How Accurate Does Historical Fiction Need to Be?
Quick answer: How accurate does historical fiction need to be? A practical standard: keep the documented record, invent in the gaps, and signal your changes honestly.
Historical fiction needs to be accurate about everything the record establishes, and free everywhere it is silent. That is the working standard most serious practitioners follow: do not move battles, resurrect the dead early or hand characters knowledge their era lacked, but invent conversations, motives and minor persons wherever history left a gap. Deliberate departures are permitted; unmarked ones are the sin. Change what you must for the story, and tell the reader in an author's note.
The question matters because both extremes fail. Fiction chained to the footnote suffocates, and fiction indifferent to the record betrays readers who, research consistently shows, absorb history from novels and screen drama more than from historians. Your invented scene may be someone's permanent mental furniture. That is power worth handling honestly.
The longer answer: three zones of obligation
Think of accuracy in three zones. Zone one, the documented: dates, outcomes, deaths, who was where. Alter these knowingly and you have left the genre for alternate history, so either stay faithful or sign the change in a note. Zone two, the texture: food, prices, medicine, law, manners, distances, what a journey or an infection actually meant. Here accuracy is craft; errors are avoidable and readers catch them. Zone three, the unrecorded: private words, feelings, motives, the servant's name history never wrote down. This is your workshop, and everything in it is legitimately yours.
Mantel's Cromwell novels model the discipline: she kept documented chronology intact and built her fiction in the silences, and said plainly that when the record spoke, she obeyed it. The inverse approach also exists honourably, but it must be loud about itself: a book that promotes its guesswork to fact, silently, is the one historians and readers rightly resent.
Cases that show where the line sits
Philippa Gregory's 'The Other Boleyn Girl' (2001) drew years of criticism not for inventing, all novelists invent, but for building drama on contested claims that many readers took home as settled history. Compare 'Wolf Hall', which stirred argument about its hostile Thomas More yet largely escaped the accuracy charge, because its interpretations sat on documented events rather than replacing them. Interpretation is the novelist's right; fabrication presented as record is the foul.
Robert Graves' 'I, Claudius' (1934) shows the oldest defence: he followed ancient sources faithfully, and those sources were themselves gossips, so the novel is accurate to Suetonius rather than to truth, an honest and fruitful position. And war fiction offers the practical norm: novelists routinely compress timelines and merge minor figures for shape, then say so. The author's note has become the genre's handshake, and readers honour writers who use it.
A working method for your own book
Before drafting, build a spine of the documented facts your story touches, with sources, and treat it as load-bearing. While drafting, when you hit a gap, invent freely but period-plausibly: the test is not 'did this happen' but 'could this have happened without leaving a trace'. When story and record collide, first try redesigning the scene, because the constraint usually improves it; the letter that cannot arrive in time is better drama than the anachronistic telegram.
If you still must bend zone one, bend it minimally, knowingly, and confess in the note: readers forgive a moved date they are told about and punish one they discover. Keep a changes file as you go, every compression, merger and guess, so the note writes itself and your own memory of what you invented stays clean. Novelists who skip this step have been known to lose track of their own fabrications.
Common mistakes on both sides
The perfectionist's mistake is research paralysis and its page-level symptom, the info-dump: chapters that prove homework instead of telling story. Accuracy is a floor you stand on, not a performance. The careless mistake is the unexamined default, assuming the past was like the present but with candles: attitudes, hygiene, risk, faith and childhood were all differently shaped, and those differences are precisely where the stories are.
The subtler trap is false balance about voices the record erased. Archives skew toward the powerful; writing enslaved people, women or the poor 'accurately' by documentary silence would erase them again. The genre's honest answer is informed imagination, grounded in what evidence exists, and several of its greatest books live exactly there.
What poets can take from this
- Keep a running changes file; your author's note is only as honest as your records.
- Let every historical constraint you hit become a plot obstacle before you consider bending it.
- Check objects, words and attitudes against your period; readers forgive interpretation, never potatoes in 1200.
- Read one historian who disagrees with your framing before you finish revising.
- When the record is silent about someone, take it as an invitation, not a prohibition.
Try it yourself
Find a one-line entry in any historical record: a name in a ship's manifest, a fine in a court roll, a burial notice. Write the fully imagined day behind that line without contradicting a single documented fact around it.
Questions
Do historical novelists need an author's note?
It has become the genre's standard courtesy. A short note separating record from invention protects readers from absorbing fiction as fact and protects the writer from accusations of deception. Most acclaimed historical novelists now include one.
Is it acceptable to change real events for the story?
Small, signalled changes, compressed timelines, merged minor figures, are accepted practice. Changing outcomes, silently, is not; at that point the book is alternate history and should own the label. The rule of thumb: bend minimally, and confess.
How much research is enough before writing?
Enough to inhabit the era's daily logic: money, food, travel, belief, law. Many novelists research wide first, draft, then research narrow to answer the specific questions the draft raises. Waiting until you know everything means never starting.
By the Writory editorial team, reviewed by working poets. Updated July 2026.
Enter a Writory poetry contest
Enter the August contest and have your own poem read and scored.