Romaji ローマ字

This post is about romaji: what it is and isn’t. More about what it isn’t than what it is.

Romaji ローマ字 (‘Rome letter’) is: a way of writing Japanese using Roman letters 

It is not:

  • -standardised
  • -English
  • -a complete way to read and write Japanese

And here I’ll go into some more detail.

Romaji is a way to read and write Japanese using Roman letters

It’s useful for words like Asahi, wagyu and Tokyo – Japanese things that people around the world want to talk about and use those words more or less as they are, with an approximation of Japanese sounds, rather than translating them. ‘Morning sun,’ ‘Japanese cow’ and ‘Eastern metropolis’ don’t have quite the same ring, I guess.

Romaji is not standardised

There are some romaji spellings that are widely accepted and known in English-speaking countries, such as Osaka, Kyoto, Tokyo, sushi, cha, sayonara, Mt Fuji, etc.

Here are a few spellings you might see for those words in various schools in Japan.

Oosaka, Ôsaka, Ōsaka, Ohsaka

Kyouto, Kyôto, Kyohto

Toukyou, Tôkyô, Tōkyō, Tohkyoh

sushi, susi

cha, tya

sayonara – sayounara, sayônara, sayōnara, sayohnara

Mt Fuji, Mt Huji

And while we’re at it, romaji could also be roomaji, rohmaji, rômaji or rōmaji. And tempura could be, in alternate-Romaji-land, tenpura. 

So which spellings are correct? Well, that’s the problem – it depends who you ask. I believe the romaji taught in most (??) middle and high schools is Hepburn. Which version of it also depends on the school, or maybe the teachers in the school, or maybe the Board of Education.

The Hepburn Romanisation system, a.k.a. ヘボン式ローマ字 hebonshiki romaji. Image: this romaji middle school page. Found via Google.

Even if teachers dedicate themselves faithfully to one particular version of Hepburn, it’s hard to stop the kids referring to the Kunrei – or is it Nihon? – charts of romaji in the back of their 108-yen notebooks. And whatever their homeroom teacher taught them 3 years ago in primary school, which was almost certainly not Hepburn, though they might have learned it at extra-curricular English lessons.

In romaji, tuki and tsuki sound the same.

My parents went to Shikoku last year and then brought back some omiyage custard cakes called hime no tuki. We talked about how it probably means ‘princess of the moon.’ Dad noticed that when read out loud,  the Japanese speakers in the room seemed to add an S after the T in tuki. Having painstakingly learned all the syllables in Japanese, Dad had learned tsu but not tu – what’s tu? Just another way of writing tsu. Because everyone knows that if you combine T and U an S sound will have to come in between, because that’s how Japanese works. Unless you don’t know Japanese. Which means you don’t know other things too, like that di, ji and zi all sound the same in Japanese and so do zu and du.

Actually, there is a good reason to keep du and di instead of just using zu and ji all the time, and that is typing. Perhaps the reason students learn romaji so long before they start learning to write and read other languages is so that they can use it to type. And if you want to type ぢ, the only way to do it is di, because ji will come out as じ every time. It’s important for spelling.

A non-Hepburn romaji chart. Image: Wisdom Bag, Yahoo JP

In the middle school where I most recently was an ALT, I don’t know which version of Hepburn they were teaching, but in the previous school it must have been some version of traditional as opposed to modified, because they learned a rule that the letter N had to change to M if it came before a bilabial consonant like P, B or M. This means, for example, that shinbun would be written as shimbun. It weirded me out when they had to write sanma さんま as samma, because you would only know how to read that if you either 1) speak a language with geminates like Italian’s ‘mamma’ with the long M or 2) already have an understanding of Japanese phonology. You need to know some Japanese phonology, too, to know that the U in shinbun or shimbun doesn’t sound like the U in the English ‘bun’ but more like the ‘oo’ in ‘toot.’ If you know no Italian or Japanese, depending on your variety of English you might read samma さんま to rhyme with stammer or Gramma (as in Grandma). This brings me to my next point.

Romaji is not English

If you are reading this post in English, then this might seem pretty obvious. To a Japanese kid (or adult) who learns romaji in about grade 3, way before they officially learn the alphabet in grade 6 or year 7, it’s not so clear-cut. When they see the letters in the alphabet it’s natural to try to read them as they were first taught, as in romaji. If you learned at 8 years old that the thing you know as チョーク is spelled tyôku, as opposed to chôku or another spelling, it’s going to be much easier to recognise and remember that than chalk. Even if you do copy out chalk 5 or 10 times in your notebook, it’s not going to stick anywhere near half as much as tyôku did. Even though it might not sound all that different, one is going to be much more comfortable, familiar and immediate than the other.

I believe this contributes to obstacles in language-learning. Much like katakana, I have mixed feelings/opinions about romaji – more about its teaching and usage than its existence. It can be a great tool for Japanese language users and non-Japanese language users to communicate. So in theory, all that should be necessary for it to trap fewer people should be something like, English teachers saying at the start of Year 7 English, hey kids, welcome to English. Let’s practice the alphabet and learn some PHONICS (hello, last year’s junior high school! To be fair, some schools do, but it’s not standardised either) and before we can properly start that, let’s take a moment to recognise that romaji and English are not the same thing, despite having a similar physical appearance. Their biggest difference is the fact that romaji is Japanese, and English is not.

Romaji is not a complete way to read and write Japanese

You may have noticed the word ‘probably’ used to talk about the meaning of the name of the custard cake package shown above. Well, hime usually means ‘princess’ (姫、ひめ), but it’s also the old name for ‘hawfinch’ (鴲 ,ひめ), which is a kind of bird. Who knows, though, it could also be the verb stem for himeru (秘める) ‘to hide.’ Then there’s the possibility that the E was supposed to be a long one, as in himei, which can mean ‘an inscription’ or ‘a scream.’ All this information is stuff that native speakers know from context or from reading the kanji, much as native speakers of English know the difference between ‘know’ and ‘no.’ But if you take away spelling and context, there’s no way to know.

When I was about high school age, I remember watching Inuyasha with my siblings. We were constantly perplexed by the name of Inuyasha’s sword, Tessaiga (鉄砕牙、てっさいが). The person doing the subtitles had transcribed the little tsu (っ) into the romaji in the subtitles* so we were reading ‘Tetsusaiga’ while hearing TESSAIGAAAAA!!! and trying to hear the extra tsu. Drove me nuts for years. Eventually I looked it up on the Internet – might have been doing some research for this blog, actually! – and discovered that it hadn’t been only us who had been troubled by this.

Even foreigners who speak Japanese fluently – even Japanese people, actually – can look kind of dumb sometimes if a Japanese name is written only in romajiIt can mean the difference between your name meaning Big-field or Little-field.

大野さん Ôno-san/Ohno-san/Oono-san ‘Ms./Mr. Big-field’

小野さん Ono-san ‘Ms./Mr. Little-field.’

In the above examples, I’ve used 3 different ways of showing that the first O is long in the Big-field name. But because it’s not standardised, students of English may learn any one of these spellings or may be taught not to indicate vowel length at all, because it’s ‘not what people are doing these days.’ The length of vowel sounds doesn’t usually affect much in English other than to give you an accent. It can affect meaning in minimal pairs like ‘ship’ and ‘sheep’ but it’s not only the length that distinguishes these words. In Japanese, though, as you see above, it can change the meaning of words. It’s the difference between snow (雪, ゆき, yuki) and courage (勇気, ゆうき, yuuki). Yuki can also be a boy’s name or a girl’s name, depending on the length of the U.

These aren’t necessarily world-changing issues, but they can create what are surely at least partially avoidable obstacles to effective communication. It has probably come up in studies. Naturally, I have the power to whinge about it, but no solution. That being said, I think standardisation would be a start, and within that, I don’t think it should be left up to the will of the writer (as it is on this page) whether they want to indicate the length of a vowel or not. That’s information that can help people understand more about your language, Japan. And if what people see and hear is consistent, it will make more sense.

*Romaji in the English subtitles: By this I mean that they’d left the sword name in Japanese, writing it in romaji rather than translating the meaning. Someone must have decided ‘Tetsusaiga’ was easier to read repeatedly than ‘Iron-Crushing-Fang.’


3 thoughts on “Romaji ローマ字

  1. Haha!! I wonder if it depends on the computer. Mine is of the Mac variety and if I just hold down the key of the vowel I want to use, it comes up with a list of options. That’s what I used for this post with no apparent issues. On old Windows computers there were accents you could do with Alt and the number pad…
    Having Googled it, it looks playing with the settings might be necessary and adding Māori or another language with macrons might work?


  2. Pingback: Dassai (not dasai) ダサイじゃなくて獺祭 | bijinjapan

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s