Ryūka Zenshū 16

The Ryūka Zenshū1 (「琉歌全集」, ‘Complete Collection of Ryūka2) is the largest collection of the ryūka style of Okinawan poetry. It was compiled in the 1960s Shimabukuro Seibin (島袋盛敏) and Onaga Toshio (翁長俊郎), and transcribed phonetically into the Shuri lect1 of Okinawan by the latter.

Ryūka (琉歌 ‘Ryūkyūan3 poems’) are the traditional poems of Okinawa. They date to at least the 1700s CE, as the term ryūka itself first appears in the Konkōkenshū (「混効験集」, compiled 1711 CE), the earliest dictionary of Okinawan. This uniquely Okinawan poetic tradition likely dates back to even before the invasion of the Satsuma domain in 1609 CE, perhaps even as far back as the 1400s CE. Ryūka as a poetic style are not confined to the capital of Shuri, with several surviving examples from throughout the Ryūkyū Archipelago. Neither are they confined to the upper classes, as several examples exist from Onna Nabe (恩納なべ), a woman who lived as a farmer in the village of Onna (恩納村) in northern Okinawa, as well as from Yoshiya (よしや), a woman sold into slavery as a prostitute in the Nakajima (仲島) red-light district of Naha (previously on the southern banks of the mouth of the Kumoji River (久茂地川 Kumoji-kawa).

Ryūka have a set meter, most typically having three lines of 8 syllables and one line of 6 syllables, for a pattern of 8-8-8-6 (or 30 total syllables). At first blush, however, almost no written ryūka appear to resemble this meter. For example, the poem I will discuss in this post, RZ 16, is written as follows:

Original Text
かれよしやいつもかれよしどめしやいるただいとの上から行ぎやい来ちやい

Transliteration4
kareyosi ya itumo / kareyosi do mesiyairu / tada ito no UE kara / Igiyai KItiyai

The poem appears to have 35 syllables, arranged in a pattern of 8-10-9-8. We know, however, that even very early in the history of Okinawan, sequences of vowels are reduced to individual vowels through coalescence (where CV1+V2 → CV3ː) and/or glide formation (where CV1+V2 → CGV2ː). For example, 上 ‘up; top’, which I transcribe as UE, would have likely already undergone a process of glide formation (and further changes) into something resembling modern Okinawan ッウィー [ʔʷíː] ‘up; top’. By way of these phonological changes, we can “fix”, or perhaps more accurately, reconstruct the poem as it was intended, which just happens to “fix” the hypermetrical lines, rendering all metrical, as below:

Reconstruction
ka.i.ju.ɕi ja ʔí.tsi.n̩ / ka.i.ju.ɕi du mi.se.ɾu / tá.da ʔi.tu nu ʔʷíː ka.ɾa / ʔń̩.dʑa.i tɕi.tɕa.i

It is likely the case that syllables, rather than morae, count when it comes to determining the number of units allowed in a line. There is some evidence from modern Okinawan that ッウィー [ʔʷíː] ‘up; top’ is two morae: it has a clear fall in pitch over the length of the vowel, something that cannot occur with one mora words. It also has a long vowel, which means it cannot also have a geminite (or “heavy”, if you prefer) onset. But in this poem, the author is at least bending the rules, if not using syllable counts over mora counts when it comes to determining whether or not a line is metrical.

Onaga inconsitently has the more conservative karijusi (rather than the more innovative kaijusi) and the more innovative cicai (rather than the more conservative cicari)While it is unclear when the syllable /ɾi/ loses its consonant, the fact that there is variation between <ri> and <i> in the original text, with a more etymological spelling of かれよし <kareyosi> and a less etymological spelling of 来ちやい <KItiyai> (from earlier *ki-te ar-i) suggests that this had already taken place.

Interlinear gloss and free translation
kariyushi=ya ‘ítsiN / kariyushi=du mise-ru / táda itu=nu ‘wíi=kara / ‘Ndza-i chicha-i
good.fortune=TOP always / good.fortune=FOC give.HON-ADN / just silk=GEN top=ABL / go\PFV-RLS come\PFV-RLS

‘Good fortune! Always give [me] the good fortune [of travelling]! Only coming and going from the tops of silks.’

This is a poem asking for good fortune in the form of travel. As a courtier in the court of the Ryūkyū Kingdom, being assigned to be on a mission to Japan, China, Korea, or the like, was among the most prestigious duties that one could be granted.


  1. Abbreviated as RZ.
  2. Shuri is generally considered a dialect (or topolect) of Okinawan. While Shuri is geographically contained to the former capital of Okinawa, Shuri, it is perhaps better thought of as a sociolect, being spoken by the descendants of the Okinawan gentry. Compare this to the status of its neighboring Naha, spoken in an adjacent geographical area, but by the descendants of the common people of the capital region. I use lect here as a compromise, rather than choosing between the more traditional terminology of calling it a dialect, or the perhaps more historically accurate term sociolect.
  3. Or Okinawan, as opposed to waka (和歌 ‘Japanese poems) and kanshi (漢詩 ‘Chinese poems’).
  4. This is intended as a strict transliteration—not a transcription—of the original text, going from the poem as it is written. No analytic devices other than slashes, to divide lines, are used.

2 thoughts on “Ryūka Zenshū 16

  1. A similar “fix” is suggested by Kawamoto in “The Poetics of Japanese Verse” – I think this is an old observation in the scholarship of Japanese poetics, but I’m paraphrasing from memory. It seems that extrametrical (ji-amari) verses in classical poetry (up to MJ at least) pratically always include a vowel onset after an open syllable, which must have been resolved as a glide or coalescence:

    Aki no ta no
    Kari-po no Ipo no
    ★Toma [wo a]rami
    Waga Koromo-de pa
    Tuyu ni nuretutu

    The abundance of historical developments like “to ari” → “tari”, etc. reinforces this reading.

    In modern pronunciation, however, readers favor just reading syllables faster, creating a three-beat-in-two effect similar to a triplet in music. As a result, modern poetry has abandoned the vowel-onset rule.

Leave a Reply to leoboiko Cancel reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.