Some context first:

I am interested in the etymology of the Romanian word gâdila/gîdila ("to tickle; the â/î variation is only graphical: it's /ɨ/, the close central unrounded vowel which in Romanian usually reflects etymologically a Latin i, e or a, a Slavic ъ or the same Turkic vowel). I see here that the origin can be Turkish (namely Ottoman Turkish). The idea of an Ottoman origin seems very odd to me because, along with other bodily-related words, it is a very familiar/basic word, unlike other clearly Ottoman terms (usually related to politics, military or cuisine); it is also associated to a (maybe) regional onomatopoeic, exclamatory expression (gâdi-gâdi!), a kind of song really, used when playing with small children and anticipating a jocundly tickling gesture with a finger. This apprehension is confirmed by the fact that the related Albanian term gudulis is also considered a possible onomatopoeia. Beside Turkish gıdıklamak (for which I don't have a wiktionary or such other link, but for which google search gives illustrative findings), the term is also present in Bulgarian, гъделичкам (gadelytchkam), but for which a Slavic root is invoked: gъdьliti, with correspondents at least in Polish.

I have no other sources on the Slavic connection, but it seemed more probable at first glance (although, considering the reliability of Wictionary etymologies, I wouldn't be surprised if the Polish/Western Slavic forms are not confirmed, and if the Old-Slavic root is just deduced by Bulgarian linguists from the extent Bulgarian form). Slavic/Slavonic imprint on Romanian is old enough to explain the origin of a basic term like "to tickle" - a scenario that can be imagined for the Albanian form. The root could thus be considered part of the Balkanische Sprachbund that has also affected Turkish.

The same logic of a Balkanische Sprachbund status seems valid even if the root is not Slavic (absent in all Slavic languages excepting Bulgarian) and, centered on the Balkan area, could be imagined of Thracian or other IE substratum origin.

I was about to become satisfied with my speculation—although the initial Romanian skepticism vis-à-vis an Ottoman origin for that intimate Romanian word can be reversed against the Balkan and IE origin of the Turkish word. But then I have discovered that the Turkish form spills to the east into Azerbaidjan as qıdıqlamaq and even Turkmenistan! I have no other sources/links, but using Google translate I have found that to tickle is gyjyklamak in Turkmen.

My initial theory crumbled, and it went into smoke completely when I looked up Google Translate for the same word in Mongolian: it's гижиг, гижигдэх (gijig, gijigdakh). This seems like a rather clear East>West transition of a form giji>gidi. It seems to me that even in Korean there is this g-j-l structure: 간지럼 ganjileom!!!

(This extraordinary large expansion to the far east even outside Turkish languages may paradoxically accommodate better my initial skepticism against an Ottoman origin, because a much older Turkic one can be more probably supposed, like an (Old) Bulgar, Cuman, or Pecheneg, something that is normal in Romanian for other Turkic terms that are too old to be Ottoman, like baci and cioban (both meaning "sheperd"), or the toponym Teleorman.)

But what about the expansion as far as Korean of this root g-d/j-l? Is this real?

These are the words, in one expanded list that goes beyond the Turkic family:

  • Romanian: gîdila, gâdila
  • Bulgarian: gadelitchkam, гъделичкам
  • Albanian: gudulis
  • Turkish: gıdıklanmak
  • Azerbaijani: qıdıqlamaq
  • Turkmen: gyjyklamak
  • Kyrgyz : kıtıktoo – kıtıgıloo
  • Kazakh: qıtıqtaw
  • Uzbek : qitiq, qitiqlash
  • Mongolian: gijig, gijigdakh
  • Korean: 간지럼 ganjileom
  • Hebrew: לְדַגדֵג <> lְdַgdֵg
  • Arabic: دغدغة <> daghdagha
  • Georgian: ტიკტიკი <> t’ik’t’ik’i

More than half of these are based on Google Translate. The last three entries are odd - they look less like verbs and more like the corresponding onomatopoeic exclamation (Romanian gâdi-gâdi mentioned above, or English coochy-coo mentioned in a comment below). It would be great if someone was able to confirm.

  • Comments are not for extended discussion; this conversation has been moved to chat.
    – curiousdannii
    Commented Jan 21, 2023 at 13:43

3 Answers 3


I think I have an answer, based on the comments under the question.

The answer is basically "no". The Romanian and Bulgarian form is common, and is of Slavic origin, and has to be considered separately, while the Albanian and Turkic forms are also to be considered separately from each other.

  • The Romanian verb a gâdila “to tickle” – which also includes regionally the form a gâdilici (accent on last i, with the same meaning) and the plural noun gâdilici (accent on second i, “itchy feeling”, plural, like here, point 2) – is the same as the Bulgarian gadelicikam, “I tickle” (since verbs have no infinitive in Bulgarian)
  • The similarity between the Bulgarian gdelicikam and the Turkic gıdıklanmak is superficial and coincidental: the k in Bulgarian is part of a suffix and the l is part of the root, while in Turkish the opposite is the case.
  • Bulgarian-Romanian form is expected (predicted) within IE family as Slavic. “To tickle” has a Proto-Indo-European root *geyd- (“to sting; prick; tickle”) which resulted in Germanic *kitōną > kitilōną (hence German kitzeln, Icelandic kitla, etc, English tickle resulting by methatesis). A linguistic rule (Grimm’s Law) states that Germanic k-t corresponds to Slavic (and other branches) g-d.
  • Bulgarian-Romanian form is to be separated within the Balkan area not just from the Turkic forms, but also from the Albanian gudulis, “to tickle” which corresponds etymologically to rhotacized Romanian gudura (only a reflexive in fact: a se gudura), meaning (of dogs) to fawn (wag its tail to show devotion), figuratively to fawn (seek favor by flattery and obsequious behavior), "get under one's skin", in fact very specifically the shaking movements of a dog’s body touching a person’s legs with the intention of gaining attention and benevolence. --- A se gudura (which can be related to the verbs a zgudui, zguduire, "to shake, tremble, convulse", of obscure origin, and a scutura, "shake, agitate, rock, remove dust", possibly of Latin origin) means literally, just about dogs: “to rub one's body against someone else's lower legs”. Just as an adult may try to humor and gain the sympathy of a small child by the primitive, almost animal gesture of tickling, a dog may try to “tickle” its master’s legs. It is a most elementary gesture of affection. Romanian gudura is a type of tickling, or the other way around.
  • The presence in Romanian of two IE-related but separate forms (gudura and gâdila) reflects the fact that the Albanian-Romanian one (gudulis-gudura) is non-Slavic, while the other (gâdila-gadelitchkam) can only be Slavic, the secondly-probable ancestry for it, that of a Balkan substratum, being already taken by gudulis-gudura.

The "Balkan" forms can be explained completely and independently within the IE and Slavic domains, no matter the significance of the fact that the Indo-European base kt/gd seems to have a formal similarity with the Turkic one (gd/qj), also the Semitic and even the Georgian (!) - given that semantically a very primitive, even ethologically-significant action may be involved.

UPDATE after reading Dan Ungureanu's Istoria limbii române, Ed. Corint 2024, which tries to show that a long list of Romanian words with uncertain but possible Slavic or substratum origin have in fact Italian and other Romance (Piedmontese, Lombard, Romansh, Occitan etc) equivalents, and are thus of late (local, dialectalized) Latin origin. On a gâdila he points out even a larger areal:

enter image description here

(For many words he has found more localized equivalents, based on rather specific innovations in time and place --namely: of late Latin & early but already Christian Romance of the North-West Italian, alpine region-- something which leads him to a revolutionary hypothesis for the origin of Romanian: more to the west, before its eventual Balkan dwelling.)

This doesn't change the overall conclusion, only removes the need for a Slavic root.

  • 1
    Correction point: The root *kitōną is Germanic (after Grimms Law), not IE. Commented Jan 17, 2023 at 11:12
  • @SirCornflakes - I will correct it. Thanks.
    – cipricus
    Commented Jan 17, 2023 at 11:13

In Korean, 간지럼 ganjireom is the noun referring to a sensation of feeling 'ticklish' but also 'itchy'; the verbal adjective / descriptive verb 간지럽다 ganjireopda is its basic lemma form.

The related verb 간질이다 ganjirida, which is analysed as 간질 + -이다, is first attested in the 18th century as ᄀᆞᆫ지러이다 gawn-ji-reo-i-da. Both are believed to be related to 가렵다 "to be itchy" garyeopda which is attested (in the spelling ᄀᆞ렵다 gaw-ryeob-da, from the 16th century). The -지- infix of 간지럽다 may be a causative infix, which is productive in modern Korean too; however I've not seen much to explain the ㄴ /n/.

This semantic family has been linked to 긁다 geulkda "to scrape", attested (as 긁다〮) in the 월인석보 Worin seokbo of 1459.

The motion of tickling someone in modern Korean generally requires a verb, such as 간지럼을 태우다 ganjireom-eul taeuda, literally to "burn/aggravate/spark an itch/tickling sensation".

This connection between tickling and itching is found in Chinese too. In modern Standard Mandarin Chinese, 癢/痒 yǎng refers to this same sensation espoused in 간지럼, and also requires a verb (e.g. 搔癢/搔痒 sāoyǎng which can mean "to scratch an itch" as well as "to tickle"; also 撓癢癢/挠痒痒 náoyǎngyang "to scratch/disturb/bother a little itch (reduplicated)").

Something that may be more interesting for those looking for such connections: colloquial Mandarin 胳肢 gézhi (Mainland Mandarin) / gēzhī (Taiwan standard Mandarin) is onomatopoeic, but is (now) attested as a verb. From one of the Health pages of the People's Daily, no less:


Rúguǒ yǒurén gézhi nǐ de yèxià huòzhě jiǎoxīn, xiāngxìn hěnduō rén dōuhuì rěn bùzhù yǎng dào xiào chū shēng

If someone tickles your armpits or the soles of your feet, (I) believe many people can't help itching so much [that they] laugh out loud.

There has been a claim that this may also be related to the Manchu gejihešembi "to tickle".

However, this word "tickle" can also be linked to the colloquial word 胳肢窩/胳肢窝 (gāzhiwō [Mainland]/gēzhiwō [TW] / keq tsr u [Shanghainese]) "armpit", literally 胳肢 "tickle" + 窩 "nest". However, the two syllables that actually go into "tickle", mean "arm" and "limb" respectively in Classical Chinese, and are both related to Tibetan ལག་པ lag pa and Burmese လက် lak. This in itself is also related to 腋 , the Classical Chinese for "armpit" (Tibetan l > Sinitic y being a common correspondence).

Could we thus say that in actual fact Sino-Tibetan "arm" *lak ⪤ *C-yak ~ yak was something that generalised through 胳肢窩 "armpit" and Manchu across to the medieval Balkans, possibly through the Han dynasty's use of "tickling torture"[citation needed] and the "Tatar yoke", as well as through 12/13th-century Mongol contact into Korean?

Despite such Altaic-baiting, I would have to agree that 'sound symbolism'-type phonemes independently formed with /k-t ~ g-d ~ g-j/ often involved in "tickling". Comparing some other Sinitic varieties:

  • Cantonese: 擳 zit1
  • Min Dong: 扱力 gĭh-lĭk
  • Taiwanese Hokkien (Min Nan): 擽呧 ngiau-ti

... and others around East and Southeast Asia:

  • Japanese: 擽る (くすぐる, kusuguru)
  • Mongolian: гижиг (gizhig)
  • Evenki: кичикадя-мӣ (kichikadya-mii)
  • Northern Vietnamese: cù
  • Central & Southern Vietnamese: chọc lét
  • Khmer: ចាក់ក្រឡេក chakkralek
  • Thai: จี้ (jîi)

... and those specifically from Austronesian:

  • Indonesian: menggelitik ('gelitik' being the root form)
  • Tagalog: kumikiliti ('kiliti' being the root, I believe)
  • Amis: kiri'
  • Malagasy: manakibokibo
  • Maori: ngāokooko, tōkenekene

We thus do see a large number of velar and alveolar/dental stops, but not exclusively. The vowels do not follow many patterns, and neither do the order and the variety of the stops.

Tibetan presents an interesting case: Written Tibetan uses ཀི་རྩི ki rtsi as an adjective "tickling", with the verb using བྱེད byed "to do". This has an orthographic 'r' but lacks a transparent link to "arm" or "armpit", and is now often written ཀི་ཙི ki tsi.

Although there is some evidence that colloquial Mandarin 胳肢 gézhi/gēzhī "tickle" is connected to Manchu, based on the fact that it is distinctly modern (from the late Ming dynasty / early Qing, e.g. in the 醒世姻緣傳 Xǐngshì Yīnyuán Zhuàn), the connection to "armpit" and to Classical Chinese 胳 "arm" and 腋 "armpit" is difficult to ignore. This remains an interesting field of research.

  • To be clear, although you mention sound symbolism right before the list, the Sinitic, Pan-Asian and Austronesian examples you give are all verbs meaning ‘to tickle’, rather than onomatopoeia used when tickling someone, right? Commented Jan 20, 2023 at 12:29
  • Yes, they are mostly verbs (although whether they are the most common verbs, that remains to be seen; cf. the Manchu situation with gejihešembi vs gejilembi).
    – Michaelyus
    Commented Jan 20, 2023 at 15:04
  1. It is widely believed that ideophones do not conform to sound-change, thus for example hahaha besides cackle, giggle, Korean kekeke and Orangutan ook-ook. This is no excuse to do away with the comparative method, but it's something to keep in mind.

  2. Proto-Germanic *kitilōną was already mentioned as inherited PIE derivative (@cipricus, above), but the response is missing details

  • a) E. tickle appears to exhibit dissimilation same as i. Ger. Zwetschke, Quetsche "quince", ii. tip, Ger. kippen "tip over", cf. tip "junk yard", Ger. Müllkippe; Ger. Tuppe "tub", Kiepe "id.; tray", Butte "id.", iii. Ger. Zipfel "tip", Tannzapfen "fir cone", Kipferl "id.", E. chip, L. cippus etc. (ignoring pine for the moment). This vexilation is robust with respect to semantics, yet far from regular.

    Suffice to say the English etymology rests on tick "light touch", which may be onomatopoeic in origin, like tick-tock, tap. See also Ger. Zecke "tick" vs. Ger. Käfer "chafer", E. coppe "spider", equivalently chicken vs. Dutch kip and indeed 雞 Old Chinese /*kˤe/ "fowl", 鳥 OC /*tˤiwʔ/, even Mi'kmaq go'qoli'gweji'j "chick".

  • b) Cli-tor-is (?) translates German Kitzler, literally which tickles, but hood is collexified and should be assumed original, for the sake of the argument. Although I have not heard of tap in this sense I have once heard Ger. Flansch ("wahrscheinlich an spätmhd. vlansch ‘Zipfel’ anzuschließen.", DWDS/Pfeifer).

  • c) see cuddle.

  1. In this view, Ger. küssen, Kuss "kiss" may possible be related. Pfeifer confirms the "sound-symbolic expression, evading regular sound-change" (DWDS: küssen; translation mine). In particular, Gothic kukjan "kiss", English cooky (cp. Fr. baiser), cooch, hooch, hooha, twat and whatnot escape the historically minded linguist.

  2. All this implies is that the comparative method as applied above is not mighty enough. See also E. giddy, allegedly from *god ("god", Wiktionary), but the etymology of this root is uncertain, to say the least.

There can be no conclusion. The question is well justified, but it is too broad. Yes, it is a "stretch".

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.