Kevin on Kraft: When Your Characters Stop Following Instructions

Article version of episode 109

[Note: you can watch or listen to the podcast version of this here.]

I missed an episode of The Alignment Show last week.

The reason is probably appropriate for a show about living your values to value your life: I got absorbed in something that matters deeply to me.

I’ve been writing The Chemistry Between Us, a romance novel I’ll publish under the pen name Devon Marin. And “absorbed” may be putting it mildly.

I’ve forgotten to eat. I’ve awakened early because Riley and Leo, the two central characters, were having a conversation in my head and I needed to capture it before it disappeared. I’ve occasionally resented the human body’s unreasonable insistence on things like bathroom breaks.

That experience also made the timing of my latest conversation with Kevin Tumlinson almost perfect.

This is the second installment of Kevin on Kraft, our occasional series of conversations about writing. Kevin has published somewhere north of 90 books at this point, although even he isn’t entirely certain of the number. More importantly for our purposes, he has spent years writing his own fiction and talking with other writers about how they do what they do.

This time, I wanted to talk about characters.

Specifically, I wanted to know when one of Kevin’s longest-running characters, Dan Kotler, stopped being someone Kevin had designed and started telling Kevin who Dan Kotler was.

Kevin’s answer? Almost immediately.

Once he knew enough about Kotler to begin writing, Kevin could put him in situations and see how he responded. Kevin describes himself as a discovery writer. He doesn’t necessarily know where the story is going before he begins. The choices made by the characters can determine where it goes next.

I recognized the experience. I, too, confess to being a discovery writer.

I began The Chemistry Between Us with an outline influenced by Save the Cat! and Romancing the Beat. I knew where I thought Riley and Leo were going.

They had opinions about that, and sometimes creative differences.

That led Kevin and me into the strange territory writers sometimes enter when fictional people begin to feel surprisingly real. We talked about imaginary friends, muses, dreams, visualization, and the experience of hearing characters speak.

But there’s a practical craft question underneath all of that: What makes a character feel real?

One of Kevin’s answers intrigued me, because it involves what you leave out.

He generally avoids detailed physical descriptions of his characters unless those details actually matter to the plot or the personality somehow. He would rather give readers enough information to activate their own imaginations.

He used Jaws as an example, a movie that illustrates the principle for books. Mechanical problems with the shark forced Steven Spielberg to show less of it than originally intended. The shark wound up more frightening because the audience’s imagination supplied what the camera withheld.

Something similar can happen on the page.

A writer may carefully describe a monster as six feet tall with particular colors, teeth, claws, and scales. But the moment we define it that precisely, we may replace whatever terrifying creature the reader had imagined with our considerably more limited description.

Character description can work the same way. That doesn’t mean physical details are useless. It means they should earn their place. A detail can reveal something about the character. It can violate an expectation. It can become significant later.

I realized during our conversation that I had done exactly that with a supporting character in The Chemistry Between Us. Early in the book, I established that she wears glasses and sometimes looks over the rims at people. Later, taking those glasses off becomes a signal. Another character recognizes it immediately: things are getting serious.

Now the glasses go beyond decoration into characterization.

Sometimes we discover the significance of something only after we’ve written it. Kevin calls this one of the dirty secrets.

You get to go back.

If you arrive at Chapter 19 and discover that something would work beautifully if only you had established it in Chapter 3, you can return to Chapter 3 and establish it.

The reader never needs to know. That isn’t cheating; it’s revision.

Kevin compared discovery writing to 3D printing. The first pass gives you a structure. Subsequent passes add layers until the finished object emerges. The reader experiences the finished story from beginning to end. The writer doesn’t necessarily have to create it in that same order.

When you write a series some discoveries happen across books rather than chapters.

For instance, Kevin described giving one of his characters claustrophobia without having developed an elaborate origin for it. Several books later in the series, he “discovered” a backstory that explained the trait. To the reader, it could look as though Kevin had known or planned the entire history from the beginning.

Spoiler: He hadn’t. He learned it as he was writing the later book. But he knew the character well enough to recognize the history when he found it.

Of course, the longer we live with characters, the harder it becomes to remember everything we’ve ever said about them. One time, a reader pointed out to Kevin that a later Dan Kotler book contradicted something established in an earlier one about whether Kotler had a middle name.

After fourteen novels, several novellas and short stories, and crossover appearances, keeping track of everything one fictional person has ever done becomes a serious information-management problem.

Of course, that led to AI. (It really is a useful tool for creating a story bible or series bible.) Kevin talked about the possibility of using AI to help maintain a character wiki and flag contradictions across a long-running series. That isn’t asking AI to invent the character. It’s using the tool to help the writer remember what the writer has already created.

(It didn’t come up in the episode, but while I don’t use AI to write my books (neither does Kevin), I used it to make a series bible for the characters and the timelines in The Sparklight Chronicles. It only has two books right now, but two more are in process. I had enough trouble with those. I can only imagine what it must be like keeping up with fourteen.)

I asked Kevin what he wished he had known about characters when he began writing fiction. Among other aspects of the answer, he repeated: don’t assume that describing more creates a more vivid character. Just use the details that tell us something. It took him a few books to learn that.

Just when we were supposed to be wrapping up, we managed to open another box.

Inside that box, we found:

• What about character arcs?

• Does a character always have to change?

• What happens when you’re writing the fourteenth book about someone? How do you let that person grow without turning them into somebody else?

And then Kevin opened another box, one I’m dying to hear him dig around in: How do you write a character who is smarter than you are?

I poked another box: How do you convincingly write someone whose characteristics and experiences differ significantly from your own?

At that point, we did the sensible thing. We stopped.

Those questions deserve their own conversation, and Kevin and I plan to pick them up in the next Kevin on Kraft.

Meanwhile, Riley and Leo are rummaging around in my head changing something I thought I had already settled with them.

I should go see what they’re up to.

---

I’m Donn King. I write about what it means to live and speak from your values, without sanding off the quirks that make a voice worth hearing. This space is where I reflect, tell stories, and stay connected with readers who value clarity, honesty, and the long work of becoming who we’re meant to be.