Let Curiosity Precede Certainty

Joan Didion says she begins each novel with no notion of character, plot, or even incident. What she had for A Book of Common Prayer was pictures, and the one that organized the others was the airport at Panama at six in the morning. In “Why I Write” (1976) she recounts imagining a woman in that airport ordering tea and insisting that the water be boiled for twenty minutes. She did not know where the woman had been or why she kept going to airports. A narrator arrived with a sentence that began “I knew why Charlotte went to the airport,” and Didion had to ask who this “I” was. Had she known the answers, she says, there would have been no reason to write the novel.

Opinions are allowed if the work is allowed to test them. Certainty becomes a problem when it arrives before anything was looked at; it has to be earned by a scene tested, a source checked, an objection faced.

Orwell did not write from doubt. In “Why I Write” (1946) he says his starting point is always partisanship and a sense of injustice, and that wherever he lacked a political purpose he wrote lifeless books. He tells how Homage to Catalonia contains a long chapter defending men falsely accused, and how a critic he respected said it had turned a good book into journalism. Orwell agreed and said he could not have done otherwise: he knew that innocent men were being accused, and without anger he would not have written the book. He had been in Spain and had seen, and his certainty came after the seeing.

Montaigne made the other stance into a form. His word essai means a trial or attempt, and around 1576 he had a medal struck showing a pair of balanced scales and the Greek skeptic’s word epekhô, “I hold back”. In “Of Cannibals” he sets aside the cosmographers and relies instead on a plain man who had lived ten or twelve years in the New World, reasoning that such a witness has no reason to embellish. His finding is that there is nothing barbarous in the people described, except that everyone gives the name barbarism to whatever is not the custom of his own country. He grants the barbarous horror of the revenge rite he reports and adds that we see their faults so clearly and are so blind to our own.

He also admits that in one conversation his interpreter was so poor that he got nothing of moment. That limit is a warning, because admiration can arrive as early as any other conclusion, and his rests largely on a man in his household, some seamen and merchants, and a failed conversation.

Begin by knowing instead, and the essay proves a conclusion reached before the first source was opened, or the novel illustrates a premise settled in week one. That can work if the writer has already done the checking. It goes thin when nothing is allowed to surprise the writer and disconfirming evidence is filtered out before it can complicate anything. In fiction, a scene written to find out what a character will do often carries information the planned version lacks, and planners can do the same inside their plans by asking, of each outlined event, what they don’t yet know about it (Practice 22). Curiosity has to end in commitments eventually, too; a deadline is one way.

Practice. Write three questions your current argument or draft cannot yet answer, taking twenty minutes to write them. Each must be something you do not know, not a rhetorical question in disguise, and each must name an act that could answer it: a document to find, a person to ask, a scene to draft in which a character is free to choose, a counterexample to look up. Then do one of the acts this week and note whether the answer moved your central claim or scene, even slightly. If every answer confirms what you already thought, suspect the questions.