Conflict & Repair

The five conflicts that actually recur, the power nobody voted for, and the four questions a repair process has to answer.

Every community will tell you it takes conflict seriously. The way to find out is to ask when its process was last used.

The five that actually recur

Money. Rarely large sums. A special levy, a member who is behind, whether the group should have bought the expensive one. Money disputes are seldom about money; they are about whether a decision felt made with you or at you.

Noise. Comically small and genuinely corrosive. The workshop at eight on a Sunday, the dog, the band practice, the child who has discovered a recorder. These escalate because both parties are entirely reasonable and neither can concede without giving up something ordinary.

Work contribution. Who is doing their share, who decides what counts, and whether the invisible work appears on any rota. Most likely of the five to be about something else — usually a feeling of not being valued, expressed as an argument about hours.

Parenting. Whose child, whose rules, whose business. Sharpened by everyone involved believing the matter is obvious, and disagreeing about what is obvious.

Romance and its ending. People who live near each other form relationships, and some end. When they do, both parties still have a vote, still eat in the same room, and still sit on the same circle. Communities are often mortified to discuss this in advance, then find that one break-up has reorganised the social map.

Notice what is not on that list: ideology, vision, founding values. Those produce arguments in year one. What produces conflict in year eight is the ordinary friction of living close to people.

Power nobody voted for

The most-quoted text here is not from the communities movement at all. Jo Freeman’s “The Tyranny of Structurelessness”, written for the women’s liberation movement in 1970 and published in 1972, argues that a structureless group is impossible. Every group develops a structure; the only question is whether it is visible.

Where structure is informal it forms around friendship networks and access to information, producing an elite that is unaccountable precisely because the group denies it exists. There is no way to challenge a power that officially is not there.

Informal power in a community is easy to spot once you look:

The founders, whose objection carries a weight nobody wrote down.
The long-timers, who can end a discussion by remembering why a rule exists.
Whoever holds the information — the accounts, the keys, the passwords, how the heating works.
Whoever has the most time, which usually means retired, or without children, or not up at six.

None of these people is doing anything wrong. That is the point.

Freeman’s remedies transfer directly: delegate specific authority to specific people for specific tasks by an agreed procedure; require accountability; rotate the tasks; allocate by ability and interest rather than by who volunteered first; and spread information as widely as possible, because information is power and a community where two people can read the accounts has two people in charge whatever its constitution says.

What a repair process has to contain

Communities collect methods — mediation, restorative circles, nonviolent communication, an external facilitator on retainer. The method matters far less than four structural questions, all of which can be asked in one conversation.

The first row is the one that matters most: a process only the powerful can trigger is unavailable to the person most likely to need it.
Dimension	The question	A weak answer
Who may start it?	The board, or a majority	Any member, alone
What happens first?	It goes to the general meeting	The two people talk, with support
Who holds it?	Whoever everyone trusts	A named role, with an outside option
What can it conclude?	A recommendation	An agreement, a rule change, or a consequence

The last row is where graduated sanctions belong. Elinor Ostrom’s study of long-lived commons found that durable systems have graduated sanctions and cheap, local conflict resolution — a first breach is not treated like a fifth, and getting a dispute heard does not require an ordeal.

What the communities that handle this well seem to have is not a better method. It is a process that changes shape when the parties are unequal.

The moves are unremarkable once named. A facilitator from outside the community rather than the person everyone trusts. Separate ways for the two sides to raise things, so the quieter one is not filing a complaint through a friend of the other. The more powerful party’s authority in the relevant role suspended for the duration, so the person deciding the rota is not also a party to the dispute. More required before a sanction lands, not less. And alongside them, a stated protection from consequences for having raised it at all — because the reason most of these disputes never enter a process is a well-founded guess about what happens afterwards.

Which makes one question worth asking a community directly: what changes about your conflict process when one of the two people is a founder? “Nothing — it’s the same for everyone” sounds like fairness, and is the answer that produces the failure above.

A separate floor sits underneath all of it, and the agreements worth reading say so explicitly. Where there is credible risk to physical, psychological or child safety, protection comes first and the process catches up afterwards. That is not a variation on repair; it is the point at which repair stops being the question.

RCOS · Layer 4

The RCOS Standard does not require the particular measures listed above, and the distinction is worth being exact about. What it requires is that a community define explicit safeguards for conflicts involving power asymmetries, dependency relationships or safety risks (§6.3.1), and that where a power differential exists between the parties, elevated safeguards must be applied (§6.3.3). Which ones is left open: external facilitation, separate channels, temporary suspension of role authority and a higher evidence threshold are offered as options a community may choose from, not a checklist it must satisfy.

That is the standard’s usual shape — it constrains the response level and leaves the technique to the group. The part not left open is the trigger: elevated safeguards follow from the condition being present, not from a judgement that a particular dispute felt lopsided, so nobody has to argue their way into being protected.

Two clauses are unconditional. Anti-retaliation protections are mandatory, covering four acts by name — raising a concern, requesting mediation, giving testimony or evidence, and taking part in a review or appeal (§6.3.2). And safety is written as an ordering rather than an exception: safety safeguards override participation rights, role continuity and operational convenience (§6.3.5). Every item in that list is something a community will be tempted to weigh against safety, which is presumably why they are enumerated.

Safety-critical conflicts get their own requirement rather than the same one: a community must define immediate protective actions that may be taken before the process has finished (§6.3.4) — again a mandatory definition, with the measures themselves a menu.

RCOS Core v0.1 — §6.3 Safeguards

When to leave instead

Repair is not always the right answer, and a topic that only counselled persistence would be doing you a disservice.

Some honest signals: the same conflict has been through the process more than once and returned unchanged; the people who could act are the people the conflict is about; you are being told your problem is your own attitude, by everyone, consistently; or you have stopped saying what you think in meetings and did not notice when you stopped.

Leaving is expensive — financially, socially, and in losing a place you meant to stay in. That cost makes it tempting to keep trying long past the point where trying is doing anything. Knowing the exit terms in advance is partly protection against exactly that, because a decision to stay is only a real decision if leaving is a real option.

Deep · why “we have no hierarchy” is the warning sign

Freeman’s argument deserves stating precisely, because it is usually softened into something much weaker.

The claim is not that informal structure is inferior to formal structure. It is that the choice is not available. Groups produce structure the way they produce norms — automatically, from repeated interaction. Declaring a group structureless does not prevent the structure forming; it forbids the group from describing it. And an undescribed structure cannot be elected, limited to a term, held to a task, or voted out.

The consequence Freeman identified is specific and unpleasant: the informal elite is not chosen for competence or commitment but for proximity to the centre of the friendship network. In her original context that meant women outside the social core were effectively excluded while everyone maintained the group was open to all. In a community it means the same thing, with the additional twist that the excluded person also lives there.

The corollary is the useful part. Formal structure is not the opposite of an egalitarian community — it is the mechanism by which one stays egalitarian. Written domains, rotating roles, published accounts and defined terms of office are all ways of making power visible enough to be shared. A community that resists all of them in the name of trust is choosing an aristocracy of the well-connected, usually without meaning to.

Worth holding alongside a limitation: Freeman was writing about political organising, not households. A group that answers every interpersonal difficulty with a new rule has its own recognisable pathology. The argument is for structure that is visible, not structure that is heavy.

Sources & further reading

1 Conflict, power and repair — the lesson this topic summarises, with the questions to ask on a visit
2 Freeman, J. — The Tyranny of Structurelessness (1970/1972) — the essay itself, free from the author
3 Ostrom, E. — Governing the Commons (1990) — graduated sanctions and accessible conflict resolution
4 RCOS Core — Layer 4: Conflict, Repair & Accountability — conflict classes, the resolution ladder, safeguards against power asymmetry, and proportional sanctions
5 Seeds for Change — free practical guides on facilitation and working through group conflict

Written by EcoHubs members, with AI assistance for drafting and editing, and reviewed by a person before publication. Facts are checked against the sources listed; anything we could not verify is marked. How this is written

Regenerative Community Operating System

Community Suitability Index

VoteCast

Seeking.Community