Jo Ha Kyū and Fu Bi Xing

ABSTRACT: Theorists and critics such as John Berger, Sabine Gross and Michel Foucault have emphasized the very different experience that occurs when a text is approached through a viewing or a reading mode. The extent to which both modes are used alters the experience again. The characteristics and history of haiku encourages extensive use of both modes. Whether reading or viewing is emphasized depends on variables that include the writing system(s) used, the language, orientation and shape on the page, and the number of lines. In order to understand the varying effects of employment of these modes, the Japanese aesthetics of jo ha kyū will be applied to original and translated haiku by English poets and haiku written by the Japanese masters. An aesthetic formulation applied to various traditional arts such as Nō drama, the current interpretation of jo ha kyū can be roughly translated as “beginning, breaking or developing, and rushing to an end”; indicating that the activity begins slowly, speeds up, and then concludes very swiftly. Today, jo ha kyū tends to be considered as applicable only to the movements that occur in temporal-based theatre. However, it was originally applied to literature and is still very pertinent to the haiku form. In addition, the older Chinese principles of fu bi xing from which jo ha kyū originated illuminate further the process of reading and/or viewing haiku.¹

__________

by Judy Kendall

The characteristics and linguistic, orthographic and cultural contexts of haiku encourage unusually extensive use of both reading and viewing modes, each of which significantly alters the experience of the text. An understanding of the varying effects of employment of these modes, in both Japanese and English, is aided by drawing on recent scientific studies of eye movements when reading and viewing, and ancient Japanese and Chinese poetic aesthetics of movement.

Haiku are strongly visual, as Jim Kacian notes,

haiku is the most painterly of poetries, given as it is to images. Yet haiku are constructed, not of pictures, but of language. They perforce must utilize the artifices of language to communicate their images, their content.²

The visual characteristics of haiku are particularly evident in Japan in the way haiku are found in calligraphy, in haiga, on tokonoma scrolls, presented as objects to be seen as much as (or more than) texts to be read. Martin Lucas puts this well,

. . . the imaginative effect of this poetry is comparable to that of Chinese and Japanese landscape painting, using a few brushstrokes to sketch a foreground subject, placing it against a spacious and indeterminate background: pines, bamboo, fishing boats, waterfalls, looming out of a mist.³

Contributing to this emphasis on visuality are aspects of the haiku’s form, scope and use of ideogrammatic as well as linear writing modes.

Both Kacian and Lucas recognise that haiku in English have inherited some of this bias towards the visual. In addition, they imply that the different ways in which the ‘artifices of language’ employed in haiku reach beyond the bounds of what is normally considered language’s terrain into the realm of pictures and even beyond that: unwritten, non-textual and even at times invisible elements contribute to the haiku’s power. In other words, an additional and important influence on the non-textual effects in haiku is the focus on what is not written, expressed or seen in the written text.

As early twentieth century Japanese writer and poet Yone Noguchi put it, the haiku is ‘a tiny star carrying the whole large sky at its back’.⁴ What is absent or not foregrounded, as in Noguchi’s ‘the whole large sky’, is crucial to the haiku’s effect, arguably even more crucial than the present or foregrounded ‘tiny star’. In physical terms, the shape and location of the haiku on the page does not only outline the text of the haiku but also the spaces which surround that text. The scrupulously minimalist confines of the haiku alert the reader’s attention to the space of the page, to what is not there, to the ‘few things [left] unsaid’. This phrase of Shinkei’s, articulated when writing to Sōgi of the aims of the renga poet,⁵ represent aesthetic values and practice that were ‘to become a cornerstone of haiku aesthetics, valued not only by generations of Japanese haiku poets but also by pioneers of haiku in English’.⁶ Thus, what is absent in haiku becomes almost tangible and visible.

This is the Japanese aesthetic of mu. Donald Richie expresses it thus,

If one fills in a corner of the paper, as did the painters of the early Sung, the unfilled portion is filled as well – filled with space, which comments on the corner, gives its body, and creates its context. Similarly, as in the formal Japanese flower arrangements, not only the sprays themselves but the space between them is considered part of the finished work. This is the concept of mu – emptiness and silence are a part of the work, a positive ingredient. It is silence which gives meaning to the dialogue that went before; it is emptiness which gives meaning to the action that went before.

This meaning, however, is one which the spectator himself must supply.⁷

Noguchi wrote, ‘The very best poems are left unwritten or sung in silence. It is my opinion that the real test for poets is how far they resist their impulse to utterance’.⁸ These sentiments have been echoed by English writers on aesthetics, with John Ruskin noting that ‘no good work whatever can be perfect, and the demand for perfection is always a sign of a misunderstanding of the ends of art’.⁹

In haiku, mu is strongly articulated through the visual continuum with which the viewer|reader engages. It is at work in the haiku’s combination of visual and textual, of reading and viewing modes, and in the gaps and shifts between these two modes of perception. Carefully modulated by specific uses of space, text and visual effects, readers are encouraged to take in visual as well as textual qualities, moving their attention from visual to textual and then back again. The ways in which these two forms of perception shift, link and intertwine are key to the haiku. They encourage and allow for momentary awareness of what is not there, not stated, in the background, significantly influencing a reader’s attitudes and responses to the haiku.

Understanding these modes in more detail offers illumination of the workings and power the haiku carries. This can be done by a study of shifts in balance between visual and textual modes of perception, that is, moments of mu, created by variations in lineation (one-liner, two-, three- or four-liner haiku), orientation (vertical, horizontal, diagonal, in columns, in stripes, etc.), shape (a square block of text, a jumping frog, a skein of geese, a never-ending circle, a spiral, etc.), orthography (the Latin (English) alphabet, kanji, hiragana, katakana etc.), and related issues of directional and non-directional reading.

These variations are connected to the language in which the haiku is written. The choice of orthography – whether linear Latin alphabet or more visually complex Japanese writing systems – affects every visual aspect: lineation, directional reading, orientation, shape and layout. Writers are keenly aware of this, as indicated by Cor van den Heuvel’s observation, ‘[t]he most common argument for [English] one-liners is that the Japanese write haiku in one vertical line or column and therefore we should write in one line also, but of course horizontally in the Western style.’¹⁰ A consideration of reading and viewing modes therefore needs to begin with a comparison of their application in relation to the textually-biased Latin alphabet and the more visually-biased Japanese writing systems.

English haiku are written in the Latin alphabet. This alphabet is designed to be read. In other words, it is not ideogrammatic. Written English text requires the reader to spell out letters to make up words and then move forward through a sentence. The semantic sense progresses in a linear fashion. For a beginner reader, as the eye travels over the text, the words are constructed letter by letter. For a more advanced reader, parallel letter recognition is used: readers ‘recognize a word’s component letters, then use that visual information to recognize a word’.¹¹ Even when known word shapes are completely destroyed by scrambling the letters, an experienced reader can still read the text without much problem, as is evident here: ‘Aoccdrnig to a rscheearch at Cmabrigde Uinervtisy . . .’.

Thus, reading text written in the Latin alphabet involves linear and cumulative movement. However, this movement is not smooth. The physical act of reading English text consists of successive jerky leaps or saccades in eye movements over a page, punctuated by fixations upon handfuls of letters at a time, usually just to the left of the middle of a word, and skipping many short and functional words. The saccades are usually forward-moving through the text but about ten per cent of them are regressive. This form of eye movement is specific to the reading mode, as neurophysologist R. H. S. Carpenter emphasises:

the saccade itself is a masterpiece of control engineering, in which the eye – a distinctly sluggish member that, left to its own devices, likes to take up to a second to settle down in a new position – is smartly accelerated and decelerated to bring the fovea to rest on its goal with a time-course that frequently lasts little more than 20 msec . . .¹²

However poems, and haiku, are not only read. They are also viewed. Andrew Michael Roberts, Jane Stabler, Martin Fischer and Lisa Otty’s fascinating study, ‘Space and Pattern in Linear and Postlinear Poetry: Empirical and Theoretical Approaches’, examines ‘the interaction of modes of reading and viewing specific to the processing of textual syntax and visual pattern’.¹³ Their analysis of ‘the results of experiments using eye-tracking, manipulations of text, memory tests and readers’ recorded responses and interpretations’ shows that ‘radical textual dislocation can override the rule that we do not fixate on blank space when reading/viewing a text.’¹⁴

Such work aims not to present a finished piece but an incomplete perception:

How pleasant —
just once not to see 
Fuji through mist.

In the words of fourteenth-century writer on Japanese aesthetics, the Buddhist priest Kenkō, what is prized is not blossom in full bloom, but ‘twigs which bear no blossoms as yet and a garden strewn with withered petals’.¹⁶ This relates to the aesthetics of mu.

Roberts et al’s observations map very well on to the experience of reading and viewing Japanese writing systems. As Carpenter has noted, the movements of the eye in reading mode are specific, accelerated and relatively uniform in direction. However, although the letter-by-letter or parallel letter recognition patterns that occur when reading the Latin alphabet are easily applied to the reading of syllabic kana, studies of the recognition and analysis of kana and kanji (syllabic and ideogrammatic) characters indicate a different trajectory for kanji, suggesting

that different processes are involved in the recognition and analysis of kana and kanji (syllabic and ideogrammatic) characters, with certain patients able to read and understand Kanji texts relatively fluently, while they are incapable of reading single words in Kana.¹⁷

In other words, the features possessed by the ideogrammatic kanji encourage the eye to view, while the syllabic kana, like the Latin alphabet, require the eye to read.

One practical reason for the emphasis on the reading mode in the Latin alphabet is that its reach of potential visual effects is minimal. Latin letters are simple. They do not consist of complex layers of radicals. They are also few. There are 26 letters to play with rather than 2,000 to 10,000 kanji or more. Such a restricted number results in far fewer visual effects and far more repetition of them. Consequently, aside from the occasional cases where typographical considerations are foregrounded in the presentation of a haiku, the visual qualities of Latin letters, such as the round fullness of an O or the jagged edges of a W, are rarely considered by writer or reader. Instead, the reading mode dominates.

In contrast, the Japanese writing systems, with their vast collection of ideogrammatic and syllabic characters, while still drawing on the reading mode, also make use of the viewing mode. In the hiragana and katakana writing systems, words are read, constructed syllable by syllable in a manner similar to that required by the Latin alphabet. However, the ideogrammatic Japanese kanji characters are seen or viewed as much as they are read. Taking a kanji in is a much more in-depth engagement than the reading of a Latin letter. A single kanji may depict one or more of several words or concepts and hold a number of visual, aural and semantic meanings and associations. It may also consist of a number of simpler kanji or radicals that themselves contain ideogrammatic, semantic and aural variations of meaning, sound and association. In addition, the reading of a kanji may in part depend on other kanji that juxtapose it on the page and alter or affect its meaning. The ambiguities that result are intensified by the fact that written Japanese does not include word divisions, and uses punctuation far more rarely than in written English.

Viewers|readers of kanji therefore have to unpick, identify and select from several associations and visual, aural and semantic readings the ones that seem most appropriate to the context, mood and semantic sense of the text in which the kanji is situated. This is not all. Such a selection process is not conclusive, since the alternative but now dismissed readings and viewings of the kanji still hover in the background, adding to the complexities that surround and are contained within the text visible on the page. The kanji, therefore, like the haiku, is not just the sum of its parts, but includes within it references that reach beyond itself, indicators of that ‘whole large sky’ to which Noguchi referred.

Confronted with kanji, the reader remains momentarily in one place, looking in depth at the kanji’s elements, and its relation to the characters between which it has been placed, before shifting onto the next. This becomes very evident when watching people write in one or other writing system. With the Latin alphabet, the pen keeps moving smoothly forward across the page, looping letters together into one relatively seamless flow. With kanji, the pen repeatedly returns to the same spot, adding up to as many as seventeen strokes (in the case, for example, of the character for a flute yaku or 龠) to create a particular character, before it progresses to the next symbol. Similarly, when reading, in contrast to the jerky linear cumulative effects associated with the reading mode, effects that predominate when using the Latin alphabet, viewing kanji requires a more holistic encompassing eloquent immediacy and circularity of vision.

Many writers, poets and artists affirm the impact of the viewing mode. Wole Soyinka sees ‘images [as] far more eloquent in any case than verbiage’.¹⁸ This echoes Picasso’s words to Antonina Vallentin, ‘A painting, for me, speaks by itself, what good does it do, after all, to impart explanations? A painter has only one language, as for the rest . . .’¹⁹ Similarly, Clyfford Still writes, ‘My paintings have no titles because I do not wish them to be considered illustrations or pictorial puzzles. If made properly visible they speak for themselves.’²⁰ As for John Berger, relative to reading, viewing is almost instant: ‘Seeing comes before words. The child looks and recognises before it can speak.’²¹ Berger also emphasises the holistic qualities of vision, ‘continually active, continually moving, continually holding things in a circle around itself, constituting what is present to us as we are.’²²

The effect of all this is a rich complexity of possible readings and viewings. This is ‘calligraphic doubling’, as Michel Foucault terms it: ‘the calligram [or kanji] that says things twice [. . . ,] the calligram that shuffles what it says over what it shows to hide them from each other’.²³ However, kanji often have kana appended to them. ‘青い’, or ‘blue|green’, is pronounced ‘aoi’, with the kanji ‘青’ sounding as ‘ao’ and the hiragana ‘い’ indicating its status as adjective. This mix of syllabograms and ideograms requires subtle shifts in mode. The reader has to view and also to read or ‘spell out’. Thus, Japanese writing systems require a dual focus on text and on image, the reader|viewer switching constantly between modes. The subtle richness that results is well captured by Sabine Gross, who observes how the use of both viewing and reading modes has dramatic effects upon the reception of a text. Any semantic meaning, for the duration of the viewing mode, is cast aside: ‘as soon as letters and words are perceived as images – and thus decoded as iconic signs – they disappear as symbolic signifiers, at least for the period in which they are scanned as and processed as images‘.²⁴

It is important to note that the break or cut occurring between kanji and kana also acts as a connector. This evokes the shift and link of renga writing. A micro-version is detectable in the haiku’s use of a kireji or cutting word, marking a shift in subject, mood, or content, an invitation to discover a link. It signals a moment of mu, occurring as the reader|viewer determines on and then puts in motion the switch from one mode of perception to another. For this moment, the reader is neither reader or viewer, but exists in space, between modes. This moment also marks a break in logic, as indicated by the British Haiku Society,

These words are like ‘verbalised punctuation’, generally with a hint of emotion or of attitude. “Haiku is grasped with all five senses, not by logic . . . in order to jump over the gap between logic and the senses.”²⁵

The break is also a link, as ‘jump’ and ‘gap’ imply, connecting closely to the aesthetic core of haiku – leaving ‘a few things unsaid’. Donald Keene reiterates this in his image of the leaping spark of the cutting word:

there should be the two electric poles between which the spark will leap for the haiku to be effective; otherwise it is no more than a brief statement.²⁶

The mix of modes encoded within Japanese writing systems automatically requires such a leap. This is not so obviously the case with haiku written in the more linear Latin alphabet. However, the Latin alphabet consists, outside of Braille, of visual signs and so in the process of reading it, viewing also has a crucial part to play.

This is evident when considering other visual variables in haiku, such as page orientation. If a haiku is written in traditional Japanese-style, without line-breaks in one vertical line down the page, this works against the intense focus on particular kanji and seems to encourage the linear cumulative effect of the reading mode. A single vertical line can be taken in at a glance, like Berger’s child looking and recognising ‘before it can speak’. However, in the traditional English three-liner, a haiku forms a small, compact, or compressed, square-ish shape on the page, with a three-part grouping of short syllabic lines that are easily encompassed in a glance. In addition, one of the line-endings, accompanied often by punctuation – usually a dash, often acts as a visual signal of the two-part division of the haiku’s content. Such a small footprint on the page reduces the number of saccades and fixations needed before the haiku can be said to have been ‘read’. A single fixation of the eye is able to take in both a group of letters and a peripheral sense of the shape and structure of the haiku, for, as Larson has noted, a single fixation includes different kinds of information-gathering:

Closest to the fixation point is where word recognition takes place. [. . .] The next zone extends a few letters past the word recognition zone, and readers gather preliminary information about the next letters in this zone. The final zone extends out to 15 letters past the fixation point. Information gathered out this far is used to identify the length of upcoming words and to identify the best location for the next fixation point.²⁷

This is indisputable in the case of one-word haiku, such as Cor van den Heuvel’s ‘tundra’,²⁸ where fixating on the word ‘tundra’ necessarily involves the whole poem, shape and structure included. In addition, the reader’s eye will be taking in the space that surrounds haiku and makes up part of its ‘point’. In other words, it is possible to view such a haiku at the same time as reading it rather than viewing it before (or after) it is read. For haiku like these, the viewing mode occurs simultaneously with the reading mode, as part of that mode. Alternation and mutuality of mode become one. It is possible to argue that such experiences of English three-line haiku lead to a more comprehensive integrated and finished effect than that achieved by the Japanese one-liner, although one could then go on to question how close this then remains to the haiku aesthetic.

Circle haiku, or cirku, raise more issues. With cirku, the emphasis is placed on viewing. As Stephen Gill makes clear, this is an English form:

I coined the term ‘cirku’ myself many years ago (ca. 1996). It was later picked up by Bill Higginson as if it might be an existing Japanese term, which of course it certainly isn’t, as they do not have the form and it is half Eng. in expression (cir- from circle). I do not claim to have invented the form as I am pretty sure someone else must have experimented with such things already, but in our Hailstone publications over the years we have always included them at my behest, so we have helped to establish it as a form these past 20 years. You will have read that in my own definition, you should be able to start or stop any line end within the circle. In my form, there is almost always lineation within the endlessly repeating circular form. Thus, at any of those line-heads (after the gaps) you are free to enter the poem.²⁹

Usually with cirku, the circle is seen before reading begins. In other words, the visual impact is so extreme as to exaggerate and extend the separation between reading and viewing mode. A similar phenomenon however can be created with any one-page poem that the reader approaches from physical distance, since in such cases the shape of the poem, including awareness of line length, line number and stanza breaks, will be perceived when individual letters are too far away or too small to read. Mario Petrucci has developed some fascinating exercises that reveal how such impressions can inform the reader in significant ways about the poem they have not yet read:

I believe the eye (as opposed to the ear) is far more interested in space than it is in metre or rhythm. Behind and beyond the sound of that inner voice we use when we read, our eyes are busy registering the shapes made by poems on the page – the promontories and vacancies of words, lines and stanzas. That’s why our eyes can instantly tell apart (most) poems from (most) prose, without having to read a word of either.³⁰

In his teaching, Petrucci presents poems on transparencies, deliberately projecting them out of focus so their shapes are visible but the text is illegible. Such demonstrations reveal how much information can be carried by shape, or ‘spatial form’ as Petrucci calls it. It is possible to see ‘a sense of uncertainty, or raggedness, in some poems. Other [readers of other poem]s pick up on a ‘martial’ or obsessive quality, usually if there are ordered ranks of lines or stanzas.’³¹

One-page poems are not usually approached in this way. However, with cirku it is unavoidable. The primacy of the viewing mode is enforced by the haiku’s relatively unusual shape and the difficult orientation of its letters. These, together with the framed circle and framing space, are easily viewed. In contrast, it requires unusual effort actually to read the text. Some letters may be written backwards or upside down and demand physical exertion on the part of the reader before they are deciphered, turning the page or head. Additionally, constituting the letters into words is a more laborious process since word divisions are usually elided (thus, incidentally, replicating more closely the experience of reading in Japanese which also dispenses with word divisions, and, often, punctuation).

Petrucci’s exercise can be applied to both Japanese and English haiku. This confirms that viewing and reading modes are used in both sets of writing systems. However, as noted earlier, due to the characteristics of the writing systems, there is a bias towards the reading mode in haiku written in English, albeit a bias that is dependent on visual variables such as size, form and layout.

Evident in the language used to describe viewing and reading modes is a shared emphasis on movement and trajectory: saccades, fixations, immediacy, circularity, linearity. To understand further the differences between these modes therefore and also the differences of effect that a change in bias between modes might have on a haiku, it is apposite to consider jo ha kyū (序破急) – the Japanese aesthetics of modulation and movement, principles which map well on to the haiku and onto a consideration of reading and viewing modes.

Although primarily applied to theatrical and temporal arts today, the principle of jo ha kyū used to relate to literature and specifically to the shift and link movement of renga, with, around 1356, the poet Yoshimoto Nijō establishing jo ha kyū as the sequential pattern in renga in his compilation of the first imperial anthology of renga, Tsukibashū (菟玖波集). Lucas amplifies this, thus:

Musical terminology was adopted to describe the development in tempo of the work, with the renga being divided into three phases, jo-ha-kyū (introduction, development, finale) corresponding, in the case of a 36-link piece, to the first 6 links, the middle 24 and the final 6. The opening section (ju) was expected to be sedate, without startling images or displays of virtuosity. The middle section (ha) was allowed to be more dynamic, with extravagant, experimental or eccentric linking being acceptable and a wide variety of images being expected. The final section (kyū) should flow smoothly, with more attention paid to the linking than to the excellence of individual stanzas.
One aspect of renga aesthetics which it is important to recognise is that a sequence was expected to be uneven. The intensity of the linking could vary from close to distant, sometimes obvious, sometimes requiring an effort of imagination to perceive.³²

Probably drawing on Nijō’s work, fourteenth-century exponent and writer of traditional Japanese Nō theatre Motokiyo Zeami then adopted and analysed jo ha kyū for Nō theatre, presenting it in his first treatise, Fūshikaden (風姿花伝, ‘The Transmission of the Flower Through (a Mastery of) the Forms’ or ‘Teaching of Style and the Flower’), as a universal concept that could be applied to the patterns of movement of all things.

As Kunio Komparu writes in The Noh Theater: Principles and Perspectives, ‘Jo means beginning or preparation, ha means breaking, and kyū means rapid or urgent’.³³ The movement of jo ha kyū involves an initial slow pace, giving an impression of expectation and exploration, which gradually speeds up, and then swiftly climaxes and terminates ‘like a flourish that dissolves as soon as it is created. [. . .] The overall effect is of an undulating wave of sound accompanying the flowing movements’.³⁴ The movement does not stop. It changes tempo, mood, and possibly direction, as it passes through jo to ha to kyū, but it remains an integral unit of movement.

Jo ha kyū is today linked with temporally-based media, such as theatre, dance and music performance, since its rhythmic premise seems supremely suited to performance and focus on movement through time in an unbroken if changing linear fashion: from start, to middle, to end. Jo ha kyū is also applicable to reception of that performance by an audience and to the prior act of composition of a work. Jo ha kyū applies to a piece on every level. It applies to performances of whole plays, individual sections of a play, a speech, a cadence, a phrase, a line (which in the poetic sections of a play consists of 5 or 7 syllables, a familiar unit in haiku), or even an individual note, word, step, breath. It is a simple progression to move from the micro-movements of the start, progression and termination of each breath of the performer to the micro-shifts in modulation that a reader|viewer|listener experiences in a haiku, whether read aloud, silently, or viewed upon the page. Zeami reflects such awareness in words that strongly resonate with Bashō’s ‘learn of the pine from the pine; learn of the bamboo from the bamboo’,³⁵ noting:

All things in the universe, good or evil, large or small, animate or inanimate, have each the rhythm of jo, ha, and kyū, [. . .] It is observed even in such things as a bird’s singing or an insect’s chirping.³⁶

A close analysis of a haiku by O Mabson Southard in American English shows how jo ha kyū might work in English language haiku,

Trailing the canoe
     out into the windswept lake —
          a pair of muskrats

Southard’s haiku starts (jo) with the movement of the canoe, which is deepened, widened or scattered (ha) by a number of individual words, ‘out’, ‘into’, ‘windswept’ and ‘lake’, and even perhaps by the ‘–’. Then, with a flourish, the whole impression is transformed (and in a totally unexpected direction) by ‘a pair of muskrats’ (kyū), although perhaps ‘flourish’ is not quite the right word for a muskrat. The haiku’s visual layout and typography instruct the reading|viewing of it. The opening word ‘Trailing’ is initialised and the first line is indented further left than any other line, as if to mark the place at which the reading should start. The poem is visually divided into three separated lines that suggest the tripartite movement of jo ha kyū. This tripartite structure is further emphasised by the successive indentations of the lines. The indentations also reiterate the order in which the lines should be read, from first line to second to last. The dash at the end of the second line indicates a break or change in subject that could be read as signalling the kyū.

However, while in one sense all these features work well with the principle of jo ha kyū, in another sense they work against it. Although the eye is encouraged to move down the page in a way that might be seen to replicate jo ha kyū, the eye is also encouraged to stop and give emphasis to the first capitalised word, then to flow horizontally across line one and lastly to break at the end of that line, before repeating this shift from pause to movement to pause in lines two and three. These successive breaks would seem to interfere with the cumulative progression that marks jo ha kyū, where each movement flows seamlessly into the next. Equally, such a seamless flow of movement appears at odds with the jerky saccades and fixations that make up the reading mode, which in fact work better with the repeated stops and returns to the margin that a three-line haiku demands.

The one-line format, in both English and Japanese, avoids the jerkiness such line breaks creates, and so is perhaps more in tune with the spirit of jo ha kyū. This is evident when considering British writer Fred Schofield’s one-liner,

quick through the brushwood shadow of a wren

Here the long stretch of unbroken words set out in one line push against any semantic breaks within the haiku that the readers might create as they slowly take in the possible semantic meaning(s). Such a movement is in keeping with a jo ha kyū-like rhythm. The pull-and-drag effect that results is very powerful, and it is an effect that is all the sharper if one considers the minute movements of the reading eye, fixing on one part of a word and then jumping to the next — like a wren’s quick almost unseen movements and/or short bursts of flight.

This becomes even more evident if one considers traditional vertically-orientated one-line Japanese haiku. An example by Kobayashi Issa makes this clear:

		 大
		 根
		 引
		 大
		 根
		 で
		 道
		 を
		 教
		 へ
		 け
		 り 

The nuances of Issa’s Japanese are evident in these four different renderings of the haiku in romāji; literal English translation; and two differing literary translations into English, by David G. Lanoue (American) and by Judy Kendall (British) and Wilhelm Wetterhoff (Swedish-speaking Finn):

dai\kon*hiki*dai\kon*de*michi\o*oshi\e(\)ke\ri

The above is a romāji rendering of Issa’s Japanese. The ‘\’ links possible related kanji (Chinese characters or character) and/or hiragana) and the ‘*’ marks the word divisions a Western reader would normally insert.

daikon / pulled-up / daikon / with / way / point

The above is a literal English translation of Issa’s Japanese. The ‘/’ marks the end of each kanji, composite kanji or collection of hiragana and kanji that make up one unit of meaning.

with a just-yanked radish pointing the way 
(David Lanoue translation)
he pulls up a daikon points the way
(Kendall|Wetterhoff translation)

In Japanese, the original Issa poem flows without a break and yet at the same time changes speed, tone and humour. It moves from the very physical and comical image of the large daikon (Japanese radish), pulled up to act as a physical signpost (jo), to a more open plane, a more philosophical indication of a way forward or a way on in life perhaps (ha), reaching a final ‘point’ (kyū). However, this emphasis on progression through the haiku fails to take account of the very deep rich contribution that the visual attributes of Japanese kanji make, paradoxically evoking the fixation-saccade movement of the reading eye. A powerful effect is created in the Japanese by the repeated strong simple open character of ‘大’. It stands out in the poem, spacious, embracing with its widening legs and extended cross-bar the white background of the page on which it is positioned. On its own, ‘大’ means ‘large’. In Issa’s haiku, ‘大’, itself a visually-expansive kanji, makes up the first part of the character for daikon, ‘大根’ (the literal reading for this is ‘large root’). This visual and semantic emphasis on size is repeated as the same kanji appears again – coming therefore both before and after the character for ‘pull’ or ‘tug’ (‘引’ or ‘hiki’). ‘引’ or ‘hiki’ is another visually expansive kanji, the open strokes of which broaden out the space contained within it. The ‘k’ of ‘hiki’ echoes the ‘k’ in the two ‘daikon’ between which it sits. The effect is to make the reader feel stuck, with the daikon, and with that act of pulling,

		 大
		 根
		 引
		 大
		 根

Interestingly, neither of the literary English translations given here takes on Issa’s device of repetition. Perhaps, in the Japanese, which already encourages the reader to pause, digging and pulling, at each kanji, a repeated character is necessary to transform that movement into a sense of being stuck. In the Kendall|Wetterhoff translation, however, that sense of hindered movement is expressed by using the one word ‘daikon’ as both object and subject, rather in the manner of a Japanese kakekotoba or pivot-word, obliging the reader therefore to hold both meanings at once). In Kendall and Wetterhoff’s version, the one word ‘daikon’ does double duty as passive object and active agent, causing a pause in the flow of the reading, a pause that parallels the effect of Issa’s repetition:

he pulls up a daikon [object] 

and

a daikon [subject] points the way

In English, any flow in reading is compromised by word divisions, which Japanese does not have. The pun effect or kakekotoba in this example of the word ‘daikon’ tries to remedy this but the effect is heavy, leaving the reader and the sense slightly stuck in the middle of the poem instead of flowing on immediately to ‘points the way’. The first three words ‘he pulls up’ could be interpreted as jo, but ‘a daikon’ referring both forwards and backwards makes a strange ha, and ‘points the way’ besides starting rather jerkily because of the repeated focus on ‘daikon’ also, then, represents an unusually long flowing ending for kyū.

In Japanese, kakekotoba are so much more common that it is necessary to repeat the character for ‘daikon’ in order to make a sufficient impact. The effect in Kendall and Wetterhoff’s English version of a deliberate heavy pause on the double-loaded daikon takes the reader far away from a jo ha kyū-like progressive momentum, held back in the middle of the poem even as it tries to point forward. Additionally, reaching the end of the poem does not take the reader any further down the page.

In Lanoue’s English version, he chooses instead to emphasise the effort involved in pulling up the daikon in the awkwardness of ‘a just-yanked radish’. He also creates this effect in his tangled repetition of sounds: the ‘j’ and ‘st’ of ‘just’ and the ‘a’ and ‘d’ of ‘yanked’ entwining closely with the ‘a’, ‘d’ and ‘sh’ of ‘radish’. This version brings the haiku closer to an approximation of jo ha kyū, the sounds and words link back to previous iterations but do not hinder that forward sweep as much as the heavier pun in the Kendall|Wetterhoff version.

The Japanese also works effectively with a sense of dual trajectory. A change from the slow start (jo) created by the appearance of two daikons is heralded by で’ (‘de’ or ‘with). The placing of ‘で’ or ‘de’ directly after the second ‘daikon’ turns that just-pulled-up daikon into an instrument of use, suggesting therefore a further action, a way forward. However, the ‘d’ of ‘de’ also reinforces the sense of difficulty in movement by connecting ‘de’ strongly with the ‘d’ in the two previous daikon, as if the ‘de’, that instrument of movement, is itself being pulled back. Nevertheless, the promise of progression is fulfilled as the haiku continues, moving with a flow and sense of release into the pointing of the way onward: ‘michi o oshie keri’ – a rush (kyū) of softer and more open sounds that end with the open, simply-drawn (and simple to read) hiragana syllabograms which seem to offer a further expanse of space stretching right down the length of the page:

		 へ
		 け
		 り

The tripartite structure presented by jo ha kyū seems particularly apt, and more fully utilised semantically, visually and orthographically, when reading a vertical Japanese one-liner. The eye can travel freely and unimpeded through the whole poem, albeit with saccades. The poem is offered without the visual breaks that would occur with the line endings in a three-line format, and is also, usually, offered without the pauses that the English use of punctuation and word divisions would signal. The result is that the eye, when reading a Japanese vertical one-liner, travels down the line unhindered by pause marks in the text, more nearly approaching the smooth cumulative progression through a work that is advocated by jo ha kyū.

This smooth movement parallels the unbroken stylised movements of a Nō or Kyogen actor sliding his feet across the stage (a style known as suriashi), as his voice moves from one part of the play, section, speech, line, to the next. Such foot movements are taught to the initiate as an embodiment of the principle of jo ha kyū. Komparu notes,

We often hear it said that Noh is an art of walking […] The pure white tabi socks gliding along and reflected in the stage, moving now quickly, now slowly, give expression to the character and will of the actor. We might even say that one can experience an entire Noh play watching only the actors’ feet.³⁹

Just as the actor’s foot never breaks contact with the ground, supposedly to avoid the distracting vibrations of the stage floor, so the audience never misses a word or a beat. The smooth progression of movement and sound is unbroken and uninterrupted by external interference, and also acts as a connector between the shifts in movement, pace and speed that do occur. The application of this to English haiku can be seen in Schofield’s ‘quick through the brushwood’. A horizontally-oriented haiku like Schofield’s achieves something of the same effect in terms of eye-movement. This is enabled by the lack of conventional markers of pause or break in English haiku. There are word divisions but there is no punctuation, line breaks or capital letters. However, given the page orientation, portrait rather than landscape, the sweep is brief, covering the shorter width of the page not its length. Progress is less obvious, less sure than in the vertical orientation. Of course on a micro-level the saccades that occur when reading contradict this impression of an apparently smooth sweep, but then, in the same way, one can argue that the tiny muscular contractions that occur when sliding the foot forward are at odds with the gliding motion that is achieved.

When the three-line format is employed in English, the application of jo ha kyū to haiku appears less straightforward. A reader is predisposed to insert a pause at a particular point or points because of visual indicators of the haiku’s shape, structure and divisions between lines and words by means of gaps, punctuation and change in letter case. Also, as noted earlier, the three-liner’s compact footprint on the page allows it to be as easily viewed as read. This too makes the linear modulating approach on which jo ha kyū depends less easy to apply.

The physical act of reading English haiku letter by letter and word by word in order to make up the poem, however, is closer to jo ha kyū than a Japanese reading | viewing experience might be. This is because Japanese haiku almost always involve kanji, which require the reader to view and take in individual concepts or characters rather than individual letters, before leaping to the next. The reader|viewer is using ‘continually active, continually moving’ vision’, looking in depth at particular kanji, and then shifting to the next.⁴⁰ Thus, the complicated rich experience of reading kanji consists of more than the sweep in a linear fashion through the poem suggested by jo ha kyū. It involves Foucault ‘calligraphic doubling’ effect. What is read is also seen. It is necessary to pause on kanji to select out of the multiple possible interpretations the ones that are apposite in the context of the emerging haiku. This is likely to involve successive adjustments and re-adjustments of the selections adopted, as the haiku is gradually taken in, and dismissed readings may still have an influence over the haiku’s general effect. The related series of eye movements do not suggest a linear cumulative sweep. Instead, the focus, or deep fixation, on particular kanji, identifying, viewing and selecting possible readings, and the subsequent leaps to other kanji mirror the saccade-fixation-saccade of the reading mode.

Further awareness of how the visual complexity of kanji contributes to the viewing|reading and interpretation of a haiku and to the movement of the eye through it, is evident in an analysis of Santōka’s famous haiku, ‘分け入っても分け入っても青い山’. To aid comprehension, here are several versions, consisting of a romāji transcription, a literal English translation and two literary translations of the haiku in English:

Written in romāji – wa\ke\i\tte*mo*wa\ke\i\tte*mo*ao\i*yama 

(‘\’ indicates possible related groups of kanji and|or hiragana, ‘*’ shows the word divisions a Western reader would probably insert);

Literal English translation – push one’s way into / despite / push one’s way into / despite|even|also|already / blue|green / mountain|hill

(‘ /’ marks the end of each kanji, composite kanji or collection of hiragana and kanji that make up one unit of meaning, ‘|’ indicates alternative readings for this word/character);

Getting further and further
Into the mountains,
But still deep blue mountains.

or

Going deeper 
And still deeper —
The green mountains.

In ‘分け入っても分け入っても青い山’, Santōka uses the kanji ‘青’ or ‘ao’, referring to the colour green or the colour blue. In the English translations, only one of the two colours remains. In addition, ‘青’ includes within its visual structure a radical that denotes another meaning. If placed on its own, the radical in the bottom half of ‘青’ can be read as ‘month’ or ‘moon’ (‘月’). No one, apart from a very early learner, would consciously think of ‘month’ when they see ‘月’, just as in English, the visual qualities of the Latin letters are unlikely to occur to readers, unless specifically highlighted in some way. However, the process of reading|viewing kanji involves the consideration of multiple aural and associative meanings. In Nelson’s JAPANESE-ENGLISH CHARACTER DICTIONARY, for example, aural readings of ‘青’ include not only ‘ao’ but ‘sei’ and ‘shō’. Possible semantic meanings include ‘blue’, ‘green’, ‘green light’, ‘pale’, ‘unripe’, ‘inexperienced’, ‘new’, ‘immature’. In this particular case, these meanings neatly coincide with some of the English metaphorical associations with ‘green’, but that is by no means always so. The top half of ‘青’ can also work as a radical if placed at the left or the bottom of the kanji in which it is included, and among others its semantic meanings include ‘jewel’, ‘jade’, ‘ball’, ‘tool’, ‘testicles’, ‘beautiful’, ‘round’.⁴³ Similarly, if ‘青’ is combined with a second character, a range of semantic meanings is possible, including ‘frost’, ‘cemetery’, ‘harmless snake’, ‘history’, ‘rice fields’, ‘grub’, ‘gray horse’, ‘cyanide’.⁴⁴

The complex readings of even the simplest kanji result in an array of dismissed meanings that sometimes remain visibly and multiply present, ready to come to the fore if the context allows. These readings contribute to the kind of movement or trajectory within the text. In the Issa haiku, the character ‘根’ or ‘root’ in ‘daikon’ (‘大根’) contributes to the overall effect of the haiku, since it emphasises difficulty of movement – the effort involved in pulling up roots in order to travel. However, the presence of ‘root’ in the poem is more visual than aural and an awareness of it requires consideration of semantic and visual roots. The reader has to delve into the kanji to pull it out, halting therefore that progressive jo ha kyū movement forward. The result is that the very act of discovery of the presence of the ‘root’ radical in the haiku emphasises the difficulty of movement that the haiku expresses.

With the complicated scenarios now emerging in terms of reading and viewing Japanese and English haiku, it is necessary to look further back for an aesthetic that reflects these shifts in movement, mode and trajectory in haiku appreciation, an appreciation that works with depth as well as with linear flow. Bashō intimated to his disciple Kyokusui that the ancestry of jo ha kyū offers some answers in his own awareness of the connections between Japanese and ancient Chinese poets:

seek the distant bones of Fujiwara Teika, follow the sinews of Saigyo, cleanse the intestines of Po Chii-i, and leap into the breast of Tu Fu.⁴⁵

Jo ha kyū can be traced to ancient Japanese imperial court dance and music, bugaku and gagaku, and gagaku takes its origins from seventh-century Chinese music. Jo ha kyū is therefore believed to have been modified from the Chinese musical principles in this period, as exemplified in the Yanyue (燕樂or court banquet music) of the Tang Dynasty (618–907 AD). These principles seem to respond more closely to the experience of in-depth readings of kanji. They include sanxu, 散序 – a first section of slow free rhythm; zhongxu, 中序 – a middle metered section usually sung at a slower speed; and po, 破 – an irregular but metered dance rhythm that gradually speeds up as it continues, then slows down, before finishing at speed.⁴⁶ The first two terms, sanxu and zhongxu correspond quite closely with jo and ha. Po however does not simply equate with kyū, but offers changes in speed and possibly changes in direction. This suggests modulations in comprehension, modulations that include more than one possibility – perhaps reflecting the movements the mind has to go through when taking in the effects of, for example, the introduction of ‘de’ in Issa’s haiku, movements that require the mind and the eye to look backwards as well as forwards. Po could also be a way of expressing the need to delve momentarily into a particular character to investigate its roots, its associations, its other possible meanings, as when dissecting the composite kanji for ‘daikon’ into its two parts of ‘large’ and ‘root’.

While sanxu, zhongxu and po fit with the movements that occur in the appreciation of haiku, they relate not to literature but to music, albeit music that is accompanied by song. However, the ancient Chinese poetic principles of fu bi xing, developed before 1000 BC, are very pertinent when considering the curious sometimes contradictory effects of ‘calligraphic doubling’ in Japanese poetry.

There is controversy as to the exact meaning of fu bi xing. L. Kip Wheeler defines it thus, the lightning image resonating with Donald Keene’s spark between electric poles:

Fu refers to a straightforward narrative with a beginning, middle, and conclusion, that stands by itself. Bi, literally “against,” implies a comparison or contrast, placing two things side by side. When one takes two different fu, and places them together, the two create a bi. This results in xing, a mental stimulation or “lightning” that pervades the mind of the reader, bringing new insight or awareness into the nature of the individual fu that compose the poem. Confucius stated that this xing is the purpose of poetry, that the point of a poem was to make the mind contemplate its subject deeply.⁴⁷

In a further consideration of fu bi xing, Ming Dong Gu remarks that ‘bi is a static, single metaphor, while xing is a dynamic, total, totalizing metaphor’.⁴⁸ Other scholars also reiterate the power of xing, as reported by Zehou Li:

among later critics we find such remarks as, ‘bi is apparent, xing is hidden’; ‘if the words have finished but the meaning lingers, that is xing’; ‘metaphorical meaning, although it is precise, is shallow, while imagistic meaning [xing], although broad, has a lasting flavor’; and, ‘As for imagistic association [xing], what one sees is one thing, what is meant is another; it cannot be pinpointed by analogy, nor thought out by reason.⁴⁹

The changes in direction, orientation, dimension and plane introduced by xing, as intimated by Ming Dong Gu and Zehou Li call up Lucas’s definition of the haiku’s poetic spell. Lucas noted that the contemporary haiku movement

looks as if it has reached something of a plateau. This plateau is a position of conformity, complacency and mere competence. And the pressures towards conformity are acute enough to make it difficult to remain true to your own original inspirations, poetic preferences and little awkwardnesses that resist hammering into shape.⁵⁰

Deploring an ‘internationally accepted formula’, Lucas remarked that ‘poets writing original haiku in English have focused on what is said and paid relatively little attention to how it is said’. Here, he is echoing Ueda’s observation that ‘an inexpert poet often ends up composing a descriptive poem so plain and trite as to evoke no feeling at all’.⁵¹ Lucas offers the concept of the poetic spell as ‘an apparently different guiding aesthetic’ which prioritises ‘Circular / Fluid’, ‘Ambiguous’, ‘Expansive / Reflective’ qualities. To these, Lucas adds ‘communication of feeling via objective description – a poem should suggest more than appears on its surface’.⁵² His aesthetic is only ‘apparently different’; these qualities match Bashō’s critical values of ‘austere beauty; karumi (literally ‘lightness’ [. . .]); yoja (‘surplus meaning’) and ‘soul’’. Such qualities also hark back to xing, seen as ‘bringing new insight or awareness [. . .] to make the mind contemplate its subject deeply’; ‘dynamic, total, totalizing’; ‘hidden’; ‘broad’ with ‘a lasting flavor’, so that ‘what one sees is one thing, what is meant is another; it cannot be pinpointed by analogy, nor thought out by reason’.⁵³

In the aesthetics so far mentioned, movement is key – shared by fu bi xing, jo ha kyū, and, indeed, sanxu, zhongxu and po. For jo ha kyū, the movement is forwards, or downwards, through the text|drama|music. For fu bi xing, there is less emphasis on progression. The movement is momentary, holistic, almost atemporal. If jo ha kyū works on the plane of the page or paper, fu bi xing takes the reader|viewer in another direction, in the manner of po – off the paper, to another plane, from a shallower to a broader view as Zhuzi yu lei suggests, away from a temporal progressive pattern of reading towards a more holistic mode of viewing, a mode that seems more akin to the experience of reading|viewing and gathering in the inherent ambiguous richness of Chinese characters.

Jo ha kyū suggests that a reader moves through the text, preferably in a long unbroken line, with the reader appreciating the cuts, twists, changes, links and, indeed, illuminations that the text might successively offer. However, the definitions passed down for fu bi xing imply a movement that lifts the reader|viewer up from the text, zooming out into the air to offer a bird’s eye perspective, a whole in one:

a quiet stretch of river
the gull’s little kick
into flight

If jo ha kyū relates to the reading mode, where one letter or word is placed after another, while fu bi xing relates to the viewing mode, where the visual effects and what is held within a letter, word, character are investigated, then xing is an accurate description of that moment when, having taken in the whole piece (fu), there is a focus on the contrast of two juxtaposing images (bi) and a zing (or xing) of lightning awareness that indicates the haiku has found its mark. Such a process is not dissimilar to Keene’s spark between electric poles, or Lucas’s darker description of poetic spell:

Through the clarity of imagery, feeling emerges: a cold, dark, sharp feeling that is at the opposite pole from sentimental assumptions of what makes a poem, far more alert, far more alive:

sharpening this night of stars distant dogs

Here, two very different moments are linked. The clarity achieved by the horizontal accumulation of imagery allows to rise (vertically) from the depths a ‘cold, dark, sharp feeling’ that brings with it the sense of an extra dimension, an added level of understanding that words alone cannot provide, describe or analyse. Similarly, when viewing one particular Japanese, or Chinese, character, the principles of fu bi xing apply. The viewer|reader identifies the character (fu); takes in the possible contrasts, associations, semantic meanings and context the character holds or is situated in (bi); and then fits them together to make a holistic whole, or, indeed, finds that the character now not only fits what it says but fits more than can possibly be articulated at once or in words at all (xing).

Both jo ha kyū and fu bi xing (as well as sanxu zhongxi po) come into play when considering the effects created by different Japanese writing systems. Visual kanji tend to represent nouns and verbs, feelings, concepts, emotions and things. The one and possibly three syllabic scripts that accompany and/or are appended to them – hiragana, katakana and furigana, spell out groups of syllables that make up particular suffixes, prefixes, connector words, indicators of pronunciation, and so on. These mostly steer the onlooker to a reading rather than a viewing mode, However, indicators of pronunciation in furigana suggest a third listening mode in which the experience of non-linear visual kanji is transformed into a temporal aural linear reading.

The long-standing British Yorks/Lancs Haiku Group meetings provide an example of aural readings of haiku. Displayed on a board, haiku are read aloud by a participant and only then discussed. The haiku tends to be read aloud twice, indicating the cruciality of this act which lifts the previously visually-perceived poem firmly into the temporal sphere. The linear performance-based principles of jo ha kyū are thus necessarily experienced during these live readings, since in a temporal medium one word tends to follow another. Viewing the haiku on the board also allows for the holistic experience of fu bi xing. The viewer|listener|reader successively switches between modes, decoding words as symbolic signifiers and as iconic signs. Whichever mode is predominant, shift is constant.

Shifts between modes occur in all encounters with haiku. Usually, in English haiku, shape, form, structure, semantics and punctuation, while stimulating a viewing mode, primarily enforce a reading mode upon the reader. With Japanese haiku the reverse is true, but the viewing of kanji is qualified, connected and placed in time, location on the page and grammatical position, by the reading of the linear syllabic scripts that accompany it. With furigana, kanji are placed in the air, in an aural reading plucked out of other possibilities. The eye, moving from kanji to hiragana|katakana|furigana to kanji, experiences in rapid succession both jo ha kyū and the apparently diametrically opposed fu bi xing. These modes of viewing|reading may not occur simultaneously but their rapid alternation works with the effects that a haiku creates; association and connection and contrast create a holistic experience in which the sudden arrival of an illuminating zing is satisfyingly deep, rich and, sometimes, very loud:

something primitive; something rare; something essential; [..]. Poetic spells don’t tell us anything, they are something, they exist as objects of fascination in their own right. You can hold them in the light and turn them about and watch each of their facets gleam.⁵⁶

Perhaps in these moments of shift, in space, not in any specific mode, is where the real power of the haiku is experienced.

Works Cited

Bashō, Matsuo. On Love and Barley: Haiku of Bashō, trans. Lucien Stryk. Harmondsworth: Penguin, 1985.

Berger, John. Ways of Seeing. London: BBC/Penguin Books, 1972.

British Haiku Society. ‘another kind of poetry’. Web. www.anotherkindofpoetry.org.uk/anotherkind/?origins. Accessed 5 September 2014.

British Haiku Society. English Haiku: A Composite view https://britishhaikusociety.org.uk/2011/02/english-haiku-a-composite-view/. Accessed 16 October 2015.

Carpenter, R. H. S. (ed.) ‘Recognition of Kanji Signs and Kana Words’. Eye Movements. Macmillan, 1991.

Carpenter, R. H. S. ‘Oculomotor Procrastination’. Eye Movements: Cognition and Visual Perception ed. by Dennis F Fisher, Richard A Monty, John W Senders New Jersey, Lawrence Erlbaum Assoc. 1981.

Foucault, Michel. This is not a Pipe Ceci n’est pas une pipe, transl. by James Harkness. University of California Press, 1983.

HailHaiku. https://hailhaiku.wordpress.com/2008/03/16/whats-a-cirku/ Accessed 21 October 2015.

Gross, Sabine. ‘The Word Turned Image: Reading Pattern Poems’. Poetics Today 18:1, 1997, 16-32.

Gu, Ming Dong. ‘Fu-Bi-Xing: A Metatheory of Poetry-MakingAuthor(s)’. Chinese Literature: Essays, Articles, Reviews (CLEAR), Vol. 19 (Dec., 1997), 1-22.

Higginson, William. The Haiku Handbook. New York: Kodansha International, 1985.

Issa haiku translations: Kendall|Wetterhof. Issa translation, from Judy Kendall and Wilhelm Wetterhof’s unpublished collection of visual renderings of Issa haiku, 2011.

Issa Haiku translations: Lanoue, David G. Issa translation. Web. https://haikuguy.com/issa/index.html. Accessed 16 September 2014.

‘jo-ha-kyu’ Shakuhachizen.com. Web. https://www.shakuhachizen.com/jo-ha-kyu.html. Accessed 10 September 2014.

Kacian, Jim. ‘Looking and Seeing: How Haiga Works’, reprinted in Long Exposure issue 3, 2015 (2002). https://www.haigaonline.com/FINALHAIGATALKVOLII.html.

Keene, Donald. Japanese Literature: An Introduction for Western Readers. New York: Grove Press, 1955, 40-41.

Kendall, Judy. ‘The delightful ambiguities in the translation of Japanese haiku: notes on Santōka’s 分け入っても分け入っても青い山’, Presence 51.

Kenkō. Miscellany of a Japanese Priest, trans. William N. Porter. London: Humphrey Milford, 1914; repr. Tokyo: Tuttle, 1983. https://djm.cc/library/The_Miscellany_of_a_Japanese_Priest_Gusa_Porter.pdf.

Komparu, Kunio. The Noh Theater: Principles and Perspectives. New York: Weatherhill, 1983.

Lanoue, David G. Issa translation. Web. https://haikuguy.com/issa/index.html. Accessed 16 September 2014.

Larson, Kevin. ‘The Science of Word Recognition’ https://www.microsoft.com/typography/ctfonts/WordRecognition.aspx. Accessed 18 February 2014.

Li, Zehou. The Chinese Aesthetic Tradition. Trans. Maija Bell Samei. Honolulu: University of Hawaii press. 2010.

Liang, Mingyue. Music of the Billion. New York: Heinrichshofen, 1958.

Lucas, Martin. ‘Haiku in Britain: Theory, Practice, Context’, 2001. https://www.haiku-presence.50webs.com/essays/essays.html. Accessed 21 October 2015.

Lucas, Martin, with Stuart Quine. ‘Haiku as Poetic Spell’. Presence 41, 46-54.

Nelson, Andrew N. The Modern Reader’s JAPANESE-ENGLISH CHARACTER DICTIONARY. Tokyo: Tuttle, 1962; second revised ed, 1988.
.
Noguchi, Yone. Selected English Writing of Yone Noguchi: An East-West Literary Assimilation, ed. Yoshinobu Hakutani, 2 vols. London: Associated University Presses, 1992.

Nonin, Takashi. ‘Getting further and further’. Web. https://terebess.hu/english/haiku/taneda.html. Accessed 1 January 2014.

Petrucci, Mario, ‘Spatial Form’. https://www.mariopetrucci.com/spatialform.htm. Accessed 21 October 2015.

Picasso, Pablo. Quoted in https://www.cleveland.com/arts/index.ssf/2012/12/the_cleveland_museum_of_art_pr.html. Accessed 14 October 2015.

Pollatsek, Alexander and Keith Rayner. ‘Eye Movements and Lexical Access in Reading’. Comprehension Processes in Reading, ed. David A. Balota, Giovanni B. Flores d’Arcais, and Keith Rayner. Hillsdale, NJ: Lawrence Erlbaum. 1990, 143-63.

Ramirez-Christensen, Esperanza. Heart’s Flower: The Life and Poetry of Shinkei (Stanford University Press, 1994).

Richie, Donald. Ozu. Berkeley: University of California, 1974.

Roberts, Andrew Michael, Jane Stabler, Martin H. Fischer and Lisa Otty. ‘Space and Pattern in Linear and Postlinear Poetry: Empirical and Theoretical Approaches’. European Journal of English Studies, 17:1, 2013, 23-40.

Ruskin, John. The Stones of Venice. London: George Allen, 1904. https://www.gutenberg.org/files/30755/30755-h/30755-h.htm.

Santōka haiku translations. Web. https://terebess.hu/english/haiku/taneda.html. Accessed 17 September 2014.

Schofield, Fred. ‘quick through the brushwood’. Wing Beats: British Birds in Haiku. Ormskirk: Snapshot Press. 2008.

Shakuhachizen.com. Accessed 14 October 2015.

Shinkei, in Esperanza Ramirez-Christensen, Heart’s Flower: The Life and Poetry of Shinkei. Stanford: Stanford University Press, 1994.

Shirane, Haruo. Traces of Dreams: Landscape, Cultural Memory and the Poetry of Basho, Stanford: Stanford University Press, 1998.

Soyinka, Wole. ‘Narcissus and other Pall Bearers: Morbidity as Ideology’, Annual Lecture, Leeds University Centre of African Studies. 8 October 2015.

Stevens, John. ‘Going deeper’. Web. https://terebess.hu/english/haiku/taneda.html. Accessed 1 January 2014.

Still, Clyfford. https://www.tate.org.uk/art/artworks/still-1953-t01498/text-catalogue-entry. Accessed 14 October 2015].

Ueda, Makoto. Matsuo Basho. Tokyo: Kodansha International, 1982.

Ueda, Makoto. ‘Zeami and the art of the nō drama : imitation, yūgen, and sublimity’. Japanese Aesthetics and Culture : A Reader, ed. Nancy G Hume. New York: SUNY Press, 1995.

van den Heuvel, Cor. (ed.), The Haiku Anthology: Haiku and Senryu in English, 2nd edn. New York: Simon and Schuster, 1986.

Watts, Alan W. ‘Haiku’. The World of Zen ed. Nancy Wilson Ross. London: Collins, 1962, 126.

Wenxin diaolong zhu zi suo yin. Chinese University of Hong Kong Institute for Chinese Studies Ancient Chinese Text Concordance Series. Hong Kong: Commercial Press, 2001.

Wheeler, L Kip. Chinese Poetry. Web. https://web.cn.edu/kwheeler/chinese_poetry.html. Accessed 5 September 2014.

Where the River Goes. ed. Allan Burns. Ormskirk: Snapshot Press, 2013.

Xu Fuguan. Zhongguo wenxue lunji. Taibei: Xuesheng shuju. 1980.

Zhong Rong. Shi pin jiao shi. ed. Lü Deshen. Beijing: Beijing daxue chubanshe, 2000.

Zhuzi yu lei. Juan 80, 3287 in Sibu beiyuao vol. 20. Taibei: Zhonghua shuju, 1965.