Twin Oaks behaves exactly as the form predicts: healthy, productive, admirable in operation, and constitutionally unable to transmit, because a bond whose integrity is its dissolution-readiness is not a bond children inherit.
Communities that held were built on the perpetual bond instead, extended past the household to the whole body.
The Architecture of Harmony orders its eleven peripheral pillars ground-up, and the record here is a check on the ordering.
Kinship sits third, among the foundational substrates, and Governance sits sixth, in political life. Every failure catalogued above failed at Kinship or at the centre, and none of them failed at Governance. Auroville’s governance design was sophisticated and lost to an authored charter. Findhorn had decades of consensus practice and lost to an absent rule. The Shakers governed themselves impeccably for a century and a half and lost to a birth rate of zero. Damanhur’s constitutional machinery is more elaborate than most nation-states of comparable population and its exposure is a single deceased author. Living successes confirm the same ordering from the other side. Bruderhof governance is unremarkable and its Kinship is total. Amish governance is thin to the point of invisibility outside the congregation, and Amish continuity is the strongest documented in the industrialised world.
Communities are decided at the centre and at the substrate.
Governance is downstream, and a movement that spends its design effort there is decorating a floor it has not laid.
Six conditions, and none of them is a preference for rural living.
A rule received. The founding document says what the community has come to see rather than what it has agreed to prefer. This is the whole difference and every other condition is downstream of it. Whether the members like the rule is not the test. Whether they are in a position to amend it, and know that they are not, is.
Entry that costs. Not a fee, which filters for wealth alone, but a real handing-over — property, an oath, a term of service. The record is unambiguous that cheap entry and cheap exit are one design flaw seen from two sides.
A calendar. Assent renewed publicly at fixed intervals, performed by the whole body together. Every durable order in the record had this. Almost no contemporary community does.
Property in trust. Held by the body for its purpose, so that dissolution destroys value rather than distributing it. A treasury that can be divided among the members at wind-up has already told you what the members are.
Land held outright. Auroville’s charter came from a legislature and Próspera’s jurisdiction came from a Honduran statute repealed in 2022; physical anchoring is not a phase a community graduates past. Title, not tolerance.
Children. The sixth is the one no framework in this literature names as structural, and the Shaker villages are the argument for it. A community that can only recruit is a community with one generation of runway, however well it is built.
And the honest limit, stated plainly, because the six conditions have a price and the price is the point. None of this is available to a body that wants leaving to be easy. Every community still standing in this record made departure costly, and a movement that wants the durability without the cost is asking the record for something the record does not contain.
Sacred Architecture ends on a list. Wherever a civilization replaces inherited pattern with individual proposition and removes the loop that used to correct it, the output degrades in ways nobody intends and everyone can feel — in a street, in a diet, in a curriculum, in a medical protocol. The list is missing an item, and it is the one that carries all the others. A people is also a pattern, transmitted or not transmitted, corrected across generations or authored once by whoever called the founding meeting. Every commune in the record was an experiment in which version was being run, and the results are in.
Buildings are the case where the answer is visible from across the road. Communities are the case where it is visible in the third generation, which is why so few of the people who start them ever see it.
See also: Architecture of Harmony, Sacred Architecture, The Network State as Harmonist Order, Community, The New Acre, Voluntary Association and the Self-Liquidating Bond, The Harmonic Civilization, Evolutive Governance