Which is partially true…. in Japan.

Elsewhere things a more complex. The growth in the last fifty years of corpus of theory and practice of English language haiku has failed to create any real consensus. Instead the field is notable for its fractiousness and the extremity of some positions taken. On the one hand it is argued that there has been a parting of the ways: English language haiku is a literature in its own right, and need pay no heed to the Japanese tradition. On the other there those who maintain that the haiku is not really a form of literature at all, and is better considered a metaphysical or contemplative tool. In the heat of claim and counter claim utterly baseless assertions are common, and humour rare. It is hardly surprising that the general reader will view such internecine squabbles with exasperation. Followed by distaste.

The zip style of prosody is offered as a counter to some of the stranger excesses of recent years. The zip is an attempt to reclaim the haiku as a form of poetry, written in cognisance of the Japanese form, but in English. First and foremost this implies the use of those techniques that are natural to the language of composition - not for ideological reasons, but because that is what haiku (or hokku, tateku, hiraku), the Japanese stanzas, have always done.

Secondly it obliges a reappraisal of the concept of teikei - strict form – with a view to applying it in English.

Though this commentary in the main deals with the example of the haiku the points raised apply equally to renga - linked verse. Indeed it can be argued that those stanza structures which are generally typical of the current English language haiku adapt with some difficulty to the differing requirements of renga and renku: a sure sign that they are different in essence to the Japanese haiku stanza which is, after all, no more than a specialised application of a standard structure.

form & freedom - Which is more of an imperative: to establish boundaries, or to test them? Put another way: the problem with tearing down the walls, brothers and sisters, is that the roof falls on your head!

The elation a poet feels when coming fresh to the concept of 'free verse' is often short lived. It rapidly becomes apparent that poetry doesn’t just require some ideas and a vocabulary, it needs a structure too, and if an off-the-peg solution is rejected, a bespoke one must be found. At the same time, poets working within the bounds of a recognised form rarely resist the temptation to tweak the collective nose of their elders and betters, or even flaunt their indiscretions with relish.

There is a paradox here - a creative tension - it would seem that freedom is not independent of formal constraint; both are part of a single system.

There are many ways to frame this conundrum: Hegelian Dialectic; Taoist Duality; Dynamic Imperative; The Whim of the Deity. Whatever we chose to call it, it’s there all the same, and the suspicion must be that we ignore it at our peril.

Indeed, when one looks at the history of the Japanese form it is not the fact that the Danrin school tended to an irregular syllable count that stands out, or that Santoka has torn up the rule book, but rather that so many of the basic structural parameters of the verse form have remained constant.

Several observations flow from this. One is that the particular stanza form, so apparently rigorous, is in fact elastic enough to contain an infinity of expression. Another is that adherence to the strict parameters of the form, or a conscious and controlled rejection of them, are a fundamental element of the aesthetics that have governed Japanese verse for over a thousand years.

Why then has the English language haiku tilted overwhelmingly toward free verse? Is this because it does not need formal parameters, or because it has abandoned the attempt to identify them?

relative extent - Or: the incredible shrinking frog.

One of the more alarming features of the English language haiku is that no one is quite sure how big it is. Translated into English the world's most published haiku, Basho's frog, is, according to Stryk, six (savor ‘em) syllables long. Whereas Yuasa has it as more than twenty (22, from memory). Clearly this creature is more chameleon than frog. And a strange one at that.

Such tremendous divergence is a warning: evidence not so much of a difference of opinion as a difference of agenda. One suspects that Yuasa, in his desire to make the poetry of his mother tongue accessible, may be guilty of a little 'explaining' in his translation. But what of Stryk… could this be an attempt to make the poem conform to a super-strict model of his own? History teaches us that the disciple was ever more zealous than the master. Unless of course this particular Master was Master Yoda:

Poet good Yoda is… Ahum!

Hmmn. Or perhaps not. In part the problem is one of relative extent, of phonic to semantic ratios. Just how many sounds does it take to communicate a given amount of information in Japanese compared to English?

This isn't as difficult to establish as it may appear. One simply assembles a few texts with definitive translations in both languages - say… the Episcopalian version of The Lord's Prayer; a couple of Articles from the Universal Declaration of Human Rights; and a paragraph or two of Confucius - then we count the syllables, divide one number by the other and we have 0.88 -1.13 or whatever.

The zip long stanza settles on 15 (mathematical geniuses will have spotted that this is indeed 0.88: 1 or 1: 1.13). There are complex ‘internal’ reasons for this, which are explored below. As for the direct ratio to Japanese, well… Prof. Shirane’s excellent essay ‘Beyond the Haiku Moment’ is freely available on the web. In it he presents seven or eight seminal haiku/hokku in romaji, and offers a translation of each. The reader is invited to count up the total number of syllables in these poems and divide it by the number of poems themselves. Yowza (15)!

And don't forget to read the essay - though devotees of R. H. Blyth are in for a nasty shock.

syllables & sounds - According to those whose business it is to know such things (and as long as we are prepared to ignore the Bushmen of the Kalahari) the languages of the world are divided into two categories: stress-timed, and syllable-timed. The gulf between them is enormous - a Montague/Capulet affair - and any attempt to bridge it must spell tragedy for both houses.

It will come as no surprise to the long suffering haijin to learn that the most dynamically stress-timed language of all is English (allegedly), and the most ineluctably syllable-timed language is Japanese (allegedly). In fact any comparison is a waste of time really because English syllables are variable in length whereas Japanese onsetsu are of a perfectly fixed and regular duration. We are talking not so much ‘chalk and cheese’ here, as ‘Voluble Fool vs. Clockwork Toy’. Allegedly.

This collective 'wisdom' has been central to the development of the English language haiku. As stress-timed metre is totally alien to syllable-timed metre (it is argued), the whole idea of metre might as well be discarded. And as syllable duration is not fixed in English there is absolutely no point in having a set number - one man’s seventeen syllables might be enunciated in a flash whereas, if we come back next year, poet X will still be only half way through her recitation.

It is only logical, therefore, to define a prosody for the English language haiku wherein the phonic properties of the language are discarded, and whose syllable count is variable - a function of semantic content only: a poem that is pure ‘meaning’.

If only the world was still flat, and the Bushmen of the Kalahari would shut up, everyone could be happy with this. But no, here comes Mr Gates and his minions, and now their infernal machines can talk. Worse: they can listen.

The problem with computers is that they’re not smart enough to be people. So they keep insisting that all human beings are 94% lettuce, which is why Japanese syllables show nearly as much variation in duration and amplitude as English syllables. And that speakers in either language automatically entrain syllable length and/or stress to conform to self generated/perceived patterns. Worse still, though the periodicity of these patterns is complex, the computers tell us that it is essentially isochronic. Disastrously, this isochronicity holds true not just within any particular language, but between languages. And, impossible though it may seem, the same basic patterns hold true for every language that has yet been analysed.

Which is what the Bushmen of the Kalahari have been trying to tell us all along - if only we would stop laughing at their funny haircuts and oddly clicking tongues - Bushmen have their poetry too. And it’s not so different from ours. The genome project has proven that we share the same Granny; it would seem she has knitted us all the same cardigan.

Speech-recognition theorists, therefore, have been left with little choice but to completely revise the stress-timed vs. syllable-timed debate. Their pay packets demand it. But what of the implications for the men in the high towers?

Well, we could do what Frank Turner and Ernst Poppel have done, and try to explain it (ask the Great Google for ‘Neural Lyre’ and don’t forget the analgesics). Or we could muster every last ounce of bombast, bluster and indignation we possess, and simply heap ridicule on those who would challenge the established 'truth'.

Alternatively, given that we are discussing a form of poetry that originated in Japan, we could ask a Japanese poet. And, if we do so, we will be told that Japanese verse does indeed use sound patterns, that these patterns are (amongst other things) mnemonic, and that individual words or groups of words are chosen because they are pleasing, or otherwise dynamic in the ear. Surprise!

Of course to make use of exactly those properties that are intrinsic to Japanese one must write in Japanese. But we also know that English has a fair size palette of sounds and patterns too. And that the fundamental appreciation of cadence and structure is no more special to one language than to another, any more than it is special to one kind of human being. Welcome to the Twenty First Century.

drawing the line - What exactly is a 'Tercet'? Is it: A/ a North European sea bird? B/ a Malaysian sports car? or C/ a three line stanza? Clue: The Japanese haiku is none of these.

Every writers' workshop in the English speaking world has, is, or will be holding a haiku evening. No matter the skill of the poets involved the output is likely to fall into one of three categories: 'the gnomic statement’, 'the world's smallest sonnet', or 'the 3 item shopping list'. Though unfortunate this is hardly surprising, for not only the dictionary but also the general academic sources describe the haiku as a verse consisting of three lines.

But it isn't. Or rather: it wasn't – not until they got their hands on it.

The Japanese haiku is written on one line. Occasionally calligraphic considerations will result in the poem appearing as three blocs of characters. But this in no way equates to three lines (with their consequent line breaks) in the sense that such have been understood by centuries of English speaking writers.

Almost by definition, to produce a poem that has any of the characteristics of a western tercet (three line verse) is to produce something which is not a haiku.

The Japanese haiku comprises blocs of five, seven, and five syllables. But these are not lines of verse, not units of meaning. They are units of phonic coherence based on patterns so fundamental to the language that the eye/ear automatically regenerates them from a text that is otherwise unpunctuated and undifferentiated.

The degree to which these are not semantic divisions may be judged from the phenomenon known as 'segment straddling' (ku-matagari) wherein a single word spans two phonic blocs - the first syllable belonging to one group, and the second belonging to the subsequent group. Significantly, to attempt this in an English tercet is to create a nonsense. Or use a lot of hyphens.

At their simplest many Japanese haiku (or hokku) present two image sets articulated by an exclamatory cutting point (kire) which is generally emphasised and inflected by the use of one of several ‘cutting characters’ (kireji). In terms of semantics - in terms of meaning - the poem often does divide as 5/12 or 12/5, but the cutting point is not dependent on the 5/7/5 pattern and the semantic division may well be 9/8, 10/7, 11/6 or an invert.

By contrast the English language haiku, when written as a 5/7/5 tercet, obliges the cutting point to appear at the end of line one or line two, and so limits the possible semantic organisation to 5/12 - 12/5 or, in the case of the 3/5/3 stress pattern standard proposed by some schools, to 3/8 - 8/3. Thus, when taken together with the inadmissibility of segment straddling, the possible structural (and, consequently, expressive) range of strict-count English language haiku is massively restricted compared to its progenitor.

This knotty problem has generally been solved by the simple ruse of preserving the three line layout whilst abandoning the concept of strict form, a practice which may be likened to throwing away the burger and eating the napkin.

If only it were possible to find a type of lineation and cadence as natural to English as the 5,7,5 and 7,7 sound clusters are to Japanese…

punctuation with pauses - Punctuation has had a bad press for at least a generation. It has been closely identified with grammar, which is synonymous with ‘rules’. And ‘rules’ are, like… uncool.

In so far as correct punctuation plays a vital part in formulating cogent sentences there is some truth in the association. Caxton got away with just the comma, full stop (period) and virgule, but as English prose in particular has become more complex the intricacy of the agreed system of organisation has also increased: like lawyers and accounts, the number of punctuation marks has trebled.

It is tempting therefore to see punctuation marks as the equivalent of mathematical symbols: characters which express a very precise qualitative or quantative relationship between known values or elements. However, on a more fundamental level, punctuation marks might be better equated to pause indicators, analogous to the rests in musical notation, and this is especially the case when text attempts to convey the cadences of speech.

The often repeated formula that Japanese haiku are the length of one breath is highly suspect, but it is certain that the original cadences arise from the oral tradition - not from text-based literature.

Some contemporary writers of English language haiku seek to emulate the Japanese form directly by writing their poems on one line, entirely without punctuation. There is a commendable logic to this approach but it does rather rest on the assumption that the senses will automatically synthesise patterns from the English text just as surely as the Japanese reader will recognise the innate shapes of that language.

A marked benefit though is that the syntax achieves more flexibility, freed as it is from the tendency of any given punctuation mark to limit rather than increase the range of relationships between clauses or image-sets.

The zip style of prosody seeks to maximise this flexibility of syntax whilst giving markers to help control the cadence of the stanza. The double space and line break are intended to pace both the phonic and semantic movement of the piece in exactly the way an orator or snake-oil salesman might. This technique is not modern. In fact it comes from the earliest written precursor of English: seventh and eighth century Anglo Saxon poetry. By pure chance it was in this same period, a mere half planet away, that the fives & sevens metre of Japanese versification was becoming established.

Fans of the X-Files or Eric van Daniken might like to consider how closely the riddle poems of the Red Book of Exeter correspond to the earliest Japanese katauta and mondo … Spooky!

expression and repression - If much modern English language haiku has taken the route of free verse, not since Stalin was freedom won at such cost. There are rules everywhere, and, as form has been abandoned, they are rules of content or style.

One set of prohibitions seems to spring from a sanguinary furtherance of Shiki’s demolition of all things Edo. We are left with a primitive formula which articulates two concrete images by means of a colon, dash, tilde or other punctuation mark. It is hard to see what relation this arrangement bears to Shiki’s own mature work, let alone Basho, Buson, Issa... but no matter, zealotry is its own reward.

Less strident, but every bit as determined, are those persons who seek to limit the haiku to the expression of beliefs associated with Zen Buddhism. Here we find the 'Haiku Moment' - a sugar coated confection found throughout Surrey and San Francisco, but oddly absent in Sapporo. Unsurprisingly this group and the former do not always see eye to eye. Both though are characterised by a very partial apprehension of a vast and, above all, varied literary heritage.

Most damaging of all is their shared antagonism toward the English language, a self-hatred which verges on the pathological. Euphony is castigated as ‘western’. Figurative language is denied as… hmmmn, 'bourgeois materialism? Humour is… denigration! The whole pathetic edifice is based on fallacy, but with just enough sense to render it, Potemkin like, solid when viewed from a distance.

It is simply untrue that Japanese verse eschews phonic techniques. Some schools, styles and historic periods might place more emphasis on one aspect than another. But they are all there. Skilful poets use them skillfully. Clumsy poets do not. Surprise!

Similarly with tropes – external reference, simile, metaphor, figurative language, all are present. And correct. Indeed there are lexicons as big as phone books filled with the minutiae of seasonal reference: their moods, and derivations. There are more poems based on the resonance of place names than most rock bands could shake an album sleeve at.

The leavening of truth that prevents the puritanical maunderings of the miserable brigade from being pure nonsense is that very short verses are easily dominated by a maladroit use of compositional techniques. The metaphor can too easily become the poem, the horse mistake itself for the rider, and when one comes from a literary tradition, like the English language literary tradition, with a scant history of short verse, such mistakes are more easily made.

But it is not the rhythm, the onomatopoeia, the alliteration that are to blame, it is the skill of the poet that is lacking. Caution is therefore sensible, outright castigation is… best left to consenting adults in private.

chamfering the square peg - At first sight the zip-style stanza structures might appear rather odd. But the reader is invited to consider which is the more idiosyncratic: the zip layout, or the fact that in 'conventional' English language renga and renku the two line ‘short’ verse will often contain more syllables than the three line ‘long’ verse.

The zip stanzas do not try to look like any pre-existing model, they attempt instead to perform in English those functions which the Japanese stanzas perform in that language.

The zip stanzas are strict form, but lineation has been separated from extent. Thus the long stanza (haiku, hokku, choku, kami-no-ku) comprises 15 syllables deployed at will over two lines, each line having a pause indicated by a double-space (caesura).

Following the ancient, and modern, convention the pause-value ascribed to each caesura is weaker than that of the line break. The nominal pause pattern is therefore weak-strong-weak, the interaction between pause structure and syntax pacing both the semantic and phonic movement of the stanza.

The long stanza is presented in such a way that the caesurae are vertically aligned [note 2009 - in practice this has proven to be a mistake. It is recommended that the layout of the two lines be more cursive; both the indentation of one line relative to the other and the caesura positioning should be arrived at instinctively. Please refer to the examples elsewhere in these pages ].

Visually the intention is to facilitate the tendency of the subconscious to read in a non-linear manner, an arrangement which facilitates fine layering and mobility of image order.

The short stanza (tanku, shimo-no-ku) is composed of eleven syllables deployed on one line incorporating two caesurae, each caesura being represented by a double-space. Occasionally, and most commonly in tanka where the two metres tend to combine, one of the short-verse pauses might require a stronger weight and is given as a line break.

At fifteen and eleven syllables respectively the proportions of the zip style long and short stanzas vary slightly in respect of the ratio of their Japanese counterparts (17 and 14 onsetsu). These quantities (15 and 11) have been adopted in order to ensure that neither English stanza fall too readily into facile rhythms. Equally it is important that the metres of the two stanzas exhibit a sufficiently differing range as to enable the macro-structure dynamics of linked verse.

The stanzas do not use overly abbreviated or otherwise compressed, notational or forced syntax. The presence of articles and principal verbs is therefore more frequent than otherwise.

All tenses and conjugations of the verb may be employed, punctuation marks are not - though some poets do make use of the two inflection markers: the exclamation mark, and the question mark (query).

The zip style has regard for the phonic properties of English, but rejects obvious versification. Similarly any figure of speech which tends to narrow the resonance of a poem is discouraged. Figurative language, uninflected simile and direct metaphor are best avoided until the poet has a comprehensive understanding of the expressive potential of the stanza structures themselves.

The zip school advances a style of prosody, not content.

Considerations such as the use of season words (kigo) and sketching from life (shasei) are entirely matters for the individual. This writer always draws haiku directly from life and, for both haiku and linked verse, adopts a high degree of seasonal reference drawn from British cultural iconography rather than the classical Japanese canon (cf. saijiki). These are personal preferences, however the poet new to haiku or linked verse might be well advised to concentrate mainly on real life experience for the former, and reserve their ambitious imagination for the latter.