Just a reminder, all. Steelheart—my new novel—is out right now! It hit #1 on the New York Times bestseller list in the Young Adult section. If you’re curious, you can read about the book here, and listen to a cool audio sample here.
For an explanation of my Wheel of Time retrospective, see the beginning of my first post, which talked about the notes, and my second post on the process. Here’s post number three. Before we begin, it should be stated that this post will contain spoilers for the entire series, ending included. If you haven’t finished, you will want to do so before reading this post.
The Gathering Storm: Writing Process
I attacked the project in earnest in the summer and fall of 2008. I realized early on that there was too much to keep in mind for me to write in a strict chronological fashion, as I had normally done in the past. For this project, I needed to take groups of characters, dump all of the information about them into my mind (like loading a program into RAM), and write for weeks on just that group. This way, I could keep track of the voices of the many characters and maintain the numerous subplots.
The hardest part of this project, I feel, was keeping track of the subplots and the voices of the side characters. This is not surprising; though I’d read the Wheel of Time many times, I was not a superfan. I loved the books, but I was not among the people who made websites, wikis, and the like for the books. I read the books to study the writing and enjoy the story; I did not spend too much time keeping track of which minor Aes Sedai was which.
I could no longer be lax in this area; I had to know every one of them. Part of Robert Jordan’s genius was in the individual personalities of all of these side characters. So I began dividing the last book (which was at that time still one novel in my mind) into sections. There were five of them. Four of these—one for Rand, one for Egwene, one for Mat, and one for Perrin—would push these four main plots toward the ending. They would happen roughly simultaneously. The other plotlines leading up to the Last Battle, and then the battle itself, were the fifth section.
It became obvious to me early in the outlining process that I was going to be writing a big book. I was well aware of what Robert Jordan had said about the final volume—you can find quotes from him on the internet where he promises it would be so large, fans would need a wheelbarrow to get it out of bookstores. I took this to heart, but knew that there was little chance Tor would let me write the book that large without cutting it.
Indeed, by late 2008, Tor had gotten word that I was promising Harriet a 2000-page book. I believe it was in January 2009 when I got the call from Harriet asking about splitting the books. I was ready for this. My first line was to tell her, “I still view this as one book, and would like to try and get it printed as one book if at all possible.” She took my arguments back to Tor, and had a long conversation with Tom Doherty. When she came back to me, she said they strongly advised a division.
I’m still not certain what would have happened if Robert Jordan had tried this. Perhaps Harriet would have persuaded him that the realities of publishing forbade a book so large. Either way, I felt I had made as strong an argument as I could—and I admitted, despite my desire to see the book as one volume as Robert Jordan had envisioned, that I would have to either discard several major parts of the outline or agree to split the novel.
I think we made the right choice. Three books gave me the chance to really dig into the project not as a one-off event, but as a process. Cutting major plotlines would have made the last book a rushed endeavor, requiring me to ignore several large threads. However, the division of the outline did create some problems, which I’ll talk about during the Towers of Midnight post.
When Harriet asked me about splitting the book, she wondered if there was a natural breaking point. I told her breaking it once wouldn’t work—but breaking it twice might. I didn’t feel A Memory of Light would work as two volumes. Looking at my outline and what I needed to accomplish, two books would either mean one very long book and one normal-sized one, or two books split equally. Both would have been awkward. The former because doing a double-sized Wheel of Time book would have the same problems as just printing the original 2000-page novel. 1400 pages isn’t much better in publishing terms. 1000, like some of the Wheel of Time books, already pushes against those limits.
The second option—two 1000-page books—was even more of a problem. If we cut it in the middle like that, we’d get the first half of all four plot sequences I mentioned above—but none of their climaxes. This (writing one book as a setup book, with the payoffs mostly happening in another book) was an experiment that Robert Jordan had already attempted, and he had spoken of the problems it created. He was a better writer than I am, and if he couldn’t accomplish such a split, I didn’t want to attempt it.
Instead, I felt that splitting the book as three books would allow us to have complete arcs in each one. Two, actually, for each of The Gathering Storm and Towers of Midnight—followed by the climactic book, A Memory of Light. So I set out to divide the plots and decide what would go where.
I knew fans would be skeptical of me taking over the project in the first place, and I knew they’d be more skeptical when we announced a three-book split. That meant I wanted my most dynamic plots in the first book. (I knew the ending would carry its own book, and was never worried about that one being dynamic enough.) In addition, I wanted to split the four sequences—Rand/Egwene/Mat/Perrin—so that we had at least one in each book that Robert Jordan had done a lot of work on. Rand and Perrin had much less material finished for them than Mat and Egwene. So it was either Rand/Egwene or Perrin/Mat for the first book.
It soon became clear that I needed to lead with Rand/Egwene. They mirrored each other in very interesting ways, with Rand’s narrative descent and Egwene’s narrative ascent. When Rand was being contemplative, Egwene’s plot had action—and vice versa. While my personal favorite of the four is Perrin’s arc, I felt his involved a lot of buildup and some less straightforward plotting as we pushed toward his climactic moments. I also decided that the plots would work with shaving off some of what Rand/Egwene were doing to save it for the second book, but I couldn’t do the same as easily for Perrin/Mat.
A book was forming in my head. Rand’s absolute power driving him toward destruction and Egwene’s specific lack of power elevating her toward rebuilding the White Tower. We needed a Mat section—I didn’t want him absent for the book—so Hinderstap was my creation, devised after Harriet asked me to be “more disturbing and horrifying” in regards to the bubbles of evil that were coming into the book.
The Egwene plot was an absolute delight to work on. Of all the things that Robert Jordan had been building for this last book (including the final chapter) before he died, I feel this was the most fully formed. Egwene’s rise and the Seanchan assault played together perfectly in classic Wheel of Time fashion, and I got to participate in unique ways, working with his notes and instructions to craft his plotlines exactly as I feel he envisioned them.
One large change I did make was splitting the Egwene dinner with Elaida into two distinct scenes, instead of one single scene. I felt the pacing worked much better this way, and it complemented the Rand sequence better with the first dinner happening, Egwene getting sent to further work, then a climactic second dinner happening where I could really bring about Egwene’s victory, all without her ever channeling.
In the Egwene sequence, I got to do the most truly collaborative work with Robert Jordan. In other places, I inserted scenes he’d written. In many others, I had to go with my gut, lacking instruction. With Egwene, I had a blend of explanations of scenes, written scenes, and Q&A prompts from Robert Jordan that made me feel as if I were working directly with him to bring about the sequence. If you want to see a full sequence in the books that I think is the closest to the way he’d have done it if he could have, I’d suggest the Egwene sequence in The Gathering Storm. (And beyond. Most of what we have for her was by his direction, inclusive of the events leading up to—and including—Merrilor.)
In taking on this project, one of my personal goals—if the series would allow it—was to focus more time on the main characters, particularly Rand. I love the middle books, with their exploration of other plots and characters, but the first book presented to us Rand, Perrin, Mat, and Egwene as our main characters. I feel that, in the true nature of the Wheel of Time, the appropriate thing to do was bring the attention back to them for the final books—and I feel Robert Jordan would have done so himself.
Rand needed to be the heart of the three novels. In pondering how to accomplish his outline, I was reminded of things I’d felt when first reading The Dragon Reborn. Rand’s anguish as a character was powerful to me, and I thought, “Surely he can’t go lower, be forced to go through more, than he’s had happen to him here.” The next few books affirmed this.
Then I read Lord of Chaos. That book breaks your heart; I found myself amazed that Rand could be brought down even lower. This progressed through the next books, with more being piled upon Rand—but the low points of Lord of Chaos are the most stark in my mind. I remember thinking, “Surely this is the bottom.”
That was why, in The Gathering Storm, I needed to attempt what Robert Jordan had successfully done twice. I needed to bring Rand even lower than the reader had assumed, expected, or even thought possible. This was in part to fulfill arcs Robert Jordan had in place, in part because of his love for the Monomyth and the Campbellian hero’s journey, but mostly because it felt right to me. Rand’s redemption, so to speak, needed to be preceded by his lowest point in the series.
This also offered me an interesting storytelling opportunity. In the original outline, Rand’s descent, his decision on Dragonmount, and his following actions as the Dragon Reborn would all happen in a single volume. In splitting the books, I could do the first part in one book, then have his actions in the second book introduce an interesting tension—the question of whether or not this new Rand was still the Rand we loved. I could prompt readers to fear that just as he became unrecognizable in the depths of his fall, he might become something unknowable in the heights of his redemption. It would make for a new kind of conflict, one I’d never explored before, through Towers of Midnight—before finally giving Rand more viewpoints in A Memory of Light to humanize him again. (Something Harriet was very glad to hear I was planning to do. Her main point regarding Rand was that he, in performing the actions he did in the last book, had to be very human in his approach to them. This was to be the story of an ordinary man who achieved something amazing, not an unknowable deity doing the same.)
I have a fondness for Aviendha, my personal favorite of the female leads in the Wheel of Time. (My favorite among the male leads is Perrin.) I wanted to see a return of Avi in the last books, as I felt we just hadn’t had enough of her lately. I also have an interesting relationship with Nynaeve, a character who I (as a young man) resented. My opinion of her is the one that grew the most during the course of my reading as just a fan, and by Knife of Dreams I absolutely loved her. I knew that with all of the crowding in the last books, she actually wouldn’t have a large part to play in the Last Battle. (Very few would be able to do so, beyond Rand/Egwene/Perrin/Mat.) Therefore, it was important to me to give her a solid and interesting sequence of scenes through both The Gathering Storm and Towers of Midnight. Her raising was not instructed by the notes, but was something I was insistent be in the books. (And along those lines, one thing Harriet insisted happen—and I was all too ready to oblige—was a meeting between Rand and his father.)
To be continued.