Talk:International Phonetic Alphabet chart for English dialects/Archive 4

recent edits
Hi Aeusoes,

Wanted to explain my recent changes. The vowels in see and sue and diphthongs in GA. Since we're showing the phonetic differences between the dialects, I thought it was important that we indicate that. (Otherwise, if we're not interested in that level of detail, we might as well represent the vowels of RP so and say as and .) Also, since the r in car etc. is phonemic in RP, it is essential that we include it, though we could follow the OED and write it  if you prefer. kwami (talk) 07:58, 4 January 2008 (UTC)
 * After I made my edit, I thought that might be a good alternative so we are in agreement there. While it's arguable about the phonemic /r/ in the coda (this is a feature that is disappearing in RP speakers with more universal linking-r taking its place), presenting information with the assumption of its presence can only provide more information than not.  If I'm not incorrect, however, my understanding is that even RP  and  are diphthongal in the same way that GA describes them to be.  Even if this is not the case, the distinction between monophthongal  and an  diphthong is of much lesser importance than between, as you suggest,  vs .  I've recently changed the vowel quadrangal image for Received Pronunciation and if we are going to get into OR phonetic detail that comes at odds with conventional transcription, looking at those quadrangals would be a good place to start.  — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  08:37, 4 January 2008 (UTC)
 * By "linking-r", do you mean [r] appears before vowel-initial words? That's what I was trying to indicate with the superscript ar, or the (r). (If it appears in some words, and not others, it's phonemic. That is, car is /ka:r/ in RP; [ka:] is allophonic.) Or do you mean that it's being lost even before vowels, the way the en in an is being lost (a apple) in GA?
 * As for the superscript, it's common to do that for phonation, affricates, diphthongs, etc., though I suppose alteration with zero would be stretching things a bit.
 * Diphthongal and  are not OR; several transcriptions indicate them, and there have been arguments that they are phonmically /ij/ and /uw/ or /ii/ and /uu/ -- that is, it's important enough for people to make base phonological claims on it. The amount of movement may be less than in  or, but it's still apparent before vowels, and to foreign students. (Again, you could argue it's not as important in  and  as it is in /au/ and /ai/.) kwami (talk) 10:20, 4 January 2008 (UTC)
 * Even if analyses considering GenAm and  to be phonemically diphthongs weren't WP:FRINGE (and I think they are), you haven't even indicated them as diphthongs. You've used the expressions  and, which suggest a palatalized /i/ and a labialized /u/. I don't think anyone has ever suggested that vowels get that kind of secondary articulation. —Angr If you've written a quality article... 10:59, 4 January 2008 (UTC)
 * I've seen several treatments of these as diphthongs. I forget who, but Ladefoged discusses some of them. He doesn't treat them as somehow oddball, only as alternate takes. As for the raised jay and double-u, that's a common convention for diphthongs, as in, or , . The problem with dropping them in-line is that they'd be the only phonemic vowels ending in phonetic consonants, and so isn't terribly appropriate. The alternate would be transcribing them as , which I've never seen anywhere. kwami (talk) 12:03, 4 January 2008 (UTC)
 * Hmm... wasn't this page supposed to restrict itself to phonemic transcription? FWIW, and  are conspicuously more "diphthongal" in RP than they are in GenAm.  More generally, this page has come to be pretty inconsistent as far as the level of detail is concerned.  For example, the Australian broad A and the NZ one sound pretty much the same to me, yet they are transcribed differently.  And RP is much more likely than GenAm to merge tore with tour.  Also, for an Englishman without the  phoneme, cure rhymes with core; in General American, speakers without the  phoneme will pronounce cure to rhyme with fur, never with core.  Jack (Lumber) 14:03, 4 January 2008 (UTC)


 * Kwami, what I mean is that in a number of non-rhotic accents, an appears in phrases like "sofa is" (sofer is) and "paw eating" (par eating) so that the intervocalic r is part of a phonological process rather than a silent consonant so that "paw eater" and "par eater" are pronounced the same.  RP, on the other hand, used to make a distinction between "paw eater" and "par eater" but, if I understand correctly, the merger is creeping into even RP speech, though it's still considered incorrect among prescriptivists.
 * I agree with Angr that superscript j and superscript w aren't good ways to indicate diphthongization since they are already used for palatalization and labialization. I'm generally not a big fan of superscript anything with diphthongs.  For what it's worth, I've seen,  in The Phonetics of Russian (1969).
 * JackLumber, the difference between the broad a of Australian and NZE ( is really minor (one of height) and may be why they sound the same to you. But you're right about tour and tore, the Wikipedia article on RP doesn't account for  and my dictionary has  as an alternate pronunciation.  — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  00:52, 5 January 2008 (UTC)


 * Ah, I see. Yes, we have rhotic hypercorrection in the States too, in places like Boston. If that's become general, it would certainly be reason to remove the ars - either that or add them to some of the other vowels.
 * You mean you saw, in The Phonetics of Russian as a description of English? That would make me feel better about using them, though I guess it would have to be in RP as well. We need to cover these transcription variants in English phonology. kwami (talk) 01:15, 5 January 2008 (UTC)
 * Jack, yes, phonemic, but to what degree of precision? If we want to be broad, we could just list, and not worry about diphthongs or vowel length at all, or . —kwami (talk) 01:15, 5 January 2008 (UTC)


 * Yes, it was in describing common ways that English speakers pronounce the vowels of Russian and the authors even say of that "a diphthong similar to this occurs in Russian in unstressed positions, that is, ."  But aren't all English monophthongs somewhat diphthongal? — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  03:10, 5 January 2008 (UTC)


 * I don't believe so. I've never seen any but the historically long vowels transcribed as diphthongs, and at least in my pronunciation the "short" vowels are just as monophthongal as the vowels of Japanese.
 * Your Russian source supports my point: /i:/ and /u:/ (and of course /o:, e:/) are diphthongal enough that they're heard as an English accent when English speakers learn another language. kwami (talk) 03:23, 5 January 2008 (UTC)
 * Ahh, but it doesn't say that it's true for only American (or only British) speakers.
 * I've also found and  in A Course in Phonology (1999) which pegs such a pronunciation to "GA and in accents of southern England."  Though these vowels (as well as  and  are realized as monophthongs or diphthongs depending on accent and Roca and Johnson call them "homogeneous diphthongs" as opposed to the other "heterogeneous diphthongs" of English .  — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  03:42, 5 January 2008 (UTC)

Why did you guy change a simple chart into a more complicated arrangement that linguist will probably understand. By making these changes you are turning wikipedia into a joke. In the process of changing things you have altered pronunciations of Australian and New Zealand vowels, which are not true and against sources for those dialects. --203.220.171.32 (talk) 03:34, 5 January 2008 (UTC)
 * Explanation made to me on Aeusoes1's talk page.
 * Note: minor formatting. I changed SA to SAE and Sc to ScE as they seem to be (from Google searches) to be acceptable short forms for South African English and Scottish English respectively. see SAE "South African English" -wikipedia and ScE "Scottish English" -wikipedia. --203.220.171.32 (talk) 05:55, 5 January 2008 (UTC)
 * Other than showing /i:, u:/ are diphthongs, what is more complicated? How are AusE and NZE wrong? kwami (talk) 06:02, 5 January 2008 (UTC)
 * Mainly it was the Australian pronunciation of tore and tour, which are correctly transcribed as and  respectively, at least according to my Macquarie Dictionary. Note: Macquarie does not use  in their transcriptions.
 * The wrong part of AuE and NZE was the use of linking-r (for want of a better word) or (ɹ), but I've had a chat to Aeusoes1 about this on his talk page (as I said above!), that's why I reverted my changes concerning the linking-r. --203.220.171.32 (talk) 06:30, 5 January 2008 (UTC)
 * I've added car as an example of the vowel in final position. Eg, "the old car is near the new car" would be for AuE.
 * Also, I've noticed the "hair, there" diphthong does not include a linking-r for the non-rhotic dialects. Shouldn't it be include? Eg, ɛə(ɹ) for RP, eː(ɹ) for AuE and eə(ɹ) for NZE. 203.220.170.212 (talk) 13:22, 5 January 2008 (UTC)
 * Fixed. kwami (talk) 17:28, 5 January 2008 (UTC)
 * BTW, I've never heard in GA. Are you sure that isn't just an iconic transcription, capturing orthographic o while retaining the diphthong, like the one we use here in Wikipedia? The transition may not cover as much vowel space as RP, but it definitely starts with some sort of unrounded vowel. Mandarin has [ou], and that sounds rather different than GA. kwami (talk) 17:41, 5 January 2008 (UTC)
 * Whether you've heard it or not is beside the point: the published sources for General American use the symbol "o": Kenyon and Knott use /o/ and Longman uses . Do you have a source for ""? —Angr If you've written a quality article... 17:45, 5 January 2008 (UTC)
 * Not offhand, though I could probably dig up /ɔʊ/. But why did you remove the /r/ from hair? kwami (talk) 19:30, 5 January 2008 (UTC)
 * Roca and Johnson (1999) say of the vowel of coat that it is in GA, before a final l it is  (or even  in South Africa),  in "the middle Atlantic and western Pennsylvania areas of the US and  in London and the southern hemisphere "undergoing fronting in younger speakers." (p. 192).  From my OR impression, it does seem to start less rounded than [o] would imply but I've never seen it transcribed  — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  20:26, 5 January 2008 (UTC)

Hi guys. I like the idea behind the recent edits but as it stands it creates a lot of inconsistency. I'm a New Zealand living in Scotland, so I hear a lot of Scots, English, New Zealand, Australian and American accents, although I haven't done any linguistics training so my ear is still learning. The things I'm concerned about are: Ben Arnold (talk) 13:27, 9 January 2008 (UTC)
 * The different transcriptions of father... I'm convinced that the New Zealand pronounciation has a different quality from the run-vowel, not just a different length. My tongue is in a significantly different place in cut and cart. I have to open my mouth more to say cart, move my tongue down at the front and up at the back. I think cut is definitely closer than cart. The Australian cart vowel seems closer again nearer NZE cat.
 * I think the New Zealand see-vowel is [ɪi] as it has been transcribed for RP and GA. The exaggerated New Zealand accent used by comedian Ginette McDonald pronounces the sound as an even more distinct dipthong, something like [ɘi]. The same may be true in Australia, although they use a very close /ɪ/, particularly in Sydney, so their /iː/ may be more pure.
 * I think the linking-r is best described in a footnote rather than as (ɹ) in the table, that would make it less cluttersome.
 * I'm not sure that the dipthongs that are transcibed with [e] in New Zealand and Australian English are in general any different from the ones that are transcribed with [ɪ] in R.P. and G.A. I think we may be creating an artifical distinction because the sources transcribed the sounds slightly differently.
 * I'm somewhat dubious about the use of [ɘ] instead of [ə] for New Zealand English. I think it's a bold call that the schwa and sit-vowel have merged in New Zealand English, as the source claims. I can pronounce about with either [ə] or [ɘ]. I would use schwa for Something About Mary and [ɘ] in he's gone walk-about, however if I use the schwa for either vowel in cricket it sounds totally wrong.

IPA is such horse ****
Why are you wasting everyone's time putting IPA all over Wikipedia? Like I want to spend 10 minutes reading a chart every time I don't know how to pronounce something. Can't we just stick with the simple phonetic pronunciation? God, you people drive me crazy. You are seriously compromising one of the best resources on the web. Think about that. Ogeez (talk) 00:14, 5 January 2008 (UTC)


 * What is the "simple" phonetic pronunciation? How else would we transcribe it? And no, you don't need to spend ten minutes every time, just a couple times, until you get it. If you can read English, which is hardly simple, you can pick up the IPA. kwami (talk) 01:00, 5 January 2008 (UTC)
 * I just noticed Ogeez is a chronic troll. kwami (talk) 05:52, 6 January 2008 (UTC)
 * Some people are strongly opposed to the IPA, like myself. It is an extremely poor indicator of pronunciation, if at all, and lacks the symbols to represent some sounds adequately. It's also run by a bureaucratic body that has taken upon itself to determine what a sound is and isn't. I think we should stick with language separate transcription systems, with the IPA being retained for advanced linguistics. To this day I have no idea where these 'sounds' come from. The IPA body is pulling phonemic distinctions out of it's arse whilst ignoring some extremely significant real ones. Notably the Czech ř, which is represented in a creepy agglomeration of diacritics and unrelated sounds and prevents the IPA from getting adopted as a phonetic system by a language of 12 million people. It used to have the symbol, but the body felt the sound isn't distinct enough. Which, if you ever hear ř is insane. This is but an example of the narrowminded arrogance of office linguists who haven't heard anything but English or French in the last decade. But enough ranting from me, I just wanted to point out it's an idiotic policy at best to use the IPA here. +Hexagon1 (t) 03:12, 6 January 2008 (UTC)

Since you evidently don't know what you're talking about (though I must admit I'm surprised you know about Czech ř), and haven't offered a viable alternative, how do you expect us to take you seriously? Or are you just trolling? kwami (talk) 05:52, 6 January 2008 (UTC)
 * I don't know about Hexagon1, but if you look at Ogeez's contributions and talk page, it's clear he's just trolling. —Angr If you've written a quality article... 12:52, 6 January 2008 (UTC)
 * I am not trolling, there is nothing more I hate than pointless arguments, but I value debate. I was merely pointing out that Ogeez may have a semi-valid point, despite the presentation. And I really have no clue what I'm talking about, good on you for catching on. :) I have qualms about the IPA but I don't know a sufficient amount about it for a debate, I was hoping someone would point out why my arguments are invalid. I might have been a wee bit too passionate, and I probably shouldn't post things like that on article talks, I realise now. :) +Hexagon1 (t) 13:10, 7 January 2008 (UTC)
 * Okay, since you're not a troll you get a civil answer. First, I think most people familiar with IPA would disagree with your major premises: (1) actually, it is a very good indicator of pronunciation, especially at the broadly phonemic level; (2) although there are a handful of sounds that are actually used linguistically and that are difficult to transcribe with the IPA, they are in fact both very few and very rarely occurring; (3) the International Phonetic Association is hardly a "bureaucratic body that has taken upon itself to determine what a sound is and isn't"; it's a group of professional phoneticians and phonologists who genuinely know better than anyone else what a sound is and isn't. I don't know all the details behind the decision to scrap the symbol representing the Czech ř, but I suspect "the sound isn't distinct enough" was not a criterion they used. Rather, I suspect they decided the sound isn't commonly occurring enough cross-linguistically, and it is adequately represented not by "a creepy agglomeration of diacritics and unrelated sounds" but by the symbol : Czech ř is an alveolar fricative trill: the [r] indicates an alveolar trill, and the diacritic indicates a closer stricture and is commonly added to symbols for approximants to indicate fricatives. Using separate transcription systems for each language would not be practical here, as doing so would require readers to learn a different system for each language discussed. (When I first started here I added a lot of information on Irish phonology and used a non-IPA system conventionally used in Irish linguistics. I quickly realized that was inappropriate and confusing in the context of Wikipedia and converted to "true" IPA.) Finally, the main reason for using IPA is that despite the numbers of people coming here to complain that "practically no one knows IPA", the fact is that more people know it than know any single other phonetic transcription system. And finally, whatever the faults of the IPA, there currently exists nothing better. It has no serious competition. —Angr If you've written a quality article... 16:08, 7 January 2008 (UTC)
 * Hexagon, together with input from some other editors I came up with a prə-NUN-see-AY-shən respelling key, which is more accessible than the IPA for a lot of people (Americans, mostly), but it was a hard sell to anyone whose native language wasn't English or who'd already taken the time to learn the IPA. Trying to use it here was a bit like trying to convert metric to imperial units. I eventually gave up, but no-one will mind if you want to use it as a secondary guide after the IPA. If you want to check it out, it's at Help:Pronunciation respelling key. kwami (talk) 18:44, 7 January 2008 (UTC)
 * Angr – I think that's a very good response, well argued and well put. I'm not particularly familiar with IPA, but I'd much rather be using something logical and universally applicable with an academic basis than ad hoc schemes which still essentially use English or American spelling. --Richard New Forest (talk) 20:01, 7 January 2008 (UTC)
 * Thank you for the response Angr, makes sense (which is more than I can say for many other arguments I've seen). Thanks! (Really.) And kwami: Interesting, although as a speaker of Australian English (and near-worshipper of the metric system) I don't think I'd be all that supportive. :) I see that that also outlines some of the benefits of IPA, in that it's not leaning towards any dialect or accent. I also guess some of my opposition may have been founded the fact most resources available are oriented towards GenAm or, if I'm lucky, RP, which doesn't really do a lot of good for me as a mainly-Australian English speaker. I apologise again for the strong initial post, I can, like any other editor, at times get a little carried away, so sorry. +Hexagon1 (t) 04:50, 8 January 2008 (UTC)
 * Actually, people like you are the reason I tried pushing the respelling key. Several people have insisted that since it's called the International Phonetic Alphabet, if you use the IPA you have to choose a specific dialect; if you want to be fair, you have to give a separate transcription for each dialect. One editor got quite angry, saying that since the key does not transcribe any actual English dialect, it's nonsense, 'original research', and needs to be deleted. That's why there are no references in the chart: It's a compromise reached by several (other) editors. All the published dialect-neutral transcriptions I know of are respellings, not IPA. kwami (talk) 06:13, 8 January 2008 (UTC)
 * Oh yeah, I understand what you mean, that's partly why my original rant talked about language dependent systems, they could be dialect-neutral. I have to admit I just skimmed over your pronunciation article without giving it much thought so I just went by what you said previously. I'll now slowly fade to the background, before I embarrass myself further. :) +Hexagon1 (t) 10:38, 8 January 2008 (UTC)
 * Oops, I got lost, forgot which talk page I was on. The guy got upset about help:pronunciation, which is a non-dialectal IPA key for English. kwami (talk) 10:45, 8 January 2008 (UTC)
 * Why not use respelling for English words and IPA for other languages? —Preceding unsigned comment added by 88.101.76.122 (talk) 21:39, 24 February 2008 (UTC)

v - voice, have, of
of? really? I always said with a normal f'. was i wrong? —Preceding unsigned comment added by 217.132.131.49 (talk) 18:55, 30 January 2008 (UTC)
 * Pretty much, yeah. I remember an intellectual exercize where one is given a passage of text and told to count the number of f's.  Usually, people miss the f's in of because they're voiced.  — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  18:57, 30 January 2008 (UTC)

œ
is this pronounced the same as "oe", like the "oy" boy? —Preceding unsigned comment added by 71.113.6.227 (talk) 01:21, 14 February 2008 (UTC)
 * Do you mean "œ" used in ordinary orthography of words like "fœtus" and "hors d'œuvre", or do you mean the IPA symbol [œ]? —Angr If you've written a quality article... 06:13, 14 February 2008 (UTC)

GA TOE
In addition to Iggy & Roca as a source for GA, the Handbook of the IPA offers three different transcriptions of the vowel in bode , , and. While these are phonemic and don't show the diphthongal possibility, there's clearly a difference between GA bode and RP bode (one that I can hear) and these representations seem to assume that the difference is the rounding of the first element. Our own article on GA indicates that there's allophonic variation between and  (though this is unsourced there). Hillenbrand (JIPA 2003) mentions that it has been "well recognized for many years that the nominally monophthongal English vowels /e/ and /o/ tend to be diphthongized, with /e/ showing an offglide in the direction of /i/, and /o/ moving toward /u/." (p 124) Are there sources that argue that GA is not  but is instead ? — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi] 09:05, 22 February 2008 (UTC)


 * I agree that GA is not the same sound as RP, though I couldn't tell you what exactly the difference is. (If I had to guess, I'd say that the GA diphthong doesn't travel as far through vowel space.) However, the first element is very clearly not rounded, at least for me when enunciated in isolation. It's not as open as the  of cup, and is very similar to the /ə/ in about. (I can't feel any difference there, or see any in the mirror.) I feel backing during the glide that I don't with . The change in rounding isn't as great as it is in, but I think that's largely due to the greater openness of [a]. None of your sources argue that the initial element is rounded; they only discuss the off-glide. (And the Handbook cannot seriously be suggesting that it is phonetically !) I suspect that the IPA letter  is only used for historical and orthographic compatibility. Or it may be due to context: Perhaps the first element assimilates more to the off-glide than it does in RP? Bode has no rounding for me, but bold does, markedly so. But then, lots of vowels shift due to adjacent consonants. If /o/ were , I'd've been able to learn to pronounce [o] by dropping the end off the diphthong, the way I was able to learn [e]. I tried, mislead by the common IPA transcription, but it didn't work — I had to learn [o] as an entirely new vowel.


 * BTW, my dialect is Southern Californian, which is reasonably close to GA. I have never noticed anything different about the /o/s on the national news or in the Midwest, though I can't say for sure there isn't any.


 * Also, I first changed the article to GA to keep it distinct from RP . Maybe RP is actually ? Or maybe the difference is GA  vs. RP ? We could just have different schwas, and therefore don't feel we can use the same transcription for both dialects. kwami (talk) 10:40, 22 February 2008 (UTC)


 * I'm also a speaker of California English and I suspect that the differences between our /o/ and that of the media is subtle (that is, I don't perceive a difference unless maybe I'm thinking about it while watching the news). Certainly the first element of the diphthong is less rounded and likely more central than, say, Spanish /o/ and a precise transcription might use  or  with special diacritics.  That HIPA and the other sources don't explicitely say that /o/ is rounded doesn't sound too exceptional; they also don't say that /u/ is rounded, perhaps because it's understood through the use of the character that it is such.  I would think that only if it weren't rounded despite their use of the  that there would be incentive to clarify.  My concern here is that we've got a bit of OR on our hands if we're relying on our mirrors and our own perceptions. If you don't have any sources to back up the "different schwas" interpretation then we should revert back to  until we find something. — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  11:05, 22 February 2008 (UTC)


 * The 'different schwas' thing was just speculation. I'm not familiar with RP. And you're right, it is OR - but still, I doubt it's . If it's rounded at all, it's rounded very slightly the way the /a/ in /aʊ/ is rounded. kwami (talk) 11:58, 22 February 2008 (UTC)
 * The Concise Oxford Dictionary 9th ed (on CD) has for . However for a broad phonemic transciption clearly  or  are preferable. &minus;Woodstone (talk) 10:36, 22 February 2008 (UTC)
 * Doesn't Oxford transcribe RP? — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  10:38, 22 February 2008 (UTC)
 * Sure. I evidently could not imagine anyone expecting sources for in GA (and misread your question). &minus;Woodstone (talk) 10:42, 22 February 2008 (UTC)

One thing for both of you Californians to keep in mind is that California English is not close to GenAm when it comes to the "tense" back vowels (GOAT and GOOSE). Those vowels have considerably fronter starting points in the West than in the rest of the country. The published sources for General American that use IPA invariably use /o/ (or if they want to emphasize its diphthongal character): this includes not only the sources Aeusoes mentioned above (I don't know what you mean by "Iggy & Roca" though; Iggy Roca is one person) but also Kenyon & Knott, Ladefoged, J.C. Wells (both Accents of English and Longman Pronunciation Dictionary), NTC's Dictionary of American English Pronunciation, dictionary.com (click on "show IPA pronunciation"), etc. The other thing to keep in mind that the reference points for vowels are fuzzy: just because the vowel is transcribed /o/ doesn't mean it has to be realized as the furthest possible back, maximally rounded Cardinal Vowel 7. We go with on this page because it's what the published sources say; if the exact realization is further front in some dialects, we can mention that in the appropriate place (i.e. the article on that dialect, not here!). That said, I do think [o] or is actually a correct transcription of the vowel a large portion of American English speakers use; in fact, most of us do have a back vowel there! —Angr If you've written a quality article... 16:46, 22 February 2008 (UTC)