Pin or pen?

« previous post | next post »

Jim T, commenting this morning on a post from back in June:

I currently work in Chicago but I'm from South Texas. My boss seems to get a real kick out of my pronunciation of the word "pen".

We have to go to him for supplies and he always make me repeat myself whenever I ask for one and laughs incessantly. He says that I pronounce the word "pen" is funny. My ignorance must shine through because although I've tried to understand the "sound" difference between "pin" and "pen", I just can't. You write with a "pen", you stick something to the wall with a "pin".

He states that I say "pin" when I should say "pen". When back home in Texas, when asked for a "pen", I've never given someone a "pin" or the other way around. So I don't understand how he hears a difference.

Jim already knows the facts here: there are distinctions between vowel categories that he doesn't make, at least in certain contexts, and can't reliably hear. He also knows, I'll bet, that there are distinctions he makes that other people can't make or hear — many native speakers of Spanish who learn English as adults, for example, can't reliably perceive or produce the difference between the vowels in bid and bead or sip and seep.

Mergers of this kind are common, both within languages and in the speech of adult learners. As Jim notes in his own case, such mergers don't generally cause much misunderstanding. Language in context is redundant enough, by cultural evolution if not by design, that misplacing a feature or two rarely results in a plausible lexical subsitution. (Of course there are many stories, and quite a few jokes, that depend on the minority of cases where it works out the other way.)

But Jim seems curious about what's going on here, so what should he do? It might help him to learn the IPA symbols for the pronunciation distinction that his Chicago acquaintances make: [pɪn] vs. [pɛn].  He already knows that "pin" and "pen" have different standard English spellings, but English orthography is variable and confused, especially in the case of vowel sounds, and so maybe seeing the distinction in a phonologically consistent spelling system would help. Or maybe it would help him to hear and practice the pronunciations of these vowels in the context of the vowel quadrilateral more generally.

He can find plenty of scholarly work on the pin-pen merger itself or on similar phenomena.  I don't know what his work environment is like, but it's possible that a couple of extended dead-pan discussions of the phonetic, historical, geographical, and social extent of other American English vowel mergers — cot-caught and Mary-merry-marry-Murry are probably the best studied — would reduce his boss's enthusiasm for the topic.

If Jim really wants to learn to hear and produce a difference between pin and pen, he has some options.  There's evidence that "high variability phonetic training" (HVPT) works, at least sometimes and to some extent — but I don't know of any suitable free implementations. (N.B. Collecting the needed recordings for HPVT in a couple of common cases, say English vowels and Mandarin tones, and writing a simple web-based interface for such training in general, would be an excellent project…)  He could consult a good dialect coach and learn to style-shift between South Texas and Chicago.

I should note in passing that Jim's description of his experience sounds a little unusual to me.

It's absolutely normal not to be able to reliably hear or pronounce a distinction that not part of your native phonological system. This is an experience that I've had every time I've encountered a new language, or a new variant of English. But usually people can hear that a minimal pair, performed side-by-side in the over-distinct facultative style of such productions, is in fact different; and usually it's fairly easy to learn to identify which member of such a pair is which.

This ability doesn't generalize to identification of unpaired examples across speakers and contexts and styles. Achieving that generalization is the goal of HVPT.  But  Jim seems to having trouble with the first step, before generalization enters the picture at all. Most likely this is because his boss is apparently motivated to be the opposite of cooperative and helpful. But if not, we should look into Jim's perceptions — and those of other Texans — more carefully.

[Update: It's common to find that discrimination is esssentially no better than identification, for a continuum of stimuli representing a consonant distinction. This is the classical "categorical perception" situation, and the classical interpretation is that all you can consciously hear — or at least all you can remember — is the identification returned by an "encapsulated" mechanism for perceiving over-learned phonological categories.

There's some controversy as to whether this is entirely true even in the classical cases; but it's always been recognized that vowel categories generally don't work this way. As with colors, pitches, light or sound intensities, etc., people can generally discriminate vowel sounds much, much more finely than they can assign them to categories.

So if it really were true that Jim T. was entirely unable to discriminate (better than chance, as usual) among stimuli on an articifial continuum from "pin" to "pen", that would be an interesting surprise.]



189 Comments

  1. John Cowan said,

    July 27, 2010 @ 9:16 am

    Well, my highly anecdotal evidence disagrees: pronouncing is easier than hearing. My wife left her native state of North Carolina at age 18, and has been living in Southern Florida (not a "Southern" accent region) and then NYC ever since. Her merger is entirely intact in ordinary speech, though she can perfectly well say [pɛn] on demand. But her ability to distinguish [pɪn] and [pɛn] in isolated over-enunciated contexts is still no better than chance after almost fifty years of hearing the distinction — I've tested it repeatedly.

    A friend of mine, somewhat younger, lives in Brooklyn and speaks just like any educated white New Yorker of his generation, complete with variable rhoticity, /æ/-tensing, and so on, except for one thing: he's really from New Orleans and has a complete pin-pen merger. When I pointed out the merger in his speech, which stands out like a sore thumb to me, he said reflectively, "I didn't know anyone pronounced pin and pen differently."

    I don't have any trouble making unaspirated voiceless stops contrast with voiced ones, but I don't speak or understand any languages that phonemicize the difference, and I often wonder whether I'd be able to hear it properly either in or out of context.

  2. Tom said,

    July 27, 2010 @ 9:19 am

    New Zealand English has this same vowel shift (e > i). If the speaker is unable to hear a difference in other speakers with say RP pronunciation then it appears there is some deafness to pronunciation at play. To be unable to see anything at fault with one's own pronunciation is normal for almost everyone.

  3. the Rising Jurist said,

    July 27, 2010 @ 9:20 am

    Pin and pen are so close in pronunciation as to be nearly indistinguishable. To really make them sound distinct, you almost have to pronounce pen like pan. I conclude that Jim's boss is an ass. He probably thinks it's funny simply because Jim is from Texas. Fortunately from Jim, people from the Midwest sound way funnier than people from the South. So, fire away, Jim!

  4. Michael said,

    July 27, 2010 @ 9:27 am

    The inability to distinguish sounds within different dialects (right word?) of one language has always struck me as fascinating. I'm British (Estuary English, touch of overlaid RP when necessary), and the cot-caught and Mary-merry-marry-Murry mergers sound weird to me.

    A francophone friend of mine was having trouble with certain English vowels, and we reduced it to one sentence to try to train him: Carly carries a curry, where the /æ/ and /ʌ/ in 'carries' and 'curry' respectively caused him the most trouble (and still do). He used to have trouble with /i/ and /ɨ/, but is better at it now.

    I bring up both points because I never realised that, even within English, and even within just US English, different dialects might have lost vowel distinctions compared to others. Fascinating.

    (I should add that I'm very far from perfect. The vowels in the French words 'sus' and 'sous' sound almost indistinguishable to me, despite the words having opposite meanings. I believe the vowels are /u/ and /y/ respectively. I've had a lot of confusing conversations with francophone friends trying to make aural distinctions my brain just don't wanna make.)

    Apologies for errors in terminology, IPA, interpretation, linguistic amateur and dilettante here.

  5. Heck said,

    July 27, 2010 @ 9:38 am

    I'm from the Deep South, and I assume the pin-pen merger is a Southern phenomenon since no one I know makes the distinction. Cot-caught must be a merger somewhere else, as I've never noticed it. I make the distinction between Mary-merry-marry-Murray, though, and I make a similar one between they're-their-there, corresponding to the first three in the M-sequence. I've pointed it out to friends, but they don't hear it, even when they can tell the difference between Mary-merry-marry. I wonder if this is a personal idiosyncrasy, or if others make this distinction, too.

  6. Skullturf Q. Beavispants said,

    July 27, 2010 @ 9:44 am

    Something I've noticed in myself about detecting a difference between merged vowels:

    I'm from Western Canada, and consequently grew up with the cot-caught merger. For probably the first 15 or 16 years of my life, I had no idea that any English speakers pronounced these words differently, but over time, I learned to detect the difference when listening to those speakers who make the distinction.

    However, even though I can detect a difference, it's very hard for me to wrap my mind around the fact that those words don't rhyme for people without the merger. I can remember reading on the internet about songs or poems that rhyme, say, "flawed" with "god", and some people complaining that those don't rhyme. My gut wants to say, "Well, sure, you can pronounce them with a subtly different sound, but they still rhyme."

  7. John Cowan said,

    July 27, 2010 @ 9:46 am

    Rising Jurist: I agree that Jim's boss is an ass. But for me, and other people who don't have the merger, pin and pen are as different as pit and pet, and for exactly the same reason. And everybody sounds funny to somebody.

    Michael: You have /u/ and /y/ reversed: sous has /u/, sus has /y/. Also, the people who merge Murray and merry do not also merge them with marry. There's a limit to what mergers can do to stressed vowels before destroying essential distinctions. Some mergers of historical distinctions have been carried through completely, like the vain-vein merger (no anglophone now distinguishes these words, though they once were different), but the great majority are made by some and not others, varying from the almost complete ate-eight merger to the very rare met-mat merger.

  8. kd said,

    July 27, 2010 @ 9:49 am

    My husband says pin & pen the same as well. Many years ago, I noted with some curiosity that when he is seeking a pen, he will request an "ink-pen" (sidenote: what other kind are there? would one ever ask for a pig-pen?). I decided that must be how folks in his neck of the woods made the distinction, since there was no way they could hear the difference between pin and pen.

  9. Nathan said,

    July 27, 2010 @ 9:56 am

    As an undergraduate I minored in linguistics. One day the instructor of our Intro to Historical Linguistics course was absent and asked a colleague to substitute/guest lecture. She talked about her research on the distribution of the cot-caught merger. In my own idiolect the merger is complete; I have no low round vowels. I found the phonetic details of the lecture baffling. At one point I raised my hand to ask how to find minimal pairs to test. How would you know which words were supposed to have this mythical /ɔ/ sound? She really didn't understand my naive question.
    I wonder if it's harder for linguists to study distinctions that are not phonemic for them.

  10. Josef Fruehwald said,

    July 27, 2010 @ 9:59 am

    This is a great example of how someone who is even powerfully socially motivated (scorn from a figure of authority) can't undo a merger. Although, it's unfortunate that Jim's getting ragged on by his boss like that. Maybe the HVPT would do the trick for him, but in the normal social life of a person, they're really out of luck.

    This reminds me a little of a news story from a while ago from Utah. Out there, they have a fill-feel merger. One day, a robber tried to knock over a cafe, placed a bag on the counter and said "Fill the bag!" The clerk then reached over and felt the bag. It was perhaps the most disambiguating context there could be, but the clerks till misunderstood the robber.

    Here's that news story:
    http://www.ksl.com/index.php?nid=148&sid=3058039

    And I highly suggest listening to the audio of the cop trying to make the distinction for the news.
    http://real.ksl.com/video/slc/2/231/23149.mp3

  11. Kirk Hazen said,

    July 27, 2010 @ 10:02 am

    For North Carolina and West Virginia, the merger has been complete for enough generations now that contextual coping mechanisms are firmly in place if confusion arises: e.g. stick pin vs. ink pen.

    Socially in the US South, the pin/pen merger often goes unstigmatized, except by occasional outsiders. What does appear on the social radar, and can be concurrent with the pin/pen merger, is the movement of the Southern Vowel Shift for these two vowels. In the Southern Vowel Shift, consider both vowels moving towards the front of the mouth and gaining an offglide to become a diphthong. So that [bit] may sound more like [bee-it] and [bet] may sound more like [bait] or [baa-it].

    If Jim T has both the pin/pen merger and the Southern Vowel Shift going on, then there may be multiple layers of social meanings happening for his inconsiderate boss.

  12. Mark P said,

    July 27, 2010 @ 10:07 am

    At first I was a little suspicious about Jim's story, because I think I can tell the pen-pin difference quite easily, despite having spent most of my life in the South, where virtually no one makes a distinction. But then I thought a little about the fact that my mother grew up in Ohio and still has traces of northern pronunciation. I also thought about the fact that a lot of people claim they pronounce certain words differently when in fact they don't; they apparently can't tell the difference in their own speech. I think I pronounce pen somewhere between pin and pen, but I suspect that I really don't, at least unless I am consciously trying to say pen differently from pin.

    Jim might try a technique I used once when someone laughed at the way I usually say "dog." She pronounced it the way she thought I said it, and I made her keep repeating it and laughing so much that the whole subject was eventually dropped.

  13. Jude said,

    July 27, 2010 @ 10:11 am

    The two sounds which I cannot distinguish as a Colorado native are the two central vowels which on the University of Iowa site

    http://www.uiowa.edu/~acadtech/phonetics/about.html# they distinguish as being different in "herd, earth, and fur" from "farmer and waterfall."

    Discovering that I had this "hearing" disability made me a more effective ESL teacher.

  14. Cecily said,

    July 27, 2010 @ 10:18 am

    @the Rising Jurist: You said "Pin and pen are so close in pronunciation as to be nearly indistinguishable."

    The whole point of this post is that whilst that is true for some people is it completely untrue for many others, including most people in England (as I am).

  15. Will said,

    July 27, 2010 @ 10:21 am

    I often pronounce "pen" like "pin", though I can distinguish the sounds if need be. However, I do not pronounce Ben and bin alike. Went sounds like "wint", but lent, pent, bent retain the e vowel. So I don't know if I really have the merger or if this is just an idiosyncratic speech pattern.
    (I'm from Iowa so I shouldn't have the merger, but my family is from all over–maybe I heard someone say "pin" for "pen" and picked up the pronunciation just for that particular word.)

  16. William Ockham said,

    July 27, 2010 @ 10:25 am

    I'm from Texas (lived here all 50 of my years except for 4 years in Nebraska during high school). I have never heard a difference in the pronunciation of pin and pen. I would have told you they were homphones before reading this. Is there really a difference?

  17. Theophylact said,

    July 27, 2010 @ 10:34 am

    Here's what I don't understand in Jim T's story:

    When back home in Texas, when asked for a "pen", I've never given someone a "pin" or the other way around. So I don't understand how he hears a difference.

    If he never makes a mistake, he must either be hearing the difference, or he must always be in a situation where the context is obvious. But in an office, surely "may I borrow a pin/pen" must nearly always be ambiguous if you can't hear the difference,

  18. Theodore said,

    July 27, 2010 @ 10:35 am

    Does Jim make a distinction between "whine" and "wine" ? Assuming his boss doesn't, he could start looking for opportunities to mock this feature of his boss's speech.

  19. Jim T said,

    July 27, 2010 @ 10:42 am

    Well I'm blown-away by all the responses. Quite honestly, I'd never thought this would be such an interesting subject and I'm now completely captivated.

    Just to toss in a bit of humor for y'all:

    My sister, who'd never really lived in Texas and in fact had been in Minnesota for years was down in Texas visiting me. We were in a Blockbuster Video store but a few rows apart. She was choosing a drink from the little refrigerator they had and after raising her voice so I could hear said "Jimmy, do you want a pop?"

    You could have a heard a pin drop as the store went quiet. After a few seconds, a gentleman that was near my sister turned to her and said, "you're not from around here are you."

    I about died laughing.

    Jim

  20. Amy Stoller said,

    July 27, 2010 @ 10:46 am

    As someone from NYC, I have always had a complete pin-pen distinction. I would never say that pen sounds like pan when thinking about my own speech and the speech I was surrounded by as I grew up, but I am aware of some midwestern accents in which there does seem to be something like a pen-pan merger – or at least a herold-Harold merger, such as I hear in Garrison Keillor. (I assume this is related to merry-marry merger, but maybe something else is going on.)

    It is certainly true that ɛ is produced with a tongue position nearer to æ than to ɪ. So aiming for pan with æ might help someone achieve pen with ɛ where formerly they had only pen and pin with ɪ.

    It seems to me there are several topics on this thread. One is whether a person has a distinction or merger in their own speech; another is whether they can hear a distinction in someone else's speech; still another is whether they can produce the distinction; and yet another is social attitudes about the whole darn thing.

    Certainly in my coaching practice I have enabled some people to produce a distinction in a new accent, where they have a merger in their everyday speech. Often they need to learn the distinction kinesthetically before they can hear it in themselves or in others.

    I will say that don't think it's at all constructive to mock anyone whose speech is different from one's own. In my opinion, Jim's problem is that his boss has a real problem – he's a bad manager.

    I do know people with pin/pen merger who solve communication issues by asking for an ink-pen when there is a danger of being thought to have asked for an identification PIN, a sewing pin, or a brooch. As a coping strategy, this works well, and is often all that is necessary. The listener might also learn to ask whether the speaker means an ink-pen or a PIN, etc.

    I do have the impression that pin-pen merger is spreading outside it's traditional boundaries.

  21. Nemo said,

    July 27, 2010 @ 10:49 am

    If Jim in fact wishes to hear the difference in how others pronounce pin and pen, he could try one of the pronunciation websites, like http://www.howjsay.com/ .

    If Jim instead wishes to shut his asinine boss up, he could surely identify similar "oddities" in his boss's speech. "Milk" might be a good one if the boss is a Chicago native — it often comes out more like "melk."

  22. Adrian Morgan said,

    July 27, 2010 @ 10:59 am

    Regarding HPVT, a friend of mine took an interest in this (after I showed her the original LL post in 2008) and was, I understand, inspired to create tools based on it for use in English language classes that she taught. (More powerpointy than webby, though, as tools go.) She was planning to do a recording session with me sometime.

    Sadly, this friend died of cancer quite recently, her most ambitious project of its kind never completed. (And a linguistics book that I lent her never finished, but that's another story.)

  23. Andrew (not the same one) said,

    July 27, 2010 @ 11:11 am

    I am confused. Does this merger apply only to 'pin' and 'pen', or would those who have it also merge 'pick' and 'peck', 'sit' and 'set', etc.?

    If the former (which is the impression I get from most of the discussion), would it not be easy enough to demonstrate how the words could be pronounced differently, simply by pointing to other words with e's and i's in them?

  24. Larry Lard said,

    July 27, 2010 @ 11:19 am

    @William Ockham

    "I have never heard a difference in the pronunciation of pin and pen. I would have told you they were homphones before reading this. Is there really a difference?"

    There are a number of pronunciation websites scattered about the web; for example http://alt-usage-english.org/audio_archive.shtml

    I find that hearing spoken passages is sometimes the only way to truly believe that some people have a split where you do not.

  25. the Rising Jurist said,

    July 27, 2010 @ 11:24 am

    @Cicily "The whole point of this post is that whilst that is true for some people is it completely untrue for many others, including most people in England (as I am)."

    Well, of course it sounds different to someone from England. You people pronounce words all wrong.

    And I realize what the point of the post is. I am one of the people for whom there is no difference between pen and pin (or cot and caught, for that matter), so I am just showing Jim some love. Southern solidarity!

  26. Nina said,

    July 27, 2010 @ 11:32 am

    I am from south Louisiana and cannot hear the distinction between "pin" and "pen." I do try to say words correctly, though, but sometimes miss it and say something like "pan."

    And, like someone else mentioned, whenever we needed to make a distinction between "pin" and "pen," we would say "stick pin" or "ink pen."

  27. kip said,

    July 27, 2010 @ 11:35 am

    @Andrew (not the same one)

    I believe it only happens before /n/ or /m/. I have a complete "pin-pen" merger, and I also have a "him-hem" merger. There is no merger for pick/peck and sit/set.

    I thought maybe the merger exists before nasals, but I can't think of a word now with "-eng" to test if the merger exists for me before /ŋ/. If "deng" was a word I would pronounce it different from "ding". Although, now that I think about it, "deng" would come out with the same vowel sound as din/den, whereas "ding" comes out closer to "dean".

  28. Melissa Fox said,

    July 27, 2010 @ 11:46 am

    Regarding the cot-caught merger, there was some considerable hilarity when a friend of mine (who has one vowel for both those words) was telling me (I have two) about a style of dancing that involves a cVller and explicit instructions (she may even have said 'orders') and so on, and I could feel the bewilderment and growing horror on my face for several seconds until I understood that for some reason her single vowel is closer to the vowel I have in "collar" than it is to the vowel I have in the word she meant, which was of course "caller".

  29. Heather said,

    July 27, 2010 @ 12:01 pm

    I was going to fumble through my own explanation, but Kirk Hazen did it better than I could. I will just add that in my experience (my family is from all over the South), the issue applies to pin and pen, peel and pill, and could apply to pet and pit if they are from an area that does not pronounce the t (or at least pronounce the t in the back of the throat instead of with the tongue at the tip of the mouth.)

    Regarding pat and pet, I saw an old Loretta Lynn commercial yesterday where she introduced her daughters, Patsy and Paggy. Paggy? Who would name her child Paggy? Looking it up on Wikipedia, turns out they are named Patsy and Peggy. But it certainly sounds like Paggy to me.

    http://www.youtube.com/watch?v=-WATWe3Ipxw

  30. slavicpolymath said,

    July 27, 2010 @ 12:06 pm

    Do not speak to Russians of sheets or beaches.

  31. Boris said,

    July 27, 2010 @ 12:07 pm

    I am originally from Russia (19 years ago at the age of 11) and I have trouble hearing the difference between the "th" sound and the "f/v" sound. I certainly make them differently myself, but I can't hear the difference. Do native speakers? Most recently, for the longest time I thought the "National Incident Commander" in the BP oil spill was named "Fat Allen" until I finally saw it in print.

  32. Josh said,

    July 27, 2010 @ 12:32 pm

    @Boris
    There are some dialects/accents in English the blur the distinction. (Mouf for Mouth, for example)
    But by and large, the th, f, and v sounds are all quite distinct in English.

    The th sound is quite difficult for a lot of non-native speakers, though, so you're not the only one.

  33. Ginger Yellow said,

    July 27, 2010 @ 12:42 pm

    New Zealand English has this same vowel shift (e > i).

    Indeed. The pin/pen merger has come up a lot recently in various contexts, and I've been struggling to get my head around it in a US context as I can only hear it in my head as a Kiwi accent. It's just not something I've happened upon (or noticed, anyway) in my travels in the US. I'm pretty sure my (Illinois born) parents distinguish the two, but I'll have to check.

  34. KevinM said,

    July 27, 2010 @ 1:00 pm

    The Spanish i/ee confusion referred to in the main post and some followups surfaced hilariously in a CD of pop songs made by Placido Domingo some years back (love Placido, but the album was execrable). He rendered the first line of the awful John Denver "Annie's Song" as "You feel up my senses…"
    Those wishing to hear the Kiwi version might watch HBO's "Flight of the Conchords." Every mention of one character's name (Brett/Britt) or any affirmative response to a question ("Yiss") will demonstrate.

  35. Jim T said,

    July 27, 2010 @ 1:02 pm

    I think that Theophylact is correct. Typically when speaking about a pin or a pen it's within a specific context.Therefore no misunderstanding would occur.

    One of my co-workers who is an Illinois native was helpful by allowing me to use him to conduct a test.

    Items: a PIN and a PEN

    Basically this was a "flashcard" type test. I would hold one of the items up and he would pronounce the word for each. After about twenty repetitions I finally heard the difference! My ears kind of went into shock. Hearing that difference actually sounded to me like a foreign language as it came across as very unusual, almost "wrong". I find it extremely difficult to describe the "shock" I experienced when I finally heard the difference.

    I then attempted to pronounce the words "properly" myself, albeit with much background laughter from my co-workers.

    Pin was fine. However "pen" started sounding like "pan" to me. My co-workers said it sounded as if I was saying "pee-uhn". I worked on this for around 15 minutes non-stop.

    I would have my co-worker say "pen" and I would attempt to mimic him. I decided to envision an "h" being adding to the word 'pen", like "peHn" and that seemed to help. Although I must admit it was a bit of a tongue twister.

    What I really find unusual is that I consider myself fairly worldly and am well educated. There are a few interesting facts:

    1. I come from a mixed marriage (in Texas that means my Mom is from Highland Park in Dallas, Texas and my Dad is from Seminole, Oklahoma). Considering that my Mamaw was a house-mom at OU, it makes for an interesting College football season.

    2. My dad was career military. From start to the current situation: born in Texas and lived there a few years (San Antonio, Ft. Hood, Park cities (Dallas). Then Manhattan and Ft. Riley, Kansas. Then moved to Italy where I "grew up" as a teenager. Then back to the USA in Oklahoma and then Ft. Sam Houston in San Antonio, Texas. Then as an adult I moved to Berlin, Germany for a few years (military but stayed off base as much as possible), the Ft. Campbell, Kentucky for almost two year, then back to San Antonio Texas for twenty years, and now Chicago for three. Note that while living in Europe we lived on the local economy, not in military housing areas.

    3. I fell in love with both Italian and German. Although certainly not fluent in either, my pronunciation of words in either language is almost flawless. I can't actually prove this to y'all but my European friends (and strangers) say I sound almost like a native speaker. I have no issues pronouncing any words in either German or Italian.

    4. Almost everyone I meet here in Chicago "knows" that I'm not from here. They usually guess the South or Texas. Although I don't really think I have an accent. I've actually been told by some Europeans I've meet that I sound like a snobby American. When I ask them to explain that further, they say I have no accent and thus aren't "provincial".

    5. Why then, all the damn trouble with pin and pen? (and a few other words of course!)

  36. SeanH said,

    July 27, 2010 @ 1:06 pm

    @Josh: for Cockney, I think what the 'th' turns into depends on where it is in the word. "Shut your mouf", for instance, but also "dahn ver pub".

  37. Lauren O said,

    July 27, 2010 @ 1:20 pm

    When back home in Texas, when asked for a "pen", I've never given someone a "pin" or the other way around.

    Shortly after moving from California (where I grew up) to Texas, I was in a meeting at work. My boss remarked, "It's so quiet you could hear a pin drop," then dropped his pen on the table to demonstrate. So there are some meaning mix-ups at least once in a while.

    Side note: Although I don't think I ever heard a Californian pronounce "pin" and "pen" the same, I definitely heard plenty of Californians say "melk" and "pellow" instead of "milk" and "pillow."

  38. Bill Walderman said,

    July 27, 2010 @ 1:31 pm

    'I have trouble hearing the difference between the "th" sound and the "f/v" sound.'

    I'm a native speaker of English and I too sometimes have trouble hearing these sounds distinctly, but I think it's due to an auditory issue, not a phonemic merger: there's no merger of these sounds in the English speech of most people I come into contract with, but the acoustical characteristics of these sounds are similar enough that it's difficult to hear the difference in less than optimal conditions. Confusion between "s" and "f" or "th" is less frequent but sometimes occurs, for me at least. Maybe someone with a background in phonetics could address this.

    But it's interesting that in Russian loan words from Greek, especially names, the "theta" sound (pronounced as an unvoiced apico-dental fricative since Roman times), which is foreign to Russian, shows up as an unvoiced labio-dental fricative /f/. For example, the Russian equivalent of the Greek name Theodoros is Fyodor. Gk. "arithmetike" was borrowed as Rus. "arifmetika," although "mathematike" became "matematika." If I'm not mistaken, Russian loan words with /f/ for Greek /th/ generally represent an older stratum of borrowings.

  39. Chris D said,

    July 27, 2010 @ 1:31 pm

    I thought maybe the merger exists before nasals, but I can't think of a word now with "-eng" to test if the merger exists for me before /ŋ/. If "deng" was a word I would pronounce it different from "ding". Although, now that I think about it, "deng" would come out with the same vowel sound as din/den, whereas "ding" comes out closer to "dean".

    I think you'll find that at least some people from Texas (I base this on my own not-very-careful observations of myself, my family, and my friends) pronounce words like "strength" and "length" with a vowel like the [i] in "tea" rather than the [ɛ] in "pet". And yes, that seems to be a tense vowel. I don't produce any vowels as [ɪ] before a velar nasal, but instead as (something like) [i]. In my pronunciation of words like "pink", for example, the vowel is tense rather than lax, and this seems to be true of my native North-Texan friends and family as well.

    This has caused a bit of amusement among my phonetically sensitive colleagues, and is, unlike the pin-pen merger, not a feature of Texas (southern?) English I have seen described, in the linguistics literature or elsewhere.

    As for Jim, I also find the "deep phonetic deafness" aspect of the story surprising. The lack of a distinction in my own speech became quite apparent to me after moving to Massachusetts, and within a few years I found myself (hyper)-correcting for the lack of distinction, quite unintentionally.

  40. Eric said,

    July 27, 2010 @ 1:33 pm

    Although, now that I think about it, "deng" would come out with the same vowel sound as din/den, whereas "ding" comes out closer to "dean".

    That's funny/interesting. I've read about the pen-pin merger and took that whole "before nasals" thing at face value, but have also noticed that even though "ring," for example, already has the same vowel as "pin" in standard Am.E., in the AAVE out here on the West Coast it is often very close to "reeng."

  41. Rodger C said,

    July 27, 2010 @ 1:33 pm

    I grew up in West Virginia with the pin/pen merger (straight-peeun vs. ink-peeun), but as I'm good at foreign languages, starting with Northern, I trained myself out of it. I also grew up pronouncing marry/merry/Mary/Murray the same, but trained myself to say only the first three the same, which I was told was standard English.

    I believe the pin/pen merger has been shown to have started in the Upper South and spread southward. I always like to say wthat when Appalachians move north for jobs, their teachers discover they can't distinguish a straight-pin from an ink-pen, and put them in speech correction classes to be taught proper discrimination by someone who can't distinguish a ham hock from a chicken hawk.

  42. Eric said,

    July 27, 2010 @ 1:44 pm

    I'd like to re-up Nathan's question from above.

    It's easy enough for someone with pin-pen merger (me) to understand the distinction others use, just by analogy with other pairs like pit-pet. But with cot-caught, I just don't even know what the vowel in the 2nd word is supposed to sound like. What's a word with that sound that I can use to get the distinction by analogy? Introspection is impossible here b/c prima facie, I'd say that each of bought-bot, fought-fot, wrought-rot are also identical. But of course, maybe in fluent speech I make a distinction that I can't hear when I deliberate about it.

    So I went to the audio archive Larry Lard linked to, where they have "Bother, father caught hot coffee in the car park." Under the USA category, every speaker save one uses a single vowel, to my ear, in the series "caught hot coffee." The exception is "American speaker with various influences," but his speech seems pretty New York-inflected to me, so I'm not inclined to take it as exemplary of a sound that, evidently, *most* Americans produce (just as I wouldn't take the Buffalo, NY pronunciation of "backpack" as exemplary of the way most people pronounce those vowels).

    So, any advice on how I can grasp the mainstream American version of this supposed, : ), caught vowel?

    BTW, Boris, I love "Fat Allen." That's awesome.

  43. Eric said,

    July 27, 2010 @ 1:47 pm

    2 different Eric's here, BTW.

  44. Lars said,

    July 27, 2010 @ 1:52 pm

    @KevinM:

    And there was that time when UN Secretary General Javier Pérez de Cuéllar spoke of the importance of "international piss and security."

  45. mollymooly said,

    July 27, 2010 @ 1:52 pm

    The pin-pen merger is one of the features of a rural/western Irish accent which is most stigmatised in metropolitan Ireland.

    @John Cowan: "the almost complete ate-eight merger"
    Since pronunciation of "ate" varies lexically as well as by accent, a different minimal pair would be better.

    @kip : 'I thought maybe the merger exists before nasals, but I can't think of a word now with "-eng" to test if the merger exists for me before /ŋ/.'

    There are "length", "strength", "penguin", "dengue"; and "enclave" if you don't fancy-French it up. Also plenty of names: Jenkins, Chernenko, Engels, Englebert, Genghis Khan, Hang Seng, Cincinnati Bengals, etc. Some of these names have variants spelt with I rather than E (in English or another language).

    Note that "England" and "English" are spelt with E but pronounced with /ɪ/ rather than the expected /ɛ/. The in- and en- prefixes (e.g. "insure", "ensure") are unstressed and hence all bets are off.

  46. Rodger C said,

    July 27, 2010 @ 1:53 pm

    I just said "caught hot coffee" to myself several times, and I'm sure I pronounce the vowel in "hot" unrounded and the other two rounded.

    I did, however, at one time in my youth, put a lot of work into pronouncing that rounded vowel as a monophthong. It seems to be unstable in American English in general. Everywhere I know of, it's either unrounded (cot/caught merger), made a falling diphthong (Southern "daowg"), or made a rising diphthong (New York "dooag").

  47. Mark P said,

    July 27, 2010 @ 2:02 pm

    @Roger C: Most people in the Atlanta TV market can now understand what the sports reporters mean when they talk about the Atlanta Hocks.

  48. J.M said,

    July 27, 2010 @ 2:04 pm

    I grew up in Dallas and never had a strong Texas accent (compared to my cousins that lived in small towns say), and what I've had has diminished because I've lived in northern cities. These two words still give me trouble. And the merger is very narrow. Ben and bin, bid and bed, even in and en sound nothing at all alike (though I'm not sure my cousins wouldn't have trouble with the first pair.)

    Whenever I think it is going to be a problem (usually that only happens with people that don't think ant and aunt aren't homophones), I think of bid/bed just before I speak. It seems to help.

  49. Ellen K. said,

    July 27, 2010 @ 2:13 pm

    For me, length and strength, and most of those other words on mollymooly's -eng list have /e/ not /ɛ/ (if one can use IPA in phonemic inscription, that is) That is, the vowel of pain. Though, saying them to myself, not quite pronounced the same, still, phonemically, the pain vowel rather than the pen vowel. Exceptions, dengue and Jenkins, which have the pen vowel (plus I'm not familiar with Seng).

  50. Tenderfoot said,

    July 27, 2010 @ 2:17 pm

    Mississippi-born, I grew up pronouncing "pin" and "pen" alike, but got corrected so often after I left the South that I now sometimes over-correct and say "pen" when I mean "pin." (Though never vice versa.)

  51. HB said,

    July 27, 2010 @ 2:42 pm

    I'm from the Ozarks and have the pin-pen merger. I've lived in the Chicago for many years now, and I can *almost* hear the difference when natives pronounce the words, but not quite. "Pen" sounds almost like "pan" when they say it.

    I can almost produce "pen" the way these folks want me to if I pretend to be speaking Spanish instead of English and put a vowel on the end of it.

    For me, the merged sound is only "e" becomes "i" before "n" or "m". Words like "pit" and "pet" are different. I wish I knew IPA so I could have a vocabulary for what I'm talking about.

  52. dw said,

    July 27, 2010 @ 2:46 pm

    @Eric, @Nathan:

    The distinction between "cot" and "caught" may be (for some North American speakers) analogous to that between "barn" and "born" or possibly "cart" and "court".

    In other words (assuming you have the horse-hoarse merger):

    * "cot" sounds like "cart" without the R-sound.
    * "caught" sounds like "court" without the R-sound.

    If you are among the small group that have merged cot-caught but split horse-hoarse then you could use a different example:

    * "Don" rhymes with "barn" without the R-sound
    * "Dawn" rhymes with "born" without the R-sound.

    For other speakers, the distinction may be principally one of length, with "caught" being phonemically long and "cot" phonemically short. I fall into the long-short group, but I grew up in England so I don't count.

  53. Flourish said,

    July 27, 2010 @ 2:50 pm

    This I/E merger has gotta be a Texas thing. My partner is Texan and he pronounces "pillow" "pellow" (which drives me nuts, by the way – I'm from California). He also does the pen/pin thing.

  54. anotherjim said,

    July 27, 2010 @ 2:53 pm

    I've received all grades of heck from my family for the last 25 years (moved from NC to PGH, PA.)

    Why do they not make fun of the way I say ten (tin)?

    I just ignore the remarks, since they don't make much sense…

  55. onymous said,

    July 27, 2010 @ 2:55 pm

    I remember the first few times people told me they pronounced "merry," "marry," and "Mary" differently, I couldn't hear any difference. But over time I've heard a few more people explain it, and listened to some clips on the internet of people who make the distinction, and now it's moderately clear to me. I might not pick it up if I'm not listening for it, but if I try it's not hard at all to hear. So it seems like to some extent it's possible to train oneself to hear distinctions one didn't grow up with.

  56. Nathan said,

    July 27, 2010 @ 2:56 pm

    @dw: As a former linguistics student, I'm used to misjudging my own pronunciations of words, but I've always had the impression that my vowel in court is the same as my vowel in coat (my English is rhotic).

    I know I've heard pronunciations of orange on TV where the first vowel sounds to me like the one in caught, but that's not the way I say it.

  57. The Ridger said,

    July 27, 2010 @ 3:02 pm

    Most people in the Atlanta TV market can now understand what the sports reporters mean when they talk about the Atlanta Hocks.

    This is funny to me (who says them the same, along with pin/pen and hot/caught and probably lots of others) because I distinctly remember being taught in school that the main reason we have all these spellings isn't because they reflect different pronunciations but rather are a way to tell words apart in reading – when the author isn't there to correct you if you chose the wrong meaning, as an interlocutor could.

  58. Nee in Germany said,

    July 27, 2010 @ 3:04 pm

    My husband and I are both Texans with the pin-pen merger, but our daughter, who is 13, has moved on to hypercorrecting "pimple" to "pemple". (We moved to Germany when she was 6, and she had only lived in Texas for 5 out of her first 6 years.) She gets very defensive when we point out the hypercorrection, although she loves correcting my terrible German vowel production. :)

  59. Larry Lard said,

    July 27, 2010 @ 3:22 pm

    @mollymooly:

    >> There are "length", "strength", "penguin", "dengue"

    You have a /ŋ/ in the first two?

  60. Rodger C said,

    July 27, 2010 @ 3:27 pm

    By the way, now that I've read John Cowan's statement that no one pronounces merry/marry/Mary/Murray all the same, I'm not at all sure I described my childhood speech habits accurately. Apparently my education has permanently confused me. Not for the first time.

  61. Jim said,

    July 27, 2010 @ 3:38 pm

    Given my first name, I've had discussions (arguments) with people who claimed that "Jim" and "gem" sounded identical. That one ends when I tell them that "Jim" and "gym" sound different, too. (There's an interesting pathway there: If "gem" and "Jim" sound the same, and "Jim" and "gym" sound the same, shouldn't "gem" and "gym" sound the same? They definitely don't to me. Maybe it's that the two pairs are like 1/4 step off, too subtle for most people, but that makes the ends 1/2 step off, more obvious.)

    I've also been accused of tight enough pronunciation to hear the "t" in "castle". Usually that takes the effect of "Are you British?"

  62. Nathan said,

    July 27, 2010 @ 3:49 pm

    @Larry Lard: I can't speak for mollymooly, but I know that many people (myself included) have /ŋ/ in strength and length; every dictionary I've checked gives that as the first pronunciation.

  63. Diane said,

    July 27, 2010 @ 3:49 pm

    When I lived in Nepal, I was always surprised by the fact that Nepalese swore they could not hear the difference between "S" and "SH" no matter how carefully I pronounced them. And yet, there are words in the Nepalese language that are clearly (to me) and consistently pronounced with an "S" and others with an "SH" and they do not mix up the two.

    Even stranger, to me, is that their alphabet actually has three letters, one of which is typically transliterated as "S" and the other two as "SH". This written distinction does not correlate very well with the distinction between "S" and "SH" as I hear it in their spoken language. And Nepalese will tell you that all three letters are all pronounced the same (I did once meet an educated person who told me that there used to be three different pronunciations corresponding to the three letters, although he could not pronounce it for me.)

    FWIW, IANAL. (For what it's worth, I am not a linguist.)

  64. Greg Morrow said,

    July 27, 2010 @ 4:06 pm

    I'm from Kentucky, with adult life in Texas, both South Midlands strongholds, and I have strong pin-pen merger. (Which correctly places me outside of Louisville, if you know your dialect clines.) If I try, I can readily produce contrasting pin/pen, and to a lesser extent hear it, but only if I try.

    I also have merry/Mary/marry merger, and I find that as confounding as Jim T does pin/pen. I can sort of understand what a second vowel in there would sound like — I have /ɛ/, and I sort of hear what a low front unrounded vowel would do (/æ/ presumably), but I can't even guess what the third vowel would be for people that contrast all three.

  65. Jerry Friedman said,

    July 27, 2010 @ 4:11 pm

    @Amy Stoller: I have the herald-Harold and marry-merry(-Mary) mergers, and I think they're exactly the same merger. Only before /r/, of course.

    @Eric (caught-cot): Try the pronunciations at http://www.merriam-webster.com/dictionary/caught and the same with "cot". Those sound correct to me or maybe slightly too high, back, and rounded. "Correct", of course, means the way I say it :-)

    @dw: Merriam-Webster's version of caught sounds about halfway between cot and non-rhotic court to me. Like Nathan, I think of court as having the vowel of coat, as I was taught in first grade. I think a lot of the Americans who use the same vowel sound in caught and court are from the New York area.

    Is there any data on how many Americans are cot-is-caught and how many are cot-is-not-caught (or CIC and CINC, as we say at alt.usage.english)? I'd guess it's about 50-50, or maybe more CIC.

  66. Jerry Friedman said,

    July 27, 2010 @ 4:14 pm

    @kd: Another kind of pen is the lead pen, which is the term some people here in northern New Mexico use for what I call a mechanical pencil and the British call a propeller pencil (I think).

  67. Christy said,

    July 27, 2010 @ 4:20 pm

    I have the pin/pen merger. I got tired of my friends correcting me so I have learned to say pehn. I always have to stop and think about it before I say it, though, or it will come out pin. I say milk, but I also say pellow, which is funny to me after having read all these comments. My dad is from Illinois, my mother from Ohio but her family sounds more like their Kentucky neighbors. My dad was in the military and I spent most of my youth in California, but on military bases. I have no idea where I got most of my accent but I suspect it is from my mother's family. I was in junior high before I learned it was wolf (not woof) and barbed (not barb) wire, or that garnet rhymed with darn it, not garNETTE. I had years of diction to overcome a childhood lisp so my words were all distinctly formed just "mis"-pronounced. I am going to listen and see if I can learn a difference between cot and caught, though I could hear a difference between caught and hot. I just can't figure out how I'm making it different…

  68. Mabon said,

    July 27, 2010 @ 4:25 pm

    Up here in Boston, some years ago we had a food critic with his own radio show who must have been from somewhere that did not have a pin/pen distinction. For every show, he offered a prize based on solving clues to a puzzle. For this one particular contest, in order to solve the clue, one had to "know" that pin and Ken rhyme — which in Boston, at least, they do not.
    Well, the contest stretched on and on without a winner, until a listener with the same dialect idiosyncrasy finally claimed the prize, much to the incredulous dismay of his dumbfounded audience.

  69. Nathan Myers said,

    July 27, 2010 @ 4:38 pm

    I grew up in Hawai'i and merged cot/caught, hock/hawk, bull/bowl and numerous others (but not pin/pen), and wondered why dictionaries displayed these words as pronounced differently. Now, though, I have acquired all the distinctions I know about, both hearing and pronouncing. My kids recognize and use them too, including some questionable ones like "dawg" and "cawffee", for fun.

    Upbringing isn't necessarily fate.

  70. Mark Etherton said,

    July 27, 2010 @ 4:43 pm

    @Jerry Friedman

    If it's the sort of metal pencil that has a lead you insert (which you turn a knob or the end of the pencil to move up as it wears away through writing) in London English it's a propelling pencil.

  71. William Ockham said,

    July 27, 2010 @ 4:52 pm

    Well, now I see why people use that IPA thing, even though IPA still means India Pale Ale to me. This stuff is hard to write about otherwise. My wife's name is Mary because she was born on Dec 22 (Mary Christmas = Merry Christmas), but I married her, not merried her. Marry rhymes with Barry, but Mary and merry rhyme with berry.

    To me, cot and caught sound completely different (likewise bot and bought, sot as in drunkard and sought). The in/en merger, on the other hand, seems totally complete. (Fin, fen) (Ben, been, bin) (kin, Ken, ken) (den, din) (gen, gin) (Chen, chin).

    I do remember an argument I once had with someone over the pronunciation of Penn Gillette's first name that was completely mystifying to me at the time. I thought the other guy was saying that the name was pronounced "pan".

  72. Xmun said,

    July 27, 2010 @ 5:24 pm

    "Mary and merry rhyme with berry"

    No they don't. "Mary" rhymes with "scary" and "hairy". Do those words also rhyme with "berry" in your speech? And are "fairy" and "ferry" homophones?

    As for pen/pin, I can record that when I came to NZ in the early 1960s as a schoolmaster, I couldn't understand why the kids would sometime say to me: "Please, sir, I forgot my pin."

  73. Neil Dolinger said,

    July 27, 2010 @ 5:34 pm

    Hopefully Jim T.'s experience was not as painful as this guy's —

    http://www.homestarrunner.com/cantsayjob.html

  74. Spectre-7 said,

    July 27, 2010 @ 5:54 pm

    I'd never thought twice about Mary/Marry/Merry merger until this past year when I discovered (much to my surprise) that my parents don't have it. This was particularly disturbing because my father's name is Barry, and I never once noticed that I don't pronounce his name the same way they do.

    It took a lot of careful listening to even pick out what was different about their pronunciation, so I'm a little better equipped now to understand how folks can be deaf to other differentiations.

  75. William Ockham said,

    July 27, 2010 @ 5:59 pm

    Xmun,

    Scary and hairy rhyme with Barry, not berry. Yes, Fairy and ferry are homophone and both rhyme with Barry as well.

  76. James Kabala said,

    July 27, 2010 @ 6:09 pm

    "For this one particular contest, in order to solve the clue, one had to 'know' that pin and Ken rhyme."

    Golly – I've read about the pen/pin merger before but always assumed that it meant people were pronouncing "pen" like "pin," not vice versa as your anecdote would seem to indicate.

  77. dw said,

    July 27, 2010 @ 6:13 pm

    @Nathan:

    I've always had the impression that my vowel in court is the same as my vowel in coat (my English is rhotic).

    Oh — I thought that only speakers with split horse-hoarse had coat + R = court.

    Does "born" without the R sound like "bone" to you as well?

    In that case I'm stumped :)

  78. dw said,

    July 27, 2010 @ 6:22 pm

    @Greg Morrow:

    My vowels in "merry" and "marry" are as far as possible phonetically identical to those in, say, "Betty" and "batty".

    My "Mary" is basically a lengthened (about double) version of "merry", with a slight schwa offfglide.

  79. Matthew Kehrt said,

    July 27, 2010 @ 6:40 pm

    @Xmun For me (from central New Jersey on the East Coast of the US), all of those words are identical except for the initial consonant cluster.

    @Eric. One way you might be able to pronounce them differently is to think of the stereotypical American imitation of a British person saying "dance". Do you normally use the vowel in that? For me, that is the vowel in "cot", while "caught" rhymes with "bought", "taught" and "fought".

  80. Nathan said,

    July 27, 2010 @ 6:49 pm

    @dw: Yes, AFAICT, bone and born are the same for me except for the r. And just to be clear, my horse and hoarse are homophones. All of these have the same vowel, /o/.

  81. dw said,

    July 27, 2010 @ 7:12 pm

    @Nathan:

    So a split "caught" would be phonetically halfway between your "cot" and your "coat".

    Sorry — I know that's not helpful :)

  82. Sili said,

    July 27, 2010 @ 7:21 pm

    Golly – I've read about the pen/pin merger before but always assumed that it meant people were pronouncing "pen" like "pin," not vice versa as your anecdote would seem to indicate.

    errrr …

    I'm not an expert, but venturing a guess I'd say that a pin-pen merger implies a kin-ken merger, so your 'observation' doesn't make any objective sense.

  83. Eric said,

    July 27, 2010 @ 7:29 pm

    Hey academic linguists, I have a nerdy question. I assume that in phonetics "field research" or whatever, lots of scenarios have several investigators listen to a speaker, make independent IPA transcriptions, and then check their transcriptions against each other. And then when the various transcriptions show some level of convergence, that's taken to be the correct phonetic description of the speech. But are there ever scenarios where the results of the investigator's transcription is checked, not against the transcriptions of other listeners/investigators, but against the speaker's own belief about her pronunciation? As someone who merges like 90% of the pairs mentioned in this thread, I'm interested in pushing a radically skeptical line: that speakers are often subjectively convinced they make a phonetic distinction (like Mary v. marry) which objective investigation would dis-confirm…

    @Matthew K.: That's funny b/c I can totally imagine John Cleese or Ian McKellen or Captain Jean-Luc Picard pronouncing cot-caught distinctly. I just can't mentally take that distinction, make it come out of the mouth of an American, and make sense of the resulting accent. In my mind, such an American would sound like they were pretending to be upper class English. I mean I get that I'm wrong but I just can't produce the distinction in my imagination.

  84. Jason L. said,

    July 27, 2010 @ 7:41 pm

    @Eric and others,

    I grew up with the hoarse-horse and the caught-cot mergers, but living in the U.K. and Boston for most of my adult life has undone some of the C-C merger (unscientific self-examination leads me to believe that the merger is weakest when I've been drinking, am uncomfortable, or am talking with people of my parents' generation). IANAL, but for me, the /or/-slash-/ɔr/ of North American English is, phonetically and phonemically, not equivalent to /o/ +/r/ or /ɔ/ + /r/ in the way that, say, /ɔt/ or /ot/ is simply /ɔ/ + /t/ or /o/ + /t/.

    For North American English speakers without the cot-caught merger who don't diphthongize ɔ, my impression is that "sore" consists of the s sound plus the or sound, while "sower" (one who sows) consists of the s sound, the long o sound , and r-colored schwa / syllabic r. "Sawer" (one who saws) is like "sower" except the central sound is the aw sound rather than the long o sound.

    This implies that you may not have the best of luck trying to hear and produce [ɔ] by analogies that involve words like "court" and "born". You may come phonetically close if you can sit on the vowel before the r starts influencing anything, but it can be hard to know what that point is if /or/ has to be treated as a single unit as opposed to ɔ/o + r. An alternative is to try to slowly transition between "cot" and "coat", especially if you can do a Minnesota/Ireland-style monophthongal /o/.

  85. Jason L. said,

    July 27, 2010 @ 7:45 pm

    @Nathan,

    Yes, AFAICT, bone and born are the same for me except for the r. And just to be clear, my horse and hoarse are homophones. All of these have the same vowel, /o/.

    Is the vowel in your "mow" and "more" the same? What's the difference between your "more" and your "mower"?

  86. Jason L. said,

    July 27, 2010 @ 8:10 pm

    @Eric,

    Here is a clip of Tom Brokaw where he says "Washington" (caught vowel) at 0:09, and "colony" (cot vowel) at 0:24: http://www.youtube.com/watch?v=bV_041oYDjg

    And here is George H. W. Bush announcing the start of the Gulf war. At 0:09, he says "conflict" followed w/i a second by "August": http://www.youtube.com/watch?v=IFrnQHaQWoA.

    Of course, perceiving two sounds as being pronounced "distinctly" for you may require a larger distinction that Brokaw or Bush are making. To be sure, "BBC English" distinguishes the two by both height and length, while GWH Bush, at least, seems to have no longer a /ɔ/ in "August" than he does a /ɑ/ in "conflict". I also get the feeling that the height difference between the BBC's "caught" and "cot" is greater than, say, Tom Brokaw's, perhaps because its /o/ is really [əʊ], so that leaves more space in the mid-back region for /ɔ/ to move to. In fact, when I affect a "British accent", I find my /ɔ/ is indeed raised so that, subjectively at least, it feels closer to my American /o/ than it does to my American /ɑ/, while I feel that my American /ɔ/ is closer heightwise to my American /ɑ/ than it is to my American /o/.

  87. Nathan said,

    July 27, 2010 @ 8:11 pm

    @Jason L.: Again, I'm pretty sure the vowel of mow and more is the same: /o/.
    mow has one syllable: /mo/.
    mower has two syllables: /ˈmo.ər/.

  88. Tim said,

    July 27, 2010 @ 8:11 pm

    In other words (assuming you have the horse-hoarse merger):

    * "cot" sounds like "cart" without the R-sound.
    * "caught" sounds like "court" without the R-sound.

    I distinguish cot and caught, and not horse and hoarse, but these examples seem wrong to me. To me, caught and cart have approximately the same vowel, and court is nowhere near either cot or caught.

    Then again, there may be something unusual about my cot/caught distinction, because, whenever I read anything about the merger that uses IPA to explain it, the symbols used to represent the distinction don't seem to match my pronunciation. Unless my understanding of IPA vowels is way off (which it might be), I think I pronounce cot and caught as /kat/ and /kɑt/. Does that seem strange?

  89. Nathan said,

    July 27, 2010 @ 8:16 pm

    And somehow I failed to answer Jason L.'s question.
    more also has one syllable: /mor/.

  90. John Cowan said,

    July 27, 2010 @ 8:22 pm

    Eric and others: I think the best trick for learning to undo the cot-caught merger is to appeal to paralinguistic sounds, which often disobey the normal phonology of one's own dialect. Even people with the full version of this merger normally distinguish Ahhhhh!, the sound of satisfaction, from Awwwww!, the sound of pity (possibly ironic). The first of these is the cot sound in the rest of North America, excepting Eastern New England and the other areas that don't have the father-bother merger; the second is the caught sound. So just put a /k/ on the front and a /t/ on the back of either of these sounds, and there you are!

    Jerry Friedman: I think merged and unmerged are about 50-50 in North America, and since Canada is all merged, that means slightly less than 50% of Unitedstatesians have the merger.

    Greg Morrow: The stressed vowel of Mary is [eɪ] for me, the FACE vowel, which is also what I hear people with Mary-marry-merry and Mary-merry-but-not-marry mergers saying. For me and other non-mergers, merry is /ɛ/, the DRESS vowel, and marry is /æ/, the TRAP vowel.

    Mollymooly: Good point; wait-weight then. I use FREIGHT for the latter lexical set.

    James Kabala: Ken and pin are merged for people with the pin-pen merger because it is also a kin-Ken merger. Typically the merged sound sounds like pin, kin, but it doesn't absolutely have to. Similarly, Western Americans typically make caught sound more like cot, whereas Eastern New Englanders and Canadians typically make cot sound more like caught, but this rule too is not absolute.

    As for those steenking or steenkeeng badges we don't need to show you, no variety of English contrasts the KIT and FLEECE vowels before /ŋ/, and so some people use their KIT vowel (which is the historic form) and some use their FLEECE vowel either consistently or inconsistently. For me, the famous line has to be transcribed steenkeeng to get the Spanish-style vowel qualities right, but for plenty of people steenking is sufficient. The same story applies to DRESS versus FACE before /ŋ/. However, there is no KIT-DRESS (or FLEECE-FACE) merger in this context, special cases like English notwithstanding.

  91. Jason L. said,

    July 27, 2010 @ 8:26 pm

    Nathan, thank you for indulging me. If you delete the first syllable of "mower", what do you get? If you remove from "more" all the sounds it has in common with "mow", what do you get?

    When saying "sore eye" casually (that is, at a level of casualness such that "I can see them all" and "I can see the mall" would sound the same), is it different from "sow rye"?

  92. Jason L. said,

    July 27, 2010 @ 8:31 pm

    @Tim,

    That sounds like the Northern Cities Vowel Shift.

    Some writers distinguish between slashes and brackets when using IPA: slashes are phonemic while brackets are phonetic. So you pronounce what would be standardly written (in an American context) as /kɑt/ and /kɔt/ as [kat] and [kɑt].

  93. Jason L. said,

    July 27, 2010 @ 8:49 pm

    @John Cowan,

    Isn't the lack of a contrast between the KIT and FLEECE vowels before /ŋ/ historically a matter of phonotactics? That is, historically at least, /i/ + /ŋ/ was simply not found in English, just as word-initial /s/ + /p,t,k/ is not found in Spanish. Yet Spanish speakers would not deny that [espaɲol] and [spaɲol] are constrastive, just as you imply that steenkeeng and stinking are contrastive for you even if for you the former is not a well-formed English word.

  94. Tim said,

    July 27, 2010 @ 8:49 pm

    Jason : Well, I've always lived in the St. Louis (Missouri) area, which apparently has been influenced to some extent by the NCVS (if the Wikipedia's map is to be believed), so maybe that explains it.

  95. Peter said,

    July 27, 2010 @ 9:04 pm

    My wife and I have had heated debates about the cot-caught merger. I see and pronounce the difference. She doesn't. We've also had some discussions involving a merger of "caulk" and "cock" [insert your own joke here], where she merges the sounds but I do not.

    As an aside, it's surprisingly hard to slow-pronounce "caulk" to be distinctive from "cock". I usually make the L sound too distinctive, like "calc" but with with "A" pronounced like "aunt" (but not the "ant" aunt pronunciation). In normal use, I pronounce "caulk" more like "caw" (crow sound) with a "K" sound at the end.

    (If discussion "cock" and "caulk" is too ribald, you can substitute with "chock" and "chalk".)

  96. Chris D said,

    July 27, 2010 @ 9:08 pm

    @Eric

    But are there ever scenarios where the results of the investigator's transcription is checked, not against the transcriptions of other listeners/investigators, but against the speaker's own belief about her pronunciation? As someone who merges like 90% of the pairs mentioned in this thread, I'm interested in pushing a radically skeptical line: that speakers are often subjectively convinced they make a phonetic distinction (like Mary v. marry) which objective investigation would dis-confirm…

    In teaching the transcription section of an introductory linguistics course, I once had a couple of students who were *convinced* that in they produced a different final vowel in words like "match" whose spelling ends in "-tch" and, say, "coach" whose spelling ends with a "-ch". The students argued that the pronunciation of "watch" contained a "t" (or rather, [t]) that "coach" lacked. They smirked incredulously when I asserted that they were wrong (their reaction is what I imagine a biology lecturer suffers from a stubborn creationist student), and proceeded to produce hyper-articulated contrasting pronunciations in which they inserted their hallucinated consonant before the word-final affricate of "watch" (that is, they produced some sort of geminated affricate at the end of "watch", while keeping the closure duration shorter for "coach").

  97. AJD said,

    July 27, 2010 @ 9:23 pm

    Eric asks:

    I'm interested in pushing a radically skeptical line: that speakers are often subjectively convinced they make a phonetic distinction (like Mary v. marry) which objective investigation would dis-confirm…

    Actually, if anything, the opposite is often the case: speakers are subjectively convinced they have a merger but objective investigation finds they maintain a distinction between two phonemes. This is called a near-merger; it's related to a phenomenon called the "Bill Peters effect", after a speaker who pronounced "cot" and "caught" differently all the time except if you asked him whether they were the same or not; when asked, he would say they were the same, and then go on to say them the same way too.

  98. Jason L. said,

    July 27, 2010 @ 9:31 pm

    @Peter,

    In my Oregon Trail-playing youth on the West Coast, I recall a non-caught/cot-merged teacher say "caulk the wagon and float it across", which we all thought was funny, because we heard her say "cock" while we all pronounced the l. The same teacher, in P.E. class, chided me as a pitcher for balking, which I heard as "bocking", but then when I looked up the word later (yes, nerd alert, I didn't know jack about baseball but took it upon myself to look up new words in the dictionary), I began pronouncing it with the l, even though it was spelled like "talking" and "walking" which had no l sound and rhymed with "shocking".

    For you, is there any l in "caulk" at all? Does it not rhyme with "hawk"?

  99. Garrett Wollman said,

    July 27, 2010 @ 10:13 pm

    Is there a definitive list of these vowel mergers? I find that I have Mary/merry/marry, father/bother, and horse/hoarse, but not cot/caught, and flicker com/calm.

  100. Rodger C said,

    July 27, 2010 @ 10:42 pm

    I pronounce the L in "caulk" just as I do in "yolk" (see elsewhere), but as it's not a word I normally use, this might be a spelling-pronunciation.

  101. Nathan said,

    July 27, 2010 @ 10:59 pm

    @Jason L.: Your last pair of questions really made me think. It revealed more detail than the broad pseudo-phonemic transcription I've been giving.

    There is a difference between my sore eye and sow rye. My vowel /o/ is usually realized as a diphthong [ow], but it seems to me there's less of an off-glide in sore eye and more than in sow rye and mower. I didn't expect something like that.

  102. Mark F. said,

    July 28, 2010 @ 12:06 am

    Eric – I'm curious – did John Cowan's comparison of "ahhh" and "awww" help at all?

    Also, I notice the "caught hot coffee" US examples don't any solidly southeastern examples, which is a good place to find unmerged cot/caught. Anyway, the cot/caught merger is really common in the US, and it's characteristic of the midwest and California, which are the most like what a lot of people see as a "neutral" American accent. On the other hand, the "Western speaker with influences from all over America" makes the distinction at least weakly, I think.

    I'm not a linguist, but I really don't think they routinely have multiple transcribers for one speaker. I don't think the field is that well funded. Anyway, certainly people can be fooled by whether they make a phonetic distinction, but I think the existence of the cot/caught distinction in many dialects of English, even in the US, is pretty well established.

    But, OK, now it's time for me to be flummoxed. What's this about an "ate-eight (or wait-weight) merger"? Do some people pronounce them differently? Who? How?

  103. Skullturf Q. Beavispants said,

    July 28, 2010 @ 12:13 am

    By the way, I generally consider myself more hip to these linguistic matters than the average person on the street, but I'm pretty sure I've never met a human being who pronounces "horse" and "hoarse" differently.

  104. Joyce Melton said,

    July 28, 2010 @ 12:18 am

    Not mentioned here is the put/putt merger that everyone in my family seems to have. Most of us are from Arkansas or Missouri or raised by people who were from there and say put as putt, which seems to crack up most Californians.

    Then again, we can make fun of them for not being able to tell the difference between don and dawn.

    We've almost all learned the difference between pin and pen and can make it at will but generally don't unless we think about it. Or want to mess with people and ask for a straight-pen or an ink-pin.

  105. Tim said,

    July 28, 2010 @ 12:23 am

    As I sit here reviewing my own pronunciation of the l in words like the aforementioned caulk or yolk, I'm coming to the conclusion that I'm consistent, but under a weird set of rules. I do not pronounce it in caulk, or rhyming words such as balk, talk, walk, stalk, etc. However, I do pronounce it after the same vowel [ɑ] when it's before before [m], e.g. calm, balm, palm, etc. From that, you'd think it's the following consonant that makes the difference, but I also pronounce it before [k] when the preceding vowel is [o], rather than [ɑ], e.g. yolk, folk, etc.

    I'm not sure whether this pattern is common around here, or if it's just me.

  106. Charles in Vancouver said,

    July 28, 2010 @ 12:36 am

    @John Cowan: With regards to undoing cot-caught… I'm from Toronto and I've been in Vancouver 4 years. When I take a satisfied "Ahhhh!" and I try to use it to pronounce "cot", it sounds like a vowel I never use in English, but matches pretty closely the vowel I use for French "aller", "sera", "va", etc. If anything it's far too much like the way I prounounce "cut". The "Awww" sound used in "caught" does strike me as an overemphasized way to avoid ambiguity with "cot" (and matches the manner in which Lady Gaga has been "caught" in her bad romance), but my normal cot/caught pronunciation is intermediate and does not seem to be the same vowel I'd use for either "awww" or "ahhh".

  107. Jason L. said,

    July 28, 2010 @ 12:40 am

    @Nathan,

    Hmmm, your response has in turn made me think. A cursory Googling shows this 30-year-old paper, which suggests that in North American English, post-vocalic /r/ and /l/ are intermediate between belonging to the nucleus and the coda. Wikipedia's article on "r-colored vowel" cites a source that asserts that "start" and "north" have as their nuclei r-colored vowels, but says that the r-coloring may arrive later than the initial sounding of the vowel.

    This to me suggests that your "sore eye" is [so˞aɪ] (the "˞" indicates r-coloring of the previous vowel) while your "sow rye" is [soʊɹaɪ]. To the extent that there's an offglide in the former, it's subsumed into r-coloring. What's your "sower I" like? My guess is that it would be [soʊɚaɪ], with perhaps the ʊ disappearing and the only difference between "sore eye" and "sower I" being that the r-coloring in "sower I" is delayed until the vowel has moved from [o] to something more central, while the r-coloring in "sore eye" starts sooner.

    Incidentally, I had thought all along that the (pre-r-colored) vowel in my /ɔr/ was a bit more open than my "normal" /o/, but after repeating all these phrases I'm no longer convinced, and I suspect that the r-coloring and offglide status led me astray. I always felt, psychologically, however, that /ɔr/ was a sound of its own, so it would make sense to say "the OR sound" in a way that saying "the OAT sound" didn't. Unless your "sower I" turns out to be identical to your "sore eye", this supports the idea that /ɔr/ is at least to some extent a separate phoneme from /o/-slash-/ɔ/ + /r/ as well as from /o/ + /ɚ/.

    The other alternatives are that the presence or absence of the ʊ offglide in the GOAT vowel is phonemic, or that /ɚ/ is really two phonemes, syllabic r-colored schwa and non-syllabic r-colored schwa (it won't suffice for there to be syllabic and non-syllabic /r/ — then "sore eye" and "sow rye" would both be /soraɪ/).

    Chances are, however, that my amateur sophistry is just that.

  108. Jason L. said,

    July 28, 2010 @ 12:56 am

    @Tim,

    I find that I pronounce the l consistently before /m/, and inconsistently before /k/, for both "-a(u)lk" and "-olk" words. "Yolk" and "folk" are /yok/ and /fok/, but "Polk" is /polk/. "Talk", "walk", "stalk", and "chalk" are /-ɑk/, but "balk", "caulk", and "Faulkner" are /-ɑlk/, with the /ɑ/ realized more like [ɔ]. The same raising and rounding of ɑ to ɔ happens before /m/. Of all these words, it may be interesting to note that the ones with the /l/ are all words I was old enough to remember learning (I don't specifically recall the occasions when I learned "Polk" and "Faulkner", or which Oregon Trail adventure was where I learned "caulk", but it was after I was old enough to start remembering learning words in general), and the ones without the /l/ are from time immemorial.

  109. Taylor Selseth said,

    July 28, 2010 @ 12:59 am

    I have the Caught-Cot merger and I though everyone said them the same until I got interested in English dialects. I tend to percieve the Caught vowel as either the Cot vowel or the Coat vowel.

  110. dw said,

    July 28, 2010 @ 1:04 am

    @Garrett Wollman

    Is there a definitive list of these vowel mergers?

    Not definitive, but you could try this Wikipedia page:

    http://en.wikipedia.org/wiki/Category:Splits_and_mergers_in_English_phonology

  111. dw said,

    July 28, 2010 @ 1:16 am

    @Skullturf Q. Beavispants

    I'm pretty sure I've never met a human being who pronounces "horse" and "hoarse" differently.

    The preceding comments on this page surely ought to make you dubious of claims of the kind "I've never met a human being who pronounced X and Y differently", where X and Y are merged in your own speech. How can you be confident you would notice if they were pronounced differently?

    Anyway, here's the Wikipedia page: http://en.wikipedia.org/wiki/Horse-hoarse_merger#Horse-hoarse_merger

  112. Jerry Friedman said,

    July 28, 2010 @ 1:49 am

    @Tim: Pronouncing calm etc. with an /l/ is very common in the U.S., and either it's increasing, or I'm suffering from the recency effect. I don't think I've heard yolk with an /l/, but I have heard folk that way. Never talk, etc.

    @Mark Etherton: Thanks for the correction.

    @John Cowan: Thanks for the statistic.

    @Garrett: John Wells's The Accents of English would probably help.

    Okay, you've all been waiting for an amateur take on the awe-or problem:

    For the great majority of English speakers (but not so much in Ireland, Skullturf), horse and hoarse are merged, though they originally had /ɔ/ and /o/ respectively. (That's the end of the part I'm sure of.) So phonemically, you can represent that sound with either symbol. That doesn't mean that either caught or coat sounds like court without the /r/; the /r/ gives you a different allophone of the vowel. Likewise you can consciously think of the sound either way.

    The merged vowel sounds pretty much the same to me in most British and American accents I've heard, and I suspect it's the original "hoarse" vowel. I suspect the original "horse" or "caught" vowel was more open and farther forward, in short, more like mine and Jason L.'s. For instance, juggernaut and nautch (dance) use "au" to represent a vowel that I think is close to (or exactly) [ɑ] and by no means so close to modern British "au" sounds. If this is right, then what happened in Britain is that not only horse but also haw moved to the hoarse vowel.

    This happened in New York and parts of the Southern U.S., too, but not in most of the rest of the U.S. (New England confuses me.) Here horsemoved toward /o/ and got the hoarse vowel, but /ɔ/ not before /r/ moved the other way, more open and less rounded, in the direction of /ɑ/, or for about half of us now, merging with it. Therefore many of us are just as baffled by the British and New York idea that court has the caught vowel as the British and New Yorkers are by the idea that court has the coat vowel.

    As evidence for this, I at least sometimes say certain words with my /ɔr/, which is not my "or" sound. Centaur, Taurus, Bryn Mawr (but the first syllable of laurel is like lore, and don't get me started on dinosaur). It's the Taurus-torus split. I don't think this is all that rare in America.

    So if [ɔ] means the vowel in RP and my "or" (pretty close, I think), then what's the vowel in my centaur and caught? I asked John Wells, and he kindly responded at his blog. Among his suggestions were [ɔ̞] (more open than the RP version) and [ɔ̜] (less rounded).

    So, finally, the summary. I think I have a long-o phoneme realized as [oʊ] in most environments and as [ɔ] before /r/ in almost all words spelled with "or", "oar", etc.; and an "aw" phoneme realized as [ɔ̞], including before /r/ in a few words. And I think this is fairly typical of caught-is-not-cot Americans outside the northeast and some places in the South.

    Apologies for the amateurism, but I'm hoping for expert correction. And maybe I'd understand all this better if I'd ever figured out Trager-Smith.

  113. Eric (The First One) said,

    July 28, 2010 @ 2:35 am

    Is this a third Eric, or the second one?

  114. Cialan said,

    July 28, 2010 @ 3:48 am

    Jim wrote:
    3. I fell in love with both Italian and German. Although certainly not fluent in either, my pronunciation of words in either language is almost flawless.

    I think it might be easier to work on one's pronunciation in a second language (L2) than in a first language (L1) because it doesn't feel like you're losing your social identity by changing your L2 pronunciation, as it would in your L1. You're just learning the new language more thoroughly when you work on your pronunciation. Also, if you have not already established pronunciation habits in the L2 different from the ones that you are trying to achieve, it of course makes it easier. Having to learn a different style of pronunciation in one's L1 feels wrong socially and is further complicated by the fact (already mentioned by others) that we often think we're pronouncing things differently than we are.

  115. Eric said,

    July 28, 2010 @ 4:20 am

    dw said [said?]: "The preceding comments on this page surely ought to make you dubious of claims of the kind "I've never met a human being who pronounced X and Y differently", where X and Y are merged in your own speech. How can you be confident you would notice if they were pronounced differently?"

    Whoa. Paging Bishop Berkeley.

  116. Eric said,

    July 28, 2010 @ 4:38 am

    Jerry Friedman said: "As evidence for this, I at least sometimes say certain words with my /ɔr/, which is not my "or" sound. Centaur, Taurus, Bryn Mawr (but the first syllable of laurel is like lore, and don't get me started on dinosaur). It's the Taurus-torus split. I don't think this is all that rare in America.

    Whoa. Paging Auguste Comte.

    Seriously, stop doing this. The one point I really want to issue from on high here is that you cannot use your "own" speech as evidence when it comes to phonetics. It's bad bad bad empiricism. If you want to know how you say things, ask lots of other people how you say them.

    Also, Turing. Page him, too.

    (second out of only 2 Eric's, so far as I can tell)

  117. Ken said,

    July 28, 2010 @ 5:23 am

    @Eric: "But with cot-caught, I just don't even know what the vowel in the 2nd word is supposed to sound like. What's a word with that sound that I can use to get the distinction by analogy?"

    To my English ears the way most Americans say "caught" seems fine. What you have lost is the "short o" sound in "cot", which is I think /ɒ/ in IPA. Even Americans who don't merge "cot" and "caught" seem – to the rest of us – almost to say "cat" instead of "cot", or to use the vowel from "car". Those who do merge seem to have made "cot" sound like "caught" – I guess that is why there are joke spellings like "dawg" and "rawk" [music] – which must have been introduced by Americans who don't merge those vowels as a comment on those who do.

    But if you want online guidance as to how they are pronounced in different accents, Google for "lexical sets" and follow the links. Or look at the wikipedia page on "IPA chart for English dialects"

    Not much point in giving sample words because if I say "cot" rhymes with "hot", "spot" and "lot" you will just reply "of course it does" and say what sounds to me like "hawt", "spawt", and "lawt". And if I say that "caught" rhymes with "fought" and "bought" it will for you to. (Though maybe nor with "port" or "tort" which for me rhyme with "caught")

    @SeanH:"@Josh, for Cockney, I think what the 'th' turns into depends on where it is in the word. "Shut your mouf", for instance, but also "dahn ver pub"."

    That's unvoiced vs. voiced.

    Not just Cockney – almost all south-eastern or eastern-English speakers (including RP) would have /f/ instead of/θ/ at the end of "mouth" in rapid or casual speech. I suspect that many Americans would. Though if you asked them to repeat it they would probably take care to say the /θ/ I think /v/ for /ð/ in "the" in a phrase like "down the pub" is probably much rarer.

  118. ElizMo said,

    July 28, 2010 @ 5:36 am

    A similar example — I'm an NZer who lives in London in a house with multiple Canadians.

    Bear, bare and beer: apparently these words are pronounced differently; however, I can't really hear a difference, and I apparently can't say them correctly, and the Canadians find this quite amusing.

    I argued that there was no need to have a difference as you'd never have a situation where you'd have confusion between the words… till I came across the "Bear Garden" in the Royal Courts of Justice, which was not the delightful drinking hall for lawyers that I was hoping for.

  119. Victor Mair said,

    July 28, 2010 @ 6:34 am

    @ Diane

    When I was in Nepal, the Nepalese would alway burst out laughing when I counted from 1 to 6 because, first of all, I couldn't distinguish between 4 and 6, but — worse than that — I made one of them sound like "fornicate." I tried very, very hard to avoid being laughed at, but just couldn't get 4 and 6 right.

    ===

    Back to pin, pan, however, I used to drive my wife Li-ching nuts by helping her practice English vowel distinctions: pin, pen, pan, pine, pain, pun, pone, poon. She could only catch about half of them when listening to me, and could accurately produce even fewer. We both had a lot of chuckles over those exercises.

  120. Jon Lennox said,

    July 28, 2010 @ 6:57 am

    @ElizMo: "beer" is definitely distinct for me from the other two (I'm from western Massachusetts, but with midwestern parents) — this is the distinction which is canonically represented as "cheer" / "chair".

    However, I would have said that "bear" and "bare" are homophones. Are there people here who distinguish these? Is this "marry" / "Mary" / "merry"? (I may have that distinction very weakly — I can imagine myself distinguishing them in careful speech, but as mentioned here it's easy to fool yourself, and it's not phonemic for me.)

  121. Victor Mair said,

    July 28, 2010 @ 6:57 am

    @Diane (again)

    I'm surprised that your Nepalese acquaintances had trouble distinguishing between S and SH, since they have *three* different S's, two of which sound to English speakers like SH, and one which sounds to English speakers like S. Perhaps it's the fact that their two SH's don't map onto our one SH that throws them for a loop.

    See the column with the S's (seventh column of the section on consonants) here:

    http://www.omniglot.com/writing/nepali.htm

  122. Mark P said,

    July 28, 2010 @ 8:20 am

    I can conceive of differing pronunciations for almost every set of words that has been discussed here, but I can't figure out how hoarse and horse would be pronounced differently.

    I know someone who distinctly pronounces the l in folk (and presumably other similar words) and it sounds odd, if not affected, to me.

  123. John Cowan said,

    July 28, 2010 @ 9:34 am

    Jason L.: Just so: phonemically there is only /ɪŋ/ in English, never /iŋ/, and for all I knew until about ten years ago, there was only [ɪŋ] and never [iŋ]. But that turns out not to be the case: there are plenty of people who render /ɪŋ/ as [iŋ], either in all words, or in specific words only, or sporadically.

    Charles in Vancouver: Quite right. My advice wasn't really meant for Canadians or Eastern New Englanders.

    Ken: You're right to say that Americans have lost short o (always excepting Eastern New Englanders). However, only a small subset of the short-o (LOT) words have merged with THOUGHT in America, the pesky CLOTH lexical set; the rest have merged with PALM. So to an RP-type ear, Americans have laht, haht, spaht, caht, rather than lawt, hawt, spawt, cawt, and clawth, crawss, cawfee for cloth, cross, coffee. The spelling dawg reflects the usage of people for whom dog is part of CLOTH (and so merged with THOUGHT), as seen by those who make it part of LOT (and so merged with PALM): the words in -og are quite variable in this way.

    Of course, for the large minority who merge THOUGHT=CLOTH with LOT=PALM, the above distinction is entirely unmeaning: for them, all the words mentioned have the same vowel.

  124. John Cowan said,

    July 28, 2010 @ 9:40 am

    Victor: The point is that while Nepalese script, like most Brahmic scripts, has three distinct letters for the Sanskrit sounds transliterated s, ṣ, ś (s, s-underdot, s-acute), Nepalese phonology has only one /s/ phoneme, which may be written in any of the above ways, having merged away the distinction. The same is true of Bengali. As is not uncommon in languages that lack contrast between /s/ and /ʃ/, the /s/ phoneme in both languages may be pronounced more like [ʃ] than [s]: this is also true of Finnish, for example.

  125. John Cowan said,

    July 28, 2010 @ 9:48 am

    Jon Lennox: Bear and bare are distinct in accents, mostly in East Anglia, Southern Wales, and Newfoundland, that lack the long mid (pane-pain and toe-tow) mergers. This merger is pretty old, and has basically run to completion except in these few areas.

    Ken: Th-fronting is basically unknown over here except in AAVE and Newfoundland, where it is usually confined to final position.

  126. Mark P said,

    July 28, 2010 @ 10:13 am

    @John Cowan: I have laht, haht, spaht, cawt, clawth, crawss, and cawfee, although my "cawfee" has changed to something closer to cahfee. I still have "dawg" although probably not as strongly as when I was younger. As far as I can tell, that is normal for my region (NW Georgia, NE Alabama).

    I noticed (and think I might have commented before) that Barack Obama pronounced dog as "dawg" (and quite strongly) when he appeared on a late-night TV show. I think I noted at that time that my spelling of dawg did not convey my pronunciation the way I intended to those who don't pronounce it that way.

  127. Jerry Friedman said,

    July 28, 2010 @ 10:18 am

    @Eric: How about the part where I said I thought a lot of other people made the distinction I was talking about? Do we need a pronouncement from on high about that?

    Anyway, with the magic of the Internet, we can do better. This is a video in which the speaker (who MYL will recognize) says stars, warm, applaud, and Bryn Mawr within the first minute. I think her Bryn Mawr vowel is the same as her applaud vowel and different from her warm and stars vowels, which is the distinction I'm saying exists. I hope others will give their judgements.

    (I listened to some other videos from Bryn Mawr before finding that one. Not one person pronounced Mawr like more, or to rhyme with flour, which I think would resemble the original Welsh pronunciation. They all pronounced it as in the video above or like mar.)

    @Mark P: I've heard only one person, an expatriate Irishman, say horse and hoarse clearly differently. His vowel in hoarse sounded to me almost like mine in tour. (This is, if anything, evidence against part of what I said in my long post.) Maybe someone will search Irish (or St. Louis?) videos for someone with a horse/hoarse distinction.

    On Nepal: There's a student from Nepal at my college. He has a "j" in his name, and I was surprised when his teacher assured me that it was pronounced /z/. Then he asked me a physics question involving a symbol that I heard as z, but it turned out to be g, the acceleration of gravity at the Earth's surface. So how should I pronounce his name?

    @Victor Mair: I hope the word for "six" was the one you made sound like "fornicate"!

  128. kip said,

    July 28, 2010 @ 10:19 am

    @mollymooly

    OK, maybe my pin/pen merger exists before /ŋ/ as well, but it's different. With pin/pen and him/hem, both come out with the short I sound. With penguin/ping, both come out with the long E sound (peeng, peengwin). Going through your list:

    * length, strength: I'm not 100% sure if I pronounce these with long E or short I. Both sound correct to me. But I definitely don't use short e like in "pet".
    * penguin: I definitely pronunce it "peengwin"
    * dengue: I've never used this word, I don't know how I would pronounce it
    * enclave: I do "fancy-French it up" to "ahnclave" but there's no /ŋ/ for me.

    As for the names:

    * Jenkins: identical to "Jinkins" (but there's no /ŋ/ for me here, so it's just pin/pen again)
    * Bengals, Engles, Englebert: "Eengles". ("Engles" rhymes with "shingles" for me.)
    * Genghis Kahn: I thought this was "Gayne-gis", with first syllable like "Wayne" but with a G?

    As for "insure"/"ensure", both start with short I for me. One place where confusion occurred for me in school: immigration/emmigration, which are pronounced the same for me.

  129. Nathan said,

    July 28, 2010 @ 10:32 am

    @Jason L.: I never thought the concept of "r-colored vowel" made much sense in rhotic English, except for the NURSE vowel. I don't ever use that sound except before /r/; it's different from both FOOT and STRUT. So through a combination of pickiness and, likely, ignorance, I would change the ˞s to ɹs in your response and call it good.

    @Ken: You said, "Those who do merge seem to have made 'cot' sound like 'caught'". But that's exactly the opposite of what I've been thinking about the merger. I thought the point was that people without the merger distinguish a rounded vowel in caught from an unrounded vowel in cot, whereas people like me use the same unrounded vowel for both. I thought that was cot. But you seem to be saying you hear Americans merge both to a rounded vowel, right?
    I had assumed "dawg" and "rawk" were created by mergers to talk about the rounded vowel of the non-mergers. Or do those words have /ɑ/? I guess the only way for me to figure out which unmerged vowel a word has is looking it up in a dictionary.

  130. Markus said,

    July 28, 2010 @ 10:41 am

    I am from Germany, my school-english pronounciation of "pen" is nearly indistinguishable from "pan", while "pin" is totally different.

  131. Carin said,

    July 28, 2010 @ 10:54 am

    @John Cowan: Your comments on Pesky CLOTH and related issues get at something that's been bugging me about my own usage, viz. that I have different vowels in dog and log, blog, cog, etc. Can you point to some literature on variable -og words? I'd love to read more about this.

    On pin-pen, I'm interested to read so many accounts of the persistence of the merger in those who have otherwise erased tell-tale regional features. I recently attended a very boring talk by a speaker – a specialist in the history of English – whose accent was so neutral-American that it could have belonged to a Radio Moscow announcer, except for what to my ear was a glaring pin-pen merger. If the talk had been more interesting or the accent less studiedly neutral, I don't think I would have fixated on all those raised vowels before nasals. In light of this discussion, I'm perhaps less astonished that someone who had obviously given so much thought both his accent and the development of English vowels exhibited that one persistent fossil of his native usage.

  132. DRK said,

    July 28, 2010 @ 10:56 am

    When we moved to Illinois, and I went to enroll my daughter, Jenny, in school, the school secretaries spelled it Ginny. I corrected them, and was informed that I pronounce her name wrong! It was years before I even realized what the woman meant, and only a chance posting on a blog (probably Language Log, come to think of it), made me aware of the pin/pen merger. Grew up a military brat with parents from Texas, thus the merger, I guess.

    Mid-westerners seem to find this particularly grating — I certainly never had this problem anywhere but the Chicago area, and my husband was in the military, so we moved around quite a bit. I can hear the difference now, but don't bother to "correct"my own speech. If we all spoke the same way, the world would lose a lot of richness and texture.

  133. Ellen K. said,

    July 28, 2010 @ 11:04 am

    @ John Cowen. In my head, it's phonemically /iŋ/, not /ɪŋ/. I agree there's not a contrast, but I don't think there's any reason to assume universal agreement on which phoneme the i sound is part of.

    [(myl) In cases like this, where there's no contrast, there's also the theory that the right answer is "neither", i.e. that the elements in question are unspecified for the distinction.]

  134. Carin said,

    July 28, 2010 @ 11:08 am

    @Nathan, the answer the second half of your question is buried in John Cowan's July 27, 2010 @ 8:22 pm post, in his answer to James Kabala: "Similarly, Western Americans typically make caught sound more like cot, whereas Eastern New Englanders and Canadians typically make cot sound more like caught, but this rule too is not absolute."

    Question for Canadians, though, prompted by other comments in this thread: do western Canadians have the both-are-unrounded merger, like western USians do, or does all of Canada have the both-sound-like-caught version?

    And a question for the linguists in the room: Is there a reason why it's analytically helpful to talk about the cot-caught merger generally, as we almost always seem to do, without distinguishing whether it's in the cot direction or the caught direction? That is, are the New England-Canada merger and the Western (N?) American merger the same phenomenon, historically? Does it matter if they are?

  135. Qov said,

    July 28, 2010 @ 12:16 pm

    This western Canadian says "caught" and "cot" both identically to "cod," except for the distinction between t and d, which I do usually articulate. I can't reliably answer for elsewhere in the country.

  136. Jerry Friedman said,

    July 28, 2010 @ 12:33 pm

    @Ken: /f/ for /θ/ is a stereotypical part of AAVE. I don't remember noticing it in my childhood, when I heard a lot of AAVE, but what is my word against Ludacris's album title Word of Mouf?

    I think it's very rare among Americans of other races, though oddly enough, I thought I heard it from someone yesterday.

  137. Peter said,

    July 28, 2010 @ 12:52 pm

    @Jason L: Wow, this post has been busy since i commented yesterday.

    > For you, is there any l in "caulk" at all? Does it not rhyme with "hawk"?

    There is no "L" in "caulk" but it does not quite rhyme with "hawk". When I am trying to draw out the enunciation to demonstrate the difference between how I say "caulk" and how I say "cock", a bit of an "L" sound creeps in there unintentionally, making it sound like "cahlc". At the same time, it's not just "caw" (says the crow) with a "K" at the end either. It's more like "cahwlk" (sorry I don't have the vocab to describe more precisely what I mean).

  138. Boris said,

    July 28, 2010 @ 1:03 pm

    My pronunciation may be colored by the fact that I'm originally from Russia (as I mentioned above), and of course, this is all subjective, but I've been living in various parts of New Jersey (first Philadelphia influenced areas then New York influenced) and pronounce Mary and Marry the same and merry differently from both (and Murray differently from the others). I haven't seen this variant discussed yet.

    I don't have the mirror/nearer merger or the furry/ferry merger. I do have the horse/hoarse merger

  139. Ellen said,

    July 28, 2010 @ 1:11 pm

    Mixed in with all this, our perception of what phoneme a certain sound is is affected by the sounds surrounding it. I can reliably tell a flapped r from a flapped d/t, even though they are the same. The accent of the speaker, and recognizing the word make the difference.

  140. Diane said,

    July 28, 2010 @ 1:16 pm

    @Victor Mair

    It may be relevant to point out that I learned Nepali mostly from people whose native language was not Nepali (mostly Dotili, a few Tharu, and a smattering of other languages of the western hills of Nepal). So I could be wrong about how Nepali words are properly pronounced.

    That said, my husband, who *is* a native speaker of Nepali, definitely can produce differences between S and SH much better than he can hear them.

  141. Tim said,

    July 28, 2010 @ 1:32 pm

    Maybe someone will search Irish (or St. Louis?) videos for someone with a horse/hoarse distinction.

    As a St. Louisan for all of my 29 years, I've never noticed anyone make a distinction between horse and hoarse, but it's not like the words appear together all that often. I may need to start listening for it. I know that I pronounce them the same.

  142. Morgan said,

    July 28, 2010 @ 2:18 pm

    "Mary and merry rhyme with berry"

    @Xmun: "No they don't. "Mary" rhymes with "scary" and "hairy". Do those words also rhyme with "berry" in your speech? And are "fairy" and "ferry" homophones?"

    Of course "Mary" and "merry" rhyme with "berry". And they all rhyme with "marry". And "fairy", "ferry", "scary", "Harry", and "hairy". And "wary". And "very", and "Carey", for that matter. As well as "Derry", "dairy", "cherry", and "extraordinary".

    Funny thing is, I have no problem hearing the distinction other people draw between "Mary", "merry", and "marry", and might even be able to reproduce it if required. But those words just aren't pronounced that way.

  143. Ran Ari-Gur said,

    July 28, 2010 @ 2:26 pm

    @Bill Walderman, wrote "Confusion between 's' and 'f' or 'th' is less frequent [than between 'f' and 'th'] but sometimes occurs, for me at least": I have friends named "Jess" and "Jeff", and even though the names sound very different in person, I've learned from experience that they're indistinguishable over digital cell-phones, with "Jess" getting heard as "Jeff". No one ever asks, "Did you say 'Jess', or 'Jeff'?" — they just know, wrongly, that it was "Jeff" — and the error is usually only detected once a pronoun gets used and suddenly they hear that "Jeff" is a woman.

    [(myl) The limited bandwidth of telephone speech (typically low-passed at about 3,400 Hz) means that the (normally robust) cues to the difference between /s/ and /f/ are mostly missing.]

  144. Jason L. said,

    July 28, 2010 @ 2:33 pm

    Jerry Friedman: I think [/f/ for /θ/] is very rare among Americans of other races, though oddly enough, I thought I heard it from someone yesterday.

    I recall a Daily Show clip (ah, here it is!) where a White South Carolinian says "troof" for "truth".

  145. Charles in Vancouver said,

    July 28, 2010 @ 2:58 pm

    My perception of Canadian accents may be a bit biased but I feel like all L1 anglophones from *cities* stretching from Vancouver to Montreal tend towards a relatively similar accent – maybe the CBC has had something to do with this. Once you pass through Quebec and enter Atlantic Canada, you're in a totally different zone (ask a Haligonian or a Newfie to pronounce "cod" and "car"!). Rural Canadians sometimes have slightly more idiosyncratic inflections, which are noticeable to the urban ear.

  146. Keri said,

    July 28, 2010 @ 3:11 pm

    I have the mary/merry/marry merger. They all sound identical to me. And rhyme with my name, Keri =)

    Once I introduced myself to someone who misheard me at first. When I repeated my name, she said "Oh, KERI. Your accent confused me." You can't have an accent in saying your own name! Your name is whatever you say it is, and if someone says it differently, they are saying it wrong! Grr. That made me angry. Plus, at that time, I had never heard of the mary/merry/marry merger and had no idea what she was talking about. I never thought I had an accent (I'm from Connecticut).

    For those who have no idea how these words are pronounced differently: As I understand it, those who have not merged these sounds pronounce Mary the same as mergers do, but the e in merry sounds like the e in pet, and the a in marry sounds like the a in mat. Well, I don't think that would help those who also have a pin/pen merger, which I don't have. I can pronounce merry/marry differently if I try, but it sounds very hoity-toity British to me. =)

  147. Carin said,

    July 28, 2010 @ 4:33 pm

    @Qov – but does your "cod" have the Awwwwww vowel or the Ahhhhhhhh vowel?

  148. David Costa said,

    July 28, 2010 @ 5:21 pm

    In the SF Bay Area, it's mostly speakers of AAEV who merge 'pin' and 'pen' — and pens are all called 'ink pens' (or ink pins) to overcome this homophony.

    I merge 'merry', 'marry ' & 'Mary', and it wasn't until I got to college that I realized that ANYONE distinguished them. Later when I was a TA for beginning linguistics, I would ask my undergrads how they pronounced the terms. Considering that the students came from all over the US (and the world) the answers were very entertaining. There were students who merged them all, ones who kept them all separate, ones who merged the 'top two', but kept 'marry' separate, and ones who merged the 'bottom two', but kept 'Mary' separate. The only common thread was complete incredulity that anyone anywhere pronounced any of them differently.

  149. Nick Z said,

    July 28, 2010 @ 5:27 pm

    It seems otiose to add to this already immense discussion, but the mentions of 'caulk' took my eye. I had always assumed the /l/ was pronounced, but having now looked at OED I see that this was obviously a spelling pronunciation. My real surprise is that so many people in these comments apparently regularly pronounce it at all. Is caulking such a common topic of conversation in America? As a non-boat owning/constructing Englishman I do not think I have ever heard it said or said it myself.

  150. David Costa said,

    July 28, 2010 @ 5:38 pm

    "My real surprise is that so many people in these comments apparently regularly pronounce it at all. Is caulking such a common topic of conversation in America?"

    Dunno how 'common' it is, but it's also the normal verb to use for patching holes in bathroom walls and such.

    [(myl) Also fixing holes that let heat out and cold in in (or the reverse) — which is why the tax-credit program for such things is sometimes called "Cash for Caulkers", in imitation of "Cash for Clunkers". This kind of caulking matters quite a bit, unless you live in California. Actually, it matters there too, but Californians like to pretend that it doesn't, which is part of the reason that Mark Twain should have said "The coldest winter I ever spent was a summer in San Francisco".]

  151. Don Killian said,

    July 28, 2010 @ 6:03 pm

    Hrm, regarding American vowels, I was just wondering how many American English speakers actually round their lips when saying what is traditionally labelled as ʊ, such as in put or book? I might be a bit of an outlier here but I don't think so, and I don't even have partially rounded lips for that.

    (And on a side note, why don't we actually have an IPA symbol for the unrounded version?)

    Also, is the æ/ɛə split really so rare? (halve and have are minimal pairs for me, for instance). No English for foreigners text book ever mentions ɛə as a variant of æ, let alone as a separate phoneme. Even for accents which don't have those as separate phonemes though, I think it's a very common sound in American English, stretching from California to Pennsylvania. Was wondering why the only mention seems to be in sociolinguistic text books.

  152. Mabon said,

    July 28, 2010 @ 6:22 pm

    Re: Mary, marry, merry and Murray.
    In my own speech (originally strong Boston accent, much diluted over my lifetime), they are all different. Mary is pronounced like "MERE-ee" (Bostonian, for sure, with a rhotic-colored vowel ), while marry has a short, flat "a" as in MASS: "MA-ree".
    Merry is "MEH-ree", and Murray is "MUH-ree".
    Note that the "r" ends up in a different syllable between Mary and marry for me, and that Murray isn't even close to the other three.

  153. Victor Mair said,

    July 28, 2010 @ 6:27 pm

    @Keri: "You can't have an accent in saying your own name! Your name is whatever you say it is, and if someone says it differently, they are saying it wrong!"

    There was a famous old professor of Chinese history at Penn named Derk Bodde. When he introduced himself, he would say, "Hello, I'm Derek Bode." People who knew him only through his writings would always do a double-take when they heard him say that. It was customary in the field of Chinese Studies to refer to him as Dirk Bod or Dirk Bod-duh; nobody called him Derek Bode. I knew Derk well and once got up the courage to ask him why he said "Derek" when his name was spelled as "Derk," and he just said, "Right, it's Derek." I didn't pursue the matter.

    Incidentally, I've heard people pronounce my name as "Meier," "Mayer," "Mayair," "Mayear," and other ways, but I've never once "corrected" them. I figure that, after all, my own pronunciation is rather arbitrary, since my ancestors back in Austria would say it very differently from me. However, if anyone asks me how to pronounce my name, and they sometimes do, then I'll point to my head and say, "It's like 'hair.'"

    I've known Americans with French surnames who come from families that pronounce them differently in different branches and places, e.g., Boucher and Naquin.

  154. dw said,

    July 28, 2010 @ 7:05 pm

    @Keri:

    The lady probably thought your name was "Cary", as in "Cary Grant".

    I have experienced Kerry/Cary/Keri confusion a few times (I have the split but live in California where most don't), but I would never be so rude as to say "Your accent confused me" :)

  155. AJD said,

    July 28, 2010 @ 8:00 pm

    Keri:

    You can't have an accent in saying your own name! Your name is whatever you say it is, and if someone says it differently, they are saying it wrong!

    I totally disagree. You have an accent in saying anything, and if someone who has a different accent wants to say the same word as you, they're going to say it differently so as to conform with their own accent. Tony Blair doesn't pronounce a /r/ at the end of his name, but if I said his name without the /r/ it would simply be an incorrect pronunciation.

    David Costa:

    There were students who merged them all, ones who kept them all separate, ones who merged the 'top two', but kept 'marry' separate, and ones who merged the 'bottom two', but kept 'Mary' separate.

    You leave out the ones you merge "Mary" and "marry" but keep "merry" separate! This came up in the class I was teaching just today; one of my students has that configuration. It seems to be the direction in which merger is taking place in the Boston area (which is very disappointing to me, as I'm from the Boston area and only about 10 years older than this student, and I have a robust three-way distinction).

    Don Killian:

    Also, is the æ/ɛə split really so rare? (halve and have are minimal pairs for me, for instance).

    To the best of anyone's knowledge, the actual split into two phonemes only exists in the immediate vicinity of New York City and the slightly broader vicinity of Philadelphia (maybe Baltimore? dunno), as well as apparently some children of New York parents in places like southern Florida. [ɛə] exists as an allophone of /æ/ in a broader set of locations, though—before nasal consonants in much or most of the U.S., and before voiced stops and voiceless fricatives for some speakers in a few cities (Albany, Cincinnati, New Orleans). For speakers with the advanced Northern Cities Shift in places like Chicago and Buffalo, [ɛə] is the only allophone of /æ/.

  156. Jason L. said,

    July 28, 2010 @ 8:02 pm

    Keri: Once I introduced myself to someone who misheard me at first. When I repeated my name, she said "Oh, KERI. Your accent confused me." You can't have an accent in saying your own name! Your name is whatever you say it is, and if someone says it differently, they are saying it wrong!

    While your interlocutor was being indelicate, I disagree with the assertion that the only correct way to pronounce your name is how you in your own dialect pronounce it. Do Tony Blair and George Bush mispronounce each others names because George's "Blair" is rhotic and Tony's "George" is non-rhotic?

    No. Names are like other words. We don't switch into different dialects when we say names of people or things that belong to another dialect. If your name were "Carrie", then a speaker who distinguishes "marry" from "Mary" and "merry" would be correct in pronouncing your name [kæri] rather than [kɛri] or [keɪri].

    In effect, you're asking other speakers to violate the phonology of their own dialects whenever they address you. "Carrie" (and "Keri") is just as much a name in their dialect as it is in yours — you don't own the name. I don't see how this is any different from people who insist that their names be written uncapitalized — in both cases, you're asking people to break their linguistic conventions and make an exception just for you.

  157. Jason L. said,

    July 28, 2010 @ 8:03 pm

    Wow, amazing how AJD and I both picked Tony Blair as an example. . ..

  158. Jon Lennox said,

    July 28, 2010 @ 8:29 pm

    John Cowan: so it therefore sounds very unlikely that ElizMo's Canadian roommates were actually mocking an inability to distinguish "bare" and "bear", but rather the inability to distinguish either from "beer".

    And to be sure, the potential for low humor in this particular merger is indeed substantial.

  159. Carin said,

    July 28, 2010 @ 10:32 pm

    @Don, I have the æ/ɛə split and didn't realize it was such a localized phenom till AJD's comment. (I grew up in DC but with a mother from Philly and dad from NYC.)

    On lip-rounding, though, isn't put without rounding putt and book without rounding buck? I can't figure out how to make the put and book vowels without rounding my lips. Or maybe those aren't the best example words to use to test how much you round your lips, since both start with bilabials, so of course your lips move apart when starting the vowel. How about the vowel of should vs. shudder?

  160. Jason L. said,

    July 28, 2010 @ 11:55 pm

    @Carin,

    The vowel of "put" and "book", ʊ, is higher than the vowel of "putt" and "buck", ʌ. It's like a relaxed version of the GOOSE vowel, whereas ʌ, realized by most Britons and most Americans as something more advanced, like [ɐ], is more like a relaxed version of the vowel in "spa" or "Saab". Depending on your speech, it may lie elsewhere: http://en.wikipedia.org/wiki/Open-mid_back_unrounded_vowel#Occurrence

    IANAL, but you might also try the reverse strategy: start with "put" or "book" and then open your jaw slightly and keep your tongue, especially the back part of it, as low as you can. You can also try starting with "put" or "book" and unrounding your lips while scrupulously keeping your tongue and vocal tract in the same position (that is, making sure you don't relax them), and then contrasting that with "putt" or "buck".

  161. Peter said,

    July 29, 2010 @ 12:38 am

    @Victor:

    Incidentally, I've heard people pronounce my name as "Meier," "Mayer," "Mayair," "Mayear," and other ways, but I've never once "corrected" them. […] However, if anyone asks me how to pronounce my name, and they sometimes do, then I'll point to my head and say, "It's like 'hair.'"

    Funny, but I pronounce "Meier" (as in Sid Meier the game designer) and "Mayer" (as in John Mayer) like "hair". "Meyer", however, I get confused on. Sometimes I pronounce it "Myer" (Hey! Oscar Mayer and John Mayer don't rhyme!) and sometimes I pronounce it "May-er" (or in the final "M" in MGM, I pronounce it like "hair").

    Do you pronounce "mayor" with one or two syllables, and if it's the former, does it rhyme with "hair"?

  162. Victor Mair said,

    July 29, 2010 @ 5:44 am

    @Peter:

    For me, "mayor" has two syllables, so it doesn't strictly rhyme with "hair," "fair," etc.

  163. James Kabala said,

    July 29, 2010 @ 9:39 am

    "I have the mary/merry/marry merger…. I'm from Connecticut."

    While many people criticized Keri's main point, I actually found the identity of her native state to be the most surprising part of her story. I am from Massachusetts myself and always thought all of New England was very much non-merger territory.

  164. Carin said,

    July 29, 2010 @ 10:10 am

    @ Jason L. – Thanks, you're right, ʊ is higher than ʌ. But I'm still skeptical about Don Killian's suggestion that lip-rounding is rare among American speakers when pronouncing book and put & sim. (Happy to be convinced, though. What do others think?)

  165. AJD said,

    July 29, 2010 @ 10:56 am

    James Kabala:

    The merger has been present at least in part in Western New England for a long time: the Linguistic Atlas of New England, which is based on data collected in the 1930s, shows complete Mary-merry-marry merger along the Housatonic in western Connecticut, and two-way Mary-merry merger throughout a lot of the rest of the western halves of Connecticut and Vermont. In Eastern New England the merger is a lot less well-established (but cf. my July 28, 2010 @ 8:00 pm comment above).

  166. Ken Brown said,

    July 29, 2010 @ 11:30 am

    Where does it all stop?

    Presumably there are limits to how many distinct and meaningful vowels a language can have? If we merge them all, we start getting words confused? There must be some minimal number of distinct vowels needed for English to make sense? (Maybe a different number from some other language)

    There must also be a maximal number of vowels that can matter, maybe a phyiological/neurological limit.

    Exagerrating for effect, if I can believe this thread (and Wikipedia) you Americans seem to be losing vowels at the rate of about four or five a century. And we Brits are manufacturing new ones, and new complex dipthngs and tripthongs, by turning "l" and "r" into vowels. (And we are also losing consonants in other places)

    Are Americans making new vowel distinctions somewhere to make up for the ones they are losing? Or is the system really becoming smaller?

  167. cDonna F said,

    July 29, 2010 @ 3:19 pm

    This fascinating conversation has reminded me of a TV program I once saw years ago. I have forgotten many of the details, but I'll describe what I remember in the hope that someone here will be familiar with the experiment described. I apologize in advance for the length and vagueness of my description.
    I believe the program was to do with current research on the development of the human brain from conception through adulthood. It was stated in the show that babies are born with the ability to distinguish between all the speech sounds that humans are capable of producing. But after a certain age (around 18 to 24 months, I think), babies lose the ability to hear, and hence distinguish, sounds that are not used in the language(s) they are usually exposed to.
    To test this, the experimenters fitted babies with electrode caps that would show when the baby's brain responded to a particular sound. Apparently there is a slight but noticeable difference in the brain's response to different speech sounds. Then they would observe the baby's brain's response to two different speech sounds, one used in English (the babies' families' language) and a slightly different sound used in one of the Chinese languages.
    The narrator stated that English-speaking people who didn't grow up hearing that particular Chinese speech sound could not distinguish it from the related but slightly different English sound. I can't recall the sounds exactly, but I think the Chinese example was something like "tch" and the English was "ch." And the babies' brains did indeed behave in the way predicted: there was a definite difference in the reaction to the two sounds in younger babies, but that difference disappeared as they grew older.

    Anyway, I bring this up for two reasons. First, I hope that someone else is familiar with this research, and can comment on it. Second, if the research is indeed valid, then I wonder if that explains in part why we can't distinguish speech sounds that we're not exposed to as children. Our brains simply failed to maintain the neural or synaptic connections necessary to allow us to hear the distinction.

  168. James Kabala said,

    July 29, 2010 @ 4:05 pm

    AJD: I overlooked your earlier comment. I am also about ten years older than my students (in a non-linguistics field) and will have to keep my ears peeled in the future.

  169. dw said,

    July 29, 2010 @ 6:30 pm

    @Ken Brown:

    Well, Hawaiian is famous for having a very small number of phonemes: according to the Wikipedia page it has eight consonants and ten vowels. By comparison the standard dialects of English have something like 24 consonants and around 16 vowels (more vowels if centering diphthongs in non-rhotic accents are included, perhaps slightly fewer in North American accents with merged cot-caught).

    I'm no expert on the subject, but it does seem as if there is an inbuilt tendency to lose phonological contrasts. When a child is learning a language, he or she tends to do the minimum necessary to function in the language community — you could call it either laziness or efficiency, according to whether you wish to put a negative or positive slant on it. If there is a contrast that can be lost without being corrected/ridiculed by peers/parents/family, then it will probably be lost.

    Look at the commenter earlier in this thread whose father made the Mary/marry/merry distinction — but who admitted that he/she never realized this until recently!

    It is certainly easy to look at words that used to be minimal pairs in some historic variety of English and are now homophones for most or all speakers — wait/weight, wine/whine, weigh/way, — and these are only words beginning with W.

  170. John Cowan said,

    July 29, 2010 @ 11:29 pm

    Ken Brown, you overestimate the amount of loss. John Wells wrote down 22 widely accepted standard lexical sets, representing 6 historic short vowels (KIT, DRESS, TRAP, LOT, STRUT, FOOT), 6 long vowels (FLEECE, FACE, PALM, THOUGHT, GOAT, GOOSE), 3 diphthongs (PRICE, CHOICE, MOUTH), and 7 r-colored vowels (NURSE, NEAR, SQUARE, START, NORTH, FORCE, CURE).

    (He has two more, BATH and CROSS, that represent different merger patterns in different regions: some pronounce BATH like PALM, others like TRAP; most accents unify CROSS with LOT, but a few with THOUGHT; no one gives them their very own vowels. Handling a wider variety of accents requires more lexical sets like TERM and FREIGHT distinct from NURSE and FACE, but never mind that for now.)

    Now let's look at the patterns of merger in a grossly oversimplified way: RP versus "General American" of the originally Western variety. The RP accent has the mergers PALM=START and THOUGHT=NORTH=FORCE=CURE (although some CURE words have gone other ways) for 18 distinct sets. In GA(W), the mergers are LOT=THOUGHT=PALM and NORTH=FORCE, for 19 distinct sets! So the loss of the LOT and THOUGHT vowels is more than compensated for by keeping all r-colored vowels but one distinct.

  171. dw said,

    July 30, 2010 @ 12:20 am

    @John Cowan

    you forget NEAR=KIT, SQUARE=DRESS, CURE=FOOT and arguably NURSE=STRUT in GA(W). So by my arithmetic your GA(W) has only 13 distinct vowels. And many RP speakers would strongly dispute that CURE=FORCE, but we'll let that go for now.

  172. John Cowan said,

    July 30, 2010 @ 7:43 am

    DW: The vowels of NEAR, SQUARE, CURE, and NURSE by definition include the /r/ in rhotic accents, so the equivalences you mention are impossible for GA(W). It's not my accent; for me the first three pre-/r/ vowels are long ones (FLEECE, FACE, GOOSE), and NURSE is vocalic /r/. I also have the hurry-furry merger, so there's no contrast between STRUT+r and NURSEt

    CURE is an oddball lexical set: it doesn't have that many words, and many accents have not merged it with any other specific set, but have broken up the words into many different sets. Fortunately I preserve it pure and undefiled. :-)

  173. Carin said,

    July 30, 2010 @ 8:48 am

    @John Cowan, tangential question: Does it really make sense to use FOOT as the example word for a set within the group of historically short vowels, since historically it had a long vowel?

  174. Ken Brown said,

    July 30, 2010 @ 8:52 am

    CURE is nothing like NORTH/FORCE in RP!

  175. PaulB said,

    July 30, 2010 @ 9:03 am

    This is fascinating. Over here in London I was aware that in the USA you use vowel sounds different from ours, varying regionally, but I'd always supposed that each variant used a set of vowel shifts that maintains distinctions between words. I'm surprised that vowel shifts can occur that create so much ambiguity.

    The only mergers discussed above that occur natively in these parts are horse/hoarse, wait/weight, and wine/whine (does anyone distinguish within the first two pairs?).

    "ate" (past tense of "eat") is usually pronounced "et". (I think it's pronounced "eight" only if intransitive.)

    I'm not aware that I've ever heard anyone pronounce the "t" in "castle".

    I disagree with Ken

    Not just Cockney – almost all south-eastern or eastern-English speakers (including RP) would have /f/ instead of/θ/ at the end of "mouth" in rapid or casual speech.

    I've been saying "down in the mouth" to myself as fast as I can, and it takes an effort of will not to put my tongue between my teeth.

  176. Jerry Friedman said,

    July 30, 2010 @ 1:00 pm

    @John Cowan: How do you use those lexical sets to describe the mergers that many of us Americans have before /r/, namely marry/merry/Mary, mirror/nearer, and hurry/furry (or only one pronunciation of slurry)?

    @PaulB: Yes, lots of us pronounce fairy the same as ferry, Rosa's the same as roses, and so forth, but there isn't that much ambiguity. For instance, I think the mirror/nearer merger causes no ambiguity at all. Of course it's not a competition, so there's no reason for me even to wonder whether the non-rhoticism of RP creates more ambiguity about porn shops and such. :-)

    And I'm still hoping someone will listen to the beginning of this speech and tell me whether Dr. Gutmann doesn't have a vowel in Bryn Mawr that's different from that of both stars and warm. (Though maybe she's not consistent about distinguishing it from stars.)

  177. dw said,

    July 30, 2010 @ 1:59 pm

    @John Cowan:

    Sorry to continue this diversion. I wonder whether we are talking at cross-purposes somehow, but my understanding is that we are attempting to count vowel phonemes in the different dialects.

    Consider the (somewhat recherche) minimal triple "hay-ring" vs. "haring" vs. "herring". In RP all three are distinct: hay-ring clearly belongs with FACE; herring clearly belongs with DRESS; thus "haring" must be a separate phoneme which becomes identified with SQUARE.

    In what you call GA(W) there would be only two sounds here: "haring" would be homophonous with "herring". So there is no need to postulate a new phoneme for the SQUARE set: it's just the DRESS set plus /r/.

    In what Wells calls "Type 1" and "Type 4" accents (e.g. Scottish, Irish, rhotic Caribbean) we would instead have "hay-ring" = "haring", which is also the historical situation in English. Again no need for a new phoneme in the SQUARE set. If I understand you correctly, your own personal accent is of this type.

  178. Todd said,

    July 31, 2010 @ 3:40 am

    When I moved from rural North Carolina to rural Georgia in 1992, my teacher and classmates got mad at me for not knowing what "gym clips" were. My best guess was that they were some sort of clips for gym shorts. They turned out to be "Gem clips," locally the leading brand of paper clips.

  179. TLO said,

    July 31, 2010 @ 6:07 am

    I was born and raised in the Pacific Northwest (I'm in my late 20s) and pronounce 'bag' and 'bang' with vowels closer (as far as I can tell) to 'bake' and 'bane' than to 'back' and 'ban.' 'Beg' is also possibly the same as 'bag.' I wonder how widespread this phenomenon is, and whether it is very recent. Do any of you share this pronunciation?
    Also, FWIW, I have the caught-cot merger and the Mary-marry-merry merger, but not the pen-pin merger.

    [(myl) Michelle Minnick Fox did some work on this phenomenon in Milwaukee, when she was a grad student at Penn almost 20 years ago. One way to sum it up is to note that for such speakers, haggle and Hegel are homophones; and bag is the same as the first syllable of bagel. I'm not sure exactly what the geographic spread of this phenomenon is, but it's certainly Out There.]

  180. Jerry Friedman said,

    July 31, 2010 @ 1:44 pm

    @TLO: Those two don't necessarily go together. I do say bank, etc., with a sound very close to that of bane, but my vowel in bag is pretty close to that in bad and lab. I think.

  181. Dan T. said,

    August 5, 2010 @ 2:44 pm

    Do some of you have no distinction between Lara and Laura, and between Sanders and Saunders?

  182. dw said,

    August 5, 2010 @ 6:28 pm

    "Sanders" and "Saunders" I would expect to be merged by cot-caught mergers.

    The merger doesn't usually apply before /r/, so Lara and Laura would normally be distinct.

  183. dw said,

    August 5, 2010 @ 6:30 pm

    …except that many, possibly all North Americans would have TRAP rather than PALM in "Sanders", which would distinguish them. Sorry for forgetting that in my previous comment.

  184. Dan T. said,

    August 6, 2010 @ 11:52 am

    Those particular names came to my mind because of some recollections regarding fan commentary about TV shows. There's a character named Lara on The L Word, and in a fan podcast about that show, the name as said by the podcasters and people who called in with comments sometimes sounded to me like "Laura", confusing me a bit because of the lack of a character named Laura on the show.

    Meanwhile, on the Disney Channel show Lizzie McGuire, the last name of "mean girl" Kate is either Sanders or Saunders, and even the writers, producers, and actors can't seem to keep it straight which it is; it appears both in speech and writing both ways. This made me wonder if a dialect-based merger was at fault.

  185. Dan T. said,

    August 6, 2010 @ 11:59 am

    There was also a DC Comics "Imaginary Story" from the 1960s (I think) in which Superman's Kryptonian mother Lara came to Earth instead of dying in the explosion of Krypton, and when she had to pick an Earthly name to go by, she called herself Laura.

  186. LaVerne said,

    September 21, 2010 @ 11:43 am

    My husband, originally from Minnesota who later moved to South Florida is usually a very nice, humble man. Someone recently mentioned that many people here in North Carolina (where my husband has lived for almost 20 years, pronounce "pin" and "pen" the same. I agreed that we do pronounce both words alike. He asked "YOU DO KNOW that they are pronounced differently and are two different words, right?" I was a little insulted and felt as though he was questioning my intelligence. We discussed for quite a while the differences and the merger of these two words along with others…I did the research and found that in a fairly large part of our country, these two words have merged, he wasn’t interested!
    I asked if he pronounced "witch" and "which" the same. I hear the difference in these two words and pronounce them differently, he doesn't!! On a different note..he pronounces Ralph “ Lauren” the same as Sophia “Loren”.

  187. Gary said,

    November 5, 2011 @ 11:04 am

    I'm an actor living in NYC who grew up in FL. I've always had a hard time with the pronunciations of these two different words. I got to Bway playing a young CT writer Eugene O'Neil & the very first line out of my mouth was "Independence Day" & I had to always be careful to think about PEN not PIN. Still get caught making the error today. :-P

  188. Will Peavy said,

    March 13, 2012 @ 4:33 pm

    I was born in Texas, and have lived in Florida for most of my life. Even if I listen very intently, I cannot hear the difference between the vowel sound in pin and pen. My wife is from New Jersey and she thinks there is a big distinction from the two. For me, both words are pronounced the same way. Even when my wife says she is pronouncing the two words differently, they sound the same to me.

  189. I can't take it anymore. - Books, authors, composition, publishers, editors, novels, print, poetry... - Page 136 - City-Data Forum said,

    November 4, 2012 @ 8:34 pm

    […] Originally Posted by Mightyqueen801 Ink pen." I suppose to distinguish from a lead pen? ROFL, but I think that's for the people who pronounce "pen" the same way they pronounce "pin". For some reason, a good chunk of the US does not seem to realize that there is a "short e" sound and says it all like a short "i". What they are really saying is "Ink pin". First time I heard this, it was a woman at work from the south. She asked me for a pin. I said, "a safety pin"? She said, "no, an ink pin". Even the robo voice on my work phone says "please inter your ID number". And I even hear anchorpersons say "Tin O'clock news." Language Log

RSS feed for comments on this post