9. Worked Analyses

Eight close readings and one revision case. The analytical apparatus is the same throughout: identify the register (§6.4), measure the reach and extent (§3.2), classify the analepsis (§3.3), audit the domains (§4), and assess the return (§6.6).

The eight are arranged in an order that is itself an argument. The first four are individual cutaways of increasing ambition; the last four are novels in which the cutaway has become the structure, and for which the deficit model has progressively less to say.

9.1 Proust, In Search of Lost Time (1913–27) — the trigger that is not a trigger

Register: 8 (involuntary memory). Reach: roughly thirty years. Extent: eventually the whole of a childhood, and then, by extension, the whole novel. Type: external analepsis, completing — though what it completes is a gap the narrator has explicitly announced as unfillable.

The passage is the most cited cutaway in European fiction and among the least accurately taught. The narrator, an adult, is given tea and a small shell-shaped cake by his mother on a winter afternoon. The taste produces an overwhelming and unidentifiable joy. He drinks again, and the effect diminishes; he attempts to recover it by concentration and fails; and only when he stops trying does the source arrive — the same taste, offered by his aunt Léonie on Sunday mornings at Combray — and with it the town, the church, the streets, the gardens, the whole of a period he had believed lost.

Four features deserve attention, and the third is the one this chapter exists to make.

The reach is long and the extent enormous, and the novel accepts the cost without argument. By the arithmetic of §3.2 this is the most expensive cell in the table, made worse: a thirty-year reach opening onto an extent measured in volumes. No diagnostic protocol licenses it. Proust’s answer is not that the cost is small but that the recovery of that extent is the book, which converts the whole subsequent work into the discharge of a single cutaway.

The narrator’s failure is part of the mechanism. He tries, repeatedly, to identify what the taste is offering, and the trying does not work. Voluntary memory, in the essay embedded in the novel, gives back only what the intellect has retained: facts about the past, arranged, dead. The memory arrives only when the effort ceases. A device whose operation requires the character to stop trying cannot be motivated in the workshop’s sense, because motivation implies a reason and this has none.

Proust’s involuntary memory is not the workshop’s motivated trigger, and that is the point. The rule at §6.6 asks the writer to connect the present object to the past material so that the reader can see why this opens that. Proust’s connection is contingent — a taste that happened to belong to those Sundays, with no significance of its own, chosen (as the novel is explicit about) for its very insignificance. The passage would be weakened by a meaningful trigger, because the argument concerns a past that survives outside the intellect, in the accidental and the sensory, and is therefore unreachable by intention. A trigger that made sense would be an act of will, and the whole force of the passage lies in the past arriving unbidden.

The consequence for the trigger rule. §6.6 and §8.1 both hold that the trigger requirement follows from narrative distance rather than standing as an independent law. Proust supplies the extreme confirmation. A narration whose subject is the operation of memory does not need to justify its departures, because the departures are the phenomenon under examination.

On translation. The passage’s power is substantially textural — the rhythm of the long sentence in which Combray rises out of the cup — and texture does not survive translation. Scott Moncrieff’s version (1922), the Kilmartin and Enright revision, and Lydia Davis’s Swann’s Way (2002) offer three noticeably different experiences of the same event. The structural analysis above holds across all three; nothing said here about the sentence rhythm would.

9.2 Brontë, Wuthering Heights (1847) — nested retrospection, measured

Register: 6 (embedded oral tale), containing 7 (found document) and further nesting. Type: mixed analepsis — it begins before the first narrative and continues into it.

The novel’s temporal architecture repays exact description, because it is usually described loosely.

The first narrative is Lockwood’s residence in the neighbourhood, beginning with his opening diary entry dated 1801 and his visit to his landlord. Within a few chapters Lockwood falls ill and asks Nelly Dean to tell him the history of the house. Nelly’s chronicle then occupies the great majority of the book.

The shortening reach is the feature to teach. This is the case §3.2 identifies as the source of momentum inside a long cutaway: the past is visibly coming to meet the present, and the reader can feel the distance closing. It is also what distinguishes the long chronicle from the static flashback, and it is available to any novel willing to let its retrospect progress.

The nesting. Lockwood’s diary contains Nelly’s spoken chronicle, which contains Isabella’s letter (quoted at length), Zillah’s report, and a great deal of reported speech at further removes. By the count at §8.7 this is a fourfold nesting, and the reader is never lost. The mechanism is voice: Lockwood is fastidious, self-regarding and literary; Nelly is practical, moralising and partial; Isabella’s letter is frightened and immediate. The reader tracks the layer by the diction, not by the tense, which removes the cognitive load that nesting would otherwise impose. Chapter 13 (The Viewpoint Shift) treats the reliability consequences; the point here is purely structural.

On the frame’s return. The novel does not simply exhaust Nelly’s chronicle and stop. Lockwood departs, returns some months later, receives the remainder of the story, and closes the book at the three graves. The frame is used, which is what distinguishes it from the frames that open and are forgotten. Chapter 10 (The Prologue) analyses Lockwood’s arrival as a functional prologue; considered from this chapter’s side, the same material is the container that makes a thirty-year analepsis structurally legible.

Where the chapter’s model fails here. By test 1, Nelly’s chronicle begins at a moment of no present-tense pressure whatever — a bored, ill man asks a housekeeper for gossip. By test 10, the past occupies perhaps ninety per cent of the book. The novel is a demonstration that when the proportion is that extreme, the correct move is not to reduce the past but to build a frame adequate to containing it, which is the recommendation at §7.1, test 10.

9.3 Ford, The Good Soldier (1915) — chronology destroyed and reassembled

Register: 9 (repeating analepsis) and 11 (reconstruction), in combination. Type: internal and external analepses in unsignalled alternation.

Dowell narrates his acquaintance with the Ashburnhams from a position after everything has happened, and he narrates it out of order — not artfully out of order, but in the order in which things occur to him, with corrections, retractions, digressions and returns. The novel’s most quoted structural remark is his own account of his method: he imagines himself telling the story aloud, at intervals, to a silent listener by a fire, saying things as they come to him. Chapter 9 §4.1 analyses the first sentence for its reliability signal; the concern here is the temporal architecture that sentence introduces.

What the reader is actually doing. Ohtsuka and Brewer’s finding (§3.6) is that readers reconstruct an underlying chronology from a non-chronological telling, at a measurable cost. The Good Soldier makes that reconstruction the reading experience. The reader is assembling a timeline — the meeting at Nauheim, the years of the affair, the sequence of the deaths, the recurrence of the fourth of August — from fragments delivered in an order determined by a narrator’s associations and evasions.

The two functions of “I don’t know.” Dowell’s celebrated admissions of ignorance are usually taught as a reliability signal, and they are; but structurally they do something else. Each is a marked gap — an explicit statement that something is missing — and per Sternberg (§3.4) an explicit gap generates curiosity where a silent one would generate nothing. Dowell’s ignorance is the mechanism by which the novel opens the slots that its later repeating analepses fill. He is not merely an unreliable narrator; he is a narrator who continuously announces where his account is defective, which is a technique of exposition as much as of characterisation.

Repetition with difference. The events are told many times. The test at §7.1 (test 9) asks what each pass adds, and in Ford’s case the answer is always the same kind of thing and never the same instance: a fact Dowell withheld, a judgement he has revised, a motive he now attributes differently, a sympathy that has shifted. The material is constant and the narrator is not. This is the model for any writer contemplating a repeating structure: the second pass must be the same events seen by a changed person, and the change must be visible.

Where the model fails. By test 7, the chronology test, the book fails comprehensively on a single reading and is designed to. By test 5, most of Dowell’s returns leave the present — the act of telling — entirely unchanged, because the present is a man in a chair. The novel demonstrates that when reconstruction is the reader’s activity, the return test measures the wrong thing.

9.4 Faulkner, Absalom, Absalom! (1936) — a past no viewpoint witnessed

Register: 11 (reconstruction), sustained for a whole novel. Reach: up to a century. Type: external analepsis; the first narrative is 1909–10.

The first narrative is thin and sharply located: Quentin Compson in Rosa Coldfield’s parlour in September 1909, then in his father’s presence, then in a Harvard dormitory in January 1910 with his Canadian roommate Shreve. Almost nothing happens in it. Everything the book is about happened between 1807 and 1869, and none of it is witnessed by any of the people telling it.

The novel’s method is reconstruction, and reconstruction of a specific and demanding kind. Rosa has a version, shaped by grievance. Mr Compson has another, shaped by the fatalism of a man who has inherited a defeated culture and a taste for irony. Quentin has fragments from his grandfather. Shreve has only what Quentin has told him, and no stake at all — and it is Shreve, the outsider, who does much of the imaginative work. The decisive scenes, including the confrontation at the gate between Henry Sutpen and Charles Bon, are constructed collaboratively by two young men in a cold room a thousand miles and forty years away, in the subjunctive: maybe, perhaps, it must have been.

What this does to the chapter’s apparatus. Every measurement returns an anomaly.

The appended chronology is worth a seminar’s discussion. A novel that ships with a timeline in the back matter has conceded that its temporal architecture exceeds ordinary comprehension. Two readings are available and both are defensible: that the apparatus is a courtesy, or that it is an admission. The more interesting question is whether a reader who consults it is reading the book Faulkner wrote — since the difficulty of assembling the chronology is a substantial part of the represented experience.

What the novel establishes for this chapter. The deficit model assumes that the past is being visited in order to explain a present. Here the present has no content to explain: Quentin’s situation is exhausted by his relation to the story he is telling. The cutaway is not a detour from the story; it is the story, and what the book is about is the impossibility of settling what happened. The formal name for this in Genette’s system is that the first narrative has been reduced to a bare occasion — a structure of which Wuthering Heights is the polite Victorian version and this is the extreme.

9.5 Morrison, Beloved (1987) — rememory, and the past that arrives as a person

Register: 9 (repeating analepsis), 10 (braided), and 12 (the unreachable past), in combination. Type: external and internal analepses, converging.

The novel’s present is 1873 at 124 Bluestone Road, outside Cincinnati; its past is Sweet Home, the escape, and the events of 1855 in the woodshed. Chapter 9 §9.5 analyses the opening sentence for its withheld referent; the concern here is the temporal architecture the novel builds on it.

Rememory. Sethe’s term, offered to Denver in an explanation that is also the novel’s theory of time: a thing that happened does not go away when it is over. The place where it happened persists; the picture of it is still there, outside her head, and someone else can walk into it. This is not metaphor within the novel’s terms — Denver is warned, in practical terms, against going to certain places.

The formal consequence is that the past does not stay behind to be visited; it occupies space and can be entered. A narrative operating on that premise cannot organise its retrospects around a present-tense character choosing to remember, because remembering is not the mechanism. The past intrudes. The novel’s departures are therefore not departures at all in the sense §1.1 describes, and the deficit model has, in the most literal way, misidentified which term is the interruption.

Caruth applied, and its limits. The account of traumatic experience at §3.7 — an event not fully registered when it occurred, returning belatedly, resistant to sequence — describes the novel’s form with unusual precision. The events in the shed are approached and veered away from repeatedly before being narrated; the approach is made by different consciousnesses at different distances; the full account arrives late, and even then partly through a newspaper clipping that Paul D cannot read. That is the shape Caruth’s model predicts.

The chapter’s caution at §3.7 applies here too, and a student should hold both halves. Caruth’s literalist account of traumatic memory has been contested — by Ruth Leys on its clinical and theoretical foundations, by Stef Craps on its Eurocentric assumptions, by Dominick LaCapra on its tendency to collapse working-through into acting-out. It is possible to hold that Morrison’s form is a precise representation of traumatic memory, and possible to hold that it is a magnificent set of literary conventions for representing it, and the choice between those positions is not settled by the novel. What is not in doubt is the formal fact: the fragmentation is the subject and not a defect of construction.

The spiral, not the circle. §8.12 sets the discriminating question for repeated returns: do they converge? Here they do, and the convergence is designed. Each approach to 1855 is nearer than the last — in information, in proximity of viewpoint, in the willingness of the narration to remain in the room. The final full account lands as the completion of a movement the entire book has been making, which is why the repetition, which by the arithmetic of §3.3 should be the most expensive structure in the chapter, reads as inevitability.

And the past arriving as a person. The novel’s most radical formal act is to convert an analepsis into a character. Beloved is the past, embodied, walking up out of the water and into the house — which means the novel can dramatise the return of what is past in the present tense, without any cutaway at all. This is the limit case of the whole chapter, and it is worth a seminar’s attention: a book that can put its past on the page as a present-tense character has abolished the interest deficit by abolishing the departure.

9.6 Ishiguro, The Remains of the Day (1989) — the whole novel as motivated recollection

Register: 5 (dramatised flashback) and 4 (iterative habitual), inside a frame that is itself register 7 (a day-by-day record). Type: external analepsis. First narrative: six days in July 1956.

The novel’s design is the cleanest available demonstration that a thin present can be a structural asset rather than a defect.

The frame. Stevens, an ageing butler, drives from Oxfordshire to the West Country in his employer’s car, ostensibly on an errand to a former colleague. The present timeline’s events, entire: he sets out; he worries about his motoring clothes; he mismanages the petrol; the radiator overheats; he is mistaken for a gentleman in a village; he walks up a hill and looks at a view; he meets Miss Kenton; he sits on a pier at dusk. Nothing in that list would sustain a chapter on its own.

The mechanism. Each day of the drive is a chapter with a place-name heading, and each generates recollection by an entirely conventional means: the landscape, the errand, the meetings, the conversations, the physical fact of leaving a house he has not left in decades. The past supplies Lord Darlington’s political career, the 1923 conference at which Stevens’s father died upstairs while he continued serving, the dismissal of two Jewish maids at Darlington’s instruction, and the twenty-year non-relationship with Miss Kenton.

Why the thin present works. Three reasons, all transferable.

1. The journey supplies a structure with a destination, so the present has direction even without incident. The reader knows where the car is going and that something will be said when it arrives.
2. The proportion is an argument. A life given entirely to another man’s household has very little present in it, and the disproportion between the two timelines states that without a word of commentary. §8.8’s discriminating question — does the book know it, and is the thinness saying something? — is answered affirmatively.
3. The present is not inert; it is a slow reveal of the frame narrator’s condition. The bantering Stevens cannot manage, the professional dignity he keeps redefining, the confusion about who he is when he is not at Darlington Hall: the present timeline is where the reader watches the cost of the past being paid.

The iterative. Ishiguro’s most underappreciated technical instrument here is register 4. Large stretches of the past are delivered not as scenes but as habitual practice — how the staff plan was drawn up, what the servants’ hall did in the evenings, what the great houses of that period expected — and the iterative is what establishes the normal against which the dramatised scenes register as exceptional. A student can learn more about the iterative from twenty pages of this novel than from any theoretical account of it.

The return that never happens. The novel’s most sophisticated move, from this chapter’s point of view, is that Stevens’s cutaways frequently fail test 5. He returns from a devastating recollection to the present unchanged, and comments on the excellence of the English landscape. The failure is the characterisation: a man who can go back to the night his father died and return with a reflection on professional standards has told the reader everything. Where the untethered return is the represented condition, §8.5 measures the technique and misses the point.

9.7 Salih, Season of Migration to the North (1966) — the embedded life story

Register: 6 (embedded oral tale), containing 7 (documents in the locked room). Type: external analepsis, delivered as speech.

The unnamed narrator returns to his village on the Nile after seven years of study in Europe, addressing his account to a listening group. In the village he finds a newcomer, Mustafa Sa’eed, who is unaccountably reserved about his own history, and who one night, drunk, recites English poetry. The narrator’s curiosity is aroused; some time later Mustafa tells him his life — Khartoum, Cairo, London, the English women, the killing of his wife, the trial, the prison years — in a single long night’s confession.

Why the embedded tale is the right register here, and not merely the traditional one.

• The teller has a motive, and it is withheld. Mustafa chooses to tell, chooses this listener, and the reader is not told why until much later. A dramatised flashback would have supplied the same events with none of that pressure.
• The listener is changed by hearing. This is the requirement §8.11 identifies for relocated material, and it is met continuously: the narrator’s own return from Europe becomes unstable in the light of Mustafa’s, and by the end he can no longer keep the two stories separate. The tale is not a container of information but an event in the frame narrative with consequences that run to the last page.
• The oral register carries the novel’s politics. The narrator addresses an audience; the story is being told in a village, in a tradition of telling, about a man whose formation was European. The form itself stages the encounter the book is about.

The locked room. After Mustafa’s disappearance the narrator eventually enters the room he kept locked, and finds it furnished as an English study, with books, papers and a portrait. This is register 7 arriving after register 6: the spoken account is followed by documentary evidence, and the evidence does not simply confirm the speech. A novel that delivers its past twice, in two registers, has made the relation between the two registers into content — which is the substitution exercise at §6.5 performed inside the book rather than in the writer’s notebook.

On translation. Denys Johnson-Davies’s 1969 English version is the one anglophone students read, and the structural claims above survive it. Claims about the texture of the narrator’s address, or about the relation between the Arabic of the village and the English of Mustafa’s education, do not.

9.8 Tolstoy, The Death of Ivan Ilyich (1886) — analepsis as the whole design

Register: 3 (summary) modulating into 5 (dramatised scene). Reach: forty-five years, shortening to nothing. Type: external analepsis with a first narrative of one chapter. Translation: Louise and Aylmer Maude.

The novella’s structure is the clearest teaching case in this chapter, because it is simple, extreme, and entirely successful.

Chapter I is the first narrative, and it is complete. Ivan Ilyich’s colleagues learn of his death from a newspaper in their chambers, and each immediately calculates what it means for his own promotion. Peter Ivanovich attends the funeral, is trapped by the widow into a conversation about her pension, is unnerved by the corpse’s expression, and escapes to a game of cards. The chapter has a protagonist, a want, an obstacle, an ending. It is a scene, not a prologue, which is why Chapter II can leave it (§8.6).

Chapter II opens the analepsis with the sentence anatomised at §5.2, and the remaining ten chapters are that analepsis, moving forward through the career, the marriage, the house, the illness, the doctors, the lie everyone maintains, the peasant Gerasim, and the final three days — until the retrospect arrives at the death the reader was told about on the first page.

The proleptic structure, and what it buys. Announcing the death first is a prolepsis in the sense of §3.3, and its effect is the one predicted there: it converts the whole subsequent analepsis into a completing structure. The reader is not asking what will happen to this man, because they know. They are asking how did a life arrive here, which is Sternberg’s curiosity rather than suspense (§3.4), and it is sustainable for the whole length because the answer is a moral argument rather than an event.

There is a second and less obvious return on the design. Because the reader has met Ivan Ilyich’s colleagues before meeting Ivan Ilyich, they read the whole retrospect knowing what his death will mean to the people around him: an inconvenience, a vacancy, an occasion for calculation. The judgement the novella is going to deliver has been demonstrated before the case is heard.

The domains, in a single structure.

The accelerating rhythm deserves separate notice, because it is the feature of the design that transfers most readily to other books. The analepsis begins in register 3 — years compressed into paragraphs, the iterative and the summary — and progressively slows, until the final chapters are approximately scene-rate and the last hours occupy pages. The reach shortens; the pace slows; the two together produce the sensation of arrival that a long retrospect otherwise lacks. This is the technique §3.2 describes abstractly, executed as well as it has ever been.

9.9 A revision case: the same material in two registers

The following applies the protocol to the series’ running example — the invented contemporary novel about three siblings and their dead father’s shop. The scene under revision is the one in which Nadia, at the door of the flat above the shop, puts her hand into the pocket of her father’s coat, still on its hook, and finds the good pen.

The canonical placement matters and should be stated before the versions. The good pen was planted in Chapter 10’s material (The Prologue), where the father pockets it on the last ordinary day instead of leaving it by the till. The line he speaks in the cutaway is quoted verbatim at the novel’s close, in Chapter 16’s material (The Ending). The prologue carries a fragment of the line, In case, as the father’s unfinished reason when he pockets the pen (Chapter 10’s revision case); this cutaway is therefore partly a repeating analepsis, and what the second pass adds is the rest of the line, the occasion, the listener and Nadia’s misreading. This cutaway therefore sits in the middle of a three-point structure: an object planted, a meaning supplied, a line returning. Everything below should be read against that.

Version A (draft)

When Nadia was nine her father let her help with the stock count for the first time. It was a Sunday in November and the shop was closed and the radiator by the door made its noise like something clearing its throat. He had set out two chairs at the counter and a mug of tea for each of them, and he showed her how to make the marks in fives, four strokes and one across, and how to cross a seven so it could not be mistaken for a one.

“You have to be able to read it in ten years,” he said.

She had asked him then why he was using the good pen. She had known even at nine that the good pen was for the ledger, that it lived by the till in its case, that Ma had had to buy a new nib for it twice that year because he kept losing it. She asked him whether he was going to lose it again, and whether Ma would be cross.

He put the cap on the pen and set it down on the counter between them, and he looked at her over his glasses in the way he had when she had caught him out at something. Outside, a bus went past and the window shook the way it always shook.

“Because someone should write it down properly,” he said. “In case anyone ever needs to check.”

She had not understood what he meant. She had thought he meant the shop — the stock, the prices, the ordinary arithmetic of it, the things a shop had to be able to prove. She had gone back to her column of fives. He had gone back to the shelves, calling out the numbers, and she had written them down in her nine-year-old’s hand while he stood with his thumb hooked in his cuff and read the labels out to her one after another until the light went.

(Approximately 320 words, and in the drafts this is modelled on it runs to two pages.)

Diagnosis

Patterns present: the unmotivated cutaway (§8.1); the childhood-explains-everything cutaway (§8.9); the untethered return (§8.5), in its most complete form, since there is no return at all.

What is worth noticing about Version A. It is not badly written. The radiator clearing its throat, the crossed seven, the marks in fives, the thumb hooked in the cuff — these are exact and characterising, and a workshop would praise several of them. The failure is entirely one of register, placement and motivation, and this is the general case: draft cutaways are far more often well-written and misplaced than they are badly written.

Two further faults are worth isolating because they are invisible from inside.

Version A converts an illumination into an explanation (§4.2). By placing the exchange in Nadia’s childhood, it proposes that this is where she learned something about her father — that the moment is formative. The canonical exchange is not formative; it is an ordinary conversation between adults, half-forgotten, whose meaning changes retrospectively. The difference is the whole of the effect.

Version A spends the line. The father’s sentence appears here, at leisure, in a scene that has been building towards it for three hundred words, cushioned by tea and radiators and a bus. The line carries the novel’s ending, where it is quoted again. Delivered like this, it has been used up.

Version B (revision)

Nadia found the pen in his coat pocket, right where she remembered him keeping it — except it wasn’t the register pen, the plain black one. It was the good pen, the fountain one he only used for the ledger, the one he wasn’t supposed to carry loose because he’d lose it. He’d lost it twice that year already, and made Ma buy a new nib.

She had asked him about it once, the second time. Why he needed the good pen for a stock count nobody was going to read twice. He’d looked at her over his glasses like she’d caught him at something. “Because someone should write it down properly,” he said. “In case anyone ever needs to check.” At the time she’d thought he meant the shop.

She had not thought, until just now, standing by the door with her hand in the pocket of a coat still on its hook, to ask him what he’d actually thought anyone might need to check.

Audit

Measurements. Register 5 at minimum length, entered through register 1. Reach: months, not decades. Extent: one exchange. Type: completing — and the gap was opened by the prologue material, where the reader saw the pen pocketed and did not know why. Trigger: the object, the strongest available (§6.6), and one that remains in her hand after the return.

What the revision did, procedurally

1. Changed the register, not the content. The same information; a scene of 320 words becomes an exchange of four lines. The compression check (§6.3) was run and returned keep the sentence, and the sentence turned out to be the father’s own.
2. Attached the cutaway to an object already load-bearing. The pen was planted two hundred pages earlier and is in Nadia’s hand at the moment of departure. The trigger did not have to be invented; it had to be found (§5.3).
3. Moved the exchange out of childhood. Restoring it to the recent past — the second time he lost the pen — removes the determinist implication, respects the canon, and makes the point sharper: this was an ordinary conversation, recently, which she did not understand at the time and understands now.
4. Shortened the reach from twenty-five years to a few months. The construction cost collapses. No radiator, no bus, no two chairs: the reader already knows this shop.
5. Rewrote the return as a changed question. Version A ends inside the past. Version B’s last sentence returns Nadia to the door of the flat holding a question she did not have when she reached into the pocket — what did he think anyone would need to check — which the present scene now has to carry. This is the changed question at §6.6, and it converts the cutaway from a pause into a turn.
6. Declined the explanatory sentence. Nothing states that the father knew he was ill, or that the pen was a signal. The reader assembles it, which is why it holds (§5.3).

What Version B has not solved

An honest account has to include this, and four things are outstanding.

The last sentence is doing a great deal of work and may be doing too much. She had not thought, until just now, to ask him what he’d actually thought anyone might need to check is the return, the changed question, the thematic statement and the emotional beat in a single construction. It is close to the density ceiling described at §5.4, and a reasonable editor would ask whether the reader could get there with less — whether the clause at the time she’d thought he meant the shop has not already done the work, leaving the last sentence to say something plainer.

The trigger is very apt. The object is the pen; the pen is the point; the memory is about the pen. §5.4 warns that an over-motivated trigger shows the machinery. The defence is that the aptness was established two hundred pages earlier by a different chapter and is therefore inherited rather than manufactured — but the defence depends on the prologue material existing, and a reader encountering this scene in isolation would find it neat.

The register is the contemporary literary default. Wry, precise close third, with the past folded into the present through the pluperfect and a fragment of remembered dialogue. It is a well-made instance of a very common thing. Chapter 9 §9.8 makes the same concession about the same character’s opening paragraph, and the concession is worth repeating here for the same reason: the revision has solved the structural problem, and has not made the prose distinctive against its own shelf.

The scene before has to motivate the hand. Nadia has learned not to touch the coat (Chapter 4’s material), and in this version the coat stays on its hook and only her hand goes into the pocket. Whatever moves her to do even that, months after March, has to be supplied by the scene this cutaway sits in, and Version B does not supply it.

One thing Version B does not attempt, and might have. The material could have been delivered at register 6 — Nadia telling Elena about the exchange, in the present, in dialogue — which would have supplied a listener, a motive for telling, and the possibility of Elena disagreeing about what their father meant. That version does not exist, and the substitution exercise at §6.5 exists in order to make writers produce it before deciding.