Repository path: workshop/experiments/E-20260729d-decision-grain/runs/materials.md · rendered 2026-09-09

materials.md

PART 1 — THE SEVENTEEN CANDIDATE RECOMMENDATIONS

C1. Declare the target register; "natural" is not a single corpus senses: naturalness, purpose-fit evidenced on: EN as target, any source full statement: A naturalness judgment is undefined until the target register is named. Period-idiomatic and contemporary-vernacular English license different sentences, and a translation that does not declare which it is aiming at is making the choice by omission.

C2. Ask what the fluency cost, not whether it is too fluent senses: naturalness, cultural-mediation, affect evidenced on: JA→EN full statement: Fluency in the target is not itself the domesticating move. The domestication lives in how the fluency is bought: fluency-by-substitution (swap the foreign item for a familiar one) assimilates source meaning away; fluency-by-scaffolding (retain the item and manufacture its sense through context) does not. So a smooth-reading translation is diagnosed by asking what was paid, not by counting its smoothness.

C3. The handling set for culture-bound items is eight-wide; choose knowingly senses: cultural-mediation, style-correspondence, consistency evidenced on: JA→EN, RU→EN, EN→FR, JA→JA, OE→EN full statement: The strategies actually observed in published practice for a culture-bound item are: retain (copy-opaque), gloss, substitute, omit, scaffold (retain and manufacture the sense through context), transparent calque (translate the foreign structure literally so its sense carries with no foreign residue), measure conversion (silently recalibrate a quantity into the target's system), and added realia (supply target-culture furniture the source never had). Two further terms name a direction and an outcome rather than a strategy: re-foreignisation (import a third language's material as the compensation) and preserved-inert (form retained, function lost).

C4. Declare the pair; sense weights are pair-relative senses: all evidenced on: 3-cell design

C5. Grammar-borne meaning transcodes into lexis and loses systematicity senses: style-correspondence evidenced on: RU→EN, JA→EN (EN-target only) full statement: Where the source encodes meaning in a grammatical category the target lacks, translators do not lose the meaning outright; they move it into lexical choices. What is lost is systematicity: what applied automatically at every occurrence becomes intermittent, applied where the translator noticed and omitted where they did not.

C6. The false friend does its damage inside a drift window senses: style-correspondence, accuracy evidenced on: OE→EN, JA→JA (diachronic) full statement: When the target holds a form the source's word could map onto, the danger is not proportional to the semantic gap. Below the window the reflex is simply usable; above it, where drift is total, nobody is tempted; the damage is done in the middle band, where taking the available form gives a locally grammatical sentence with the wrong meaning. The corollary is counter-intuitive and is the sharper half of the claim: having no reflex and having a totally drifted one are the same condition, so lexical overlap only decides whether a tempting wrong candidate is on the table — semantic distance, not overlap, is the operative variable.

C7. Published translators do not handle a forked class uniformly senses: consistency, cultural-mediation evidenced on: 5 pairs

C8. Archaism buys inheritability back — foreignisation is a capability, not only a stance senses: style-correspondence, purpose-fit evidenced on: OE→EN

C9. Long source periods are split in published practice senses: style-correspondence, naturalness evidenced on: JA→EN

C10. Measure the lead's contamination before selecting material senses: — evidenced on: all full statement: Measured against pairs of published translations of the same passage, a lead translation is the most central text in its cell — the largest share of its wording appearing somewhere else in the cell — and is symmetrically close to both published texts. Centrality of this kind is what a translator landing near the centroid of plausible renderings produces, with no recall involved, and it fits the data better than memory does. Asymmetry, not overlap, is the signature of recall.

C11. Check a published pair for dependence before using it as a baseline senses: — evidenced on: LA→EN, DE→EN, RU→EN full statement: Two translators working independently from one source do not normally produce long runs of identical target-language wording. Where they do, textual dependence is the better explanation than period, form, or the source forcing one rendering — and dependence is common enough in published pairs that any use of a published pair as an independence baseline must be checked first, not assumed.

C12. One self-revision pass buys naturalness without moving accuracy senses: naturalness, accuracy, style-correspondence evidenced on: JA→EN

C13. Measure whether a comparison's two arms are identifiable without reading, with a sign count rather than a correlation senses: — evidenced on: all full statement: Before a paired comparison is judged, measure whether a rule that reads nothing — count the words, count the characters — can say which text is the treatment. Report it as the accuracy of the better of the two constant rules (always pick the larger, always pick the smaller), which is a sign count. A correlation between preference and the size of the difference is a different measurement and does not substitute: identifiability and cue-use are independent properties, and a correlation can only see the second.

C14. Never pool a paired comparison across length-sign strata — and never read the stratified figure as an estimate below ~22 items per stratum senses: — evidenced on: all

PART 2 — LOG A: 21 decisions from one translation

A1. Target register, and nothing in the project could settle it. Sand 1846: an essayistic narrator, formal but not ornate, addressing a general reader. The project's naturalness sense is anchored at exactly two poles, 1922 period-idiomatic and 2008 contemporary-vernacular, and this text belongs at neither. I chose an unmarked modern literary English — no contractions in the narration, no deliberate archaism, "byre" and "kine" rejected as importing a British-dialect colour the French has not got. Alternatives live: a period-1846 English (Trollope-adjacent), which would have made the long periods easier and the argumentative passages heavier; a plain contemporary register, which would have cost the sentence shapes. Declared under purpose-fit above.

A2. éducatibilité → "educability." Marked, faintly technical in French; "educability" is marked and faintly technical in English. Rejected: "perfectibility" (a different eighteenth-century doctrine Sand is not invoking), and the paraphrase "man's capacity to be educated" (loses the abstract noun that the sentence is about).

A3. l'homme → "man."** Generic masculine singular. "Human beings" would modernise the argument's shape; the surrounding chapter is about the labouring man as a class, and Sand's own generic is masculine. Cost: the English generic "man" now reads as a period marker in a translation I have otherwise declared unmarked, so D1 and D3 pull against each other and I let D3 win locally.

A4. arène → "arena." Sand calls a ploughed field an arena because she is still looking at Holbein's Dance of Death, where the ploughman's field is a stage. Flattening to "expanse" or "field" would remove the figure. Kept, at the cost of a momentary oddity in English.

A5. tableau → "picture."** Holbein's ploughman is a woodcut, not a painting; "painting" would be an error of fact introduced by the translation, "engraving" an over-precision the French declines. "Picture" holds the same width as tableau.

A6. The suspended direct object in the landscape sentence. French: le paysage … encadrait de grandes lignes de verdure … ce large terrain. The object arrives after the instrument. English wants "framed that wide stretch … with great lines of greenery." I kept the French order ("framing with great lines of greenery — … — that wide stretch of strong brown earth"), which is strained English. Reason: the alternative "was framed by" reverses the agency Sand gives the landscape, and the natural reordering loses the delay. This is a style-correspondence purchase paid for in naturalness, made knowingly.

A7. tiède → "mild."** No English adjective is simultaneously weather-neutral and tactile: "tepid" and "lukewarm" are pejorative in English and neutral in French; "warm" overshoots; "soft" is a different sensation. "Mild" is the idiomatic collocation for a day and is the flattest of the options. Loss recorded: the skin-level warmth tiède carries.

A8. exhalait → "breathed out." Chapter II sustains a personification of the earth as a body (fendre le sein de cette terre jalouse, just before this excerpt). "Gave off" or "emitted" would be the unmarked choices and would break the figure at its second occurrence. Kept the bodily verb.

A9. celui d'Holbein → "Holbein's ploughman." The pronoun's antecedent (le laboureur d'Holbein) is outside the excerpt. Supplying the noun is an addition to the text, made solely because the passage is excerpted; in a whole-work translation the pronoun would stand. Flagged as an artefact of excerpting rather than of the pair.

A10. areau → "swing-plough of ancient pattern," and this was the hardest decision in the passage. areau is the Berry regional form of araire, a wheelless swing-plough, and Sand has used charrues (wheeled ploughs) two sentences earlier for the other men's implements. The contrast is doing work: the old man's tool is archaic where his neighbours' is not. Four live options: (a) copy-opaque "areau", which manufactures an opacity for the English reader that Sand's reader did not have, since her word is glossed on the spot by de forme antique; (b) "ard", the exact English term, which is archaeologists' vocabulary and would read as a misprint; (c) "plough of ancient form", which loses the noun and therefore the charrue/areau contrast entirely; (d) "swing-plough," the standard English agricultural term for the wheelless implement, which preserves the contrast. Chose (d), and kept "of ancient pattern" so that Sand's own gloss survives. Cost, and it is the real one: areau is regionally marked in French and "swing-plough" is not marked at all in English. The regional colour is gone; only the technical contrast is saved.

A11. de ces vieux travailleurs → "those old workers."** Sand applies a human-labour noun to the oxen, in a chapter arguing about human labour. "Beasts", "creatures", or "that old working kind" would each remove the crossing. Kept the noun even though "workers" for oxen is odder in English than travailleurs is in French.

A12. frères → brothers, italics kept. Transparent calque. The French is a regionalism, italicised and glossed by Sand herself (comme on les appelle dans nos campagnes); the calque is fully transparent, so retaining the French string would add opacity the source does not have. The italic is retained because it marks the word as quoted local usage, not as foreign.

A13. comme on les appelle dans nos campagnes → "as we call them in our country districts." French on + nos makes the narrator an insider to the region. English "as they are called" would lose it; "we" keeps it. See D19, where the same on occurs without the possessive and I decided the other way — the inconsistency is deliberate and is the closest thing in this log to a consistency fault I chose to commit.

A14. Animal pronouns: "he/his" throughout, never "it."** French grammatical gender forces nothing here (bœuf is masculine, which is grammar, not personification), but the passage's entire argument is that the ox has a friendship and a grief. English "it" would silently refute the paragraph. Held across all seven references.

A15. camarade d'attelage → "yoke-fellow."** English has a settled idiom with the same affective colour, and it re-sounds the yoke, which recurs twice more. Rejected "harness-mate", "team-mate" (modern sporting sense).

A16. étable → "cattle-shed."** étable is specifically for cattle (as against écurie for horses), so "stable" is inexact. "Byre" is exact but pulls a British-regional register the French does not have (see D1). "Cow-house" is exact and reads as a gloss. "Cattle-shed" is precise, unmarked, and slightly duller than the French.

A17. déplorables mugissements → "piteous bellowing." False friend: déplorable here is lamentable / to be wept over, not English "deplorable" (= reprehensible). One of two false friends in the passage.

A18. terres plus fortes → "heavier … ground." The second false friend, and an agricultural one: terre forte is the technical term for heavy, clayey soil. "Stronger soil" would be a plain mistranslation that reads perfectly well in English, which is what makes it dangerous.

A19. Le bouvier dira and the register of the quoted speech. The future is gnomic — the oxherd will tell you, i.e. this is what oxherds say — and English "will tell you" carries it. More consequential: Sand does not mark this speech dialectally, though she marks peasant speech elsewhere in the book. I rendered it plain. The temptation was to supply rustic colour ("There's a pair o' bullocks gone"), on the model of a cultural-mediation handling that translators demonstrably use; I declined, because there is nothing in the source to repay and the addition would be invention. Also within this decision: paire de bœufs → "a pair of oxen", not "a yoke of oxen", which is the English farming idiom but would sound the yoke motif a third time where Sand does not.

A20. c'est ce qu'on appelle → "these are what are known as." Impersonal on with no possessive, unlike D13, so I kept it impersonal at a cost in fluency. fraîchement liés → freshly bound, italics kept, transparent calque again. Rejected: "newly broken", which is the English farming idiom but names a different operation — breaking is subduing an animal, lier is tying animals into a team. Substituting it would have been fluent and wrong.

A21. conduisait / gouvernait kept distinct: "was driving" / "was managing."** Sand uses two verbs for handling a team within four sentences and uses gouverner again later in the chapter for the child. English could level both to "drove"; I held the distinction on the chance that it is deliberate, knowing I cannot tell from the excerpt whether it is.

PART 3 — LOG B: 42 decisions from another translation

B1. Target register. Ōgai 1916, historical orthography, quasi-classical: an essayistic-historical narrator in span A, free indirect interior in span B, both measured and unornamented. I chose an unmarked English with a slightly formal, period-neutral cast — no contractions in narration, no deliberate archaism, but a tolerance for "to be sure", "and there an end", "it fell out that", which are period-available without being costume. Alternatives live: a 1916-contemporary English (Edwardian), which would have matched the source's date but not its plainness; a flat contemporary English, which would have cost the sentence rhythms in span B.

B2. 高瀬舟 → "the Takase boat", not "the Takase River boat" and not italicised.** The title object. The source names it flatly and immediately explains it; the English can do the same. Rejected: "Takasebune" (opaque where the source is not), "the barge" (wrong vessel), "the Takase River boat" (used by one published title, and it doubles "Takase" against the next clause's "Takase River").

B3. 同心 → "dōshin, a constable" on first use, "constable" after. Realia with no English equivalent: a low-ranking Tokugawa police officer under a 与力 under the city magistrate. Four options were live: (a) bare transcription throughout, which makes the reader carry an unglossed term through five paragraphs where the source's reader carried nothing; (b) transcription with an in-text gloss on first use, then the gloss word — chosen; (c) "constable" throughout, which loses the term entirely; (d) a footnote, which this project's artifacts do not use. Cost:** the source word is not marked or foreign for its reader, and mine is, once. I bought the term's existence at the price of one moment of opacity [comparison to a decision in the other log redacted for blinding].

B4. 遠島 → "banishment to the islands."** A specific Tokugawa penalty (transportation to a penal island). "Exile" is too grand and unlocated; "transportation" is the exact English legal term for the British analogue and would import Botany Bay. "Banishment to the islands" is a transparent calque that stays inside the source's own frame, and it lets 島 recur later as "the islands" without a new noun.

B5. 牢屋敷 → "the prison house."** 牢屋敷 is the walled jail compound. "Gaol" is British-period colour the source has not got (the D1 rule); "prison compound" over-specifies. "Prison house" is dull and exact.

B6. 京都町奉行 → "the Kyoto city magistrate."** The 町奉行 was administrator, police chief and judge at once. "City magistrate" carries the judicial sense and drops the administrative one. Rejected: "the Kyoto Town Magistrate" (capitalised office-title English, wrong register), "the city commissioner" (loses the court, which the next paragraph needs for 白洲).

B7. 默許 → "a tacit permission," and the sentence's three-beat close. The source ends with a bare apposition: 所謂大目に見るのであつた、默許であつた。Two flat predicates in series, the second a single noun. English resists ending on "it was a tacit permission" twice over, so I used a dash: "a matter of looking the other way — a tacit permission." Cost: the source's paratactic thud becomes a subordinated afterthought.

B8. 科 → "offence," consistently, against 罪人 → "criminal."** The source distinguishes 科 (the act, the charge) from 罪 (the guilt). English "crime" would level them. Held across the span.

B9. 心得違 → "a lapse of judgement," with "what was called." A period term for a culpable misunderstanding of one's own situation — closer to mis-taking one's position than to a moral failing. Rejected: "a misunderstanding" (too light), "an error of judgement" (modern administrative). The source's 所謂 flags the term as quoted, and I kept the flag: "what was called."

B10. 相對死/情死 → "what was then termed a mutual death — a lovers' suicide." The hardest realia decision in span A. 相對死 (aitaijini) is the Tokugawa legal term; 情死 (jōshi) is the ordinary word, and Ōgai supplies the second to gloss the first — 當時相對死と云つた情死. So the source itself performs a gloss, and the English can perform the same gloss with the same structure. Rejected: shinjū (the word an English reader might actually know, and the wrong register — it is the theatrical word); "double suicide" (accurate and flat, but it kills the two-term structure Ōgai built).

B11. 入相の鐘 → "the vesper bell." Unresolved at the moment of writing and written anyway, per R06 §2. 入相の鐘 is the temple bell at sunset. "Vesper bell" is the settled English literary phrase and is Christian monastic vocabulary placed on a Buddhist temple. Alternative live and recorded: "the evening bell", which is neutral and loses the fixed-phrase quality. I wrote the marked one and flagged it. See the R04 log, where the revision returned to this.**

B12. Tense: the source shifts into the present and I followed it. Span A runs in past narration (であつた) and then, in paragraph 3, moves to the present: 語り合ふ … 繰言である. Ōgai's frame slides from historical report into a habitual present and back out. English can do this and it reads as a stumble unless it is committed to, so I committed: "talk over their lives the whole night through. Always, always it is the same going over…" then back to "The constable … was able to learn." Cost: an English reader may take it for an error. This is a style-correspondence purchase paid for in naturalness.

B13. 白洲 → "the white sand of the magistrate's court."** The gravelled area where the accused knelt to be examined. Options: transcription (opaque and unearned — unlike 同心 the word recurs once), "the courtyard" (loses the whole image), "the white sand" alone (meaningless in English), the chosen expansion. The expansion is four words where the source has two characters, and it is the kind of in-text glossing that accumulates into a translation that explains itself.

B14. 口供 / 口書 kept distinct: "the formal deposition" / "the written statement."** 口供 is the oral statement given at examination; 口書 the written record of it. English has no settled pair and could level both to "testimony". Held, because the sentence's whole point is the difference between hearing and reading — the official does one or the other and learns nothing either way.

B15. 眷族 absorbed.** 親戚眷族 is a two-word compound where both halves mean kin. English has no idiomatic doublet here that is not archaic ("kith and kin" is wrong in sense and colour). Rendered "the kin and household". Recorded as a loss: the source's formulaic weight is gone.

B16. 不覺の涙 → "the tears that escaped him."** 不覺 = unawares, against one's will, the tears that get past you. "Involuntary tears" is right and clinical; "unbidden tears" is right and poetic; the relative clause is plainer than either and keeps the agency where the source has it, in the tears.

B17. 桁を違へて → "shifted the decimal place." 桁 is the digit-place; 桁を違へて考へる is to think at a different order of magnitude — Shōbei is trying to make Kisuke's 200 mon commensurable with his own household economy by rescaling. Written as "shifted the decimal place", which is exact and anachronistically mathematical for 1916 and for a Tokugawa constable. Alternatives live: "however he changed the scale of it", "shift the order of magnitude". Left as written per R06 §2, and flagged. See the R04 log.**

B18. 扶持米 → "his stipend rice."** The rice allowance a retainer was paid in. Not "salary" (erases the rice, which is the point of the household's tightness), not "rice stipend" (which I nearly wrote; "stipend rice" keeps the noun order of the compound and reads as a term rather than a description).

B19. 手一ぱいの生活 → "a life lived to the last inch."** Idiom for hand-to-mouth, at full stretch. Rejected: "hand-to-mouth" (implies not balancing, and the previous sentence has just said the accounts do balance), "a life lived to the hilt" (means the opposite in English).

B20. 意識の閾 → "the threshold of consciousness," and this is the clearest loss in span B. Ōgai is importing Herbart's Schwelle des Bewusstseins into 1916 Japanese, where it reads as a foreign scientific term dropped into an interior monologue — the sentence changes texture around it. The English phrase is native, standard and unmarked: the same words that are an import in Japanese are ordinary psychology in English. There is nothing to buy the markedness back with; italics would mark it as foreign to English, which is exactly wrong. Taken as a loss and left plain.**

B21. 譃 → "But that is a lie," and the whole paragraph turns on it. Shōbei proposes the easy explanation (Kisuke has no dependants) and rejects it in three characters. The stored source text had this word missing until this session (RS-20260728h), reading 「しかしそれはである」. Had I translated the stored text I would have had to invent a predicate. Options once the word was restored: "But that is a lie" (blunt, and the source is blunt), "But that will not do" (softens a self-accusation into a demurral), "But it is false". Chose the blunt one; the sharpness is Shōbei catching himself**, not an argument.

B22. 係累 → "no one depending on him." 係累 is the encumbrance of dependants. "No ties" is the idiom and is wrong in emphasis: Kisuke has no ties in the sense of nobody to keep, which is what the surrounding argument needs. Rejected also "no encumbrances" (period-legal colour).

B23. The four-step 蓄 cadence, and I kept the repetition English usually trims. 病がなかつたら／食つて行かれたら／少しでも蓄があつたら／もつと多かつたら — four conditionals in the same frame, each "if only". English style advice would vary the verb. Held all four on "he thinks, if only…", with the fourth marked by "again". Cost: the paragraph is more insistent in English than it is in Japanese, because English tolerates exact repetition worse.

B24. 蓄 → "laid by."** Not "savings" (a modern financial noun that makes Shōbei's thought about an instrument rather than a hoard), not "a store" (agricultural). "Laid by" is verbal and period-neutral and carries the four repetitions without becoming a term.

B25. 踏み止まる → "halt." Literally to plant the feet and not go further. "Stop" is too weak, "draw the line" imports a boundary Shōbei is not drawing, "hold back" implies restraint against desire, which is nearer but adds effort — the point is that Kisuke is simply standing there. "Halt" twice, once in the general sentence and once in "the one now halting before his eyes."

B26. 毫光 → "a light was streaming." 毫光 is specifically the ray from the 白毫, the tuft between a Buddha's brows: an iconographic term. Options: "a halo" (the Western iconographic word, instantly legible, and it puts a Christian saint's attribute on a Buddhist image); "an aureole" (same problem, plus register); "a light was streaming" (the plainest, loses the Buddhist term entirely and keeps only the phenomenon). Wrote the plain one. Cost, and it is the mirror of D3:** here I refused the naturalising word and got nothing back for it, because English has no Buddhist iconographic vocabulary in common use.

B27. 睜つて → "opened his eyes wide."** The restored character (RS-20260728h). The verb is emphatic — to stare wide-eyed. The stored text read 「驚異の目をつて」, a fragment. "Opened his eyes wide upon Kisuke, in wonder" splits 驚異の目 into the action and the affect, which the Japanese fuses into one noun phrase.

B28. 入相の鐘: "the vesper bell" → "the evening bell", and the reason came from the other span. The draft (D11) wrote the marked English fixed phrase and flagged it. Revising with both spans in front of me, the objection sharpened from "Christian word on a Buddhist bell" to something checkable: span B ends on 毫光, a Buddhist iconographic term I had already refused to naturalise (D26). Opening the frame with a Christian canonical hour and closing the reflection with an un-naturalised Buddhist ray sets the two spans pulling in opposite directions on the same axis. Cost, and it is real: 入相の鐘 is a fixed poetic phrase in Japanese, and "the evening bell" is not fixed in English, so the correspondence of stock-phrase to stock-phrase is what I gave up to buy the consistency. This decision was not available to the draft**, which wrote span A before span B existed.

B29. 桁を違へて: "shifted the decimal place" → "allowed for the difference in scale". D17 wrote the anachronism and flagged it: decimal-place notation is not available to a Tokugawa constable or to 1916 Japanese, and 桁 is the abacus column. Rejected on revision: "shifted the decimal point" (same fault, worse), "juggled the figures" (breezy, and Shōbei is not juggling). Loss:** the concreteness of 桁 — a physical column of beads — is gone, and "difference in scale" is an abstraction where the source has an object.

B30. 同心 → "constable": the draft's internal inconsistency, examined and resolved in favour of the draft. D3 chose "constable"; D5 rejected "gaol" for 牢屋敷 on the ground that British-period colour is barred by D1. Those two are in tension and the draft did not notice. Resolved by a distinction the draft could not make: "constable" is the received term in English-language Tokugawa scholarship** for 同心, i.e. a technical rendering, whereas "gaol" is a spelling with regional colour and no technical content. Kept. Recorded because it is a decision the draft made without knowing it had made one.

B31. 毫光: "a light was streaming" → "light was streaming".** The indefinite article makes it a discrete object — a lamp, a beam — where 毫光 is radiance. One word deleted; the smallest change in the revision and the one I am least able to argue for.

B32. 多數を占めてゐた: "made up the greater number" → "made up the greater part of them".** The draft's phrase floats without a referent; the source's subject is the class of banished criminals just named.

B33. 窺ふことの出來ぬ境遇である: "could not glimpse" → "cannot glimpse", restoring a tense the draft flattened. The source's sentence ends in the present** (である), completing the slide into habitual present that D12 identified and committed to. The draft followed the slide for two clauses and then dropped back into the past for the last one. Revising against the source caught it. This is the clearest case in the pair of a revision repairing a fault the draft's own logged principle should have prevented.

B34. 頃: "at the hour when" → "at about the hour when".** 頃 is approximate and the draft made it exact.

B35. Span B's tense: the draft flattened it systematically and the revision let the flattening stand, on purpose. Paragraph 25 runs in the Japanese present throughout (出納が合つてゐる, 生活である, 過してゐる, 潜んでゐて, 頭を擡げて來るのである) and the draft rendered all of it in the past. So the draft followed a tense shift where it was local and marked (span A, D12/D33) and normalised one where it was systematic (span B), without noticing that it had done two different things. Kept the past, because English free indirect discourse in a past-tense narrative conventionally uses past for exactly this ongoing-state material, and the Japanese present here is the corresponding target-language convention rather than a marked effect. What the revision actually produced was not a change but the knowledge that a decision had been made**, which is the honest description of most of this pass.

B36. Paragraph 26's present tense, kept, and now with a reason.** 「…なられさうにない」 stands in the present in a paragraph closed by an explicit quotative 「と、庄兵衞は思つた」, so it is quoted thought and licensed in a way paragraph 25's present is not. The draft rendered 26 in the present and 25 in the past and was right by accident; D35 and D36 together are the reason.

B37. 同心を勤める人にも: "Men who serve as constables have all sorts of natures" → "Those who serve as constables are, like other men, of various dispositions". The draft dropped the も, which is the sentence's hinge: constables too are various, i.e. do not become uniform by holding the office. Rejected: "even those who serve as constables" (over-marks it).

B38. 人の一生: "a human life" → "the whole of a human life".** 一生 is the span, not the fact. The paragraph is about how far a man goes before he stops, so the whole-life sense is load-bearing.

B39. 只漠然と: "Vaguely" → "In a vague way".** Fronted adverb in one word reads as a comment on the prose; the phrase reads as a description of the thinking.

B40. 大目に見る → "looking the other way": examined and kept.** The live alternative was "turning a blind eye", which is a fixed English idiom and would have answered D28's stock-phrase-for-stock-phrase argument. Rejected because it carries a specific attribution of wilful concealment; 大目に見る is indulgence, not concealment. Recorded because it is the one place where D28's own principle pointed one way and I went the other.

B41. Consistency sweep on the realia, all held: 遠島 "banishment to the islands" (3×), 罪人 "criminal" (6×), 科 "offence" (3×), 同心 "constable" (5× after the one glossed first use).** No instance was varied for English elegance. The one term I checked and could not make consistent is in D42.

B42. 護送 / 宰領, the draft's open question, closed as unresolvable from these spans. D-final of the R06 log flagged that the source uses 護送 (paragraph 5, and again in the compound 護送の役) and 宰領 (paragraph 4) for what looks like the same office. Checked against the whole work: 宰領 occurs once, 護送 three times, and one instance is too few to establish a contrast. Rendered "taken in charge by" and "escort duty" respectively — different English for different Japanese, with no claim that the difference means anything. Left open rather than levelled** [comparison to a decision in the other log redacted for blinding].

C15., C16. and C17. below are three further candidate recommendations, numbered continuously with C1-C14 above.

C15. At a culture-bound item, four ordered tests decide the handling; apply the outcome to the whole class senses: cultural-mediation, style-correspondence, consistency evidenced on: JA→EN, RU→EN full statement: A culture-bound item is a word or phrase naming something the source culture has and the target culture does not have under the same description. At such an item, do not choose among the eight handlings by feel. Apply these four tests in order and take the first that fits. (1) Is the item's sense already stated at the site by the source itself — an apposition, a relative clause, a repeated morpheme that is itself translatable? Then retain the item (transliterate a name, keep the term) and translate the source's own explanatory material. Add no gloss of your own and do not substitute; the sense arrives at no cost to you. (2) Is the item a quantity, and is its value rather than its cultural identity what the sentence uses? Then convert the measure into the target's system and do nothing else. (3) Does the target already hold an established borrowing of the item, or an exact equivalent from a practice the two cultures share? Then use it; no fork arises. (4) Otherwise ask whether the item is load-bearing for the passage's argument or plot, or whether it is furniture. Load-bearing: retain the item and scaffold it — manufacture its sense from the surrounding context in your own words — and do not substitute a target-culture near-equivalent, however ready one is. Furniture: retain it copy-opaque and spend no gloss on it, or omit it where the sentence does not use it. Standing constraint: whichever handling a test returns for one member of a class of items, apply that handling to every member of that class in the same text.

C16. At a culture-bound item, three ordered tests decide the handling; apply the outcome by sentence position senses: cultural-mediation, style-correspondence, consistency evidenced on: JA→EN, RU→EN full statement: A culture-bound item is a word or phrase naming something the source culture has and the target culture does not have under the same description. At such an item, do not choose among the eight handlings by feel. Apply these three tests in order and take the first that fits. (1) Is the item the first culture-bound item in its paragraph? Then retain it copy-opaque, so that the paragraph's foreign texture is established at its head and the reader is oriented before any mediation is offered. (2) Is the item a noun of three or more syllables in the source, or a proper name? Then gloss it — keep the source form and supply a short in-line explanation. (3) Otherwise substitute the nearest target-culture item. Standing constraint: where two items of the same class fall in one sentence, handle the earlier by test 1 and the later by test 3, so that the class's range is visible to the reader rather than flattened.

C17. Where more than one rendering is live, take the shorter one senses: naturalness, style-correspondence evidenced on: all full statement: At any decision where the translator has more than one candidate rendering in view, count the English words in each candidate and write the one with the fewest words. Where two candidates have the same number of words, count the characters and write the shorter. Where they are the same length in both, write the one whose first word comes earlier in alphabetical order. Apply this at every such decision without exception.