People imagine that writing a novel is exclusively an exercise on the inventive and imaginative faculties and sense of construction. Few realize that it is difficult to produce a good novel without a certain amount of research.
In my novel The Girls of Slender Means I merrily portrayed two lovers on the roof of a London girls’ club, in June, under constellations, among which I mentioned Orion’s belt. More than 20 years later, a reader who had come newly to the book wrote to tell me that Orion could not be seen in London in June. This reader, a lady from the Isle of Sark, was a well-versed astronomer. I changed the text of the novel for all the subsequent editions, and have become close friends with my astronomer reader. Normally I ascertain all facts of this nature. If I were to say, for instance, that on 5th May, 1911, X took a taxi in Golder’s Green because it had started to rain, I would first find out if it really was raining there that day. There is always the chance that someone from Golder’s Green whose fifth birthday it was that day will read my words and refute them. “I remember it well. I had a party. It was a sunny day, no rain.”