The theory below is limited to what the chapter’s methods use, and each strand starts from a question a writer asks at the desk.
Why does a detail on page one weigh more than on page 150?
Because readers treat certain positions as meaningful. Peter J. Rabinowitz, in Before Reading (1987), describes the tacit rules of notice by which readers decide which parts of a text matter. Among the strongest are the rules of privileged position: titles, epigraphs, first and last sentences, and the beginnings and ends of chapters are read as load-bearing, and whatever is placed there is assumed to be there for a reason. A writer who mentions the weather on page one has told the reader, whether she meant to or not, that weather matters in this book. Detail that would pass unnoticed on page 150 becomes on page one a promise, and unredeemed promises accumulate into a sense that the book does not know what it is about. The discipline is not fewer details but details the writer is willing to be held to, and names are the costliest case, since a character named on page one is promised to matter. The positive use is placement doing work that description would otherwise have to do. In the first chapter of Frank Herbert’s Dune (1965) the Reverend Mother holds a poisoned needle, the gom jabbar, at the boy Paul’s neck as part of a test; at the novel’s climax his sister Alia kills Baron Harkonnen with it, and the weapon needs no introduction, because the opening gave it its weight.
Early material also governs how later material is read. In Solomon Asch’s experiments on impression formation (1946), the same list of traits produced a more favourable judgement when it began with intelligent than when it began with envious, and Menakhem Perry’s “Literary Dynamics” (1979) carried the finding into reading as the primacy effect: the reader builds a hypothesis from the first material and fits later material to it, resisting revision until the contrary evidence is overwhelming. For the writer this cuts both ways. An impression made by accident is expensive to correct: if page two makes the protagonist read as passive, fifty pages of agency will read as inconsistency. An impression made on purpose is a resource. Jane Austen’s Emma (1815) calls its heroine handsome, clever and rich, and says that she seemed to unite some of the best blessings of existence; most readers adopt the flattering frame, and the one verb lets the later correction be shown to have been signalled.
What does the reader need before she can care?
Enough to build a picture of what is happening. Research on how readers understand stories holds that they do not store sentences but construct a situation model, a mental representation of the state of affairs a text describes. Rolf Zwaan and Gabriel Radvansky (1998) propose that readers track five dimensions of it, the protagonist, space, time, causation and intention, and that each shift in one of them costs effort to process. Catherine Emmott’s Narrative Comprehension (1997) adds the contextual frame, the reader’s running record of who is present in a scene and where, and argues that texts which do not bind characters to frames early leave readers disoriented. The five dimensions correspond closely to the obligations set out in The Taxonomy: whose experience this is (character), where (world), when and how fast (form and world), what is making things happen (stakes) and what someone wants (character and stakes). The obligations can be read as a craft description of what a reader must have before a situation model exists at all. The practical inference is that an opening may withhold some dimensions to make a mystery, but must supply enough that the withheld ones are felt as absences. A named character in an unnamed place is intriguing; an unnamed pronoun doing an unspecified thing in an unspecified place is noise.
Does the reader arrive with nothing?
No, and the qualification governs how the chapter’s tests are used. Hans Robert Jauss’s horizon of expectations (Toward an Aesthetic of Reception, translated by Timothy Bahti, 1982) is the set of generic, formal and thematic expectations a reader brings from earlier reading, and the opening pages are where a text confirms, adjusts or violates it. That is why genre can be signalled cheaply: the reader supplies the schema, and one well-chosen noun can orient someone who has read forty books of the kind. Genette’s Paratexts (translated by Jane E. Lewin, 1997) describes the rest of what the reader brings: title, cover, jacket copy, shelf, reviews. Yet the paratext is not the author’s to control, nor permanent, nor always present. Covers change between editions, jacket copy is written by marketing departments, and openings are excerpted, photocopied and read in workshops without any of it. So the test that asks the writer to read her opening as if she knew nothing about the book (test 5: Zero context in Diagnostic Tests) survives as a deliberately austere idealisation. It does not describe how books are read. It isolates what the prose does on its own, and an opening that fails it has a dependency that the writer should choose knowingly.
Where should the backstory go?
Wherever it buys most. Meir Sternberg’s Expositional Modes and Temporal Ordering in Fiction (1978) treats exposition not as a place in a novel but as a problem of distribution: every narrative must convey its antecedents, and the variables are when and in what order. Exposition can be preliminary, delivered before the action, which gives clarity at the cost of suspense; delayed, supplied after the action has begun; or distributed, scattered and never gathered. Sternberg’s distinction between suspense, uncertainty about what will happen, and curiosity, uncertainty about what has already happened, explains why delay is productive: the reader who does not yet know why Josef K. was arrested, or what happened at 124, is pulled forward by a question about the past. Front-loaded backstory fails not because backstory is dull but because delivering it early spends an asset. Backstory and Flashback (Chapter 7) treats the cost of each later return to the past.
The dates of the convention
The advice in this chapter describes current conventions, not laws. Horace praised Homer for plunging his audience in medias res, into the middle of things, instead of beginning with the egg from which Helen was hatched, but Homer’s audience already knew the Trojan War, and Homer recovered the antecedents later by retrospection; the rule was about sequence, not omission. Nineteenth-century novels such as Père Goriot (1835) and Middlemarch (1871–72) favoured preliminary exposition and a narrator who surveys a society before entering a scene. Modernist novels plunged into a consciousness and distributed orientation. Contemporary openings attack faster and show their voice earlier, under institutional pressures that Noah Lukeman’s The First Five Pages (2000) describes from inside publishing, and his book is best read as an account of a filtering institution, not of an aesthetic principle. Some of the openings that have lasted longest in English would fail a modern first-page screen, and Bleak House is admired partly because of its opening of fog, not in spite of it.
See also: the Reference theory, with transportation and absorption, Suvin’s novum and Delany’s account of the science-fictional sentence; the history of opening conventions.