There is a moment every learner of Japanese goes through. You see the word 東京 written on a sign, and it is beautiful, and you have no idea what it sounds like. The characters carry meaning and sound at the same time, but only if you already know them. Before you have studied kana and kanji, they are elegant drawings that tell you nothing about pronunciation.

Romanization serves as the bridge. It writes the sounds of Japanese in the Latin alphabet, the same twenty-six letters you are reading right now, so that someone who has never studied the language can sound out a word and get close to how a Japanese speaker would say it. And the system that does this job, the one your textbook uses, the one on Japanese road signs and train stations and passports, the one the entire internet defaults to, is Hepburn romanization.

The system is not new. An American missionary named James Curtis Hepburn popularized it in 1867, in the first Japanese-English dictionary, though the system itself was developed by the Rōmaji-kai (Romanization Club), a society of Japanese and foreign scholars of which Hepburn was a member. What is new is that in December 2025, the Japanese government made Hepburn the official national romanization standard, the first revision of the national rules in about seventy years. The system the world was already using is now the system Japan officially recommends.

What Hepburn romanization actually is #

Hepburn romanization is a system for transcribing Japanese sounds into the Latin alphabet, designed so that an English speaker reading the result will pronounce the word close to how a Japanese speaker would. That is the entire design principle, and every property of the system flows from it.

The principle comes with a trade-off. Hepburn is phonetic for English speakers, not systematic for kana. The syllable し is written "shi" and not "si," because "shi" is closer to how し sounds to an English reader. This is the core decision that defines Hepburn and drives every difference from its competitors. A system that maps kana to romanization in a regular, one-to-one way writes し as "si," which is cleaner as a rule but misleading as a pronunciation guide. Hepburn chose pronunciation over consistency.

There are three variants worth knowing. Traditional Hepburn is the original system from the 1867 dictionary, which romanized some sounds in ways that have since been refined. Modified Hepburn, also called Revised Hepburn, is the modern standard: it regularized a few edge cases and is the version used in textbooks, passports, and most dictionaries. The third variation concerns macron usage, particularly whether the long "i" gets a macron (sensei vs sensē), which is a matter of style rather than a substantive difference.

The history: a dictionary, a missionary, and a society #

James Curtis Hepburn was an American Presbyterian missionary and physician who arrived in Japan in 1859, one of the first Westerners to settle in Yokohama after the country opened to foreign trade. He was a doctor by training, but his lasting contribution was lexicographic. Over eight years, with the help of Japanese assistants, he compiled the first Japanese-English dictionary, published in 1867. The romanization system in that dictionary was a practical tool: it let English speakers look up Japanese words they could not yet read, and it let them pronounce what they found.

The dictionary went through multiple editions. The third edition, published in 1886, adopted the system refined by the Rōmaji-kai, the Romanization Society, a group of Japanese and foreign scholars who had been working on standardizing romanization. This is the version that became the standard and is essentially what we now call Modified Hepburn.

Hepburn was not alone. His chief assistant was Kishida Ginkō, a Japanese scholar who contributed substantially to the dictionary and the romanization system. The collaboration is worth noting because it means Hepburn romanization, despite bearing an American name, was a joint Japanese-American project from the start.

The system spread through dictionaries, textbooks, and eventually into official use. By the mid-twentieth century it was the romanization on passports, road signs, and train station names. But it was never the official system of the Japanese government. That role belonged to Kunrei-shiki, a more systematic romanization made the national standard in 1937 and reaffirmed in 1954. For roughly seventy years, Japan had one official system and one practical system, and they disagreed. In December 2025, that changed.

How the system works #

The basic syllables #

Japanese has five vowel sounds: a, i, u, e, o. They are the same five vowels as Italian and Spanish, with no ambiguity: each letter maps to one sound, always. The consonant-vowel syllables build on them: ka, ki, ku, ke, ko; sa, shi, su, se, so; ta, chi, tsu, te, to; and so on through the full gojūon table of forty-six basic syllables.

The interesting part is where Hepburn diverges from a systematic mapping. Five syllables break the pattern:

• し is "shi," not "si"
• ち is "chi," not "ti"
• つ is "tsu," not "tu"
• ふ is "fu," not "hu"
• じ is "ji," not "zi"

These are the forms that make Hepburn pronounceable to an English reader. "Shi" sounds like し. "Si" would sound like the English word "see," which is wrong. The trade-off is that these forms break the one-to-one correspondence between kana and romanization, which is exactly what competing systems like Kunrei-shiki fix.

Long vowels and macrons #

Japanese distinguishes between short and long vowels, and the distinction changes meaning. In Hepburn, long vowels are marked with a macron: a line over the letter. ō, ū, ā, ē, ī.

• Tōkyō (東京), the long "o" in both syllables
• Ōsaka (大阪), the long "o" at the start
• kyū (九, "nine"), the long "u"
• obāsan (おばあさん, "grandmother") vs obasan (おばさん, "aunt"), the macron is the only difference, and it changes who you are talking about

Macrons matter because they carry information. The common simplification is to drop them in casual contexts: Tokyo, Osaka, Kyoto. This is acceptable for proper nouns that are already known to the reader, but it strips pronunciation information for learners. A learner reading "Tokyo" and sounding it out in English will say something closer to "toe-key-oh," which is not how any Japanese speaker says it.

Passports and long vowels #

Japanese passports do not use macrons. A long "o" is written as "oh," "oo," or "ou" depending on the kanji, and a long "u" becomes "uu." Someone named 大野 gets "Ohno" on their passport, and 久保 becomes "Kubo" without the long vowel marked at all. This is a practical compromise: macrons are not available on every form and do not appear in every database, so the passport system uses letter combinations that any system can handle. It is not Hepburn in the strict sense, but it is the closest approximation the passport format allows. The result is that the same person's name can look different on their passport, in a textbook, and on a road sign, which is one of the frictions the 2025 standardization is meant to reduce.

The syllabic n and the apostrophe #

The syllable ん, written as "n" in Hepburn, is a nasal sound that can appear at the end of a word or before a consonant. Most of the time it is straightforward: san (さん), shinbun (新聞, "newspaper"), nihon (日本, "Japan").

The apostrophe enters when ん comes before a vowel or a "y" sound. Without it, the romanization is ambiguous. kin'en (禁煙, "no smoking") and kinen (記念, "commemoration") are different words. The apostrophe separates the "n" from the following vowel so you know it is a consonant, not part of a syllable like "ne." The same rule applies to names: Jun'ichirō vs Junichirō, where the apostrophe tells you the "n" is not part of a "ni" syllable.

This is the rule that most casual romanization ignores. It is also the one that causes the most ambiguity, because without the apostrophe, two different Japanese words can romanize to the same string.

The particles は, を, へ #

Japanese has three grammatical particles that are pronounced differently from how their kana would suggest. Hepburn writes what you hear:

• は is "wa" when used as a particle, not "ha" (its reading as a regular syllable)
• を is "o" when used as a particle, not "wo"
• へ is "e" when used as a particle, not "he"

This reflects a historical shift. The kana spelling preserves an older pronunciation that no longer exists, but Hepburn follows modern pronunciation. A sentence like わたしは is written "watashi wa," because that is what it sounds like.

Doubled consonants #

The small っ, called sokuon, indicates a geminate consonant: a consonant held twice as long. In Hepburn, this is written by doubling the following consonant.

• kitte (切手, "stamp"), the "tt" is held
• matte (待って, "wait"), the "tt" is held
• kippu (切符, "ticket"), the "pp" is held

The doubled consonant is a real phonetic feature, not just a spelling convention. If you do not hold the consonant, you get a different word: kite (着て, "wear") vs kitte (切手, "stamp").

Hepburn vs. Kunrei-shiki vs. Nihon-shiki #

Hepburn is not the only romanization system for Japanese. Two others matter: Nihon-shiki, designed in 1885, and Kunrei-shiki, a simplified version of Nihon-shiki that became the official Japanese standard in 1937 and was reaffirmed in 1954.

The difference comes down to a single design decision and its consequences. Hepburn optimizes for pronunciation by English speakers. Nihon-shiki and Kunrei-shiki optimize for a systematic, one-to-one mapping between kana and romanization. One kana, one romanization, always, with no exceptions.

The table clearly shows the trade-off. Hepburn gives you "shi," "chi," "tsu," "fu," and "ji"; these are forms an English reader can pronounce correctly on the first try. Kunrei-shiki gives you "si," "ti," "tu," "hu," and "zi"; these are forms that are regular and predictable from the kana, but that an English reader will mispronounce.

Why was Kunrei-shiki the official system for so long? It was because it is more systematic, reflects Japanese phonology more accurately, and is easier for Japanese people to learn as a set of rules. It maps the kana table onto the alphabet in a consistent grid. Why did Hepburn win in practice? It was because it is what the rest of the world uses. Textbooks published outside Japan use Hepburn. Passports use Hepburn. Road signs use Hepburn. Train stations use Hepburn. The internet uses Hepburn. The official system and the practical system disagreed for seventy years, and in December 2025, the Japanese government resolved the disagreement by making Hepburn official.

What practitioners say #

The trade-off between the two systems is not abstract. Linguists and teachers who have worked in Japan describe the consequences directly.

Neil Kubler, a linguist who has written textbooks on both Chinese and Japanese, points out that Kunrei-shiki has a real pedagogical advantage: "Kunreisiki is better for representing the Japanese sound system from the point of view of native speakers. Kunreisiki is also better for serious Japanese learners as it greatly simplifies the description of Japanese grammar: cf. matu/matte/matimasu (all with the same verb stem mat-) vs. Hepburn matsu/matte/machimasu (with 3 different verb stems)."

Joseph Tomei, who has taught in Japan for over thirty years, describes the real-world fallout: "The fight between Hepburn and Kunrei was so bitter and inconclusive that the government, to try and split the baby, had Hepburn for 'outward facing' uses (i.e. when foreigners might use/read it) while Kunrei was for 'inward use'. So transport signage was in Hepburn, but some government documents and other uses was in Kunrei. This was a particular problem when Japanese started getting credit cards, because Shoji Chiba on a passport would be Syozi Tiba and be rejected."

The 2025 standardization is, in part, a response to exactly this kind of confusion.

Why Japan Made Hepburn Official in 2025 #

On December 22, 2025, the Japanese government issued a cabinet notification establishing Hepburn romanization as the national standard. It was the first revision of the national romanization rules in approximately seventy years.

The reasoning was straightforward: Hepburn was already the de facto standard in passports, road signs, train stations, and international communication. The official system, Kunrei-shiki, and the practical system, Hepburn, had diverged, creating inconsistency between what the rules said and what everyone actually did. Government documents used one system; the signs on the street used another. Making Hepburn official aligned the rules with reality.

What this means for learners is simpler than what came before: the system your textbook uses, the system your dictionary uses, and the system on Japanese street signs are now the same system. You no longer need to translate between official and practical romanization. One system, one set of rules, everywhere.

How Fink uses Hepburn #

Fink puts a pronunciation line in the Insights panel under every Japanese translation, and the system it uses is Hepburn romanization with macrons for long vowels. If you translate a sentence into Japanese and open the Insights panel, the pronunciation line shows you how to say it, written in letters you already know.

Why Hepburn and not IPA? Because a learner who cannot yet read kana does not need a third unfamiliar system between themselves and the text. The International Phonetic Alphabet is a powerful tool, but it requires learning a new set of symbols before you can read a single word. Hepburn uses the letters you already know and gives you sounds to work with immediately. You can read Tōkyō and get close to the right pronunciation without learning anything first. That is the whole point of the system, and it is the reason Fink chose it. For more on when IPA is the right tool and when romanization is better, see Understanding IPA.

The exact scheme is named explicitly in the prompt: "Hepburn romanization, with macrons for long vowels." Never "the pronunciation." A model asked vaguely will invent a different convention every time, and a pronunciation system that changes between requests is worse than none, because you cannot build a habit on it. Naming the system is what makes the pronunciation line consistent, and consistency is what makes it useful.

How to use romanization when learning Japanese #

Romanization is a scaffold, not a destination. Here is how to use it well:

1. Read the romanization before checking the definition. When you look up a word, sound it out from the Hepburn transcription first, then check the meaning. This builds the connection between sound and word.
2. Say it aloud. Pronunciation is a physical habit. Reading a sound silently does not build it; saying it does.
3. Use romanization as a bridge to kana, not as a replacement. The goal is to outgrow it. Once you can read hiragana and katakana, the romanization becomes a check rather than a crutch, and most learners stop needing it after a few months.
4. Notice the recurring patterns. After a while, you start to see that し is always "shi," that long vowels always get macrons, that the particle は is always "wa." The rules become automatic, and the romanization fades into the background.

This is the same loop described in How to use a translator to learn a language: look up a word, read the pronunciation, say it, notice the pattern. Romanization is what makes that loop work for Japanese, because it is the step that connects the written form to the sound without requiring you to know the writing system yet.

The short answer #

Hepburn romanization is the most widely used system for writing Japanese in the Latin alphabet, popularized in 1867 by an American missionary so that English speakers could pronounce Japanese words correctly, though developed by the Rōmaji-kai, a society of Japanese and foreign scholars. It prioritizes pronunciation over systematic kana-to-romaji mapping, which is why it writes し as "shi" and ち as "chi" even though those break the regular pattern. Japan made it the national standard in December 2025, aligning official rules with what the world was already doing. Fink includes Hepburn romanization in its insights for every Japanese translation, because someone learning Japanese needs sounds they can read, not a third writing system to learn first.

Questions people ask about Hepburn romanization #

Is Hepburn romanization the same as romaji? Romaji means "Roman letters" and refers to any romanization system for Japanese. Hepburn is one specific system, the most widely used one. Kunrei-shiki and Nihon-shiki are also romaji. So "romaji" is the category, and "Hepburn" is one member of it.

Why does Hepburn use "shi" instead of "si"? Because "shi" is closer to how し sounds to an English speaker. "Si" would be pronounced like the English word "see," which is wrong. Hepburn prioritizes pronunciation over systematic kana mapping, and this is the defining example of that choice.

What's the difference between Hepburn and Kunrei-shiki? Hepburn is phonetic for English speakers: し is "shi," ち is "chi," つ is "tsu," ふ is "fu." Kunrei-shiki is systematic: し is "si," ち is "ti," つ is "tu," ふ is "hu." Kunrei-shiki was the official system in Japan from 1954 until December 2025, when Hepburn replaced it as the national standard.

Do I need to learn romanization to learn Japanese? No, but it helps in the early stages. Use it as a bridge to kana, not a replacement for it. Most learners stop needing romanization after a few months, once they can read hiragana and katakana reliably.

Why does Fink use Hepburn instead of IPA for Japanese? Because someone who cannot yet read kana does not need a third unfamiliar system. Hepburn uses letters they already know and gives them sounds to work with immediately. IPA would add a layer of symbols to learn before the actual language. For languages written in the Latin alphabet, Fink does use IPA, the choice depends on the script.

What is Modified Hepburn? Modified Hepburn, also called Revised Hepburn, is the modern standard version of the system. It regularized a few edge cases from the original 1867 dictionary: most notably, it uses "n" before consonants instead of "m" (so しんぶん is "shinbun," not "shimbun"), and it uses macrons for all long vowels consistently. Traditional Hepburn used "m" before b, m, and p, so 新聞 was "shimbun." Modified Hepburn fixes this to "shinbun." When people say "Hepburn" today, they almost always mean Modified Hepburn.

Where is Hepburn romanization used? Road signs, train station names, passports, textbooks published outside Japan, most dictionaries, and the internet at large. Since December 2025, it is also the official national romanization standard of Japan. If you have seen Japanese written in Latin letters, you have almost certainly seen Hepburn.