The Answer That Ends Your Discomfort Is Usually the Wrong One

Picture a production incident at two in the morning. The error rate on the checkout service is climbing. Six people are on the call, half of them woken up by the page, all of them staring at the same dashboard.

Ten minutes in, someone says it has to be the deploy that went out at four that afternoon. You can hear the call relax. A cause has been named, and the work is obvious now. Somebody starts the rollback.

The rollback takes forty minutes. The error rate keeps climbing the whole time. The real cause is a certificate that expired on a vendor’s endpoint at 1:52 a.m., and nobody looked at it, because by 2:10 everyone already knew the answer.

The people on that call were good engineers, and they grabbed the deploy because naming it let six tired people stop feeling lost. The answer that ends your discomfort is usually chosen for that reason, and the evidence never got a vote.

The Apparent Contradiction: Decisive People Answer Fast

We promote people who answer fast. The executive asks “what’s the plan?” in the first hour of a crisis, and the leader with a plan looks strong. “I don’t know yet” sounds like someone who is not ready for the job.

Yet the fast answer is often the slowest route to the right one. The rollback that ate forty minutes is the small version. The bigger versions are the reorg that fixed the wrong problem, the hire made because the role had been open too long, and the architecture picked in the first meeting and defended for two years.

Most people see two options here. You are decisive, or you are the person who keeps asking for more data while the building burns. The poet John Keats saw a third one. In a December 1817 letter to his brothers, he described the quality he thought made Shakespeare great. He called it Negative Capability, “when man is capable of being in uncertainties, Mysteries, doubts, without any irritable reaching after fact & reason.”

“Irritable reaching” is the exact feeling on that incident call, where the reaching was about the irritation more than the evidence. As I use the term here, negative capability is the ability to keep a hard question open until the evidence picks the answer, and to close it on the date you set.

This is a different problem from the uncertainty of working for years with no proof it will pay off: that one asks for endurance over years, and this one asks for judgment in the few minutes between a hard question and the first answer that feels good.

The Deeper Truth: Relief Picks the Answer

The psychologist Arie Kruglanski spent decades studying what he calls the need for closure, the desire to have some firm answer, any answer, rather than stay confused. In a 1996 paper with Donna Webster, he described two moves that follow. People seize on early information, and then they freeze on it. Once frozen, they defend the answer and stop taking in anything new. Time pressure and fatigue make both moves stronger, and two in the morning has plenty of both. So does most of leadership. Nobody chooses for the first answer to show up, but everyone on that call chose to act on it, and fatigue explains that choice without excusing it.

Medicine has studied the same failure under the name premature closure: accepting a diagnosis before it has been checked, and stopping the search. When Mark Graber and his colleagues reviewed diagnostic errors in internal medicine in 2005, premature closure was the most common thinking error they found. Doctors are trained to resist it and still do it, because a diagnosis ends the not knowing for the doctor as well as for the patient.

The first answer feels like progress because it delivers relief. You were buying an end to the discomfort, and an answer came with it.

What you are holding at that moment is what the Greeks called doxa: opinion, a belief you hold without having tied it to its reasons. In Plato’s Meno, Socrates says even a true opinion is like the statues of Daedalus, which were said to be so lifelike that they ran away unless someone tied them down. A true opinion runs off the first time it is challenged, because you do not know why it is true. Tie it down with the reason and it becomes knowledge, which the Greeks called episteme. (The comparison of doxa and episteme goes deeper on the difference.)

So when an answer you grabbed for relief turns out to be right, you got lucky. On the next incident the same reflex will grab the wrong cause just as fast and feel just as sure.

The obvious objection is the senior engineer who glances at a graph and knows. They are often right, and their experience is real. Experience improves the first guess without checking it, and a well-trained guess about a familiar failure is still a guess when the failure is new. The test in the moment is one question: can you say what you would expect to see if your answer were wrong? If you cannot name anything, relief picked the answer, and your next move is to go find out.

AI makes this sharper and more useful at the same time. Answers are free now. Paste the error into a model and ask for the cause, and you get a confident, well-organized paragraph in four seconds. Ask it for a verdict and it gives your irritable reaching a better-written answer to grab. Ask it for the alternatives, and for what would rule each one out, and it becomes the best partner you have for keeping a question open.

The Integration: How to Decide When You Don’t Know Enough Yet

Some questions do not deserve to be held open. If a choice is cheap and easy to undo, decide now and learn from what happens. Picking the meeting time, the name of a branch, or which of two decent candidates gets the second interview should take seconds. Relief is a fine reason to pick when being wrong is cheap and you will find out fast. Holding those open passes for rigor while you put off deciding.

The questions worth holding are the expensive ones and the ones that are hard to reverse: a cause during an outage that is still running, a reorg, a senior hire, a platform you will live on for five years. For those, holding the question open is concrete work, in four steps.

Write the question down in one sentence before anyone proposes an answer. “Why is the checkout error rate climbing?” is a different question from “how do we roll back the deploy?”

List at least three rival answers. With one answer on the table you end up defending it, and with three you can test them against each other.

Name the evidence that would rule out each one. For each candidate, what would you expect to see if it were wrong? Answering that gives the room something to check while it waits.

Then set the time when you will decide with whatever you have. On an outage that might be five minutes. On a reorg it might be two weeks. The date keeps the open question from becoming an excuse.

Side by side, the two kinds of questions look like this:

Use the model for this. Ask Claude for three possible causes, ranked, with the one log line or metric that would rule out each, then have it make the strongest case for the cause you like least and tell you what you would see if your favorite explanation were false. A model does not get tired at two in the morning, and it has no stake in the deploy being the answer. Tell it to keep the question open and it will, faster than any room of people can.

Rerun the incident call with that habit. Ten minutes in, someone on the call, maybe the engineer who joined last month, says: “Three candidates. The four o’clock deploy, the payment vendor, the database. Here is the check for each one. Five minutes.” The deploy check comes back clean. The vendor check shows the expired certificate in under four minutes, about half an hour before the rollback would have finished. Saying that sentence takes no title, and the newest person on the call can say it as well as the incident commander.

The sentence worth practicing is short: “I don’t know yet. Here is what will tell us, and here is when I will decide.” It settles a room better than a confident guess does, because people can see what to check, and a guess loses the room the first time a check comes back against it. If your team keeps skipping straight to solutions, start one step earlier and define the problem before anyone solves it.

The Mastery: Negative Capability as a Decision Skill

The stuck feeling itself has a Greek name, aporia, which means being without a way through, from a- (without) and poros (a passage). In the Meno, Meno complains that talking with Socrates is like touching a torpedo fish, the stingray that numbs whatever it touches. Socrates leaves him confused and unable to answer.

Socrates does not apologize for that. A little later he walks a young enslaved boy through a geometry problem. The boy gives a confident wrong answer, then a second one, and then admits he does not know. Socrates tells Meno the boy is better off now than when he was sure, because now he will want to find out.

That is the deeper reason to refuse the comfortable answer. The rollback minutes and the bad reorgs are real costs, but the boy gains something else: he now wants the truth more than he wants to feel settled. A leader who holds questions open only to save time will drop the habit the first time a guess happens to work.

For a leader, negative capability is a decision skill, and patience is only part of it. The deadline is what makes holding a question open count as judgment. Rilke told a young poet to “love the questions themselves,” which is good advice for a poet and dangerous advice for anyone who owns a production system. A question held open with no date turns into drift, the same way a belief that never gets questioned turns into something you do not actually hold. So there are two ways to fail. One closes the question too early, on relief. The other never closes it, out of fear of being wrong. The skill is holding a question open as long as the stakes call for and no longer, which is why the date matters as much as the rivals.

The leader who has this skill is easy to spot. They can stand in the room with the question still open, say so out loud, and keep people looking. The room calms down anyway, because someone has named what would end the wait.

This part does not get easier. I still feel the pull every time a hard question lands and the first answer shows up a few seconds later with its relief attached. What practice gives you is the habit of noticing the relief and treating it as a warning instead of a verdict.

Four moves for this week:

1. Write the question before the answer. The next time something breaks or a big decision lands, put the question in one sentence at the top of the doc or the channel before anyone proposes a fix.
2. Make the model give you rivals. Ask for three candidate answers and the evidence that would rule out each one. Never ask a model for the answer to an expensive question until you have asked it for the alternatives.
3. Name the evidence that ends it. For every candidate, write what you would expect to see if it were wrong. Go check that first.
4. Put a date on it. Say when you will decide with what you have, and then decide on that date.

Final Thoughts

The urge to answer shows up at two in the morning and in the boardroom, and it will keep showing up for the rest of your career.

The relief that comes with the first answer is out of your hands, and what you do in the half second after it is yours. The answer you just grabbed made you feel better. Before you act on it, check whether the evidence picked it too, and if you cannot tell, hold the question open on purpose, with rivals on the table and a date when you will choose.

If you want to build the judgment to sit in a hard question without grabbing the first way out, MasteryLab is built for that kind of work.

FAQ

What is negative capability?

Negative capability is the ability to stay in uncertainty and doubt without grabbing at the first explanation that makes the discomfort stop. John Keats coined the term in an 1817 letter to his brothers. For a leader it works as a decision skill: hold the question open, list the rival answers, name the evidence that would rule each one out, and decide by a set date.

Why do people jump to the first answer?

Because not knowing is uncomfortable, and the first answer that fits makes the discomfort stop. Psychologist Arie Kruglanski calls this the need for closure: under pressure, people seize on early information and then freeze on it, defending it against new evidence. Time pressure and fatigue make it stronger.

How do you make a decision when you don’t know enough yet?

First decide whether the question deserves to stay open. If the choice is cheap and easy to undo, decide now and learn from the result. If it is expensive or hard to undo, write the question in one sentence, list at least three rival answers, name the evidence that would rule out each one, and set a date when you will decide with whatever you have.

What does aporia mean?

aporia is the Greek word for being stuck with no way through, from a- (without) and poros (a passage). In Plato’s dialogues it is the moment someone realizes they do not know what they thought they knew. Socrates treated it as progress, because a person who knows they are stuck will finally start looking.