Talk:Sound correspondences between English accents

Table is wrong

In Australian English, the o in no is pronounced very differently to the o in cold.

The vowel sound in soul, roll, cold, fold, mould is very different to the vowel sound in no, toe, hoe, soap, goat, wrote.

Someone should note this in the table. Also, folk and soak rhyme in Australian English. —Preceding unsigned comment added by 58.171.228.189 (talk) 07:55, 10 December 2008 (UTC)

Yeah, it's [ɔ] in much of the US. It was overlooked in the table. kwami (talk) 09:00, 10 December 2008 (UTC)

Unless you pronounce "called" and "cold" the same, it's not the [ɔ] sound. "oʊ/o" is most appropriate in my opinion ("cold" is like saying the first part of the "no" diphthong).

Halt

Halt has the same vowel as hold in New Zealand English, the two words only differ in the voicing of the final consonant.

If I understand correctly, hold is pronounced something like [hɒɯd] in New Zealand English as a result of l-vocalisation. This would seem to imply that halt should be transcribed phonemically as /hoʊlt/, with /oʊl/ pronounced by New Zealanders as [ɒɯ] like gold or mold. This is not the way RP speakers would transcribe it. What I don't know is whether this makes halt a special-case with two pronunciations or whether halt is an example of a lexical set that is pronounced differently by RP speakers and New Zealanders. If it's a special case then it makes a bad choice for the table. If it's part of a lexical set then we should add a new row to the table.

Ben Arnold (talk) 07:38, 1 January 2009 (UTC)

I can't speak for NZ English, but here is what the Australian Concise Oxford Dictionary (1988; ISBN 0-19-554619-9) says about ‘halt’ (p. xiii):

In RP ŏ [i.e. IPA-for-english ɒ] is a recognized variant of the sounds also pronounced as aw and o͡r [i.e. IPA-for-English ɔ and ɔr] when followed by l or r. In Australia this variant is the dominant pronunciation in such words as alter, assault, fault, salt, coral, or floral.

Incidentally, ‘hold’, ‘gold’ and ‘mo[u]ld’ are poor example words, in that some people pronounce these (and ‘bold’, ‘cold’, ‘fold’, ‘sold’ and ‘told’) as -ōld (i.e. like holed, goaled, etc.) (and indeed this is the only pronunciation that said AusOED gives for each these eight words), while other people (including I) pronounce each of them as -ŏld (like the vowel of ‘bond’, ‘conned’): so sold & soled are distinct in my accent. It's not just -old words, but also e.g. ‘bolt’, ‘colt’, ‘dolt’, ‘jolt’ (-ōlt according to AusOED, -ŏlt for me). Thus (combined with Ben's mentioning both /oʊ/ (typically associated with ō) and [ɒ] (typically associated with ŏ)), I'm not certain whether Ben claims that NZE ‘halt’ is hŏlt or hōlt. (I believe he claims the latter.)

Said AusOED doesn't list hōlt as a pronunciation for ‘halt’, it gives only hawlt and hŏlt.

Pjrm (talk) 09:19, 17 April 2010 (UTC)

What's up with the gaps?

There are gaps in the Irish English column, and the entire South African English column is blank. Why? -- Army1987 – Deeds, not words. 15:58, 4 January 2009 (UTC)

Because no one has found (possibly, no one has looked for) reliable sources on which to base the information. —Angr 19:09, 4 January 2009 (UTC)

Wouldn't it make more sense to remove the SAE column altogether until such a source is found, then? (As for IrE, which source mentions some sounds and not others?) -- Army1987 – Deeds, not words. 20:19, 4 January 2009 (UTC)

I disagree, leaving the gap is a good way to "invite" someone to fill it. If you remove the column the information will probably never be added. I'll post a request for help over on the South Africa Project page. Roger (talk) 16:28, 7 January 2009 (UTC)

Should we add a column for Indian English, Caribbean English, and any other dialect of English we can think of, then? -- Army1987 – Deeds, not words. 13:15, 13 January 2009 (UTC)

If it is a "standardised" dialect with at least one widely accepted authorotative dictionary - Yes. Roger (talk) 14:17, 13 January 2009 (UTC)

Is South African pronunciation reflected by at least one widely accepted authoritative dictionary? —Angr 14:39, 13 January 2009 (UTC)

Yes. The Oxford Dictionary of South African English [1] It is compiled by the The Dictionary Unit for South African English at Rhodes University in Grahamstown.[2] Unfortunately I don't own a copy, nor do I know IPA well enough, otherwise I could have made a start at filling in the "offending" column. Roger (talk) 16:18, 13 January 2009 (UTC)

Unfortunately neither Amazon nor Google Books has a preview of the inside contents. I'd be interested to see if it even contains pronunciation information (I notice Amazon's product description says nothing about pronunciation), and if it does, to what extent it differs from RP - because while South African English differs phonetically from RP considerably, the phonemic differences (which is what would be shown in a dictionary, and on this page), are I believe very slight. —Angr 16:34, 13 January 2009 (UTC)

(<-) ...the key difference being the South African bit-kit split. This article has more serious issues, however. It clearly violates WP:OR, since

• Articles may not contain any new analysis or synthesis of published material that serves to advance a position not clearly advanced by the sources. This article attempts to do something that has never been done before. There has never been such a thing as an "IPA chart for English dialects."
• Much of the content is actually flat-out OR. For instance, no previously published IPA transcription for American English allows for Canadian raising of PRICE. And Irish /r/ is in most cases a retroflex approximant and not a flap.
• Some columns, such as Irish, Canadian, and Welsh, are unsourced and will always be. Does Irish English have an authoritative dictionary? I think not. Is Irish English standardized? Not at all; there's a lot of regional and social variation. Is Irish English worth studying? Absolutely! As far as dictionaries are concerned, the Canadian Oxford Dictionary does use IPA, but its IPA scheme is not intended to be an accurate phonetic representation of the sounds of Canadian English. Likewise, the Australian scheme devised by Harrington, Cox, and Evans is not used in the Macquarie Dictionary, which sticks to the more traditional (and phonetically inaccurate) RP transcription. Therefore, for the purpose of this article, what dictionaries do is irrelevant.
• Last but not least, the table is difficult to read.

I'm Jack^(Lumber) and I approve this message. 01:24, 15 January 2009 (UTC)

Well, while I agree with you that there's no single set of rules that will apply to every English speaker within a given population, I believe that there are certain trends that make, say, New Zealand English different from General American English. On my desk right now is Handbook of Varieties of English, edited by Bernd Kortmann and Edgar W. Schneider (Berlin: Mouton de Gruyter, 2004). It is a thousand-page book, very informative, that phonologically analyzes all the dialects in this chart and many more. I think it's a fine resource for this chart, and I'm planning to add South African English and fact-check the other dialects when I have time, if that's all right with everyone else. Hopefully you'll agree with me, because I think this is a VERY valuable chart (and, by the way, I don't find it difficult to read at all).66.71.70.66 (talk) 01:44, 5 February 2009 (UTC)

Yeah, Handbook of Varieties of English is a wonderful book. I was thinking about using it too. You can view a lot of it online. Thegryseone (talk) 02:26, 5 February 2009 (UTC)

Then again, the HOVOE features detailed phonetic charts for every single dialect, while the purpose of this page was to show phonemic transcriptions for English. I'm Jack^(Lumber) and I approve this message. 00:14, 6 February 2009 (UTC)

Actually, according to Phonetics, HOVOE would be considered a phonological description (or "phonemic" to use the old school term). Phonology deals with abstract systems of sounds and gestural units and their allophones. In addition, the first book has "Phonology" written on the side. What's confusing is that allophones are written using what's called "narrow phonetic" transcription. However, showing the different allophones of a phoneme is a part of phonology and not the study of phonetics. Thegryseone (talk) 11:36, 7 March 2009 (UTC)

True, that's what I was thinking too. It seems that South African English (White South African English) isn't very well-studied. This makes sense if you think about it, because very few people speak it compared to say, American English. Thegryseone (talk) 00:37, 6 February 2009 (UTC)

Indeed, the Handbook has lengthy articles as well as IPA charts for several North American dialects, including New England, the South, Newfoundland and so forth. I assume your "this makes since" was actually supposed to read "this makes sense." It makes sense, since you have the pin-pen merger. :) I'm Jack^(Lumber) and I approve this message. 02:54, 6 February 2009 (UTC)

LOL...you cot me or should I say...caught me. Very clever indeed, Jack. I always enjoy a bit of linguistics humor. See that's proof that I wasn't lying to you. That damn merger screws up my spelling quite often. Immigrate and emigrate are the worst. On a related note, I read on some science Web site that people with the pin-pen merger and people without it have different patterns of brain activity. Ah, here it is. Thegryseone (talk) 03:31, 6 February 2009 (UTC)

But people with the kit-bit split are the brainiest of all. LOL! I don't agree that South African English is not well studied, there are several South African universities with very active English departments and don't forget the dictionary unit at Rhodes University in Grahamstown. I think the problem we have/had in this article is that IPA is not very well known or widely used in South Africa. (I for example had never heard of it until I found Wikipedia!) Roger (talk) 13:39, 9 February 2009 (UTC)

I stand corrected. What the problem is is that no one knows how to transcribe South African English phonemically. It seems that no one has ever done it before. That's what I was getting at. So while it may be well-studied, we don't know how to transcribe its vowel phonemes. Thegryseone (talk) 18:45, 9 February 2009 (UTC)

John C. Wells has, in Accents of English 3: Beyond the British Isles. He uses all the same symbols as for RP, except for /ə/ for the words like bit, and /əɪ/ rather than /eɪ/ for the FACE vowel. —Angr 19:06, 9 February 2009 (UTC)

Why can't we use that as a source then? Thegryseone (talk) 19:16, 9 February 2009 (UTC)

(<-) Well, Wells's tables are not really intended to give an account of what the various dialects sound like, which is what our article is supposed to do, I guess. I mean, we basically have Wells and the HOVOE at opposite ends of the spectrum, and what we have to do is figure out an intermediate approach--but that would be original research. This means we're stuck in a rut. I'm Jack^(Lumber) and I approve this message. 02:48, 10 February 2009 (UTC)

Sort by similarity

Is there any way to order the table columns by related dialect, or number of shared features? This would help clarify similarities and contrasts in related dialects.

I imagine that RP and US English would be the two major poles, dialects with transitional features between them, and outliers beyond RP. I realize this is not a 1-axis range, and the relationships are complex, but there are obviously dialects with more shared features, and putting these next to each other will help the reader.

I'm no expert, but here's a start for discussion. —Michael Z. 2009-02-09 22:41 z

1. SAE
2. NZE
3. AuE
4. WaE
5. RP
6. IrE
7. ScE
8. CaE
9. GA

For example, Trudgill (in International English) says we basically have two major types of accents, an "English" type (England, Australia, New Zealand, South Africa), and an "American" type (U.S., Canada), with Ireland being somewhat intermediate and Scotland being by itself. I'm Jack^(Lumber) and I approve this message. 02:51, 10 February 2009 (UTC)

I think that's a great idea! I took the time to actually compare how similar all the English variants are. South African English is more similar to Australian than New Zealand English. Scottish is more similar to RP and Welsh than other varieties. Irish could go before RP and SSE could then go between RP and CaE (SSE aligns the most with GA, CaE and RP; which makes this order perfect). IrE is the hardest to find a good spot for (due to high similarity with very dissimilar variants). This is what I think it should look like:

1. NZE
2. SAE
3. AuE
4. WaE
5. ScE
6. IrE
7. RP
8. SSE
9. CaE
10. GA

---

Rough proposal for a new table. Added a Southern American English because in my opinion it's different enough from GA to deserve it's own spot. in India, if someone can name a place that has english as the majority native language, it should be added. I think it's a nice order because it keeps the shared features fairly close together, and also has a good indication of area (brittish isles → america → southern hemisphere). if you compare it with æ or eɪ, it's also pretty close. Not just relating to this specific table, but the article page seems extremely cluttered. Are the reduced vowels really necessary to have here? They have their own article. Consonants should be split into normal and marginal(retroflex t, flapped t, trilled r, wh, ch, etc) because there is still a general "correct" way to pronounce them, whereas the vowels are a lot more variable.

Thank you. Southern American English in its several but related varieties is probably spoken by at least twice the combined populations of Australia, New Zealand, Scotland, Ireland, Wales, and (white) South Africa, or double the number of Canadians, and half again as much as the population of England itself, yet is always ignored as if it didn't exist.—dshep/2010.01.22

Here are Kenyon & Knott's (A Pronouncing Dictionary of American English) entries for the Southern pronunciations of the keywords on the chart (in order): æ/ɑ/ɑ/ɔ/ə/ɪ/ɪ/i/e/ɛ/ɜ,ɝ/ə,ɚ/ʌ/ʊ/u/aɪ/ɔɪ/o/aʊ/ɑ:/ɪə(r,ɛə(r/æə(r,ær/ɔə/oə/ʊə(r/jʊə(r. I believe the proposed chart below is more accurate for the examples shown, but Wikipedia policy insists upon published sources rather than individual observation, however better the latter may be. The above list does not convey an impression of there being any real difference (from GA) that matters, perhaps because of the keywords chosen, but anyone hearing a Southerner speak instantly notices a (and sometimes profound) distinction.—dshep/2010.01.26

Dialect	"oʊ"	"ɒ"	"aɪ"	"ɜr"
RP	əʊ	ɒ	aɪ	ɜː
WaE	ou/oː*	ɒ	ai	ɜː
ScE	oː	ɔ	aɪ	ʌɾ ɛɾ ɪɾ
IrE	oː	ɑ~ɒ	ɔɪ	ʌɾ ɛɾ ɪɾ
CaE	o~oʊ	ɑ~ɒ	aɪ, əɪ	ɝ
GA	oʊ	ɑ	aɪ, (ʌi)	ɝ ʌɹ
SaE	oʉ (?)	ɑ	äː, əɪ	ɝ
AuE	əʉ	ɔ	ɑe	ɘː
NZE	ɐʉ	ɒ	ɑe	ɵː(ɾ)
SAE	œʉ	ɔ	aː	øː

First actual attempt at contributing to a Wikipedia article :) 83.80.137.241 (talk) 19:14, 5 November 2009 (UTC)

Indian English

I have been somewhat concerned that Indian English has not been included in the chart, seeing as a significant amount of Anglophones are from India and of these, the majority speak the national dialect and have an accent that suits. Would it be necessary to include a separate chart for consonants, because I am aware letters like /d/ and /t/ are retroflexed if Indian English is to be included?

R. John Lloyd (talk) 21:21, 11 March 2009 (UTC)

Is there a large enough (and sufficiently "connected") population of mother-tongue English speaking Indians for there to be a reasonably consistent "standard" phonology? Roger (talk) 21:31, 11 March 2009 (UTC)

Yes, surely Indian English is a well-established dialect, and I agree that we need to include it. Other important omissions are northern English, southern rural English and Estuary English/Cockney – all very distinct and spoken by large numbers of people. I think the dialect columns should cover both consonants and vowels in one chart, as consonants do vary too (for example, th/t/d in Irish English, f/v for th in Estuary/Cockney, the differences in use of dark l). Some earlier discussion on related points in Talk:IPA chart for English dialects/Archive 5#English regional English Richard New Forest (talk) 22:33, 11 March 2009 (UTC)

AFAIK, Indian English varies dramatically according to the native language of the area, so which would we choose as representative? kwami (talk) 08:24, 23 January 2010 (UTC)

There are three factors that may make it fairly simple to transcribe Indian English as a single entity

1. Although there are many languages, there are quite a few areal features that have to do with pronunciation (such as the presence of retroflexes).
2. British English has been taught in India for several hundred years.
3. There is probably a standard variety of English in India that, like GA or RP, is viewed at as most prestigious.

I don't know a whole lot about Indian English, but I'm sure there are resources out there that can help enlighten us. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 08:30, 23 January 2010 (UTC)

If there's a standard variety of English pronunciation in India that is viewed as most prestigious, it's RP. I've met Indians who have never stepped foot in Britain but whose pronunciation might have been cultivated at Eton and Oxford. +Angr 09:17, 23 January 2010 (UTC)

You could say the same thing about the United States. American English speakers tend to feel that British varieties English are "better" so that they even see Cockney as more correct. So I guess the question isn't what's most prestigious but what carries the most prestige in the context of Indian English. What's the variety featured on national media? — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 18:50, 24 January 2010 (UTC)

AFAIK, IE varies from region to region according to the phonology of the local language, and this might hold for broadcast English as well. kwami (talk) 23:40, 24 January 2010 (UTC)

Northern Irish

Seeing as England, Scotland, Wales and the Republic of Ireland (judged by the flag) are all on here, shall we add Northern Ireland as well? There is a description of it in Wells, Accents of English.

By the way, what is the evidence that most people in the Republic of Ireland use /ɔɪ/ in the PRICE words? Wells gives /ai/ here (page 419), and I think that he's right. Epa101 (talk) 19:57, 18 March 2009 (UTC)

There is an argument for including it, but aren't there are other, more distinct dialects to include first? Perhaps if there is room after some of those discussed above? Richard New Forest (talk) 20:29, 18 March 2009 (UTC)

Indian English sounds more important, although I can't say that I know a thing about it. As for the "regional" dialects of England, I'd say that they are less important and are changing with age very rapidly. Northern Ireland is much more different from RP than any region of England: also, it is constitutionally distinct as well. Epa101 (talk) 22:37, 18 March 2009 (UTC)

Note also that the dialect boundary between Northern Irish English and Southern Irish English does not correspond to the political boundary. People from Counties Donegal and Monaghan, and maybe Cavan as well, speak with a Northern accent although they're from the Republic. —Angr 22:46, 18 March 2009 (UTC)

That's interesting, although it's the same thing for the boundary between England and Scotland; people from Berwick-Upon-Tweed and the surrounding area sound Scottish (Wells, p.351). Does anyone know of any other comprehensive descriptions of Northern Irish English other than that in Wells? Epa101 (talk) 15:25, 21 March 2009 (UTC)

I don't agree that Northern Irish is "much more different" from RP than English regional dialects – perhaps somewhat more than most. Anyway, surely it's quite a bit closer to standard Scots than, say, Liverpudlian or Cockney is to RP. There are those who'd say it is Scots not English, so would not belong on this page at all... Richard New Forest (talk) 19:59, 21 March 2009 (UTC)

As far as I am aware, there is no official marker between English and Scots as languages. English is an official language in Northern Ireland, and no linguist claims that RP is spoken in Northern Ireland by anyone, so it seems justified to include it as an English dialect. I agree with you that it is quite similar to most Scottish accents, but I don't think that's an argument against its inclusion. I can't see how it can be justified to give every nation in the British Isles a column except Northern Ireland.

As regards Scouse or Cockney, what are the sources for the modern accents? I am a bit worried with English regional dialects that the information will be descriptions of very old dialects, as most of the sources that I've read have been. For example, on Britannica's article for the English language, it claims that Northerners use /i/ in blind, find, mind and replace initial /tʃ/ with /k/, both of which died out forty years ago. Epa101 (talk) 13:07, 22 March 2009 (UTC)

Personally I'd agree that English and Scots are dialects of the same language. However, many Scots speakers regard their vernacular as a true language, and object to calling it a dialect of English – their (quite logical) reasoning is that it developed continuously in parallel with English, but (unlike all other dialects of English) at no stage was it spoken in England itself. However you look at it, Scots proper is a lot more than a mere accent, being pretty much incomprehensible to many standard English speakers, having a considerable separate lexicon and an established orthography (not to mention its own Wiki: Scots Wikipædia). It's probably as different as many recognised languages are from each other, and I suspect that the reasons that it's not generally regarded as such are largely political. It reminds me of that well-known definition of a language as "a dialect with an army".

As to the age of sources, I think that's something we just have to work with, as with all other Wikipedia sources. If all we have is early 20th century sources, then that's what we'll have to give. In any case, though regional English accents have certainly changed over the last century or so, surely they remain at least as distinct as GA, New York, Australian, South African etc. Richard New Forest (talk) 19:01, 22 March 2009 (UTC)

Whether Ulster Scots is a dialect of Scots or of English (or a separate language altogether) is sort of beside the point - both Ulster and Scotland have, in addition to Scots, local accents of Standard English (Northern Irish English and Scottish English respectively), which is presumably what this page would be reporting on anyway. —Angr 13:13, 23 March 2009 (UTC)

Good point. Richard New Forest (talk) 20:38, 23 March 2009 (UTC)

Leaving aside the dialects within England for a moment, does anyone object to the creation of a Northern Irish column based on Wells' description of it? I'll wait for at least a week before proceeding.

Another problem with having English regional dialects is whether there will be enough space. If we are to add Indian English and Northern Irish English, the table would be quite big already. Perhaps a separate table would be needed or perhaps even a separate page. Epa101 (talk) 19:48, 23 March 2009 (UTC)

If we had consonants and vowels in one table, there would be plenty of width available for half a dozen more accents at least. Richard New Forest (talk) 20:38, 23 March 2009 (UTC)

Rather than put them in the same table, I think we should arrange the consonant table and vowel table vertically rather than horizontally. That way both of them have room for horizontal growth (though the consonant table presumably doesn't need much). —Angr 06:20, 24 March 2009 (UTC)

I have made an effort to add NI English to the table, but the format has become very difficult now. When I pressed preview, I realised that it had all gone wrong as soon as merged cells came into it. I'll try again tomorrow. Epa101 (talk) 21:29, 1 April 2009 (UTC)

Possible source for some references

I think that this book might be quite useful in providing citations. In addition, I think that the IPA table for the North of England in this book demonstrates why it would not be practical to include a Northern English column on this article's table.

At the time of writing, the phonetics for Indian English is not included in the book preview, which is unfortunate as that would have been a useful addition to the table. Epa101 (talk) 14:09, 12 April 2009 (UTC)

Links

One of the external links: http://www.arts.gla.ac.uk/IPA/IPA_chart_(C)2005.pdf Has been moved or removed. 82.181.193.196 (talk) 08:33, 25 May 2009 (UTC)

Corrected link to [3], from the new IPA site. Drydic guy (talk) 10:29, 25 May 2009 (UTC)

What is a row?

I have what I think is a pretty basic question to which an answer is both necessary to interpret the table and not given in the article: what do the rows of this table represent? When I see different symbols within a row, does this mean the nationalities represented by the columns use different sounds, or different *symbols*? For instance, the symbol "eə(ɹ)" occurs only for English as spoken by the English themselves and by New Zealanders. Is this a sound unique to these nationalities? Or is it a symbol unique to these nationalities, representing the same sound as one or more of the other symbols in the same row?Paalexan (talk) 02:29, 17 June 2009 (UTC)

Each row is a sound that is distinct in some dialect or other. That is, if you wanted to design a perfect English orthography, which would work equally well for any dialect, you'd need a letter (or digraph) for each row. The IPA symbols are supposed to represent the pronunciation of that sound in each of the dialects listed, so if the IPA is different between two dialects, say for the ee sound in feet, then it is pronounced differently in those two dialects. Or at least ideally: since our sources are different for different dialects, some of the differences may represent differences in transcription rather than actual pronunciation. kwami (talk) 02:45, 17 June 2009 (UTC)

Yod vowels

I have moved the /juː/ row to be under /uː/ , just as /jʊər/ is under /ʊər/. I question whether these should be included at all. The strongest reason to have them is just to remind people to include in their transcriptions a sound that is somewhat opaquely represented in English writing. I am also going to label them as part of the GOOSE and CURE lexical sets respectively, which is how Wells classified them. CURE = /ʊər/ but not /jʊər/ is especially misleading. CURE is itself a rather unfortunate name, since that has the NURSE vowel for me. But it's standard and those pre-r vowels are so unstable, it would be hard to find a satisfactory substitute. Brock (talk) 19:01, 4 July 2009 (UTC)

/juː/ is different from /uː/ in that it's actually closer to /jʉː/, at least a distinction like that used to exist in AmE. http://www.npr.org/templates/story/story.php?storyId=5220090 Temporal Fugitive (talk) 20:24, 22 April 2010 (UTC)

Update: According to the Atlas of North American Dialects, most AmE dialects front /uː/. "...In all but a few areas,the first part of this vowel has shifted strongly towards the front of the mouth, like French u or even like English i in bit... Except in the South, this fronting does not occur before /l/, so there is a strong contrast in the vowels of too and tool."

It makes the most sense to list both /u/ and /ʉ/ for the /uː/ dia-phoneme in GA and CaE and just /jʉː/ for the /juː/ dia-phoneme. Seriously, the word "computer" pronounced with a /ju/ (as in /j/ + the /u/ in "tool", "pool") would sound really awkward to most AmE speakers. Temporal Fugitive (talk) 01:04, 26 April 2010 (UTC)

That's not how it's typically transcribed, though. Most dialects pronounce /ʌ/ as [ɐ] but because it's so rarely transcribed that way, we have gone with the more widely known transcription. It's the same way with the vowel of boot. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 01:59, 26 April 2010 (UTC)

I agree, and more generally, this page is not the place for discussing details of English dialectology or phonology. These observations can be better made at American English or one of the articles on a specific dialect of American English. +Angr 06:01, 26 April 2010 (UTC)

missing splits

apparently need kit-bit split for SAE and you-ewe for British Isles peripheral. kwami (talk) 20:31, 31 August 2009 (UTC)

too ambitious?

Is this page just promising too much? It's really an attractive idea, but when it comes down to it, phonetic variation in English is really just too complex to allow for such a comparative chart. Even limiting it to a few varieties isn't likely to work because people will just get pissed off that their variety is left out. One thing that might be done to simplify things is to cut out categories, such as marginal sounds. I'm not sure that glottal stops should be considered marginal anyway. They're sure common in many dialects as an allophone for /t/. mnewmanqc (talk) 15:59, 1 September 2009 (UTC)

Unicode representation

This is a very useful chart. Another use for this would be to line this up to Unicode representations of the characters used and I for one can see applications for this in generating TTS applications etc. It probably would clutter up what is already an already over-encumbered page and I propose to create a separate page with the bare minimum of information against Unicode representations. Any thoughts? Sjc (talk) 19:21, 25 September 2009 (UTC)

I don't understand. What would be the point of this extra page? — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 22:00, 25 September 2009 (UTC)

It would be a useful reference point for a subset of Text-to-Speech & Speech recognition programmers who use IPA as one of the base points for TTS construction. Expect a lot of future activity in this area. It may well be more appropriate to place this somewhere in Wikibooks with a linkie though. Sjc (talk) 08:42, 26 September 2009 (UTC)

But aren't the characters already in unicode? — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 16:48, 26 September 2009 (UTC)

I haven't articulated this clearly and you're quite right to question my objective. Having gone away and had a good think about what was needed, what this boiled down to was a one-page (or sub-optimally one section) per implementation mapping containing 3 basic columns: Pan-English, national, and examples. On reflection, as I said previously, while this more obviously belonged in Wiktionary or possibly Wikibooks, I probably will not do it as page but redo this as an XML schema, DTD & document since its primary intended consumer is more likely computer than human and make it available via Wikicommons. Sjc (talk) 06:27, 5 October 2009 (UTC)

BATH vowel

In Southern England the vowel of bath and most /æ/ before fricatives is the same vowel as father. What would the compromise IPA transcription be for this BATH vowel? I suspect and would favor /æ/ as it's more common and more conservative. This is relevant for English words which may need to be transcribed for clarification or demonstration and for foreign words with the Israeli city Jaffa being a possibly affected. I was prompted to ask this by the transciption of photograph as /ˈfoʊtəɡrɑːf/ in the English orthography article. Jackessler (talk) 00:11, 30 September 2009 (UTC)

Well, we have a few choices. We can transcribe such words twice, once for each vowel sound; we can choose one over the other, as you suggest; or we can introduce a new letter—<a> has been suggested—to indicate a vowel that varies per register. This hasn't been discussed at any length. If the variation is largely predictable, the way that /j/ dropping is predictable in GA, then we should go with your suggestion; exceptions can be marked by double transcribing. If it is not all that predictable, then we can go one of the other two routes; objections might be raised about introducing a new distinction to remember, and <a> suffers the problem that it is often used to mean specifically /æ/ or specifically /ɑː/, though I think I've managed to eliminate most cases of that which transclude an IPA template. kwami (talk) 00:29, 30 September 2009 (UTC)

James Murray in the first edition of the Oxford English Dictionary used just such a symbol to represent the variable sound of the vowel in words such as 'pass' and 'chant'. Unfortunately the symbol he used was the script a (=ɑ) which, if you didn't read the notes would assume to be IPA /ɑ/. —dshep/2010.01.22 —Preceding unsigned comment added by 98.88.124.138 (talk) 03:14, 23 January 2010 (UTC)

(I'm copying this over to the WP:IPA for English key, where it is more relevant.)

Ordering

The general page about IPA directs people to this page for a guide to pronouncing IPA symbols. It might be more useful for that purpose if the symbols were in the same order as the Latin alphabet characters they resemble, or if the tables could be converted to one of those sortable tables where you can click on the column headers to sort them. With the symbols being in no predictable order, a person has to scan the entire table for each glyph they want to look up, instead of being able to hone in on the glyph they're looking for with a nice mental binary search. --208.79.47.40 (talk) 01:32, 4 December 2009 (UTC)

Never mind, there are already a few other pages that have the symbols in the same order as the Latin alphabet characters they resemble, such as Help:IPA and Wikipedia:IPA_for_English. In fact, those are the two pages linked at the top of the IPA main article. The link to this page isn't until further down in the article under the "Description" section. --208.79.47.40 (talk) 01:44, 4 December 2009 (UTC)

Other dialects

There could also be an attempt to add other versions of North American English, including Southern American, African-American, and Newfoundland English. These differ considerably from either American or Canadian English. I would argue against including Indian English, since as I understand it, its phonology differs upon the influence of the local languages, and for most speakers it is a second language. This site should limit itself, and I think demonstrating vowel realization on the basis of genetic descent alone is a good rule of thumb. Also, concerning the Englishes of the Caribbean, while being spoken natively by the population, it hasn't been demonstrated that they descend entirely from a parent version of English or through some "creolization" process, or through areal influences. If Caribbean phonology can be generally summarized, then it could be helpful to add a column, but otherwise, for the sake of clarity, I would leave it out. Also, we need to decide if we are describing spoken dialects or national "standards," which are really either just media standards or stereotypes. The choice would make a great deal of difference, as one is sociolinguistically viable, while the other may or may not be. —Preceding unsigned comment added by 128.135.24.92 (talk • contribs) 05:14, 15 February 2010

In my opinion, we should include all countries that have a significant number of native English speakers. http://en.wikipedia.org/wiki/List_of_countries_by_English-speaking_population#List_in_order_of_native_speakers

At least the ones where the majority of the population is composed of native English speakers (Jamaica, Trinidad and Tobago, Singapore, Guyana). Temporal Fugitive (talk) 20:43, 22 April 2010 (UTC)

rhotic schwa in Scottish

Can Scottish English end words in /ər/, i.e. LETTER words? According to this chart it can, giving [əɾ]; but Scottish English#Phonology seems to imply that it has [ə] for this set. Lfh (talk) 10:15, 2 April 2010 (UTC)

At first, I was confused by that too. It just means that the sound before the R is a schwa. Scottish isn't supposed to drop Rs. Temporal Fugitive (talk) 01:56, 23 April 2010 (UTC)

Table layout

What do you guys think about something like this:

Dia-phoneme æ Lexical set TRAP Examples lad, bad, cat

Dia-phoneme ɑː Lexical set PALM Examples father

Dia-phoneme ɒ Lexical set LOT Examples not, wasp

RP ScE SAE WaE SCE æ a æ a ɛ AuE CaE GA IrE NZE æ æː æ æ ɛə eə ɪə ɑ/æ ɛ

RP ScE SAE WaE SCE ɑː a ɑː aː ɑ AuE CaE GA IrE NZE aː ɑ/ɒ ɑ ɑː ɐː

RP ScE SAE WaE SCE ɒ ɔ ɔ ɒ ɔ AuE CaE GA IrE NZE ɔ ɑ/ɒ ɑ ɑ ɒ

Doesn't have to be a replacement for the large table, but it could be a supplement. The large table is good at showing mergers, this isn't. Editing and adding things to these mini-tables is a lot easier. The big table can be a little confusing, too. Temporal Fugitive (talk) 00:40, 23 April 2010 (UTC)

So what would be the point of these additional tables? I don't see any new information.

Also, does GA really exhibit the bad-lad split? — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 00:59, 23 April 2010 (UTC)

No, it's not the bad-lad split. It's æ-tensing. Non-southern speakers tense æ in certain environments (pre-nasal, pre-g, and/or phonemically). Inland North tenses æ in all environments. Some accents (e.g. Western) use the whole spectrum. Tensed æ variants are quite widespread in Broadcast speech. As for the purpose of these tables, I personally think that it's actually easier to compare just one element. I often find myself scrolling up or down to check which English variant I'm looking at. The big table is in a standstill, too. It's impossible really hard to add additional columns without messing up the table. Temporal Fugitive (talk) 12:00, 23 April 2010 (UTC)

I don't think æ-tensing should be shown in this chart. The chart is supposed to be of phonemes, and æ-tensing in General American (as opposed to Philadelphia & New York accents) is not phonemic. +Angr 12:26, 23 April 2010 (UTC)

You're right. Likewise, Canadian raising isn't phonemic. Yet, these are very important aspects of US and Canadian English. I think we should keep them. At least the most common tensed æ variant (the one that is and has been in the main table for a long time). Otherwise it wouldn't be a very good representation of these English variants (isn't that what we want to do, create the most accurate comparison of English variants?) Temporal Fugitive (talk) 12:53, 23 April 2010 (UTC)

Doesn't Canadian English have the father-bother merger? --Atemperman (talk) 18:45, 29 April 2010 (UTC)

Yes, it does. My mistake, sorry! What do you think about the concept of these tables? Their contents aren't really that important right now anyway. Temporal Fugitive (talk) 20:05, 29 April 2010 (UTC)

Irish Rs

http://en.wikipedia.org/wiki/Irish_English "With some local exceptions, /r/ occurs postvocally, making most Hiberno-English dialects rhotic.[4] The exceptions to this are most notable in Drogheda and some other eastern towns, whose accent is distinctly non-rhotic. In Dublin English, a retroflex [ɻ] is used (much as in American English). This has no precedent in varieties of southern Irish English and is a genuine innovation of the past two decades. Mainstream varieties still use a non-retroflex [ɹ] (as in word-initial position). A uvular [ʁ] is found in north-east Leinster.[5] /r/ is pronounced as a postalveolar tap [ɾ] in conservative accents. Mícheál Ó Muircheartaigh and Jackie Healy-Rae are both good examples of this."

I think it's safe to change all Irish Rs to ɹ. —Preceding unsigned comment added by Temporal Fugitive (talk • contribs) 01:25, 23 April 2010 (UTC)

But we're using ‹r› broadly to stand for any r-like sound in English, not narrowly in the sense of an alveolar trill. See the consonants table, for example. +Angr 05:24, 23 April 2010 (UTC)

Only in the diaphonemic column, in all the other columns we're being more phonetically accurate. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 06:06, 23 April 2010 (UTC)

Phonemic or phonetic? Let's read the main article.

"This chart lists the diaphonemes of English, in bold, followed by their most common phonetic values, in plain text. For the vowels, a separate phonetic value is given for each major dialect."

As it stands and has stood for the longest time, the chart is clearly labeled as representing PHONETIC VALUES of phonemes. A purely phonemic chart would be useless as phoneme symbols are uniform across the globe (when using the IPA), unless your only point were to show which dialect has what mergers and/or splits. By listing a different than standard symbol for whatever dialect, we are in fact giving it a phonetic value! Thus, all common phonetic values for each phoneme in each dialect in the chart should be included. —Preceding unsigned comment added by Temporal Fugitive (talk • contribs) 21:52, 12 June 2010 (UTC)

I agree. Is there some disagreement about this? — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 21:57, 12 June 2010 (UTC)

That's pretty much what we have. What are we missing? — kwami (talk) 22:27, 12 June 2010 (UTC)

I'm not saying that we're missing anything. Although I could contend that /ʌ/ is misrepresented, as that is not its phonetic value for many dialects. I know that's the way it's usually transcribed because that's what it used to sound like, but sound change doesn't change the phonetic value of the symbol itself, which still represents and will represent a different sound in non-English IPA usage as it is supposed to represent an absolute, very well defined sound (kind of mars the I in IPA if we insist on using this symbol, doesn't it?).

Anyway, Wikipedia contributors who blindly think that this chart is phonemic gave impetus for creating this section. That was my original point :) Temporal Fugitive (talk) 23:02, 12 June 2010 (UTC)

Oh, I see. You're talking about this edit summary. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 23:06, 12 June 2010 (UTC)

Yes! That and a couple posts on this page (e.g. read Angr's comment in the "Table Layout" section). Temporal Fugitive (talk) 23:14, 12 June 2010 (UTC)

This page was never supposed to be a phonetic chart. Back in the day, it used to feature the most common phonemic charts for RP, General American, and General Australian, according to several reliable sources. The page has since grown out of control and it's now mostly unsourced. We can't really turn this page into a phonetic chart--we would simply have no sources whatsoever. The charts found in the sources provided--such as Kenyon & Knott, Harrington, etc.--are phonemic and not phonetic.

Furthermore, exact phonetic values may vary widely, and vowel allophones are not shown *anywhere* in the table except for the GenAm TRAP vowel. By the way, the very phoneme /æ/ has a few more allophones besides [æ] and [eə]--not just in General American. I'm [dʒæˑkɫɜmbɚ] and I approve this message. 23:32, 13 June 2010 (UTC)