Every New Year’s Day an ancient and august ceremony takes place in The Palace of The Emperor of Japan in which a tanka poem written for the occasion by The Emperor is sung. Yes, you read that correctly, the poem is sung, for tanka in Japan are often sung, as opposed to recited.
There are two reasons I begin with comments on tanka rather than this essay’s core subject, haiku. In the first place, whenever I mention the singing of tanka to writers of that genre they are invariably surprised. Indeed, I have yet to meet a writer of tanka in English who has heard a Japanese tanka being sung.
And it is not only The Emperor’s tanka that is sung in Japan, singing tanka is a norm. Hyakunin Isshu is a popular Japanese card game in which contestants kneel opposite each other with one hundred cards on the ground between them. Each card has the last two lines (the second verse) of a tanka poem written on it. A quizmaster, so to speak, sings the first verse of one of the tanka and the contestants’ task is to find the card with the final two lines of that verse and remove it to claim a point. As the game progresses there are fewer and fewer cards and the contest becomes increasingly tense and exciting. Played in homes throughout Japan, this game has a serious aspect too, with regional and national games televised. The name of the game, Hyakunin Isshu, refers to the fact that the game consists of one poem each, by one hundred poets. Most Japanese viewers will be able to identify a large number of the final verses on hearing the opening verse as it is sung.
Tanka is a modern name for the very old poetic form of waka. And from the waka, with its five line structure of 5-7-5-7-7 morae (commonly referred to as ‘syllables’ in English), the renga developed. The first verse of the renga became the hokku, which in turn, developed into the haiku. A consistent feature of this entire line of progression over the best part of a thousand years, is the use of lines of 5 or 7 morae and, in particular, the verse structure of three lines of 5-7-5 morae. This is the second reason I open with tanka: the classic 5-7-5 structure of the haiku is a key and consistent feature of Japanese poetry all the way back to its emergence in the form of waka.
Where there is song, there is rhythm. When we consider the long tradition of singing tanka, and the close connection of haiku to this old form, our attention is drawn to the rhythm of haiku poems.
There is a difficulty about the use of the word ‘syllables’ when talking about Japanese literature. A syllable is a linguistic phonic unit, that is to say that it is a unit of sound. In the English language syllables can be short or long sounds. For example, ‘I breathe’ is comprised of two syllables, of which, ‘I’, is short and ‘breathe’ is long. The sound units of the Japanese language are called morae (plural for mora) and morae are all of the same phonic length. Syllables and morae are essentially the same thing – units of sound. There is a distinction in that morae are ‘syllables’ of the same phonic value. This is an important point as using specific numbers of morae to structure poems delivers a rhythm or metre that cannot be achieved by using a syllable count in English with its long and short sounds.
The English language is not so much a ‘syllabic’ language but one based on beat – an accentual-syllabic language. The rhythms of literature in English derive not from the numbers of syllables used (as in Japanese) but from the syllabic accents, or beats. For example, the iambic pentameter is based primarily on the number of beats that a line contains – in its case, five (penta coming from the Greek for five). Each line is comprised of five sets of two syllables with the beat falling on the second one, but it is the beat that is our focus not the syllable count.
Many haiku writers in English dismiss the so-called syllable count of the Japanese for writing haiku in English. However, no alternative structure has become widely accepted.
Some linguists describe morae as phonetic units of which there is one mora in a short syllable and two in a long one. It is tempting then to consider writing haiku in English in the 5-7-5 form, counting short syllables as one unit and long ones as two. The difficulty with this is that the rhythm of a long syllable is not the same as the rhythm of two morae. However, it is likely to be closer to the Japanese form of haiku that counts in morae than an approach that counts all syllables as if they constitute one mora each.
We tend to see other languages as variations of our native tongue and generally try to fit them into its template. This can be a source of confusion. For example I recall being asked by a friend in Japan if I would like to go to Makudo? (It turns out that Makudo is an abbreviation). I hadn’t heard of this and my hesitation prompted my Japanese pal to repeat the name in full – shall we go to Makudonarudo? Now I was truly puzzled, where or what is Makudonarudo? They put me out of my misery by pointing down the street to the yellow arches of a McDonald’s. Yes, those world-renowned arches of Makudonarudo. We went there and had two Biggu Makku.
Makudonarudo is not the Japanese word for McDonald’s, as such, rather it is the way Japanese speakers say McDonald’s. It is an attempt to use their limited repertoire of 101 sounds to say a word that their stock of sounds does not cover. And we do the same. We also alter the way we read and say Japanese words (especially when we read them in our alphabet) to try and fit them into our linguistic template. When I lived in a rural area of Japan called Akaiwa (A-ka-i-wa) I found it difficult to recognise what my English speaking friends were saying when they referred to it as they would say it with a pronunciation, tone, and rhythm that was not familiar to me. They would ask me how things were in Akaywa and I would wonder where they were talking about. This is not simply a matter of mispronunciation, but more importantly, it is also about misunderstanding the rhythm of another language when we attempt to speak it. Every vowel in Japanese is pronounced separately. If we merge two vowels when speaking (such as the a and i in Akaiwa) then not only do we mispronounce but we also change the rhythm of what we are saying.
5-7-5 in everyday life
In Japan we encounter the 5-7-5 morae pattern, so associated in The West with haiku, in all kinds of everyday situations.
One of Japan’s national newspapers, Mainichi Shinbun, has been eliciting road safety slogans from its readers every year since 1966. They publish at least 20 of these annually and in some years have printed more than 50 entries. There are no rules for the participants to follow other than the creation of a slogan that might help reduce death and injury to road users. I have translated the following two examples:
sumaho motsu sono isshun ga jiko no moto
using your smart phone
in this moment
accidents are caused
awatenna mukashi wa minna aruiteta
in the old days
we all walked
What is curious about these slogans is that, just like haiku, they are almost always in the form of 5-7-5 sounds and consist of three phrases. They look like haiku. They have the rhythm of haiku, and therefore sound like them. Yet haiku they are not. Nor are they intended to be. They are slogans.
Spoken phrases of five or seven syllables work well in Japanese and in Japan we hear them everywhere. The 5-7-5 format can even be identified in everyday announcements such as in train stations or department stores. As with the slogans above, there is no intention for these utterances to be haiku (or senryu for that matter), it is simply that this is a very natural way to communicate in the Japanese language.
Some Japanese history teachers use three haiku-like verses to help students remember the chronology and leadership characteristics of the Samurai chiefs: Nobunaga, Hideyoshi and Ieyasu. The first of these, Nobunaga, was a violent dictator who ruled by the sword. His successor, Hideyoshi, was a skillfull diplomat who could cleverly manipulate people into doing his will. Ieyasu was known for his patience, which served him well as he quietly waited for Hideyoshi to die of natural causes before stepping into his place. The three haiku-like verses I mentioned above all share the same upper and lower lines for ease of recall and the middle lines differ to indicate the nature of each of the leaders. Here they are with my translations:
nakanunara koroshite shimae hototogisu
if it doesn’t sing
just kill it
nakanunara nakasete miseyou hototogisu
if it doesn’t sing
I will make it want to sing
nakanunara nakumade matou hototogisu
if it doesn’t sing
I will wait until it sings
In Japanese we hear variations of phrasing in combinations of five morae and seven morae such as 5-7 or 5-7-7, 7-7, and of course the haiku and senryu format of 5-7-5. When we look at other poetic structures in Japanese literature we most often find that they too are based on variations of morae phrases of either five or seven. Perhaps the most famous of these forms to westerners is the tanka with its structure of 5-7-5-7-7.
There are two fundamental dynamics that combine to deliver rhythm in haiku. They are not the only two, but the two I will talk about here are the principal rhythmic forces in typical Japanese haiku.
The first of these is the inherent rhythm, or pattern, contained in the moraic nature of the Japanese language. The second is the use of the Japanese aesthetic known as ‘ma’.
Ma is a particular and ubiquitous aesthetic that is used in all forms of art in Japan. Ma is often translated as ‘emptiness’. It is about what is not there in a work of art. It can be the empty space in a visual image or an Ikebana flower arrangement. When skilfully used ma can bring a vibrancy, an energy, to a work of art. The empty space that can act to highlight what is there, the silence that amplifies music.
In haiku there is ma in the rhythm – the silences at the end of each of the first and second lines. Professor of English at Kobe College, Hirai Masako, states:
When reading haiku, you should pause after 5 syllables, and again after the next 7 syllables, in order to enjoy its rhythm.[iii]
These pauses, typically occurring as Hirai sensei describes, are expressions of ma. They are employed at the line ends as a deliberate contribution to the overall rhythm of the haiku. Furthermore, they are not usually the same length, indeed, typically the first pause is three times longer than the second. This is described by Kan’ichi Abe in a short article entitled ‘Rhythm: Fixed Form of 5-7-5’.[iv] He cites an example by ‘a scholar’ who, using the symbol 0 to indicate a pause
illustrates the similarities between a tune with eighth-notes and haiku
Ha ha ki gi ni 0 0 0 / ka ge to i u mo no 0 / a ri ni ke ri 0 0 0.
If we use digits 1 to 4 to indicate the beat of eighth notes, bracket the beats that are paused, and then arrange these into three lines as a haiku, we may see the rhythm as beats.
1 2 3 4 1 (0 0 0)
1 2 3 4 1 2 3 (0)
1 2 3 4 1 (0 0 0)
This is the rhythmic pattern of a typical Japanese haiku of 5-7-5. When viewed in this way we can see its shape more accurately. With the duration of the pauses (the ma) accounted for, we gain the perspective that each of the three lines of the poem are the same length. They cover the same temporal ground. For a musician this has a similar effect to adding a time signature to sheet music. Without a time signature, a music score becomes difficult to decipher. Musicians can interpret it in various ways but not be able to understand the intentions of the composer. This is the essential problem faced by English speaking haikuists if they are not aware of the nature of haiku rhythm in Japanese. Like all rhythms, we need to hear it to appreciate it. However, many haikuists I know, who write in English, tell me that they have never heard a Japanese haiku being read aloud. It is for this reason that I read haiku in Japanese aloud at my haiku workshops so that the students can hear both their tone, and importantly, their rhythm.
Let’s take a look at a well-known haiku by Buson and see if we can develop a sense of its rhythm using the above approach.
na no hana ya
tsuki wa higashi ni
hi wa nishi ni
field of rapeseed –
the moon in the east
the sun in the west[v]
In this first step the romanized version of the haiku is on the left and my translation to the right. Next we will see the romanized version broken into its sounds, or morae:
na no ha na ya
tsu ki wa hi ga shi ni
hi wa ni shi ni
Now to add the pauses in brackets:
na no ha na ya (0 0 0)
tsu ki wa hi ga shi ni (0)
hi wa ni shi ni (0 0 0)
It is worth noting that a pause of three beats is allowed for after the final line. For the Japanese, this silence, this space, or ma, comes at the end to complete the haiku.
Haiku are, in essence, an oral tradition. They come into their own when read aloud, hence Hirai sensei’s instruction as to how to read them, pausing in the appropriate places to deliver their rhythm. To ‘enjoy’ it, as she puts it.
In the Japanese language a poet is said to ‘read’ or ‘make’ haiku as opposed to ‘write’ them. They use the verb, to write, when referring to the physical act of writing. Given this, and the importance of rhythm in haiku, I favour the word ‘compose’ when talking about the act of ‘making’ haiku. This has the added advantage of implying musicality, an important dimension of haiku poetry.
Putting in the Pauses
This is not an exhaustive examination of rhythm in Japanese haiku. There are other factors such as the use of kireji (sometimes referred to as ‘cutting words’ in English) or haiku that have more than 17 morae (jiamari haiku) or of less than 17 morae (jitarazu haiku). I have focused here on what I regard as the principal factors of rhythm in Japanese haiku – the 5-7-5 structure and the use of the aesthetic ma in the form of pauses at the line endings.
Writers of haiku in English have long tended either to favour the 5-7-5 structure (sometimes exclusively) or to reject it (sometimes completely). The idea that the 5-7-5 structure does not work in English is based on syllables in English being of differing lengths and morae in Japanese being sonically uniform. However, there are some fine haiku written in English using the 5-7-5 syllable format. It is more the case that differences between syllables and morae are complicating factors rather than excluding ones.
An outright rejection of the 5-7-5 structure of Japanese haiku when writing in English can overlook its role in providing rhythm in haiku. And rhythm is a crucial aspect of all poetry.
It is my hope that we can rethink this question from the perspective of the role of 5-7-5 in the rhythm of haiku and consider ways to strengthen the rhythm of haiku in English.
The use of ma in haiku is much more clear-cut in my view. Silence is available in all languages. The use of pauses as described in this essay can be incorporated into haiku in English (some already are). It is my experience as both a writer and editor of haiku that employing ma in this way invariably adds rhythmic strength to haiku.
One area that requires careful attention in this regard is the use of enjambment in haiku. This is a popular technique in English poetry generally and has increasingly entered the haikuists’ toolbox. Where two lines of a haiku are enjambed, there is the risk that when read aloud one line will run into the other, so to speak. This will eliminate any pause between those lines. However, it is possible to use enjambment in such a way that there is a natural pause at the end of the first line in such a case. Here is an example from the Irish Nobel Laureate Seamus Heaney:
But this year I face the ice
With my father’s stick.[vi]
The second and third lines are enjambed. However, when read aloud there is a pause at the end of the second line of the type that occurs in a typical Japanese haiku. It is as if there is an invisible comma there, and this pause sounds natural when recited.
With or without a judicious use of enjambment, the rhythm of haiku in English can benefit from the use of ma. Forgive me for referring to the title of my own first haiku collection in my closing remark. We may do well if we Let Silence Speak.[vii]
Sean O’Connor, 28th January 2019
i Nishi, Gene. (2000) Systematic Japanese: A Simple, Effective Method for Self-study. Shufutomono Publishing. Tokyo. p. 13.
ii Homonyms are words with the same sound but different meanings and heteronyms are words that are spelt the same but differ in meaning.
iii Hirai, Masako, ed. Now, To Be! Shiki’s Haiku: Moments for Us Today. Osaka: U-Time Publishing, 2003. p. 14.
iv Abe, Kan’ichi. ‘Rhythm: Fixed form of 5-7-5.’ In Japanese Haiku 2001. Modern Haiku Association eds. Tokyo: Gendai Haiku Kyokai, 2001. p. 253
v My translation.
vi There are at least three versions of this haiku by Seamus Heaney. This one is from an article by Brad Leithauser, In Praise of Concision, published in The New York Times, 17/2/2013. The other versions I have seen differ in the second line but do not have the effect of altering the point being made here.
vii O’Connor, Sean (2016) Let Silence Speak. Alba Publishing. London.
— Sean O’Connor
Presence 63, March 2019
Available as a PDF