You meet HSI agent Jared in a Bilton chapter. He's driving through Chicago. The car goes through a series of streets. Then Bilton notices the Rubik's cube dangling on the keychain. Three chapters later you meet Jared again — the Rubik's cube guy. You don't remember Jared's name. You remember the Rubik's cube. That's exactly what Bilton wants.1
The IRS agent who comes back into the story later: he reads every email three times because he read somewhere that you only remember 20% of what you read, so if you read three times you remember 60%. Every time he reappears, he's re-reading. Oh, the IRS guy.1
This is the character-anchor-trait method. Give every character a quirky recurring behavior that the reader's brain can use as a memory handle. The trait doesn't reveal psychology — it operates as cognitive infrastructure. You're not writing depth; you're writing retrievability.
Bilton can't read Russian novels. He can't keep up with the names.1 The patronymic plus first name plus nickname plus diminutive structure means a single character generates four different lexical handles, and the reader has to track which is which while the actual story unfolds. The reader gets exhausted and gives up.
American narrative non-fiction has a related problem. American Kingpin features FBI, HSI, Secret Service, IRS agents, plus Ulbricht's roommate, girlfriend, mom, college and high school friends. Without anchor traits the reader would have to track twenty proper nouns and would fail. With anchor traits the reader tracks twenty traits and never thinks about the proper nouns again.
The trait is the working memory; the name is the legal record.
Three criteria emerge from Bilton's practice:
Idiosyncrasy. The Rubik's cube is specific. He liked puzzles would not work. The trait has to be small, weird, concrete, and visible — the kind of thing a person would notice and remember about a real human being. Generic interests don't work; specific habits do.
Recurrence. The trait must appear multiple times across the book. Jared's Rubik's cube isn't mentioned once at his introduction; it appears whenever Jared appears, on his keychain, in his car, on his desk. The IRS agent isn't described as a triple-reader once; every email scene has him reading three times.
Visibility. The trait must be observable from outside, not psychological. He was insecure is not a trait; it's a diagnosis. He read every email three times is a trait. The reader sees the behavior, not the explanation. The explanation belongs to the character; the behavior belongs to the reader's memory.
Bilton gives a second example: Jeff Bezos's laugh. People always talk about he's just got this rolling loud laugh. Bezos would never tell you that, but people who've been in his presence are always like, his laugh, his laugh, his laugh.1
This is the externally-reported anchor trait. You can't always observe the trait yourself; sometimes you have to construct it from what other people remember about the person. The mechanism is the same. The reader meets Bezos and you give them the laugh. Every Bezos appearance after that, the laugh is there — and the reader has a memory handle.
Same with FBI agent Tarbell from American Kingpin: his coworkers told Bilton about Tarbell's "would you rather" jokes (would you rather sleep with your mother or sleep with your father; would you rather be half your height or double your weight). Bilton tested it with Tarbell directly. Tarbell played along. The jokes became Tarbell's anchor.1
The character-anchor-trait method does not produce psychological depth. The Rubik's cube doesn't tell you why Jared became a federal agent; it doesn't tell you about his marriage or his fears. That work happens elsewhere — in scenes, in interviews, in the choices Jared makes under pressure.
Anchor traits are scaffolding. They let the reader hold the cast in working memory so the actual character-development work can land. A psychologically complex character with no anchor trait is unmemorable; an anchor trait without psychological complexity is a cartoon. The discipline is to do both.
Bilton and Gelb's character-core-urge are doing the same craft from different ends. Gelb (documentary) builds character by identifying the single deep want that organizes a subject's behavior — the chef who must be the best, the artist who must be seen. Bilton (prose) gives the reader anchor-traits as cognitive scaffolding so the deep want, when it emerges, has a character-shape to attach to. Gelb's work is upstream; Bilton's is downstream. They are sequential moves in the same craft. A narrative-non-fiction writer using both: identify the core urge (Gelb), then build the anchor trait that the reader can use as the urge's external avatar.
Bilton converges with source-material-as-character on the principle that character is built from external observation, not internal speculation. Both refuse to invent interior states the writer hasn't witnessed or had reported. The anchor trait method is one specific implementation of this broader discipline.
Character-anchor-traits maps onto two adjacent domains where the same memory-handle architecture appears under different names: behavioral-mechanics (the baseline-trait that interrogators establish for each subject) and psychology (the cognitive working-memory chunk-architecture that determines what the brain can track during long sustained attention).
Behavioral-Mechanics: Six-Minute X-Ray Elicitation Suite (Hughes) — Hughes teaches investigators to establish a baseline behavioral signature for each subject in the first minutes of contact: this person taps their fingers when relaxed, this person looks left when remembering, this person's voice goes up when nervous. The baseline is the comparison-point that lets the investigator detect deviation later. Bilton's anchor traits work on structurally the same logic, in reverse: the writer establishes the trait early so the reader has a baseline to recognize the character against in later scenes. The structural identity: both crafts establish early behavioral signatures that serve as cognitive comparison-points for later observation. The interrogator detects truth from deviation; the reader recognizes character from recurrence. The insight neither domain generates alone: the anchor-trait method is structurally identical to baseline-establishment in behavioral interrogation — the writer and the interrogator are both designing memory-anchors that let third parties recognize the subject across multiple interactions. A novelist who studied Hughes would write more retrievable characters; an interrogator who studied Bilton would notice subjects' anchor-traits faster.
Psychology: Portable Memory Devices — External Memory Anchors — the cognitive psychology literature on external memory anchors (mnemonic devices, association handles, episodic-memory cues) shows that the brain attaches identity to peripheral concrete details more durably than to central abstract ones. Faces are recognized via peripheral features (the mole, the haircut, the glasses) more than via overall facial structure. Names are forgotten faster than weird quirks. Bilton's anchor traits exploit exactly this constraint — give the reader peripheral concrete handles (Rubik's cube, three-time email re-read, rolling laugh) and the reader's brain encodes the character at high durability. Give the reader name + age + occupation and the reader encodes nothing. The structural identity: the brain's identity-recognition architecture privileges peripheral concrete detail over central abstract description. The insight neither domain generates alone: Bilton's craft intuition for what makes characters memorable maps directly onto cognitive psychology findings about episodic-memory architecture — and the practical consequence is that narrative non-fiction writers who lean on demographic description (40-year-old white male from Seattle who works at a tech company) are producing characters their readers cannot remember, while writers who lean on idiosyncratic concrete habits are producing characters the reader will recall years later.
A third connection: anchor traits operate the same way the iceberg principle operates — give the reader the visible 1/8 and trust their brain to construct the underwater 7/8.
The Sharpest Implication
If the brain's working memory for character is built on peripheral concrete handles rather than central abstract description, then every character introduction that opens with demographics (age, occupation, hometown) is failing on first contact — and most published narrative non-fiction opens character introductions exactly this way. The uncomfortable implication: the standard journalistic profile structure (Sarah is a 34-year-old software engineer from Portland) is structurally unmemorable, and the writers who have learned this are eating the lunch of writers who haven't. Bilton's discipline turns the introduction inside out: lead with the idiosyncratic concrete trait (the Rubik's cube on the keychain) and treat the demographic detail as background that the reader will pick up incidentally.
Generative Questions