Can You Know a Book Better Without Reading It? — Semiotic

Home · Getting Started · Charts · Features · Playground

This page requires JavaScript for interactive chart demos.

AI / Machine-readable docs: llms.txt · llms-full.txt · Full AI Reference · schema.json · API Reference · Examples

View on GitHub · npx semiotic-ai for CLI access

Machine-readable page content

Canonical: https://semiotic.nteract.io/examples/distant-reading

Can You Know a Book Better Without Reading It?

A remake of the distant-reading essay

This interface compares four novels through chapter signals, narrative phases, and character flows. Use those views to locate structural changes, then return to the text with a more specific question.

1818

Frankenstein

Mary Shelley

Its strongest shape is not the monster. It is nested testimony that turns inwardness into motion.

24chapter bins

74,500modeled words

Ch. 23peak pressure

10volatility

The book as a signal field

Pick a signal, then scan the chapter map. A tall mark identifies a chapter where that signal is concentrated and gives you a place to ask what changed.

Interior close-upChapter 11

Reflection, self-argument, memory, shame, and inward narration. Score: 80. Dominant texture:interior. Estimated local weight: 3,058 words.

Interior through the book

A chapter-level trace lets the reader see where the chosen signal gathers, breaks, and returns.

Where that signal lives

The same chapter scores are collapsed into narrative phases, trading local detail for a quicker structural read.

A distant reading still has characters

The Sankey traces who frames whom, which forces mediate action, and which places organize the book. It summarizes modeled relationships rather than plot events.

QuestionWhen does confession become pursuit?

Source note

Modeled as 24 chapter-level bins from the 1818 public-domain novel.

Corpus fingerprints

Each dot is a chapter. The books keep their color, so clustering shows when different novels briefly occupy the same signal territory.

Visualization changes the first question

The data model is intentionally small: four local signal arrays, phase summaries, and a network per book. These views prepare questions for close reading; they do not interpret a passage or settle what a novel means.

JSX

import { LineChart, BarChart, SankeyDiagram } from "semiotic" import { BOOKS, signalRows, phaseRows } from "./data/distantReading" const book = BOOKS.find((d) => d.id === "frankenstein") const activeSignal = "interiority" <LineChart data={signalRows(book, activeSignal)} xAccessor="chapter" yAccessor="value" yExtent={[0, 100]} /> <BarChart data={phaseRows(book, activeSignal)} categoryAccessor="phase" valueAccessor="score" orientation="horizontal" /> <SankeyDiagram nodes={book.network.nodes} edges={book.network.edges} nodeIdAccessor="id" sourceAccessor="source" targetAccessor="target" valueAccessor="value" />