It occurred to me recently that one way to apprehend the difference between the sounds of Japanese and the syllables of English is to think of baseball. The pattern of 5-7-5 sounds (not syllables) in a Japanese haiku produces a poem of a particular heft or weight. Think of that as being like a baseball. And I do mean a baseball—a specific size of ball used in American professional baseball leagues. But if you write 5-7-5 syllables in an English haiku, you end up producing a bigger ball—like a softball—because of differences in language. Japanese words are typically many short staccato syllables, so a Japanese haiku reaches 17 sounds using fewer words and less content than what is provided in an English haiku in 17 syllables. In fact, I recently read an observation by Kit Pancoast Nagamura, a Japan-based haiku poet who hosted NHK’s “Haiku Masters” television show for three years, that if you write 17 syllables in English, you can easily write enough content to fill two haiku in Japanese (she said this in her book, Grit, Grace, and Gold: Haiku Celebrating the Sports of Summer, Tokyo: Kodansha USA, 2020, xvi). In the same book, Nagamura adds that “after nearly 30 years in Japan, I’ve come to think that briefer poems, far shorter than 17 syllables, better capture the piercing brevity which is the hallmark of traditional haiku” (xvii). Thus, an English-language haiku will typically reach the weight and size of a “baseball” with fewer than 17 syllables, whereas insisting on 5-7-5 syllables nearly always produces a larger “softball” size of poem. As a result, a 17-syllable haiku in English may be said to be “obese” compared with the leaner weight of a Japanese haiku. That leaner haiku isn’t “hefty” at all, but does have a particular heft.
Although it helps to know Japanese, one can still get a feel for this difference in heft by listening to haiku in both Japanese and English, hearing the distinction that 5-7-5 syllables in English nearly always takes longer to say, never mind that the English also contains more words, concepts, or images. Sensitivity to the words and images shared in Japanese haiku will give you an additional sense of each poem’s individual heft, and therefore a sense of the heft of the genre itself if you pay attention to many haiku in Japanese, even if only through translation.
You can also get a feel for the difference by using Google Translate on any Japanese text other than haiku. If you count the sounds of the Japanese—properly counting vowels with macrons such as ō or ū as two sounds and the “n” sound as an additional sound at the ends of words (and within certain other words)—you’ll see that the English nearly always has fewer syllables. To illustrate with a specific example in prose, a note in Japanese on the Haiku International Association (HIA) website says the following:
The romaji for this text is as follows, and counts out to 78 on (sounds, counting “HI” spelled out as “e/i/chi/a/i” for “aitch, eye”), confirmed by a native Japanese speaker:
Kokusai haiku kōryū kyōkaide wa, kikan-shi HI ya kochira no hōmupēji oyobi denshi mēru o tsukai, kaiin no minasama ni go an'nai ya go renraku o shite orimasu.
In comparison, below is the website’s own translation of the same text in English. Here they are translating the content, with no need to accommodate any particular syllable count (thus, no intentional padding or chopping), which comes out to only 59 syllables. This includes two instances of “HIA,” each counted as three syllables in English, one syllable for each letter, which does not even occur in the Japanese, making the English longer than it could have been because the first acronym could have been omitted and the second one could have been replaced by “It,” reducing this content to just 54 syllables. Also notice, simply visually, how the English takes up more space on the page, despite having significantly fewer syllables compared to the number of Japanese sounds:
HIA, the Haiku International Association, has no account with either Instagram or Facebook, and does not maintain a presence on either of those platforms. HIA publishes a quarterly magazine called HI.
The difference in this specific example is about 76 percent—the English needs only 76 percent of the syllables (compared to the number of sounds in Japanese) to get the same information across (or just 69 percent if the HIA acronyms are revised in the English). Pick any Japanese novel or other prose passage and you will always see a similar ratio if you can compare that content to an English translation. William J. Higginson has written that about 12 to 14 syllables in English is equivalent to the 17 sounds of Japanese, a ratio of about 71 to 82 percent. Others have proposed 10 to 12 English syllables, a ratio of about 59 to 71 percent. The Tanka International organization in Tokyo advocates for a maximum of just 21 syllables in English as being equivalent to the 31 sounds of a traditional Japanese tanka, a ratio of about 68 percent. It therefore cannot be escaped that, if one is writing 5-7-5 syllables in English, the poem will simply be longer, fuller, and with excess heft (which we might call obesity) compared with a typical haiku in Japanese.
The following anecdote may dramatize the difference. John Stevenson has said he once attended an international Playback Theater conference in the mid-1990s. At the event he shared a 5-7-5 English-language haiku (not one of his own) with some of the Japanese participants who also wrote haiku. In response, they asked, “Why is it so long?”
Now let’s also consider a haiku example—one of my own that is indeed long—which received an honourable mention in the 74th Bashō Haiku Contest in 2021, held in Iga City, Japan (birthplace of Bashō). Here’s my poem, followed by the contest’s Japanese translation (see complete contest results):
hush of first snow—
a single candle burns
on the mahogany pulpit
hatsuyuki ya kyandoru hitotsu saidan ni
In English, my poem is 4-6-8 syllables, admittedly on the long side. In Japanese, the translation is 5-7-5 sounds, and includes the “ya” cutting word to match the cut indicated by the em dash in English (“ya” is an essentially meaningless term used only in haiku to indicate a pause or emphasis, using up one of the 17 Japanese sounds, leaving what remains to be even leaner). However, the 18-syllable length in the original English is too long to fit into the 5-7-5 pattern in Japanese, and not just because of its 18th syllable. So, of necessity, the Japanese translation had to cut words. Loosely, in Japanese, the poem is “first snow— / one candle / on the altar.” That may be a better poem, which further bolsters my point about the leanness of haiku in Japanese compared with English. The simplicity of the Japanese translation shows how wordful and perhaps even wordy my original poem is in English. My main point, though, is not which is better but how the English had to be shortened to fit the sound pattern in Japanese. The translation loses “hush” (which may well be implied, so that’s okay), as well as “mahogany” and changes “pulpit” to “altar.” These changes, except for the use of “altar,” demonstrate why a poem as long as 17 syllables (and one syllable longer in this case) is too long compared with Japanese. Specifically, the translation cuts five entire syllables from the English to make the poem fit into the 17 sounds of Japanese. In English, to put it simply, 17 syllables is heftier than the 17 sounds of a Japanese haiku. And, by heftier, I do not mean meatier or richer, but simply obese.
Of course, one could beg the question: Does a haiku in English have to match or approximate the weight of a haiku in Japanese? In other words, do they both need to play with the same ball size, even if the sounds in Japanese cannot be directly translated into 17 syllables in English? One has to say no in order to justify 5-7-5 syllables as a target for haiku in English. This is partly why I say that 5-7-5 syllables in English is a violation of the Japanese haiku form, not a preservation of it. But, if you choose to say that 5-7-5 syllables is your target for haiku in English, seemingly accepting it as a given, then you have made a compromise. And, if you allow yourself to make that compromise (and it definitely is, because the poem is in reality a different length), then why isn’t the “compromise” (from a 5-7-5 writer’s point of view) to write haiku with fewer than 17 syllables equally allowable? On the surface, then, it seems possible to argue that one option or the other is a compromise, and that may well be the case. However, I would suggest that 5-7-5 syllables in English is the only compromise compared with Japanese. That’s because it produces a softball, a ball of a different heft, a choice to write “obese” haiku, no matter how one might try to argue otherwise. Furthermore, having 17 syllables is an issue without even accounting for the fact that the syllables can vary so greatly in English compared with Japanese (compare “strengths” and “radio,” for example). As has been pointed out before, the word “haiku” itself is two syllables in English, but counts as three sounds in Japanese. That alone should serve as a convincing metaphor for the problem of considering 5-7-5 syllables in English to be equivalent to Japanese haiku form. But no, 5-7-5 syllables is a ball of markedly different size, and thus a departure from the lean art of Japanese haiku. A poem shorter than 17 syllables in English is therefore not “minimalist” at all, but hews closer to the heft of haiku in Japanese. And, thus, a poem as long as 17 syllables in English is better understood as “maximalist.” If one chooses to accept the compromise of 5-7-5 syllables in English, that’s always a personal choice, but it remains a compromise, and one is still obliged, for the sake of haiku as a literary art, to employ other techniques that matter more than filling a bigger bucket.
The fact that 17 syllables produces a “softball,” which is larger than the “baseball” produced in Japanese haiku, is one reason why the vast bulk of contemporary haiku translations in English do not conform to a pattern of 5-7-5 syllables. As with writers of original haiku in English, these translators also seek to hit other targets usually present in the original Japanese, such as the effective use of seasonal reference (kigo or season word), a two-part juxtapositional structure (equivalent to using a kireji or cutting word to divide the poem in two), and the use of primarily objective sensory images based in the five senses, among other targets. In addition, the discipline of hitting all these targets is far greater than merely counting syllables, which is the most trivial of haiku’s disciplines. Indeed, if all you aim for in English is to dutifully count out your 17 syllables, at best you end up producing a softball, not a baseball. It’s a bigger ball, easier to hit, and less dense, and thus used in bush-league sports, not professional American leagues. Furthermore, many Western attempts at haiku, such as widely seen on the Internet and social media, miss so many other targets that they don’t even produce a softball, but poems with little substance or density, a Wiffle ball. I do not mean to suggest that these Wiffle-ball haiku are childish; rather, I am trying to clarify simply that their writers are producing a different heft of ball, as are those who are producing softballs.
Can you still have fun with a Wiffle ball or a softball? Absolutely. A problem arises, however, if poets playing with these balls believe they are playing baseball. If they think so, it would seem that their poems will strike out nearly every time and their players won’t even know it. Furthermore, their poems won’t translate smoothly or completely into the traditional Japanese form, as shown with my “hush of first snow” poem. What works in the sandlot game won’t fly in the big leagues or even the minor leagues. Nevertheless, perhaps these more amateur balls are stepping stones to the bigger leagues, and it’s no wonder that the vast majority of the leading haiku poets writing literary haiku in English have graduated through Wiffle balls and softballs to the making of real baseballs. Quite simply, they have a feeling for the real heft of haiku.
Michael Dylan Welch has been investigating haiku since 1976. He is a director for the Haiku North America conference, founder and president of the Tanka Society of America, cofounder of the American Haiku Archives, and founder/director of the Seabeck Haiku Getaway and National Haiku Writing Month. His haiku, essays, and reviews have appeared in hundreds of journals and anthologies in at least 22 languages. Michael lives in Sammamish, Washington, where he enjoys racquetball, skiing, travel, and reading. His website is www.graceguts.com.