Poetry Magazines - The best collection of poetry magazines online - Simon Armitage

Last autumn The British Museum Press proposed to me the exciting challenge of producing a book of Japanese haiku. Laura Brockbank, the editor who commissioned it, wanted a ‘gift book’, attractively presenting Japanese works of art from the Museum’s collection, face to face with haiku that ‘spoke’ to the works of art. No exact representation or illustration was to be aimed at, and when this happened, it may have been by chance.

     We decided to include three versions of each haiku - the original in Japanese characters, the transliterated version in Roman characters, and an English translation.

     As I could not cover all these bases myself, we needed a Japanese collaborator. We asked Akiko Sakaguchi, a BHS member living at present in the UK, to provide the Japanese versions and contribute from her expertise on various points. Laura, who did the picture research, and resources available to her at the Museum, completed the team. The purpose of this article is to record some of the useful questions we asked ourselves, and explain how we overcame particular problems.

     Question 1 How were the Japanese versions to be created? There were several possibilities: brushed calligraphy, a computer font that simulated calligraphy but in a standardised way, and a computer font that was plain and simple, like a newspaper might use. We chose a ‘calligraphy font’ from the five or six Akiko was able to show us (it is called HGS Gyoshotai).

     Question 2 How would the Japanese versions be laid out? One way (used for example in Blyth’s Haiku) is to give the Japanese words in an unbroken horizontal line, reading left to right. This is ‘safe’, because it avoids decisions about breaking the text into meaningful phrases, and the Western reader can follow it, but it is a gross distortion of the accepted Japanese way of doing things.

     One unbroken vertical line is another possibility. This is ‘safe’ too, and has the merit of resembling a Japanese tanzaku, but with the English text flowing one way, and the Japanese text another, it makes uneconomic and untidy use of space. We opted therefore for a three-line vertical representation (persuaded perhaps by the pleasant results of this arrangement in Chiyo-ni by Patricia Donegan and Yoshie Ishibashi, pub. Tuttle 1998), but we decided to go further in the direction of shikishi simulation. Whereas the Tuttle designer chose a uniform lay-out, with a regular number of syllables per line always on a downward incline from right to left, Akiko was encouraged to do what she would have done if she had had a brush in her hand - design each shikishi individually, with irregular configurations of lines, variable numbers of syllables per line, and the number of lines most usually three, but not uncommonly four. As she herself puts it, “Haiku have their own lives, one of those lives being a visual one.” She often tried three or four different arrangements of the characters before deciding on the one that pleased her most. This was an original and creative way of doing things. Now it is for everybody to judge for themselves how much more interesting the Japanese versions can be if the aesthetic principles of Japanese calligraphy are given free rein. Here are three examples to illustrate how the calligrapher’s mind may work.

Fig 1 inazuma no naka inazuma no hashiri keri

lightning
running down inside
lightning

Inahata Teiko

Here, Akiko saw the opportunity to make the calligraphy something like the lightning flash which is illustrated in the accompanying picture. I have superimposed lines on the shikishi to show what was in her mind.

Fig 2 shiratsuyu no tama fungaku na   kirigirisu

grasshopper -
do not trample to pieces
the pearls of bright dew

Issa

The lone-standing word is tama, which is an example of kake-kotoba (or ‘pivot-word’) - a word that is spoken once, but which the experienced reader recognises as having a double meaning, i.e. the reader understands shiratsuya no tama and then tama fungaku na, The calligraphy is able to signal this by isolating tama, whereas the original cannot repeat the word without exceeding the syllable count (while the double meaning is altogether beyond the transliteration). Also, the word tama means ‘dewdrop’, and isolated like this it may actually look like one.

Fig 3 mienu me no   ho no megane no tama mo fuku

this one eye sightless
but on that side also
I polish my glasses

Hino Sojo

An effect intriguing to the Japanese eye is created by ending each of the first three lines in the character (no): we have the effect of a comical face wearing glasses, which illustrates the haiku.

Question 3 Were there any further problems in rendering the Japanese versions? Yes, there were. Certain characters used, for example by Basho, are considered archaic and have been replaced with others in computer fonts. We had no choice but to adopt the modern alternatives. On other occasions there is more than one acceptable character for a particular word (typically, one in kanji and the other in hiragana.’ This tended to come to light when we found RH Blyth in the 1950’s had used a different character from the one preferred in the ‘authorised’ encyclopaedias of haiku, known as Haibungakudaijiten (‘The Great Dictionary of Haiku Literature’ ), pub. Kadokawashoten, 1999; Haiku no kaishaku to kanshojiten (‘Encyclopaedia of Interpretation and Appreciation of Haiku’), pub. Kasamashoten, 2000; Basho taisei (‘The Complete Works of Basho ’, pub. Sanseido, 1999); and Buson zensho (‘The Complete Works of Buson’, Vol 1 pub. Kodansha, 1992). Here we followed the advice of the encyclopaedias.

Question 4  How should the transliterations be spaced? The ‘safe’ way might be to print all the transliterated words in one continuous line. You can see the effect in Fig 1. Apart from looking forbidding, it denies the reader any insight into how a Japanese speaker might normally phrase the haiku when delivering it aloud. We felt it would be helpful, as many other anthologists including Blyth have done, to insert wide spaces between phrases (typically of five, seven and five syllables), as in Fig 2. However, once again there seemed to be good reasons for not making 5-7-5 an inflexible rule. Fig 4 illustrates why it would not have been sensible to do so.

Fig 4 umi kurete   kamo no koe   honoka ni shiroshi

the sea darkens -
the voices of the wild ducks
are faintly white

Basho

In this haiku, honoka ni might seem to be required in the middle section, to lengthen it, but in speech it is inseparable from shiroshi. In the calligraphy version, a pleasing aesthetic effect (the characters for shiroshi might be a bird in flight?) is made by separating honoka ni from shiroshi without associating these words more closely with kamo no koe.

     In short, although a one-to-one correspondence between the shikishi and the transliteration can often be achieved, in particular cases standardisation would result in some loss, either of aesthetic effect in the shikishi or representation of the oral poem. The transliteration aims at pleasure for the tongue and ear, the calligraphy at pleasure for the eye. These cannot always be made to coincide.

Question 5 Are there any conventions affecting the choice of colours one can print in? When it was suggested that the shikishi be printed in red, Akiko explained that in Japan black is considered the only acceptable colour. Red, we learned, is suitable only for corrections, and the Japanese eye has long been trained to see black and white as the ideal combination of two colours which comprehend all others.

Question 6 Did existing transliterations truly represent the spoken word? Occasionally a case can be made for transliterating a Japanese character in more than one way. We opted for the one closest to the sounds made in everyday speech. An example occurs in Fig 5. We could have used yasumemu as an alternative to yasumen, but the latter gives a better idea of what you will hear most Japanese speakers actually say.

Fig 5 uguisu ni    temoto   yasumen  nagashimoto

Question 7  There are many respected sources for translations of haiku, but are they always reliable? Kaj Falkman, in various previous issues of Blithe Spirit, usefully challenged some of the ‘established’ translations, and Akiko and I made it our policy to review all existing translations of whatever provenance. For example, we came across an example of ‘genderisation’. Blyth’s version of the haiku in Fig 6 is on the left. We progressed through the version in the middle and finally decided on the one on the right.

Fig 6 ogi nite  shaku wo toraseru   botan lana

There is nothing in the Japanese original to indicate that it was a male who did the measuring. Not to know whether the doer was male or female pushes us into use of the passive voice, against one’s inclination, but in neither the second nor the third version is this too flagrant. The weakness of the second version is that it permits a pathetic fallacy, attributing volition to the peony. In the final version, there is ambiguity: does the compulsion come from inside the poet or inside the blossom? Also, the final version has a semblance of ‘cutting’, which the other two versions lack.

     On the ‘vocabulary front’ we came across some peculiarities of Japanese, which led us to differ with previous translators. On the one hand, Japanese makes some unexpected distinctions (eg different words for folding fans and round, non-folding fans, a fact which was disregarded by Blyth, who uses ‘fan’ indis-criminately). On the other hand, the language used not to distinguish clearly between frog and toad, or between cricket and grasshopper. Words for these could once be used interchangeably. This, of course, gave us a certain freedom in linking some haiku to some pictures (a haiku that had previously been translated with the word ‘cricket’ in it might just as well be rendered with ‘grasshopper’, though this might offend those who had got used to a ‘traditional’ version.) Works of art have sometimes acquired ‘traditional’ titles; we found ourselves pairing our grasshopper and cricket haiku with a picture of a grasshopper formerly labelled ‘cicada’.

     To satisfy design requirements it was necessary to harmonise punctuation. In particular, Blyth’s heavy use of punctuation and capital letters clashed with the minimal punctuation and use of lower case that are now fashionable. To achieve consistent treatment we had to obtain permission from the original publishers ‘to bring Blyth into line’. (We sincerely hope this has not disturbed his peace in the Elysian Fields, where punctuation is hopefully a matter of trivial concern.)

Question 8 How should the collection be organised - by season, by topic, or whatever? We opted for the traditional arrangement by seasons of the year, but as a result we encountered several problems. One of these is Western expectations: for us New Year is a short period surrounded by winter, for the Japanese it is (or was, when it was in February) the precursor of spring. For them it is also the first season. We decided it would suit the poetic effect we were trying to create for the typical British reader if we placed the few New Year haiku at the end of the book. Another problem was to satisfy  the publisher’s desire,  as far as possible,  to have a book of four equal seasonal sections. The fact is that spring and summer haiku (or the seasonal topics that inspire them) are far more numerous than those from autumn and winter (Blyth’s Haiku devotes nearly 700 pages to the first two seasons, well under 500 to the last two.) This disparity is echoed in the images that Japanese artists like to portray,  which are broadly common to both poetry and art.

     Sometimes, however, artists use an image more loosely than the saijiki permits. You have, for example, to be careful with cats. Any hint of them being in love, and they’re in spring. But in a poem where a cat is unable to catch a leaf with its claws, it is the leaf, not the cat that determines the season, the cat is merely an accessory to the fact. If the leaf is falling through the air, we have an autumn haiku; if it is lying fallen on the ground, it is a winter haiku.

     This became critical on one occasion when we felt we had established a credible ‘fragrance’ between two haiku on one spread, and between at least one of them and the accompanying work of art, and yet our combination was flawed. Fig 7 is the example.

Fig 7 kyonite mo   kyo   natsukashi ya  hototogisu

even in Kyoto
when I hear the cuckoo
I long for Kyoto

Basho

haru hitori    yari nagete   yari ni ayumi yoru

alone in the spring -
hurling a javelin, and then
walking after it

Nomura Toshiro

Do you agree that listening to a cuckoo’s call, echoing back yet at the same time going further and further away, yields a sensation similar to throwing a javelin and walking after it? Maybe not, but that is what ‘fragrance’ is like. A hint that requires the reader’s sensibility to bring it to fruition.

     The hototogisu or Japanese cuckoo is a summer kigo or ‘season word’. We found a picture of it flying over summer flowers. But inadvertently we had already fallen into the assumption that the cuckoo, as in Britain, heralded the spring. We had paired it with an exemplary modern haiku, for which no picture i.e. of a javelin thrower, would ever be forthcoming. There would have been no difficulty in abandoning this second haiku and moving the first one to the summer section. But we chose not to. In a case like this we thought poetic effect, in particular trying to exemplify ‘fragrance’, was more important than sticking in a  pedantic way to the prescriptions of the almanac.

Some further points

Akiko opened my eyes to the unexpected allusiveness of certain haiku. The publishers wished to include a picture of rabbits in tall grass. The haiku in Fig 8 was the only one we could find that corresponded closely enough to it.

Fig 8 meigetsu ya   usagi no wataru   suwa no umi

the harvest moon -
rabbits go scampering it
across Lake Suwa

Buson

A dried-up lake in autumn? Whatever was Buson on about? In fact, this is a fantasy haiku. Buson recalled the legend that the wily fox is supposed to cross Lake Suwa in winter, running over the ice. And seeing the full autumn moon, which as we all know has (in the Far East) a rabbit in it, he saw the reflection in the water and imagined that rabbit passing over it. Well, well.

     It was desirable to meet the constraint of providing (within a mere 71 haiku, as it turned out to be) examples not only of the major poets (Basho, Buson, Issa, Shiki) but also of a broad range of other poets, down to our own times. The number of poets eventually reached 25. In the modern period we wished to be as even-handed as possible between representatives of the groups that call themselves ‘traditionalist’, ‘modern’ and ‘avant-garde’.

     The book closes with a section of biographical notes. Useful material for these can be gleaned from various sources (Higginson, Ueda, Shirane, Yuasa, etc) but none of these authorities deals with all the poets, and they don’t always agree about what they do tell you. Not only do poets get born and die in different years, but they may have different names (Chiyo-ni or Chiyo-jo? Kamijima or Uejima Onitsura? Akutagawa Ryonosuke known for short as Akutagawa by some and as Ryonosuke by others, and by others not shortened at all). This is not to say that any of these authorities was absolutely wrong; they probably used the best information available to scholars at the time they composed their works. We went by the ‘authorised bibles’ (see above).

     I suffered some initial worry that the works of art would narrow the opportunities for presenting a generous range of haiku, since the quest for beauty in art might overwhelm the quest for truth in haiku. I didn’t want to add weight to the quite prevalent idea that haiku only deals with ‘pretty things’. There was also a risk, since only one of the available works of art was of a later date than the mid-nineteenth century, and the majority from an earlier period, that we would give the impression that haiku had fossilised (and was therefore doubtfully relevant to our times). Laura helped us to alleviate these problems by allowing us to include a number of modern haiku which had no direct connection with the accompanying work of art, but carried some oblique reference to another haiku which did illustrate the picture. Fig 7 as already noted, is an example of this. To dispel any misconception that haiku is universally ‘genteel’ it was also desirable to include one or two ‘earthy’ subjects. For example, one of the most lush pictures selected, entitled The Chinese Beauty Yang Guifei, is accompanied first by a haiku of ‘the beautiful sort’ (the one about the peony, Fig. 6 above) and then by Issa’s

the young girl
blows her nose
in the evening glory

Our overriding principle was to provide the reader, newcomer to haiku or someone deeply read in the field,  with a genuine enjoyable poetic experience, and we feel confident we have achieved it.