I.0 · What Power Is Made Of

Taken here: the monolith. Given back: the supply chain. Reads as: naivety.

Power looks like a substance. It appears to be held, the way a rock is held, by people who have it and not by people who do not. This is the single most useful error anyone ever made on your behalf, and every institution you have dealt with depends on your continuing to make it.

It is not a substance. It is a delivery. It arrives daily, from you and from several million other people, in the form of turning up, filling in, signing off, handing over, and not saying anything. It has to arrive again tomorrow or there is nothing there. No ruler in history has produced their own power; every one of them has been supplied with it, and the supply has always come from the governed.

This is not encouragement. It is a logistics observation. If power were a substance, the correct response would be to acquire more of it than the other side, which is a contest you lose. Since it is a delivery, the question becomes narrower and much more answerable: what exactly is being delivered, by whom, and what would happen if some of it stopped.


I.1 · The six things they need supplied

Strip an apparatus down and six inputs are required. Every one of them is produced by somebody who could, in principle, stop.

InputWhat it isWho supplies it
AuthorityThe belief that the thing has the right to decideeveryone who does not laugh
PeopleBodies who obey, in numberstaff, members, the enrolled
SkillsThe specific competences it cannot do withouttechnicians, clerks, drivers, coders, nurses
IntangiblesHabit, deference, the feeling that this is how things areculture, and your own reflexes
MaterialMoney, property, transport, communications, the buildingpayers, owners, workers
SanctionsThe capacity to punish, and the willingness to carry it outpolice, courts, managers, and their consciences

Not one of those cells says the ruler.

The practical consequence is that resistance is not a matter of confronting the apparatus at its strongest point. It is a matter of identifying which of the six is thinnest, who supplies it, and what it would take for enough of them to stop. The pillar that falls is almost never the one that looked weakest from outside; it is the one whose people were closest to being asked properly.


I.2 · Why withdrawal is harder than it sounds

Withdrawal distributes the cost; it does not make the cost disappear. One clerk who stops may be dismissed. Four hundred clerks who stop may force a decision. They may also receive four hundred letters explaining that each case has been considered individually. The problem is how to make withholding effective while supporting the people who bear its costs. Parts II through VII work on that problem. Headcount alone does not solve it.

The supply is habitual, not decided. Most of what is delivered each day is not consented to; it is simply done, by people who have never been asked and would find the question strange. Habit is the largest of the six inputs and the least discussed, because there is nobody to negotiate with.

The apparatus can outlive its authors. An arrangement assembles from decisions made at different times, by people with different powers, and comes to be presented as something nobody is authorized to change. Each office can explain the part that belongs to the next office. That does not mean nobody benefits, nobody decides, or nobody is responsible. It means responsibility has acquired a forwarding address. You can raise a concern. Raising a concern is a process.


I.3 · What the other side is made of

Two halves, and the distinction decides where effort goes.

The material half is the part that would still be there if everyone stopped believing in it: the granary, the enclosure, the ledger, the server, the payroll, the lease. Control of the means of survival is the oldest instrument on record and the least improved upon, because it does not require anyone to agree with anything.

The story half is the part that exists only in heads: the outside threat, the permanent emergency, the golden past being restored, the single undivided people whose will is being expressed, the dissent that is reframed as a security matter. Each of these is a device for closing the range of thinkable options, and each is cheap to run compared to the material half, which is why any apparatus under pressure produces more of them.

The error worth avoiding is attacking only the story. It is the more visible half, the more offensive half, and the half that generates the most satisfying arguments. It is also the half that regenerates overnight. A change of vocabulary is not a victory.


I.4 · The commitment

One thing has to be said plainly now, because everything after it depends on the commitment and none of it works without.

Nothing in this handbook is a method for hurting anyone.

Not as squeamishness. Three reasons, and they are not equally strong. A reader is owed the difference, because two of them hold whatever the evidence turns out to say and one of them does not.

Some of it cannot be taken back. A person who is injured is injured. A person who is killed stays dead. A conviction follows somebody into every job application for twenty years. This requires no dataset and wins no argument about success rates; it is simply the first thing true about harm, and it is the firmest ground the commitment has.

Somebody else pays for it. The second is about who decides. One person’s escalation does not stay with that person. It arrives at whoever is standing nearby, whoever is arrested afterwards, whoever has a visa or a custody arrangement or a supply of medication that does not survive a night in a cell. None of them chose it, and the strength of the feeling behind the choice does not make it theirs.

And it makes the argument, though less supported than one might wish that non-harm is the better way. The apparatus of the six inputs is vulnerable in a particular way: when a disciplined refusal meets a disproportionate response, the disproportion is visible, and visibility can move the people in the right-hand column of that table — the ones who supply the sanctions, and their consciences.

That is a mechanism. It is not the only one and it must not be allowed to become the only one. Withdrawal of cooperation, disruption of delivery, the arrival of people who were not previously involved, the quiet shifting of institutional support — none of those need anybody to be hurt first. A handbook that made harm the engine of its own argument would be describing something other than §I.1, and would be recommending something worse.

Stated at its actual strength: where there is footage of your side doing damage, the asymmetry is harder to read and the situation converts more easily from a problem the apparatus cannot solve into one it is staffed, funded, trained and popularly supported to solve. How much harder, and how much more easily, depends on who is watching and what they were already inclined to think. Part V §V.7 sets out how contested that ground is. What holds without qualification is that the conversion is far cheaper to perform than to reverse, that it is usually performed in an afternoon by whoever is angriest, and that the people who pay for it are not usually the ones who chose it.


One more thing, and it should not be left to be inferred.

None of this is a promise that discipline will keep you safe. It will not.

The best evidence available finds no measurable difference in fatality rates between people who kept strictly to nonviolent methods and people who did not, once a state had decided on severe repression. Those records are drawn from cases already known to have gone badly, so they cannot tell you how often it goes that way — but they can tell you that behaving well is not what decides it. The states in them did not require an accurate account of anyone’s conduct in order to act, and Part VI §VI.3 sets out what is done with the outrage afterwards. It is managed. There are five standard ways and they are used because they work.

Backfire is a political consequence. It is not a shield. It does not prevent the injury, reverse it, or compensate the person who took it, and a movement that speaks about it as though it did is spending people it has not asked.

Anyone deciding whether to be in a road, a building or a room should be told this in these words before they decide, and should be able to decide differently later without owing anyone an explanation.


I.5 · Why this is a manual and not a manifesto

A manifesto tells you what should be the case. There are at least two in the Adamic corpus, so the author’s objection cannot be mistaken for abstinence. The difficulty begins when everybody agrees and the cupboard is still locked.

A manual tells you what to do on Tuesday. It is a lesser document and a more useful one. It can be wrong in ways a manifesto cannot, because it makes claims that fail visibly, and this one is written to be checkable: every part ends with the finding that would show it was mistaken.

What follows is a working tongue, twenty rules, a curated set of methods, an account of how movements are played, and a loop for finding out whether any of it is working. None of it is original. Most of it is four hundred years old and the rest is forty thousand.

What this assumes

Every method in Part V assumes something this handbook has not said until now.

A records request works where records have to be produced. A written-basis request works where reasons have to be given. Attributed minutes matter where an attributable decision carries weight, and work-to-rule bites where somebody has to answer for the work not being done. Take those conditions away and Band A is a list of ways to be ignored politely.

So, plainly: this is a handbook for resisting inside institutions where records, obligations, scrutiny and routes of challenge still have some practical force. Not all of it. Not reliably, and not fairly. Some.

That condition is local and it is not established by the name of the country. It differs between two institutions on the same street, and between two people inside one of them. The colleague whose right to remain depends on their employment and the colleague with a pension and a union are not standing in the same place, and the second one should be slower than they usually are to explain the procedure to the first. Whether the condition holds is worked out where the work is, by the people doing it, before the methods are chosen.

One further limit, because Part V cites research. What that research studied was campaigns to bring down governments and end occupations. It can tell you something about how those worked. It cannot tell you the odds on a records request, and anybody who quotes you a percentage for one has made it up.

FALSIFIER. If a regime can be shown to have retained power for a decade while a decisive majority of the suppliers of all six inputs actively withheld them, consent theory is wrong and this handbook has no foundation.