The essay
It is late, and a printmaker is bent over a block of linoleum with a blade no wider than an eyelash. The print is nearly done: a woman on a balcony, laundry on a line, a city behind her. What the printmaker is fixing is a single curve where a sleeve meets a wrist. It is half a millimetre too stiff. Nobody at the exhibition will notice. Most people will spend four seconds on this print, and they will look at the face.
She cuts it anyway. Then she pulls a test proof, looks at it under the lamp, and decides it is better.
Here is the question underneath that small scene. It comes up for anyone who makes things: if nobody sees the work that went into something, how do you know whether that work was any good, and whether it was worth doing? Applause is the usual answer. It is also usually late, often missing, and not very precise.
1. Invisible repetitions
Most skilled work is built from repetitions that never appear in the final result. A composer plays the same four bars at different tempos until one of them stops sounding like an exercise and starts sounding like a decision. A programmer runs the same failing test thirty times, changing one variable each time, until the bug stands still long enough to be caught. Someone learning a language says one awkward sentence aloud on the walk to work every day for a week. On the eighth day it comes out without thought.
None of this shows up in the concert, the release notes or the conversation. That is how it is supposed to work. The repetition gets absorbed into the result, the way scaffolding comes down once the building stands.
But not every repetition is practice. Some of it is just a loop. Polishing the same paragraph for the sixth time can be craft, or it can be a polite way of avoiding the next paragraph. The difference is not effort. The difference is whether each pass is asking a question. Is this tempo too slow? Does the bug appear when the list is empty? Repetition with a question builds skill. Repetition without one mostly builds fatigue.
2. Process versus public outcome
The public sees the last frame. A good logo looks so simple that it seems obvious, and that obviousness is exactly what the forty rejected versions paid for. A well-designed form feels like nothing at all. Nobody admires a door that opens the way they expected.
So it is tempting to swing the other way and say the process is what really counts. That is only half true. A bridge that falls down does not earn points for its thoughtful design meetings. An app that loses your data is not redeemed by its tidy codebase. Process matters because it is how good outcomes become likely, not because it replaces them.
A fairer way to hold both: the outcome is what the work owes the world, and the process is what the work owes the next piece of work. A careful process that produced a mediocre result still leaves you better equipped. A lucky result from a careless process leaves you with a hit you can't repeat.
3. Feedback helps; applause is a blunt instrument
Applause measures a mix of things: the quality of the work, yes, but also timing, mood, who shared it, whether the platform happened to show it, and whether the audience was the right one. A brilliant essay posted on a busy news day can vanish. A mediocre one can travel because of a good headline. Applause tells you that something happened. It rarely tells you what.
Feedback is different because it points. "The third paragraph lost me" is more useful than a hundred likes, even when the reader can't say why. It tells you where to look, not what to do. Good feedback is specific, comes from someone who understands what you were trying to do, and leaves the decision with you.
There are real exceptions. For a stand-up comedian, laughter is not a vanity metric; it is the test the joke has to pass. For a shop, sales are information. When the work's purpose is to move an audience in a particular way, the audience's response is evidence. The trouble starts when applause is used as a verdict on work that was never trying to be popular. A quiet field guide, a careful refactor and a difficult first novel can all be valuable and still not be applauded much.
4. What counts as evidence of quality
If applause is unreliable, you need other evidence, and a surprising amount of it is available without an audience. Three kinds are especially useful.
Clarity of purpose. Can you say in one sentence what the work is for and who it is for? A song for a wedding and a song for a protest are judged by different standards. Without a clear purpose there is nothing to measure against, which is why vague projects feel endless.
Quality of execution. Does it do what it set out to do? Some of this is testable. Does the code pass the case that used to break it? Can a stranger follow the recipe without asking a question? Does the print hold together from across the room and up close? Comparing today's version with last month's is one of the most honest tests there is.
What you learned. Even work that never gets shown can leave behind something portable: a technique, a mistake you now spot early, a better sense of your own taste. That is real value, though it is not the same as the work being good.
Self-assessment has an obvious weakness: we are generous judges of our own work on good days and harsh ones on bad days. A useful check is to ask whether you would be comfortable showing the piece to someone whose standards you respect, and whether you could explain your choices to them. If you could, you have evidence. If you would rather they didn't look closely, that is evidence too.
5. When sharing actually matters
None of this is an argument for working in a cupboard forever. Work meant for people eventually needs people. A designer who never tests with users is guessing. A programmer who never shows code to another programmer will keep certain blind spots for years. Sharing early catches mistakes while they are cheap. Sharing the finished thing closes the loop. It turns private skill into something useful to someone else, and it lets the world correct you.
It also matters who benefits from the unseen work. The idea that "the work is its own reward" is lovely when you say it to yourself and suspicious when someone else says it to you. If a team relies on your careful invisible fixes, those fixes should be visible to the people who decide pay and credit. Valuing hidden work does not mean agreeing to keep it hidden, or to do it for free, when someone is profiting from it.
A rough rule: share when the work's purpose involves other people, when you're stuck in a way a fresh eye could fix, or when it's good enough that keeping it private is mostly fear. Hold it back when it's still changing too fast for feedback to land, or when it was only ever practice.
6. Standards without perfectionism
The printmaker's sleeve raises a fair worry. If every invisible detail deserves care, when does anything get finished?
Perfectionism treats every detail as equally important. Standards don't. A standard starts from the work's purpose and asks which details serve it. The curve of the sleeve matters because the whole print is about a gesture, a woman reaching for a line of laundry. A stiff wrist makes the gesture false. The exact number of bricks in the background does not matter in the same way, and she knows it.
Two habits help. First, decide before you start what "done" will look like, while you're still calm, rather than deciding at midnight when everything looks wrong. Second, notice when your changes stop making the work better and start just making it different. That is usually the moment to stop. Calling something good enough is not lowering the bar. It is saying clearly where the bar was.
Back to the balcony
So, was the printmaker's late-night cut worth it? Yes, though not because effort is sacred or because hidden labour is noble by default. It was worth it because she could say what the print was for, she could see that the change made it better at that, and she will cut the next wrist faster. Her evidence came from the work itself, not from the audience.
Applause will come or it won't, and when it does it will be blurry. You don't have to wait for it to know where you stand. When recognition is absent, there is one principle you can use:
Judge the work by the clarity of its purpose, the quality of its execution and what you learned from it. Then make one decision: publish it, improve it, or move on.
All three are good outcomes. None of them needs permission.