From May of 2013 to the end of August 2015, I completed thirteen books. It wasn’t so much that I had a system, or anything like that; I just had plenty of time on my hands and a burning drive to use that time for writing. The point here is not to brag, because fast writing is not a guarantee of quality writing. The point is I had reached a state where I was confident in my ability to finish what I started.

Then things changed. In August of 2015, I was having trouble with the middle of the Willow North novel (which will probably be a trilogy, in the end), so I decided to work on something else for a while. I’ve done that before, writing an entire novel while giving my subconscious time to work out problems with a different one. And I’d had plenty of people ask me what happened to Zara after the end of Servant of the Crown. So I thought it would be interesting to tell that story. I worked out the bones of a plot, created a new culture for the continent south of Tremontane, and dug in.

It fought me the whole way. I made it to about 85,000 words before realizing it was a really stupid book. There were some great characters, but Zara herself was dull. My husband insisted that she didn’t read like an 86-year-old woman, which I think now is true–at the time I resisted that feedback because I couldn’t face the fact that the problem was with the whole book. But the 85K mark represents the point where, having written the same chapter three different ways, I realized the book was a dud. That none of the endings I’d forced into existence–and it took force–had any resonance.

Thirteen novels is enough to make you feel invincible. I was incredibly demoralized by this setback, not least because I had no idea what had gone wrong. Later analysis suggested that I’d written myself into a corner and that Jacob was right about Zara not behaving like an old woman, but at the time it was just frustrating. And I admit to being prideful. Most authors have at least one trunk novel–the kind you finish and lock away in a trunk because for whatever reason, it’s not good enough to see the light of day–and there was no reason I should be different. So I locked Voyager of the Crown in its own file (I am too superstitious to just delete it) and went on to write the four interrelated short stories that became Exile of the Crown.

But I couldn’t get the idea of a Zara novel out of my head. Willow was still giving me trouble, the only other idea I had was stalled out, and I got to a point where I hadn’t written anything in weeks, which was a nightmare. So in January of 2016 I started planning a new book. It was an exciting opportunity to explore Veribold and to finally allow Zara to reunite with her family, and I was looking forward to it.

It was worse than the first one. At 47,000 words I had to admit it was another failure. Once again Zara didn’t behave like an old woman, and the plot was just stupid. This time I was quicker to realize the problem, but it was every bit as demoralizing. I wanted to tell this story and it was clear it wasn’t working out.

At this point, there should probably be some kind of revelation. After all, if I wanted it badly enough, I should be able to figure out the problem, right? Problems are just opportunities in disguise, right?

In this case–no.

For someone who depends heavily on outlines, I’m also remarkably dependent on instinct. I can feel when the shape of a story is working and when it isn’t. And I could tell these were not books I was going to be able to save, no matter how much I wanted to, because they were fundamentally broken. I had to admit to failure. And then I had to move on.

So what makes the difference between a total loss and a temporary setback? I’m still not sure. Wondering Sight, my alternate-Regency-era fantasy with psionics, had a very rocky start where I was working with the wrong plot, but I fixed that and the book turned out fine. Willow North’s book, which I’m currently working on, turned out to have trouble related to the balance between the three plotlines–also fixable. So it’s not as if a problem with writing is always a sign that you should give up. With Zara’s books, I eventually realized that a major part of the problem was that Zara was never intended to be a POV character. When I created her, I was experimenting with ways to make side characters powerful, and part of that experiment was not letting any of the story be told from her perspective. The final story in Servant of the Crown, “Long Live the Queen,” was a departure from that, as were the stories in Exile of the Crown, but Zara was not meant to carry the weight of a novel.

The other problem was that much as I liked Zara, I didn’t actually want to write her story. I was doing it because I felt it would be popular. And that’s not the best reason for writing–trying to game the system. There’s a fine line between having an audience in mind when you write and writing not because you love something, but because you think it will sell. The latter is perfectly acceptable if you’re that kind of writer. Turns out I’m not.

Nobody sets out to fail. I’d rather either one of those books had worked out. But if the alternative was ending up with an awful but completed book, I’d rather take the failure.