"Canadian raising": expert opinions needed

[Buran]

So, the first question: do you notice this difference in vowel quality in your own speech and/or in other Canadian lects, and which diphthongs does it affect for you?

It's noticeable upon observation, but I don't think anybody who isn't told about it would notice it on their own. It affects /aj/ and /aw/ (changing them to /ʌj/ and /ʌw/), but not /ɔj/. My dialect is that of Vancouver, specifically near Arbutus Ridge.

(2) Described as above, CR looks like a rather uninteresting allophony. However, Canadian English also features intervocalic alveolar flapping, with intervocalic realizations of /t/ and /d/ becoming indistinguishable between two vowels (betting=bedding). Yet the realizations of the preceding vowels can remain different due to CR, so certain words still sound differently, although the difference is in the quality of vowel rather than the consonant. Examples:

- for /aɪ̯/: rider vs. writer;
- for /aʊ̯/: clouding vs. clouting;
- for /ɔɪ̯/ I can't find a minimal pair, a quasi-minimal one: avoiding vs. exploiting.

How does your lect treat the words in these pairs?

For me, the distinction is mostly length, actually, but this may be in combination with the raising. I perceive "wide" as /wa:jd/ and "white" as /wajɾ/, though phonologically, it's probably more like [waˑjd] and [wʌjɾ]. "Ice" and "eyes" is one where length is very distinctive, at least to me; it sounds like /ʌjs/ and /a:jz/. The raising is much more perceptible here as well.

The vowels /ɪ/, /ʊ/, /ə/, and /ɛ/ exhibit a less noticeable length contrast in my lect, and only in monosyllables. "Bit"/"bid", "soot"/"should", "but"/"bud", and "bet"/"bed" have a smaller difference in length than "wheat" (/wijɾ/)/"weed" (/wi:jd/), "root" (/ruwɾ/)/"rude" (/ru:wd/), "rate" (/rejɾ/)/"raid" (/re:jd/), "wrote" (/rowɾ/)/"road" (/ro:wd/), "bat" (/bæɾ/)/"bad" (/bæ:d/), and "rot" (/rɑɾ/)/"rod" (/rɑ:d/). "Wetting" and "wedding", as well as "butting" and "budding", are homophones for me.

In a number of dialects (including Travis B.'s lect) in certain environments the distribution of raised vs. non-raised diphthongs appears to be sensitive to morphological structure of the word, breaking the neat correlation with historical consonant voicedness. Some cases of this type:

- morphologically obscure words like spider don't have the raising and are grouped with writer rather than rider;
- morphologically obscure words like idle don't have the raising and are grouped with title rather than tidal.

The above examples are about unexpected vowel quality before a voiced /d/; there are some curious examples with diphthongs before voiceless consonants, too:

- eyeful (no raising) vs. Eiffel, rifle (with raising);
- high school (with raising) vs. high score (no raising).

What vowels does your lect have in the above examples (spider and idle, in particular)?

For me, "spider" is /spajdr/, "idle" is /ajdl/. "Eyeful" is /ajfl/, "Eiffel" is /ajfl/, and "rifle" is /rajfl/, although perhaps the vowel is closer to /ɐj/, so these may be moving towards /ʌj/. "High school" is treated as a compound, explaining the presence of the raising, while "high score" is two separate words.

Interesting note on length in my lect: nasals and liquids seem "invisible" to length distinctions- the consonant after the nasal/liquid determines vowel length. In "carp" (/kɑrp/)/"carb" (/kɑ:rb/), for example, the /r/ is "invisible"; vowel length is determine by /p/ vs. /b/. In "fount" (/fawnt/) and "found" (/fa:wnd/), the /n/ is "invisible".

[Basilius]

Thank you, Adjective Recoil!

Before I ask any more sensible questions...

[waˑjd] and [wʌjɾ]

I suspect you've been mistyping "ɾ" (alveolar flap) for "ʔ" (glottal stop)? One would expect [wʌjʔ] for white.

For me, "spider" is /spajdr/, "idle" is /ajdl/

I think you do flap your /d/'s in this position, don't you?

Also, the real issue is whether spider rhymes with rider or writer (assuming the latter two are different); likewise, if idle rhymes with title or tidal.

[Ser]

words like spider don't have the raising and are grouped with writer rather than rider

In other words, they have the raising. Right?

[Basilius]

words like spider don't have the raising and are grouped with writer rather than rider

In other words, they have the raising. Right?

Yes. Must sleep more

[Buran]

Thank you, Adjective Recoil!

Before I ask any more sensible questions...

[waˑjd] and [wʌjɾ]

I suspect you've been mistyping "ɾ" (alveolar flap) for "ʔ" (glottal stop)? One would expect [wʌjʔ] for white.

It's not quite a glottal stop- my tongue does end up in alveolar position. Maybe it's best described as a tap, or /t/ with no audible release. However, when speaking rapidly, this tap does change into a glottal stop ("whitewash" is /wʌjʔwɑʃ/, but "white" is /wʌjt/, the /t/ with no audible release). It is flapped before syllabic /r/, and it is a glottal stop before syllabic /n/ (but not /m/; "bottom" is /bɑɾm/. In rapid speech, after /n, /t/ turns into [ʔ] (and I think it's going to disappear entirely before long), though only if /t/ is part of the coda and not the onset of the next syllable. Coda /t/ generally is eroding into glottal stop, so in rapidly spoken "Did you buy a plant at the store?" is something like /dʒba:j ə plænʔ ɛʔ ðə stɔr?/.

For me, "spider" is /spajdr/, "idle" is /ajdl/

I think you do flap your /d/'s in this position, don't you?

Also, the real issue is whether spider rhymes with rider or writer (assuming the latter two are different); likewise, if idle rhymes with title or tidal.

I don't flap the /d/ in "spider", but I do in "idle", and before syllabic /l/. "Spider" rhymes with both "rider" and "writer", but with "writer", it sounds a bit off.

[LeCiagoPanda]

About the t/glottal stop thing, I do roughly the same thing as Adjective Recoil; normally it ends in an alveolar position, but the glottal stop is mostly allophonic with this sound. I have a mainly Torontonian accent with Vancouver influence, so I'd say "Did you buy a plant at the store" like /dɪdʒə bɑ:j ə plæn æʔ ðə stor/, the same way I say /Tərɑno/ for "Toronto". I believe this is a uniquely Torontonian trait, however, where /nt/ and /nd/, if followed by a vowel, become [n].

Just for contrast, I pronounce all final unvoiced plosives in a similar way, with no audible release, and with a glottal stop which clips or cuts them short. Final *voiced* plosives, however, do not seem to have the glottal stop and instead carry on slightly longer, but also have no release.

EDIT: in retrospect, Adjective Recoil seems to be right here: "heart" and "hard" are distinguished by the vowel length, so I think the shorter vowel length in "heart" was making me think the final "t" was shorter, rather than the vowel. I still, however, feel like there is some other difference between final /t/ and /d/, I just can't figure out why.

I pronounce the diphthong in "spider" like "idle" and "rider", and I flap the "d" in all of them. I flap pretty much every intervocalic /t/ or /d/, even if the second "vowel" is just a syllabic /r/ or /l/.

I pronounce the diphthong in "exploiting" and "avoiding" the same way.

I'd say I'd pronounce the diphthong "clouting" like /ɛʊ̯/ rather than /ʌʊ̯/ if I"m interpreting the supposed "Canadian" pronunciation correctly.

I say "eyeful" without raising, but "Eiffel" and "rifle" with raising, I don't say rifle roughly like /rajfl/.

"Highschool" and "high score" both have no raising.

So, in summary; yes, I do notice the difference in vowel quality, including in people in Vancouver. For the second question, the minimal pairs were only distinguished by their diphthong, with the exception of the third entry, where the diphthongs were the same. Finally, I say "spider" like "rider" and "idle" like "tidal".

[Buran]

About the t/glottal stop thing, I do roughly the same thing as Adjective Recoil; normally it ends in an alveolar position, but the glottal stop is mostly allophonic with this sound. I have a mainly Torontonian accent with Vancouver influence, so I'd say "Did you buy a plant at the store" like /dɪdʒə bɑ:j ə plæn æʔ ðə stor/, the same way I say /Tərɑno/ for "Toronto". I believe this is a uniquely Torontonian trait, however, where /nt/ and /nd/, if followed by a vowel, become [n].

In rapid speech, for me, "Toronto" does turn into /tərɑnow/; maybe the assimilation of /t/ into /n/ has to do with stress?

[Ser]

Who the fuck says /təˈrɑnoʊ/? It's /ˈtrɑnoʊ/.

[Buran]

Who the fuck says /təˈrɑnoʊ/? It's /ˈtrɑnoʊ/.

You know what, you're right. Actually, it's almost like /tsrɑnow/, if I say it fast enough.

[jmcd]

About the t/glottal stop thing, It might be something like [ʔt].

Don't worry, we don't shoot people for necro'ing threads here.

I take it be fine if I reply to this then?

It's not just Ireland.Travis mentioned Scotland; N. England and Scotland have lots of dialects whose typical realizations are narrower than [aɪ̯] (including [ɛɪ̯], [eɪ̯] and what is usually transcribed [ʌɪ̯]).

And there was not any major Scottish migration, to my knowledge, to the US, following the American Revolution. And what there was before that generally ended up settling in areas which are not now part of the NCVS (think the South and Appalachia).

There was indeed much less immigration from Scotland to America than from Ireland but the main areas they immigrated to were New England and other areas in the north. The immigration to the South and Appalachia was mostly Northern Irish /"Scots-Irish" which is not really the same thing.

Also, the Scottish immigration to Canada is more significant percentage-wise (where only English and French ancestry outdo it and not by that much) so the Scottish influence would be more likely on Canadian raising; on the Northern side of the border.

[Basilius]

It seems that Adjective Recoil's and LeCiagoPanda's lects are different in some points, so it makes sense to discuss them separately for the time being...

In this message, I am going to talk about Adjective Recoil's lect, and my questions are primarily addressed to A. R., although comments on the treatment of the same examples in other people's lects are of course welcome.

Interesting note on length in my lect: nasals and liquids seem "invisible" to length distinctions- the consonant after the nasal/liquid determines vowel length. In "carp" (/kɑrp/)/"carb" (/kɑ:rb/), for example, the /r/ is "invisible"; vowel length is determine by /p/ vs. /b/. In "fount" (/fawnt/) and "found" (/fa:wnd/), the /n/ is "invisible".

Yes, this is important, and some caution is needed here. Compare:

I dunno. The article I posted the link to somewhere upstream mentioned raised vowels in ninth and pint for Canadian English (or a variety thereof?), i. e. basically what you have in your lect?

Okay, in my dialect, ninth and pint have short but unraised vowels, i.e. not all short historical /aɪ̯/ not separated from conditioning consonants by morpheme boundaries are raised in my dialect.

No, I do not have raised vowels before nasals or /l/ anywhere (and hence not in whilst).

That is, we know for certain that in a dialect with CR, vowel shortening before certain consonants (clusters) and CR proper can be *two different things* which shouldn't be confused.

BTW, Adjective Recoil, your transcription above seems to mean that fount has shortened but not raised vowel for you (i. e. your lect is essentially like Travis B.'s in this point). Is this indeed the case?

The vowels /ɪ/, /ʊ/, /ə/, and /ɛ/ exhibit a less noticeable length contrast in my lect, and only in monosyllables. "Bit"/"bid", "soot"/"should", "but"/"bud", and "bet"/"bed" have a smaller difference in length than "wheat" (/wijɾ/)/"weed" (/wi:jd/), "root" (/ruwɾ/)/"rude" (/ru:wd/), "rate" (/rejɾ/)/"raid" (/re:jd/), "wrote" (/rowɾ/)/"road" (/ro:wd/), "bat" (/bæɾ/)/"bad" (/bæ:d/), and "rot" (/rɑɾ/)/"rod" (/rɑ:d/). "Wetting" and "wedding", as well as "butting" and "budding", are homophones for me.

It is interesting that your examples of homophone pairs are about vowels showing only "a less noticeable length contrast" in final syllables. Let's check if this is just a coincidence.

Are the words in the following pairs exact homophones for you:

* seeded :: seated
* ladder :: latter
* plodding :: plotting
* waded :: waited
* coding :: coating

- ?

For me, "spider" is /spajdr/, "idle" is /ajdl/.

It is flapped before syllabic /r/, and it is a glottal stop before syllabic /n/ (but not /m/; "bottom" is /bɑɾm/.

I don't flap the /d/ in "spider", but I do in "idle", and before syllabic /l/.

OK, it seems that analysis of such sequences of alveolar articulations may be complicated, and perhaps the actual pronunciation may vary depending on speech tempo &like...

Let's check the important bit then.

- Are rider and writer homophones for you?
- Are the consonants written -t- and -d- identical in your pronunciation in these words? If not, what is the difference?
- Are the stressed vowels in these words identical? If not, what is the difference?
- Are the sounds (or sound sequences) written -er identical in these words? If not, what is the difference?

"Eyeful" is /ajfl/, "Eiffel" is /ajfl/, and "rifle" is /rajfl/, although perhaps the vowel is closer to /ɐj/, so these may be moving towards /ʌj/.

Do you mean that the vowels in eyeful and rifle are the same? Or do you mean that it's only the vowels in Eiffel and rifle that are "moving towards /ʌj/"?

[Buran]

Interesting note on length in my lect: nasals and liquids seem "invisible" to length distinctions- the consonant after the nasal/liquid determines vowel length. In "carp" (/kɑrp/)/"carb" (/kɑ:rb/), for example, the /r/ is "invisible"; vowel length is determine by /p/ vs. /b/. In "fount" (/fawnt/) and "found" (/fa:wnd/), the /n/ is "invisible".

Yes, this is important, and some caution is needed here. Compare:

I dunno. The article I posted the link to somewhere upstream mentioned raised vowels in ninth and pint for Canadian English (or a variety thereof?), i. e. basically what you have in your lect?

Okay, in my dialect, ninth and pint have short but unraised vowels, i.e. not all short historical /aɪ̯/ not separated from conditioning consonants by morpheme boundaries are raised in my dialect.

No, I do not have raised vowels before nasals or /l/ anywhere (and hence not in whilst).

That is, we know for certain that in a dialect with CR, vowel shortening before certain consonants (clusters) and CR proper can be *two different things* which shouldn't be confused.

BTW, Adjective Recoil, your transcription above seems to mean that fount has shortened but not raised vowel for you (i. e. your lect is essentially like Travis B.'s in this point). Is this indeed the case?

The diphthong in "fount" is short and unraised ([aw]). Perhaps the /n/ controls the raising, while the /t/ controls length? Maybe this has to do with English's sonority hierarchy?

The vowels /ɪ/, /ʊ/, /ə/, and /ɛ/ exhibit a less noticeable length contrast in my lect, and only in monosyllables. "Bit"/"bid", "soot"/"should", "but"/"bud", and "bet"/"bed" have a smaller difference in length than "wheat" (/wijɾ/)/"weed" (/wi:jd/), "root" (/ruwɾ/)/"rude" (/ru:wd/), "rate" (/rejɾ/)/"raid" (/re:jd/), "wrote" (/rowɾ/)/"road" (/ro:wd/), "bat" (/bæɾ/)/"bad" (/bæ:d/), and "rot" (/rɑɾ/)/"rod" (/rɑ:d/). "Wetting" and "wedding", as well as "butting" and "budding", are homophones for me.

It is interesting that your examples of homophone pairs are about vowels showing only "a less noticeable length contrast" in final syllables. Let's check if this is just a coincidence.

Are the words in the following pairs exact homophones for you:

* seeded :: seated
* ladder :: latter
* plodding :: plotting
* waded :: waited
* coding :: coating

They are not homophones, but only because of the length of the vowel in the first syllable. It's probably to do with stress. I tried saying "gravity" and "gravidy"; in both, /ɪ/ exhibits no noticeable length contrast, and both are homophones. Looks like t > d /V_V.

For me, "spider" is /spajdr/, "idle" is /ajdl/.

It is flapped before syllabic /r/, and it is a glottal stop before syllabic /n/ (but not /m/; "bottom" is /bɑɾm/.

I don't flap the /d/ in "spider", but I do in "idle", and before syllabic /l/.

OK, it seems that analysis of such sequences of alveolar articulations may be complicated, and perhaps the actual pronunciation may vary depending on speech tempo &like...

Let's check the important bit then.

- Are rider and writer homophones for you?
- Are the consonants written -t- and -d- identical in your pronunciation in these words? If not, what is the difference?
- Are the stressed vowels in these words identical? If not, what is the difference?
- Are the sounds (or sound sequences) written -er identical in these words? If not, what is the difference?

"Eyeful" is /ajfl/, "Eiffel" is /ajfl/, and "rifle" is /rajfl/, although perhaps the vowel is closer to /ɐj/, so these may be moving towards /ʌj/.

Do you mean that the vowels in eyeful and rifle are the same? Or do you mean that it's only the vowels in Eiffel and rifle that are "moving towards /ʌj/"?

"Rider" and "writer" are not homophones, because of length and raising ([ra:jdr] vs. [rʌjdr]). /t/ and /d/ are essentially identical. In both words, the sequence written "er" is a syllabic /r/ for me.

"Eiffel" and "rifle" rhyme perfectly, while "eyeful" does not (because of length); only "Eiffel" and "rifle" are moving towards /ʌj/; "eyeful" is unaffected.

Another note: I think a more general sound change, stop [-voice] > stop [+voice] /V_V, is coming into effect in Vancouver English. A few months ago, I said "bugged" (/bəgɪd/) for "bucket".

[Basilius]

The diphthong in "fount" is short and unraised ([aw]). Perhaps the /n/ controls the raising, while the /t/ controls length?

It looks like they do.

Then I'd expect a similar distribution of features in pint (and ninth) vs. find. Is my guess correct?

Also, what do you have in whilst and wild?

<...> It is interesting that your examples of homophone pairs are about vowels showing only "a less noticeable length contrast" in final syllables. Let's check if this is just a coincidence.

Are the words in the following pairs exact homophones for you:

* seeded :: seated
* ladder :: latter
* plodding :: plotting
* waded :: waited
* coding :: coating

They are not homophones, but only because of the length of the vowel in the first syllable.

Wow! But this makes a whole lot of difference!

Then, the following question is crucial.

* higher :: hire

Do these sound exactly the same?

[Basilius]

LeCiagoPanda: it would be very important now to hear your comments on the examples we've discussed for Adjective Recoil's lect.

Especially on these:

* seeded :: seated
* ladder :: latter
* plodding :: plotting
* waded :: waited
* coding :: coating

* higher :: hire

[Drydic]

While I'm not a Canuckistani, I'd be rather surprised if whilst was commonly used up north. I rarely if ever hear of NAE dialects using it.

[Buran]

The diphthong in "fount" is short and unraised ([aw]). Perhaps the /n/ controls the raising, while the /t/ controls length?

It looks like they do.

Then I'd expect a similar distribution of features in pint (and ninth) vs. find. Is my guess correct?

Also, what do you have in whilst and wild?

In both "pint" and "ninth", the vowel is [aj]. "Find" has [a:j]. I don't have a consistent pronunciation for "whilst", since I use it very rarely. "Wild" has [a:j].

* higher :: hire

Do these sound exactly the same?

They are homophones.

[Kereb]

The diphthong in "fount" is short and unraised ([aw]). Perhaps the /n/ controls the raising, while the /t/ controls length?

It looks like they do.

Then I'd expect a similar distribution of features in pint (and ninth) vs. find. Is my guess correct?

Also, what do you have in whilst and wild?

<-- Canadian.

I also have the unraised diphthong in "fount". Length I don't think is an issue for me here.
HOWEVER, I have a raised diphthong in "pint" and an unraised one in "find".

"whilst" and "wild" both unraised ... I can say "whilst" raised as well without it sounding especially odd, but that's at least partly due to the fact that "whilst" isn't heard here one way or the other.

[Kereb]

removed double post

[Bristel]

I'm from Seattle, and I'm not sure if Washingtonians have the "rising", but I say:

leg /leɪg/ instead of /lɛg/
egg /eɪg/ instead of /ɛg/

I default to /kreɪg/ for Craig, instead of /krɛg/

Is this case a Canadian rising, or something western US? Or is this the general tread for GA?

[Kereb]

No that's a different thing and I think it's probably more widespread than Canadian Raising. It is potentially but not necessarily comorbid with CR though.

EDIT: like I've always thought of it as a more "American" pronunciation but it does occur around here. I don't have "aig" myself but I remember a kid in high school who did have it, and had just noticed it, going around one day asking everybody which way they said "egg". Results were mixed!
A brief googling turns up some people on yahoo answers also attributing it to Americans, along with some Americans saying they don't have it, and of course the predictable moans of those without it decrying it as "wrong".

[Drydic]

It is indeed more widespread than Canuckistani Raising. To the point where I didn't realize until I read that post that it affects Craig (ie that people somewhere actually say [kr\Eg]), though I was well aware of leg and egg before I even knew anything about linguistics.

[Basilius]

(Apologies for slow replying...)

Re-reading Adjective Recoil's messages, I'm realizing that CR is not contrastive in his lect. What is contrastive is (secondary) vowel length (or rather, vowel clipping).

CR is a rather trivial allophony: "clipped" diphthongs are raised, long diphthongs aren't; there's one well-defined group of exceptions: words where the diphthong precedes a nasal + voiceless stop (fount, pint); in these, the diphthong is clipped but not raised. Interestingly, this group of exceptions is also seen in Travis B.'s lect.

However, the secondary vowel length contrast is not trivial and is worth further exploration.

Also, I'm realizing that nothing of what Travis has written in this thread seems to contradict a similar interpretation or raising in his lect. (Unfortunately, Travis didn't mark vowel length in his transcriptions...)

Then, this thread should properly be renamed to "Secondary Vowel Length Contrasts in NAE".

However....

<...>This is slightly off-topic, but it may have been somewhat different. I am used to there still being a remaining preceding vowel length difference between unstressed intervocalic /t/ and /d/, indicating that the two phonemes are not truly neutralized, even when unstressed intervocalic /t/ is fully voiced.

I'm not.

And,

I also have the unraised diphthong in "fount". Length I don't think is an issue for me here.

Anyway, whether or not CR is contrastive by itself, seems to depend on how a specific lect treats secondary vowel length distinctions. So it appears that we must concentrate on vowel lengths...

[Basilius]

Also,

<-- Canadian.

Which part of Canada? It looks like there's not so much uniformity...

[Kereb]

Also,

<-- Canadian.

Which part of Canada? It looks like there's not so much uniformity...

Vancouver. And I actually grew up here, which is somewhat rare, so I've been told a couple of times that I have an accent in my own hometown. Using my own family and Shit I've Read as a guide, features of the (vanishing) variety of this area are things like: pretty strong Canadian Raising affecting all of /aɪ/ /aʊ/ /ɑr/, retraction of /æ/, monophthongising the shit out of /oʊ/ and /eɪ/, and a couple other things.

[Buran]

So it appears that we must concentrate on vowel lengths...

I wouldn't discount the raising. In a few hundred years, it may have developed into something far more conspicuous than a relatively subtle allophony (for example, completely different vowels). However, I do agree that in my lect at least, vowel length, and not raising, is contrastive.