The Three Main Systems of Japanese Romanization

Romanization means transcribing a language into Roman letters -- as in the English alphabet. The Japanese word for romanization is ローマ字 (roomaji - Roman letters). There are three main systems (shiki) of ローマ字. Nihongo Web uses the more phonetic Hepburn system, called Hebon shiki in Japanese.

Hebon-shiki (Hepburn system ヘボン式) was devised by an American missionary doctor in the 1860s to transcribe the sounds of the Japanese language into the Roman alphabet.  It is widely used today both in the English-speaking world and in Japan.  It is also widely used in English library catalogs. 
 

Kunrei-shiki (訓令式; also known as ISO3602, or  as 文部省 Monbusho system), is a compromise between Hepburn and Nippon-shiki roomaji. The Hepburn system stresses conformity with English phonology, and Nippon-shiki stresses one-on-one correspondence with the kana system.

Kunrei-shiki is also sanctioned by the Japanese Ministry of Education (monbusho 文部省) and is mostly used within Japanese schools.  Kunrei-shiki is also known as ISO 3602, as it was approved by the International Standard Organization as a common roomaji method for the language.

Nippon-shiki (日本式, or Japan-style) conforms strictly to the kana writing system, and is easier for Japanese natives to interpret than the more anglicized Hepburn system. If you want your Japanese word processor to generate づ instead of ず, you can use this system and type "du."

Many Japanese word processing applications conform either to Nippon-shiki or Kunrei-shiki system. Windows IMEs can use all three systems.

Nihongo Web uses very little romanization, but when it does, words will be rendered just as they are written in kana. Below are a few extra pointers in order to avoid common romanization mistakes.  If you are used to using Japanese word てんぷらprocessing programs (i.e. more familiar with Kunrei and Nihon-shiki than Hepburn), these may be especially useful:

          For the syllabic nasal, n is always used preceding b, m, and p.

o      新聞 (しんぶん)  > shinbun not  shimbun

o       天ぷら (てんぷら) > tenpura  not  tempura

                     Double consonants are represented by repeating the following consonant.

o       結婚 (けっこん)  > kekkon

o       発達 (はったつ)  > hattatsu

o       EXCEPTION TO THIS RULE: When c is the following consonant, insert t instead of repeating c.

o       坊ちゃん (ぼっちゃん)  > botchan not  bocchan

o       抹茶　(まっちゃ)  > matcha not  maccha HOWEVER Windows renders this word only by typing maccha

        Letters are represented truthfully to the pronunciation. For example, a particle 「は」is represented as wa, not ha;「を」as o, not wo; 「へ」as e, not he. Also,「こんにちは」is represented as konnichiwa, not konnichiha.

      Elongated hiragana vowels in Nihongo Web correspond to the writing system. For example, the word for teacher or doctor (先生 > せんせい) is represented as sensei, not sensee as in some systems devised to reflect pronunciation. Also, some systems render elongated o as oh. Nihongo Web stays true to the writing system by using either ou (more common) as in とうきょう toukyou and きょうと kyouto or oo as in おおきい ookii and とおい tooi.  Some systems use a dash over the vowel, but that does not help students know if the elongated o is to be written as ou or oo.

For more on the big mess called roomaji (romaji/ro-maji/romazi/...) see Wikipedia's article on Romanization Systems

Back to the Hiragana Tutorial