Airworthy, not perfect: finishing your work

Look at the last piece of written work you handed in and split the hours at the point where it was essentially finished. Getting there took a certain number of hours. Getting from there to what you actually submitted often took as many again. Almost none of that second half is visible in the result, and nobody marking it will ever know which sentences you wrote eleven times. That distribution is the problem, and it is not a problem of discipline.

Where the second half of the hours goes

A draft exists after a weekend. Then the reading you add to be safe, three sources that mostly confirm what you already wrote. The introduction, rewritten because the first sentence sets the tone. Formatting. A read-through for errors, then another, at which point you are no longer finding errors but confirming their absence, which has no end.

None of it is fake work. Each change is a genuine improvement, which is why the process is hard to interrupt from inside. What you cannot see from inside is the exchange rate. The improvement is real and worth two percent, the evening is real and worth three hours, and a subject you have not opened is quietly building up. The price of the last tenth is almost never the time itself. It is whatever the time was taken from.

Thoroughness has a direction

Thoroughness points at the work. It asks whether the argument holds, whether the source says what you claim it says. Those questions have answers, and once they are answered the checking stops by itself. Perfectionism points at you. Its question is whether this is good enough that showing it will not be embarrassing, and no amount of work on the text can answer that, because the text was never what it was about.

A usable test is to name what you are looking for before the pass begins. If you can name it, that the third section repeats the second, you are being thorough and you will stop when the list runs out. If the answer is a feeling, something is off, it is not there yet, the pass is not measuring the work and will not settle anything. Another marker is where the hours land, since disproportionate effort on parts nobody assesses separately is rarely about quality. It is about the part of the job that feels controllable.

Released against a standard, not against an ideal

An aircraft is not released for flight because it is flawless. It is released because it is airworthy, and airworthiness is a written standard rather than an impression. There is even a document listing what may be broken without stopping the flight: certain lights, one of two galley ovens, each with conditions attached and a deadline for repair. Nobody at the gate is asked whether the aircraft is perfect. Somebody checks it against a list, signs, and it goes.

A standard has an upper edge and somebody else wrote it. For your work that standard usually exists in writing too, in the assignment brief, the marking criteria, the question you were actually asked. Most people read it once at the start, in a hurry, then spend six weeks measuring against a private standard both higher and vaguer than the real one. Reading the criteria properly at the halfway mark replaces a feeling with a list.

Deciding in advance what finished means

Write down what will be true when the work is done, before you start, in two or three sentences. Six sections, each making one claim and supported by one source. The calculation checked once, not four times. Written in advance, while you have no attachment to the text, this is easy to do honestly. Written at the end, it turns into a negotiation with yourself, and you lose it.

Finishing then becomes a check rather than a judgement, and the last stage gets a time budget instead of a quality target. Two hours for the final pass, on a named evening, and then it is submitted. A budget can be enforced. Until it is good cannot, because that definition moves as you approach it. It also helps to write down what is allowed to stay imperfect: a sentence you do not love, a section that is correct without being elegant.

Reading, which has no natural end

Writing at least has an object that can be finished. Reading does not. Every paper cites six more, and no point exists at which the material announces that you now know enough. The standard has to be imported, because the subject will not supply one.

Two ways to import it. Set the quantity in advance: the three texts the lecturer named plus anything they cite twice, or two afternoons. Or set an end condition and stop once a new text is mostly telling you things you already know. Reading is also the most comfortable way of not finishing that academic work offers, since it is quiet, defensible, and it postpones the moment when you have to produce something that can be judged.

Handing something in with the seams showing

One experiment is worth running, on a piece of work that matters but is not decisive. Take it to your written standard and not one step past it. Before you submit, note what you expect to happen: the mark you predict, the criticism you expect.

Then compare that with what comes back. What you are buying is information about your own exchange rate, and nobody else can give it to you, because only you know what the earlier versions cost. Sometimes the mark is lower than it would have been. Knowing by how much is the point. A smaller version of the same thing is showing a draft to another person while it is still visibly rough.

When the last ten percent is the grade

Some work genuinely lives in its final tenth, and treating it like an ordinary assignment is the mistake in the opposite direction. A thesis where the last chapter carries whatever contribution you are making. A proof or a piece of code where one small error invalidates everything after it.

Three questions separate those cases from the rest. Will the work be looked at again after it has been marked. Is the distance between correct and excellent explicitly rewarded in the criteria. Does a small defect destroy the whole thing rather than costing a proportional amount. Where one of these holds, the extra passes are not perfectionism, they are the work itself, and stopping early is the error. The awkward part is that a perfectionist can find one of the three in a two-page seminar summary, which is why the category belongs at the start, on the day the work is still abstract, since that is the last moment your judgement about it is worth much.