The estimable John Emerson of Idiocentrism has a post pointing out that the Inuit and the Turks have similar words for kinds of boats, both of which have wound up in English:

The word “kayak” came into the European languages in the seventeenth or eighteenth century, probably brought from Greenland by Dutch or Danish whalers. Some version of this word is now used in most European languages for any boat built on the model of Inuit (Eskimo) skin boats… Through Turkish, in the 1500’s the “caique” finally appeared in Italian as the name of a boat found on the Adriatic, and the name spread from there to the other European languages, finally reaching Sweden in the 1700’s. There the boat names “caique” and “kayak” met – albeit as the names of boats of entirely different kinds.
The fact that the Turks and the Inuit both had boat names pronounced something like “kayak” seems at first to be a pure coincidence of the type that cranks love and linguists dread. However, a good case can be made that the Inuit and the Turkish words were etymologically related, and that the word probably originated in Turkish.

His post makes a case for that relationship, and he has added a recent query:

According to commentor Ruth at GNXP, Turkish qayiq also means “ski”, and is an inflected word derived from the root kay– “slide”. She wonders whether the Inuit qayaq is also an inflected word. If not, it would seem to be a borrowing from Turkish into Inuit, whereas if it’s inflected in both languages, perhaps there was a common Turkish-Inuit ancestor.

Anybody know enough Inuit to be able to answer this?

Incidentally, the linked GHXP post mentions that John’s book Substantific Marrow is out, with sections on “The Back Door of Europe” (on the Baltic-Black Sea corridor), “The State”, and “Love or Money”; he adds: “By Christmas I should have a third book out, about philosophy, economics, and temporality. Sometime next year my book on Inner Eurasian history should be out; this book should be of interest to many people here. I’m going to be spending the next several years gathering and finishing up stuff I’ve been working on since about 1985.” Sounds promising!


  1. I dunno, I just have trouble getting Inuit…

  2. Also the Japanese kaji 舵 (rudder, helm)
    http://starling.rinet.ru/cgi-bin/response.cgi?single=1&basename=\data\alt\turcet&text_recno=914&root=config and then click “Altaic etymology”
    Another correspondence is dāg (mountain) (proto-Turkic)
    and (nuna)tak (ice mountain).

  3. SnowLeopard says:

    I’ve only ever seen ‘qayaq’ as the generic word for “boat” in various Inuit languages, but I can check a few sources when I get home and report back. I’m aware of only a few other apparent cognates to Turkish, such as words for “mother” and “father”, but lots of common words seem to have no match. Turkish also seems to have a radically different grammar from Inuktitut/Yup’ik/etc., so I’d be skeptical of a deeper connection.

  4. John Emerson says:

    If Turkish is like Mongol, and I suspect it is, qayiq would be a deverbal noun derived from the verb qai / qay “slip, slide”. Agglutination is unlike inflection in that the root is a full word in itself, and is not truncated or transformed as it is when inflected.
    Disclosure: I don’t know Turkish, and only a little Mongol.

  5. There is a strong assonance with “coracle”, apparently from welsh (cwrwgl, believe it or not). These were used in pre-roman times; could this be english->roman->turkish?

  6. John Emerson says:

    In an early version of this piece I worked on “coracle” “carack” and “cog”, early European names for types of ship.
    I am aware that this is the kind of amateur speculation that drives linguists nuts. The Turkish-Inuit connection, however, is pretty reasonable.
    I looked up the Yakut, and (as LH has suggested) they apparently came to their present location rather recently (since 1000 AD, maybe since 1500 AD.) However, before ~1000 AD what we call Mongolia was Turkish (Uighur). The Mongols were north and east of there. How far north and east, I don’t know.

  7. SnowLeopard says:

    Fortescue et al.’s Comparative Eskimo Dictionary with Aleut Cognates offers the Proto-Eskimo form “qayaR”, with R apparently a continuant form of q, as the ancestor of the word “qayaq” in virtually every modern Eskimo language listed. ‘Qayaq’ isn’t obviously derived from ‘qai’ (Yup’ik, flat surface), or ‘qaiq’ (Inupiaq, to be smooth, flat), unless perhaps by way of ‘qai’ (Inupiaq, to come), ‘qaivartuq’ (Yup’ik, comes to the surface), ‘qaivoq’ (Greenlandic, gets there, is visiting). (Various sources.) I don’t know what else the original questioner may have meant by the word “inflected”.
    For those hunting for parallels, Jacobson’s Yup’ik Eskimo Dictionary gives ‘aana’, ‘aata’, as the Yup’ik for mother, father, which is not too far from what I guess to be acceptable Turkish words: ‘ana’, ‘ata’.

  8. peter desmond says:

    some summers ago, in andalusia, i was at a party talking to an inuit school principal from greenland, who had come to southern spain for the rugby (lots of english expats). we stepped out on the balcony for a cigarette. the village has had electricity for 30 years now, but one still gets a clear view of the night sky.
    not having previously met anyone inuit, especially in a medieval spanish morisco/mudejar hill town with whitewashed buildings and step streets, i cast about for something to say. seeing the big dipper midway up the sky, i pointed to it and said, “is that a constellation for you?” (it was a pretty safe conversational gambit. no one lives farther north than the inuit!)
    he got all excited, pulled me back inside, poured another glass of rioja, found a pen and a paper napkin, and drew me a picture of the inuit constellation.
    you know how europeanoids, or at least americans, think of the three stars as the handle and the four stars as the bowl? we typically think of it as scooping UP water. the inuit version of the constellation is exactly upside down from the way americans perceive it.
    the three stars (which in imagination are extended beyond the four stars as their foundation) represent the deck of the kayak.
    the four stars are a little upright post on the deck around which you wrap the rope attached to your harpoon. neat, huh? the stuff you learn at parties.

  9. Etienne says:

    Interesting. Assuming it’s a loanword, are there any similar such words in the languages spoken in Eastern Siberia (i.e. in the languages we may assume to have been contiguous to Proto-Turkic and Proto-Eskimo-Aleutian)?
    Also, I doubt we could establish which language the word originated from (and let’s bear in mind that the original donor language might be another (possibly extinct) language unrelated to Turkic or Eskimo-Aleutian): attempts to identify this on morphological grounds (i.e. the language where the word is relatable (language-internally) to other words/roots is assumed to be the original donor language) must take into account such things as folk etymology (whereby words are re-analyzed in a historically unjustifiable way) as well as the fact that over time connections between historically related words in a single language can disappear (LADY and LOAF are historically related, but this historical fact is not synchronically visible to native speakers).

  10. David Muscat says:

    The word kajjikk in Maltese means a particular light fishing boat. The word occurs as well in Sicilian. We must have got the word from Sicily but I can’t think of any connection with the Turkish word.

  11. The word caique passed from Turkish into various Mediterranean languages. Malta was never under Ottoman rule but the Ottoman’s controlled the E. Mediterranean during some periods.

  12. John Emerson says:


  13. caffeind says:

    If the words for “dog” in most of the languages mentioned are related, “boat” isn’t that hard to believe either.

  14. otterman says:


  15. Don’t know if this has anything to do with the matter at hand, but kajakas is Estonian for seagull (no etymology, alas):
    Proto: *kaja (-ka)
    English meaning: a k. of bird
    German meaning: Möwe; Larus
    Finnish: kajava, kajakka, kajova ‘Möwe; Larus’
    Estonian: kajakas (gen. kajaka), kajak

  16. marie-lucie says:

    If the words for “dog” in most of the languages mentioned are related, “boat” isn’t that hard to believe either.
    Just because words for the same thing are similar does not mean that the languages are related, as the same word can be “borrowed” (rather, adopted) into a variety of languages. For instance, words for ‘coffee’ everywhere sound similar, because they have been adopted into many languages from a single independent source – linguists would not say that those words are “related” since the source is from outside (most of) the individual languages, and has not developed from a “parent” of those languages.
    I don’t find “boat” comparable to “dog” or even to “coffee” as there is such a wide variety of boat types, made of different materials and built on different principles, which could have developed independently in different places (the extreme differentiation of dogs through breeding is relatively recent).
    About the Finnish word for ‘gull’: many, many words for birds originate from imitations of the cry of the bird in question. Raucous cries such as those of gulls or crows often have “ka” in them. But it is good to include such an example here because people often conclude from the existence of resemblant words with a definite meaning (here “boat type”) that this meaning must be the original one for the sound of the word, anywhere and everywhere. But the same combination of sounds can form words with very different meanings as well, in all sorts of languages.

  17. marie-lucie says:

    “kayak” as an “inflected word”:
    What is meant here is “derived word” – a word built on another word, as for instance in English “friendly” which is derived from (= built on) “friend” by the addition of -ly. There is no such thing as an “inflected word”, only an inflected form of the same word: for instance, “friends” is an inflected form of the word “friend” (indicating the plural), not a separate word including “friend” in it, as “friendly” is. Similarly, “friendliest” is an inflected form of the word “friendly”, not a separate word.

  18. Alper Ersoy says:

    “kayık” is indeed an inflected word. The root is “kay-“, which is a verb that means slip/slide. However, the commentor Ruth at GNXP is wrong, “kayık” does not mean ski in Turkish. That is “kayak”, which is also inflected from the same root.
    About the suffix “-ık”, it has a past tense meaning that something has happened or already in a state that is defined by the root: in this case it means slipped/slided. Therefore it also makes an adjective as well as a new noun. While the noun meaning represents the boat itself, the adjective meaning defines the shape of it.
    However the other suffix “-ak” is used to make a new noun that represents a tool in order to accompolish what the root means (which must be a verb). In this particular example: root is to slide or to slip. So “kayak” is a tool that you slip/slide with.

  19. Alper Ersoy says:

    Sorry, i wanted to mean “derived word”, not inflected. marie-lucie is right.

  20. Alas, this paper by José Andrés Alonso de la Fuente pretty much explodes this nifty idea.

    It turns out that Turkish kayık is either from the root kād- ‘turn, bend’ or the root kad- ‘wood’ (both with regular change /d/ > /y/) plus the semantically vague suffix -ğVk, where V is a vowel depending on vowel harmony. On the Inuit side, the underlying form is qan- ‘come/go’ + -yaq ‘locative’, as is shown by the Eastern Inuit form qainnat ‘kayaks’. What’s more, the Aleut word for ‘kayak’ has a different etymology altogether, so the word must postdate the Eskimoan-Aleut split. So the central /y/ is coincidence, and in fact the whole similarity is coincidental altogether.

    The paper also has a fine list of coincidences from various languages:

    (a) Japanese 名前 na-mae and English name or German Name;

    (b) Chinese (Mandarin) 餐厅 cāntīng ‘dining room’ and English canteen;

    (c) Jaqaru [Aymaran] aska and English ask;

    (d) Sanskrit da- and Cora [Uto-Aztecan] da- ‘give’;

    (e) Spanish lengua < Latin lingua and Hopi línga /leng’i/ ‘tongue’;

    (f) Mbabaram [Pama-Nyungan] dog and English dog;

    (g) Thai rim and English rim;

    (h) Teda [Nilo-Saharan] kulo and Spanish culo /kulo/ ‘anus’;

    (i) Songhay [Nilo-Saharan] mana and Quechua mana ‘no’;

    (j) Turkish tepe ‘hill’ and (Classical) Nahuatl tepē-tl ‘mountain’;

    (k) Turkish iyi and Japanese いい ii ‘good’;

    (l) Romanian fiŭ ‘son’ < Latin filius and Hungarian fiú ‘son, boy’ < Proto-Finno-Ugrian *poji ‘boy’ (cf. Finnish poika);

    (m) French feu ‘fire’ < Latin focus ‘heart, fireplace’ and German Feuer ‘fire’ < Proto-Germanic *fūr-i (cf. Old English fȳr) < Proto-Indo-European *pūr-;

    (n) Aleut uku- ‘to get sight of, to find’, uku-x̣ta ‘to see’ vs. Old Russian oкo ‘eye’ < Proto-Indo-European *ókʷ- ‘eye; to see’, also involved in another classic example:

    (o) Modern Greek μάτι ‘eye’ (apocopate of the diminutive ομμάτιον, from όμμα ‘eye’ < *óp-ma < PIE *ókʷ-mn̥) and Malay ma-ta
    ‘eye’ (we could add Dura [Tibeto-Burman] mata ‘look, match’ too);

    (p) Spanish mucho < Latin multus and English much < Old English mycel < Proto-Germanic *mikil- (cf. Gothic mikils(a) ‘great, many’) < Proto-Indo-European *meǵ(ʰ)- ‘big’ (cf. Latin magnum);

    (q) Toponyms: Colima, one of the states of Mexico, and Kolyma, river (and a village) in Northern Siberia.

  21. That is a nice list.

  22. -(a) Japanese 名前 na-mae and English name or German Name;

    It’s one of the words used to support Nostratic theory.

    It goes like this:

    Proto-Japanese: *ná(N)
    Meaning: name
    Old Japanese: na
    Middle Japanese: nà
    Tokyo: nà(mae)
    Comments: The verbal correlate for *naN ‘name’ in PJ is *nǝm- ‘to pray’. The noun goes back to the suffixed form *ĺi̯ŏ́m(o)-ŋa (note the recurring *-ŋ-suffix in the reflexes: MKor. nì’jàkì also goes back to *nìŋa-ki
    Proto-Altaic: *ĺi̯ŏ́mo(ŋa)
    Meaning: name; spell, divination

    Going one level up, we arrive at

    Uralic: *nime (? *lime)
    English meaning: name


    Indo-European: *(e)nomen-,
    Meaning: name

    Proto-Germanic: *naman, *namna-n; *nōmian- vb.
    Meaning: name
    Old English: nama, -an m. `name; noun’
    English: name

  23. Nobody thinks Japanese is Uralic, so I don’t know what “one level up” means. As I pointed out when we discussed this, the n-coincidence of IE and Uralic may be very old borrowing or a faint trace of relationship, but the only core Altaic analogue is Mongolian ner, which resembles the IE and Uralic forms only in its first consonant.

  24. Part of my post did not make it for some reason.

    Anyway, Proto-Altaic is reconstructed as *ĺi̯ŏ́mo(ŋa) with meaning: name; spell, divination

    Turkic: *jom, *jom(ŋ)ak
    Meaning: 1 tale, legend 2 luck, omen 3 word 4 riddle
    Mongolian: *dom, *domag
    Meaning: 1 magic 2 legend
    Tungus-Manchu: *nim-ŋā-
    Meaning: 1 to shamanize 2 fairy-tale
    Korean: *ni’jaki
    Meaning: tale, story
    Japanese: *ná(N)
    Meaning: name

    Nostratic form *ĺVm(ŋ)V

    Indo-European: *(e)nomen-,
    Meaning: name
    Altaic: *ĺi̯ŏ́mo(ŋa)
    meaning: name; spell, divination
    Uralic: *nime (? *lime)
    meaning: name

  25. David Marjanović says:

    Also worth mentioning is Hittite laman “name”; you can’t actually reconstruct the “initial” n all the way down to PIE.

    Indo-European: *(e)nomen-

    As usual, the Moscow School cites its PIE reconstructions from Pokorny. Today, at least as long as the Hittite l is ascribed to later dissimilation, this stem is reconstructed as e-grade *h₁néh₃-mn-, zero-grade *h₁n̥h₃-mén- – which brings us to the next question: that’s a mn-stem, formed from a root *h₁neh₃- and the widespread suffix *-men-. There’s no trace of this segmentation elsewhere in Nostratic as far as I can see (unless that Altaic *-ŋa thing is related). Either the whole comparison is wrong, or the word was reanalyzed as a mn-stem somewhere on the way to PIE.

    …I suppose it could have become one of those singulative n-stems as an intermediate step? **h₁léh₃m-n-?

    For those pesky laryngeals we’d probably need an Afro-Asiatic or maybe a Kartvelian comparison.

  26. David Marjanović says:

    BTW, Oleg Mudrak now thinks that Proto-Altaic (and Proto-Turkic) *ĺ was actually [ɬ] rather than [lʲ].

  27. I take the easy way out and consign Nostratic to the realm of interesting speculative fantasy.

  28. No reason to be so absolute about it. Weak evidence deserves at least a weak degree of belief.

  29. David: A summary of PIE > Hittite sound changes says “PIE *n before two nasals in the same word > H /l/”, but of course the evidence given is this very word! Hopefully there is other evidence as well.

  30. David Marjanović says:

    Hopefully there is other evidence as well.

    Even if there is, it doesn’t tell us in which direction the sound change went – unless n occurs before two nasals in other Anatolian languages, of which I have no idea and which may be impossible to tell from the scarce documents.

    I take the easy way out and consign Nostratic to the realm of interesting speculative fantasy.

    Too easy for me. As G. Starostin (2009) put it on p. 12 of the pdf:

    “Certainly, cases where the word for ‘name’ has been borrowed are known; likewise, for the word ‘water’; likewise, cases where two or more random items on the 100-wordlist have been borrowed from a single source. But in most, probably even all, such cases borrowing of such basic items has only become possible due to a concentrated “bombardment” of the recipient language by lexical items from the donor language — “bombardment” which, obviously, begins with a large number of technical and cultural terms. A situation under which two ancient languages “meet”, exchange terms for ‘name’, ‘water’, ‘give’, and ‘sinew’, and then part company borders on the ridiculous, and at least requires extra proof.”

    Now, this is just about IE and Uralic. But similar phenomena are all over the Nostratic landscape. I’m sure that much will have to change from the current reconstructions of Proto-Nostratic (or indeed Proto-Afro-Asiatic and Proto-Altaic), not limited to the fact that the Moscow School systematically errs on the side of inclusion because old Pokorny did; there’s just too much evidence to explain it all away.

  31. No doubt, and I’m not saying I categorically reject all such comparisons. But the work that’s been done in that direction is largely so credulous as to be next to useless, and while it may well be that more careful research, using truly comparable lexemes that have been rigorously reconstructed, will wind up convincing me, I’ll wait till that happens before bothering to shift my preconceptions, which grow heavier with every passing year.

  32. While on this subject, neither kayak nor qayiq are related to Spanish cayuco ‘small boat, dugout’, derived from cayo ‘island, key’, itself reportedly from Taíno.

  33. russell armbruster says:

    Here’s my theory on the words kayak and caique and their similar meaning.Both words could have be derived from imitating the sound of the boat or the oar in the water.This could explain the palindromic sound of the words.Also the word kick was derived from a kick in the water which was a shorter stroke in the water.

  34. Adele weaver says:

    how do you say kayak in Inuit language?

  35. qayaq, as the thread above tells you. Tolle* et lege.

    *Here in the post-classical sense ‘scroll up’.

  36. I am a Turkish, and we still say “kayık” for these small boats, usually wooden.


  37. It took my attention When I occasionally saw the word KAYAK in internet They use it in english for small boats and I ended up here. 🙂

    Common words in remote languages won’t prove They’re related or they have a common ancestor. There are thousand of Arabic words used commonly in Turkish. You would sometimes think Turkish could be related to Arabic but it’s not. not even close. one thing I know is kayık or kayak is definitely originated in Turkish. if they are borrowed to Turkish from another language then this language has to be prehistoric and already extinct. Because these words are like oldest words used in Turkish and I can’t think any other word that I’m sure it’s Turkish about.

  38. January First-of-May says:

    Because these words are like oldest words used in Turkish and I can’t think any other word that I’m sure it’s Turkish about.

    Sadly it’s not that simple. One could hardly think of a word more Russian than хлеб, for example, yet it is actually an early medieval Germanic loan.

  39. John Cowan says:

    Is the /-b/ a, what’s the word for it, anti-spelling pronunciation, then? I’d expect Germanic /-f/ to correspond to Russian /-p/, since Russian did not have /f/ then, and indeed it is pronounced [-p], so what accounts for the written form?

  40. The inflected forms have [b].

  41. JC: there’s no F in the German word Laib=loaf, and the genitive is Laibes.

  42. David Marjanović says:

    The [b] is also present in FYLOSC, which lacks final devoicing.

Speak Your Mind