Onomatopoeia in everyday Japanese

« previous post | next post »

Of all the languages I know, Japanese is the richest in onomatopoeia (and poorest in swearing).  Here's a brief introduction to reduplicative sound symbolism.

‘Pachi pachi’ or ‘kachi kachi’? Japan launches foreigners’ guide to tricky world of onomatopoeia

As foreign population reaches record levels, the western prefecture of Mie has compiled a guide for those who need it

Justin McCurry in Osaka
The Guardian (Tue 14 Nov 2023)


It is a linguistic trap few learners of Japanese have avoided: declaring yourself pera pera (fluent in a language) when you’re really peko peko (hungry); or breaking into applause (pachi pachi) when the dentist asks you to kachi kachi (bite repeatedly).

Navigating the rich and varied world of Japanese onomatopoeia can result in laughter and mild embarrassment, but the words can also be a quick and effective way to get through to a friend or colleague.

As Japan’s foreign population reaches record levels, lifted by the arrival of more people to fill a gaping hole in the labour market, volunteers in the western prefecture of Mie have compiled a guide to commonly used onomatopoeic words for language learners.

The book, entitled E Kara Oto ga Wakaru Hon (understanding sounds using pictures) was the idea of Masao Hara, the deputy head of a nonprofit in the prefecture whose interactions with non-Japanese convinced him the guide would come in useful.

It contains a host of words that can be used in everyday situations, such as a visit to the doctor, who might hear of their patient’s throbbing (zuki zuki) ankle or pounding (gan gan) headache, a piri piri (stinging) insect bite or muzu muzu (scratchy) throat.

The book contains a list of 100 words divided into categories —  from actions and emotions to the weather and descriptions of inanimate objects.

Wan Fang, a Chinese resident, said the guide had already made her job at a supermarket a little easier.

“When I was told that the floor was tsuru tsuru in Japanese, I didn’t know what it meant, but when I saw the illustration in the book, I instantly understood that tsuru tsuru means the floor is clean or slippery,” Wan told the newspaper.

The print run of 1,000 copies is expected to find a keen readership among students attending local Japanese language schools – as of January, Mie was home to 31,000 foreign residents. Nationwide, the non-Japanese population reached a record 3.2 million last year, according to the immigration services agency.

The guide only scratches the surface, however. There are said to be more than 1,000 onomatopoeia in Japanese – enough to make most language learners come over all fura fura (dizzy).

Not all onomatopoeia reduplicative.  Some of my favorite colloquial expressions in Mandarin consist of other types of onomatopoeia, for instance, pàng dūdū 胖嘟嘟 ("chubby") and shǎbùlèngdēngde 傻不愣登的 ("daffy; clueless; dazed; stupid")


Selected readings

[Thanks to Don Keyser and June Teufel Dreyer]


  1. Peter Taylor said,

    November 15, 2023 @ 9:25 am

    Are any of these actually onomatopoeia?

  2. satkomuni said,

    November 15, 2023 @ 10:49 am

    Peter Taylor, you're right. There is another word for these reduplicative "conceptual" sounds, words to "imitate an idea" rather than "onomatopoeia"'s sound imitation, and I have been trying for a long time to remember what the word is. It is of similar length to "onomatopoeia" and also starts with o or oe or similar.

  3. Quinn James said,

    November 15, 2023 @ 11:10 am

    @satkomuni you might be thinking of the term "ideophone," covered in the "Waza waza" link in Selected Readings.

  4. Andy said,

    November 15, 2023 @ 11:18 am

    @satkomuni: Ideophones.

  5. rpsms said,

    November 15, 2023 @ 11:19 am

    Probably thinking of Ideophones, which are a subset of onomatopoeia (I think)

  6. Jonathan Smith said,

    November 15, 2023 @ 2:33 pm

    paku-paku 'gobble-gobble' or sth., whence "Pac-man" or so they say.

    "phonestheme" could refer to the sound sequences involved — I suppose these Japanese words are as a whole phonesthemic. And "sound symbolic" / "ideophonic" as noted above. Onomatopoeia is I suppose to be regarded as a (relatively iconic) variety of sound symbolism.

  7. cliff arroyo said,

    November 15, 2023 @ 3:50 pm

    The term mimetic adverbs also used to be used.

  8. Chris Button said,

    November 15, 2023 @ 3:55 pm

    @ Cliff Arroyo

    Yes, a book I have on the topic provides an English translation of its title as a "A Practical Guide to Japanese-English Onomatopoeia and Mimesis"

  9. Jim Breen said,

    November 15, 2023 @ 5:28 pm

    Strictly speaking, they are mimetic terms, but they are often referred to as onomatopoeia. In the JMdict/EDICT project we lump them together and tag them as "on-mim". There are over 1,200 entries tagged at the moment; most of them are cases of mimesis. They are usually adverbs; many with the と (to) particle, and many are used as verbs with the する (suru – to do) auxiliary.

    A good reference is Hiroko Fukuda's 1993 book "Jazz Up Your Japanese with Onomatopoeia". She covers about 250 terms; most of them mimetic.

  10. Jonathan Smith said,

    November 15, 2023 @ 5:55 pm

    It seems that "mimesis"/"mimetic" became associated with the Japanese phenomenon in particular at some point — maybe not the greatest general-purpose term as sounds can't be imitative of, say, the way things are or feel in any concrete sense, and the sound associations exploited are often (or often seem to be) language-specific rather than tied to universal features of the human sensory apparatus.

  11. AG said,

    November 16, 2023 @ 3:56 am

    I've always enjoyed the fact that one of these is the literal sound of silence. (shiiiin, シーン)

  12. Michael Watts said,

    November 16, 2023 @ 6:03 am

    I have to second Peter Taylor's implicit question – why are any of these terms being referred to as "onomatopoeia"?

    Similar terms are also labeled "onomatopoeia" in the Waza-waza link, and there is no explanation of why.

  13. Chris Button said,

    November 16, 2023 @ 9:51 am

    Giseigo / giongo — onomatopoeia
    Gitaigo — mimesis

    I suppose that's the best we can do. It reminds me of the whole debate over whether "etymology" can be used for "jigen":

    Gogen — etymology
    Jigen — character etymology (I have no problem with that, but there was a debate on LLog about it a while back, and I feel many would go with "character origins" or something and not allow the word etymology to be (ab)used in this way).

  14. Forte said,

    November 16, 2023 @ 9:01 pm

    Any link to download the guide book?

  15. Michael Watts said,

    November 16, 2023 @ 11:26 pm

    The wikipedia page on "Ideophones" does not fill me with confidence. It appears to rest on the theory that ideophones do not obey the constraints that apply to "ordinary words". It gives the example that "zigzag" might at first appear to be an ideophone, but it cannot be one because it participates in the normal English inflection of verbs, whereas if English grammar called for "The rabbit zigzag zigzag across the field" instead of "The rabbit zigzagged across the field", that would be evidence of ideophones in English.

    This feels to me more like an "explanation" of the fact that certain Japanese words have a reduplicated form.

    But it does raise an obvious question: we have the example "waza-waza". It is clearly an adverb, referring to the manner in which an action is performed. Does it obey the Japanese rules of adverbial inflection? I don't know the answer, but adverbs don't do a whole lot of inflection, so I tend to suspect that the answer is yes. At that point, we've (hypothetically) lost the only piece of evidence that would have called waza-waza an "ideophone" instead of a "word".

    Attacking the idea from the other direction, here is a lyric from The Nightmare Before Christmas:

    is it rotted and covered with gook?

    [The word rhymes with "look".]

    I once wrote to Mark Liberman asking about the phenomenon here, but there was no response. The lyric is of interest to me because it is very easy to understand despite the fact that I have no other exposure to the word "gook", and I assume this is the case for most English speakers. "Gook" is present in the relevant sense in wiktionary [cited to 1983!] and in Merriam-Webster, though not in the Cambridge dictionary, but I can assure you that my experience of hearing the word in the song was that I had never encountered it before. This did not impede understanding because the contextual cue from the rest of the lyric is so strong.

    I tend to agree that the sounds used to form this nonce word are not coincidental. This would make it a good candidate for the concept "ideophone", except that it's also clear that it obeys all the normal English rules of being a noun. By wikipedia's standard, that means it cannot be an ideophone.

    Can wikipedia's definition actually be defended? And if not, what is the meaning of this category supposed to be?

  16. Michael Watts said,

    November 16, 2023 @ 11:29 pm

    (Note also that "The rabbit went zigzag zigzag across the field" is perfectly normal English, but seems to conflict with wikipedia's analysis.)

  17. AG said,

    November 17, 2023 @ 5:33 am

    minor point, but "gook" as in gunk or slime is totally familiar to me (west coast US, born 1975)… google "gook all over it" to see many examples of usage unrelated to the nightmare before christmas.

  18. Chris Button said,

    November 17, 2023 @ 10:18 am

    @ Michael Watts

    I think ideophone or mimesis works fine for gitaigo.

    As for verbing them, I think it is largely irrelevant since the Wikipedia article was clearly not written with a consideration of the differences between English and Japanese in how verbs are formed. That's to say you can't just add a suffix like English -ed on the end in Japanese and call it a verb.

    Compare say "pera pera hanasu" (fluently "pera pera" speak "hanasu") with "blabbered"

  19. Rodger C said,

    November 17, 2023 @ 10:57 am

    AG, to me (West Virginia, b. 1948) "gook" in this sense sounds very old-fashioned. Plus, of course, it's also an insult for Asians, so perhaps especially avoided on the West Coast.

  20. cliff arroyo said,

    November 17, 2023 @ 11:20 am

    "gook? The word rhymes with "look"."
    "insult for Asians"

    My version of 'gook' as a slimy substance rhymes with 'kook' and not 'cook'.
    As an insulting term for Asians I've read it but not sure if I've heard it (I'm almost certain I've never heard it in real life, maybe in a movie… anyway while reading my mental pronunciation for it also rhymes with 'kook'.

  21. Scott P. said,

    November 17, 2023 @ 11:55 am

    My version of 'gook' as a slimy substance rhymes with 'kook' and not 'cook'.
    As an insulting term for Asians I've read it but not sure if I've heard it (I'm almost certain I've never heard it in real life, maybe in a movie… anyway while reading my mental pronunciation for it also rhymes with 'kook'.

    Whereas in my ideolect 'gook' as slimy rhymes with book or cook and 'gook' the slur rhymes with 'kook'.

    And yes, I (and folks I interact with) use the word with its former meaning all the time.

  22. Jonathan Smith said,

    November 17, 2023 @ 3:48 pm

    Wikipedia says that "[w]hile English does have ideophonic or onomatopoetic expressions, it does not contain a proper class of ideophones", which seems right; English words like zigzag and the like are ideophonic i.e. exploit phoneme-level sound-meaning relationships, but don't constitute a discrete and highly coherent lexical class as with Japanese gitaigo. Granted there are tricky cases… in say "running helter-skelter…" we get very close to the Japanese-type situation where sound-symbolic words are stuck on "adverbially".

    Re: terminology, the point is not to translate the Japanese technical term gitaigo, for which 'mimetic (adverbs)' or whatever is fine (even if say 'manner-mimetic' would be a closer calque), but to discuss the phenomenon of sound symbolism generally.

  23. Chris Button said,

    November 17, 2023 @ 5:21 pm

    @ Jonathan Smith

    I don't think applying English syntactic structure to Japanese works very well (one of the reasons I avoid studying syntax like the plague)

    Regarding translating terms, phonaesthetic is far less applicable than mimetic (which is the standard translated terminology)

  24. Chris Button said,

    November 17, 2023 @ 5:22 pm


  25. Chris Button said,

    November 17, 2023 @ 5:26 pm

    Sorry "phonaesthetic" without the "o"–got it right the first time.

  26. Chris Button said,

    November 17, 2023 @ 8:06 pm

    If anyone's interested in a similar phenomenon in another language, it's worth taking a look at Peri Bhaskararao "The process of chiming in Tiddim Chin" (1989).

  27. Michael Watts said,

    November 18, 2023 @ 5:26 am

    but don't constitute a discrete and highly coherent lexical class as with Japanese gitaigo. Granted there are tricky cases… in say "running helter-skelter…" we get very close to the Japanese-type situation where sound-symbolic words are stuck on "adverbially".

    But that's what I was asking. Do they constitute a discrete lexical class in Japanese? What makes them different from other adverbs / adjectives / nouns / verbs?

    And is kachikachi, which by its gloss "bite repeatedly" appears to be a verb, really part of a discrete lexical class with wazawaza, which isn't one?

    Jisho does not have a "bite repeatedly" gloss for kachikachi, but it does list 2 glosses which are adverbs and 3 which are noun/adjectives [all English glosses are adjectives, but the part of speech is listed as simultaneously adjective and noun], all 5 of which are marked "onomatopoeic or mimetic word".

    In fact, the jisho entries for kachikachi do severe damage to wikipedia's theory that ideophones are characterized by failing to participate in the normal structure of the language – the part of speech listings are pretty explicit about the grammatical structure that the senses of kachikachi require, being "adverb taking the 'to' particle," "adverb (fukushi)," "noun which may take the genitive case particle 'no'," and "na-adjective (keiyodoshi)."

  28. Jonathan Smith said,

    November 18, 2023 @ 3:50 pm

    The group seems coherent enough that it would be intuited as a class, thus books collecting 1000's of such items? But this is only my vague impression. Some obvious largely shared features appear to be (1) phonological template moraAmoraB-moraAmoraB (i.e. "base" + "reduplicant"); (2) no etymologies in the traditional sense rather arising ad hoc via exploitation of sound-meaning relationships; (3) no inflection as with the core Japanese adjectives and verbs; (4) "boot-strapping" into syntax via devices like ABAB [to] action verb 'verb [with a] ABAB', ABAB + forms of suru 'to do/go ABAB', etc.

  29. Jonathan Smith said,

    November 18, 2023 @ 3:53 pm

    @Chris Button "I don't think applying English syntactic structure to Japanese works very well" >> The Point of the discussion above — "word classes" should be defined on internal grounds

  30. Chris Button said,

    November 18, 2023 @ 9:09 pm

    @ Jonathan Smith

    It's actually the loose throwing around of the words "adjective" and "adverb" that is bugging me. The word classes cannot simply be borrowed over like that (beyond easy definitions in learners' textbooks). But I am no syntactician; I jist speak Japanese poorly. If we're lucky someone knowledgeable on the matter might chime in.

  31. Chris Button said,

    November 19, 2023 @ 12:33 am

    To put it very crudely:

    “-i” adjectives are inflecting verbs
    “-na” adjectives are non-inflecting nouns

    I agree you do need to define things on internal grounds, but you need to do it without applying an external framework.

    Meanwhile, I’m sticking to phonology :)

  32. Jonathan Smith said,

    November 19, 2023 @ 11:21 am

    Hmm… re: "adverb," I avoid it above… re: the two types of "adjective," the question is hardly so simple and anyway irrelevant here. But one does note in "Japanese adjectives are lie" a parallel to your phonological "vowels are a lie," etc., i.e., an abiding attraction to the theoretical/(pseudo-)metaphysical as opposed to the factual/descriptive.

  33. Chris Button said,

    November 19, 2023 @ 12:08 pm

    @ Jonathan smith

    "vowels are a lie,"

    I am quite sure I have never said such nonsense! I think you might be confusing surface phonetics with certain underlying phonological analyses.

    "Japanese adjectives are lie"

    That depends on your definition. But what I am saying is by no means particular to Japanese at all, and I certainly cannot take credit for the insight!

    and anyway irrelevant here

    Actually, I think it is the whole point, but maybe we are talking at cross purposes now.

RSS feed for comments on this post