This week I submitted chapters three and four of Raging Alone to the writing group. There were some good points made but one of the ones that the entire table brought up was the names.

Like Sand and Blood, Raging Alone is focused around a small, isolated clan in the desert. The language and the names were inspired by people I've worked with over the years with a healthy borrowing from French and Japanese. However, they appear not to be names that are easy for English readers to understand.

Once I figured that out, I made a point of having Sand and Blood start off slowly to introduce characters and conventions gradually instead of throwing a lot of names quickly at the reader. In some ways, it ended up coloring that entire piece; more than once I was told it started off too slowly which dragged out Rutejìmo's flaws.

With Raging Alone, though I got much of the same thing. However, since I just finished a round of Author Intrusion, I realized I could quantify my efforts using what I just finished writing wrote.

Metadata

I keep information about the chapter in the top of each file in a YAML “front matter”.

---
title: Sulking
characters:
  primary: [Desòchu]
  secondary: [Kiramíro]
  referenced: [Gemènyo, Hyonèku, Kiríshi, Chyojímo, Hikòru]
---

> There is no more graceless being than a teenage boy coming into his age. --- Nagufiga Makinàfu, *Sexuality Among the Sands*

Like most typical days in Shimusogo Valley, the hot midday blew winds...

The key part is the characters section which lists the point of view (primary), characters in the chapter (secondary), and those who are talked about but not present (referenced).

This is also what is used to populate the footer of the chapters such as this one at the bottom.

Author Intrusion

Using the last post for setup, I can use JMESPath to create a comma-separated list of characters that are in the chapter.

If I just wanted a list of secondary characters, I would use: characters.secondary[] | sort(@) | join(', ', @) which says get all the items from characters.secondary, then sort them, then combine them together.

If I want all the characters from all three categories (primary, secondary, and referenced), I can just replace secondary with * to get characters.*[] | sort(@) | join(', ', @).

Related to that, I can also get the count of character in the chapter with: characters.*[] | length(@).

Analysis

Running the tool on the first five chapters of Sand and Blood, I get:

$ author-intrusion chapter list -f file.num -f title -f "characters.*[] | length(@)" -f "characters.*[] | sort(@) | join(', ', @)"

Well, seen this way, I didn't do a very good job of “gradually” introducing character names at all. One might said I did an absolutely terrible job of introducing the constructed language. I guess that points out that a subjective belief verses objective analysis.

Now, I don't distinguish characters who are there but not named so the list is smaller. Below is a quick test of the seven names in the first chapter. I use expr - 1 because there is the entry in the matter which I don't want to include and readers don't see.

sand-and-blood$ for i in Chimípu Ganósho Gemènyo Hyonèku Rutejìmo Tejíko Yutsupazéso;do echo -n "$i: "; echo $(expr $(grep $i chapter-01.markdown  | wc -l) - 1); done
Chimípu: 2
Ganósho: 0
Gemènyo: 1
Hyonèku: 22
Rutejìmo: 23
Tejíko: 0
Yutsupazéso: 3
sand-and-blood$

Above you can see that Ganósho and Tejíko were not listed by name but the others did. Regardless, what I thought was a gradual introduction was closer to me throwing a bowlful of names at the reader and saying “here you go!”

Let's see if I did better with Raging Alone:

Well, I've done a lot better job of reducing how many names are thrown at the reader but it also points out I'm still throwing a lot of them in the first chapter. Getting that number down to four or five would make it an easier introduction and points out a possible place to trim down the story. In the writing group, we found a way of cutting out two or three names out of “The Call” which would make it much more manageable.

Now, chapter five is the “inciting event”, so where I would normally split a chapter, I would probably want to add it after chapter five so we get to the “big event” faster. But that's for another blog post.

Conclusion

Sometimes, its hard as an author to see problems in the story while writing it. One of the reasons I created Author Intrusion was to help me focus them the problem I struggled to see. In this case, I thought I was being gradual with names but once it was pointed out, it was obvious I was not.

I also realized there are some additional checkers I could include to help identify these problems while writing:

• #17: Introduce a checker that counts a number of entries (a JMESPath such as one against characters) against a threshold
• #18: Ensure that every chapter has an epigraph (my style guide)
• #19: Check that a character in narrative is identified in the metadata.