Talk:English-language vowel changes before historic /r/

From Wikipedia, the free encyclopedia
Jump to: navigation, search

Names of mergers: words or IPA[edit]

@Wolfdog: I personally like the word-based merger names. It's nice for the mergers to be named after words that are affected by them. Having IPA in section headings also irritates me. Could you explain what you mean by saying that merger names that use words are "conjectural"? It seems that some of the names are indisputably correct: Marymarrymerry, for instance. — Eru·tuon 06:34, 29 August 2016 (UTC)

  • I agree that the word-based names sound more pleasant and I'd stick with them if they were all well-established names. Yes, a few are... such as the Marymarrymerry merger (though I can also find articles that call it the marrymerryMary merger). However, a name like "squarenear merger", for example, is not readily found in scholarly sources; we could just as easily call it the cheerchair merger (a name I myself initially used based off the ANAE's description and examples) or countless other names, such as the beerbare merger, fearfair merger, etc., etc. However, by using IPA, we don't need to invent our own names for every type of documented merger, which would risk becoming original research as well as being constantly vulnerable to arbitrary renamings.
In response to your irritation over having IPA in section headings, I'll ask: why do you feel irritated? Are you specifically irritated that I used the IPA code/template "IPAc-en", or just that I used IPA in general? I'd be happy to go into more detail about either, though I won't be able to discuss much with you if you find IPA inappropriate for section titles simply as a matter of taste or aesthetics. Wolfdog (talk) 06:55, 29 August 2016 (UTC)
What's the problem with the name "squarenear merger"? These keywords are part of Wells's lexical sets, which are in widespread use among scholars. I'm surprised that anyone interested in phonetics would think that it's OR to use such a name.
When it comes to IPA in section headings, I'm totally indifferent. As long as we're able to link to a certain section, I don't really care how we call it - as long as it's appropriate (and IPA alone certainly is!) Mr KEBAB (talk) 10:49, 29 August 2016 (UTC)
There's no inherent problem with such a name. My concern is with consistency. I'm fine with going forward using Wells's lexical sets if that's what our consensus is (though it seems like Mr KEBAB is OK with retaining IPA names), but then will all merger names use his sets? For example, will we rename the low back merger of words like cot and caught the LOT-THOUGHT merger? Will we call the merger of words like horse and hoarse the NORTH-FORCE merger? Will the merger of con and Kahn be the LOT-PALM merger? Will the weak-vowel merger now be called the commA-horsES merger? I'm just asking that we make a decision that provides consistency. On the other hand, I would argue that IPA is the perfect tool for consistency, while Wells's sets can only provide some consistency (e.g. what would we call the flour-flower merger using Wells's sets?) Wolfdog (talk) 15:23, 29 August 2016 (UTC)
Yes, I am OK with retaining the IPA, just as I'm OK with using established names for mergers, as well as names derived from Wells's lexical sets.
Actually, the cot-caught merger is indeed called "THOUGHT-LOT merger"... by Wells himself, in the AoE (from page 473 onwards). One of the drawbacks of using that name is that it may seem to imply that there's a distinct LOT vowel in North American English, when that's obviously not the case (it merges with PALM in most NA accents, and with THOUGHT alone in Boston).
My take on that is that we don't really have to be consistent, unless there are WP rules that would require that. When there's an established name for a certain merger, we should strongly prefer it over a name that uses Wells's lexical sets, unless it itself is quite popular. In the last case, the choice is completely arbitrary.
AFAIK Wells doesn't provide a way of distinguishing "flour" from "flower", but maybe that's just my bad memory. Who knows.
Maybe using IPA is the easiest solution after all... but then again, common names must be listed at the beginning of each section, otherwise the article would be incomplete. Mr KEBAB (talk) 17:49, 29 August 2016 (UTC)
My irritation is mainly aesthetic, and I guess it isn't really discussable. But there are two other concerns.
Using IPA makes the section headings less accessible to readers who haven't studied the IPA. They can see the tooltips when they're browsing the page, but if they are looking at the Table of Contents (which doesn't have tooltips), they will find the merger names incomprehensible.
The IPA system is sometimes misleading. It is based on the phonemes of old-fashioned RP. It does not describe the phonological systems of other dialects, where their phonemes differ from old-fashioned RP, and therefore gives a misleading picture of mergers that happened in other dialects.
For instance, the merger of north and force involved /ɔː/ and /ɔə/ in RP; in North American English, it actually involved /ɔr/ and /or/. The different phonemic representations imply different processes: smoothing in RP, loss of height distinction (tense–lax neutralization) in NAE. This fact is obscured by the use of the diaphonemic symbols. (Of course, it is also obscured by both processes being described in the same section.)
With Mary, the situation is worse: /ɛər/ is based on old-fashioned RP, which doesn't even undergo the merger. A more accurate symbol might be /e(ː)r/. The merger was, like north and force, a loss of height distinction.
With north/force/thought and cure, I think the merger probably involves two monophthongs /ʊː/ (/ɵː/) and /ɔː/ (/oː/). Using /ʊər/–/ɔər/ (setting aside the question of the /r/) is anachronistic: it's old-fashioned RP.
The IPA system is really only a substitute for the lexical sets. It is not an actual transcription of the vowels that underwent the mergers. So, I think using the lexical sets or minimal pairs/triads would make more sense. Perhaps the fact that the IPA symbols are sometimes inaccurate could be explained in each section, but there would still be a segment of the readership who would find that confusing or even nonsensical. — Eru·tuon 23:38, 29 August 2016 (UTC)
You may be right about taking people that don't know IPA into consideration, we seem to have not taken that into full consideration...
You need to keep in mind that phonemic IPA does not have to be phonetically accurate, and we should expect our readers to be aware of that (they're smart, I hope!) The fact that we choose symbols that are more appropriate for the conservative variety of RP rather than GA is just an arbitrary decision, and a quite smart one, due to the fact that the former variety has more vowels.
Bear in mind that both NORTH and FORCE also appear prevocalically, and in that environment the /r/ is mandatorily pronounced in RP (and other non-rhotic standards of English pronunciation). Mr KEBAB (talk) 01:20, 30 August 2016 (UTC)
The Wikipedia system is diaphonemic, not phonemic. So, it departs more from the phonetic and phonological realities of existing varieties of English than a phonemic transcription system would. It can't be seen as an honest attempt to represent the actual phonological system of any existing variety of English.
I don't think readers or even editors will necessarily be aware of the distinction between phonetic, phonemic, and diaphonemic transcriptions. People have many levels of phonetic and phonological education. Some of them are "dumb" from our perspective and will be confused by our diaphonemic transcriptions. I have seen evidence of this.
There was recently a discussion where someone incorrectly stated that John had the same phonemes, /ɒn/, in all English dialects. But there is no /ɒ/ phoneme in GA; the Wikipedia transcription is diaphonemic, and the diaphonemic symbol /ɒ/ represents the phoneme /ɒ/ in RP and the phoneme /ɑ/ in GA. The two can't be said to correspond, because there is a different number of low vowels in GA and RP, with different phonological features distinguishing them, and they do not occur in the same lexical sets.
And there are the cases where someone has created a recording of a vowel, like [ʌ], using the (RP or GA) vowel that is represented by the diaphonemic symbol /ʌ/, rather than the official phonetic pronunciation of the symbol, which should be fully back and sounds odd to my ears when it's done right.
Not sure why you mention /r/. I'm aware of linking r, and my concerns have to do with vowels. — Eru·tuon 02:14, 30 August 2016 (UTC)
@Erutuon: That's true, but it doesn't matter in this case. Phonemic or diaphonemic is still not phonetic, and even in case of phonetic transcription you can't expect full accuracy all the time (well, unless you're making it yourself ;)).
Hmm, maybe I'm overestimating our readers, which is sad to hear.
Maybe I was being pedantic - the point was that intervocalically as well as word-finally immediately before a word that starts with a vowel, /r/ is retained and so transcribing those vowels as /ɔːr/ and /ɔər/ is actually correct (and so would be transcribing THOUGHT as /ɔːr/ since, as you probably know, failing to pronounce intrusive /r/ sounds fake in RP, and may lead to suppressing the linking /r/ where its presence is called for (see Lindsey (2013))). Mr KEBAB (talk) 16:46, 30 August 2016 (UTC)
@Mr KEBAB: It would be kind of funny to see thought transcribed /θɔːrt/, but it would sort of be consistent... ^.^ — Eru·tuon 01:14, 31 August 2016 (UTC)
@Erutuon: Yeah... maybe I didn't make myself clear enough (I wouldn't use it prevocalically preconsonantally). Mr KEBAB (talk) 01:47, 31 August 2016 (UTC)
The term "near-square merger" is fairly common in the scholarly literature. AJD (talk) 04:03, 30 August 2016 (UTC)
(And in general I do think we should use names that are fairly common in the literature when possible. Sometimes there's more than one commonly-used name for a merger, like "horse-hoarse merger" and "north-force merger", in which case we can use either or both. I think naming phonemic mergers in terms of IPA symbols is more likely to be confusing than helpful, perhaps unless such a name is actually commonly used. AJD (talk) 04:07, 30 August 2016 (UTC))
Thanks for everyone's thoughtful input. So, going forward: the consensus seems to be that the order of naming protocol is that we go with a common name first, and if there is none we go with names based on Wells's sets second, and if there is none (or if it would still be unclear) we can use IPA names. Doable? My other major concern, aside from consistency, was that anyone could make up a name that is not actually well-sourced, like the fern-fir-fur merger or steer-stir merger which seem to be names that have been originally invented by WP editors. Wolfdog (talk) 11:59, 30 August 2016 (UTC)
@Wolfdog: You're welcome. If you're concerned about made-up names, just add the page to your watchlist and moderate it. Don't worry too much about preventing vandalism or other unwanted edits (if that's what you mean), you can't do much about it. It's Wikipedia after all... Mr KEBAB (talk) 16:46, 30 August 2016 (UTC)

Are there any "official" names for the tire, tower, tar–related changes? I think Geoff Lindsey called it smoothing, but that's a general word for the category of change (reduction of a triphthong to a long monophthong). And it might not be appropriate for the American English version, which probably retains the /r/... — Eru·tuon 01:14, 31 August 2016 (UTC)

I'm not aware of any, but why do you think that "smoothing" is an inappropriate name for such changes in AmE? I'm not sure if it implies non-rhoticity, but it doesn't look like that. Mr KEBAB (talk) 08:01, 31 August 2016 (UTC)
Well, I was thinking maybe it was a corollary of the monophthongization /ai/ > /aː/ (or something like that) that happens in Southern American English. Perhaps I am wrong. I haven't read anything about the AmE version of the sound change, only what Geoff Lindsey says about SSB smoothing. — Eru·tuon 08:27, 31 August 2016 (UTC)
@Erutuon: No idea if that's the case, but maybe you're right. It'd be good to have sources on that, speculation is fine only as far as talk pages are concerned. Perhaps the Atlas of North American English is the answer? Maybe you or @Wolfdog: could check your local libraries and look through that book? Mr KEBAB (talk) 17:08, 31 August 2016 (UTC)
@Mr KEBAB: Unfortunately, I searched in the online catalog, and my local library doesn't have it. Fortunately Wolfdog found a reference. — Eru·tuon 20:06, 31 August 2016 (UTC)
@Erutuon: Thanks for looking it up. Mr KEBAB (talk) 20:07, 31 August 2016 (UTC)

This one is on a different topic: I think we need to enclose Wells's terminology in the sc template. What do you think? Mr KEBAB (talk) 02:21, 31 August 2016 (UTC)

Do you mean the symbols for lexical sets: writing NORTHFORCE with small caps, but horsehoarse with italics? If so, I agree. — Eru·tuon 07:56, 31 August 2016 (UTC)
Yes. The reason I propose that is that that's what's done in AoE and on many WP pages. It'd be nice to see consistency in that case, and using the normal caps is not a very good option. Mr KEBAB (talk) 08:01, 31 August 2016 (UTC)
I agree that this makes sense too. The ANAE calls the phenomenon that causes the tire-tar merger "/ay/ glide deletion," "/ay/ monophthongization," or "Stage 1 of the Southern Shift." This may not be a perfect merger, by the way, and, actually, it's never explicitly mentioned as a merger in the ANAE as far as I can see, though there is evidence of the tire-tar merger, or something like it not just in Southern American English, but also some Midland American English: "A considerable amount of glide deletion is found just north of the red isogloss [the Southern United States], in Midland cities close to the South. However, in these communities /ay/ glides are deleted only before resonants (nasals and liquids), in time, nine, tire, mile, etc." Erik Thomas, who studies Southern American English, also describes how many Southern speakers "show merger of FIRE with START, resulting in [ɒɚ~ɑɚ~ɒː~ɑː]. This merger is highly stereotyped and, consequently, is most typical of older, working-class, and less educated speakers" and soon after also mentions that the "merger of POWER with START occurs infrequently, mostly among the same groups who merge FIRE with START." Wolfdog (talk) 19:25, 31 August 2016 (UTC)
Great find Wolfdog, thanks! Mr KEBAB (talk) 19:36, 31 August 2016 (UTC)

Could someone explain the poor-pour merger to me?[edit]

Unfortunately, I feel like a fool. All my life I have heard poor and pour pronounced the exact same way: [pɔːɹ]. And also, I have heard pored also pronounced that exact same way. However, I have heard shore and sure pronounced differently, the former as [ʃɔːɹ] and the latter as [ʃʊɹ]. So, could somebody please explain to me what the distinction between poor and pour is? Thank you. (talk) 03:34, 17 November 2016 (UTC)

Better yet, if somebody could add an audio on to the page as an example of a speaker without the poor-pour merger, as such occurs for the hurry-furry merger and the Mary-merry-marry merger, that would be great. Thank you. (talk) 03:38, 17 November 2016 (UTC)
Just put the vowel of sure that you describe into poor and you've got the poorpour distinction. — Eru·tuon 04:36, 17 November 2016 (UTC)

Thank you, sir/ma'am. (talk) 04:11, 18 November 2016 (UTC)

Request to add a table[edit]

I think a possible table of the words in "horse" and "hoarse" classes may be helpful. I have the merger myself, and I find that the current summary isn't very easy to follow. Could we possibly do that? Thank you.LakeKayak (talk) 17:50, 30 December 2016 (UTC)

It seems nobody has an opinion. However, if a table is to be added, then someone who is familiar the horse–hoarse distinction should do the job. And I am not. Therefore, I would be the least-qualified, per se, to do the task.LakeKayak (talk) 20:32, 31 December 2016 (UTC)

I added the table. Any help with the table is appreciated. Thank you.LakeKayak (talk) 21:48, 1 January 2017 (UTC)

Request of changing the name Nurse merger[edit]

Currently, the page refers to the merger of the vowel in fern, fir, and fur. However, the page Phonological history of English vowels refers to the name of the page as the "fern–fir–fur merger." For one, I would prefer consistency in the naming. And for another, the term "nurse merger" seems rather vague. It doesn't describe which vowels were merged. So, I suggest the name of the section be changes to "fern–fir-fur merger." Thank you.LakeKayak (talk) 20:00, 1 January 2017 (UTC)

It does, but you have to know the history of English vowels. "Fern–fir–fur merger" is less ambiguous only if your accent distinguishes them - and the vast majority of native speakers don't. They have to guess that the vowels in question are DRESS, KIT and STRUT. Mr KEBAB (talk) 04:54, 2 January 2017 (UTC)
The spellings of the three words give them a pretty good clue. But we should try to use names that are used in reliable sources - I don't know whether this one is. W. P. Uzer (talk) 09:53, 2 January 2017 (UTC)
@Mr KEBAB: My own dialect does not distinguish the vowels. And I still find the name ambiguous. As you said, sir, in order to understand the name, you have to know the history of English vowels. However, I feel that the name of the merger should identify which vowels are merged, whether or not you know the history of the English vowels. Thank you. LakeKayak (talk) 12:35, 2 January 2017 (UTC)
Also, can we at least be consistent from page to page with naming?LakeKayak (talk) 12:56, 2 January 2017 (UTC)
I think that if we want to use a truly unambiguous name, we should go for "DRESS-KIT-STRUT merger before /r/". It's hardly OR, as Wells's lexical sets are widely used. Mr KEBAB (talk) 17:57, 2 January 2017 (UTC)
It's only the vowels in those words that are merged, not the words themselves, of course, so that title would need modifying. Wouldn't it be better just to quote those lexical sets at the start of the article or section? The vowels ɛ, ɪ, and ʌ or ʊ seem to have merged to /əː/ rather than /ɜː/ before r in British English. Do they merge to /ɜː/ in American English? Dbfirs 21:06, 2 January 2017 (UTC)
@Dbfirs: Unfortunately, little to no distinction is made [ɚ] and [ɝ] in American English. However, if I had to guess, I would say they merge towards [ɜ].LakeKayak (talk) 21:36, 2 January 2017‎ (UTC)
Thanks. The article is obviously written from an American viewpoint. Should we include the British version? Dbfirs 21:47, 2 January 2017 (UTC)
Go ahead. I have no objection.LakeKayak (talk) 21:55, 2 January 2017 (UTC)
Thanks. I've done that. We still haven't decided on the title. Dbfirs 22:15, 2 January 2017 (UTC)
The transcription /ɜː/ is meant to be RP. It was probably designed for older RP, which often had a pretty open er vowel (open-mid to near-open). Nowadays, the generic schwa symbol /əː/ might be more accurate, though I'm sure the pronunciation varies quite a bit. The American merged sound is a r-colored vowel mid-ish central vowel, as mentioned by LakeKayak or a syllabic approximant with a bunched tongue (which I think means that a part of the tongue a bit back from the tip comes close to the roof of the mouth somewhere in-between the alveolar and velar regions). So symbols that are used for the American sound include ⟨ɝ, ɚ, ɹ̩, ɻ̩⟩. — Eru·tuon 23:57, 2 January 2017 (UTC)
Thank you for the clarification. We do seem to use /ɜː/ when we mean /əː/ in Wikipedia. I was taking my pronunciations from those in our IPA chart and the representation in the big Oxford English Dictionary which uses /əː/. I'll listen carefully to see if anyone in England still uses /ɜː/ for the merged vowels. Dbfirs 08:57, 3 January 2017 (UTC)
@Erutuon: I beg your pardon, sir. Do you happen to know a possible name for the merger that would be less ambiguous? Any help is appreciated. Thank you.LakeKayak (talk) 00:20, 3 January 2017 (UTC)
Sorry, I don't have any great ideas for what the merger should be called. I kind of like fernfirfur, but if it's a name that's invented by us, best not to use it. — Eru·tuon 05:05, 3 January 2017 (UTC)
Thanks, anyway.LakeKayak (talk) 15:17, 3 January 2017 (UTC)
@Dbfirs: Wells's lexical sets (DRESS, KIT, STRUT, etc.) do represent vowels. See lexical set. Mr KEBAB (talk) 08:02, 3 January 2017 (UTC)
Yes, my point was that we would need to include the term "lexical sets" or "vowels" in the title, not just the words. People who have read Wells would understand what we meant without explanation, of course. Dbfirs 08:57, 3 January 2017 (UTC)
If Wells calls it the NURSE merger, then it seems most reasonable to call it that, rather than invent our own name. The details about what it means will be in the text. W. P. Uzer (talk) 11:19, 3 January 2017 (UTC)

In the section "Nurse merger," it does say that a variant to the name, that is sometimes used, is "bird–term–nurse merger." Could we use that name?LakeKayak (talk) 15:35, 3 January 2017 (UTC)

As nobody seems to object, I am going to change the title to "bird-term-nurse merger."LakeKayak (talk) 17:20, 4 January 2017 (UTC)

Done.LakeKayak (talk) 17:22, 4 January 2017 (UTC)

IPA vowels[edit]

I appreciate that IPA phonemes are only intended to be approximate, and that there will be considerable regional variation on their interpretation, but the Oxford English Dictionary is now using /əː/ for the vowel in the Bird–term–nurse merger. I realise that /ɜː/ has been traditionally used, but the sound of this vowel in the sound file of our IPA chart would mislead anyone wishing to learn British English. Is it the sound recordings that are wrong, or does Wikipedia need to catch up with modern usage? Dbfirs 09:31, 5 January 2017 (UTC)

Probably we should give both symbols, with an explanation, until such time as all leading dictionaries switch over to the OED's new (and IMO preferable) convention. I believe the same applies to /a/ in place of /æ/, and possibly some others. W. P. Uzer (talk) 14:27, 5 January 2017 (UTC)
By the way, what is the situation with American English? Is there any reason there, other than tradition, to use /ɜ/ or /ɜː/ for the NURSE vowel rather than /ə/ or /əː/ or /ʌ/? W. P. Uzer (talk) 14:32, 5 January 2017 (UTC)
@W. P. Uzer: Actually, there is. The vowel in "nurse" is still pronounced with the historical [ɜ] in many varieties of American English. Therefore, it is still the current sound used.LakeKayak (talk) 16:22, 5 January 2017 (UTC)
So is it that the vowels of NURSE and STRUT are more open than the unstressed schwas of buttER and sofA? How do American scholars tend to phonemize these various sounds? W. P. Uzer (talk) 10:31, 6 January 2017 (UTC)
@W. P. Uzer: The vowel in STRUT is represented as any one of these three notations in American English: [ʌ], [ʌ̈], or [ɐ]. But all three of them would be pronounced more open than [ə]. In my own dialect, /ʌ/ still retains its traditional value, so I myself can't speak for all dialects. I can only go by what the page General American says.
As for the vowel in NURSE, unfortunately, there is little difference made between /ɝ/ and /ɚ/. However, listening carefully to my own speech, I do notice that I pronounce the "u" in "nurse" more open than differently than the "e" in "butter."
Unfortunately, American scholars like Labov don't always use the IPA. In particular, Labov uses a separate system because his study is the evolution of vowel classes. Anyway, it can be hard to look at how American scholars transcribe the vowels as a way to determine how the vowel is being pronounced.LakeKayak (talk) 17:17, 6 January 2017 (UTC)
Thanks for the information. I was also wondering whether American linguists regard some or all of these sounds as being cases of the same phoneme (if they talk about phonemes, that is). In descriptions of RP we always seem to have /ɜː/ (or /əː/) distinct from /ə/ - I assume because of contrasts like 'convert vs. 'concert - and of course /ʌ/ distinct because we don't have the hurry-furry merger (and I think I remember Wells gives unending vs. an ending for /ʌ/ vs. /ə/). W. P. Uzer (talk) 09:44, 7 January 2017 (UTC)
Also, for British English, I feel that we could possibly identify /ɜ/ as older and /ə/ as "younger" (due to lack of better term). That would provide the most accurate information without misleading people.LakeKayak (talk) 16:25, 5 January 2017 (UTC)
That's not accurate.
Note that when you transcribe the NURSE vowel with the symbol ⟨ə⟩, the length mark must be preserved (⟨əː⟩), as this vowel is always distinct from the COMMA vowel /ə/.
I see some confusion between phonemes (written between slashes) and allophones (written between brackets) in this discussion. The NURSE vowel can be transcribed either /ɜː/ or /əː/. You can't talk about 'the difference between /ɜː/ and /əː/', as these are different symbols that represent the same vowel phoneme, and they say nothing about its actual phonetic realization. See phoneme and allophone.
In General British, this vowel varies phonetically between close-mid [ɘː], mid [əː] and open-mid [ɜː], just as the schwa does. In older GB it could be as open as [ɐː], but such a pronunciation is nowadays considered almost comical. It does make a lot of sense to consider this vowel a mere lengthened version of the schwa. See Cruttenden, Alan, ed., Gimson's Pronunciation of English, 8th ed., pp. 135–136.
The reason that the NURSE vowel is transcribed with ⟨ɜː⟩ is not because it's its traditional phonetic value, but because over 20 years ago ⟨ɜː⟩ was suitable for any central vowel, not just specifically the open-mid one (see Bauer, Laurie, Linguistics Student's Handbook, p. 134). That's also the reason for which the revised phonemic transcription of Australian English uses the symbol /ɜː/, even though the Australian NURSE vowel is about close-mid. Mr KEBAB (talk) 14:44, 11 January 2017 (UTC)

Then, I made an error. Thank you, sir. I'll remember that for next time.LakeKayak (talk) 19:42, 11 January 2017 (UTC)

Cure–force merger[edit]

Currently, the section reads as follows:

In Modern English dialects, the reflexes of Early Modern English /uːr/ and /iur/ are highly susceptible to phonemic merger with other vowels. Words belonging to this class are most commonly spelled with oor, our, ure, or eur; examples include poor, tour, cure, Europe. Wells refers to this class as the cure words, after the keyword of the lexical set to which he assigns them.
In traditional Received Pronunciation and General American, cure words are pronounced with RP /ʊə/ (/ʊər/ before a vowel) and GenAm /ʊr/.[14] But these pronunciations are being replaced by other pronunciations in many English accents.
In southern English English it is now common to pronounce cure words with /ɔː/, so that moor is often pronounced /mɔː/, tour /tɔː/, poor /pɔː/.[15] The traditional form is much more common in the northern counties of England. A similar merger is encountered in many varieties of American English, where the pronunciations [oə] or [or]⁓[ɔr] (depending on whether the accent is rhotic or non-rhotic) prevail.[16][17]
In Australian and New Zealand English the centring diphthong /ʊə/ has practically disappeared, replaced in some words by /ʉː.ə/ (a sequence of two separate monophthongs) and in some by /oː/ (a long monophthong).[18] Which outcome occurs in a particular word is not always predictable, but for example pure, cure and tour come to rhyme with fewer, having /ʉː.ə/, while poor, moor and sure come to rhyme with for and paw, having /oː/.

Although the section is referred to as "cureforce merger," the section never even explicitly says that the two classes are merged. (In other section like the "Horse–hoarse" merger, it explicitly states that the two classes are merged.) Therefore, could somebody elaborate on the section? Thank you.LakeKayak (talk) 02:49, 13 January 2017 (UTC)

I have a problem with "The traditional form is much more common in the northern counties of England." Is it really much more common? AFAIK the speech of many younger northerners is getting closer and closer to the local southeastern standard ('Estuary English'), as far as e.g. th-fronting, l-vocalization and t-glottaling are concerned (but BATH still equals TRAP). I'd expect /tɔː, pɔː/ to be the predominant forms for such speakers, but I could be wrong. Some citations would be useful... Mr KEBAB (talk) 17:10, 14 January 2017 (UTC)
I certainly hear "Estuary English" here in the north, but this is mainly spoken by "offcomers" from the south. Some of the younger generation are losing their local accents, but the /ɔː/ pronunciation is certainly not common amongst youngsters where I live (and I'm not in the extreme north). (There are still quite a few of us older generation left in the north.) I agree that we could use a citation for the claim, but, until we find one, I'd be happy to remove "much". Dbfirs 18:46, 14 January 2017 (UTC)
The claim was added by a speaker of Yorkshire dialect, Epa101, on February 11th, 2008‎, and he might be able to help with a reference. Dbfirs 19:00, 14 January 2017 (UTC)
Ok, thanks for the clarification. By saying 'young' I mean 30-40 or younger, not just 20. Mr KEBAB (talk) 21:14, 14 January 2017 (UTC)

I have a different question. However, I have no objection to the change that User:Mr KEBAB requested. For my question, in accents with the cure–force merger, do "cure" and "force" even rhyme? I have no idea.

I myself seem to have this partial merger. On one hand, I pronounce "poor" and "pour" as perfect homophones, both as [pɔːɹ]. On the other, I pronounce "cure" and "force" as [kjʊɹ] and [fɔːɹ], respectively. Also, I don't perceive "shore" and "sure" as homophones. I pronounce the former as [ʃɔːɹ] and the latter as [ʃʊɹ]. Can anybody help me out?LakeKayak (talk) 20:29, 14 January 2017 (UTC)

I haven't requested a change yet, just asked for clarification. Mr KEBAB (talk) 21:14, 14 January 2017 (UTC)

Then, I made an error. I apologize.LakeKayak (talk) 22:19, 14 January 2017 (UTC)

Hello. First, it's quite a surprise to be contacted about an edit that I made almost nine years ago. When I made this edit, I was only just starting to learn about phonetics.
I don't have a straight forward reference for the claim. Looking through my copy of Urban Voices (1999), I notice that Newcastle, Hull and Sheffield all have FORCE and CURE clearly distinct. In the West Wirral, this is less the case but then it's debatable whether this dialect counts as "Northern English".
A Handbook of Varieties of English lists both the monophthongal and diphthongal varieties in the chapter on the North of England. (Looking through this book, I'm not sure if any dialect in England actually has NORTH, FORCE and CURE as three distinct phonemes as is the case in parts of Scotland and of North America.)
However, from what I've learnt in the last nine years, I know that some northern varieties have [at least in the past had] a FORCE-CURE merger but with a diphthong. For example, more and moor were homophones as /mʊə/ as you can see on page 234 in Joseph Wright's description of his Yorkshire dialect. Although there are few young people who would pronounce more as /mʊə/ now, I nonetheless feel that my edit was misguided in portraying Northern English as not having a FORCE-CURE merger, as it seems as if the distinction in vowels between the word "force" and the word "cure" was actually introduced to the North [or at least to WR Yorkshire] by dialects further south.
Overall, I have no objection to deleting the sentence. Epa101 (talk) 23:32, 14 January 2017 (UTC)
P.S. as an aside, TH-fronting and L-vocalisation have been in parts of the North for a long time. In the 1892 Wright book, see page 91 for TH-fronting and page 40 for L-vocalisation.
Thank you for that interesting link with the merger going the other way (on page 235). Slightly further north, the old dialect where I live had (and still has) /mʊə/ for moor but had /mɪə/ for more (now rare except amongst those who deliberately preserve the dialect). I still think that the merger is less common in the north, but I'm happy to remove "much". These wide generalisations always have exceptions. Dbfirs 00:30, 15 January 2017 (UTC)
No problem. Nice to talk about it. /mɪə/ was the form for north Yorkshire as well. I'm not particularly knowledgeable on etymology. Wright suggests that more historically had the same vowel as in the GOAT set (e.g. bone, boat, both) and I think that I'm right in saying that these words historically had /ɪə/ in Cumbria/north Yorkshire but /ʊə/ in south Yorkshire. Moor has a different root. Epa101 (talk) 01:08, 15 January 2017 (UTC)
/mɪə/?! Is it a possible Dutch/German influence (meer and mehr, respectively)? Mr KEBAB (talk) 01:19, 15 January 2017 (UTC)
It could also be from the Scottish mair. Epa101 (talk) 01:42, 15 January 2017 (UTC)
Yes, the so-called Scots language is descended from old Northumbrian, just as my local dialect is, so they share a common origin. The word goes back thousands of years with various forms: "α. OE mara, OE mare (feminine and neuter), eME mære, eME maren, ME maure, ME mayr (north.), ME mere (transmission error), ME morow (transmission error), ME–15 mar (chiefly north.), ME–15 mare (chiefly north.); Eng. regional (north.) 16 18 mare, 18– maar, 18– mair, 18– meear, 19– mear; Sc. pre-17 maer, pre-17 maire, pre-17 mar, pre-17 mayir, pre-17 mayr, pre-17 mayre, pre-17 meair, pre-17 mer, pre-17 17 mare, pre-17 17–18 meir, pre-17 17– mair, 18 mear (Shetland), 18 mehr, 18 mere, 19– maer (Shetland); also Irish English 18– mair, 18– mare.
β. ME moar, ME mour, ME moyr, ME moyre, ME–15 moare, ME–15 moore, ME–15 (17 Eng. regional (Lancs.)) moor, ME–16 mor, ME– more, 17 muore (Eng. regional); Sc. pre-17 moir, pre-17 moire, pre-17 moor, pre-17 mor, pre-17 mour, pre-17 moyr, pre-17 17– more."

per the OED which also gives the etymology as "inherited from Germanic: Cognate with Old Frisian māra , Middle Dutch mēre (Dutch has the double comparative meerder ), Old Saxon mēro (Middle Low German mēr , also mēre , mē ), Old High German mēra , mēro (Middle High German mēr , mē , German (arch.) mehr- in mehres , neuter, mehre , plural; compare the double comparative forms Old High German mērōro , mēriro , Middle High German mērer , mērre , early modern German mehrer greater, more, German mehrere (plural) several), Old Icelandic meiri , Norn (Shetland) mire , mere (also adverb mire ), Swedish mera , Danish mere (the Swedish and Danish forms are the neuter adjective used adverbially), Gothic maiza < an adjectival base parallel to the adverbial base of mo adv.1 Compare the following adverbial forms in West Germanic languages, either influenced by or secondary developments from the adjectival forms: Old Frisian mār, mēr, Middle Dutch meer (Dutch meer), Old Saxon mēr (Middle Low German mēr, mēre), Old High German mēr (Middle High German mēre, mēr, German mehr)."
If I said "es t'gitten any /mɪə/" to any of the farmers in the area who have been here for many generations, I would be understood perfectly. Dbfirs 08:31, 15 January 2017 (UTC)

I am a little curious. Does this merger occur in American English?LakeKayak (talk) 01:56, 15 January 2017 (UTC)

It doesn't seem to be a part of most American accents but it is part of African American Vernacular English. Epa101 (talk) 22:15, 30 January 2017 (UTC)

Thank you.LakeKayak (talk) 23:02, 30 January 2017 (UTC)

Pure-poor split?[edit]

This question is a little off topic, but it is still related. At times, when looking online, along with the poor–pour merger, I see the pure–poor split. Can anybody explain exactly what the split is? (I do realize they were at one time, instated on this page. However, looking in the history, the summary didn't help very much.) Thank you.LakeKayak (talk) 02:49, 18 March 2017 (UTC)

I didn't know they were ever homophones (though I know nothing about the history of pronunciations). Pure is /pjʊə/ and poor is /pʊə/ here in northern England, both two syllables (with a slightly different /ʊ/, but that might just be a local variation). I am aware that there are lots of regional variations, and that British BBC English seems to be changing, possibly under the influence of "Estuary English". Having looked at the history, this is not the split that was previously in the article, so it probably isn't helpful. Dbfirs 07:37, 18 March 2017 (UTC)
The yod is, AFAIK, ignored, so the name doesn't imply homophony (and remember that /jʊə/ and /juː/ are usually analyzed as consonant + vowel sequences). I'll try to explain it, but you must find a source yourself (try Australian English phonology#Bibliography). In Australia, the historical /ʊə/ was split into either /ʉː.ə/ (a disyllabic sequence) or /oː/ (a mid back monophthong). An example of a minimal pair is tour /ˈtʉː.ə/ and tore /ˈtoː/. Probably the same applies to New Zealand English. I'm not sure about pure, it could have either vowel, though it's probably /ˈpjʉː.ə/. You need to ask someone from Australia. Mr KEBAB (talk) 13:23, 18 March 2017 (UTC)
Actually I'll do it myself: @Pelagic: Mr KEBAB (talk) 15:22, 18 March 2017 (UTC)
Spot-on with /ˈpjʉː.ə/, @Mr KEBAB:. I don't have sources, but first-hand (Australian) experience is that pure keeps the y-sound (unlike suit), and is disyllabic like tour and fewer. Poor is like pour, pore, tore or port/ˈpoː/ (or /ˈpɔː/? I'll have to trust the experts that Australian or is higher than in Southern British).
The u-sound in pure is the same as in suit, kook, boot, boom choose (though the last two are longer); as opposed to sook, shook, cook, book, put. When I say pure, fewer or sewer slowly, I can feel my tongue drop straight down from the u-position to the schwa-position with no forward movement, so the centralised notation of ⟨ʉ⟩ rather than ⟨ʊ⟩ in [ʉˑ.ə] seems to fit well.
Pelagic (talk) 17:33, 19 March 2017 (UTC)
Thanks! The exact transcription of the mid back vowel doesn't matter, not least because we're using phonemic slashes, rather than phonetic brackets, which necessarily makes our transcriptions more abstract. You can use either symbol. See Australian English phonology#Examples of vowels to compare the more modern transcription with the traditional one (and, in my experience, Australian /oː/ is pretty much exactly like the southern English /ɔː/, it's only the symbol that differs). Mr KEBAB (talk) 18:12, 19 March 2017 (UTC)
Does this split also occur in American English? Most of the American English pages list /ʊər/ and /jʊər/ as separate phonemes, where the former is typically merged with /ɔːr/ and /ɔːr/.LakeKayak (talk) 18:18, 19 March 2017 (UTC)
I'm an American from Florida and I have a split. "poor" is [pɔːr] and "pure" is [pjɝ]. "tour" is [tuː.ɚ] for me, rhyming with neither "poor" nor "pure". Fish567 (talk) 23:59, 19 March 2017 (UTC)
@Fish567: I'm from New Jersey and I also have a split. I pronounce "poor" as [pɔːɹ], "pure" as [pjʊɹ], and "tour" as [tɔːɹ]. However, does anybody have a source that addresses the issue in American English? (This is more for my own sake than for the page's.)LakeKayak (talk) 01:12, 20 March 2017 (UTC)
Maybe I misspoke. Whether or not we have a source, it can be left out of the article. However, I just want to know that this is an actual term used by linguists. Some terms that I have heard, one of which being, "from-rum merger" aren't.LakeKayak (talk) 01:38, 1 May 2017 (UTC)
Merriam-Websters online dictionary mentions such a split. <<Many speakers do not have the dipththong

\ur\ and have merged it with either \ər\ (when it follows palatal consonants such as \sh\, \ch\, or \y\ in words like sure, mature, or obscure) or \•or\ (in other environments). Similarly, many speakers of r-dropping dialects have merged \ur\ with \ər\ and \or\ in the same respective environments.>> Fish567 (talk) 01:47, 25 June 2017 (UTC) ──────────────────────────────────────────────────────────────────────────────────────────────────── This will do. Thank you, Fish567.LakeKayak (talk) 17:35, 26 June 2017 (UTC)

Horse-hoarse merger[edit]

Does anybody else feel this way? On the page Northern American English, it says that the north was one of the last regions to be horse–hoarse merged. While this page says (or at least used to) that some Western and Southern dialects remain unmerged. This leaves the question of who initially had the merger.

I was able to find this answer in the ANAE, but before I insert anything, I want to a second opinion. Thank you.LakeKayak (talk) 19:57, 18 May 2017 (UTC)

Pure-pier merger?[edit]

I live in Minneapolis, Minnesota, US; however, my accent also has some elements of northern Minnesota's Iron Range dialect. I have a tendency to pronounce words that have a consonant followed by [ju] which is then followed by [ɹ] as though they had an [i] sound. An example is "pure" being pronounced as [pjiːɹ], so it sounds very similar to "pier." At first I thought this might just be an oddity in my speech pattern, but my sister has the same change. I was wondering if this is at all common anywhere, or if it's on the page and I just couldn't find it. — Preceding unsigned comment added by Franxz (talkcontribs) 14:24, 20 June 2017 (UTC)

I have the same feature. It's not mentioned on the page as such, but it must originate from the curenurse merger: /kjuɹ/ > /kjəɹ/ > /kiɹ/. (For folks who don't understand the IPA: kyoor > kyurr > keer.) — Eru·tuon 22:31, 20 June 2017 (UTC)

External links modified[edit]

Hello fellow Wikipedians,

I have just modified 2 external links on English-language vowel changes before historic /r/. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

You may set the |checked=, on this template, to true or failed to let other editors know you reviewed the change. If you find any errors, please use the tools below to fix them or call an editor by setting |needhelp= to your help request.

  • If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
  • If you found an error with any archives or the URLs themselves, you can fix them with this tool.

If you are unable to use these tools, you may set |needhelp=<your help request> on this template to request help from an experienced user. Please include details about your problem, to help other editors.

Cheers.—InternetArchiveBot (Report bug) 06:38, 21 September 2017 (UTC)

Boston accent and Mary–marry–merry merger[edit]

Flower-flour merger?[edit]

Even if this kind of merger happened, the eponymous example seems flawed, because flour is etymologically identical with flower, so it is highly unlikely that the two words were ever pronounced different. The difference in spelling has only been introduced to keep their different meanings separate, not to indicate any kind of different pronunciation. --Ubel (talk) 22:29, 26 January 2018 (UTC)

Yes, I agree that this particular case is more like a recent split than a merger. The example words are a convenient shorthand for the supposed /aʊr/–/aʊər/ merger which I don't recognise at all. For example, I think hour was originally pronounced /aʊrə/ and the position of the /r/ and /ə/ were reversed to become /aʊər/, and it is only very recent generations of people who have started saying /aʊr/ as one syllable, perhaps influenced by the spelling. The word sour was spelled "sower" as early as 1561, and this suggests that it was pronounced as two syllables, so the merger, if it happened, was probably in late Middle English. There can be no modern accents that lack the merger, but some modern accents have recently split the pronunciations. However, I'm not an expert in phonetics, so I'll leave it to someone else to judge whether my claim is valid. Dbfirs 22:19, 27 January 2018 (UTC)