Back to the table

102 Genre Hf Historical Fiction Genres

Historical Fiction

Invented stories set in a real past.

What it is

A historical novel is a made-up story set in a real past, far enough back that the author had to research the period instead of remembering it. The cast is invented, or made of real people given invented inner lives, and they live inside a period that actually happened. You ask the reader to accept the result as roughly what being there felt like. The story is fiction. The world is fact.

Readers of the genre want two things at once. They want a story that stands on its own, and they want the feeling of having visited somewhere real. Because the period is real, they read with a low hum of doubt the whole way through. Every detail either earns their trust or spends it. So this genre polices things other genres never bother with: what people believed, how they spoke, which events may be bent and which may not, and whether the author owns up to the changes. The shelf has been one of the longest-lived in publishing. If you take one thing from this page, take this: the reader's trust is the foundation the plot stands on, so decide your changes, make them on purpose, and admit them in a note at the back.

Every working definition of the genre draws a line in time, and they all land near the same spot. The Historical Novel Society was founded in Britain in 1997 by Richard Lee, a bookseller and editor, and it publishes his definition. The test is fifty or more years between the events and the writing, or an author born after the events, who can reach them only through research. The society's reviews policy uses the same test: the setting must be fifty-plus years before the book was written. Lee's own page admits the limits are arbitrary and were drawn because some line was needed. From there the society's coverage extends to alternate history, time-slip, historical fantasy, and novels spanning several periods.

Sarah Johnson, the society's reviews editor for many years, explained the reason for the line in a talk at the Associated Writing Programs conference in March 2002, now on the society's site. The line marks the difference between research and personal memory. She also admitted that any fixed number misfires. To a reader born in the 1960s, the Second World War is history. To a reader who remembers the 1940s, maybe not.

The Walter Scott Prize for Historical Fiction, founded by the Duke and Duchess of Buccleuch and first awarded in 2010, uses sixty years rather than fifty. Its rule says most of the storyline must sit at least sixty years back, in honor of the subtitle of Waverley. More on that book in a moment.

For you as a novelist, the number matters less than the reason behind it. You had no window to look out of. Everything in the book had to be found, rebuilt or imagined, and readers know that. So they bring expectations unlike those for any other genre on this table.

Two neighbors need separating right away. A novel that was contemporary when written, by Jane Austen or Thomas Hardy, is a period novel to you and never a historical novel, because its author had the window. A world that never existed is not historical either, however medieval its castles look. The History page explains how invented worlds get their pasts, and ends with a short section on what flips when the past is real. This page picks up where that section ends.

Waverley; or, 'Tis Sixty Years Since came out anonymously in Edinburgh and London on July 7, 1814, and by the usual account it founded the modern historical novel. The founding move is in the subtitle. Chapter one explains it: Walter Scott picked a date that old people still alive could remember, yet far enough back to be truly different, and he said so before the story began. That promise is still the genre's basic contract. The cover, the title and the dateline at the head of chapter one all tell the reader this was real, and I have done the work. Everything else on this page hangs on keeping that promise.

Writers talk about "getting it right" as if it were one job. There are at least three, and each fails in its own way.

Accuracy means the checkable facts are true: dates, battles, who reigned, when the railway arrived, the price of a loaf, the length of a crossing. You can look these up. An error here is a plain mistake.

Authenticity means the book feels true. People think, want, fear and speak the way their time did, and the reader believes they have been there. A book can have every date right and still ring false from end to end, because its people are modern people in costume. A book can take liberties with the record and still feel true; most of the best ones do. James Alexander Thom, a frontier novelist with three decades of documentary fiction behind him, keeps a motto on his site: "once upon a time it was now". The past was somebody's present, lived without knowing how it would turn out. Authenticity means returning the reader to that not-knowing. Thom's book, The Art and Craft of Writing Historical Fiction, is listed below; it covers archive and field research, rebuilding a period's speech and physical surroundings, and welding verified fact to an invented plot.

Anachronism means an object, word, idea or attitude out of its time. The obvious kind is props: a wristwatch on a man in 1850, a zipper in 1890. The damaging kind is words that did not exist yet or meant something else, social freedoms nobody had, and beliefs nobody could hold. Readers find a misplaced mind far worse than a misplaced object.

Keep the three apart and the work gets simpler. Accuracy is a research job for the calendar. Anachronism is a revision pass. Authenticity is the craft itself, and most of this page is about it. Sentence-level texture, the smell of tallow or the weight of wet wool, lives on the Sensory Detail page. Learning a period's rules through research is treated on Worldbuilding, and the location itself on Setting.

"Historical fiction" is one label on the shelf, but it covers several households, each with its own readers and its own added promises. Seven of them come up again and again.

Three neighboring genres have pages of their own. Historical romance follows the rules of Romance: the love plot decides the ending, and the period has to do real plot work. The historical mystery, from Ellis Peters's Cadfael novels (the first, A Morbid Taste for Bones, came out in 1977) to today's large indie market, takes its contract from Mystery and its research bar from this page. Alternate history, such as Robert Harris's Fatherland (1992), with Germany victorious in the war and a Europe living under that outcome, is counted as historical fiction by the Historical Novel Society, but it runs on the logic of Science Fiction: change one thing and follow it through.

Why it matters

Readers extend a kind of trust they give no other genre. They know the period is real and that part of what they are reading is invented. They cannot see the seam. They rely on you to have placed it in the right spot. Most of the praise this genre gets, and most of the complaints, trace back to that one fact.

Readers are quietly checking six things. None of them is written down anywhere, and readers rarely put them into words, but reviews keep coming back to the same six.

Check one: the period does real work. The story should not be movable to another century with a change of costume. If a mother fighting to keep her children would play out just as well in 2025 as in 1739, the period is decoration. The test is whether the era's own rules, its law, money, medicine, religion and distances, generate the obstacles. The Setting page makes the same case for place.

Check two: the big facts hold. The war ends when it ended. The king dies when he died. This readership includes people who read a great deal of history and some who teach it. Wrong dates and impossible journeys get noticed. Johnson said in 2002 that regular readers show no mercy to obvious errors, since each one throws the whole of the research into doubt. One slip and everything else is suspect.

Check three: people think like their time. The hardest of the six, and it gets its own paragraphs further down.

Check four: the author knows invented from recorded. Readers accept huge amounts of invention: whole households, whole love affairs, conversations nobody wrote down. What they resent is invention dressed as record. The worst cases involve a real person's good name or a real atrocity.

Check five: there is a note. Readers expect one, look for it, and sometimes read it before the book.

Check six: the story comes first. The commonest complaint about a well-researched novel is that nothing happens. A research dump is a lecture with a cast. The Exposition page covers how to feed research in.

The usual trust-breakers are easy to list. A modern conscience that is never punished. Costume dialect nobody ever spoke. A real person doing something terrible with no support in the record and no note. A famous event quietly moved. Research delivered as tour-guide commentary. And an ending that rewrites a real outcome without saying so, which is alternate history under the wrong label.

Dialogue is where most books fail, and they fail in two opposite directions. The older error is tushery. In May 1883, writing from Hyères to his friend W. E. Henley, Robert Louis Stevenson called his own The Black Arrow, an adventure of the Wars of the Roses, "a whole tale of tushery". Sidney Colvin, who edited the letters, added a footnote: "tushery" was the pair's private word for romances in the manner of Scott's Ivanhoe, where everyone says tush, by my troth and what thinkest thou. Stevenson was mocking himself, and the joke still works. A newer, informal word for the same thing, "forsoothery", gets passed around among historical novelists; nobody knows who coined it. Tushery fails because nobody ever talked that way. It was the nineteenth century's idea of fifteenth-century speech, and it reads as costume.

The modern error is the reverse. Picture a seventeenth-century parlor where the characters say okay, I need some space and that's not fair on me. Each of those is either a plain anachronism or modern feelings-talk. Readers hear both kinds of error just as loudly.

Most working novelists aim for the same register. Register means the style of speech, and the one they want is clean, a little formal, free of modern idiom, with no jolts in either direction. You get there by subtraction rather than decoration. Strip out modern slang, therapy-speak and recent coinages. Write plain sentences. Then add a handful of real period terms, only where the scene needs the thing named, and let context teach the reader what they mean. A big etymological dictionary gives entry dates; look up any word you doubt. Scott's postscript set the principle two centuries ago: show the Scots through their habits, manners and feelings, never through exaggerated dialect.

Go further back and the job becomes translation. A novel set in Rome cannot be written in Latin, and it cannot be written in the manner of an Edwardian translation from the Latin either. Robert Graves's I, Claudius (1934) presents itself as the emperor's autobiography, built on Suetonius and Tacitus, and Claudius speaks clear, dry, modern English. It works because the voice never wavers and nobody in it holds a twentieth-century idea. The general craft of speech is on Dialogue and Character Voice.

The deepest anachronism is your own beliefs inside a character's head. It usually comes from good intentions. You want your heroine clear of her century's cruelties, so she quietly rejects the hierarchy everyone around her accepts, finds slavery obviously wrong without ever meeting an argument for it, and holds a mild modern skepticism toward her own religion. Readers call this presentism: present-day values applied to the past and poured into it. They spot it instantly, and it removes the very thing they came for, a mind that is not their own. There are honest ways through.

Let dissent cost something. Every period had dissenters, and some held views that look modern now, but they usually paid for them. A questioning character can be the most interesting person in your book, provided the price shows: a lost position, a broken engagement, a trial, loneliness, fear. Costless dissent is just your voice.

Use the era's own words for the feeling. A resentful eighteenth-century wife would not reach for twenty-first-century self-help language. Her words might be religion, duty, property, honor, the body. Finding her words instead of lending her yours is most of the job.

Let a sympathetic person hold a repugnant view. The most unsettling historical fiction lets you like a character without the book agreeing with them. Those scenes are hard to write, powerful to read, and truthful.

Do not swing to caricature either. A cast made only of its era's prejudices is just as flat. People of any era are mixed and inconsistent, and kind inside cruel systems. The Setting page says the same.

The book still keeps a moral view. Events carry it, along with the prices people pay and whom the reader gets to see. What you do not need is a protagonist whose conscience happens to match yours. Theme covers how a book holds a view without stating it.

Nearly every historical novel has real people in it, if only in the distance. The more central they are, the more care you owe them. The first decision is distance. Scott's method, and the genre's for roughly a century after him, was to keep fictional people at the center while real ones pass by. The prince appears, the battle happens, the hero is invented. That leaves you free to invent without putting words in real mouths. The modern biographical novel reverses it: the real person is central, which gains closeness and loses freedom.

The second decision is where invention is allowed. Fair ground is the silences: private talk, thoughts, feelings no record kept. The History page calls this the genre's usual home. Disputed ground is contradicting the record, or adding damaging material about a real person with nothing to support it: a crime, a betrayal, a cruelty, an illness. The more recent the events, the higher the stakes, because there are more living relatives and more survivors. The Second World War boom turned this from theory into live dispute.

The case that made the argument public was The Tattooist of Auschwitz. Heather Morris's 2018 novel was drawn from the real lives of two people, Lale Sokolov and Gita Furman, and its source was Sokolov's own testimony to Morris. Late in 2018, Wanda Witek-Malicka, a researcher at the Auschwitz Memorial, fact-checked it in Memoria, the Memorial's magazine. Her list of errors ran from Gita's prisoner number up to whole episodes that historical sources could not confirm, and her verdict was that the book was unsuitable as a guide for anyone trying to understand the camp's history. Morris answered in an interview with ABC News in Australia in September 2019. She framed the book as fiction resting on a true story; her aim had been a story, never the story of the Holocaust; and readers who wanted the facts had factual accounts to go to. Each side holds up on its own premises. Two lessons stand. A cover line saying "based on a true story" changes the contract. And when the atrocity is real and the victims are real, your readership includes people for whom the facts are memory rather than material.

Three rules of thumb cover most cases. Rule one: invention goes into the gaps, never against the record, and any contradiction you cannot avoid gets declared. Rule two: invented wrongdoing is the slowest invention to allow yourself. Readers forgive an invented love affair freely and an invented crime almost never. Rule three: the claim and the book must agree. A book that calls itself fiction may invent. A book presented as the true story has given up much of that license. The Gregory example below shows the cost.

The note at the back is standard, and readers expect it. Call it a historical note, an author's note or an afterword. The best form does four jobs in one to three pages: which characters are real and which invented; what you changed and why (a battle moved by a month, two real people merged into one, a one-line event grown into a chapter); what is truly unknown, so the reader can tell gap-filling from fact-moving; and two or three sources for the curious.

A note is not an audit. It need not catalog every minor liberty, it is not a bibliography, and it should not argue in advance against reviewers. Bernard Cornwell's notes are the model. Each one hints at which characters are real and which are invented, with no claim to be complete. Give enough that the seam can be trusted, and stop there.

The note goes at the back, for a reason Scott gave in 1814. His own note at the end of Waverley is titled "A Postscript, which should have been a Preface", and he put it last because prefaces go unread. Front matter tells the reader how to read before they have read anything. At the end, the answers land on questions the reading has already raised. Putting it last also protects the reading itself. A reader who learns on page one who is invented reads those people differently.

The trade moves in waves, so read this part as a sketch. This section was checked on October 3, 2026. Publishing shifts, and historical fiction shifts faster than most. Take what follows as a picture of how the trade works, verify the details before you act on them, and consult current lists rather than summaries.

At the time of writing, the dual-timeline novel sits near the top of the genre for commercial visibility. Its strength is in book-club and women's fiction. A present-day woman finds a letter, a house, a photograph or a secret, and her chapters alternate with those of a woman of the past. The appeal is that the modern strand guides the reader and supplies a puzzle while the past strand carries the feeling. The risk is built in: the historical strand's ending is already fixed when the present strand opens, so the present strand has to be discovery, and reading a diary is not enough. Multi-POV describes the trap. Inciting Incident explains why each strand needs its own disruption: one question, asked twice.

The Second World War wave has run since the middle of the 2010s. Since then the war has been one of the biggest commercial settings. Genre round-ups have called it well-trodden ground for some years now, and a new book needs an angle nobody has taken: a place, a profession, a group. Readers have wearied of the same war novel. The war itself still draws them. The same kind of novel has now spread to later settings: the First World War, the 1920s, the 1950s and the 1960s.

Reading groups choose this genre more than almost any other. Publishers answer with reading-group guides in the back of the book or online; Lisa Wingate's site sends readers to her publisher's book-club page. A book with something to argue about, a moral choice, a buried history, a real injustice, can earn years of word of mouth.

Standalones and series split the shelf. At the literary end and the two-timeline end, single books dominate, each sold one at a time on its own premise and its author's name. The series ends, the mystery, Saxon, Roman, naval and military shelves, sell on a hero who comes back every book. The Tudor and royal shelf sits between, often as linked standalones that share a court and move from woman to woman. The craft consequence is plain: a series lead has to outlive the war, while a standalone lead may die. The Series page helps with the choice.

Self-publishing's stronghold is the series end. The typical indie lines are Roman and Viking adventure, naval novels in the wake of O'Brian and Forester, Regency and Victorian mystery, and multi-generation sagas. Revenue comes from ads aimed at series readers and from the biggest online retailer's subscription reading scheme, whose terms change often, so check the current rules yourself. The Historical Novel Society reviews indie novels beside traditionally published ones, with a dedicated indie reviews team that finds titles through retailer sites, reader sites and indie award shortlists. The route is on Self-Publishing.

Discovery works differently at each end. At the literary end, prize shortlists lead, and they carry more weight here than in most genres, with newspaper reviews behind them. Across the genre: round-ups, reading-group picks, library holds, the society's reviews and annual previews, and series reading orders. Adaptations send readers back to the books in large numbers. Covers are so consistent within each household (a woman seen from behind on a cobbled street; a redcoat in smoke; a crown) that you can read the subgenre from across a store. Packaging is a contract; see Reader Promise.

The Historical Novel Society keeps the genre's records. Its quarterly, Historical Novels Review, reviews new titles from the big houses, small presses and indie authors in every issue, with an Editors' Choice mark for the strongest. Its conferences run every two years, across the United Kingdom, North America and Australasia, with craft sessions and agent pitch meetings. It gives awards for new and short fiction. Its lists of forthcoming titles are the fastest way to see what is coming and which periods are crowded.

The Walter Scott Prize is the genre's chief literary honor. Its first winner, in 2010, was Wolf Hall. The shortlists since then make a fair map of the literary historical novel, and the prize's site lists every year's, which makes the best reading list for the literary end. The judges' taste differs from readers' taste, but the shortlists show the range of the form.

For market size, go to the measurers. Skip the blog posts. In the United States, Circana BookScan measures print sales at the checkout; it was NPD BookScan until the 2023 renaming. Publishers Weekly runs BookScan's numbers through the year and adds a yearly round-up. Internationally, Nielsen BookData gives comparable figures. You will find no market numbers on this page, on purpose. A printed figure goes stale within the year, and only the current one is any use.

The story borrows its shape from whatever it is telling. A war or military adventure usually takes Three-Act Structure, with the campaign or the battle as its climax, and the record itself works as a Ticking Clock with a known end. A biographical novel follows the hinge moments of a life rather than one plot, and it is strongest when one stretch of that life carries its own Dramatic Question. A saga is a chain of linked arcs across the generations, one Character Arc each, held together by a Theme. A dual-timeline novel is two plots woven together, and the weaving itself is the skill; see Multi-POV and Subplot. Emma Darwin's Get Started in Writing Historical Fiction, listed below, gives research, picturing the past and period voices a chapter each, then moves on to structure, the shapes open to a historical story, and the subgenres.

One problem belongs to this genre alone: the ending is often known. The ship sinks. The queen is executed. The war is lost. Suspense has to move onto people instead: who survives the sinking, how the queen's sister saves herself, what the family's choices cost. A reader who is ahead of the characters is normal here. Stakes and Tension explain how suspense works under those conditions.

Real-world examples

Waverley (1814): the ordinary man who wavers into a rising

Waverley; or, 'Tis Sixty Years Since was published in three volumes on July 7, 1814, by Archibald Constable in Edinburgh and Longman in London, with no author's name on it. Edward Waverley, a young English officer, arrives in Scotland just before the Jacobite rising of 1745, befriends a Highland chief and the chief's sister, and drifts (his name says it: he wavers) into Charles Edward Stuart's rebel army. He goes through the Battle of Prestonpans and comes out the other side. The book is free online at Project Gutenberg, short enough to sample in an evening, and two short passages in it are of outstanding use to a historical novelist.

Chapter one is an argument about genre. Scott explains his subtitle by listing the ones he did not use. Calling the book "a Tale of other Days" would have promised a Gothic castle with an uninhabited wing and an aged servant. "a Romance from the German" would have meant abbots, secret societies and trap-doors. "a Sentimental Tale" would have brought an auburn-haired heroine with a harp. "a Tale of the Times" would have meant society scandal. His story is dated sixty years back from November 1, 1805, which rules out both chivalric romance and a tale of present-day manners, and his subject, he says, is "more a description of men than manners". Scott was setting the promise for a kind of book that had no shelf yet, through the one piece of packaging he controlled; the Reader Promise page explains why that matters. Sixty years meant a period gone but not yet legend, remembered by the old and invisible to the young.

The final chapter is the earliest historical note. Scott titled it "A Postscript, which should have been a Preface" and put it last because prefaces go unread. In it he names the true incidents: a government officer and a Highland gentleman exchanging protection, which he marks "literally true". His account of the Battle of Preston came from eyewitnesses and was checked against John Home's history of the rising. His characters, he says, are types built from the era's common habits and not likenesses of individuals, and he avoided caricatured dialect, showing the Scots through their habits, manners and feelings. All four parts of today's note are already there: the real, the invented, the sources and the method. Two centuries before either idea had a name, the founding promise and the first historical note were both on the page.

The hero is ordinary by design. Waverley is no great man. He is impressionable, romantic and passive for long stretches, and history happens around him and through him: the prince passes, the battle is fought, the rising fails, and he survives and marries. Georg Lukács, the Hungarian critic, wrote about the historical novel in the 1930s while exiled in Russia, and argued that Scott's innovation was a mediocre protagonist placed at the collision point of opposed forces, so that a historical clash is lived through a man stuck between the camps. Perry Anderson, revisiting Lukács in the London Review of Books in 2011, summed the Scott hero up as a protagonist of no great distinction through whom the clash gains an individual focus. You need no theory to use the device. An ordinary invented person forced to pick a side is the most adaptable tool the genre has. The real events go untouched and the story stays your own.

Sharpe's Eagle (1981): a fictional feat inside a real battle

Bernard Cornwell spent ten years at BBC television, finishing as head of current affairs in Northern Ireland. In Belfast he met Judy, an American, who could not relocate to Britain, so he moved to the United States. He was refused a Green Card, and writing was a living that needed no permit. He had long wanted to write a series about a Napoleonic-era British soldier. He admired C. S. Forester's Hornblower novels and found no army counterpart to them, so he wrote one.

Sharpe's Eagle was published by Collins on February 9, 1981. Richard Sharpe is a rifleman promoted from the ranks, and he has led the series through its whole long run since, though the books were not written in the order the story happens. ITV adapted them for television with Sean Bean as Sharpe. The novel is set in the Talavera campaign in Spain in July 1809, and at its climax Sharpe and his sergeant, Patrick Harper, take a French Imperial Eagle, a regiment's standard, in the middle of the battle.

Historically, no eagle fell to any British soldier at Talavera. Cornwell's historical note says so outright. The first eagle the British really captured, at Barrosa in 1811, he saved for Sharpe's Fury (2006). The same note credits the letters and diaries of the period's soldiers for much of the book's detail.

Cornwell is the model for the military adventure, and for the bargain the whole genre runs on. Three principles show in that one note. First: fictional hero, fictional feat, real battle. The battle happens when and where it happened, its course is followed, and the real commanders act as the record says. Invention stays inside the fictional character's part. Military readers accept this completely, because the frame is true and the feat is plausible: someone could have done it, and here is what it would have taken.

Second: the note marks the boundary. A few pages, real against invented, and you can still see the same thing on his site today. His notes are part of what his readers buy. The question pages on his site are full of readers quoting them back at him. Each note tells you which characters are real. Far from weakening the book, the note strengthens the next one, because readers trust the next battle, knowing he will say where he bent it.

Third: research as grain rather than lecture. The letters and diaries show up in how the soldiers talk, what they carry, what they fear and what they grumble about. The pleasure of seeing how things were done, the same one a Thriller runs on, is applied here to an army of 1809. The loyal, knowledgeable military readership exists because of it, and the series end of the genre rests on it.

The Other Boleyn Girl (2001): the silent sister and the cost of a claim

Philippa Gregory holds a doctorate from the University of Edinburgh on the popular fiction of eighteenth-century circulating libraries. Her novel came out in 2001, narrated by Mary Boleyn, Anne's sister, who was Henry VIII's mistress for a short time before Anne rose to be queen. It won the Romantic Novelists' Association's Romantic Novel of the Year in 2002, became a BBC television adaptation in 2003 and a film in 2008, and grew into a long series about the women of the Plantagenet and Tudor courts. More than any other single book, it made the Tudor novel seen through a woman's eyes a mass-market form.

The book is useful to you in two ways: for the opportunity it found and for the risk it ran.

The opportunity was a woman the record barely touches. Little is known about Mary, and that silence made her the right narrator. England's most famous marriage had her at its heart, yet her thoughts went almost entirely unrecorded. She can be present at every famous scene, and her thoughts can be almost wholly invented. That template is now everywhere: sisters, ladies-in-waiting, overshadowed wives.

The risk lies in what the book does to the people around her. It goes beyond filling silences. Anne is drawn cold and ruthless, and the novel gives weight to the charges of adultery and incest on which Anne and her brother George Boleyn were executed in 1536, charges most historians call fabricated. A deformed fetus and details like it rest on no contemporary source; the historian Claire Ridgway of The Anne Boleyn Files traced them to a later, anti-Anne Catholic writer. The objection went public. David Starkey, in The Daily Telegraph in 2013, called Gregory's work "good Mills and Boon" and said historical novelists should not be taken seriously as historians. Susan Bordo, in The Creation of Anne Boleyn (2013), observed that Gregory's claim to be faithful to history irritated scholars more than the errors did.

The lesson is about the claim, since inventing was never the fault; every historical novelist invents. Liberties presented as liberties are safe ground. A claim of fidelity, plus damaging material about a real person with nothing to support it, draws the same criticism every time. Readers loved the book and kept buying across the series, and some came away believing things about Anne that are untrue, which is the cost the historians pointed at. Decide what your book claims before you write it, and make the note match.

Hamnet (2020): built in the gaps of the record

Maggie O'Farrell's Hamnet was published by Tinder Press in Britain on March 31, 2020, and by Knopf in the United States in July 2020. It won the Women's Prize for Fiction in 2020 and has since been adapted for the stage and for film. Its subject is Shakespeare's son Hamnet, who died in 1596 aged eleven, and the boy's mother. O'Farrell has said her motive was bafflement and sadness at how nearly absent the boy is from the biographies.

The book is the real-woman biographical household at its most deliberate, and three of its choices are worth copying.

O'Farrell builds in the gaps. Only a handful of documents mention Shakespeare's wife, and fewer mention the son. His cause of death is unrecorded; the novel gives him plague, which is plausible and unproven. The mother gets an inner life no document could hold. Nothing in the book contradicts what is known. The History page describes writing into silence; here it runs across a whole book.

She takes names from the record over the familiar ones. The mother is Agnes rather than Anne, because her father's will used Agnes. The choice is defensible without being settled: spelling was loose, and other documents say Anne. Either way it signals a woman seen fresh rather than the biographies' familiar figure.

She refuses the famous name. The playwright is never named in the book. He is the tutor, the husband, the father, the player, defined by his relation to the people at the center. The Folger Shakespeare Library's blog, at the film's release, made the point that this one choice keeps Agnes central; otherwise the most famous writer in English literature would take the book over. The advice carries: if your novel is about someone beside a celebrity, ask how much of the celebrity the reader needs. The answer is usually less than you think. Every page he occupies is a page your protagonist does not.

Notice the ingredients. A real woman, a real loss, a famous man in the background, one ending the reader knows and one they do not. Those are the commercial end's ingredients, used at the literary end.

Before We Were Yours (2017): the two-timeline novel done right

Lisa Wingate's novel was published by Ballantine on June 6, 2017. One strand is Memphis in 1939, where Rill Foss, aged twelve, and her four younger siblings are taken from the family's shantyboat on the Mississippi and end up in an orphanage run by the Tennessee Children's Home Society. The other is Aiken, South Carolina, in the present day, where Avery Stafford, a federal prosecutor from a wealthy political family, has a chance meeting with an elderly stranger and begins to doubt the family story she was raised on.

The history underneath is Georgia Tann of Memphis, whose operation, from the 1920s to 1950, sold children taken from poor families to rich adopters under cover of the Tennessee Children's Home Society. Wingate's page quotes Publishers Weekly: the novel was 2017's third-longest-running bestseller. It won the Southern Book Prize and became a reading-group staple.

Three things make it the commercial two-timeline model.

Each strand does what the other cannot. The 1939 strand carries peril and grief: children in an institution, with no rescuer coming. The present strand carries discovery: a powerful woman with a great deal to lose learns where her family's comfort came from. The reader is ahead of Avery almost from the start, so the suspense in the present is about what she will do once she knows, and never about what happened. The split is correct, and weaker two-timeline books get it wrong; Multi-POV explains how.

The history drives the plot. Tann's trade causes everything in both strands. The period is working at full strength, and the book is no love story with a backdrop. Readers take in a real injustice through characters they have come to love, which is exactly what reading groups want to talk about, and why the book became a staple with them.

The consequences ran off the page. In 2019 Ballantine published Before and After, a nonfiction book by Wingate and the journalist Judy Christie, telling the stories of Tennessee Children's Home Society adoptees. Some of them had discovered their own histories through the novel, and reunions with birth families followed. No recent case documents so thoroughly what happens when a novel's history reaches the people who lived it. The case also carries a reminder about ethics: write about a recent atrocity and the survivors and their descendants are among your readers.

An everyday scenario: the smog novel

This one is made up; only the event is real. Imagine you are ninety thousand words into a draft called Visibility Nil, set during the Great Smog of London, the five days in December 1952. Your heroine is Joan Alderley, twenty-three, a telephonist in an ambulance control room. The buses have stopped and the ambulance crews cannot see the curb. Your research has been thorough. Your beta readers are fond of Joan, and each of them says, separately, that something about the book feels off. One puts it this way: it "reads like a modern novel in a vintage frock". So you run the six checks, in the order given above.

Is the period carrying the plot? Mostly yes. The fog is the obstacle: no buses, no visibility, ambulances unable to find addresses, people dying at home. That part is strong.

Do the big facts hold? Yes. The fog lasts its five days, in December 1952, the buses stay off the road, and no date in the draft is out of place. The research was thorough, and here it shows.

Does Joan think like 1952? No. She privately believes the officials are lying about the death toll and the cause, says so to her supervisor in chapter four, and the supervisor agrees. Her mother's churchgoing draws quiet contempt from her and nothing worse than irritation in return. A married colleague is beaten by her husband; Joan tells her the beating is a crime and she should leave, and she leaves. None of this is impossible for a woman in 1952. But all of it costs Joan nothing, and all of it arrives complete, in the phrasing of 2026.

The same check has a second question. Does anyone know what only we know? Yes, and here is the deeper fault. The scale of the deaths and the coal-smoke cause only came out in the weeks and months after the fog. Joan "knows" both while she is in it. The knowledge is your hindsight, lent to her.

What has the writer done to real people? In chapter twenty a real government minister dismisses the deaths. The private dialogue is invented. The remark sounds callous, and nothing in the record supports it.

Is there a note? No.

Does the story come first? Not yet. Joan knows the answer from the start, so the middle of the book has nowhere to go. Chapter after chapter shows the fog in careful detail while nothing is at stake, and the research begins to read as research.

Now the revisions. Joan's knowledge is cut down to ground level: too many calls about old men who cannot breathe, undertakers overwhelmed, rumors, nobody willing to give a number. Her suspicion becomes a question instead of a certainty, and the suspense shifts to whether anyone will listen. The quarrel with her mother over faith becomes a real conflict, with Joan half a believer herself and afraid of what she would lose if she stopped. Her advice to the colleague becomes what a woman of her class and era might plausibly say, and the colleague's choice becomes hard, with a price attached. The minister keeps only his public words, which are quotable and sourceable, and the private scene goes to an invented civil servant. You draft two pages of back matter under three headings: what is real (the fog, the dates, the public statements, the later findings); what is invented (Joan, her telephone exchange, the ambulance unit); and what nobody could have known yet in 1952. The premise has not changed. The book has gone from a woman of 2026 touring 1952 to a woman of 1952 making a discovery.

Try it yourself

Allow one session for a book you are planning, or about a day for a draft you have written. Keep the results in a document beside the manuscript.

  1. Write the period sentence. One line: This story could only happen in ______ because ______. The second blank names the law, belief, technology, distance, illness or institution of the period that stands behind your main obstacle. If you cannot finish the sentence without the word "atmosphere", the period is decoration. Move the obstacle into the period, or change the period.

Only in ______ because ______.

  1. Draw the line between record and invention. List every real person and real event in the book. Beside each, write one of three words: recorded (you follow the record), silent (you invent where the record is empty) or changed (you depart from the record). Every changed item needs a one-sentence reason. A changed item that blackens a real person beyond what the record shows needs a source, or a rethink.

Real people: ______ (recorded / silent / changed). Real events: ______ (recorded / silent / changed).

  1. Keep a license list while you draft. Every moved date, every pair of people merged, every shortened journey, every invented scene with a real person gets one line: what, where, why. This list is the raw material for your note, and keeping it as you go is far easier than rebuilding it from a finished manuscript.

License list: what ______ where ______ why ______

  1. Run the hindsight sweep. Read your viewpoint characters' thoughts and dialogue for one thing only: knowledge they could not have had. Outcomes, later discoveries, terms not yet coined, present-day ways of describing feelings. Mark each one. Turn certainties into suspicions and conclusions into questions.

Hindsight flags: chapter ______ ______

  1. Do the word pass. Pick ten dialogue words you are unsure of and check each one's entry date and its meaning at the time. Then search the manuscript for modern idiom: okay, process used about feelings, space used about relationships, reach out, deal with it. Cut or replace each one. Add no old-fashioned words to make up for them.
  2. Write one belief you do not share. Choose a sympathetic character and a view that was normal in the period and is wrong to you. Write one scene in which they act on it, with no mockery from the narrative. If you find it hard to write, you have found the spot where the period was being hidden from the reader.

One belief I do not share, held by ______: ______

  1. Draft the note now. One page in three parts: what is real; what is invented or changed, and why; what is unknown. Add two or three sources. Drafting it before you revise exposes the liberties you cannot explain.

Note: real ______ / invented or changed ______ / unknown ______ / read next ______

  1. Read three recent shelf-mates. Pick your household first: saga, war, Tudor or royal, real-woman biographical, dual timeline, naval or military, or literary historical. Then find three books in it published in the last five years that sold well enough to be known. Read their first chapters and their historical notes, and write down each one's page-one promise and its back-of-book admission.

Common mistakes and misunderstandings

Mistaking the research for the book. Months in the archive breed an urge to use everything, and the result is a novel that stops every few pages to explain how a mangle works. Research earns its place only when it changes what a character can do, wants or fears. The rest is the submerged part of the iceberg, holding up the visible part without ever being seen.

The costless modern heroine. The genre's most frequent reader complaint. Dissent existed everywhere; costless dissent existed nowhere. When her modern views never cost her anything, readers hear your voice instead of hers.

Costume dialogue. Prithee, mayhap and 'twas produce a pageant, never the past. Aim for speech that is clean, slightly formal and free of modern idiom. A few real period words will carry the load.

Undisclosed changes. Moving a battle, killing someone early, putting a famous person in a room history never put them in: all allowed. Doing it secretly is the problem. Readers forgive the change and resent the concealment, and one hidden change they discover makes them distrust every other fact in the book.

"Based on a true story" is a claim, not a decoration. The wording on the cover sets the expectations. If the book is mostly invented, say "inspired by" and name the inventions. If it tells a real person's story, the bar for accuracy rises, because the people who were there, or their families, may read it.

Invented guilt for a real person. A crime, a cruelty or a betrayal hung on a real person without support in the record is the gravest liberty the genre allows and the one most likely to rebound on you. Give invented sins to invented sinners.

Expecting the note to excuse everything. A note explains liberties. It cannot excuse carelessness. "I changed the date of the battle for dramatic reasons" is a fair line. "I am aware that people did not talk like this" only admits a fault you should have fixed.

Treating fifty years as law. The line moves forward every year and now reaches the 1970s. Today's version of Johnson's reader who remembers the 1940s is someone who remembers the 1970s. The Historical Novel Society itself calls its line arbitrary. Let research and reader distance decide which shelf you are on, and check how comparable books are actually sold.

Confusing the period novel with the historical novel. A Victorian novel was contemporary when it was written. Use it as a model by all means, but remember that its author could assume what you must discover, and its first readers knew things today's readers do not.

Letting the known ending kill the suspense. A book whose only question is whether the ship sinks has no question left once the reader knows the answer. Shift the doubt onto people: who gets a seat in the lifeboat, what they did to get it, what it cost. Every good novel of a famous event does this.

Treating the genre as lesser. The first version of this error says the research is the hard part and the story can be borrowed from history. The second says the book is a costume romance and the research can be skipped. The finest books avoid both. They honor the reader's trust, spend their invention on people, and draw the past truthfully.

Related elements