Since it's written in Hiragana, I presume it is likely not a recent loan word. However, its pronounciation bears resemblance to "軲轆", a Mandarin word meaning wheel--similar to ぐるぐる's meaning of "round" or "round-and-round".

According to Wiktionary, the etymology of "軲轆" is apparently quite ancient, deriving from Tibetan or even Proto-Indo-European. Does ぐるぐる also share Proto-Indo-European roots (despite Japanese not being part of that language family)?

  • 1
    Is there a reason to think it's not an ideophone?
    – Draconis
    Commented Jun 25, 2020 at 20:54
  • 1
    軲轆 gūlu [kʰulu] (Mandarin, colloquial) 1. wheel; 2. to roll. ぐるぐる [ɡɯ̟ᵝɾɯ̟ᵝɡɯ̟ᵝɾɯ̟ᵝ] "turning round and round". Do you really mean the two words really sound similar? Did you know that all those Japanese reduplicated 2×2 syllable words always spelled in Hiragana are onomatopoeia, sound symbolism, words that phonetically imitate, resemble, or suggest the sound, the feeling, or a thing that it describes. Have you read this article? en.wikipedia.org/wiki/Japanese_sound_symbolism
    – Yellow Sky
    Commented Jun 25, 2020 at 21:56
  • This should be asked at the Japanese Language site. :)
    – curiousdannii
    Commented Jun 26, 2020 at 0:38
  • 2
    I'd say it's on-topic here. Commented Jun 26, 2020 at 6:33
  • 1
    Is there evidence that the similarity is entirely coincidental? If ぐるぐる is onomatopoeia, does 軲轆 also share certain onomatopoeic roots? It does not seem objectively obvious that something turning makes a “guruguru” or ‘gulu” sound. What I’m interested in is whether there is a connection in the onomatopoeia itself, assuming that it is entirely onomatopoeic. Commented Jun 26, 2020 at 13:37

1 Answer 1


Chinese gūlu < *kʰaːroːɡ is probably not onomatopoeic, especially if it came from (PIE) *kʷékʷlos "wheel" (related to English "circle") as Bauer suggests.

Japanese guruguru "round and round" is standard onomatopoeic reduplication, like boroboro "old and tattered", wakuwaku "excited" etc. It's not very old at all; the oldest citation in the Kokugo Daijiten is 1672 (in guruguru-wage, a wrapped bun hairstyle).

guruguru must be related to gururi (1603) and kururi (1151), both with a meaning of "turning around, going around". An intermediate form may be kururi-kururi (ca. 14c) by dropping the -ri, cf. guru-tto < gururi-to. The morpheme in meguru (8c), "surround, go around", is probably the same, and with normal Japanese vowel mutation, also the one in korogaru "roll", koro-koro "rollingly" etc. (note kururi and korogaru are written with the same kanji, 転, Mandarin zhuǎn).

In summary, guru-guru is a late instance of a whole family of Japanese words circling around a core morpheme kuru/koro, with the meaning of "around, round and round, roll".

A connection of this productive "around" morpheme to Chinese *kʰaːroːɡ "wheel" (and thus to PIE) cannot be discarded, but I don’t think there’s any concrete evidence for it. The old Japanese orthographies for this morpheme didn’t use the characters 軲轆, but rather kana, or semantic characters like 回. This means that, if kuru comes from *kʰaːroːɡ, the transmission would have to have been ancient and prehistoric, so that by the point writing arrives, the connection was not apparent to Japanese speakers anymore.

Another possibility is that kuru is a native J morpheme but its productivity was later reinforced by gūlu in some form. But again, if that happened, I would expect more historical writers to have represented it with 軲轆 (or at least 轆 -ro/-ru, which is used in Japanese, marginally); but that didn't happen.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.