Chapter 14. Five Diagnostics, and a Final Farewell

Across thirteen chapters, we followed the process of building a character from zero to output: from seed to self-objectification, from the everyday-slice scenario through twisting, drama, and identification, from flaws and trauma to goals and conflict, then through the relationship network, immersion, five-media output, and nonstandard characters. All twelve stages are complete.

This chapter combines two tasks: five diagnostics that test whether the finished character passes the acceptance threshold, and a reckoning and farewell for the lecture as a whole.

Does the Character Contribute to the Work?

The final act of a character build is to ask whether the person contributes to the work. A successful build does not automatically produce a useful dramatic character. Whether the person occupies a distinct place and possesses a distinct texture inside the work is a separate question.

We use five diagnostics drawn from criticism and creative practice. Passing all five gives strong evidence that the character has a place in the work. Depending on the work, one or two failures may be acceptable; some failures, however, require immediate redesign.

Diagnostic 1: Bechdel Test

The Bechdel Test originated in a 1985 comic by Alison Bechdel.

Three work-level conditions:

The work contains at least two women.
They speak to one another.
Their conversation concerns something other than a man.

All three must be met. This is a low baseline, and passing does not guarantee good representation of women. Failure is nevertheless a warning that the work may not be giving women enough dramatic space.

M is a woman, and her cast includes a female junior forensic pathologist. A scene in which they discuss an autopsy or working together passes. This is a test of the whole work, not of one character in isolation.

Diagnostic 2: Mako Mori Test

The Mako Mori Test takes its name from the character in the 2013 film Pacific Rim. Mako possesses an arc of her own: she seeks to become a Jaeger pilot after losing her family, rather than existing merely as a tool for a male protagonist.

The original three conditions:

At least one woman appears.
She has her own narrative arc.
That arc is not reduced to serving another character.

Although created to evaluate women characters, we generalize it here and apply it equally across genders and identities.

Is the character's five-act transformation in Stage 11 their own change rather than a device for someone else?

M's arc is her own acceptance of her mother's absence. E's arc is the reproduction of trauma after failing to save a seven-year-old hostage. J's arc is the decision to reject a bestselling mentor and preserve their own stance. All three pass. A process that strongly establishes an independent arc makes a pass more likely.

Diagnostic 3: Sexy Lamp Test

The Sexy Lamp Test was proposed by Kelly Sue DeConnick.

One condition:

If the character can be replaced by a sexy lamp without changing the plot, the character fails.

It tests active contribution. Replace M with a lamp and no one discovers a wedding ring on the corpse, chooses to put it in a bag, or violates procedure. M's decisions move every event, so M passes.

A character whose removal leaves the story unchanged must be redesigned immediately. The test asks whether the character is the subject of events or merely their object. Ripley in Alien passes easily: replace her with a lamp, and the decisions that govern survival disappear.

Diagnostic 4: Walks-into-a-room Test

The fourth is an informal industry tool.

When the character first enters a scene, can the audience identify who they are and what they want within five seconds?

Three tools support that recognition: one appearance factor, a physical anchor, and speech. All three should operate in the first appearance.

For M:

- Appearance factor: a white shirt under autopsy clothing. Pass.
- Physical anchor: calluses inside the two joints of the left index finger. Pass.
- Speech: a short, exact line such as “Twice preserves it.” Pass.

M passes. Failure means the first appearance is inefficient and the audience cannot recognize the person quickly enough.

Diagnostic 5: One-Line Test

A well-built character can be read without a long explanation. WALL-E speaks very little, yet his work among Earth's rubbish, the objects he collects, and the way he looks at EVE reveal loneliness, love, and curiosity. When action and gaze speak for the person, the character approaches completion.

One condition:

Describe the character in one line. Beyond the word “protagonist,” does the line contain a trait that distinguishes this person from every other character?

“A forensic pathologist with thirty years of experience discovers her missing mother's wedding ring on a corpse” is strong. “The protagonist has a dark past” is weak. “A cold contract killer disobeys an order for the first time when faced with a seven-year-old hostage” is strong. “A capable detective” is weak.

The first line of the Stage 12 integrated sheet must pass. M's line—“A forensic pathologist with thirty years of experience finds her mother's lost ring on a corpse and confronts the strength she deliberately concealed”—cannot be transferred to another character unchanged.

Diagnostic 6 (Supplementary): Pixar Truth Test

The supplementary diagnostic returns to the Pixar Truth Test from Chapter 9. It asks whether the opening Lie and ending Truth face each other like a mirror, and therefore whether the ending is causally consistent with the beginning.

M's opening Lie: “If I do not learn what happened to Mother, she may still return alive.”
M's ending Truth: “Even without learning what happened to Mother, I can live with her absence.”

The statements mirror each other, so M passes. If the ending Truth does not answer the opening Lie, the causal chain of the work is broken.

Integrated Review of Five Plus One Diagnostics

The six results for M are:

M's diagnostic results:
- Bechdel: Pass (multiple women in the work)
- Mako Mori: Pass (an independent arc)
- Sexy Lamp: Pass (events depend on M's decisions)
- Walks-into-a-room: Pass (appearance, anchor, and speech work immediately)
- One-Line: Pass (an irreplaceable one-line description)
- Truth Test: Pass (Lie and Truth mirror each other)

All six pass.

E, a morally gray figure, produces a different pattern:

E's diagnostic results:
- Bechdel: Fail (almost no women in the war setting). A warning requiring review of whether the bias is intentional.
- Mako Mori: Pass (an independent arc of self-forgiveness concerning C's death)
- Sexy Lamp: Pass (E's choices—refusing orders and attempting rescue—move events)
- Walks-into-a-room: Pass (shoulder burn and cold speech identify E in five seconds)
- One-Line: Pass (“A cold contract killer disobeys an order for the first time before a seven-year-old hostage.”)
- Truth Test: Pass (the Truth is clear even though E never reaches it; deliberate tragic non-arrival)

Diagnostics are mirrors, not a checklist demanding every circle. Bechdel may be inapplicable to a one-person story. E's failure asks whether the absence of women is conscious or merely overlooked. Two diagnostics, however, are mandatory:

Sexy Lamp failure: redesign immediately; the character contributes nothing to the work.
One-Line failure: rewrite the hook; the character is not distinguishable from others.

A character who fails either should not enter the work unchanged.

Close-up · Katniss Everdeen in The Hunger Games Katniss almost textbook-perfectly passes the diagnostics. Her bow, mockingjay pin, and the one line “the girl who volunteered in her sister's place” identify her immediately; her decisions drive events; and she possesses her own arc from protecting her sister to becoming a symbol of revolution. A strong protagonist passes naturally without being designed merely to satisfy tests. Diagnostics are mirrors, not goals to fill.

Who Does What?

AI is strong at a first diagnostic pass. Given the integrated sheet and cast, it can quickly assess all six.

The producer decides whether to accept or redesign the character. Responsibility for tone, concept, and audience remains human. Even if an automated assessment passes, the producer redesigns when the result does not fit the work.

Reckoning of the Whole Lecture

Ursula K. Le Guin wrote in the introduction to The Left Hand of Darkness that fiction is not prophecy or proof but a kind of thought experiment: place people in a condition and ask what follows. Character building is similar. We do not force a predetermined answer onto a person; we design a strong question and watch how the person answers.

The fourteen chapters moved as follows: Chapter 1 established the stance; Chapter 2 mapped twelve stages; Chapter 3 defined four levels and five preparatory modules. Chapters 4–6 covered seed, the five mirrors, and the everyday-slice story. Chapter 7 handled twisting, drama, and identification; Chapter 8 flaws, deficiencies, and trauma; Chapter 9 goals and conflict. Chapter 10 built the relationship-network matrix; Chapter 11 immersion through questions, letters, diaries, address, speech, and five acts; Chapter 12 five-media output; Chapter 13 nonstandard characters and diversity; and Chapter 14 diagnostics and reckoning.

The tools are numerous, but they matter only atop two stances: treat the character as a human being, and let the producer lead the work with an AI work partner. The first refuses to reduce the person to data; the second separates rapid first output from human decisions, compression, and acceptance judgments.

Acknowledging the Limits

This process is not the only correct method. The twelve stages are a guide, not a command.

Some writers move directly from a seed to a 1,500-character story and skip the mirror. Others begin with an image in the mind rather than randomized cards. Both can produce living characters. A creator may borrow one part of this process or none of it. The lecture argues not “follow these steps exactly,” but “such a process is possible.”

The time budget of eight to fifteen hours assumes AI assistance. A workshop that works entirely by hand may need two or three times longer. That is not inefficiency; manual work has its own depth. The stated budget simply does not transfer unchanged.

Points of Contact with Companion Lectures

This lecture concerns building a new character from zero.

Two companion tasks stand beside it.

First, preparing foreign-language texts for translation: deeply reading and organizing an existing work before translation. If this lecture is about making, that task is about reading.

Second, building an integrated library for a worldbuilding IP: gathering the finished characters and every document about their world in one place. If this lecture creates one person's folder, that task creates the library containing every folder.

The three form one line: build, read, gather. This lecture addressed the first.

To the Colleagues Who Worked Together

Many hands contributed to turning the process into fourteen chapters: defining its stance, arranging twelve stages, building the five measures and five-media formats, and honestly separating producer and AI responsibilities. The principle “with affection rather than anger” first emerged in that collaborative workplace.

Thank you also to readers who followed all fourteen chapters. Take any part that supports your work, or take none. I hope, however, that you keep hold of two stances: treat a character as a human being, and let a producer lead the work with an AI work partner.

Closing: For Those of Us Who Say Producer, Not Writer

We return to Chapter 1. We called the person who builds a character not merely a writer but a producer. A writer writes text directly; a producer is responsible for content from seed to output. The roles may overlap or separate.

The process follows the producer's flow of consciousness: not “How do I write one beautiful line?” but “How do I lead one character from zero to output?” Each stage belongs to that flow.

This distinction mattered throughout. Every chapter separated what the producer must decide from what an AI work partner can draft. Honest separation supports both efficiency and quality.

We have taught the work to the end of fourteen chapters, but the work itself does not end here. The character enters a story; the story is published; readers meet the character; and readers speak the character's name again. This lecture cannot accompany every later stage. If it gives you tools capable of reaching them, it has done its work.