I’m not going to get into the politics of the mess in the north Caucasus except to say that there are no good guys, but I have to get a minor linguistic gripe off my chest: all the news broadcasts are talking about “ah-SET-ee-ə” and the “ah-SET-ee-ənz.” What’s next, cro-AT-ee-ə? ve-NET-ee-ən art? I realize none of the broadcasters and reporters have ever heard of Ossetia before, but you’d think the patterns of English spelling would clue them in to its proper pronunciation, ah-SEE-shə. I suppose it’s another case of hyperforeignification, like “bei-ZHING.”

Incidentally, Ossetian (as every schoolboy knows) is an Iranian language, and the Ossetian name for Ossetia is Iryston, based on Ir, the self-designation meaning ‘an Ossetian’ (well, actually it specifically refers to the majority group of Ossetians, and the minority Digors resent the use of that name for the whole people, causing some Ossetes to identify with the medieval Alans and call Ossetia “Alania,” but let’s set that aside—if you’re interested in the messy politics of Caucasian ethnic nomenclature and the Alans, read “The Politics of a Name: Between Consolidation and Separation in the Northern Caucasus” [pdf, html] by Victor Shnirelman); it used to be thought that Ir was derived from *arya- ‘Aryan’ and thus related to Iran, but Ronald Kim denies this in “On the Historical Phonology of Ossetic: The Origin of the Oblique Case Suffix,” Journal of the American Oriental Society, Vol. 123 (Jan. – Mar. 2003), pp. 43-72 (JSTOR); the relevant discussion is on p. 60, fn. 42. Kim says it may be from a Caucasian language, or it may be descended from PIE *wiro- ‘man.’ (The word Ossetian is based on a Russian borrowing of the Georgian term Oseti.)

Update. A couple of weeks later, having heard the SET pronunciation approximately six thousand times and my favored pronunciation not once, I am giving up and reconciling myself to the fact that, for whatever stupid reasons, the new pronunciation is firmly established and I might as well accept it. Already I hear it with indifference, and soon I’ll probably start saying it myself; in a few years I’ll look back at this post with amusement and try to remember what it was like to experience the shock of the new. Such is language change.


  1. I doubt it’s “hyperforeignification” as much as it is a direct borrowing of the Russian pronunciation. The newscasters probably just asked a Russian speaker how you say it. (Note the reduced unstressed initial ‘o’ too.)
    I’ve always wondered how broadcasters come by their pronunciations of exotic locales. I particularly winced at Qatar, or gutter, or cutter, from 2003-4.

  2. I tend to pronounce Peking-as-was “Byjingo”. People seem to understand.

  3. John Emerson says

    Time to recommend Ronald Wixman, Language Aspects of Ethnic Patterns and Processes in the North Caucasus, University of Chicago Geography Research Series no. 191, 1980.
    The book is mostly about Soviet language policy, but outlines the amazing linguistic complexity of the area, especially Azerbaijan.
    IIRC there’s a considerable area where Avar is the lingua franca, but most people speak even more obscure languages at home.

  4. For some reason AHD has “oh” at the start, not “ah”, but the spelling would suggest “ah”, at least for Americans.

  5. languagehat: In Arabic, Ossetia is spelled أوسيتيا, which I imagine must be pronounced oh-SEET-ee-ə. Similarly, Croatia is spelled كرواتيا and I know that’s pronounced cro-AT-ee-ə.
    Ian: The Qatari dialect actually pronounces Qatar close to gutter. Certainly closer than kuh-tar. (!?)

  6. KCinDC: Maybe I’m atypical, but it never occurred to me that it might be pronounced with ah instead of oh.

  7. @Omar: The only English word that comes to mind beginning with ‘oss’ is ‘ossified’, which may be the source of KC’s impulse to pronounce it with ‘ah’.

  8. Actually, I think I might’ve pronounced it with “oh”, but when I saw LH recommending “ah” and then thought about the “ss” in the spelling, “ah seemed more correct.

  9. “BeiZHing” instead of “BeiJing” has long annoyed me as well. “Jing” as in “jingle” is not a difficult sound for native American English speakers. But whatreally gets me going is the pronunciation “carryohkey” for “karaoke”–“kahrahOKAY”–and “footahn” for “futon” [pronounced “fooTONE], also easy sounds in English.

  10. The newscasters probably just asked a Russian speaker how you say it.
    That’s quite possible.
    Omar: Well, sure, that’s what I’d expect in Arabic, but not in English.
    oh- vs. ah-: AHD has only the former, Merriam-Webster only the latter. Odd.

  11. Omar, I’m less concerned with the q>g transformation than I am with the emphatic T>tongue flap. Plus, it sounds like “gutter” when you say it like that.
    One of the great linguistic wonders, for me, is how closely related Yaghnobi and Ossetic are. My Iranian-linguist friend says it’s thanks to the Scythians.
    Let’s keep our ears open over the next week: languagehat’s post is now one of the first Google results for “ossetia pronunciation.”

  12. At last, I may do some good in the world!

  13. If only people could take the “zh”-saying skills they use incorrectly in “Beijing” and apply them to “Solzhenitsyn”, I’d be much happier. I got tired of hearing “Solzenitsyn” from various newscasters last week.

  14. Reading the comments on this thread I can’t help wondering if making IPA a compulsory subject in schools around the world wouldn’t be a huge step to improving clarity of communication. I am not even sure if the “ah-SET-ee-ə” mentioned above is the same as what I hear as “Oh-SET-ee-ə” from local newsreaders here in Zild. I don’t know IPA, and am not fussed on having to learn it, but it would help lift discussions like this out of the Qatar.

  15. John Emerson says

    I might add that another name for the Alans was As or Aas, and my guess is that Ossete is “As” + suffix. This would support the “Ah” pronunciation.
    I’ve posted this before, but: Alans foremerly in the Roman service were a factor among the Vandals and the Suevi, and an important factor among the Bretons when they came to the continent (Bachrach: History of the Alans in the West). Alans in the Mongol service were a presence in China, where they probably merged with the local populations. The Ossetes are the ones who remained at home.
    And there’s even a theory that King Arthur was an Alan or Scythian from Roman Britain. I haven’t read it carefully but it’s not completely fanciful.

  16. OK, so the html was stripped from my last post because I had to C&P it after the commentsa box went down for a while. If anybody wants to hear how NZ newsreaders say the word in question, it’s here:

  17. I got tired of hearing “Solzenitsyn” from various newscasters last week.
    I must have missed that one. Or I’m desensitised by “Karadich”.
    And there’s even a theory that King Arthur was an Alan or Scythian from Roman Britain.
    Is that the same one according to which he was a Sarmatian or is it a new theory?

  18. michael farris says

    Ossetian with t doesn’t bother me at all (maybe I think of it as a Russian loan?)
    But if this is going to devolve into awful exoticising pronunciations by journalists I just heard a new one that set my teeth on edge (a sure sign that it will soon take over.
    Tibetian (maybe rendered Tih-BEH-chan)
    The speaker in question sounded native (maybe British this and some other quirks have burned everything else out of my mind) and said it over and over and over again.
    Also, “Tibetian” gets a scarily large number of google hits, mostly it seems from native speakers.

  19. “Also, “Tibetian” gets a scarily large number of google hits, mostly it seems from native speakers.”
    I’ve heard it too. I’m sure it’s nothing to loose sleep over.

  20. I’m pretty sure Bush has used “Tibetian”, along with “Grecian” and several other anomalous “-ian” forms.

  21. Ah, it was on Language Log.

  22. "King" Arthur Krown says

    This would support the “Ah” pronunciation.
    Only in American. Aas (meaning ‘hill’) would be like Brit. ‘horse’ in Norway, and Brit. ‘orse is like Oss… It makes a good case for IPA, though I’d rather die than actually learn it.
    Doc Rock: what really gets me going is the pronunciation…
    Perhaps you’re not a linguist, Doc, but how does the linguists’ irritation with mispronunciation of foreign words square with the linguists’ irritation with prescriptivism? I’ve only just stopped telling my daughter to not split infinitives, and now I have to correct carry-okie and make her say kahrahOKAY? She’ll look like a complete wanker at school and get beaten up. Don’t linguists want to simply stand back and watch the rest of us make a huge mess, and then write a paper about it?

  23. “It makes a good case for IPA, though I’d rather die than actually learn it.”
    Thanks for the support!

  24. Slightly off the topic of Ossetia, there’s a useful article in today’s Guardian about internet-pickupable radio stations
    It says Philosophy Bites is a 10- to 20-minute show where philosophers are interviewed on “bite-sized” topics… The episode with Hugh Mellor, discussing the nature of tense and time, is mesmerisingly convoluted.

  25. “Proper” pronunciation? How’d you figure that then? The OED certainly accepts both, and it seems pretty reasonable to me. I could understand your concern if it was mangling the native Ossetic word, but since the English comes from Russian and Georgian anyway I don’t see the issue.

  26. I am not now, nor have I ever been a member of the linguistics fraternity–I am a translator and language teacher (literature degrees) and a damned “prescriptivist” due to six years of Latin under a martinet!

  27. John Emerson says

    Bulbul: Same theory.
    Rightly or wrongly, I use “Scythian” generically, to mean all northern Iranians of the steppe, regardless of whether they’re Sarmatian, Sauromatians, Alans, etc. I think that “Alan” might be the most correct name for the period after ca. 300 AD. By that time the Northern Iranians were just one rather lesser people among many, and the Huns and/or Goths were dominant in the area. They did seem to play a considerable role in the Roman military, and unlike the Germans and Huns they seem to have stayed consistently loyal. (Arminius [= “Herman the German”] was able to defeat the Romans because they thought he was on their side).

  28. One forgets, on this side of the pond, that o tends to be ah in American. Ossetia definitely starts with a short o in English. I hear it as O- SET- y – ah. (Not knowing IPA, I can’t render the clipped o sound, like the English form of octave or oblate.
    Graham – are you lurking ? What did the Pronunciation Unit rule ?
    A.Crown – your ref to the Guardian piece lured me in and shows how gloriously time-wasting the ‘net can be !

  29. michael farris says

    I haven’t heard Solzenicyn, but I have heard a lot of newscasters say Soldženicyn, often the same ones that affect Beižing (bangs head against wall).

  30. Nicholas Ostler says

    Well, Language Hat is in good – or at least well-acquainted -company. In a BBC interview on Radio 4 two days ago, Michael Saakashvili repeatedly pronounced the name as [ɔsi:ʃǝ]. Phew! And it is noticeable that all the Google instances of “Ossetia pronunciation”, where they refer to English, give results consistent with this – evidently unknown to the news fraternity. But what are we to make of ‘Ingushetia’ – [iƞguʃi:ʃǝ], anyone? That may be what’s really putting the journalists off.

  31. Krown, A.J.P. says

    I wonder, John Emerson, if you’ve ever considered writing something like Gombrich’s 1935 children’s A Little History of the World, except for grown ups. It would be quite fun for the rest of us.

  32. John Emerson says

    I am working on something about Central Asia. I have self-published Substantific Marrow, which includes pieces on the Crimean Goths, the Kalmyks, the Caucasian Albanians, et. al.

  33. The OED certainly accepts both, and it seems pretty reasonable to me.
    I’m not sure why the “certainly,” but now that I check, yes, it does accept both, unlike any American dictionary I have. Interestingly, the OED gives the -ti- pronunciation first for the UK and second for the US, so perhaps there’s yet another cross-Atlantic divide, though all the broadcasters I’ve heard have been Yanks. It does not seem reasonable to me, because it directly contradicts the spelling conventions of English. As for its coming from Russian, that’s neither here nor there; samovar, taiga, and steppe are from Russian, and we don’t pronounce any of them anything like the Russians do. (Not to mention names like Rodchenko and Khrushchev.)
    “King” Arthur: Yeah, that’s Doc Rock’s personal peeve about karaoke; no linguist would have any problem with the standard American pronunciation (nor do I, though I grew up partly in Japan).
    I wonder, John Emerson, if you’ve ever considered writing something like Gombrich’s 1935 children’s A Little History of the World, except for grown ups. It would be quite fun for the rest of us.
    This is a great idea.

  34. John,
    I see. My knowledge of those Iranian peoples is comes mostly from Soviet historiography which make a distinction between Sarmatians and Scythians, largely, I suspect, for some ideological reason. Alans, if I recall correctly, were always mentioned in one breath with Sarmatians. Ah well, let’s add that to the list of stuff I should read up on.

  35. John Emerson says

    Bulbul, for the last 20 years or so Western and Russian archeology have been converging. During the Cold War communication was limited, which was horrible since a lot of the best sites were in Soviet areas. (Much the same is true of the archeology of China).

  36. Emers, another theory is that King Arthur came from Kelso. Pronounced “Kelsae” locally; but then locally “Jedburgh” is pronounced “Jeddart”, and “Kirkcudbright” is pronounced ….., all of which make Byjingo pretty easy.

  37. Krown, A.J.P. says

    I bought it (as an object)– how can anyone not. The titles of the pieces looked so good, like a list of ice-cream flavors.

  38. so how do the ossetians say it? ah or oh?
    if bush said grecian i’m sure it’s not his fault but the fault of the hair colouring for men.

  39. as for exotification of names vs. pronouncing them correctly vs how it’s said in english even if wrong,
    i live in nanjing. nan is a rising tone (#2). jing is a high tone (#1). shanghai is shang4 hai3, dropping and then dipping. when back in america, that’s how i say it out of habit. but i often have to correct myself to saying shaeng-hi with no tones since thats how people here do, and nanjing with no tones at all. otherwise i look like a fool despite saying it correctly. alternatively, in the states i say muslim as moos-leem with a voiceless S. in china that got me weird looks among my expat friends so i have to “correct” that to muhz-lim.
    so if communication is the purpose, then by all means im all for messing up how we say things. i drop my kh glottals when saying arabic words to people who dont speak arabic, as im sure they think khaled is “call-ed”.
    pardon the lengthy unfocused comment.

  40. mollymooly says

    The Russian pronunciation is fine for North Ossetia, but using it for South Ossetia should be controversial.

  41. mollymooly says

    The US/GB difference is similar in “Tunisia” and “Parisian”.

  42. Mollymooly, what about “Polynesian”, “Asian”, “ambrosia”, “decision”, “lesion”, “contusion”, etc.?

  43. mollymooly says

    Personally, ambrosia is -zee-uh, not -zhu. The others, not.
    Wikipediaalso lists:
    /siː/ /ʃ/ cassia, CassiusA2, hessian
    /tiː/ /ʃ/ consortium
    /ziː/ /ʒ/ Frasier, Parisian, Malaysia
    /siː/ /ʃ/ cassia, CassiusA2, hessian
    /ziː/ /ʃ/ transientA2, nausea
    glacier /ˈglæsiə/ or /ˈgleɪsiə/ vs /ˈgleɪʃɚ/
    and on a different note
    /diː/ /dʒi/ cordiality
    /ʃ/ /ʒ/ AsiaB2, PersiaB2, versionB2
    /z/ /ʃ/ Dionysius

  44. http://www.guardian.co.uk/sport/2008/aug/11/olympics2008.olympicsandthemedia
    Returning to Beizhing, this week’s Media Supplement of the Guardian newspaper has a piece about the BBC’s problems of getting a uniform pronunciation from their reporters. For what it’s worth, they are supposed to pronounce it with a hard J.

  45. “Plus, it sounds like “gutter” when you say it like that.”
    And that may well be intentional. The majority of Englsih-speakers who use that pronunciation are military, and there is a well-established custonms of derogatory names for places. It extends to American places too – Fort Lost-in-the Woods, Wild Chicken, etc.
    “Gutter” may be bad, but it’s better than “Asscrackistan”.

  46. Bulbul, IIRC, the difference between Sarmatians and Scythians was mostly geographic: it’s assumed they spoke basically a group of closely related dialects, but the Sarmatians lived further to the southwest, in western and southern Ukraine.
    The Ossetians are indeed the mediaeval Alans, a Scytho-Sarmatian tribe.
    It’s interesting how out of the huge territory where it was once spoken, the language is now only spoken in a small area in Caucasus. Pretty much like Celtic, which was once spoken in half of Europe and now it’s spoken just in a few rural corners in the British Isles and Brittany.

  47. ah- versus oh-:
    Again we come up against this difference between most Americans and most in the Rest of the World – interesting because of how it affects interpretation of exchanges in a cosmopolitan blog like LH.
    Australians and British would not pronounce Ossetia with either ah- or oh-. Left to our own devices, we never realise o as ah; and we distinguish a “short” o as in got and a “long” o as in goat. Now, clearly some kind of a “short” o is apt for Ossetia, and our o may in fact be as inaccurate as American ah. But that ah strikes us as strange, because we have ah pretty well exclusively as a realisation of a (in father, after, etc.). Personally, the examples I find strangest are Kohsohvoh and cahzmohs. Then I reflect that we ourselves say Kosovoh (with inconsistency) and cozmos (with presumably improper voicing).
    Compare also American and AusBrit realisations of ou in couscous and the like: American has the u of tuna (as they say tuna); AusBrit the u of put. That one is harder to account for than the difference with o, and the AusBrit practice does seem to represent most non-English u better than the American practice does.
    The Australian Olympic commentators are as woefully parochial and as criminally insouciant as ever. Some even pronounce it both ways in quick succession, thinking that will absolve them from the charge of catastrophic ignorance (if they think at all on the matter, which is unlikely). At least our prime minister Kevin Rudd gets it right, being famously fluent in Putonghua.

  48. BTW, I just noticed that in an English-language interview, the Georgian president, Saakashvili, was pronouncing it Oh-see-shə.

  49. I was distracted yesterday or would have commented sooner in reply to Paul and Nicholas Ostler.
    When I was in charge of the Pronunciation Unit, up to the end of 2001, we recommended o-SEE-sha (topped and tailed with schwa) and ing-goo-SEE-sha, but the journalists almost all refused to accept this, despite our pointing out Croatia, Venetia, montbretia (garden plant) and all the adjectives ending in -tial. Their argument was that the Russians used ‘-SET-ia’. As we are talking about non-Slavic peoples, this, to me, was irrelevant, but I see that the Oxford BBC Guide to Pronunciation, edited by my successors Lena Olausson and Catherine Sangster, succumbs, and recommends oss-ET-i-uh (with a note: “An older anglicized pronunciation, uh-SEE-shuh, is given in various reference works. Our research has indicated that the pronunciation we recommend, which is closer to the Russian, is more acceptable”) and ing-guush-ET-i-uh (no note – the -uu- is a [to me strange] attempt to represent the ‘oo’ of ‘book’ without resorting to diacritics). I consider this simply a cop-out, because they can’t get the journalists to follow them, although BBC Radio newsreaders would do so with no problem. My view was that “established anglicisations” should be followed, but as no one had heard of Ossetia or Ingushetia before, except for Kremlin watchers and Caucasian specialists, there were no established anglicisations of the names. Therefore it was the Pron. Unit’s job to lead, and not simply to bow to ignorance.
    I notice that since the Olympics have actually started, almost all BBC personnel have begun to say ‘bay-JING’. I wonder if some panjandrum has ordered this: the Unit recommended it from as early as 1979, but nobody took any notice (apart from the aforesaid newsreaders).

  50. As a BBC employee I can confirm that we have indeed all been issued with an email from upper management about how to pronounce Beijing (whether or not we have anything to do with the Games..).
    It annoys me that they decide to get in such a state about this when day-to-day consideration of pronunciations is such a low priority. The current use of Ossetia is an interesting example.

  51. Widsith –
    the only other time that senior management made a stand on pronunciation was the ludicrous decision to fly in the face of all the evidence and insist on the pronunciation ‘awl-thorp’ for Althorp (Earl Spencer, the owner of the property had made it clear that ‘awl-trup’ was correct) when Diana was killed. I received threatening letters from my manager when I wouldn’t back down, and it all contributed to my accepting voluntary redundancy three years later when my own departmental management systematically undermined my position, and quite blatantly lied to me about the department’s future.
    A far cry from the 1980s, when the then Assistant Director General, Alan Prothero, said at an editorial meeting: “We have a Pronunciation Unit – USE IT!”

  52. Thanks very much for your inside information, Graham, and I can’t begin to tell you how much I admire a man who’s willing to risk his job in a principled stand over pronunciation!

  53. Doc: How do you justify pronouncing “futon” as /foo-tone/?

  54. The best approximation for an English speaker would be “ftohng”.

  55. I say FOO-tahn, and anything else would sound pretty pretentious to me.

  56. michael farris says

    I would love to be able to (honestly) write that I rhyme it with ‘button’ (futton) but my pronunciation is basically the same as hat’s (hard to say for sure about that final vowel, I do rhyme it with crouton, so the final -on is just like the word ‘on’.

  57. One kid I knew in high school kept calling it a “photon”.

  58. Actually there is so much difference between the US and Japanese products that two different names are entirely appropriate. If they start selling the American-style futon in Japan, they can always call it a “fuutan”.

  59. Blahbetty Blahblah says

    I also have trouble squaring the descriptivism with irritation at oh-set-EE-ah.

  60. Michael Farris, I rhyme “futon” with “crouton” as well, but the vowel sound in the second syllable is like the one in “con”, not “on” (which rhymes with “lawn”). But maybe you’re from a part of the US where “ah” and “aw” are collapsed.

  61. “but the vowel sound in the second syllable is like the one in “con”, not “on” (which rhymes with “lawn”)”
    Noetica gave a detailed post earlier about “AusBr” English, but neglected the variety spoken on the no-penal side of the Tsaman. Here “con” and “on” have exactly the same vowel sound. I think that this is likely true of British English and Strine too.

  62. michael farris says

    KCinDC, my vowel system is basically lowest common denominator Usian; con, on and lawn all rhyme for me. For years I had no idea what was going on when people made fun of dialects on the basis of ah-aw (dawg, coffee talk* etc).
    *the Mike Meyers Saturday Night Live skit, I have no idea which ah-aw sound is “supposed” to be in ‘coffe’ so the substitution or whatever was going on was lost on me.

  63. mollymooly says

    Interesting that, unlike -tia and -sia, there is no coalescing of -dia (Arcadia, encyclopaedia, Cambodia, obsidian). Unless you count Acadia>Cajun, or BrE meeja. And the French say Cambodge.

  64. Anyone remember the panorama of French history at the beginning of Raymond Queneau’s Les fleurs bleues?
    “Les Huns préparaient des stèques tartares, le Gaulois fumait un gitane, les Romains dessinaient des grecques, les Sarassins fauchaient de l’avoine, les Francs cherchaient des sols et les Alains regardaient cinq Ossètes.”
    There’s a great site which glosses all the wordplay here. It speculates that the “cinq Ossètes” might be an allusion to the film “Cléo de cinq à sept” by Agnès Varda.
    And Queneau knew about the Alans, the Ossetians and the Nart sagas. There’s a reference to the latter slightly further on: “les Alains [regardaient] d’un air narte.”
    “Tant d’histoire pour quelques calembours” indeed.

  65. Mollymooly, don’t forget “Injun” and “idjit”.

  66. Kevin Beane says

    I’m not a linguist and most of these comments are frankly a bit over my head. So, my opinions might either be ill-informed, or a re-statement of a previous comment that was worded beyond my understanding.
    That said, when I read “ossetia,” my mind is pronouncing the same wrong way the newscasters are.
    After reading this, I started to ask myself why. I have come up with a couple of theories.
    a) The double s shouldn’t throw us, but somehow it does. It somehow makes me want to treat the T as double, too. If it were spelled Osetia, I am pretty sure I would pronounce it the way the actual word should be pronounced.
    b) Croatia and Venetian aren’t enough to drill it in our heads correctly, and I have a great, amusing example of a similar phenomemon.
    I’m from Akron, Ohio. As an adult I moved to Elkton, Maryland. A friend of mine, born and raised in Elkton, saw Akron written and called it, “ack-RAHN,” which to me sounds like the name of a villainous robot.
    Of course, the actual pronounciation/stress is no different from Elkton, which she, an intelligent college graduate, had been saying all her life!

  67. John Emerson says

    Many thanks to JCass.

  68. Stuart:
    Here [in NZ] “con” and “on” have exactly the same vowel sound. I think that this is likely true of British English and Strine too.
    Sure, they do. I did neglect to mention the o in for. That is even nearer to the “foreign” vowel in Kosovo, is it not? Since this o is available in most varieties of English, it is curious that we don’t all use it for a very common foreign o. Why do Americans not use it in Spanish plurals like enchilados (rather than -ohss, so odd to others’ ears)? Why don’t they use it for both vowels in cosmos, instead of using two vowels neither of which faithfully mimics the repeated omicron of the original? We Restniks do similar things too, as I mentioned earlier. Why do we not use a single repeated vowel for the single repeated “foreign” vowel of Kosovo?

  69. To Noetica:
    Me no think such word exists.
    You talking about enchiladas?
    Enchilada. Ends in “a.”
    Enchiladas. Ends in “as.”
    Never hear of no enchilados, amigos y amigas.

  70. It is la tortilla which is being chilied, therefore the feminine ending.
    Next question: explain how a beer can be named “Negra Modelo”.

  71. michael farris says

    Since were talking about mexican food (very cruel to mention enchiladas while I’m in Poland which does not have fresh corn tortillas at all), maybe it’s a good time to point out that ‘refried beans’ is a bad translation of frijoles refritos.
    The re- in the original doesn’t mean ‘again’ but is rather a prefix found in many (most? all?) varieties of informal spoken Spanish that means ‘very’ (kind of like -isimo but ). To make re- stronger follow it by -que- and to make reque- stronger, follow it by -te- to make requete- as in requetecansado (totally worn out). To be completely ridiculous add requete- to super and place it before another adjective (as a separate word) or a noun.

  72. Enchililad?s:
    Ach, stupid me. Of all the nouns I could have chosen. OK, take as an example Marcos, which I heard pronounced on the radio by an American with -ohss just as I finished posting last time. Why do they do that? After all, it sounds very little like the Spanish, and the o of for is available and is almost certainly going to be a better approximation even allowing for variation across idiolects. And then, what about cahzmohss, after all? That is an interesting “snap to grid” choice for a nation to make, from the many available alternatives.
    It’s a topic that attracts little interest. Ah well. I’ll keep observing in such a way, in any case.

  73. Noetica, even Americans who use [ɒ] in “for” don’t have it as a phoneme but as an allophone of /ɔ/ (assuming they have that as a phoneme). You might as well ask why they don’t use an unaspirated [p] at the beginning of “Beijing” — after all, they can say it in “spin”.
    For me the vowels before /r/ are difficult to match to the vowels used in other contexts.

  74. I’m surprised that no one’s suggested that Ossetia with the pronunciation [o/ɒsɛtɪja] is derived from Oseti-a, ie from the Russian borrowing of the Georgian term. The ethnonym Ossetes gives a clue to the stem.

  75. mollymooly says

    I want some of those superfried beans!

  76. Fine, KCinDC. Of course the phonotactics is sometimes relevant. But more often, I think, is a whole lot of socially and educationally driven practice.
    How (once more) about cahzmohss? How about Marcohss? What is it about the ending -os, whether occurring in a Greek masculine singular or a Spanish whatever, that determines -ohss to be the generally accepted American pronunciation? Why is the choice of oh extended to pronunciation of “foreign” o very generally indeed? Such choices are not usually phonotactically determined. The a in Marcos is realised as ah, so there is pressure for differentiation in realising the o, wherefore there is pressure for -ohss. Same for adiohss and tacohss. But this explanation cannot readily be extended beyond such specific cases, and does nothing to account for cahzmohss, nor even burritohss. Nor Orinohcoh (which M-W Collegiate gives like this, with the vowel of law – equivalent to the vowel of for? – for the first o). Certainly it does nothing to explain couscous with the u of tune rather than the freely available u of put.
    I make similar enquiries concerning British ways with a in French, Italian, and Spanish words.
    (Please excuse loose representation of pronunciations: accessible, but not IPA-kosher.)

  77. marie-lucie says

    (Marcohs, etc)
    Normal weakening of unstressed vowels in English would make Marcos sound like Marcus, with a schwa – so English speakers trying to approximate Spanish pronunciation have to put a stress on the o in order to avoid a schwa, and lengthen the vowel: therefore they say not MAR-cus but MAR-COHS, as in the name Margo, which has two syllables, both stressed.
    A few years ago Canadian radio announcers started to differentiate the pronunciations of the -er and -or endings in the names of many professions, so we hear for instance ambAssadOHr with a secondary stress on the final syllable. I am not sure if this is done in other countries.
    About “j”: For a while there were problems in the Ivory Coast, so the name of the city of Abidjan was mentioned frequently on the radio. The dj of course indicates the complex sound that is written j or ge or dge in English (as in the initial and final sound in judge or George), but English Canadian announcers pronounced it as if it was written in French Abijan.
    Some people are just not very good at approximating sounds they are unfamiliar with. I knew a Japanese girl named Kiwako. Her Canadian landlady called her Koo-ay-koh and similarly distorted the names of other foreign tenants.

  78. …so English speakers trying to approximate Spanish pronunciation have to put a stress on the o in order to avoid a schwa, and lengthen the vowel: therefore they say not MAR-cus but MAR-COHS…
    Fine, Marie-Lucie. And American pronunciation norms do work hard to avoid a profusion of schwas, unlike AusBrit norms. British in particular takes collapse into schwa as a mark of refinement and high culture.
    However, Restniks do not typically say a schwa in Marcos and the like, and do not do that lengthening you speak of. This might give us pause. We use the “short” o of our on, got, and often, or of everyone’s for (which in fact M-W Collegiate equates with the vowel of law, in its reading of American pronunciation). The question that exercises me concerns this difference between Americans’ and Everyone Else’s practice with foreign words. Is it because our “short” o (quite near to many foreign vowels written as o) is more salient to us and somehow more available, while the nearest equivalent in American speech (the vowel of for and law) is somehow less available to them, and their more available “short” o is disqualified as confusible with the way foreign a is sounded? Perhaps. Strictly, though, length itself seems less relevant. So is stress, per se, less relevant.
    And then there’s couscous, still unaccounted for.
    It is interesting to compare the several realisations of typical foreign words listed in Webster’s 3rd International (W3I) with the far more restricted realisations offered in M-W Collegiate (MWC). I think this is only partly accounted for by the more comprehensive plan of the W3I. There is independent reason to think that pronunciation norms are now more closely prescribed than in the heyday of the unupdated W3I. Broadly social and educational pressures, as I say (and we might muse about the qualifier collegiate, here). The same sort of tightening might be detected in the evolution of SOED, I think.

  79. I while ago I found this post here from the Language Log archives that proposes an explanation for the pronunciation of Beijing with a post-alveolar voiced fricative rather than with the corresponding affricate. (sorry, I don’t know how to do the symbols in html.)
    And here is another post from the Linguism blog that addresses the English fricative/affricate dilemma.
    What I was *really* trying to find, but couldn’t, was a piece I read a while back explaining the allophonic variation of the voiced post-alveolar affricate—because I believe it stated that for some American English speakers, there is no allophonic variant in the form of a corresponding fricative…think of the word ‘garage’; some pronounce it with the affricate, others the fricative. And this may be yet another thing that compounds the pronunciation issue for words like ‘Beijing’. You add sociolinguistic reasons (as suggested in the links above) and the confusion escalates.
    If anybody has a link to information on that, I’d love to read it! Thanks!

  80. Oh, and here’s another one from Language Log…it’s a follow up from the earlier post (first link in the comment above). While this doesn’t directly explain why Americans have so much trouble pronouncing ‘Beijing’, it does suggest solid reasons for the many varied forms of this place name by speakers of Chinese. As Bill Poser over there says, It’s all very simple, really.

  81. Certainly it does nothing to explain couscous with the u of tune rather than the freely available u of put.
    I am not sure why you find this so puzzling. I do not think “short u” is any closer to the Arabic vowel than “long u,” and in any case why does the fact that the Americans and British happen to have settled on different representations of a foreign word require some overarching explanation?
    Me, I find the U.K. pronunciation of words like pasta and macho bizarre and unaccountable—they directly contradict the foreign pronunciations, when the -a- of father is freely available—but I don’t see how it can be explained in a scientifically satisfactory way.

  82. Couscous:
    It is not anything specific in couscous that I find in need of explanation, as if Americans and British were aware of the Arabic vowel they needed to represent – or even had ever heard it uttered. I chose couscous as a concentrated example. The same difference is evident with these, in which Americans do lean (or would lean) strongly toward the tune vowel and British toward the put vowel:
    zuppa inglese
    osso bucco
    opera buffa
    And so on. Sources: MWC and SOED, and some personal recollection.
    Macho and pasta:
    I have already suggested that widespread British practice with such words is questionable. (“I make similar enquiries concerning British ways with a in French, Italian, and Spanish words.”) Pasta, plaque, Jacques Chirac (all with the vowel of hat except the final a in pasta) are resisted in Australian speech. We find them as odd and as inexplicable as American -ohss. As for macho, some of us use the vowel of hat, but many here do even worse. Some of us, and some British, say maKIZmo; and oreGAHno is standard. Americans are generally far better at pronouncing and incorporating Spanish than we are, for good geographical and historical reasons – which makes their ways with Spanish o and u even harder to account for.

  83. couscous arrives into English via French and seems to follow the standard mapping of vowels from there.
    cosmos with both vowels the same is a perfectly acceptable American pronunciation. That there is a competing tendency to “foreignize” all -os endings doesn’t seem too surprising. Nor the one to make it schwa. Here I think the key is whether you perceive this as an ordinary English word or still scientific Greco-Latin jargon.
    I’m not entirely clear what the proposed alternative for Romance /o/ is. Generally, it seems to end up /o:/, with no regard for source length distinctions, if any. That’s American [oʊ], sometimes roughly described as the “pure” Romance vowel plus a glide. Fast-track language courses tell you to try to clip off that glide to sound less American. I can see how other dialects’ [əʊ] or [əʉ] is a lot worse, but I don’t think that’s what you’re objecting to.

  84. michael farris says

    Noetica, I’m not sure what the point is, surely American tune (rhymes with moon) is closer to Spanish and Italian u than the put vowel (in any dialect) is.
    That said, I don’t perceive any special need to mimic foreign pronunciations (I pronounce bouzouki with a schwa in the first syllable) or to make various national standards of English converge in this or other questions of pronunciation.
    I think the best explanation for such changes is the idiosyncratic one. Foreign words of the kind mentioned come into English at different times in different dialects by different methods of diffusion. I’d be surprised if there was more unanimity. There are also mental, subconscious algorithms for assigning consonants, vowels and stress that differ somewhat from dialect to dialect.

  85. Noetica, I don’t understand why it’s puzzling that Americans don’t use a phoneme that you think is better but that doesn’t exist in their speech. We don’t have /ɒ/, so we’re hardly going to use it in domesticating a foreign word. Are you really suggesting we should be using /ɑ/ (our “short o”) in those words?
    Also, “burritos” and some of the other words you mention have /z/ at the end, not /s/, because they’re treated as normal English plurals, not Spanish ones. Do you actually pronounce “burrito” with /ɒ/ at the end? Do any nonforeign words have final /ɒ/ in your dialect?

  86. As has come up a few times before, some varieties of ENE (Bostonian) have [ɒ] for the merged cot-caught vowel. These distinguish father-bother, so [a] is also available.
    Also noted here before is that for some dialects with cot-caught merger, Los Angeles and Las Vegas have the same first syllable. Grammatically, that sounds like trouble, but I guess in practice it isn’t much.

  87. Yes, I shouldn’t have implied that I spoke for all Americans, but the vast majority don’t have that phoneme. The mention of “Los Angeles” does add a wrinkle to Noetica’s complaint, since I think most Americans who don’t have the cotcaught merger do use /ɔ/ there (pronouncing “Los” like “loss”). But then they use /o/ in “Los Lobos” and maybe even “Los Alamos”.

  88. There’s a funny coincidence: in Circassian “an ossetian” pronounced the Russian way (“ossetin”) means “one who pays tribute” (and this correspond to the recent history).

  89. John Cowan says

    steppe [is] from Russian, and we don’t pronounce [it] anything like the Russians do.

    Yes, well, about that: what’s mysterious here is not the Russian but the German (which is the source of the English spelling). [stɛp] strikes me as a much better approximation of [sʲtʲepʲ] for folks who don’t do palatalized consonants than [ʃtɛpə] is. The [ʃ] is automatic and unavoidable, but how does a monosyllable become a disyllable? Is it because German(ic) final [p] is unexploded, whereas Russian final [pʲ] isn’t?

  90. David Marjanović says

    (High?) German final plosives are never unreleased. (I find unreleased plosives one of the most difficult features of English to learn.) German Steppe must come from the French spelling steppe; *step would look weird in French, and it wouldn’t look grammatically feminine (which it is in French and German).

  91. David Marjanović says

    (High?) German

    I forgot that unreleased and even glottalized final plosives are so common in northern Germany that Theo Vennemann used them, together with the English and West Jutland ones, as evidence for the glottalic theory of Proto-Indo-European.

  92. Trond Engen says

    The Barbarians are never more than a steppe away.

  93. David Marjanović says

    ^ Thread won.

  94. Yes, that in itself justifies leaving old threads open indefinitely.

  95. marie-lucie says

    It’s great to be able to revisit old threads. I notice that it is not often that old data or opinions need to be revised.

  96. I first heard it pronounced [ɔsi:ʃǝ] and Saakashvili used this pronounciation when speaking English so of course it must be acceptable. But extremely relevant is it came to English from Georgian (via Russian). The ‘-eti’ part of the word ‘Ossetia’ is actually a suffix in Georgian denoting ‘the land where the X people live’. Hence the Svans live in Svaneti for example. In Georgian it is of course pronounced as [eti], so I can’t bring myself to change the [t] to [ʃ].

  97. The pronunciation in Georgian is irrelevant to the pronunciation in English. Of course Svaneti has a /t/; it lacks the final -a that triggers the change to [ʃ] in English.

  98. And though I’m just a steppe parparian,
    In two weeks I’ll pe Tocharian,
    And it shows.
    Anything Koes.
         —Kevin Wald

  99. I just ran into an English example of the Romance re- ’emphatically’, redoubtable ‘formidable’, whose origin is this ré- (the French form) plus an older sense of douter ‘fear’. The b is of course a bit of English pedantry, as if directly from Latin dubitare ‘doubt, hesitate, ponder’.

  100. Does redouble count?

  101. January First-of-May says

    I *think* the two recent posts were supposed to go in another thread.

    On-topic, I’ve never heard of Ossetia being pronounced (even in English) in anything but the Russian way (though the simplified vowel in the first syllable, if the OP’s description is correct, does suggest “asked a Russian speaker how they say it”, as opposed to a more regular English development).
    I’d probably have said something like “ə-SEE-tee-ə”. My English idiolect is weird (I really should make a blog post about it before it normalizes too much – there’s all sorts of weird things in there; one of them was previously described in a Language Log comment… though the parts that have to do with “poinsettia” appear to be out of date, now it’s more like “pə-in-SET-tee-ə”).
    Similarly, I would not say “cro-AT-ee-ə” but “cro-AH-tee-ə”; I’m not sure on “Venetian” – I’d probably end up saying something like “ve-NEE-see-ən” (as if “Venice-ian”), but the presumably correct pronunciation “ve-NEE-shən” is also a possibility.

    I also say “Bei-dzin” (because Бэйцзин), when I don’t say “Pekin”. And “futon” (whatever it is – a type of Japanese clothing?) rhymes with “put on” (stress on second syllable), and I also have second-syllable stress in “Qatar” (so it’s pretty much “kah-TAR”).

  102. Athel Cornish-Bowden says

    The pronunciation in Georgian is irrelevant to the pronunciation in English.

    Agreed. How many of us know enough about Georgian pronunciation to take it as a model for pronouncing words of Georgian origin in English? In some cases (not this one) we’d need to be able to cope with long consonant clusters with no intervening vowels.

    Of course Svaneti has a /t/; it lacks the final -a that triggers the change to [ʃ] in English.

    And in French, of course, with a trigger of -e to change [t] to [s]: compare “démocratique” with “démocratie”.

  103. I (uncharacteristically) refrained from remarking on this in another recent thread, but no longer:

    Does English practice generally respect that Hungarian postconsonantal -y denotes palatalization? As in György Ligeti or Zoltán Kodály? (I guess the closest approximations with English phonemes would be something like [dʒerdʒ ‘ligeti] and [‘zolta:n ‘koda:j]).

    I know I have heard [kʰodaly] even on the classical channel of Danish public radio.

  104. I’ve always heard [‘koda:j] from American classical announcers, so it must be in all the pronunciation handbooks they’re trained on and doubtless keep at their desks. Can’t say about György — I’ll have to listen for it, not that his music is played very much.

  105. So which is older in English, “redoubtable” or “redoubt”? I would have guessed the latter, but I’m not sure. (And I don’t have OED access from home.)

  106. The former goes back to 1421 (H. Luttrell in H. Ellis Orig. Lett. Eng. Hist. 2nd Ser. I. 84: “Redoubtabel and souverain Lord”), the noun only to 1608 (F. Vere Comm. 4: “Because there were upon it certain small redoubts held by the enemie, we took along with us two small field-pieces”) — but there’s an adjective redoubt meaning ‘revered; noted, distinguished; formidable’ that goes back even earlier, to 1417 (in H. Ellis Orig. Lett. Eng. Hist. 2nd Ser. I. 54: “Righte excellente, righte gracious, and our righte redoubt and righte soveraiyne leige Lord”).

  107. whose origin is this ré- (the French form)

    There is no accent on the -e- in French redoutable.

  108. I’ve just read the whole thread (quickly), and unless I skimmed over it, I don’t think the matter of “Qatar = Gutter” ever got past the suggestion it was an insult. Actually I first heard it from regional specialists on PBS, and I believe it’s the case that qaf is normally pronounced [g] in Gulf Arabic. At any rate it’s no worse than calling the place Catarrh, is it?

  109. Oh, as far as I know it’s perfectly fine; “gutter” and “cutter” are the most accurate available pronunciations in English. It’s just unfortunate that both have distracting soundalikes. (Cf. Uranus.)

  110. Indeed. Having dodged the Scylla of urinous and the Charybdis of your anus, what’s left: you’re a noose, perhaps?

  111. Surprised that Yaghnobi isn’t discussed in more details on LH. A new genetics study claims that they possess almost unaltered Sogdian DNA – while the Tajiks have a mysterious substantial admixture of a South Asian-related DNA.

  112. I know that when Russians came to Central Asia, India was the main trading partner for Bukhara – and I think for today’s Tajikistan too (not a source for admixture, but still implies contact).
    I do not know how it worked practically, though.

  113. Dmitry Pruss says

    Supposedly Al-Biruni in his treatise on the 1000s India lamented that the Hindus started to treat all foreigners with suspicion after their cities were conquered and people, sold as slaves to the greater Persia by the Turks. In the Islamic societies, the slaves eventually dissolved in the populace, which is why today’s Muslim Egyptians or South Arabians have substantial Sub-Saharan African DNA component, but Copts and Jews, none.

  114. Yes, but :

    “Similar to Zhabagin et al. work [65], the present study shows no impact of the Arab cultural expansion in Central Asia on the Indo-Iranian speaker’s genetic diversity, despite the first one leading to a shift in language for Tajiks. We also do not see a gene flow from Iran despite the Persian cultural expansion which led to a language shift from an east-Iranian language to a west-Iranian in Tajiks—when Yaghnobis kept their east-Iranian language [66].”

    I am not sure I understand the first part (“the first one”?).

Speak Your Mind