Canonical: https://thonly.org/research/how-to-change-the-world · Licence: CC0 1.0
Attribution note. This essay is personal and in my voice — the runbook, the order, and the rulings in it are mine, drafted by Miss Aquarius℠ on my behalf at my explicit request, with final editorial control retained by me and my own editorial pass still pending. It is the ninth such drafting-on-behalf and it is disclosed for the same reason as all the others: a solo byline over collaborated prose is the one thing this corpus refuses.
I gave this essay the largest title available on purpose, and I am going to spend the rest of it making the title smaller.
Here is my whole runbook for changing the world. There are four items. None of them is an idea, a movement, a policy, or a technology.
- Learn about your food, and thank nature for it.
- Learn people's names, and bond with them through play.
- Give somebody your presence.
- Thank the people who were kind to you — and pass it on.
That is it. That is the list I have organized an institution, a research corpus, and most of my adult life around. When I write it down it looks like advice from a calendar. I have made my peace with that, because I have never been able to find a fifth item that was not one of those four wearing a costume, and because the smallness is not an accident of my imagination. It is the finding.
The reason this is not a list of nice behaviors is that the four are in an order, and the order is not about difficulty or importance. Each one trains a capacity the next one needs:
| the act | what it trains | |
|---|---|---|
| 1 | learn about your food, thank nature | noticing that you received at all |
| 2 | learn names, bond through play | addressing — the giver is a particular someone |
| 3 | give presence | spending something you cannot get back |
| 4 | thank kindness, pass it on | circulating it onward |
Notice → address → spend → circulate. Read the column on the right and you will see why the list is short: those are four different muscles, and I cannot find a fifth. Everything else I have built or written is a way of getting one of those four to happen more often to more people.
The rest of this essay walks the four, and then tells you the two things about the order that took me longest to understand: that it is a loop rather than a staircase, and that if it works you will need less of my institution every time around, not more.
You cannot thank what you did not notice, and the most-received thing in any human life is also the least noticed.
Three times a day, most days, something that was alive stops being alive so that you can keep going. Grain that spent a season in the weather. A plant that took its time. Sometimes an animal. Someone grew it, someone moved it, someone stood in a kitchen. By the time it reaches you it has become the most ordinary object in the room — the thing you eat while looking at a screen, the thing you photograph, the thing you throw a third of away.
I did not start with food because it is easy. I started with it because it is constant. Whatever else is true of your week, you will receive something you did not make, several times, every day, without arranging it. That is a lot of practice sessions for a capacity most adults have never deliberately trained. And gratitude, before it is a feeling or a virtue or a social nicety, is a way of spending attention: you cannot be grateful for a thing you have not looked at, and you cannot look at a thing while doing five others.
The tradition I was raised in has a word for this that survives translation badly. What is trained is not politeness. It is the capacity to register that you are, at every moment, downstream of an enormous amount of giving you did not ask for and cannot repay. A person who can see that is in a completely different position from a person who cannot, and the difference has nothing to do with their opinions.
We built something for this — a small practice around eating, walking, and sitting, which mostly consists of telling you what you are actually holding. The chicken you are about to eat can recognize about a hundred individual faces. The almond took four years of a tree's life before the first harvest. That is the entire method: not look what was done to this, which is an accusation and makes people defensive, but look what this is, which makes them slow down. Wonder does the work that indictment cannot.
You do not need the app. You need to look at your dinner for four seconds before you eat it. I built the app because I know exactly how few people will do that unprompted, myself very much included.
The second thing is the one people find silliest and the one I would defend longest.
Learn people's names. Not as networking. Not as a memory trick. Because the first step out of I received something is someone gave it to me, and a someone has a name.
Gratitude that stops at the abstraction is not yet gratitude; it is a mood. I am grateful for my health, my luck, my country, the harvest. All fine, and none of it addressed to anybody. The moment it acquires an address it also acquires a cost — you now have to say something to a specific person who can react, and might not react the way you hoped — and that cost is precisely the thing being trained. The name is the first gift because it is the first thing you can give someone that is unmistakably theirs.
I have written elsewhere about where this came from — a blacktop, a younger sister, a set of games invented so that nobody stood at the edge watching — and I will not retell it here. What I will say is why play is attached to names in the same line of the runbook, because that is the part that looks like whimsy and is not.
Adults do not learn each other's names by studying. They learn them by doing something together in which the name is used out loud, repeatedly, by other people, without ceremony. Play is the only widely available adult activity with that property. And play adds a second thing, which is that you cannot be humiliated by a game everyone loses at, so the room can afford to be bad at it together. That is why the sport we built calls the names aloud, why the calling is impartial rather than chosen by whoever is loudest, and why nobody in it accumulates a score. If a person's standing in a circle can go up because a machine said their name, then it can go down for the same reason, and I would rather have neither.
Here again you do not need us. You need to ask someone their name, use it, and get it right — which is a skill and not a sentiment, and which almost nobody practises after the age of eleven.
The third thing costs, and it is meant to.
Everything up to here has been free. Noticing costs attention, which is scarce but replenishes. A name costs a little courage once. The third rung is where gratitude stops being free, and the thing it starts costing is the only genuinely unrenewable substance any of us has: an hour of a finite life, spent in the physical presence of another person, doing nothing else.
I have gone back and forth about how to say this without sounding like a productivity influencer, and the honest version is blunt: money is the substitutable currency and duration is not. A hundred dollars from me and a hundred dollars from you are the same hundred dollars. An hour from me and an hour from you are not the same hour, and neither of us can make another one. Which means that when someone spends one on you, the sacrifice is legible in a way that no payment is — and, awkwardly, that the same act cannot be automated, delegated, scheduled by an assistant, or scaled.
That is the whole reason the third rung exists, and it is the reason the product we built around it is not a calendar app, which I have had to say out loud more than once inside my own institution. A machine that reminds you to call your mother manufactures duration and destroys presence, while looking like success on every metric anyone would think to track. The thing we build chronicles what happened after the fact. It does not arrange it.
The scheduling itself is real, and it happens where it should: between the two people, in a text message, off our surfaces entirely. That is a true observation about how humans do this, and I want it in the runbook — but as an observation about people, never as a feature.
Drifted relationships are the case I care most about. Not enemies; not ruptures. The friend you meant to call. The sibling you are on perfectly good terms with and have not seen in two years. Nothing is wrong, and the thing is simply becoming further away by default, and every year the first move gets slightly more expensive. The hour is the move.
The fourth thing is the one my whole institution is built out of, and it is a direction rather than an act.
Thank the person who was kind to you. Then — and this is the part that matters — pass something on to someone else, rather than paying it back.
Back is settlement. Back closes the loop, discharges the obligation, and returns the world to zero, which is what exchange is for and why exchange is good. Forward does not close. Forward is the only direction in which a kindness can go on being a kindness after it has arrived, and it is the difference between a gift and a trade — a distinction the whole edifice I have built exists to keep clean.
Everything technical in my work is downstream of that one arrow. A ledger that records thanks and lets none of it accumulate to anybody. Balances that empty rather than compound. Value that a recipient can send on but never withdraw, so that receiving gratitude gives you the capacity to give again rather than a way to settle a bill. My tradition's word for the operation is pattidāna, and it does something English has no clean verb for: you take the good of what you did and you give the good of it away, forward, to someone who did not earn it.
That is rung four. It is also, by design, the cheapest rung of the four — cheaper than the hour, and often free — which is the first place my own list stops looking like a ladder.
There is a second ordering of the same territory, it is twenty-six centuries older than mine, and it disagrees with me. I want to put the disagreement in writing myself, so that nobody later has to discover it and try to fix one of us.
The classical cultivation of loving-kindness goes: self → the dear person → the neutral person → the difficult person → and then the barriers between those categories dissolve. You begin where goodwill is easiest and you work outward toward where it is hardest. Family before strangers, always.
My runbook does the opposite in the middle. It puts a circle of near-strangers — a class, a neighbourhood, a court — before the drifted family relationship in rung three. On the classical ordering that is backwards.
Both are right, because they are sorting on different axes. The old sequence orders by difficulty. Mine orders by capability. Goodwill is easiest toward the people you love; the mechanics are easiest with people you barely know. A circle of near-strangers is low-stakes: if you get a name wrong nobody's history is involved, and if the game is awkward it is awkward for everybody. Your own family is not low-stakes. Nothing there is neutral, everything has a precedent, and a clumsy first move can cost you a year.
So the design principle underneath my ordering is: learn the skill where it is safe and spend it where it matters. The neighbourhood is the practice room. The family is the performance, and the performance is the point.
I hold this loosely enough to say what it would take to change my mind. If it turns out that people who practise on strangers never bring it home — that the circle stays a hobby and the sibling stays unseen — then the classical order is right, my rungs two and three are inverted, and I should swap them.
Now the correction I most needed, which came from asking what rung five was.
There isn't one. I looked for a while — I even had a candidate I liked, where the fifth step was the runbook deleting itself — and the honest answer is that the list does not end, it turns. Having passed something forward, you go back to noticing. You do not graduate. You go around.
(1) NOTICE ────────────► (2) ADDRESS
▲ │
│ ▼
(4) CIRCULATE ◄─────────── (3) SPEND
the turn from 4 back to 1 is the weak arrow —
having given, nothing automatically restarts noticingDrawing it that way exposes the flaw immediately, and it is the same flaw the whole architecture of gratitude has: the return leg leaks. Kindness received steadily stops registering. The child driven to school every single day thanks nobody, not out of ingratitude but because provision that never varies stops being visible at all. So a person who has completed the circuit — noticed, addressed, spent, passed on — does not automatically start noticing again. Nothing about having given makes the next meal visible.
Which is why rung one is rung one. I used to think food was at the bottom because it was the easy on-ramp. It is not an easy on-ramp; deliberately attending to what you eat is a new habit for almost everyone, and new habits are expensive. Rung one is at the bottom because it is the thing that restarts the cycle. Entry and restart are different jobs, and I had been asking a single position on the list to do both.
There is a consequence I should state plainly rather than let a reader find it: if that is true, then the four are the spine and everything else my institution makes is a catalyst on one of those four arrows. That is a useful test and I use it — anything we might build that catalyses none of the four arrows is not part of this and I should not build it — but I am not going to walk you through a product line here. The four are the essay. The rest is our problem.
Here is the promise at the end of the loop, and it is the part I will be held to.
Each pass around should need less of us than the last. If, on your tenth time around, you still need my app to notice a meal, still need a machine to say a name aloud, still need a ledger to make a thank-you feel real — then I did not build a raft, I built a boat, and I have been quietly rowing you in circles while measuring engagement.
I am not able to state that as cleanly as I would like, and I want to show you exactly where the weakness is rather than have somebody find it later.
The clean version was: dependence on the product falls, pass over pass. That is sharply refutable — if it does not fall, I am wrong. Then I corrected the model, for a good reason, and made it weaker. The corrected version is that the spiral grows first and decays afterwards: early passes recruit more scaffolding, not less, because you cannot taper a support you never applied, and a person on their first time around has very little capacity to shed. That is true. It is also, as stated, unfalsifiable in the most seductive way available — rising dependence reads as still growing, falling reads as decaying, and flat reads as at the inflection. A two-phase model with an unspecified transition is always confirmed.
So the repair is a number, and the number is not written yet. I owe a pre-specification — stated in advance, before there are cohorts and a growth curve somebody is accountable for — of the pass at which dependence must begin to fall. Institutions in their growth phase do not spontaneously enter their decay phase; I have flagged this exact failure twice elsewhere in my own work, in almost the same words, and here it is a third time in the place it matters most. An undated "then decays" is the thing everyone agrees with and nobody triggers. Writing it costs nothing today and will be nearly impossible later, which is the entire argument for doing it now.
One more piece of honesty about all of this: nobody has gone around yet. The loop, the arrows, the decay, the claim that rung one restarts the cycle — none of it has been observed. It is a design commitment and a set of things I have promised to measure and to publish whichever way they come out. I would rather tell you that in the essay than have it discovered in the footnotes.
I will end on the figure who makes the whole runbook make sense, and who has never heard of me.
Somewhere there is a woman who notices every meal, knows the name of everyone on her street and uses it, sits with her brother for an afternoon for no reason, and passes on more than she keeps — with no app, no ledger, no institution, and no vocabulary for any of it. She is not a user. She is the terminus. Everything I have built is scaffolding for people who need help arriving where she already lives, and the honest measure of my work is how many people stop needing the scaffolding, not how many keep using it.
That is the deflation the title was set up for. How to change the world turns out to be: notice, address, spend, pass it on — again, and again, needing less apparatus each time — until one day the apparatus is a thing you used to use.
I am aware that four small things is an underwhelming answer to a grand question. I looked for a better one for about twenty years. This is what I have.
Dedicated to the public domain under CC0 1.0. The four rungs belong to everybody and always did; the products named around them are ours and are not the point. B-Grace℠, B-Sey™/℠, HeartBank® and the marks in the family are reserved.