So it was this multi-perspective, multi-character book, and it went through all of these different manifestations. I'm not sure there was a single moment where I thought to myself, Oh, I need to write about Margaret Cavendish. She just kept taking over the book I thought I was writing.

It's interesting, editing can be so immersive for me that I've noticed that the authors I edit have a pretty profound effect on how I hear language for a while.

I had all these sparkles I'd collected and wanted to work in, but when I originally started writing it and it was originally this novel about all these people set in 1666, what I was so interested in was the New Science.

I would be researching seventeenth-century garden design or I would be doing something with Pepys, but I just kept using all of it to write about Margaret Cavendish. It took me a long time to realize that I just wanted to write a book about her. Years.

As a publisher what you are trying to build is a long life for a book, to help it find its readers in many different ways, whether or not it made this list or got that review, etc. I'm sure some of that thinking has been useful to me as a writer as well.

The way I've talked about my research process is that it was like magpies. I was just sort of moving through all these books and when something shiny would pop out I'd be like, Ooh, I love it! and I'd pluck it out. It's fun to figure out how to use those bits you really love - like I'd read about gold shoes with cork heels. Obviously, Margaret would have to wear those shoes.

There was a stage inside it and a crank on the outside that would rotate something, like a tiny tree carved of cork, onto the stage, and then the thousands of little mirrors would multiply that one tree so that the viewer would see an infinite forest instead.