Everything in this chapter is an instrument for noticing. None of it is a rule, and the evidence for that claim is in the chapter itself.
Don Quixote proceeds by free goals. The Count of Monte Cristo runs on clean wins for hundreds of pages. The Remains of the Day is a stalled sequel four hundred pages long, built on false dilemmas, narrated by a man who never forms a goal of his own. Catch-22 breaks the chain deliberately, and the breaking is the argument. The Stranger refuses the decision beat. The Handmaid’s Tale cannot supply it. Crime and Punishment commits, on a mechanical reading, the gravest fault the model recognises — a sequel that swallows five-sixths of the book — and is among the two or three most powerful novels ever written. A model that rules these out has not identified defects in them. It has identified its own boundaries, which is the most useful thing a model can do.
Three points follow from that list.
First, the six beats describe one way of building momentum, not the thing itself. They are dominant in commercial fiction, common in literary fiction, and absent from a great deal of admired writing — the episodic, the meditative, the polyphonic, the encyclopaedic, the novel of a single day. Momentum can also be produced by curiosity about the past, by formal pattern, by rhythm, by image, by the accumulating pressure of a voice. The chain is the cheapest route and the most teachable. It is not the only one, and a writer who believes it is will produce books that move well and mean little.
Second, the model’s provenance is not a scandal but a specification. It was made in 1965, in Oklahoma, out of a market that paid by the word and required legible forward motion. Its virtues — agency, escalation, external goals, decisions that foreclose — are that market’s virtues, and they encode a set of assumptions about what a person can do to the world. Those assumptions are not universal, and §8.7 concedes the sharpest case. Knowing where an instrument was made tells you what it will measure well and what it will measure badly.
Third, and most usefully: the model’s real value in revision is diagnostic rather than corrective. It does not tell a writer what to do. It tells them what they have made. A draft that fails the disaster test in every chapter is not a broken thriller; it may be a novel of consciousness that has not yet admitted what it is. A draft in which no criterion ever changes is not necessarily deficient; it may be a satire, in which the refusal of change is the point, and which needs to be rewritten towards that rather than away from it. The most common productive outcome of running §7’s protocol is not a repaired chain. It is the writer discovering, sometimes with relief and sometimes with dismay, what kind of book they have been writing.
Which suggests a final question, to be asked after the ten tests and in preference to any of them:
If this book does not run on the chain, what does it run on — and is that engine actually running?
A book that can answer needs nothing else in this chapter. A book that cannot has found the real problem, and it was never a problem of beats.