ABCDEFGHIJKLMNOPQRSTUVWXYZ
1
言語
(Language)		号番号(Vol/No)発行日種別(Type)著者(Author)題名(Title)キーワード(Keyword)要旨(Abstract)ページ(page)
2
日本語35巻1号2025年6月20日招待論文・書評食野 真太郎ホイト・ロング（著）秋草俊一郎・今井亮一・坪野圭介（訳）（2023）『数の値打ち―グローバル情報化時代に日本文学を読む』フィルムアート社1-4
3
EnglishVol.35, No.1June 20, 2025Book ReviewMESHINO ShintaroLONG Hoyt (Author), AKIKUSA Shun’ichiro; IMAI Ryoichi; TSUBONO Keisuke (Translators) (2023). The Values in Numbers: Reading Japanese Literature in a Global Information Age. Tokyo: Film Art Inc.1-4
4
日本語35巻1号2025年6月20日招待論文・書評髙久 文枝・松井 佑樹石黒圭・吉甜・佐野彩子（共編）（2024）『スマホは辞書になりうるか―日本語学習者の辞書引きの困難点と指導法』明治書院5-9
5
EnglishVol.35, No.1June 20, 2025Book ReviewTAKAKU Fumie; MATSUI YukiISHIGURO Kei; JI Tian; SANO Ayako (2024). Can a Smartphone Function as a Dictionary?: Difficulties and Teaching Strategies in Dictionary Use by Learners of Japanese. Tokyo: Meiji Shoin.5-9
6
日本語35巻1号2025年6月20日招待論文・書評王 雨金愛蘭（著）（2024）『外来語の基本語化 現代新聞「叙述語彙」への進出』大阪大学出版会10-14
7
EnglishVol.35, No.1June 20, 2025Book ReviewWANG YuKIM Eran. (2024). The Basic Integration of Loanwords: The Expansion of Loanwords into
Predicate Vocabulary in Contemporary Newspapers. Osaka: Osaka University Press.		10-14
8
日本語35巻2号2025年9月20日論文B松下　達彦，新井　智大，岩下　智彦，田中　祐輔，田中　牧郎，河内　昭浩，柏野　和佳子，山崎　誠小中学校の学年別語彙配当を定めるための頻度データと基本語彙表の統合と比較
国語教育，学年別語彙配当，コーパス，基本語彙表，基準頻度，
頻度レベル
日本の国語教育では漢字の学年配当が重視されるが，学習単位としては「語」のほうが重要である．本研究では，小中学校の学年別語彙配当表を作成・提案するため，児童向けコーパス（教科書，小学生新聞，児童書など）の語彙頻度と，田中（1956），池原（1957），阪本（1958），児童言語研究会（1962），阪本（1984），中央教育研究所（1984），浜本（1990）等，既存の教育基本語彙表の収録語をUniDicの短単位に統一して一つのデータベースとして統合し，教育基本語彙表の頻度レベルを分析した．語彙頻度に重みづけして基準頻度を計算し，各種の語彙表に付された学年レベルごとに，頻度分布を中央値，四分位数，箱ひげ図などで確認した．その結果，小学1・2年の語は高頻度レベル，中高校生の語は低頻度レベルに分布する一方，小学3～6年では頻度分布に明確な違いが見られないことが明らかになった．この結果を踏まえ，学年別語彙配当の決め方について検討した．15-30
9
EnglishVol.35, No.2Sept. 20, 2025Paper BMATSUSHITA Tatsuhiko; ARAI Tomohiro; IWASHITA Tomohiko; TANAKA Yusuke; TANAKA Makiro; KAWAUCHI Akihiro; KASHINO Wakako; YAMAZAKI MakotoIntegrating and Comparing Frequency Data and Basic Word Lists for Establishing Vocabulary Allocation by School Year in Primary and Secondary EducationJapanese language education at primary/secondary schools, Year-level vocabulary allocation, Corpus, Basic word lists, Baseline frequency, Frequency levelIn Japanese language education, the school-year-based allocation of kanji is often emphasized, yet words represent more fundamental learning units. This study aims to develop and propose year-level word lists for primary and secondary education. To this end, we compiled word frequency data from children’s corpora—textbooks, children's newspapers, and juvenile literature—and integrated them with entries from existing educational word lists (Tanaka 1956; Ikehara 1957; Sakamoto 1958, 1984; Jigenken 1962; Central Educational Research Institute 1984; Hamamoto 1990), standardised into UniDic short-unit forms. Frequency levels were analyzed, and weighted baseline frequencies were calculated using data from multiple corpora. The frequency distributions by assigned to each year level in each word list were then examined using medians, quartiles, and boxplots. Results showed that words for Years 1–2 appeared at high frequency levels, those for secondary at low levels, while words for Years 3–6 showed no clear frequency differences. Based on these findings, we discuss principles for allocating vocabulary by year.15-30
10
日本語35巻2号2025年9月20日招待論文・書評佐々木　稔谷中瞳（著）（2024）『ことばの意味を計算するしくみ　計算言語学と自然言語処理の基礎』講談社
31-35
11
EnglishVol.35, No.2Sept. 20, 2025Book ReviewSASAKI MinoruYANAKA, Hitomi. (2024). How the Meaning of Words is Calculated: the Introduction to
Computational Linguistics and Natural Language Processing. Tokyo: Kodansha.		31-35
12
日本語35巻3号2025年12月20日巻頭言入江 さやかテーマ特集「特定の場面の計量的研究」について37
13
EnglishVol.35, No.3Dec. 20, 2025ForewordIRIE SayakaForeword to the Special Section of the "Quantitative Study of Language Used in Specific Situations"37
14
日本語35巻3号2025年12月20日特集・招待論文 A小林　雄一郎，岡﨑　友子大規模コーパスに基づく接続詞の連続パターンのレジスター分析
接続詞の連続，書き言葉と話し言葉，レジスター分析，日本語コーパス本研究の目的は，日常的な言語使用から専門的な場面までの多様なレジスターを対象として，接続詞が連続するパターンのレジスター分析を行うことである．分析コーパスには現代日本語書き言葉均衡コーパス，日本語日常会話コーパス，日本語話し言葉コーパス，昭和・平成書き言葉コーパス，昭和話し言葉コーパスを用いた．そして，対応分析，階層型クラスター分析，ランダムフォレストなどの統計手法で，コーパスごと，レジスターごとに接続詞の連続パターンを比較した．その結果，書き言葉と話し言葉による使用頻度の大きな違い，さらに，話し言葉における対話と独話による頻度パターンの違いが見られた．38-53
15
EnglishVol.35, No.3Dec. 20, 2025Invited Paper (A)KOBAYASHI Yuichiro, OKAZAKI TomokoRegister Analysis of Sequential Patterns of Conjunctions Using Large-Scale Corpora
sequential patterns of conjunctions, written and spoken language, register analysis, Japanese corporaThis study aims to conduct a corpus-based register analysis of the Japanese language, targeting a wide range of registers from everyday language use to specialized domains. Utilizing multiple corpora, including the Balanced Corpus of Contemporary Written Japanese (BCCWJ) and the Corpus of Spontaneous Japanese (CSJ), the research compares the sequential patterns of conjunctions across different corpora and registers. Employing several statistical methods such as correspondence analysis, cluster analysis, and random forests, the study identifies linguistic features that distinguish written and spoken Japanese across various registers. The findings reveal significant differences in usage between written and spoken language, as well as variation between monologue and dialogue data.38-53
16
日本語35巻3号2025年12月20日特集・招待論文 A河瀬　彰宏漫才台本の構成分析―M-1グランプリにおける高評価パターンの抽出―
物語論，談話構成，N-gram，系列解析，多重比較補正本研究は，観客から高く評価される漫才の構成を明らかにするために，M-1グランプリ2015-2019年大会の計50演目を対象に，台本を機能単位に分割し，その並びの頻度について演目ごとの長さの違いを補正して比較した．分析の結果，高評価の演目では，ボケ直後をツッコミで素早く回収する構成が多く，低評価では，話題の引き延ばし，ボケの継ぎ足し・反復が多かった．同一機能の連続出現の影響を除いて再分析をしたところ，これらの傾向は維持され，短い周期でツッコミを挟む運びが高評価に資することがわかった．54-69
17
EnglishVol.35, No.3Dec. 20, 2025Invited Paper (A)KAWASE AkihiroStructural Analysis of Manzai Scripts:
Extracting Patterns Associated with Higher Ratings in the M-1 Grand Prix
Narratology, Discourse structure, N-gram, Sequence analysis, Multiple-testing correction (FDR)
This study analyzed 50 manzai routines from the 2015–2019 M-1 Grand Prix to identify structural patterns associated with higher audience evaluations. Transcripts were segmented into functional units, and sequential patterns were compared using length-normalized N-gram frequencies. High-rated routines more often resolved a boke (a deliberate deviation) quickly with a subsequent tsukkomi (an explicit corrective call-out) whereas low-rated routines exhibited prolonged set-up and additive or repetitive boke. A sensitivity analysis that collapsed consecutive identical labels (run-length encoding) yielded the same qualitative result: the advantage of short cycles that insert tsukkomi promptly persisted, while differences attributable to long identical runs diminished. These findings indicate that maintaining a short lag from boke to tsukkomi, and avoiding extended stretches of set-up or repeated boke, characterizes routines that receive higher evaluations.54-69
18
日本語35巻3号2025年12月20日特集・招待論文 A麻　子軒品詞・語種構成から見る話題の文体的特徴―J-TOCCを用いたコレスポンデンス分析―
話題，場面，文体，J-TOCC，品詞，語種，コレスポンデンス分析『日本語話題別会話コーパス：J-TOCC』に収録された15話題を対象に，品詞構成と語種構成という古典的な文体の計量指標を「話題」という場面に適用し，コレスポンデンス分析を通じて相互の対応関係を検討した．その結果，「03旅行」「07学校」など経験を語る話題では具体的な時空を出発点として話が展開される傾向があるため，名詞が多用されている一方，「12夢・将来設計」「13マナー」のような社会的・抽象的な話題では，行為という観点から話が進むため，動詞が話題の中心になりやすいことが分かった．さらに，「10動物」「01食べること」といった主観的な印象に基づく話題では形容詞系を起点として談話が展開されることが観察された．語種の観点でも，新しい概念や海外起源の要素が多い「08スマートフォン」「04スポーツ」では外来語，社会的・抽象的な「12夢・将来設計」「15日本の未来」では漢語，日常生活に基づく「10動物」「13マナー」では和語が相対的に優勢となるプロットが得られた．70-85
19
EnglishVol.35, No.3Dec. 20, 2025Invited Paper (A)MA Tzu-HsuanStylistic Characteristics of Topics from Parts of Speech and Word Type Composition: A Correspondence Analysis Using J-TOCCTopic, Register, Style, J-TOCC, Parts of speech, Word type, Correspondence analysisThis study examines 15 topics included in the Japanese Topic-Oriented Conversation Corpus (J-TOCC) by applying two classical stylistic quantitative indicators—parts of speech composition and word type composition—to the context of “topics,” and investigates their correspondences through correspondence analysis. The results show that in topics such as “03 Travel” and “07 School,” which involve recounting experiences, conversations tend to develop from concrete spatiotemporal points, resulting in frequent use of nouns. In contrast, in more social and abstract topics like “12 Dreams and Future Plans” and “13 Manners,” discourse progresses from the perspective of actions, making verbs more central. Furthermore, in topics based on subjective impressions, such as “10 Animals” and “01 Eating,” discourse was observed to develop from adjectives. From the perspective of word types, topics involving new concepts or elements of foreign origin, such as “08 Smartphones” and “04 Sports,” featured more loanwords; social and abstract topics like “12 Dreams and Future Plans” and “15 Japan’s Future” tended to favor Sino-Japanese words; and everyday life topics such as “01 Eating” and “06 Housework” showed a relative dominance of native Japanese words.70-85
20
日本語35巻3号2025年12月20日特集・招待論文 A小磯　花絵『日本語日常会話コーパス』を活用した「から」「ので」の選択に関わる要因分析
日常会話，会話コーパス，「から」，「ので」本稿では，『日本語日常会話コーパス』を用い，原因・理由を表す「から」と「ので」の選択に関わる要因を検討した．分析の結果，雑談から会議・会合へと場面の改まり度が高くなるにつれて「ので」の使用率が増加すること，また丁寧体の使用率が高い会話ほど「ので」が選ばれやすいことが明らかになった．さらに，家族との会話では「から」が圧倒的に優勢である一方，同僚や仕事関係者，客など，対人配慮が求められる相手に対しては「ので」の使用が増加する傾向が確認された．以上の結果は，「から」と「ので」の選択が，場面の改まり度や話者間の関係性といった社会的・状況的要因に関わるものであり，対人配慮が求められる場面や相手ほど「ので」が選好されることを示すものである．86-101
21
EnglishVol.35, No.3Dec. 20, 2025Invited Paper (A)KOISO HanaeAnalyzing Factors Affecting the Use of kara and node with the Corpus of Everyday Japanese Conversationeveryday conversation, conversation corpus, kara, nodeThis study investigates the factors affecting the choice between the Japanese causal connectives kara and node, using the Corpus of Everyday Japanese Conversation (CEJC). The analysis demonstrates that the proportion of node increases as the level of formality rises, from casual conversation to meetings, and that node is more likely to occur in speech styles employing polite verb endings. Furthermore, while kara is overwhelmingly dominant in conversations with family members, node tends to be favored in interactions with colleagues, business partners, and customers, where considerations of politeness and social distance are more salient. These findings suggest that the choice between kara and node is systematically conditioned by social and situational factors such as the degree of formality and the nature of interlocutor relationships, and that node is especially preferred in contexts requiring interpersonal consideration.86-101
22
日本語35巻3号2025年12月20日特集・招待論文 A李　暁燕日本の学校プリントに潜む暗黙知の計量的可視化―外国人保護者の理解を妨げる要因―
学校プリント，暗黙知，外国人保護者，学校文化，多文化共生本研究は，日本の小学校で配布される学校プリントが外国人保護者の理解を妨げる要因の解明を目的とした．810件のプリントを対象に，カテゴリ別語彙，依頼表現，共起ネットワークを分析した．その結果，第一に「生活習慣」「行事」「金銭」「学用品」といった語彙カテゴリは，時間厳守，行事参加，金銭負担，持ち物準備といった暗黙の規範を反映していた．第二に，「ください」「お願いします」といった依頼表現が頻出し，保護者に対する協力要請が繰り返し行われていることが示された．さらに共起分析からは，「学校」「時間」などを中心とする語彙群が，学校文化の基盤的構造を形成していることが示唆された．これらの知見は，学校プリントに潜む「見えないルール」が教育的包摂を阻む要因となることを示し，文化的背景を補足するAI支援の有効性を示唆する．102-114
23
EnglishVol.35, No.3Dec. 20, 2025Invited Paper (A)LI XiaoyanMaking Tacit Knowledge Visible in Japanese School Printouts: Barriers to Understanding for Foreign Guardians
School handouts, Tacit knowledge, Foreign guardians, School culture, Multicultural coexistence
This study aims to clarify the factors that hinder foreign guardians’ understanding of school printouts distributed in Japanese elementary schools. A total of 810 printouts were analyzed in terms of request expressions, vocabulary categories, and co-occurrence networks. The results revealed, first, that request expressions such as “kudasai (please)” and “onegaishimasu (we ask for your cooperation)” appeared frequently, indicating that repeated cooperation requests are made to guardians. Second, vocabulary categories such as “daily routines,” “events,” “money,” and “school supplies” reflected implicit norms related to punctuality, event participation, financial obligations, and preparation of belongings. Furthermore, the co-occurrence analysis suggested that vocabulary groups centered on words such as “school” and “time” form the basic structural foundation of school culture. These findings indicate that the “invisible rules” embedded in school printouts can become barriers to educational inclusion and highlight the potential effectiveness of AI support systems that supplement such cultural background.102-114
24
日本語35巻3号2025年12月20日特集・論文 A井関 龍太歌詞の品詞構成が反映する文体特徴の探索的研究―東京都小中高校歌の計量的分析―
品詞構成，MVR，文体，歌詞，校歌，順列エントロピー，統計的複雑度文章の文体特徴を品詞構成の観点から捉える指標としてMVRがあるが，MVRと具体的な文体特徴の対応関係に関する実証的・理論的検討は十分であるとは言えない．本研究では，校歌の歌詞という変異の幅が比較的限られた文章を対象として，品詞構成と文体特徴との対応関係を探索的に検討した．東京都の小中高の学校の校歌を分析したところ，MVRは小学校で中学・高校よりも高いことが確認された．ただし，その違いはM率のわずかな差によるものであった．具体的な文章の違いを検討したところ，N率の高い歌詞は倒置法や体言止めを多用した詩文的な形式をとる傾向が強く，V率が高い歌詞は述語で区切る物語的な形式をとる傾向があることがうかがわれた．M率はこれらのパターンを乱すような働きをしており，いずれかの表現形式への傾きを和らげているようであった．これらのことから，校歌の歌詞という文脈において品詞構成の違いがどのような表現の実現形態と結びつくのかを論じた．115-130
25
EnglishVol.35, No.3Dec. 20, 2025Paper (A)ISEKI RyutaAn Exploratory Study of Stylistic Features Reflected in the Part-of-Speech Composition of Lyrics: A Quantitative Analysis of Tokyo Elementary, Junior High, and High School Songspart-of-speech composition, MVR, style, lyrics, school song, permutation entropy, statistical complexityMVR (Modifier Verb Ratio), an index that captures the stylistic features of texts in terms of part-of-speech composition, has not been sufficiently examined for its correspondence with specific stylistic features. This study explored the correspondence between MVR and stylistic features using school-song lyrics, which are texts with a relatively limited range of variation. Analysis of song lyrics from elementary, junior high, and high schools in Tokyo confirmed that MVR was higher in elementary schools than in junior high and high schools. However, the differences were due to slight differences in the M (Modifier) rates. At the individual text level, lyrics with a high N (Noun) rate were more likely to take a poetic form with many inversions and noun-phrase endings, whereas lyrics with a high V (Verb) rate were more likely to take a narrative form with predicate endings. The M-rate seemed to disrupt these patterns and neutralize the inclination toward one of the expressive forms. These results suggest that the parts-of-speech composition of school-song lyrics relates to the form in which the lyrics are realized.115-130
26
日本語35巻3号2025年12月20日特集・招待論文 B居關　友里子子どもの日常会話における丁寧体使用―『子ども版日本語日常会話コーパス』に基づく量的・質的検討―
『子ども版日本語日常会話コーパス（CEJC-Child）』，子どもの言語使用，スピーチスタイル，丁寧体本研究は，国立国語研究所が構築を進めている『子ども版日本語日常会話コーパス（CEJC-Child）』を用い，子どもの日常会話における丁寧体使用を量的・質的に検討したものである．CEJC-Childモニター版に収録された調査対象児10名による，約50時間分の会話データから述語が動詞・形容詞である発話を抽出し，普通体・丁寧体の区別を行い丁寧体使用率を算出した．その結果，丁寧体は2歳から2歳半頃以降に安定して観察された一方で，丁寧体使用率が月齢に比例して上昇する傾向は見られず，むしろ収録場面の特徴に大きく左右されることが示唆された．そこで，丁寧体使用率の高い会話を中心に場面に着目して質的分析を行ったところ，複数の調査対象児に共通するものとして，丁寧体が，普段の自分とは異なる者として発話していることを表示する場面や，情報を確固たるものとして相手に提示する場面で使用されていることが確認された．131-143
27
EnglishVol.35, No.3Dec. 20, 2025Invited Paper (B)ISEKI YurikoPolite Form in Children’s Everyday Conversation: A Quantitative–Qualitative Study Using the Corpus of Everyday Japanese Conversation for Children
Corpus of Everyday Japanese Conversation for Children (CEJC-Child), Children's Language, Speech Style, Polite Form
This study investigates the use of the polite (desu/masu) form in children’s everyday conversations by employing the Corpus of Everyday Japanese Conversation for Children (CEJC-Child), currently being developed by the National Institute for Japanese Language and Linguistics. From approximately 50 hours of conversational data produced by ten target children in the CEJC-Child monitor version, utterances whose predicates were verbs or adjectives were extracted and classified according to plain vs. polite form, and the rate of polite form usage was calculated. The results showed that polite form began to be used consistently from around 2 to 2.5 years of age; however, the rate of polite form usage did not increase in proportion to age. A subsequent qualitative analysis of conversations with relatively high rates of polite form revealed two recurring patterns across multiple children: polite form was employed when speaking as someone other than one’s usual self, and when presenting information as certain and authoritative to the interlocutor.131-143
28
日本語35巻3号2025年12月20日特集・招待論文 B立部　文崇，藤田　裕一郎初級日本語授業における教師発話の構成―機能分類を元にした分析―
日本語教育，教師発話，授業談話分析，教師の即興的調整本研究は，初級日本語授業における教師発話の機能と構造に注目し，IRF構造（導入・応答・フィードバック）と主分類（教室運営／指導内容）の観点から，2432発話を定量・質的に分析したものである．質問形式やフィードバックの焦点に合わせて，教師発話は状況に応じて柔軟に調整されており，言語支援と関係調整の両面で重要な役割を果たしていた．教師の即興的対応や相互行為の調整を可視化することにより，日本語教育における授業談話研究の深化と，教師養成・研修への応用可能性を示唆する．144-159
29
EnglishVol.35, No.3Dec. 20, 2025Invited Paper (B)TATEBE Fumitaka; FUJITA YuichiroThe Structure of Teacher Speech in Beginner Japanese Classes: An Analysis Based on Functional Classification
 Japanese language education, teacher utterances, classroom discourse analysis, teachers’ improvisational adjustmentsThis study investigates the functions and structures of teachers’ utterances in beginner-level Japanese language classes, based on a quantitative and qualitative analysis of 2,432 utterances. The analysis adopts the IRF (Initiation–Response–Feedback) framework and a dual categorization: classroom management and instructional content. Teacher utterances were flexibly adapted depending on question types and feedback orientation, playing vital roles in both linguistic scaffolding and relational coordination. The results show that teachers actively shift their utterance strategies to suit interactional needs, especially in balancing instructional clarity and learner engagement. By visualizing teachers’ improvisational responses and adaptive feedback, the study offers insights into how teacher talk functions as a tool not only for instruction but also for classroom dynamics. These findings suggest that teacher utterances should be analyzed as context-sensitive acts that mediate understanding and foster participation. The study also highlights the value of fine-grained utterance tagging and IRF-based discourse analysis in Japanese language teacher education. It demonstrates the potential for using real classroom data to support professional development and reflective practice, providing a foundation for future research on interactional competence and pedagogical flexibility in language classrooms.144-159
30
日本語35巻3号2025年12月20日特集・論文 B新井　庭子概念を中心とする教科書テキストの計量的分析―分析項目の提案と実践―
教科書，概念体系，用語，定義，分類本論文は，学校教育という特定場面の言語表現において，知識基盤を伝える上で中核的な役割を担う教科書を分析対象とし，概念についての説明を計量するための尺度を導入するとともに，その尺度で学年進行に伴う変化を可視化する．具体的な分析対象として小学5年生から中学2年生までの理科教科書を選び，分析項目として索引語数（異なり用語数），定義表現の数，分類表現の数を採用・導入して分析した結果，索引語数は学年とともに段階的に増加することがわかった．小学校では中学校に比べ，1つの用語あたりに費やされる文数・語数が多いことが明らかになった．また，小学校は中学校に比べ，定義される用語の平均的な割合が明確に低いことが示された．概念に関する明示的な説明である定義表現と分類表現の数は，小学6年生から中学1年生にかけて約4倍に増加する結果となった．160-175
31
EnglishVol.35, No.3Dec. 20, 2025Paper (B)ARAI TeikoA Quantitative Analysis of Conceptual Explanations in Science Textbooks: Proposal and Implementation of Analytical Metrics
School textbooks, conceptual terminology, definition analysis, classification framework
This study analyzes textbooks as the primary medium for conveying a knowledge base in the context of compulsory education. It introduces a quantitative measure for assessing conceptual explanations and visualizes changes across grade levels. The analysis focuses on science textbooks from Grade 5 to Grade 8, examining three key elements: term types, the number of definition expressions, and the number of classification expressions. The findings reveal that elementary school textbooks devote more sentences and words per term compared to middle school textbooks. While term types were analyzed as a surface-level cue for concepts, the number of definition expressions—representing explicit conceptual explanations—did not align with term types. Elementary school textbooks demonstrated a significantly lower proportion of explicitly defined terms compared to middle school textbooks. Furthermore, the number of definition and classification expressions, which provide explicit conceptual explanations, increased approximately fourfold from Grade 6 to Grade 7. The study contributes to quantitative linguistic research on educational texts and offers a foundation for evaluating the clarity and accessibility of conceptual explanations in school curricula.160-175
32
日本語35巻4号2026年3月20日特集・論文 A李 羽喆パラグラフ・ライティングにおける語彙産出の可視化
人間教師, ChatGPT, DeepSeekの評価を比較して		処理水準モデル, 産出的な語彙学習，形態素解析，生成 AI 評価と人間評価本研究は，中国の大学におけるパラグラフ・ライティング活動を通じ，日本語学習者の語彙産出の縦断的変化を量的・質的に可視化し，加えて人間教師と生成AI（ChatGPT・DeepSeek）による評価の一致度を比較検討したものである．語彙産出の量的変化として，使用語数の増加が見られ，特に一文あたりの語彙密度が高まるという表現の多様化が確認された．一方，質的評価では，人間教師は最も幅広く柔軟な評価を示し，個々の学習者の語彙使用や感情表現を多面的に捉えていた．これに対し，ChatGPTは慎重かつ保守的な評価傾向を示し，DeepSeekはやや人間教師に近い傾向を持ちながらも柔軟性に欠けた．これらの結果は，語彙の「質」の向上が単純な「量」の増加には比例せず，学習者が語の産出を通じて自己表現を試行錯誤する非線形的な過程であることを示唆する．語彙教育においては，生成AIの定量的評価の利点を活かしつつ，学習者の内的成長や表現の豊かさを汲み取る人間教師の共感的かつ柔軟な支援の重要性が改めて示された．177-192
33
EnglishVol.35, No.4March 20, 2026Paper (A)LI YuzheVisualizing Vocabulary Production in Paragraph Writing:
A Comparative Analysis of Evaluations by Human Teachers, ChatGPT,
and DeepSeek		levels of processing, productive vocabulary learning, morphological analysis,
AI and human evaluation		This study examined quantitative and qualitative changes in vocabulary production among
Japanese learners through a paragraph writing activity conducted at a university in China. It also
compared evaluations by human teachers and GenAI (ChatGPT and DeepSeek). Morphological
analyses showed a gradual increase in learners’ vocabulary use and sentence length, particularly
noting an increase in the use of native Japanese words (Wago). Qualitative analyses indicated
that human teachers provided the most flexible and nuanced assessments, effectively capturing
subtle differences in learners’ vocabulary choices and emotional expressions. While the GenAI
assessments correlated with human evaluations, ChatGPT tended to provide cautious and
conservative ratings, limiting its ability to capture subtle expressive nuances. DeepSeek’s
assessments were somewhat closer to those of human teachers but still lacked sufficient
flexibility. These findings highlight the importance of human empathy and nuanced
understanding in assessing vocabulary skills, underscoring that effective evaluation should
integrate attention to learners’ emotional expressions and personal growth in addition to
quantitative measures.		177-192
34
日本語35巻4号2026年3月20日特集・論文 A李 琦ジャンルによる換言の接続表現の使い分け
―BCCWJ，CSJ，CEJCの出現頻度別の特徴―		ジャンル，換言の接続表現，メディア，分離性，関与性，計画性，論理性本研究では「すなわち」「つまり」「要するに」「要は」という4種の換言の接続表現を対象とし，書籍・講演・雑談の3ジャンルにおける出現頻度の特徴を量的・質的に分析した．その結果，各表現はいずれもジャンル間で出現頻度に有意な差異を示すことが明らかとなった．とりわけ，「すなわち」は書籍，「つまり」は書籍と講演，「要するに」は講演と雑談，「要は」は雑談で突出して使用されることが確認された．これらの分布を質的に検討した結果，書き言葉と話し言葉を単純に文字・音声の二分法で考えるだけでは不十分で，別の観点と組み合わせて説明する必要があることが明らかになった．本稿では，Chafeが提唱した分離性と関与性という概念を用い，換言の接続表現の使い分けを説明した．また，その背後として，書き言葉と話し言葉をめぐる複合的な要因の作用が示唆された．193-208
35
EnglishVol.35, No.4March 20, 2026Paper (A)LI QiGenre-Based Variation in Paraphrastic Connective Expressions:
A Frequency-Based Analysis Using BCCWJ, CSJ, and CEJC		genre, paraphrastic connectives, media, detachment, involvement, planning, logicalityThis study examines four paraphrastic connective expressions in Japanese: sunawachi, tsumari, yosuruni, and yō wa, through a quantitative and qualitative investigation of their frequencies across three distinct genres: books, lectures, and casual conversations. The analysis reveals statistically significant differences among the genres, showing that sunawachi is predominantly used in books, tsumari appears frequently in both books and lectures, yosuruni is common in lectures and casual conversations, and yō wa is especially prominent in casual conversations. A subsequent qualitative examination indicates that a simple binary classification of written versus spoken language, defined solely by textual or auditory media, is inadequate to explain these variations. Instead, this paper adopts Chafe’s concepts of Detachment and Involvement to elucidate how these paraphrastic connectives are employed, suggesting that multiple, interrelated factors shape their use in both written and spoken contexts.193-208
36
日本語35巻4号2026年3月20日特集・論文 B陳 宇鍇専門家コミュニティと一般向けの科学技術日本語における
レジスターバリエーション―科学技術系の学術雑誌とnoteを対象としたMulti-Dimensional分析―		科学技術日本語，レジスター，Multi-Dimensional分析，因子分析，
サイエンスコミュニケーション，言語変異		本研究は，これまで科学技術日本語を専門家コミュニティ内で使用される日本語として分析する視野から一歩踏み出し，一般向けの科学技術日本語も射程に入れ，この二つの場面における言葉遣いの特徴を解明することを目的とする．具体的には，科学技術系の学術雑誌と科学技術系ハッシュタグを含むnote記事における日本語テキストを，それぞれ専門家コミュニティと一般向けの日本語と捉え，Multi-Dimensional分析を実施した．結果として，学術雑誌の言葉遣いでは「抽象表現」「漢字表現」が多く含まれており，note上の言葉遣いでは「インタラクティブ・主観表現」「叙述的表現」「平易・丁寧表現」に偏っていることを明らかにした．209-223
37
EnglishVol.35, No.4March 20, 2026Paper (B)CHEN YukaiRegister Variation in Technical Japanese across Expert Communities and the General Public:
A Multi-Dimensional Analysis of Scientific and Technical Academic Journals and the Note Platform		technical Japanese, register, multi-dimensional analysis, factor analysis,
science communication, language variation		Technical Japanese has been predominantly examined within the context of expert communities. However, its use in communication with the general public remains underexplored. This study aims to clarify the language patterns of technical Japanese in these two distinct contexts. Specifically, the study considers Japanese texts from scientific and technical academic journals as representative of expert communities, while texts from note (a Japanese blog platform) are regarded as representative of writing for the general public. After collecting both types of texts, a multi-dimensional analysis was conducted to examine their language patterns. As a result, the findings indicate that academic journals tend to contain more abstract expressions, as well as kanji expressions. In contrast, the language usage on note is more likely to exhibit interactive and subjective expressions, narrative expressions, and plain and polite expressions.209-223
38
日本語35巻4号2026年3月20日論文 A蘇 振軍日本語書き言葉における定式表現分析の試み
―n-gramモデルを用いて―		定式表現，コロケーション，表現文型，n-gramモデル本研究は，日本語の書き言葉における基礎的な定式表現の種類や使用割合を明らかにすることを目的とし，『現代日本語書き言葉均衡コーパス』における3つの「出版サブコーパス」を対象に，n-gramモデルを用いて3-gramから7-gramの高頻度語列を分析した．その結果，3レジスターに共通する定式表現には語順が固定化された表現文型と，実質語が複数組み合わさるタイプが多く，特に3-gramと4-gramにおいては大きな割合を占めることがわかった．また，その多くはコロケーションや文法コロケーションに分類された．さらに，2つ以上の実質語からなるコロケーションは，［名詞／助詞／動詞］の組み合わせが目立ち，より詳細な意味を表現する傾向が確認された．最後に，定式表現の割合については，コロケーションのような固定定式表現の割合が高いが，表現文型のような半固定定式表現も少なからずあり，書き言葉における語彙の規則性と柔軟性が共存していることが示唆された．224-239
39
EnglishVol.35, No.4March 20, 2026Paper (A)SU ZhenjunAn Attempt to Analysis of Formulaic Language in Written Japanese: Using the n-gram Modelformulaic language, collocations, patterns, n-gram modelThis study aims to clarify the types and usage frequency of basic formulaic language in written Japanese by analyzing the publication sub-corpus from the Balanced Corpus of Contemporary Written Japanese. We used the n-gram model to analyze high-frequency n-grams ranging from 3-grams to 7-grams. The results reveal that the formulaic language common across the three genres largely consist of patterns and combinations of multiple content words, with 3-grams and 4-grams occupying a significant proportion. Furthermore, many of these expressions are categorized as collocations. Additionally, formulaic language consisting of two or more content words tend to feature combinations of [noun/particle/verb], which express more detailed meanings. Finally, in terms of the proportion of formulaic language, fixed expressions such as collocations are found to occupy a high percentage, while semi-fixed expressions, such as patterns, are also present to a notable degree. These findings suggest that both the regularity and flexibility of vocabulary coexist in written Japanese.224-239
40
English35巻4号March 20, 2026論文 BDELFARIYADI Fahri; UTSUGI AkiraVariations of [ai] and [ee] in Modern Japanese:
Using the Corpus of Everyday Japanese Conversation		vowel sequences ai, ee-form, variants, modern Japanese, language variationIn everyday Japanese, vowel sequences such as [ai] potentially undergo coalescence, resulting in the ee-form. While studies have explored this language variation, including dialect-focused studies as well as a case-specific investigation, comprehensive analyses of its factors remain insufficient. This study analyzed the factors influencing the ai-ee form by extracting data from the Corpus of Everyday Japanese Conversation. Statistical analyses, such as regression analysis, were also employed. The findings revealed that the speaker’s sex played a crucial role in predicting the appearance of the ee-form. Particularly, the ee-form tends to emerge in male speakers’ utterances. Regarding intra-linguistic factors, the parts of speech acted as the most influential variable, with adjectives and auxiliary verbs likely to undergo the ee-form.240-255
41
日本語Vol.35, No.42026年3月20日Paper (B)デルファリヤディ ファーリ，宇都木 昭現代日本語における「アイ」と「エー」の変異
―日本語日常会話コーパスによる分析―		連母音アイ，エー形，変異体，現代日本語，言語変異日本語の日常会話においては，連母音アイが母音融合によりエー形へ変化することがある．この変異については現在まで方言に焦点を当てた研究や特定の事例を対象とした調査が行われてきたが，その要因については充分に論じられてきたとはいえない．本研究では，日本語日常会話コーパス（Corpus of Everyday Japanese Conversation）を用い，エー形の変異に関与する言語内外の要因についてロジスティック回帰分析を含む統計的分析を行った．その結果，エー形の出現においては話者の性別が最も重要な要因であることが明らかになった．特に，男性話者がエー形を使用する傾向があった．また，言語内部要因としては品詞が最も強い影響をもち，主に形容詞および助動詞に出現しやすい傾向にあった．240-255
42
日本語35巻4号2026年3月20日学会参加報告真田 治子"QUALICO2025（国際計量言語学会大会）
― 2025年6月26日～6月28日，於マサリク大学教養学部（チェコ）―"
256-260
43
EnglishVol.35, No.4March 20, 2026
Conference Report
SANADA Haruko"QUALICO 2025:
International Quantitative Linguistics Conference
Held at the Masaryk University in Brno, Czech Republic from June 26 to June 28, 2025"
The International Quantitative Linguistics Conference 2025 (QUALICO 2025) was held at the Masaryk University from June 26 to June 28, 2025. It was organized by the Department of Czech Language at the Faculty of Arts, Masaryk University, together with the International Quantitative Linguistics Association (IQLA). Three keynote lecturers were invited, and 34 papers for talks and five papers for a poster session were presented. Participants from 18 countries attended, including EU countries such as the Czech Republic, Spain, and Italy, as well as Canada, the U.S., Brazil, China, and Japan. Of the 34 papers, six were presented by Japanese researchers or by a foreign student studying in Japan. Additionally, a paper dealing with Japanese data was presented by researchers whose native languages are not Japanese. Most papers focused on classical or fundamental topics such as quantitative indices and valency theory; linguistic laws such as the Menzerath-Altmann’s law and stylometry (including authorship attribution); and large language models (LLMs). Several research presentations focused on AI. The premise of research in this field is the growing presence of AI-generated text in society and academia. Studies have examined how this affects human society and politics, including developing methods for AI detection (identifying whether content was generated by AI) and comparing corpora containing AI-generated texts with those composed of human-authored texts. The IQLA Council Business Meeting was held and new board members were selected. The next conference in Barcelona has been planned for 2027. A call for papers for the next conference will be announced on IQLA’s website.
256-260
44
日本語35巻5号2026年6月20日論文 B西内 沙恵味覚を表す多義的形容詞における語形の偏り―単義語との比較及び類似度評定から―多義語，語形，活用形，コーパス調査，類似度調査，現代日本語形容詞，味覚
本研究では多義語の語義ごとに特定の語形に偏って現れる使用実態に対して，既存のデータベースを援用し，要因の究明を試みる．対象は先行研究で扱われる味覚を表す現代日本語形容詞とした．まず，先行研究によって示された，語の意味と用法の関連を再現するために，『現代日本語書き言葉均衡コーパス』の分類語彙表情報（加藤・浅原・山崎2019）を活用し，語義ごとの活用形・語形の出現頻度を調べた．次に，偏りが確認された語形に焦点を絞り，語義と語形の結び付きを調べるために，類似の語義を持つ単義語の出現傾向と比較した．最後に，類似度評定（西内2025）を用いて，語形と用例の類似度に基づき分析を行った．調査の結果，語義ごとの活用形・語形の偏りが一部再現できた．再現できた語形の偏りには，単義語の語形の出現傾向と一致しない場合があったことから，語義の意味的な動機付けによって語形が偏っているわけではないと分析した．さらに，類似度の高低から，語義と特定の語形の関連に意味的な動機付けによらない結び付きを示した．261-276
45
EnglishVol.35, No.5June 20, 2026Paper (B)NISHIUCHI SaeTendencies in Word Forms of Polysemous Adjectives Expressing Taste:
A Comparison with Monosemous Words and Similarity Judgments
polysemous words, word form, inflectional form, corpus study,
similarity judgments, contemporaneous Japanese adjectives, taste
This study aimed to investigate the factors underlying the phenomenon in which specific word forms of polysemous words tend to appear preferentially, depending on their respective senses. In this investigation, we focused on contemporaneous Japanese adjectives that express taste, as addressed in previous research. First, we examined whether certain senses of polysemous words are more likely to appear in specific inflectional forms and word forms by analyzing the frequency of these forms according to sense, using the annotation of Word List by Semantic Principles for the Balanced Corpus of Contemporaneous Written Japanese. Next, in order to examine how specific word forms are associated with particular senses, we conducted a corpus-based comparison of the occurrence tendencies in monosemous words. Through this analysis, we confirmed that specific word forms are not necessarily linked directly to particular senses. Furthermore, we analyzed the degree of similarity between specific word forms and usage examples using a similarity judgment survey. The results showed that although biases in inflectional forms were observed depending on the senses, the occurrence patterns of word forms in monosemous words did not align with those of particular senses in polysemous words. This finding indicates that the bias in word forms is not attributable to the senses themselves. On the other hand, the analysis of similarity ratings suggested that in some cases, specific word forms were indeed associated with certain senses of polysemous words. 261-276
46
日本語35巻5号2026年6月20日招待論文・書評大下 由乃丸山岳彦（編）（2025）『コーパスで学ぶ日本語学　日本語の文法・音声』朝倉書店
277-281
47
EnglishVol.35, No.5June 20, 2026Book ReviewOSHITA YoshinoOSHITA Yoshino: MARUYAMA Takehiko (ed.) (2025).
Japanese Grammar and Phonology,
Series on Corpus Corpus-based Learning of Japanese. Tokyo: Asakura ShotenShoten.		277-281
48
日本語35巻5号2026年6月20日招待論文・書評于 拙麻子軒（著）（2024）『無生物主語他動詞文の日中対照研究―大規模均衡コーパスと多変量解析を用いた新たなアプローチ』関西大学出版部
282-286
49
EnglishVol.35, No.5June 20, 2026Book ReviewYO CjyetMA Tzuhsuan . (2024). A Comparative Study of Japanese and Chinese Inanimate
Subject Transitive Verb Sentences: A New Approach Using Balanced Corpora and
Multivariate Analysis. Suita: Kansai University Press. 282		282-286
50
日本語35巻5号2026年6月20日解説林 直樹データで語る日本語（2）―方言データの構築・管理・分析から活用まで―
方言，音声，語彙，文法，談話，言語地図，言語意識，言語行動本稿では，方言研究を音声・語彙・文法・談話・言語地図・言語意識・言語行動に大別し，それぞれの領域におけるデータの構築方法や分析方法，注意点について述べる．それぞれの領域で扱われるデータは異なることが多いため，その構築・管理・分析においても注意すべき点が異なる．また，音声のように分析に際して特定のソフトウェアを用いることもあるため，その扱い方についても簡単に示す．これらの解説を通じて，計量的手法を適用した方言研究の現状についても概観する．287-300
51
EnglishVol.35, No.5June 20, 2026TutorialHAYASHI Naoki Revealing Japanese Through Data (2): How to Build, Manage, Analyze, and Utilize Dialect Data
Dialect, Phonetics, Vocabulary, Grammar, Discourse, Linguistics Atlas, Language Consciousness, Language Behavior
This paper broadly categorizes dialect research into phonetics, vocabulary, grammar, discourse, linguistic atlas, language consciousness, and language behavior. It describes data construction methods, analytical approaches, and key considerations for each domain. Since the data types handled often differ across domains, the points requiring attention in data construction, management, and analysis also vary. Furthermore, since specific software may be required for analysis in certain domains (e.g., phonetics), I briefly outline how to handle such tools. Through these explanations, I also examine the current state of dialect research employing quantitative methodologies.287-300
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100