I’m currently writing a time travel adventure story that has an American character from 2015 visiting England in 1943.

To make the voice of the characters distinctive and authentic, I not only have to capture the differences between American and English usage, but between modern American usage and 1940’s British usage.

I have found Google NGram viewer an invaluable tool in tackling this challenge. Take this graph for example, which shows that my 2015 American character will probably say “pants”, but my 1940s British character will more likely say “trousers” (but a 1940s American might use either).