“What about the sister?”

(Note: This post is the thirty-eighth installment in my author’s commentary for The Icon Thief, covering Chapter 37. You can read the earlier installments here. Be aware that major spoilers follow.)

I still remember where I was when I figured out the killer’s identity in The Icon Thief. Early in the writing process, I was casting about for a striking opening scene, and finally came up with the idea of a woman’s headless body found mummified in the sand under the boardwalk at Brighton Beach. The image was inspired by the setting, and by a few memories of my own—I’ll never forget seeing a dead dog preserved in the sand at a beach in Goa—but at first, the image was all I had. I didn’t know who this woman was, or who had killed her, or why. I just knew that it would give me a convenient hook for the procedural story I wanted to tell, which would bind together the novel’s various strands and parcel out information to the reader in as painless a way as possible. We know the beats of a murder investigation from countless other books and movies, and I hoped that grounding a third of the narrative in that familiar pattern would grant me greater freedom in other parts of the novel, where the shape, by design, isn’t immediately obvious. And that’s more or less how it worked out.

Of course, now that I had a victim, I also needed a murderer, as well as a way to tie the case back to the larger art world story I was trying to tell. The idea of the girl’s death serving as a form of blackmail, with my villain using it to persuade someone on the inside to serve as an accomplice to the heist that occupies much of Part I, came to me right away. At that point, the motive behind the murder wasn’t particularly important: it was just a convenient plot device. All that really mattered was that the killer be someone who could grant my thieves access to the mansion, so I created the part of Zakaria Kostova, the assistant to my wealthy oligarch, for this express purpose. Kostova only appears in a handful of scenes—among other things, he’s the buyer who outbids Maddy at the art auction in the opening chapter—so I didn’t put much thought into his character when plotting out the initial section of the book. In the end, I wrote the first half thinking that I could figure out why Kostova had killed the girl later. A jealous argument was all it would take.

But then something strange happened. While I was developing the character of Archvadze, my oligarch, it occurred to me that he needed a girlfriend. (I was probably inspired by the partners of certain other oligarchs who have made a splash in the art world.) Accordingly, I worked up the character of Natalia Onegina, whom Maddy encounters at the party in the Hamptons. Here, again, she was little more then a plot device: the party is supposed to promote the opening of Natalia’s upcoming gallery, which gave Maddy a reason to be there. I liked Natalia a lot, and thought she was a memorable presence, but didn’t expect her to play much of a role in the story after that. Which brings me to the epiphany I mentioned above. I was standing at my bathroom sink, splashing water on my face—I think I might have just finished shaving—when it suddenly hit me: What if Natalia and Karina Baranova, the dead girl, had been sisters? The idea appeared out of the blue, fully formed, as if it were a fact of nature just waiting to be discovered. And that’s when I knew that Natalia had killed Karina Baranova.

When I look back at Chapter 37, in which Powell makes this connection, it’s hard for me to imagine that the story could have gone any other way. Yet it’s worth emphasizing that I’d already written half the book with a different killer in mind, and in the end, I wound up changing surprisingly little. (A similar thing happened with a crucial plot point in City of Exiles, but for the details, you’ll need to wait for the commentary on that novel, if I ever get around to it.) And it’s a reminder that fiction can always surprise you, however carefully you may have planned it in advance. As regular readers know, I’m an obsessive outliner, and at first, it might seem as if that would reduce the degree of surprise a book can hold for its author. What I’ve found, instead, is that it actually increases these moments of serendipity. When I’m making an outline, I’ve got the plan for the book—or at least one large section of it—in my head at all times, and it serves as a magnet for gathering up ideas and moments of inspiration that might otherwise be lost. That idea was in the air all along, but it only fell into place once there was a matrix in which it could grow. And the rest of the novel would never be the same…