N1 2010-07 — audit report

71 questions (問題1–13, the non-listening half). Generated from the OCR and correction audit trails.

問題13 (the コンクール information-retrieval task) is published officially in the 日本語能力試験公式問題集 第一集 (2012); its text and answers are confirmed against the official version. The rest of this sitting rests on re-typesets.

OCR confidence

Source: retypeset-digital — quality 5/5: Crisp retypeset digital text, fully legible at 1782x2400. Header watermarks (日语轻松考 / 祝亲考级 通过~通过~通过~ / 'Yuuki Bùi' signature on page 2) sit in margins and do not overlap body text. Ads on pages 1, 14, 15. No furigana. Being a third-party retypeset, transcription typos are possible (e.g. kana substitutions in listening options).

Every section was transcribed twice independently (a full-page pass and a double-resolution half-page pass) and reconciled. Of 13 sections, 1 agreed on all content between the two passes (reconciled automatically, differing only in transcriber notes) and 12 had at least one content difference resolved by a third adjudication pass that re-read the page images. 63 leaf-level differences were examined in total.

Residual OCR risks (27):

• 問題1: no human/agent spot-check was performed for this section (A/B byte-agreement on content was treated as sufficient)
• 問題2: None significant. The print inserts layout spacing around the （　　） blanks (e.g. 課長が （　　） してくれた); both transcribers dropped this layout spacing consistently per the blank convention, so no information is lost.
• 問題3: Q16: the source retypeset underlines 張り合っている (～ている form) while all four options are plain て-forms (競争して etc.); this mismatch is faithfully transcribed but may differ from the original official exam's underline extent.
• 問題4: Questions 22 and 24 were verified against the page images at standard resolution only, not at the 4x zoom used for the spot checks; a subtle one-character correlated error there is unlikely (crisp digital retypeset) but not formally excluded.
• 問題4: The source is a third-party retypeset, so even a perfect transcription may differ from the original 2010-07 exam wording.
• 問題5: Q28 「A会社の元社員」 reads oddly (A社 would be more natural); kept as printed per conventions but it may be a retypeset artifact of the original exam wording.
• 問題5: The source is a retypeset, so even perfectly transcribed text may silently differ from the official exam; nothing further can be verified from these images alone.
• 問題6: Q40 may diverge from the official 2010-07 exam: the retypeset prints 『りかい』 (hiragana) and 『わかる』, while one transcriber reports circulated official copies have 『うかる』/『合格する』. Transcribed as printed per convention; worth cross-checking when establishing the answer key, since the answer-key source may correspond to the official wording.
• 問題6: Q40's ★ in the second slot is unusual for this question type; it is clearly printed that way, but if the retypeset altered the sentence (see above) the slot position could also differ from the official exam.
• 問題7: The source is a third-party retypeset; 友たち (paragraph 3) and the past-tense 親しみやすくさえあった (question 42 option 2) are suspected deviations from the original exam text but are transcribed exactly as printed, so the final text may differ from the authentic 2010-07 paper.
• 問題7: All correctAnswerIndex values remain null by convention; no answer-key information was adjudicated here.
• 問題8: The exact codepoint of the printed title dash is unrecoverable from pixels; － (U+FF0D) was chosen because the ink is measurably shorter than 一/ー in the same lines, but some other short-dash glyph from the Chinese retype font cannot be fully excluded.
• 問題8: The URL is letterspaced monospace, so its spacing is partly a normalization judgment; the kept space after www. reflects a measured extra empty cell, but the original exam URL almost certainly had no space (and plausibly lowercase furniture-miki).
• 問題8: 荒野健彦 and 「ちゃんねる」 (without the 2) are faithfully transcribed retypeset readings that likely diverge from the original exam text.
• 問題9: The 注-marker parentheses in passage (2) are printed half-width in this Chinese retypeset; the final uses full-width （注） per the punctuation convention. If the project later decides printed paren width should be preserved, this is the one spot to revisit.
• 問題9: The PDF has no usable text layer (extract-text.py returns nothing for these pages), so all adjudication rests on the renders; the renders are crisp digital type, so confidence is high.
• 問題9: Passage (2) paragraph breaks around その原因は…。 were agreed by both transcribers and look correct in the renders, but dense layout means paragraph segmentation was verified visually only.
• 問題10: The colon in the 注 glosses was transcribed as full-width ：; at the available resolution a half-width colon with surrounding spaces cannot be 100% excluded, but full-width is consistent with the retypeset's punctuation elsewhere.
• 問題10: The source is a Chinese retypeset, so spellings like ご存知 may differ from the original JLPT printing; transcribed as printed per convention.
• 問題11: Long stretches of passage body text (passage B paragraphs 2-3, passage A paragraph 2 interior) were verified only at moderate zoom; a correlated typo shared by both transcribers in an unchecked stretch remains possible.
• 問題11: The half-width-space rendering of （ ① ） is a judgment call on a printed gap of ~0.55 em; the stated conventions cover only empty blanks, so a different adjudicator could defensibly choose full-width.
• 問題11: The 注 glosses are placed at the end of passageB per printed location even though both glossed words occur in passage A; consumers rendering passages separately should be aware the glosses for passage A's vocabulary live in the passageB field.
• 問題12: A and B content fields are byte-identical, so correlated transcriber errors were the main risk; the full passage and all four questions were re-read from the images, leaving little residual exposure, though single-character slips in unchecked whitespace-level details (e.g., exact full-width punctuation variants like ： in the 注 glosses) remain conceivable.
• 問題12: The retypeset itself contains likely typos (賞励金, 情報供与, 脅される, 続け高い without て) which are transcribed as printed; if the original exam text is ever consulted, these spots may differ from the true original.
• 問題13: The parentheses around the URL (http://www.kiyomori-shi.jp) were transcribed as full-width （）; at the available resolution full-width vs half-width parentheses around Latin text is hard to distinguish, but both transcribers agreed and the glyph spacing in p12-b.png is consistent with full-width.
• 問題13: The 賞 label line and the table are transcribed as a separate <p>賞</p> followed by <table>; the printed table sits indented to the right of the 賞 label, so an alternative reading could attach the label differently, but the agreed representation is faithful to the layout conventions.
• 問題13: Whether the gap between a 応募規定 item number and its text is exactly one full-width space is a judgment call on column-aligned source layout; both transcribers agreed on one full-width space per the normalization convention.

Source-text quirks the OCR transcribed verbatim (printed typos/recall artifacts; the ones judged to be retype errors were corrected in the layer below) — 37 noted at OCR time:

• 問題3: Question 15: verified at high zoom that the printed underline covers なじんで (its right edge extends fully under で, ending before き), matching the て-form options.
• 問題3: Question 16: the printed underline covers all of 張り合っている (including ている), but the four options are plain て-forms (競争して etc.) rather than ～している; possibly a retypeset inconsistency. Transcribed as printed; confirmed against the high-zoom crop.
• 問題4: Question 20 is the only one with a two-column option layout (options 1 and 2 on the first printed line, 3 and 4 on the second); the option-number mapping was verified against the printed numbers. Questions 21-25 have one option per line.
• 問題5: Q28: stem printed as 「Ａ会社の元社員が１１日」 with full-width Ａ and １１, normalized to half-width per conventions (A会社, 11日); 「A会社」 is slightly unusual phrasing (more commonly A社) and may be a retypeset artifact — transcribed as printed.
• 問題5: Q29: option 4 is printed 「申しあげたく」 with あげ in kana, in contrast to option 3 「差し上げたく」 with kanji — transcribed as printed.
• 問題6: Question 40: the ★ is printed in the second of the four slots (『わかる』 ＿ ★ ＿ ＿), unlike questions 36-39 where it sits in the third slot. Verified at high zoom by both transcribers and the adjudicator; transcribed as printed.
• 問題6: Question 40: ジョン asks about the word printed in hiragana as 『りかい』 and アリ answers with 『わかる』. Verified glyph-by-glyph at high zoom and transcribed as printed, but flagged: one transcriber reports widely-circulated copies of the official 2010-07 exam have 『うかる』/『合格する』 here, so this may be a retypeset alteration or hiragana substitution.
• 問題6: Adjudication: A and B agreed on all content fields; the only diffs were note wording. All five stems and all twenty answer options were re-verified character-by-character against the double-resolution renders (p04-b.png, p05-t.png).
• 問題7: Paragraph 3 prints 「友たちのような父親」 (友たち, missing the dakuten) while paragraph 4 prints 「友だちのような父親」 (友だち) — almost certainly a retypeset typo; transcribed as printed (verified at high zoom by both transcribers and in adjudication).
• 問題7: Question 42 option 2 clearly prints past-tense 「親しみやすくさえあった」, which sits oddly against the present-tense passage and could be a retypeset alteration; transcribed as printed (verified at high zoom).
• 問題7: Cloze blanks are printed in the passage and instruction as （41）-style full-width parenthesized numbers, not boxed; converted to 【41】 markers inside the passage per convention, kept as printed in the instruction.
• 問題8: Passages (2) and (4): the retypeset prints note markers （注1） etc. immediately BEFORE the glossed expressions (e.g. 自分がどこに（注2）アクセスしているか、巨大（注3）匿名掲示板（注4）「ちゃんねる」), unlike official JLPT superscript-after typesetting; transcribed as printed.
• 問題8: Passage (2): the bulletin board is printed as 「ちゃんねる」 with no leading digit, both in the passage and in the （注4） gloss; the original exam very likely had 「2ちゃんねる」, so this looks like a retypeset omission. Transcribed as printed.
• 問題8: Passage (2) citation second author prints as 荒野健彦 (verified at high zoom); unusual name, possibly a retypeset corruption of the original credit. Transcribed as printed.
• 問題8: Passage (3): the URL is printed in spaced monospace with a full-width colon (http : //www. Furnituremiki.co.jp/); normalized to contiguous ASCII except that the gap after www. is one full empty monospace cell (pixel-measured 24px vs 10-12px after the other periods), so a genuine printed space is kept: http://www. Furnituremiki.co.jp/. The capital F and missing hyphen are as printed (original was plausibly furniture-miki).
• 問題8: Passage (4): 「過激な競争に勝ち抜けません」 is as printed (過激, not 過酷/過当) — slightly unusual collocation, possibly a retype quirk. Also そのバランスを測り uses 測り as printed.
• 問題9: Question 57 stem is printed 「どのよう役割」 (missing な; likely retype error for 「どのような役割」) — transcribed as printed; confirmed against the image at high zoom.
• 問題9: Question 57 option 3 is printed 「着るの人柄や個性を引き出すことで」 (likely retype error for 「着る人の人柄」) — transcribed as printed; confirmed against the image at high zoom.
• 問題9: Passage (1) author is printed as 斉藤茂太 (the simplified 斉; commonly written 斎藤茂太); transcribed as printed.
• 問題9: Passage (1) sets 修業 in full-width curly double quotes: “修業”; kept as printed.
• 問題9: Passage (3) prints 「正しいか。正しいくないと、言うべきです」 — 正しいくない is a suspected retype error for 正しくない; the extra い was confirmed against the image at high zoom and is transcribed as printed.
• 問題9: Passage (3) final sentence reads 「それなのにそれを消費者と呼ぶのは」; 消費者 might be a retype error for 消費, but it is clearly printed 消費者 — transcribed as printed.
• 問題10: Passage prints ご存知 (not ご存じ) in まだ感情機能をもつに至っていないのはご存知のとおりです; transcribed as printed.
• 問題10: Options for questions 61 and 62 are printed two per line (1/2 on the first line, 3/4 on the second); mapping verified against the printed option numbers.
• 問題10: The passage uses なにかを in paragraph 3 (hiragana) while the 注2 gloss uses 何かを (kanji); both transcribed as printed.
• 問題11: Question 63 option 4 is printed as 学歴は実在と関係がある。 — 実在 is suspicious here (実力 would make more sense, and option 2 already uses 実在する) but it is clearly printed; verified against the double-resolution image.
• 問題11: Question 65 option 1 ② part is printed as 学歴と言う区分も大切だ with 言う in kanji, unusual for this grammaticalized use (passage B itself prints 学歴による区分といった方式 in kana); likely retypeset orthography quirk, transcribed as printed and verified against the image.
• 問題11: Passage A paragraph 2 prints 行なわれている (okurigana な) and paragraph 3 prints 変わりない (not 変わらない); both as printed and image-verified.
• 問題11: Passage B paragraph 2 prints まだ解る with kanji 解; as printed.
• 問題12: Passage paragraph 2: printed インセンティブ（賞励金） — 賞励金 is a suspicious retypeset spelling (standard would be 奨励金 or 報奨金); transcribed as printed.
• 問題12: Passage paragraph 5: printed 「情報供与」 — breaks the 〜給与 naming pattern of the surrounding coined terms (出会い給与, 技磨き給与, ふれあい給与, おそば給与), possibly a retypeset typo for 「情報給与」; transcribed as printed.
• 問題12: Passage paragraph 8: printed 生活が脅されるほどの額 — original text likely 脅かされる; transcribed as printed.
• 問題12: Question 69 option 3: printed アルバイトを続け高い技術や人脈を得る — possibly missing て (続けて高い); transcribed as printed.
• 問題12: The source citation （松永真理『シゴトのココロ』による） is printed before the four 注 glosses; the printed order is preserved (citation <p> precedes the 注 <p>s). Verified against the page image.
• 問題13: The instruction says 下のページは…, even though the document is printed below on the same page; the original exam presumably said 右のページ. Transcribed as printed.
• 問題13: In 応募規定, the line 写真部門に出品する写真は… is printed with no item number, separated from item 2 by a blank printed line; the original exam likely numbered it 3, so this looks like a dropped number in the retypeset. Transcribed unnumbered as printed; the blank line is not represented.
• 問題13: 表彰式 line: （予定）is immediately followed by 最優秀賞 with no space or punctuation, as printed.

Corrections applied (33)

Evidence-based reconstructions of the original exam text. This section is the canonical correction log for the sitting.

• 問3 q14 question: 「<u>ルーズ</u>なところ」 → 「<u>ルーズな</u>ところ」
  • The correct choice だらしない is already adnominal; leaving な outside the tested span produces the ungrammatical replacement だらしないなところ. The full target is ルーズな.
  • evidence: structural: 問題3 choices replace the underlined span.
• 問7 q41 passage: 「「友たちのような父親」である」 → 「「友だちのような父親」である」
  • 友たち is a dropped-dakuten artifact for 友だち. The same passage prints 「友だちのような父親」 for the identical concept later, so the two cannot both be original; 友だち is the only standard form (confirmed in published Japanese via Google Books). Clear internal inconsistency makes 友たち the corruption.
  • evidence: Internal inconsistency confirmed in output/N1_2010-07.json: mondai7.passage contains both 「友たちのような父親」 and 「友だちのような父親」 for the same referent. Google Books shows 友だちのような父親 as the standard published form. The jlpt-freq witness reproduces the same self-contradiction (line 196 友たち, line 200 友だち), confirming a shared retype artifact rather than intentional spelling. [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
• 問9 q57 question: 「衣服にどのよう役割があると」 → 「衣服にどのような役割があると」
  • どのよう modifying 役割 is ungrammatical; the adnominal requires な (どのような役割). A dropped-な retype artifact.
  • evidence: Grammar: どのよう before a noun requires な. The jlpt-freq witness (line 403) prints 衣服にどのよ う役割 with a space where a character was dropped, directly supporting a lost な between よう and 役割. [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
• 問9 q56 passage: 「正しいか。正しいくないと、言うべきです」 → 「正しいか。正しくないと、言うべきです」
  • 正しいくない is an impossible non-word; i-adjective negation drops the final い (正しい → 正しくない). This appears in the author's own assertive prose (…と、言うべきです), which must be grammatical, so it is a retype artifact inserting an extra い, not exam content.
  • evidence: Grammar: i-adjective negation is 正しくない, never 正しいくない. The jlpt-freq witness (line 390) reproduces the same 正しいくないと error, indicating a shared upstream retype artifact of a clear non-word. [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
• 問9 q57 answers[2]: 「着るの人柄」 → 「着る人の人柄」
  • 「着るの人柄」is ungrammatical; intended 「着る人の人柄」(the wearer's character). A 問9 reading-comprehension option, not a 問題4 distractor, so the awkwardness is a retype error rather than deliberate.
  • evidence: Linguistic: 着る+の+人柄 is not a valid construction. Lineage-shared retype typo, uncontradicted by any witness (all copies derive from the same retypeset). [manual retype-error pass, 2026-06-20]
• 問12 q66 passage: 「賞励金」 → 「奨励金」
  • 賞励金 is not a word; the gloss for インセンティブ is 奨励金 (奨励＝encouragement). A 賞→奨 single-character retype error.
  • evidence: Linguistic: 奨励金 is the standard term; 賞励金 has no dictionary entry. Lineage-shared, uncontradicted by witnesses. [manual retype-error pass, 2026-06-20]
• 問12 q66 passage: 「情報供与」 → 「情報給与」
  • The passage coins a series of terms ending in 〜給与 (技磨き給与, ふれあい給与, おそば給与); 情報供与 breaks the pattern. A 給→供 retype error; intended 情報給与.
  • evidence: Internal pattern: the surrounding coined terms all end in 給与. [manual retype-error pass, 2026-06-20]
• 問12 q66 passage: 「生活が脅される」 → 「生活が脅かされる」
  • 生活を脅かす (to endanger one's livelihood) is the idiom; 脅す (to coerce/threaten a person) does not take 生活 as subject. 脅かされる is the intended passive.
  • evidence: Linguistic/collocational. Lineage-shared, uncontradicted by witnesses. [manual retype-error pass, 2026-06-20]
• 問3 q16 question: 「<u>張り合っている</u>」 → 「<u>張り合って</u>いる」
  • 問題3 tests the underlined word against the four options, which are bare て-forms (競争して, 応援して, …). For the substitution to be grammatical (お互いに競争している), the tested span is 張り合って with いる outside; the retype extended the underline over ている.
  • evidence: Question logic: the options are て-forms, so the underline must be 張り合って, not 張り合っている. [manual retype-error pass, 2026-06-20]
• 問8 q47 passage: 「ちゃんねる」 → 「2ちゃんねる」
  • 巨大匿名掲示板「ちゃんねる」refers to 2ちゃんねる (2channel), the well-known huge anonymous BBS; the retype dropped the leading 2. Applied to both mentions.
  • evidence: Real-world referent: 巨大匿名掲示板 = 2ちゃんねる. [manual retype-error pass, 2026-06-20]
• 問8 q49 passage: 「バランスを測り」 → 「バランスを図り」
  • バランスを図る (to strike a balance) is the idiom; バランスを測る is unidiomatic with …試行錯誤を続ける. はかる has several kanji (図/測/計/量) and the retype chose the wrong one.
  • evidence: Homophone-kanji (はかる) error; the idiom is バランスを図る. [manual retype-error pass, 2026-06-20]
• 問9 q57 passage: 「それを消費者と呼ぶ」 → 「それを消費と呼ぶ」
  • The sentence contrasts 生産 (production) with this term (それなのに…と呼ぶ); 生産⇔消費 is the parallel. それ refers to the act, so 消費者 (a person) does not fit — 者 is a retype addition.
  • evidence: Semantic parallel 生産/消費; それ (the act) cannot be a 消費者. [manual retype-error pass, 2026-06-20]
• 問11 q65 answers[0]: 「学歴と言う」 → 「学歴という」
  • The grammaticalized 〜という (the category called 'academic background') is conventionally kana; passage B of the same item prints といった in kana. 言う is a retype kanji-conversion.
  • evidence: Internal parallel: passage B uses kana といった for the same construction. [manual retype-error pass, 2026-06-20]
• 問13 q70 passage: 「<p>写真部門に出品する写真は」 → 「<p>3　写真部門に出品する写真は」
  • The 応募規定 list runs 1 (清森市内…), 2 (絵画部門…), then 写真部門… with no number; the retype dropped item number 3. Restored to match the sequence.
  • evidence: List sequence 1, 2, [missing] → the 写真部門 item is 3. [manual retype-error pass, 2026-06-20]
• 問8 q0 passage: 「（注3）匿名掲示板（」 → 「匿名掲示板（注3）（」
  • （注N）note marker repositioned to follow its glossed term (the retype placed the marker before the word).
  • evidence: Marker-only move; the marker-stripped text is byte-identical. [note-marker relocation pass, 2026-06-20]
• 問8 q0 passage: 「（注4）「2ちゃんねる」な」 → 「「2ちゃんねる」（注4）な」
  • （注N）note marker repositioned to follow its glossed term (the retype placed the marker before the word).
  • evidence: Marker-only move; the marker-stripped text is byte-identical. [note-marker relocation pass, 2026-06-20]
• 問8 q0 passage: 「（注1）琴線に触れず、」 → 「琴線に触れず（注1）、」
  • （注N）note marker repositioned to follow its glossed term (the retype placed the marker before the word).
  • evidence: Marker-only move; the marker-stripped text is byte-identical. [note-marker relocation pass, 2026-06-20]
• 問8 q0 passage: 「（注2）場当たり的なネーミングで」 → 「場当たり的なネーミング（注2）で」
  • （注N）note marker repositioned to follow its glossed term (the retype placed the marker before the word).
  • evidence: Marker-only move; the marker-stripped text is byte-identical. [note-marker relocation pass, 2026-06-20]
• 問8 q0 passage: 「（注3）何はさておき自」 → 「何はさておき（注3）自」
  • （注N）note marker repositioned to follow its glossed term (the retype placed the marker before the word).
  • evidence: Marker-only move; the marker-stripped text is byte-identical. [note-marker relocation pass, 2026-06-20]
• 問9 q0 passage: 「（注2）“修業”」 → 「“修業”（注2）」
  • （注N）note marker repositioned to follow its glossed term (the retype placed the marker before the word).
  • evidence: Marker-only move; the marker-stripped text is byte-identical. [note-marker relocation pass, 2026-06-20]
• 問9 q0 passage: 「（注3）ステップアップするに」 → 「ステップアップする（注3）に」
  • （注N）note marker repositioned to follow its glossed term (the retype placed the marker before the word).
  • evidence: Marker-only move; the marker-stripped text is byte-identical. [note-marker relocation pass, 2026-06-20]
• 問9 q0 passage: 「（注）おしはかるこ」 → 「おしはかる（注）こ」
  • （注N）note marker repositioned to follow its glossed term (the retype placed the marker before the word).
  • evidence: Marker-only move; the marker-stripped text is byte-identical. [note-marker relocation pass, 2026-06-20]
• 問10 q0 passage: 「（注1）裏打ちして、」 → 「裏打ちして（注1）、」
  • （注N）note marker repositioned to follow its glossed term (the retype placed the marker before the word).
  • evidence: Marker-only move; the marker-stripped text is byte-identical. [note-marker relocation pass, 2026-06-20]
• 問10 q0 passage: 「（注2）駆り立てる動」 → 「駆り立てる（注2）動」
  • （注N）note marker repositioned to follow its glossed term (the retype placed the marker before the word).
  • evidence: Marker-only move; the marker-stripped text is byte-identical. [note-marker relocation pass, 2026-06-20]
• 問10 q0 passage: 「（注3）計算づくで結」 → 「計算づくで（注3）結」
  • （注N）note marker repositioned to follow its glossed term (the retype placed the marker before the word).
  • evidence: Marker-only move; the marker-stripped text is byte-identical. [note-marker relocation pass, 2026-06-20]
• 問12 q0 passage: 「（注1）ウォッチできる「」 → 「ウォッチできる（注1）「」
  • （注N）note marker repositioned to follow its glossed term (the retype placed the marker before the word).
  • evidence: Marker-only move; the marker-stripped text is byte-identical. [note-marker relocation pass, 2026-06-20]
• 問12 q0 passage: 「（注2）いかんともしがたいも」 → 「いかんともしがたい（注2）も」
  • （注N）note marker repositioned to follow its glossed term (the retype placed the marker before the word).
  • evidence: Marker-only move; the marker-stripped text is byte-identical. [note-marker relocation pass, 2026-06-20]
• 問12 q0 passage: 「（注3）ビッグな報」 → 「ビッグな（注3）報」
  • （注N）note marker repositioned to follow its glossed term (the retype placed the marker before the word).
  • evidence: Marker-only move; the marker-stripped text is byte-identical. [note-marker relocation pass, 2026-06-20]
• 問12 q0 passage: 「（注4）フリーに」 → 「フリー（注4）に」
  • （注N）note marker repositioned to follow its glossed term (the retype placed the marker before the word).
  • evidence: Marker-only move; the marker-stripped text is byte-identical. [note-marker relocation pass, 2026-06-20]
• 問8 q47 passage: 「<p>（注1）ネット社会の匿名性は完全に透明人間の心理です。」 → 「<p>ネット社会（注1）の匿名性は完全に透明人間の心理です。」
  • （注N）note marker repositioned to follow its glossed term (the retype placed the marker before the word).
  • evidence: Marker-only move; the marker-stripped text is byte-identical. [note-marker relocation pass, 2026-06-20]
• 問8 q47 passage: 「自分がどこに（注2）アクセスしているか」 → 「自分がどこにアクセス（注2）しているか」
  • （注N）note marker repositioned to follow its glossed term (the retype placed the marker before the word).
  • evidence: Marker-only move; the marker-stripped text is byte-identical. [note-marker relocation pass, 2026-06-20]
• 問11 q63 passageA: 「（注1）眉をひそめる人もいるかもしれません」 → 「眉をひそめる（注1）人もいるかもしれません」
  • （注N）note marker repositioned to follow its glossed term (the retype placed the marker before the word).
  • evidence: Marker-only move; the marker-stripped text is byte-identical. [note-marker relocation pass, 2026-06-20]
• 問11 q63 passageA: 「（注2）ほどほどでかまわない」 → 「ほどほど（注2）でかまわない」
  • （注N）note marker repositioned to follow its glossed term (the retype placed the marker before the word).
  • evidence: Marker-only move; the marker-stripped text is byte-identical. [note-marker relocation pass, 2026-06-20]