Most of this is already good. Most of it is easy, and funny, and the best part of both our days. This is the place we look after the rest, so that the good part gets to be the whole story rather than the bit in between.
Where she came from, what she carries, what she is fiercely good at, and what she is quietly working on.
Where he came from, what he does when things get hard, and what he wants out of a life with her.
The promises we have already made each other. Short, on purpose.
Everything we are actually working on, and how far through each piece we are.
The time we are setting aside. Coffee and evenings out as much as hard conversations.
What we have read, what it explains, and how much of it to trust.
Two people doing their own work, next to each other, on purpose. Not one project and one manager. Not one patient and one nurse. Two whole people who like each other enormously and want to be better at the small number of things that get hard.
There are two pages about us as people, one for each, and neither of us writes in the other's. Everything else is shared. Commitments holds only what we have actually agreed and stays short. Workspace holds everything still in motion, broken into pieces small enough to finish. Schedule is when, and it has as much coffee in it as it does hard conversation, because the good part is not a reward for getting the difficult part right. It is the point.
Neither of us uses the other's page as evidence. Nothing written honestly about yourself comes back later as "well you said yourself that you...". These pages only stay true if they are safe, and they are the most useful thing here.
This is a portrait, not a file. It has no boxes to fill in, because a person is not a worksheet.
She grew up having to run things no child should have to run, in a house that gave her neither safety nor slack, and she came out of it competent in a way that most people never have to become. She reads a room before she is through the door. She sees the practical answer while everyone else is still describing the problem. She is fast, direct, and unafraid of a hard conversation, and a great deal of what works in both our lives works because she is the one who saw it and moved on it.
The cost of that is a nervous system that never got taught how to be looked after. Being cared for lands strangely on her, because for most of her life the arrangement ran the other way. When something goes quiet she does not read it as calm, she reads it as the beginning of being left, and she goes toward it hard, because going toward things hard is the strategy that kept a small girl safe. That is where the fierceness comes from. It is armour built by someone who genuinely needed armour, and it has never been the same thing as cruelty, whatever it feels like from the inside afterwards.
Her deepest fear is that she is turning into her mother. She is not, and the evidence is the fear itself. She stopped. She got frightened. She went and read the entire research literature and built a document to try and fix it, which is not something a person who has stopped caring how she affects people does. That is what breaking a cycle actually looks like from the inside: unglamorous, self-critical, and largely invisible to the person doing it.
She is also very funny, extremely good company, better read than she lets on, and the person I want to tell things to first. Ninety percent of this is not a problem to be managed. It is the best part of my life.
This is Nick's reading of you and it is certainly partly wrong. Replace any of it, or all of it, with your own. The version of this page that matters is the one you write.
The same, from the other side. Short, because the working material lives elsewhere now.
I grew up in a house with enormous warmth and no edges, where nobody was ever criticised and nobody ever set a rule, and in the absence of structure I became the one who supplied it. I was the parent to my sister. I learned that a family stays afloat when somebody is easy to be around, and I got very good at being easy to be around. It made me steady, patient, and genuinely hard to rattle, which are the parts of me I like.
It also means I do not say what I want. I bend early, I bend cheaply, and I keep a quiet count of it that nobody can see. When something gets difficult my first instinct is to lower the temperature by taking myself out of the room, and I know now that what I experience as protecting us lands on her as being left. I would rather be honest about the whole of that than only the flattering half: some of the going quiet is kindness, and some of it is that disappearing works and I have used it.
What I want is simple and I have never said it in one place. I want to marry her and have children with her. I want a life with a lot of unremarkable good days in it, where we are two separate people getting on with our own things and glad to find each other at the end of the day. I want to be trusted with the soft part rather than only the capable part. And I want to be the kind of partner she does not have to manage, which means saying things at the time, in daylight, at normal volume.
Only what we have actually agreed. Short on purpose. Anything still being worked out lives in the Workspace.
Nobody makes a good decision about how to fight while they are already fighting, because the part of you that makes good decisions is exactly the part that goes offline. The only moment either of us has real authorship over our behaviour is beforehand. So we decide beforehand, together, while we are fond of each other.
A time-out without a stated return time does not read as a break. It reads as abandonment. The return time is the bridge across both of us at once: it turns he has gone blank and I am disappearing into I am calming down, not leaving, and I am coming back to you at half four. Same act, opposite message.
| Off limits | Why this one |
|---|---|
| Contempt | Sneering, mockery, the voice, the eye-roll. The strongest single predictor of divorce in the research, and it does not only hurt: it disables the ability to receive a repair attempt at all. Once someone is being sneered at, no joke lands and no apology reads as sincere. |
| Character assassination | Complain about a behaviour. Never attack the person. You cannot un-hear it. |
| Kitchen-sinking | Throwing in every past grievance turns a conversation into a threat, which triggers flooding, which ends any chance of resolution. |
| Threatening the relationship | Your partner is the person your nervous system relies on for safety. A threat to the relationship is not heard as a hurtful thing said in anger; the body reads it as a survival threat, and it teaches you permanently that the bond can be revoked when someone is angry. |
| Invoking each other's parents | It fuses the safe person with the original source of harm. Devastating and self-fulfilling. |
| Silence as a weapon | Nick's addition. Going unreachable is his version of the same move and it does the same damage. |
The stop signal, the floor object, the repair vocabulary, the if–then plans: all still being worked out, all in the Workspace. Things graduate onto this page once they are settled, not before.
Everything not yet settled, in pieces small enough to finish. The bar shows how much of each has actually been filled in.
Wants, not needs. Not what would fix anything. What you would choose if nothing were broken.
A limit is a fact about you, not an accusation about the other person.
Named in the calm so they do not have to be discovered in the middle of something. Each one comes with what you will do instead of the old behaviour, because a limit without an alternative is just a demand.
What each of us is doing regardless of what the other does. Not a trade.
What each of us is working on alone, what help looks like, and what does not help.
The point is not disclosure. It is that neither of us is guessing, and that guessing is where most of our worst fights actually started.
Anything either of us has got wrong about the other. Flagged means rewritten, not defended.
A way of taking a topic all the way to done, so it stops being available as ammunition.
Sancha's document borrowed a word from Boszormenyi-Nagy: the ledger, the running childhood account of what was given and what was owed, the one that never balances and never closes. This is the opposite thing, deliberately. A ledger where items close.
Nothing between us is ever formally over. Nothing is filed, so when we are flooded all of it is available, because none of it was ever put away. That is why a fight about a Saturday plan can reach back eighteen months. It is not kitchen-sinking on purpose. It is that the kitchen sink is the only storage we have.
A topic taken to resolution and written down as resolved is not available as ammunition. Not because we are being disciplined, but because there is a record, we both signed it, and reaching for it means contradicting something we agreed while calm.
| State | Meaning | What we may do with it |
|---|---|---|
| Open | Named and written down, not yet discussed properly. | Raise at a Ledger session. Not in a fight. |
| Talking | Started, not finished. | Continue next session. Either of us may add in writing between sessions. |
| Landed | Each of us can state the other's position in a form the other accepts. Understanding done, decision not yet. | Move to a decision, or accept there is not one and close anyway. |
| Closed | Finished. Either resolved, or accepted as a difference we stop relitigating. | Nothing. Off the table. |
| Parked | Deliberately postponed, with a date. | Nothing until the date. Parking with no date is avoidance, which is Nick's speciality, so watch him on it. |
The company, the hours, the cost, and what we each expected it to be.
Worth splitting the operational questions from the emotional ones before the afternoon starts, because otherwise they get fought as one thing and neither gets answered.
Nick's item, opened by Nick.
His to work on and hers to understand, not the other way round. More likely to close if it starts from what it feels like inside rather than from an incident, because starting from an incident turns it into a trial of that incident.
The one that needs the most protection, not the most avoidance.
Last time it came up it ended the relationship. So: the floor from the very start rather than only once it is hot, and a hard stop at two hours regardless of where it has reached. An unfinished conversation both people walked out of intact beats a finished one that cost a fortnight.
Where the line sits between this is mine and this affects us.
Nick's item. The proposed test is whether something directly affects the other person, applied honestly in both directions. Get the principle agreed in the abstract before testing it on an example, because starting from an example turns it into a fight about that example and the principle never gets settled.
The two physical things the whole protocol runs on. Neither chosen yet.
The signal has to be neutral and slightly silly, something neither of us could possibly hear as contempt, chosen together while calm. That is what makes it a repair attempt that cannot be misread. The floor is a physical object: whoever holds it is the speaker, and the other person cannot be speaker without it.
What actually lands on each of us, and what looks like repair but reads as an attack.
The most counterintuitive finding in the research: what separates couples who last from couples who do not is not how hard they fight, it is whether a repair attempt gets received. Almost everyone makes them. In some couples the clumsy joke lands and the temperature drops; in others the identical joke reads as mockery.
Nick's primary repair attempt is going quiet. In his mind it is de-escalation. It lands on Sancha as the still face: present but unreachable. So his most sincere repair attempt is structurally guaranteed to arrive as its opposite. The fix is not for him to stop needing space. It is to make the space legible, which is what the stated return time does.
Written together, out loud, in one sitting.
If a sentence cannot be agreed, both versions go down and the disagreement becomes a Ledger item rather than a fight.
Twenty minutes is physiology, not etiquette.
The stress chemistry has to physically clear and that takes real time. Twenty minutes is a floor; aim for thirty. Do not come back at minute eight because the silence is unbearable, because that is exactly the moment the agreement exists to protect. Ceiling of a day, beyond which it stops being regulation and becomes punishment. Still flooded when the time comes? Extend it out loud. "I am still not there, can we make it half eight" is following the protocol, not breaking it.
The signal, then: "I am too flooded to do this well. I am not leaving you. Let's come back to it at half eight." Not: "I need space," followed by a door.
Breathe with a longer out-breath. Leave the space. Do something absorbing. Cold water on the face works fast and is a hard-wired brake, not a relaxation technique.
What kills the break: rehearsing, cataloguing, phoning a friend to re-run the prosecution, drinking. Your body cannot tell the difference between fighting and vividly imagining fighting. That is exactly how twenty minutes becomes three days: two people go to separate rooms, each builds a better case, and they reconvene more flooded than they left.
You come back with your heart rate down and your curiosity back on. If you come back with a better argument, you did it wrong.
Do not just start talking again. Unstructured re-entry is how couples flood each other twice.
No problem-solving yet. The purpose at this stage is to understand and be understood, not to reach a solution. Premature problem-solving feels like being steamrolled. And validation is not agreement: it means you can see that their reality makes sense from where they stand.
Seven prompts, filled in after a break, to organise what you want to say before you say it.
One or two sentences each. Brevity is not optional: the capacity to listen to a partner who is upset with you runs out at about four sentences.
Why our agreements do not stick, and the one fix with real evidence behind it.
It is not character. A vague intention needs exactly the faculty that flooding disables. An if–then plan hands the behaviour to the situation rather than to willpower in the moment, and the effect size in the research is several times larger. Convert every vague agreement into one. Specificity is the active ingredient.
So that nothing sits unanswered by accident.
Who we are seeing, who we still need to find, and who is chasing it.
Individual work and couples work are separate tracks and should not share a therapist. The strongest evidence for couples work is with attachment-based approaches where the therapist is trained in the model rather than borrowing from it, and the best predictor of whether any therapy works at all is the fit with the person in the room, not the method.
A time-out you have never rehearsed will not deploy under fire.
Same reason an evacuation you have never drilled turns to chaos. Under real flooding neither of us will rise to the occasion; we will fall to our level of practice. So it gets practised cold, monthly, on something trivial.
That is correct and expected. The awkwardness is the pathway being built. Two adults play-acting a stop signal over something trivial while both slightly embarrassed is doing real work, and it is the cheapest thing here.
Every six weeks. Is any of this working, and what do we delete?
Drag to move. Click to edit. Double-click a day to add. Export any item, or the whole month, into your own calendar.
Not all of this is work, and the ratio matters more than any single item. What keeps couples steady is the reservoir of good time, not the absence of hard time. Booking only the difficult conversations would make this worse rather than better.
One topic from the Workspace, taken as far as it goes. Floor from the start, agreed close condition, hard stop even mid-sentence.
Out of the house, no agenda, no Workspace, no logistics. If either of us reaches for something from the Ledger, the other says "that is Wednesday" and it goes back in the box.
Booked, not improvised, and treated as unmovable. The pleasant part is the first thing to erode under strain, because it is the only part with no deadline attached to it.
Booked whether or not anything is wrong that month. Going only when things are bad teaches both of you that the appointment means things are bad.
Appreciations first, always. Then what went right. Then one issue, and only one. It is almost impossible to attack someone while you are thanking them, which is why the order is fixed.
Working the backlog toward closed. One or two items, never five. Stop on time; anything unfinished moves to Talking.
Practise the fight, cold, on something small and genuinely real. It will feel ridiculous, which is the point.
Deliberately scheduled separateness, so that space becomes something either of us can take without it reading as withdrawal. This is the whole point of two people side by side rather than one merged unit.
Money, diaries, house, travel. It exists so that admin stops leaking into evenings and stops being the trigger for something that was never about admin.
Is any of this working, and what do we delete? The likeliest failure is the Ledger becoming a place things get filed to avoid them.
On the shared calendar only so nothing gets booked over it. What happens in it is not shared unless she wants it to be.
Same terms. Visible as a block, not as a subject.
Dragging something is allowed and expected. Cancelling is not. If a week is impossible, move it. A skipped State of the Union becomes two skipped, and then it is gone, and then everything routes through fights again because there is no other channel.
Where all of this comes from, and how much weight each piece can actually bear.
The primary document is Sancha's brief, Why We Fight Like This. This is the short index of what it draws on, so that when something here sounds like an assertion you can check whether it is a well-replicated finding or someone's clinical framework.
| Framework | Who | What it explains here, and how solid it is |
|---|---|---|
| Flooding | Gottman | Why repair plans evaporate mid-fight: past roughly 100bpm the prefrontal cortex is partly offline and neither person can reach their reasonable self. Four decades of observational data. Solid. |
| Repair attempts | Gottman | That survival depends not on how hard you fight but on whether a repair attempt gets received. The single most useful finding here. Well supported. |
| The Four Horsemen | Gottman | Criticism, contempt, defensiveness, stonewalling, with contempt as the strongest single predictor of divorce. Well supported. |
| Pursue and withdraw | Christensen & Heavey | The demand and withdraw pattern, more common in distressed couples and predictive of decline. Well replicated. |
| Attachment and EFT | Sue Johnson | That the fight is rarely about its content and usually about whether you are there for me. Strongest evidence base of any couples method, with the caveat that outcomes track how faithfully the therapist follows the model. |
| The still face | Tronick | Why going quiet is not neutral but one of the most distressing things a nervous system can receive. Among the most replicated findings in developmental psychology. |
| Hot and cool systems | Metcalfe & Mischel | Why a plan made calmly is unavailable to a flooded person. The reason this is a protocol rather than a set of good intentions. |
| Implementation intentions | Gollwitzer & Sheeran | If–then plans substantially outperforming goal intentions across a large meta-analysis. The most directly actionable science here. |
| Parentification | Boszormenyi-Nagy, Jurkovic | The relational ledger and destructive entitlement. A clinical framework rather than an experimental finding, but the descriptive fit is unusually close. |
| Earned secure attachment | Main, Pearson, Roisman | That a difficult childhood is not destiny, and that what predicts breaking the cycle is being able to tell a coherent story about it rather than having avoided the adversity. The most hopeful finding in the whole literature. |
| The adaptive child | Terry Real | That under threat the old survival strategies take the wheel. A framework rather than a measured phenomenon, but a precise vocabulary for something real. |
| Lapse, not relapse | Marlatt | That the catastrophic reaction to a slip does more damage than the slip. Why breaking the protocol once is not permission to abandon it. |
Research and information, not a clinical assessment, not a diagnosis, and not a substitute for care from people who know us. Everything here points toward that rather than replacing it.