Category Archives: Essays

What the Heck are “kireji” (cutting words)

Just what are “cutting words”? You do know that they are an important part of the definition of Japanese haiku, but it is very hard to find much information about what they are and why they are so important. The most often comment you read about them is that they have no equivalent in the English language. Having lived in Japan as long as I have, you run into a lot of things that you are told can’t be understood outside of Japan, but as time has goes on you realize that in most cases this is not true. So, with this attitude I will wade into the blank space of trying to explain what “cutting words” are and how they work in haiku.

First, we have to know what “cutting words” are. Haruo Shirane’s “Traces Of Dreams” (pg.100) gives a list of them:

“Kana, mogana, zo, ka, yo, ya, keri, ran, tsu, nu, zu (su), ji, se, re, he, ke, ikani, shi

Kana, mogana, zo, ka, yo, and ya are sentence-ending particles or adverbial particles, keri, ran, tsu, nu, zu (su), ji are auxiliary verb; and se, re, he and ke imperative verb endings. Ikani (how) is a speculative verb, and shi and adjectival ending.”

It’s unfortunate that Shirane didn’t offer any translations or explanations of the particles or of the auxiliary verbs, so to find out what how they might translate into English I had to go look in the classical dictionary of Japanese that I have. Even though the dictionary is in Japanese only, it is very helpful because not only does it explain the words, it also gives example of how these ancient words are now spoken in Japan.

Before moving on, it should be pointed out that the literal translation of “kireji” is not “cutting words” as generally assumed by English speakers, but rather is “cutting letters”, the list Shirane gave us actually is not of words per se, but simply alphabet letters which if used at at the right place will force a breath pause, i.e. a cut, in the diction of a sentence. Let’s start with the sentence ending particles:

“Kana” is an utterance that shows the speaker’s wonder or admiration, and works like an exclamation point does in English. (The modern Japanese equivalent is “….da naa” or “…..koto da.”) If the speaker isn’t looking at something when saying it the implication is that they are wondering if about something rather than admiring something they have in sight, which is the usage of it in spoken Japanese today.

“Mogana” is an utterance that shows the speaker’s hope or desire, the phrase of “if only….” captures the feeling it implies. (“…ga areba ii naa” is the expression used now.)

“Zo” is an utterance that adds emphasis to what the speaker has said, much like how a strong accented tone on one word can expresses one’s resolve or define what someone thinks about something. (still in modern usage.)

“Yo” is an utterance that can soften or harden the tone of expression that the speaker is using. We use soft tones in a variety of ways to add or take off implied meaning when we speak. (still in modern usage.)

“Ya” is an utterance that adds an emotional lilt to what has been said, like we do when we either speak in hushed or hurried tones to either hide or show our emotions. (still in modern usage depending on region……”da na”…….”da yo”)

Now, that we understand what these sentence ending particles are, the first thing we are confronted by is how they contradict Western ideas about haiku being objective without any subjectivity shown by the writer. All of the above are the exact opposite, they are personal utterances which clearly show the subjective emotions of the writer. And since they are utterances, they imply a first person narrative speaker. Any Japanese haiku which uses them means that the reader knows that haiku was written by an “I” speaker.

Not only does this contradict established Western ideas about the aesthetics of Japanese haiku, it also shows how the convention of writing haiku with no punctuation marks is also counter to what is in the Japanese. After all, these are sentence ending particles, and a Japanese reader understands them just as we do when reach the end of a sentence or a phrase. And, since some of them are exclamations, the Japanese reader reacts to them the same way we do when we come to an exclamation mark at the end of a sentence.

If you think of this in the context of how making “cuts” into sentences
is an integral part of writing haiku, then why did haiku in the west go off in another direction and evolve into the convention of not using punctuation at all except for the nondescript line dash? Isn’t punctuation how we cut and embellish language, building up diction to give expression an unspoken dimension? Punctuation: a tool that allows us to “cut” sentences in a variety of ways; why writers writing in such a short form would simply throw it to the wayside is so baffling, what a mystery!

Before moving on to the auxiliary verbs, it’s probably better to understand how verbs are conjugated in Japanese before understanding on how they function as “cutting letters.”

Japanese verbs are always conjugated at the end, just like we conjugate the past tense of such verbs like “walk,” where the past tense is “walked”. Of course, we also put auxiliary verbs in front the the verb to say “want to walk,” “can walk,” etc… However, the Japanese always conjugate at the end of the verb by changing the end, “aruku” is “walk”, “aruita” is “walked”, “arukitai” is “want to walk”, “arukeru” is “can walk”‘ etc… for every verb tense.

This is how the “cutting letters” that are auxiliary verbs work, they are end verb conjugations that not only cut breath pauses into phrases, they also indicate different verb tenses. Knowing this, we can proceed to explain what verb tenses they indicate. (The modern Japanese equivalents for them are in the brackets)

“Keri” has three entries in the dictionary. First it states that it indicates an action that happened in the past (…..ta, ….te ita). Second, it indicates an sudden awareness of something that the speaker now notices with admiration or wonder (da naa, atta naa), and lastly it shows the speaker expressing wonder at the recollection of a past action that someone or something else did (….ta so da, ….ta to iu koto da).

The first entry is easy enough to understand because it is the simple past tense, but since the last two also indicate recollections or reactions to things that happened in the past, it means that it would translate to using a perfect tense, and since it also implies an emotional state, it would also have to be punctuated with an exclamation mark. A perfect tense with an exclamation is generally how we express emotions to things that have happened in the past.

The similar and often used verb end of “nikeri” should also be mentioned. It indicates the memory of something that has happened in the past (…te shimatta no datta), and the admiration and wonder of some that has occurred in the past and is continuing in the present (….te kita naa, ….ta koto yo). This is what using the a perfect continuous tense indicates in English.

The dictionary entry for “ran” say that it is is a present tense speculation and is a inflection of the verb ending of “ramu” that appeared in the middle of the Heian Era. “Ramu” has four entries, the first says that it is a speculation of something that is happening which the speaker is not directly involved in doing (ima goro wa ……te darou). Next, it is noted as being a speculation about the reason why for something is now happening (….da kara, ….te iru no darou, …te iru no wa, …da kara darou). Thirdly, it is a euphemism for reporting hearsay information (….you na, ….to ka iu). Lastly, it is used the same way the verb end of “mu” is, as a simple speculative statement (…darou).

Japanese has no true future tense, rather they usually skirt around it by using the simple present tense or by attaching a speculative “perhaps” on the end of sentences to indicate that it is an action that they are not 100% sure of happening. Since none of us really knows what will happen in the future, it is hard to beat the logic behind of indicating the future in this manner. Of course, the English language is a bit less pragmatic when it comes to talking about the future, but the use of “darou” indicates the future tense with this verb ending.

“Tsu” first entry states that it indicates the completion of an movement or an action (…..ta ….te shimau), secondly it also indicates of being sure of an movement or action coming into fulfillment or realization (“kitto ……. darou). Lastly, it is used to indicate two actions following each other or happening at the same time (…..tari…….tari).

“Nu” is another ending that shows an action being completed and its entries are similar to “tsu” (…….ta ……..te shimau), (kitto…..darou), (….tari…..tari) with the only exception being that it strengthens the certainty of an action or movement being completed. In another separate entry it also can express a negative statement.

The difference between these two dictionary entries is the degree of certainty of the action being completed. To be honest about it, I’m not sure how to explain the difference. My sense of it is that “nu”‘is the simple past tense and that “tsu” is like the modern usage of “….tsutsu” (which means “while” or “under way”)
that indicates the movement of an action already started as it moves to an understood conclusion, but I could be mistaken in believing this.

“Zu (su)” is the negative verb form (….nai).

“Ji” is also the negative, but it also indicates a spectulation (….nai darou) as well as an expression of will or volition (….nai tsumori da, …..takunai), which is the future negative.

“Se, re, he, and ke” are the imperative endings for the verbs that end with “su, ru, fu and ke”.

“Ikani” is the verb modifier “how” (donna fu ni…., donno you ni)

“Shi” is an adjective ending similar to how we use the “y” to make adjectives, “sleepy, easy et al.”

As with the spoken particles we find again that what is in the Japanese is the total opposite of what is in the English. The verb usage in English language haiku is severely limited and generally only is in the present tense, and verbs are never used to express intention, recollection or speculation. Again, with haiku being such a short form, you would think that the writer would want to use all verb usages available as a way to deepen and amplify expression.

Of course, the problem with “cutting words” is that translators, both east and west, simply don’t translate them. Since they are part of the definition of what a haiku is, and add such an array of ways that the writer can express themselves, it has been a huge loss in having the convention of translations into English ignore them become the norm. This blogpost hopes to fill this void a bit so that haiku might be understood more as the poetry that the Japanese themselves enjoy with it and for the writer in English who finds value in exploring the ways that the Japanese use it to create such poetry.

Then again, the question of why you need to see what the strategies for poetry in Japanese haiku is before you write with the full abilities of your own native language is something that the individual writer must ask themselves.

HSA Definition: A Definition That Isn’t One

A couple of years ago I met a Japanese who was writing something about how English evolved as a language so I lent him some of the history of English books that I had and he in return gave me a book titled “A Haiku Path.” It is book that was published by the Haiku Society of America in 1994 to celebrate the first 20 years of its existence.

Chapters 4 and 5 in the book are titled “The HSA Definitons: What is a Haiku?” and “The HSA definitions: Haiku, Senryu, Hokku, Haikai” which give a detailed account of the correspondences and meetings that led to the writing and establishing of a definition of haiku that the society used from 1973 until 2004 when it slightly changed the definition.

These two chapters clearly show that the HSA definition was not a definition about what haiku was as a form, rather is shows that the definition was formed with the clear intent to claim one style of writing (i.e. objective) as the only style that haiku could ever be, and this claim that this style was the only valid one was wrongly based on the idea that this is how the Japanese always wrote haiku.

In the April of 1971 Anita Virgil wrote to Harold Henderson to ask about why he had in 1970 asked members to gather definitions of haiku, hokku and haikai, and Henderson soon after returned a letter that said he wanted “to get all haiku enthusiasts to agree on some one definition of the word and to get all English dictionaries to define it.” Henderson, who was battling cancer at the time, asked Virgil to be the force behind the idea due to his poor health. Virgil replied that she thought it was of the most importance that he was the one who first penned a definition because “no one else can commend the attention and respect of such haiku enthusiasts” and “no one’s opinion bears the weight of authority…….than yours does…..dictionaries and schools depend on “authorities.”” (pgs. 44-45)

This prompted Henderson to write back on May 15, 1971 a reply that contained this definition of haiku (pg. 46):

“1. A short Japanese poem recording a moment of emotion in some way connected with nature.  Usually, but not always, consisting of 17 ‘jion’ (symbol-sounds)
2. An English adaptation usually consisting of 17 English syllables, or less.”

The first thing to note in this definition is that it definitely states that the English adaptation is copying the Japanese model, that the writer in English is writing like a Japanese if they “record a moment of emotion in some way connected with nature.” And the second thing to note is that this is not a definition of a form, but is rather a definition for the state of mind that the writer should apply when writing a haiku.

Henderson wrote this earlier in the same letter (pg. 46):

“It is hard enough to define poetry (see our various dictionaries); a definitive definition of haiku is probably impossible. When I asked Professor Yagi (a long time Japanese acquaintance of Henderson) he answered, “There is no definition of haiku in Japanese.  Haiku are what the poets make them.” ………..obviously haiku is not a form.”

It is obvious that Henderson is confused about what he is trying to define. Sure, there is no one definition of poetry, but every verse form of poetry has a definition. Sonnets and terza rima etc….all have definite definitions of what they are as a form. What you can’t define is how a writer will decide to think and write in them.

Henderson’s statement that haiku is a not a a form is categorically false. Everyone knows that the classical definition of its form in Japanese is that it is 17 letters, has a defined seasonal word and a defined cutting letter in it. Henderson didn’t understand Professor Yagi’s reply to him because he didn’t understand the difference between a definition of poetry and a definition of a verse form. Yagi is telling him that the Japanese use the haiku as a way to express themselves and there is no defined mindset of how the writer must think in haiku.

Despite the obvious problems, in the June 16, 1971 HSA meeting, Henderson’s definition was slightly amended and presented with the urging of comments upon them from HSA members (pg 49):

“1) A Japanese poem, a record of a moment of emotion, in some way way linking Nature with human nature. Usually, but not always, consisting of 17 ‘on-ji (sic) (Japanaese  “symbol-sounds”
2) An English adaption, usual consisting of 17 syllables or less.”

By the October 18, 1971 meeting there were some members who had trouble with the use of “emotion” in the definition. Michael McClintock wrote (pg 52):

“I feel that the word “emotion” represents an emphasis and conceptual limitation that is fundamentally misleading, and which neither Japanese haiku nor it’s English adaptation convincingly sustain. I think “A Japanese poem, a record of a moment’s keen perception or experience” is as inclusive as it should be, and as exclusive as it must be, for accuracy.”

William J. Higginson, who as was at the meeting, noted he was also bothered by the word “emotion” and when “someone recalled the definition of poetry as” “emotion recollected in tranquility” (Wordsworth)”  Higginson commented “that few modern poets hold to that definition.”” (pg. 52)

Anyone that was interested in poetry in the 20th century will immediately recognize that problems McClintock and Higginson had with the use of “emotions” were based on nothing more than the attitudes towards what kind of poetry should be written in America at the time because the definition of emotions, and to what extent that the poet should express their own personal emotions, were the battleground markers that T.S. Eliot set when he attacked Wordsworth’s subjective poetic style against the objective style of writing that he thought was right for poetry.

That the problem was between “subjective writing” and “objective writing” is shown clearly in the reasons why Anita Virgil proposed that the “of emotion” in the original be replaced by McClintock’s “keenly perceived” at the December 13, 1971 meeting (pg 53):

“I have found his “record of a moment of emotion” can be read with incorrect or misleading emphasis. If stress is laid on “moment of emotion,” the way is still open to subjectively written haiku. Yet, it cannot be denied that haiku is initiated out of emotional response and should be able to evoke similar response from its reader. I eliminated the word “emotion” from my version of the definition…….(the use of “record” in the definition) encompasses the objectivity we desire in haiku and also tucked in the word is the preserving of an event of the present.”

It is now too obvious that the intent behind this definition is now to drive any kind of subjective writing out of haiku by having the society officially deign that Japanese haiku never was any kind of writing except objective statements that were written in the present tense. When Virgil offered this definition she too so connected it by also opening her definition with “a Japanese poem.”

It is pretty easy to prove that these ideas about Japanese haiku are completely wrong, all you have to do is look at what “cutting words” provide for the Japanese writer. They are either different verb tense conjugations, which throws the idea that haiku is always present tense right out the window, or they are spoken particles of speech by which an “I” speaker makes a personal emotional statement out things they write about, which makes the idea that the Japanese always only wrote objectively a very ridiculous one.

The book now severely cuts down Henderson’s part of the conversation in the process (and it’s a loss to not see what his objections were), but it noted that “a barrage of correspondence went on between Virgil and Henderson reflecting the frustrations of the Definitions Committee members in their long year-long search for common understanding, and for establishing priorities within the definition work” and “several letters from Henderson to Virgil in April which expressed dissatisfaction with the work the committee had done so far.” It also notes “that on the 26th he sent to the committee members, Virgil and Higginson, a four-page letter stating that he thought Addendum 4 (dictionary definitions prepared by Virgil and Higginson that were presented on Mar 13, 1972) was the “wrong approach” and raising a number of objections to the definitions themselves” and how “on the 28th he wrote Virgil relating details of a long conversation he had with Higginson the night before which left him “with the impression, among others, that we do not agree on the meaning of some words used in our ‘definitions.'” (all quotes pg. 62)

The disagreements came to head when in a letter dated May 2, 1972 Virgil wrote to Henderson telling him that she has “given all I can to the creative aspect of this project” and that she is “withdrawing” from project. In the letter she argues that “a close scrutiny of the elements in our version provides that the reader will be able to experience it himself if the poet records the essence of a moment keenly perceived. It is the moment (the limited-in-space-and-time thing of which you speak in your letter to me, not the poet’s feelings about what occurred) that the haiku seeks to re-create. If he does just that, gives the objects just-so, the happening, the reader will be permitted to have the experience FOR that is all he has given………..My main concern has not been with acceptance but with attaining a definition which I believe functional and true. I believe the definitions contained in this last contribution from Bill (Higginson) and me are just that.” (pg. 63)

Again, anyone with in an interest in poetry in the 20th century will immediate recognize that Virgil’s argument is echoing T.S. Eliot’s famous ideas about an “objective correlative” which is defined by the online Merriam-Webster dictionary as “something (a situation or chain of events) that symbolizes or objectifies a particular emotion and that may be used in creative writing to evoke a desired emotional response in the reader.” And her ideas about the role of the poet suppressing their feelings echoes Eliot’s himself in the 1919 essay “Tradition and the Individual Talent” where he stated “Poetry is not a turning loose of emotion, but an escape from emotion; it is not the expression of personality, but an escape from personality.”

Her withdrawal only lasted a week and it is obvious that Virgil won the battle against “the only authority who could commend the attention and respect of all” because at the May 15, 1972 meeting Henderson tentatively proposed a definition of haiku in which “keenly perceived” had replaced “emotion.” The official HSA definition now is simply a rehash of ideas that came to define 20th century poetry in America.

By the January 15th of the next year a letter was placed in the minutes that was to be the official HSA letter that was to be sent to editors of dictionaries concerning what the society had deemed the correct definition of “haiku” to be. It had Harold Henderson’s name as the lead authority on it, with Anita Virgil and William J. Higginson as the Committee of Definitions members below.

It was a definition that claimed validity by being based on the Japanese and it was one that defined haiku as a type of poetry that could be fulfilled by only one certain type of sensibility. There was absolutely nothing about what the actual form of what haiku was except a weak reference to it being three lines and 17 syllables which the note after made superfluous. (pg. 82):


1. A unrhymed Japanese poem recording the essence of a moment keenly perceived, in which Nature is linked to human nature. It usually consists of seventeen “jion” (Japanese symbol-sounds)
2. A foreign adaption of 1. It is usually written in three lines of five, seven and five syllables.

Note to 2
That part of the definition which usually begins “It is usually written” places a heavy weight on the word “usually.” We depend on that that word to provide latitude for variations in syllable count and in number of lines or other external aspects of “form” providing they meet the primary stringent requirements expressed in the first part of the definition. Through 17 syllables is still the norm in English language haiku, it is more and more common for a haiku to consist of fewer syllables. Rarely is a haiku longer than 17 syllables.”

There can be no doubt now that this definition is not interested in giving a definition of haiku as a verse form, rather it is trying to install one way of writing as the only way to create haiku. Everything in the definition is subservient to “meeting the primary stringent requirements expressed in the first part of the definition.” And, of course, the “stringent requirement” is that the writer must only write in an “objective style.”

With this given the seal of HSA approval, editors gladly assumed that the HSA knew what it was talking about and this definition showed up in print everywhere. My own experience makes it easy to makes  to understand what happened in the world of haiku after, how with a now accepted definition that expressly forbid any other type of haiku except one that was “a moment keenly perceived” (which later became better known as ‘a haiku moment’), haiku editors ruthlessly drove out anything other type of writing in the name of being “pure” to what haiku was in Japanese.

In the early 1990s the Haiku Society of America now faced the dilemma of having a definition that was now known to be based on a severely flawed idea about Japanese haiku, due to Haruo Shirane’s book, “Traces of Dreams: Landscape, Cultural Dreams and the Poetry of Basho” showing how skewed the Western perception of Matsuo Basho was. So they reconvened a committee to rewrite the definition, which, interestingly enough, Anita Virgil refused to join by stating she saw no reason to do so (Harold Henderson had died in 1974 and William Higginson was a chairman of the second committee) but the committee died away without making any progress.

With the development of the internet there was more pressure about how English language haiku had mistaken ideas about Japanese haiku, and in 2003 an another committee was formed to work on a new definition. (Higginson again was a member.) By 2004 they had come up with a second definition, which feigned allegiance to the Japanese by talking about some of the ways they write them in the note after, and simply reaffirming all the arguments used for changing Harold Henderson’s original by inserting “imagistic language” for “keenly perceived”:


Definition:  A haiku is a short poem that uses imagistic language to convey the essence of an experience of nature or the season intuitively linked to the human condition.

Notes:  Most haiku in English consist of three unrhymed lines of seventeen or fewer syllables, with the middle line longest, though today’s poets use a variety of line lengths and arrangements. In Japanese a typical haiku has seventeen “sounds” (on) arranged five, seven, and five. (Some translators of Japanese poetry have noted that about twelve syllables in English approximates the duration of seventeen Japanese on.) Traditional Japanese haiku include a “season word” (kigo), a word or phrase that helps identify the season of the experience recorded in the poem, and a “cutting word” (kireji), a sort of spoken punctuation that marks a pause or gives emphasis to one part of the poem. In English, season words are sometimes omitted, but the original focus on experience captured in clear images continues. The most common technique is juxtaposing two images or ideas (Japanese rensô). Punctuation, space, a line-break, or a grammatical break may substitute for a cutting word. Most haiku have no titles, and metaphors and similes are commonly avoided. (Haiku do sometimes have brief prefatory notes, usually specifying the setting or similar facts; metaphors and similes in the simple sense of these terms do sometimes occur, but not frequently. A discussion of what might be called “deep metaphor” or symbolism in haiku is beyond the range of a definition. Various kinds of “pseudohaiku” have also arisen in recent years; see the Notes to “senryu”, below, for a brief discussion.)”

The real problem with the 2004 redefinition, the definition that is posted on the society’s webpage today, is that it does not own up to the mistakes of the past. It’s rather hard to see it as anything but something to retroactively add cover to the type of haiku that got written under the mistaken auspices of the first definition because the only way to understand what the term “imagistic language” means is to see it as a rephrasing of Eliot’s “objective correlative” theory, which the second definition explains as “the original focus on experience captured in clear images.” There is no way to understand the phrase “metaphors and similes are commonly avoided,” except as being seen in the context of Anita Virgil’s statement about how haiku can not be the “poet’s feelings about what occurred.”

The Haiku Society of America’s stated goal is “to promote the writing and appreciation of haiku in English,” but as long is it continues to hold to a definition that only allows for one certain style of poetry it utterly fails in its mission to do so. And it cannot do this until it comes up with a definition that allows for the writing of a “subjective” haiku that doesn’t use “imagistic language.” And it cannot do that until is comes up with a definition that views haiku solely as a verse form rather than a certain style of poetry. Perhaps the first step in this would be for it to review what Harold Henderson’s rationale for writing “a moment of emotion” was and include it in a new definition.