Talk:Canadian raising

I'm resurrecting this half-decade old debate, because the dogmatic use of IPA on Wikipedia has gotten out of hand.

Many defend IPA phonetic symbols as the only precise way to convey pronunciation over a written medium. That may well be true, but should we consider educating people "precisely" to be more important than educating them at all? Because that's the real dilemma facing Wikipedia's dogged insistence on the use of IPA: For the vast, vast majority of readers, the IPA is—both literally and figuratively—Greek to them. It is a completely opaque wall of arcane squiggles that brings learning to a screeching halt. Is this the price we're willing to pay for perfect precision?

If I sound a wee bit frustrated, it's because I am: I've made this argument time and time and time again, and I'm pretty sure I'm on the right side of it. Wikipedia is supposed to be "for the people". Therefore, an education in linguistics and the knowledge of a complex phonetic alphabet should not be necessary to, say, casually look up the differences between American and Canadian English.

"Rhymes with..." examples, though slightly less precise than the IPA, are far more accessible—and, since the IPA is defined on this very site by the use of such rhyming examples, it strikes me as a little silly to say they're inadequate. Why not simply skip the middle man and put the rhyming words in the article body, instead of requiring readers to painstakingly cross-reference every single symbol against the IPA page? Eunomiac (talk) 00:43, 21 October 2013 (UTC)

"Rhymes with..." examples, though slightly less precise than the IPA - nothing could be further from the truth! Take a look at this page and read about the various vowel mergers and vowel splits. That's one thing. Here's the other thing: the exact quality of vowels varies considerably across dialects. And, of course, realizations from one dialect may overlap the realizations in another one. Even in speakers within the same city. For instance, Cockney soup may sound identically to RP soap - [səʊp].^[2] "Rhymes with..." my question is: in which accent?

— Peter238 (v̥ɪˑzɪʔ mɑˑɪ̯ tˢʰoˑk̚ pʰɛˑɪ̯d̥ʒ̊) 16:33, 25 November 2014 (UTC)

This article is a perfect example of what I'm talking about: articles.latimes.com/1985-04-03/local/me-28522_1_europe-vacation. It reads: "Lewis insisted Tuesday evening after returning from New Zealand, courtesy of the airline, that, "They didn't say Auckland. They said Oakland. They talk different."" Nonsense, they said Auckland in the NZ accent, which in his US accent sounds like Oakland. Unless we know IPA, such inaccurate and anecdotal descriptions of accents will be omnipresent. — Peter238 (v̥ɪˑzɪʔ mɑˑɪ̯ tˢʰoˑk̚ pʰɛˑɪ̯d̥ʒ̊) 08:34, 5 December 2014 (UTC)

Firstly (which ryhmes with Worstley) I found this article incomprehensible. I did not study linguistics and was hoping for some accesible information. I am Canadian and have lived in many Canadian cities and Spider and Rider always ryhme. Canadians don't say A Boat for about and we don't say aboot. I can say the prounciation but can't represent it in written form. I hope this article is on a list for rewriting. Hopefully by someone who can make it understandable for the average reader. Thanks

I'm not sure I've heard any accent that says aboot, even the stereo typical native canadian accent or the guy with the plaid, suspenders and the gun. . .--69.157.65.108 21:43, 11 July 2006 (UTC)

another Canadian - from ontario and have heard voices coast to coast (but maybe not them all)

The point the article makes (successfully, in my view) is that we don't say "a boot", but that anglophones without Canadian raising think we do. When a sound doesn't exist in your dialect, you're liable to grab for the closest approximate sound that does. For such people, "a boot" or "a boat" is the closest approximate, though since we have the sound, we don't understand how we could possibly be saying either of these.

For this Virginian, the way my relatives from Northern Minnesota say about is immediately reognized as boat. But they way they say boat is much closer to my bought than to my boat. The way they say bought is the same as how they say bot and the same as how I say bot. But I have no idea where people got the idea that they say a boot for about which are as distinct for them as they are for me. —Preceding unsigned comment added by 71.63.55.81 (talk) 01:02, 30 August 2009 (UTC)

Further, a lot of people have mentioned the spider/rider business. I don't know how I would classify my pronunciation of "spider", but it's not quite the same as "rider". It's more clipped, and the 'd' is closer to being voiceless (i.e. a 't'), almost like "spiter" but not quite. According to the West/Central Canadian English article, the sound of the middle consonant in both is an alveolar tap.

Saying "spider" with the ending of "rider", thus emphasizing the 'd', sounds very drawn-out to me, almost New Yorkish. --Saforrest 18:32, 31 July 2006 (UTC)

I agree, the article is very incomprehensible. All of the examples use that strange set of symbols, and if you don't understand them and the other linguistical jargon, it renders the article basically useless (I don't go to Wikipedia for indepth technical info, I go to it for an overview, like an encyclopedia) Anyways, I agree, as a Canadian I was disagreeing with all the examples, like "about -> a boat." I read the link to voice clips and came up with this: http://www.yorku.ca/twainweb/troberts/sounds/house.au . I guess it's just Americans exaggerating, to me it doens't sound like "a boat", but maybe it does to others. Maybe it's the same deal with the spider/rider buisness, but I can't understand the article enough. If we could come up with some phrases, I could record myself saying them and an American could as well, so we could have a bit of contrast. Anyways, I think this article really needs to be more understandable for a more, uh, casual (i guess) reader. --JakeParker 01:49, 14 August 2006 (UTC)

Do you feel the same about Quantum field theory? The fact of the matter is that some articles are going to require a casual reader to click on the interwiki links to understand. This happens to be one of those articles. AEuSoes1 03:46, 14 August 2006 (UTC)

I wouldn't compare this article to quantum field theory. While the QFT article may not be written to be understandable by the average man, it should be written to be comprehensible to someone with knowledge of quantum mechanics. Writing it for joe-anybody or even someone who only knows newtonian physics would be wrong, because it would be redundent (like you said, interwiki articles are there for that purpose.) However, in my opinion, people shouldn't have to understand all the IPA symbols to get anything from this article. Even though this is categorized under phonology, it is also listed under "Canadian English" and "Languages of Canada". Canadians are the ones who seem confused on the talk page. I'm not saying we should remove the IPA content and linguistic jargon, but the article shouldn't depend soley on them. It would be no harm to add better examples and clarify existing ones so that casual readers can get something from the article as well. To summerize, QTF needs only be readable to those with knowledge of quantum mechanics, but people other than linguists may be interested in this small peculiarity of Canadian accents. --JakeParker 18:31, 14 August 2006 (UTC)

The fact of the matter is that there's no better way to talk about pronunciation than with the IPA, especially with something that refers to a more technical aspect of speech. I think most people, especially Canadians, are confused because the article sounds like it's talking about something significant when it's really a very subtle phonetic aspect. I explained Canadian raising in person to someone who had it in her speech and she could barely hear it even after she understood the concepts. AEuSoes1 18:56, 14 August 2006 (UTC)

I have observed during my many years in the praries exactly what Saforrest mentioned about the loud and lout pronunciation in all the western Canadian provinces. —Preceding unsigned comment added by Wikilord17 (talk • contribs) 23:12, August 28, 2007 (UTC)

For my part, I have noticed that my pronunciations of "fire" and "spire" do not rhyme. I apply "Canadian raising" to "fire" but not to "spire". This doesn't fit any of the cases described in the article. Am I atypical? I lived my first four years in North Dakota, but my parents were from Montana (father) and Chicago (mother). Eric Kvaalen (talk) 19:08, 7 April 2008 (UTC)

Raising of /ay/ before /r/ in some or all words has been documented in some parts of the North and Northeast, but (as far as I know) not very seriously examined as a variable. As for me, I apply raising in both fire and spire but not wire; I'm from Beverly, Massachusetts. AJD (talk) 20:09, 8 April 2008 (UTC)

Yes, I also do not raise in the word "wire". By the way, my mother's parents were both from Boston (Dorchester). Is the raising we're talking about a feature of the Boston accent? Eric Kvaalen (talk) 16:05, 10 April 2008 (UTC)

I don't know—I don't think any research has been done on this. AJD (talk) 18:42, 10 April 2008 (UTC)

Could fire and spire be raised in non-rhotic Boston accents or does the /r/ have to be there for the raising to take place? Thegryseone (talk) 05:16, 8 October 2009 (UTC)

What is the documentation that write and ride have exactly the same vowel in Canadian English and do not have similar raising? I'm an American English speaker whose childhood was split between the Great Lakes area and the Mid-Atlantic. I find that I distinguish between writer and rider and between pouter and powder, not by raising a vowel but by lengthening the vowels before the d variant in each pair. Thus, writer has the vowel of write and pouter has the vowel of pout, while the vowels of rider and powder are slightly drawled or lengthened compared to those of writer and pouter. Similarly, budder has a slightly lengthened vowel, compared to butter. However, I find that my pronunciation of ride also has a slightly longer vowel sound than my pronunciation of write and that my pronunciation of bud has a slightly longer vowel sound than my pronunciation of butt, indicating that the difference my speech is making is to make a slight lengthening of the vowel before a nasal regardless of whether the nasal is in the middle (medial position) or at the end of the word. Thus, this is not analogous to Canadian raising. But it makes me ask if Canadian speakers are actually using the same vowels in pairs such as write/ride or if it just sounds like they do because the words are also differentiated by nasalization. -- Bob (Bob99 (talk) 22:55, 28 January 2013 (UTC))

Maybe I'm just reading it wrong, but there seems to be something missing between the first and second paragraphs, detailing the vowels used by non-Canadian speakers that are used to approximate Canadian vowels, and which have the effect of exaggerating the Canadian raising. I don't know if it has been removed or was never there at all, but perhaps that section should be cleaned up a bit?

--65.95.137.181 (talk) 17:09, 25 May 2009 (UTC)

I'm from way south in Maryland, and I have the impression that raising /aɪ/ to [ʌi] before nonvoiced consonants is nearly standard in "garden-variety" American English (you know, the speech that is without accent :-)). Is it not extremely common for speakers throughout the U.S. to follow the pattern of [aɪ]/[ʌi] in all pairs like:

eyes/ice

ride/right

strive/strife, and

tribe/tripe?

It even seems to me to be productive: it would apply for such a non-word pair as glizhe/glishe. Note that I am not asserting universality; as one example, I'm well aware that for some speakers tribe rhymes with drab. But is my intuition about the frequency of this raising south of the Canadian border way off?—PaulTanenbaum (talk) 03:11, 16 September 2009 (UTC)

Oh, and by the way, in the great rider/spider debate, for me the former is pronounced with [aɪ] but the latter with [ʌi]. So unless I were speaking very slowly and carefully, I would not pronounce spider the same as I would the verb phrase "spied 'er."—PaulTanenbaum (talk) 03:21, 16 September 2009 (UTC)

That is explained as a boundary effect. But if that's the reason, why do spider, tiger and cider, eider, Snider, Niger, Geiger not have the same vowels? Could this be a frequency effect, such as the number of syllables in every (2), memory (2-3), mammary (3)? — kwami (talk) 21:24, 10 September 2010 (UTC)

My speech has "writer" and "rider" distinct. I grew up in Tucson AZ from age 4, nowhere near Canada ... go figure. (Before that, in New Haven CT. Possibly the distinction was learned that early and kept?) In the "great spider/rider debate" as someone termed it, naturally these words don't rhyme for me. The "spider" class has "spider", "tiger" and variably "cider" (a semi-common term in my speech, not as common as "spider" or "tiger"), while "geiger", "Snyder" etc. belong to the "rider" class, which makes me think this is indeed a frequency effect.

Another strange case: "high school" (as an idiomatic term) is raised, but "high school" (interpreted literally) is not.

A similar frequency effect seems to apply in my pronunciation of written "o" before the sounds that ought to trigger the lot-cloth split in non-final syllables. In other words, my speech has the cot-caught distinction (despite having grown up in Tucson), and the "caught" vowel occurs consistently in monosyllabic words before /f/, /s/, /θ/, /ŋ/, /g/ except a few exceptions (e.g. "Goth", "scoff"), but quite variably before these same sounds in non-final syllables. "Offer", "often" and "coffee" have the "caught" vowel, but "proffer" and "Moffett" the "cot" vowel; "foster" usually the "caught" vowel, but "roster" and "Cossack" and "ossify" the "cot" vowel; "donkey" and "dongle" have the "caught" vowel, but "bongo", "Congo" and "Mongol" the "cot" vowel. It seems like words marked as [+proper-name] or [+Latinate] or [+foreign] are consistently without the shift, but others sometimes have the shift, esp. common words.

Finally, the same raising occurring in "writer" occurs with /ar/, hence "carter" and "carder" are different in my speech. (Likewise this raising is quite audible in words like "start" and "dark".) But this does *NOT* apply to any other sounds, hence "pouter/powder", "butter/budder", "metal/medal" are all the same. It seems like some people have generalized the lengthening before final voiced sounds into morpheme-final positions as well, similarly to what routinely happens in phonemic splits (e.g. the shifting of "o" to the "caught" vowel in my speech occurs in "frosting", "tosser", "longing", and anywhere else there's a clear morpheme boundary). Does this mean these people have incipient phonemic vowel length? Benwing (talk) 01:35, 20 February 2013 (UTC)

ʌɪ isn't referenced on the IPA page. — Preceding unsigned comment added by 184.147.122.14 (talk) 11:41, 11 February 2014 (UTC)

I can't even find out what ʌɪ means when I Google it. I literally have no idea how to pronounce this, and so I can't tell what the difference between rider and writer is. 99.236.50.122 (talk) 05:47, 1 August 2014 (UTC)

WTF? ʌɪ is two letters, of course it's not on the IPA page. 129.97.124.176 (talk) 21:26, 14 November 2014 (UTC)

I live in Ontario and virtually no one here exhibits Canadian raising except the extreme geriatric (over 90). Everyone else speaks general american. The "aboot/aboat" diphthong is something of a myth, all Canadians use the vowel and "loud" to say about. Ride and price always rhyme too- no raising there either.

Please delete this article. Canadian raising is extinct. Now just let the concept go.. — Preceding unsigned comment added by 99.224.205.57 (talk) 22:40, 17 October 2014 (UTC)

Your anecdotal evidence needs to be substantiated by sources before it can justify the deletion of an entire article. I wouldn't claim to have any authority on this matter, not being a Canadian or a dialectologist, but I have certainly heard Canadian raising used by Canadians under 90 in various places in media. If Wikipedia were based on anecdotal evidence, my anecdotal evidence would perhaps disprove your nearly-universal claim. As it is, anecdotal evidence is worth nothing, and you need a source that supports your claim. (Furthermore, even if Canadian raising no longer exists, the article could not be deleted merely for that reason: the sound change existed at one point, and there would still need to be an article covering it.) — Eru·tuon 03:16, 18 October 2014 (UTC)

Ok I don't know if this conversation is still going on, but here's my two cents. I am 40 years old and grew up/still live in the general BC south-west coast area. We saw a LOT of American television growing up, where most people spoke with no accent, but southerners and people from the Bronx DID (bearing in mind that this was '70s and'80s TV).

For me, "about" rhymes with "a lout".

"Spider" rhymes with "rider" and tiger and cider, eider, Snider, Niger, and Geiger, but not with "writer." I was having a hard time imagining how they would NOT rhyme, but I think I've figured it out.

"Fire" and "spire" and "wire" all rhyme.

"Offer," "often," "coffee," caught," "proffer," "Moffett," "cot," "foster," "roster," "Cossack," "ossify," "donkey," "dongle," "bongo," "Congo," and "Mongol" ALL RHYME.

"Carter" and "carder" are different in my speech as well, as are "pouter/powder", "butter/budder", and "metal/medal": the "r" before the "d" in "carder" is longer and more pronounced than in "carter," and the "d" more pronounced that the "t"...I think.

And "ride" and "price" do NOT rhyme for me: "price" is a quick word, while "ride"...spends some time on the "i".

And PLEASE, the article is interesting, but I'm a mom with 3 kids, I have neither time nor energy nor desire to devote to learning the funny symbols, I'm just interested in the differences in how people speak. Please KEEP the funny symbols, but please also add more "rhymes with" examples. Otherwise, I have to just gloss over those parts of the article. I can't remember what "fronting" is. I'm just interested. And please apply my inaccurate use of the word "rhyme" for the purpose I intended it.

Thanks :)

Anysweetname (talk) 08:19, 28 November 2014 (UTC)Anysweetname

Stuff like "rhymes with" is to a large extent unacceptable in such articles - click here to read why. — Peter238 (v̥ɪˑzɪʔ mɑˑɪ̯ tˢʰoˑk̚ pʰɛˑɪ̯d̥ʒ̊) 16:26, 28 November 2014 (UTC)

Is the second vowel really lax? The only pronunciation of school that I'm aware of has /uː/, not /ʊ/. — Peter238 (v̥ɪˑzɪʔ mɑˑɪ̯ tˢʰoˑk̚ pʰɛˑɪ̯d̥ʒ̊) 09:07, 24 February 2015 (UTC)

It's a detailed phonetic transcription of my pronunciation; it sounds to me like I pronounce /uː/ as [ʊ] here, since the syllable is unstressed. Maybe it's OR to include such a detailed transcription based only on my pronunciation; if so, I will modify the transcription. However, in that case several other symbols would have to be changed, like [ä] to [a]. — Eru·tuon 09:24, 24 February 2015 (UTC)

TBH I've never seen reports about /uː/ laxing (while keeping it distinct from /ʊ/) in North America. I'd change that to simple [u]. I've also forgotten about the full–fool merger, but that is non-standard in probably all places but New Zealand. — Peter238 (v̥ɪˑzɪʔ mɑˑɪ̯ tˢʰoˑk̚ pʰɛˑɪ̯d̥ʒ̊) 09:52, 24 February 2015 (UTC)

/uː/ is certainly not laxed before /l/ in my dialect, but here I'm pronouncing school with vowel reduction, since it's the second part of a compound and the first part, high, is stressed. It seems that Stress and vowel reduction in English § Reduced vowels in the close rounded area reports reduction of /uː/ to [ʊ̈]. Are you saying that that type of reduction isn't reported for North American English? (It could be that I am reducing inappropriately in the recording — even as a native speaker.) — Eru·tuon 10:06, 24 February 2015 (UTC)

Let's invite @RoachPeter: to the discussion. I'm not sure what to think. I certainly don't hear a central vowel there, it's rather back and it seems to me that it's the same as the onset of [uə] in the second word. I've read only a tiny portion of literature that's available (there's new stuff released every day), so I couldn't know whether it is or isn't reported for North American English. The only thing I know is that I don't recall reading about it, at least not until reading the link that you provided. I'm not sure about the last sentence, if it's your normal pronunciation then there is nothing incorrect about it. — Peter238 (v̥ɪˑzɪʔ mɑˑɪ̯ tˢʰoˑk̚ pʰɛˑɪ̯d̥ʒ̊) 10:23, 24 February 2015 (UTC)

After another listen, it sounds like you're right and the two vowel qualities are similar. However, I think they may be near-close rather than close (judging from the soundfiles here), in which case the second word would be transcribed [ˈhäɪˌskʊˑ.əɫ]. My judgement may still be faulty, though; perhaps they really are close, or at least begin slightly closer than near-close. — Eru·tuon 22:31, 24 February 2015 (UTC)

I remember that you said that your /ʊ/ is laxer and more front than [ʊ], i.e. [ɵ]. That certainly opens up the possibility of /uː/ being somewhat more open than fully close. Anyway, North American /uː/ is most often between cardinal 18 ([ʉ]) and 8 ([u]), i.e. advanced back. The default sound of IPA [ʊ], which is not a cardinal vowel by the way, has more or less the same backness as the usual North American /uː/. (By the way, I've read some reports about the Southern UK /uː/ that it may be laxed to the point where /ʊ/ and /uː/ are distinguished only by length and sometimes a slight diphthongization of the latter. I'll link to them if I find them again.) All in all, I'd stick to [uː] in both of your transcriptions, unless you really want to use [u(ː)ə] to transcribe the second word. — Peter238 (v̥ɪˑzɪʔ mɑˑɪ̯ tˢʰoˑk̚ pʰɛˑɪ̯d̥ʒ̊) 22:57, 24 February 2015 (UTC)

All right, I'll change the vowel quality symbols to [u̟]. There's a clear vowel lengthening of /uː/ in the second word, which should be noted; it seems to co-occur with greater syllabicity of the /l/. NAE vowels are prototypically short, according to the dialects chart, but sometimes lengthened. Maybe short and half-long should be replaced with half-long and long; not sure. — Eru·tuon 23:46, 24 February 2015 (UTC)

Thanks. It's not that NAE monophthongs are prototypically short, it's that vowel length is merely allophonic (see pre-fortis clipping) in both GA and in GCan (see Wells (1982), volume 3). The same is true for traditional RP, because all its monophthongs have different quality (/ə/ and /ɜː/ may or may not have the same quality.) In modern RP however, SQUARE (realized as [ɛː]) contrasts with DRESS (realized as [ɛ]) only through length. So, as you can see, the phonemicity of vowel length in English in general is variable. — Peter238 (v̥ɪˑzɪʔ mɑˑɪ̯ tˢʰoˑk̚ pʰɛˑɪ̯d̥ʒ̊) 00:11, 25 February 2015 (UTC)

Oh, I see. That sounds right. This difference isn't very explicitly described on Wikipedia, which is unfortunate. — Eru·tuon 02:29, 25 February 2015 (UTC)

See Vowel_length#Short_and_long_vowels_in_English. It could use some sourcing, but as far as I can see the whole section is correct. — Peter238 (v̥ɪˑzɪʔ mɑˑɪ̯ tˢʰoˑk̚ pʰɛˑɪ̯d̥ʒ̊) 09:56, 25 February 2015 (UTC)

I have just removed the section on possible origins. There were two reasons for this, or really two which boil down to one. It seemed to be opinion-based. If a linguist made this assessment, I could justify the presence of the assessment on the page. However, as it was not sourced, I could not see the merit of its presences on the page.LakeKayak (talk) 22:47, 18 May 2017 (UTC)

@LakeKayak: The source is probably one of Roger Lass's papers, but we need a full citation. Mr KEBAB (talk) 00:07, 19 May 2017 (UTC)

Since ⟨ʌ⟩ used to transcribe the STRUT vowel is merely a conventional notation, it seems to me it should be ⟨ɐ⟩ instead especially given we're talking about actual phonetic quality of the vowel and not phonemic construction, unless actually referring to a back vowel. This goes to other articles discussing Canadian raising too. Nardog (talk) 03:15, 5 June 2017 (UTC)

This one I actually agree on. However, Wolfdog has reported the vowel to be a back vowel.LakeKayak (talk) 18:05, 5 June 2017 (UTC)

I've read a bit about the issue and the quality of [ʌ] in the raised diphthongs is variable. I don't have a source now, but from what I've read it varies from [ɐ] to [ə]. In some regions of Canada, you can even have [oʊ] for the raised MOUTH. Mr KEBAB (talk) 18:50, 5 June 2017 (UTC)

What the exact sound of the raised vowel is hardly matters here. What matters is what one means by ⟨ʌ⟩ or ⟨ɐ⟩. If one means a central vowel, or rather if the cited source says a central vowel, then one shouldn't use ⟨ʌ⟩, simple as that.

For the record, I guess one could employ ⟨ʌɪ⟩ and ⟨ʌʊ⟩ as phonemic symbols that represent any realization of raised vowels since flapped rider and writer contrast, but in such case one needs to explicitly specify them as such prior to using any. Nardog (talk) 12:05, 7 June 2017 (UTC)

@Nardog: I'll try to find a source. Mr KEBAB (talk) 12:15, 7 June 2017 (UTC)

@Nardog: Descriptions vary.

According to the vowel chart in Rogers (2000) (the one you can see in the article), the phonetic quality in Canada is [əɪ, əʊ], yet he transcribes them with ⟨ʌj, ʌw⟩.

According to Wells (AoE vol. 3, p. 494), in Canada they are [ɐi, ʌʊ]. He transcribes them with ⟨əi, ʌʊ⟩.

According to Gordon (2004, in Handbook of Varieties of English, p. 295), in Inland Northern American English they are [ɜɪ, ɜu].

Boberg (2004, in Handbook of Varieties of English, p. 360) provides a detailed description of the situation in Canada. Here's the quote:

Even among those Canadians who show consistent Canadian Raising, its phonetic implementation is not uniform across Canada. Most Canadians have two principal allophones of /aɪ/ (raised to lower-mid position before voiceless consonants and low-central or low-back elsewhere) and three of /aʊ/ (raised before voiceless consonants, fronted to [aʊ] or [æʊ] before nasals, and low-central elsewhere). One of the few phonetic variables that divides Canadians regionally is the articulation of the raised allophone of /aʊ/. In Ontario, it tends to have a mid-central or even mid-front articulation, sometimes approaching [ɛʊ], while in the West and Maritimes a more retracted sound is heard, closer to [ʌʊ]. Among some speakers on the Prairies and in Nova Scotia, the retraction is strong enough to cause some tokens of raised /aʊ/ to merge with /oʊ/, so that couch and coach sound the same, and about sounds like a boat (though never like a boot, as in the American stereotype of Canadian Raising).

Let's keep ⟨ʌɪ, ʌʊ⟩. Mr KEBAB (talk) 15:07, 7 June 2017 (UTC)

@Mr KEBAB: Thank you for your research, but I don't see how you got to that conclusion. ⟨ʌ⟩ means open-mid back by default, so one cannot use it to mean anything else, unless in phonemic transcription or with a caveat. Nardog (talk) 06:39, 8 June 2017 (UTC)

@Nardog: You can use ⟨ʌ⟩ when you mean [ʌ̈], with the diacritic implied by the context. The IPA doesn't specify how much centralization is involved when you use that diacritic, so [ʌ̈] can denote a fully central vowel. Mr KEBAB (talk) 11:10, 8 June 2017 (UTC)

@Mr KEBAB: "with the diacritic implied by the context" My point exactly. Especially when looking at something like § Result, I don't think the use of [ʌ] (sans dieresis) is appropriate. Nardog (talk) 00:04, 9 June 2017 (UTC)

@Nardog: Absolutely. Go ahead and edit that information according to the quote from Boberg (2004) if you feel like it. Outside of that section though (at least whenever the exact phonetic quality of the diphthongs isn't discussed), there's little reason to keep the diacritic. I suppose there wouldn't be much harm in transcribing the diphthongs with ⟨əɪ, əʊ⟩ instead if you really wanted to, but I'll leave that to other users to decide. I think ⟨ɐɪ, ɐʊ⟩ fail to convey that the first elements are considerably raised, but that's just my POV. Mr KEBAB (talk) 00:44, 9 June 2017 (UTC)

@LakeKayak: Could you clarify this edit summary? Flapping does not occur intervocalically in determine. The description of the environment is more complex than "intervocalic". — Eru·tuon 23:40, 11 June 2017 (UTC)

@Erutuon: For one, I am not so sure that it only occurs in unstressed syllables as the previous wording made it out to be. For "what on earth" may be pronounced as [wʌɾ.ɔn.ɝθ] even when stressed. I will admit from hearing my own speech that I do tend to say just as frequently [wʌʔ.ɔn.ɝθ] when stressed, but I tend to use the latter more when annunciating every syllables, result in a more staccato speech. As a result, the t in the "what" no longer becomes intervocalic.

As for determine, the t may actually be in prevocalic so that the word is syllabified as /də.tɝ.mɪn/ rather than /dət.ɝ.mɪn/. Prevocalic-flapping is rare. While the description of the environment may be more complex than intervocalic, I feel that the description is also more complex than "between vowels in unstressed syllables".LakeKayak (talk) 20:09, 12 June 2017 (UTC)

@LakeKayak: The wording said "before unstressed syllables", not "in unstressed syllables". That would apply to the t in what on earth, as it is before the unstressed syllable of on.

I agree that the t of determine is in the onset of the syllable, but even so, it is still intervocalic, so it would be flapped if the environment for flapping is simply "intervocalic".

I'm not sure I'm satisfied with the wording "intervocalically before an unstressed syllable", as it requires a particular syllable division (that is, community being divided as /kəˈmju.nət.i/ rather than /kəˈmju.nə.ti/), which may or may not have any evidence besides this rule. (To be "before an unstressed syllable", the alveolar stop has to be in the previous syllable.) But at least it is probably more accurate than simply "intervocalically". — Eru·tuon 21:26, 12 June 2017 (UTC)

@Erutuon: Which is why I wrote "between a vowel and an unstressed vowel" at Help:IPA for English. But that felt a bit clunky so I went with "between vowels before unstressed syllables" this time, even though it might be slightly less accurate. I think the word before can be used somewhat loosely, because "before or at the beginning of unstressed syllables" or "between a vowel and the nucleus of an unstressed syllable" sounds almost redundant or too convoluted. I certainly wouldn't object to changing it to something more technically accurate, though. Nardog (talk) 05:43, 13 June 2017 (UTC)

@Nardog: I guess your wording at the IPA help page is more accurate. It's still not quite complete: flapping occurs in party [pɑɹɾi], after /ɹ/ rather than a vowel, and sometimes in winter, before /n/. But at least it's correct: it doesn't encompass all cases, but it does not include any cases in which flapping absolutely cannot occur. — Eru·tuon 18:36, 13 June 2017 (UTC)

Post-nuclear /ɹ/ is definitely a vowel. (No, I'm not joking.) Flapping after /n/ seems to be not quite the same process, based on its narrower distribution, I think. AJD (talk) 19:41, 13 June 2017 (UTC)

@Erutuon: I'm late to the party (no pun intended), but yes, the vowel (nucleus) of the first syllable of [pɑɹɾi] is [ɑɹ]. Since [ɹ] is an approximant, it can very well be considered a semivowel (Ladefoged & Johnson 2010:234), so one could (theoretically) employ the transcription [ɑɚ̯].

Flapping of /nt/ is a totally different phonomenon because not only is flapped /nt/ nasalized, it is just /nt/ and not /nd/ that gets flapped; gender doesn't become [ˈdʒɛɾ̃ɚ]. Regardless, I don't think it's worth mentioning in this article because it doesn't affect raising. Nardog (talk) 20:17, 13 June 2017 (UTC)

@Nardog: The analysis of /ɹ/ as a semivowel (which could also be transcribed /ɚ/) does make sense to me, even apart from the context of flapping. Under that analysis, the environment in party would be united with the other examples. — Eru·tuon 23:17, 13 June 2017 (UTC)

@Mr KEBAB: RE:"between vowels is prevocalic". Can you please clarify? I thought before between vowels was intervocalic and not prevocalic.LakeKayak (talk) 18:54, 13 June 2017 (UTC)

@LakeKayak: Intervocalic is both postvocalic and prevocalic. You can't have a consonant the simultaneously precedes and doesn't precede a vowel. (Also, 'intervocalic' means 'between vowels'). Mr KEBAB (talk) 18:55, 13 June 2017 (UTC)

I made a mistake there. I fixed it now. Sorry about that.LakeKayak (talk) 19:00, 13 June 2017 (UTC)

@Mr KEBAB: First, I wish to apologize for . I made a bonehead edit. Secondly, how do you feel it should best be worded? I think flapping may be a wider environment that unstressed syllables, but I seem to be having trouble finding the best wording.LakeKayak (talk) 19:18, 13 June 2017 (UTC)

@LakeKayak: I wasn't the one who reverted you in that diff (see ). I'd go with 'In most dialects of North American English, syllable-final /t/ and /d/ are pronounced as an alveolar flap [ɾ] whenever they precede unstressed vowel-initial syllables (often including the initial syllable of a following word), a phenomenon known as flapping.' You get the general idea, the wording may not be the best. Mr KEBAB (talk) 19:31, 13 June 2017 (UTC)

That works. Thank you.LakeKayak (talk) 19:39, 13 June 2017 (UTC)

The best description of the environment for /t/-flapping I can come up with is "between two vowels, where the second vowel is either unstressed or word-initial." (Note that you can get flapping before a stressed syllable only across a word boundary, as in "at anchorage".) AJD (talk) 19:41, 13 June 2017 (UTC)

Can somebody verify that flapping can occur in a stressed syllable only across a word boundary? I have some doubt of whether or not this is true.LakeKayak (talk) 02:23, 14 June 2017 (UTC)

Well, it occurs in a stressed syllable even without any word boundary involved: the previous example of [ˈpʰɑɚɾi], which, if it's syllabified /ˈpɑɹt.i/ according to the analysis of Wells described below, would have the /t/ in a stressed syllable. That it is! [ˌðæɾəɾˈəz] (my pronunciation) would also have flapping in a stressed syllable with a word boundary after it. — Eru·tuon 05:44, 14 June 2017 (UTC)

@Mr KEBAB: "Syllable-final /t/ and /d/" is still not perfect because, as Erutuon has previously stated, it cannot account for /t, d/ that are analyzed as syllable-initial (/kəˈmju.nə.ti/). I also think "often including the initial syllable of a following word" wrongly implies that it occupies the majority or at least a fair proportion of all instances of flapping. (Besides, do we really need to describe flapping that thoroughly in this article? The current one is quite a run-on sentence.) I think Ajd's suggestion is pretty reasonable. Nardog (talk) 20:17, 13 June 2017 (UTC)

I support Ajd's phrasing too. It happily avoids the thorny question of syllable boundaries. — Eru·tuon 21:29, 13 June 2017 (UTC)

@Nardog and Erutuon: What is the evidence that /-nə.ti/ is a correct syllabification? The reason Wells syllabifies it /-nət.i/ is not only because the /t/ can be flapped, but also because it can be glottalized in southern England, /ə/ is shortened and /t/ is articulated more weakly than it would be had the correct syllabification been /-nə.ti/. At least that's how I understand it. I don't view this issue as that "thorny".

Yeah, I asked to check my sentence, and that's not what I meant to imply. What I meant to imply was that flapping accross word boundaries happens less frequently than word-internally, where it is almost completely mandatory in General American. But I think your version is better, yes. Mr KEBAB (talk) 22:39, 13 June 2017 (UTC)

Well, using glottalization as evidence for syllabification seems equally as ad-hoc as using flapping for the same purpose, since they occur in some of the same situations. But perhaps there are other reasons for this syllabification, so it's not really ad-hoc. — Eru·tuon 23:13, 13 June 2017 (UTC)

@Erutuon: Yeah, I just said it: as far as I understand Wells's analysis, he writes /-nət.i/ because /ə/ is clipped and /t/ is articulated more weakly due to its syllable-final position (in which all consonants are weaker than when they are syllable-initial). I think the answer is the second reason: the flap and the glottal stop are just products of further lenition of /t/. But this is complicated by the fact that in southern England you can just as often hear the more posh pronunciation [kəˈmjʊʉnətsɪi] (esp. by females), with a strongly affricated /t/. I don't know. Mr KEBAB (talk) 23:21, 13 June 2017 (UTC)

@Erutuon: I originally removed the information because I thought original research was being added onto the page. I was a little puzzled by it. Also, it seems that this user may be using multiple IP addresses: one, two, and three. However, if you feel otherwise, then we can discuss it through.

• First, regarding this:

"(Also note that in six of those nine words, /aɪ/ is preceded by a coronal consonant; see above paragraph. In five [or possibly six] of those nine words, the syllable of the morpheme after the syllable with /aɪ/ contains a liquid)

While I initially identified it as original research, I now think that it is simply far too trivial.

• Second, I was a little surprised by this:

Raising may also be influenced by what that same study calls "biphone neighborhoods." So if a given speaker pronounces fight as [fʌɪt], then other words beginning with the sequence /faɪ/, e.g., fine, file, may also come to have a diphthong with a raised first element in that speaker's idiolect. Fight would then be the "neighborhood anchor" of the /faɪ/ biphone neighborhood. A word might become the neighborhood anchor due to its frequency or for some other reason.

• I told the IP user that it probably would fit best a new section altogether. As it discusses data from a recent and specific finding, the section could start with "Canadian raising has been studied at large in Hall's survey, during which he/she found that ...". That's all.LakeKayak (talk) 23:19, 28 June 2017 (UTC)

I believe that IPv6 addresses unpredictably change by default. (Kind of annoying on Wikipedia, where that is the only identifier we have.) So the editor was not intentionally changing their IP address to conceal their identity: they have no control over their IP address. Probably it's best to discuss here rather than posting at the talk page for one of their IPs (which are liable to change).

I agree that the added content is unwieldy and needs restructuring. However, it would be better to do that yourself rather than simply deleting the content. — Eru·tuon 03:31, 29 June 2017 (UTC)

I think the thing about liquids may actually be relevant. I raise /ai/ before most voiceless consonants and for me, I've noticed that the only exceptions where it's raised before a voiced consonant are when it directly precedes a syllabic liquid. --2601:246:C602:67B3:F520:5E6A:3B4B:E98E (talk) 22:29, 4 November 2017 (UTC)

• • • • Debate over validity****

Many Canadians dispute the existence of Canadian Raising. Recent studies are needed to confirm the existence of vowel raising in modern day Canada. — Preceding unsigned comment added by 184.145.55.244 (talk) 15:47, 2 September 2020 (UTC)

Do you have any sources for that claim? Unbeatable101 (talk) 01:44, 23 April 2021 (UTC)