Nanori (名乗り): The Name-Only Kanji Readings That Break the On/Kun Rules

Nanori (名乗り) are kanji readings that appear only, or almost only, in personal names and certain place names. They sit outside the tidy on'yomi and kun'yomi split a learner has just memorized.¹² You can know every kanji in a Japanese name and still be unable to pronounce it, because the reading belongs to the bearer rather than to any dictionary entry.

Overview

Kanji name readings behave differently from ordinary vocabulary readings. Standard kanji dictionaries publish a separate "Names" column alongside on'yomi and kun'yomi, and the readings listed there are often longer than the everyday list combined.³⁴

This article places nanori inside the broader reading system and sketches the small set of patterns that make most names recognizable. It then draws the boundary against jukujikun and ateji, and ends with the daily-life consequences: business-card furigana, the 2025 family-register reform, and the social norm of asking.

What Nanori Is

Nanori as the Third Reading Bucket

名乗り (なのり) refers to kanji readings that appear only, or almost only, inside personal names and certain place names. The term itself comes from the verb 名乗る, "to declare a name."¹²

Standard Japanese-English kanji dictionaries split per-character readings into three labeled lists: on, kun, and name (名乗り / Names). They treat nanori as a distinct category from kun.³⁴ Linguistic typology, which compares writing systems by type, likewise separates nanori from kun, on, jukujikun, and ateji. The reason is that nanori readings are licensed by name use rather than by a lexical entry, or ordinary dictionary word.⁵⁶⁷

漢字辞典かんじじてんでは、音読おんよみと訓読くんよみのほかに「名乗なのり」の欄らんがあります。³⁴
"Kanji dictionaries include a 'nanori' column in addition to on'yomi and kun'yomi."

For many characters, the nanori list is longer than the on plus kun list combined. High-frequency name kanji such as 一, 真, 生, 義, and 之 each carry well over a dozen nanori entries in the major reference dictionaries.³⁴⁸

Nanori as a fossil layer

Many nanori readings are old-fashioned pronunciations of the character that have since fallen out of ordinary vocabulary. The dictionary category is, in part, a fossil layer of earlier Japanese preserved by name use.⁵⁹

The reading system a learner has already met (on'yomi vs. kun'yomi) is the anchor. Nanori is best thought of as a third bucket beside those two, not as a quirk inside one of them.

What Counts and What Does Not

A reading is treated as nanori when major kanji dictionaries list it under 名乗り / Names and it does not appear under the dictionary's ordinary-vocabulary readings. The boundary is editorial, so dictionaries disagree on borderline cases.³⁴¹⁰

Rare kun'yomi that happen to be common in names are not nanori. Reading 中 as naka, for example, is the standard kun and is simply being used inside a name. It never moves out of the kun column.³⁸

「中なか」を「なか」と読よむのは訓読くんよみであって、名乗なのりではありません。³⁸
"Reading 中 as naka is a kun'yomi, not a nanori, even though it is common in names."

The line between "obsolete kun" and "nanori" is not firm. Readings labeled nanori in one dictionary are labeled rare kun in another.⁸

「希き」を「のぞみ」と読よむのは、辞書じしょによっては名乗なのりに分類ぶんるいされます。³¹⁰
"Reading 希 as Nozomi is classified as a nanori in some dictionaries."

Kodansha Learner's lists nanori under a dedicated N label for each character, after on and kun, and excludes readings the editors judge too idiosyncratic.³ Nelson's New Japanese-English Character Dictionary uses the heading "Names" in a similar way.⁴ Daijirin and Kōjien work at the word level rather than the character level, so they do not maintain a dedicated nanori list per character. Instead, they label name-only readings inline.¹²

Why Names Break the On/Kun Rules

Names Predate the Tidy Reading System

Personal names were written with kanji from the earliest attested Japanese inscriptions, well before modern dictionaries separated on and kun in a standard way.¹¹⁹ Man'yōgana, the early use of kanji for their sound regardless of meaning, supplied the mechanism. It let arbitrary readings attach to name-writing centuries before any on/kun distinction was codified.¹¹⁵

名前なまえの漢字かんじは、辞書じしょの整理せいりよりずっと古ふるくから使つかわれています。¹¹⁹
"Name kanji have been in use long before dictionaries organized the readings."

元服 (genpuku), the coming-of-age ceremony documented from the Nara and Heian periods, involved the bestowal of an adult name. In the warrior class, this adult name was sometimes called 烏帽子名 (eboshi-na).¹²¹³ The bestowed name was treated as the bearer's own. That makes it reasonable to describe a name-as-personal-property pattern that long pre-dates any modern dictionary.

元服げんぷくのとき、新あたらしい名前なまえが与あたえられました。¹²¹³
"At the genpuku ceremony, a new name was bestowed."

Many readings labeled nanori in modern dictionaries come from this pre-standardization era. They survive in names even when the corresponding ordinary-vocabulary reading has shifted or disappeared.⁵¹²⁹

Names Carry Meaning the Sound Alone Cannot

In Japanese naming practice, kanji are selected for their meanings and associations (strength, light, hope, harmony). The reading is selected, partly independently, for euphony, or how good it sounds. This two-axis selection is the structural reason the on/kun system cannot predict name readings.¹²⁸¹⁴

親おやは、漢字かんじの意味いみと音おとの両方りょうほうを考かんがえて名前なまえを決きめます。¹²¹⁴
"Parents choose a name by weighing both the kanji's meaning and its sound."

The same target sound can therefore be written with many different kanji. The same kanji can also carry many different target sounds, depending on which axis the parents prioritized.⁸¹⁴

This two-axis freedom is not a modern invention. Pre-war family registers already contain readings that no ordinary dictionary lookup could resolve.¹⁴

The Legal Layer: Jinmeiyō Kanji and the Koseki

Article 50 of the Family Register Act (戸籍法第50条) restricts which kanji are legally usable in given names on the koseki, Japan's family register. The allowed set is the jōyō kanji list (the 2,136-character set fixed by the 2010 Cabinet Notification) plus the supplementary jinmeiyō kanji list of 863 characters, for a combined total of 2,999.¹⁵¹⁶¹⁷

子供こどもの名前なまえに使つかえる漢字かんじは、常用漢字じょうようかんじと人名用漢字じんめいようかんじに限かぎられています。¹⁵¹⁶
"The kanji usable in a child's name are limited to the jōyō and jinmeiyō lists."

The Ministry of Justice maintains the jinmeiyō list, and its post-war composition has been expanded repeatedly in response to parental petitions.¹⁷¹⁸

Surnames are not on this list

The jōyō + jinmeiyō constraint applies to given names only. Surnames are inherited and may contain characters outside both lists. Only the given-name slot faces the legal filter at registration.⁸¹⁸

A 2023 amendment to the Family Register Act, 戸籍法等の一部を改正する法律 (令和5年法律第48号), was promulgated on 9 June 2023 and took effect on 26 May 2025. It added 振り仮名 (furigana) as a registered item on the koseki for the first time.¹⁹²⁰

令和れいわ七年ななねん五月ごがつ二十六日にじゅうろくにちから、戸籍こせきに名前なまえの振ふり仮名がなも記載きさいされるようになりました。¹⁹²⁰
"From 26 May 2025, the family register also records the furigana of the name."

Under the post-2025 system, the registered reading must be 一般に認められているもの ("generally accepted as a reading of the kanji used"). Ministry of Justice guidance lists examples of disallowed categories. These include readings opposite in meaning to the kanji, readings easily confused with a different name, and readings unrelated to the kanji's meaning or sound.¹⁹²¹

漢字かんじの意味いみと反対はんたいの読よみ方かたは認みとめられません。¹⁹²¹
"A reading opposite in meaning to the kanji is not accepted."

Existing residents have a transition window from 26 May 2025 to 26 May 2026 to file their declared reading. After that date, the municipality may register the reading administratively from the notice it has already sent.¹⁹

振ふり仮名がなの届出とどけで期間きかんは一年いちねんです。¹⁹
"The notification window for the furigana is one year."

The restriction is on the reading, not on parental creativity as a whole. Permissible readings still include partial-character readings (心愛 Kokoa), silent-character readings (彩夢 Yume), and compound-as-a-whole readings (飛鳥 Asuka), provided the parents can justify them at registration.²¹

Common Nanori Patterns

Shortened Kun'yomi

The most productive nanori pattern is a kun'yomi with its trailing syllables dropped. Most family-name nanori arise this way.³⁸ 飯 ii in 飯田 is the standard worked example: the surname is read Iida, with 田 supplying its standard kun da.³²²¹⁰

飯田いいだは、飯いいを「いい」、田たを「だ」と読よみます。³²²
"Iida is read with 飯 as ii and 田 as da."

Many "Iida-style" family names pair a clipped nanori in the first slot with a regular kun in the second.²²¹⁰

訓読くんよみの最後さいごの音おとが落おちて、名前なまえの読よみになることがあります。³⁸
"The final sound of a kun'yomi can drop and become a name reading."

Archaic or Variant Kun'yomi

A second pattern preserves earlier kun readings that have fallen out of ordinary vocabulary. These readings are fossilized, not invented.⁵⁸⁹ 真 masa is the textbook case: the modern kun is ma (as in 真心 magokoro) and the modern on is shin, but masa survives as a productive name element across the family of 雅 / 政 / 正 / 真.³¹⁰

真しんを「まさ」と読よむのは、古ふるい訓読くんよみが名前なまえに残のこった例れいです。³¹⁰
"Reading 真 as masa is an old kun'yomi surviving in names."

信 nobu, 義 yoshi, and 之 yuki behave similarly. Their nanori reflect kun readings attested in older Japanese and now confined to the naming layer.³⁴⁸

Meaning-Driven Readings

A third pattern licenses a reading from the kanji's semantic field, or range of meanings, rather than from any historical kun. A related native word is read on the character because the meaning lines up.⁸¹⁴ 希 read as Nozomi in a female given name is the textbook case. 希 means "hope, rare" and its standard readings are ki, ke, and mare, but nozomi, the native noun meaning "hope," is licensed as a nanori on the meaning axis.¹³¹⁰

希きを「のぞみ」と読よむのは、意味いみから導みちびかれた名前なまえの読よみです。¹⁸
"Reading 希 as Nozomi is a name reading derived from the kanji's meaning."

Sasahara (2006) treats this pattern as the main engine behind much of the post-war expansion of nanori inventories.⁸

Name-Only Morphemes

A small set of nanori syllables appears only as final elements of given names and never as ordinary vocabulary. The well-known cases are -suke (之介, 之助), -rō (太郎, 一郎), and -ko (子, final in many female given names).⁸¹⁴ These behave as bound elements, or pieces that cannot stand alone, in the name-formation system rather than as alternative readings of any free word.⁸

太郎たろうの「ろう」は、名前なまえの最後さいごにしか使つかいません。⁸¹⁰
"The rō in Tarō appears only at the end of names."

Some high-frequency name characters carry dozens of nanori in a single dictionary entry. 之 alone lists more than thirty in standard Japanese kanji references.³⁴¹⁰

之ののような字じは、名前なまえの中なかにしか出でてこない読よみをたくさん持もっています。³⁸
"Characters like 之 carry many readings that only appear in names."

The Kirakira Edge Case

キラキラネーム (kirakira names) is the journalistic term for modern given names whose registered reading has little or no relationship to the standard readings of the chosen kanji.²¹²³ The standard worked example is 光宙 read Pikachū (the kanji denote "light" and "space / heaven"). A second often-cited case is 月 read Raito, with the kanji meaning "moon."²¹²³

光宙こうちゅうと書かいて「ピカチュウ」と読よむ名前なまえは、キラキラネームの代表例だいひょうれいです。²¹²³
"The name written 光宙 and read Pikachū is the textbook kirakira example."

Kirakira names operate by the same separation between character and sound that licenses nanori in general. What makes them distinctive is the distance between the chosen reading and any dictionary-attested reading of the kanji.⁸²³ The same mechanism, sound-driven kanji choice, is also the engine of ateji loanwords. That places kirakira names at the junction of nanori freedom and ateji-style sound-first selection.²³²⁴

The 2025 koseki reform is the legal mechanism that constrains how far this distance can go. The "generally accepted reading" rule prevents, for example, a meaning-opposite reading like 高 (high) registered as hikushi ("low").¹⁹²⁰²¹

2025年ねん5月がつの改正かいせいで、極端きょくたんな読よみ方かたは登録とうろくできなくなりました。¹⁹²¹
"With the May 2025 amendment, extreme readings can no longer be registered."

Nanori vs. Jukujikun vs. Ateji

Three Things That Look Alike

Nanori, jukujikun, and ateji are three different categories. They are distinguished by scope (single character vs. whole compound) and by direction of fit (kanji-to-sound vs. sound-to-kanji).⁶²⁴⁷

• Nanori is a per-character reading licensed by the name context; the reading attaches to the individual kanji.³²⁴
• Jukujikun is a per-compound reading: kanji are chosen for their meaning, but the reading is assigned to the whole word, not to its parts.²⁴⁷¹⁰
• Ateji is a sound-first selection: kanji are chosen for their sound, with the meaning of the chosen characters often incidental; canonical loanword examples include 寿司 sushi, 倶楽部 kurabu "club," and 珈琲 kōhī "coffee."²⁴⁷

Category	Scope	Direction of fit	Canonical example
Nanori (名乗り)	Single character, in a name	Meaning or history first, sound matched	飯 ii in 飯田 Iida322
Jukujikun (熟字訓)	Whole compound	Meaning first, reading attached to the whole	飛鳥 Asuka, 今日 kyō, 大人 otona24710
Ateji (当て字)	Single or compound	Sound first, meaning of kanji often incidental	寿司 sushi, 倶楽部 kurabu, 珈琲 kōhī247

The three can overlap in edge cases. A name written 飛鳥 Asuka behaves as both a nanori and a jukujikun-class reading. Dictionaries label such cases inconsistently.²²²⁵¹⁰

当あて字じは、漢字かんじの意味いみより音おとを優先ゆうせんします。²⁴⁷
"Ateji prioritizes sound over the meaning of the kanji."

When a Name Is Actually Jukujikun

Some readings traditionally listed under nanori, especially in toponyms (place names), are better analyzed as jukujikun: the reading binds to the compound as a unit, not to either character individually.²⁴⁷¹⁰ In practice, the two categories overlap in toponyms rather than dividing along a hard line. Dictionaries sort the same place names into either column depending on editorial policy.⁸²⁴

飛鳥 Asuka is the standard case. Neither 飛 nor 鳥 carries a, su, or ka as an attested per-character reading. The compound reading derives from the makura-kotoba, or fixed poetic epithet, 飛ぶ鳥の (tobu tori no), historically attached to the older spelling 明日香.²⁵¹⁰

飛鳥あすかを「あすか」と読よむのは、字じごとの読よみでは説明せつめいできません。²⁴²⁵
"Reading 飛鳥 as Asuka cannot be explained character by character."

Sasahara (2010) treats this border explicitly. He sorts place-name readings into a nanori-and-jukujikun continuum rather than a strict either/or.²⁴

Reading Japanese Names in Daily Life

Furigana Is Not Optional on a Business Card

Japanese business cards (meishi 名刺), official application forms, school rosters, and hospital intake sheets routinely carry a dedicated furigana field for the name. The field is practical, not decorative.¹⁹²⁶

名刺めいしには、漢字かんじの上うえに振ふり仮名がなが書かいてあります。¹⁹²⁶
"Business cards carry furigana written above the kanji."

The 2025 koseki reform makes furigana a registered item of the family register itself. It formalizes a convention that daily-life paperwork had already enforced.¹⁹²⁰

Asking is part of the meeting script

Teaching reference works treat asking how to read a name on first meeting as a normal, polite act. The asker is seen as competent rather than ignorant.²⁶ The phrasing varies, but the question is built into the introduction routine even between Japanese speakers.

When You Do Not Know, Ask

No algorithm reliably resolves an unknown name reading from kanji alone. Name-reading lookup tools list multiple nanori candidates per character. Combining candidates across two characters multiplies the search space, and the right answer is selected by the bearer, not by a lookup procedure.⁸²⁶

辞書じしょだけでは、読よみ方かたは決きめられません。⁸
"Reading cannot be settled from a dictionary alone."

When a name is unfamiliar, the socially correct procedure is to ask the bearer. Japanese speakers do this with each other routinely, and the question is built into the meeting-someone-new script.²⁶

日本人にほんじん同士どうしでも、初対面しょたいめんでは名前なまえの読よみ方かたを確認かくにんします。²⁶
"Even Japanese speakers confirm name readings with each other on first meeting."

Good to know

The Same Name Kanji, Two Families, Two Readings

A single family-name kanji string can carry different readings in different families. 中田 may be Nakata or Nakada, and the choice belongs to the family rather than to the region.⁸¹⁴ Lookup tools list both candidates without resolving them. The bearer's family is the source of truth.⁸

中田なかたは、家いえによって「なかた」とも「なかだ」とも読よみます。⁸¹⁴
"中田 is read either Nakata or Nakada, depending on the family."

Place Names Use Nanori Too

Nanori extends to toponyms, or place names. 飛鳥 Asuka is the standard linguistic example, with the reading derived from the makura-kotoba 飛ぶ鳥の attached to the older spelling 明日香.²⁵¹⁰ Smaller-scale local readings (small-town surnames, hamlet names) often resist prediction by non-local readers. The daily-life rule for the learner is the same as for personal names: look up, and ask if the lookup fails.²²²⁶

飛鳥あすかは奈良県ならけんの地名ちめいで、「あすか」と読よみます。²²²⁵
"飛鳥 is a place name in Nara Prefecture, read Asuka."

Why IMEs Still Convert Kira-Kira Names

Japanese input-method editors (IMEs) convert from the declared reading to the registered kanji because the koseki reading, not the dictionary reading, is what counts for legal identity. A registered name typed with its registered reading converts to its registered kanji, whether or not the reading is "in" the dictionary.¹⁹⁸

辞書じしょにない読よみでも、本人ほんにんが使つかっていれば名前なまえとして通用つうようします。⁸
"Even a reading not in the dictionary works as a name if the bearer uses it."

This is a workflow accommodation, not an endorsement of pure invention. The post-2025 reform constrains which readings reach the koseki in the first place.¹⁹²¹

The Reading Was Always the Owner's Property

In classical Japanese practice, the 名乗り was the name the holder declared, and the reading travelled with the person rather than with the character.¹²¹³ 元服 (genpuku) bestowed an adult name. In the warrior class, the bestowed name was the eboshi-na (烏帽子名), and in the aristocratic record it was sometimes called the imina (忌み名).¹²¹³

The situation in which the koseki reading, not the dictionary reading, defines a person's name is continuous with this older arc. The 2025 reform makes the reading a formally registered item, not a wholly new one.¹⁹¹²

現代げんだいの戸籍こせきは、その伝統でんとうを法律ほうりつの形かたちにしたものです。¹⁹¹²
"The modern koseki gives that tradition legal form."

See also

• Go-on, Kan-on, Tō-on: The Historical Layers Behind a Kanji's Multiple On'yomi
• Hyōgaiji (表外字): The Kanji Beyond Jōyō and Jinmeiyō
• The History of Kanji: From Oracle Bones to the Jōyō List

References

Footnotes

1. 三省堂編修所 編. 『大辞林 第四版』. 三省堂, 2019. Entry: 名乗り (なのり). ↩ ↩² ↩³ ↩⁴ ↩⁵

2. 新村出 編. 『広辞苑 第七版』. 岩波書店, 2018. Entry: 名乗り (なのり). ↩ ↩² ↩³

3. Halpern, Jack (ed.). The Kodansha Kanji Learner's Dictionary, Revised and Expanded. Kodansha USA, 2013. "Guide to the Dictionary", section on name readings (名乗り) and per-character nanori listings. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³ ↩¹⁴ ↩¹⁵ ↩¹⁶ ↩¹⁷ ↩¹⁸ ↩¹⁹ ↩²⁰ ↩²¹

4. Nelson, Andrew N. (rev. Haig, John H.). The New Nelson Japanese-English Character Dictionary. Tuttle Publishing, 1997. Introduction on reading categories and per-character "Names" lines. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸

5. Frellesvig, Bjarke. A History of the Japanese Language. Cambridge University Press, 2010, pp. 158–164 (kun readings, name attestations) and pp. 280–285 (modern reading layers). ↩ ↩² ↩³ ↩⁴ ↩⁵

6. Bjarke Frellesvig and Stephen Wright Horn (eds.). The Oxford Handbook of Japanese Linguistics. Oxford University Press, forthcoming editions; chapters on the writing system distinguish kun, on, jukujikun, ateji, and nanori as separate reading categories. (Where this chapter is unavailable, ⁵ covers the same material.) ↩ ↩²

7. Shibatani, Masayoshi. The Languages of Japan. Cambridge University Press, 1990, pp. 128–135 (the writing system and reading categories). ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸

8. 笹原宏之. 『日本の漢字』. 岩波新書, 2006, chapters on name kanji, jinmeiyō revisions, and reading variation across families and regions. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³ ↩¹⁴ ↩¹⁵ ↩¹⁶ ↩¹⁷ ↩¹⁸ ↩¹⁹ ↩²⁰ ↩²¹ ↩²² ↩²³ ↩²⁴ ↩²⁵ ↩²⁶ ↩²⁷

9. Habein, Yaeko Sato. The History of the Japanese Written Language. University of Tokyo Press, 1984, pp. 22–35 (early Japanese personal names and their kanji writing). ↩ ↩² ↩³ ↩⁴ ↩⁵

10. 日本国語大辞典 第二版. 小学館, 2000–2002. Entries: 名乗り, 飛鳥, 飯, 真, 希. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³ ↩¹⁴ ↩¹⁵

11. Seeley, Christopher. A History of Writing in Japan. University of Hawaiʻi Press, 2000, pp. 41–48 (man'yōgana, early personal names) and pp. 137–145 (script standardization). ↩ ↩² ↩³

12. Plutschow, Herbert E. Japan's Name Culture: The Significance of Names in a Religious, Political and Social Context. Japan Library / Routledge, 1995, pp. 53–78 (genpuku, nanori, eboshi-na, imina). ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹

13. 加藤友康 ほか 編. 『年中行事大辞典』. 吉川弘文館, 2009. Entry: 元服 (genpuku); discussion of 烏帽子名 and adult-name bestowal. ↩ ↩² ↩³ ↩⁴

14. 佐藤稔. 『読みにくい名前はなぜ増えたか』. 吉川弘文館, 2007. Survey of unusual name readings and their pre-war and post-war ancestors. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸

15. 文化庁. 「常用漢字表」(平成22年内閣告示第2号), 2010. The 2,136-character base list against which jinmeiyō is the legal supplement. https://www.bunka.go.jp/kokugo_nihongo/sisaku/joho/joho/kijun/naikaku/kanji/ ↩ ↩²

16. 法務省民事局. 「子の名に使える漢字」案内. https://www.moj.go.jp/MINJI/minji86.html ; 戸籍法第50条 (the article restricting given-name kanji to jōyō plus jinmeiyō). ↩ ↩²

17. 法務省. 「人名用漢字一覧」. https://www.moj.go.jp/MINJI/minji86.html (jinmeiyō kanji list, total 863 characters as of the most recent revision). ↩ ↩²

18. 円満字二郎. 『人名用漢字の戦後史』. 岩波新書, 2005, on the post-war jinmeiyō list, its expansions, and the social pressure that drove them. ↩ ↩²

19. 法務省. 「戸籍にフリガナが記載されます」. 法務省民事局. https://www.moj.go.jp/MINJI/furigana/index.html (改正戸籍法施行日: 令和7年5月26日 / 26 May 2025). ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³ ↩¹⁴ ↩¹⁵

20. 法務省. 「戸籍法等の一部を改正する法律」(令和5年法律第48号), 公布: 令和5年6月9日, 施行: 令和7年5月26日. https://www.moj.go.jp/MINJI/minji04_00088.html ↩ ↩² ↩³ ↩⁴

21. 日本経済新聞. 「キラキラネームに歯止め 改正戸籍法、来年5月施行」, reporting on the Ministry of Justice guidance for the May 2025 reform and the three categories of disallowed readings (反対の意味、誤認、無関係の音). (Used only for the explicit three-category framing; the underlying rule is in ¹⁹ and ²⁰.) ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹

22. NHK放送文化研究所. 『NHK日本語発音アクセント新辞典』. NHK出版, 2016. Entries on place-name and surname readings, including 飛鳥 (あすか) and 飯田 (いいだ). ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷

23. 笹原宏之. 「キラキラネーム」関連解説, in 『漢字の歴史』補論. 岩波書店, 2014 (academic background on the kirakira phenomenon and its relationship to nanori freedom). ↩ ↩² ↩³ ↩⁴ ↩⁵

24. 笹原宏之. 『当て字・当て読み 漢字表現辞典』. 三省堂, 2010. Reference work distinguishing 当て字 (ateji), 熟字訓 (jukujikun), and 名乗り (nanori) with worked examples. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹²

25. 飛鳥資料館. 「飛鳥 (あすか) という地名について」. 奈良文化財研究所. https://www.nabunken.go.jp/asuka/ (place-name etymology, the 飛ぶ鳥の "tobu tori no" makura-kotoba). ↩ ↩² ↩³ ↩⁴ ↩⁵

26. Backhouse, A. E. The Japanese Language: An Introduction. Oxford University Press, 1993, pp. 67–80 (kanji readings and the distinction between dictionary readings and name readings). ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷