Gabriel Betteredge, house-steward of an English country estate, has been asked by his employer’s family to write down everything he remembers about the disappearance of a large diamond. Before he begins he consults his copy of Robinson Crusoe, which he has been consulting for thirty years as a guide to every difficult decision. He then spends some pages on his history with the book, the household, his daughter, his late wife, and his views on the relation between Sundays and shaving. The diamond does not appear for some time. The crime takes considerably longer.
Jack Reacher, a former military policeman travelling without luggage or a fixed address, is eating breakfast in a diner in a small American town when the local police arrive and arrest him. The first sentence puts him in the diner. The second brings in the threat. By the end of the first paragraph the reader knows where they are, who the protagonist is, and what is immediately at stake.
Both openings are the best of their kind, and each is close to the exact opposite of the other. What each one buys, and what each one costs, is a more useful question than which is better.
The eighty-page opening buys a relationship. The narrator’s reliability is constructed slowly, the world acquires texture, and the reader develops an interest in a particular consciousness. By the time the plot arrives, the reader has lived with the narrator for some hours and knows how to weigh what he says. This is structural insurance for everything that follows, paid for in advance.
What it costs is every reader who never reaches page eighty. The reader who decides at page thirty that nothing is happening is a reader the book cannot reach. In a reading economy where the sample chapter and the first ten percent decide most purchases, that is a real risk to take. The novel may be brilliant but the reader who closed it at page thirty will never find out.
Calling the long opening leisurely is wrong, because it is doing work. The Robinson Crusoe habit is not padding and it is not local colour. It is a test the reader runs on the narrator without being told they are running it. A man who consults a novel for guidance in every difficulty is absurd, and the reader registers the absurdity immediately. A man who admits to the habit in the first pages of a document he has been asked to write for a legal purpose is also, less obviously, transparent. By page ten the reader has learned that this narrator will report things that make him look ridiculous. That is precisely what the rest of the book requires of him, and there is no faster way to establish it. The reader watches him do it.
The one-paragraph opening buys the reader. Someone opening a thriller wants the contract settled immediately: this is the world, this is the protagonist, this is the stake. The contract is delivered, the reader is engaged, and the reader is still present in chapter ten when the plot becomes complicated. The opening has paid for the rest of the book.
What it costs is that insurance. A reader given a contract but no relationship is in the book for the plot rather than for the narrator, and if the plot weakens they will leave. If the narrator turns out to be complicated in ways the contract did not promise, they may feel misled. The fast opening makes promises the rest of the book has to keep, and those promises are about action and stake rather than depth.
The fast opening has its own version of the same trick, which is less often noticed. The arrest in the diner is not there only to start the plot. It establishes that the protagonist is the kind of man to whom this happens and who does not appear to mind very much, and it does that in the space it takes to describe a breakfast. Both openings are characterisation. One does it by accumulation and one does it by incident. The choice between them is a choice about how much of the reader’s patience the book can spend before it has earned any.
There is a middle position, and it is where most good contemporary novels sit. The Secret History opens with a confession of murder: the reader is placed inside the action, the voice is established, and specific promises are made about what the book will do. It does not run for eighty pages and it does not reach a crisis in three sentences. It uses the first chapter to set the voice and the central event, and then slows down and does the relationship work that the long opening would otherwise have done. Atonement manages something similar with its first long set piece: the dinner and the play and the visit to the fountain, which does contract work and relationship work at the same time. The reader is given enough plot to want the rest and enough texture to trust the writer. The calibration in both is very precise.
The standard explanation for the long Victorian opening is that it was a function of serial publication, and that no novelist paid by the instalment would have chosen otherwise. That is partly true, and the economics did support the long opening. But many of the great long openings were written by novelists with options. Eliot did not write the prelude to Middlemarch for the serial economics: she wrote it because it was doing structural work she could not do anywhere else. James, in the late novels, did not build his accreting openings because the magazines required them, but because the consciousness he wanted to render needed the time. The long opening is a tool that some writers chose when faster ones were available, because the slow one did something the fast one could not.
The mirror-image objection is that the one-paragraph thriller opening is shallow, and that the form does not produce literature. This is snobbery, and it does not survive contact with the books. A brisk first paragraph does not commit a novel to a brisk four hundred pages. Some of the best crime fiction of the past thirty years opens fast and then runs slow, and the slow running is where the literary work happens. The opening is a door. It tells you very little about the size of the house.
The contemporary novelist has both openings available, and choosing between them should follow from what the rest of the book needs. A novel that depends on the reader’s deep trust in a particular narrator needs more opening than the fast model allows. A novel that depends on the reader being in motion from the first line needs less than the Victorian model provides. Defaulting to either pole because it is what the form is expected to do is the only mistake.
One issue is that the fast opening has become close to mandatory for anything sold as a thriller. Publishers expect it, agents expect it, readers expect it: a manuscript that arrives with a long opening will be asked to compress. The compression is often right. It is also close to automatic, applied whether or not the long opening was doing structural work, and the writer who chose it deliberately is not always given the chance to say why. The reflex is a small flaw in the current state of the craft.
The remedy is mostly the writer’s. Anyone who chooses the long opening has to know precisely what it is buying and be ready to argue for it. That argument will sometimes be lost. It is still worth making, because the alternative is a form in which every book opens the same way regardless of what it needs.
I spend more time on openings and endings than on any other part of a book. The opening I am most pleased with took more than a dozen rewrites. It opens fast enough to make a contract and then slows deliberately in the next pages to build a deeper relationship with the narrator. Whether the calibration is right is not for me to say.
Betteredge and his Robinson Crusoe would not survive a contemporary editor. That is a loss, and we should be clear about what has been lost: not charm, and not leisure, but a particular kind of trust that can only be built at length. The fast opening cannot buy it. The slow opening cannot survive without a reader willing to wait. Most novels today are written for a reader who will not.