Craft advice says that a transition must orient the reader quickly, and does not say why, how quickly, or what orienting consists of. Five bodies of work answer, each opening here with the question it answers at the desk.
How large is this gap, and does the reader know?
The frame is Gérard Genette’s treatment of duration in Narrative Discourse (1972; translated by Jane E. Lewin, 1980): the relation between the time an event takes in the story and the space it takes in the telling. The Purposes of a Scene (Chapter 1) sets out his four movements — scene, summary, pause and ellipsis — as a decision about the speed of a unit. This chapter works at the seam between speeds, and needs chiefly the fourth. An ellipsis is story time that passes while the telling spends nothing on it. A transition, in these terms, is an ellipsis whose edges have been finished.
Genette distinguishes an explicit ellipsis, which declares the lapse (three days later), from an implicit ellipsis, which the reader infers from evidence (new light, new clothes, a healed cut), and from a hypothetical ellipsis, which cannot be located and is discovered only later, when something the narrative skipped comes to light. The explicit kind is cheap, and its risk is monotony. The implicit kind costs nothing at the join but borrows against the new unit’s first sentences, which must now carry evidence of elapsed time as well as place and person. The hypothetical kind, the material of Ford Madox Ford’s The Good Soldier, makes the narration seem not in control of its chronology when used on purpose, and the author when used by accident. An ellipsis is also definite when its length is stated, and qualified when the text says what filled it (three days later, after the funeral); most workshop advice about joins asks for a definite interval or warns against too much qualification.
Genette’s category of frequency adds a bridging tool most drafts lack. The singulative narrates once what happened once; the iterative narrates once what happened many times (Every Thursday he counted the till twice). An iterative sentence crosses an indefinite stretch of time while characterising, because a habit is a fact about a person as well as a calendar, and a writer who uses only the time-stamp has one bridging tool where she could have two. One more term is held in reserve: paralipsis, Genette’s word for the omission of information the narration’s own conventions should have supplied. A gap that skips the event the book is about, while presenting itself as an ordinary interval, is a paralipsis disguised as an ellipsis, and the failure patterns call it the false ellipsis.
What is the reader doing in the first words after the white space?
Catherine Emmott’s Narrative Comprehension (1997) proposes that readers keep contextual frames, mental representations of the current episode holding who is present, where and when. Characters are bound into a frame, and the frame in use is primed. Three of her operations describe a transition. A frame switch abandons the primed frame and builds another; frame recall retrieves a frame built earlier, as when a novel returns to a thread; frame repair corrects a frame built on a wrong assumption.
The consequences are practical. A frame is not cleared by white space; it stays primed until a cue dislodges it, so a reader crossing an unanchored gap does not experience a blank but reads the new sentences into the old scene. The cost of a bad transition is therefore repair, not confusion: the reader has gone confidently the wrong way and must walk back. Recall is much cheaper than switching, so one name returns a reader to a thread she knows and a paragraph of reminder is wasted on her. And a place with a person bound into it primes a scene, so Elena let herself into the shop anchors more strongly than The shop was cold.
Why does one gap feel heavier than another?
Discourse psychology measures the load. The event-indexing model of Rolf Zwaan, Mark Langston and Arthur Graesser (1995), with Zwaan and Gabriel Radvansky’s survey of 1998, holds that readers track five dimensions of a narrative — time, space, protagonist, causality and intentionality (whose goal is in play) — and slow down measurably at a sentence that breaks continuity on any of them. Zwaan’s experiments on time (1996) found the effect graded: an hour later costs more to process than a moment later. The finding this chapter leans on is that the costs add up across dimensions. A cut to another character in another building pursuing another goal changes four dimensions and costs about what four separate shifts would; a jump of five years that keeps one woman in one house changes one. The dimension count — the number of the five dimensions a gap changes — is therefore a better predictor of how much orientation a gap needs than the length of time it skips. Causality and intentionality, the dimensions craft advice forgets, explain why a join can settle when and where and still read as a non sequitur.
The experiments use short constructed texts in laboratories and describe tendencies in processing, not literary value. And the cost buys something. Work on event segmentation by Jeffrey Zacks and Barbara Tversky (2001), and by Nicole Speer, Zacks and Jeremy Reynolds (2007), suggests that readers parse experience into events at these discontinuities and consolidate memory there; a join is where the previous scene gets filed.
Where can I put exposition without stalling the story?
Meir Sternberg’s Expositional Modes and Temporal Ordering in Fiction (1978) treats exposition as a problem of distribution: every story must convey its antecedents, and the only variables are when and in what order. Just after a break a reader accepts summary without resistance, because she has been told that something was skipped and is looking for an account of what happened; the paragraph that stalls a scene in its middle reads, after a break, as the transition doing its job. The gap is the natural home of distributed exposition. Sternberg’s distinction between suspense (uncertainty about what will happen) and curiosity (uncertainty about what has happened) adds a rule for intervals: make an interval mysterious only if the reader’s curiosity about it is something the book means to spend later. If nothing will be paid out, say how long it was.
Is “cut to” a good way to think?
Most writers picture transitions in film terms. David Bordwell and Kristin Thompson’s Film Art describes continuity editing, the conventions that make classical cuts unobtrusive, whose establishing shot, match on action and cross-cutting have prose relatives. Sergei Eisenstein’s montage supplies the other half, the cut as collision, demonstrated by the Kuleshov effect, in which one face is read as hunger, grief or desire according to the shot beside it; that is the parent of the match-cut. The analogy breaks in three places. A film cut delivers a whole frame — place, hour, weather, faces, clothes — in a twenty-fourth of a second, while prose delivers it one phrase at a time in the order the writer chose. Film orientation is free, since the rain is in the shot whether or not it matters, while every clause of prose orientation is paid for and read as a claim that the thing named matters; prose has no free background, which is why the establishing paragraph is a promise, not neutral scene-setting. And a viewer cannot go back where a reader can and does, so an unclear prose join produces the backtrack. Cut to is good advice about sequence and bad advice about cost: a writer thinking cinematically will under-budget the arrival.
The dates of the convention
The unbridged cut is young. The eighteenth- and nineteenth-century novel spoke its gaps: Henry Fielding opens each book of Tom Jones with an essay, and Dickens’s narrator in Bleak House turns aside from Chancery, in its second chapter, for a glimpse of the world of fashion on the same miry afternoon. The blank line standing alone as a narrative signal became ordinary in the twentieth century, and the unacknowledged cut the Anglo-American default after modernism; the later files describe that convention, not a law. The orientation budget — the principle that every gap imposes a cost varying with the dimensions it changes, and that the right expenditure is the least that crosses this gap — is therefore offered as a revision instrument. It assumes that orientation is a cost and attention scarce, and the coda names the novels that refuse the assumption.
See also: the Reference theory, with tables of Genette’s types, Emmott’s operations and the five dimensions, and the history in three phases; The Opening Pages (Chapter 9), on Sternberg and on the reader’s rules of notice.