Stroke

Stroke grew up in a small village, in the heron family who had always kept the village records. They signed the important papers, and they were trained to know the handwriting of every soul in the village — whose careful loops, whose hurried scrawl, whose fancy signing swirl.

It was slow, step-by-step work, checking each mark against the ones they knew. And by the time Stroke was six heron-years old, she'd learned the lesson at the heart of it: a gut feeling was never good enough. Anyone could feel sure. Only checking every tiny feature, patiently, one at a time, could make you actually sure — or honestly tell you that you weren't.

One day, when she was twenty-two, Stroke walked to SleuthLab, where Inspector Vex was waiting.

"What is document analysis?" he asked.

Stroke stood tall. "Compare specific features — not just how it looks. Letter shapes. Slant. Pressure. Spacing. If many specific things match, you can be very sure. If only a few match, it's probably a different writer — or the same writer in different conditions."

Inspector Vex nodded slowly. "You are appointed."

In the lab, Stroke teaches the steps for document analysis. First, compare specific features — letter shapes, slant, pressure, spacing, and swirls, not the overall "look." (She'll tap her glass: "Does the t-crossbar join the stem in the middle, at the top, or not at all? Do the loops close full, or stay open?") Second, collect enough examples — writing changes with your mood, your speed, even your paper, so you need many samples to learn how a person usually writes. Third, use the same conditions when you can — similar paper, similar pen, similar words, or the differences will fool you. Fourth, look at the ink and paper too — what's the ink made of (their friend Drop can help), how thick is the paper, is there a watermark? Fifth — and she warns about this most — don't trust your gut: "If you want the writing to match, you might see a match that isn't there. So check every detail, one by one, and try not to know who wrote it beforehand — that's called being blind, and it keeps you honest." Sixth, think about class-versus-individual (as their friend Loop teaches): some features only narrow the suspects; others point to one person. And always, junior-forensics scale — "Who wrote the anonymous note teasing the class clown?" not huge scary crimes. "It's the method that matters," she says, "not the guess."