Talk:Standard Canadian English

/oʊ/ vs /o/ and /eɪ/ vs /e/

@Kbb2: I'm not sure why you insist that, for GOAT, /o/ is preferable to /oʊ/ and, for FACE, /e/ is preferable to /eɪ/. I know this has traditionally been the case, but can you provide some recent sources to back this up? Leading linguists on Canadian English like Bill Labov and Charles Boberg prefer the latter. Boberg's "Regional Phonetic Differentiation in Standard Canadian English" (2008), for example, suggests that, phonetically, FACE comes closest to being monophthongal in the Prairie Provinces and even then the phonetic transcription is slightly diphthongal [eːʲ]. A monophthong is similarly unsupported for GOAT. So, for a phonemic transcription, the diphthongs are better supported if we use the phonetic research as a basis. At the same time, how about the argument that using the diphthongal representation merges the standard Canadian and American varieties into a single uniform system? Isn't there some merit to that as well? I'm interested to hear any alternative arguments to these. Wolfdog (talk) 22:48, 8 April 2020 (UTC)

I agree with Wolfdog. The monophthong notation is indeed somewhat common in North America, Kenyon & Knott being perhaps the biggest influence, but I don't see why that should trump the uniformity with the far better established Gimsonian notation, especially given ⟨e⟩ represents DRESS in that system. Nardog (talk) 01:59, 9 April 2020 (UTC)

@Wolfdog: Because there's no cot-caught nor north-force contrast in this variety of English, which is why four needs to be transcribed with GOAT (this also proves that it belongs to the monophthong series as far as phonology is concerned, as in General American). ⟨o⟩ is a more optimal choice for that. ⟨e⟩ for FACE is just for consistency. Plus, let's not forget that these are commonly realized as centering diphthongs before coda /l/. There's a firm distinction between phonemic and phonetic transcription in IPA, and GOAT has a wide allophonic range in Canada (I mean within accents that don't monophthongize it, not among different regions - though that is probably true as well).

Other languages in which phonetic diphthongs are typically transcribed as mid monophthongs include Afrikaans ([iə iɵ uə] as ⟨eː øː oː⟩), Northern Dutch ([ei øy ou] as ⟨eː øː oː⟩) and Swedish ([eə øə oə] as ⟨eː øː oː⟩). Northern Dutch /oː/ has almost the same allophony as GOAT in Canadian English, with [ou] occuring in most positions, [oː] before /j/ and [oə] before /r/ (though I hear [oː] before [onset] intervocalic /r/ ["onset" is superfluous here, per Booij], especially when it's alveolar rather than uvular).

Also, in Montreal, where MARRY contrasts with the merged MERRY and MARY, I bet the latter is closer to [e] and so could be identified as FACE, which is another reason to write it with ⟨e⟩ (unless I'm mistaken and the vowel is closer to open-mid - yeah, that could be the case instead, considering how open TRAP can be in Canada).

EDIT: By the way, I would have absolutely no problem with switching over to ⟨e o⟩ in all articles about North American English for the sake of consistency. Remember than ⟨e⟩ is already used for both DRESS and FACE, depending on the accent (compare the representation of vowels in Received Pronunciation with our transcription of Scottish English). ⟨eː⟩ is similarly semi-ambiguous (only if you don't bother to learn the conventions) between SQUARE (in Australia and South Africa) and FACE (in Welsh and some Northern English), depending on the accent. It can even denote the merged SQUARE–NURSE vowel (in Scouse). Kbb2 (ex. Mr KEBAB) (talk) 02:45, 13 April 2020 (UTC)

I appreciate the explanation of your point of view, and I understand the logic, but I can keep raising further questions (and I'm sure so can you back to me): Why is the context of a following /r/ an overwhelming enough example that it wins the argument? -- that it proves [GOAT] belongs to the monophthong series, overwhelming any evidence to the contrary? Phonology is not exactly a field that's conducive to absolutes. Along the same lines, I'm surprised to hear you insisting that four needs to be transcribed with GOAT. Nothing needs to be done one way or another in transcription, and, in fact, I've cited notable scholars that certainly do not transcribe it that way. We can keep bickering, or we can actually use the leading scholars/sources to clear this up once and for all: a strategy deliberately aimed at rendering our continued bickering unnecessary. Like I said, Labov and Boberg use the diphthong; Nardog concurs, citing Gimson (or at least his system). Do you have sources that definitively trump this? Wolfdog (talk) 21:45, 14 April 2020 (UTC)

@Wolfdog: It's the context of the following /r/ (in the case of GOAT alone) and the coda /l/ (in the case of both FACE and GOAT) and the fact that (correct me if I'm wrong) just as in General American, stressed syllables of loanwords that contain mid front unrounded and mid back rounded vowels in the source language are typically approximated by FACE and GOAT, rather than any other monophthong (it's also the fact that the close-mid sounds [e o], especially when elongated, will always be heard as FACE and GOAT by native speakers of NAE). Especially GOAT behaves awfully like Dutch /oː/ (I mean when it comes to phonology) in that regard. FLEECE and GOOSE should be transcribed with ⟨ɪi ʊu⟩ if we want consistency, as they're also phonological monophthongs that surface as diphthongs in many if not most positions (especially when stressed). But they don't before /r/ nor coda /l/, so it's better to use ⟨i u e o⟩. It seems to me that the difference between Scottish and NAE FACE and GOAT is phonetic, not phonological as they're underlying monophthongs in both dialect groups - again, just as in Dutch. They might've become phonological diphthongs in RP, though, partially due to the fact that neither /eɪ/ nor /əʊ/ can precede /r/ within the same morpheme in RP (the closest vowels to those that can precede /r/ in that context are /eə/ and /ɔː/, which are realized as [ɛː oː]).

In the case of NAE, the issue of whether SQUARE (in the mary-merry-marry merged dialects) and NORTH/FORCE are best analyzed as FACE + /r/ and GOAT + /r/ or DRESS + /r/ and THOUGHT + /r/ depends on the analysis, but analyzing NORTH/FORCE as /ɔr/ when there's no /ɔ/ in the dialect outside of r-contexts and NORTH/FORCE are merged is uneconomical (this, I believe, is a compelling reason to write NORTH/FORCE with ⟨or⟩ in all dialects of NAE that have the merger for the sake of consistency, save for those which have contrastive New York-style THOUGHT that is itself very close). Also, in dialects that contrast MARRY with the merged MARY/MERRY, it's sometimes much more logical to analyze the latter as /er/ as the vowel is tense (this is probably a better analysis for the MARRY/MARY/MERRY-merged dialects anyway). In those dialects too the preferred transcription should be ⟨e o⟩. Kbb2 (ex. Mr KEBAB) (talk) 04:09, 17 April 2020 (UTC)

I'm not sure what more to say. AGAIN, I hear and understand all the logic of what you say perfectly clearly -- it all makes sense. AGAIN, at this point, we need to use (and compare) actual sources. It's not about logical argumentation at this point; it's about providing sources that bolster one transcription or another: showing what they actually say. In terms of absolutely no problem with switching over to ⟨e o⟩ in all articles about North American English for the sake of consistency., why don't you bring this question up at Talk:American English or, even better, to incorporate our Canadian debate, at Talk:North American English, so we can hear some new voices and see how the larger editing community feels? You and I can only argue each of our sides to each other so much before we (or, at least, I) go insane. Let's hear from new people. And let's hear about what actual sources say (so far, you have Wells partly on your side [for GOAT but not FACE], and I've got Labov, Boberg, and Trudgill on mine). Wolfdog (talk) 12:27, 17 April 2020 (UTC)

Actually, it appears that even Wells has switched from /o/ to /oʊ/ since 1982's Accents of English. Check out his more recent works like Longman's Pronunciation Dictionary or Sounds Interesting. He there uses /oʊ/ to represent the American phoneme, re: switching over to ⟨e o⟩ in all articles about North American English. Wolfdog (talk) 16:12, 24 April 2020 (UTC)

The NORTH-FORCE vowel

@Wolfdog: Wells treats it as GOAT + /r/ in cot-caught-merged General American (in the case of speakers that don't have a distinctive FORCE vowel), and so does Ladefoged in his description of California English. Which is obvious, given the fact that FORCE is GOAT + /r/ in Scotland and parts of Ireland. I mostly hear that vowel as closer to [o] anyway in most of North America (unless I'm deaf). Kbb2 (ex. Mr KEBAB) (talk) 02:36, 13 April 2020 (UTC)

Yes, I think you're hearing it perfectly correctly. I appreciate you bringing scholars into this. Again, though, most scholars I know have moved past Wells' 1982 phonemicization of GOAT in GenAm, if that's what you're referring to. But how about this: we can just avoid discussing the phonemics of the FORCE vowel on the page altogether (which I tried, and you reverted). The only important point we're making is that Florida, tomorrow, etc. belong to the FORCE set. You and I agree, so we can leave it at that. That can be explained without the need for phonemic transcriptions at all. And let's say our reader wants to know how Canadians actually produce that sound; we could even provide that phonetically AND with a citation ([ɔɹ] according to a 2004 Boberg source and quite possible [oɹ] in other sources). Again, no need provide any phonemic transcription whatsoever. You replacing this with the merged horse–hoarse vowel is written /or/, despite its definitive tone, I can just as easily remove due to its lack of a citation. However, we could write this: the merged horse–hoarse vowel is prononuced [ɔɹ] (Boberg 2004) with a citation and everything! Wolfdog (talk) 22:36, 14 April 2020 (UTC)

@Wolfdog: But how about this: we can just avoid discussing the phonemics of the FORCE vowel on the page altogether I don't see that as a reasonable proposal. I also don't understand what you mean by most scholars I know have moved past Wells' 1982 phonemicization of GOAT in GenAm - it's nothing to "move past" as it's not an obsolete (whatever that would mean) analysis. Can you cite those scholars?

I can just as easily remove due to its lack of a citation. Not really, unless Wells isn't a suitable source (e.g. because he says something other than what I think he says). Remember that FORCE was historically just GOAT before /r/: /foːrs, ɡoːt/ (compare Scottish /lɔt, θɔt, nɔrθ, fors, ɡot/). Treating the merged north-force vowel as /or/ (/oʊr/, if you prefer that), especially when the cot-caught and the father-bother mergers are present is strongly backed by historical phonology.

EDIT: Ok, "not really" is an unfortunate wording. I haven't provided an inline citation, that's true. Kbb2 (ex. Mr KEBAB) (talk) 15:53, 15 April 2020 (UTC)

Labov and team (including the Canadian Boberg of course) use the diphthong. Aren't these the major players/influencers on the modern North American English phonetics scene? Peter Trudgill also uses the diphthong. In fact, you can find here that he literally says "Canadian English has /ou/ in words such as borrow, sorrow, sorry. What more do you want? There's an exact quote in a credible source. Wolfdog (talk) 15:50, 16 April 2020 (UTC)

@Wolfdog: In this section we're discussing the phonemics of the NORTH-FORCE vowel, not whether GOAT should be transcribed with ⟨o⟩ or ⟨oʊ⟩. All I want is a response that's on topic. Kbb2 (ex. Mr KEBAB) (talk) 15:53, 16 April 2020 (UTC)

You've got my response: if you're going to represent FORCE phonemically, use /oʊr/. Wolfdog (talk) 15:55, 16 April 2020 (UTC)

Re edit summary

@Wolfdog: I said "per WP:BRD". What is there to be unsure about? Kbb2 (ex. Mr KEBAB) (talk) 14:37, 25 April 2020 (UTC)

@Kbb2: The continued removal of a valid source: Boberg, Charles (2008). "Regional Phonetic Differentiation in Standard Canadian English". Journal of English Linguistics, 36(2), 140. https://doi.org/10.1177/0075424208316648. Wolfdog (talk) 15:41, 25 April 2020 (UTC)

@Wolfdog: I reverted the whole thing per WP:BRD, disregarding the details. Kbb2 (ex. Mr KEBAB) (talk) 16:59, 25 April 2020 (UTC)

Father-bother merger

• lost:
• ha ha:

Hello everyone. It's not all the Canadian people have a father-bother merger. This Canadian speaker pronounced the vowel of "lost" and "ha ha" differently. 138.229.19.202 (talk) 13:51, 9 May 2020 (UTC)

That's no proof that there's no father-bother merger in Canada. Plus, lost, as a CLOTH word, has THOUGHT in those North American dialects that distinguish THOUGHT from LOT/PALM - see lot-cloth split. Kbb2 (ex. Mr KEBAB) (talk) 08:15, 23 May 2020 (UTC)

That's neither here nor there given the cot–caught merger is pretty much complete in Canada. Rather, the problem with the OP's assertion lies in thinking that a phonological merger means a merger in phonetic quality when in fact it means a loss of an opposition. Nardog (talk) 16:17, 23 May 2020 (UTC)

@Nardog: Those vowels were distinct in the past, so even if what the OP said was true, it'd mean that THOUGHT is distinct from PALM and it'd say nothing about LOT.

Agreed. Kbb2 (ex. Mr KEBAB) (talk) 14:11, 24 May 2020 (UTC)

Vowel length

@Schmoobalizer Can you add a source for your claim that vowel length is the main distinction between raised /æ/ and /ε/? If you know of such a source, and you can also verify that /i/, /u/, and /ɒ/ are also longer than /ɪ/, /ʊ/, /ɛ/, and /ʌ/, then every instance of each vowel in the article should be marked phonemically for length. Tyrui (talk) 00:09, 9 June 2023 (UTC)