7.0 How to use the protocol
The ten tests below are for revision, not composition. Applied during drafting they induce paralysis; applied to a finished draft they are the most efficient triage instrument in this chapter.
Three procedural notes:
- Run them in order. Tests 1–3 are structural and their failures invalidate the others: there is no point tuning voice in an opening that has no scene in it.
- Record the answer, not the verdict. The output of a test is a sentence describing what the opening does, not a pass or fail.
- A failed test is a question. Every one of these tests is failed by at least one novel discussed in this chapter. The test’s job is to make a choice visible so that it can be confirmed as a choice.
7.1 The ten tests
Test 1 — The contract test
Can I point to the exact words where each contract — voice, genre, character, world, stakes, form — is opened, within the first two pages?
Perform the annotation described in §5.5. For any contract with no marking, ask the follow-up: deliberate deferral or oversight? Austen defers character in Pride and Prejudice; that is a choice underwritten by an overwhelming surplus in voice. A draft that defers character and has no distinctive voice has not made a choice; it has made an omission.
Test 2 — The scene test
Complete the sentence: “In these pages, ___ is trying to ___, and ___ is in the way.”
If the sentence cannot be completed, the opening is not yet a scene, whatever else it may be. This is a test the five-contract model alone can pass while still failing, because contracts can all be technically opened by a passage in which nothing happens. Chapter 1 (The Purposes of a Scene) covers the remedy in full.
Test 3 — The momentum test
Within the first page: does something change, or does someone want something?
Note the disjunction. Change or desire suffices; both is better; neither is a static opening. A useful refinement is to ask when the first verb of intention or alteration appears, measured in words from the start. In the published openings this chapter discusses the answer is usually under fifty, which is an observation from its examples, not a count.
Test 4 — The voice test (the substitution test)
Could this first paragraph be moved into a different novel, of a different genre, by a different author, without anyone noticing?
The strongest form of the test is practical: retype the draft’s opening paragraph alongside three published openings with the attributions removed, and ask a reader to identify which is which and describe each voice in two adjectives. A draft that generates the adjectives clear and fine has failed. Distinctiveness, not competence, is what is being measured.
Test 5 — The zero-context test
Read the opening as though you know nothing about the book — no title, no cover, no jacket copy, no author.
Per §3.7, this is a deliberately austere idealisation. Its practical form: give the first two pages to a reader who knows nothing, and ask them to answer, in one sentence each — Who is this about? Where and when? What kind of book is this? What do you want to know? Any question that cannot be answered marks a dependency on paratext.
Test 6 — The promise test (the second-page test)
Does what follows deliver on what the opening promised, in tone, pace and stakes?
Practically: read the opening, then read pages 30–40, and ask whether they belong to the same book. This is the only test in the protocol that cannot be run on the opening alone, and it catches the most expensive single error in the repertoire (§8.4).
Test 7 — The promise-inventory test
List every detail given a privileged position in the first two pages. Which are redeemed later?
Per §3.2, everything named early reads as significant. An inventory of ten first-page details of which two recur is not richness; it is eight unredeemed promises. The remedy is usually deletion rather than the addition of eight subplots.
Test 8 — The density test
Does any paragraph exceed the density ceiling (§5.4)?
Read the first page aloud. Where the reader’s breath fails, or where a listener asks wait — what’s happening?, the ceiling has been crossed. Reading aloud is not an ornament to this test; it is the test.
Test 9 — The formal-declaration test
Does the opening declare the book’s structural rules — person, tense, viewpoint plurality, apparatus?
If the novel uses multiple viewpoints, footnotes, documents, dual timelines or a shifting tense, ask on what page the reader first learns this. If the answer is later than the end of the first chapter, expect resistance.
Test 10 — The abandonment test
Where would an indifferent reader stop?
Read the opening as an unsympathetic stranger and mark the exact sentence at which attention would lapse. This test is unpleasant and unusually informative: it converts the vague sense that an opening “drags” into a coordinate. In this chapter’s estimate, from drafts rather than a count, the mark falls more often at the third paragraph than at the first sentence, where many drafts stop performing and begin explaining.
7.2 Reading the results
| Failed tests | Likely diagnosis | First move |
|---|---|---|
| 2 and 3 | The opening is description or exposition, not scene | Find the action underneath and start there |
| 4 alone | Competent, undistinguished prose | Rewrite in the character’s diction, not the author’s |
| 1 and 7 | Scattered focus; too many promises, too few kept | Cut named entities; concentrate detail |
| 5 alone | Paratext dependency | Supply the missing frame inside the prose |
| 6 alone | Mis-promise — the most expensive failure | Change the opening to match the book, not the reverse |
| 8 alone | Over-revision; the page has been polished into airlessness | Restore some single-purpose sentences |
| 9 alone | Structure arrives as a surprise | Declare the pattern in the first chapter |
A warning about test 6. When an opening and a book disagree, writers instinctively revise the book toward the opening, because the opening is the more finished object and often the more exciting one. This is almost always the wrong direction. The opening is two pages; the book is four hundred. Revise the smaller object.