Pronouncing it by the book

« previous post | next post »

A correspondent who had better remain nameless tells me that while dining among mostly strangers at the birthday dinner of an old friend he encountered a young woman who had an accent that he absolutely could not place anywhere on the globe. It seemed almost British, and yet not really. Eventually he just asked. She was from Northern California, but had been born in the Midwest, and she acknowledged, "Everyone always assumes I'm British or something just because I'm more careful to pronounce words properly. It only sounds unusual because everyone simply ignores how words are spelled anymore." Everyone else at the table simply nodded as though that made all the sense in the world.

My correspondent did not reply; he was simply dumbstruck. But as he listened to her some more, it became clear that she was indeed pronouncing many words precisely as they were spelled, letter by letter. She even managed to pronounce each individual vowel in diphthongs (pro-u-no-u-nce, and so on). She seemed to have no schwa — no reduced vowels, as in the first syllable of potato. Potato would have a first syllable identical with Poe for her; photography would begin with foe, commodity would begin like communist. This isn't how most people speak. But she thought that was a serious failing on their part (though apparently she did at least accept that there were some final letters that were intended to be silent: once would have only one syllable).

Later in the evening the topic of her pronunciation came up again, and she explained that people had "grown lazy". She differed from them simply in that she took great pains to pronounce things "properly, as they are written". She apparently regarded the speech of all other Americans (and, one can only assume, all other speakers of human languages), to be "lazy".

Did this anecdote from my correspondent surprise me? Not enormously. I have known people who insisted they could hear the difference between meat and meet. (I thought I could show such a person the error of their ways by saying, "OK, which one is this? [mi:t]". But the joke was on me: the person just said "That was [mi:t]!", and I had no way to explain what had just happened; it would turn into something about me being deaf to the alleged difference.)

People do hold the most extraordinarily strange views about writing and speech, apparently thinking of the latter as a poor and thoroughly imperfect reflection of the former, as if letters had come first and primitive man had slowly learned to turn them into utterable sounds.

The young woman in the anecdote above had apparently gone some way in the direction of pronouncing every letter (I'd like to hear her for myself), pronouncing letters that most people omit. She might be one of the few people (there are a few) who pronounce the p of psychology. Yet somehow I am skeptical about the notion that she pronounced every pronounceable letter. The k of knife, for example? The l of would? The g of impugn? The g and h of light? I imagine that my correspondent may be exaggerating a little, just as (I suspect) the young woman herself is.

It is possible to convince people of the truth of certain empirical claims about their own pronunciation. It involves recording them and making them listen to the playback. The linguist David Crystal was doing a workshop for British teachers and found that one of them regarded intrusive [r] as an abomination. That is, she regarded it as utterly wrong and unacceptable and coarse to pronounce the idea of it as the idearof it, as millions of British speakers do. So he had her say a few phrases like the idea of Africa and Asia on the other hand, and played her back. There were some clear cases of intrusive [r], and when she listened she could hear them herself. For a moment there was silence, and then the woman simply burst into tears.

David Crystal tells that embarrassing story with no relish at all. It's a dangerous game, monkeying with someone's perception of something as precious to them as their native language and how they speak it. Be careful out there.



165 Comments

  1. Ralph Hickok said,

    November 9, 2010 @ 2:06 pm

    I had a friend at Harvard who spoke in much the same way. When I first met him (in the Lowell House poolroom, of course), I asked where he was from and he said "Oxford," and I accepted that without comment. I eventually learned that his hometown was Oxford, Ohio.

  2. Mel said,

    November 9, 2010 @ 2:06 pm

    I'm guilty of pronouncing some letters most people don't (like the 't' in 'often'–off-ten), because I read precociously and voraciously. It's especially common with words I read before I heard them.

    I'm sort of…baffled…by the idea that English can be pronounced entirely as written, though.

  3. Jim said,

    November 9, 2010 @ 2:10 pm

    Heh, I had to think a minute about the commodity/communist distinction. In my head they sounded the same, until I tried actually saying them out loud. That schwa's a sneaky character!

    [I'm sort of presupposing some basic phonetics here, and not everyone gets the hang of it (as some of the comments below show!); but you're exactly right: you have to say it out loud (naturally, not artificially and defensively as if challenged over your pronunciation accuracy), and actually listen to what comes out. Then you realize that commodity, with the stress on the second syllable, has a first syllable that virtually never sounds anything like the first syllable of communist. Lesson one. And lesson two is that the vowel you get in commodity is called "schwa" (a term from the Hebrew grammarians), and has no letter in English that uniquely corresponds to it. Linguists write it as [ə] (I'm assuming you have a browser that recognizes Unicode!). —GKP]

  4. Dw said,

    November 9, 2010 @ 2:12 pm

    We invest enormous resources in teaching children written language and very few in even discussing spoken language. It's hardly surprising that some people get the idea that written language is "correct" while spoken language is some kind of debased variant.

  5. Ben C said,

    November 9, 2010 @ 2:25 pm

    My wife is from Salt Lake City, and I've brought up interesting facts about dialect over the course of her marriage, and it's still difficult for her to believe that I'm not criticizing her speech when I do this. I think it's human nature to believe that language should have some kind of default form.

  6. Mark P said,

    November 9, 2010 @ 2:27 pm

    What some people have trouble learning is that their perceptions are not reality; they think they pronounce words in a certain way and are convinced that they hear them that way, too.

  7. Jens Fiederer said,

    November 9, 2010 @ 2:37 pm

    I share some similarities with the young woman, although I have no illusion that my way is the right way. I have been mocked for pronouncing an "l" in "salmon" and the second "b" in "bomber" (a term for a kind of sandwich in our area, not an instrument of war). Part of that might be from growing up in Germany, where most of our letters are pronounced (although I was later completely surprised to learn that the final consonant in "Rat" and "Rad" are pronounced exactly the same way. Part of it might be that I have encountered most words first by reading them rather than hearing them in speech – in fact, there have been several instances of vocabulary coalescing (I just made that term up, there might be a proper term for it) where I found out that a word I knew from reading and a word I knew from hearing were actually the same word (but the spelling was an odd one).

  8. Colin Reid said,

    November 9, 2010 @ 2:38 pm

    What I find particularly odd is the idea that someone who pronounces every letter and doesn't centralise unstressed vowels would sound British, even to Americans. 'Foreign', yes, but surely Americans have heard enough British accents to expect to hear some schwas? Are there any significant populations of native English speakers in the world that don't centralise unstressed vowels?

    As to the claim that English speakers have 'grown lazy', how modern a feature of English is centralisation of unstressed vowels reckoned to be? Did it happen after the great vowel shift for instance?

  9. Coby Lubliner said,

    November 9, 2010 @ 2:42 pm

    There are some registers in spoken English in which vowels that are reduced in ordinary speech are given a full pronunciation. A good many lawyers, for example (including the current President and Secretary of State of the USA), routinely pronounce the indefinite article a as /e/, typically realized as [ɛɪ]. I remember that Christopher Darden, who prosecuted O. J. Simpson, consistently referred to him as the [di'fɛndænt].

    There are also styles of singing (hymns, classical music, old folk songs) in which many, if not all, unstressed vowels are pronounced as full.

    Whether this is the result of reading or of oral tradition is hard to tell.

  10. Dw said,

    November 9, 2010 @ 2:47 pm

    @Colin Reid:

    Centralization of unstressed short vowels dates back to Middle English. However some dialects, including the prestige speech of southern England, short I resisted centralization. This is why today some accents distinguish pairs such as "except" and "accept" while others merge them.

  11. Dw said,

    November 9, 2010 @ 2:51 pm

    @Colin Reid:

    In answer to your other question, some Caribbean accents generally resist centralization. Wells's "Accents of English" vol. 3 would have more details, but I don't have it in front of me.

  12. John Cowan said,

    November 9, 2010 @ 2:52 pm

    Of course, there are a few people in Ireland and the West Country who still distinguish between meat and meet, just as there are a few folks from the 'folks who even yet lack the long-mid mergers and so differentiate between pain and pane, tow and toe. It takes simply forever for distinctions to be utterly lost everywhere in this highly pluricentric language of ours.

    [Yes; but the person I was referring to definitely said [mi:t] for both words. —GKP]

  13. octopod said,

    November 9, 2010 @ 2:58 pm

    Ha! When I started reading this I thought it might be about me, because just last night a linguist whom I met in a bar asked me about my "accent". (We concluded, though, that I do not have this kind of hypercorrection accent but rather something which one might call an "opera accent".)

    I tend to mentally tag accents with decreased centralization as West African, fwiw, or sometimes Caribbean as above.

  14. Quintesse said,

    November 9, 2010 @ 3:07 pm

    @Jens Fiederer: Being Dutch I was actually thinking the same thing. Because we tend to pronounce most letters I had to think a moment about the "commodity/communist" example. And the funny thing is that even though most Dutch speak very well English there are many examples where people will actually pronounce the K in knife ;)
    (And of course in another day and age people did actually pronounce the g in light, the Dutch and the Germans still do)

  15. gribley said,

    November 9, 2010 @ 3:08 pm

    Guy Deutscher claims that the "laziness" of speakers in dropping syllables and otherwise simplifying pronunciations is a major force in language change, in his excellent The Unfolding of Language. His theory of erosion and accretion seems to me a very plausible theory of how some structures, like grammatical endings, evolve; but I can't tell whether it's his own pet theory or is widely held.

    The speaker in question, of course, is presumably one of those types who don't believe in language change. I guess we're just lucky she doesn't insist on speaking classical Latin at the dinner table.

  16. Spell Me Jeff said,

    November 9, 2010 @ 3:14 pm

    I once made the horrifying mistake of commenting to an acquaintance that her accent sounded a bit formal to me. (It was not quite Thurston Howellish, but heading that way.) My friend explained (with a touch of indignation) that as a child she'd had some sort of speech impediment, and that the way she spoke now was the result of many years of therapy.

    I wonder if the woman in this post is actually covering for something similar.

  17. Jens Fiederer said,

    November 9, 2010 @ 3:23 pm

    @Quintesse the "g" in light? In German we have "Licht" ("c" instead of "g") but the "c" is not pronounced separately, it is part of our "ch" sound…I don't know about Dutch (I read an Asterix and Obelix comic book once that was written in Dutch and understood a good portion, so it's not DREADFULLY far from German).

    Never been tempted to pronounce "k" in "kn", although that is natural to a German, but HAVE pronounced the "p" in psychology, the "k" sound in Xerox, and the "t" in "tzar".

  18. Olga said,

    November 9, 2010 @ 3:24 pm

    This reminds me of a phonetics assignment I gave a few years ago. Students were given English words (written) and told to transcribe them as they would pronounce them. One of the words was "indictment". You can guess what happened.

  19. jadagul said,

    November 9, 2010 @ 3:43 pm

    I'm something of a hypercorrecter myself–for instance, I do pronounce the "p" in "psychologist", and usually silent "g"s at the beginning of words ("gnomic", for instance). I distinguish cot-caught and marry-merry-Mary, and sometimes gilt-guilt.

    It's partly pure pretension and partly a result of several years of debate, followed by several years of singing; there are certain types of hypercorrected pronunciation that make you much easier to understand even though they sound ridiculous in common speech. (The clear example here is final "d"s and "t"s; the way most English speakers pronounce them makes them almost completely inaudible in adverse circumstances).

    And I can attest that Americans often mispeg this as a British accent, basically because many Americans think "British accent = RP accent = impeccably correct English."

    Aside to Coby Lubliner: in classical singing, which is about the only one I'm qualified to comment on, we have a tendency to resist vowel reduction just because a schwa is much harder to make sound good than most other vowels.

  20. vanya said,

    November 9, 2010 @ 3:44 pm

    A very interesting young woman. All the examples given show her simply pronouncing unstressed vowels and dipthongs with different values than a normal American, there are no examples with silent consonants. I would guess that she must have internalized when young that silent consonants and "silent e" are normal and correct. It's the vowels that lazy people pronounce wrong. This makes sense to me given the way we were taught the rules of English in grammar school – I remember spending a lot of time on silent e, silent "gh", etc. but never ever did we learn what a schwa was or that unstressed vowels can have different values. Probably lots of Americans go around thinking that poe-tay-toe is correct and that "schwa" is not a "real" sound.

  21. Ferdinand Cesarano said,

    November 9, 2010 @ 3:48 pm

    "…everyone simply ignores how words are spelled anymore"

    This person should have paid less attention to overprecise pronunciations, and more attention to grammar.

    This affirmative "anymore" (in place of, perhaps, "nowadays") strikes me as a major clunker. "Anymore" is part of a negation of a verb. It would require a verb to negate (such as, for example, "pay attention (to)"); it would also require either a negative subject ("no one", "nobody") or a negative word before the verb ("not", "scarsely"). So, "no one pays attention anymore" or "people don't pay attention anymore".

    Odd that this (spurious) precision in pronunciation and this slovenly grammar occur in the same person. These two traits, taken together, make me think that this person uses study aids such as flash cards, but does not read much.

  22. Jonathan Mayhew said,

    November 9, 2010 @ 4:00 pm

    I'm curious, Jadagul. How would you distinguish "gilt" from "guilt"?

  23. empty said,

    November 9, 2010 @ 4:01 pm

    I have definitely noticed that many British speakers give the (unstressed) first syllable the full vowel treatment rather than a schwa in words like "control". Likewise the unstressed final syllable in words like "bollard" and proper names like "Maynard" and "Concord".

  24. John Thayer Jensen said,

    November 9, 2010 @ 4:04 pm

    Regarding 'intrusive r' – I grew up in California, have lived in New Zealand for the last 37 years. My understanding of the phonology of New Zealand word-final non-high non-front vowels is that they have a final -r glide.

    Interesting confirmation of it has been my older son's acquisition of some such words when he learned them for the first time in New Zealand. He was 8 when we moved here. His mother's speech is pretty standard American west coast. Mine is pulled a fair bit towards New Zealand English. My son, fairly early on, decided on his mother's pronunciation as his standard (my other three children all have more or less New Zealand speech patterns). The word 'spa' – used in New Zealand in the phrase 'spa pool' to refer to a 'hot tub' – he first learned here. He pronounces it 'spar' with a fully rhotacised 'r'. I assume he (correctly) phonologised the vowel, and gave it his phoneticisation.

    My wife has a Dutch friend whose first name is Alida. She, and all her friends, use the shorten form 'Alie' – with a "long A" – and my wife pronounces it "Arlie" – with rhotacised 'r'

    And I once heard a New Zealand comedian on television faking an American accent. He pronounced the word 'bra' (brassiere, that is) 'brar' with a full -r at the end.

    My wife does some tutoring of reading. She had to learn new minimal pairs. "Law" and "lore" are homonyms.

    But if you try to talk to New Zealanders about their speech, if they are fairly literate, they deny this homonymy. You have to listen to what they say.

    jj

  25. jo said,

    November 9, 2010 @ 4:06 pm

    @jadagul (via @Coby Lubliner)

    Intuitively I'd agree that classical singing an example of a 'speech' genre that resists vowel reduction. However, I do think that Jadagul's suggested cause for it might be a case of backwards reasoning. That is, I don't agree that there is anything uniquely difficult about singing a schwa — classical singing in French surely has frequent examples. I would be more inclined to think that "schwa is hard to sing" is a post hoc rationalisation of a resistance to vowel reduction that exists for some other reason. In fact, I think you can often hear the justification (from non-linguists) that some sound or combination of sounds (e.g. /kn/) is "hard to say" as a the reason for avoidance of something in one language (e.g. English) that is perfectly sayable as part of other languages (e.g. German).

  26. ggustafs said,

    November 9, 2010 @ 4:11 pm

    I think you need to be very careful about attributing the way the woman's accent to her stated (or conscious reason). This is probably just her defensive reaction to the intrusive question and felt embarrassed. It is very likely that there is another reason, anyone who is around classically trained singers, will recognize the this in not unusual in their conversational speech. Additionally, for native Northern Californians it is quite common to hear "potato" and "photography" to as described

  27. Mark P said,

    November 9, 2010 @ 4:13 pm

    @Olga – in my previous life as a newspaper reporter I did a story about a Mennonite community in east central Georgia. In a comment on a story he had heard on the radio one man said that someone had been "in-dict-ed." I might have assumed it was from not hearing the word spoken, but, as I said, he was commenting on a radio report. This community was fairly isolated, including running their own schools using the few college graduates from their community as teachers. They allowed the use of some modern technology, like radios, but not others, like television.

  28. David Houghton said,

    November 9, 2010 @ 4:15 pm

    @Ferdinand

    In some dialects 'anymore' isn't a negative polarity item anymore. This is certainly the case in West Virginia. I believe it is common throughout the South.

  29. Dan T. said,

    November 9, 2010 @ 4:17 pm

    If you can pronounce the "k" in "knish" and the "l" in "Salman Rushdie", you should be able to pronounce the "k" in "knife" and the "l" in "salmon".

  30. Ben said,

    November 9, 2010 @ 4:25 pm

    I love her flagrant use of "positive anymore"; not the most acrolectal trait (though a charming one). I hope the quotation is accurate in this regard.

    @Cory Lubliner: "Defend-ant" is lawyer shtick; it's their way of saying, "You can tell we're experts because we pronounce an everyday word strangely." It's like doctors and "sontimeter." And not only does the president often pronounce the indefinite article as [ɛɪ], he also often uses "a" where one might expect "an." I don't know if that's a lawyer thing or just a quirk of his idiolect, but it's interesting.

  31. JLR said,

    November 9, 2010 @ 4:29 pm

    Can you imagine if someone tried this with French?

  32. Jonathan Mayhew said,

    November 9, 2010 @ 4:34 pm

    Has there been a LL post on positive "anymore"? I find it charming too when I hear it, because of its subtle implication of negation.

  33. David Fried said,

    November 9, 2010 @ 4:35 pm

    Cory,

    The pronunciation of "defendant" that you cite is very common, almost standard, in the criminal bar. I think it's an idiosyncracy or bit of jargon common to the profession, not an instance of a more general accent or manner of pronunciation. I have no real explanation for it, but in my days as a prosecutor I often thought it had a slightly pejorative ring. Civil attorneys rarely use it, in my experience (I certainly don't). I would be interested in a study to determine whether the defense bar avoids it.

    The only other instance of lawyers' pronouncing a common word with an idiosyncratic stress is the pronunciation "pre-CEDE-ent" in the fixed phrase "condition precedent." Of course when lawyers refer to "legal precedents" they use the ordinary pronunciation with stress on the antepenult.

  34. Mark F. said,

    November 9, 2010 @ 4:35 pm

    My guess is the thing that most made her sound British is that she probably didn't flap her intervocalic t's. That's a pretty big tell for the UK/US difference.

    I like to listen to the Jodcast, Jodrell Bank's astronomy podcast, and I've noticed that they and I both make a distinction between "meteor" and "media", but it's at a different place in the word. I pronounce the 't' and 'd' identically, but the ends of the words differently. For the Jodcast regulars, the opposite is true.

  35. Thomas Thurman said,

    November 9, 2010 @ 4:41 pm

    And I once heard a New Zealand comedian on television faking an American accent. He pronounced the word 'bra' (brassiere, that is) 'brar' with a full -r at the end.

    It's surprisingly difficult to see past mergers in one's own dialect. I'm from southern England, and I've noticed myself that I'm unable to affect a rhotic accent because I can't tell on the fly which vowels are r-coloured. (If I stop and think about it, I can tell from the spelling— but not on the fly.)

    In similar vein, I help run a website about the Shavian alphabet, which has separate letters for the vowels in "cot" (

  36. Thomas Thurman said,

    November 9, 2010 @ 4:42 pm

    …and "caught". Several people who have the cot/caught merger have complained quite bitterly about this pair of letters, because it turns out that for the life of them they can't tell which one to use.

    In other news, including Shavian letters breaks the comment form on Language Log.

  37. Sandy Nicholson said,

    November 9, 2010 @ 4:57 pm

    The thing that particularly struck me about what the woman said was not the essence of her claim but the way she put it (assuming the quote is verbatim):

    It only sounds unusual because everyone simply ignores how words are spelled anymore.

    Leaving aside the fact that I tend to pronounce it ‘any more’ ;o) I was thrown by the peculiar turn of phrase. Had she said ‘no one cares how words are spelled anymore’, that would have been fine. Had she said ‘everyone simply ignores how words are spelled nowadays’, I’d have been happy. (Actually, I’d have concentrated more on the message and worried about that instead.)

    I was about to ask why no one else had commented on this seemingly odd use of anymore in a positive, declarative clause (a performance error, perhaps?), but the OED seems to offer an out for the woman: in sense 2 (chiefly Irish English and N. Amer. colloq.), any more can indeed mean something like ‘nowadays’. Am I the only one who was thrown by this?

  38. GeorgeW said,

    November 9, 2010 @ 5:10 pm

    @John Cowan: How are 'meet' and 'meat' pronounced by those who make the distinction?

  39. DRK said,

    November 9, 2010 @ 5:14 pm

    Given that that the British are the people who gave us "Woostersher" for Worcestershire and "Chumley" for Cholmondeley, it seems a little funny to hold them up as paragons of full pronunciation of words as written.

    Her turn of speech as described sounds so odd and forced that I wonder if it is the result of some sort of self-imposed speech therapy, perhaps for stuttering.

  40. Thomas Thurman said,

    November 9, 2010 @ 5:25 pm

    @Sandy Nicholson: See the comment above yours by Ferdinand Cesarano.

  41. richard howland-bolton said,

    November 9, 2010 @ 5:27 pm

    @ GeorgeW
    I suspect that meat sounds more like 'mate' (as 'tay' for tea)

    @ DRK
    and 'windm' for Wymondham
    and my personal favourite 'hazebrugh' (ending with a schwa) for Happisburgh

  42. Peter G. Howland said,

    November 9, 2010 @ 5:29 pm

    What an interesting coincidence. A few months ago here in the American southwest, a visitor at a neighbor’s house across the street came over to ask me if I could help guide her through the installation of new line in their power weed trimmer. As I was in the process of telling her what little I knew of the operation, she interrupted to ask what part of England I was from.

    The woman, on the mature side of 40, was clearly from England herself and when I explained that I was born in Brooklyn, schooled in southern California, wrote technical proposals at Stanford Research Institute, produced architectural plans in Idaho, bucked hay on a farm in Montana, and now do nothing more than enjoy my ensuing geezerhood here in “Preskit” , Arizona, she was astonished. As was I for being “accused” of being a Brit simply for attempting to communicate clearly.

    It’s simply an unconscious habit of mine to structure my sentences carefully, articulate clearly and enunciate distinctly when speaking to someone about a technical matter or an issue of importance to me. And of course, like everyone else, I don’t have an accent (!)

  43. The Ridger said,

    November 9, 2010 @ 5:39 pm

    Ben says "And not only does the president often pronounce the indefinite article as [ɛɪ], he also often uses "a" where one might expect "an." I don't know if that's a lawyer thing or just a quirk of his idiolect, but it's interesting."

    I pronounce "a" as schwa unless it's emphatic, but I too often find myself saying "a" where "an" is expected. In fact "a [ɛɪ]" and "an" are both emphatic, and "a schwa" is more or less the unmarked, unemphatic, neutral normal… And so it is for lots of people I know.

  44. dw said,

    November 9, 2010 @ 5:39 pm

    @empty:

    But in other words it's the other way round. For example, in "python", "Amazon", "jaguar" and no doubt many other words it's the Americans that retain the unreduced vowel, while the Brits have a schwa (or syllabic N).

  45. dw said,

    November 9, 2010 @ 5:46 pm

    British television shows exported directly to the US have historically tended to be documentaries and period dramas rather than gritty urban soap-operas. Similarly, British actors who go to Hollywood have tended to the RP end of the spectrum, at least until recently (Michael Caine is, I suppose, the most obvious exception).

    No doubt this is part of the reason that careful, consciously "correct" speech may be interepreted by Americans as "British" or "English".

  46. Alan Gunn said,

    November 9, 2010 @ 5:49 pm

    "The only other instance of lawyers' pronouncing a common word with an idiosyncratic stress is the pronunciation "pre-CEDE-ent" in the fixed phrase "condition precedent." Of course when lawyers refer to "legal precedents" they use the ordinary pronunciation with stress on the antepenult."

    There's nothing idiosyncratic about the way lawyers pronounce "precedent" when it's an adjective meaning "having come before," as in the phrase "condition precedent." It's analogous to the difference between pronouncing the noun "arithmetic" and the adjective "arithmetic" (as in "arithmetic fraction," with the emphasis on the third syllable). It probably seems odd to non-lawyers because people who aren't lawyers seldom use the adjective form of "precedent"; indeed, lawyers themselves don't use it much except in connection with the historical but ridiculous distinction between conditions precedent and subsequent.

  47. Mark F. said,

    November 9, 2010 @ 5:50 pm

    Sandy Nicholson — The "positive anymore" is one of the ways you can tell someone is from the Midwest in the US, although its use is spreading.

  48. Layra said,

    November 9, 2010 @ 5:58 pm

    @Johnathan Mayhew:

    Oddly enough, I can't hear the difference myself, but I pronounce "guilt" with a more rounded vowel than "gilt".

  49. Layra said,

    November 9, 2010 @ 5:58 pm

    That is, I can't hear the difference but I can tell that my lips are doing different things.

  50. Chris Kern said,

    November 9, 2010 @ 5:59 pm

    Yeah, although I don't personally use it, that usage of "anymore" is unremarkable enough to me (as a Midwesterner) that I didn't even notice it in the original post.

  51. marie-lucie said,

    November 9, 2010 @ 6:02 pm

    JLR: Can you imagine if someone tried this with French?

    That person would end up pronouncing more or less as in Old French (9th to 14th century at the outmost (?)), except that OF also pronounced "ch" and "j" or "ge" as in current English. But the now normally silent letters (esp. in the endings -s and -ent [3rd person plural]) were pronounced.

    jo: singing a schwa — classical singing in French surely has frequent examples

    It is not just "classical" singing, pop singing also does it, but with more of a choice of pronouncing schwa or not at the end of a phrase, as in spoken Standard French.

  52. Troy S. said,

    November 9, 2010 @ 6:05 pm

    I wonder if she preserved the wr digraph in "written". I try do this sometimes, mostly just to annoy my wife. It's fun watching her roll her eyes when I try to order a wrap. It's also quite difficult to pronounce correctly. And an almost sure sign that any word that has it is pure Old English.

    Incidentally, I've been seeing "racked"a lot in print where I expect "wracked". Which one is the eggcorn?

  53. ~flow said,

    November 9, 2010 @ 6:07 pm

    now i will start to teach all my fellow german speakers only ever to say [ix] and [ax] for ich and ach. after all, those are spelt the same! also, szene should be [stse:ne:]. just to think how much fun [p-ho:to:gra:p-hi:] can be! ancient greek would look up to us!

    once i was in the library with an old friend of mine who was born and raised in taiwan. she has an education as a lawyer but works a lot as an interpreter. in the east asia section, she took that book from the shelf and for some reason it was important how that title would have been written in the romanized catalog—was it filed under 'ting' or 'ding'? now there stood a native speaker of mandarin who couldn't answer a pretty simple phonological question about her mother tongue; as far as i remember, it was really a simple character.

    likewise, i had a native chinese teacher from the mainland. a colleague of her let us in on the 'secret' that she often was not quite clear about the tone of a given word. she pronounced the word alright, and her non-native colleagues had no trouble which of the only four plus one tones of putonghua she had just been using. she had published a book on chinese grammar at one point in her life, so she must have had some kind of background.

    these two anecdotes lead me to believe that much of what we think we know about pronunciation is really what we derive from writing and formal education in general. schoolkids in taiwan and the mainland do get taught phonetic scripts with tones, but they have very little incentive to use them as grownups. similarly, people in germany care very little about the [iç] / [ax] distinction as they don't have to express that detail in the orthography.

  54. fog said,

    November 9, 2010 @ 6:26 pm

    If I had never found Language Log (and a few other influences) I can completely imagine myself talking like that. I pronounce the first R in February even though I never made a conscious decision to do so. I would never hear someone else say Febuary and think it was wrong, and I didn't even realize there were two different pronunciations until I was having unexpected spelling troubles one day.

  55. marie-lucie said,

    November 9, 2010 @ 6:28 pm

    One place where vowels are being differentiated is with the endings -er and -or in nouns designating persons: at least among radio or TV announcers, =or is often pronounced stressed and as in for in words like mentor and ambassador (not in honor or labor, for instance). Perhaps this is because most -er nouns derive transparently from verbs, as in baker and teacher (but not butcher, for instance), but -or nouns do not? Certainly a virtual verb stem *ment is implied in the noun mentee corresponding to mentor.

    Obama's pronunciation of the article "a"

    Pronouncing the article a like the letter "a" in isolation is certainly not peculiar to Obama's speech: it makes the article more salient, as well as slowing down speech a little. In addressing an audience (especially a large one) during formal proceedings, in a place where clear, distinct pronunciation and careful choice of words are very important to being understood on the spot, every detail which helps the audience process the words is important. If he says a instead of an, it could be because a with that pronunciation requires a slight pause, and an does not.

    [(myl) Are some people complaining about Obama using unreduced "a"? Curious, because this was also a complaint about his predecessor — see here, here and here.]

  56. PeterSzigetvari said,

    November 9, 2010 @ 6:31 pm

    @GeorgeW: you probably do the distinction in pairs like _great_ vs _grease_. The people John Cowan talks about have the same contrast more extensively. (See Wells' Accents of English, pp. 194f.)

  57. peterv said,

    November 9, 2010 @ 6:52 pm

    It's the pronunciation of letters that aren't in the written word which I find interesting in American English. Young people on the phone commonly pronounce the letter y at the start of "hello", and the letter m at the start of "bye". Does anyone know a reason for this?

  58. dw said,

    November 9, 2010 @ 7:18 pm

    @peterv

    I would imagine that the intrusive /m/ in "bye" is the result of the lips closing before the voicing has been turned off.

    How could one pronounce a "letter y" at the start of "hello". Is the /h/ not pronounced at all?

  59. Spell Me Jeff said,

    November 9, 2010 @ 7:36 pm

    "Hello" pronounced "Yello," but if my experience is common, there are restrictions.

    Y is not used when the word is articulated as a question. "Hello?"
    Y is not used if the stress pattern would homophone with "yellow"

    That is to say, "Yello" is ok when the 2 syllables have roughly equal stress.

  60. Ellen K. said,

    November 9, 2010 @ 7:40 pm

    Jo, Jadagul did not say that a schwa is hard to sing, but rather that it's hard to make sound good. Two different things.

  61. John Thayer Jensen said,

    November 9, 2010 @ 7:42 pm

    @Thomas Thurman:

    Yes – my wife has the same (short) vowel in 'cot' and 'caught' When she was trying to tutor New Zealand children, that pair is different – but they pronounce 'caught' and 'court' the same, because, as I think, they are phonologically identical – both have the 'r'

    Once, when I had lived in New Zealand for less than two years (I was a linguistics lecturer at the time), I was asked to give a talk to our Annual General Meeting on New Zealand phonology. The AGM talk is always a little informal, and lubricated with cheese and wine, so I suppose I did all right, though I knew very little except from personal observation. But it was great fun to develop the talk.

    That was late 1974, I think :-)

    jj

  62. groki said,

    November 9, 2010 @ 7:44 pm

    @peterv, @dw

    I am long past "young people" so no longer a native speaker, but I conjecture elision, as in:

    "yello" = "y[es h]ello" and "mbye" = "[mm-h]m, bye"

    that's how I interpret them, anyway. I think I dimly (or dimly think I) recall hearing formulations like these in standup routines over the years.

  63. Don said,

    November 9, 2010 @ 7:46 pm

    I am in a chorus that just performed Beethoven's Ninth. Our director repeatedly emphasized that we needed to sing schwas to correctly pronounce the German text — e.g., in the second syllable of menschen.

  64. Xmun said,

    November 9, 2010 @ 7:48 pm

    Have you all heard about the clergyman who pronounces the t in "epistle"? I know of him only by hearsay, but I knew someone once who pronounced the t in "mortgage". She was born and bred in Glasgow and had kept her Scottish accent despite living for many years in NZ. Did she also pronounce the post-vocalic r? I can't remember, but I don't think so.

  65. dw said,

    November 9, 2010 @ 8:00 pm

    @peterv:

    Sorry, I should have read your post more carefully. An intrusive [m] at _the beginning_ of bye is a very easy phonetic phenomenon to explain. To pronounce the [b] in "bye" the speaker has to
    * bring the lips together (if they are not already together) and release them
    * raise the velum (soft palate) so that the air does not pass into the nasal cavity
    * activate the vocal cords.

    If the vocal cords are activated prematurely, while the lips are together but before the velum is raised, then a [m] will be the result.

  66. J. Goard said,

    November 9, 2010 @ 8:07 pm

    My mom, raised near Chicago, uses positive "anymore", and doesn't seem to fully grasp what I (California) find strange about it.

  67. Danmcc said,

    November 9, 2010 @ 9:33 pm

    I once had to watch an enormously frustrating conversation between two native English speakers. One was Australian and one from the United States (forget exactly where). The Australian was bemoaning the fact that she didn't pronounce her words properly and posited that this might be because she had never been taught proper pronunciation in school.

    She proceeded to ask the American to teach her how to correct her speech. He, confident that he didn't have an accent, started to school her in pronouncing things according to his own dialect. Memorably one piece of advice was "Just pronounce it the way that it sounds". Both were genuine in their believe that he had the "correct" pronunciation and that she "should" speak more like him.

    My head nearly exploded.

  68. jadagul said,

    November 9, 2010 @ 9:33 pm

    Ellen, Jo, Don: Ellen nailed it. A schwa is just as easy to produce as other sounds, if not easier; but it's really hard to keep it from sounding flat and nasal. So in mediocre choirs especially it gets avoided all over the place. (Don, I actually have that problem in one of my choirs: I'm good enough to want to pronounce the schwas in German but most of the people in this choir just can't do it so the conductor is having us give them more value).

    You can see the same phenomenon with the "i" vowel (sorry, don't know IPA though I really should learn). It's hard to sing without letting it go really nasal. A lot of classical singers let it drift into "ih" instead. Most pop singers actually let it drift towards an "ey"; if you want a really clear example of this, listen to Lady Gaga's "Telephone" which always bugs the crap out of me ("I'm kinda biz-zey, I'm kinda biz-zey").

    Jonathan Mayhew, Layra: I place the vowel in "gilt" somewhat higher and further back in my mouth than that in "guilt." In "gilt" the corners of my lips pull out to the side a bit; in "guilt" they come forward and round. Sort of like the difference between a Spanish and an American "r".

  69. Matt McIrvin said,

    November 9, 2010 @ 10:08 pm

    As a bookish kid, I learned many words by sight without knowing how to pronounce them. I wonder if something like that is a contributor to this woman's "careful pronunciation" habit.

  70. Joe McVeigh said,

    November 9, 2010 @ 10:24 pm

    Geoff says: "She apparently regarded the speech of all other Americans (and, one can only assume, all other speakers of human languages), to be "lazy"."

    While she may assume that, I just wanted to point out that Finnish speakers do indeed say every letter.

    [Purleeeze! Let's not talk like dimwits. It's not about pronouncing the letters! Not even in Finnish it isn't. (Think of the Pirahã, who have no writing at all and never did; what do they pronounce?) Writing systems are a 5,000-year-old cultural invention for recording some aspects of speech in permanent visible form. Speech came first (probably at least 50,000 years earlier). What the Finns have done is invent a writing system that fits the pronunciation of their language almost perfectly. That's nothing to do with "saying every letter". It's to do with spelling conventions that match the sounds that people utter. Do try to get this straight. This is Language Log, you know. —GKP]

    At least, sort of. In Standard Finnish, there's only one word, according to one of my Finnish teachers, where native speakers do not pronounce every letter – ruoat (plural food).

    Of course, Standard Finnish is only spoken in Parliament, official speeches, and setting like that, but that's one of the first rules they teach you when you learn Finnish – say every letter. I wish I could tell you more, but I'm not fluent in Finnish. I just lived in Finland for about seven years. I do know that some of my friends at the University of Helsinki read this site. Maybe they can shed some more light on this topic. Don't know if they read the comments though.

    Also, how did your friend not laugh out loud at this woman. That would have been my first reaction. And then I would have asked her about her views on how laughing should be pronounced.

    [So you would have been rude and tried to embarrass her. That is why you weren't invited to the dinner. —GKP]

  71. Ray Girvan said,

    November 9, 2010 @ 10:24 pm

    There were some clear cases of intrusive [r], and when she listened she could hear them herself

    I've never dared do the actual experiment, but I've similarly met people who insisted they don' use glo'al stops.

  72. John said,

    November 9, 2010 @ 10:29 pm

    @peterv: Given that these are young people, I think you have to consider the possibility of ironic pronunciations.

    I've heard 'Yello' for at least 50 years (US, Upper Midwest, New England) and think it might have been a tag line for a comedian in a variety or vaudeville show. Current use (outside the name of a band) could be considered 'retro' in an amusing way.

    'Mbye' may be a twist on animated TV series South Park and its character Mr MacKey who pronounced "okay" as "'mkay".

  73. Ray Girvan said,

    November 9, 2010 @ 10:33 pm

    @ John Thayer Jensen He pronounced the word 'bra' (brassiere, that is) 'brar' with a full -r at the end.

    In some of the stronger local accents down here (Devon) they do. A couple of years back, the speech of the delegate from Honiton in the annual town criers' competition contained, as part of its spoof celebration of Honiton's textile industry, the memorable line:

    We made Madonnerrr's brarrr, for all to marrrvel at

  74. Jim said,

    November 9, 2010 @ 10:34 pm

    I've been queried about being British for much the same reason: people hear dialectic precision and map it to the only thing they can come up with. (Too much 'Enry 'Iggins, gov!)

    "Often" is one of those words, and I put the slightest hint of the "t" in "castle" as well.

    I like Layra's comment about words sounding the same but you can tell you are making different lip formations. Definitely true with cot/caught and pen/pin, and my favorite example trio: gym/gem/Jim

  75. ella said,

    November 9, 2010 @ 10:43 pm

    @John Thayer Jensen

    My son, fairly early on, decided on his mother's pronunciation as his standard (my other three children all have more or less New Zealand speech patterns). The word 'spa' – used in New Zealand in the phrase 'spa pool' to refer to a 'hot tub' – he first learned here. He pronounces it 'spar' with a fully rhotacised 'r'. I assume he (correctly) phonologised the vowel, and gave it his phoneticisation.

    This is a very common phenomenon in r-ful children of r-less parents. My parents (and I) speak Southern BrE (well, I code switch, but that's another matter), but my brothers have lived their whole lives in Canada, and have always spoken with a rhotic accent as a result. One of their most memorable hypercorrections was when they were 4 or 5 and decided that the word 'dawn' was pronounced 'dorn'

  76. David P. said,

    November 9, 2010 @ 10:46 pm

    Henry James, the novelist (American, but moved to England), conversing with a niece visiting from America. The niece said something about "jewels".
    James: "That's pronounced 'jew-els' ".
    Niece: "I'm afraid we don't pronounce our vowels the way you do."
    James: "That's pronounced 'vow-els' ".
    Niece: "Oh, Uncle Henry! Don't be cruel!"
    James: "That's pronounced 'cru-el' ".

  77. Kragen Javier Sitaker said,

    November 9, 2010 @ 10:46 pm

    If she really pronounced words the way they were spelled — e.g. "pronounce" as [pronounse] or [pronounke] instead of [prənæʊns] as in RP or (what I hear as) [prənæuns] in American Anchorperson — I think most people in the US would peg her as having an incomprehensibly thick Mexican accent, not a British accent.

  78. Robert Furber said,

    November 9, 2010 @ 11:51 pm

    >> I thought I could show such a person the error of their ways by saying, "OK, which one is this? [mi:t]". But the joke was on me: the person just said "That was [mi:t]!",

    I have to admit I burst out laughing at this. It's oddly reminiscent of Abbott and Costello's "Who's on first?".

  79. onymous said,

    November 10, 2010 @ 1:21 am

    I'm inconsistent about pronouncing the 'p' in words that begin with 'ps'. This is something I would never have done until college, when I suddenly found myself with the need to say the names of the Greek letters 'psi' and 'xi' and distinguish them on a semi-regular basis. Since then the 'ps'- pronunciation has started to creep into other parts of my speech, but occasionally I notice myself doing it and it just sounds weird. And saying 'xylophone' as 'ksylophone' seems like a step too far.

  80. Lucy Kemnitzer said,

    November 10, 2010 @ 1:42 am

    I don't see anything grammatically remarkable about "everyone ignores how words are spelt anymore." I think I might be more surprised by (but certainly not offended by) something like "I seem to be walking slowly anymore," but a phrase with a negative meaning (everybody ignoring being about the same as nobody paying attention to) just doesn't even stand out from the rest f the conversation.

    It just seems odd to call it a major clunker.

    [Let's be clear about this point (several people have mentioned it above): some American dialects use anymore in positive contexts (People are just so rude anymore) and others restrict it to negative contexts (People just aren't polite anymore). It's not a mistake to use anymore the way the young woman in the anecdote did, if you speak the first type of dialect. Millions do likewise. The second type of dialect is regarded as more standard (it matches what you'd find in most serious writing for publication, and also matches British English), but that's just a historical accident. —GKP]

  81. Zeno said,

    November 10, 2010 @ 3:03 am

    I've been puzzled at the return of the "t" in "often," which had been silent in my youth but today is often audible in people's speech. It's always struck me as an example of hypercorrectness, and I do not doubt that the young lady in the story says "oft-en" instead of "offen".

    Recently I was amused to hear a Christian radio talk-show host struggle with "Christendom." I'm sure he knew better than to pronounce the "t" in "Christmas," but he insisted on trying to do so in "Christendom." It sounded awful.

  82. C Thornett said,

    November 10, 2010 @ 3:39 am

    I have a private snigger when I hear a person who pronounces secretary as seketry denouncing the slovenly speech of younger people, Americans, the working class…anyone who isn't a middle class, educated RP speaker, in fact.

    RP in fact involves elisions that many 'lesser' accents don't make.

  83. Cialan said,

    November 10, 2010 @ 3:39 am

    I must admit that I've found myself differentiating between "then" and "than" in an attempt to try to keep others around me from losing the spelling distinction between them. I'd imagine that "then" and "than" are pronounced with different vowels when stressed but that both merge into a schwa when unstressed. If my guess is correct, I suppose I'm stressing them rather than just changing the vowel sound. Does anyone else have a comment or theory about the "then"/"than" merger?

    [There isn't any merger in most styles of speech and most dialects. Than is very often pronounced with schwa (it is a preposition, and mostly unstressed), but then is usually pronounced with a full vowel, the one you hear in get, regardless of whether it introduces a clause (Then we went to lunch) or acts like a substitute for a preposition phrase meaning "at that time" (He wanted it done right then). On top of that, it's extremely hard to construct a context where there could be an ambiguity, because then and than do such different jobs. So I am quite surprised you feel a need to over-enunciate the vowels to keep the two words from (as you fear) collapsing together. I think you might just be mistaken about how they sound in your own speech. —GKP]

  84. xyzzyva said,

    November 10, 2010 @ 3:48 am

    @onymous
    For xylophone, instead of my preferred /ˈzaɪlɵfoʊn/, or the difficult /ˈksaɪlɵfoʊn/, I often hear people go with /ɨɡˈzaɪlɵfoʊn/, which bugs me to no end. Other ‹x›-initial words get the same treatment, e(x)specially Xavier /ɨɡˈzeɪvjər/.

    It still throws me off that most European languages that maintain orthographical ‹ps› actually pronounce the /p/, though Spanish very oddly* does not.

    *That is, it's odd that Spanish keeps the null letter; it's not odd that they don't pronounce it, since that cluster would violate phonotactic constraints.

  85. michael farris said,

    November 10, 2010 @ 5:02 am

    A few years before my first linguistics class I had become interested in linguistics and at one stage was trying to reconcile the differences between what I was hearing people say and what dictionaries and typical IPA use said they say (the sources I had access to didn't countenance leveling /t/ and /d/ intervocalically among other things).

    At some point I thought "maybe they're right" and modified my speech in a couple of ways, consistently differentiating between rider and writer and winner and winter. Within a week or so at least three people asked if I was English. I gave up and went back to my old pronunciation.

    A few years later in an English composition class one student took open and obvious pride in her hyper-careful pronunciation and usage (including no contractions and an attempt to follow the shall/will rule). I thought she sounded unutterably odd but I understood where she was coming from. There was no diplomatic way to offer another point of view (if anything the teacher seemed to encourage her).

    Of course there are languages with a more harmonious fit between orthography and pronunciation. In Polish a 100% reading pronunciation (if it includes proper voicing assimilationh) will sound stuffy sometimes but in formal contexts it would be okay and and a 90 % (approx) reading pronunciation will always sound okay.

  86. the other Mark P said,

    November 10, 2010 @ 5:31 am

    My wife is sometimes accused of being English because she speaks clearly. In fact she has a fairly standard New Zealand accent ("caught" = "court", but certainly not "cot"). Apparently the speaking slowly and clearly over-rides the actual accent for many people.

  87. Julie said,

    November 10, 2010 @ 6:39 am

    My first linguistics class was taught by an instructor who spoke nearly unintelligible English, and had no clear concept of the native dialects of his California students.

    The second was taught by a Virginia native (wonderful teacher!) who used "positive anymore." It was the first time any of us had heard the phrase. That was more than 30 years ago. Now I hear it on the radio pretty regularly. I think it's spreading.

    A friend of mine had an Irish father and was given an English tutor (to make sure he wouldn't grow up Irish–in California!). Makes for an interesting blended accent.

  88. outeast said,

    November 10, 2010 @ 7:01 am

    Funny that a native speaker should have been so preoccupied with crushing the schwa from her diction. Back when I taught EFL I used to actively drill my students in the schwa – I found that overclarification of vowel sounds was often a significant factor in making an accent hard to understand.

    [Absolutely right. For me, African accents — pretty much all of them — are among the absolutely most difficult to understand, and it is noteworthy that speakers of African languages show a very strong tendency to restrict themselves entirely to the pure vowels [i e ɛ a ɔ o u] — absolutely no schwa. That plus a very different pattern of stress and intonation just about wrecks intelligibility for me. When a Congolese general or a Nigerian politician is talking over a phone line on the BBC World Service, I pretty much end up with not a clue about what they've said. —GKP]

  89. Jongseong Park said,

    November 10, 2010 @ 7:32 am

    From the anecdote as told by your correspondent, I only took away that the young woman was not reducing vowels that are normally reduced. There is no mention of her pronouncing normally silent consonant letters.

    [I have only my anonymous informant's description for any of it. But I imagine the young woman would be the kind of speaker who said Wed-nes-day (not wenzdi), and Chriss-t-mass (not crissmus), and streng-th-s or strenk-th-s (not strenks), and so on. —GKP]

    I wonder which strong vowels she uses for 'from', 'of', 'on', and 'been', given that in American English the 'spelling' pronunciations are not usually heard for these words even when they are stressed.

    On a general note, most people simply have never had the occasion to carefully examine their own pronunciations. No one actually pauses to draw up one's personal phonemic inventory and determine which sounds are used in which order for each word. Whatever internal representation we use in pronouncing the words we know eludes direct contemplation. By contrast, the spelling of a word is visible and reinforced every time we read or write that word, so it is much easier to recall.

    For the longest time, it didn't occur to me that 'any' was pronounced /eni/ in my own idiolect as in most varieties of English. I had to sound out 'any' and 'penny' in my mind to realize that they rhymed. I've been pronouncing it that way, expecting others to pronounce it that way, but due to the interference of spelling, it had never occurred that the vowel spellt 'a' was the same as the 'short e' sound. Before someone pointed this out, I'm not sure that I even noticed that 'any' had an irregular spelling, that the vowel I used for 'a' in 'any' was different from the vowels I usually associate with 'a'. Such is the primacy of orthographic representation in the mind.

    [Excellent points. It really is a disaster to take the letters in English words seriously, as opposed to listening to what gets said. You'd tend to pronounce any as Annie, and make mother rhyme with bother, and end thumb with a b, and so on for thousands of other mispronunciations. —GKP]

  90. Julie said,

    November 10, 2010 @ 8:23 am

    Okay, that explains why the instructor I had was unintelligible. He was, in fact, African. There were times when the entire class was looking around at each other in dismay, having no idea what he was talking about.

  91. Troy S. said,

    November 10, 2010 @ 8:45 am

    @Jeongson Park: You're right about people not being very introspective about their own ideolects. In college phonology course, I realized I don't even know the "correct" pronunciation of my own name. Is that initial consonant cluster palatalized or not? Maybe I should ask Mom…

  92. micdeniro said,

    November 10, 2010 @ 9:01 am

    Lawyers pronounce defendant the way they do so they won't spell it defendent when they submit briefs.

  93. ix said,

    November 10, 2010 @ 9:33 am

    As a foreign speaker, I can't really tell what the 'r' is doing in "the idea of it". Would it be pronounced similar to "the I dear of it"?

    The dutch, FWIW, speak okay English, but definitely have one of the strongest accents I know. Nothing compared to Indian people, but still. Which has got me wondering about India. I think African dialects are actually more understandable than Indian ones.

  94. marie-lucie said,

    November 10, 2010 @ 9:50 am

    Any like penny? I (living in Canada) don't know anyone who rhymes them, and neither of them rhymes with Annie.

  95. Bill Walderman said,

    November 10, 2010 @ 9:56 am

    ". . . the Shavian alphabet . . . has separate letters for the vowels in "cot"
    …and "caught". Several people who have the cot/caught merger have complained quite bitterly about this pair of letters, because it turns out that for the life of them they can't tell which one to use."

    Isn't this a good argument for traditional English spelling, however distant from any particular speaker's pronunciation it may be?

    "On a general note, most people simply have never had the occasion to carefully examine their own pronunciations."

    myl promised us a post on how people perceive their own pronunciation in vowel merger situations, as opposed to the objectively measurable characteristics of their speech . . .

  96. Nicholas Waller said,

    November 10, 2010 @ 10:24 am

    @ marie-lucie – I am pretty sure I (BrE) rhyme any and penny, though Annie would not rhyme either way… if it didn't sound like penny, any might be spoken more like "a knee". Run together and without the "k" being pronounced… But I am sure I pronounce Mary, Marry and merry differently.

    I went to the Tehran Book Fair a few times as a publisher's rep in the early 90s. There was a local student who came up to speak and who had a very good command of English in terms of vocabulary and grammar, and you could understand what he was saying. He was very pleased to find some native English speakers to demonstrate his skill to (discussing the then-relatively recent stuff about Salman Rushdie, for instance).

    He had presumably learned from some old and precise set of tapes, or some old recordings of someone with an odd accent, and I was trying to explain carefully to him – without crushing his pride or insulting him – that his diction, which completely lacked an Iranian accent, nonetheless sounded peculiar, artificial, over-elaborate and unusual, like a parody of 1930s English or something. I didn't know technical (ha) terms like schwa, but tried to suggest to him that to fit in to a normal conversation with English people he would have to rough it up a bit, knock the edges off, run words together, elide things, be more idiomatic. He sounded more alien (in an sf sense, as opposed to foreign) than someone with a thick Iranian accent and less vocabulary and less knowledge generally.

  97. Joe McVeigh said,

    November 10, 2010 @ 10:28 am

    @GKP – Thanks for extrapolating my comment. But Geoff, it is about pronouncing the letters, at least for your anonymous woman and for the Finns.
    "…because I'm more careful to pronounce words properly. It only sounds unusual because everyone simply ignores how words are spelled anymore."
    and
    "But as he listened to her some more, it became clear that she was indeed pronouncing many words precisely as they were spelled, letter by letter."
    and
    She differed from them simply in that she took great pains to pronounce things "properly, as they are written".
    That's what I was getting at. For second language learners, it can be about pronouncing the letters too. Most adults learn the sounds and writing of a second language at the same time, assuming that language has a writing system. It seems to me that this woman was trying to do the reverse of what Finns have done, i.e. match her speech to her writing system. She would see the Finnish speech system as perfect (or not "lazy") because of her backward belief about how speech systems should work.
    And because of the Finnish writing system, it still has to do with "saying every letter." When native Finnish speakers, as well as Finnish learners, come across a word that they don't know, they already know how it's pronounced. I would assume this woman believes she could do the same with English.
    Have I got it straight now? I'm trying. I heard this is Language Log.

    [OK, thank you for trying. But it is not about pronouncing the letters. It's about the letters matching very nicely the sounds in the words as they are pronounced. It's spelling that can be more or less faithful to speech, not the other way round. —GKP]

  98. Faldone said,

    November 10, 2010 @ 10:31 am

    To add another datum to the "positive anymore" saga: I was born and raised in Chicago and spent significant amounts of time in the Navy, Boston, Flagstaff AZ, and southern California before settling in Southern Tier/Finger Lakes New York. My memory has it that I never heard the positive anymore until I got to New York state. I have a fairly high level of confidence in this memory since I remember being somewhat amused by the usage.

  99. John F said,

    November 10, 2010 @ 10:39 am

    Pronunciation is to language, as fashion and identity are to clothing. Things change; and what's acceptable and normal in one place is silly and/or possibly obnoxious elsewhere. The mappings between sounds and letters are incredibly fluid even within one accent and there are even inconsistencies in a syllabic language like Japanese (ha/wa; desu->dess/dess(-uh)).

    By way of a cultural analog, at one time shaved heads, Doc Martens and turned up jeans with braces were a sign of solidarity between working class blacks and whites. Nowadays you hear of a skin head and you think neo-nazi.

    I do get very annoyed at BBC presenters, though, who leave out the trailing l and r. What is "Ah Q'aeda"?

  100. BReed said,

    November 10, 2010 @ 10:40 am

    Without a schwa, I'd find it much more difficult to proounce the terminal syllable in "pretentious." But then, I'm from Montana.

  101. GeorgeW said,

    November 10, 2010 @ 10:48 am

    @Bill Walderman:

    I think good arguments can be advanced for spelling reform (i.e. more phonetic) as well as for traditional spelling.

    As you say, our orthography preserves some lexical distinctions like 'cot-caught' (where they are merged). Also, our orthography maintains transparent relationships between words such as 'sane' and 'sanity.'

    On the other hand, we do expend a lot of educational energy teaching spelling where, in a shallower system, the time might be more productively spent elsewhere.

    Maybe, smarter spell checkers would be the answer so they would sense the appropriate form for homophones like 'right – rite – wright – write.'

  102. The Ridger said,

    November 10, 2010 @ 10:59 am

    @ Marie Lucie: Any like penny? I (living in Canada) don't know anyone who rhymes them, and neither of them rhymes with Annie. I (living in Maryland and having grown up in Tennessee) rhyme them, though not with Annie, though I know many who do rhyme "any" and "Annie".

  103. Dan T. said,

    November 10, 2010 @ 11:22 am

    I rhyme "any" with "penny", and can vaguely recall some nursery rhyme that does this too, so it's certainly not unique to me.

    Rhyme set 1: any, penny, Jenny, Kenny
    Rhyme set 2: Annie, Danny, fanny, uncanny

  104. I.D. Mercer said,

    November 10, 2010 @ 11:40 am

    I grew up in Western Canada. As a child, I rhymed "any" and "many" with "mini" and "tinny", and I remember hearing people who weren't in my household who did as well. But then I noticed that most speakers of North American English rhyme them with "penny", so I switched.

    (As the above shows, I don't have the pen/pin merger.)

  105. Ray Girvan said,

    November 10, 2010 @ 12:06 pm

    @ ix: "the idea of it". Would it be pronounced similar to "the I dear of it"?

    Yes. In the UK, it's said that way by both rhotic speakers (i.e. mostly in the south-west) and non-rhotic ones with r-intrusion.

  106. Jongseong Park said,

    November 10, 2010 @ 12:07 pm

    To clarify, I was talking about the strong form of 'any'. At least for me, introspection is even trickier when it comes to weak forms. I would guess that the first vowel in 'any' is more readily reduced than the corresponding vowel in 'penny', which would hardly ever be reduced to a schwa except in cases like 'halfpenny'.

    If I must guess, those who rhyme 'any' in its strong form with neither 'penny' nor 'Annie' might be influenced by this difference in propensity to vowel reduction.

    It is impossible to talk in absolutes about vowel reduction in English. The way I picture it mentally, the realization of reduced vowels corresponding to different underlying strong vowels may be thought of as probability clouds that largely overlap but are not identical. The first vowel in 'photography' would be pictured as a cloud that overlaps largely with the schwa cloud of 'about', but there would be one end of the cloud representing realization which wouldn't generally surface as the schwa in 'about'.

  107. Michael said,

    November 10, 2010 @ 12:16 pm

    I agree that the woman in question sounds like an irritating goof. But I will admit that I recently heard a podcast where the reader (a Brit with a hint of RP), pronounced the th in clothes, a subtle thing I have since heard in other UK speakers, and I quite like the sound of it. Like most native English speakers (I'm Canadian), I say /cloz/, but I have thrown the th in a few times recently just for kicks. It certainly doesn't draw any attention.

  108. Michael said,

    November 10, 2010 @ 12:18 pm

    @ Marie-Lucie
    I've lived in Canada most of my life, and until this thread had no idea that any and penny didn't rhyme. Haven't heard anyone here who doesn't rhyme them.

  109. Ken Brown said,

    November 10, 2010 @ 12:34 pm

    the other Mark P said: "My wife is sometimes accused of being English because she speaks clearly. In fact she has a fairly standard New Zealand accent ("caught" = "court", but certainly not "cot"). "

    "caught" = "court" is normal in most English English as well, including RP.

    dw said: " British actors who go to Hollywood have tended to the RP end of the spectrum, at least until recently (Michael Caine is, I suppose, the most obvious exception)."

    Hmmm…. not so sure… Stan Laurel, Charlie Chaplin, Cary Grant, Robert Newton, Charles Laughton, Richard Burton, Sean Connery? They didn't speak RP and often acted in versions of their natural accent. Connery almost always, he made his east-coast Scots serve for RP, various American, and Irish accents almost without modification. Cary Grant's famous "mid-Atlantic" accent often sounds nearer to Bristol than to Baltimore to me.

    And actors like Anthony Hopkins, Bob Hoskins, Jude Law, Gary Oldman, Catherine Zera-Jones, Tim Roth, don't have RP accents in real life – although they often seem to act in RP or American voices – so anyone who hears them in interviews ought to be aware of their non-RP accents. (And yes, these nearly all seem to be men. I wonder why?)

    And any American born since the 1930s is likely to be familiar with the accents of British pop and rock singers – the Beatles at any rate, and quite likely the Rolling Stones, the Who, the Bee-Gees (who sing in high-pitched Manchester accents) and who knows how many since.

  110. Terry Collmann said,

    November 10, 2010 @ 12:40 pm

    ix: "The dutch, FWIW, speak okay English, but definitely have one of the strongest accents I know." Really? Most BrE native speakers, I think, would say Dutch people have pretty unaccented English, the only really obvious give-away being a slight "pebble in the mouth" quality about the vowels and the pronunciation of "st-" as "scht-".

  111. Leo said,

    November 10, 2010 @ 1:13 pm

    The Dutch, along with the Scandinavians, are very well known for their excellent English. It's one of the things that we in England find so amazing and frustrating – you spend months or years trying your hardest to acquire even an elementary grasp of a foreign language, then you go to Amsterdam and every waiter and shopkeeper is tetralingual.

    There certainly is a distinctive Dutch accent in English though – ask Austin Powers.

    Re. African accents – certainly they are often hard for Brits to understand, but isn't it the case that many anglophones in Africa are not really "native" speakers? It would be unfair to compare such a person's accent with a Briton's or American's.

  112. I.D. Mercer said,

    November 10, 2010 @ 1:18 pm

    For a Dutch accent when speaking English, listen to the guy on the ING Direct commercials.

  113. Thomas Thurman said,

    November 10, 2010 @ 1:20 pm

    the corresponding vowel in 'penny', which would hardly ever be reduced to a schwa except in cases like 'halfpenny'.

    Unless you're like me and pronounce "halfpenny" without a vowel there at all, as /ˈheɪpni/. Not that I've had much of a reason to say it since 1984.

  114. Coby Lubliner said,

    November 10, 2010 @ 1:33 pm

    micdeniro: Lawyers pronounce defendant the way they do so they won't spell it defendent when they submit briefs.

    I used to fantasize that Christopher Darden (from whom I first heard this pronunciation) would say to Marcia Clark, "That's a lovely pen-dant you're wearing."

    My point was that vowel reduction in English is, as other commenters have noted, register-dependent, unlike those languages (e.g. Russian, Bulgarian, Catalan, Portuguese) in which it is a fixed aspect of the language. The presence or absence of this aspect may be the chief difference in pairs of languages (Bulgarian and Macedonian, Portuguese and Galician) that might, under other linguistic criteria, be regarded as variants of the same language.

    A good brief discussion of vowel reduction in English can be found in (where else?) Wikipedia.

  115. GeorgeW said,

    November 10, 2010 @ 3:13 pm

    Can anyone point to a reasonably accessible source that gives a good account of the British rhotic accent, r-intrusion and the like?

    I am unclear on the rules for 'poor' > /pua/, yet 'idea' > /idear/. When is the /r/ deleted and when is it inserted?

  116. Lane Greene said,

    November 10, 2010 @ 3:23 pm

    I for one am not as kind as David Crystal, and think it's hilarious to imagine this woman being forced to tears by the realization of her inane prejudice against intrusive-r for what it was.

  117. Thomas Thurman said,

    November 10, 2010 @ 3:34 pm

    GeorgeW: there is no "r" sound in "idea" in RP as such, nor in "of", but when two vowels in separate words are run together back to back, as in "idea of", /r/ appears between them.

  118. GeorgeW said,

    November 10, 2010 @ 3:38 pm

    @Thomas: Thanks.

    So, an /r/ is inserted when a word ending in a vowel is immediately followed by a word beginning with a vowel?

    Where is /r/ deleted? In all environments?

  119. Alces said,

    November 10, 2010 @ 3:43 pm

    @GeorgeW: It depends on context. Generally non-rhotic accents delete coda /r/, i.e. before a consonant or at the end of a word. No words actually have a coda /r/ in the phonology. But when a word ends in a vowel other than /i/, /u/ or a diphthong, and the following words begins with a vowel, /r/ is inserted between as sandhi, to prevent a vowel cluster from occuring.

    The reason for this is that historically, /r/ was deleted in coda position. So 'dear' on its own would become /diə/, but when a vowel followed to start the next word, it stayed /diər/. But maintaining this rule requires that words still have to contain /r/s in their underlying phonology–otherwise why can you say '/hiə/+/r/ I am' but not 'the /aɪdiə/ of it'?. To solve this, /r/s were inserted after any word that had a vowel that could be left by the loss of a coda /r/ at the end, even if they had never had an etymological /r/.

  120. GeorgeW said,

    November 10, 2010 @ 3:58 pm

    Alces: Thanks to you as well.

    So, this differs from the Bostonian /r/ insertion? I am old enough to remember JFK's 'Cuba' /kuber/. I am quite sure that he inserted the /r/ with no following word.

    Also, some Southern AmE. dialects insert /r/ word internally like 'Chicago' > /chicargo/, Washington > /warshingon/.

    I guess these operate under different rules.

  121. Leo said,

    November 10, 2010 @ 4:21 pm

    Sometimes, when speakers who are naturally non-rhotic try to adopt a rhotic accent, they make mistakes and insert /r/ where it would never occur in the target accent.

    For example, Ian Curtis (from Greater Manchester) adopts rhoticity in "Love Will Tear Us Apart" – he correctly includes the /r/ in "apart", but also inserts it into "flawed", singing flɔrd, as though it were "floored". (In the line "Is my timing that flawed?")

  122. Nathan Myers said,

    November 10, 2010 @ 4:31 pm

    The creole called "Hawaiian Pidgin" entirely omits schwa from its English adoptions. Locals parodying tourists who try to speak pidgin overemphasize the improper schwas for laughs.

    The only hypercorrection I hear regularly that grates is "off-ten", which has occurred on NPR entirely too much lately. I first heard positive "anymore" in 1981 from a Pennsylvanian. It still sounds peculiar.

  123. Bill Walderman said,

    November 10, 2010 @ 4:41 pm

    "Also, some Southern AmE. dialects insert /r/ word internally like 'Chicago' > /chicargo/, Washington > /warshingon/."

    I grew up on the upper west side of Manhattan (less fashionable or decidedly unfashionable then), but I'm rhotic and my mother was from Oklahoma. In fourth grade (1955-6) I was humiliated by my teacher (Mrs. Oxley) for spelling "wash" the way I pronounced it: "warsh."

  124. GeorgeW said,

    November 10, 2010 @ 5:06 pm

    @ Bill Walderman: I grew up in Florida (it was more Southern at that time) where some speakers inserted the /r/. I can recall being very confused when learning to spell. Where in the hell is the in 'Chicargo?'

    But, I think this only occurred in certain words (like 'warsh'). I don't think there was systematic /r/ insertion according to any phonological rule.

  125. AlexK said,

    November 10, 2010 @ 6:11 pm

    A point can be made about this if you consider French. In French, many words have endings such as -au, -aux, -eau and even the infamous -eault. Yet all of them are pronounced as [o], and never will you find an L1 speaker pronouncing them as diphtongs or triphtongs, let alone prnouncing the -x or -lt.
    A popular hypothesis (although I don't know how sound it is) claims that spelling was intentionally made complicated by the medieval literate classes to protect their monopoly over reading and writing.
    This neatly illustrates the fact that spelling and pronunciation are only connected to a limited extent, and that spelling did not precede pronunciation.

  126. Bob Ladd said,

    November 10, 2010 @ 6:57 pm

    @George W.: It's possible that JFK said Cuba with a final /r/ even when not followed by a vowel, but it's also possible that rhotic Americans heard Cuba -r- is and just noticed the /r/ on the end of Cuba without understanding that there was a system to when the /r/ did and did not appear.

    Also, it's possible that JFK was trying to make his accent less Bostonian and more General American, and not always doing it right. Any non-rhotic speaker trying to put /r/s into their speech is going to get them in the wrong place sometimes, for reasons discussed by various people above. My father grew up non-rhotic on Cape Cod (in the 1920s and 30s) and deliberately became rhotic when he went to college. But he still occasionally screwed up – at one point in my childhood we lived on a street called Hawthorn Avenue, which gave him no end of trouble. In his native speech both syllables of Hawthorn had the same vowel, and to pronounce it in an authentic rhotic way he had to put an /r/ after the second vowel but not the first one. He didn't always manage to get it in the right place.

    Something similar may have happened to JFK when he was talking about Cuba.

  127. Dw said,

    November 10, 2010 @ 7:17 pm

    @GeorgeW

    The Wikipedia article is decent:

    http://en.wikipedia.org/wiki/Linking_and_intrusive_R

  128. Bill Walderman said,

    November 10, 2010 @ 8:01 pm

    "this only occurred in certain words (like 'warsh'). I don't think there was systematic /r/ insertion according to any phonological rule."

    That's my impression–it's a lexical, not a phonological, variant, although I'm at a loss to think of a rhyme for "wash" that could be subject to a phonological rule. But I also grew up thinking that the capital of the US (and the city of my birth and the city where I'm writing this) was Warshington, DC.

  129. Ellen K. said,

    November 10, 2010 @ 8:10 pm

    Regarding the use of "anyone" in the quote in the original post, for me, it's a matter of register. It sounds colloquial, not something you'd say in careful speech. Thus it seems a mismatch with her reported way of pronouncing English. Like, not that it's not good English, just that it's not something you expect in the kind of English she's speaking.

    That said, I didn't actually notice it when I first read the post. But then, it was a conversation at a party, and I was reading it in my own accent (middle part of the US), so no reason it should have sounded exceptional.

  130. GeorgeW said,

    November 10, 2010 @ 9:06 pm

    @Dw: Thanks. For some reason, I never think of Wikipedia unless the issue is rather trivial. It is amazing what is available there.

  131. xyzzyva said,

    November 10, 2010 @ 9:18 pm

    @Bill Walderman
    Was your warsh /'wɑɹʃ/ or /ˈwɔɹʃ/?
    I've always thought /ˈwɑɹʃ/ would make more sense as a rhoticized /ˈwɑːʃ/, but have only heard /ˈwɔɹʃ/, which I assume is by analogy with war /ˈwɔɹ/, ward /ˈwɔɹd/, etc.

  132. James Kabala said,

    November 10, 2010 @ 9:25 pm

    Ken Brown: Few Americans know what RP is, and it seems that most Americans find any British accent (except a really strong Cockney or rural accent) to be upper-class-sounding even if it would never be regarded as such in Britain itself.

  133. michael farris said,

    November 11, 2010 @ 3:59 am

    I grew up in smalltown and rural Florida too (with southern and cowboy influence). I do remember the worsh example though I don't think I ever said it that way. On the other hand I did grow up saying torlet for toilet though I don't know how widespread that was.

    @James Kabala: I agree. Most Americans have extreme difficulty distinguishing accents from the British Isles, roughly they can disitnguish

    1 stereotypical Scottish
    2 stereotypical Irish
    3 stereotypical cockney (is it still a going concern in the UK?)
    4 all purposh "British"

    Anything that isn't obviously 1,2 or 3 is liable to be perceived as 4.

  134. Leo said,

    November 11, 2010 @ 5:54 am

    Michael Farris – stereotypical cockney is on the way out. But the word "cockney" is still colloquially applied to the speech of young Londoners today, which is quite different. Try the search term "Multicultural London English".

    Historically, "cockneys" were working class and from East London. But people from other parts of the country sometimes mistakenly think "cockney" is just a synonym for "Londoner", which doesn't go down well – John Major is not a cockney. Of course it works both ways – people from Middlesbrough do not appreciate being called geordies.

  135. ajay said,

    November 11, 2010 @ 9:08 am

    people from Middlesbrough do not appreciate being called geordies.

    Indeed not. The correct epithet is "Smogmonster".

  136. James Kabala said,

    November 11, 2010 @ 9:14 am

    I associate "turlet" with Archie Bunker. (Is "torlet" the same? – the Bunker version included a vowel change as well as an r.)

  137. Maureen said,

    November 11, 2010 @ 9:17 am

    I think a lot of Americans recognize the different levels of UK class and region differences, whether or not they can accurately identify their source, just as a lot of casual American anime watchers can tell when somebody's speaking in a different regional accent than what they're used to hearing.

    But it's hard to imitate such accents if you don't hear them very often; and Americans have no particular reason to feel that Bristol is less attractive or classy of an accent than Oxford. So practically speaking, there's no difference for most of us. Just all different kinds of pretty. :)

    Of course, the flipside is that for some Americans, all UK accents are equally hard to understand. I know people who watch Doctor Who with the captions on. (And honestly, as fast as people talk, there's good reason for it. Watching the old Doctor Who used to be relaxing late night sf viewing. Now it's something you have to be rested up for.)

  138. Ben Hemmens said,

    November 11, 2010 @ 9:34 am

    For a language so used to a complex relationship between spelling and pronunciation, it occurs to me that an exception is often made for family names in the US, where often enough either the pronunciation (Gallagher, Berkeley) or the spelling changes (e.g. variations on -itsch, -itz, -ich or -itzky) to make things more phonetic. The slavic -ic / itz /itzky names are rarely modified in German-speaking areas, so that here you have to know whether your individual -ic is really an -ich, -itch or -itz, your individual -ici is an -itzy or an -ich, and your individual -icki is an -ich or an -itzky.

    Maybe it's just a matter of era: a hundred years ago, a Gerschovitz might not only have become Gershwin in the US but possibly, let's say, Kürschner in Vienna, whereas nowadays there are many recent immigrants from the former Yugoslavia. But Gerschovitz to Gershwin is about assimilation, not about making the spelling phonetic.

    So it strikes me as interesting if the adaptation of names for phonetic reasons is more prevalent in English, the language with the generally less regular pronunciation, than in German.

  139. michael farris said,

    November 11, 2010 @ 10:40 am

    @James Kabala, no the first syllable rhymes with more, oar and door and doesn't sound like Archie Bunker (who did say turlet) at all.

  140. KevinM said,

    November 11, 2010 @ 5:24 pm

    From a lawyer & musician:
    Re: Lawyers' pronunciation. Somebody wrote "The only other instance of lawyers' pronouncing a common word with an idiosyncratic stress is the pronunciation "pre-CEDE-ent" in the fixed phrase "condition precedent."" Actually, this is only an apparent variant pronunciation of a common word. "Precedent" (the noun) is indeed a common word. "Precedent" (used as a postpositive adjective) is not. Condition precedent, I believe, is one of many legal fixed phrases left over from Norman French (body politic, court martial, notary public).
    As for the other lawyer example (pronouncing defend-ant as if the last syllable were the insect): Never underestimate the lawyer's concern that his or her words be transcribed accurately by a court stenographer. I suspect many exaggerated or hypercorrect legal pronunciations originated that way.
    Classical singers, above. For sure, a schwa can be sung. It can even sound musical, but it helps if the following consonant is a singable one (n, say, as opposed to k). Thus, in the Ninth Symphony example you gave, your conductor had you sing Mensch'n. But as a singer you'd draw out the natural short e in "prophet."

  141. Peter G. Howland said,

    November 11, 2010 @ 6:10 pm

    @Thomas Thurman, and others…

    After studying the various pronunciation comments posted here, I decided to attempt a test of my own locutions for ‘cot’ and ‘caught’, because to my instinctive internal ear they seemed quite different. This direct contemplation of one’s own pronunciations, as Jongseong Park so eloquently pointed out in his Nov. 10 @ 7.32 a.m. comments, is hard to do. Nevertheless, over the next several hours, as I stumbled around the house attending to various minor chores, I made up a number of short silly sentences containing both words and uttered them aloud while attempting to listen carefully to the sounds I produced.

    And sure enough, when “He was caught with a stolen cot” popped out I heard the difference…

    caught = “cawt”…with the ‘aw’ as in “awful”
    cot = “caht”…with the ‘ah’ as in “otter”

    …and there wasn’t an ‘r’ in there anywhere! My ‘caught’ is not ‘court’ as it is in John Thayer Jensen’s New Zealand experience. But please, I intend no pretension of “correctness” in this observation. It’s merely a demonstration of an instance in my own idiolect (AmE of the western US).

    To those of you fluent in IPA transcription, I apologize for my primitive phonetic representations here. Maybe some day I’ll take a stab at learning these mysterious symbols and their meanings, but at my age I’m not sure there’s enough time left for such a monumental task!

  142. Rhodent said,

    November 11, 2010 @ 7:05 pm

    @~flow:

    The issue your Taiwanese friend was having wasn't an inability to answer a simple phonological question; it was a matter of not knowing which romanization was being used. The [t] sound will be written with a 'd' in pinyin romanization but with a 't' in Wade-Giles.

  143. Peter G. Howland said,

    November 11, 2010 @ 7:40 pm

    @Troy S. – In reference to your ‘racked / wracked’ question, there is no “eggcorn” here. They are simply homophones; different words, different meanings, same pronunciation.
    In my personal collection, “Homophones: Over 2,740 Opportunities to Spell Things Wrong in American English” I offer the following examples:

    – He wracked his brain to determine if the billiard balls were racked correctly.

    – The wrack and ruin of the pillagers left his bed destroyed. His rack was ruined.

    You can take it from here…it’s in B-flat with an augmented relative minor at the bridge.

  144. Max Pinton said,

    November 11, 2010 @ 10:19 pm

    Peter G. Howland, that's the tricky thing about "as in" examples: in my accent (AmE, Pacific Northwest), the "aw" in "awful" and the "ah" in "otter" are the same. I think the IPA is ɑ.

    Speaking of Shavian, I ran into the same problem with Gregg Shorthand, which makes vowel distinctions that I don't. That was a real stumbling block for me.

  145. John Cowan said,

    November 12, 2010 @ 3:38 am

    David Fried: In addition to defendant rhyming with ant, there is also juror rhyming with more, at least in the New York City courts. And it isn't just lawyers who use it: court officers do too. I have sat many a day in the ju-roar room.

    Troy S.: In most cases, wrack should etymologically be rack. In particular, when you are racked with guilt or racking your brain, you are metaphorically on or employing the rack, the historical instrument of torture. There is also a separate word wrack, a variant of wreck, which gives us wrack and ruin, but rack and ruin is a plausible eggcorn and does show up. The result is that many people prefix all racks with a w just for insurance.

    GKP: In Ireland, Annie and any are indeed homonyms, leaving would-be spelling reformers who want to be fair to all accents in doubt whether to respell it enny, inny, anny. (I prefer enny, since the current spelling is obviously broken, suggesting as it does a rhyme with zany, and enny is far and away the majority pronunciation.)

    GeorgeW: I have listened to Kennedy's speech very carefully, and I do not believe that he says "Cuber" except in the expected cirumstance of a vowel following it. So the /r/-insertion rules are the same in non-rhotic Boston as in non-rhotic London. (There is no /r/-insertion rule in AAVE, however.)

    As for warsh and Warshington, they are lexically specific in the dialects that use them, not part of a general /r/-insertion pattern. However, it's true that rhotic speakers who pick up learnèd words from non-rhotics sometimes insert an /r/ when they should not: Jim Scobbie, the Scottish (and rhotic) phonologist, reports that he says Chicargo, inaurgural, unaurthorised and idear. In North America, on the other hand,idear is also lexically specific, essentially a separate form of the word idea, since many completely rhotic speakers use it, including at least two politicians who have spoken of new idears quite consistently.

  146. PaulB said,

    November 12, 2010 @ 7:55 am

    marie-lucie: mentor is not the best example because the -or is not etymologically a suffix – the word comes from Μέντωρ, an older friend of Odysseus.

    Michael Farris: the accent commonly heard nowadays in the South-East of England is known as "Estuary English" (named for the Thames Estuary). It's somewhere between cockney and RP.

    Ken Brown: popular singers often use different pronunciations for singing than for speaking, partly for the same reasons as classical singers, and partly in imitation. Mick Jagger in particular often sings in something (to my ears) like a southern US accent. (His usual spoken accent is "Mockney".)

    I'm surprised that anyone can detect a Manchester accent in the Bee-Gees' singing – to me it sounds more like Neighbours than Coronation Street.

  147. Uly said,

    November 12, 2010 @ 12:59 pm

    I rhyme "any" with "penny", and can vaguely recall some nursery rhyme that does this too, so it's certainly not unique to me.

    Simple Simon
    Met a pieman
    Going to the fair.
    Said Simple Simon
    To the pieman
    "Let me taste your ware."
    Said the pieman
    To Simple Simon
    "First, let's see your penny."
    Said Simple Simon
    To the pieman
    "I have not got any."

  148. Julie said,

    November 15, 2010 @ 6:55 am

    @Peter G. Howland:
    If you live in the western US, you have often heard "cot" and "caught" pronounced alike, even if you don't do so yourself. Both pronunciations can be heard all over the West. I pronounce them alike; my husband doesn't. Both from California.

    Oh, and like I.D. Mercer, I don't rhyme "any" with "penny," although he does. I rhyme "any" with "tinny," and always had a little trouble with Simple Simon over that issue.

    My fully-rhotic father sometimes says "idear," but I think it's an intentional mockery.

  149. Peter G. Howland said,

    November 15, 2010 @ 5:10 pm

    @ Julie – Your observation is absolutely right about “cot & caught” seeming to be pronounced the same throughout the western US. At a recent family gathering here in Arizona of a group ranging in age from 12 to umpty-ump and collectively from the southern California area, I had everyone recite the “He was caught with a stolen cot,” phrase out of earshot of one another. And sure enough, all fourteen of them pronounced cot/caught in nearly the same way. With the exception of Jade, a very bright 7th grade girl whose grandmother is fluent in native Italian and with whom she has frequent contact. Jade simply wanted to know why someone would steal a stupid cot (!)

    In general I attribute this pronunciation phenomena to the inherent lack of concern for enunciation so typical of SoCal folks where speech precision just isn’t all that important. I mean, sure, we all understand each other, but any more, not only am I always the oldest person in the room, I’m also the one who always talks funny.

  150. Geoffrey K. Pullum said,

    November 15, 2010 @ 6:09 pm

    I noticed there were 149 comments on "Pronouncing it by the book" at this point. Could someone log in and contribute one more, so I could hit the round number of 150 one time? Oh, wait a minute, never mind; I just realized that this comment will do it. I've got my 150. I've rung the bell.

  151. xyzzyva said,

    November 15, 2010 @ 8:44 pm

    Not to un-ring GKP's bell, but I felt the need to correct my earlier posting:
    Wash is diaphonemically /ˈwɒʃ/, not /ˈwɑːʃ/, which makes the r-inserted form /ˈwɔɹʃ/ seem much more sensible than /ˈwɑɹʃ/.

    This cot-caught/father-bother merged speaker has yet again made the error of relying on the orthography to keep track of the distinction.

  152. John Cowan said,

    November 17, 2010 @ 11:28 am

    Peter G. Howland: Unfortunately, the "SoCal" explanation won't warsh. The cot-caught merger affects not only, as you say, most of the American West (except for a few pockets here and there and the considerable number of older speakers who predate it), but also the whole of Canada, a place not known for its "lack of concern for enunciation".

    And believe me, people like me think "any more" in positive sentences just as strange as you find the merger to be.

  153. Graham Asher said,

    November 17, 2010 @ 5:05 pm

    It should be

    Said the pieman
    To Simple Simon
    "Show me first your penny."
    Said Simple Simon
    To the pieman
    "Alas, I have not any."

    In particular, the last line must be corrected from the lame "I have not got any.". It doesn't scan otherwise.

  154. Graham Asher said,

    November 17, 2010 @ 5:08 pm

    Yes, I know that was pedantic. But I suppose I belong to the last generation of English children brought up on traditional nursery rhymes. Quite a lot about the money of the time, too… 'you owe me three farthings, say the bells of St. Martins…'. Funny to have to explain to one's children what a farthing was.

  155. abby said,

    November 17, 2010 @ 11:24 pm

    I've been criticized by my sisters for saying the word "worship" as it is spelled, "Wor" and "ship" when they insist it's pronounced more like "wurship." I'm still not convinced who's correct.
    Unrelated to this, I've had an alarming amount of people inquire after my British accent, an accent which I think would be hard to develop living in Western Washington (the state, not capitol). I really don't understand this, but it keeps happening.
    Unrelated to both of these ideas, there is the sad fact that whenever I see the word "daschund" (is that even the right spelling?) I think "dash-hund" not, as I have been informed is the correct pronunciation, "doxon."

  156. Darrell said,

    November 20, 2010 @ 3:44 pm

    When I was on a grand jury in Kings County, New York – that is, Brooklyn – I was shocked to hear every single assistant district attorney pronouncing "petit larceny" as if it began with the words "pet it" rather than "petty". I became so unsure of my pronunciation that I had to look it up. It never occurred to me that they might be doing this for the sake of the transcription.

    @Abby: Dachs-Hund, German for "badger-dog." The problem is the orthographic cluster CHS, pronounced like an X, running together with the H of Hund.

  157. abby said,

    November 24, 2010 @ 4:10 am

    @Darrell: the petit thing is horrible.
    And I know that the dachs-hund is technically right, but it just shocked me so much when I found out in my adolescence that the pronunciation which I had assumed was a lie.

  158. Peter G. Howland said,

    November 24, 2010 @ 5:12 am

    Yeah, yeah, I know…I’ve already run over the brim with my commenting privileges for this post, but I’ve got just one last thing. Really.

    My use of the phrase “but any more” (note two-word form) in my 11/15@5:10 was not perceived by me as being imbedded in either a positive or negative construct. It simply referred to relative time and was what automatically dribbled out of my dancing fingertips as a concise, informal way of saying, "things weren’t always like this, but in the inexorable march of time I have come to the realization that any more I am…blah, blah, blah."

    I don’t mind being snarked at for obvious foe paws (Eggcorn alert! Eggcorn alert!) but this is not, in my view, a grammatical “clunker”.
    I have no qualifications for a linguistic defense of this particular use of the phrase, yet as a cultural norm of my internal idiolect I don’t think its use requires an apology. I am, however, sorry if it makes your teeth grate, your shoulders shudder, or your spleen get all wiggly with the creeping fantods.
    ….
    And to whom the holiday obtains, y’all have a nice Thanksgiving. I just hope you’re not the one who gets "cawt" having to sleep on the raggedy-ass nearly homophonous "caht" just to accommodate your stay-over leeching relatives.

  159. Alon Lischinsky said,

    November 25, 2010 @ 6:12 am

    @xyzzyva: Spanish has regarded mute ⟨p⟩ in the ⟨ps⟩ cluster as optional for quite a while now. The Academy regularly changes its collective mind about etymological spellings; Greek ⟨th⟩ was done away with in the 17th century but reinstated in the 1720s (only to be eliminated again a century later), and simplification of ⟨ps⟩ to ⟨s⟩ was recommended in the 19th century, only for the preference to be reversed in the 1990s.

    The situation is even worse with medial ⟨bs⟩, where simplification is recommended for etymological ⟨obs⟩ (thus oscuro is preferred to obscuro, although the latter is still allowed), but adamantly forbidden for etymological ⟨abs⟩ (*asoluto is not recognised as a valid alternative to absoluto), even though the same phonotactic rules led to /b/-elision in the spoken language.

    A non-inconsiderable number of speakers actually pronounce initial /ps/ in careful or formal speech, even though the cluster violates the normal phonotactic constraints; the same speakers also tend to use /ks/ for initial ⟨x⟩. I've encountered the phenomenon more frequently in Latin America (where ⟨x⟩ is very rarely voiced as /(g)z/) than in the Peninsula.

  160. Frans said,

    November 29, 2010 @ 3:52 pm

    @Quintesse the "g" in light? In German we have "Licht" ("c" instead of "g") but the "c" is not pronounced separately, it is part of our "ch" sound…I don't know about Dutch (I read an Asterix and Obelix comic book once that was written in Dutch and understood a good portion, so it's not DREADFULLY far from German).

    I would transcribe German licht as /li:çt/, while I would transcribe Dutch licht as /lɪxt/. I hear that English light can also be pronounced as /lɪxt/ in Scots.

    Never been tempted to pronounce "k" in "kn", although that is natural to a German, but HAVE pronounced the "p" in psychology, the "k" sound in Xerox, and the "t" in "tzar".

    I used to pronounce a /p/ in psychology because that's the way it is pronounced in Dutch.

  161. Kate Gladstone said,

    February 19, 2011 @ 1:40 am

    This phenomenon needs a name. For a reason discussed in another linguistics blog's thread on the matter — http://david-crystal.blogspot.com/2007/06/on-imaginary-pronunciations.html — I suggest that we call it "Holofernizing."

  162. Tom Paine said,

    March 4, 2011 @ 10:55 am

    I'm surprised that nobody else here shared my immediate thought that the young woman in question combined two classic signs of high-functioning autism or Asperger's syndrome: an exotic fixation on a topic that seems natural to the subject but bizarre to others, and speech that is (somewhat condescendingly termed in the official literature) "stilted" or "pedantic."

  163. Rose Eneri said,

    September 27, 2011 @ 11:29 am

    I recall the Seinfeld episode in which Elaine refers to the Eagle's song Witchy Woman. Jerry is puzzled and cannot place the song until Elaine sings it, "Ooo ooo, witchay woman". To which Jerry says, "Oh! You mean WitCHAY Woman".

    My husband is from New York City and has a sister named Rona. He habitually says things like, "Remind me later to call Roner". And when I met my neighbor, who is from Boston, he had to say his name (Cahl) about 6 times before I realized it was Carl.

    @GeorgeW – I too find it frustrating when examples are not provided to show the distinction made by speakers who no not have a merger that I do have.

  164. Frank Workman said,

    June 29, 2012 @ 10:15 am

    I did not take time to read of the comments, so mine may be a repeat. If so, please accept my apology. My comment comes in the form of a question: Why will many pronounce the indefinite article, "a" in everyday speech, "uh," (as it should be), then, pronounce it, "aye," when they are reading?

    My reason for logging on to this site this morning was, on one of our local channels this morning, both, the anchor man and the weather man, while reading the text on their teleprompters, used the latter pronunciation, and they sounded like students in "aye" third grade reading class.

    I know that some linguists say that the article, "a" is pronounced, "aye" when emphasis is needed to express the idea that only one (and no more) subject complements are meant. Or, has the English language mutated to the place of accepting either pronunciation? If so, I am more out of touch than I realized

  165. Frank Workman said,

    June 29, 2012 @ 10:24 am

    Uh, sorry! I failed to proof my comment….Paragraph 1, line 1, the word, "all" should be inserted between The words, "read," and "of."

    In paragraph 2, line 2, the comma should be omitted after the word, "both."

RSS feed for comments on this post