Eyjafjallajökull fail

« previous post | next post »

OK, how do YOU pronounce Eyjafjallajökull?

Andy Newman and Bao Ong, "Iceland Volcano Spews Consonants and Vowels", NYT 4/16/2010, offer a selection of attempts by people they asked in the Times Square subway station. For example:

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

They end with "the lowdown from a native speaker at the Icelandic consulate, who would give only her first name, spelling unknown but pronounced Becca":

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

Becca's final, hyper-carefully-articulated performance is presented separately here:

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

Jóhann Heiðar Árnason has contributed a more fluent Ogg Vorbis pronunciation file to Wikipedia. A transcoded .mp3 version is here:

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

A youtube clip leads with Jóhann's version, and then offers a cavalcade of broadcast news presenters, whose attempts are, well, somewhat more confident than those of the Times Square passers-by:

And Neal Conan on NPR's Talk of the Nation (broadcast April 19), has clearly done his homework (though he still gets the first vowel wrong), but he still has a bit of trouble getting from one end of the word to the other:

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

The Icelanders are having a good laugh about it all:

Mark Memmot, "Can You Say 'Eyjafjallajokull'? Icelandic Volcano's Name Is A Tongue-Twister", NPR 4/16/2010, offer an "AP audio clip of Rognvaldur Olafsson, chief inspector of Iceland's Civil Protection Agency, saying "the Eyjafjallajokull volcano" rather rapidly:

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

I've attempted to cut out just the volcano name itself in this clip:

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

(And apparently his name, with the proper diacritics, should be Rögnvaldur Ólafsson.)

The BBC's pronunciation guide entry (AY-uh-fyat-luh-YOE-kuutl (-uh)), and the IPA given in the Wikipedia article ([ˈɛɪjaˌfjatlaˌjœːkʏtl̥]), both seem a bit more enthusiastic about the last vowel than Jóhann and Rögnvaldur were. And I'm skeptical of some of the other vowel transcriptions as well.

Another (apparently native) hyperarticulated (and clipped in recording) pronunciation is here:

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

(This last one has a distinctly different (backer) third vowel than the others.)

Those of you with IPA skillz, how would you render the native pronunciations given in this post?

And another question. In English, we don't generally try to imitate exactly the native pronunciation of foreign proper names, any more than speakers of other languages do for their own borrowings. But especially for unfamiliar names, where there isn't a traditional anglicization, we generally make a sort of half-way attempt. What should the attempt be for Eyjafjallajökull?

[See here for some further (and more serious) discussion by David Shaw, who gives this account of the morphological decomposition:

Eyja is the genitive of ey – meaning 'island'. It's actually the same word as the 'is' in our word island, which explains why the 's' is silent (it comes an erroneous spelling modification in the 15th century). But ey=island is very familiar to us from the names of many an island, like Anglesey (Englishman's island), Sheppey (sheep's island) or Islay (island's island!), also the famous 'new-born' island of Surtsey, just off the south coast of Iceland.

fjalla is the genitive of 'fjöll' (they love their umlauts, those Icelanders) – meaning mountain – and is closely related to the Yorkshire word for mountain/hill – fell. Non-Yorkshire souls (do people from outside 'God's own county' have souls?) might have heard of the outdoor pursuits of 'fell walking' and 'fell running' (insofar as they've heard of anything from Yorkshire that isn't pudding or terriers).

jökull – meaing glacier – is the diminutive (!) form of an Old Norse word meaning 'piece of ice', but, etymologically speaking, it is the same as the 'icle' part of our word icicle, which it still sounds a bit like.

David's suggested anglicization: Eh-ya-fyat-la-yuh-cuttle.

Erlendur in the comments offers this correction: "Eyja is in fact the genitive of eyjar – the plural of ey, so it means islands. Fjalla is indeed the genitive of fjöll, however it means mountains as fjöll is the plural form of fjall." So I guess Eyjafjallajökull means "glacier of the island mountains"? ]

[For some Eyjafjallajökull volcano science,follow the link.]

[As to where those pre-stopped laterals come from, see Pétur Helgason, "Phonetic Variation as a Source of Historical Sound Change: Examples From a Data Base":

In analysing spontaneous speech data from a female speaker of Central Standard Swedish, I have encountered two phonetic processes that are reminiscent of historical sound change in Icelandic. First, this informant frequently preaspirates her unvoiced stops, which in Icelandic is phonologically obligatory. Second, in her speech there are frequent occurrences of emergent [d] between vowels and laterals (i.e. pre-stopped laterals), a fact which bears a striking resemblance to the historical change ll > dl in 16th century Icelandic (the timing is debatable, though). These facts strongly suggest that sound changes where preaspiration arises and [d] emerges between a vowel and a lateral have their roots in the detailed phonetic variation of online speech.

But these variations must be tendencies that are somehow in the cultural DNA of Scandinavian languages, since I've never seen such things in English or any of the other languages I've worked on. ]

[Sarah Hawkins, marooned in Romania, writes:

There are preaspirated stops in Geordie English (Newcastle on Tyne) as well as quite a few current Scandinavian accents. The connection for Geordie is of course obvious, but it would be interesting to see if it's there in say Minnesota, which has some distinctive aspiration, to my ears.

And for the stop before a liquid, I've seen them here and there. I have also seen EPG evidence in a Madrid Spanish speaker of consistent use of complete stop at the onset of what should phonologically just be an alveolar trill (i.e. Spanish /r/).

It's amazing how much we (well, anyway, I) don't know about English and other well-studied languages…]

{Update 4/19/2010 — Another native-speaker performance, from ABC News, is here:

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

And for more analytic detail about those voiceless lateral fricatives, see this morning's post from John Wells, and my own "A little Icelandic phonetics".]



127 Comments

  1. Sverre said,

    April 17, 2010 @ 1:20 am

    I find Jóhann's pronunciation to fit very well with the IPA transcription given. Rögnvaldur (should be spelled Rögnvaldur Ólafsson) speaks really fast, though.

  2. Daniel von Brighoff said,

    April 17, 2010 @ 1:28 am

    All talk of pronunciation aside, the pleonasm "Eyjafjallajökull glacier" (jökull already means "glacier") is on its way to becoming a peeve for me. But I suppose if I can accept "The La Brea Tar Pits", I can learn to live with this.

  3. Troy S. said,

    April 17, 2010 @ 1:46 am

    I'm glad to hear Farsi isn't the only language with unpronounceable consonant clusters at the ends of words. It seems to me the final /tl/ gets reduced to an aspirated /t/ or maybe metathesized to /lt/ on that last link.

  4. Sili said,

    April 17, 2010 @ 1:59 am

    As I Dane I'm ashamed of how far off I was in my attempt before listening. I had no clue <ll> was /tl/, nor that <u> was fronted.

    At least I do recall from scholl – lo, these many years ago – that "jökull" means glacier (/'glæʃɪə/), even if I hadn't heard of this particular one.

  5. Alissa said,

    April 17, 2010 @ 2:19 am

    The final /l/ is voiceless, so it's there but hard to hear. Also, u is phonemically /ʏ/, though it seems to get reduced quite a bit in fast speech. I don't have the skillz to actually do a transcription though.

  6. Robert T McQuaid said,

    April 17, 2010 @ 3:39 am

    Call it Eyjafjalla glacier, and reduce the problem by 37%.

  7. Dierk said,

    April 17, 2010 @ 3:48 am

    I decided to go with 'icelandic volcano'. Close runner-up [I live in Northern Germany, we are inundated with completely meaningless and unimportant "news" on grounded flights]: Stupid non-news.

  8. Simon Cauchi said,

    April 17, 2010 @ 3:55 am

    Eyjafjallajökull: yet another name I can't pronounce, like Popocatépetl (the pronunciation of which, given in the Oxford Dictionary of the World, is obviously anglicized).

    glacier (/'glæʃɪə/): who says? According to my dictionary, it's /'glæsɪə/ or /'gleɪsɪə/

  9. Army1987 said,

    April 17, 2010 @ 5:55 am

    On Italian TV, they don't even *try* to pronounce its name, they just say "a volcano in Iceland".

  10. Army1987 said,

    April 17, 2010 @ 5:59 am

    And the vowels in the native pronunciation don't sound *that* fronted to me; I'd go with jɞkʉ̞tl̥ or something like that.

  11. Anthony Hope said,

    April 17, 2010 @ 6:09 am

    None of the audio clips in this post are accessible from Mobile Safari on the iPhone because the iPhone OS doesn't support Flash. Please consider an alternative next time.

    [(myl) I certainly hope that Steve Jobs heeds your plea, or that the world's creators of operating systems and browser software someday get it together to create an audio-output technique that actually works on all platforms. For the moment, flash is by far the most widely-available and reliable alternative, I'm afraid. (And please, don't tell me about the HTML 5 audio tag, at least not without actually trying it in the real world.)]

  12. Jason! said,

    April 17, 2010 @ 7:43 am

    As retribution, we should get a volcano in Philadelphia and listen to how Icelanders pronounce it as opposed to natives. Most of the time I hear it, the locals pronounce it "Fluff-ya". I'm betting similar things are happening in Iceland.

    [(myl) Good point. Hard to arrange the volcano part, though. Also, what with the morphemes being originally Greek and all, foreigners mostly think they know how pronounce our city's name.]

  13. martin jamison said,

    April 17, 2010 @ 7:56 am

    ayurvedic yogurt

  14. MattF said,

    April 17, 2010 @ 8:00 am

    I can see I'll never be a linguist– the notion of pronouncing that string of letters never even occurred to me.

  15. language hat said,

    April 17, 2010 @ 8:39 am

    the pleonasm "Eyjafjallajökull glacier" (jökull already means "glacier") is on its way to becoming a peeve for me.

    So your theory is that all English-speakers should acquaint themselves with the vocabularies of all other languages so as to avoid any possible "pleonasm" (to use that word in a very odd sense)? Do you avoid saying "the Alhambra" because al- means 'the' in Arabic? If not, why not?

  16. Peter Taylor said,

    April 17, 2010 @ 9:25 am

    myl wrote :

    (And please, don't tell me about the HTML 5 audio tag, at least not without actually trying it in the real world.)

    Used with fallback?

    [(myl) Sorry, the (extraordinarily complicated) code-sample at that site fails to play on the machine I used to write this post, and on which I'm writing now (using a current version of Firebox on a current version of Ubuntu Linux).]

  17. DaveK said,

    April 17, 2010 @ 9:43 am

    Go with the tabloids. You just know that the NY Post, Phila. Daily News, etc, will have headlines like BIG EY STILL SPEWING or something to that effect.

  18. dearieme said,

    April 17, 2010 @ 9:49 am

    In Britain it seems to have become Mt Unpronounceable.

  19. rootlesscosmo said,

    April 17, 2010 @ 10:19 am

    Say it to music:

    http://www.youtube.com/watch?v=GITm8J8LfWc&feature=player_embedded

    (courtesy Alex Ross) (cross-posted to Language Hat)

  20. Meg said,

    April 17, 2010 @ 10:19 am

    And now please do a post about Icelandic morphology. There are very interesting word forms in the news of last week, most of them wrong. Seems like they are using copy&past without checking wikipedia first. ;-)

    eyjafjalla: 2951results –> genitive
    eyjafjöll: 182 results–> nominative
    eyjafjallajökli: 2 results–> dative (one correct, congrats aftenposten, though a little weird in a non-Icelandic text…)
    eyjafjallajökul: 139 results –> accusative

  21. Stephen Jones said,

    April 17, 2010 @ 10:24 am

    I certainly hope that Steve Jobs heeds your plea, or that the world's creators of operating systems and browser software someday get it together to create an audio-output technique that actually works on all platforms

    Firefox won't play the Flash file though it will play in IE tab.

    You keep asking for browsers to have media players included. Are you happy with the amount of bloat this would entail?

    [(myl) Less support would be required for audio players than for static image display, or for the rendering of complex fonts (of course, that's still not working right across the board either)…

    But anyhow, whatever it takes, consistent audio and video presentation across browser/OS combinations would be a long-overdue step forward.]

  22. [links] Link salad slumbers beneath an Icelandic glacier | jlake.com said,

    April 17, 2010 @ 10:38 am

    […] OK, how do YOU pronounce Eyjafjallajökull? […]

  23. Daniel von Brighoff said,

    April 17, 2010 @ 10:39 am

    So your theory is that all English-speakers should acquaint themselves with the vocabularies of all other languages so as to avoid any possible "pleonasm"

    Not all speakers. But it's not asking too much for those reporting on an event in Iceland to learn a thing or two about Icelandic, surely? In theory, journalists should have better access to experts than the general public–if they only cared to make use of it.

    I see your "the Alhambra" and raise you a "Mount Fujiyama". Everyone's got a point at which these sorts of tautologies start to annoy them; my threshold's probably just a bit lower than yours.

  24. Nathan said,

    April 17, 2010 @ 10:40 am

    Chumley.

  25. Pau Amma said,

    April 17, 2010 @ 10:44 am

    Re soundclips: maybe you could also give links to the naked MP3s?

    [(myl) Here you go:

    Times Sq 1 / Times Sq 2 / Times Sq 3 / Begga 1 / Begga 2 / Jóhann Heiðar Árnason
    Rögnvaldur Ólafsson 1 / Rögnvaldur Ólafsson 2 / Last one

    ]

  26. Stephen Jones said,

    April 17, 2010 @ 10:56 am

    Less support would be required for audio players than for static image display, or for the rendering of complex fonts

    The support for playing one format would not be that bloated, but nobody is happy with that (FIrefox already plays Ogg files natively but it is pretty hard to find them, and there's no serious demand for IE to play them natively). The situation is not as bad as for the video codecs (I do agree with you that flash is best for video, though I would say .mp3 is the best supported for audio) but to play the various number of audio file formats on the web would require a considerable amount of programming, and I fail to see the advantage over having a plug in).

    [(myl) What matters is the behavior. I want to be able to put a simple formula in an html page that will cause audio to be played reliably for everyone who interacts with the page (with the usual control via keypress etc.), without users having to worry about how they've configured their system, which non-standard plug-ins if any they've installed, etc.

    I'd be a lot happier than I am now if this were true for EVEN ONE audio format and EVEN ONE method of controlling it — but it isn't,and so I'm not. The best solution that I've been able to find is to use embedded flash players, which don't work on iPhones or iPads, thanks to Mr. Jobs, and perhaps not on some other mobile platforms as well, I don't know.

    I think that you're exaggerating the difficulties of providing for the standard range of audio formats — there are several open-source libraries, with excellent coverage, whose binary size is small compared to the things that people routinely load into programs these days. And we're used to being able to display .gif, .jpg, .png, .tiff, etc. etc. in html pages without worrying about who has what browsers, oeprating systems, plug-ins, preferences, etc.

    But if (say) .mp3 files would work reliably across browsers and operating systems, I'd stop complaining. As it is, the industry deserves a collective D-, in my opinion, for leaving things in such a mess after more than 15 years.]

  27. Cali said,

    April 17, 2010 @ 10:57 am

    In Spain, as in Italy, it's pronounced "Volcano in Iceland". I do think it's the most pragmatic way; it won't be news for much longer than a few weeks (I hope).

  28. Erlendur said,

    April 17, 2010 @ 10:58 am

    Sorry for my nitpicking, but I must point out that David Shaw's explanation is not entirely correct. Eyja is in fact the genitive of eyjar – the plural of ey, so it means islands. Fjalla is indeed the genitive of fjöll, however it means mountains as fjöll is the plural form of fjall.

  29. language hat said,

    April 17, 2010 @ 11:10 am

    But it's not asking too much for those reporting on an event in Iceland to learn a thing or two about Icelandic, surely?

    Yes, of course it is; what an absurd idea! You think, then, that reporters covering the Chechen Wars should learn a thing or two about Chechen? And you didn't answer my question about "the Alhambra." Do you use it, or have you arbitrarily decided that your principles don't apply to it? And do you avoid discussing places with whose language(s) you are unfamiliar? Of course you could memorize Wikipedia's List of tautological place names and avoid (for instance) the sin of talking about the "Paraguay River," but as Wikipedia itself warns, "This list is incomplete."

    Everyone has a right to their own crotchets and inconsistencies, but to try to impose them on the rest of humanity, or criticize those who don't go along with them, is a Bad Thing.

  30. Ryan Denzer-King said,

    April 17, 2010 @ 11:25 am

    For the slower pronunciation I get [ejafjatłajøkʰʏtł]. For the more rapid version it sounds more like [ejafʲatłaʏkʏtł]. As for an Anglicization, I'd go for [ej.jəf.jat.lə.jow.kʰətl].

  31. Stephen Jones said,

    April 17, 2010 @ 11:30 am

    What matters is the behavior. I want to be able to put a simple formula in an html page that will cause audio to be played reliably for everyone who interacts with the page (with the usual control via keypress etc.), without users having to worry about how they've configured their system, which non-standard plug-ins if any they've installed, etc.

    Which amounts to having an audio player included in the browser.

    Now I suspect you may be right to say that this would be a better solution than having a plugin, but the plugin architecture is already there. I also suspect the unreliability is more hardware, specifically network, related than a question of unreliable software. If you preloaded the file it would probably always play, but waiting for a 30MB mp3 file to download before you see the page isn't everybody's idea of fun.

    I still maintain that Flash is best for videos (Apples refusal to allow Flash is pure bloody mindedness that will cost it dear) but I very much doubt it's more reliable for soundfiles.

  32. Stephen Jones said,

    April 17, 2010 @ 11:32 am

    And we're used to being able to display .gif, .jpg, .png, .tiff, etc. etc. in html pages without worrying about who has what browsers I think you'd have a problem viewing them in Lynx :)

  33. Adrian Morgan said,

    April 17, 2010 @ 11:50 am

    Those of you with IPA skillz, how would you render the native pronunciations given in this post?

    Well, I wish you'd numbered the recordings for ease of reference in answering this question. But anyway, I'm going to take the question seriously and pretend I'm a student in your undergraduate phonetics class. My answers are honest best attempts, and errors are likely to be ones I could learn from. I'm rounding off most phones to those found in my native English because that's how I hear them.

    The recording following "Becca's final, hyper-carefully-articulated performance":
    [eijɐfjɐtlɐjʉgʊtˡ]

    The recording following "Jóhann Heiðar Árnason has contributed" is beyond my competence, so I'm going to skip it.

    The recording following "I've attempted to cut out":

    [ejəfənʉt]

    The recording following the paragraph beginning "Another (apparently native) hyperarticulated (and clipped in recording) pronunciation":

    [eijɐfjɐtlɐˌjʉkwətˡ]

    Suggested anglicisation for print: "EI-ya-fjat-la-YUU-quot", accompanied by footnote explaining lateral release.

  34. Bobbie said,

    April 17, 2010 @ 12:49 pm

    If we are going to pronounce places names "as the natives do" then we need to say Deutschland, not Germany; Kohn instead of Cologne; and my city of Norfolk Viriginia as Naw-fuhk, not Nor-Folk or North-Fork.

  35. Bobbie said,

    April 17, 2010 @ 12:49 pm

    Sorry, I should have written Koln.

  36. Jerry Friedman said,

    April 17, 2010 @ 12:58 pm

    I like Ryan Denzer-King's suggested anglicization, but I'd be tempted to suggest [ʊ] for the last two vowels.

  37. [Surftipp] Language Log und der Eyjafjallajökull « [ʃplɔk] said,

    April 17, 2010 @ 1:21 pm

    […] Language Log und der Eyjafjallajökull Heute nur ein schneller Surftipp: Mark Liberman hat beim Language Log über die Aussprache von Eyjafjallajökull (den lustigen, aschespuckenden isländischen Vulkan) […]

  38. Dan T. said,

    April 17, 2010 @ 1:41 pm

    Wikipedia insists on using only Ogg audio, despite it being less widely supported than MP3 (for instance, it doesn't work on iPods/iPhones), because Ogg is open-source, open-standard, while MP3 is patent-encumbered.

    Audio and video formats unfortunately have a byzantine array of different camps supporting or opposing them for a variety of commercial and ideological reasons, getting in the way of compatibility. The big companies have their proprietary platforms they want to push, while some of the geeks are opposed in principle to proprietary stuff and want an open standard. Then there are the pushers of a proprietary standard who refuse to support somebody else's proprietary standard because it competes with theirs, which is where you get the Apple / Flash wars.

    [(myl) Exactly the same things could be said about image formats, but (luckily for the fate of the internet) the industry stumbled into a solution. Unfortunately, this didn't happen for audio. BS excuses to the side, it's basically just a case where sensible people didn't get a standard established before various larger players got involved, and the various larger players proceeded to play out a version of the Prisoner's Dilemma in which (nearly) everybody defaulted. It now looks like a solution might emerge within a few years, but I wouldn't underestimate the ability of the various parties involved to screw things up again.]

  39. Stephen Jones said,

    April 17, 2010 @ 2:00 pm

    Sorry, Mark, but what you're saying doesn't make much sense to me.

    There are pretty well as many image formats as sound formats. In practice two image formats, .gif and .jpg rule supreme on the net because they come out standardized. Nothing to do with the owners of either patent.

    Every plug in and media player plays .mp3s yet there are still problems. As far as I can tell the problem is not proprietary formats but download size or the way the software deals with them.

    [(myl) We seem to agree on the premises but disagree on the conclusions. You claim that html access to audio is a screwed-up mess because there are lots of formats and it would be complicated for browsers to deal with them. I point out that there are lots of image formats as well, and it's complicated to deal with them, but html treatment of images is pretty good; so your excuse for the bad state of html audio makes no sense. (And anyhow, reasonable treatment of even one format would be a big step forward.) You agree with me about image formats, and at this point, I'm puzzled about what point you're making.]

  40. Suzanne Kemmer said,

    April 17, 2010 @ 2:26 pm

    About those pre-stopped laterals, Mark wrote:

    "But these variations must be tendencies that are somehow in the cultural DNA of Scandinavian languages, since I've never seen such things in English or any of the other languages I've worked on."

    You can hear some pretty close analogues in Yorkshire English. When I first met my husband M.B., who is Yorkshire born and bred, I noticed that he pronounced 'full' with a quick alveolar stop or flap before the velarized l. I used to tease him and imitate his pronunciation as 'fuddle' pronounced with a Yorkshire short u (as in 'put'). Same with 'pull', 'wool' and other words that end in short-u + l. I've noticed this pronunciation with many other speakers from the north of England although I don't know the distribution of this dialect feature.

    A sound change that came from Scandinavia? Medieval language contact, or even a later spread? As Bernard Comrie once told me (he comes from that coast), people have been going back and forth across the North Sea a long time, right into modern times. Despite the Sprachgrenze between West and North Germanic that has long divided the languages, sound changes/areal features could have affected Yorkshire long after the Viking period.

    [(myl) Fascinating.]

  41. George Amis said,

    April 17, 2010 @ 2:58 pm

    The anglicizations suggested by Ryan Denzer-King, Jerry Friedman and Adrian Morgan are certainly easier for native English speakers, but they're still pretty alien. I think most US and British TV and radio announcers would have a much better chance of making themselves understood with a more full-blown English version, like Eyjafjall (or even Eyafell) Volcano. But I suppose it's much too late for anything like that.

  42. Beth G. said,

    April 17, 2010 @ 3:36 pm

    Terribly interesting, Suzanne. Icelandic isn't the only place pre-stopping of sonorants shows up: Faroese has it, as does Manx and Cornish (and some dialects of Northern English?). Realizations of pre-occlusion, as I've heard Celticists call it, varies between those languages. It only occurs before dental sonorants (/l/ and /n/) in Icelandic and Faroese but apparently with /m/ as well in Cornish; in Icelandic the inserted stop is voiceless while it sounds voiced to me in Faroese.

    I know nothing about the histories of Manx and Cornish, so this is baseless speculation, but pre-stopping would appear to be an areal feature. At least, it would be odd if this rather unusual phenomenon arose independently in more than one of those geographically not-to-distant languages.

  43. Simon Cauchi said,

    April 17, 2010 @ 3:50 pm

    Bobbie should have written Köln.

    How sensible the Spanish and Italian news readers are!

  44. Hershele Ostropoler said,

    April 17, 2010 @ 3:55 pm

    So whence /-tl-/ from ? Same place as Welsh?

  45. Stephen Jones said,

    April 17, 2010 @ 4:21 pm

    Not sure I do disagree with you about image formats. I convert everything to .jpg or .gif because other formats don't display consistently.

    Another point is that images render all at a time. Sound files have to render sequentially.

  46. Bloix said,

    April 17, 2010 @ 4:54 pm

    The pronunciation isn't particuarly hard. The problem is the orthography.

    First of all, if this were an English name for a geographical feature, it would be probably three words – "Mount Isle Glacier" most likely. If you saw "Mountisleglacier" you woudn't know how to pronounce it, either. So the convention of running the words together freaks us out right away.

    Secondly, the 'J' throws us off. But if you've got any sense of any Germanic language you're likely to guess that J is close to English Y. I did, and I sounded it out as "AYE-yah FYAH-la YIR-kul."

    But I wasn't happy with that because I had a sneaking feeling that the double L might have some special pronunciation. And indeed it does – apparently it's "TL," which is unguessable.

    So the correct pronunciation is "AYE-yah Fyat-lah Yir-kutl."

    I say apparently because I hear that final TL only in the example from Rognvaldur Olafsson. What I hear in all the others is something between Yirkik and Yirkich. Perhaps it's an aspirated T, which English doesn't have at the end of words.

    If you were to say, "Aye-yah Fyat-lah Yir-kut," which is easy, you'd be doing a good enough job to be understood by Icelanders. But there's no way that we English speakers can get those sounds out of the orthography. The spelling messes with our heads and we decide it's unpronounceable.

  47. Bloix said,

    April 17, 2010 @ 5:06 pm

    Ah, I see that Troy S. has already suggested that the final TL becomes an aspirated T.

  48. Anna R. said,

    April 17, 2010 @ 5:13 pm

    I noticed msnbc has come up with 'Eyjafjöll Glacier' in place of Eyjafjallajökull. Seems very sensible to me. In fact, I'm very impressed. It seems they went to the trouble of finding out what the nominative pl. is.

    Anna

    PS #1: About your friend at the Icelandic consulate: 'Begga' is a common enough nickname in Iceland, a short for Berglind, Bergljót, Bergdís and probably other names as well.

    PS #2 Most Icelanders pronounce 'Eyjafjallajökull' the way Rögnvaldur Ólafsson and Jóhann Heiðar Árnason do, for sure. Rural people may say the word a tad more slowly. But 'the hyper-carefully-articulated performance' sounds unnatural to say the least.

  49. David Eddyshaw said,

    April 17, 2010 @ 5:38 pm

    Scots Gaelic has preaspirated stops too; basically the consonants written with the symbols p t c are voiceless and preaspirated after a vowel word internally , and (post) aspirated initially. The consonants written b d g are in fact voiceless and unaspirated.

    I don't know all that much about Icelandic but the consonantal system seems to be pretty similar in this respect. Given the high degree of contact in mediaeval times between the Gaels and the Norsemen (to say nothing of the high contribution of Gaels to the Icelandic settler population) this might well be more than coincidence.

  50. JimG said,

    April 17, 2010 @ 5:50 pm

    It may indeed be worth learning to say the name. The micro-particles of ash up in the jet stream could stay aloft for years and continue to hazard air transport, which didn't exist when Krakatoa blew.
    http://www.nytimes.com/2010/04/16/opinion/16winchester.html
    How many words do we have for maritime and surface transportation of passengers and freight?

  51. JKD said,

    April 17, 2010 @ 6:08 pm

    I see your "Mount Fujiyama" and raise you "the hoi polloi"….

  52. James C. said,

    April 17, 2010 @ 6:18 pm

    If I were going to borrow the word into Tlingit then I’d probably have something like Éiyaxwadlayúkutl /éːjaxʷatɬajúkʰutɬ/, which is pronounced like [éɪ.jə.xʷə.tɬə.jʊ́.kʷʰʊtɬ]. Except for the substituted /xʷ/ for /fj/ (Tlingit lacks labial consonants) it’s not a bad approximation, I think. The lateral affricates are pretty obvious to my ear. The penultimate stress is converted to tone which is typical; I’m not sure why I hear tone on the initial syllable. (I’m not a native speaker of Tlingit, just a learner and linguist who’s worked a lot on loans.)

  53. George Amis said,

    April 17, 2010 @ 6:19 pm

    Apologies to Robert T. McQuaid, who suggested Eyajfjalla Glacier before I suggested something almost identical.

  54. David Eddyshaw said,

    April 17, 2010 @ 6:28 pm

    Indeed, only the hoi polloi say Mount Fujiyama

    http://www.japaneselifestyle.com.au/travel/mount_fuji_name.html

  55. Eimear Ní Mhéalóid said,

    April 17, 2010 @ 6:29 pm

    Suzanne, urban Dublin speech often shows similar effects (parodied by fictional character Ross O'Carroll Kelly when he has working class Dubliners pronounce the name of the "Herald" as "Heddild"). And, of course, in the Viking era Dublin was an important Viking kingdom with strong links to York. All very speculative, but it's fun to speculate.

  56. Breffni said,

    April 17, 2010 @ 7:14 pm

    Eimear, I can see how that might be related, but isn't it restricted to /ɹəl/ syllables ("herald") and /ɹl/ ("curl", "world") clusters? Between /ɹ/ and /l/, in other words. I can't think of counterexamples.

  57. David Conrad said,

    April 17, 2010 @ 7:23 pm

    I had been saying it "Ay-ya-fyall-uh-yikool". I'm glad to know the correct pronunciation although I can't quite get my tongue around it. It sounds to me like "kuts" or "kutsch" in the last syllable in the recordings from the native Icelanders, so it must be an aspirated t.

    [(myl) I think that the final orthographic 'll' is pronounced [tl], where the [l] should have an open circle under it, indicating voicelessness (various browsers and operating systems are likely to render the underdot in various inappropriate places, so I won't confuse you by trying to put it into the IPA string…). This seems to be pronounced as a [t] with a lateral (i.e. l-like) voiceless release. Treated as single segment, it could be treated as a lateral affricate — but this is like the difference between (say) [tʃ] treated as a stop+fricative sequence vs. as an affricate.]

    I'm glad we've given them something to laugh about. They can laugh about our pronunciation while we laugh about their banking system.

    [(myl) It's not clear to me who has had the last laugh about Icelandic "banking".]

  58. Breffni said,

    April 17, 2010 @ 7:28 pm

    Dierk: "completely meaningless and unimportant 'news' … Stupid non-news" – how glad I am not to live on your planet, where the closing of vast tracts of airspace for an indefinite period, with millions of people stranded, counts as banal.

  59. Are languages important? - Page 7 - World Literature Forum said,

    April 17, 2010 @ 7:36 pm

    […] find it hard to understand the language of their ancestors. Harry with sound files Language Log Eyjafjallaj

  60. Buck Ritter said,

    April 17, 2010 @ 8:41 pm

    LanguageHat said: "Everyone has a right to their own crotchets and inconsistencies, but to try to impose them on the rest of humanity, or criticize those who don't go along with them, is a Bad Thing."

    Why? Why do you, in your capacity as a linguist, attempt to ridicule if not to purge anyone who does not march in lockstep with your anarchism? Do you not see how childish such "descriptivist" nihilism is?

  61. Mike Farrell said,

    April 17, 2010 @ 9:01 pm

    Fascinating post and thread! Thanks, all. (Even if I did skip over the 'puter stuff–it made my head hurt.)

    As everyone knows, it's phil-DELPH-ya.

  62. W. Kiernan said,

    April 17, 2010 @ 9:26 pm

    Really? Two ells spelled out equals a tee and a cymbal tap spoken? Hell, that isn't linguistics, it's cryptography.

  63. Hellga said,

    April 17, 2010 @ 9:46 pm

    A of y'all a yokel.

  64. Claire Wahrhaftig said,

    April 17, 2010 @ 10:09 pm

    Every locale has its own special way of pronouncing itself. Here in San Francisco the locals (like myself, and Sen. Dianne Feinstein) say Samp'ncisco. Then again our native born Austrian Governor's pronunciation for our state is always good for parody, Kal eee for neee yah, yes it is fery fery cherman.

  65. Julia S. said,

    April 18, 2010 @ 12:38 am

    I myself would not die rather than say "Paree".

    One of many ways in which I am unlike Winston Churchill.

    On the other hand, I don't go around saying "Paree" when I'm speaking English…I don't say "the hoi polloi" myself because I know Greek and so it sounds weird to my ear, but I don't freak out when other people say it. Much.

  66. dw said,

    April 18, 2010 @ 12:55 am

    @Troy S:
    I'm glad to hear Farsi isn't the only language with unpronounceable consonant clusters at the ends of words

    English is no slouch when it comes to consonant clusters. I can't imagine that there are many non-native English speakers who could easily pronounce, say, the /ksθs/ cluster in "sixths". Even most native English speakers simplify it.

  67. Stephen Jones said,

    April 18, 2010 @ 1:59 am

    It's not clear to me who has had the last laugh about Icelandic "banking".

    It seems to me the 'bankers' have, seeing they seem to be living very comfortably in Iceland or the UK.

  68. Jarek Weckwerth said,

    April 18, 2010 @ 6:47 am

    Regarding volcanoes and banking, reportedly there's already a joke about this. I'm not sure if the link works, so let me just copy from Iceland Review Online:

    Britain: Iceland are you crazy?!? Why did you send us volcanic ash ? Our airspace has shut down.

    Iceland: What? That's what you asked for isn't it?

    Britain: NO! We said cash! Cash you dyslexic idiot. CASH!

    Iceland: Woooops…

    There's also a comment that fits the "linguistics as misreported by the media/laypeople" thread:

    To the British and Dutch Governments: There is no C in the Icelandic alphabet, so when you ask for Cash, all you get is…

    Letters, sounds, who cares.

    Original link

  69. Vulkāniskā epopeja « Mente et malleo said,

    April 18, 2010 @ 10:14 am

    […] Vulkāniskā epopeja Publicēts Aprīlis 18, 2010 Uncategorized Leave a Comment Pēdējā laika notikumi pasvieduši lielisku tēmu ar ko sākt blogu. Karstākais notikums ziņās gan tiešā, gan pārnestā nozīmē ir vulkāna izvirdums Islandē zem Eyjafjallajökull ledāja. Nosaukumus šoreiz pat nedomāju tulkot, tāpat nesanāks, bet pareizais veids, kā to izrunāt atrodams šeit. […]

  70. Tom O'Brien said,

    April 18, 2010 @ 11:26 am

    You might try "The Amazing Slow Downer" software from ronimusic.com. I use it to slow down folk music I'm trying to learn by ear; it preserves the pitch. It might make some of the high-speed mp3s easier to understand.

  71. Garrett Wollman said,

    April 18, 2010 @ 11:43 am

    As far as the sound files issue goes, just what is wrong with providing plain links? Nearly every system is capable of handling them correctly (even if the developer hasn't licensed all the necessary patents to play the files back legally in the U.S.). I note that the RSS version of the post (which is nornally the only one I read unless I want to comment) has them, but unfortunately unconnected to where they are referenced in the text. (Writing as a user of yet another platform which will never have Flash.)

  72. Panu Höglund said,

    April 18, 2010 @ 1:28 pm

    I'm glad to hear Farsi isn't the only language with unpronounceable consonant clusters at the ends of words

    Icelandic, unpronounceable?

    Anyone calling Icelandic unpronounceable should be sentenced to a ten-years long 16-hours-a-day course of Georgian.

  73. Army1987 said,

    April 18, 2010 @ 1:56 pm

    Popocatépetl
    I had to try three or four times before being able to say [popoka'tepetɬ], and the affricate at the end wasn't the problem. So much for "CV is the easiest syllable structure to pronounce".

  74. Stephen Jones said,

    April 18, 2010 @ 2:30 pm

    As far as the sound files issue goes, just what is wrong with providing plain links?

    And next?

  75. Charles Perry said,

    April 18, 2010 @ 2:55 pm

    Claire — I lived in the Bay Area for 18 years (10 of them in what is known as The City in those parts), and I always heard San Francisco pronounced as Særsisko, with a nasalized "æ."

  76. Toni Keskitalo said,

    April 18, 2010 @ 3:28 pm

    Preaspiration of stops, Northern Sámi has it. It has also developed a homorganic stop in front of nasals inside a word. There's no prelateralisation of /ll/ as far as I can tell; there was probably no /ll/ in Proto-Sámi. A few examples, Finnish – NS.

    käte- (that's a stem; dictionary form "käsi") – giehta 'hand'
    rykiä – rahkat 'to cough' (final t is preaspirated, too)
    halla – suoldni 'frost' (a Baltic loan, originally /ʃalna/ [ascii-ipa /Salna/, FUT *šalna])
    jää – jiekŋa (protoform was /jæŋe/ [FUT *jäŋe])

  77. Simon Cauchi said,

    April 18, 2010 @ 4:46 pm

    @Army1987. My brother, long resident in North America, asked me what tree that was in our back garden. I told him it's a pohutukawa. He tried, but couldn't manage to pronounce that CVCVCVCVCV word.

  78. liamascorcaigh said,

    April 18, 2010 @ 5:59 pm

    Since we're seeing it every time we turn on the TV and the ash column rises to untold heights why not just call it "The Eyeful Tower"?

  79. Will Steed said,

    April 18, 2010 @ 7:45 pm

    Given that we have cognates of all of the parts, as you say, why not call it Eyfellicle – ['ej.fel.ik.l]?

    One of the reporters on the youtube video did pretty well with ['ejafatlaj@kul]. They should be commended for a thoroughly pronounceable anglicisation.

  80. for review: Pronouncing Foreign Names in English « Fully (sic) sandpit said,

    April 18, 2010 @ 9:02 pm

    […] a comprehensive post on Language Log, describing the difficulty of Anglicising a difficult and long foreign […]

  81. So how do you pronounce the icelandic volcano Eyjafjallajökull #ashtag | Volcano Help said,

    April 19, 2010 @ 6:39 am

    […] listen and learn from the natives. Here is a great guide to how you pronounce Eyjafjallajökull: http://languagelog.ldc.upenn.edu/nll/?p=2257 Related Posts:RT @periodicvideos: Our volcanic ash in jet engine video if you missed it […]

  82. Jongseong Park said,

    April 19, 2010 @ 8:39 am

    There's a discussion on Korean Wikipedia about how to render this name in hangul, the Korean alphabet.

    Four years ago I came up with a proposal for rendering Icelandic names in hangul. It has no official recognition, although the National Institute of the Korean Language followed the scheme (with one minor modification, which I follow for the current version of my proposal) for the official hangul transcriptions of the names of Icelandic athletes at the Beijing Olympics.

    People were trying to figure out how to render Eyjafjallajökull in hangul using my system, but got it wrong. According to my proposal, it should be 에이야피아들라예퀴들. In Revised Romanization of Korean, that is eiyapiadeullajekwideul. Korean doesn't have /f/, so it gets mapped to the aspirate p, and eu represents the epenthetic vowel in Korean. The wi is in fact the standard approximation for the close front rounded vowel (which is in fact an alternate pronunciation in Korean). Currently the Wikipedia article uses an ad hoc rendering based on the IPA transcription.

    For users of the roman alphabet, this may be a pronunciation issue, but for others, it's a spelling issue as well. And judging by the examples given on the discussion page, chaos is reigning in the Korean media with dozens of imaginative renderings of the Icelandic name.

  83. Army1987 said,

    April 19, 2010 @ 9:05 am

    No problem with pohutukawa. The fact that all the C's are different probably helped me a lot. (But for some reason, before getting to the end of the post I had taken -oh to be a digraph for a long vowel or diphtong, rather than /o.h/.)

    My suggested Anglicization: /'eI jA: fi at lA: ,jO: kUt/ (where eI = FACE, A: = PALM, i = happY, a = TRAP, O: = THOUGHT, U = FOOT).

  84. Woordlog » Eilandenbergengletsjer said,

    April 19, 2010 @ 10:37 am

    […] stemloos is, hoor je die bijna niet. Het lijkt dus of het woord eindigt op een t. De discussie bij Language Log over de uitspraak van Eyjafjallajökull doet bijna evenveel stof opwaaien als de uitbarsting […]

  85. Icelander said,

    April 19, 2010 @ 10:43 am

    You an perhaps try this on for size:

    EYa-fYAdla-YUkudl

    That looks like a rather decent approximation to this Icelander's eyes.

  86. Jane B said,

    April 19, 2010 @ 10:47 am

    And contributing with my mixed background to the intellectual discussion…

    Eh? The pronunciation is all, the dust is yet!

    At least it's easier to say!

    Loved the discussion, but my eyes glossed over when it got to computer-speak. I'm glad it's not just me.

    Being on dial-up, I got the 'speakings' in buffered parts. I live next to the Icelandic Consulate in my city and still missed that ending! My my!

  87. language hat said,

    April 19, 2010 @ 11:55 am

    Anyone calling Icelandic unpronounceable should be sentenced to a ten-years long 16-hours-a-day course of Georgian.

    Georgian? Hah! I see your Georgian and raise you a Circassian!

  88. Alles Asche | Daumenschraube.ch said,

    April 19, 2010 @ 12:36 pm

    […] waren und ihren Urlaub zwangsverlängern mussten. Gelernt habe ich auch etwas, denn dank dem Language Log weiss ich nun, wie man Eyjafjallajökull korrekt ausspricht. So tönts halt, wenn man Wikinger […]

  89. A volcano in Iceland called … « Nigel "Teacher" Caplan said,

    April 19, 2010 @ 1:28 pm

    […] on hand to explain how to say it (John Wells on his blog, and Mark Liberman on Language Log, and here again). The BBC's famed pronunciation unit — of course — has its own advice. NPR in the […]

  90. Eyjafjallajokull links / Liens à propos de l'Eyjafjallajokul | Islande 2010 said,

    April 19, 2010 @ 4:37 pm

    […] Une page spécialisée sur 'le volcan Islandais' comme les lâchent l'appellent / A specialized page about the 'Icelandic Volcano', as cowards name it :languagelog.ldc.upenn.edu/nll/?p=2257 […]

  91. Kijk vaker omhoog « Drabkikker said,

    April 20, 2010 @ 4:46 am

    […] omhoog te kijken: zoals u vernomen zult hebben loopt momenteel de Eyjafjallajökull (dat spreek je zo uit) de boel onder te assen, en dat gaat interessante dodelijk saaie zonsondergangen opleveren de […]

  92. Mark F. said,

    April 20, 2010 @ 6:29 am

    Re "the hoi polloi" — According to the MWDEU, the people who first used "hoi polloi" in English knew Greek and still chose to say "the hoi polloi". This makes sense because "hoi" will not be recognized in English as an article in the same way that "le" or "el" will be. I think it's better to put in the "the" than to leave it out.

    As for the actual topic of the thread, I don't think it's just the orthography that makes it hard to pronounce. Syllables in English don't start with /fya/, and (although Mark has beeing trying to help us with it) the ll's are trouble.

  93. Don’t Feel Bad, Icelanders Can’t Pronounce Their Volcano Either « American Elephants said,

    April 20, 2010 @ 7:21 am

    […] Language Log helpfully offers up audio of five native speakers pronouncing the name of the volcano, and, equally helpfully, they're all over the map! I count at least four different pronunciations.  So quit worrying, even the Icelanders can't pronounce the name of their own mountain. I am beginning to think it was named by a cat walking on a keyboard. Perhaps the world will come to speak of it in Potteresque terms — "the Mountain which must not be be named." […]

  94. Army1987 said,

    April 20, 2010 @ 1:27 pm

    @Daniel von Brighoff & JKD:
    All in!

  95. Neal Cornett said,

    April 20, 2010 @ 1:44 pm

    I live in Owensboro, Kentucky (USA), where, according to many of our fellow countrymen, we can't pronounce our own language, let alone someone else's. The recordings I've heard of Icelanders saying it sound to me as though they're talking about a lady named Ava Layvitz. I guess I just can't hear all the sounds with my southern ears. Curiously, a recording I found on the net of a Norwegian lady pronouncing it sounded pretty much as it's spelled (if the j is like an English y). My original attempt was AYuh-feeAla-YAY-kul, which ought to amuse somebody.

  96. Andrew Woode said,

    April 20, 2010 @ 5:06 pm

    I was wondering what a modern English cognate to Eyjafjallajökull would be, including a reflex of OE gicel, cognate to jökull, as the last element? It would be a more elegant way of producing a pronounceable version for international use.

  97. NGN Daily: Iceland’s Eyjafjallajökull Volcano | NetGreen News said,

    April 20, 2010 @ 8:16 pm

    […] came across this nifty site that will try to make you an expert at pronouncing Eyjafjallajökull. Go here to start brushing up on your Icelandic volcano pronunciation of Eyjafjallajökull. Share […]

  98. Link love: Language (16) « Sentence first said,

    April 22, 2010 @ 6:19 am

    […] song. Ogham stones of Wales. Foreign Accent Syndrome. Words for hens and chickens. How to pronounce Eyjafjallajökull. Public profanity and presumptions of purity. The development of language, in cartoon form. A short […]

  99. Army1987 said,

    April 22, 2010 @ 12:05 pm

    @Andrew Woode
    "(ic)icle", as stated in the post. So, let's make that "Islefellicicle", shan't we?

  100. Jongseong Park said,

    April 22, 2010 @ 1:19 pm

    Ooh, I want to play this game. What would a modern English cognate of Reykjavík be? The first element, meaning 'smokes' is possibly cognate to 'reek', but cognates of vík ('bay', 'inlet') survived in only a few English place names derived from Old Norse as 'wick'. Most instances of 'wick' in English place names of course mean 'dwelling' though and derive from Latin, not Old Norse… In any case, I propose 'Reekwick'.

  101. Florian Blaschke said,

    April 22, 2010 @ 3:47 pm

    A few comments to the typological convergencies around the Atlantic. Prestopping belongs to a number of innovations that have appeared in Scandinavian dialects on both sides of the Atlantic: Icelandic, Faroese, and southwestern dialects of Norwegian (but only L is prestopped there, not N).

    We do not know very much about the Norwegian-derived Scandinavian dialects in the British Isles but the little we know about Shetland Norn indicates that it was intermediate between Icelandic, Faroese and southwestern Norwegian and took part in those innovations. Interestingly, the innovations in question are attested only in the late medieval period (if memory serves right), long after the colonisation period, so they would seem to have spread through contacts across the ocean. Essentially, there seems to have been a dialect continuum/chain ranging from Icelandic via Faroese and Norn to (south-)western Norwegian (the point of origin of the colonisations, apparently), even in the post-medieval period. This indicates that the ocean was not such a great obstacle to communication as it is usually imagined to be – the same is true of the North Sea, by the way, where the languages around it have seen convergent developmental trends for centuries, especially in the Hanse age -, and contact continued after the emigration from Norway (we do know that from historical sources, as well). Also, the conservativity of Icelandic tends to be overestimated to an extent and in any way somewhat artifical, Old Icelandic always being present as a model much like Attic Greek to speakers of later forms of Greek and available as a source of constant renewal of archaic forms. In any case, Icelandic pronunciation definitely has changed considerably (thanks to the First Grammarian, we are informed about the phonology of 12th century Icelandic very well), and prestopping belongs to those distinctive changes.

    Another distinctive feature of Icelandic pronunciation is preaspiration, and again this is shared with Faroese, (western and northern, IIRC) Norwegian dialects, and even (northern, IIRC) Swedish dialects. Perhaps the feature was once more widespread in Scandinavia – I'm not sure if it belongs to the same "package" of innovations, and how old it is in Scandinavian. While there does not seem to be direct evidence, from the aforementioned the probability that Norwegian-derived dialects in the British Isles had preaspiration too very high.

    Interestingly, Einar Haugen mentions in his "The Scandinavian Languages" that Scandinavian survived in Caithness as long as the 15th century and on the Outer Hebrides as long as the 16th, curiously now the last stronghold of Scottish Gaelic. (The Hebrideans therefore have a case to make that they are really descendants of Vikings rather than Celts, and in any way it is striking that even though Scottish Gaelic is nowadays perceived as a complicated exotic language that nobody except perhaps a few freaks can be bothered to learn, it was still influential enough in the 16th century that the Hebrideans indeed did bother to learn it and even give up their native Scandinavian – presumably an "easier" language, although actually, Icelandic with its complex inflecting morphology including four nominal cases and various strange umlaut phenomena looks like a nightmare for the average speaker of Modern English).

    In this light, the fact that Scottish Gaelic is distinctive by featuring preaspiration – especially on the islands, but it is widespread on the adjecent mainland as well – looks like a holdover from what is essentially a Scandinavian substrate rather than superstrate or adstrate. I cannot remember reading anything about prestopping in any local dialect of Scottish Gaelic, but its occurrence in Manx surely makes you wonder. To this point, I hadn't heard of prestopping in Cornish, I would be interested in learning more about it.

    As has been pointed out, both preaspiration and prestopping (of nasals) also occur in Sami languages (not only in Northern Sami, by the way), and cannot be explained as an ancient feature there (neither seems to be reconstructed for proto-Sami by Sammallahti). The ultimate origin of these areal features is obscure; perhaps they go back to (one of) the substrate layers that Sami has assimilated while spreading through Fennoscandia, especially in the north and west (curiously, however, modern Finnish, for which a Sami substrate has to be assumed, has the clusters "ht" and "hk" that sound very similar to preaspirated stops). It is possible that at least some Scandinavian dialects have acquired the features from the substrate directly instead of from Sami. In any event, this scenario would fit the distribution outlined so far quite well.

  102. Florian Blaschke said,

    April 22, 2010 @ 4:03 pm

    My use of the word Atlantic was a little generous and potentially misleading; in fact, of course, the area in question covers only the shores of (little more than) the Norwegian Sea. Sorry.

    As for an Anglicisation of Eyjafjallajökull, I prefer Will Steed proposal farther above, Eyfellicle, to Islefellicicle, since icicle never means glacier. Quite curious that speakers of English have such misgivings about the adaption of an Icelandic name when the Icelanders are famous for adapting foreign words and even translating them – even placenames, as has been pointed out. Bad conscience on the part of English speakers regarding foreign and especially exotic and little-used languages, anyone?

  103. Mike Farrell said,

    April 22, 2010 @ 4:09 pm

    How to name a volcano:

    http://theoatmeal.com/comics/volcano_name

  104. Florian Blaschke said,

    April 22, 2010 @ 4:46 pm

    Oh, I forgot what Suzanne Kremmer said about Yorkshire English; Northern English dialects have had a lot of influence from Norwegian in the lexical and presumably also in the morphological field, so why not in the phonetic/phonological field also?

    It reminds me of the adjective Liverpudlian for Liverpool; I've always wondered about that one. On the other hand, the -pool is supposed to derive from Welsh pwll, so perhaps that's the explanation for the "dl".

    Anyway, I wonder – does Icelandic have voiceless lateral fricatives or even affricates or only voiceless lateral approximants?

    There's even a gentleman who pointed out on the talk page for "Icelandic language" on Wikipedia that some Icelandic unaspirated stops sound glottalised to him (at least he reported perceiving [k']) and wondered if they were ejectives, and even produced the abstract of a phonetic paper that seemed to confirm his impression. I would like to know more about this – any phoneticians who can help? I wondered if the sounds in question might not be more similar to the Eastern Armenian voiceless unaspirated stops, which sound lightly glottalised to me but are said to be tense and not true ejectives. (I think it was in Fortson's Introduction to Indo-European linguistics where this was mentioned.) There's also the fortis consonants in Swiss German which are suspected by some to be actually tense. And of course the famous Korean stiff-voiced stops, although there it's not clear whether these are tense, stiff, or faucalised.

  105. Luisetta Mudie said,

    April 23, 2010 @ 2:56 am

    English rendition? Islafell Fire'n'Ice

  106. SPROGMUSEET » Eyjafjallajökull said,

    April 23, 2010 @ 1:42 pm

    […] Om det helt nøjagtigt udtales tl eller hvad, diskuteres i teknisk detalje på Language Log her og her og på John Wells's phonetic blog […]

  107. Slappe y’er og ustemte l’er – i islandsk og dansk – ə said,

    April 24, 2010 @ 9:42 am

    […] oppe at vende, lige bortset fra, indtil nu, denne blog. Jeg vil nøjes med en henvisning til Language Log og Sprogmuseets behandling af emnet – der er rigeligt med links til at surfe videre derfra. […]

  108. Margaret Paulson said,

    April 24, 2010 @ 11:14 am

    Saying the"The Island Mountain Glacier" glaciar is like…Daniel von Brighoff saying The Sahara Dessert= the The Greatest Desert desert.
    Any others?

  109. Pronouncing Foreign Names in English – Fully (sic) said,

    April 25, 2010 @ 9:30 am

    […] a comprehensive post on Language Log, describing the difficulty of Anglicising a difficult and long foreign […]

  110. Claire Bowern said,

    April 25, 2010 @ 5:25 pm

    Regarding prestopping — it's found in several Australian languages (both phonemically and sub-phonemically), and also in various languages in Amazonia, though there mostly with nasals.

  111. bread & roses said,

    April 27, 2010 @ 1:20 am

    This post was a lot of fun. I kept trying to nail the pronunciation, which put the word going through my head, which attached itself to a melody- so here's one anglicization of the name in song (to the tune of "John Kanaka")-

    http://www.youtube.com/watch?v=6J9xp7DPQwg

  112. Gerry said,

    April 27, 2010 @ 7:12 pm

    The Australian TV program Media Watch had a go at volcanic pronunciations the other day: http://www.abc.net.au/mediawatch/transcripts/s2882760.htm

  113. Florian Blaschke said,

    May 1, 2010 @ 10:32 pm

    I've never heard of that before, Claire. Interesting.

  114. Ey… Eyfygico….EYjaFajalla..Eyjafjallajökull? ARGH!! « Suspension of disbelief. said,

    May 7, 2010 @ 10:18 pm

    […] reading: A more in-depth explanation of the Eyjafjallajökull's name. […]

  115. The end of the world has to wait « Jedes Töpfchen ein Deckelchen said,

    May 15, 2010 @ 6:00 am

    […] Spiele zu unserer Hochzeit geben  – wir hatten in der Tat kurz darüber nachgedacht, einen "Wie man den Namen des Vulkans ausspricht"-Contest machen , aber mangels Isländischer Gäste bekamen wir keine Jury zusammen. Es bleibt also […]

  116. Den der vulkanisbreen « Mellom turrfisken og veden said,

    May 16, 2010 @ 12:37 pm

    […] Massevis av folk der ute i verda har blogga om uttalen. […]

  117. sterna said,

    May 18, 2010 @ 3:04 pm

    For a standard English pronunciation of the name, I'd skip the t in the final tl. There are some double Ls in Icelandic that are pronounced ll, and indeed the accusative form of Eyjafjallajökull is Eyjafjallajökul, so that rendering would be perfectly clear to an Icelandic speaker. I think it's not as ear-bleedy as a mangled stop and lateral release would be, too.

  118. Gunnar said,

    May 27, 2010 @ 9:47 pm

    Wow! A very interesting linguistics lesson for a Norwegian semi-hobby linguist!

    Ey (øy) is a perfectly good Norwegian word as well, as are fjall (fjell) and jøkull. But in my dialect, I would not be pre-aspirating any stops, though some Norwegian dialects do ("thick l"). I was surprised by the insistence on the /tl/ in both fyar/tl/a and joeku/tl/, but following the explanations here, that makes sense.

    On my sole visit to Iceland, I found that I could read most signs, but not understand much spoken Icelandic. My reading of signs was benefiting from education on Old Norse – otherwise I probably would not have stood a chance. Of course, Norwegian, Swedish and Danish are all the same language, although the Swedes can't spell and the Danes can't pronounce :-)

  119. Leonardo Boiko’s Diary said,

    June 11, 2010 @ 1:22 pm

    […] watching the poor anglophones trying to pronounce such a simple, logical word as Eyjafjallajökull (1, […]

  120. Krakatoa -Days of Darkness (AD 535-AD 546) | Vulkane Krakatau Indonesien said,

    August 7, 2010 @ 6:35 pm

    […] Language Log » Eyjafjallajökull fail – previous post | next post ». OK, how do YOU pronounce Eyjafjallajökull? Andy Newman and Bao Ong, "Iceland Volcano Spews Consonants and Vowels", NYT 4/16/2010, offer a selection of attempts by people they asked in the Times Square subway … […]

  121. John F said,

    September 2, 2010 @ 5:18 am

    I live in a part of Northern Ireland that uses glottal stops more than most but I've never really noticed until now, that sometimes people from many different English speaking regions might have a barely heard glottal stop in words like full and wool, i.e. fu'ul, wu'ul, which might be related to the letters 'll' being pronounced like 'dl'. Though I haven't studied this, so my mind could be deceiving me.

    I saw a documentary a month or two ago about the volcano, and the English-speaking presenter was very careful to pronounce the name so it ended 'yokootla'. I didn't know about the 'll' thing, so I found it strange, but I did notice the native Icelanders on the programme pronounced it more like the English spelling would indicate.

  122. Amazing Pictures Not From Your Last Party said,

    December 15, 2010 @ 10:02 am

    […] go do that.  And then practice saying "Eyjafjallajokull."  You never know when you might need to say it again. Share and […]

  123. John C. said,

    February 24, 2011 @ 10:56 am

    This is just too funny to me. I took a linguistics class in college and this brought back a lot of memories. I'm not even sure there is a correct pronunciation in English.

  124. Website said,

    November 18, 2011 @ 3:13 pm

    Wonderful goods from you, man. I've understand your stuff previous to and you're just too wonderful. I actually like what you have acquired here, really like what you are saying and the way in which you say it. You make it entertaining and you still take care of to keep it sensible. I can't wait to read much more from you. This is actually a tremendous website.

  125. Óttar Ísberg said,

    December 9, 2011 @ 11:32 am

    The best pronunciation was by the guy described as:

    Another (apparently native) hyperarticulated (and clipped in recording) pronunciation is here:

    It is obviously native and it doesn't seem to be hyper-articulated, just clearly spoken.

    The
    "Update 4/19/2010 — Another native-speaker performance, from ABC News, is here:{Update 4/19/2010 — Another native-speaker performance, from ABC News, is here"
    is also good. But if you listen carefully, you'll notice that the latter pronounces the k softer than the former.

    From David Shaw:

    "jökull – meaning glacier – is the diminutive (!) form of an Old Norse word meaning 'piece of ice', but, etymologically speaking, it is the same as the 'icle' part of our word icicle, which it still sounds a bit like."

    The affix -ull is not diminutive. See Grænlandsjökull for an example.

    The ending -ull occurs in a number of places, e.g., þögull (silent).

    His suggested transliteration is good, however.

    As for preaspiration, Eyjafjallajökull is not a good example, since it contains none.

  126. Special Guest Blogger–Ruby Duvall! « Passionate Reads said,

    July 10, 2012 @ 10:44 am

    […] weird in historical context. Do you remember the April 2010 eruptions of Eyjafjallajökull, that impossible-to-pronounce volcano in Iceland that shut down European airports with its ash […]

  127. Ms. Lefevre said,

    December 6, 2012 @ 4:51 pm

    I just wanted to thank you for the audio files. We are reading about volcanoes in my sixth grade class, and without this resource, I wouldn't have been able to answer the question, "How do you say that word?"

RSS feed for comments on this post