CHOSEN

The Belonging Project, book 2 - Jamie Anson

Chosen The Hidden Question We Spend Our Lives Answering

Jamie Anson Peer Review Manuscript · September 2026

Part I

THE QUESTION Three hundred thousand years ago, there were humans. Not something vaguely human. Not monkeys waiting to become us. Us. Homo sapiens. Our species has been around for roughly 300,000 years. Imagine being one of them. You don't have a job. You don't have money. You don't have a country, a credit score, a mortgage or a retirement account. There is no grocery store, hospital, police department or university. No Instagram. No Google. No Uber Eats. Nobody is asking you to reset your password because it needs one capital letter, two numbers and a fucking hieroglyphic. There are people. That's basically your world. The people around you know where the water is. They know which plants will feed you and which ones will kill you. They know how to make tools, find animals, protect children, carry the injured and survive a winter. If those people stop wanting you around, there is no apartment across town and no LinkedIn post announcing that you are excited to explore new opportunities. For an enormous part of human history, other humans weren't merely one pleasant part of life. They were part of the infrastructure keeping you alive. Then humans did something extraordinary. We built more infrastructure. Agriculture. Cities. Writing. Money. Kingdoms. Hospitals. Corporations. Governments. Universities. Insurance. Airplanes. Refrigerators. Therapy. Tinder. Excel. The human animal did not disappear. We gave it pants. We put it inside an office tower and gave it Microsoft Excel. And then we became extremely confident that we understood it. WE ARE EARLY In 1949, António Egas Moniz received half of the Nobel Prize in Medicine for helping develop an operation called a leucotomy. If you have no idea what that means, good. The word makes it sound cleaner than it was. Doctors deliberately cut connections in the front part of a person's brain in an attempt to change severe mental symptoms. 1949. Not 1499. We had airplanes, radio, antibiotics and nuclear physics. Einstein was alive. These were not people living in caves. They had universities, journals, experts, conferences and credentials. They believed they were practicing modern science. Walter Freeman later popularized a faster American version in which an instrument shaped like an ice pick could be passed through the thin bone above the eye and into the front of the brain. Some patients were reported as improved. Others were left with seizures, disability, major personality changes or death.

The lesson is not that every doctor involved was cruel. Many were trying to help people for whom there were few effective treatments. That almost makes the lesson worse. Smart, educated, well-intentioned humans can do something grotesque while sincerely believing they are the advanced ones. Every generation has a dangerous little fantasy: We finally know what the fuck is going on. Then the next generation arrives. That does not make modern psychology, neuroscience or economics equivalent to lobotomy. It means they are unfinished, because every science is unfinished. Somebody noticed attachment. Somebody noticed loss aversion. Somebody noticed conformity, rejection, psychological safety and relational uncertainty. Somebody saw a pattern, found a way to measure it, and invited other people to try to break it. I think there may be another pattern worth measuring. Not a replacement for the others. Not a theory that explains everything. Two numbers that may be hiding in plain sight. The first is value: how much would losing this particular relationship or group matter to me? The second is security: how likely do I currently believe it is that I will still belong here tomorrow? Here, membership does not mean enrollment. It means an expected continued place in a particular relational world: this family, friendship, marriage, team, workplace, community or group. And security, S, is deliberately narrow. It is the person's estimate that this defined place will continue across a defined period or event. It is not automatically the same thing as status, job security, emotional responsiveness, psychological safety or moral worth, even though those things can affect the estimate. The second word matters: believe. I do not need to be about to lose you. I need to think I might. Human beings cannot directly inspect another person's future intentions. We estimate. A delayed text, an invitation, a laugh, a criticism, a promotion, a new friend, a hand on a shoulder, a closed bedroom door. Tiny events become evidence. Maybe that is why the same event can feel completely different to two people. If I believe my place with you is safe, a disagreement can remain a disagreement. If I believe my place is fragile, the same disagreement can feel like the first step toward exile. In ordinary English, the whole book keeps asking two questions. Do you choose me? Do you still choose me? Psychology already knows belonging matters. That is not the claim. Researchers have spent decades studying attachment, acceptance, rejection, conformity, relationship uncertainty, identity, commitment and psychological safety. Good. The question here is narrower: can we separately measure how much a particular membership matters and how secure that membership currently feels? Then, when something happens, can we measure how much

that security estimate changes? And does knowing those things help predict what the person does next? Maybe. Maybe not. If the answer is no, throw the idea away. A mathematical model can tell us what follows from its assumptions. It cannot prove that the assumptions describe eight billion human beings. If an existing theory already predicts the behavior just as well, use it. If these two numbers cannot be measured separately, drop them. If they add nothing, let them die. Better a dead idea than a protected one.

THE ANSWER I HAD NO REASON TO TRUST On December 29, I ate 0.99 grams of fresh Golden Teacher mushroom. I am telling you this near the beginning because hiding it until the end would be a cheap trick. The first version of this idea did not arrive from a spreadsheet. It arrived while I was under the influence of psilocybin. My life was difficult. Relationships, business and family were consuming enormous amounts of attention. I was drinking more. Using THC more. Scrolling more. I felt nauseated and sat staring at my laptop. Then a thought arrived with embarrassing simplicity: close the laptop. The answer is in the real world. I did. The room remained the same room, but my experience of it changed. Everything felt warm, connected and safe. I thought about my dog, my parents, my children, pride, care and belonging. The word that appeared was love. Then my brain did what it always does. It tried to turn a feeling into a model. Love plus one. Fear minus one. Two liquids in a cup. Pour in love and fear gets displaced. But I could not simply order my own cup to fill. In the picture in my head, I had to pour into other people and trust that love could move through the system. It was ridiculous. It was also one of the clearest experiences of my life. An hour later the acute effects faded. The question did not. Was it psilocybin? Obviously. Was it also memory, exhaustion, expectation and everything else happening in my life? Almost certainly. Was it God? I have no instrument that can answer that. More important, I had just been handed an answer I had absolutely no reason to trust. The experience proved nothing. It gave me a question. Altered states can generate vivid connections; vivid is not the same thing as true. If the question was useful, it should survive contact with everything humans had already learned. If it was bullshit, I wanted to find the wall. So I started looking for places it should break.

Part Ii

What Will You Do To Remain One Of Us?

ONE IN NINE Vienna in the 1840s had a problem that should have been impossible to ignore. The General Hospital ran two maternity clinics. In 1846, 4,010 women gave birth in the First Clinic and 459 died. That is 11.4 percent. In the Second Clinic, where midwives were trained, 105 of 3,754 women died, about 2.8 percent. The difference was not hidden in some statistical fog. Women knew. Some begged not to be admitted to the First Clinic. A woman could arrive at a hospital looking for help and discover that the identity of the people assigned to help her changed her odds of leaving alive. Ignaz Semmelweis was a young Hungarian doctor working in the First Clinic. He chased explanation after explanation. The position women gave birth in. The priest who walked through the ward. Crowding. Air. Fear. None of it satisfactorily explained why one clinic was so much deadlier than the other. Then a colleague, Jakob Kolletschka, died after being cut during an autopsy. The illness looked horribly familiar. Semmelweis began to suspect that doctors and medical students were carrying something from corpses to living women on their hands. Germ theory was not yet available to him as the clean explanation we inherit today. He called the suspected material cadaverous particles. He ordered doctors and students to wash their hands in a chlorine solution before examining women. In 1848, after the washing rule had been used consistently, 45 of 3,556 women in the First Clinic died. About 1.3 percent. The clinic that had killed roughly one woman in nine in 1846 was now killing roughly one in eighty. The numbers did not require a microscope. Something enormous had changed. The familiar version of this story is comforting because it has a hero and idiots. Semmelweis discovers handwashing. Arrogant doctors refuse to listen. Science eventually wins. Real history is messier. Semmelweis did not have modern germ theory. His explanation was incomplete. Medical theories of disease were different. His communication became combative. Some contemporaries did accept parts of his practice. The point is not that every doctor who resisted him was stupid or evil. The point is more uncomfortable: evidence can arrive before a group has a safe way to absorb what the evidence means. History becomes stupid when we turn everybody who disagreed with the eventual winner into a cartoon villain. Imagine what acceptance demanded. It did not merely say, Here is a better soap. It implied that the hands of physicians, the people trained to heal, had helped carry death from autopsy rooms to mothers. A factual update could also be an identity wound. That does not prove belonging caused the resistance. It gives us a question worth carrying through the rest of this book: what happens when changing your mind changes what your membership, competence, status, or moral identity means? Evidence may be free. Believing it can be expensive. What makes being wrong more expensive than staying wrong? Dead ideas are cheaper than dead mothers.

THE LINE IS OBVIOUS Solomon Asch had a problem with an earlier kind of conformity experiment. If you put people in a dark room and ask them to judge how far a tiny point of light moved, there is no obvious answer. People influence one another partly because the world itself is ambiguous. Asch wanted to remove that excuse. He wanted a situation where a person could look directly at reality and know the answer. So he used lines. One card showed a single line. Another showed three comparison lines. One was obviously the same length as the first. The task was so easy that people working alone almost never got it wrong. Then Asch put a real participant in a room with a group of men who were secretly working for the experimenter. One by one, the group members gave their answers aloud. On ordinary trials they answered correctly. Then came a critical trial, and every one of them calmly chose the same obviously wrong line. Now imagine the timing. You have eyes. The answer is right there. The first man says the wrong thing. Strange. The second says it too. Then the third. The fourth. The fifth. Nobody is laughing. Nobody looks confused. By the time the answer reaches you, the experiment has changed. It is no longer merely a vision test. There are now two pieces of evidence in front of you. The card says one thing. Every other human in the room says another. Across Asch's 123 participants, the wrong majority pulled responses toward it on 36.8 percent of the critical trials. That number is often mangled into the claim that 37 percent of people were conformists. It was 36.8 percent of critical responses. About a quarter of the participants never yielded at all. On most judgments, most people still said what they saw. The experiment is interesting because humans were neither sheep nor perfectly independent. They moved. The interviews afterward make the movement more human. Some people who yielded came to doubt their own perception. Others believed the group was wrong but did not want to stand apart from it. Others tried to explain the discrepancy away. That distinction matters. A person can say the same wrong answer for very different reasons. One mind changes its belief. Another keeps the belief and changes the public behavior. The standard explanation is social influence, and it is a good one. We do not need to rename Asch. The useful question is narrower: does the cost of standing alone change when the person believes disagreement could cost them a valued place? If disagreement is guaranteed to have no social consequence, does conformity fall? If dissent threatens exclusion from a group the person deeply values, does it rise? Now the idea can be separated from the original effect. The line was obvious. That was Asch's genius. He stripped away uncertainty about the object and left uncertainty about the people. Our later experiments can ask whether the price of standing alone changes when continued membership is secure.

THE BORING GROUP Initiation rituals are weird because the stated task often has almost nothing to do with the group. Eat this disgusting thing. Stay awake. Run until you vomit. Memorize nonsense. Wear the humiliating outfit. Let older members insult you. Pay the fee. Prove you want it. From the outside, the ritual can look like a tax on joining. Psychology has several ways to explain why difficult initiation can increase attachment. One is effort justification. If I suffer to obtain something, admitting that the thing is worthless makes the suffering harder to explain. So the mind may raise the value of the prize. That idea sits comfortably inside cognitive dissonance theory and does not require CHOSEN. But initiation also changes the social meaning of membership. A group that is difficult to enter can signal that membership is scarce, that other members paid a similar cost, and that leaving would abandon an investment. The ritual creates a story: people like us went through this. You are one of us because you did too. The danger is obvious. If pain becomes evidence of loyalty, cruelty can masquerade as culture. Hazing survives partly because every generation can point backward. I endured it. Therefore it mattered. Therefore you should endure it. The suffering becomes part of the membership credential. Suffering does not automatically make belonging secure. Sometimes the opposite happens. Abuse can make membership feel conditional on continued submission. A person may cling harder while feeling less secure. Value and security can move in different directions. That is exactly why separating the two variables matters. A group can become extremely valuable while membership feels extremely fragile. That combination may be combustible. If I have sacrificed friendships, money, sleep, reputation and dignity to join, the group is now expensive to lose. If I also believe one wrong move can get me expelled, the social stakes of obedience become enormous. That high-value, low-security combination is one of the cases the concentrated experiments near the end of this book should test. The most dangerous group may not be the one that makes you suffer to enter. It may be the one that makes you suffer to enter and then spends the rest of your membership reminding you that you can still be thrown out.

Part Iii

When Information Gets Expensive

THE FINE The daycare experiment in Haifa is useful precisely because nobody needs a theory of civilization to understand it. Parents were supposed to collect their children by closing time. Some arrived late, which meant a teacher had to stay. So six daycare centers introduced a fine. Ten Israeli shekels if a parent arrived more than ten minutes late. Four other centers served as comparisons. If price is punishment, the prediction seems obvious. Make lateness more expensive and get less lateness. The researchers called the paper A Fine Is a Price. The title contains the whole grenade. They got more. Late pickups increased significantly after the fine appeared. Then the researchers removed the fine. Lateness did not simply snap back to its old level. Something about the meaning of the interaction had changed. Money works. The interesting part is what the money changed besides price. The researchers argued that penalties enter relationships that already contain unwritten expectations. Before the fine, being late could mean making another person wait for you. After the fine, a parent could interpret the same act as a service with a price. Ten shekels changed information about what kind of situation this was. Human beings do not respond only to the physical event. They respond to what the event means inside a relationship. If I am ten minutes late to meet a stranger, a friend, my child, a judge, and a person I desperately want to impress, the clock records the same ten minutes. My brain does not. Context changes the social prediction. This matters far beyond daycare. Every rule teaches people what kind of relationship they are in. Every incentive says something about what counts. Pay children for every book and you may teach them what reading is for. Measure employees only by a metric and you may teach them what the company actually values. Charge for behavior that used to violate a social norm and you may accidentally put it on the menu. We keep trying to change human behavior by changing the price. Sometimes the price was never the thing controlling the behavior. But this experiment is also a warning for us. It does not prove belonging was the hidden variable. Social norms, guilt, pricing, expectations and incomplete contracts already explain a great deal. Merely pointing at a weird result and yelling belonging is not a theory.

THE PERSON WHO SAYS NO Amy Edmondson became interested in a puzzle inside hospital teams. You might expect the best teams to make the fewest mistakes. In early work, she encountered something stranger: some teams that looked better also reported more errors. One possibility was horrifying. Maybe the better teams were not making more mistakes. Maybe they were better at admitting them. That insight developed into the idea of psychological safety, which Edmondson defined as a shared belief that a team is safe for interpersonal risk taking. In a 1999 study of 51 work teams, psychological safety was associated with learning behavior. The phrase has since become corporate wallpaper, which is what happens to a useful idea after enough PowerPoint decks. Strip away the jargon and the mechanism is brutal. If telling the truth can cost me status, respect, promotion or membership, I have two problems instead of one. I have the actual problem, and I have the problem created by revealing it. A nurse notices a medication error. An engineer sees a defect. A junior analyst finds a hole in the forecast. A child knows Dad is wrong. A friend thinks the wedding is a mistake. The information exists. The question is whether the social system can receive it. This is where secure belonging and psychological safety overlap, and the overlap has to be acknowledged rather than marketed as novelty. Edmondson already gave organizational science a way to think about interpersonal risk. CHOSEN's narrower addition is to ask whether the person's estimate of continued membership is one component of that risk, and whether measuring it improves prediction. Those are not the same claim. I can fear humiliation without fearing expulsion. I can feel secure in my job but afraid of looking stupid. I can feel socially loved and still worry about losing authority. Status, competence and membership are related but separable. So build the experiment to separate them. Tell participants that their place on the team is guaranteed, but that dissent may lower their status ranking. Tell another group that status is unaffected, but repeated dissent may cost future team membership. Give both groups the same hidden information and the same financial incentives. Which threat suppresses speaking up more? If status wins, write status. If membership wins, write membership. If both matter, estimate both. The point is not to make every person who says no into a hero. The person may be wrong. The point is to build rooms where being wrong and being disagreeable are not confused with being disposable.

THE TRUTH THAT DIDN'T SURVIVE THE TRIP Hierarchy is not the enemy. Airplanes, hospitals, armies and companies need roles. The problem begins when truth has to compete with belonging, status or future opportunity. If telling you what I see threatens my place with you, you have built a system with an incentive to become blind. That is exactly why the case matters. Organizations rarely destroy information by placing it in a shredder. They change its social cost. A concern has to travel from the person who sees it to the person who can act on it. At every step, somebody has to decide whether speaking again is worth being difficult, slowing the schedule, challenging a superior, contradicting a customer, or becoming the person everyone wishes would shut up. Every layer can edit reality without anyone consciously deciding to lie. A technician softens a sentence. A manager removes uncertainty. A vice president asks for stronger evidence. A room interprets silence as agreement. Then the person with the most authority can become the person with the least access to unfiltered reality. On the night before Challenger launched, engineers at Morton Thiokol were worried about rubber seals called O-rings in the shuttle's solid rocket boosters. The forecast was unusually cold. During a teleconference, Thiokol engineers presented concerns about how the seals performed at low temperatures. The engineering recommendation was not to launch below 53 degrees Fahrenheit. The next morning was far colder than that. The recommendation did not survive the organization. After discussion and a management caucus, Thiokol reversed its position and recommended launch. Challenger lifted off on January 28, 1986. Seventy-three seconds later the vehicle broke apart. Seven people died. The official investigation also found that some of the people making the final launch decision did not know the full recent history of O-ring trouble. They did not know that Thiokol had initially recommended against launch, or how strongly some engineers still objected after management changed the recommendation. The truth existed. It just did not survive the trip. It is tempting to turn the story into a morality play about cowardly managers silencing brilliant engineers. The Rogers Commission record is more complicated. The data were incomplete. People interpreted the evidence differently. NASA and contractor processes had normalized concerns over time. Communication moved through layers. Pressure existed without requiring a villain to announce, I would like to kill astronauts today.

PULL THIS CORD In 1984, General Motors and Toyota reopened a car plant in Fremont, California, as a joint venture called NUMMI. The old GM plant had become notorious for terrible labor relations and terrible cars. More than 85 percent of the new NUMMI workforce had worked in the old plant. Toyota did not solve the problem by finding a different species of American worker. It changed the system around many of the same people. One of the most visible changes was a thin cord hanging along the assembly line. The andon cord. In Toyota's system, a worker who saw a problem could pull it. The first pull summoned help. If the problem could not be fixed before the car reached a certain point, the line stopped. That sounds like a manufacturing detail until you understand what an assembly line represents. A stopped line is visible. Expensive. Everybody knows where the interruption happened. In a system built around keeping production moving at all costs, admitting a problem can make the person who noticed it look like the problem. This American Life's reporting on NUMMI captured workers describing how different this felt. At the old plant, defects could roll onward and be repaired later. Under the Toyota system, workers were expected to expose the defect now. Rick Madrid described watching a bolt problem get addressed during training in Japan and realizing that fixing it immediately simply made sense. One bolt changed his attitude. The cord did not create courage through motivational posters. It changed the rules surrounding bad news. Pulling the cord was not sabotage. It was the job. The system tried to make discovery of an error valuable before the error became expensive. This is a concrete version of the principle hiding underneath the Challenger story. Information does not travel merely because somebody possesses it. Institutions create prices for transmission. The price can be time, embarrassment, rank, career opportunity, conflict, or membership. NUMMI is not a clean experiment proving secure belonging. Toyota changed many things at once: training, teamwork, job classifications, quality processes, labor-management relations and production methods. We cannot point to the cord and claim causality for the entire turnaround. But the cord gives us a design question. What would the social equivalent look like? Not a literal rope in every marriage and boardroom. A rule that says: finding the problem early earns protection, not exile. The person who interrupts the process remains one of us. The strongest version is not speak up. Every company already says that. The strongest version is to make the response to bad news observable. The first person who pulls the cord teaches everybody else whether the cord is real. Culture is what happens to the second person after everyone watches what happened to the first.

THE WALLET WITH MORE MONEY In 2019, researchers carried more than 17,000 wallets into institutions in 355 cities across forty countries. Some contained no money; others contained money. In virtually every country, wallets containing more money were more likely to be reported. More money increased the material reward for dishonesty, yet honesty increased. The researchers did not conclude that belonging had defeated utility. Their evidence pointed instead to concern for the owner and the growing psychological cost of seeing oneself as a thief. Put more money in the wallet and stealing becomes more profitable, but it may also become harder to tell yourself that you are not really stealing. Humans carry reputations in other people's heads, and in their own. That is why the wallet belongs here. It is not clean evidence for CHOSEN. It is evidence against a simpler mistake: humans do not walk through the world with one scoreboard labeled CASH. Material incentives matter, but so can identity, norms and what an action means inside a relationship.

THE TRUST PROBLEM Trust is one of those words that becomes less precise the longer you stare at it. I trust my surgeon. I trust my friend. I trust the bridge. I trust that my child will eventually tell me what happened. Those sentences do not describe the same mechanism. For this book, I care about a narrower piece: what trust does to uncertainty between people. Two people never have direct access to each other's future behavior. We infer. We reveal information. We make promises. We watch what happens. Trust changes how much uncertainty we can tolerate while continuing to act together. That is why I became fascinated by speed. In sales, you can feel an interaction moving. A customer asks a hard question and gives you the real objection. Good. Now both people are working on the same problem. Another customer smiles, says everything sounds great, and hides the actual reason they will never buy. The conversation can look pleasant while going nowhere. Calling that trust velocity is tempting. It sounds scientific. That is exactly why we should not use the phrase casually. Velocity has units. Distance divided by time. If we cannot define what is moving, in what direction, across what scale, measured by whom, then velocity is a metaphor wearing a lab coat. So define the problem before naming the number. Suppose two people begin with different private models of a decision. We can measure how long it takes before the relevant differences are accurately represented in a shared record. We can measure how many turns are required before each person can state the other's concern. We can measure how quickly hidden information appears. We can measure whether the resulting action predicts what both parties said they wanted. Now trust becomes experimentally useful. Manipulate whether disclosure is socially safe. Give one group a history showing that admitting uncertainty is rewarded. Give another a history showing that uncertainty reduces standing or membership. Then introduce the same hidden objection. Does it surface earlier? Does the group reach an accurate shared model faster? Notice what we are not measuring. Agreement. Two people can understand each other perfectly and still disagree. In fact, forcing agreement can destroy the measurement by teaching people to hide the disagreement. Trust, in this narrow sense, may be what allows humans to move through uncertainty created by other humans without requiring certainty first. Secure belonging may be one ingredient because it lowers the cost of revealing information that could create friction. But competence, honesty, incentives and past reliability obviously matter too. The equation is not ready. Good. We have something better than a slogan: a measurement problem. We have also reached the point where the book needs to come back down to earth. Hospitals, rocket boosters and factories are useful because the stakes

make the machinery visible. Most of us encounter the same uncertainty somewhere quieter: across a bed, at a dinner table, in the few seconds after somebody we love looks at someone else.

Part Iv

Do You Still Choose Me?

LOVE YOU FOREVER Robert Munsch's Love You Forever looks, at first glance, like an extremely simple children's book. A mother holds her baby. The child grows. The relationship changes. The mother ages. Eventually care moves in the other direction. Millions of readers encounter it as a story about a parent's love surviving time. The origin makes the simplicity heavier. Munsch has written that he and his wife had two babies born dead, and that the story began as a song for them. For a long time, he could not sing it without crying. His regular publisher did not want the eventual book, so he took it elsewhere. On his official site, Munsch says the book has sold about fifteen million copies. A private expression of loss became something strangers used to talk about permanence. The book is not a scientific study. Its popularity does not prove that humans possess a hidden membership variable. It gives us language for the thing we are trying to measure. A child changes constantly. The infant who depends on you becomes the toddler who screams no, the teenager who closes the door, the adult who leaves, the person who builds a life you did not design. If love requires sameness, parenting is a catastrophic business model. What many people seem to hear in the story is a promise that survives those state changes. Your behavior can alter what happens between us without erasing the deeper fact that you remain part of my world. That promise is not the same as unlimited access. This matters enormously. Secure belonging does not mean a parent must accept abuse, fund addiction, tolerate violence, ignore boundaries or prevent an adult child from leaving. It does not mean the child owes the parent closeness. Love can survive a locked door. This is one place ordinary language is better than our equations. I can love you and not trust you. I can trust your competence without loving you. I can want you in my future and still refuse to be in the same room today. The variables overlap without collapsing into one another. The measurable question is not whether a person has unconditional love in some cosmic sense. It is whether the person expects the relationship to survive specific kinds of change. Disagreement. Failure. Distance. A new partner. A new sibling. A different belief. A boundary. That expectation may be what secure belonging feels like from the inside: this event can change us without automatically ending us. The strange beauty of the Munsch story is that the promise does not stop with the original relationship. Care moves forward. A person who was held becomes a person who holds. Love is finite in attention and still capable of propagation.

That may be the most hopeful form of abundance in this book. Not infinite time. Not infinite resources. A pattern that can be given away without disappearing from the person who taught it.

THE LOVE CHEMICAL THAT ISN'T We went looking for the molecule of love. What we found was a live argument. Oxytocin became irresistible because it appears near so many experiences humans care about. It is deeply involved in childbirth and milk release. Researchers have also studied it in attachment, social recognition, stress, cooperation and trust. In 2005, Michael Kosfeld and colleagues published a famous experiment in Nature. Participants received oxytocin through the nose or a placebo and then played an economic trust game. The oxytocin group transferred more money to another person. The title was wonderfully clean: Oxytocin increases trust in humans. Popular culture did what popular culture does. Oxytocin became the love hormone. A complicated research program became a molecule with a personality. Then the clean story got messier. A registered replication published in 2020 found no overall increase in trusting behavior in its main social-contact condition, although an exploratory analysis hinted that effects might differ across people. Other research has likewise made the simple universal story harder to defend. The responsible conclusion is not that oxytocin is irrelevant to social behavior. It is that one famous result does not justify calling a molecule love or trust. Oxytocin still matters biologically. Social relationships still affect bodies and brains. But there is no scientific permission slip for calling one molecule love or trust. The useful lesson is that social behavior reaches into biology while resisting our desire for one-variable explanations. This is not science failing. This is science working. A clean claim met null results and then a possible boundary condition. Maybe later work will narrow it further. Maybe the boundary result will fail. The correct response is not to pick whichever paper gives us the prettiest story. It is to keep updating the story. CHOSEN should expect the same treatment. Your brain is meat. Every moment of tenderness, jealousy, trust, embarrassment and grief happens while living tissue inside your skull exchanges electrical and chemical signals. That does not reduce love to chemistry any more than understanding acoustics reduces music to air pressure. It gives us another level at which to ask what is happening.

DOES CHOOSING THEM MEAN UNCHOOSING ME? Imagine two people you love call at the same time. Both need you. You can only go to one. Choose. That tiny problem reveals something uncomfortable. Love may feel infinite as an idea. Attention isn't. Time isn't. Money isn't. Your body can only be in one place. There are twenty-four hours in a day no matter how enlightened you become. Picture a magnifying glass over an ant. Sunlight that was harmless when spread across a wider area becomes intense when concentrated. Move the glass and the heat moves with it. Human attention can feel like that. When somebody falls in love, the beam narrows. Hours disappear into conversation. Sleep becomes negotiable. Food becomes optional. The rest of the world gets quieter. At its best, concentrated love feels like being on fire. You can also burn the ant. Intensity is not the same thing as security. In fact, if one person becomes your entire social world, the relationship can become a single point of failure. Every movement of the glass matters too much. Humans survived in groups for a reason. Two people can love each other deeply and still benefit from friends, siblings, parents, neighbors, teammates and communities. Two people can do many things. Two people have a harder time taking down the woolly mammoth. THE TERRIFYING THING ABOUT LOVE Jealousy is usually described as an ugly emotion. Sometimes it is. It can become controlling, violent and destructive. But underneath jealousy is often an intelligible question. Could this person become more important to you than I am? Or even simpler: Does choosing them mean unchoosing me? That question appears in romance, but sex does not own it. Children feel it when a sibling arrives. Friends feel it when a best friend finds another best friend. Employees feel it when the boss hires a new star. Adult children feel it when parents remarry. Humans track attention because attention predicts allocation. Who gets the call? Who gets the seat? Who gets the secret? Who gets forgiven? Who gets chosen when choosing becomes expensive? THE PIE There really is a pie. Your time is finite. Your attention is finite. Some relationships will receive more than others. Pretending otherwise does not make us loving. It makes us bad accountants. But allocation and belonging are not the same thing. You can spend Saturday with one child without expelling the other from the family. You can help a friend through a crisis without announcing a permanent friendship ranking. You can disagree with your partner without ending the relationship. Secure belonging is confidence that temporary changes in attention, agreement, proximity, status or allocation do not automatically threaten continued membership.

That is the distinction jealousy often cannot see in the moment. You chose them becomes you didn't choose me. Those are not the same sentence. THE ANT IS NOT THE SUN Another mistake is making one relationship responsible for producing all the light. Romantic culture loves this idea. You complete me. You're my everything. I only need you. It sounds beautiful right up until the relationship moves. A resilient social life has redundancy. Not because every person is interchangeable. They aren't. The point is almost the opposite. Different people hold different parts of us. Nobody has to be equally close. Nobody has to receive equal attention. But perhaps nobody has to be kicked out of humanity either. Love may be scarce in attention while still capable of multiplying through behavior. Fear may work the same way. I can receive security from you and carry some of it into the next room. I can watch you choose someone else without believing that you've stopped choosing me.

WHAT EXACTLY IS CHEATING? Cheating sounds like a sexual category until two people try to define it without using the word cheating. Is it sex? Kissing? A dating app? Porn? Secret messages? An emotional relationship? Dinner with an ex? Hiding money? Lying about where you were? Different couples draw different lines, and some couples deliberately remove sexual exclusivity while treating honesty as nonnegotiable. That variation reveals something useful. The injury is not always the physical act. Cheating may be the secret violation of the rules under which another person believes they are being chosen. That definition is not permission to redefine betrayal after the fact. The rules have to be genuinely shared. If one person thinks the relationship is exclusive and the other privately decides exclusivity is primitive, philosophy does not cancel deception. Consensual non-monogamy is useful here because it separates variables that monogamous culture often bundles together: sex, romance, love, commitment, honesty, exclusivity, family, primary attachment and belonging. Two people can agree that sex outside the relationship does not threaten membership. Another couple can experience a flirtatious text as catastrophic. The physical event does not contain its social meaning by itself. Jealousy therefore deserves more precision than insecurity. Sometimes jealousy is a reasonable response to evidence that an agreement is being violated. Sometimes it is a false alarm. Sometimes it is an attempt to control another person's freedom so completely that uncertainty disappears. The question is not how do we eliminate jealousy? It is what signals allow freedom and security to coexist, and why the same allocation of attention can mean nothing in one relationship and the beginning of exile in another.

THE THING YOU CAN NEVER KNOW You cannot directly know whether another person will choose you tomorrow. That is not pessimism. It is a property of having separate brains. You can know what they did yesterday. You can hear what they promise. You can observe patterns. You can inspect a contract. You can share passwords, locations, bank accounts and DNA. None of those gives you tomorrow. This creates one of the central problems of social life. We have to act on estimates of minds we cannot inspect. Trust is partly the decision to move while the estimate remains incomplete. Control offers a seductive solution. Reduce the other person's options. Make departure expensive. Make disagreement costly. Surround the relationship with rules so tight that uncertainty appears to vanish. But the more choice you remove, the less their continued presence tells you about preference. Imagine two doors. Behind the first, your partner can leave at any time and repeatedly stays. Behind the second, leaving would destroy their finances, friendships, housing and relationship with their children. In both rooms the partner is physically present. The evidence means something different. This is why power and love can imitate one another from far away. Both can produce proximity. Both can produce compliance. Both can produce the sentence I'm not going anywhere. Only one requires the person to retain a meaningful alternative. Power says: you can't get rid of me. Love says: you wouldn't want to. Of course real relationships contain obligations. Parenting, marriage, caregiving and community create duties that should not evaporate whenever desire changes. Freedom is not the absence of responsibility. The point is epistemic: forced behavior tells us less about voluntary preference. The estimate of being chosen can therefore never reach absolute certainty without destroying the condition that makes choosing meaningful. There will always be some uncertainty left. Secure belonging may not be the elimination of that uncertainty. It may be the ability to live inside it without treating every independent movement as evidence of abandonment.

THE ESTIMATE Here is the smallest version of the mechanism I think we are looking for. You have a relationship. The relationship has value to you. You also have an estimate, conscious or not, of whether the relationship will continue. Something happens. You update. Your partner puts the phone face down when you enter the room. Your boss removes you from a meeting. Your friend starts a group chat without you. Your child stops answering. Your mother sounds disappointed. A new employee gets invited to lunch. A sibling receives praise. None of these events contains a breakup, firing, exile or replacement. Each can nevertheless become evidence about the probability of one. Now take the same event and change the prior estimate. In a relationship that has survived twenty years of conflict, a delayed text may barely move the number. In a relationship that nearly ended last week, the same delay can detonate the afternoon. This is why the magnifying glass is not simply counting minutes of attention. Twenty minutes can carry more information than eight hours. Eight ordinary hours at work may say nothing about a marriage. Twenty minutes of secret whispering at a party may say a great deal, because the event is interpreted as predictive. The brain is not asking only, how much attention did I receive? It may be asking, what does this allocation of attention predict about my future place? We do not yet have a validated CHOSEN meter. That is fine. Researchers routinely study things they cannot touch directly by using multiple imperfect measures. Self-report. Choice behavior. Reaction time. Attention. Physiology. Longitudinal outcomes. The job is to see whether the measures converge. Start with value. How costly would it be to lose this relationship? What resources, identity, affection, status, history and future plans are tied to it? Then security. How likely does the person believe continued membership is across a defined period and defined class of events? The definitions have to be specific. My security in a marriage is not identical to my security in a poker group. My confidence that I will still be employed next month is not the same as confidence that colleagues like me. A vague belonging score will recreate the exact problem we are trying to solve. Imagine a temporary-team experiment with four conditions. First make the team matter more to some participants than to others: perhaps future rewards or access to a valued second task depend on staying with the team. Separately manipulate security: tell half the participants that their place is guaranteed and tell the others that future membership will be reconsidered after the next round. Then give everyone the same ambiguous social event, such as being left out of a team decision. Measure S immediately before and after the event. Then measure what people actually do: speak up, conform, withdraw, retaliate, repair, disclose

information or leave. At the same time, measure the established alternatives - relationship-specific attachment where relevant, rejection sensitivity, relational uncertainty, psychological safety, status threat, commitment and material incentives. CHOSEN earns something only if V, S, their interaction or the post-event change in S predicts new behavior beyond those established measures. Then test updating. Present a standardized negative event. Measure S before and after it. Does delta S predict behavior after controlling for relationship- specific attachment, relational uncertainty, rejection sensitivity, sociometer measures, commitment and investment, status threat, material incentives and ordinary loss aversion? Better yet, compare preregistered models on data they have not already seen. If yes, we have a variable. If no, we have a beautiful paragraph. The estimate drives behavior, not necessarily reality. That may be the entire tragedy. A person can be surrounded by love and behave as though exile is imminent. Another can be seconds from losing everyone and feel completely secure. Same red light. Different update. That is the whole mechanism in miniature. Value tells us how much losing the relationship would matter. Security tells us how plausible the loss currently feels. The event supplies new information. The update changes what the person does next.

Part V

When The Answer Feels Like No

WHAT HAPPENS WHEN EVERYONE LEAVES AT ONCE? Most losses arrive one at a time. A job ends but home remains. A relationship ends but friends remain. A friendship changes but family remains. The rest of the social map keeps telling you that you still exist somewhere. Then there are periods when several maps seem to disappear at once. Work changes. A relationship collapses. Family becomes complicated. The phone that once contained a dozen easy conversations becomes a list of people you are no longer sure you can call. I know that room. There is a particular kind of panic in looking at a phone and realizing that the question is not who do I want to call? It is who can I call without making things worse? The device is full of names. The feeling is still empty. I had spent years understanding business through numbers. Revenue. Conversion. Retention. Forecasts. Hundreds of pages of financials. Almost no pages of feelings. Then the social balance sheet changed faster than I knew how to model. I knew what I thought I needed. Family. That sentence can become dangerous when urgency enters it. If belonging feels threatened, connection can stop behaving like a preference and start behaving like an emergency. You reach harder. Explain more. Ask for reassurance. Try to close distance quickly. The intensity itself can frighten the people whose reassurance you need. Now the loop begins. Threat produces urgency. Urgency produces intensity. Intensity produces withdrawal. Withdrawal becomes evidence that the threat was real. This is not a universal theory of relationship breakdown. It is one pattern I can recognize because I have lived close enough to it to feel the machinery. The scientific question is whether perceived membership threat reliably produces behaviors that accidentally increase the threat. Personal experience is where a hypothesis can begin. It is not where the hypothesis gets to win.

THE EMPTY ROOM Ostracism researchers have spent decades studying what happens when people are ignored or excluded. Some experiments are almost comically small. A person is left out of a simple ball-tossing game. A computer simulation stops sending them the ball. Nothing material is taken. No one is physically injured. Yet exclusion reliably produces distress and threatens basic psychological needs in the laboratory. The absurdity is the point. The nervous system does not require a divorce decree or a termination letter before it begins responding to social disappearance. Sometimes all it needs is the ball going to everybody else. Kipling Williams and other researchers built a large literature around ostracism because the phenomenon is both ordinary and powerful. People report threats to belonging, self-esteem, control and meaningful existence. The immediate reaction can occur even when the exclusion comes from strangers or from a group the person does not particularly admire. That does not mean every ignored text is a biological emergency. Laboratory effects are not destiny. People differ. Context matters. Chronic exclusion is not equivalent to a few minutes in a game. But the research gives us a useful baseline. Social absence itself can be information. Our question begins with what happens next. Does the excluded person estimate that membership is temporarily interrupted or permanently threatened? Does a credible signal of continued membership change the reaction even while the immediate exclusion remains? An empty room can be physically empty. It can also be a room full of people who no longer send the ball back.

THE LABEL A child brings home a bad grade. The paper says 58 percent. The number is local. Then an adult says, You are lazy. The event has become a label. Or an employee misses a target and hears, You are not leadership material. A behavior has become a forecast about the person. Labels are efficient because they compress. Doctor. Criminal. Mother. Immigrant. Genius. Addict. Employee. Failure. Each word can carry a truckload of expectations before the person has done anything in front of us. Psychology and sociology have long studied labeling, stigma and expectancy effects. The careful lesson is not that a word magically transforms a human. It is that categories change what observers notice, what behavior they expect, and sometimes what opportunities they provide. The problem changes when the label predicts membership. Good student is not merely a description if the child believes being a good student is the admission ticket to love. Strong employee is not merely performance feedback if weakness appears to threaten continued place on the team. Loyal friend becomes dangerous when disagreement counts as disloyalty. Then the person is no longer optimizing only for the behavior named by the label. They may be protecting the membership attached to it. This is one reason conditional belonging can produce brittle performance. If failure means I did something badly, I can inspect the failure. If failure means I am no longer one of us, the evidence becomes radioactive. Secure belonging changes the grammar. Your behavior can fail without your belonging failing. That sentence is not an excuse from consequences. A surgeon who repeatedly harms patients should lose the job. A violent partner can lose access. A student can fail the course. Membership in humanity is not membership in every role. The distinction is between evaluating an action and making every evaluation a referendum on whether the person still counts.

WHAT HAPPENS WHEN YOU TELL SOMEONE THEY BELONG? That limitation makes the science better. Secure belonging does not require a world with no red lights. It requires that every red light not automatically mean exile. Your idea can fail without you failing. Your behavior can need correction without your membership disappearing. You can hear no without hearing leave. In a 2011 experiment, Gregory Walton and Geoffrey Cohen worked with 92 college students. Students were randomly assigned to different conditions, which is one of the basic tricks researchers use to make a fair comparison. The exercise was brief. It suggested that awkward moments, rejection and other social problems in college can happen to many students and can get better with time. In other words: a bad day here does not automatically mean you do not belong here. For Black students in that study, the intervention was associated with improved academic outcomes over the following years, along with reported health and well-being benefits. The intervention did not give students money, tutoring, easier classes or new professors. It worked on interpretation. A bad day can be interpreted in different ways. I failed this test can mean I need to study differently. It can also mean people like me do not belong here. The event is identical. The prediction attached to the event is different. If belonging feels secure, negative information may remain local. The professor disliked my paper. My friend is busy. My partner is angry. If belonging feels fragile, the same information can spread through the entire social model. Then came a much larger test. In 2023, Walton and dozens of collaborators reported a randomized experiment with 26,911 students at 22 U.S. institutions. The intervention increased first-year full-time completion especially for students in groups that had historically progressed at lower rates, but only in contexts where those groups had real opportunities to belong. The message could not manufacture an environment that contradicted it. Most of this book has been asking how humans get into trouble. Here is a happier question. What happens when you make belonging safer?

THE SCOREBOARD We keep score constantly. Money is a scoreboard. Grades are a scoreboard. Followers, promotions, medals, invitations, citations, seats at the table and who gets called first are scoreboards. The problem is not that humans measure. The problem is forgetting what a measurement leaves out. A company can maximize quarterly profit while burning out the people creating it. A family can maximize quiet by making one child afraid to speak. A country can improve an average while a minority absorbs the cost. A relationship can maximize time together while destroying freedom. Every metric creates a temptation to confuse the visible number with the objective. This book began with the suspicion that some decision models underweight the value and security of social membership. But adding belonging to the scoreboard does not solve the scoreboard problem. A cult can maximize belonging. A gang can maximize loyalty. An authoritarian organization can create extraordinary internal solidarity. Belonging is therefore not the final scoreboard. A group can create extraordinary internal solidarity while harming people outside it. We will come back to that problem when the circle gets larger.

THE MODERN FOREST Exile used to have obvious geography. Outside the fire. Outside the village. Outside the walls. Modern exclusion often has Wi-Fi. Nobody invited you. Your friends made another group chat. Your husband stopped touching you. Your teenager closes the bedroom door. The boss stops asking for your opinion. Everybody laughs at a reference you do not understand. Your best friend gets a new best friend. Your mother says she is disappointed. Your child stops calling. None of these events is literally the Pleistocene forest. That comparison becomes stupid if taken too far. You are not going to be eaten by a leopard because you missed brunch. But the brain still has to interpret social movement. Who is moving closer? Who is moving away? Which changes are temporary? Which predict replacement? Which relationships survive conflict? Modern life may actually make some estimates harder because the evidence is continuous. Read receipts. Follower counts. Last active. Location sharing. Likes. Stories viewed. The number of people invited to something you can watch from your couch. Technology did not invent social comparison. It industrialized the evidence stream. Secure belonging does not mean becoming blind to that evidence. Sometimes the evidence is accurate. A relationship really is ending. A job really is disappearing. Friends really do move on. The goal is calibration. A red light should mean red light, not automatically the end of the road. Winning is not eliminating every signal of distance. It is learning which signals predict actual loss, which do not, and how to respond without creating the outcome you fear. The modern forest is full of notifications. So far the book has mostly asked what happens when one person wonders, do you still choose me? The next movement changes scale. What happens when the thing a person is protecting is not only an individual relationship, but a family, team, faith, nation, cause or imagined future? The question is larger and more dangerous: where has the person drawn the boundary around me and us?

Part Vi

The Bigger Game

US FOR NO REASON Humans can manufacture an us out of almost nothing. Social psychologists call one demonstration the minimal group paradigm. Put people into categories using trivial or even arbitrary distinctions, then let them allocate points or rewards. Under some conditions, people begin favoring their assigned group even though the group has no history, no shared struggle and no meaningful future. That finding is easy to over-romanticize. It does not mean every arbitrary label instantly creates hatred. It does mean the machinery for categorization is cheap to start. Give the mind a line and it may begin drawing around it. Think about how quickly this happens outside laboratories. Sports teams. Schools. Departments. Neighborhoods. Android and iPhone. People who put pineapple on pizza and people who believe those people should be tried at The Hague. Most of these identities are harmless because the stakes are low. The mechanism becomes dangerous when resources, status, fear or moral worth attach to the line. What matters for our question is what happens after the line exists. If I identify with a group, membership can become valuable. If membership becomes valuable, signals about whether I remain inside can acquire motivational force. The minimal group result alone does not prove any of that. It merely shows how little raw material humans sometimes need to begin constructing an us. The deeper problem is that the boundary can become morally lazy. Once there is an us, it becomes easier to keep a scoreboard that contains only us. Our jobs. Our safety. Our children. Our victory. Their cost disappears into the margin. That is why the destination of this book cannot be stronger tribes. Humans are already spectacular at tribes. The harder problem is preserving the warmth and coordination of us without needing somebody else to become less human. Maybe civilization is partly the project of making larger and larger groups feel real enough to count. Family. Village. City. Nation. Humanity. Each expansion asks the nervous system to care about people it cannot personally know. The line can be drawn for almost no reason. The work is deciding what the line is allowed to erase.

THE BOYS AT CAMP In the summer of 1954, researchers brought boys to Robbers Cave State Park in Oklahoma for what looked like summer camp. The boys were divided into two groups that developed their own identities, names and norms. Then the researchers arranged competition between them. The groups became the Eagles and the Rattlers. Friction escalated. Insults, raids and hostility followed. The famous final act came when the researchers introduced problems neither group could solve alone. A water supply problem. A truck that needed to be pulled. Goals that required cooperation. Hostility decreased as the boys repeatedly worked together. It is a beautiful story because it appears to contain a recipe. Divide people and make them compete, get enemies. Give enemies a shared problem, get cooperation. Reality is less clean. The Robbers Cave studies were highly engineered field experiments, and the research program had earlier attempts and methodological complications that simplified textbook retellings often omit. The boys were selected, observed without full awareness of the research, and placed inside situations the researchers deliberately shaped. We should not pretend somebody discovered the universal operating system for peace at summer camp. Still, the idea of a superordinate goal, a goal that groups cannot easily achieve separately, is powerful. It changes the geometry of the problem. The other group can move from obstacle to resource. But a shared enemy is also a powerful organizer. Two groups can unite beautifully by deciding a third group is the problem. Cooperation inside the circle tells us nothing by itself about the size of the moral scoreboard. So agreement by itself cannot be the goal. A conspiracy can agree. A firing squad can coordinate. A company can cooperate perfectly while dumping costs onto people who are not represented. The harder question is what the us is trying to accomplish, who is counted, and whether disagreement can still enter the system. Robbers Cave gives us the optimistic half: shared problems can reorganize relationships. The rest of this book has to earn the harder half: can humans build a shared problem large enough that solving it does not require inventing another them?

WHO ARE YOU WILLING TO DIE FOR? The defensible hypothesis is smaller and stranger. Humans can assign enormous value to outcomes that continue beyond the individual body. The scientific problem is to understand when, for whom, how strongly and why. If belonging helps determine which other lives enter the calculation, measure that. Do not worship it. There is a seductive conclusion available here: maybe every selfish act is secretly service to a larger us. We do not get to take it. If a person hoards food, we cannot invent a future plan to feed the tribe and declare the theory saved. Hidden motives require evidence. Otherwise CHOSEN becomes a machine that turns every possible observation into proof of itself. The terrifying part is that intense cooperation inside the beam can coexist with terrible treatment of people left outside it. Research on parochial altruism has documented versions of this uncomfortable combination. Humans may not become destructive because they stop caring. Sometimes the circle they care about becomes too small. This is where the magnifying glass becomes useful again. Pressure may narrow concern, but narrowing does not necessarily mean self. A starving parent may focus almost entirely on a child. A terrified soldier may focus on the people beside him. The beam gets smaller. The target can still be somebody else. The question I cannot shake is simpler: what exactly is the thing the person is trying to preserve? Sometimes the answer is obviously the body. Sometimes it is a child. Sometimes a family. Sometimes an identity, a people, a faith, a country or a future the person will never see. Before we call an action selfish or selfless, perhaps we should ask where the actor has drawn the boundary of self. That does not mean every martyr, soldier, parent, hunger striker, rescuer or terrorist is running one psychological program. Motives differ. Coercion matters. Ideology matters. Kinship matters. Honor matters. Despair matters. Psychologists studying identity fusion use the term for a powerful feeling that the personal self and a group are deeply joined. In multinational research, people who were strongly fused with a group reported greater willingness to make extreme sacrifices for it. Small groups such as family were especially powerful, and larger groups became more sacrifice-worthy when they felt more family-like. Evolutionary biology has been wrestling with versions of this problem for a long time. William Hamilton showed mathematically how costly behavior toward relatives can make evolutionary sense when the benefit to relatives, adjusted for relatedness, is large enough. The individual body can pay a cost while genes shared with relatives still benefit. Even biology had to become more careful about what self-interest means. And yet people go on hunger strikes. Parents give scarce food to children. Soldiers have died protecting other soldiers. People have accepted prison, torture and death rather than betray a cause, a family, a faith or a group. None

of that proves humans are secretly saints. It does prove that the body staying alive is not always the winning variable in a human decision. The body is extremely persuasive about staying alive. Hold your breath and notice how quickly philosophy loses the argument. Skip food long enough and hunger begins rearranging your attention. Put your hand on something hot and the nervous system does not schedule a committee meeting. It moves you. There is an obvious problem with saying survival comes first. Humans sometimes choose not to survive.

ORDINARY MEN On July 13, 1942, men from Reserve Police Battalion 101 entered Józefów in occupied Poland. They were not an elite killing unit in the way people sometimes imagine perpetrators of mass murder. Many were middle-aged reserve policemen from Hamburg, workers, clerks, salesmen and tradesmen. Their commander, Major Wilhelm Trapp, told them they had been given a terrible assignment. They were to round up the Jewish population. Able-bodied men would be selected for labor. The others were to be shot. According to postwar testimony examined by historian Christopher Browning, Trapp also offered an exit. After a pause, one man stepped forward. His company commander berated him, and Trapp stopped the commander. Then roughly ten or twelve more men stepped out as well. Others later found quieter ways to avoid or leave the killing. The option was not imaginary. It was simply not taken by most. By the end of the day, roughly 1,500 Jewish people, including women, children and elderly people, had been murdered. The battalion later participated in further mass shootings and deportations. Nothing about this chapter should make that violence feel abstract. The victims are not props in a psychology demonstration. Why did the men participate? There is no honest one-word answer. Browning emphasized conformity, deference to authority, career concerns, brutalization by war, ideology and the reluctance to leave comrades with the dirty work. Daniel Goldhagen later argued that deeply rooted antisemitism deserved much greater explanatory weight. Historians have debated the balance. The Nazi state, years of antisemitic propaganda, institutional authority, war, dehumanization and individual choice all matter. Belonging is not allowed to swallow the Holocaust. If CHOSEN ever produces the sentence they killed because they wanted to fit in, burn the page. The case matters for a more precise reason. Some men appear to have experienced refusal not only as refusal of an order, but as separation from comrades. If I step out, somebody else has to do what I refuse. What will the men think of me? Am I weak? Am I leaving them alone with it? Those questions can coexist with ideology and obedience. Human motives stack. That stacking is exactly what a serious model must handle. A person can hate the victim, fear authority, want promotion, avoid shame, follow peers and protect membership at the same time. The existence of one motive does not acquit the others. Explanation is not exoneration. A clean experiment cannot recreate Józefów, and it should not try. But ordinary group tasks can test the mechanism. Give participants a chance to refuse a harmful or unfair action. Hold authority and material incentives constant. Change whether refusal is framed as abandoning teammates or as an accepted individual choice. Measure refusal. If social framing changes behavior, that tells us something limited and real.

The moral lesson is not that anyone can be made to do anything. Many people refused, resisted, hid victims, deserted, lied to authorities or accepted punishment. Human beings are capable of conformity and defiance. The scientific question is what changes the price of each. The historical question is darker. How do institutions arrange a world in which cruelty becomes the cheaper social move?

EVERYBODY THINKS THEY ARE RIGHT That gives us a clean question for later testing: hold the evidence constant and change the social cost of changing your mind. Does secure membership make updating easier after identity, status, incentives and prior confidence are accounted for? If not, belonging loses ground. If it does, never make agreement the price of belonging becomes more than relationship advice. It becomes a rule for building better information systems. Now make updating socially expensive. Imagine that accepting new evidence means your friends think you are a traitor, your parents are ashamed of you, your community rejects you, or your profession has to admit it hurt people. The evidence has not changed. The price of believing it has. Humans make this harder because beliefs are not stored in isolated jars. They connect to people. A belief can tell me who my friends are, whether my parents respect me, what community welcomes me, whether I am a good person, and whether the last twenty years of my life made sense. Changing one proposition can threaten an entire social map. One useful way scientists describe this is updating. You begin with an estimate. New information arrives. A good system changes the estimate by an amount appropriate to the quality of the evidence. If the weather forecast changes from ten percent rain to ninety percent, taking an umbrella is not hypocrisy. It is learning. A belief is not the same thing as certainty. You can think rain is likely without knowing it will rain. You can trust your partner without possessing tomorrow's surveillance footage. You can believe a company will succeed, a treatment will work, a child is telling the truth, or a strategy will improve a country. Human beings constantly act before the future has agreed to reveal itself. Almost nobody wakes up intending to spend the day defending ideas they know are false. We believe our beliefs. That is the problem.

THE LAST NO Nine people say yes. One person says no. Most groups know what to do next. Convince the problem person, outvote them, ignore them, shame them or remove them. Then celebrate alignment. The last no is not necessarily the enemy of alignment. It may be the most valuable piece of information left in the room. The cultural change is still enormous: never make agreement the price of belonging. The moment people have to choose between telling the group what they see and remaining one of us, the group has begun corrupting its own information. Alignment therefore cannot mean unanimous identical thought. It means we understand the objective, understand the competing models, understand the remaining disagreement, and know what evidence could move us. Sometimes that produces agreement. Sometimes it reveals a real conflict of values that cannot be merged. Sometimes time runs out and somebody with legitimate authority has to decide. The rule has to work in reverse. If nine people say no and one says yes, investigate the yes. Semmelweis was not correct because he was lonely. The Challenger engineers were not correct because they objected. Minority status is not evidence. Majority status is not evidence either. Maybe the dissenter misunderstood everybody else. Maybe everybody else misunderstood the dissenter. Maybe the dissenter has a fact nobody else has. Maybe they are afraid. Maybe everybody else is afraid. Maybe the group is optimizing for one outcome while the dissenter is protecting another. Before removing the no, find out what produced it. No is information. Not truth. Not a veto. Information.

THE SCOREBOARD HAS TO CONTAIN EVERYBODY Maybe the most dangerous human sentence is not I do not care about anyone. It is I am doing this for my people, followed by a definition of people small enough to make everybody else's pain disappear. This does not solve moral philosophy. The equation gets ugly almost immediately. Good. Ugly is where the work is. Whenever somebody presents a plan that makes us better off, ask the irritating question: who is us? So nobody gets kicked out of humanity has to become more than a sentiment. It is a constraint on the scoreboard. My child can matter more to me than a stranger without the stranger's suffering becoming zero. My country can have special obligations to its citizens without foreigners becoming imaginary. A boundary can change responsibility without changing personhood. Every moral equation has a cheap trick available: change who counts. A policy can create extraordinary flourishing if we measure only the winners. A company can produce wonderful returns if the suffering it creates never appears on its financial statements. A family can look peaceful if the one person absorbing all the pain is told to stay quiet. That does not solve the hard questions. It creates them. What counts as flourishing? How do we compare present suffering with future benefit? What happens when helping one person harms another? What about animals? What about people who have not been born yet? What about pain that produces growth, meaning or protection from greater harm? Good. Those are not reasons to abandon the objective. They are the remaining math. Suppose we could agree on one objective before arguing about every strategy underneath it. Reduce unnecessary pain and suffering. Increase human flourishing. Keep learning how to do both better.

THE PROTOCOL The whole thing can be compressed further: choose the goal, share the models, protect the information, try the safest useful thing, measure what happened, change your mind, try again. If that sounds stupidly simple, good. The difficult part was never memorizing the instructions. It was building groups where humans can actually follow them. First, align on the objective as far as you honestly can. In this book the proposed objective is to reduce unnecessary suffering, increase human flourishing, keep learning how to do both better, and refuse to make anybody's experience mathematically disappear. That objective itself remains open to criticism. If somebody has a better one, put it on the table. Second, expose the plans. Not slogans. Plans. What do you want to do? What do you predict will happen? Who benefits? Who pays? What are you assuming? What would make you change your mind? A belief that cannot name evidence capable of changing it is no longer functioning as a prediction. Third, investigate the disagreement. Treat yes and no as information. Ask what each person sees, what they value, what they fear losing, and which facts differ between their models. Do not demand emotional agreement where factual understanding is enough. Fourth, merge what can be merged. Human arguments are often packaged as bundles. One plan contains a good diagnosis and a terrible remedy. Another contains a bad diagnosis and a useful safeguard. Stop requiring ideas to travel with the tribes that produced them. Turn competing beliefs into competing maps instead of competing tribes. Steal the useful parts from everybody. Fifth, act at the speed reversibility allows. If an experiment is cheap, safe and reversible, uncertainty is often an argument for trying it, not for endless paralysis. Endless certainty-seeking is still a decision. If an action can kill people, destroy ecosystems, start wars, permanently remove rights or create damage that cannot be repaired, uncertainty is an argument for much greater care. Sixth, measure the result against the shared scoreboard. Did suffering fall? Did flourishing rise? For whom? For how long? What did we fail to count? What new problem did the solution create? Seventh, update without exile. If your plan lost, you are still here. If mine lost, I am still here. Ideas do not need constitutional rights. People do. Eighth, repeat. There is no final meeting where humanity receives a laminated card saying congratulations, you have completed civilization. The protocol is the willingness to keep correcting the map while keeping the people who drew the bad parts inside the conversation whenever safety allows. The protocol is intentionally boring. Grand theories often become dangerous at the exact moment they become too beautiful to question. So ours needs to function less like scripture and more like a lab notebook.

Part Vii

Try To Kill The Idea

WHERE CHOSEN DOESN'T WORK A theory that explains everything explains nothing. If every act of conformity means belonging, every act of rebellion means searching for better belonging, every kindness means secure belonging and every cruelty means threatened belonging, close the book. We have built astrology. Humans do things for reasons that have little to do with social membership. Hunger can dominate a decision. So can pain, addiction, sexual desire, habit, curiosity, sensory pleasure, neurological disease, exhaustion and plain error. Status is not identical to belonging either. Researchers have argued that status can operate as a fundamental motive in its own right. Sometimes money really is the best explanation. Sometimes a person takes the higher-paying job because they want the money. Sometimes somebody leaves a group because autonomy matters more. Sometimes a dissenter accepts exile because truth or principle is worth the cost. CHOSEN gets weaker, not stronger, if I rename every one of those motives belonging. The useful tests are the ones where competing explanations predict different outcomes. Hold the financial reward constant and change only information about social membership. Hold status constant and change whether disagreement threatens continued inclusion. Hold the negative event constant and change whether the person has one reliable ally. Then measure choices, attention, stress, memory, cooperation, truth telling and whatever else the competing theories predict. If belonging security changes nothing once the other variables are controlled, the hypothesis takes a hit. If our proposed trust measurement cannot be defined consistently, throw it away. If a simpler model predicts behavior better, use the simpler model. I do not want a theory that cannot lose. I want to know where it loses. There is another failure mode worth making explicit. Sometimes people knowingly choose themselves. Not as an investment in the tribe. Not because survival will help the children later. Not because some invisible community benefits. They want the money, the sex, the drug, the status, the comfort, the revenge, or the extra hour of sleep. Any theory of humanity that cannot tolerate ordinary selfishness has confused aspiration with description. There are also cases where belonging can make outcomes worse. A tightly bonded group can protect a lie. A family can enforce silence. A team can punish whistleblowers. A movement can make cruelty feel virtuous. Secure belonging inside a small circle does not automatically create concern outside it. Sometimes the strongest us creates the sharpest them. That means the moral destination of CHOSEN cannot be more belonging at any cost. It has to distinguish secure membership from compulsory conformity, and intense local care from a scoreboard that erases outsiders. The theory should lose whenever manipulating belonging security adds no predictive power beyond simpler variables. It should lose when an alternative model predicts behavior more accurately with fewer assumptions. It should lose

when our proposed measures cannot be reproduced by independent researchers. And it should lose if the alignment protocol makes groups worse at finding truth. Those are not disclaimers hidden in the back of the book. They are invitations. If you can kill the theory cleanly, please do. A dead theory is useful if it clears space for a better one.

THE FAILED PROPHECY For seventy years, one of psychology's most famous stories about belief went like this. In 1954, a Chicago-area woman named Dorothy Martin said she was receiving messages from superior beings. A catastrophic flood was coming. Believers prepared for rescue by flying saucer. Researchers led by Leon Festinger infiltrated the group because they wanted to watch what happened when the prophecy failed. The flood did not come. The classic account, published in 1956 as When Prophecy Fails, said the most committed believers did not simply abandon the belief. They found ways to explain the failure and became more active in spreading the message. The case became a foundational story for cognitive dissonance: when reality collides with a belief tied to major commitment, the mind may work very hard to protect the investment. It is a fantastic story. There is now a serious problem with it. In 2026, historian of psychology Thomas Kelly published an archival reexamination arguing that the canonical account is materially wrong. Using newly unsealed records, he concluded that the group had actively proselytized before the failed prophecy, then recanted and dissolved after it failed. More seriously, Kelly argues that the researchers covertly manipulated events and that the authors knew central claims in the published account were false. That is a strong accusation from one recent archival paper, not a final vote of the field, but it changes what an honest retelling has to say. I love this problem. Not because it destroys cognitive dissonance. It does not. Decades of research on dissonance do not stand or fall with Dorothy Martin. I love it because it demonstrates the rule this book needs more than any particular result: the story does not get tenure. Imagine if I had written the older version beautifully. Midnight approaches. Believers remove metal objects so they can board the saucer. The clock passes. No ship. No flood. Then a message arrives explaining that their faith saved the world. The group suddenly seeks publicity. Perfect narrative. Perfect mechanism. Perfect chapter. And perhaps materially wrong. That is the danger of narrative nonfiction. Reality is under no obligation to preserve our scene structure. A story can be memorable enough that later writers repeat the interpretation without reopening the file. Eventually the explanation feels like the event itself. This book is especially vulnerable to the same mistake because belonging is such an intuitive lens. Once you start seeing it, every stubborn belief can look like membership defense. That is exactly when we need to become suspicious of ourselves. So keep the failed prophecy in the book, but invert its job. It no longer proves that groups protect belief. It proves that we have to let new evidence embarrass old stories. Kelly may be challenged by later scholarship. If he is, update again.

A theory of belief updating that refuses to update itself is performance art.

THE TWO BLANK SPACES The point is not to protect the answer. It is to improve the question. Attachment theory, relationship-specific working models, relational uncertainty, sociometer theory and the investment model already describe important pieces of this territory. If they explain the behavior fully, they win. CHOSEN does not need ownership of a variable; it needs the variable, if it exists, to be measured honestly. A first crude model says that a decision may depend partly on the value of continued membership, the person's current estimate that membership will continue, and the expected effect of an event on that estimate. That is not yet physics. It is a shopping list for measurement. The cleanest test independently manipulates V and S, gives everyone the same social event, measures S again, and compares preregistered predictions against established alternatives. If V and S cannot be separated experimentally, or the larger model does not predict new data better, the decomposition takes a direct hit. That is the novelty claim now: not that psychology forgot belonging, not that nobody modeled relationship security, and not that earlier theories were too static. The possible contribution is the deliberate separation and joint measurement of the value of continued membership and the estimated security of that membership, plus the change in security after a particular event, inside decisions where competing theories make different predictions. S is not simply attachment security with a new haircut, and V is not relationship value pretending nobody studied value before us. The question is narrower: can continued-membership security be measured separately from value, and do those measures add reliable predictive information after the established alternatives are already in the model? If not, use the better model and let these variables disappear.

Part Viii

Who Are You Choosing?

NOBODY GETS KICKED OUT OF HUMANITY Nobody gets kicked out of humanity sounds soft until you try to apply it to somebody you despise. Then it becomes a constraint. It does not mean everybody belongs in my house. It does not mean everybody gets my trust, forgiveness, money, attention, vote, friendship, body or time. It does not mean a violent person stays near the person they harmed. Boundaries are not exile from humanity. This matters especially for survivors. I do not get to turn a theory of belonging into a moral invoice and hand it to somebody who has been abused. Nobody owes an abuser forgiveness, reconciliation, understanding, access or love. You are allowed to close the door. You are allowed to lock it. The principle operates at a different level. The person on the other side of the locked door remains a human whose suffering and flourishing still exist in the moral calculation. Consequences can be severe without requiring dehumanization. That distinction is harder than hatred because hatred simplifies. If the other person becomes a monster, I no longer have to think about competing goods, rehabilitation, deterrence, safety, proportionality or what produced the behavior. The scoreboard gets cleaner because I deleted a row. Keeping the row does not tell us what consequence is correct. It forces us to admit the consequence has a cost. The same rule applies upward. Being one of us does not purchase immunity. A loved child can require consequences. A loyal employee can be fired. A friend can lose access. Secure belonging is not a promise that every role lasts forever. The largest version of belonging is therefore almost strangely thin: you count. Your experience does not become zero because I oppose you. Maybe humanity's next move is not learning to stop forming tribes. We may be incapable of that, and tribes can be wonderful. Maybe the move is learning to build an us without requiring anyone outside it to become nothing.

THE BOOK I HOPE SOMEBODY BEATS Human beings have been leaving notes for one another for a very long time. Stories. Laws. Scriptures. Equations. Constitutions. Experiments. Songs. Children's books. Operating manuals. We keep trying to tell the next person what we learned before we disappear. Some of those notes changed the world. None finished it. That should make us humble, but it should not make us passive. The fact that no map contains the entire territory is not an argument for driving without one. This is my map. Being chosen matters. That part is not radical. The unresolved question is the one we started with: can the two numbers, and the updates between them, predict anything we were missing? I think those estimates may change the meaning of ordinary events. A criticism can be information or exile. A new friend can be harmless or replacement. A mistake can be local or become evidence that I no longer belong. Same event. Different update. I think groups destroy information when disagreement threatens membership. That idea overlaps with established work on psychological safety, conformity, ostracism and social identity. The job is not to rename those fields. The job is to test whether perceived membership security adds useful predictive power after those variables are already in the model. I think trust may affect how quickly people expose uncertainty and build enough shared understanding to act. I do not know whether that deserves an equation. If we cannot define the units, reliability and predictive advantage, it does not. I think the boundary around us may help explain why humans can be breathtakingly generous and breathtakingly cruel. But I do not think every sacrifice is belonging, every selfish act is secretly altruism, or every atrocity is conformity. The theory has to survive the cases where hunger, status, ideology, money, sex, habit, fear, disease, temperament or plain old self-interest explains more. I think a civilization worth building needs a scoreboard that does not make suffering disappear merely because the person experiencing it is outside our favorite circle. I do not know the perfect formula for flourishing. I do know that changing the denominator until your plan looks good is bullshit. Most of all, I think the next step is measurement. Run the discriminating tests. Make the competing models predict different outcomes before the data arrive. Then replicate them. Run it in another country. Run it with people who do not look like the first sample. Preregister the analysis. Let researchers who hate the theory design the strongest test. Publish the null result. Find the boundary condition. Find the embarrassing case.

If membership security turns out to explain nothing after existing constructs are measured properly, fantastic. We can stop talking about it. If it explains something small and reliable, quantify the small thing. If it explains something large, make the model better. If trust velocity remains a metaphor, retire the phrase. If somebody finds a cleaner variable, use theirs. Then improve the protocol too. Find a better definition of flourishing. Find the people our scoreboard forgot. Find the situation where discussion becomes paralysis. Find the situation where speed becomes recklessness. Find the case where secure belonging strengthens the wrong group and makes outsiders less safe. Bring back the evidence. I do not want CHOSEN to become another belief people defend because admitting it was wrong would cost them membership in the CHOSEN tribe. That would be a spectacularly funny way to miss the point. So if you find an error, you do not owe the book loyalty. The goal is not for CHOSEN to be right forever. The goal is for the next person to start from farther ahead. Choose everyone. Listen to the no. Test the plan. Measure the suffering. Measure the flourishing. Change your mind. Try again. And if you can build a better map, please do not protect mine out of politeness. Beat the fucking book.

A NOTE ON EVIDENCE This book deliberately mixes established findings, historical interpretation, personal experience and new hypotheses. They do not deserve the same confidence. The central CHOSEN proposal is a measurement hypothesis: decisions may become more predictable if we separately measure the subjective value of continued membership, the estimated security of that membership, and the change in that security estimate after a specific event, then compare the resulting model against established alternatives. It is not an established scientific law. The belonging foundation includes Roy F. Baumeister and Mark R. Leary, “The Need to Belong: Desire for Interpersonal Attachments as a Fundamental Human Motivation,” Psychological Bulletin 117 (1995); John Bowlby's work on attachment and internal working models; R. Chris Fraley's work on global and relationship-specific attachment representations, including Fraley, Heffernan, Vicary and Brumbaugh's 2011 ECR-RS measure; Ximena Arriaga, Madoka Kumashiro, Jeffry Simpson and Nickola Overall, “Revising Working Models Across Time,” Personality and Social Psychology Review 22 (2018); Leanne Knobloch and Denise Solomon's work on relational uncertainty, including “Intimacy and the Magnitude and Experience of Episodic Relational Uncertainty Within Romantic Relationships,” Personal Relationships 9 (2002), and the later relational turbulence framework; Mark R. Leary and colleagues' sociometer theory; Caryl E. Rusbult and colleagues' investment model of commitment; James A. Coan and David A. Sbarra, “Social Baseline Theory: The Social Regulation of Risk and Effort,” Current Opinion in Psychology 1 (2015); and Amy C. Edmondson, “Psychological Safety and Learning Behavior in Work Teams,” Administrative Science Quarterly 44 (1999). Key experiments and cases discussed include Solomon E. Asch, “Studies of Independence and Conformity,” Psychological Monographs 70 (1956); Uri Gneezy and Aldo Rustichini, “A Fine Is a Price,” Journal of Legal Studies 29 (2000); Alain Cohn, Michel André Maréchal, David Tannenbaum and Christian Lukas Zünd, “Civic Honesty Around the Globe,” Science 365 (2019); Gregory M. Walton and Geoffrey L. Cohen, “A Brief Social-Belonging Intervention Improves Academic and Health Outcomes of Minority Students,” Science 331 (2011); Gregory M. Walton and colleagues, “Where and with whom does a brief social-belonging intervention promote progress in college?” Science 380 (2023); and William B. Swann Jr. and colleagues, “What Makes a Group Worth Dying For?” Journal of Personality and Social Psychology 106 (2014). For Challenger, the manuscript relies on the Presidential Commission on the Space Shuttle Challenger Accident and related historical records. For NUMMI, it draws on This American Life's reporting and interviews with workers, union representatives and managers. For Ignaz Semmelweis, the mortality figures and chronology are drawn from historical and medical scholarship on the Vienna General Hospital. Robert Munsch's account of Love

You Forever, including the two babies born dead, the song that grew from that grief, the publisher change and the book's approximate fifteen-million-copy sales total, comes from his official website. The oxytocin chapter reflects the influential 2005 Nature paper by Michael Kosfeld and colleagues and Carolyn Declerck and colleagues' 2020 registered replication, which did not reproduce a general increase in trusting behavior in its main social-contact condition. The chapter therefore treats the original result as an important finding in a mixed literature, not as evidence for a universal 'love' or 'trust' molecule. The chapter on When Prophecy Fails reflects Thomas Kelly's 2026 archival paper in the Journal of the History of the Behavioral Sciences. Kelly argues that the group proselytized before the failed prophecy, recanted and dissolved afterward, that researchers covertly manipulated events, and that the original authors knew central published claims were false. Because this is a recent archival challenge, the chapter reports Kelly's claim directly without pretending the scholarly record is closed. Historical cases involving mass violence are not presented as demonstrations that belonging alone caused the behavior. The discussion of Reserve Police Battalion 101 follows Christopher R. Browning's Ordinary Men; Browning describes roughly a dozen men out of nearly 500 stepping forward at Józefów, while the chapter also acknowledges the broader debate over ideology, antisemitism, authority, conformity, career concerns, coercion and wartime institutions. The discussion of sacrifice also draws on W. D. Hamilton's work on inclusive fitness and research on identity fusion and parochial altruism; these mechanisms are presented as alternatives or complements, not as proof of CHOSEN. Where later evidence changes any story in this book, the story should change. That is not a defect in the project. It is the project.