N1 2011-12 — audit report
70 questions (問題1–13, the non-listening half). Generated from the OCR and correction audit trails.
OCR confidence
Source: mixed — quality 3/5: Cleanly rendered retypeset Word-to-PDF text, fully legible with no scan noise — but the retype has a high typo density, roughly 3-8 errors per question page: kana confusions (あつ for たつ, ぞんな for そんな, にわ/でわ for には/では, だれる for される, ごる for ころ, もちった for もらった), katakana garbles (カワンクー for カウンター, アモン.デユール[]の『ウルフ.シデイ』ハ), dropped characters (ようや for ようやく), stray characters (け before an option), wrong 問題4 header text, and ☆ substituted for the official ★ in 問題6. These are dangerous for a readings/vocab test where exact kana matters, so OCR output needs cross-checking against official sources. The two answer-key pages are low-resolution scanned tables with a 佳禾网校 watermark and Word artifacts, but every cell is readable. Mixed full-width/half-width digits in question numbering throughout.
Every section was transcribed twice independently (a full-page pass and a double-resolution half-page pass) and reconciled. Of 13 sections, 4 agreed on all content between the two passes (reconciled automatically, differing only in transcriber notes) and 9 had at least one content difference resolved by a third adjudication pass that re-read the page images. 157 leaf-level differences were examined in total.
Residual OCR risks (34):
- 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: no human/agent spot-check was performed for this section (A/B byte-agreement on content was treated as sufficient)
- 問題3: no human/agent spot-check was performed for this section (A/B byte-agreement on content was treated as sufficient)
- 問題4: no human/agent spot-check was performed for this section (A/B byte-agreement on content was treated as sufficient)
- 問題5: The Q33 scene-label parenthesis width was the only A/B content disagreement; the half-width-vs-full-width call rests on reading 「(インタビューで)」 as half-width ASCII parens in the render (clearly narrower and tighter than the full-width ( ) blanks on the same page) plus the convention that Japanese punctuation is kept full-width. If the convention is instead read as strictly transcribe-as-printed for punctuation width, B's half-width form would be the correct output.
- 問題5: Q34 stem: the small 。 after 「辛いらしい」 renders slightly compressed at line-fit; transcribed as the ordinary full stop 。 by both transcribers, which the high-zoom crop supports.
- 問題5: The stray け before Q35 option 4's number is excluded from the option text by both transcribers (it precedes the option number); if downstream consumers want a strictly positional transcription, the け would need to be represented somehow, but there is no convention slot for inter-option debris.
- 問題5: This is a prep-school retype with a demonstrated typo density; typos that happen to form plausible Japanese (and were typed identically by both transcribers) would not surface in the diff and may survive undetected despite the six spot checks.
- 問題6: The instruction-line star treatment (<u>☆</u> as printed vs normalizing to ★) is a convention judgment call; the slot-run ★ convention was read as applying only to in-stem slot runs. If downstream tooling expects a literal ★ in every mondai6 instruction, this section will differ.
- 問題6: Q39's option set (もの/増税/のない/避けよう) is hard to assemble into a fully grammatical sentence even allowing for the missing に; the retype may have corrupted one of the options themselves, but the print is transcribed faithfully.
- 問題6: Q40's 客観的して is ungrammatical as printed; the intended original (客観視する or 客観的にする) cannot be recovered from this source.
- 問題6: This is a re-typed prep-school document, so even characters both transcribers and the adjudicator agree on may differ from the genuine exam paper.
- 問題7: The exact codepoint of the fallback-font glyph after デユール is uncertain; ∐ (U+2210) is taken from the PDF text layer and matches the rendered serif glyph, but the prep school may have intended Ⅱ (U+2161).
- 問題7: The 28 leading full-width spaces before 有栖川有栖 were dropped as paragraph-leading indent per convention, losing the printed right-alignment (noted in notes); the 12 spaces before 草々 were kept since they are line-internal.
- 問題7: This retype is extremely typo-dense; the transcription is faithful to the retype, not to the original 2011-12 exam wording.
- 問題8: Comma-glyph classification (half-width ASCII , vs full-width ,vs 、) relies on visual judgment of small low marks; A's notes state these were also cross-checked against the PDF text layer, which I treated as authoritative where the rendered glyph was ambiguous.
- 問題8: Several retype typos (にようる, スリーン, おかけいますが, ステースタ, よな, doubled 、, duplicated 上手に) are preserved verbatim per the transcribe-exactly convention; a downstream consumer expecting clean text should be aware these are deliberate.
- 問題8: B consistently 'corrected'/normalized typos and commas, so its disagreements were treated as convention violations rather than alternative readings; every content diff resolved to A and was independently confirmed against the high-res images.
- 問題9: Q55 option 1 contains what looks like a full-width comma 「,」after「使用していやが」in the scan rather than the usual「、」; per convention Japanese punctuation stays full-width and both transcribers read「、」, so「、」was kept, but the exact comma glyph is uncertain.
- 問題9: Passage (2) first sentence has a stray leading full-width space「 に置いた6本」marking a likely word dropped in the retype; rendered as a single U+3000. Whether the original had additional dropped characters there cannot be recovered from the scan.
- 問題9: 「NHK文化センター」: the PDF render inserts spaces around the Latin run (NHK 文化センター); A's spaceless form was kept consistent with the digit/letter normalization convention, but the surrounding-space handling is a judgement call.
- 問題9: Passage (3): the ② circled marker is absent from the passage body (only the underline is printed) while Q57's stem cites ②; this asymmetry is in the source, not an OCR error.
- 問題10: The retype is heavily corrupted; some retained 'typos' (もごと, とういう, どうのように, すろ, 考え方法, 世界え, たらえ, 必至, 等々として, 関わのない) are judged from the renders but could in a few cases be render/antialias artifacts rather than the source — each was confirmed at 3x–5x zoom, so confidence is high but not absolute.
- 問題10: Citation internal spacing was normalized; the exact run of full-width spaces between 二郎 / 『現代の哲学』 / による is not preserved.
- 問題10: Whether the passage prints '、' or a bare gap after '視野に収め' was resolved to a comma via a 4x zoom; the mark is faint.
- 問題11: Passage B '分かる用になる': the glyph is clearly 用 in the render, but it is a source typo for 人; flagged in notes and transcribed as printed.
- 問題11: Passage B 'とか 好きだなと' uses a full-width space mid-sentence that is almost certainly a layout artifact of the retype; kept as printed.
- 問題11: The three 注 glosses are physically printed as one block on the following page beneath an unrelated passage; assigning 注1 to A and 注2/注3 to B follows the reference numbers, but the original print does not split them per-passage.
- 問題12: Paragraphs 3 and 6 were so badly corrupted in BOTH independent transcriptions that they provided little cross-check value; the final wording rests on my own reading of the double-resolution halves alone. The most uncertain single token is '昔、' opening para3 sentence (A read '也、', B read '巣も、'); the image is legible as 昔 but the print is a small font.
- 問題12: The tail after the モンシロチョウ sentence was previously truncated in the JSON despite being printed on the same source page; restored from the page image, Gemini 3.5 Flash OCR of the cropped region, and the plain-text witness. Exact printed line wrapping is not represented.
- 問題12: The instruction's printed digit spacing ('1 2 3 4') was normalized to '1・2・3・4' to match the rest of the corpus; if strict literal transcription is later required this single field would change.
- 問題13: The two low-dot glyphs .(U+FF0E) after 必要です/送信してください。 and the low comma ,(U+FF0C) before 同フォーラム are typographically subtle; transcribed exactly as the rendered glyphs but a different normalization policy could fold them to 。/、.
- 問題13: The 「開発部門の応あわかる募には」 sequence is genuinely garbled in the source Word retype; transcribed as the visible glyphs (応・あ・わ・か・る・募) rather than a guessed intended reading.
- 問題13: せいかつ (hiragana) and the final 「試作品されます」 read as source retype errors; left as printed per the no-correction 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) — 87 noted at OCR time:
- 問題1: Q1: source prints 「時間があつにつれて」 — almost certainly a retype typo for 「たつにつれて」; transcribed as printed.
- 問題1: Q2: source prints 「高校生のごる」 (likely typo for 「高校生のころ」) and 「海外で動きたい」 (likely typo for 「働きたい」); transcribed as printed.
- 問題1: Q3: source prints the garbled word カウンター as カワン + half-width ク (U+FF78) + horizontal bar ― (U+2015) in the PDF text layer, i.e. カワンク―; transcribed as カワンクー (full-width ク + 長音符) since the print is visually a katakana word with a long-vowel mark.
- 問題1: Q6: source prints 「ようや景気回復の」 — likely retype typo for 「ようやく」; transcribed as printed.
- 問題1: Q1: 「時間があつにつれて」 is a suspected retype typo for 「時間がたつにつれて」; transcribed as printed.
- 問題1: Q2: 「高校生のごる」 is a suspected retype typo for 「高校生のころ」, and 「海外で動きたい」 for 「海外で働きたい」; transcribed as printed.
- 問題1: Q4: 「におわれた」 is printed in kana; the original exam presumably has 「に追われた」. Transcribed as printed.
- 問題1: Q6: 「ようや景気回復の」 is a suspected retype typo for 「ようやく景気回復の」; transcribed as printed.
- 問題2: Q7: printed 「反対だれた」 — almost certainly a retype typo for 「反対された」; transcribed as printed.
- 問題2: Q8: printed 「ぞんな軽装で」 — likely retype typo for 「そんな」; transcribed as printed.
- 問題2: Q10: printed 「この言葉にわ、」 — likely retype typo for 「この言葉には」; transcribed as printed.
- 問題2: Q13: option 4 printed 「シェプ」 (handakuten confirmed at high zoom) — likely retype typo for 「シェア」; transcribed as printed.
- 問題2: Question 7: stem ends 「反対だれた。」 — verified at high zoom (だ with dakuten); almost certainly a retype typo for 「反対された」. Transcribed as printed.
- 問題2: Question 8: stem begins 「ぞんな軽装で」 — verified ぞ with dakuten; suspected retype typo for 「そんな」. Transcribed as printed.
- 問題2: Question 10: stem reads 「この言葉にわ、」 — suspected retype typo for 「には」. Transcribed as printed.
- 問題2: Question 13: option 4 is 「シェプ」 — verified at 8x zoom (handakuten, プ not ブ); suspected retype typo for 「シェア」 (the word in the genuine exam). Transcribed as printed.
- 問題3: Instruction prints 「の言葉に意見が最も近いものを」 — 意見 is a suspected retype typo for 意味; transcribed as printed.
- 問題3: Q17 stem begins 「ぞの結果」 — suspected retype typo for 「その結果」; transcribed as printed.
- 問題3: Q14 answer 2 prints 「平凡」 with no trailing な, while answers 1/3/4 all end in な — possible retype omission; transcribed as printed.
- 問題3: Q19 answer 3 prints 「繊細」 with no trailing さ, while answer 4 is 「豊富さ」 — possibly a retype omission of さ; transcribed as printed.
- 問題3: Instruction line reads 「の言葉に意見が最も近いものを」 — almost certainly a retype typo for 「意味が最も近いものを」; transcribed as printed.
- 問題3: Question 17 stem begins 「ぞの結果」 — likely a retype typo for 「その結果」 (dakuten clearly printed); transcribed as printed.
- 問題3: Question 14 option 2 is 「平凡」 with no trailing な, while options 1, 3, 4 are な-adjective forms (的確な・積極的な・否定的な); possibly a retype omission, transcribed as printed.
- 問題4: The printed instruction reads 「次の読み方として最もよいものを」 — almost certainly a retype error for the standard usage-question instruction 「次の言葉の使い方として最もよいものを」 (読み方 is the 問題1 phrasing). Transcribed as printed.
- 問題4: Q21 option 1: 「夕食をとる時間なまちまちだ」 — suspected retype typo for 「時間がまちまちだ」. Transcribed as printed.
- 問題4: Q21 option 2: 「まちまちにつ包んでもちった」 — garbled retype; 「につ包んで」 likely for 「に包んで」 and 「もちった」 likely for 「もらった」. Transcribed as printed.
- 問題4: Q23 option 1: 「すくに」 — suspected typo for 「すぐに」 (dakuten dropped). Transcribed as printed.
- 問題4: Q23 option 2: 「配布だれる」 — suspected typo for 「配布される」. Transcribed as printed.
- 問題4: Q25 option 4: 「歩いていから」 — suspected typo for 「歩いていたから」. Transcribed as printed.
- 問題4: Instruction line reads 「次の読み方として最もよいものを」 — almost certainly a retype error for the standard 問題4 usage instruction 「___の言葉の使い方として最もよいものを」; transcribed as printed.
- 問題4: Q21 option 1: 「夕食をとる時間なまちまちだ」 — 「時間な」 is a suspected retype typo for 「時間が」; transcribed as printed.
- 問題4: Q21 option 2: 「まちまちにつ包んでもちった」 — garbled retype (likely intended something like 「別々に包んでもらった」); transcribed as printed.
- 問題4: Q23 option 1: 「すくに」 suspected typo for 「すぐに」; transcribed as printed.
- 問題4: Q23 option 2: 「配布だれる」 suspected typo for 「配布される」; transcribed as printed.
- 問題4: Q25 option 4: 「歩いていから」 suspected typo for 「歩いていたから」; transcribed as printed.
- 問題5: The instruction line's choice separators are printed as full-width periods 「1.2.3.4」 rather than the standard middle dots (the same retype uses 「1・2・3・4」 for 問題4 on the same page). Digits normalized to half-width per convention; the full-width periods kept as printed.
- 問題5: Q26 option 4 is printed 「を通じる」 (verified at high zoom) — suspected retype typo for 「を通じて」 (the other three options are all て-forms).
- 問題5: Q27 stem prints 「致します」 in kanji where the official exam prints 「いたします」 — transcribed as printed.
- 問題5: Q33 uses full-width angle brackets <…> around the speakers' utterances instead of 「…」. Transcribed as printed.
- 問題5: Q33 stem 「田中監督語自身」 — suspected retype typo for 「田中監督ご自身」. Transcribed as printed.
- 問題5: Q33 options 1 and 3 end with a full stop (「お聞きいただけません。」, 「お話になるでしょうか。」) while options 2 and 4 have no trailing punctuation; option 1 additionally lacks the expected final か (verified at high zoom) — suspected typo for 「お聞きいただけませんか」. Transcribed as printed.
- 問題5: Q34 option 2 「めでにならないとすれば」 (verified at high zoom) — suspected retype typo for 「までにならないとすれば」 (compare option 1). Transcribed as printed.
- 問題5: Q35 stem prints a full stop after the first clause: 「…意義は理解出来ても。長年…」 (verified at high zoom) — a comma is almost certainly intended. Transcribed as printed.
- 問題6: Source prints ☆ instead of ★ throughout this section. In the question stems the slot run is printed as an underscore run around ☆ (e.g. __☆_); transcribed as ' _ _ ★ _ ' per convention. In the instruction line the printed star is an underlined ☆ with no surrounding slot run, transcribed as printed: <u>☆</u>.
- 問題6: Instruction prints 1234から (no interpuncts, unlike the official 1・2・3・4から); digits normalized to half-width ASCII per convention, the missing separators are as printed.
- 問題6: Q36: 'というわけでわないだろう' — わ is almost certainly a retype typo for は (というわけではないだろう); transcribed as printed.
- 問題6: Q38: scene label printed with half-width parentheses '(会社で)', and dialogue wrapped in full-width angle brackets <…> instead of the usual 「…」; transcribed as printed.
- 問題6: Q38: 'しなくれいいよ' — suspected retype typo for 'しなくていいよ'; transcribed as printed.
- 問題6: Q38 answer 4: '預けてくればいい' — possibly missing れ (預けてくれればいい); transcribed as printed.
- 問題6: Q39: '手始め自動車税' — suspected retype typo for '手始めに自動車税'; transcribed as printed. The option set (もの/増税/のない/避けよう) reads awkwardly and may itself be a flawed retype.
- 問題6: Q40: '自らを客観的して精神活動である' is ungrammatical as printed — likely a retype garble of something like '自らを客観視する精神活動'; transcribed exactly as printed.
- 問題7: Instruction line has retype typos, kept as printed: 読ん出 (likely 読んで) and かぁ一つ (likely から一つ); the choice list is written 1.2.3.4. with full-width periods (digits normalized to ASCII per convention).
- 問題7: Lead-in line typo kept as printed: 野間氏似当てて (likely 〜に宛てて).
- 問題7: The Amon Düül line is heavily garbled in the retype and kept verbatim: お送り下アモン.デユール∐の、『ウルフ.シデイ』ハ、…借りたもの(のばす)です — ASCII periods instead of ・, large ユ in デユール, a serif fallback-font glyph best matching ∐ (U+2210, presumably a mangled Ⅱ of Amon Düül II), half-width シ (U+FF7C) in ウルフ.シデイ, and katakana ハ in place of the particle は. The PDF text layer extracts the ∐ out of reading order (on its own line), but visually it sits immediately after デユール, before の.
- 問題7: Very high retype-typo density throughout the letter; transcribed exactly as printed. Other suspected typos kept verbatim: お知らしたとおり (お知らせした), 引つ越し (large つ, twice), 大阪にくることがれば (missing あ), 着てもらえるなら (来て), お送り下 (お送りした), (のばす) (のはず), 二十年以上もか借りて, ということのありませんように (ことが), 引つ越し後二整理 (後に整理), 借り他ままの本レコード (借りたままの本やレコード), ぼろと出てきて (ぼろぼろと), 見当 足らない本 (見当たらない; full-width space kept as printed), 貸しをたか (貸したか), 神戸で合った時 (会った), 一万円だけか (だっけか?), 時間がたつのが早く驚きます (早くて?), and question 43 option 1 借りているのっは (借りているのは).
- 問題7: 前略 is followed by a half-width space before ご無沙汰 (kept as printed; confirmed in the PDF text layer).
- 問題7: 草々 is right-aligned via a run of 12 full-width spaces after 奥さんによろしく。, kept as printed (count confirmed against the PDF text layer); 有栖川有栖 is right-aligned via 28 leading full-width spaces, which were dropped as paragraph-leading indent.
- 問題7: CD printed full-width, normalized to CD per convention. Quoted phrases use half-width ASCII square brackets [ ] as printed (the retype uses them where the original presumably had 「 」).
- 問題8: Source is a Chinese prep-school retype (clean digital text, plain question numbers); transcribed exactly as printed, retype typos included. Ambiguous glyphs cross-checked against the PDF text layer and the double-resolution half-page renders.
- 問題8: Suspected retype typos transcribed as printed: passage (1) 「脳のお記憶」 (likely 脳の記憶); passage (1) citation ends 「…とはどういうことか にようる)」 (likely による); question 46 option 3 「上手に上手に活用し」 (duplicated 上手に); passage (2) 「床用洗剤スリーン」 (likely クリーン); passage (2) 「大変お手数をおかけいますが」 (likely おかけしますが); passage (3) 「取り込んでいることが」 and question 48 option 3 「取り込んでいる」 (likely 取り組ん); passage (3) 「ステースタ」 (likely ステータス); passage (3) 注 「近くにあって、、人の生活」 (doubled 、); passage (4) 「仙人のよな人」 (likely のような).
- 問題8: Mixed comma glyphs as printed, confirmed against the PDF text layer and high-res images: full-width ,after 実際には/とか in passage (1), in the question 46 stem, and after につきましては in passage (2); half-width ASCII commas throughout passage (4), in its 注3 gloss, and in question 49 option 3 (while passage (4) option 4 and the 想像力が発達している line use full-width 、, confirming the mixed-comma source).
- 問題9: Passage (1) is a clean digital retype with several typos transcribed as printed: 「持ている」(for 持っている), 「それが大変な間違いた」(missing だ), 「ノスタルシー」(シ, while the 注 spells ノスタルジー), 「なのであるうか」(for であろうか). Body uses curly double quotes “役にも立たぬ”, “わからない” as printed.
- 問題9: Q50 stem printed as「大変な間違いだたあるが、何が間違いだったのか。」— garbled retype (standard would be「間違いだったとあるが」); the phrase「大変な間違い」is NOT underlined in this retype (neither stem nor passage), so left unmarked.
- 問題9: Q52 option 4 reads「絵は人の生き方を表したをもである」in the print — a garbled retype; transcribed exactly as printed (neither transcriber's reading matched).
- 問題9: Passage (2) first body line confirmed against the double-res p07-t scan as: 「…用意された大型保温テントに66箱の段ボールが運びこまれた。 に置いた6本のボンベから、濃度60%の二酸化炭素ガス流し込まれた。」— note 大型 (not 冷温), 66箱 (not 6畳), 運びこまれた, a stray leading full-width space before「に置いた」(likely a dropped word in the retype), and a missing が/を after ガス. All transcribed as printed.
- 問題9: Passage (2) typos transcribed as printed: 「卵までか駆除できます」(stray か), 「で増する可能性」(odd verb form), 「最大のメリットとうい」(for という), citation「8月17日付く夕刊」(付く, with a full-width space after the opening().
- 問題9: Passage (2)(注1)reads「ガスや煙で殺菌殺虫を行う」and(注2)uses「書棚」as printed.
- 問題9: Passage (3) paragraph 3 confirmed as「…象徴していた内容と、複雑な思考を経て…文章との違いが、そう感じさせるのだ。」(内容と, not 内容を/内容が; 文章との違い). 「一つ二つで危うく連れなって」transcribed as printed (odd wording).
- 問題9: Passage (3) Q56 option 1「ほんやりしている」(for ぼんやり) and Q58 option 3「考え内容以上」transcribed as printed.
- 問題10: This retypeset is unusually error-ridden; transcribed exactly as printed including obvious typos. Notable typos retained: Q59 stem 'もごと'(→ものごと)、'とういう'(→という)、'どうのように'(→どのように); Q60 stem 'どうのような'(→どのような); Q61 option 3 '解明すろこと'(→解明すること); Q62 option 2 '考え方法'(→考え方); Q62 option 3 '世界え'(→世界観); Q62 option 4 'たらえられる'(→とらえられる).
- 問題10: Passage typos retained as printed: '自分に関わのない'(→関わりのない); '等々として'(printed として where 'に' would read more naturally); '人生観・世界観の根本的にほかならない'(→根本にほかならない).
- 問題11: Passage B opens '真にアートの価値が分かる用になるためには' — 用 is a retype typo for 人; transcribed exactly as printed (用). Both A (人) and B (種) misread this glyph.
- 問題11: Passage B sentence '愛着も出てくるだろうし、身近において毎日みていることで、いろんな刺激を受けていくはずだ。' verified against the double-resolution image; both transcribers mis-transcribed the latter half (A dropped the middle clause and wrote 湧いて; B truncated to '見た回数も増える、').
- 問題11: Passage B '…とか 好きだなと心を動かされる作品…' — full-width space before 好きだな and the joining particle と are both present as printed.
- 問題11: Passage B '…さらに、どうして自分はこれが好きなのか、一歩踏み込んで…' — どうして as printed (A read 自分はどこ, B read なぜ).
- 問題11: Passage B paragraph after (中略) begins '数万円にしろ、数万円にしろ' — 数万円 is repeated verbatim, almost certainly a retype error (likely intended 数千円にしろ、数万円にしろ). Transcribed exactly as printed.
- 問題11: Q64 stem ends without a question mark ('述べていることは何か'); transcribed as printed.
- 問題11: Q64 option 2 reads 'アートの価値と、買い続けることで…' where the case particle と is suspect (likely intended は, parallel with options 3 and 4); transcribed as printed.
- 問題11: Passage A citation has a full-width space: (宮津大輔 『現代アートを買う』による). Passage B citation: (小山登美夫 『その絵、いくら? 現代アートの相場がわかる』による) with full-width ? and a full-width space after it, all as printed.
- 問題11: Passage A's retype gives the reading for 堰 inline as 堰(せき)(注1); normalized to
<ruby>堰<rt>せき</rt></ruby>(注1)to match JLPT ruby style. - 問題12: Both transcribers heavily corrupted passage paragraph 3 and paragraph 6; re-read from the double-resolution halves. Paragraph 3 actually reads '…店先などには巣をかけない。…嫌なスズメはやってこない。昔、ツバメがたくさん巣をかけると、店は繁盛するといわれた。話は逆であって、繁盛している店にツバメが集まってくるのである。' Paragraph 6 actually reads '…人工のものであろうとも、そしてそこをたくさんの人間がうろうろしていようとも、平気で棲みついてしまう。カラスやツバメのように、人間がいることをむしろ利用しているものだって、けっして少ないとはいえない。…人間がいるおかげで豊富な食物がたやすく手にはいるようになった。…' — both transcribers had dropped/garbled the middle clauses here.
- 問題12: The source page prints the passage tail containing アメンボ and 富栄養化, followed by the 日高敏隆 citation; a prior pass stopped before those lines. Restored the missing tail and citation from the rendered page and text witness.
- 問題13: Instruction reads 「右のページは、「機器・道具コンテスト」の案内である。…」 (verified against p12-b high-res); the コンテスト案内 document is the right-page facing document while the two question stems sit on the left page.
- 問題13: Q69 stem verified against p12-b: 「次の四人のうち、現段階でアイデア部門の特定課題の応募条件を満たしているのは誰か。」 The 名前/職業/応募作品/提出可能なもの table header and 山田 row are on p12; 鈴木/チェン/川村 rows are on p13.
- 問題13: Source is a Chinese prep-school re-typeset; several retype typos transcribed exactly as printed: 応募方法 line 「開発部門の応あわかる募には、…」 (garbled, verified glyph-by-glyph on p13-b); 「せいかつを便利にする…」 (hiragana せいかつ); Q70 stem 「適切化」 (likely 適切か); Q70 option 2 「アイデアぶもん」 (likely 部門 in kana); final clause 「同フォーラムで表彰され、試作品されます。」 (likely intended 試作品が展示されます).
- 問題13: Punctuation transcribed exactly as printed (source mixes glyphs): 「必要です.応募はすべて」 and 「送信してください。.応募フォーム」 use the low .(U+FF0E) seen on p13-b; 「アイデア部門の入賞作品は,同フォーラム」 uses the low comma ,(U+FF0C), distinct from the 、 used elsewhere (e.g. 開発部門の入賞作品は、「2012…」). The .after 送信してください。 was missed by both transcribers and restored from p13-b.
- 問題13: URL printed with trailing slash: (http://www.benrika.com/) (verified on p13-b). 応募期間 dash transcribed as ―― (two U+2015) per convention.
Corrections applied (88)
Evidence-based reconstructions of the original exam text. This section is the canonical correction log for the sitting.
- 問1 q1
question: 「時間があつにつれて」 → 「時間がたつにつれて」- あつ is a non-word; the verb is 時間がたつ (time passes), garbled in retype
- evidence: jlpt-freq witness L4: 時間がたつにつれて [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問1 q2
question: 「高校生のごる」 → 「高校生のころ」- ごる is a non-word retype typo for ころ (when)
- evidence: jlpt-freq witness L7: 高校生のころ [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問1 q2
question: 「海外で動きたい」 → 「海外で働きたい」- 動き is wrong kanji; context is 働く (to work)
- evidence: jlpt-freq witness L7: 海外で働きたい [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問1 q6
question: 「ようや景気回復の」 → 「ようやく景気回復の」- ようや is a truncated retype of the adverb ようやく
- evidence: jlpt-freq witness L19: ようやく景気回復 [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問2 q7
question: 「反対だれた」 → 「反対された」- だれた is a garbled retype of the passive された
- evidence: jlpt-freq witness L23: 反対された [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問2 q8
question: 「ぞんな軽装で」 → 「そんな軽装で」- ぞんな is a non-word; demonstrative is そんな (dakuten retype error)
- evidence: jlpt-freq witness L26: そんな軽装 [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問2 q10
question: 「この言葉にわ、」 → 「この言葉には、」- にわ is a phonetic retype of the topic particle には
- evidence: jlpt-freq witness L32: この言葉には [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問2 q13
answers[3]: 「シェプ」 → 「シェア」- シェプ is a non-word retype of シェア (share)
- evidence: jlpt-freq witness L43/L761: シェア [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問3 q14
answers[1]: 「平凡」 → 「平凡な」- missing な; other options are な-adjective forms
- evidence: jlpt-freq witness L47: 平凡な [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問3 q17
question: 「ぞの結果」 → 「その結果」- ぞの is a non-word; demonstrative is その (dakuten retype error)
- evidence: jlpt-freq witness L54: その結果 [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問3 q19
answers[2]: 「繊細」 → 「繊細さ」- missing さ; parallel option 4 is 豊富さ
- evidence: jlpt-freq witness L64: 繊細さ [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問4 q21
answers[0]: 「夕食をとる時間な」 → 「夕食をとる時間が」- 時間な is ungrammatical; subject particle が mistyped as な
- evidence: jlpt-freq witness L70: 夕食をとる時間がみんなまちまちだ [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問4 q21
answers[1]: 「につ包んでもちった」 → 「に包んでもらった」- につ包んで / もちった are garbled retypes of に包んで / もらった
- evidence: jlpt-freq witness L70: まちまちに包んでもらった [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問4 q23
answers[0]: 「すくに」 → 「すぐに」- すくに is a dakuten-dropped retype of すぐに
- evidence: jlpt-freq witness L78: すぐに配布される [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問4 q23
answers[1]: 「配布</u>だれる」 → 「配布</u>される」- だれる is a garbled retype of the passive される
- evidence: jlpt-freq witness L78: どこにいても配布される [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問4 q25
answers[3]: 「歩いていから」 → 「歩いていたら」- missing た and final form is conditional たら; original is 歩いていたら (tightened to/from: witness shows たら not たから)
- evidence: jlpt-freq witness L84: 歩いていたら、友達を見失ってしまった [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問5 q26
answers[3]: 「を通じる」 → 「を通じて」- other three options are て-forms; retype changed て to る
- evidence: jlpt-freq witness L88: を通じて [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問5 q33
question: 「田中監督語自身」 → 「田中監督ご自身」- 語 is a retype typo for the honorific prefix ご
- evidence: jlpt-freq witness L107: 田中監督ご自身 [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問5 q33
answers[0]: 「お聞きいただけません。」 → 「お聞きいただけませんか」- missing final interrogative か; trailing period not in original
- evidence: jlpt-freq witness L109/L830: お聞きいただけませんか [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問5 q34
answers[1]: 「めでにならないとすれば」 → 「までにならないとすれば」- めで is a retype typo for までに
- evidence: jlpt-freq witness L113: までにならないとすれば [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問5 q35
question: 「理解出来ても。長年」 → 「理解出来ても、長年」- sentence-internal full stop before continuing clause is a retype error; should be a comma
- evidence: jlpt-freq witness L115: 理解できても、長年 [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問5 q35
answers[3]: 「受入れないではないか」 → 「受け入れないではないか」- dropped okurigana け; verb is 受け入れる
- evidence: jlpt-freq witness L118: 受け入れないではないか [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問6 q36
question: 「というわけでわないだろう」 → 「というわけではないだろう」- でわ is a phonetic retype of ではない
- evidence: jlpt-freq witness L128: というわけではないだろう [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問6 q38
question: 「しなくれいいよ」 → 「しなくていいよ」- しなくれ is a garbled retype of しなくて
- evidence: jlpt-freq witness L134: しなくていいよ [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問6 q39
question: 「手始め自動車税」 → 「手始めに自動車税」- missing particle に; 手始めに (as a start)
- evidence: jlpt-freq witness L136: 手始めに自動車税 [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問6 q39
answers[1]: 「増税」 → 「増税を」- dropped object particle を
- evidence: jlpt-freq witness L137: 増税を [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問6 q40
question: 「自らを客観的して精神活動である」 → 「自らを客観的に観察して行う精神活動である」- 客観的して is ungrammatical; に観察して行う dropped in retype
- evidence: jlpt-freq witness L138: 自らを客観的に観察して行う精神活動である [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問7 q41
passage: 「野間氏似当てて」 → 「野間氏にあてて」- 似当てて is a non-word; should be にあてて (addressed to)
- evidence: jlpt-freq witness L145: 野間氏にあてて [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問7 q41
passage: 「お知らしたとおり」 → 「お知らせしたとおり」- お知らした is a non-word; verb is お知らせする
- evidence: jlpt-freq witness L148: お知らせしたとおり [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問7 q41
passage: 「くることがれば」 → 「くることがあれば」- dropped あ (ことがあれば)
- evidence: jlpt-freq witness L148: くることがあれば [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問7 q41
passage: 「もの(のばす)です」 → 「もののはずです」- (のばす)です is a garbled retype of のはずです
- evidence: jlpt-freq witness L150: 借りたもののはずです [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問7 q41
passage: 「二十年以上もか借りて」 → 「二十年以上も借りて」- stray か; reads 二十年以上も借りて
- evidence: jlpt-freq witness L151: 二十年以上も借りて [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問7 q41
passage: 「借り他まま」 → 「借りたまま」- 他 is a wrong-kanji retype for the particle た
- evidence: jlpt-freq witness L153: 借りたままの本 [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問7 q41
passage: 「本レコード」 → 「本やレコード」- dropped connective や (books and records)
- evidence: jlpt-freq witness L153: 本やレコード [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問7 q41
passage: 「ぼろと出てきて」 → 「ぼろぼろと出てきて」- dropped ぼろ; mimetic is ぼろぼろと
- evidence: jlpt-freq witness L153: ぼろぼろと出てきて [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問7 q42
passage: 「誰に貸しをたか」 → 「誰に貸したか」- 貸しをた is a garbled retype of 貸した (lent)
- evidence: jlpt-freq witness L155: それを誰に貸したか [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問7 q43
passage: 「神戸で合った時」 → 「神戸で会った時」- 合った is wrong kanji; should be 会った (met)
- evidence: jlpt-freq witness L157: 神戸で会った時 [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問7 q44
passage: 「時間がたつのが早く驚きます」 → 「時間がたつのが早くて驚きます」- dropped て; reads 早くて驚きます
- evidence: jlpt-freq witness L158: 時間がたつのが早くて驚きます [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問8 q46
passage: 「脳のお記憶」 → 「脳の記憶」- stray お; reads 脳の記憶
- evidence: jlpt-freq witness L179: 脳の記憶 [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問8 q46
passage: 「にようる)」 → 「による)」- にようる is a garbled retype of the citation による
- evidence: jlpt-freq witness L180: 『「頭がいい」とはどういうことか』による [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問8 q46
answers[2]: 「上手に上手に活用し」 → 「上手に活用し」- 上手に accidentally duplicated in retype
- evidence: jlpt-freq witness L184: 上手に活用し [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問8 q47
passage: 「おかけいますが」 → 「おかけしますが」- おかけいます is a non-word; verb is おかけする
- evidence: jlpt-freq witness L200-201: お手数をおかけしますが [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問8 q48
passage: 「取り込んでいることが」 → 「取り組んでいることが」- 取り込む is wrong; context is 取り組む (work on)
- evidence: jlpt-freq witness L213: 取り組んでいることが [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問8 q48
passage: 「ステースタ」 → 「ステータス」- ステースタ is a garbled retype of ステータス
- evidence: jlpt-freq witness L214: ステータス [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問8 q48
passage: 「、、人の生活」 → 「、人の生活」- doubled reading comma is a retype error
- evidence: jlpt-freq witness L218: 近くにあって、人の生活 [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問8 q48
answers[2]: 「取り込んでいる」 → 「取り組んでいる」- 取り込む is wrong; context is 取り組む
- evidence: jlpt-freq witness L222: 積極的に取り組んでいる [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問8 q49
passage: 「仙人のよな人」 → 「仙人のような人」- よな is a truncated retype of ような
- evidence: jlpt-freq witness L227: 仙人のような人 [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問9 q50
passage: 「持ている」 → 「持っている」- missing small tsu; should be 持っている
- evidence: jlpt-freq witness L258: 持っている人々 [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問9 q50
passage: 「それが大変な間違いた」 → 「それが大変な間違いだ」- 間違いた is ungrammatical; copula is だ
- evidence: jlpt-freq witness L260/L916: それが大変な間違いだ [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問9 q51
passage: 「ノスタルシー」 → 「ノスタルジー」- dakuten dropped; the passage's own 注 spells ノスタルジー
- evidence: jlpt-freq witness L265/L275: ノスタルジー [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問9 q51
passage: 「なのであるうか」 → 「なのであろうか」- あるうか is a garbled retype of あろうか
- evidence: jlpt-freq witness L265: なのであろうか [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問9 q52
answers[3]: 「絵は人の生き方を表したをもである」 → 「絵は人の生き方を表したものである」- をも is a garbled retype of もの
- evidence: jlpt-freq witness L294: 絵は人の生き方を表したものである [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問9 q53
passage: 「二酸化炭素ガス流し込まれた」 → 「二酸化炭素ガスが流し込まれた」- dropped subject particle が
- evidence: jlpt-freq witness L300: 二酸化炭素ガスが流し込まれた [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問9 q53
passage: 「卵までか駆除」 → 「卵まで駆除」- stray か; reads 卵まで駆除できます
- evidence: jlpt-freq witness L303: 卵まで駆除できます [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問9 q53
passage: 「書庫内でカビ発生し」 → 「書庫内でカビが発生し」- dropped subject particle が
- evidence: jlpt-freq witness L313: カビが発生し [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問9 q53
passage: 「図書館内で増する」 → 「図書館内で増殖する」- 増する is a non-word; dropped 殖
- evidence: jlpt-freq witness L315: 図書館内で増殖する [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問9 q53
passage: 「メリットとうい」 → 「メリットという」- とうい is a garbled retype of という
- evidence: jlpt-freq witness L318: 最大のメリットという [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問9 q53
passage: 「8月17日付く夕刊」 → 「8月17日付夕刊」- stray く after 付; reads 8月17日付夕刊
- evidence: jlpt-freq witness L322: 2010年8月17日付夕刊 [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問9 q54
answers[0]: 「虫やカビ増殖したから」 → 「虫やカビが増殖したから」- dropped subject particle が
- evidence: jlpt-freq witness L338: 虫やカビが増殖したから [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問9 q55
answers[0]: 「使用していやが」 → 「使用していたが」- していや is a garbled retype of していた
- evidence: jlpt-freq witness L347: 使用していたが [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問9 q55
answers[0]: 「薬品を使用するになった」 → 「薬品を使用するようになった」- dropped よう; idiom is ようになった
- evidence: jlpt-freq witness L347: 薬品を使用するようになった [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問9 q55
answers[1]: 「新しい受け入れた」 → 「新しく受け入れた」- 新しい should be adverbial 新しく
- evidence: jlpt-freq witness L349: 新しく受け入れた [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問9 q55
answers[3]: 「有害なガス使用していた」 → 「有害なガスを使用していた」- dropped object particle を
- evidence: jlpt-freq witness L353: 有害なガスを使用していた [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問9 q57
passage: 「一つ二つで危うく連れなって」 → 「一つ二つで危うく連なって」- 連れなって is a garbled retype of 連なって
- evidence: jlpt-freq witness L366: 一つ二つで危うく連なって [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問9 q58
answers[2]: 「考え内容以上のこと」 → 「考えた内容以上のこと」- dropped た; should be 考えた内容
- evidence: jlpt-freq witness L402: 考えた内容以上のこと [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問10 q59
passage: 「ほんとに見渡しうる」 → 「ほんとうに見渡しうる」- dropped う; adverb is ほんとうに
- evidence: jlpt-freq witness L408: ほんとうに見渡しうる [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問10 q59
passage:<rt>かか</rt></ruby>わのない→<rt>かか</rt></ruby>わりのない- dropped り; should be 関わりのない
- evidence: jlpt-freq witness L412: 自分に関わりのない [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問10 q59
passage: 「人生観・世界観の根本的にほかならない」 → 「人生観・世界観の根本的知にほかならない」- dropped 知; phrase is 根本的知
- evidence: jlpt-freq witness L431: 世界観の根本的知にほかならない [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問10 q59
passage: 「(二郎 『現代の哲学』」 → 「(渡邊二郎『現代の哲学』」- author surname 渡邊 dropped; philosopher is 渡邊二郎
- evidence: jlpt-freq witness L433: 渡邊二郎『現代の哲学』 [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問10 q59
question: 「科学的にもごとを見るとういうことを、筆者はどうのようにとらえているか。」 → 「科学的にものごとを見るということを、筆者はどのようにとらえているか。」- もごと→ものごと, とういう→という, どうのように→どのように retype garbles
- evidence: jlpt-freq witness L436: 科学的にものごとを見るということを、筆者はどのようにとらえているか [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問10 q60
question: 「どうのような傾向か」 → 「どのような傾向か」- どうのような is a retype garble of どのような
- evidence: jlpt-freq witness L447: どのような傾向か [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問10 q61
answers[2]: 「人間全体を解明すろこと」 → 「人間全体を解明すること」- すろ is a garbled retype of する
- evidence: jlpt-freq witness L452: 人間全体を解明すること [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問10 q62
answers[1]: 「主体的に考え方法を示せること」 → 「主体的に考える方法を示せること」- 考え方法 is ungrammatical; should be 考える方法
- evidence: jlpt-freq witness L454: 主体的に考える方法を示せる [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問10 q62
answers[2]: 「科学的な世界えを視野に」 → 「科学的な世界を視野に」- stray え; reads 科学的な世界を
- evidence: jlpt-freq witness L455: 科学的な世界を視野に入れられる [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問10 q62
answers[3]: 「客観的にたらえられる」 → 「客観的にとらえられる」- たらえ is a garbled retype of とらえ
- evidence: jlpt-freq witness L455: 客観的にとらえられる [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問11 q63
passageA: 「この自分満足の部分」 → 「この自己満足の部分」- 自分満足 is a non-standard retype of 自己満足
- evidence: jlpt-freq witness L461: 自己満足の部分 [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問11 q63
passageB: 「アートの価値が分かる用になる」 → 「アートの価値が分かるようになる」- 用に is a wrong-kanji retype for ように
- evidence: jlpt-freq kaisetsu L1006: 真にアートの価値がわかるようになるためには [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問13 q69
passage: 「せいかつを便利にする」 → 「生活を便利にする」- hiragana せいかつ is a retype error; document uses 生活 elsewhere
- evidence: jlpt-freq witness L575: 生活を便利にするオリジナルな機器 [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問13 q69
passage: 「開発部門の応あわかる募には」 → 「開発部門の応募には」- 応あわかる募 is a garbled retype of 応募
- evidence: jlpt-freq witness L582: 開発部門の応募には [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問13 q70
question: 「応募するのが適切化」 → 「応募するのが適切か」- 適切化 is ungrammatical; sentence asks 適切か
- evidence: jlpt-freq witness L564: どの部門に応募するのが適切か [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問13 q70
answers[1]: 「アイデアぶもん」 → 「アイデア部門」- ぶもん in hiragana is a retype error; 部門 in kanji elsewhere
- evidence: jlpt-freq witness L566-567: アイデア部門の自由課題 [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問12 q65
passage: 「町人や人家に」 → 「町や人家に」- 町人 does not fit 都市鳥 context; should be 町や人家
- evidence: jlpt-freq witness L500: 町や人家に「適応」した [correct-retypes workflow (propose + adversarial verify), 2026-06-17]
- 問5 q33
answers[2]: 「お話になるでしょうか。」 → 「お話になるでしょうか」- JLPT 問題5 answer options are bare clause fragments without sentence-final periods; options 1, 2, and 4 in this set carry no trailing 。 (option 1 was already normalized). The stray 。 on option 3 is a retype artifact, not part of the option text.
- evidence: Category typo, confidence medium. [retype-corrections-pass2 (propose + adversarial verify), 2026-06-20]
- 問9 q53
answers[2]: 「所有していて和紙製の本などの害虫駆除」 → 「所有していた和紙製の本などの害虫駆除」- The て-form 所有していて cannot attributively modify the following noun phrase 和紙製の本; the slot requires the past attributive 所有していた ("Japanese-paper books that individuals/groups owned"). The passage itself uses the parallel 個人や団体が所蔵していた…書籍. て is a retype error for た.
- evidence: Category typo, confidence high. [retype-corrections-pass2 (propose + adversarial verify), 2026-06-20]
- 問11 q63
passageB: 「買うと言うステップ」 → 「買うというステップ」- Here という is the grammaticalized quotative-nominal complementizer ("the step of buying"), conventionally written in kana, not 言う. The retype used kanji 言う.
- evidence: Category orthography, confidence high. [retype-corrections-pass2 (propose + adversarial verify), 2026-06-20]
- 問11 q64
answers[1]: 「アートの価値と、買い続けることで分かるものである。」 → 「アートの価値は、買い続けることで分かるものである。」- The case particle と strands the noun phrase with no second conjunct, leaving the option ungrammatical. The parallel options use 価値は (opt4) and 価値とは (opt3); the topic は yields the intended "the value of art is something understood by continuing to buy." と is a retype error for は.
- evidence: Category typo, confidence medium. [retype-corrections-pass2 (propose + adversarial verify), 2026-06-20]
- 問12 q65
passage: restored omitted passage tail and source citation after 「モンシロチョウ(注2)は日なたを失った。」- The source page continues with the concluding モンシロチョウ sentence, the アメンボ/富栄養化 paragraph, the final key-point sentence, and the 日高敏隆 citation. The prior JSON stopped mid-page before those printed lines, which left 注3 and 注4 apparently orphaned.
- evidence: rendered source PDF page /10; Gemini 3.5 Flash OCR of the cropped tail; jlpt-freq witness L525-L534. [ruby-followup-ocr-fix, 2026-07-08]
- 問12 q65
notes[1]: 「チョウの一種類」 → 「チョウの一種」- The rendered source page prints 一種, not 一種類.
- evidence: rendered source PDF page /10 and Gemini 3.5 Flash OCR crop. [ruby-followup-ocr-fix, 2026-07-08]
Answer Confidence
69 high · 0 medium · 1 low · 0 unresolved (of 70).
Generated from answer-audit.json at build time. Questions listed below are below high confidence or have source disagreement.
Answer Sources
in-doc-key(key) — in-document key, input/N1_2011-12/riyu-qingsongkao-answers-and-script.pdfjlpt-freq/extracted_text/N1/2011/2011.12_n1-2011.12真题+答案+听力原文.txt(key) — github.com/5Mcv4He/jlpt-freq extracted_text/N1/2011/2011.12_n1-2011.12真题+答案+听力原文.txtjlpt-freq/extracted_text/N1/2011/2011.12_2011年12月解析.txt(key) — github.com/5Mcv4He/jlpt-freq extracted_text/N1/2011/2011.12_2011年12月解析.txtsolver-1(solver)solver-2(solver)solver-3(solver)
Questions Needing Attention
| Question | Chosen | Confidence | Notes | Votes |
|---|---|---|---|---|
| q36 | 4 | low | in-doc-key=2, jlpt-freq/text/N1/2011/2011.12_n1-2011.12真题+答案+听力原文=2, solver-1=4, solver-2=4, solver-3=4 | |
| q40 | 1 | high | in-doc-key=1, jlpt-freq/text/N1/2011/2011.12_n1-2011.12真题+答案+听力原文=1, solver-1=4, solver-2=1, solver-3=1 |