How to write historical fiction

You want to set an invented story inside a real past, and the genre's definition is more precise than its reputation suggests. Wikipedia's article defines historical fiction as a genre in which "a fictitious plot takes place in the setting of particular real historical events", set in the past and attending to "the manners, social conditions and other details of the depicted period". The plot is invented; the world is not. That split is the whole job: a story that could happen anywhere does not become historical by wearing old clothes.

The definition and the threshold

Where the past begins is contested, and the article gives the three answers in circulation. The Historical Novel Society counts works "written at least fifty years after the events described". Sarah Johnson's test is set before the middle of the last century and written "from research rather than personal experience". Lynda Adamson's is a period "at least 25 years before it was written". Pick a threshold honestly for your own book; readers will apply one anyway. The genre's modern founder is Sir Walter Scott, whose Waverley (1814) set the pattern, and the critic Gyorgy Lukacs drew the bar the best historical fiction still clears: Scott "was the first fiction writer who saw history not just as a convenient frame" but as a distinct social and cultural setting. Period as frame or period as place: that choice is the genre's first decision.

Research without drowning

The article's two research exemplars worked at different scales, and both are instructive. Georgette Heyer "collected reference works and kept detailed notes on all aspects of Regency life", building a private reference library for one narrow period. Tolstoy used primary and secondary sources and interviewed survivors of 1812, until War and Peace became partly "ethnography fictionalized". Craft consensus on how much is enough is simple to state and hard to keep: you cannot research everything, so budget it. Research deeply where the plot touches, research what the viewpoint character would actually notice (food, money, street noise, who may be spoken to and how), and let the rest stay unmentioned rather than guessed. Guesswork is where anachronism breeds.

Then record what you establish. A fact fixed in chapter three, a coin's value, a journey's length, who is at war with whom, is canon from that sentence on, and the glossary entry on canon explains what that commits you to. This is a story bible workload like any invented world's, with one difference: your bible has two sections, one for what really happened and one for what you invented, and the second must never contradict the first without saying so.

Anachronism and accuracy discipline

The article is direct about the audience: works are "sometimes criticized for lacking authenticity", and "this tension between historical authenticity and fiction frequently becomes a point of comment for readers and popular critics". Alexandre Dumas's Queen Margot had its historical accuracy questioned, and readers are still debating it. The discipline that keeps you out of those reviews is borrowed from invented-world craft: treat the real setting as a rules system. Write down what people in your period can do, know, say, eat, own and expect: distances measured in travel time, the speed of a letter, what medicine believes, what the law allows. Then, when the plot needs an exception, you will feel the rule bend, and you can bend it on purpose. Deliberate license, visible in the choices, reads as voice. Accidental license reads as error, and historical readers keep receipts.

Real people, invented people, and the subgenre map

The article shows the dial's full range. La Princesse de Cleves (1678) sets one invented heroine among a nearly all-historical cast; Colleen McCullough's Masters of Rome invents nobody; postmodern novels mix history with invented events and open fantasy. All are legitimate; choose your ratio before drafting, because it decides what your ending is allowed to do to real lives. The shelves are broad: historical mystery (Ellis Peters's Cadfael Chronicles popularized it), historical romance and Regency (Heyer again), nautical fiction, fictional biography, and alternate history, defined as fiction in which "one or more historical events happen differently than they did in reality". Nor is the tradition Western property: three of China's Four Classic Novels are historical fiction set in the distant past, The Tale of Genji argued fiction could give a "fuller and therefore 'truer'" history a millennium ago, and Jurji Zaydan's twenty-three historical novels shaped Arab readers' sense of their own past.

Where historical meets web fiction

Web fiction's most historically rooted shelves are Chinese. Wuxia is defined as fiction about "the adventures of martial artists in ancient China", and modern wuxia "is largely set in ancient or pre-modern China"; its founding text is Water Margin, one of the historical classics just named. The catalog's wuxia versus xianxia guide draws the border between the mortal tradition and its transcendent descendant, and the cultivation fiction guide covers the ladder-climbing form serial readers know. Both transplant historical imagination into serial form, and they set the bar your own historical serial is read against: keep the period's manners on the page the way Heyer kept her notes, and the past will hold.