Book Anatomy: ask any classic book about the craft that made it great

Ask any classic book about the craft that made it great. and get answers that cite chapter and paragraph. 3 free questions a day. Take one great book apart and see how it works. Each book is a hub of 180 lenses in 12 groups: beats, chapter hooks, cast, dialogue, pace, viewpoint, motifs and more. Counts come from the full text. AI readings cite chapter and paragraph, and every quote is checked against the text.

Browse by genre · Browse by lens · Your scrapbook · Plans (free beta)

What's new in Book Anatomy

• 2026-10-15: Ask the book now sits at the top of every book, with example questions
• 2026-10-15: Answers now check every chapter they name against the passage they cite
• 2026-10-15: 175 lenses on every book, in 12 groups
• 2026-10-15: Now 97 classics, including 12 plays

Browse by genre

Thriller and adventure

Crime and mystery

Romance

Gothic and horror

Science fiction

Fantasy and children's

Literary

Plays

Browse by lens

See one lens across every book. The top 20 first:

• Beat positions Both
• Chapter length chart Count
• Dialogue share per chapter Count
• Cast and presence map Count
• Chapter ending hooks AI
• Viewpoint per chapter Both
• Want and need of the lead AI
• Scene log AI
• Setups and payoffs AI
• Speech fingerprints Count
• Sentence length per chapter Count
• Emotional arc shape Count
• When the lead and the first dialogue arrive Count
• Mystery clue list AI
• Antagonist AI
• Tension curve AI
• Subplot thread map Both
• Motif tracker AI
• Signature words Count
• Commercial facts card Research

All 180 lenses

Grouped as in our plan. Count lenses are exact; Both mixes counts with AI reading; AI lenses cite only checked quotes; Research lenses are coming later.

Story and structure

• A1. Where do the big beats fall as a % of the book: inciting incident, midpoint, low point, climax? Both
• A2. What is the inciting incident and how many words in does it land? Both
• A3. How many acts or parts, and how long is each? Both
• A4. Scene log: goal, conflict and outcome for every scene. AI
• A5. Which scenes shift a value (safe to unsafe, love to hate) and which stay flat? AI
• A6. What changes at the midpoint and how does the lead's plan change? AI
• A7. How do the subplots weave? Thread map by chapter. Both
• A8. Setups and payoffs: what is planted, where, and where it pays off. AI
• A9. How many try and fail cycles before the goal is met? AI
• A10. How do the stakes rise chapter by chapter? AI
• A11. Do scenes link by "but" and "therefore", or only "and then"? AI
• A12. Map to the 15 Save the Cat beats. AI
• A13. Map to Yorke's five acts. AI
• A14. Map to the hero's journey stages. AI
• A15. Crisis choices: what hard choice does the lead face in each act? AI
• A16. Reveals and twists: where they fall and how they were hinted. AI
• A17. Timeline: story time against page time, gaps and flashbacks. Both
• A18. Sequences: which runs of chapters form one mini story? AI
• A19. Ratio of plot chapters to breather chapters. Both
• A20. Luck count: where does chance help or hurt the lead? AI

Character

• B1. Cast list with first appearance and share of mentions. Count
• B2. Who is on the page in each chapter (presence heat map)? Count
• B3. What does the lead want, and what do they truly need? AI
• B4. What false belief does the lead hold, and how is it tested? AI
• B5. Arc type: positive, flat or negative? AI
• B6. Key arc moments with chapter and line. AI
• B7. The antagonist: goal, methods, and when they show their hand. AI
• B8. How is each main character first shown (first line, first act)? Both
• B9. Who appears with whom? Network chart of shared scenes. Count
• B10. Which relationships change most over the book? AI
• B11. Foils and mirrors: who reflects the lead? AI
• B12. Backstory wound: how and when is it revealed? AI
• B13. Agency: how often does the lead act rather than react? AI
• B14. Physical description: how much, when, which traits repeat. Both
• B15. Names, nicknames and titles, and how their use shifts. Count
• B16. Mentors and sidekicks: their role and how they exit. AI
• B17. Repeated tags, gestures and tics per character. Both
• B18. Moral choices that define the lead. AI
• B19. Balance of speaking parts by gender. Count
• B20. Who changes and who stays fixed? AI

Dialogue and voice

• C1. Share of words in dialogue, whole book and per chapter. Count
• C2. Each character's number of lines and share of speech. Both
• C3. Speech fingerprint: line length, pet words, questions, exclamations. Count
• C4. Dialogue tags: "said" against other verbs, and adverbs on tags. Count
• C5. Action beats used in place of tags. Both
• C6. Subtext: where people say one thing and mean another. AI
• C7. Dialect and class markers, and how they are spelt. Both
• C8. How exposition is carried in talk. AI
• C9. Longest speeches and what each one does. Both
• C10. Who wins each argument, and how power shifts in talk. AI
• C11. Banter and humour patterns. AI
• C12. Blind test: can you tell who speaks without the tag? AI
• C13. Catchphrases and verbal tics. Count
• C14. Interruptions, silences and broken off lines. Count
• C15. Voice in letters, diaries and documents. Both
• C16. Forms of address between characters and how they change. Both

Pace

• D1. Chapter length in words, as a chart. Count
• D2. Mean sentence length per chapter. Count
• D3. Paragraph length per chapter. Count
• D4. Scene against summary: how much is shown live? AI
• D5. Action against reflection, per chapter. AI
• D6. Tension score per chapter (1 to 5) as a curve. AI
• D7. Emotional arc of the whole book, matched to the six shapes. Count
• D8. Chapter endings: cliffhanger, question, reveal or quiet close. AI
• D9. Chapter openings: in scene, in summary, or after a time jump. AI
• D10. Do chapters and sentences get shorter near the climax? Count
• D11. Dialogue density against pace. Count
• D12. Time jumps and how they are signalled. Both
• D13. Info dump spots: long runs with no action or speech. Both
• D14. Reading time per chapter. Count
• D15. Micro tension in quiet scenes. AI
• D16. Section breaks per chapter. Count

Viewpoint and narration

• E1. Person and tense: first or third, past or present. Count
• E2. Whose viewpoint is each chapter? Both
• E3. Share of the book told by each viewpoint character. Both
• E4. Narrative distance: close inside a head, or far off? AI
• E5. Strict viewpoint or head hopping? AI
• E6. Is the narrator reliable? Where are the cracks? AI
• E7. Frames and tales within tales. AI
• E8. Narrator asides and addresses to the reader. Both
• E9. Documents used: letters, diaries, reports, by type and sender. Both
• E10. What the narrator hides, and where the reader knows more than the lead. AI
• E11. Free indirect style: where the narrator slips into a character's voice. AI
• E12. How thought is shown: tagged, italic or blended. Count
• E13. What triggers each change of viewpoint? AI
• E14. Hindsight: where an older narrator comments on events. AI

Openings and endings

• F1. First line and first paragraph: what each does. Both
• F2. What the first page sets up: voice, world, question. AI
• F3. How many words before the lead appears? Count
• F4. How many words before the first line of dialogue? Count
• F5. The hook of chapter one. AI
• F6. Prologue: its job and its length. Both
• F7. How fast is the main story question posed? AI
• F8. Final chapter: which beats resolve what. AI
• F9. Last line: what it leaves in the mind. Both
• F10. Loose ends tied and left open. AI
• F11. Epilogue: its job. Both
• F12. Echoes between the opening and the ending. AI
• F13. Length of the wind down after the climax. Count

Genre craft

• G1. Genre and subgenre, and how well the book fits. AI
• G2. The genre's must have scenes: present or missing? AI
• G3. The genre's conventions: which are used? AI
• G4. Mystery: clue list, red herrings, fair play test. AI
• G5. Thriller: clocks, chases and set pieces. AI
• G6. Romance: the meeting, the barriers, the grand gesture. AI
• G7. Horror and gothic: how dread builds, when the monster is seen. AI
• G8. Science fiction: the new idea and its rules. AI
• G9. Fantasy: moments of wonder, and the cost of magic. AI
• G10. Adventure: stages of the journey and physical trials. AI
• G11. Literary: inner life and the moment of insight. AI
• G12. Children's: age of the lead, level of peril, humour. AI
• G13. Plays: entrances, exits and stage time per character. Count
• G14. Where the book bends or breaks its genre's rules. AI
• G15. Tropes present, with where they appear. AI

Prose style

• H1. Readability score per chapter. Count
• H2. Vocabulary range. Count
• H3. Signature words: what this author uses far more than others. Count
• H4. Adverb and adjective rates. Count
• H5. Share of weak verbs such as "was" and "had". Count
• H6. Passive voice rate. Count
• H7. Variety of sentence openings. Count
• H8. Rhythm: how much sentence length varies. Count
• H9. Similes and metaphors: how many, and of what kind. Both
• H10. Senses used: sight, sound, smell, touch, taste. Both
• H11. Repetition, lists of three and other patterns. Both
• H12. Old words that may trip modern readers. Count
• H13. Showing against telling, with examples. AI
• H14. How much description comes before action. Both
• H15. Filter words such as "saw", "felt", "heard". Count
• H16. Shifts in tone. AI
• H17. How the comedy works. AI

Setting and world

• I1. Places named and how often each appears. Count
• I2. Setting by chapter, as a map. Both
• I3. Setting shown in blocks or woven into action? AI
• I4. Weather and light used for mood. Both
• I5. The social world: class, money, rules. AI
• I6. Rules of the world (magic, science) and when each is shown. AI
• I7. Lore: dripped in or dumped? AI
• I8. Real history used, and how. AI
• I9. Journeys: distance and time taken. Both
• I10. Objects that matter, where they appear. Both
• I11. Density of period detail. AI
• I12. Setting as an enemy. AI
• I13. Indoors against outdoors. Both

Theme and motif

• J1. Where is the theme first stated, and by whom? AI
• J2. The theme question, and how each subplot answers it. AI
• J3. Motif tracker: recurring images by chapter. Both
• J4. Symbols, and how their meaning shifts. AI
• J5. Theme words and their frequency over the book. Count
• J6. The moral argument: who stands for which side? AI
• J7. The title: its meaning and echoes in the text. Both
• J8. Epigraphs, allusions and borrowed texts. Both
• J9. Opening image against final image. AI
• J10. Pairs and contrasts, such as light and dark. Both
• J11. Core values at stake: life, love, justice, truth. AI
• J12. What the book leaves open on purpose. AI

Reader effect

• K1. Where does suspense peak? AI
• K2. Open questions: how many are live at the end of each chapter? AI
• K3. Surprise against suspense: which does the author use where? AI
• K4. Emotional peaks, with the lines that cause them. AI
• K5. Why do we care about the lead so early? AI
• K6. Humour beats. AI
• K7. Spots where readers may get lost. AI
• K8. What modern readers may find dated or hurtful. AI
• K9. What rewards a second read. AI
• K10. Most quoted lines. Research
• K11. Page turner score: the mix of hooks, short chapters and open questions. Both

Commercial facts

• L1. Total word count. Count
• L2. Chapter count and average length. Count
• L3. Length against modern norms for the genre. Both
• L4. First published as a serial or a volume? Research
• L5. Serial instalment breaks and how each ends. Both
• L6. Sales and reviews in its day. Research
• L7. Adaptations: film, TV, stage, count and type. Research
• L8. Sequels and series. Research
• L9. Modern books in the same mould (comps). AI
• L10. Who it would be shelved for today. AI
• L11. Content notes. AI
• L12. A modern pitch line and blurb. AI
• L13. Title style against genre norms. AI

Books in copyright

For these we give our own notes on structure only. No text, no quotes, and every figure is cited or marked as an estimate.

• The Da Vinci Code, Dan Brown (our notes on structure only)
• The Girl with the Dragon Tattoo, Stieg Larsson (our notes on structure only)
• Harry Potter and the Philosopher's Stone, J.K. Rowling (our notes on structure only)

Trails

• The ticking clock: A deadline turns a hunt into a race. The tighter the clock, the shorter the chapters tend to feel.
• The puzzle as plot engine: The plot moves one clue at a time. Each solved clue opens the next door and raises the stakes.
• The outsider sleuth: The detective comes from outside the closed world. They see what insiders miss, and insiders resent them.
• The hidden heir and chosen orphan: A young lead with lost parents learns they are part of a bigger story. The reveal gives the plot a personal stake.
• Many voices, one case: The story is built from several viewpoints or documents. The reader puts the truth together.

Also

Compare books (free beta): pick 2 to 5 books and get shared patterns and lessons to borrow.

Method

Every classic on the shelf passes two tests: the author (and any translator) died in 1955 or earlier, so it is free in the UK, and it was first published in 1930 or earlier, so it is free in the US. We use clean public domain plain text with the publisher's header and licence removed. We split chapters and paragraphs and give each paragraph a stable ID. Count lenses are exact for our edition. AI lenses are our reading, made once with a small AI model and stored with a version, and we drop any quote that does not match the text word for word. Across the shelf so far 7045 quotes passed and 2236 were dropped in the chapter pass. In the full lens pass the AI could only cite lines from that checked bank: 17678 citations passed a second check and 2 were dropped.