The Undecipherable Poem, No. 9 of the Manyoshu


Avery Morrow, 2004


The Manyoshu, or "Compilation of Ten Thousand Leaves", is the earliest known work of poetry in Japanese history. Written in the mid-700s, it postdates the chronicles of Shinto myth called the Kojiki and Nihongi, the record of provinces called the Fudoki, and possibly the record of Shinto law called the Yengishiki (the date for this last work is undetermined). However, its order chronologically does not detract from its significance as a powerful creative work by perhaps hundreds of authors. It contains 4,516 poems, most of which were written in the 600s or 700s C.E. And out of all these poems, one of them has never been deciphered.

It is a confusing question why the Japanese would need to translate their own literature. We can compare this to our own ancient writing such as Beowulf and the Anglo-Saxon Chronicle, which are unreadable to any English speaker except in translation. These, however, have all been translated by deriving an Old English vocabulary. The Japanese have a far more momentous problem. They did not create a writing system of their own until after the Manyoshu was written; although the poems are in Japanese, they are written with Chinese characters and surrounded by ancient Chinese notes and inscriptions. At the time that the Manyoshu was written, Japan was adopting Chinese words and characters, and used Chinese characters to represent Japanese syllables; but by roughly 820 C.E., only a hundred years later, this system had been entirely forgotten, and the poems were gibberish. So, for over 1000 years, scholars have attempted to reconstruct the original Japanese from a hopelessly complicated writing system.

Most of the time, the Chinese characters summon up certain Japanese syllables, and the original, beautiful language can be recalled. Sometimes, more work is needed to determine how to interpret the characters. But only in one case out of the 4,516 are the original characters entirely incomprehensible, a veil of eighth-century logic behind which lies the poem. And strangely enough, this occurs not in the middle of the collection, but right at the start; poem number nine.

Two modern translations of the Manyoshu into English exist: Ten Thousand Leaves (Harold Wright, 1979) and The Manyoshu (Nippon Gakujutsu Shinkokai, 1965). Both explain the difficulties of translation, and that study of the Manyoshu continues, and that recently Japanese classicists have discovered the Manyoshu uses eight vowels rather than five, but neither one touches on No. 9. I would like to warn the reader that both of these translations use an original numbering system which handily omits Kokka Taikan No. 9 entirely and thus avoids the problem of its translation. The comprehensive work Japanese Court Poetry (Brower and Miner, 1961), also discusses the linguistic problems on pages 80-81 but does not mention No. 9. Based on this incomplete survey, I reached the conclusion that this paper is the first formal essay in English to bring up the problem of No. 9.

N.B., February 2006: While browsing at the library, I found an old translation of the Manyoshu that does attempt a translation of No. 9 (although rather clumsily), but I forget the name.

The Partial Translation

The latter half of No. 9 has been deciphered. I quote it here in full, using the version by the Japanese Text Initiative:

Original text: 莫囂圓隣之大相七兄爪謁氣 / 吾瀬子之 / 射立為兼 / 五可新何本
Modern kanji: 莫囂円隣之大相七兄爪謁気 / 我が背子が / い立たせりけむ / 厳橿が本
Modern reading: ????? / ?????? / わがせこが / いたたせりけむ / いつかしがもと

I know little of modern Japanese and nothing of ancient Japanese, so my attempt at understanding who or what Seko and Itsukashi are was rather futile. I was faced with a problematic task; No. 9 had been omitted by all the English translators for the same reason that I was looking into it, so there was some meaning to the ancient Japanese that I needed to find out without a proper translation. Luckily, some writings existed on the World Wide Web, and with the help of a machine translator I read a short monologue on it by a horror writer, which excerpts a translation from a modern Japanese text.

The Mansion of Ghastly Characters by Kiichirou Kurasaka
fetched from Rhino Mountain on 26 March 2005
My translation (likely inaccurate)
活字中毒、読むものがないとやってられない、という意味で弱いことは言うまでもないが、 「文字それ自体」にも非常に弱い。 It goes without saying that poison of the print will not do, if nobody reads it. The "character by itself" is exceedingly weak.
映画『セブン』の感想でも書いたが、万葉集に、上半分が未だに全く解読不能の歌がある。Although I already wrote about this in my comments on the movie se7en, there is a poem in the Manyoshu of which the upper half cannot be deciphered.
わたしはこれがこわい。To me, this is frightening.
莫囂円隣之大相七兄爪謁気 わが背子がい立たせりけむ厳橿がもとWyrg gende acbire madentag wher myne Seko once stode, at the rootes of Itsukashi.
初めてこの歌を見た瞬間、背筋に鳥肌が立った。When I first read this poem, goosebumps stood up on the back of my spine.
後半部分は「わが君がお立ちになったという神聖なかしの木のもとよ(旺文社対訳古典シリーズ桜井満訳注)」という意味。The second half of this poem means, "I am at the base of the sacred oak tree where they say you stood" (Obunsha Publishing's Bilingual Classics series; translation notes of Sakurai Mitsuru).
ということは推測するに、恋人が旅立ったことを歌った訳で、前半部分には大して意味がないか、あってもその相手への想いを詠んでいるとしか思えない、なのにわたしはこの歌がこわい。 Guessing at the meaning, if he sings because he quests for his lover, even if the first half means nothing, this poem is frightening to me.
それはもう、手がつけようがないから怖いのだ。あまりにも意味不明なのに、そこには何か、自分にはどうしても読み取れない確固たる意味がある。判らないから想像だけがふくらむ、それが怖い。 Since some hand has already written it, it is dreadful. Although too much is incomprehensible, there is a meaning to it which can never be read. Since one cannot understand it, the imagination swells. It is a thing to be feared.

Contemporary Discussion

Google is the tool of an unscrupulous reporter. However, without a better database to use, and having exhausted my library, I was forced to turn to it for information on No. 9.

I was unable to finy any discussion of No. 9 on the Web in English, but I did find a single posting on USENET's sci.lang.japan from 1995. It was written not by a Westerner but by a Japanese native to correct his own translation of old Japanese. I quote it here, with some corrections to the author's English:

Subject: maguwai and mokkori
Date: 1995/06/11
Message-ID: <3re0v3$>
newsgroups: sci.lang.japan

In my last post concerning maguwai ... I wrote:
> as far as I recall now "maguwai" implies rotation rather than ......

I need to correct this.   I checked with Manyoshu for the waka I had had in mind 
and got the following points.

Volume 1, No. 9, Manyoshu, is [a] Tanka by Nukatano Ohokimi.  
It is said that [nobody has] yet been successful to read the first 12 Manyogana
(actually kanji).  However, a Korean ... Dr. I Yonhi (since I do not know 
how to spell her name, I just spelt [it] in romaji) recently elucidated 
the unk[n]own part [using the] Old Korean language.
(Japanese was originally a dialect of Korean, but by the era this poem was written in, over 2000 years had passed since the journey from Korea.)
According to her elucidation, the first 5 Manyogana are read as:

mage dongruriji (old Korean) [, which] me[an]s "Magu wo mawaseyo" (modern Japanese).

It is quite erotic as you can imagine.  Magu means "mokkori" (the limited usage), and is I believe
from "maguwai". ["mokkori" means "a tent in one's bedsheets"]

Muneo Saito

What an anticlimactic solution to this mystery! Rather than a love poem, according to Dr. Yonhi's translation, No. 9 is simply an erotic poem, and the sacred tree Itsukashi has an obscene double meaning.

Luckily, I did not abandon my quest upon reading this. A myriad of Japanese commentaries on the Web do not agree with, or even mention, the translation of Dr. Yonhi. Here's a complete summary of the search results:

Search resultQuoteTranslation (inaccurate)
Fun with the Manyoshu!しかし、この歌の解釈は定まったものが無い様です。特に前半の読みについては色々な説が出されていて、いまだに決着がついていません。 However, a fixed translation of this poem has yet to be decided. There are various opinions about the reading of the first half, and no firm conclusion.
My Trip to Chokoku-ji Free Image Hosting at
三重のやまみつつ ゆけ吾が瀬子が い立たしけむ 厳橿が本
[Apparently carved by the owners of a hot spring]
The triple mont / I will go where my Seko / once stood / at the roots of Itsukashi
King Nukata's Manyoshu *難読歌。「莫囂圓隣之 大相七兄爪謁氣」が読めない。 A difficult poem. "莫囂圓隣之 大相七兄爪謁氣" cannot be translated.
[A long list of competing theories follow]
Seabed Village's Manyoshu 「いろいろ読めるんだね('ヘ`;マイッタ。そっから後ろのほうはどうなの?」

「そっちも諸説あるんだけど、『吾が背子がい立たせりけむ厳橿がもと』 で大方定まってるわ。意味は『かつてあなたが立っていた橿の木のも と』という感じね(。・_・。)」

「なるほどねえ。それを踏まえて、サラちゃんは上の句の訳、どれが適 当だと思う?(・ω・)ノ」

"So many different ways to read it! ('ヘ`;) I give up. How's the other end go?"

"It reads, 'wher myne Seko once stode, at the rootes of Itsukashi'. There are various theories, but it's almost settled upon: 'at the oak tree where you were once standing'.(。・_・。)"

"Okay, I get it! So the poem is about you, Sara-chan? I do think it suits you! (・ω・)ノ"

The genocide of the Jews in ancient Japan表音羅列文が作られたその過程が飛ばされおり、そして、わが背子、からが隠蔽文になっています。
The phoenetic enumeration of this sentence was concealed with the red herring, "myne Seko" etc.
In other words, these 12 characters express the Hebrew language directly in kanji. [Yeah, sure.]
However, only 4 phrases of 莫囂圓隣 are hiding proper nouns.
If we investigate the meaning of each of the kanji of 莫囂圓隣之大相七兄爪謁氣, we get the following: [ridiculous derivation follows]
The Mystery Poem of the Manyoshu ま、コーヒーでも飲みながらやってみてください。クロスワードパズルよりはおもろいでしょ。ただし、深みに嵌まると大変ですよ。 Please drink your coffee and think about it. It's more interesting than a crossword puzzle. But don't think about it too hard, that'd be serious.
ITT we add "... BUCK NAKED!" to the end of sentences in world history莫囂圓隣之大相七兄爪謁氣 吾が瀬子が 射立たせりけむ いつかしがもと 全裸でWyrg gende acbire madentag wher myne Seko once stode, at the rootes of Itsukashi ...BUCK NAKED!

As you can see from this sampling, there is much foolishness and many crackpot theories about No. 9, which gives one the idea that it has never been properly translated. So, I put this question to sci.lang.japan and asked whether there was any one proper reading, and whether Dr. Yonhi was correct. I received a reply in Japanese, which was not what I was expecting, but I attempted to translate it.

Kouji Ueshiba's response to my questionTranslation (inaccurate)
漢字あるいは漢字を基とした万葉仮名で書かれている万葉集の和歌を朝鮮語で読もうとする試みはかなり前からあります。There have been many attempts to read characters of the Manyoshu's poems in Korean.
『人麻呂の謎』とかいった題名で人麻呂の和歌を朝鮮語で解釈しようとした試みもあったように思います。I believe the book about the attempt to translate Hitomaro's poems in Korean was called "The Hitomaro Code".
物珍しさもあって私もチラッと見たことはありますが、まあまともに取り扱うべきものとも思えませんでした。[Like you,] I was also curious about it, so I kind of skimmed it, but it didn't seem to explain it very well.
なお、9番の歌は、全く読めていない訳ではなく、最初の2句についてだけ意見が大きく分かれており、3句以降はほぼ類似した読みが与えられています。In any case, as for the reading of No. 9, opinions are divided only about the first two lines, and for the next three phrases the readings are mostly similar.
また、読めていないのはこの歌だけでなく、他にも何首かあります。千年以上写し写しで伝わってきた4000首を越える和歌がこれだけキチンと読めればそれはそれで十分なような気もするのですが・・・Moreover, there are other poems [to deal with] that cannot be read. If these 4000 tanka transmitted by copies of copies over 1000 years can be read this well, I kind of think that's good enough...
上 柴 公 二Sincerely, Kouji Ueshiba


The problem of Manyoshu No. 9 has never before appeared in English literature, but it is well-known in Japan. After 1000 years of study, professional classicists have come to no conclusion about it, which leads to offbeat theories such as making the kanji into Korean or Hebrew. It is likely that, like so many other mysteries of our past, there will never be a definitive solution which turns this sad half-a-poem about a lost lover into a complete and fluid song.

This is unfortunate, but as Mr. Ueshiba writes, there are 4,515 poems of the Manyoshu which have been properly translated, and it's much easier to enjoy those! It's impressive that Japanese classicists have been able to reconstruct so much of the ancient texts and poems. So, don't get too bothered about it.