How Do We Make Decisions Without a Permanent Ruler?

From a ruler to a decision process

From a ruler to a process: who actually decides? In a hierarchical model the answer is often simple: the person at the top decides. That model can be fast, but it creates structural dependence on the judgment, character, and information of one person or a narrow circle. When a position becomes a permanent entitlement to the final word, a community is no longer managing only tasks; it is also managing a concentration of power.

The alternative is not chaos. A better shift is from “who is the ruler?” to “what is the process, who is affected, who may participate, who executes the decision, and how can it be reviewed or corrected?” Authority then becomes a function tied to a task rather than a permanent status.

• a decision should have a clear scope and time horizon;
• participation should track affected interests, responsibility, and legitimate roles;
• execution should be separated from an unlimited right to set the rules;
• the process should include review, correction, and where appropriate recall.

The goal is not that nobody ever leads. The goal is that leadership does not become ownership of other people’s decisions.

Define the decision and the circle of affected people first Many bad group decisions fail before discussion even begins because nobody is clear about what is being decided. One person is debating the goal, another the means, a third the values, and a fourth their distrust of someone in the room. If the object of decision is undefined, no voting rule can repair the confusion.

Before choosing a method, write the question in a form that can be accepted, rejected, or amended. Then identify who is directly affected, who bears costs, who holds special knowledge, and who will be responsible for implementation. This does not mean everyone always has an identical role in every detail; it means differences in role should be justified and visible.

1. What exactly is the decision?
2. Who will experience its consequences?
3. What information is missing?
4. Who can implement the decision?
5. When and by what criteria will we review it?

Deliberation is not the same thing as a decision A community can talk for hours and still not know what it decided. Deliberation and decision are connected, but they are not identical. Deliberation improves understanding of reasons, consequences, and alternatives. A decision rule tells us when the process is mature enough to move into action.

Research on deliberation stresses that its value is not limited to producing agreement. A valuable outcome may be a clearer understanding of where and why disagreement remains. If a community demands visible harmony, it may lose precisely the information carried by a persistent minority.

Good deliberation is not successful because everyone ends with the same view. It is successful when people better understand the problem, one another’s reasons, and the consequences of the available options.

Consensus, consent, and voting

Consensus: when very high common commitment is needed Consensus can make sense for decisions that deeply change a community’s identity, long-term commitments, or foundational rules. Its strength is that it pushes a group to take objections seriously and search for an option that everyone can at least genuinely accept. In small groups, such a process can increase shared understanding and willingness to implement the result.

But consensus has costs. It can be slow, give hidden power to the most persistent blocker, and create pressure for people to suppress legitimate disagreement in the name of harmony. Research on group decision making has found that norms encouraging critical thought can improve decision quality more than norms primarily rewarding the preservation of consensus.

• use it for foundational and difficult-to-reverse decisions;
• define in advance what counts as consensus and what counts as a reservation;
• allow recorded dissent without shaming;
• do not treat silence as genuine agreement.

Consent: good enough and safe enough to try For practical and reversible decisions, consent can be more useful than full consensus: “I do not have a reasoned objection showing that this proposal is currently unsafe or unfit for our shared purpose.” A member may personally prefer another option and still allow a trial.

This is especially useful when a decision can be time-limited and evaluated after a pilot period. Instead of seeking a perfect solution, the group looks for a safe-enough next step. An objection should not become a personal veto without reasons; it should connect to purpose, known risk, a protected right, or an agreed criterion.

When a decision is reversible, a well-designed experiment is often better than an endless search for perfect unity.

Voting: when a clear and timely conclusion is needed Voting is simple, transparent, and useful when a group must choose among clearly defined options within a reasonable time. With two options, majority rule has strong egalitarian properties: each vote counts equally, and movement of votes toward an option cannot perversely move the outcome against it. With more than two options, aggregation becomes much less straightforward.

Condorcet cycles and Arrow’s theorem warn us that there is no magical voting system that satisfies every attractive condition across every possible preference profile. A voting result is therefore not a revealed “will of the community.” It is the output of a particular rule, agenda, and set of options.

The Green Book also offers a useful criticism of plebiscites. Gaddafi argues that a “yes” or “no” can hide reasons, conditions and alternatives, so a binary outcome alone does not express the full political will of participants. R58 does not turn this into a rejection of referendums; it converts the criticism into a procedural rule: on important questions, voting should be preceded by space for reasons, amendments and genuine alternatives. A vote is the conclusion of a process, not a substitute for one.

• allow adequate information and deliberation before the vote;
• state the threshold clearly: simple, absolute, or qualified majority;
• do not use bare majority alone to override foundational rights;
• publish the vote distribution and important minority reasons.

Delegation, sortition, and limited mandates

Delegation: temporary trust, not a transfer of sovereignty It makes little sense for every member to decide every technical detail. Communities therefore delegate: an individual or small group may be trusted to develop a proposal, negotiate, implement, or decide within a limited field. This is division of labor, not necessarily hierarchy.

Healthy delegation needs at least five safeguards: a clear mandate, limited scope, a deadline, reporting, and revocability. Liquid or delegative democracy models add the possibility that people vote directly on some issues and temporarily delegate on others. Experimental work, however, cautions that overdelegation can discard information and concentrate influence.

The Green Book goes further on representation, arguing that no one should permanently exercise sovereignty in place of the people. THY-REALITY does not adopt this as an absolute ban on representatives or delegates; it uses it as a stress test: does the delegate hold a narrow and reviewable mandate, must they report back, can they be recalled, and can authority return to those affected after delegation? When the answer becomes no, delegation starts to slide from division of labour into substitution of the people by the office-holder.

Delegated authority is borrowed authority. If it cannot be reviewed, revoked, or time-limited, it can quickly harden into a position of power.

Sortition: spreading access to temporary roles Sortition, or selection by lot, is useful when several members are suitably qualified, when we want to reduce campaigning for a position, or when a representative mini-public should examine a problem in depth. Modern citizens’ assemblies often combine random selection with demographic stratification.

The OECD has documented hundreds of representative deliberative processes in which randomly selected citizens receive time to learn, hear evidence, and deliberate. But sortition is not automatically fairer than every alternative. It needs a clear purpose, support for participation, protection for vulnerable minorities, and a meaningful connection to actual decisions.

• it can fit advisory groups, oversight panels, and some temporary offices;
• it is not a substitute for competence where a role requires specialist training;
• randomness reduces some forms of capture but does not erase all inequalities of influence.

Limited mandates, rotation, and recall If a function carries significant power, it should have an end point. Limited mandates and rotation reduce the risk that a role merges with a person’s identity and temporary authority becomes private property. Recall gives a community a way to act when the holder clearly violates the mandate.

Term limits also have costs. Research on legislative term limits reports losses of experience and institutional memory in some settings and weaker productivity in others. Good design therefore does not rotate everything at once or discard knowledge. A better combination is time-limited power with durable knowledge transfer.

• a mandate should state its beginning, end, and scope;
• roles can overlap in time so knowledge is transferred;
• records and documentation belong to the community, not the office holder;
• recall should have a clear threshold and safeguards against impulsive personal retaliation.

Minorities, vetoes, and conditions for safe deliberation

Minorities, vetoes, and foundational boundaries A majority does not become morally infallible by being a majority. When a decision touches basic bodily security, dignity, freedom of conscience, or the normative floor established in R57, raw numbers cannot be the only source of legitimacy. The community needs constitutional boundaries: areas where even 90 percent cannot simply vote the remainder into subordination.

At the same time, a universal personal veto lets one member hold the entire group in permanent deadlock. Vetoes therefore make sense only in narrowly defined areas: foundational rights, direct personal danger, or violation of the community’s basic agreement. Ordinary operational decisions need other mechanisms for recording and addressing minority positions.

Protecting a minority does not mean the minority always decides. It means the majority may not cross certain moral and procedural boundaries.

Information, facilitation, and psychological safety The decision rule is only the final centimeter of a process. Quality often depends on whether people receive important information and whether they dare to share it. Research on psychological safety links lower interpersonal risk with voice, learning, error reporting, and collaboration.

A facilitator must therefore not become a hidden ruler. Their job is to protect the process: time, turn-taking, clarity of proposals, the right to object, and equal procedural treatment. If the facilitator sets the substantive agenda, filters unwelcome arguments, or selectively summarizes, the process may be horizontal in form but not in fact.

• make key information available before the decision;
• allow people to ask questions without ridicule or punishment;
• disclose conflicts of interest;
• separate process facilitation from advocacy for a substantive outcome.

Combined processes, emergencies, and digital tools

A combined process: deliberate → propose → test → decide The most robust systems often use a sequence of methods rather than one method. A community can first gather information and deliberate, then ask a small group to formulate a proposal, let the wider group test objections, and use voting for unresolved alternatives. After a defined period, the decision is reviewed again.

1. Define the question — what are we solving, and what are we not?
2. Gather information — facts, experience, costs, risks, and affected interests.
3. Deliberate — actively look for reasons that should change the proposal.
4. Form a proposal — clear, short, and implementable.
5. Choose the rule — consensus, consent, voting, delegation, or another agreed method.
6. Record the decision — who does what, by when, with what resources.
7. Review consequences — is it working, and did it create unexpected harm?

Legitimacy then does not end at the moment of a vote. Decision making becomes a learning cycle.

Fast decisions and emergencies A fire, a broken water line, a violent incident, or an immediate safety threat does not allow two hours of circle discussion. A community without a permanent ruler therefore also needs pre-agreed emergency mandates. In advance, it defines who may act temporarily, within what limits, and how quickly the decision must return for review.

The crucial difference between an emergency mandate and permanent authority is that urgency does not expand power without an end. When the justification for the exception disappears, the exceptional authority ends as well. This preserves speed without turning every crisis into a route toward permanent hierarchy.

In an emergency we delegate speed, not unlimited sovereignty.

Digital tools: assistance, not a substitute for legitimacy Online polls, shared documents, forums, proposal platforms, and delegative voting tools can lower the cost of participation. They enable asynchronous work, traceable amendments, and involvement from people who cannot always attend in person. But a technical ability to click is not democratic legitimacy.

A digital process must answer questions of identity, privacy, access, security, manipulation, algorithmic transparency, and error correction. If a platform invisibly ranks proposals or concentrates delegated votes, it may simply move hierarchy from the meeting room into software.

• the tool should be understandable to nontechnical members;
• important decisions should have a human explanation, not only a numerical result;
• keep a fallback process for people without reliable digital access;
• collect personal data only to the extent needed for the process.

Example: a community decides on a shared budget

Imagine a community of 30 households with a limited fund and three possible projects: repairing a water catchment, creating a shared space for children, or improving a path. If the group begins with “which project is best?”, interests quickly collide. A stronger process first gathers costs, urgency, the number of people affected, and the risks of delay.

The group can then deliberate about phased or combined options. If three legitimate alternatives remain, it may use a previously agreed voting rule. The selected project receives a limited implementation mandate, and a report returns to the community after three months. New information can trigger correction without turning revision into personal defeat.

The goal is not for everyone to receive their first choice. The goal is a process fair and intelligible enough that someone who voted differently can still say: I did not win, but I understand the decision and recognize the process.

Twelve questions for auditing shared decision-making

1. Is it clear what we are deciding?
2. Are people substantially affected by the decision able to participate?
3. Do participants have access to the key information?
4. Can disagreement be expressed safely?
5. Is the decision method known in advance?
6. Does the threshold fit the weight and reversibility of the issue?
7. Are there foundational boundaries a majority cannot vote away?
8. Are delegated powers limited and reviewable?
9. Is responsibility for implementation clear?
10. Does the decision have a review date or correction path?
11. Can a minority record its reasons without punishment?
12. Would we still accept this process if we expected the outcome to go against our first choice?

Exercise: write a one-page “constitution of decision-making”

Write a one-page protocol for your community. Not a legal labyrinth, but a short agreement a new member can understand. It should distinguish types of decisions and state which method applies to each.

• routine operational decisions;
• reversible experiments;
• major financial commitments;
• changes to foundational rules;
• emergency decisions;
• delegated mandates and recall.

Add a rule for recording decisions, a review mechanism, and one minority safeguard. Then test the protocol on a real small decision. If nobody can use it without the author explaining it, it is still too complicated.

Conclusion: authority should be a function, not a permanent status

R58 does not claim that every horizontal group is wise or every leader dangerous. It makes a narrower claim: a community can separate functional leadership from a permanent right to rule. Someone can coordinate a project, facilitate a meeting, or make an emergency call without becoming the group’s permanent ruler.

Good collective decision making is therefore not one method but an architecture: clear powers, appropriate thresholds, voice, limited delegation, foundational boundaries, records, review, and learning. Consensus is not always best. Voting is not always tyranny. Sortition is not always fairer. Delegation is not always hierarchy. Everything depends on design, context, and correction.

A free community is not one in which nobody ever leads. It is one in which no leadership role becomes an untouchable right to rule others.

R59 will build on this foundation by asking the next question: who counts as a member of the community, how does someone enter, what does membership require, and how can it end peacefully?