Talk:Phonemic differentiation/Archive 2

Weak vowel merger

Quote-The weak vowel merger is a merger of schwa with unstressed /ɪ/ (sometimes written as /ɨ/). As a result of this merger the words affect and effect are homophonous, in accents without the merger they are distinct. The merger is complete in the Southern Hemisphere accents and variable in General American and Hiberno-English.

What about the merger of unstressed oo sound in book as in tonight and today with the schwa sound? What merger is that? I think it's worth mentioning.

Are there any accents where those are consistently pronounced [tʊˈnaɪt] and [tʊˈdeɪ] (as opposed to individual speakers who pronounce them that way as a form of hypercorrection)? If not, it's not really a merger. If so, is there published evidence of accents where the pronunciations with [ʊ] are used? If not, the information is unverifiable and can't be included. --Angr/_comhrá 04:27, 4 May 2005 (UTC)

I don't have exact data, but the existence of this opposition is strongly suggested by Charles Kreidler (in: Charles Kreidler: Describing Spoken English, An Introduction, London/New York: Routledge 1997, ISBN 0-415-15095-7, p. 110) suggesting that in the words "July prudential accurate masculine popular" (italics by Ch.K.) there may occur both the pronunciations /u/ and /ə/. On the previous page, he mentions that "the weak final vowel of value, issue, menu and the pre-vocalic vowel of casual, virtuous, evacuate is either /uu/ as in true or /ʊ/ as in foot, or it is neither of them" (all highlightings by Ch.K.).

He continues with the remark that in the samples "obey Korea accolade potato yellow" (italics by Ch.K.), some speakers have /o/ in all instances, while others have /ə/ in the first three and /ou/ in the last two.

So while these mergers seem to occur, I can't tell where they do.

Another remark: Am I right in assuming that the use of the symbol ɨ in this context is not meant to refer to the close central unrounded vowel [ɨ] but is just used in order to have a sign that is different from both [i] and [ɪ]? J. 'mach' wust • tskʃpræːx 10:15, 4 May 2005 (UTC)

I'm uncomfortable calling any of these phonemic mergers. The differences you're describing sound to me like they're differences in the realization of unstressed vowels, not in the conditioned merger of phonemes in some dialects as opposed to others. Just because some people say [ˈmæskjəlɪn] while others say [ˈmæskjʊlɪn] doesn't mean they have different phonemic representations of the word; it just means they have different rules (or constraint rankings if you do OT) concerning the realization of the phoneme /u/ in unstressed open syllables. --Angr/_comhrá 12:37, 4 May 2005 (UTC)

I have no problem with considering these phonemic mergers. It's not phonemic mergers in the phonological sub-system of stressed vowels, for sure, but phonemic mergers in the phonological sub-system of unstressed vowels, as in the minimal pair of /əˈfekt/ 'affect' vs. /iˈfekt/ 'effect'. Maybe /ˈkʊki/ 'cookie' vs. /ˈkʊku/ 'cuckoo'? J. 'mach' wust • tskʃpræːx 13:57, 4 May 2005 (UTC)

Um...who says /ˈkʊku/ for "cuckoo"? I've only ever heard this word pronounced ['ku: ku]...Tomer ^TALK 09:56, May 16, 2005 (UTC)

Erm... me and a few tens of millions of other Brits (well, /ˈkʊk.u:/ actually) -- oh, and cuckoos of course... Picapica 09:46, 5 Jun 2005 (UTC)

OK fáin. Yú dém brits énd yär skrúí pärnansíêixänz! :-p Ai l let it sláid djas qis wans!  :-p P.S. Stáp títxigh qä bärds háw tä mispärnàuns igglix!!! :-p Tomer ^TALK 12:31, Jun 5, 2005 (UTC)

For some of the words Kreidler mentions I'm not even sure it's a dialect distinction so much as free variation. I myself can certainly say either [dʒəˈlaɪ] or [dʒuˈlaɪ] and either [pɹəˈdɛnʃəl] or [pɹuˈdɛnʃəl]. (I'm also having trouble interpreting his comment "the weak final vowel of value, issue, menu and the pre-vocalic vowel of casual, virtuous, evacuate is either /uu/ as in true or /ʊ/ as in foot, or it is neither of them", since saying something is either A or B or neither of them isn't saying anything at all.) I'm also not convinced that English has separate phonemic systems of stressed and unstressed vowels, at least I'm not convinced that that's a widespread enough opinion that it can be reported here without constituting original research. The vowels that contrast in cookie and cuckoo contrast in stressed syllables as well, so there's no reason to set up different subsystems for them. --Angr/_comhrá 15:15, 4 May 2005 (UTC)

I'm adding it as the tomato-tonight merger. We can still debate whether or not it's a merger, Steve

We better discuss it first! I'm sorry for the out of context quote. You're right, Angr, the quote I've chosen is doesn't make any sense in isolation. Basically, Charles Kreidler follows the same reasoning like John C. Wells in the SAMPA recommendations for transcribing English: "The vowels /i:/ and /u:/ in unstressed syllables vary in their pronunciation between a close [i]/[u] and a more open [I]/[U]. Therefore it is suggested that /i/ and /u/ be used as indeterminacy symbols." Charles Kreidler is suggesting exactly the same thing, only that he is using different symbols (/ii uu/, /ɪ ʊ/ and /i u/). J. 'mach' wust • tskʃpræːx 20:20, 4 May 2005 (UTC)

Since we don't know yet whether there is any dialect of English where the weak u-vowel is not in free variation with schwa, I've move the following to the talk page:

Tomato-tonight merger

The tomato-tonight merger is a merger of the unstressed /ʊ/ sound in book and the schwa sound that occurs in most dialects of English. In dialects that have the merger tomato and tonight both start with /tə/, but in accents that don't have the merger tonight starts with /tʊ/. In many accents, unstressed /ʊ/ and /ə/ can occur in free variation.

J. 'mach' wust • tskʃpræːx 20:37, 4 May 2005 (UTC)

Cambridge dictionary lists July http://dictionary.cambridge.org/define.asp?key=43083&dict=CALD only with the weak u-vowel, so there are some accents where it's not in variation with the schwa sound.

Cambridge dictionary lists occupant with both options, weak-u and schwa sound http://dictionary.cambridge.org/define.asp?key=54853&dict=CALD .

Goaty-goatee merger

To my knowledge, there are dialects that merge the pronunciation of the words goaty and goatee. I don't know which dialects. Who knows about it? Or should I just go ahead and start this section without saying where this merger occurs? J. 'mach' wust • tskʃpræːx 10:20, 4 May 2005 (UTC)

Different how? I don't know the word goaty, but if asked to pronounce it I guess I'd say [ˈgoti] (or [ˈgoɾi] if it were a familiar word), while goatee is [ˌgoˈti]. In other words, the difference between them is stress, not phonemes. --Angr/_comhrá 12:37, 4 May 2005 (UTC)

Ach, phonology vs. stress! Suppose one defends a phoneme /ɨ/: Then this might be considered a minimal pair of /ɨ/ and /i/. I would however agree that the main distinction of the two is better described in terms of stress. Yet still, I think this article is the most appropiate place to mention this merger, since the main purpose of this article seems to be the description of pronunciation mergers (we might consider moving this article to pronunciation merger). J. 'mach' wust • tskʃpræːx 13:13, 4 May 2005 (UTC)

"Phonology vs. stress"? Stress is part of phonology, isn't it? I'm definitely against expanding the article to include any sort of pronunciation merger regardless of whether it concerns phonemes or not. The article is already difficult to maintain, having a way-too-long list of splits and mergers in all imaginable English accents, and being constantly subject to unverifiable, unsourced additions by anonymous users (such as the "hippy-hippie merger" above). And, not to belabor the point, I'm definitely against adding anything that isn't widely known and discussed in the linguistic literature. --Angr/_comhrá 15:29, 4 May 2005 (UTC)

I don't understand: On one hand, you're saying stress is part of phonology, but on the other hand, you're also saying that the difference between goaty and goatee is not phonemes, but stress. J. 'mach' wust • tskʃpræːx 18:12, 4 May 2005 (UTC)

Phonemes are the contrastive segments (vowels and consonants) of a language. Phonology is the entire sound system of the language, involving not only vowels and consonants, but the distinctive features that make up phonemes, as well as stress and/or tone, syllable structure, and intonation. --Angr/_comhrá 19:24, 4 May 2005 (UTC)

That makes sense! Thanks a lot for the explanation! J. 'mach' wust • tskʃpræːx 20:39, 4 May 2005 (UTC)

This is something I've been aware of for some time. Among dictionaries which use IPA for pronunciation, there has been a division when it comes to the final unstressed "-y" / "-ie" (and presumably other unstressed "i"s). Some use /ɪ/ while others use /iː/. Apparently there is a range of realisations for such unstressed "i"s so it's a tough call. Now some newer dictionaries, and I believe the more recent SOED is among them, have chosen to add /i/ for this case: "heap" /hiːp/, "hip" /hɪp/, "hippy" /ˈhɪpi/. This makes good sense for my own idiolect in which the "short" vowels /æ/, /e/, /ɪ/, /ɒ/, /ʌ/, /ʊ/ cannot occur word-finally but "long" vowels and unstressed vowels can. This is also the approach we've generally taken up on Wiktionary. — Hippietrail 12:34, 25 May 2005 (UTC)

The phenomenon you mention is discussed at happy tensing. --Angr/탉 12:57, 25 May 2005 (UTC)

As for a goaty goatee merger, that's widespread in at least Melbourne in my experience. Both words get stressed on the first syllable, and so both endings are unstressed [i], and so both t's (often but not always) get voiced. Felix the Cassowary 10:45, 2 Jun 2005 (UTC)

force-cure vs. pour-pure

I don't understand the difference between these two mergers. Is the difference in pour underlying the Horse-hoarse merger? J. 'mach' wust • tskʃpræːx 09:23, 14 May 2005 (UTC)

I hadn't noticed they were both there. You're perfectly right, they're the same merger. I'll fix it. --Angr/_comhrá 09:38, 14 May 2005 (UTC)

They're not the same merger. For me, poor and pour are homonyms, but cure rhymes with fir, not for. I'm restoring them back as they were.

They are the same merger, but they don't apply to the same words for all people. In particular, Americans are likely to use [ɝ] after palatals like [j], and so many say [kjɝ] for cure. But for the people who pronounce it [kjɔː], it's all part of the same merger. --Angr/_comhrá 05:28, 15 May 2005 (UTC)

Maybe you've spent too long away from home, Angr.  :-p Most Americans say [ki:ɹ] or ['ki ə] for "cure"... I've been all over "this great land of ours", and I've never heard "kyaw" for "cure". I must admit, I find little to support your [kjɔː] claims. West of the Connecticut River especially, such a pronunciation would be regarded in the US as "foreign", and east of the Connecticut, from my experience, it would be regarded as pedantic, and certainly not characteristic of common speech anywhere in the US. For most people in the US, cur rhymes with fur, and cure (almost) rhymes with fear. I don't know if there are different mergers going on here or what, but clearly there's a disconnect between the observations. I have never heard a dialect of AmE that rhymes cure with fir (although I've often heard "for" pronounced as if it were spelt "fir"), nor have I ever heard of a pour-pure merger. The pour-poor merger, however, is almost universal in my experience, although some speakers make a conscious attempt to differentiate between the two, pronouncing them pu:^wr (or even p^wɝ:) and pu:r, respectively. I honestly believe these differentiations to be hypercorrection. As I may have mentioned at some point in the past, I live in a something of a hyper-literate town. Tomer ^TALK 07:19, May 15, 2005 (UTC)

I never claimed [kjɔː] was an American pronunciation. I think it's a British (and probably Southern Hemisphere) pronunciaton. But I've certainly never heard cure pronounced keer by anyone, American or otherwise. The American pronunciations I'm familiar with are [kjʊɹ] and [kjɝ] (to rhyme with fur). Tomer, do you have any sources for the [kɪɹ] pronunciation? I think it's really interesting that there are people who pronounce it that way; I'd like to read up on it. Anyway, I think I'm going to combine all the sections about the outcomes of EME [uːr] into one section, where we can discuss the fact that some people merge it with [oɹ] (in at least some words), some people merge it with [ɜɹ] (in at least some words), and, apparently, some people merge it with [ɪɹ] (in at least some words). But whatever we say it's important to provide sources for all our claims; all the Wikipedia articles on English accents and dialects suffer from too much unsubstantiated "the people around here say it this way; the people over there say it that way" and not enough citation of verifiable sources. (I confess I've been guilty of doing that too.) --Angr/_comhrá 10:57, 15 May 2005 (UTC)

You'll notice I haven't modified the article to reflect my unsourced claim.  :-p Unlike you, apparently, I've heard lots of people say "keer". In fact, most of the time words are shown written pseudophonetically in dictionaries, they use the system you are relying on. I think it's outdated. the [u:] -> [ju:] has long since split into two (or more) different realizations: primarily, /ju:/ initially and /i:^w/ following consonants. Where the /j/ has modified the preceding consonant, however, it's sometimes anyone's guess what vowel the original [u] will be realized as. Sounds like something someone shirley (sic) has written a doctoral dissertation on, but I haven't found it yet... Tomer ^TALK 17:01, May 15, 2005 (UTC)

I'm pretty sure using [iw] in words like beauty, cute, etc., is actually older than using [ju]. At any rate it was almost certainly more like [iw] in Early Modern English. Using [ju] after consonants is the innovative pronunciation. But a shift of [iur] to [ir] is probably an innovation. I've never heard this dialect, so I'm very curious about it. Does it make homophones of pure and peer? Does curious rhyme with serious? What about after coronals? Is endure a homophone of endear? Is sure a homophone of sheer? --Angr/_comhrá 17:55, 15 May 2005 (UTC)

So come to Wisconsin. :-) Usually, yes, pure and peer are homophonous. Curious and serious rhyme. Endure is "in 'du:r" (the "i" there is wrong, but I don't have time right now to find the IPA symbol for a syllabalized "n"...), while endear "in 'di:r". Sure is sh.r (syllabalized r), while sheer is shi:r. Shire, on the other hand, is "sh/\ j.r", where the /\ is a short a (a "uh", not to be confused with schwa). Shire and shier do not rhyme tho, as the /a/ in shier is [a:], rather than [/\]. Tomer ^TALK 18:54, May 15, 2005 (UTC)

Here in California most people have [kj3˞] for "cure," not [kɪɹ]. I have [ɪnd3˞] for "endure" and [ɪndɪɹ] for "endear." Tomer, I definitely wouldn't say <<Most Americans say [ki:ɹ] or ['ki ə] for "cure">>, altho apparently some dialects may have that. Svenska84 19:56, 16 Jun 2005 (UTC)

Golf/Gulf merger

The prolific anon asserts that this occurs in "some dialects of AuE". First off, I'm not aware of any great differentiation between various "dialects" of AuE. Idiolects are one thing, but not really fair fodder for an encyclopedia article. AuE is often, in fact, characterized as quite homogeneous. That said, it's ridiculous to assert that this merger, if it actually is one, (is there any evidence that the vowels were actually distinct between these words? has this "merger" affected any other instances of Cul-/Col- words?) is found only in Australia. I've never heard anyone who speaks English as a native language, distinguish between gulf and golf. Maybe I just don't talk enough about gulfing down by the golf (sic), but this looks rather suspect to me. Tomer ^TALK 09:51, May 16, 2005 (UTC)

I'm not in a position to comment on dialects in Australian English, although I have heard that there are some differences between South Australia and the accents of the eastern states. But I can definitely testify that I make a distinction between golf [gɑlf] and gulf [gʌlf]. Webster's Third New International gives the following pronunciations for golf: "[gɑlf], [gɔlf], also [gɑf] or [gɔf], sometimes [gʌlf]", and for gulf "[gʌlf, gʌʊf]". (Webster's Third is rather fond of giving the option of vocalizing a coda L before another consonant; for milk it gives [mɪlk, mɪʊk].) And you're right, even for people who do pronounced golf [gʌlf] this isn't a merger unless the distinction between the phonemes /ɑ/ and /ʌ/ is always lost, or at least always lost in a similar environment to this one. --Angr/_comhrá 12:39, 16 May 2005 (UTC)

I speak AusE. I pronounce these two the same. I think that this is pretty standard across AusE. It seems to me that the environment would be a coda consisting of /l/ plus another consonant. Jimp 14Jun05

I speak northern US (some undiscerning ears and critical redneck hick (full insult intended for the insult they levy against me in saying so) from "down south" even think I have a Canadian accent in such words as out (which I clearly pronounce as owt, not oot)), and have never heard anyone, despite my wide travels throughoWt the US, heard anyone differentiate between "golf" and "gulf", and highly suspect that anyone who does make any such differentiation is exhibiting a literate affectation rather than a learned pronunciation. Tomer ^TALK 09:40, Jun 15, 2005 (UTC)

Of course you've heard people distinguish golf and gulf. You just haven't noticed it. AJD 15:19, 15 Jun 2005 (UTC)

Tomer, you may want to pay a little more attention next time you travel around the US, as many people easily and naturally differentiate between "gulf" and "golf." I'm from California and grew up saying [gɑɫf] for "golf" and [goɫf] for "gulf" (I have underlying /ʌ/ for "gulf" but the presence of the velar [ɫ] brings that vowel back and rounds it in my dialect). This is definitely not a learned distinction but one I've quite naturally made since as far back as I can remember. To my ears, someone who doesn't distinguish them definitely sounds like they're speakers of another dialect, which is fine of course, but I just wanted to show you that your perceptions based on your dialect cannot necessarily be applied to all of America or North America. As for people thinking you're Canadian, Wisconsin is an area affected by Canadian raising so if people hear something that "sounds Canadian," I don't think they're doing it to be mean (as if being Canadian were a bad thing??!) but are basing it off of what they really hear. First off, no Canadian Raising speaker (whether they're Canadian or American) says "oot"--it's something like [ʌʊt] or [əʊt], never [ut]. You may not notice it in your speech but many people form the Inland North especially around the Great Lakes region do have Canadian Raising even tho they may perceive that only Canadians are doing it. I've known people from Minnesota and Wisconsin (and have been to both states) and they have all had Canadian Raising to some degree. Svenska84

Yessir yessir 3 bags full. I didn't claim that peeps from the Great Lakes region don't exhibit various degrees of Canadian Raising, simply that it is not a universal phenomenon, neither as "complete in all speakers" nor as "complete in effect on pronunciation". That said, it's entirely possible that I simply don't hear the difference between [ɑɫ] and [oɫ], as they're not phonemic for me as far as I'm aware. This is why I avoid making edits to the article based simply on my personal experience, prefering instead to piss and moan here. As for the pronunciation of "out", I don't think anyone but Terrence and Phillip says [u:t], but [əʊt] is a pronunciation which I can tell you, categorically, is never found among native American speakers of Great Lakes accents...it is, in fact, a pronunciation which upper midwesterners quickly and accurately identify as Canadian, although those who know anything about it call it "Toronto Valley accent", rather than the rather ignorant appelation "Canadian", since it is found only rarely west of about London, Ontario, and then primarily in the speech of "imports" or by younger speakers influenced by Canadian television broadcasts. Finally, if you look at any authoritative discussion of American dialectology, you'll find that Wisconsin is split into nearly equal parts of at least 3 different dialects, the differences between which are generally well-known to people in Wisconsin. (I almost said "Wisconsinites", but since I'm not originally from Wisconsin, I had to edit myself quickly there... :-p) Tomer ^TALK 09:51, Jun 17, 2005 (UTC)

Back to its existence in AuE, there's a general merger IME between the gulf and golf vowels before /l(tdns)/ and perhaps /lf/ (as the short O vowel, hence [mOlt@plAi]), and between the bald and bold vowels before the same group (as the long O vowel, hence they're both [bOuld]). In the case of bald/bold, there are some people who avoid the merger in that particular case, it seems, but I've never heard 'fault' pronounced (by AuE speakers) as [fo:lt] (i.e. 'forlt'). (I didn't realise there was any avoidance till someone said to me last week 'oh [bo:ld], I thought you said [bOuld], and I said, 'I did, that's how it's pronounced ...?' and discussion ensued.) These are avoided in derived forms, so called and cold are still distinct, and the distinction between [Ol] and [Oul] is weak enough anyway that it is very nearly a four-way merger, and may be in some speakers. I can't vouch for these being AuE-only except by virtue of seeing different pronunciations from my own recorded in dictionaries (including the Macq., which is very British in many pronunciations). Felix the Cassowary 07:19, 19 Jun 2005 (UTC)

Do these dictionaries provide evidence of this merger, then? --JHJ 7 July 2005 17:04 (UTC)

No ... They just have different phonetic descriptions from what I'd expect based on my pronunciation. e.g. the Maquarie Dictionary has 'multimillionaire' as /%mVlti"mIlj@nE@/, which looks and sounds, umm, very British. So that's evidence an uncommon merger has taken place, but not evidence of how general it is, nor does it constitute written evidence. A study (Cox, F. & Palethorp S. (2003). 'The border effect: Vowel differences across the NSW–Victorian Border, Proceedings of the 2003 Conference of the Australian Linguistic Society: 1–14) on the Victorian merger of /æl/ and /el/ did not find any evidence of an effect of /l/ after low back vowels, but (a) it was in a hVl context i.e. not where /o:/ (RP /O:/) or /a/ (RP /V/) are affected, and (b) it was focussing on height and frontness, but the difference between short o /O/ and the starting point for long o /@u\/ before an /l/, i.e. [Ou] are effectively the same, so they may've missed it. (/u\/ = u-dashed.) Interesting that you have some of the mergers too. Felix the Cassowary 7 July 2005 23:41 (UTC)

I (from Sheffield, UK) have some similar mergers: the /ɒ/, /ɔː/ and /oʊ/ vowels all merge in the environments you're talking about (so, for example, salt, fault, golf and bold all have the same vowel), but /ʌ/ isn't affected, so gulf and golf are distinct. If the /l/ is at the end of a word or followed by a /d/ or /z/ from an inflectional ending, then I feel /ɒ/ and /oʊ/ merge (doll and dole) but /ɔː/ stays distinct (so cold and called are different). I'm not aware of any published description of this. --JHJ 7 July 2005 17:04 (UTC)

Bad-lad split

As far as I know the sounds [æ] and [æː] are allophones of the phenome /æ/, not separate phonemes in Australian English. Correct me if I'm wrong, but one of the ways of telling the difference between phenomes and allophones is to use minimal pairs, isn't it? And if so, then there arn't any minimal pairs in Australian English for these two sounds. – AxSkov (T) 13:47, 24 May 2005 (UTC)

I have e-mails from linguists who are native speakers of both Australian English and English English who claim to have minimal pairs between the noun (and denominative verb) /kæːn/ and the auxiliary verb /kæn/ and between the noun /spæːn/ and the past-tense verb /spæn/, or near-minimal pairs between long /bæːd/, /sæːd/ and short /læd/, /fæd/. Apparently the two can contrast only before voiced obstruents and nasals, though, and not everyone does it, which is why I hedged with "some varieties of" English English and Austrlian English. There's more about it in the works cited as well (Wells, Burridge, Leitner). --Angr/_comhrá 14:07, 24 May 2005 (UTC)

The differentiation between /kæn/ and /kæ:n/ isn't limited to BrE and AuE. The noun and its related verb are usually pronounced with /æ:/ or with æ-tensing in Wisconsin, but the auxiliary verb is unaffected by æ-tensing, and in unstressed positions is usually pronounced /kn̩/ (although not in "can't", as I've mentioned before, in which it is usually pronounced /kæⁿʔ/). Tomer ^TALK 17:27, May 24, 2005 (UTC)

There's a fellow at Yale named Erich Round, from Melbourne, who's written a paper arguing that the Australian pattern is some sort of combination of the RP pattern and the Mid-Atlantic pattern. He's sent me a PDF of his paper (which I haven't read yet), but unfortunately (for him and for us if we wanted to cite it) he hasn't been able to get it published. --Angr/_comhrá 17:41, 24 May 2005 (UTC)

Is it possible for me to get a copy of this PDF? or is such a question from a nobody on the Internet considered inappropriate? Felix the Cassowary 11:27, 2 Jun 2005 (UTC)

If you send me an e-mail and give me your e-mail address I'll send you a copy. --Angr/_{tɔk tə mi} 11:42, 2 Jun 2005 (UTC)

Hal-hell merger

In my idolect (raised in Melbourne Australia, lived in various places in and out of Australia since), I merge /el/ and /æl/ into /æl/. In fact I can only pronounce the Israeli national airline as /ˈælˈæl/. I haven't been able to find other people talking of this but I'm sure it's not rare in Australia. Does anybody know of anybody who has commented on this phenomenon and if there's an accepted name for it? — Hippietrail 12:53, 25 May 2005 (UTC)

All I know about it is the fact that at List of words of disputed pronunciation it's asserted that younger people (under 30) in Melbourne may pronounced the name of their city [ˈmælbən]. --Angr/탉 13:11, 25 May 2005 (UTC)

This is widespread only in Victoria, apparently, but it very definitely exists. But I can't hear the difference between el and al, so while I pronounce 'celery' and 'salary' alike, when I was talking to a Sydneysider the other week we were both bemused, he by the fact that I couldn't hear the difference (without trying) and I that he made it. It was reported in an article in The Age a few years ago. (I also have had experience with people merging 'shell' and 'shall', but due to various other splits there's not many homophones at end-of-word.) This is an unsourced comment; consider it how you will. (BTW, I don't merge 'Hal' and 'hell', the first one I say with a longer vowel, rhyming it with 'pal' (as in mate).) Felix the Cassowary 11:19, 2 Jun 2005 (UTC)

There aren't many words in English that end in /æl/ to begin with, because Middle English short /a/ (the usual source of Modern English /æ/) became /ɔː/ before a dark L: ball, call, fall, gall, hall, mall, pall, tall, wall. I'm not sure why shall didn't do this. The other words that end in /æl/ are mostly nicknames (Hal, Sal) or words that entered the language later as slang (pal). There's been a lot published on the phonology of Australian English, so it shouldn't be hard to find a source for this. However, since it's apparently only found in Melbourne, it seems too regionally specific for this page. It should be mentioned at Australian English instead. --Angr/_{tɔk tə mi} 11:39, 2 Jun 2005 (UTC)

Celery = /'sɛɫ ri/, Salary = /'sæ lr̩ i/. Comments please. Tomer ^TALK 09:49, Jun 15, 2005 (UTC)

True except apparently in Melbourne where for younger speakers at least (maybe older ones as well) they're both /ːsæləri/. Seems to be a sound change in progress, as well as dialect differentiation in Australia, which seems to have more regional accent differences than it used to. --Angr/_{tɔk tə mi} 10:29, 15 Jun 2005 (UTC)

Pull-pool merger

In my idiolect (see above), these words are still minimal pairs but I'm positive my pronunciation varies from what I find in dictionaries. I'm also positive my case is usual among Australians. Dictionaries give "pull" /pʊl/ and "pool" /puːl/. The latter is unnatural for me. The closest I can come is "poo'll" /puːəl/. My natural pronunciation for "pool" is /pʊːl/ but I've been unable to find a similar transcription in the discussions I've come across. Does this match other people's views? — Hippietrail 12:53, 25 May 2005 (UTC)

Well there is not quite a merger here, as the 'oo' in "pool" is longer than the 'u' in "pull", and as the vowels are not homophones then there's not a merger . I pronounce "pool" as /pʊːl/ and "pull" as /pʊl/, and I don't know anyone who pronounces it as /pʉːl/, this pronunciation seems strange. I think you would be hard pressed to find a dictionary that includes this pronunciation, whether it be /pʊl/ or /pʊːl/. I do agree with you that /pʊːl/ is the usual Australian pronunciation for "pool" and it suggests to me that the Australian dictionaries need to revise some of their pronunciations. – AxSkov (T) 13:47, 27 May 2005 (UTC)

It's not a merger. It's simply that there is an allophone of /u:/ before /l/, which is [ʊː] (actually, I think, the allophone is closer to [u:], and the normal pronunciation of /u:/ is more centred, u-dash in the IPA). Though I'm not sure if native untrained speakers are 'allowed' to hear allophones, but it's very definitely rule-based. In any case, the choice of characters with the IPA in the Macquarie Dictionary is an abomination and ought to be changed, but mostly for other reasons---I think it's ligitimate to conflate the vowel of 'fool' and 'food'. Felix the Cassowary 11:10, 2 Jun 2005 (UTC)

Not only ligitimate but not to conflate the vowels of fool and food would be an incorrect phonemic transcription because they are allophones. Dictionaries use phonemic transcriptions and for good reason. Though, I would like to see Mac. Uni.'s Dictionary better reflect Aussie pronunciation. Jimp 8 Jun 2005

That's exactly what I said :) Felix the Cassowary 07:09, 8 Jun 2005 (UTC)

cot-caught merger/restricted phoneme

Someone wrote:

"Sometimes a phonemic merger causes a certain phoneme to become a restricted phoneme. For example, the cot-caught merger causes /ɔ/ to become a restricted phoneme that can only occur before /ɹ/, so for example, card and cord are still distinct in accents with the merger as /kɑɹd/ and /kɔɹd/ but /ɔ/ has merged with /ɑ/ elsewhere making /ɔ/ a restricted phoneme. "

This is inaccurate, to say the least. The vowel in 'cord' is clearly a phonemic /o/, not /ɔ/. (cf. tear/tore/torn, shear/shorn, etc. with steal/stole/stolen and speak/spoke/spoken, not "steal/stawl/stawlen" or "speak/spawk/spawken".)

Benwing 3 July 2005 06:21 (UTC)

That's not evidence on its own of course. Those series were set up in pre-modern times, using pre-modern phonemes. In modern times and with modern phonemes, 'tear' has a different phoneme from 'shear' and in non-rhotic dialects, the phoneme in 'tore' RP /O:/ or AuE /o:/ is clearly different from that in 'stole' RP /@u/ AuE /@u\/ [Ou]. That doesn't say anything about the difference in American English, mind; just that your evidence/argument is not good.

Are there any cot/caught merging dialects that still maintain the horse hoarse distinction?

Felix the Cassowary 3 July 2005 07:15 (UTC)

Scottish English and Eastern New England merge cot/caught but keep horse/hoarse distinct. However they merge them as /kɔt/, not as /kɑt/. AFAIK all accents that merge them as /kɑt/ have also merged horse/hoarse. It's definitely attractive to consider that cot/caught and horse/hoarse-merging accents simply have no phoneme /ɔ/ at all, and Wells does mention the possibility (meaning we can put it in without being accused of original research), but the opinion that the phonemic representation of horse/hoarse in merged accents is /hɔrs/ is so firmly entrenched in linguists' minds it will be difficult to dislodge it. --Angr/_{tɔk tə mi} 3 July 2005 07:23 (UTC)

engle-angle merger?

There is some discussion on California English of the raising of front vowels before /ŋ/. In fact, in at least my dialect, there is a merger of /æ/ and /ɛ/ before /ŋ/ (to /e/). Does anyone know of anything in the literature describing this? Does it merit a mention here? Is there a name for it? Can anyone think of something better than engle-angle? maybe Bengal-bangle? Cheng-Chang? Nohat 3 July 2005 16:44 (UTC)

I first heard of this from you, several months ago, when you mentioned pronouncing bank /beŋk/, but I don't know of any literature on it. /ɛŋ/ is a pretty rare sequence in English anyway (the only native words I know of that have it are length and strength), so minimal pairs will be hard to find. What does "engle" mean? --Angr/_{tɔk tə mi} 3 July 2005 19:53 (UTC)

Mebbe he meant Engels.  :-) Tomer ^TALK July 3, 2005 20:04 (UTC)

Engle is a common name-morph, as in the names Englebert, Englehart, Engleman, Englewood, Engleking, etc., which sometimes also occurs as a name by itself; see e.g. [1]. Nohat 06:26, 20 July 2005 (UTC)

dull-bull merger?

Another merger I realized recently that I have, but which is definitely not overwhelmingly common, even around here in California, is the merger of /ʌ/ and /ʊ/ before /l/ to /ʊ/. Thus dull and bull are rhymes. There don't appear to be any minimal pairs for this distinction, so I suppose it should be hardly surprising that there is a merger for some people. Is there anything in the literature about this one? Does it have a name? Can anyone think of anything better than "dull-bull"? "null-full"? "bull-bulk"? I have trouble finding these because I have no intuition about which words have /ʊ/ and which have /ʌ/, unlike the above where which is which can be determined by the spelling. Nohat 3 July 2005 16:57 (UTC)

• To help you out, although I know around here (Wisconsin) pronounce what you are calling /ʊl/ as a syllabic l, i.e., the only vowel is the l itself:
  • /ʌ/ : dull, null, hull, bulk, sulk, hulk, pulp, etc.
  • /ʊ/ : bull, full, pull, etc.         Tomer ^TALK July 3, 2005 18:42 (UTC)

Tomer, which vowel do you have in pulpit? I have /ʊ/ but I've heard people say it with /ʌ/. --Angr/_{tɔk tə mi} 3 July 2005 19:46 (UTC)

I gots /ʌ/. Tomer ^TALK July 3, 2005 20:05 (UTC)

It doesn't seem to be predictable at all when it's /ʊ/ and it's /ʌ/. Is there any cue, whether synchronic or diachronic, to indicate which vowel it is? Or do speakers without this merger simply have to memorize which word has which vowel? Nohat 3 July 2005 21:22 (UTC)

I know of no cue. I didn't rely on memorization, it's just the way I learned English, and the way which I'm accustomed to hearing it spoken around me. I have a question for you now. When you say Fulton or Folsom, what vowels do you use before the l? Tomer ^TALK July 3, 2005 21:33 (UTC)

Well, you must have memorized it if there's no cue. Fulton = /fʊltən/, like full+/tən/; Folsom = /folsəm/, like foal+/səm/. However, the /o/ is probably pronounced more like [ɔ], but I don't think I have any phonemic /ɔ/s in my native phoneme inventory. Nohat 3 July 2005 22:05 (UTC)

It's not exactly memorisation, because you don't learn them from spelling. Its like saying you have to memorise which words have /U/ and words words have /i/, because even though there's the spelling, you don't rely on it to specify how you talk normally. Most -ull words you'll learn from speech, rather than spelling. Felix the Cassowary 4 July 2005 00:43 (UTC)

It's memorization, it's just not active memorization (sitting down with flash cards and studying them). Instead it's the automatic memorization that makes up all of L1 acquisition. --Angr/_{tɔk tə mi} 4 July 2005 05:22 (UTC)

Heh. Or in my case, L3.  :-p Tomer ^TALK July 4, 2005 05:45 (UTC)

• This merger is mentioned near the bottom of one of the draft chapters of the Atlas of North American English, which has a link from the fill/feel and fell/fail section: [2].--JHJ 4 July 2005 12:06 (UTC)

Yeah, it lists four mergers at the end of that chapter:

• The bull-bowl merger
• The hull-hall merger
• The bull-hull merger
• The hull-hole merger

unsigned comment left at 19:36, 4 July 2005 by User:205.188.117.14

Of which 0 people had the hull-hole merger --Angr/_{tɔk tə mi} 4 July 2005 19:48 (UTC)

I've done some (obviously original) research around my office and the bull-hull merger does not appear to be very common, even among natives of the same area as me. However, I have definitely found some other people that have it. They have been people who acquired English in the south bay. People from San Francisco and the east bay don't have this merger. My sample size is pretty small though, although the test is fairly easy: I just ask "do dull and full rhyme"? Nohat 06:32, 20 July 2005 (UTC)

Fleece merger

A question, for those who know more linguistics than me, about the "fleece merger". This page says that some speakers in northern England retain a distinction, with [I@] in the meat group and [i:] or [@i] in the meet group. There's some evidence, some of which has appeared on Yorkshire dialect and accent that it's more complicated than this, in that some (but not all) words in the meat set actually have [EI] (or similar) as in eight.

I have a copy of a book about traditional Sheffield dialect (Orreight mi ol' , by Don Alexander) which uses the spelling "eigh" for the vowels of eat and meat but the spelling "eea" (which I would think means [I@] or similar) for the vowels of team and cream. There's also a web page about Calderdale accents (linked from Yorkshire dialect and accent) which states that eat and meat rhyme with eight. (These accents generally have a weight/wait distinction, so I don't think there's a suggestion that meat and mate are homophonous.)

Any idea what's going on here? Is there a historic reason for eat and team to have different vowels?

--JHJ 20:18, 19 July 2005 (UTC)

Possibly: team and cream have original long vowels, going back to Old English tēam and French crème respectively, while eat (< OE etan) and meat (< OE mete) have vowels that were originally short but lengthened by open syllable lengthening. Maybe there are northern dialects where lengthened short vowels became /ɛɪ/ just like old /ɛɪx/, while old /ɛː/ became /ɪə/. Or maybe it depends on whether the following consonant was m or t. --Angr/_{tɔk tə mi} 22:27, 19 July 2005 (UTC)

It could be just random; as far as I know, there's no explanation for why "great" and "meat" no longer rhyme except for partial lexical diffusion. Similar things seem to have split mass/crass from pass/class/glass/grass in British English. Benwing 03:56, 20 July 2005 (UTC)

Well, in some dialects of British English (they all rhyme for me), but I know what you mean. Anyway, thanks to both of you for your prompt answers. I suspect it could actually be a mixture of all three explanations.--JHJ 16:20, 20 July 2005 (UTC)

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".

+

• 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)

i don't know what your point is. this article is about describing what *is*, not what *should be*. Benwing 05:50, 3 August 2005 (UTC)

There is apparently some troll/spammer that goes round to various English-related forums and posts links to that SSWD webpage, asking people's opinion on Schoonmaker's opinion. This is probly just one more example. I'm not sure if either he or Schoonmaker has a point, other than getting us all to read his webpage. Felix the Cassowary 07:08, 3 August 2005 (UTC)

I don't know what this guy is smoking, but I speak a variety of American English with a large number of mergers, some of them apparently quite rare (like NULL-FULL), but everyone that I know has a FURRY-FERRY distinction. I certainly have a MARY-MERRY-MARRY merger, but not a MURRAY-MERRY merger. I grant that probably somewhere some people have this merger, but it's certainly not very widespread and the rest of this rant is obviously ignorant ramblings. As for "spelling reform", I think all the phonemic differentiations listed in this article make for quite a nice argument against it, at least any of the completely phonemic systems that some "spelling reform" advocates advocate. Nohat 07:25, 3 August 2005 (UTC)

I took a look at that site and it's actually quite hilarious. It's the workings of a person, who, when faced with the fact that some people speak differently from him, decided to invent a system for writing English that enshrines the particular regionally marked variety that he speaks as the new "standard". It's quite humorous, really, the idea of maintaining the MERRY-MARY-MARRY distinction, which for a majority of AmE speakers is completely merged, and for the most of the rest, partially merged, but creating, out of whole cloth, a MURRAY-MERRY merger, that almost nobody has, simply because it doesn't fit easily into his transcription system. Some of the explanations in his dictionary [3] are also quite hysterical (not to mention inconsistent and capricious): "Banal" is an unusual and pretentious word. An unusual and pretentious pronunciation suits it..."Epoch" should not be confused with "epic", so either [ˈɛpɑk] or [ˈipɑk] is preferable to [ˈɛpək]. A long-E seems appropriate for a word that refers to a long time. The assertions behind supposed "correct" pronunciations make Charles Herrington Elster (of The Big Book of Beastly Mispronunciations) look erudite and linguistically informed (two things which if you've read the book you will easily come to the conclusion that Mr. Elster is not). New Yorkers sure are full of themselves. [flɑrɪdə] and [ɔrəgɑn], indeed! Nohat 08:36, 3 August 2005 (UTC)

FWIW, I understand that some area of Pennsylvania or Phillidelphia or some area that starts with a P in the US has a curry-merry merger. I also understand that American English has either had a curry-furry merger, or never split the two in the first place. So that'd make for a furry-merry merger, bizarre as that seems to my ear. Felix the Cassowary

(From Tense-lax neutralization, it seems that In the Philadelphia accent the [marry-merry-Mary] contrast is preserved, but merry tends to be merged with Murray; likewise ferry can be a homophone of furry, so I wasn't too far off. Felix the Cassowary 10:05, 4 August 2005 (UTC))

Yeah, the MERRY-MURRAY merger seemed vaguely familiar. One thing for certain, though, is that it is decidedly not is a feature of GenAm. Which of the MARY-MERRY-MARRY mergers is part of GenAm is debatable, certainly, but MERRY-MURRAY is defintely regionally marked (for Philly, apparently). On the latter, I'd say that the full 3-way distinction is (very marginally) marked, but any of the merger combinations, whether maintaining a 2-way distinction or no distinction at all, is not marked. But there are those who disagree. Canepari for example calls maintaining the 3-way distinction "neutral", which is his term for unmarked GenAm. On a totally separate topic, I'm curious which "various English-related forums" this troll has been visiting. Cheers! Nohat 07:49, 5 August 2005 (UTC)

Book-buck

I'm no linguist but I chanced on this article and I'm intrigued. My girlfriend (born in Northern Ireland and raised in northwestern England) pronounces "book" and "buck" close-to-identically (I think it's /ʊ/: in IPA), whereas I (brought up in southern England) use /ʊ/: in "book" and /ʌ/: in "buck". Similarly, for her "pussy" (as in cat) and "pussy" (as in "full of pus"") are the same, whereas I make a distinction. Her attempt at my "southern" pronounciation of the latter verges towards "passy".../ə/: or even /æ/:. Is this a recognised example of phonemic differentiation or have I missed the point? Dave.Dunford 05:10, 13 October 2005 (UTC)

Yes, it is. See Foot-strut split for the one you're talking about. --Angr/_{tɔk tə mi} 06:59, 13 October 2005 (UTC)