Somewhere in the last week you said something like I'm terrible at this, or he'll never change, or that's just how it is here. You didn't weigh it up. It arrived already finished.
A closed idea doesn't feel like an idea. It feels like a description of how things are. Every one of them started as a question that stopped being asked — and closing is so ordinary that it happens to everybody, all day, mostly for good reasons.
A working guide to noticing when one of your own ideas has stopped taking in new information — and to opening it again, if it's worth opening.
The problem it addresses is that closure is invisible from the inside. An idea you've stopped examining doesn't feel unexamined. It feels like a description of how things are, and it goes on quietly deciding what you attempt, what you rule out, and what never occurs to you to ask. Nothing flags it. There's no sensation of having stopped.
So the aim isn't to hold fewer views, or to doubt everything, which is its own kind of trap. It's to be able to tell which of your ideas are still open — and to have something to do about the ones that aren't.
It's built on two sets of language patterns published in 1975, one for recovering what a thought left out and one for deliberately leaving things open. You don't need to know anything about either to use what's here.
They have to. You can't reconsider everything, so conclusions harden into assumptions and assumptions sink out of view.
From inside, a sealed idea is indistinguishable from a fact. Certainty is a feeling, not evidence — and it's the only tell you get.
Precise ones recover what got skipped. Vague ones hold the space open long enough for something new to arrive.
Both models rest on one distinction. Deep structure is everything you know about an experience — images, memory, felt sense, most of it never in words. Surface structure is the few words that form.
In between, you delete most of it, distort what's left, and generalize single instances into rules. That's not a defect; it's what makes thinking fast enough to be useful.
The cost is that the compression is one-way. What surfaces reads as complete. The exceptions were dropped, the source of the rule was dropped, the alternatives were never listed — and none of that absence is visible in the result.
So it stops being a claim you could examine and becomes the floor you stand on. Most of it never reaches words at all, which is part of why it goes unchallenged — putting it into words is the first thing that makes it examinable. Ask what closed, and it usually reopens.
"Nobody ever listens to me."
Two people, in one meeting, last Thursday, talked over a proposal about hiring. Others in the room did listen. It has happened before, with those two. It matters because being talked over is how the last job ended.
The tell is certainty. Not the content of the thought — the feeling that it's obvious. Obviousness is what closure feels like from the inside, which is why it's such a poor guide and such a good signal. When you notice you're sure, you've found something worth opening.
Efficiency explains why thoughts compress. It doesn't explain why reopening one is unpleasant, or why people defend positions they've stopped believing. Something else is going on, and the existentialists described it better than the cognitive account does.
Every settled idea does you a favor: it removes a decision. I'm not the sort of person who does that closes a question you'd otherwise have to keep answering. The certainty isn't the point — the relief from having to choose is.
Which means reopening it isn't free. You get the question back, and the obligation with it. That's the resistance, and it isn't stupidity or laziness. It's a reasonable person declining a cost.
Bad faith — reporting a choice as though it were a fact about yourself. I'm just shy converts something you keep doing into something you irreversibly are, and a thing you are cannot be held against you.
His counter-claim was that there is no fixed essence underneath doing it. Whether or not you accept the metaphysics, the observation stands: "this is just how I am" is a sentence with a job, and the job is closing a question.
This is why the questions on this page can feel unwelcome rather than useful. A good question doesn't hand you information; it hands you back a choice you'd successfully put down. The nagging sense that you'd rather not look at something is not a sign the enquiry is going badly. It's usually a sign you've found a live one.
The usual textbook sorts these into deletions, distortions and generalizations. More useful: sort them by what they close, and what reopens it.
Notice what these have in common. Not one of them is a false statement — most are broadly accurate. They're unfinished thoughts presented as finished ones, and that's a different problem, which is why arguing with them fails and questioning them works.
There is a seventh, and it can't be given an example. By definition you can't quote a thought you've never put into words. It's the widest kind — not a claim about this situation but about how situations go — and it does more work than the six above precisely because it has never been stated clearly enough to be doubted.
The only way to reach one is from above. Take something you're sure about, and ask what would have to be true for it to make sense. Whatever answers is the thing you've been standing on.
You won't catch any of these as you say them. Nobody does, including people who have taught this for thirty years. You catch them ten minutes later, or the next morning, reading something back. That's not you failing at it — that is the skill, and afterwards turns out to be soon enough.
This section is set the other way up on purpose. The two models are inverses of each other, and it seemed worth printing them that way — everything you just read runs in one direction, and everything below runs in the other.
The Milton Model is usually taught as hypnotic language. Its real function is simpler and stranger: language vague enough that you can't arrive at your usual answer, so something else has a chance to turn up. Precision closes a question. Ambiguity holds it open.
— . , . .
Six clauses, each doing something. Hover one to see what.
Notice that knowing the mechanism did not switch it off. That is the honest thing about this half: it isn't a trick that stops working once explained, because it isn't a trick. It's language with room in it, and the room is real whether or not you can name the device.
Meta reopens a thought by demanding what it skipped. Milton keeps it open by refusing to specify. One recovers what closure discarded; the other delays closure long enough for something new to arrive. Used alone, the first can become interrogation and the second can become fog. Used together they are a way of holding a question — which is the actual skill, and much rarer than having answers.
This decides whether any of the above works, on yourself or anyone else. The words barely change; the function reverses completely.
Not because it's unkind, though it is. Because it's rhetorical — it already has the answer and is demanding agreement. It performs the shape of a question while doing the work of a conclusion, so nothing opens. The thought closes harder, now with a bruise attached.
Would you be interested in an answer you didn't expect? That's the whole diagnostic, and it applies equally to what you ask other people. Wanting to find out is the mechanism — everything else is technique. If you know what the answer had better be, you're using the vocabulary of enquiry to restate a verdict.
Open-endedness is not permanent doubt, and a page arguing for questioning owes you the limit. Both failures feel like rigor from inside.
Arriving fast and feeling certain. The familiar failure.
Questioning that won't terminate. Feels like effort, produces nothing.
Openness used to avoid deciding. Feels like wisdom.
Socrates is usually quoted as saying he knew nothing. He didn't say that. What the Apology gives is narrower and far more useful: what he did not know, he did not think he knew. Not a claim of universal ignorance — a claim about calibration. He held plenty of views. What he held separately was the line between the two states.
That line is the whole subject. Ideas don't close by becoming false; they close by migrating quietly from the second column to the first, where they stop being checked. Everything above is a way of noticing the migration.
So: close for action, stay open for evidence. Decide, commit, act on it — and keep the belief revisable rather than the decision. Someone who won't commit isn't open-minded, they're absent. What you're aiming at isn't fewer conclusions; it's conclusions that still have a door in them.
Not on your worst thoughts. On ordinary ones, where the stakes are low enough that you can hear the machinery.
Where this comes from: the questions are the Meta Model, the openings are the Milton Model — two sets of language patterns published in 1975, the first for recovering what a thought left out, the second for deliberately leaving things open. Writing a thought down before examining it comes out of cognitive restructuring; addressing yourself in the second person comes out of self-distancing research. The questioning itself is much older than any of them.