Appendix E — Short Forms

The chapters are written for the novel, and most of their instruments assume its length: room to plant and pay, to establish what is ordinary before departing from it, to let a thread go dormant and bring it back. This appendix says what the instruments become in the two shorter forms of prose fiction, the short story, a narrative of roughly two to eight thousand words that a reader finishes at one sitting, and the novella, a narrative of roughly twenty to forty thousand words, long enough for a novel’s range of feeling and too short for its machinery. The twelve chapters revised from the first edition each treat the short forms in their Reference text, in a section on the unit across forms, and the appendix draws its account of those chapters from there, with a link to each. The six chapters new to this edition do not treat the short forms, and what the appendix says of them is its own application of their instruments, marked as such.

Two forces

Two forces account for nearly everything the chapters say about the short forms, and a writer who keeps them in mind can work out the rest.

The first is proportion. Most units of a novel carry a fixed cost and a benefit that grows with the length of the book that follows. A prologue’s doubled start, a transition’s forty words, a cutaway’s suspended question and a viewpoint’s cold start cost about the same in any book, but they are repaid over the pages after them, and a short story has few pages after anything. A forty-word bridge is a fraction of a per cent of a novel of ninety thousand words and one per cent of a story of four thousand, and a story may have six of them. The arithmetic is the reason the chapters keep finding that a unit a novel can afford is unaffordable in a story.

The second is the absent baseline. A novel teaches its reader what is ordinary in its world, how a character usually speaks, what a room usually holds, before it departs from the ordinary, so that a deviation registers. A story has no time to build that reserve. Every detail is read as significant from the first line, every exchange must carry its subtext without a history of the speaker’s habits behind it, and the reader leans on the conventions of the form, which a story’s writer must know she is borrowing.

The novella meets a third problem, which the chapters state in nearly the same words each time: it is usually written by novelists. It has a novel’s range of feeling and a fraction of its length, so a novelist’s habits, a three-page approach to a scene, a five-thousand-word chapter, three viewpoints, a six-scene subplot, consume it. Its characteristic solution is concentration: one place described fully and everything else bare, one transition system held throughout, one viewpoint or a frame, a subplot two scenes long. The Death of Ivan Ilyich (1886), which opens on the announcement of its hero’s death and then retraces his life, is the chapters’ model case, and it recurs because its design solves several of the form’s problems at once.

Part I: the scene and its materials

The scene (The Purposes of a Scene, Chapter 1). In a story a single scene may be a third of the whole, so there is no room for scenes that establish the normal; the normal is established inside the scene that breaks it, usually by iterative clauses (every Sunday he would…) set in the scenic narration. The novella’s solution is the compressed scene, rendered dramatically but with heavy internal ellipsis, so that the reader has the texture of a full scene at half its cost.

Scene and sequel (Scene and Sequel, Chapter 2). A story usually holds one scene and sequel pair, and often only its second half: the disaster happened before the text begins, and the story is the sequel. The decision beat carries a disproportionate weight, since the form’s characteristic ending lives there, a decision taken and made irreversible or refused and left visible. The novella’s characteristic shape is one disaster and a long sequel; its characteristic failure is a novel’s chain squeezed into a novella’s length, pairs that are all scene and no absorption.

The dialogue exchange (The Dialogue Exchange, Chapter 3). A story has no baseline of a character’s habits, so its subtext must be legible from the first exchange, and it gets there by convention: the reader of a literary story already knows that a conversation about hills is not about hills. Two speakers are the norm, phatic and ritual talk cannot be afforded, and the closing of the gap between what is said and what is meant, which in a novel is a climax, is often the story’s ending. The novella wants exchanges that begin at their second or third turn.

The descriptive detail (The Descriptive Detail, Chapter 4). With no baseline every detail reads as chosen, so a story tolerates less unmotivated detail than a novel, not more, and a motif must make its three appearances in a few pages, with sharper variation between them. The form’s classic way of letting in air is a single concrete object given no explanation. A story’s title often supplies the world detail a novel spends a paragraph on. The novella’s natural solution is one dominating place or object described more fully than a novel would allow, which is why so many novellas are remembered by a place.

Part II: the joins

The transition (The Transition, Chapter 5). A story has few joins, and each is in a privileged position, read as meaningful. It cannot establish a system of transitions, and the reader has not lived in any frame long enough to recall it cheaply, so every switch is a construction. The hard cut is the story’s default and the summary bridge its greatest risk, and the white space between a story’s sections is read as structural, closer to a stanza break than to a novel’s scene break. The novella can do what neither other form can: hold a single transition technique across its length, so that what would be a metronome at ninety thousand words becomes a form at thirty.

The chapter break (The Chapter Break, Chapter 6). A story is read at one sitting, so its section breaks offer no permission to stop and are priced by nothing but what happens across them; The Transition (Chapter 5), not The Chapter Break (Chapter 6), governs them. Numbered sections announce a formal seriousness. The novella works with many short chapters, which give it the pulse of a long poem, or with none, a single continuous movement; what fails is the middle course, five or six chapters of novel length, which make a short book feel like an abbreviated long one.

Backstory and flashback (Backstory and Flashback, Chapter 7). A story cannot amortise the cost of a cutaway over the fifty pages after it, so the return must come at once, and a dramatised flashback of eight hundred words in a story of four thousand gives a fifth of the whole to the past. The registers that suit the form deliver the past without stopping the story: a clause of implication, a participial aside, the iterative, and the corrected assumption. The last is the form’s advantage, since a reader holds a whole story in mind and a final revelation about the past recolours every sentence before it. The novella can make the retrospect its whole design rather than an interruption, as Tolstoy does, because the reader’s investment in the present has not yet become something to protect.

Non-linear and multi-timeline architecture (Non-linear and Multi-timeline Architecture, Chapter 8; the appendix’s application). The chapter’s instruments shrink rather than vanish. A story told in two times usually has no room for strands with questions of their own, so its second time is more often a cutaway, judged by Backstory and Flashback (Chapter 7), and its order chart has a handful of rows; what remains essential is orientation, which in a story must work within the first clause of each section. The novella can carry two strands, and its brevity makes a short interval between them easy to hold, which is one reason the frame, a present telling of a past, suits it.

Part III: the architecture of the whole

The opening pages (The Opening Pages, Chapter 9). A story may open at its central situation rather than approach it, and its opening is also its middle, so contract work and plot work must happen in the same sentences, with nothing deferred. Its length is itself a formal declaration, and its title is often a working part of the opening. The novella can open on its ending, as The Death of Ivan Ilyich opens on the death and the colleagues’ reaction, declaring its moral apparatus in the first pages with no approach at all.

The prologue (The Prologue, Chapter 10). A story has, in the main, no room for a prologue: the costs are fixed and the benefits have nowhere to accrue. Where a story uses a separated preliminary unit, it nearly always serves the authority domain, a found letter, a frame narrator, an editorial note, because that payoff is immediate, and its title often does a prologue’s work in a few words. The novella should compress the schedule on which a prologue’s distance is closed, with a first fragment in the first tenth rather than the first quarter and two fragments rather than four or five; or it can open a frame and never return to it, as The Turn of the Screw does, because the interval is short enough for a reader to hold an unclosed frame without strain.

The middle (The Middle, Chapter 11; the appendix’s application). A story has no middle in the novel’s sense, no long stretch between commitment and ending in which options can be narrowed by stages, so the option list collapses to a single narrowing, usually the story’s one turn. The plateau, the chapter’s first failure, is the story’s characteristic failure in miniature: a second scene that restates the first at the same stakes. A novella’s middle is short enough to hold one turn at its centre and not two, and its clock is usually shorter than a novel’s, a day or a season rather than a year.

Arc and theme (Arc and Theme, Chapter 12; the appendix’s application). A story seldom has room for an arc located by a run of decisions, so it more often shows the single decision at which a character’s rule for choosing changes, or the moment at which she refuses to change it; Scene and Sequel’s finding (Chapter 2) that the decision beat carries the story’s ending is the same observation from the other side. Theme in a story is carried by fewer answers to its question, often one character’s and the world’s, which makes the sermon, every answer pointing the same way, easier to fall into and easier to see.

The viewpoint shift (The Viewpoint Shift, Chapter 13). The story’s default is a single fixed viewpoint, and the default is close to a requirement: a second vantage consumes an investment the story cannot rebuild, two shifts read as instability rather than pattern, and the story is very often about what one person cannot see. The exceptions are the story told from two vantages in sequence, the second reframing the first, and the story that widens at its end, as “The Dead” moves from Gabriel Conroy’s mind to the snow falling over Ireland. In a story every character but one is minor, so allocation is nearly all character compression. The novella works with one fixed viewpoint or with a frame, which gives it two positions and a listener at the cost of one additional vantage, and which is why so many novellas are frames.

The subplot (The Subplot, Chapter 14). The short story is the form in which a subplot is nearly always wrong: at fifteen per cent of a story it is too little for a chain and too much for a motif, it cannot go dormant without breaking the story, and a second chain implies a second value shift competing for the ending. Stories use a compressed analogy instead, a second situation glimpsed rather than narrated, or a story told inside the story. The linked collection is the exception, each story a thread and the book the braid, with relations of analogy and contrast rather than cause. In the novella the rule is to prefer a second chain two scenes long, one early and one late, over one six scenes long; the reader builds the middle of it at no cost.

The reveal (The Reveal, Chapter 15; the appendix’s application). The short forms favour the closing reveal, because a reader holds the whole text in mind and a late disclosure can recolour every sentence before it, which is the corrected assumption of Backstory and Flashback (Chapter 7) at the scale of a whole story. The double-reading table is short and its rows close together, so a giveaway is easier to commit and a cheat easier to find. The early reveal is rare in a story, which has no second half for a successor question to drive.

The ending (The Ending, Chapter 16). The story’s ending has fifteen pages of deposit to spend rather than four hundred, which is the worst ratio of context to closural demand in prose. Almost no aftermath is available, plot resolution is often waived for the terminal image or incompletion, the title is often part of the ending, and the last sentence may be the only closural device. The epiphanic close, a recognition rendered as an image in the last paragraph, has been the default since Dubliners, and a writer should choose it against alternatives rather than inherit it. The novella is the natural home of the arrested ending and of death as a structural terminus, both of which discharge the most pressure in the least space; its aftermath should be about half the proportion of a novel’s.

Part IV: beyond the book

Series and genre (Series and Genre, Chapter 17; the appendix’s application). A story keeps or declines its genre’s core promise as a novel does, and its brevity makes a decline harder to declare, since there is less room for the early sign. The series has its short-form counterpart in the linked collection and the story cycle, in which recurring characters or a place carry the reader’s promises from story to story; each story should close its own question, as each volume of a series should, and leave open only what the collection has shown to belong to the whole.

The structural edit (The Structural Edit, Chapter 18; the appendix’s application). The ladder holds, but its middle rungs come close together. A story’s book map may have a dozen rows, its allocation is usually settled by the single viewpoint, and its shape is one turn and an ending, so a revision of a story reaches the joins and the sentences sooner, and the sentences carry a larger share of the work. What does not change is the order: what the story promises and asks comes first, and its last sentence, the most revised in most stories, is revised last.

The forms at a glance

The table gathers, for each chapter, the instrument’s chief change in each short form. It is a summary of the sections above and of the Reference sections they draw on.

Read down the novella column and the third force is visible in most rows: the form succeeds by choosing one thing and doing it throughout, which is the reverse of the novel’s habit of variety, and the reason a novelist’s habits are the novella’s commonest fault.

The Craft Behind the Story — Appendix E, Short Forms. Introduction · Glossary