Pausing Over Pronunciation.

A nice piece by Anne Curzan on not being sure how to say a word; she begins by describing reading aloud to students from a quote and seeing the word islet coming up:

Torn about the status of the “s,” I decided to try to turn this moment of pronunciation panic into a teachable moment. I stopped when I got to the word, and I said to the class, “How do you all pronounce that word?”

There was a noticeable pause. A few ventured, “Eye-let?” Then a couple of students said they thought they had heard “iss-let.” One student from Florida confirmed that this pronunciation occurs in Florida. Others admitted that they weren’t sure they had ever said the word out loud.

We checked a couple of standard dictionaries and found just one pronunciation: “eye-let.” So now we know what is considered standard. (That said, I’m not convinced that the pronunciation with an /s/ won’t make enough inroads in American English to become a standard variant. I’ve suggested to the editors at American Heritage that we track it on the usage ballot.)

Learning the standard pronunciation, however, seems to me not the most important benefit of the pause. We were also able to have a conversation not only about some of the vagaries of English spelling but also about the way our status as an “educated speaker” can feel up for evaluation when we hit some of these tricky words we’re not sure how to pronounce. Can we actually say, “I’m not sure how to pronounce that” without getting laughed at?

I was thinking about this story a couple of days ago because I mentioned to a colleague that I had just recorded a radio segment about the pronunciation of the word niche. He exclaimed, “That word always gets me! I am never sure how to pronounce it.” We commiserated over our shared angst when confronted with this word. Does “neesh” sound too French and too pretentious? Does “nitch” make us sound unsophisticated?

If you’re thinking “nitch” must be the new, “bastardized” pronunciation, you are wrong. Many standard dictionaries include both pronunciations. And according to the American Heritage Dictionary of the English Language (fifth edition), the current pronunciation “neesh” is a 20th-century innovation, as the word was “Englishified” (my term) to “nitch” soon after it was borrowed from French in the 17th century.

My colleague then added, “And then there’s homage! I don’t know what to do with that one either…” I agreed: there’s the issue of where the stress goes as well as whether to say the initial /h/. I added the word forte to the how-should-I-pronounce-that mix. […]

Some of those words probably figure on almost everybody’s mental list of worrisome pronunciations; I think I’ve settled on “nitch” and “for-tay,” but the rejected variants buzz around my brain reproachfully. It is well, as always, to try not to judge either oneself or others harshly in these matters.

Comments

  1. clique.

  2. Why not to split the baby and pronounce it nish.

  3. To me, homage is /ˈɑmədʒ/ if you’re American and /ˈhɒmədʒ/ if you’re British, and /oʊˈmɒːʒ/ is hommage (which is italicized, as a foreign word).

  4. An old friend, who I have not seen in person in many years, has a five-year-old daughter named “Isla.” I see lots of pictures of her on Facebook, but just the other day, I clicked on a movie of her and her dad, which included sound. I was quite surprised to find that her name was pronounced EYE-luh. I (who always say “islet” as EYE-let) would never have dreamed that the S in her name was silent.

  5. When saying ‘concerto’ do I use English ‘t’ or Italian? I speak Italian so using the English one feels like sound of nails on blackboard. Same with ‘latte’.
    I use forvo.com and howjsayit.com for correct pronunciations.

  6. I actually do say “nish”, and as far as I know I always have. On the other hand, Gale (who is a “nitch”-sayer), has never noticed that I say it differently from her — or possibly has never heard/noticed me saying it at all. To be fair, I had to ask her to confirm her pronunciation for me. This is a bit of a ritual in my family: I hold up a Post-It with one or more written words on it and ask people to pronounce them.

  7. One word I seem to be out on is ‘sachet’, which I like to pronounce /’sæʧət/, although most of the people around me seem to say /sæ’ʃei/.

    I think I’ve mentioned before that I once pronounced ‘cognisant’ as /kɒg’naizənt/, only to be corrected to /’kɒgnəzənt/ by a Japanese speaker (blush).

    Also ‘facile’ as /fə’si:l/ instead of /’fæsail/ (groan).

  8. Brett: “Islay”, on the other hand, is pronounced “Eye-luh”, not “Is-lay” or even “Eye-lay”. No one’s quite sure why: normally when you see an island with “-a” or “-ay” at the end it’s derived from a Scandinavian name (“oy” = “island”) but apparently not in this case. The silent S seems to be a recent addition (influenced by “island” presumably) – in older maps it’s Ila or Yla.

  9. “Neesh” rather than “Nitch” seems fairly recent in BrE; I’ve never heard a BrE speaker say anything but “neesh”, but pre-war it was common enough for the limerick about the young lady from Chichester* to work.

    *”…who made all the saints in their niches stir.”

  10. I have never, ever heard anyone say ‘fort’ for ‘forte’. The OED still sermonises: “As in many other adoptions of French adjectives used as nouns, the feminine form has been ignorantly substituted for the masculine; compare locale, morale (of an army), etc” Though we don’t pronounce the final e’s of the latter words of course. The opposite has happened with ‘dilettante’ and ‘rationale’, pronounced as if from French, without the last e.

    On this topic, something I’ve long wondered about — where does the phantom American medial e in ‘erudite’ (“eree-udite”) come from?

  11. Conrad, isn’t that that just pronouncing the “u” as “yoo” rather than “oo”? I’d’ve thought that would be more likely for non-Americans than for Americans.

  12. I recently fell in love with the word chichi. What’s the vote on that one? I say SHEE-shee.

  13. Keith: it might be that, but that doesn’t explain why. I can’t think of any parallels. Eriudite is in my experience standard US, but never heard it in UK.

  14. [ˈ]l[ɛ̃|ɑ|̃æn|ɔn][ʒ|dʒ][ə|Ø][ˈ][[ʁ|ɹ][i|e]

    Instead of saying it, I usually just mime.

  15. There is a general recession of the /j/ in /ju/ after a consonant in English, but it takes different forms in different Englishes and is affected by stress. Stressed /rju/ was long ago reduced to /ru/ in all varieties: nobody says “ryude”, for example. (In Welsh English, the /ju/ that is historically /ɪw/, as in new, has remained so even after /r/, however.) Similarly, /tju dju nju lju/ have become simple /tu du nu lu/ in North America, whereas the first two are moving toward /tʃu dʒu/ elsewhere; /sju zju/ are now /su zu/ throughout North America and to a great extent elsewhere as well.

    Almost all of the unstressed forms have followed similar patterns, except that /tj/ /dj/ /sj/ /zj/ have palatalized in almost all varieties except the most conservative. However, North American English has less vowel reduction than other Englishes, probably because it was heavily influenced by L2 and L2-influenced varieties where words were more likely to be pronounced as spelled. However, /ɛrjudaɪt/ remains hard to say, and so it unpacks into /ɛriudaɪt/.

  16. I say /’erədayt/ and avoid the whole mess. (I think; I haven’t said it in quite a while.)

  17. mollymooly: hence, ‘unmentionables’?

  18. The glided-[u] is on its way out in Southern American English. There have been various studies that demonstrate this.

    Most Southerners have retained the on-glide in words like [kjut] ‘cute’ or [bjukalik] ‘bucolic,’ (at least those who say bucolic). However, I am a linguistic dinosaur who still says [nju] for [nu] ‘new’ and [njuz] for [nuz] ‘news.’ But, I recognize that most normal, well-adjusted Americans have embraced modernity and moved on to the unadorned [nu and nuz].

    It seems to me, in unscientific observations while watching BBC news and listening to my English sister-in-law speak that it is retained in British English to a greater degree than American English.

  19. I’ve never been embarrassed about learning the pronunciation of a word I’ve only ever read. Curse of a reader, is my explanation. Know what it means, how to spell it, but. These are also words I don’t usually say out loud,or if I do, it’s with exaggerated enunciation, or clearly the wrong way. In surgery, it’s the gigli saw. Everyone says it “gig lee.” Then we all correct each other – I’m told it’s jē′ylyē. No one can say that with a straight face, though. Sounds affected.

  20. Ajay: The pronunciation /’aɪlə/ for Islay is a half-hearted imitation of its Scottish Gaelic name Ìle /’iːlə/. It may come from a Celtic anatomical term (related to Latin īle ~ īlium ‘flank, groin’), perhaps referring to the island’s shape or location. This is of course highly speculative (“Methinks it is like a groin”). Even more so are other published etymologies, like the one by Richard Coates (2009), which derives the name from a Semitic source as ‘holy/pure island’ (root *ħll) or “better even” as ‘god’s island’ (if from *’il(-āh), as in El, Allah, Elohim, etc.). I have found a copy of the paper here. It’s all to be taken with a rather big grain of salt.

  21. he pronunciation /’aɪlə/ for Islay is a half-hearted imitation of its Scottish Gaelic name Ìle /’iːlə/.

    Isn’t it rather a perfectly normal development, with long /i/ becoming /aɪ/ in English like every other long /i/?

  22. Yes, if the name was Anglicised before the Great Vowel Shift. But then it isn’t clear why the final schwa wasn’t dropped. The modern English pronunciation is some kind of Anglo-Gaelic compromise, with orthography also playing a role (the -s- is of course folk-etymological, after isle, as, for that matter, is the -s- of island).

  23. D.O. (2nd comment): “Why not to split the baby and pronounce it nish.”

    Isn’t that exactly what most English-speakers do with a lot of foreign words, particularly names? We don’t pronounce ‘Debussy’ to rhyme with ‘the hussy’, as if it were English, but we also don’t try to pronounce it as a Frenchman would, with all the subtleties of vowel and consonant sounds, even if we know how – unless we’re being pretentious. We generally just say Deb-yuh-see, and leave it at that. Similarly with the V sound in Wagner, saying ‘Proust’ to rhyme with ‘goosed’ rather than ‘moused’, and so on. There’s a lot to be said for ‘splitting the baby’ when using foreign words in English.

  24. Yes, if the name was Anglicised before the Great Vowel Shift. But then it isn’t clear why the final schwa wasn’t dropped.

    Good point.

    There’s a lot to be said for ‘splitting the baby’ when using foreign words in English.

    I entirely agree!

  25. When I pronounce “erudite” (which is not often) I say “air-you-dight”.

    In my youth I heard a student ask the late great English topologist Frank Adams whether the word “homotopy” should be pronounced “HOMotopy” or “hoMOtopy”. He replied that the correct answer, based on the Greek origins of the word, was intermediate between those two. So this was a careful man. On the other hand, in speaking of mathematical things named after Hurewicz he pronounced the “u” like “you”.

  26. This isn’t worth a post on its own, but speaking of pronunciation, I just read a Browning poem called “The Statue and the Bust” and was taken aback by the end rhyme:

    Though the end in sight was a vice, I say.
    You of the virtue (we issue join)
    How strive you? De te, fabula!

    It would seem Browning expected the Latin tag to be pronounced “Dee tee, fab-yoo-lay!” Which is even more extreme than I thought traditional pronunciation got.

  27. ‘Proust’ to rhyme with ‘goosed’ rather than ‘moused’

    Except in this double dactyl:

    Said Agatha Christie
    To E. Phillips Oppenheim:
    “Who is this Hemingway?
    “Who is this Proust?

    “Who is this Vladimir
    “Whatchamacallum, this
    “Neo-post-realist
    Rabble?” she groused.

  28. Brett – Britons would have known it was “eye-la” because of Isla St Clair http://en.wikipedia.org/wiki/Isla_St_Clair (last name pronounced “sinclair”, of course)

  29. On the other hand, in speaking of mathematical things named after Hurewicz he pronounced the “u” like “you”.

    The ultimate tongue-twister in the field of historical linguistics is the name of an Indo-European morphological phenomenon known as the Kuryłowicz vṛddhi.

  30. Am I the only person with a different vowel sound in isle (and Islay, and I think islet) from the one in island (and eye, and eyelet)?

  31. Interesting. Can you describe the difference?

  32. I’ve heard one teacher say that she deliberately has her students read out loud material/literature that is difficult, and doesn’t help them with pronunciations: she hopes that this will instill respect for a level of literacy above their own.

  33. Sounds like deliberate humiliation of students to me.

  34. J. W. Brewer says:

    Or the students might learn how to get past difficulties like not knowing how to pronounce an unfamiliar word on the page by just bluffing in a confident tone of voice, plausibly assuming their audience is by and large no less ignorant than themselves and they’ll probably get away with it unless they sound uncertain. It only takes a modicum of cynicism to think that that’s a valuable life skill for teachers to impart (whether intentionally or otherwise).

  35. I can imagine it working that way, if the teacher made sufficient effort, but I can more easily imagine it leading to snide comments on the part of students who knew the word in question and raucous laughter on the part of the class, all this to be filed under “what does not kill us makes us stronger.” I despise that sort of cruel-to-be-kind approach to education (or anything else).

  36. J. W. Brewer says:

    That Browning poem is chock-full of “rhymes” that do not rhyme for me. I suspect that *some* of them reflect genuine differences between my pronunciation and his (and/or that of an assumed standard/posh BrEng speaker of his time who was the hypothetical ideal reader of the genre in which he wrote) but that many others would have been taken even at the time as within the then-acceptable bounds of “poetic license,” but do not actually evidence that the words were pronounced rhymed in ordinary discourse. It’s terza rima, so the rhymed words are supposed to come in sets of three. If we set aside those that might be “fixed” if a reduced final vowel were pronounced unreduced (e.g. call/funeral/befall), we have lots of problematic triples like the following:

    breath/underneath/sheath
    multitude/stood/subdued
    prove/love/alcove (3 different vowels for me!)
    mouth/South/youth
    sigh/infallibly/by [cf Blake’s famous eye/symmetry, I guess]
    love/above/remove
    was/glass/pass
    said/aid/fade
    warm/charm/arm
    suns/bronze/once
    wonder/ponder/yonder
    his/bliss/this
    uttermost/lost/ghost

    A man that will do this is capable of anything, and nothing should be inferred (beyond what we might already know about how Englishmen of his generation had typically been taught at school to pronounce Latin) about how he might naturally pronounce “fabula” in non-poetic contexts.

  37. Good point! But I still wonder if that was how he meant it to be pronounced.

  38. J. W. Brewer says:

    Now we just wait for John Cowan to explain how for each of the triples I gave there has been an extant historical variety of English in which all three entries were pronounced rhymed . . . (my only hope is that there has never been a variety in which *all* of them worked as rhymes that Browning could have hypothetically had in mind when he wrote).

    For e.g. rhyming “love” with “prove” which has high ancient authority to commend it, I’ve never quite been sure what poets/readers did (or expected) once the words no longer rhymed in actual speech. Do you pronounce them the normal way (where they don’t rhyme but will nonetheless be deemed for whatever reason acceptably “poetic”), or do you change the normal pronunciation of one to force a rhyme, a la “come live with me and be my Louvre / And we will [something something] pleasures prove.”

    There is a more general problem here in that we can’t uncritically use rhymed poetry as evidence of past pronunciation without also knowing what did and didn’t count as a “rhyme” according to the literary conventions of the time.

  39. Re “De te, fabula!”; traditional pronunciation of classical languages in school contexts (as opposed to as incorporated into English usage) did often use the English long and short vowels consistently to represent the Latin and Greek ones; so a final -a pronounced as ‘ay’ is not impossible. However, it would be wrong here, as surely the ‘fabula’ is a vocative (though the ‘te’ preceding it is an ablative), so it would have to be ‘fay-byoo-ler’ (in non-rhotic British terms).
    I remember once hearing a distinguished (and by then retired) academic referring to a Greek concept of ’em-PIRE-ee-ay’ ,using what must have been the pronunciation of his school days. It took me a moment to realise this was the traditional pronunciation of ἐμπειρία ( modern εμπειρία ), complete with English vowels and spurious Latinate stress.

  40. Re “De te, fabula!”; traditional pronunciation of classical languages in school contexts (as opposed to as incorporated into English usage) did often use the English long and short vowels consistently to represent the Latin and Greek ones; so a final -a pronounced as ‘ay’ is not impossible. However, it would be wrong here, as surely the ‘fabula’ is a vocative (though the ‘te’ preceding it is an ablative), so it would have to be ‘fay-byoo-ler’ (in non-rhotic British terms).

    The first part is correct, which is why I thought it was extreme but not wrong. The second part is not; you are misled by the comma (surely not Browning’s but added by some editor) — the tag is de te fabula narratur ‘the story applies to you’ (literally ‘is told about you’), so fabula is nominative, not vocative.

  41. Sorry, didn’t think to check for the original quotation; but of course a Latin first declension nominative singular would also have a short vowel, so it would still be a ‘false quantity’ to give it an -ay pronunciation.

  42. Right, but I can see it being treated as long by position or something. It’s weird, in any case.

  43. In the traditional English pronunciation of Latin, Latin vowel length is used to determine stress, but it is not used to determine vowel quality (miscalled “length”), which is done on the basis of English spelling conventions. Since all English final vowels are tense (except historic -y), you would indeed expect /diti/ for de te and /fæbjəleɪ/ for both fabula and fabulā.

  44. Ah, very enlightening! And googling tells me there is a long Wikipedia article, but it seems hopelessly confused and confusing, jumbling together the pronunciation of English words borrowed from Latin with the pronunciation of Latin words in English. It does mention an “older tradition (in use in the 17th-18th centuries)” that “had made all final a’s ‘long’, regardless of their Latin length,” and I guess that’s what’s going on here.

  45. I think they pretty much are the same pronunciation, since when Latin words were borrowed into English, they were borrowed with the pronunciations with which they were already spoken as Latin words. The alternative to /eɪ/ for final -a would of course be /ə/, as in alumna, so I wouldn’t be surprised by /fæbjələ/ either.

  46. “. . . since when Latin words were borrowed into English, they were borrowed with the pronunciations with which they were already spoken as Latin words.”

    Alumni? Most people I know say /alʌmnaɪ/.

  47. Which is the traditional pronunciation of the Latin word in English. But I’m glad you brought it up, because we’ve come to a stupid moment in history when as far as I can tell most English speakers (Americans, anyway) say /alʌmnaɪ/ for both alumni (traditional pronunciation) and alumnae (using the reformed, classicizing pronunciation of the final vowel, rather than the /i:/ [“ee”] that traditional pronunciation demands), creating havoc with what appears utter perversity. I mean, either traditional or reformed pronunciation gives you a nice contrast, /alʌmnaɪ/ vs. /alʌmni:/; only by mixing and matching do you get this ambiguity. Why??

  48. David Marjanović says:

    Am I the only person with a different vowel sound in isle (and Islay, and I think islet) from the one in island (and eye, and eyelet)?

    Is that that unexplained Scottish split? (I think it’s called the TIE-PIE split…)

    I’ve heard one teacher say that she deliberately has her students read out loud material/literature that is difficult, and doesn’t help them with pronunciations: she hopes that this will instill respect for a level of literacy above their own.

    She’s in deep denial then.

    (Realitätsverweigerung in German, “reality refusal”.)

  49. David Marjanović says:

    Why??

    Simple lack of knowledge of which written form means what?

    Add number confusion to the gender confusion: I’ve seen women refer to themselves as “an alumni”.

  50. There’d be ambiguity about the gender of the alums anyway, unless the listener knew whether the speaker was using traditional or reformed pronunciation.

  51. “Simple lack of knowledge of which written form means what? Add number confusion to the gender confusion: I’ve seen women refer to themselves as “an alumni”.

    I would also add ‘alumni’ as a singular, masculine or feminine. Frankly, I have to think a second before I use ‘alumni’ as a singular and depending on audience, I might go ahead.

    However, it is confusion only if speaking Latin. At some point a borrowed word becomes nativized and takes on native phonology and morphology. ‘Alumni’ may now be an English word. And, of course, those classical Romans probably didn’t use it just like the Greeks or PIEs from whom they borrowed it. And, the PIEs would have been rolling over in their graves at the misuse of their language.

  52. Well, I am Scottish – but I’m pretty sure tie and pie are the same!
    Best minimal pairs we’ve found so far are Fife/five and life/live (as in wire).

    Linguist friend suggests ‘life with [ʌɪ] and live with [ae]’ but I don’t remember enough IPA to do anything other than believe him. The five/eye vowel feels like its deep down in my mouth while the Fife/isle one feels almost between my lips, if it helps…

  53. ‘Proust’ to rhyme with ‘goosed’ rather than ‘moused’

    Except of course in Monty Python’s “All England Summarise Proust Contest” in which, the host explains, contestants have twenty seconds to summarise the plot of “In Search of Times Past”, once in swimwear and once in evening dress.

  54. However, it is confusion only if speaking Latin.

    I’m not sure what you mean. I said nothing about speaking Latin; I was talking entirely about speaking English. The English pronunciations of the English words alumni and alumnae should be /alʌmnaɪ/ and /alʌmni:/; it doesn’t matter which is which so long as there’s a distinction. The fact that there is no distinction is not the result of some inevitable sound change, it’s the result of perverse mix-and-match. There is no inherent reason why it should be impossible to tell, hearing “There were a couple of /alʌmnaɪ/ there,” whether they were male or female. It’s not the end of the world, obviously; English will get along perfectly well without this useful but not necessary distinction. I still think it’s weird and stupid.

  55. @LH: My only point (not well stated) is that many AmE speakers (who don’t know Latin) are not confused since the word has been pretty well nativized in English. I think ‘alumnus’ is well understood as one singular form used in more formal (AmE) contexts (‘alumni’ in less formal). ‘Alumna,’ I think, is pretty much limited to academic or very formal contexts.

    (Or maybe I just run in rather plebeian circles)

  56. My only point (not well stated) is that many AmE speakers (who don’t know Latin) are not confused since the word has been pretty well nativized in English.

    I’m still not following you, since there are two words, not one. (Again, Latin is irrelevant here; we’re talking purely about English.) Unless you’re claiming that alumna is literally not an English word? It’s certainly used in all published material referring to female graduates, and I hear women say “I’m an alumna” with no sense that they are being pretentious. But then I run in rather academic circles. It may be the case that in the Common Man’s parlance there is only the word alumnus with its plural alumni (/alʌmnaɪ/), but the fact remains that when I hear people who use the female form alumna pronounce the plural, they say /alʌmnaɪ/, and I’m quite sure they know it’s written alumnae. There’s some sort of leveling going on, which is fine, that’s how language works, but this is an instance where language is going out of its way to efface a useful distinction.

  57. But the distinction is only useful if you actually want to separate the alumnae and alumni, far more often they’re bunched together so having one pronunciation saves everyone the bother of having to say the word twice. This isn’t a distinction that english normally makes, after all.

  58. True, and I imagine my dismay is 99% oldfartism.

  59. J. W. Brewer says:

    Are there enough other Latinate words ending in -ae lexicalized into English that a non-specialist will have any general clue of what pronunciation the spelling ought to cue? I do tend to agree that if the gender distinction between alumni and alumnae is no longer useful in practice, the way to deal with that is just to have one word, not have two differently-spelled words with theoretically different meanings that are pronounced unnecessarily homophonously.

    You can find constructions out there on the internet like “Alumni of Mount Holyoke,” where “alumnae” would seem (to me) especially apropos. I don’t know how actual MH alumnae feel about that (and no doubt they do not all have identical views), but presumably (at least for those alumnae under 55 or 60 who chose to go there after single-sex schools had already become the exception rather than the rule) having gone to Mount Holyoke in the first place might imply some view that distinctions of gender remained salient and should not be deemed lexically obsolete.

  60. I thought I would google “he is an alumni” vs. “he is an alumnus” to see the relative hits. However, Google came up with a pertinent article that caught me eye:

    http://www.macleans.ca/education/uniandcollege/why-youll-never-be-an-alumni/

    Maybe the decline in ‘alumnus/alumna/alumni’ and increase in ‘graduate’ supports your point about confusion.

    Also, FWIW, I think the various alum-words are, at least in the U.S., limited to college education. We don’t use the terms for graduates of high school, trade school, military training, etc.

  61. While in the FWIW category, ‘alumnis’ got 318,000 ghits.

  62. Unsurprising but interesting!

  63. I think “minutiae” has entered nonspecialist English, but it seems to be usually pronounced as if spelled “minutia”, and probably spelled that way too.

  64. J. W. Brewer says:

    https://www.youtube.com/watch?v=nfslFnlPCY4 is a video compilation of interview snippets with various (I assume) current students (i.e. future alumnae) at Wellesley College (it’s an all-female student body, for non-U.S. readers who may not know of its fame . . .), who when asked how to pronounce “alumnae” offer a diverse range of answers.

  65. On the American pronunciation of foreign scholars’ names, the other day I was teaching and was at a loss to explain why I pronounce Walter Benjamin with a y sound for the j, but not a v sound for the w. That’s how my college profs always pronounced it too, but it doesn’t make a whole lot of sense.

  66. It makes sense to me; given names are much more basic parts of our vocabulary than family names. It’s very hard for an English-speaker to look at the given name “Walter” and think of it with a /v/ sound, whereas it’s much easier with, say, Bruno Walter. At least for me.

  67. I always think of my own German middle name, Woldemar, with a /v/; granted, it’s not a typical name for any sort of anglophone.

  68. Yeah, I meant typical English names whose pronunciation is burned into your brain. Although now that I think of it, it holds for any name that got burned into your brain at an impressionable age; I can’t think of Pedro Ramos (who pitched for my Senators and Twins when I were a wee lad) as anything but “Pee-dro Ray-mos,” because that’s how English-speakers pronounced it in those barbaric times.

  69. David Marjanović says:

    my own German middle name, Woldemar

    The o, though, is only explicable as American.

  70. only explicable as American

    Not at all. My grandfather, Woldemar Schultz, was born in Germany in the general vicinity of Heringen on the Hesse-Thuringia border, and always spelled his name so. Other Woldemars. He pronounced the o long, and although I assimilate it a bit to English phonology, I still make it [ˈvoʊldəmɑɻ].

    But I will confess to pronouncing “Walter Benjamin” in English fashion, and though my grandfather’s surname comes out German-style as [ʃʷʊlts], I make the name of Charles Schultz plain American [ʃɫ̩ts].

  71. But I will confess to pronouncing “Walter Benjamin” in English fashion

    Now that I think of it, I probably would have done so too if I hadn’t been living in New Haven (and eventually working in the kind of bookstore where his name would have been bandied about) during the period when he was becoming A Thing.

  72. David Marjanović says:

    Interesting. de:wikipedia confirms that the o exists, but doesn’t try to explain it either.

  73. Except in this double dactyl:

    Back in the days when Mary Ann Madden ran the New York Magazine competition (which I now discover has made a comeback) and the challenge was to create double dactyls, she disqualified anyone who even tried to slip in:

    Higgledy Piggledy
    Vladimir Nabokov….

  74. I note that “Voldemar” is an Estonian form, and I do have at least one Estonian relative who married into my ethnic-German-from-Russia family and could have influenced the form of my grandfather’s name. As for Benjamin, I really know nothing about his work, only about his death, and then only because of his connection with Hannah Arendt.

    Prince Woldemar of Lippe (1824-1895) is one of the famous Woldemars:

    In 1892 along with the other German sovereigns [he] attended a gathering in Berlin with the German Emperor William II. After the Emperor described the other sovereigns as his vassals. Prince Woldemar took exception and interrupted the speech to say, “No, Sire, not your vassals. Your allies, if you like”. This was seen as the coup de grāce to the Emperor’s ambition to become “Emperor of Germany” instead of just “German Emperor”.

    The True Woldemar

  75. For some reason I hung my mother’s birthplace on my grandfather here. Bogus. I don’t know where in the Russian Empire he was born exactly, but probably in or near modern Ukraine.

    I have just finished reading (but not watching) the Wallander stories, and I have decided to agree with Kenneth Branagh by saying /ˈwɔləndər/ and not [valˈlanːdər]. Doing it the other way makes for too much trouble, too many confusions. My wife is both reading and watching, and I have given her a crash course in pronouncing Toneless Swedish to avoid too much manglement and misunderestimating. The books have different translators, so there are lots of inconsistencies like Skåne vs. Scania; we call it [ˈskoʊnə] around here.

  76. I was just telling my brother (who has rewatched the Apu trilogy with undiminished enthusiasm) that Satyajit Ray’s given name is pronounced (something like) SHOT-to-jit in Bengali but that it would be ludicrous to try to use that in speaking English.

  77. Eli Nelson says:

    What an interesting thread! I’m glad it’s been brought up again. In fact, I was recently wondering about the pronunciation of erudite. For a long time, I supposed that I should pronounce the word “erudite” with /rj/, because I mentally grouped the historical simplification of /rj/ to /r/ with my American elision of /j/ after other coronals like /l/ and /n/, and /j/ after these sounds is retained when the cluster follows a stressed vowel (for example, value and menu are still pronounced with /j/ by American English speakers). Then I came across this:

    “The traditional pronunciation of erudite, with -uu- or -uh- in the second syllable, is a subtle and refined reminder of the word’s origin — e- (without) -rudite (rudeness). The popular pronunciation, ER-yuh-DYT, with the y inserted, ignores the rude dwelling within erudite and illogically transforms the short Latin vowel into a long English one.”
    –Charles Harrington Elster, The Big Book of Beastly Mispronunciations

    Now, Elster just seems to be talking out of his ass when he tries to explain the reasons for the r-less pronunciation. There is of course no principle that Latin roots in English words don’t change pronunciation when the stress moves to a prefix—consider eluent /ˈɛljuːənt/, from Latin luere “to wash,” which also gives us *ablution* /əˈbluːʃən/. And Latin vowel quantity, as mentioned earlier in this thread, is almost always completely irrelevant to English pronunciation.

    However, reading this prompted me to look up the pronunciations that are given for this word in various dictionaries, and Elster does seem to be right about one thing: for erudite, the pronunciation without a yod seems to be more standard, and possibly more traditional (I don’t know enough about the history to be sure about that). This was certainly a surprise to me!

    I think the actual reason why erudite behaves differently from eluent might be the odd status of intervocalic “r” in English syllabification. In non-rhotic accents, there’s the well-known problem of how to syllabify words like “ferry”: If we do /fɛ.ri/, we violate the constraint against syllable-final /ɛ/; if we do /fɛr.i/, we violate the constraint against syllable-final /r/. In rhotic accents, the tense-lax constrast is often neutralized before /r/ by merging historically lax vowels before /r/ with tense r-colored counterparts: this creates a difference in vowel quantity between /ɛːr/ and /ɛl/ which might account for their differing effects on a following /j/.

    My hypothesis that the lack of /j/ in erudite is due to some general phonological process like this seems to be supported by the following data: I haven’t been able to find any words spelled with “ru” where a pronunciation with /rj/ is mandatory (even when the “u” is in an unstressed syllable). It’s always possible, or required, to use /r/ instead of /rj/ in words like corrugate, cherubim, virulent, erubescent, neurula, purulent, querulous, spherule, sporule, ferrule.

    Actually, even going beyond words spelled with “ru,” I can’t think of any words in Standard English where a pronunciation with /rj/ is mandatory (setting aside compound words like lumberyard, dooryard where British English speakers would not have consonantal /r/). Words spelled with “riV” or “ryV” generally seem to be pronounced with /ri/, even in contexts where words with “liV” or “niV” would generally be pronounced with /j/: compare carrion on the one hand to stallion and companion on the other, or Sirius, delirious on the one hand to bilious on the other. Even the loanword Aryan,whose spelling with “y” seems like it would strongly suggest /j/ to English speakers, may be pronounced with three syllables.

  78. I’ve always thought bilious had three syllables and rhymed with supercilious.

  79. I grew up using a yod-less pronunciation of “erudite,” but at some point in high school or college, I switched to one with a yod, because that seemed to be how everyone else said it.

  80. I prefer erudite without yod. Although I’m rhotic, I do align with the non-rhotics regarding the tense-lax distinctions, so the choices, as I see them, would be a) to make the first syllable “air”, a sound which is never conveyed by Latinate e for me,* b) to say the second syllable with a true /rj/ onset, which I’d consider quite alien, or c) to break it into four syllables – “erry a dite” – which is actually how I first acquired the word as a child.

    *On that topic, there’s also err, which has always been [ɚː] for me, though most Americans seem to prefer [ɛɚ] – aligning (for them) with error, errant etc. By the same token, deterrent and deterrence shift to [ɚː] to align with deter. I tentatively prefer [ɛ.ɹ] in those two – though with some ambivalence, because even to me it sounds a little affected.

  81. marie-lucie says:

    Scottish English: five/fife with different vowels:

    This is very frequent in Canada, and the vowel of the second word (higher and more fronted than that of the first word) is attributed by American scholars to a relatively recent phenomenon they call “Canadian raising”, rather than to a survival of the Scottish pronunciation of many early immigrants.

    pronunciation quandaries

    I have been familiar with spoken academic English for decades, but it looks like I still have a lot to learn.

    erudite : It never occurred to me that there might be a yod in this word. Perhaps I have heard it but not particularly noticed it.

    bilious and supercilious : don’t these words rhyme for most speakers? with “-il-ious”?

    niche : I am OK with hearing “nish” in English. Here in Halifax there is a bar/restaurant which used to be called Nîche, which looked very pretentious. A few months ago the name was changed to simply Niche.

    In French la niche can mean either ‘an indentation in a wall (often at a street corner) meant to receive a statue of a saint’, or ‘a doghouse’. Neither meaning is particularly suitable for a bar or restaurant.

  82. @m-l: Well, Canadian raising does differ from the Scots vowel length rule – it’s generally allophonic (though for some, particularly in the Upper Midwest, it has adopted some phonemic qualities), whereas the Scots version is morphophonemic (contrasting tide and tied, for example, which are not distinguished in Canada). It also extends far beyond areas of significant Scottish settlement, occurring in the aforementioned Upper Midwest, the Pacific Northwest, New England (I have it), the Delaware Valley and even coastal Virginia. It is similar, though, and I think it’s likely that the Scots played a role in its spread and perhaps its genesis in Canada. Looking at some other features of CanEng – the merger of /ɒ/ and /ɔː/, often to a rounded vowel, the retraction of /æ/ and its partial annexation of the lexical domain of /ɑː/, the near-monophthongal qualities of /eɪ/ and /oʊ/ – it’s hard not to get an impression of general Scottish influence.

  83. Merriam-Webster has only the two-syllable pronunciation for bilious, which surprised me, but for supercilious the first pronunciation has five syllables (as I’d pronounce it) and the second has four (with the last i being /j/). So it seems that MW thinks most (or at least many) Americans don’t rhyme the two words. I do, but only because I pronounce bilious in a way MW doesn’t recognize, with three syllables.

    For punctilious, MW doesn’t have a pronunciation with /j/ at all.

  84. Eli Nelson says:

    @Keith Ivey:

    I’ve always thought bilious had three syllables and rhymed with supercilious.

    Well, it’s true that is another pronunciation in use. I’m not really sure why there are some words where “i” can be syllabic even when it comes after a short vowel + a single consonant and before a fully unstressed syllable, while there are other words where it is almost always pronounced as a glide in this same phonological enviroment.

    But overall, it seems like words spelled with “li” are pronounced with /lj/ a lot more often than words spelled with “ri” are pronounced with /rj/.

    @Lazar:

    On that topic, there’s also err, which has always been [ɚː] for me, though most Americans seem to prefer [ɛɚ] – aligning (for them) with error, errant etc. By the same token, deterrent and deterrence shift to [ɚː] to align with deter. I tentatively prefer [ɛ.ɹ] in those two – though with some ambivalence, because even to me it sounds a little affected.

    I have also thought about how to pronounce err and deterrent/-ence, and come to basically the same conclusions. I wonder if the pronunciation of deterrence might have been influenced by the word occurrence, since the verbs occur and deter rhyme.

    For other words also, it’s not really clear to me how the suffix -ent normally affects the pronunciation of preceding syllables. In most cases, I think it is preceded by a “long vowel” when it follows a single vowel letter and a single consonant (such as in potent) but there are certainly a number of exceptions (possibly due to French influence) such as clement, present, and (for some people) patent. I’m also not sure whether people usually use long or short vowels in words such as parent, apparent, transparent, or if there is variation (I don’t make the distinction here, so I don’t have any intuitions of my own). And many people use a short vowel in inherent, despite the long vowel of inhere, and the usually long vowels of adherent and “coherent.

    My own original pronunciation was /ɪnˈhɛrənt/, although now I’m bothered by the inconsistency and considering switching. Then again, it’s not exactly possible to avoid inconsistency in English: it’s baked into the spelling (which is much less changeable than pronunciation) for many sets of words, such as conceit, deceit and receipt. (To me, the inconsistency of the spelling of the last word compared to those of the previous two words is much more irritating than the fact that none of them are spelled phonemically).

  85. The -parent words all take /æ/ for me; in RP they all take /æ/ except for parent itself, which takes /ɛǝ/.

    But yeah, a few years ago I consciously changed from /ɛ/ to /ɪɚ/ in inherent to align it with the other -herent words.

  86. I prefer erudite without yod.

    Me too.

    err, which has always been [ɚː] for me

    Me too.

    By the same token, deterrent and deterrence shift to [ɚː] to align with deter. I tentatively prefer [ɛ.ɹ] in those two – though with some ambivalence, because even to me it sounds a little affected.

    Now, that’s a bridge too far! Just thinking about it feels weird to me.

  87. I’ve always thought bilious had three syllables and rhymed with supercilious.

    Apparently there are others who say it that way, but nope, it’s not standard and in fact I don’t think I’ve heard it.

  88. !

    Very surprising to me to think of bilious as having fewer than 3 syllables. (CanE speaker.) On the other hand, I think it’s one of those words that I’ve essentially only ever read.

  89. I have /lj/ in “supercilious”, and probably most words where it’s at all plausible. (Trisyllabic “Tokyo” sounds wrong in my head, though not so much in my ear.) The slender consonants of Irish have left a vestige in Irish English. There are those who have e.g. the derided /kjɑr/ for “car”, and conversely those with e.g. /fæbələs/ for “fabulous”, which could be a hypercorrection.

  90. Trisyllabic “Tokyo” sounds wrong in my head
    Wait, what, are these guys doing it wrong?

  91. But the main point is that nobody has /rj/, which is why Ryukyus is very hard for anglophones to say. The change of /rju/ to /ru/ and other /rj/ to /ri/ has run to completion.

    I say /ɛr/ and /ɛrɚ/ and /ɛrənt/ and /ɛrənd/, probably as a consequence of having no trace of Mary-merry merger, although err is a much more bookish word than the others and my pronunciation is probably by analogy. I have /æ/ in all the -parent words including parent itself.

  92. I say /ɛr/

    Huh. For me, that would be phonotactically disallowed.

  93. How do you say “air”?

  94. /ɛɚ/. That is to say, the first two phonemes in error or errant can’t form their own word for me. JC said he’s unmerged (like me), but maybe his phonotactics are a little more liberal here.

    For similar reasons, I’m always caught a bit off guard by the nicknames Larr and Jerr which sometimes pop up for Larry and Jerry.

  95. Clearly for the people who use them they are homophonous with lair and *jair: you have to have a full Mary-marry-merry merger to make sense of them.

    For me the supposed stricture against final lax vowels just doesn’t exist: when I say marry slowly it is /mæ.ri/, not /mær.ri/ (much less /meɪr.ri/ or /meɪr.i/, and the “Eh” of conscious indifference is simply /ɛ/.

  96. Yeah, I’m with you on that point. (My basic analysis of the tense-lax neutralizations is that they eliminate /V.rV/ in favor of /Vɚ.V/ – the reason why merged merry and Sirius generally sound to me like Mary and serious. Merriam-Webster, despite their aversion to IPA, is one of the few dictionaries that gets this right.) But you said you had /ɛr/ for err, not rhyming with hair or burr?

  97. marie-lucie says:

    JC: Ryukyus is very hard for anglophones to say

    This makes me think of the pronunciation of Hyundai. The spelling suggests that it should start like huge (not yuge) , but I have only heard it as Hundai (as in the Huns). But my sample may be very restricted: anyone else uses or has heard a different pronunciation?

  98. Aayte, here’s another one for you English dialect mavens. How widespread is the monosyllabic /ʌrndʒ/ ‘orange’? I heard it used by one person (born 1970s, working class, upstate NY), and saw it in an old jokey dialect book about Appalachian hillbillies. Otherwise, nowhere.

  99. I’m not familiar with /ɚːndʒ/ “urnge”, but I’ve heard several accounts of /ɔɚndʒ/ “ornge” from the non-Northeastern parts of the country. I’d guess that it tends to co-occur with disyllabic /ˈflɔɚ.də/ for Florida. (They’re /ˈɒː.ɹəndʒ/ and /ˈflɒː.ɹə.də/ in my New Englandish lect.)

  100. Arnge, not urnge, is how I remember hearing it.

  101. Oh. Yeah, I guess that could occur too (with the choice of vowel influenced, perhaps, by a broad Northeastern preference for the Tory-torrent distinction or by some upland echo of the horse-hoarse distinction) – though again, I’m not familiar with it.

  102. Eli Nelson says:

    @marie-lucie:

    This makes me think of the pronunciation of Hyundai. The spelling suggests that it should start like huge (not yuge), but I have only heard it as Hundai (as in the Huns). But my sample may be very restricted: anyone else uses or has heard a different pronunciation?

    I would pronounce it with /hj/, except in all of their ads it is clearly pronounced /ˈhʌndeɪ/. I don’t know why they chose to use the non-standard romanization “ai” for the last syllable; whenever I read it I always feel like it should be /aɪ/ since I (and, I think, most Americans) most often encounter word-final “ai” in the romanized forms of words from Japanese such as “samurai.” The usual romanization of the Korean vowel is “ae,” which also seems more intuitive to me as an English spelling (it would fit with the English spelling of “Sundae”).

    @Y:
    Like the others, I have never heard of /ʌrndʒ/ as a monosyllabic pronunciation of “orange”; only /ɔrndʒ/ and /ɑrndʒ/ (which can be realized in various ways phonetically).

  103. David Marjanović says:

    /ɛɚ/

    What, in two syllables?

    disyllabic /ˈflɔɚ.də/

    Here, too, I’m not sure what you mean. [ɚ] is supposed to be a retroflex vowel, and thus a syllable nucleus of its own. Any attempt to use it as the nonsyllabic part of a diphthong should simply result in the retroflex approximant [ɻ], a consonant.

  104. It could have been /ɑrndʒ/. I don’t distinguish vowels at that corner of the vowel space as well as an English speaker without paying special attention, and that was a word I heard in passing more than ten years ago. I do recall it being unmistakably open and unrounded.

  105. What, in two syllables?

    No more so than any other diphthong. I follow the minoritarian but not unknown practice of identifying the rhotic phonemes [ɑɚ], [ɛɚ], [ɪɚ], [ɔɚ] and [ʊɚ] in NAmEng – reflecting the fact that the vowels in unmerged Mary and merged Mary, merry, marry cluster with each other and don’t behave like the checked phoneme in better or in unmerged merry. The nature of [ɛɚ] is, in my view, that the rhotic element has a longer duration than in [ɛɹ] and – crucially – is identified with the preceding syllable, not the following one. There can be differences of quality too, but I think they’re of lesser importance. Here’s a recording of me saying [ˈmɛɹi] (unmerged merry), [ˈmɛɚi] (Mary; merged merry, marry), [mɛɚ] (mare), [mɛɹ] (*merr – a disallowed syllable).

    As I alluded above, a lot of sources would have you believe that the tense-lax neutralizations involve changes in the pronunciation of mare or near, which is, frankly, bad phonology. What I consistently hear from the rest of the country is “fairy” for ferry, “spear it” for spirit – not a difference in fair or spear.

  106. Eli Nelson says:

    Well, to me my “near” sounds more like /niɚ̯/ than /nɪɚ̯/. But this is an area like the vowel in “king,” where different people have different perceptions, and the phonetic realization may be somewhere in the middle in some cases. It doesn’t really matter whether you use /ɪɚ̯/, /ɪr/, /iɚ̯/, or /ir/ to represent the vowel of merged speakers. (The last may need to be distinguished from the sequence of sounds in “key-ring,” but that can be treated as a matter of syllabification, like “nitrate” vs. “night rate”).

    I agree that for me rhoticized vowels generally pattern more like diphthongs than like series of lax vowels + consonantal /r/.

  107. Lazar, do you merge mare and mayor?

  108. No, those are [mɛɚ] and [ˈmeɪ.ɚ] for me.

    (One other note: in my speech, at least, the onsets of [ɛɚ] and [eɪ] are basically identical – a true mid vowel – but since [ɛ] and [e] are both equally unsuitable, I keep them by convention. Maybe it’s a bad habit.)

  109. Eli Nelson says:

    Actually, all of my rhoticized vowels seem generally closer in quality to my unchecked/tense vowels than my checked/lax vowels:

    [ɑɚ̯] sounds like “r” after the normal vowel /ɑ/ that I use in “fall.”
    [e̞ɚ̯] sounds like “r” after the smoothed allophone of /e̞ɪ̯/ I use in “fail,” or the raised allophone of /æ/ I use in “gram” or “graham” (which are merged). I agree with Lazar that this seems lower than a true IPA [e], but also different from the quality I use for the English phoneme /ɛ/.
    [o̞ɚ̯] sounds like “r” after the smoothed allophone of /oʊ̯/ I use in “foal,” or the first part of the diphthong /o̞ɪ̯/.
    [iɚ̯] sounds pretty much like “r” after the normal vowel /i/ (as in “feel” or “real,” both monophthongs).
    [uɚ̯] sounds like “r” after the normal vowel /u/ (as in “fool” or “fuel,” both monophthongs). (It’s not naturally present in my speech, since I use [o̞ɚ̯] in POOR and [ɚ] in CURE/LURE/SURE. But I have acquired it in words like “tour” and “dour”).

  110. Heh. I used to say “dour” to rhyme with “tour,” but at some indeterminate time, well in the past, I decided it sounded silly and since then have rhymed it with “sour.”

  111. “nitrate” vs. “night rate”

    Also the Nye trait, which is a pottery decoration style apparently known only to linguists, not to potters or archaeologists.

  112. Rodger C says:

    I grew up saying, I think, [ɔɚndʒ].

  113. Tim May says:

    For “bilious”, the OED lists only the trisyllabic pronunciation for British English. For US it gives both (with the trisyllabic first, for whatever that’s worth).

  114. Huh, must be out of date.

  115. Yup, entry is from 1887.

  116. David Marjanović says:

    Here’s a recording of me saying [ˈmɛɹi] (unmerged merry), [ˈmɛɚi] (Mary; merged merry, marry), [mɛɚ] (mare), [mɛɹ] (*merr – a disallowed syllable).

    Interesting. I can’t hear any differences in the rhotics. The differences I hear are very slight, and they’re all in the /ɛ/; exaggerating these differences would lead to [ˈmɛɹi], [ˈmeɹi], [mɛɹ], [mæɹ].

    But then, my native Upper German leaves me unusually badly equipped to deal with the concept of “syllable”.

    As I alluded above, a lot of sources would have you believe that the tense-lax neutralizations involve changes in the pronunciation of mare or near, which is, frankly, bad phonology. What I consistently hear from the rest of the country is “fairy” for ferry, “spear it” for spirit – not a difference in fair or spear.

    With this I agree completely.

    in my speech, at least, the onsets of [ɛɚ] and [eɪ] are basically identical – a true mid vowel –

    Oh, good! I was wondering if I was remembering it right. 🙂

  117. Tim May says:

    Yup, entry is from 1887.

    The definition may be, but the US pronunciations certainly are not, as the first edition didn’t have them. I don’t think the second did, either.

  118. Tim May says:

    (That is, the earlier editions had pronunciations, but only for RP.)

  119. Ah, well, I still say it’s out of date, because I’ve never heard an American say it that way and M-W and AHD give only the two-syllable pronunciation.

  120. Kichiji says:

    In the advertisements I’ve only ever heard one pronunciation of Hyundai. It’s always pronounced hyunn-day (hj^ndej). I used to think that this was a badly mangled pronunciation, and that it should be pronounced hyoon-die (hju:ndaj) but it turns out that the hangul in standard romanisation would be hyeondae, implying that it is pronounced hyunn-deh (hj^nde) which is pretty close to hyunn-day. EDIT: I’ve just watched a bunch of Hyundai commercials on youtube, and I’ve heard all of hunn-day, hyunn-day, and hi-unn-die, but hyunn-day is still the most common. I’m guessing that the alternate pronunciaton of hunn-day is a simplification, like yod-dropping.

Speak Your Mind

*