Talk:English-language vowel changes before historic r

From Wikipedia, the free encyclopedia

Flag
Portal
 English-language vowel changes before historic r is within the scope of WikiProject Australia, which aims to improve Wikipedia's coverage of Australia and Australia-related topics. If you would like to participate, visit the project page.
??? This article has not yet received a rating on the quality scale.
??? This article has not yet received a rating on the importance scale.
Assessment comments

Contents

[edit] Americentricity

Article might be better titled "American-language vowel changes before historic r" since it is so American centric.

Just a suggestion.

The thing is that it's North American dialects which are most affected by these vowel changes. Such changes in other dialects do occur and are covered here. --Jimp 01:29, 11 October 2006 (UTC)

[edit] Florida oranges

Of course it's a tense-lax neutralization. It's exactly the same process as all the others: "short o" before r merges with "or", in exactly the same way as "short e" before r merges with "air", "short i" before r merges with "ear", and so forth. The only reason it's not "in effect" a tense-lax neutralization synchronically is that, subsequent to the neutralization, "short o" underwent a merger with "ah". The result of this merger was that there now exist tokens of "short o" (original "ah") before r contrasting with "or", as in starry~story, and so the synchronic prediction of the merger is violated—but that's the result of a secondary development, the father-bother merger. Tense-lax neutralization is still how it must have come about. Sorrow, borrow, and tomorrow are actual exceptions, since they do retain original "short o" before r, but that just makes it a conditioned tense-lax neutralization (applying everywhere except before "long o"); it doesn't prevent it from being a tense-lax neutralization at all. (Alternatively, that itself could also be the result of a secondary development after TLN: "or" descends to "short o"+r before "long o".) AJD 06:54, 27 May 2005 (UTC)

Originally (in about the 17th century) there was /ɒrV/ (V is any vowel) in forest, borrow, etc., /ɑːrV/ in jarring, starry, etc., /ɔːrV/ in Laura, Taurus, etc., and /oərV/ in glory, story, etc. These remained distinct in RP until the 20th century, when the horse-hoarse merger applied to /oərV/ making it /ɔːrV/, so today in RP we have /fɒrɪst/, /dʒɑːrɪŋ/, /lɔːrə/, /glɔːri/. In America, there was a loss of phonemic length and a loss of /ɒ/, which mostly merged with (now no longer phonemically long) /ɑ/ (the father-bother merger), except before /r/, where the picture is more complicated:
  • In some accents, original /ɒrC/ (C is any consonant or a word boundary) became /ɔrC/ (hardly a merger since centaur was probably the only word with /ɔːrC/ before the loss of phonemic length), so /nɒrθ/ > /nɔrθ/. Later most of these accents underwent the horse-hoarse merger.
  • In other accents, /ɒrC/ merged with /ɑrC/ (the card-cord merger)
  • In some accents (Canadian), original /ɒrV/ merged with /ɔrV/, so you have /fɔrɪst/, /bɔro/, /lɔrə/, etc.; later the horse-hoarse merger changed /glori/ to /glɔri/, adding more words to this group.
Actually I think it's more accurate to say the horse-hoarse merger changed /ɔr/ to /or/, but I'll stick to the conventional transcription since this isn't the place to air my personal theories.
  • In other accents (Northeastern), original /ɒrV/ merged with /ɑrV/ so you have /fɑrɪst/, /bɑro/, /dʒɑrɪŋ/, etc.
  • Still other accents (GenAm) started out like the Canadian pattern but then changed /ɔro/ to /ɑro/ by dissimilation; sorry went along in order to keep the same vowel as sorrow.
The Canadian pattern is a tense-lax neutralization, yes, but the later GenAm dissimilation isn't, and what the "Florida oranges" section is discussing is the later dissimilation, isn't it? --Angr/tɔk tə mi 07:42, 27 May 2005 (UTC)
The GenAm dissimilation isn't a TLN, but as you say, it started out like the Canadian pattern, which is a TLN. Which means that the present-day GenAm orange class has its origins in TLN just as much as the mirror, hurry, marry, and merry classes do. That's all I was trying to say in my last edit to the article. I think the main focus of the section should be the original change, not the later assimilation: it's tense-lax neutralization that's the focus of the article here, after all. I'll attempt to make it clearer when I've got a few minutes.
By the way: it's not quite fair to say "In other accents (Northeastern), original /ɒrV/ merged with /ɑrV/ so you have /fɑrɪst/, /bɑro/, /dʒɑrɪŋ/, etc." This is just exactly the father-bother merger, and therefore Northeastern accents (Boston) not subject to the father-bother merger didn't undergo it. AJD 15:16, 27 May 2005 (UTC)
I'm not sure it is necessarily the father-bother merger, because in NYC, for example, father and bother don't necessarily rhyme. I think the most traditional kind of NYC accent has [fɑəðə] vs. [bɑðə], doesn't it? Certainly non-rhotic Clark [klɑək] is different from clock [klɑk]. --Angr/tɔk tə mi 22:49, 27 May 2005 (UTC)
Are you sure that, in NYC dialects where father isn't merged with bother, sorry is still merged with sari? At any rate, this is aside from the main point, which is whether the GenAm orange class has its origins in TLN; since you do seem to agree with me on that, I'm going to change the article back (and try to restructure it to be clearer on that point). AJD 23:36, 27 May 2005 (UTC)
I'm not sure about the New York accent, but the Boston accent quite definitely does not have the father-bother merger (see Boston accent). The "Florida-oranges" lexical set, in Boston, uses the "bother" vowel: /ɒ:rV/. Given the Boston accent's rather complex phonemic relationships to other dialects (/ɒ/ being used in "cot", "caught", and "horse", but not in "hoarse"), I think that if it's going to be included in the chart, it should get its own box rather than being listed alongside either RP or NYC/Philadelphia. I'm going to be bold and edit the chart and corresponding text. --The Lazar 21:55, 28 January 2006 (UTC)

For me, the sound in "Florida oranges" is [or], not [ɔr], and that's the common pronunciation where I live (southern California). I think this should be noted in the text, although I'm having trouble figuring out exactly where or how to put it. -Branddobbe 23:40, September 3, 2005 (UTC)

[edit] misnamed node

This node is misnamed, since it refers to tense-lax neutralization (TLN) in general, but in fact applies only [a] to English, [b] to neutralization before /r/. Similar TLN occurs before 'ng' in English, and to some extent before 'sh'. You could also argue that, e.g. the caught-cot merger is a tense-lax neutralization.

Furthermore, there are various other examples of neutralization before /r/ on the "Phonemic differentiation" page or linked from it: horse-hoarse, card-cord, pour-poor, fern-fir-fur, near-square, nurse-square, tower-tire, etc. Whether these are "tense-lax neutralizations" or some other kind of neutralization before r seems a rather silly and artificial distinction to make. I'd strongly suggest renaming this node to "English Neutralizations before r" and combining the relevant information from the Phonemic differentiation page into this. It might also make sense to remove the separate pages on hoarse-horse and card-cord (with redirects to this page) and combine them, since they're quite short and there is little sense in having some of these mergers in separate entries and some together; nor is there much sense in splitting every merger into its own entry.

When this is done, the page called "tense-lax neutralization" should actually reflect its title, and cover this in a properly general fashion -- e.g. discuss the connection between tense-lax and long-short, note the tendency in various languages to convert length to a tense-lax distinction (e.g. German, Proto-Romance, Polish), with the often further result of tense-lax mergers (e.g. Proto-Romance, Polish); the tendency for neutralization in closed syllables; etc. Also, why does English so consistently have neutralization before r? This should be discussed.

If there is consensus, i'll consider making some of these changes myself if no one else would rather do it.

Benwing 1 July 2005 18:09 (UTC)

In principle, you're right. I think the reason these changes are called "Tense-lax neutralizations" is because that's what they're called at http://students.csci.unt.edu/~kun/tln.html --Angr/tɔk tə mi 1 July 2005 20:57 (UTC)
I see the point about the article having a general title but only referring to English, although googling the phrase "tense-lax neutralization" (and ignoring Wikipedia mirrors) does find a few other examples referring to this phenomenon. (The variant spelling with "neutralisation", which of course I tried first, gives no hits at all.) I do think there's some merit in maintaining this group of mergers as a distinct group; they have a similar regional pattern (they all occur across most of North America, excluding parts of the north-eastern US, but do not appear in the British Isles or the southern hemisphere countries, as far as I'm aware) and all involve mergers before intervocalic /r/ only.--JHJ 4 July 2005 17:09 (UTC)
I have to disagree with both you and Angr's in-principle agreement. Not all the mergers you mention occur before /r/ for instance, IIUC the nurse-square and tower-tire mergers usually happen in non-rhotic dialects, and I think that pour-poor mergers are also predominately seen in non-rhotic ones, with rhotic ones either preferring to keep them distinct or merging 'poor' with 'fur'; many of the mergers have definitely occurred since the creation of non-rhotic dialects as well, even if they've happened in both (horse-hoarse). So there's two different kinds of mergers: Those before /r/, and the decrease of centralising diphthongs and triphthongs (AuE maintains an independent SQUARE phoneme, but it is (usually) pronounced [e:] ... this is a member of the same family, from AuE's/AuE linguists' perspective, as our poor-paw merger). Also, there are TLNs before /l/ (I understand some western US speakers merge sale and sell). Felix the Cassowary 2 July 2005 10:42 (UTC)
The diphthongization before /r/ was probably the proximate cause of the mergers; it wasn't so much a merger of /niːrər/ and /mɪrər/ as of /nɪərər/ and /mɪrər/. So it might be more accurate to call it a merger of centering diphthongs with lax vowels, rather than a merger of tense vowels with lax vowels before /r/. For mergers before /l/, see the fill-feel and fell-fail mergers at Phonemic differentiation as well as the second of the Foot-goose and full-fool mergers, and of course centering diphthongs are common before /l/ too (hence the homophony of real and reel even for people who don't merge fill and feel). --Angr/tɔk tə mi 2 July 2005 11:28 (UTC)
I disagree. Especially, I don't think this should be merged with articles on other mergers before /r/. TLN as defined in this article is a well-defined and unitary phenomenon. It makes no more sense to merge, say, hoarse-horse merger into this article than it would make sense to merge æ-tensing into Northern Cities Vowel Shift: they have properties and causes in common and overlap in some of their effects, but they are distinct phenomena for all that.
As for the title, it's just that "tense-lax neutralization" is the only name I know that clearly refers to the entire family of mergers: the only concise alternative names that come to mind are things like "marry-merry-Mary merger", which only refers to one aspect of TLN. Can you suggest a better title? AJD 6 July 2005 00:35 (UTC)

OK, perhaps "before r" isn't quite right; maybe "before centering vocalics" is more correct. But definitely [a] this page is badly misnamed; [b] there is data elsewhere that needs to be merged here; [c] these are all basically incidents of the same phenomenon. English velar /r/ and /l/ are phonetically quite similar to a schwa, which is why you see similar changes in both rhotic and non-rhotic dialects (and explains the origin of the non-rhotic change; note the corresponding change of /l/ to /w/ in Cockney etc., and similar changes in German, French, Brazilian Portuguese, etc.).

BTW can someone help clear up some facts about "Canadian raising"? This is claimed to be a largely northern phenomenon in the U.S., but in various informal surveys I've done, every person I've encountered, from all over the U.S., distinguishes 'writer' from 'rider'. [me too, and I grew up in Arizona from age 4 -- likewise, 'spider' and 'spied [h]er' are not the same for me; the former has the diphthong of 'writer'.] Benwing 3 July 2005 06:04 (UTC)

Do you mean before central vowels? There are central vowels and there are centering diphthongs, but I'm not familiar with the term "centering vocalics". It's true coda liquids (/r/ and /l/) tend cross-linguistically to become vocoids (vowels or glides). As for "writer"/"rider", is it the vowel quality that's different for you (as it is for Canadians), or just the vowel quantity. I grew up in Texas from age 9, and for me writer and rider are identical, but I can convince myself the diphthong in spied 'er is longer than the diphthong in spider even though the qualities are identical. I have the same distinction between a longer diphthong in shyness and a shorter diphthong in minus. It all has to do with morpheme boundaries, not a phonemic contrast between long and short diphthongs, though. --Angr/tɔk tə mi 3 July 2005 07:14 (UTC)

For me there is a quality difference between the vowels in writer/rider; probably a small quantity difference as well, although in fast speech the quantity difference may disappear, while the quality remains. the quality difference is that the first vowel in the diphthong in "writer" tends towards schwa while in "rider" the diphthong starts with the vowel of "cot". the exact same difference occurs between spider/spied 'er and in your example minus/shyness. Note that for me, "highness" can have both pronunciations, but with different meanings: in "your highness", it rhymes with "minus"; in "highness" = "height", it rhymes with "shyness". the "raised" diphthong of "writer" also occurs in "bicycle" and "nitrogen", and *optionally* in "hydrogen". it also seems to me that it is optional in "pilot". i don't see how this is anything but a phonemic split, unless you allow for really abstract theories of phonology -- at which point you might as well assert that the sound "engma" is "not really" a phoneme, since you can dispense with it at some level of abstraction.

and you're right my terminology was muddled. /r/ and /l/ in English, esp. at the end of a syllable, are highly vocalic in their phonetic nature. they are also pronounced centrally in the mouth, due (among other things) to their strong velarization in this position -- although in AmEng /r/ and /l/ have a fair amount of velarization everywhere. Maybe "centering off-glides" is a better term. the basic point is that regardless of terminology there is one essential mechanism causing all of these vowel reductions before these particular phonemes, which does not (in general) lead to reductions elsewhere. (the reductions before /n/ are due to a separate phenomenon, i.e. nasalization, which often causes raising -- although lowering in French, God only knows why).

Benwing 8 July 2005 09:36 (UTC)

All this sounds eminently plausible to me: centering diphthongs arise before velarized coda liquids, and then can merge with the nearest lax vowel. But can we find a published source that analyzes these mergers this way? We can't add original research here, after all. --Angr/tɔk tə mi 8 July 2005 09:48 (UTC)

Well, it doesn't seem like we need any authority to simply *group* things in a particular way, e.g. grouping changes before schwa/r/l based on simple observation that they tend to occur a lot, and in similar ways. After all, every day Wikipedians make decisions on grouping, and it is not considered original research or needs a published source. We can put in a theory explaining why these changes happen if/when we find such a published theory.

Benwing 8 July 2005 09:58 (UTC)

[edit] Furry-ferry merger

Here's what L. Craig Schoonmaker http://www.geocities.com/sswordday/ has to say about the ferry-furry distinction:

  • Quote-THIS is the famous "distinction without a difference", except that there are about 4 times as many -erry's as -urry's. And please note that Dictionary gives woor.ee, foor.ee, and hoor.ee (that's the sound that the U with a 'hat' (circumflex accent) shows: short-OO), which I have not heard so regard as bizarre. Either they heard wrong or they're on drugs.

+

  • Dictionary, oddly, is sometimes just plain wrong. For instance, "water" is not shown there as ever being pronounced "wut.er", but I listened very carefully to reports of water-main breaks on TV stations in the New York Tristate Metropolitan Area (the broadcasting capital of North America), and wut.er is plainly the pronunciation educated people in this area give that word. The SSWD project, of course, cannot offer "water" precisely because it has more than one common pronunciation.

+

  • If you put together the -erry's and the -ery's pronounced the same, you get a MASS of words with ER as the crucial spelling, but if you try to use -ury rather than -urry, you get a completely different sound. So I think we'll go with -erry. But I appreciate your views. Cheers.
  • Quote-UR, ER, OR, and AR may be pronounced with tiny differences by SOME speakers in SOME dialects as to SOME words. I went to your URL for the Cambridge dictionary, which offers TWO bizarre transliterations (which may or may not be rendered in standard IPA but is opaque to me -- IPA transliterations tend to proceed from the positions of vocal apparatus of the linguists who speak them in preparing to write them; SSWD is concerned about what people HEAR, and if they hear no difference between, for instance, vaann and venn for French "vin", it doesn't matter to them whether the person saying it forms the word one way, because the listener hears it the same no matter which way a speaker might articulate it). Most to the point, the Cambridge dictionary shows TWO pronunciations, British dialect and American standard.

+

  • I then went to the Merriam-Webster URLs for the other words and clicked on the speaker icon to listen to the pronunciations rendered, in American English, and found no distinction worth making. All those words would rhyme PERFECTLY as most people regard things. Of course, we could avoid the problem altogether by saying that there are two different pronunciations for "worry", so the word can't be changed!

+

  • For most ordinary , for whom the SSWD project is intended, not for linguistics specialists, there is between a great many word pairs or groups, no difference worth 'worrying' about. There are a lot of overeducated people who have bugaboos about tiny matters of no consequence, and will argue them endlessly, to everyone else's tedium. I'm not about to argue the linguistic equivalent of how many angels can dance on the head of a pin, a subject that may have fascinated some medieval theologians but nobody else.

+

  • The SSWD project is about NEEDED change, and preferably changes that people can readily apply to things they HEAR. One transliteration for a small range of actual sounds is convenient, and all spelling is convention. Few speakers of standard English distinguish in sound between "ferry" and "furry". Having a distinction in spelling for these two HOMONYMS is useful. As to which spelling you favor for a reform of "worry", I have noted that you favor "wurry".

You are talking nonsense. They don't sound even similar to most British speakers. 89.243.97.14 (talk) 00:07, 15 February 2008 (UTC)

+

  • The problem may be only that a following-R tends to alter the quality of the vowel before it, for some speakers more than others. I have not yet offered this word (which you plainly render "wurd" and I render "werd") and might select "wurry", on the basis that some people might see it as parallel to "merry", which they pronounce like "Mary". Or I may not offer it at all, since, as some people regard things, it has two pronunciations so cannot be changed if a change would antagonize some significant body of speakers. I am asking for more comments. Cheers.
  • Quote- YES, I noted that in checking "merge", some dictionaries use the U with a hat as the vowel. But in any case, that is the ER sound, as shown plainly by the sample words in Dictionary.com's own pronunciation key: "urge, term, firm, word, heard".

+

  • As for "ont", I suggested that because "ant" is a homophone we can eliminate from a language filled to overflowing with homophones, and seems to those of us who say "ont" -- meaning a large proportion of the best-educated people in the U.S. and almost everybody in Britain, Australia, New Zealand, the Caribbean, etc. -- that calling a person by a homophone for an insect is arguably disrespectful. I have no power to impose anything, and the SSWD site is designed mainly to make people think. As for "tord", too-waurd is a spelling pronunciation, and as with ev-er-y and other spelling pronunciations (which my Random House Unabridged labels so people know better than to use them), spelling reformers can properly advise people that tho they think they are being careful to be correct, they are actually being wrong.

+

  • The distinction between "ferry" and "furry" is, I repeat, not "worth making. All those words would rhyme PERFECTLY as most people regard things." People who try to draw needless distinctions and force people to try to supply only one of essentially interchangeable spellings do spelling reform a disservice. This is not the distinction between "merry" rhyming with "berry" and "merry" rhyming with "Mary". It is TRIVIA that ordinary people do not waste time on and cannot justify wasting educational time and money on. If you see a poem in which one line ends with "ferry" and the next appropriate line ends in "furry" or "worry" or "cherry" or "very", will you be startled by an appalling lack of rhyme? If so, you are one in perhaps 15,000 people.

+

  • Native speakers of English cannot and do not make the short-E as in "bed" and follow it with R in the same syllable and come out with anything like what most people say for "very", "berry", etc. Following-R changes the quality of many vowels in its same syllable.

+

  • Make all the silly and PRETENTIOUS distinctions you want. Ordinary people concerned with communication rather than language as an arcane study to itself will not trouble to heed you. 205.188.117.14 04:30, 3 August 2005 (UTC)
This page is about what is, not what should be. Benwing 07:34, 4 August 2005 (UTC)
What does it matter what L. Craig Schoonmaker has to say? The man is as much a linguistics expert as is a jellyfish. Jimp 05:58, 11 January 2006 (UTC)

[edit] planning on moving/merging this page

i am planning on merging this page with the info in Phonemic differentiation that pertains to mergers before historic /r/. tentative name is "English-language pre-rhotic mergers". yes, i am aware that some of these mergers have occurred in non-rhotic dialects, but that does not change the basic facts of the situation. suggestions for a different name are welcome. Benwing 07:26, 4 August 2005 (UTC)

I don't like "pre-rhotic": some of the mergers I think you're talking about (near/square, for example) occur mainly in non-rhotic dialects, and it just seems wrong to call them "pre-rhotic" mergers. "Mergers before historic /r/" might be better, but I'm not sure. Also, I'll repeat my opinion that the family of sound changes currently discussed on this page should be kept as a distinct group. They all involve short (or "lax", or whatever) vowels before intervocalic /r/, and they have very similar geographic distributions (most of English-speaking North America, but not New York or Boston, and not the British Isles or the southern hemisphere). --JHJ 11:34, 4 August 2005 (UTC)

'English-language pre-rhotic mergers' and 'English-language centralising diphthong simplifications' makes sense. I don't see how it doesn't change the basic facts of the situation. (I spelt 'centralising' with an S because that's how I spell and I didn't think about it till now, but of course 'centralizing' would make equal sense and I'm not interested in disputing the point.) Felix the Cassowary 12:30, 4 August 2005 (UTC)

JHJ -- I think that the attempt to group exactly what's on this page, and nothing else, is stretching it. Neither the marry/merry nor hurry/furry mergers are tense-lax neutralizations, since both vowels are lax. nor does the "florida oranges" section describe such a neutralization. hence, the entirety of TLN on this page is mirror-nearer and merry-mary, not a reasonable separation.
I'm not bothered about whether it's called TLN (though that does seem to be a term used for the phenomenon in question). I'm talking about the geographic distributions (to me, TLN as currently described on this page is one - or maybe four - of the most notable phonological differences between typical "British" accents and typical "American" accents) and the fact that they're all mergers before intervocalic /r/ only. You seem to want to group these mergers with things like the nurse/square merger of Scouse, which is really just a merger of two long vowels: fur /f3:/ and fair /fE:/ merge as /fE:/.--JHJ 08:55, 5 August 2005 (UTC)
I think "English-language mergers before historic /r/" might be best, as it describes most clearly the situation. Benwing 00:40, 5 August 2005 (UTC)

If you can put that slash at the end of the title, then I'd accept it (not like it, mind), otherwise I absolutely hate it on the grounds that I absolutely hate articles that have a title technical limitations header. I thought the hurry/furry was a TLN; for me, furry's in the same category as Mary (which is long, but corresponds in this case to US tense). Indeed, I thought the distinction between /@r/ and /3r/ in American was [-tense] and [+tense]. (But that still doesn't make the marry/merry merger fit here.) Felix the Cassowary 02:04, 5 August 2005 (UTC)

Hmm. Let's take out the slashes. "English-language mergers before historic r"? Do you like "... pre-rhotic mergers" better? Benwing 02:27, 5 August 2005 (UTC)
No, it's not pre-rhotic, so the 'historic' is important IMHO... Felix the Cassowary 02:41, 5 August 2005 (UTC)

[edit] Furry-ferry merger

Is this merger just a near merger and not an actually phonemic merger? Something that I've read suggested that. 64.194.44.220 02:34, 13 December 2005 (UTC)

[edit] Moved from Talk:Bear-beer-burr merger

This is a personal blog? http://caxton.stockton.edu/classwork/discuss/msgReader$44 Sorry for putting with the references if it is. Foogol 21:06, 18 December 2005 (UTC)

It looks like one to me. It's a teacher talking about something she's noticed in her students' speech and writing, isn't it? Anyway, even if it's not a personal blog, it's still not a useful source since she's just reporting on something she read about in an introductory linguistics textbook (Fromkin & Rodman, which is usually used in freshman Intro to Linguistics courses). If you get a copy of F&R out of the library and track down their source for the phenomenon in AAVE, that will be a useful source. --Angr (t·c) 21:10, 18 December 2005 (UTC)

[edit] Original research

This looks like original research. Although this phenomenon is mentioned at http://www.rehabmed.ualberta.ca/spa/phonology/features.htm (scroll down to "Vowel Centralization Before /r/ (VC-r)"), that says it is "reported in Memphis data". No source is provided discussing this as a phenomenon of rap or of the AAVE of St. Louis. --Angr (t·c) 21:10, 18 December 2005 (UTC)

It looks to me as if it's partly based on a Language Log posting: http://itre.cis.upenn.edu/~myl/languagelog/archives/000697.html --JHJ 12:02, 19 December 2005 (UTC)

Another source that refers to the phenomenon is at http://www.ausp.memphis.edu/phonology/#Vocalic. --Angr (t·c) 10:56, 29 December 2005 (UTC)

[edit] Articles for Deletion vote

Articles for deletion This article was nominated for deletion on 17 December 2005. The result of the discussion was NO CONSENSUS.


[edit] Previously under Talk:Horse-hoarse merger

I think that is actually useful information that you've removed, angr. So, I've reverted it and added quotation marks.

I never said it wasn't useful information. I said it was a copyvio (copyright violation), which it is. You cannot copy text wholesale from other websites and paste it here. Even inside quotations marks it was far too long to count as fair use. But we can add a link to the page it comes from so the interested reader can learn more at the original page, which is what I've now done. --Angr/tɔk tə mi 04:43, 31 May 2005 (UTC)

What about dialects like mine, where the merge has taken place but the pronunciation is with "o" instead of "ɔ"? -Branddobbe 00:04, September 4, 2005 (UTC)

I think in most accents with the merger, the resulting vowel is closer to [o] than to [ɔ], but for better or for worse linguists have agreed to transcribe the merged vowel as [ɔ]. In other words, transcribing merged ho(a)rse as /hɔrs/ rather than /hors/ is more a matter of consistency with tradition than of phonetic accuracy. --Angr/tɔk tə mi 05:45, 4 September 2005 (UTC)
Do you have sources for the agreed-upon transcription of the merged vovel? I also come from a region where horse/hoarse are homophones pronounced with a vowel much closer to [o] or [oə], and I know from experience that I differentiate the pronunciations [hɔɹs] and [hʌɹs]. Given my impression from undergraduate linguistics courses that linguists rarely agree upon anything, it strikes me that it might be more accurate to mention the various pronunciations of the merged aspect while noting that a/the predominant linguistic tradition transcribes the vowel as [ɔ]. JohannVII 10:36, 19 January 2007 (UTC)
I don't have a source where a bunch of linguists have gotten together and explicitly said "let's transcribe the merged horse/hoarse vowel as [ɔɹ]", but if you look at books and dictionaries that use an IPA transcription (Ladefoged, Longman's pronunciation dictionary, Kenyon & Knott, etc.) all or almost all of them do use [ɔɹ]. And for that matter, other dictionaries that use their own idiosyncratic transcription system will use their own equivalent to [ɔ] in these cases as well (e.g. Merriam-Webster transcribes the merged ho(a)rse vowel as "hôrs", not as "hōrs", where "ô" is the vowel of "bought" and "ō" is the vowel of "boat"). I'm confused though by your second sentence: you come from a region where they're homophones, but you personally distinguish them? Or do you mean you distinguish merged "ho(a)rse" from "hearse"? —Angr 11:12, 19 January 2007 (UTC)

[edit] Toe-tow merger

Has this (the horse-hoarse merger) anything to do with the Toe-tow merger? If so, let's merge the articles. Jimp 22Dec05

I don't think so. The toe-tow merger was earlier; in fact, the horse-hoarse merger is a merger of EME "short o" (/ɒ/ with the outcome of the toe-tow merger before /r/. --Angr (t·c) 06:24, 22 December 2005 (UTC)

In EME, you have something like

  • toe /to:/
  • tow /tOU/
  • hoarse /ho:rs/
  • horse /hOrs/

(where /O/ is the vowel of cot)

So toe and hoarse had the same vowel (and indeed they still do in some accents, like some Scottish ones). The H/H merger merges /O/ and /o:/ before /r/ (to /O:/ in non-rhotic accents), and the T/T merger is a complete merger of /o:/ and /OU/. I don't know whether any words had /OUr/ or, if they did, whether there are any accents that distinguish it from the reflex of /o:r/ (hoarse). Norfolk accents, I believe, have H/H merged but not T/T.

I'd suggest merging toe-tow merger with pane-pain merger under Wells's name "Long mid mergers". I think this article is long enough to stay separate, but if it's merged anywhere it should be to English-language vowel changes before historic r.

--JHJ 11:48, 22 December 2005 (UTC)

I agree with JHJ. Toe-tow merger can be usefully merged with pane-pain merger as Long mid mergers, and horse-hoarse merger is long enough (and sociolinguistically important enough IMHO) to remain an independent article. --Angr (t·c) 12:22, 22 December 2005 (UTC)
Yes, that all makes good sense. And merging Pane-pain merger with Toe-tow merger to Long mid mergers makes more sense than merging it with Rain-rein merger to Vane-vain-vein mergers did. Perhaps words like four, source, etc. may have had /OUr/ as opposed to the /o:r/ of such words as perhaps oar, roar, etc. just judging by the spelling. Wonder whether such a distinction would still be made (if it ever had been) by some. Jimp 16:59, 28 December 2005 (UTC)

[edit] Previously under Talk:Phonological history of English back vowels before historic r

Merging all this here was a good idea but I think I have a better one. Instead of Phonological history of English back vowels before historic r and English-language vowel changes before historic r here's what I suggest. Split the latter into English-language vowel changes before historic coda r and English-language vowel changes before intervocalic r (or some such titles). Then merge this article with the latter of these new articles. The problem is what to do with the Phonological history of English back vowels before historic r intro. Jimp 17:53, 8 January 2006 (UTC)

Why is Phonological history of English back vowels before historic r a separate article anyway? Why not just have one single Phonological history of English vowels before historic r article? --Angr (tɔk) 18:05, 8 January 2006 (UTC)
Indeed, it makes no sense to me either. This was my other idea: merge them back together. Let's do it. Jimp 10 January 2006
The following discussion is an archived debate of the proposal. Please do not modify it. Subsequent comments should be made in a new section on the talk page. No further edits should be made to this section.

The result of the debate was don't move. —Nightstallion (?) Seen this already? 07:50, 23 March 2006 (UTC)

[edit] Requested move

81.154.24.25 has requested that this page be moved to American English vowel changes before historic r. This section is for discussion of that suggestion.

The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made in a new section on this talk page. No further edits should be made to this section.

[edit] a major topic is missing here

Starting about the time that English incunabula were being printed (late 15th cent.) there begins to appear evidence that many words formerly spelled with er now began to be spelled with ar.

This is mentioned in Wells (Accents I:203), in Prins (History of English Phonemes, 1974:154-6) and especially in Wyld (History of Modern Colloquial English, 1956:212-222), where he connects the early appearances especially with "S.E. and E. Midland texts". The Wells book mentions some common examples such as star - far - heart which were not only previously spelled with -e- but also are easily related to German Stern - fern - Herz. In the first two cases, the spelling also changed, but in the case of heart it is not clear whether the new spelling reflects the new pronunciation which is clearly different from Middle English herte. Wells speaks of "quite a few instances" and Wyld lists several pages full of examples along with citations and sources.

Along with the many words where the spelling changed to the new ar and has remained so, there were many other words in which the spelling did not changed, but in which the pronunciation must have also been ar. This assumption rests of large amounts of private correspondence from the 15th to the 18th centuries, some of which Wyld refers to: when people wrote "I hard it was so" the assumption is that they were spelling to reflect their own pronunciation, a habit much more in vogue several centuries ago. The public printers, on the other hand, maintained --- for some reason -- the older er---or in some cases such as learn,an ear---in this second group of words. Wyld states that the new ar pronunciation was spreading more and more in the 16th century, with, for example, Queen Elisabeth using less of them when young but more in her writings from later years. Prins (p.156), gives a short list of examples (certain - desert - mercy - person - sermon - servant - verdict - vermin) and states that they were all spoken with ar in the 18th century. He singles out these words as the kind which had earlier acquired the new ar spelling, but had later, "under French influence", reverted to the old er spelling. The American Encyclopaedic Dictionary 1895, which I have just started examining, lists several "obsolete" variants with an ar spelling for words such as merchant and perfect.

To further complicate the story, with the rise, in the late 18th and early 19th centuries, of spelling pronunciation and the new idea that there was only one correct spelling for each word, there began a large-scale reversion to the er spellings that had often been preserved by printing houses; this in turn eventually led to almost all such <er> [ar] words acquiring the ur ( ɝ ) sound that we now hear in mercy - certain etc. Wells's "spelling pronunciation perhaps playing a rôle" seems like quite an understatement when one considers the large scope of this reversion. The only survivors I know of in American English are sergeant and heart, hearth, while RP still maintains the ar sound in words such as clerk - Berkeley - Derby. The proper noun Clark still has the old ar since its spelling had already changed. Another doublet is parson, which used to be the same as person, with the latter, more common word undergoing the reversion.

As a youngster in the US in the 50s, I occasionally hear such expressions as Marcy me! ( = mercy), What in tarnation... (old doublet related to eternal), "you consarned.. (&@#%$)" ( = concerned), That'll larn ya ( = learn ). More interestingly, the retention of the old ar in popular ("rustic") speech was also found in words which traditionally seemed to have had a long -e- vowel, and which today have been standardised with the er (not ɝ ) pronunciation: there's the old song about how Davy Crockett "killed him a bar when he was only three"" (=bear), and the even better known doublet thar for there, as in "in them thar hills" and the whaler's "Thar she blows!". When checking in the OED for early variants of some of the above words, I found some examples such as fern and herd where the earlier ar form was referred to as Scottish. A number of such ar variants are also cited in the Concise Scots Dictionary 1985; somehow I feel a connection between these Scots citations and the Scotch-Irish flavor of hillbillies,Davy Crockett, and sailors (Aye aye Captain!- Aye is northern English for "yes"). I wonder if these ar forms, although beginning in the South, ended up being preserved more in the nonstandard northern dialects?

The question of who said what and where is very difficult: was there another dialectal tradition where short er did not change to ar, and did this accent later become more dominant? From where do we get this mixture of long er and short er reflexes? In the same small section (4.19) Prins first says that short er merged with ir and ur, but then he says that "er frequently became ar". This is confusing --- I'm a specialist in Tibetan, not English phonology, but I'm interested in this problem due to its connection with parallel dialectal developments and the power of spelling pronunciations: can anyone offer suggestions? Also, I think at least some of this material belongs in the main article. Jakob37 11:54, 7 April 2007 (UTC)

[edit] Mary-marry-merry merger

One of the best-known pre-rhotic mergers is known as the Mary-marry-merry merger[4], which consists of the mergers before intervocalic /r/ of /æ/ and /ɛ/ with historical /eɪ/ [5]. This merger is quite widespread in the American West, Inland North, Midland, and in Canada (cf. sample 1). A merger of Mary and merry, while keeping marry distinct, is found in the South and as far north as Baltimore, Maryland, and Wilmington, Delaware; it is also found among Anglophones in Montreal[6]. In the Philadelphia accent the three-way contrast is preserved, but merry tends to be merged with Murray; likewise ferry can be a homophone of furry. See furry-ferry merger below. The three are kept distinct generally outside of North America, as well as in the Philadelphia accent, the New York-New Jersey accent, the Boston accent, and the accent of Providence, Rhode Island[7] (cf. sample 2).

– From the article

Which are /æ/, /ɛ/ and /eɪ/ respectly, in the dialects that the three-way contrast is preserved?

Yes. AJD 14:44, 16 August 2007 (UTC)

[edit] Southern American English

How is the "short o" before intervocalic r pronounced in the South? I think it is similar to NYC and Philly. 208.104.45.20 (talk) 00:29, 18 January 2008 (UTC)