Japanese Names in English: Read Them and Write Your Own

Names are where romanization stops being an abstract system and starts being about a specific person, which is exactly why they are the place it goes wrong most often. A name has one correct spelling in English letters — the one its owner uses — and no rule you can apply will reliably produce it.

That cuts both ways. Reading a Japanese name out of kanji is genuinely hard, because the characters do not contain the reading. Writing your own name in Japanese is much easier, because the rules are mechanical and there is a standard answer for almost every English name. This article covers both directions, and the sounds that cause all the trouble.

The family name comes first

Japanese names are written 姓 (sei, family name) then 名 (mei, given name) — the opposite of English order. 田中 太郎 is Tanaka Tarou: family name Tanaka, given name Tarou.

For a long time, English-language publications reversed this to fit English convention, printing "Tarou Tanaka". In recent years several major English-language outlets have switched back to Japanese order without a comma, so both "Tanaka Tarou" and "Tarou Tanaka" appear in the wild today.

The practical consequence is that when you meet a Japanese name in English, you may not be able to tell which part is the surname. The reliable tells are context — how the person is referred to later in the piece, or how they are listed in a database — rather than the shape of the words. If you are writing about someone and it matters, check their own preferred order rather than guessing.

Why most Japanese names romanize cleanly

Here is the encouraging half. Once you know how a name is read, converting it to romaji is entirely mechanical — kana map to Latin letters one for one, with no exceptions and no judgement calls.

田中 is read たなか: ta, na, ka → Tanaka. 鈴木 is すずき → Suzuki. 高橋 is たかはし → Takahashi. You are not translating anything; you are transliterating, and transliteration of kana is a lookup table.

The hard part is upstream of that, and it is worth understanding precisely, because it explains why no tool can be fully trusted with names.

The reading is not in the characters

This is the single fact that makes Japanese names hard, and it has nothing to do with romanization.

Most kanji have several possible readings, and a name can use a reading that appears nowhere in ordinary vocabulary. 東 can be higashi, tou, or azuma. 一 can be ichi, kazu, or hajime. Nothing about the character tells you which one applies — the reading is a property of the name, chosen by the family, not derivable from the writing.

So a name like 東海林 might be read Shouji or Toukairin depending on the family. Both are correct. A dictionary-based tool given only the kanji has no way to prefer one, which is why name readings have to be stored rather than computed, and why any converter — this site's included — will occasionally get a name wrong. It is not a bug you can fix with a better algorithm; the information simply is not present.

A second trap sits on top of it: long vowels. 佐藤 is さとう, which romanizes to Satou — and you will also see Satō, Sato and Satoh, depending on which romanization convention the person or the publication follows. 伊藤 is Itou, 加藤 is Katou, 大野 is Oono. The vowel is genuinely long; the disagreement is only about how to spell length in Latin letters.

Writing your name in Japanese

The other direction is straightforward, and it runs on exactly the same machinery that turns "computer" into コンピューター. Your name gets pushed through the Japanese sound system and comes out in katakana.

Four rules cover nearly every English name:

L becomes an r. Japanese has no l, so both English liquids land on the same row. Laura becomes ローラ (Roora), Lisa becomes リサ (Risa), Emily becomes エミリー (Emirii).

A final consonant gains a vowel. No Japanese syllable ends in a consonant except ん (n), so a name ending in a hard consonant gets a vowel appended to carry it. It is quiet — say it and drop it. Mark becomes マーク (Maaku), Tom becomes トム (Tomu), Smith becomes スミス (Sumisu).

V is unstable. Japanese has a kana for v — ヴ — but it is a late addition and many words still use the b row instead. Victoria can be ヴィクトリア or ビクトリア, and both are accepted.

Long vowels take a ー. A long vowel in your name is written with the dash: マーク, エミリー, ローラ.

Written in Japanese order, with the family name first, "Emily Smith" becomes スミス・エミリー — the interpunct ・ separates the two names, the same way it separates words in a foreign phrase.

One caution on choosing a spelling: there is often more than one accepted rendering of a name, and people have preferences. デビッド and デイヴィッド are both "David". For a personal name — a business card, a residence document — use the spelling the person has already chosen for themselves rather than the one a tool produces.

Reading a Japanese name you have never seen

The order that wastes the least time:

1. Look for a romanized or kana version first. Search results, social profiles and business cards usually carry one. It is the authoritative answer and it costs nothing to check.
2. Failing that, romanize what you can. If the name has kana alongside the kanji, the kana give you the reading directly and the rest is mechanical.
3. Treat a computed reading as a guess. In a dictionary-based converter a name it does not recognise will be marked rather than invented — an honest blank is more useful than a confident wrong answer.
4. Remember that some names are simply ambiguous. 東海林 read either way is not a mistake by anyone. If a reading matters, ask.

For a name a tool does not know, the romaji converter will still romanize whatever kana it can find and flag the rest, which is usually enough to reconstruct the reading by hand. What is romaji explains where the letters come from, and the interactive kana chart covers the characters your name will be built out of.