How to Write Your Name in Katakana: Foreign-Name Transcription Rules with Examples
Writing your name in katakana comes down to one shift in mindset: spell the sounds of your name, not its English letters. Fit those sounds into Japanese mora and choose the closest available kana for each.1 The 1991 Cabinet Notification on loanword notation supplies the kana inventory behind these rules. The same system that handles common-noun loanwords also handles foreign personal names.23
Overview
Why foreign names use katakana
In modern Japanese, katakana writes 外来語 (gairaigo, foreign loanwords), foreign place names, and foreign personal names.234 The 1991 Cabinet Notification 外来語の表記 defines itself as the reference for "the writing of loanwords in general social life." Its appendix lists worked entries for both place names (marked 地) and personal names (marked 人).23
Visually, katakana signals that a token came from outside the native vocabulary. This gives the reader a parsing cue and tells them to read the spelling as phonetic, not morphemic.41 NINJAL's public Q&A describes 外来語の表記 as the よりどころ (reference) for converting non-native sound to kana. The Notification's 第2表 is reserved for spellings that aim to stay as close as possible to the source pronunciation or spelling.234
Kanji 当て字 (ateji) for foreign names appears in historical Sinographic names (孔子, 釈迦) and a small set of Western names rendered with phonetic kanji (亜米利加 → America, 倫敦 → London). But for modern Western personal names, the dictionary-default writing system is katakana, not ateji.3 Hiragana is not used for foreign names in standard prose.34
What this article covers (and what it does not)
In scope: Western, chiefly English, personal names written in katakana. This includes the rules that generalise: mora-fitting, vowel insertion, l→r, th→s/z, v→b/v, double consonants, long vowels, extended-katakana fallbacks, and the most common edge cases.23415
Three areas are out of scope. Kanji 当て字 for foreign names is a separate spelling practice with its own conventions.3 Chinese and Korean names usually follow their own community-internal conventions (kanji on'yomi or hangul-based transliteration) rather than the gairaigo rules.3 The legal mechanics of how the residence card, 住民票, and 戸籍 record names are administrative-law topics, not spelling topics. This article stays on the orthography side of that line.67
The core principle: spell sounds, not letters
Hear your name, then fit it to katakana
Japanese is a mora-based script: every kana other than ン and ッ represents one CV unit (a consonant plus a vowel), optionally with a glide.89 An English name has to be re-cut into Japanese mora before kana selection begins.15
The workflow has three steps. Say the name aloud the way a native speaker says it. Segment that audio into Japanese-style CV mora. Then pick the closest available kana for each mora.15 Beginners often skip the first step. That is why romaji-driven conversions (Charles → チャルレス) produce broken katakana.1
This is a structural principle, not just a heuristic. Japanese has fewer phonemic contrasts than English: no /l/–/r/ contrast, no /θ/–/ð/, no contrastive /v/, no /f/ outside フ, no /ti/–/di/, and no /tu/–/du/ in the base inventory. Transcription therefore has to project the source phonology onto the smaller Japanese inventory.189 The kana you pick are determined by sound, not by the source spelling.41
マイケル・ジャクソンはアメリカの歌手でした。110
"Michael Jackson was an American singer."
Why the result is an approximation, not a translation
Foreign-name katakana is a transcription. It is not a translation, and it is not a one-to-one transliteration of the source spelling.41 Two careful Japanese transcribers can render the same English name slightly differently, and both spellings can be defensible. The source phonology offers more contrasts than the target inventory can match.341
文化庁's reference text states this directly. 外来語の表記 explicitly permits two licensed spellings for the same word in many cases. バイオリン and ヴァイオリン both appear in the cited 慣用 pairs, with Table 2 offering the closer-to-source option and Table 1 the more naturalised one.23 The same logic applies to personal names. デビッド and デヴィッド both correspond to "David," with the 第2表 form keeping closer to the /v/.31
NHK's broadcast-standard reference and major Japanese dictionaries record canonical spellings for well-known names. But they record conventions, not derivations: ヘミングウェイ, ベートーヴェン, シェークスピア exist because they crystallised in print, not because a rule mechanically produced them.310 For a learner's own previously unspelt name, the rules give a defensible first pass; the spelling that ends up on the residence card and bank account is the operative one thereafter.67
The substitution rules
L becomes R
Japanese has no phonemic contrast between /l/ and /r/. The ラ-row (ラ リ ル レ ロ) and its yōon extensions cover both English L and English R in transcription.18 The rule is stated plainly: Japanese does not have separate l and r sounds, and l- is normally transcribed using the kana that are perceived as representing r-.1 London becomes ロンドン by this rule.1
Worked names: Laura → ローラ (l→r, plus the long ō from English /ɔː/), Lily → リリー (l→r twice; final -y as long ī marked with ー), and Carl → カール (r-coloured /ɑːr/ collapses to a long vowel via ー; the L does not surface).111
Th becomes S, Z, or T depending on the sound
English /θ/ (voiceless th, as in think) maps to the サ-row.1 English /ð/ (voiced th, as in this) maps to the ザ-row.1 The Wikipedia entry on transcription gives "ゴッサム Gossamu 'Gotham'" for /θ/→s. It gives "マザー mazā 'mother'" for /ð/→z.1
Some older transcriptions use T (タ-row) in legacy spellings. テレサ for "Theresa" survives in published reference works alongside the more recent サ-row attempts.310 Where multiple spellings are attested, the canonical katakana spelling for that specific name overrides the general rule. Beth → ベス and Heather → ヘザー are the dictionary forms for these two names.111
V becomes B (often) or ヴ (sometimes)
/v/ is not a native Japanese phoneme. Speakers typically realise both spellings with a [b]-initial.812 The 1991 Cabinet Notification gives writers two licensed options. The 第1表 option is the ハ-行 + 濁点 (バ・ビ・ブ・ベ・ボ), which naturalises /v/ to /b/. The 第2表 option is the ヴ-row (ヴ・ヴァ・ヴィ・ヴェ・ヴォ), used for closer-to-source transcription.2312 The 第2表 carries the explicit heading 「原音や原つづりになるべく近く書き表そうとする場合に用いる仮名」.23
For personal names, convention decides the choice, not a fixed rule. David → デビッド (Table 1 ハ-行 default) and デヴィッド (Table 2 ヴ-row option) coexist. Victor → ビクター and ヴィクター coexist. The original /v/ is preserved orthographically in the Table 2 form even where the speaker still pronounces [b].312 The full Table 1 / Table 2 split for /v/ lives in the companion article on extended katakana.
ヴィクター・ユーゴーは有名なフランスの作家です。312
"Victor Hugo is a famous French writer."
F, W, and other gaps that the extended katakana solve
The base 五十音 has only フ in the F-column (phonetically [ɸu]). In modern use, only ワ and ヲ remain in the W-row.8 To write English /f/ and /w/ before vowels other than the surviving base cell, writers use the extended-katakana digraphs codified in the 1991 Notification: ファ・フィ・フェ・フォ for the F-row (第1表) and ウィ・ウェ・ウォ for the W-row (第2表).231
Worked names: Fiona → フィオナ (Table 1 フィ + base オ + base ナ), Frank → フランク (single-mora フ before a consonant cluster), and Wendy → ウェンディ (Table 2 ウェ + base ン + base デ + small ィ).111 The Table 1 versus Table 2 inventory and the historical motivation for it sit in the companion article on extended katakana for loanwords.
Ti and Tu need ティ and トゥ
In the base 五十音, チ is read as /tʃi/ ("chi") and ツ as /tsu/ ("tsu"). Neither cell covers English /ti/ or /tu/ directly.18 The 1991 Notification supplies ティ (テ + small ィ, 第1表) for /ti/ and トゥ (ト + small ゥ, 第2表) for /tu/.231 The extended F-row and W-row digraphs use the same small-vowel mechanism.
Worked names: Tim → ティム, Timothy → ティモシー, and Tina → ティナ. The older naturalised forms チム and チモシー are licensed by 第1表 substitution but rarely used in current print.3110 English first names that begin in /tu/ are rare; Tuesday → トゥーズデー is the typical pattern when the word is written.1
ティムさんは来年から東京に住みます。31
"Tim will live in Tokyo from next year."
Other common substitutions
Here is a short reference for less common substitutions.
| Source sound | Convention | Worked names | Source |
|---|---|---|---|
| /si/ | シ (the base inventory has only シ /ʃi/; スィ is stylised, not in 第1表 or 第2表) | Sid → シド, Sylvia → シルビア, Cindy → シンディ | 31 |
| /hu/ and /hju/ | フ or yōon ヒュ for the /hj/ glide | Hubert → ヒューバート | 1811 |
| /kw/ | クァ-row (第2表) or クア・クイ・クエ・クオ (Table 1 fallback) | Quincy → クインシー | 23111 |
| /ks/, /gz/ (English x) | クス between vowels or finally (often with sokuon); グザ・グジ for voiced /gz/ | Max → マックス, Felix → フェリックス, Xavier → ザビエル (via Spanish) | 1 |
Vowel insertion: why Japanese pads consonants
Japanese syllables end in a vowel or ン
The Japanese mora is structured as (C)V, as the moraic nasal ン, or as the first half of a geminate consonant ッ.89 Every consonant other than syllable-final ン therefore needs a following vowel when written in kana.89 English names that end in a consonant other than /n/ require an epenthetic, or "echo," vowel after the final consonant.59
Loanword-phonology studies treat the default epenthetic vowel as /ɯ/, the high central vowel realised in the katakana cell ウ.9 Quantitative studies of Japanese loanword adaptation report that more than 70% of epenthesis tokens use the default epenthetic vowel /ɯ/. The vowel's phonological underspecification, devoiceability, and short duration make it the cross-linguistically expected choice for default epenthesis.9 This is the linguistic basis for the "add ウ" rule that beginner textbooks teach as a heuristic.5
Which vowel gets inserted
The default is U (ウ-row) after most consonants: Tom → トム, Chris → クリス, Jim → ジム, Sam → サム.159
After T and D, the default becomes O (オ-row). The base mora ツ and ヅ would distort the consonant by adding an affricate or merging with /zu/.5 The convention is therefore Pat → パット, Brad → ブラッド, Ted → テッド.15 After CH and J, the default becomes I (イ-row), because the relevant mora are チ and ジ themselves. Mitch → ミッチ ends in チ, and Madge → マッジ ends in ジ.5
A source consonant /m/ before /b/, /p/, or /m/ becomes the moraic ン rather than receiving its own epenthetic vowel: computer → コンピューター, sample → サンプル, summit → サミット (with sokuon ッ before /m/ rather than ンム).5 The same ン handles word-final /n/ directly, with no echo vowel: Ben → ベン, John → ジョン, Brian → ブライアン.15
Consonant clusters get split
English permits onset clusters such as /br/, /sp/, /str/, /skr/, /spl/ and coda clusters such as /-nks/, /-mps/, /-rld/. No Japanese mora can carry these clusters.89 In transcription, each consonant in the cluster gets its own vowel, usually default ウ unless the rules above override it. The exception is an /m/ or /n/ that surfaces as ン.159
Worked examples from the source include spring → スプリング, scratch → スクラッチ, and kick → キック.1 The same mechanism produces Brian → ブライアン, Chris → クリス, Stephen → スティーブン, Springfield → スプリングフィールド, Jennifer → ジェニファー.111
Double consonants: the small ッ
When a stop consonant is "held"
The 1991 Notification lists 促音 ッ in 第1表 as a productive element of katakana loanword spelling, alongside 撥音 ン and 長音 ー.23 In loanwords, ッ appears before a voiceless obstruent (p, t, k, s, ch, sh) where the source language has either a geminated consonant or a short stressed vowel followed by a stop.1513
The canonical rule is that ッ in katakana is used mainly in loanwords to approximate foreign consonant clusters, especially after a short, stressed source vowel before a stop.113 The same mechanism produces Jack → ジャック (short /æ/ + /k/), Beck → ベック (short /ɛ/ + /k/), Patrick → パトリック (medial /tr/ becomes ト + リ, and final /k/ takes ッ + ク), and Becky → ベッキー (medial ッ + キ for the doubled /k/).1511
ッ counts as one mora. ジャック is three morae (ジャ + ッ + ク), not two.2313 The underlying mora-doubling mechanic is the same one that drives 小さい「つ」 in native words, applied to gairaigo spelling.
Where ッ does not appear
The general distributional constraint, drawn from the Sokuon reference, is this: ッ never appears at the beginning of a word or before a vowel. It rarely appears before a syllable that begins with n, m, r, w, or y, and it does not appear before voiced consonants (g, z, d, b) except in loanwords, distorted speech, or dialects.13
The "except in loanwords" carve-out matters for foreign names. A foreign name whose source language has a doubled or short-stressed voiced consonant may still take ッ in transcription (Eddy → エディー or エッディー coexist; Buddy → バディー or バッディー coexist). But learners should follow the dictionary-form spelling for a given name, not a self-derived rule.11310 In stable transcriptions, ッ spreads predictably only across voiceless obstruents. Everywhere else, the writer follows the listed spelling.1310
Long vowels: the chōonpu ー
English long vowels and stressed syllables get ー
外来語の表記 makes the chōonpu ー the default device for long vowels in loanwords: 「長音は，原則として長音符号「ー」を用いて書く」 ("Long vowels are, as a rule, written using the long-vowel mark 'ー'").2 The Notification's own example list includes エネルギー, オーバーコート, グループ, ゲーム, and パーティー.2 The rule extends without change to personal names, which the appendix lists in the same alphabetical 用例集 as common nouns.23
For names, ー appears in three patterns:
- English long vowels and diphthongs that are not in the エイ/オウ exception class: Mary → メアリー (final /-iː/ as ー), Lee → リー, Joan → ジョーン (/oʊ/ as ョー).2111
- R-coloured syllables that collapse to a long vowel in Japanese: Carl → カール, Mark → マーク, George → ジョージ.2111
- English -er, -or, -ar word endings, which the Notification explicitly covers: 「英語の語末の－er, －or, －ar などに当たるものは，原則としてア列の長音とし長音符号「ー」を用いて書き表す」. Heather → ヘザー and Victor → ビクター both follow this rule.2111
When you do not use ー
Short, unstressed English vowels do not take ー. The workflow is to map them to the corresponding short Japanese vowel and close the syllable with the next consonant or with the moraic ン.15 Tom → トム (short /ɒ/ → オ-row, no bar), Sam → サム (short /æ/ → ア-row), Ken → ケン (short /ɛ/ → エ-row + final ン).1511
English spelling is unreliable as a length signal. Tom and Tome are written almost the same in English but transcribe to トム versus トーム. Kim and Keem both transcribe to キム because the surface vowel length sits inside the same Japanese short-vowel cell.1 The chōonpu records perceived vowel duration, not the source language's romanized spelling.241
A note on the エイ vs エー split
外来語の表記 gives an explicit exception to the chōonpu default: 「「エー」「オー」と書かず，「エイ」「オウ」と書くような慣用のある場合は，それによる」 ("When there is an established convention of writing 「エイ」「オウ」 rather than 「エー」「オー」, follow that convention").2 The clause allows both spellings for the same source sound, leaving the choice to convention.2
For personal names, the same split applies. Faye → フェイ (エイ for the /eɪ/ diphthong feel), Kate → ケイト or ケート (the Joshu reference table lists ケート, but ケイト is widely seen for the diphthong-preserving option), and Jane → ジェーン (エー is conventional for this name in published reference works).211011 The same rule, anchored to 慣用, decides which spelling is "right" for a given name. There is no derivable mechanical answer. The detailed エイ versus エー split for general loanwords sits in the companion article on long vowels in katakana.
ジェーンさんは三日後に帰ります。2111
"Jane will return three days from now."
Name order: how to fill in a Japanese form
The default for foreign names
On Japanese administrative forms for foreign residents, the dominant convention is family-name-first. This matches the order recorded on the 在留カード (residence card) and the 住民票 (resident record).67 Japanese system forms list 姓 (family name) first, then 名 (given name). Foreign residents are expected to enter their name in the same order it appears on their residence card.6 The residence card itself records the romanized name. The katakana rendering is a reading aid (フリガナ) added on forms that ask for it.67
The convention covered here concerns how foreign residents fill in their own names on Japanese forms. A separate question is how Japanese names should be ordered when romanized in English. The 関係府省庁申合せ of 25 October 2019 directs that 公用文 (official documents) write Japanese names 姓→名 in the Roman alphabet from 1 January 2020 onward.14151617 The two policies share an order, but for different reasons.
Reading the form labels
氏名 (shimei) labels the full-name field on virtually all Japanese forms. 姓 (sei) is the family name slot, and 名 (mei) is the given name slot.6 The convention is to enter the surname in 姓 and the given name in 名. This mirrors the order on the residence card.67
フリガナ or カタカナ above a name field asks the writer to render the same name in katakana. This tells the form-processor how to pronounce it.6 For foreign residents, this is the operative katakana spelling. It follows the same word order as the romanized name, and Japanese institutions match the spelling registered there rather than any "linguistically correct" alternative.67
ミドルネーム (middorunēmu) is a separate field on some official forms, such as immigration applications. For routine paperwork, it is often folded into the 名 slot.6
Middle names
The katakana middle dot ・ (中黒, nakaguro, U+30FB) is the standard separator for multi-part foreign names written in kana. The Transcription into Japanese reference states that foreign phrases and names are sometimes transliterated with an interpunct separating the words. This mark is called a 中黒 (middle dot), as in ビル・ゲイツ (Bill Gates).118 The Japanese punctuation reference states the general rule: ・ separates foreign words and names when they are written in kana.18
For a three-part name like John David Smith, the conventional katakana rendering is ジョン・デイビッド・スミス (given · middle · family). The middle dot goes between each pair.118 On a form where 姓 and 名 are separate fields, the family name (スミス) goes in 姓. The given-plus-middle (ジョン・デイビッド) goes in 名, with the middle dot kept inside the 名 field.186
The middle dot is its own punctuation mark (U+30FB). It is not an ASCII period, a hyphen, or the European interpunct (U+00B7). It is also distinct from the chōonpu ー (U+30FC).18 On Japanese input methods, the IME produces it directly. Copy-pasting ASCII bullet characters or full stops as substitutes produces broken text.18
Romaji forms
Forms that ask for the name in Latin letters in addition to katakana conventionally use Modified Hepburn romanization. The Library of Congress adopted this system in 1989 as one of its ALA-LC romanizations.1920 Modified Hepburn uses macrons (ā, ī, ū, ē, ō) for long vowels. It also uses n rather than m before /b/, /p/, /m/ (so 群馬 is Gunma, not Gumma).1920
The 25 October 2019 関係府省庁申合せ governs Japanese names in 公用文, not foreign names. For foreign residents, the romanization on the form should match the residence card, regardless of order.141516 When the form is for a Japanese name, the 申合せ-directed order is 姓→名. The family name is often capitalized in full (for example, YAMADA Taro) so the field's order is unambiguous.17
Worked examples: ten Western names start to finish
Single-syllable and short names
| Name | Katakana | Mechanic in play | Source |
|---|---|---|---|
| Tom | トム | default ウ-row epenthesis after /m/ | 1511 |
| Sam | サム | default ウ-row epenthesis after /m/ | 1511 |
| Ben | ベン | final /n/ as moraic ン, no epenthesis | 1511 |
| Kate | ケイト or ケート | エイ/オウ exception or default ー | 2111 |
| Ken | ケン | final /n/ as moraic ン | 15 |
Names with L, R, and TH
| Name | Katakana | Mechanic in play | Source |
|---|---|---|---|
| Laura | ローラ | l→r (ラ-row), long ō for English /ɔː/ | 111 |
| Carl | カール | r-coloured /ɑːr/ becomes long ā via ー; final L does not surface | 111 |
| Heather | ヘザー | voiced th→z; final -er → ア-row long vowel per 1991 rule | 2111 |
| Beth | ベス | voiceless th→s; default ウ-row close on the final S | 111 |
| Thomas | トーマス | voiceless th→t (legacy), long ō, default ウ-row close on final S | 311011 |
Names with V, F, and W
| Name | Katakana (Table 1 / Table 2) | Mechanic in play | Source |
|---|---|---|---|
| David | デビッド / デヴィッド | v→b (Table 1) or v→ヴ (Table 2); sokuon ッ + ド for /d/ ending | 311211 |
| Victor | ビクター / ヴィクター | v→b or v→ヴ; final -or → ア-row ー per 1991 rule | 231211 |
| Fiona | フィオナ | F-row 第1表 digraph フィ; base オナ closes | 2111 |
| Wendy | ウェンディ | W-row 第2表 digraph ウェ; デ + small ィ for final /diː/ | 23111 |
| Frank | フランク | base フ before consonant cluster; final ン + ク | 15 |
Names with consonant clusters and double consonants
| Name | Katakana | Mechanic in play | Source |
|---|---|---|---|
| Chris | クリス | onset /kr/ → ク + リ; default ウ-row close on final S | 1511 |
| Brian | ブライアン | onset /br/ → ブ + ラ; final /n/ as moraic ン | 1511 |
| Jennifer | ジェニファー | yōon ジェ; coda /-fer/ → ファー (F-row digraph + ー) | 2111 |
| Patrick | パトリック | medial /tr/ → ト + リ; final /k/ → ッ + ク (sokuon + ク) | 151311 |
| Jack | ジャック | short /æ/ + /k/ → ッ + ク (sokuon doubling) | 151311 |
Long-vowel names
| Name | Katakana | Mechanic in play | Source |
|---|---|---|---|
| Lee | リー | l→r; long /iː/ → ー | 2111 |
| Mary | メアリー | long final /-iː/ → ー | 2111 |
| Steve | スティーブ | onset /st/ → ス + ティ; long /iː/ → ー; v→b | 2111 |
| George | ジョージ | onset /dʒ/ → ジョ; long /ɔːr/ → ー | 2111 |
| Joan | ジョーン | /dʒoʊ/ → ジョー; final /n/ as moraic ン | 2111 |
Good to know
Don't romanise, then katakana-ize
The most common beginner mistake is converting the romaji spelling letter by letter (Charles → チャルレス) instead of listening to the spoken name. The sound-driven answer is チャールズ: long /tʃɑːr/ + /lz/ → チャー + ルズ.
Romaji is a writing convention for Japanese sounds. It is not an intermediary between English spelling and katakana. The 1991 Notification frames katakana loanword spelling as 「外来語」表記, the rendering of foreign sounds. NINJAL's Q&A reinforces that personal-name transcription targets the source phonology, not the source spelling.241
One name, two licensed spellings is normal
The 1991 Notification explicitly authorises both ハ-行 (Table 1) and ヴ-row (Table 2) for /v/. It also authorises both ウ-row (Table 1 fallback) and ウィ・ウェ・ウォ (Table 2) for /w/, and both エー and エイ for English /eɪ/.23 David → デビッド and デヴィッド coexist. Mary → メアリー is standard, but メアリ also appears. Jane → ジェーン and ジェイン coexist.
The practical rule is consistency once you have chosen a spelling. Japanese institutions match the spelling registered on the residence card or first official document, not the form that is linguistically closest to the source.67
Center dot ・, not ASCII period or hyphen
The katakana middle dot is U+30FB, a dedicated punctuation mark. It has a fixed role in foreign names (ジョン・スミス) and multi-word transliterations (パーソナル・コンピューター).18 The ASCII period (.), the ASCII hyphen-minus (-), the European interpunct (·, U+00B7), and the chōonpu ー (U+30FC) are distinct characters. Substituting any of them produces broken text in Japanese typography.18
The 1991 Notification governs this entire problem
外来語の表記 (内閣告示第2号, 28 June 1991) is the prescriptive reference for writing 外来語, including foreign place names and foreign personal names.23 Its 前書き defines it as the よりどころ for writing contemporary Japanese loanwords in general social life.24
Its 第1表 fixes the conventional digraphs (シェ, ジェ, チェ, ティ, ディ, ファ, フィ, フェ, フォ, ツァ, ツェ, ツォ, デュ). Its 第2表 fixes the closer-to-source digraphs (ヴ-row, ウィ・ウェ・ウォ, クァ-row, トゥ・ドゥ, ツィ, テュ・フュ・ヴュ, イェ). Its 留意事項 covers 撥音 ン, 促音 ッ, 長音 ー, and the エイ/オウ exception.23 Personal names use the same machinery as common-noun loanwords. The appendix marks personal-name entries with 人 alongside common-noun entries to make the parity explicit.3
Japan's name-order policy for Japanese names switched in 2020
The 関係府省庁連絡会議 decided on 25 October 2019 that Japanese names in 公用文 (official documents) would be written 姓→名 in the Roman alphabet. The change took effect on 1 January 2020.141516 This reverses the older 名→姓 convention adopted for Roman-alphabet contexts in the Meiji period.
The 2019 directive applies to Japanese names. The convention for entering foreign names on Japanese forms is also 姓→名, but for the unrelated reason that all Japanese forms list 姓 first. It is governed by form design, not by this directive.1767
See also
- Why "Tokyo" Is Two Syllables in English and Four Morae in Japanese: Loanwords as a Timing Drill
- Long vs. Short Vowels in Japanese: The Distinction Beginners Miss
- Wago, Kango, Gairaigo, Konshugo: The Four Vocabulary Strata of Japanese
- When Native Japanese Words Are Written in Katakana: Emphasis, Onomatopoeia, Scientific Names, and Other Stylistic Uses
- Common Romaji Mistakes That Mislead Pronunciation
- Lookalike Katakana: How to Tell the Most-Confused Kana Apart