Simply Haiku: A Quarterly Journal of Japanese Short Form Poetry
Contents Archives About Simply Haiku Submissions Search
Autumn 2006, vol 4 no 3

A column by Robin D. Gill


Line in Haiku (& waka)

To parse or not to parse and how to
HIC   robin d. gill

I   Uninformed Translators?

In Rise, Ye Sea Slugs! (2003), I wrote that

“In Japanese, a haiku is a one-line poem touching upon seasonal phenomena, natural or cultural, about seventeen syllabets . . . ”

So, when I came across the following words of Steven D. Carter in his recent Simply Haiku interview (by Robert Wilson, Vol.4, No. 2, Summer 2006), I naturally thought it might be about me:

“Unfortunately, there are uninformed translators who have claimed that poems were usually recorded in one line, which is patently false. (That they can be recorded in one line does not mean that they always were.) I want to draw attention to the truth of the matter, pure and simple. After all, lineation is one of the few tools of a translator, and giving it up for no good reason is silly.”

I also recalled the line from that song “You’re so vain (you probably think this song is about you),” for I do not know that Carter has seen my work. Regardless, the paragraph ticked me off for several reasons:

1) The fact that poems are usually recorded in one line, does not mean they always are – this is too obvious to bear stating and in no way makes the former claim “patently false.”

2) The matter is not pure and simple but complex – for a start, we need to know what “recorded” means.

3) The manner in which the originals are usually recorded should not dictate the translator’s presentation of the poems to begin with.

Reading between the lines, however, I later came to see that someone (probably plural) must have come down hard on Carter’s innovative manner of presenting waka – which I am favorably disposed to – and his strident tone doubtless reflects defensiveness rather than belligerence. I suppose this is not surprising, for people tend to get very aggravated about form. Or, that is my impression, for I vaguely recall ferocious argument about one-line translations of haiku, though that might have had something to do with the personalities of those involved.

II   What you see in Japanese

Before discussing the translation of waka or haiku, let me try to describe how the originals of both are found and, at the same time explain why the meaning of “recorded” matters.

Most journals, the logs kept for renga or haikai meets and the books reproducing large numbers of these for many aspiring poets and other readers, do indeed record the poems in single lines, one next to the other, with the name of the poet below or between. This goes for most waka collections (though, I confess to not seeing as many reproductions as I would like to in order to make such a judgment), too, though the 31-syllabet waka are sometimes recorded in two lines, both because the paper could be too narrow for a large brush to fit a waka easily in a line and because waka do tend to break between the first 17 syllabets and the final 14, though this break was as often as not indicated by a slight space rather than a new line. With haikai, the 17 syllabet hokku was almost never broken into two lines. If it was, it was simply a matter of miscalculation rather than poetic intent. When books came to be printed in greater numbers during the twentieth century, not only did the practice of putting poems into single lines continue, but the tiny print permitted vast numbers of poems to be squeezed on a page. With haiku, some books have double columns and with senryu (the same 17 syllabets), one may even find three columns – I have a pocket-sized book with 60 senryu per page!

On the other hand, when we look at the artistic presentations of waka or hokku, either as pure calligraphy or together with a picture (haiga), we find poems parsed in more ways, and sometimes more minutely than anything found in English outside of experimental modern poetry. The breaks could indicate pauses or they could reflect a purely aesthetic choice – some characters look better at the end of a vertical line than the top, etc. They could create fairly uniform blocks of characters or they could be scattered almost at random between parts of the picture. In Carter’s Traditional Japanese Poetry, we find reproductions of 31 syllabet waka split into anywhere from four to ten lines, or clumps of writing on the picture (usually, portraits of the poet) cards. One haiga by Issa reproduced in Blyth’s Haiku (Spring) shows a ku divided into five lines: Garden butterfly / child crawl-when / flies / crawl-when / flies – you can see he is reproducing the crawling progress, none of which is in the picture, which is only a low aerial of his hut (My direct translation to show what was happening. Also, if you have said book, Blyth’s “The house and Issa too” is a mistranslation, for it means, “the house, too, by Issa.”). Note that the vertical lines in both it and the other Blyth reproduced next to it go left to right rather than the common right to left! As might be expected, waka or ku brushed onto long thin strips of paper (tanzaku) tend to have one or two lines and those on rectangular or square cards (irogami) tend to have anywhere from three to seven.

III   Haiku as Plastic Art

We have all seen Japanese pictures with writing dispersed, or placed here there and everywhere, usually in a manner pleasing to the eyes. Today, we find this freedom carried over to cartoons. To my mind, this is largely a function of the direction of the writing. As Japanese can be written vertically, a story line which moves horizontally may be accompanied far better than can be done with horizontal writing. In Rise, Ye Sea Slugs! I hypothesized in considerable detail that this verticality also made the single line presentation of a haiku work better in Japanese than in English, because the vertical line resembled an object whereas a single horizontal line was nothing but a horizon, but I did admit to reading one horizontal single line haiku which, I felt, worked: it was by Allen Ginsberg and the mosquitoes blown across a marshland matched such a design. And now I think it would look wonderful if the new ability to move font up and down a few points could loosen the line up a bit (wait until you see what I do in Cherry Blossom Epiphany – hurricanes permitting, out this fall)!

Yes, it is odd. Japanese is better with just one line and in any number of them! Such freedom is fine, and I find it extremely enjoyable. However, is it not tantamount to saying that Japanese poetry is not lined?

There is an interesting parallel here to something else true for Japanese writing but not for Occidental languages. It does not make spaces around words. Of course, it can, and in some of the oldest poetry where the lack of a proper mix of Chinese characters and phonetic letters (the tendency at the time being to use one or the other) make it hard to quickly read, one can find space within lines (i.e., clusters), but generally speaking, you find no spaces. This makes it easier to fit Japanese around pictures, for no one would complain of the breaks. In English, on the other hand, where we are used to certain places for breaks, they cannot occur anywhere else without irritating us. In translation, the worst case I know of is with waka. The full Kokinshu collection translation by Rodd and Henkenius (1984/1996) is full of odd spaces within lines and absurd enjambments (such as a single “the” sitting like an orphan at the line’s end!). It looks horrid and I sympathize with friends who claim to have tossed the book right out of their libraries. The problem is that it is done for purely syllable-counting reasons with no thought for aesthetics. Steven D. Carter, in his newest work (Just Living: Poems by the Medieval Monk Tonna, 2002; reviewed by Robert D. Wilson, Simply Haiku, Vol.4, No. 2, Summer 2006) does odd things with space, but does so artistically, so I am all for it (I wish he would pay a bit more attention to the beat and less to the syllable count, but that is another argument).

While I am all for Carter’s playful broken lines, it should be recognized that he is recreating a form of artistic presentation and not the more economic ways of recording the poems. Yet, I do not mean to minimize the importance of such design issues. The style does affect the reader’s grasp of the meaning. I would only add that such an approach in no way invalidates a book introducing twenty waka per page in single line translation, if such exists, or, for that matter, the vice-versa. Models of both types can be found in the original. The same thing can be said for haiku. It is as valid to have them broken artistically into many pieces on single pages as to squeeze them in fifty to a page, one line each.

IV   Personal choices in lineation

Personally, I favor dense books with many poems per page, for, like the majority of people in the world I am poor (one reason I do not yet own Carter’s newest) and cannot afford to buy, ship or store many books. If I see one I like, I am happy to copy it, myself, onto a piece of paper and hang it on the wall. And, I cannot help wondering if our world can afford the luxury of the type of book we have come to associate with poetry.

Yet, most one-line translations just do not work. As Carter notes, “lineation is one of the few tools of a translator.” What he means is that the translator is disadvantaged in many ways, and the line, like italics (a terribly under-used tool English has that Japanese lacks) or the title (something Carter does not add, Harold Stewart of the rhymed couplets always added, and I sometimes add) comes in useful. How? Japanese have much surprise in their poetry, largely accomplished by ambiguity and punning English cannot match without line-breaks. (I will elaborate on this all-important matter when the HIC series is enlarged into a book. Meanwhile, you may note how lines work in translations of Japanese poetry by comparing how they read to what happens when it is rewritten in a single line.) At any rate, since single-line translations transmit less information in English, they would require more notes and notes are something we never have enough of (Usanians seem to be strangely allergic to them).

Why do I favor three lines arranged symmetrically? To give me the object (pronounced in the French way) feeling I get when seeing a Japanese haiku written in a single vertical line of calligraphy. My biggest problem is the damn punctuation. The hyphens and the dot, dot, dots, so common in English haiku and translation, look horrible and even the exclamation point and question mark look awkward, aesthetically speaking. More so than in my earlier books, I am now dropping almost all such punctuation and leaving it to the reader to decide what is better as an exclamation and what is improved by an interrogative. When you consider that Japanese must often reread a ku to get it and that the common caesura, kana (or ya) can include either a “!” or a “?” or neither feeling, so that the Japanese reader also must decide, this could be justified as “accurate” translation though the English major may throw a fit! But, I do not do this consistently, for sometimes I grow afraid the reader won’t get it.

But, center balanced is not the only answer. I am also experimenting on a step-approach, where the result is something between a single line and several. I experimented with it on some ku by Geerte Verbeke appended to Fly-ku! On the whole, I like three-steps, but since the grammar within not a few haiku in Japanese seems to split in two rather than three, I would not even complain were someone to do some in two lines, and others in three.

I also have developed a new way to present waka, not so revolutionary, perhaps, as Carter’s creative parsing, but different in two ways from any previous translations I know of. I find it very satisfying on the eyes and superb for catching both the wit that is found in most waka and the spirit, which is more lyrical/romantic than haikai. It will not be described here at this time, for I want it to surprise readers of Cherry Blossom Epiphany, which will include a hundred or so waka.

One very common thing I can’t stand is no-caps on ku or waka that are not center-balanced. Left-margin lines (even with indents) need capital letters. True, Japanese does not have them, but the top of a vertical line is automatically the head, while this is not true for a horizontal one. We have writing going right-left and left-right, but bottom-up is almost unheard of.

V   A Paradox that is no problem

What puzzles me the most is that every artistic instinct in me favors asymmetric over symmetrical balance, yet I find myself a proponent of centering, something aided, I might add by the fact the middle line is plumpest. But that is only for print. If I were to combine my ku or my translations of ku with paintings, I would certainly ditch the center-balance for a free approach that would vary with the design. And, none of this is dependent upon anything I know about what Japanese may do or not do.

Send comments anytime to and to see how much fun translation can be visit my site and check out the paraversing links.

The writer apologizes for no illustrations, no examples, and an altogether rushed job. Excuses include: proofing and indexing 3000 translated haiku in a 740 page book while suffering from snow-blindness (If you have no sunglasses do not let the family force you to go to Key West!), and keeping a watch on 2 houses, one without shutters (facing hurricanes, not good) and one with a cat, while the computer blacks out and otherwise misbehaves. But, how good my head feels! I dug into the dresser for a clipper bequeathed from my father this morning and the crown I shaved down to a uniform 5/8”now feels the air from the pair of fans, with the guards removed (dangerous, but they work and look better), that keep me from overheating.