Chapter 11. Making decisions

A working draft, November 2026. This book is being written while The Grove is being built; expect it to change. Checked against the build as it stood on 1 November 2026.

In this chapter

• What the boss's call, majority voting and consensus each cost
• How consent and the advice process combine, with clocks in place of meetings
• How a proposal moves through shape, advice, the vote, a timelock and its review
• Why there's no expert veto or override, and what happens when a forum is stuck
• What people can and can't do when the building's on fire

Say the word consent to someone who has spent a few years in a collaborative workplace and watch their face.

They're picturing a circle of chairs. A proposal that took three weeks to write. A round where everyone speaks, then another round, then a break, then someone who missed the last meeting asks to start again. A decision that finally lands in the fourth session, worn down until nobody objects because nobody much cares any more. Or, just as often, no decision at all: the item rolls over to next month's agenda, and the month after, until the problem fades or the person who raised it leaves.

The sceptic's conclusion is a fair one. Either consent means endless meetings, or it means nothing gets decided. Probably both. Far better, surely, to let someone with authority make the call and get on with the work.

What the usual ways cost

The boss making the call has real strengths. It's quick. Someone is clearly accountable. Nobody has to sit in a circle.

But a decision made at the top is made far from where the information lives. The person who notices that the onboarding process loses a third of new clients is rarely the person who can change it. They can mention it, write it up, hope it reaches the right desk. Most people try that two or three times. After that they stop, and the organisation loses what they could see. Changes arrive by announcement, so the people who live with them had no hand in them. And once a decision is made, nobody has much reason to check whether it worked, because finding out it didn't would cost the person who made it.

The alternatives most groups reach for carry their own costs. Majority voting is fast, but it leaves a losing side who go along without owning the result. Consensus, where everyone has to agree, gives any single person a way to hold everyone else in place, often over a preference rather than a real problem. It rewards the loudest and the most patient. The tired go quiet. The people who care most about the work get worn down by the process of changing it.

For the business, all of these end in the same place: decisions that are slow, or wrong, or never checked. For the people, they end in a quiet lesson that noticing things isn't worth the trouble.

A different question

In the 1970s, a Dutch engineer named Gerard Endenburg was running his family's electrical engineering company and looking for a way to govern it without the usual fights between owners and workers. Drawing on the work of the educator Kees Boeke, he settled on a small change of question that turned out to matter a great deal.¹ Instead of asking does everyone agree?, the group asks does anyone have a reasoned objection?

That's consent. It isn't everyone saying yes. It's nobody being able to name a harm. A proposal doesn't have to be anyone's favourite. It has to be good enough to try and safe enough that the group can learn from it. Teachers of sociocracy have spent decades refining how that works in practice.²

A second strand came from a different direction. Frederic Laloux, studying organisations that had done away with bosses, described what he called the advice process: anyone can make a decision, as long as they first seek advice from the people it affects and the people who know the area well.³ Advice isn't permission. It's there to be listened to, not obeyed.

Put those together and something shifts. Anyone can raise a tension, and anyone with a seat where a change is decided can propose it. Advice is open and wide. Consent is narrow and reasoned. And there's a third ingredient that neither strand needed to name, because both grew up in rooms full of people: time. Meetings exist largely because coordination used to need everyone present at once. If a clock can do that job instead, the circle of chairs becomes optional.

How a change happens

In an organisation on The Grove, every change to the structure is a proposal: a new part of the organisation, a changed purpose, a seat filled or handed on. Stepping down from a seat needs no proposal. It's a right. And every proposal starts from a tension, a sensed gap between how things are and how they could be.

Anyone holding a seat in the forum that will hear the change can raise one. So can whoever holds the branch a piece of work sits on, raising it from that work. Anyone at all, seat or not, can surface a tension through a sensing in the Field or through feedback, and someone with a seat can take it up.

Say you hold a seat, and you've noticed that client questions keep landing in the wrong part of your organisation and bouncing back. You draft a proposal. It carries the tension in your words. It carries the change itself, written so precisely that the system can enact exactly what it says. It says what you expect to happen, in plain words and as up to four of the organisation's own measures, so the review has something real to look at later. It carries what the idea stands on: the work or evidence behind it. And it sets its own periods for advice and review.

While you draft, Wise-Eye can read it and offer what's missing: who would answer for the work, a measure, what would show it worked, what it would cost, a first step. If you shape it together in conversation, it asks lightly who this touches and whether they've been heard. You can take its offers, or not. It's a prompt, never a gate.

Then the proposal moves through its stages, each one on a clock.

Shape. Another person in the forum that will decide it, never you, checks that it's clear and that it does what its words say. Not whether it's a good idea. Just whether it's honest about itself. (A forum is the group of people who govern a part of the organisation: for a branch, the stewards of its parent and its siblings, with its own steward taking part, as Chapter 9 described.) If they think it isn't clear, they send it back and say what it needs. You can rework it, or stand by it, in which case the next forum up checks it instead. If nobody looks at it for three days, it rises to the next forum up. If nobody agrees its shape in time, it lapses. A founder working alone has nobody else to check, so their proposals pass this stage at once.

Advice. The proposal is put to two kinds of people: those with skill in what it touches, and those it directly changes. A change to the purpose, or to the agreement about how we're here together, is put to every member. This lasts between three and fourteen days, or until everyone asked has answered. Advice never holds a proposal up. The one exception is someone who depends on a locked asset the proposal would move; their word does carry weight.

The vote. The people who decide either consent or raise a concern. It passes when those consenting carry enough of the forum's trust and no concern stands. How much is enough depends on how far the change reaches: a fifth, two fifths or two thirds. The vote stays open for two days so everyone deciding can see it, unless they've all consented sooner. A concern holds the proposal until it's met, either by reworking the proposal or by the person lifting it. A concern has to say what must be answered before you could consent. We ask that it names a harm, not a preference: "I'd prefer it another way" isn't a concern. For now the system checks only that something is written. Nobody's trust is ever counted for them. But silence isn't a concern: once the two days are up, a change can pass without everyone's word. And if a concern is still standing when time runs out, ten days from its latest version, the proposal lapses: things stay as they are, the concern stays on the record, and the tension behind it is still shown.

Made. High-impact changes wait out a timelock first, three days, with a circuit breaker: a concern raised in that time stops it. So there's a pause before anything that can't easily be undone. When one proposal carries several changes, they're made together or not at all.

Reviewed. When its day comes, four weeks later by default and open for a week, everyone who took part is asked: did it land, what did we learn, should we keep it, change it or stop it? Evidence from the organisation's own records sits alongside, including the measures the proposal named at the start. If the review recommends a change, that becomes work for the forum. The system never raises a proposal of its own.

Any rework at any stage resets the proposal: it's checked again and advice is asked again. A rework that only answers the advice it was given goes back, for a day, to those who gave it. What people consent to is what they were advised on. There's no room for a bait and switch.

Who decides

The people who decide are the governors of the forum that holds the change, which is the forum around the branch's parent: the stewards of the parent and of the branch's siblings, with the branch's own steward taking part as a person. Whoever holds the changed branch's seat decides alongside the others, as a person, not from outside. If that forum has fewer than two people in it, the decision moves up until one has. A founder alone, or a founder whose members hold no seat, decides alone. Their voices are weighted by the trust their peers have placed in them for that kind of matter, which Chapter 12 explores. The organisation-wide pool of skill advises; the local forum decides.

Wise-Eye takes part all the way through, but never decides. It gives the checker a note on whether the change matches its words. It offers its own advice, clearly marked as its own. If you ask, it will rework your proposal to fit the advice you received. And at review, it summarises what happened alongside the evidence.

There is no expert veto, and no safety override. That's deliberate, and it's the part people push back on hardest. The reasoning is developmental: an organisation that lets a call be made, go wrong and correct itself learns more than one that's protected from ever trying. The backstops are the ones already described: the consent of the people directly affected, the review that comes round whatever happens, and trust, which peers can move towards whoever keeps turning out to be right.

When it's stuck

For the rare case where a forum is evenly split, the design includes a jury drawn by lot. It's a last resort, used only when reshaping has been tried, time has run out, and leaving things as they are would itself do harm. The jury doesn't rule on whether the proposal is good. It rules only on whether an objection is legitimate. Apathy never goes to a jury: if people simply haven't engaged, the proposal lapses. The jury isn't built yet.

Tensions that don't go away

A lapsed proposal doesn't make its tension disappear. Organisations on The Grove show tensions of two kinds side by side. Some the work reveals by itself: work bouncing between branches, work done from the wrong place, work nobody holds, overlaps such as the same outside person turning up in two branches' work. Others are raised: the tension behind a proposal that lapsed, was withdrawn, or whose review said it didn't work.

Since 1 November 2026, an unresolved tension rises to the next forum up after fourteen days, and on up a forum at a time. If it reaches the root, it says so plainly: the whole organisation hasn't answered this yet.

None of this needs a meeting. It runs asynchronously, on a clock, in the places people already work. People can still meet if they want to. Nothing waits for them to.

Much of this hasn't been tried with real people yet, and the lengths of the clocks are our best first guesses.

And when the building's on fire?

Anyone who has run anything asks it within minutes. The roof is leaking onto the stock. A client's details have gone somewhere they shouldn't. A key supplier folded overnight. Who grabs the wheel?

In an organisation on The Grove, nobody grabs the wheel of the organisation itself. There is no administrator, no override, no emergency path around the stages above. But most of what a fire needs was never a change to the structure. It's work, and work moves without a proposal.

Here is what someone can do today. Whoever holds a branch can act on its work at once: take it, answer it, ring whoever needs ringing. Work that comes in marked as unable to wait goes to the top of what its holder sees and into their day. Anything at high stakes is held for a person, with a careful draft from Wise-Eye, which never sends it alone. Money never moves through the system at all. It stays in the organisation's own bank, so someone who has to pay a plumber at midnight simply pays, and records it after. With a receipt, and under any limit the organisation has set for one person's word, it counts on their word, open to question by any member for a week. The books never refuse a payment for lack of funds; whatever the budget can't cover shows as owed. And when work is at risk of missing its day, whoever holds the branch above can step in and take it, even out of someone's hands, or approve an answer that was waiting on one person who isn't there.

What they can't do is just as plain. Nobody can change the structure in a hurry. A change passes only as fast as everyone involved answers: a check, advice that closes once everyone asked has replied, a vote that closes once every decider has consented. A high-impact change then still waits out its three-day timelock. The circuit breaker on that timelock is a brake, never an accelerator: a concern can stop a change in those three days, but nothing can hurry one through. Nobody can be taken out of a seat without a proposal. And nothing is pushed. No page, no alert, no phone ringing at three in the morning. If the fire starts while everyone's asleep, it waits in the place they'll look first when they wake.

That's a real cost, and we don't pretend otherwise. In September we sketched a break-glass mode: one named person given command over one area's work for a few hours, easy for others to cancel, unable to touch the purpose, the money or anyone's right to leave, and owing a full account afterwards. We haven't built it. Part of it, sending the commander's work to people's phones, sits against a later decision that nothing goes out to a member. Whether to build the rest is open.

Why live without it for now? Because an override built for emergencies tends to become the way things are really decided. Whoever holds the key gets used to turning it, and everyone else learns to wait for them. What we kept instead is a bet: that most fires are work rather than structure, and that the people nearest the work are best placed to act. Once the smoke clears, any change to the structure goes the ordinary way, and its review asks whether it helped. An emergency that keeps coming back is usually a missing branch, and the answer is to grow one.

How we decided

We weighed majority voting, consensus and consent, and kept weighted consent. Majority voting leaves losers. Consensus hands everyone a veto over preferences. Consent asks for reasons.

We split advice from deciding. Advice is open and routed to whoever has the skill, wherever they sit. Deciding is local and belongs to the forum that holds the change. We put it to ourselves as: the global field advises; the local field decides.

We dropped an older rule that kept a branch's own people out of decisions about their own branch. It had been meant to stop people protecting their patch. In the end we removed it for the same reason we refused an expert veto: people learn by deciding about the things they live with, and by living with the results.

We narrowed who can block, twice. First to the people directly affected, and then further, to the deciders at the vote and to people who depend on a locked asset. Everyone else advises, and advice is heard.

The jury started life as a more general tool. In October we saw how casually it had crept into the design, and narrowed it to the last resort described above. The idea of unresolved tensions rising through the organisation came from outside voices in November, people who pointed out that a lapsed proposal could quietly bury a real problem.

What's still uncertain: whether peers really will move their trust towards good judgement fast enough to stand in for an expert veto, and whether a standing concern will feel like accountability or like obstruction to the people around it. We think the review and the rising tensions will keep things honest. We'll know when people use it.

What you might take from this

Consent doesn't have to mean meetings, and it doesn't have to mean paralysis. Put a clock on each stage, keep advice wide and blocking narrow, and require that every objection names a harm, and decisions start to move.

Anyone can raise what needs changing, and nothing changes without surviving scrutiny.

Notes and further reading

For a different route to the same idea, Holacracy's integrative decision-making, which tests objections against whether they would cause harm: Robertson, B. J. (2015), Holacracy: The New Management System for a Rapidly Changing World, Henry Holt. On synergy and co-creation: Hubbard, B. M. (1998), Conscious Evolution: Awakening the Power of Our Social Potential, New World Library.