The story came to me in the form of the first paragraph, and then the realization that of course it would be one of those rare short stories from multiple points of view. So then it was just a matter of following the threads of that idea, combined with the thought that something had gone terribly, terribly wrong. And that sometimes after something has gone terribly wrong, it’s not so easy as picking up the pieces and starting over . . . not if something irrevocable has occurred.

The first part of the story came to me in early 1999, when I was working at Tor, in the Flatiron Building. From the window of my little work area, I could look across the street and see this massive apartment building, all the windows and lights flicking on and off all day as people went about their lives. There was one window with curtains that moved back and forth behind the glass, which always struck me as odd, since the rest of the windows had flat shades and blinds that never moved.

I began with the idea of a horror story set during a mountaineering expedition and all the detail, Kanchenjunga, its five peaks, Crowley, all of that came out of the research. I just kept discovering these weird facts and coincidences and everything slotted into place. It felt like I’d stumbled onto a story that wanted to be told, which has never happened before or since.

I found myself contemplating what it might mean to combine the picaresque with necromancy. I find necromancy an entertaining profession, I love eighteenth century-style narratives, and I adore shady heroes, so it seemed a natural combination.

As a writer I don’t always like to answer all the questions a story presents, or every question readers might have. Of course, I want the logic of the story to operate . . . but I want to leave room for a few mysteries, too. To me, that’s a key element to much of the horror fiction I enjoy—the questions that keep readers coming back to a piece, especially the ones that might make them see it a little differently the next time they read it.