The Line on the Hull

It began on a road in Croatia, in 1991. I saw nothing then.

I spent eight years in the British Army learning how not to die too quickly, or at least to outlast the people arranging it. I learned where danger sits in physical space: high ground, dead ground where a man can move unseen, the places fire comes from before you hear it. What the Army didn't teach me was how to read people. Nobody offered that course.

It was late autumn 1991. Croatia had been at war for months. I was driving, with a translator beside me. The road near Vukovar should have been safe. There were no reports of fighting there. But the road was empty in a way that should have told me something, the kind of emptiness where the absence of people is itself information. Where have the farmers gone? Why are there no tractors? Why is the village we just passed through utterly still at eleven in the morning? I noticed these things. I registered them. I did not act on them.

I could read terrain. That training still worked. I could look at a junction and see where a position would be set up, where the angles of fire would converge, where the dead ground lay. A ridge either provides cover or it doesn't. A road either has escape routes or it's a trap. The geometry told the truth.

I saw smoke ahead and slowed the car. At a junction sat a vehicle marked PRESS. The word that is meant to keep you alive. A journalist, like me. I stopped and asked the driver if the road ahead was safe.

He barely looked at me. His eyes moved across my face the way you'd glance at a road sign you'd already passed, registering the presence, discarding the content. There was no hostility. Hostility I'd have read correctly. This wasn't that. There was something else. An absence. The place where interest in another person's survival should have been was simply empty. Like looking at a window where a light should be on and finding darkness.

'Yes', he said.

One word. Flat. Uninterested. Not a lie, exactly. Something worse than a lie. A lie at least acknowledges the existence of the person being lied to.

I drove on. Half a mile later, I turned a bend and found the source of the smoke. A burning car with a body lying beside the open passenger door. I braked and threw the car into reverse. That's when the bullets came. One punched a hole through the windscreen, roughly eight inches from my head. Two more went into the body of the car. The gearbox screamed in reverse. We escaped. Nobody was hurt.

I pulled over once we were out of range. My hands wouldn't stop shaking. I could operate a camera in a firefight. I could think clearly under mortar bombardment. But my hands shook in that car and I didn't understand why. I kept coming back to his face. Not the gunfire. His face. The quality of his looking at me and not seeing me. My gut had tightened when he looked at me. My body fired the alarm. My social training switched it off. Be polite. Assume good faith. Don't judge a fellow professional on a feeling you cannot articulate. And I was new to journalism, a freelance when almost no one in television news was, still half in awe of the correspondents who had the staff jobs. Who was I to distrust one of them on a feeling I could not name? I drove on because I trusted words over a gut feeling. He said the road was safe. My gut said something was wrong. I chose the words.

He had either failed to warn me or sent me into it. I never learned which. I never saw him again. But the pattern was the same either way. My survival was irrelevant to his calculation. I was not a person to him. I was an obstacle, or an opportunity or nothing at all.

The road wasn't safe. He didn't care.

For nearly thirty years that was a story about a callous man. It is not. There are four things in it I could not see at the time.

The first is what was missing in him. Not malice. Malice I would have read. What was missing was weight: the information that we might die in the next few minutes arrived, and it moved nothing. That absence has a shape, and once you have seen it you can see it again.

The second is what switched off in me. Not judgement. An alarm far older than the Army, and what switched it off was a word painted on a car. PRESS. It is exactly the kind of sign we are built to trust, and it was on the wrong man. I read the sign and not the face, and I went on reading signs for years. Behind the sign was a belief most of us hold and few of us examine: that there is good in everybody. It is right most of the time. That is what makes it worth using against us.

The third is that he was not rare. I met him again in Bosnia, in Chechnya, in Afghanistan and in Iraq. He turns up in the record too: the aid chief who, after an earthquake, paid for sex in a villa his charity rented. Three of my colleagues were murdered by the man they had hired to take them across a border, who wanted what they were carrying. He had been trusted the way I trusted the sign on the car. The pattern does not kill by itself. It kills through people. I read people for a living for twenty-four years, in places where misreading one could get me killed, and none of that experience reliably stopped me trusting the wrong man. I am the first exhibit. The man at the junction was not the exception in my working life. He was one of a series.

The fourth is what he took. It took me longest to see, because it looks like nothing. A proper answer would have cost him a minute, an explanation and some responsibility for what happened to us next. He spared himself all three. We paid for that on the road ahead, and it could have been our bodies beside the burning car. Nobody would have called it taking. That is the whole pattern at its smallest: one man spares himself a small cost by taking something from people he never has to see, their time, their money or their safety, and they pay for it.

Give the same man more power and nothing changes except how much he takes and how many people pay for it. At the junction the cost fell on the two of us in the car. In a boardroom it falls on thousands of workers and pensioners. At the top of a state it falls on a country. What we fear in a leader is not a different kind of person. It is the man at the junction, with a bigger lever.

A story like that one comes with its own explanation. A few bad people exist, they are rare, our institutions mostly hold them, and when one slips through, a reporter finds him. That account has served us for a lifetime. It is also why we are not looking at where this comes from. We think it is an aberration. It is not. It is the oldest relationship we have, between the few who find taking easy and the many who assume nobody does, and it is a part of us that we choose not to see. You have met the man at the junction many times. You may not have seen him.

In 1993, a year into the Sarajevo siege, I filmed soldiers of the United Nations Protection Force selling black-market goods to the civilians they had been sent to protect. Men on the most visible humanitarian mission on earth were running a trade. The force sent to protect the besieged was extracting from them.

It is an easy story to file: men behaving badly in a bad place. But the blue helmet was the word on the car, at the scale of a city. A whole besieged population trusted the sign, because that is what the sign was for, and the men wearing it were not monsters. They were what any garrison becomes when the sign is trusted and no one outside is watching.

The pattern has names for its parts and none for the whole. This pamphlet gives it one: extraction. Extraction is using other people's trust to draw power, money and attention to yourself, while the cost lands on them. A few people find it easy, because they are little moved by what happens to others. Call them the cold few, and the rest of us the warm many. The pattern is what happens when the cold few find a place where nobody is watching.

We could call those few something worse, narcissists or psychopaths. Those are labels for a clinic, and we are not in one. We are naming what a person does, not what a person is, and this is not a list of who to remove. The cure is not a hunt for people.

The divisions we were handed are real, and some of them do terrible harm: left against right, one colour against another, one faith or sex against the next. They are also worked at, paid for and kept going, because a country arguing with itself is not looking up. And there is a division we are not shown, between the few and the rest of us. The few go on taking from us while we argue: the pension, the cladding that burned, the certificate signed by the party being certified. We have not been failing to look. We have been looking the wrong way.

Freedom is maintained, not won. It is seldom seized all at once. It is taken a piece at a time, and some of it we give away, a convenience at a time.

The misreading

Ask what has gone wrong and half the country will say populism: one man who says only he can fix it, and a crowd that believes him. The other half will say the people who ran things for themselves for years and stopped listening. Both are describing the pattern at its loudest, and it is where we look. But the man on the balcony is only the one in plain view. The same relationship runs through every country at every scale, in the boardroom, the parish council and the hospital ward, and almost none of it makes the news. In each of them someone draws the money, the credit and the attention to himself and leaves the cost with everyone beneath him, and what he does there changes more lives than anything on the screen. The pattern survives by not being noticed, and nothing changes until it is.

We have seen this before. A leader who is the state and the state that is the leader. Officials who compete to agree with him. The treasury run as a private account, the face on the money, the supporters who will hear nothing said against him. The costume changes. The shape does not. In 1941 the world was down to about a dozen democracies. We built them back, and now the tide is going out again. By the end of 2025 autocracies outnumbered democracies, ninety-two to eighty-seven. On the count kept by V-Dem, the Swedish institute that measures the world's democracies, about three in four of us lived under autocracy. And for the first time in over fifty years the United States itself was reclassified, from a liberal democracy to an electoral one. Many Americans fear it will fall further.

I spent a career covering the arrivals, the strongman on the balcony, the tanks, the square. What we almost never covered was the years before, when the institutions that might have stopped him were being quietly softened. By the time we got there with the cameras, the story was mostly over.

So the present gets read as a sudden emergency: a dangerous man has seized a democracy, and he can be beaten at the next election. That is not wrong about the man. It is wrong about where he came from. The protections came down long before he arrived, and by the time he is in power you are looking at the result, not the cause.

That is why the usual remedies bring relief and little change. Win the next election, and the structure that produced him produces his replacement, and it wrote the rules of the election too. Prosecute him, and the prosecutors and the courts have been worked on by the same hands. Wait for the next government to put the institutions back, and it will find little left to put back: the people who ran them gone and the rules rewritten.

The pattern has no politics of its own. Some of its largest modern cases were built by the communist left, Stalin's terror and Ceaușescu's Romania, not by anyone's right wing. It takes whatever colour the room is painted.

We do not merely tolerate the takers. We have always half-admired them. We hang the word 'Great' on the most successful and stand them on columns: Napoleon in Paris, and Alexander, the man who burned his way from Greece to the Indus and whom two nations still claim. The pattern did not only capture our institutions. It captured our sense of greatness.

The cold few have always sought power. The rest of us have always pushed back, not always in time and never permanently, and where we held them, we did it in ways that did not depend on anyone becoming a better person. When the system works, it makes the chair harder for the most dangerous to reach, and it holds whoever sits there. When it fails, who sits in the chair is all anyone talks about, and the talk goes in circles, because people do not change much. Three lines carry the answer. Pattern, not people. Architecture, not character. Contain, not blame.

And it is not only at the top. It is the household where the children learn to read a mood before they can read a book. It is the neighbour the whole street has learned to walk around. It is the manager two floors up whose department loses people and nobody writes down why. None of that reaches the news, and none of it is counted with the rest, because we file it by setting: a family matter, an office problem, a national scandal. Filed that way it never adds up, and the adding up is the point. The answer is not to vote out whoever is in. It is to contain the pattern, and to contain it everywhere, because what is left uncontained lower down is what the man at the top later stands on.

What was built, and what went

The protections that hold a society together were not given. They were built, each after a particular catastrophe, by people who had counted the bodies. Samuel Plimsoll's line was made law in August 1876 and painted on every hull, so that an owner could not load a ship past the point of safety and drown the crew for the insurance. Not a plea for better men, but a line anyone on the dock could see, so the overloading showed before the ship went down.

And the first version failed, which is the sharper lesson. The 1876 Act let the shipowner paint his own mark, and some painted it on the funnel. It took fourteen years to put right, and the fix was not a better class of owner but an independent table of limits anyone could look up, with the line assigned by a body that had no stake in the cargo. The mark only worked once the owner no longer drew it. One question fits everything that carries a signature. Who drew the line?

Fifty years of dismantling does not look like dismantling. It looks like a series of reasonable adjustments. A clause in a finance act. A cut to the inspectorate. A quiet change to what counts as a conflict of interest. Most of them, on their own, are too small to notice. The few that are big enough get past us because we are looking somewhere else. Nobody is in charge of the whole of it. Each removal is cheap, and the next one is cheaper, because the check that would have objected is the thing that went.

What it leaves behind is rot. The train is late and dirty, there is sewage in the water, and two hours on the telephone buys back something already paid for. The rot is not only in the institutions. It is in us. The coldness of the few works its way into the many, and we find ourselves harder, quicker to grab and slower to trust, without ever choosing it. The few do not multiply. The behaviour does. And the rot takes the takers with it in the end, because the floor they stand on is the one they are pulling up.

You do not see the rot until something falls through it. Fifty-seven killed when two trains met head-on at Tempe, in Greece, in 2023, on a line whose new signalling had been certified on paper and did not exist on the track. And most of what it takes never reaches the news. Two per cent a year in fees, the figure on the pension statement too small to notice, compounds over a working life into up to half the pot. While the country watches the daily provocation, the next clause is drafted and the next protection thinned. By the time the noise is loud enough to march against, the guardrail that would have held is already gone.

What we already know

Most people assume the cold few are rare. They are not. Call it about one person in twenty. In a room of twenty that is, on average, one person, and across a working life it means people we have answered to, dealt with and been related to, not a handful of monsters somewhere else. Picture one and most of us picture a dictator, which is like picturing a tsunami when somebody says water. The ones who matter are in the supermarket queue and across the table at work, and one of them cooked us dinner this month and we complimented the food.

Nobody has counted exactly what this pamphlet describes. The clinic counts disorders, and a cold hand on a form is not a disorder, so the clinic's count is only a floor, and the floor moves with the test. At its strictest it is about six people in a thousand. Across the disorders the clinic recognises, it is between three and six in a hundred. Count instead the people who feel little for others, with no diagnosis at all, and the research runs to one in twenty and beyond. One in twenty is a working figure and a cautious one, the best the evidence allows. What it is not is small. If it is wrong, the rest of this pamphlet still stands. None of the cases that follow depends on a number.

What the number counts is not a disease. Cold and warm are two ways of being human, both ancient, running on one piece of evolved hardware. Almost all of us sit in the warm middle. At the cold end are a few who take far more than they give and are moved by little of what others feel. A different setting of the same machine. Our history is the story of containing them through structure, not of treating them as ill.

None of this is one man's theory. Some of the best minds of the last half-century have mapped parts of it, and mapped them well. Elinor Ostrom catalogued the containment that worked, village by village. Daron Acemoglu and James Robinson showed how extractive institutions capture a whole country. Mancur Olson showed why a few with a great deal at stake beat a great many with a little at stake. Ostrom, Acemoglu and Robinson each won a Nobel prize for the work, and the brain science later in these pages is not one laboratory's result. So the knowledge is not missing, and it is not mine. It lies scattered across fields that seldom read one another, and it has never reached the people it is about.

What is new here is the joining up, and one departure. None of those scholars treats the number of takers as constant, and what turns on that is whether the work ever ends. If the takers are a rare few, you fix things by removing them, and vigilance is a campaign that ends. If every generation brings roughly the same share of them again, containment is not a reform you can finish. It is upkeep, like a road or a sea wall, a thing you fund forever or lose.

Those are two different worlds. The mark itself is needed either way. A line drawn by the owner fails whether or not the owner is cold, and most owners are not. How many takers there are decides how long we need the mark, not whether we need it.

The few and the mechanism

Behind all of it are people, and they come in two kinds. A few took pieces out on purpose, though no one designed the whole, and no one could have. The cold few are not a club. They work together for exactly as long as their interests overlap and not a day longer, and the structure they leave behind takes shape the way water finds its way to the bottom of a building, one path of least resistance at a time, with no plan and no need of one. The rest are the maintenance crew: the operators who run what they inherited, sit in every party and would be startled to be told they were part of anything. Most of us have served on that crew somewhere.

Extraction is not a politics. It captures politics of every colour, and the method belongs to no party. It runs wherever a public good meets a private interest: the consultants who bill every government of either colour, the revolving doors of the drug industry and private equity, the defence contracts that go to the same firms whoever wins.

The structure is not one mechanism but many, locked together. The door that revolves between the regulator and the regulated. The inspectorate cut until it cannot inspect. The clause nobody reads. The contract that puts the profit in one place and the loss in another. And the certificate signed by the party being certified, the easiest of them to see: a signature on a form that says a thing is safe when it is not. Take Grenfell Tower, London, in 2017. Seventy-two people died in a fire that climbed nineteen storeys in under twenty minutes. The cladding had been certified to a standard the installed product did not meet. The testing house had been sold off and lived on the fees of the firms it tested, and a developer could shop for its own inspector. The people who signed the papers did not live in the tower. The people who died in it never saw the papers.

Or Boeing's 737 MAX, certified in a process the regulator had largely handed over to Boeing, where a committee of Congress later found a culture of concealment. One employee asked another, in writing: 'Would you put your family on a MAX simulator trained aircraft? I wouldn't.' Somebody in that building could see the passengers. It made no difference to what was signed. Across two crashes, 346 people died. Boeing was trusted to check Boeing.

Britain's Post Office ran an accounting system called Horizon, rolled out to branch counters from 1999. Before it arrived, some of the Fujitsu employees who built it had already found that it could produce data which was false. The public inquiry has since found that a number of Post Office employees, senior and less senior, knew or at the very least should have known the same. The inquiry also found that throughout the lifetime of that first system the Post Office 'maintained the fiction that its data was always accurate'. Of Horizon Online, the version that replaced it in 2010, the finding is flatter: employees of both the Post Office and Fujitsu knew. The High Court later found twenty-nine bugs, errors and defects in the software, and that the first version was 'not remotely robust'. On the strength of its figures the Post Office prosecuted and convicted around a thousand sub-postmasters and their staff for money that had never gone missing. Convictions have since been quashed in their hundreds, many by Act of Parliament. At Boeing the person who could see the passengers was not the person who signed. Here the knowledge and the decision sat inside the same organisation, and the prosecutions went on for years.

Each of these is a line drawn by the owner. The harm needs neither malice nor incompetence. It needs only that one person signs and another carries the consequences, and that the two are kept far enough apart that the one who signs never sees who will pay. At the Post Office the people who knew could see exactly who would pay, and it went on anyway.

Why we go along

A cold few can steer a warm many because of how we are wired. Trust is the technology that built everything: trade, law, cities, medicine, the ability to sit in a metal tube at 35,000 feet eating a bad sandwich. Between two people who trust each other, and inside a group that does, the brain's alarm falls quiet. The amygdala, the brain's watch for threat, is turned down. Katalin Gothard's laboratory has seen it in monkeys, and other laboratories have seen the same calming in people. It is what makes cooperative life possible at all. The laboratory's finding stops there. The next step is mine. Among strangers, at the scale of a country, that trust is fed signals that only look like closeness: the word on the car at the junction, the familiar voice on the screen, the brand in the kitchen for forty years, the candidate who shakes your hand once at a rally. The wiring goes along, because going along is what it evolved to do. We are not fools. Our alarm has been switched off by a copy of the very thing it was built to trust. The vulnerability is not a flaw in the design. It is the design. And the same wiring points to a way out. The brain that learned to go along may learn to refuse, and the equipment for it is already in every head.

The people who follow the strongman hardest are not cold but captured, running the same warm wiring as the rest of us, turned by people who studied how to turn it. Trusting a good con artist is an ordinary human response, and nothing to despise. Nor are the rest of us outside it. We did not see the cold few either. We let the institutions be hollowed out, and we did not notice that nobody was looking after anything. We were the maintenance crew. If his hardest followers have been captured, so have the rest of us, more quietly: by our own small accommodations and the daily looking away. None of that makes us the same as the few who do the taking. It does mean the responsibility for where we are is shared, and that blame has no part in a better future. It only distracts us from building one. We are in this together.

Vertical, not horizontal

Our quarrels with each other need no plan. The few who take from all of us only have to pay for the channel, the campaign and the feed, and the quarrels keep themselves going, with the rest of us pointed at each other. The division that does the work runs the other way, top to bottom: the cooperating many, and the few who feed on all of them. This is not a conspiracy. It is self-interest followed many thousands of times over, each taker, and each operator who goes along, doing what pays, none of them needing to know the others.

The pattern is in front of you too. Not in Croatia. In your own life.

Picture a care home. The carer on the night shift came from Manila on a visa and sends half her wage home. She left because at home a few families have held the money and the power for generations, and the work that pays is abroad. The old man she washes voted to cut immigration, and tells her so, and she is gentle with him anyway. He sold his house to pay for his room. And both are being taken from by the same fund. It bought the home, sold the building and now rents it back, at a price that comes out of her wage and his fees. Neither of them has ever heard its name. The other half of the country is not the enemy. The same structure is taking from them too.

So the discipline is: contain, not blame, and regard the person. Blame is what the divided halves do to each other, and every hour of it is energy we spend defending the thing that takes from us. And when the blame is aimed upward instead, the discipline is the same. The few are real, but hunting them changes nothing: as long as taking still pays, others will step into their place.

The enemy is the pattern, not the people running on it, and our appetite for a villain, a name, a face, is a large part of how the pattern stays unseen. We have an old word for it, evil, and we never stop telling stories about it. The word has never helped, because it looks for monsters, and the man at the junction was not one. Nobody is angry with a shark. You put a net across the beach so it cannot reach the swimmers, and the oldest net we have is the one we call morality.

The watch we let fall

We have held the few before. For most of our time on earth we did it by living close together. In a village there is nowhere to hide a cold streak, so the person who takes more than they give is known, and named, and checked. Settled peoples built the same thing again and again. England had frankpledge, ten households answerable for one another, so that a man who took from his neighbours answered to nine other households who knew his name. If one of them stole a pig and ran, the rest were fined for failing to produce him. That concentrates the mind. Try proposing it at a parish council. China bound households in tens. So did the Inca, on the other side of the world, who had never heard of either. Different peoples, no contact between them, and the same answer to the same problem, which was human nature. They had something we have mislaid, a working knowledge of what a human being is, and they built for it. And they had time. A village could watch a man for thirty years. An institution decides about him in a forty-five-minute interview. We did not lose the eye. We lost the years. Time mattered because the watching shaped what it watched. A cold streak that shows at once gets checked, so the kind that lasts is the kind that looks warm at first: the charming neighbour, the pillar of the parish. It shows over years, not in an interview. That is also why nobody can count the cold few exactly, and why nobody should try to diagnose them across a room.

They had names for the work. The wardmote was the assembly that inspected standards and held defaulters to account, and the ale taster checked the strength and the price of the beer. Both survive in the City of London, the wardmote as a meeting with the local councillors and the ale tasters as ceremony. Nobody marked the day the checking stopped.

None of it was gentle. Frankpledge as the Normans ran it was inspected by the sheriff twice a year and the fines travelled upward. Lords and clergy and men of property were exempt. It bound the people with the least. It worked, and nobody at the top was in it.

When the machines came and the village broke, we built new eyes for the same task: the inspector, the regulator, the free press, the court that could not be sacked. Those new eyes are the watch these pages describe, and they are being put out one at a time. Now we have mislaid the town square as well, broken into a billion private feeds, each a warm-lit room we would rather not leave, and we call it progress because our tools are extraordinary. We did not grow wiser. We grew faster, and we let the watch fall. We kept the parchment, Magna Carta under glass in London and the Constitution under glass in Washington, and let the words that did the work fall out of use. We think we are better than the people who wrote them. In one respect they were better than us: they knew what they were writing against.

What we can build

The people who rebuilt the world after the Second World War had grief. We have knowledge. They built after they had counted the dead. We can build before.

Aviation is the strongest case we have. Boeing's self-certification is the exception, a line drawn by the company it was meant to hold, and the rest of the trade does not work that way. In the late 1950s commercial jet aviation suffered roughly forty fatal accidents per million departures. By the 2020s the rate had fallen by more than 99 per cent, and not by finding better people. Every crash is investigated independently. Pilots report their own errors without fear of punishment. Black boxes keep their own account of the flight and hand it to the investigator, not the airline. The regulator must act on the findings or explain in writing why it has not. Take those protections away and aircraft would still fly. It would be a different trade, and a far more dangerous one.

I have flown under both systems. Into Sarajevo on the airlift we called Maybe Airlines, diving for the runway at the last moment: military flying, and every incident went into a log. And out of a clearing in Mozambique, where our tail rotor had clipped a tree and kept us there for two days. I never heard another word about it.

Iceland applied the same design to its banks after the 2008 crash. It created an independent special prosecutor, convicted dozens of its senior bankers, chief executives among them, and did not collapse a second time. Nobody chose the names in advance. The office went where the record led. And the design holds with no institution at all. Alcoholics Anonymous, with nothing to seize and no leaders to buy, was built so that extraction gets no handhold: nobody in charge, every group paying its own way, no member's name in public. It has held since 1935. The design is the lesson, not the meetings.

None of these holds by itself. Each was built on purpose, and each can be quietly taken apart the moment the watching stops. Containment is a watch, kept up when nothing seems wrong, which is exactly when the pattern counts on no one looking.

The Madison Pledge

If aviation is the proof, the Pledge is where a voter can start. It takes its name from James Madison, the father of the American constitution, who put the whole idea in one sentence: 'If men were angels, no government would be necessary.' Madison saw that enlightened statesmen will not always be at the helm, so the design must not assume they will be. The Madison Pledge is a Plimsoll line for politics: a mark you can see from the shore, past which the loading sinks the ship. It does not ask a candidate to be a better person. It asks for a mark the candidate does not draw alone. Seven commitments, short enough to print on a card, asked of every candidate at every election: empathy, evidence, transparency, accountability, skin in the game, independence, precedent. Three of them read like this. Transparency: say who funds you, publish your evidence and print who paid on the first page. Independence: take no money from any business or lobby whose profits depend on your decisions. Skin in the game: do not exempt yourself from what you impose on others. Each is an act that leaves a record, not a claim that can be polished. The signature is not the mark, because a signature costs nothing. The mark is what the candidate does afterwards, on the record, over a whole term, checked against the card by people the candidate does not control. The Pledge does not depend on who signs. It sets a watch, and anyone can keep it.

Any candidate can ignore the Pledge. Ignoring it is an answer too, and it goes on the record beside the signatures and the refusals. Nobody was ever going to reform the cold few, and the Pledge does not try. They are held by the structures the rest of us keep: the auditor, the inspector, the clerk, the person whose name is on the form. Hold the operators and you hold the machinery the few depend on. The full seven, with what each demands, are set out at the back of this pamphlet.

Britain has already tried the pledge without the mark. After cash-for-questions in the nineties, a committee under Lord Nolan wrote seven principles of public life, the Nolan principles, and they have applied to everyone holding public office ever since. They were in force, and framed on office walls, through the expenses scandal that took members of every party fourteen years later, and through everything after it. The principles were not the problem. The people bound by them were left to judge themselves, and a principle a man holds himself to is a mark he paints on his own hull.

And a pledge needs a keeper. A public register, held by someone with nothing to gain, recording the refusals as plainly as the signatures. The question fits on a doorstep: 'Will you sign this, or will you explain why you will not?' The answers, the refusals and the silences are all information.

The work

No one else is coming to do this. We have looked for the people whose job it is to keep the protections standing, and found only ourselves, the people they protect.

Noticing is not enough. The work is rebuilding what was dismantled in the quiet, and it will not wait. The firms that built the feeds we scroll are now building artificial intelligence, the machine that learns. The choice is whether it is built to extract or to contain, and almost every line drawn so far has been drawn by the firms themselves.

It can look too big to take on. Think of the pattern as knotweed. One root system runs under everything, which is the daunting part and also the hopeful one, because it is beaten not by a single heroic cut but by many people working their own patches, season after season, and never stopping.

Start on your own. The next time something is wrong and you catch yourself explaining it away, stop. Write down the date, what happened and what you felt before you could say why. Three lines, on paper. One entry is a note. A notebook of them is evidence.

Then look at the marks you already live with. The hygiene rating in the restaurant window. The MOT certificate. The energy rating on a house. The credit rating on a bond. The company accounts, audited by a firm the company pays. Ask one question of each: who drew the line? You will not need this pamphlet to answer it.

Put one protection back where you work: a real check, a named responsibility, a door that cannot be quietly closed. Ask, once, out loud, in a meeting where it is not expected: who signed this, and who carries it if it goes wrong. This is all the moment asks, and it will ask again tomorrow. You do not have to lead. It asks that when something is wrong, someone stands up and says so, and the rest of us back them, so no one who stands up is left standing alone. The backing is the easier half, and it is where most of us come in. What it costs is small beside the cost of the silence, which lands on all of us and only grows.

Ask the candidate whether they will sign, and keep asking.

We do not win this by becoming what we fight. Blame is the pattern's own coin and cruelty is the air it breathes. The one move it has no reply to is the one it cannot make itself. Grace. Dignity held towards a person who has not earned it. Nothing here is harder. The pattern can copy a sign, a helmet, a certificate, anything that is only a surface, but grace has no surface. It is a cost paid by the giver, and the pattern does not pay. Contain someone without hatred and they cannot be made a martyr, and we have not turned into them. Grace is not only mercy. It is a guard. Any power built to hunt the cold few is the first thing the pattern captures, so the answer is architecture, never a list of names.

Stop the taking and life gets better for all of us. Richer, safer, freer, kinder, not a dream but the plain sum of what we stop throwing away. It is the pension with the two per cent still in it. It is cladding that was what the paper said it was, and a tower with seventy-two people still at home in it. It is less fraud and less conflict, and the money and attention they swallow spent instead on problems too big for any one country. Everything we have built so far, we built despite the taking. A world without the taking is the floor, not the ceiling. That is what the watch is for. It is the world most of us think we already live in.

Until the lines are drawn by someone other than the owner, and kept, the watch starts with each of us, in our own lives. Contain the pattern. Regard the person.

Stop blaming. Start mending.

Before you go

This pamphlet has no publicity machine behind it. If it was worth your time, give it to one person who is sure the problem is the other lot, with one question: who drew the line? Then, if you can, leave an honest review. One hand at a time.

This pamphlet is free to read and download at trueregard.com/hull. It may be reproduced and passed on unchanged, free of charge.

The full argument, the science, the containment record and the strongest case against this reading are in the book, The Extraction Pattern, at trueregard.com. Its foreword is by the neuroscientist Katalin Gothard, whose laboratory is one of those that did the work.

From The Extraction Pattern, Chapter 3: The Two Operating Systems

The book's third chapter is about the two ways a human being can be built, and about the neuroscientist who found the colder one in his own brain scan.

In October 2005, on an afternoon at the University of California Irvine medical campus, James Fallon, a fifty-eight-year-old neuroscientist, sat at his desk with two stacks of brain scans in front of him. One stack was forensic. PET scans of murderers, the worst of them, sent over the years by colleagues who worked for the defence in death-penalty cases. The other stack came from a separate Alzheimer's project for which he had used his own immediate family as the control group: his mother, his wife, his three children, his siblings. He had been working through both stacks for hours.

He came to the scan at the bottom of the family pile and stopped. It looked exactly like the most abnormal of the killers' scans he had spent the morning describing. The orbital cortex, just behind the eyes, was dimmed in the way the worst forensic cases were dimmed. The amygdala signal was wrong in the way the worst forensic cases were wrong. He assumed the processing had failed. He went and checked. It had not. He came back and accused his lab technicians of switching the scans on him as a prank. The technicians said they had not switched the scans. He sat back down and peeled the tape off the subject-identification code at the corner of the printout.

It was his own.

He was, on the day he found his own scan in the killers' pile, a tenured neuroscientist with a charmed life. Married to his school sweetheart of forty years. Three grown children. Friends, colleagues, professional standing. His mother phoned him a few months later and told him there were murderers in his family tree on his father's side. She had been waiting for him to ask. The genealogy went back to colonial Rhode Island and a man called Thomas Cornell, hanged in 1673 for the murder of his mother. He counted seven alleged murderers in the line. Genetic testing of his own DNA produced more bad news. He carried high-risk versions of the genes that research on aggression and low empathy had tagged as warnings. The brain. The genes. The genealogy. By every measure the science had at the time, the desk in front of him held his own diagnosis.

He died on 20 November 2023, aged seventy-six, surrounded by his family. He had been a scientist, not a killer, for the entire run between the desk and the deathbed. When the journalist Joseph Stromberg asked him in 2013 what had made the difference, he gave the answer that became the line everyone who quotes him quotes. 'I was loved, and that protected me.'

He had gone and asked, years before he died. He put the question to his wife, his mother, his brothers and his colleagues, and the answers were not the ones that line prepares you for. Every one of them, he said, including psychiatrists who knew him well, told him they had always known what he was. His wife Diane told him he did not connect to people, that he was cold, that he was glib and marvellous at parties and loved strangers and world hunger, and that to the people actually close to him he was quite a disappointing person to be around. His brother Mark, asked years later by a radio reporter, put it more simply. 'It's a game for him to manipulate a room and a situation.'

Fallon's own inventory, given freely and often, was worse than theirs. He said he lied constantly and enjoyed it. He said he manipulated people for sport and called it intellectual bullying. He said he treated his wife and children the way he treated somebody at a bar. He skipped family funerals to drink in a beach bar. And on a trip to Kenya in the 1990s he took his visiting brother into the Kitum Caves on Mount Elgon, where two people are known to have contracted Marburg virus, without telling him what was in there. 'I just told him, I said, don't touch the ground.' The brother worked out what had happened two years later. He has not trusted Fallon since.

None of that put him in a dock. That is the point of him.

The line about being loved comes from the same interview as a sentence nobody quotes. After the scan he had decided to behave better, and he explained why. 'I'm not doing this because I'm suddenly nice, I'm doing it because of pride, because I want to show to everyone and myself that I can pull it off.' He told Diane he was using his own narcissism to beat his own psychopathy. Her verdict, in his telling, is the most exact sentence anybody has written about this configuration. She said she appreciated the effort even though he was not sincere.

So the question the scan raises is not what saved him. He was not saved. The question is why a man whose scan looked like the worst forensic cases on his desk spent fifty years hurting the people who loved him and hurt nobody else at all.

Read on in The Extraction Pattern. trueregard.com/buy

The Madison Pledge in full

A checklist, not a manifesto. Seven things a citizen can ask of anyone who wants power.

Empathy. Meet the people your decisions affect. Account for the impact.
Evidence. Do not push a policy that has failed before without publishing, with evidence, why this time is different.
Transparency. Say who funds you, publish your evidence and print who paid on the first page.
Accountability. State the targets and the dates in advance. Step down from that job if you miss them.
Skin in the game. Do not exempt yourself from what you impose on others.
Independence. Take no money from any business or lobby whose profits depend on your decisions.
Precedent. If another democracy has solved this, adopt their fix or explain why it will not work here.

That is the list. It is not a loyalty test. It is seven things you are entitled to ask.

The register of signatures and refusals is at trueregard.com/pledge. For now I keep it. That fails this pamphlet's own test, and it should be the first entry on the register: a line drawn by its owner, held until a better home can be found, and handed over the day there is one.

How this could be proved wrong

A claim worth trusting is one that could be proved wrong. This argument has two halves, and the first can be tested now. An institution that answers a scandal by changing its people and not its structure will be captured again. Change the structure and it holds, even if the new people are no better. If that is false, the argument is wrong. The bet is on the register at trueregard.com, with a case and a date. Boeing answered two crashes and 346 deaths, and then a door plug blowing out of a 737 in January 2024, by changing the people at the top, twice, and not the structure. Its own employees still sign off its work on the regulator's behalf. The regulator took back part of that power after the crashes and more after the door plug, and handed back the right to issue certificates for new aircraft in July 2026. The bet is that before the end of 2031 a further failure will be found in what Boeing certified for itself.

The second half, that every generation brings roughly the same few, cannot be tested yet, because the study that would settle it has not been run. It should be, and soon. The book sets out one more test, the one that could sink the whole argument. Take the senior people that public inquiries have named. Set beside them people in the same kind of job at organisations that were looked at and found sound. Then count, from the public record, how many on each list had already been found at fault before. If the argument is right, the first list will have at least twice as many. If the two lists come out the same, it was the job and not the person, and the argument is wrong.

The case against this reading is in The Extraction Pattern too. Extended citations and sources are in the Evidence Library at trueregard.com/sources. The register of what would prove this argument wrong, the field guide and the live Madison Pledge are kept at trueregard.com. The book is at trueregard.com/buy.

Frontline Club Books

Frontline Club Books, founded in 2026, is the imprint of the Frontline Club and publishes work in the Club's tradition of independent journalism. The Frontline Club, at Norfolk Place in Paddington, was founded by Vaughan Smith in 2003 as a home for independent journalism in Britain, and is dedicated to the memory of the colleagues of Frontline News TV who did not come home.

Vaughan Smith is a journalist, former war reporter and the founder of London's Frontline Club.

How this was written

I had no research department and no commissioning editor of my own. I used artificial intelligence, Anthropic's Claude, to research, to check, to argue with and to help me draft. The thinking is mine, the reporting is mine, the judgement on every line is mine, and so are the mistakes. I spent my journalism behind a camera. I know the difference between the person pointing the equipment and the person whose story it is.

In memory

Nicholas della Casa (1960–1991)
Rosanna della Casa (1960–1991)
Charles Maxwell (1952–1991)
Rory Peck (1956–1993)
Carlos Mavroleon (1958–1998)
Roddy Scott (1971–2002)
James Miller (1968–2003)
Richard Wild (1978–2003)

The eight colleagues of Frontline News TV, killed in the field. They were documenting what power wanted hidden.

Imprint

First published 2026 by Frontline Club Books

Paperback ISBN 978-1-0676425-4-9

Frontline Club Books Limited

Ellingham Hall, Hall Road, Ellingham

Bungay, Norfolk NR35 2EN

trueregard.com

This pamphlet is about patterns of behaviour and the structures that contain them. It does not diagnose.

Read it here

Take it with you

Hear it