Finish According to Purpose

If a writer’s only completion rules are “when I can’t stand it anymore” and “when it’s due,” the writer will eventually stop, and will have learned nothing about whether the Work was ready, because neither rule measures the Work. One of them is fatigue and the other is the calendar.

Paul Valéry, in a 1933 essay on his poem “Le Cimetière marin,” wrote that to lovers of disquiet and perfection a work is never finished, a word that means nothing to them, but abandoned. The abandonment, he went on, whether it hands the work to the flames or to the public, and whether it comes from weariness or from an obligation to deliver, feels to such people like an accident. The line is usually quoted without its first half, and quoted alone it becomes a consolation in which nobody finishes and everybody quits. Valéry was describing one temperament. On our reading, his weariness and his obligation to deliver are the same two things that end a work when nothing else governs the ending.

A third condition is available. The Work is finished when further changes no longer serve what it is for. Flawlessness can’t be the test, because no manuscript gets there, and a Work can always be made different, as Chapter 7 has already said. A purpose, once stated, generates tests. “Could it be better?” always gets a yes. “Does the reader now understand why the sister left?” and “Can every figure in chapter three be traced to a source I could produce?” can actually be answered, and once they are, whatever changes remain are preferences.

Hemingway’s many rewrites of the last page of A Farewell to Arms (Practice 83 gives his own account) are often told as a story about perfectionism. As a case of finishing it is a narrow one. The ending’s function was not in doubt, and the open question was one page and one criterion, the right words. He put the effort where the standard was unmet and stopped when he judged it met. Whether he judged correctly is for readers to decide.

When a film or television script reaches production it is issued as a production draft, and later changes travel on colored pages, in a common order of blue first, then pink, yellow, green and goldenrod and on through further colors, so that everyone on the set can see which version is current. The writer’s sense of readiness doesn’t usually set that date; the schedule, the budget and the people waiting for pages do. Such a script isn’t frozen. It is in a named state with a procedure for change, which is a fair description of what finished can mean for any text, and sometimes the calendar really is part of the purpose. An analysis of an election that arrives after the vote has failed at the one thing it was for.

Practice. Write a finish sheet in about half an hour. Purposes move, and if the Work changed in the drafting, criteria written at the start will finish a book that no longer exists, so restate the purpose first (Chapter 2). Discovery writers often can’t state criteria until a draft shows what the Work is, and for them the sheet belongs after the first retrospective look at structure, not before. It has three headings. Must be true lists conditions tied to the purpose and the reader’s promise, each one checkable. May remain unresolved lists the ambiguities and imperfections you have decided to accept, with a reason beside each. Belongs elsewhere collects ideas for a later edition or another Work. Date the sheet. When revision starts to feel endless, consult it rather than your mood. An editor or collaborator can check the Work against it too.