To the Lighthouse: YES, Tanka Has a Bipartite Structure

Below are three excerpts (note: I added one more excerpt, Oct. 4) from scholarly references posted here in response to the following claim made by M. Kei, Editor of  Atlas Poetica:

The ‘fiveness’ of tanka is well entrenched in English, so much so that little attention has been given to understanding exactly what a ‘phrase’ is, and equally little attention given to how larger structures within a tanka function. Some accept as a given that tanka has a bipartite structure, but others point out that tanka has traditionally been written in one, two, three, four, and even five part structures (note: who are the “others?” No textual evidence or scholarly references given here). Some translators have experimented with fragmenting tanka even further (cf. Steven Carter) into six, seven, eight, even nine or ten lines Even when a two part structure is used, the position of the break has changed as fashions in tanka have changed (note: a poem’s structure is not only defined by its lineation or stanza break, but also by its theme, imagery, ... sometimes even by its footnote. For example, just think about the structural role that the footnote to Ginsberg’s “Howl” plays). There is an intuitive grasp of the3-2 structure in which three lines of observation are followed by two lines of subjective response, dubbed the “haiku+2” model of tanka, but little articulation of exactly how and why this is done, aside from the erroneous belief that tanka is a haiku with something extra added on. There is no awareness of when and why the break changed from the end of line 2 to the end of line 3, or what ramifications that has for the structure and reading of tanka. Even less knowledge is available regarding tri-partite structures, such as the very ancient 5-7,5-7,7 structure, nor any understanding of how that might lend itself to modern works in English. While some attention is given to the “rush of five lines down” (tanka without breaks), equivalent attention has not been paid to four and five part tanka….

-- M. Kei, “The Labyrinth of Tanka,” Modern English Tanka, 2;3, Spring, 2008, p. 208.

Excerpts:

I “Kokoro as Ecological Insight : the Concept of Heart in Japanese Literature” by Eric Thomas Sherlock, MA Thesis, Asian Studies, University of British Columbia, 1984, p. 107)

Since ancient times there had been a custom of having one poet write the first three lines of a waka (kami no ku), while another poet finished the poem by adding the final two lines (shimo no ku). This practice eventually became extended, through multiple authorship, to a hundred stanzas of alternately three and two lines. The new verse form, which became known as renga, first originated among court poets as a form of amusement and relaxation after an evening of serious waka [ancient name for tanka] composition. By the Ashikaga period, however, renga itself had achieved maturity as a serious poetic form.

II “`'Anti-Poeticism' in Japanese Poetic Tradition”   by   K. Yamanaka, Conference paper, 2004 International Comparative Literature Association, Hong Kong.

Traditional poetic forms are mostly variations on the basic design shown in (1):

(1) i. Verse design

5 ○×○×○＃＃＃｜

7 ○×○×○×○＃｜             b

5 ○×○×○＃＃＃∥

7 ○×○×○×○＃｜      a

7 ○×○×○×○＃｜　           c 

(note: b refers to 5-7-5, a 5-7, and c 7-7)

ii. Scansion and rendering

Owing to two-beat isochronic progression, a mora-long caesural pause (#) is inserted when word boundary falls on any of the odd syllables as in /hika-ri #-nodo-keki/ or /haru-no #-hi ni- ##/, thereby lending rhythmic variations tothe basic pattern.

Standard waka has this five line scheme. By repeating the unit (a), any length of longer poems (chôka) can be obtained; by alternately composing (b) and (c) among plural poets for definite numbers of times renga and its later development haikai is produced (two stave dialogue, ten stave sequence, thirty-six stave sequence, fifty, one hundred, etc.) . It will be easy to observe that detachment of the upper stave [or hemistich] (b) makes a hokku, or its modernized variation haiku, which, incidentally, is a formal innovation in that it lacks the usual end-signal in seven syllables , as well as in freedom of diction and relaxed canons.....

Earl Miner observes in his masterly introduction to Japanese linked poetry that “the tendency of Japanese poems to be short is well known, as is also the tendency for the short to become shorter. It is less well known that a countervailing tendency leads to the integration of shorter into larger wholes” (Miner 1979: 9). This integrative move was officially acknowledged when the anthology Kin’yôshû (1127) first included a book on linked poetry, but it is unmistakable that a long practice of exchanging love poems or capping the upper stave made by one person with a happy sequel by another eventually produced this novel poetic form.

But the “larger wholes” did not mean simple progression by a plotline or on one continuous theme. An inviolable rule was that no individual stave should have an overt semantic connection except with its immediate predecessor. In order for this peculiar poetic sequence to be viable, the first prerequisite was evidently formal that a waka stanza had a bipartite structure with an obligatory caesural pause nearly at the middle (5-7-5#7-7). But the second requirement was a special principle of progression. Early attempts were patently verbal, continuing by puns, antanaclasis – use of the same word in a different sense –, antonyms, chiastic arrangement of semantically related words and so on, but later it was discovered that refined association of ideas, and still later correspondence could function as more artistic cohesive ties.

Another important corollary from the bipartite structure was something that deserves to be called an aesthetic. Already in the 11th century, it was generally understood that poetic themes can be formulated in paired concepts like “fallen petals in the garden”, “a skein of geese migrating at night”, “increased love by absence”, etc. (cf. Teika’s “A Compendium of One Hundred Poetic Themes” ). It was an awareness more conceptual than imagistic at this stage, but subsequent spin-off of, and symbiosis with pictoral art of various forms paved the way for the imagist aesthetic associated with later hokku poetry.

Notes:

2 Earl Miner (1979: 10) states the difference between renga and haikai very succinctly with an outfit metaphor. “When renga learned to dress up properly, haikai  stepped forth as a barefoot equivalent.” “Barefoot” means relaxed canons and freedom of diction.

3 More exactly, there are two more variations on this pattern: 5-7-7; 5-7-7 sedôka and 5-7-5-7-7-7 bussokusekika.

4 As for prosody, the problems to be solved are why eight beat rhythmic structure (quadruple time), for one thing, out of several other alternatives and why it is implemented with seven and five syllables for another. The hypothetical two beat time as the basis of rhythm appears to be acceptable but the remaining question of odd number fulfilment seems to be still open. See Kawamoto 2000, especially pp. 224-27 for a detailed discussion on this point.

5 “Fujikawa hyakushu-dai”. Major categories are bi-thematic (Object & place, Time & thought, Place & thought, Time & place), while some others are single like Love, Time, Reminiscence, Grievance and Ceremonial.

6 Originally, Japanese painting started as visual representations of the universe of discourse rather than a pure mimetic art, and only later, when imagistic poetics reached its apex in Bashô’s haikai renga (=renku ‘linked verse’) did ut pictura poesis come to the fore as a guiding principle both in production and reception.


III “Waka and Form, Waka and History” by Mark Morris, Harvard Journal of Asiatic Studies, 46:2, Dec., 1986, pp. 583-4.

The early waka sentence, the line in motion of the Manyoshu does not yield in any direct fashion to pure complexity. What will define the formal strategies of waka poets is the possibility of playing off complexity against a linear ground, a ground always available for the creation of effects that may stress simplicity but that can do more. The Shinkokinshu (c. 1205) is best known for a nominally weighted and strongly segmented division of the waka structure into two unequal halves, a 5-7-5//7-7 shape which scans as an emphatic seven-five rhythm; the upper and lower segments tend to concentrate upon distinct nuclei of form and image. Yet Shinkokinshu- poets could still invest linearity with new energy. One good example is a poem attributed to the daughter of Fujiwara no Shunzei:

kaze kayou nezame no sode no hana no ka ni
kaoru makura no haru no yo no yume
(SKKS 2:112)

(In the scent of the flowers blown past my sleeves
as awake I lie, dream of a spring night, of a
fragrance on my pillow.)

Here the nominal-style of the Shinkokinshu is brought together with two strings of noun-connecting no; the result is far more elusive than the sequence of noun-no-noun articulated above in KKS 15:797. We encounter what might be termed a hyper-linear form of waka. The poet has utilized linear flow not as a linguistic force to be resisted or even terminated but as a fact of language that poetry can turn against language. The phrases hurry past swiftly, interlocking and offering few breaks for marking out predictable scansion, and never anchored or resolved by final predication. The reader can only try to hang onto bits of clear primary signifiers as they sweep along yielding fragments of larger meanings and connotations that never quite coalesce on the surface of the poem.62 To divide the kaze kayou poem into short lines and to fit them out with the stability of a stanza may, for a scholar or translator, be unavoidable. But many readers would probably agree that to do so is to thwart precisely that which makes the poem go.

(note: Of the later poetry collections, the most highly admired was the eighth imperial anthology, Shin kokinshu (The New Collection of Ancient and Modern Poems), complied in 1205 and comprising 1978 poems written exclusively in the waka form ... Still emulating the traditional ideals, [the poets] were able to reinvigorate the thirty-one-syllable form through innovative use of such devices as classical allusion, wordplay, and symbolism. Through classical allusion, they could create novel comparisons or contrasts that would expand the meaning of their poems. Through wordplay, they could bring together two disparate ideas or images and thereby evoke unusual associations. Through the use of symbols that vaguely referred to things remote and unearthly, they could transport the reader's imagination. Poetic techniques such as these had been used in earlier poetry, but never to this extent or with such dramatic effect... excerpted from Makoto Ueda's Modern Japanese Tanka, pp. xiii-xiv)

Examples:

driftwood

at the foot

of a sand dune

nobody around

I talk to it

use her lap

for a pillow --

even then

my thoughts are

all for me

Takuboku Ishikawa

Even one who claims

to no longer have a heart

feels this sad beauty:

snipes flying up from a marsh

on an evening in autumn.

Saigyo

From one darkness

into another darkness

I must go.

Light the long way before me,

moon on the mountain rim!

Izumi Shikibu

it may be my angel --

this small sparrow

      I shot

then returned home

smelling the gunpowder

Shuji Terayama

wondering for years

what would be

my life's defining moment

     an egret staring at me

     me staring back

Jeanne Emrich

a gnat’s smudge

on my forearm --

      the smallest death

      i have known this year

      but typical

William Ramsey