I’ve written about fish names fairly often (e.g., whitefish, Fishbase, bream), and now I must do so again, even though I otherwise have no contact with our cold-blooded cousins. I was reading Ben McGrath’s “The White Wall” (about the Iditarod) in last week’s New Yorker when I came across this description of the dogs’ meal: “Next came a round of frozen sheefish steaks, followed by beef, and, for dessert, chicken skins—pure fat.” I thought “sheefish” might be a typo (the New Yorker not being what it used to be), but no, there it was in Webster’s Third New International Dictionary: “sheefish [shee (prob. native name in Alaska or northwest Canada) + fish]: inconnu.” The vague etymology was frustrating (it’s not in the AHD, which would have had a better one, and the OED won’t get around to redoing the S’s for years), but “inconnu” took me aback—that’s the name of a fish? Sure enough, the OED has “A game fish, Stenodus leucichthys, belonging to the family Salmonidæ and found in Alaska and north-west Canada,” with citations going back two centuries. And the etymology? “French, unknown.” Which is baffling at first glance, but of course what they mean is that inconnu is a French word meaning ‘unknown.’ And why is a fish called “unknown”? Apparently it’s unknown. (Alas, inconnu isn’t in M-W Collegiate or the AHD, and the Wikipedia article, Nelma, lists it as an alternate name but doesn’t attempt an explanation.)


  1. Could it be related to Russian “yukola” (possibly of Finno-Ugric origin but incorporated, with variations, into many native languages of Northern Far East and America’s Far North. “Sig” is one of the main yukola-making types of salmonid fish, encompassing cisco, sheefish, & whitefish of N America. Sheefish may or may not be related to the word “sig”, although it’s one of the sig species. But specifically, its Russian name is нельма (which even made it into a Latin ), and it’s indeed the #1 yukola-making fish.

  2. Maybe the inconnu was first found near Nome.

  3. Oh, here we are:
    Dubbed the “poisson inconnu” (unknown fish) by explorer Alexander Mackenzie’s voyageurs in the 19th century, the game fish is also known as the “connie,”, “coney” or “sheefish.”
    btw the Latin name is Stenodus nelma

  4. Odd that it should be nicknamed connie, which suggests that the original anglophone pronunciation was “in-CONN-ue”.

  5. JC, yes, that’s why jumped at yukola at first, which also stresses the penultimate syllable.
    Apparently sheefish is known as sii in Iñupiat Eskimo (also in this online disctionary

  6. sorry fused link. Iñupiaq dictionary

  7. And what about nelma? Russian?

  8. FWIW, Random House/ gives the following etymology for “sheefish”:

    1785–95; shee (

  9. Well, that was odd. Take 2:

    1785–95; shee (< a Subarctic Athabaskan language; compare Eastern Kutchin ṣyuh ) + fish

  10. Y, etymology of nelma is anybody’s guess but the word has been attested in Russian since at least XVIth c. when it appeared in the Domostroy. Since the relevant Domostroy passage is a menu choke full of fishes, I copy it in its entirety for LH’s pleasure:
    А в Великой же мясоед после Христова Рожества в стол еству подают рыбную; селди паровые, селди свежые застуженые, лещи паровые, спины белые рыбицы, спины лосоши, спины нелмежи, спины семожьи паровые, стерляди паровые, сиги, лодоги паровые, ухи назимые, короваи, поросята, утки телные, ухи шафранные, ухи черные, ухи мневые, молочика, печенцы мневые, ухи щучи с перцом, ухи окуневые, ухи плотичи, ухи лещевые, ухи карасевые, тавранчюк белужеи, тавранчюк стерляжеи, ухи с мешечки, ухи столчаники, ухи стерляжи, ухи судочи, ухи черевца стерляжи. Просолнаго: белая рыбица, лососина, нелмина, стерлядина, осетрина, головы стерляжи, головы щучьи с чесноком, кружек, стерлядина, щуки отварачеваные, окуни, плотицы росолные, лещевина, щучина росолная с хреном, щучина жывопросолная, трубы, схабы белужи, селди жареные, осетрина шехонская, осетрина косячная, осетрина длинная, шти.

  11. Whoa, the folk etymology of “kangaroo” supposedly meaning ‘I don’t know’ actually came true for this fish!

  12. I copy it in its entirety for LH’s pleasure
    Thank you!

  13. mattitiahu says

    Growing up in the Canadian North that one always threw me for a loop. I always supposed some voyageurs catching one and having no idea what to make of it. But surely they could have come up with a better name…

  14. A search of Google Books also turns out the following entirely useless tidbits:
    – An essay on British dialects, discussing ‘shee-fish’ vs. ‘hee-fish’.
    – An 1826 book, where ‘shee-fish’ refers to the wels catfish ( Silurus glanis L., aka sheatfish, per Wikipedia). That one is a European catfish, unrelated to the subarctic salmonid. I don’t know what ‘wels’ or ‘sheat’ are.
    – An 1895 issue of Fieldiana, the first mention in print of the Alaskan fish (“During the fall, the most important subsistence activity for members of Kotzebue society was fishing through the ice for tom-cod, sheefish, and smelt.”)

  15. It takes a certain savoir faire to name a fish inconnu.

  16. marie-lucie says

    The obvious thing would have been to ask local people what they called it: for instance, different varieties of Pacific salmon are called things like coho, sockeye, chum, among others.

  17. Dmitry, my Russian is near-nonexistent, but doesn’t your 16th century menu include sig as well as nelma? Did it mean some other kind of sig?
    I wonder if sig comes from a Yeniseian language, and if so if it’s cognate to the Eastern Gwich’in ṣyuh given above.

  18. Actually nelma is mentioned twice in the Domostroy’s Afterfeast of Theophany menu: also in спины нелмежи, a pretty archaic adjective form.
    Y – sig in a strict sense is a European whitefish, Coregonus lavaretus, but all related fish species are also called sigs in a broader sense (the same was as we call all salmon species “salmon”, although in a strict sense it’s only a European salmon). Vasmer explains that sig-fish is known by pretty much the same name in Western Slavic as well as in Nordic and Baltic languages (Danish “sig”, Icelandic “sikr”, Polish “sieja” etc.), and in Finnish too (“siikа”); long-distance, wide-ranging borrowing is unlikely, so the word must be older than the split of Germanic/Baltic/Slavic languages.

  19. Quite irrelevent, but perhaps amusing or intriguing to some:
    Maybe the inconnu was first found near Nome.
    Recognizing anon’s little joke, I followed his link to
    An alternate theory is that Nome received its name through an error: allegedly when a British cartographer copied an ambiguous annotation made by a British officer on a nautical chart, while on a voyage up the Bering Strait. The officer had written “? Name” next to the unnamed cape. The mapmaker misread the annotation as “C. Nome”, or Cape Nome, and used that name on his own chart;[2] the city in turn took its name from the cape.
    The British cartographer was the Hydrographer of the Navy, and the British officer was William Bligh, lieutenant in charge of chartmaking on Cook’s voyage of 1779.
    This ‘alternate theory’ is the one I learned as the origin of the city’s name.

  20. “tom-cod, sheefish, and smelt” – Their smelt must be different than the ones we have here, because ice-fishing for those would really not be worth the bother.

  21. John Emerson says

    My guess is that the “Wels” in “Wels catfish” is related to Wales, Walloon, Gaul, Vlach, etc and means “foreign” or strange”. These are HUGE 500-lb + catfish found in the Danube etc, capable of eating small children though I’m not sure there are documented cases.
    The “whitefish”, as I have said, has five commercial / popular names (cisco, tullabee, whitefish, lake herring, chub) which can all be applied to about five closely related species.

  22. John Emerson says
  23. John Emerson says

    Coregonus. This link adds “bloater” to the popular names, and about 70 species to the 5 species I mentioned.

  24. John Emerson says

    Wikipedia can answer all questions.
    “Who’s on first”.

  25. I started it, so I’ll finish it. Wels is not from Wales: it’s a German borrowing. German Wels is from the same root as Wal ‘whale’, though I haven’t found out the history of this doublet. They sure are big though.
    Sheat-fish is another German borrowing, from Scheidfisch, i.e. sheath-fish, apparently ‘because it resembles the sheath of a sword’.

  26. I once read the suggestion that Wales, Walloon, Vlach … don’t mean simply “foreign” or strange” but meant something like Latinised/Romanised Foreigners. So the Irish, Slavs and Balts don’t get the title, but Romano-Britons and others do. It’s a fine hypothesis because one good counter-example would disprove it.

  27. John Emerson says

    It may have switched according to context, like “Tajik”, which now means Persians in C. Asia, but in the past meant…. several other things, including (as I remember) Arabs in C. Asia.

  28. The source is Proto-Dene (Athabaskan) *šiˑš, apparently via an eastern Northern Dene language where the coda has fallen off. Denaʼina has šiš, Deg Xinag has šʳešʳ, Lower Tanana has šʳišʳ and Gwichʼin has šʳiah, all meaning ‘whitefish, inconnu’. I also have a listing for Southern Slave sih as ‘fish sp.’, which is probably the same thing. Dogrib (Tłįchǫ) has łìh ‘whitefish’ which may be the same thing but I don’t know why > ł would happen.

  29. Also for salmon names see my essay: Salmon Names.

  30. marie-lucie says

    James C., your article on salmon names made me homesick for the Nass Valley – having become used to “real salmon” I am spoiled for ever. I tried to add a comment with the local English names, but after trying 5 times and failing to prove I was not a robot, I gave up! As concerns the local native names, none were similar to Tlingit.
    I don’t know why *š > ł would happen.
    Does this apparent correspondence occur elsewhere? It is possible that it has not been noticed, or it has been dismissed, for a perceived “lack of phonetic plausibility”, but I can assure you that there are examples of such (at least in some other language families). On the other hand (assuming this is a true correspondence, not a coincidence), it does not mean that one of the two sounds has changed into the other: both could be independently derived from yet a third.

  31. It certainly wouldn’t surprise me: they are acoustically quite similar.

  32. James C.: Thanks very much for your informative comment and post; you should go back to writing the blog! If you do, let me know and I’ll add it to the sidebar.
    I tried to add a comment with the local English names, but after trying 5 times and failing to prove I was not a robot, I gave up!
    My deepest apologies! Until I change platforms and get some real anti-spam measures, like captcha, all I can suggest is John Cowan’s trick of adding <b></b> between a couple of letters. I know, for instance, that the string “exit” is verboten (who knows why?), so I used that trick to quote it there; if one of the words in your comment had that string, you’ll need to do the same.

  33. marie-lucie says

    LH, this problem had nothing to do with you. I (m-l) am the one who tried to prove I was not a robot according to instructions, by trying to copy those strange and strangely distorted words that are intended to serve as gatekeepers. Usually I can do it right away or at most on the second attempt, but *five* times trying unsuccessfully is too much! It is not your (LH’s) fault, or James’s, it must be whatever he is using for his blog (I think I commented on that blog once or twice long ago and I don’t remember problems then, so maybe the blog support has new rules now).

  34. Oh, for once it’s not a problem at my end! Whew.

  35. marie-lucie says

    LH, I see the source of the misunderstanding: my sentence could be ambiguous, but I meant I was trying to add a comment with salmon names on James’s blog, not here.

  36. It takes a certain savoir faire to name a fish inconnu.

    Don’t you mean a certain je ne sais quoi?

  37. David Marjanović says

    I don’t know what ‘wels’ or ‘sheat’ are.

    German Wels “catfish”, by default Silurus glanis.

    I once read the suggestion that Wales, Walloon, Vlach … don’t mean simply “foreign” or strange” but meant something like Latinised/Romanised Foreigners.

    Possibly from the Celtic tribe recorded by the Romans as Volcae.

    long-distance, wide-ranging borrowing is unlikely, so the word must be older than the split of Germanic/Baltic/Slavic languages.

    Why can’t Old Norse have borrowed it from Baltic and/or Slavic? Germanic probably isn’t very closely related to Balto-Slavic, though not much research has been done on that.

    Does this apparent correspondence occur elsewhere?

    AFAIK, all Athabaskan languages have both sounds as phonemes, and they generally correspond 1 : 1, so when they instead correspond with each other in this word, that needs an explanation.

Speak Your Mind