Writing a historical novel is a strange process, for it has to violate that old rule, “write what you know.” The writer of contemporary fiction can attempt to stick close to her own reality, but no matter how much research the historical novelist does, she writes about a foreign, even alien world. That’s one reason I do it and why so many readers enjoy it; it’s a form of time travel.

I wouldn’t want to be thrown back to live in the past, but how wonderful to be able to slip back as an observer. If we could, however, I’m sure that even those of us who do the most meticulous research would discover we have many things wrong.

I’ve been researching Regency England for thirty years, but a recent book gave me new insights into the life of a governess then. Far from always being a downtrodden nobody, she could be firm about her conditions of employment, eat with the family, and even entertain gentlemen. There are notes from that book here. A Governess in the Time of Jane Austen

Therefore, though we must strive for accuracy we also have to learn when enough is enough to write the book. Remember, no one can ever know everything, and besides, research never stops.

Another trap for the historical novelist is a belief that accuracy equals excellence. We are interpreters of the past for readers in the present, and good interpretation cannot always be exact. Consider the French phrase “comme deux gouttes d’eau. In complete accuracy it is “like two drops of water”, but an interpreter would render it, “like two peas in a pod.”

So we should use the language that conveys the truth.

Consider this dialogue. The setting is England, 1816, a gentleman’s study.

“Those figures are wrong, dad. I haven’t spent so much.”

“Then use the computer to check them, but you’re not taking the car again until it’s straightened out.”

If I wrote that most readers would think me an idiot, imagining a PC and an automobile and hearing “dad” as modern. However, they are all correct. Some young men did call their father “dad” (Piers Egan, Life in London); a computer was a clerk who specialized in calculations; and car was a common term for a carriage.

We have to be accurate at a deeper level than mere facts.

The same principle applies to words with the potential to confuse. I work carefully around the levels of a house because in England the first floor is the one above the ground floor, whereas in North America it is the ground floor. Having someone jump out of the first floor window means different things to different readers.

What about obscure period words? They add flavor to a historical novel, but we must be cautious. If I put one in to show off, I kill my darling, as William Faulkner advised. Why? Because a word that means nothing to the reader will pull her out of the book, and I want her there, immersed in another time and place not paging through Webster’s.

This is a great page about historical language. You can’t say that.

Am I recommending inaccuracy? Never. But remember the art of translation. I do the deep research and then distil and refine what I know in order to create an enthralling visit to the past. I don’t dump information on the page to show that I know it. As with an acrobat, the easier I make it look, the better I’ve done my job.

How do you approach writing about the past? How is it different to writing about the present, or about imaginary places?