Diagnose Before Rewriting

A diagnosis names where the trouble is, what a reader would experience there, and what is not yet known. It doesn’t name a solution. “The reader does not believe the refusal” is a diagnosis. “Add a flashback to the refusal” is a repair, and it forecloses the others, an earlier choice, a different refusal, a cut. In an argument, “The reader accepts the claim in section two and rejects it in section four” is a diagnosis, while “add three more examples” is already a decision about the cause. Going to the remedy first feels like progress, and the first plausible remedy is often the one that preserves the most existing text.

One effect has many possible causes, and a repair fits only one. A reader’s disbelief at a refusal might come from missing setup, an earlier inconsistency, a pace so fast the stakes were never absorbed, or a character denied enough interior, and each requires a different change. Rewriting without choosing among them is a gamble that may work by accident, and an accidental success teaches the wrong cause.

Many writers know a plainer form of the same defect. A writer senses that a scene is wrong and rewrites it line by line, until the sentences are better and the scene is no more alive; the sentences were often not the problem. Speed is a likely reason, since rewriting can be done immediately and diagnosis has to be thought about. Chapter 7 gives a four-part revision task, made up of the observed problem, its effect, the intended change, and the evidence that would show improvement, and this practice is the step before the third part.

Jane Austen began what she called Chapter 10 of the second volume of Persuasion on July 8, 1816, and dated the end of Chapter 11 July 18. Then she replaced both. The canceled chapters survive in manuscript at the British Library, as far as is known the only surviving holograph pages of a novel she planned and completed for publication. In them Anne and Captain Wentworth come together through Admiral Croft’s blundering intervention, a rumor of Anne’s engagement to her cousin, and a denial. Her nephew James Edward Austen-Leigh, drawing on family recollections, wrote in his A Memoir of Jane Austen (title page dated 1870) that the novel had been brought to an end in July with the reconciliation of the hero and heroine in a scene at Admiral Croft’s lodgings, but that the performance did not satisfy her. She “thought it tame and flat.” She canceled the condemned chapter, he says, and wrote two others, entirely different, in its stead, and Cassandra Austen’s memorandum puts the new writing between July 18 and August 6. The printed novel reaches the reconciliation by another route, Wentworth’s declaration in a letter. If the family account is accurate, a judgment about effect led to a change in the event itself.

Sometimes an experiment reaches a diagnosis that thought can’t. A discovery writer may not know why a scene is flat until a different version of it has been drafted. A rewrite done as a test carries a hypothesis (if the scene starts a page later, does the tension return?) and is judged by whether the hypothesis held. A fix has no hypothesis to check.

Practice. In twenty minutes, write a problem statement on a single page for one troubled passage. It has three sentences: where the trouble occurs, what a reader experiences there in observable terms, and what is not yet known about why. Do not use the words should, add, cut, or tighten. Leave it for a day. Then, in twenty more minutes, write down two different possible causes before proposing any change, and decide what you would look at to tell them apart.