Chapter 2. Moves Through Six Stages

In the previous installment, we named this place: . This time, I would like to show what happens inside it, and in what order.

When we take up a new work, the first question that comes to mind is simple: “Where do we begin to get hold of this?” We have six stages for answering it. Each stage has a defined set of materials to produce, and a defined amount of time for producing them.

It is like building a new house. You do not do everything at once. First you inspect the site and draw a plan; then you raise the frame; then you fill in the walls; then you bring in the furniture; and finally you clean. Ignore the order, and you may bring in the furniture only to tear down the walls again. Preparatory work for a book is the same. The work stays intact only when what is made, and in what order, have been decided.

Let us begin with the whole picture.

Stage 1 — Read the work and describe its character. Stage 2 — Divide it into roughly eighteen segments and extract keywords. Stage 3 — Link events into chains of causality. Stage 4 — Make a note for each character. Stage 5 — Build some thirty kinds of materials around the work’s world. Stage 6 — Cross-check every material.

Applied to Pride and Prejudice, these six stages took about fifteen hours. Stage 5 took the longest. Let us look briefly at each stage.

Stage 1 — Read the Work and Describe Its Character

The first task is to read the book. We read through it, read it closely, grasp the flow of the whole work, and write a one-page document. We call this document, which never exceeds two pages, the work profile.

It records five things: the period the work treats; its genre; its cultural setting; its prominent motifs; and what seems most difficult in translation.

The profile for Pride and Prejudice came out as follows: a social novel of Regency England and a romance built on a marriage plot. It has a strong use of free indirect discourse (meaning that sentences in which the narrator slips into Elizabeth’s inner life occur throughout the novel). Twelve motifs—class, marriage, family, letters, servants, food culture, clothing, and others—operate strongly. Its many period expressions make vocabulary difficult, and its long sentences make grammar difficult as well.

This single document becomes the standard for every decision that follows: what materials to make, where to spend more time, and which traps to mark in advance all derive from it.

Stage 2 — Divide It into About Eighteen Segments and Extract Keywords

Next, we divide the whole work into groups of chapters. If we divide it too finely, there will be too many tables; if the groups are too large, too many changes of meaning enter one table. In most cases, dividing a work into fifteen to twenty segments is appropriate.

We divided the sixty-one chapters of Pride and Prejudice into eighteen segments. Segment 1 comprises Chapters 1–6 and Segment 2 Chapters 7–9; this is how the whole novel becomes eighteen groups. We cut at narrative breaks: points where the place changes, time jumps, or an event concludes.

We make one table for each segment. Every character, place, object, concept, and institution newly introduced in that segment receives a row. Bingley, Darcy, the Bennet family, and Netherfield first enter the table for Segment 1; Hunsford, Rosings, and Colonel Fitzwilliam enter the table for Segment 9.

A row in the table is not merely a word. It states what that keyword means at that point in time. In the table for Segment 1, Darcy is “a proud, taciturn, wealthy gentleman.” When the same Darcy reappears in the table for Segment 12, he is recorded as “a master showing Elizabeth his estate, transformed into gentleness.” The person and name are the same, but their meaning at that point is different. That is why we make eighteen tables.

There is one strict rule: we do not write spoilers. We may not write under Darcy in the Segment 1 table, “He is actually a good person and marries in the end.” At the time of Segment 1, that information does not exist. The time of the material must match the time of the work if the material is to be honest. A translator working on Segment 1 must be able to open only the meaning available at Segment 1.

Stage 3 — Link Events into Chains of Causality

We assign serial numbers to every moment in the work when meaning changes. In Pride and Prejudice, we identified forty-nine events in the main text. The first is the news reaching the Bennet family that Bingley is coming to Netherfield; the last is the weddings of the two sisters, Elizabeth and Jane, and the settling of their subsequent relationships.

Each event is marked with its preceding event: which event is the result of which other event. Event 28, Darcy’s first proposal, is the result of Event 27, Colonel Fitzwilliam’s unguarded disclosure; Event 27 in turn results from Event 19, Bingley’s sudden withdrawal to London. This is how events are tied into a chain of cause and effect.

Events occurring before the novel begins are indexed separately. Darcy’s birth, the building of Pemberley, the period when Wickham’s father worked on Darcy’s father’s estate, and Collins’s ordination do not appear directly in the main narrative, but they determine ages, relationships, and the roots of conflict. We identified forty-one such background events in Pride and Prejudice.

The forty-nine main events, forty-one background events, and the smaller links between them give us an event master index that organizes the whole time of a work.

Stage 4 — Make a Note for Each Character

Next come character notes. We create a separate document for every character whose name appears. For Pride and Prejudice, we prepared notes for thirty-four people.

The core of a note is its development in event order. It tracks how the character changes as the work passes through them, alongside the event numbers assigned in Stage 3. Elizabeth’s note records “Act 1: Formation of prejudice; Act 2: Confrontation and rupture; Act 3: Turning point; Act 4: Reassessment; Act 5: Fruition,” together with the event numbers.

The protagonist’s note is long and detailed. It includes inner evolution across five acts, arcs of key relationships, and even gradations of tone to watch in translation. Supporting characters’ notes are more concise. A bit part may end in one or two paragraphs.

But we do not omit a bit part. Mrs. Reynolds, Pemberley’s housekeeper, appears in only one scene in the whole work. Yet the five minutes in which she testifies about Darcy overturn the reader’s entire perception. A minor character can be the decisive witness to a scene; that is why we made her note as well.

Stage 5 — Build Some Thirty Kinds of Materials

This is the most labor-intensive stage. We separate and write some thirty kinds of documents for every context surrounding the world of the work: period, class, custom, and language.

The next installment will explain in detail how these thirty-odd kinds are classified. In short, there are four layers: eight types made for every work; several added according to the work’s genre; one or two added according to its period; and around ten added according to its central motifs.

For Pride and Prejudice, that total came to thirty-five kinds. For other works, Romeo and Juliet comes to about eighteen and Moby-Dick to about twenty-four. The procedure is the same, but the kinds of materials vary with the character of the work.

Stage 5 is the longest. It took about seven hours for Pride and Prejudice, an average of twelve minutes per sheet. The greatest amount of time is spent deciding what each material must say. Designing the structure of a material is harder than filling it in.

Stage 6 — Cross-Check Every Material

Finally, we hold every material against the others. Do the character notes and event index agree? Has the same word been rendered differently in Korean across documents? Is every character’s name written the same way everywhere? Does every preceding-event reference point to an event that actually exists?

Discovered errors are handled by grade. Errors that directly affect translation accuracy are corrected immediately: if a character’s name is written differently in one place, a translator may render the two forms as two different people. Errors affecting traceability of the materials, such as an event number pointing to the wrong place, are corrected before release. Items needing structural reinforcement, such as a missing description of a minor character, are supplied in the next work cycle.

For Pride and Prejudice, Stage 6 verification produced about ninety checklist items, five of which required immediate correction. The rest were sorted by grade and carried forward to the next task.

What Exists When Stage 6 Is Finished

When the six stages are complete, a standardized work folder exists for a single work. The folder bears the work’s name. Inside are thirty to forty character notes, one event master, some thirty volumes of setting materials, one folder guide, and one verification report.

A translator begins the main work with this folder at hand. When a word, a character, or a period custom stops them while translating a chapter, they find the answer in the folder. A resource whose answer cannot be found within thirty minutes is not a resource but merely a page; this is why the folder guide serves as a map of where to find what.

The preparatory work for one volume of Pride and Prejudice took about fifteen hours. The time we reduced—from six months to thirteen weeks, from thirteen weeks to two weeks, and from two weeks to a simplified two hours—did not become possible by removing people and inserting machines. It is the result of separating into six stages the precise points at which human judgment must arrive.

The next installment enters the longest Stage 5. We will examine the classification logic that divides the thirty-odd kinds of materials, how that classification settled into a total of thirty-five kinds for Pride and Prejudice, and why.