日本語表記

Generally, a good transcription system for any language must satisfy two criteria:

1. Allow a lossless mapping of any word from and back to the original. I.e. you must be able to transcribe any word back to the original alphabet, given only its romanized version. 2. Come as close as possible to capturing the proper pronunciation. I.e. allow a layman reader reading the transcribed version to pronounce the words close to their correct pronunciation.

There is no Japanese transcription system satisfying both these criteria.

The most common way of transcribing Japanese into English suffers from a key problem in that it ignores long vowels, transcribing them as short. Therefore Sumou is transcribed shortly as Sumo. This is a serious problem enough, since the distinction between long and short vowels is a key one in Japanese, that I think the advantage of using the common transcription of Japanese is far outweighed by the disadvantage of it being very bad.

Generally Sumou is transcribed Sumo, ignoring the "u" at the end, making the "o" a long vowel. It is not a problem in a common word like Sumou but becomes a problem in general transcription.

Furthermore, as a sidenote, "Sumo" has through it's usage become a native English word, in the sense that it differs from Sumou, it's original. It is used in English language in a differing context and a differing meaning. Like other words that have become English language property - Tsunami, Karaoke, Judo... Therefore it seems to me that using Sumou exclusively, relating to what I consider true Sumou, makes even more sense.

Other methods of marking the long vowels generally use macrons or other diacritical signs above the vowel in question. This is also a problem since they're not readily usable on most computers...

There is one "transcription" method available readily in Japanese, that is the use of it's hiragana alphabet to transcribe all words. Within this transcription Sumou is transcribed as su-mo-u and so, marking the exact pronunciation.

Therefore I use the following transcription method which satisfies the above criteria and is very legible and is a direct translation of the Japanese "transcription" thus making every word directly retranscribable back to Japanese.

<table border="1"> <tr><td>あ
a</td><td>い
i</td><td>う
u</td><td>え
e</td><td>お
o</td><td>ヴァ
va</td><td>ヴィ
vi</td><td>ヴ
vu</td><td>ヴェ
ve</td><td>ヴォ
vo</td></tr> <tr><td>か
ka</td><td>き
ki</td><td>く
ku</td><td>け
ke</td><td>こ
ko</td><td>が
ga</td><td>ぎ
gi</td><td>ぐ
gu</td><td>げ
ge</td><td>ご
go</td></tr> <tr><td>きゃ
kya</td><td>きゅ
kyu</td><td>きょ
kyo</td><td>ぎゃ
gya</td><td>ぎゅ
gyu</td><td>ぎょ
gyo</td><td>くゎ
kwa</td><td>ぐゎ
gwa</td></tr> <tr><td>さ
sa</td><td>し
shi</td><td>す
su</td><td>せ
se</td><td>そ
so</td><td>ざ
za</td><td>じ
ji</td><td>ず
zu</td><td>ぜ
ze</td><td>ぞ
zo</td></tr> <tr><td>しゃ
sha</td><td>しゅ
shu</td><td>しょ
sho</td><td>じゃ
ja</td><td>じゅ
ju</td><td>じょ
jo</td></tr> <tr><td>た
ta</td><td>ち
chi</td><td>つ
tsu</td><td>て
te</td><td>と
to</td><td>だ
da</td><td>ぢ
dji</td><td>づ
dzu</td><td>で
de</td><td>ど
do</td></tr> <tr><td>ティ
ti</td><td>ディ
di</td><td>ちゃ
cha</td><td>ちゅ
chu</td><td>ちょ
cho</td><td>つぁ
tsa</td><td>ぢゃ
dja</td><td>ぢゅ
dju</td><td>ぢょ
djo</td><td>づぁ
dza</td></tr> <tr><td>な
na</td><td>に
ni</td><td>ぬ
nu</td><td>ね
ne</td><td>の
no</td><td>にゃ
nya</td><td>にゅ
nyu</td><td>にょ
nyo</td></tr> <tr><td>は
ha</td><td>ひ
hi</td><td>ふ
fu</td><td>へ
he</td><td>ほ
ho</td><td>ば
ba</td><td>び
bi</td><td>ぶ
bu</td><td>べ
be</td><td>ぼ
bo</td><td>ぱ
pa</td><td>ぴ
pi</td><td>ぷ
pu</td><td>ぺ
pe</td><td>ぽ
po</td></tr> <tr><td>ひゃ
hya</td><td>ひゅ
hyu</td><td>ひょ
hyo</td><td>びゃ
bya</td><td>びゅ
byu</td><td>びょ
byo</td><td>ぴゃ
pya</td><td>ぴゅ
pyu</td><td>ぴょ
pyo</td></tr> <tr><td>ま
ma</td><td>み
mi</td><td>む
mu</td><td>め
me</td><td>も
mo</td><td>みゃ
mya</td><td>みゅ
myu</td><td>みょ
myo</td></tr> <tr><td>や
ya</td><td> </td><td>ゆ
yu</td><td> </td><td>よ
yo</td></tr> <tr><td>ら
ra</td><td>り
ri</td><td>る
ru</td><td>れ
re</td><td>ろ
ro</td><td>りゃ
rya</td><td>りゅ
ryu</td><td>りょ
ryo</td></tr> <tr><td>わ
wa</td><td>ゐ
wi</td><td> </td><td>ゑ
we</td><td>を
wo</td></tr> <tr><td> </td><td> </td><td> </td><td> </td><td>ん
n*</td></tr> </table>

*Syllabic n ん is transcribed as n before consonants but as n' before vowels and y.

Geminate consonants are transcribed by doubling the consonant following the っ, with the following exceptions:

<table border="1"> <tr><td>っち
tchi</td><td>っちゃ
tcha</td><td>っちゅ
tchu</td><td>っちょ
tcho</td></tr> </table>

The only serious problem this method of transcription poses for the English speaker is the usage of double "o" at the beginnings of words where the Japanese words have them, leading to possible mispronunciation, since "oo" is generally read in English as a long "u" not "o".

Thus Oozeki and Oosaka look problematic...

However there is no other way I think to work around this with consistency, so this small drawback of the transcription is to be kept in mind.


.