Deliberate play and editing novels with Claude AI.
Do you find it as satisfying as I do to learn a new term that describes something you do?
Especially when you already had a term for it and the new term sounds better, and makes you feel better about yourself!
It turns out that what I thought were “clever procrastinating tactics” were actually “deliberate play”. So you can see why I’m feeling pretty smug today.
Deliberate Play
In the book “Hidden Potential” (which I’m reading at the moment), Adam Grant says “deliberate play often involves introducing novelty and variety into practice”. So I can’t use the term to describe when I’m struggling with writing a piece and my mouse, as if it had a will of its own, opens a new browser tab and sends me to LinkedIn.
But I CAN use it when I’m struggling with a piece of fiction writing and I pause my writing to head over to ChatGPT for feedback. I’m still honing my writing skills, but I’m doing it by introducing something that is easier and more fun for me to do. (I wonder if going to my jigsaw puzzle when I’m struggling with a scene is deliberate play, or simply play…)
Of course I then need to go back to writing on my own at some point, or else deliberate play, just becomes “play-crastinating”. (Adam didn’t come up with that one, I did. That’s what happens when you place your fingers on your keyboard and get out of your head.)
My New Developmental Editor
A developmental edit in a novel looks at the whole structure of the piece, its pace, character arcs and other high-level aspects of writing fiction. As a writer, it’s important to have an editor doing this, because it’s difficult to distance yourself from your own work. (Also, in the case of many writers like myself, we don’t have the deep knowledge that editors do.)
I’m not at the stage yet of contracting a professional to do this for my piece - it would be a waste of their time and a waste of my money. But I’ve finally finished most of the first draft of the cozy and I’m ready to assess it as a whole.
I have plenty of books with generic advice on editing, so I will dig them out for guidance, but I would also like some specific pointers on how to approach my own piece. So I turned to Claude AI.
”Claude?” I hear you say. “What’s wrong with ChatGPT?”
Nothing, but I find Claude’s responses more down to earth and less “clowny”, even though it can totally aim to please too. It’s also got a nicer and shorter name, which helps when you’re writing about it.
The first step to “engage Claude as an editor” was to compile all the words into a PDF in Scrivener, a great writing tool. I then fed the PDF into Claude (all 69,868 words of it) and started asking it questions.
As I’ve mentioned before (and thanks April for picking up on this!) one of the great things about working with genAI on artistic projects is that it helps dig into your subconscious. In order to get decent feedback, you have to ask decent questions. And it’s easier to ask good questions about your work to someone else (or somebot else), than to yourself. So as well as obtaining suggestions about how to improve the piece,from Claude, I created my own editing plan. And it didn’t feel like work.
I’d engaged in “deliberate play”.
Hints and Mistakes
On the whole, I was very impressed at how specific the feedback was to the piece, and how obvious it was that Claude had read the whole document. This in itself is amazing - it took about half an hour to get all the information I needed from it, and I was able reformat the Word version of the script for printing while waiting for the answers.
But the title that Claude has given to our chat has reminded me to stay alert:
“A Mysterious Disappearance at the Chiswick Park Studio”.
Well, there are no “mysterious disappearances” in the piece, just two murders. And the main character does live in Chiswick Park, but the studio is in Soho… It also referenced the wrong chapters some times when suggesting additions.
Let’s leave that to one side and focus on what did work.
Main Character Development
After asking Claude to check whether the subplots had all been tied up, I addressed something that I’m worried about: my main character is too much like me.
A few times in my writing, I’ve had my character react in ways that I would react myself ie conscientiously. That is a problem when you have a main character who is supposed to be carrying out her own detective investigation, going into places she shouldn’t be going into, and probing into people’s lives. I also wanted to have fun while writing the character, but if all she does is make the same choices I would make, there won’t be any surprises for myself, or the reader.
To help with this, I came up with a trait early on in the piece for my character: taking notes on her phone constantly about who she was meeting, her first impressions of them, etc. I realised I had dropped this at some point, so I asked Claude to identify more times when this could happen throughout the piece.
All 8 suggestions it gave me were worth following up. I liked this one, for example, which could take place after the second murder: “In Chapter 13, after John's death, Mia could take notes about the similarities between John and Ricky's deaths while she's in the studio.”
This is a more active way of getting her to decide to look for poison in the kitchen, than her just thinking about it.
I then asked:
”It would be good to make the note taking a bit of a "super power" for Mia. How do you think I might do that?”
One of the suggestions I liked was: “Mia's notes could contain a crucial piece of evidence that she didn't even realize she had until she reviews them. A seemingly random observation could hold the key to unlocking the mystery.” I like it because if I, Pilar, make notes about something, I rarely go back to them!
As our strengths can easily turn into our weaknesses, I asked Claude:
”I'm not sure if I want Mia to be as conscientious. Could you give me some ideas of how she could be less conscientious, like some of the actions or decisions she could make (while keeping the note-taking); or how to make the note-taking a liability at some point? (eg maybe she makes notes cause she's a bit scattered brain and someone finds her notes...???)”
I liked this one:
Mia's note-taking could be a compulsive habit that she sometimes does at inappropriate times, like during a sensitive conversation or while someone is sharing personal information. This could lead to misunderstandings or hurt feelings.
And I’m tempted by this one…
Mia's notes could be disorganized and scattered, reflecting a somewhat chaotic thought process. She might have trouble finding the information she needs when she needs it, leading to confusion or delays in her investigation.
This one is also interesting:
Mia could be so focused on taking notes that she misses important social cues or opportunities to gather information in the moment. Her reliance on written notes could sometimes hinder her ability to read people or situations effectively.
Secondary Characters
I also asked Claude to tell me if my main character’s best friend was too much like her - she is - and whether the character of Abuela worked.
In the novel, Mia often shares her findings with the picture of her grandmother she has by her bedside. I wanted to have a literary device to help me get some of her thoughts “out there” and to give more time-depth to the piece. However, I’m aware that at the moment, the character is not sufficiently developed to be interesting. Unfortunately, Claude agrees…
The character of Abuela, as presented in the current draft, does run the risk of feeling like a literary device rather than a fully developed character. Mia's conversations with her deceased grandmother often serve as a way for Mia to process her thoughts and feelings about the case, but Abuela herself doesn't have a strong presence in the story.
However, I think the concept of Abuela as a guiding force in Mia's life has a lot of potential. Here are a few suggestions to make Abuela feel more integrated into the story and less like a literary device.
”Mia's conversations with her deceased grandmother” is the type of clue I look for to let me know that the bot has indeed scanned and “understood” the piece.
Creating The Editing Plan
Without realising it, I had my edit planned!
Even if all of Claude’s answers were useless (which they aren’t), this would have been an incredibly valuable exercise. I know what I need to be looking out for - having extra guidance from Claude is a bonus. But as a last prompt, I asked Claude to draw up a plan for the editing stage. (This needed to be the final prompt as there is still a cap on messages in Claude, even at Premium level.)
In case you are editing your own piece, here’s an edited version of the advice. I’ve left most of the detail out, as it’s general writing advice.
1. Review the story's structure and pacing.
2. Analyze character arcs and development.
- Implement the suggestions for making Bea and Abuela more well-rounded and integrated into the story.
3. Assess subplots and loose ends:
4. Evaluate the climax and resolution. (I asked it a few questions about this during the session.)
5. Examine themes and motifs.
- Identify the central themes of the story (e.g., the blurred lines between fiction and reality, the power of observation) and ensure they are consistently developed throughout the narrative.
(This one needs another session where we work on theme and how to incorporate it. I love it how theme emerges subconsciously through the writing.)
6. Polish dialogue and prose.
7. Get feedback and revise.
Throughout the developmental editing process, it's important to keep the big picture in mind. Focus on the overall story arc, character development, and thematic resonance rather than getting bogged down in minor details or line edits. Make notes as you go, and don't be afraid to experiment with different ideas or approaches.
Remember, the goal of a developmental edit is to refine and elevate the core elements of your story. By focusing on the structure, character, themes, and overall impact, you can take your draft to the next level and create a more engaging and satisfying reading experience.
If I wanted Claude to continue doing the hard work, I could go back to this list and ask it for more specific guidance relevant to my piece.
But really, that would be moving from “deliberate play” to “deliberate delegation of the hard work”, which I’m not ready to do yet!
I’ve still so much to learn.
Let me know if you have any questions about this or any of the other processes and experiments I talk about in this newsletter.