Words with /Qr/ in RP

How do you pronounce these words:

foreign
Oregon
origin
Florida
forest
horrible
quarrel
warren
warranty
borrow
sorry
sorrow
tomorrow

(This is mostly for Americans, but others are free to answer of course. I'm mostly interested in the first vowel of each of these words.

foreign - /for@n/
Oregon - /or@g@n/
origin - /or@dZIn/
Florida - /flor@d@/ (that's my location)
forest - /forIst/
horrible - /hor@b@l/
quarrel - /kwor@l/
warren - /wor@n/
warranty - /wor@nti/
borrow - /bAroU/
sorry - /sAri/
sorrow - /sAroU/
tomorrow - /t@mAroU/

I have:

"foreign" : ["fO:RI~:n] or ["fO:Rn=:]
"Oregon" : ["O:R@:ga~:n]
"origin" : ["O:RI:dZI~:n]
"Florida" : ["fL\O:RI:4@:]
"forest" : ["fO:RIst]
"horrible" : ["hO:R@:bM:]
"quarrel" : ["kwO:RM:]
"warren" : ["wO:RI~:n] or ["wO:Rn=:]
"warranty : ["wO:RI~nt_hi:] or ["wO:Rn=t_hi:]
"borrow" : ["ba:Ro:]
"sorry" : ["sO:Ri:]
"sorrow" : ["sa:Ro:]
"tomorrow" : [t_h@~:"ma:Ro:] or [4@~:"ma:Ro:] or [t_hu~:"ma:Ro:] (formal)

Note though that my THOUGHT vowel is not [O] but rather [Q]; [O] is rather a positional allophone of /o/ before /r/.

With the exception of "Oregon", I pronounce all of those with [Q:], which is my merged cot-caught vowel.

foreign ["fQ:r\@n]
Oregon ["O@`@%gQ:n]
origin ["Q:r\@dZIn]
Florida ["flQ:r\@4@]
forest ["fQ:r\Ist]
horrible ["hQ:r\@b5=]
quarrel ["k_hwQ:r\5=]
warren ["wQ:r\@n]
warranty ["wQ:r\@nti]
borrow ["bQ:r\7U]
sorry ["sQ:r\i]
sorrow ["sQ:r\7U]
tomorrow [t_h@"mQ:r\7U]

Well, I say them the same way as SpaceFlight, with the exception of tomorrow, which has [Or].

I'm Aussie so I use the same phoneme as found in RP (the LOT vowel) only it's raised to more of an [O] realisation (THOUGHT is also raised: [o:]). Very uninteresting.

That's basically what I expected, but I've read that the [ar\] form is spreading to other accents. Lazar, have you heard any Bostonians using [ar\] for these words? I have [or\] for all of them, like most Canadians. (I'm transcribing it with [o] instead of the usual [O] because it doesn't make sense to me to add an /O/ phoneme that only occurs before /r/, yet doesn't contrast with /o/ and is phonetically somewhere between the two. I think Lazar mentioned somewhere that he follows the same transcription system.)

<< I think Lazar mentioned somewhere that he follows the same transcription system.>>

No, I think you're thinking of Travis actually. ;-) I do prefer to transcribe a separate /O@`/ phoneme, especially because the non-rhotic variant /O@/ here contrasts with /o/. Basically, this makes it much more convenient to compare rhotic and non-rhotic accents, so in other words, my reason for this transcription would be interdialectal compatability. (Like many phonemic issues, it's a somewhat arbitrary choice though.)

<<Lazar, have you heard any Bostonians using [ar\] for these words?>>

No, because that would entail a general merger of the FATHER and BOTHER classes. In other words, the thing that distinguishes Northeastern accents is that they're tory-torrent-unmerged, so they use the same vowel in "horrible" that they do in "cot". Thus New Yorkers, and any other tory-torrent unmerged people outside of New England, use [A] in "horrible", because they also use [A] in "cot". But Bostonians have [Q] in "cot", so thus they also use [Q] in "horrible".

I know someone that moved to the NW from Indiana, and here's how he said those words:

foreign - /for@n/
Oregon - /or@g@n/
origin - /or@dZIn/
Florida - /flor@d@/
forest - /for@st/
horrible - /hor@b@l/
quarrel - /kwor@l/
warren - /wor@n/
warranty - /wor@nti/
borrow - /bAro/
sorry - /sQri/
sorrow - /sQro/
tomorrow - /t@mOro/

I thought it was interesting that he had /Q/ for "sorry" and "sorrow", /A/ for "borrow", and /Or/ for "tomorrow".

>><< I think Lazar mentioned somewhere that he follows the same transcription system.>>

No, I think you're thinking of Travis actually. ;-) I do prefer to transcribe a separate /O@`/ phoneme, especially because the non-rhotic variant /O@/ here contrasts with /o/. Basically, this makes it much more convenient to compare rhotic and non-rhotic accents, so in other words, my reason for this transcription would be interdialectal compatability. (Like many phonemic issues, it's a somewhat arbitrary choice though.)<<

My problem with the notation /O@`/ when describing my own dialect is that, aside from my dialect not having such breaking except for /aIr/, /aUr/, and /OIr/, my dialect distinguishes /aIr/, /aUr/, and /OIr/ from /aI@r/, /aU@r/, and /OI@r/. IMD, when not followed by vowels, the two series are indistinguishable, but when followed by vowels the two series are distinguishable as such breaking only occurs in closed syllables and is not affected by morpheme boundaries.

Also, interdialectal compatibility is not as much of an issue with my dialect as it is not close enough to any dialects for what you mention to be much of an issue; rather, the only dialects for which interdialectal compatibility would be an issue are other rhotic NAE dialects from the standpoint of my own dialect. Also, while I use phonemic notation that assumes the NCVS, it maps closely enough to the notation normally used for writing GA for it to be easy to compare with phonemic transcriptions for dialects with more GA-like phonologies. (Also, my dialect is much closer to GA at the phonemic level than it is at the phonetic realization level.)

Note, however, that I have backtracked notationally on using /R/ to simply using the nondescript /r/ because it seems that it is common for individuals here to have postalveolar approximants in a general fashion in onsets and only really use uvular approximants consistently in codas (with my own idiolect being more progressive in favoring uvular approximants in all positions except after coronals and labials).

Actually, forget about /OIr/ versus /OI@r/, as it seems that /OIr/ really does not exist in my dialect (and only /OI@r/ exists, acting like /aI@r/ and /aU@r/). I was thinking in parallel to /aIl/, /aUl/, and /OIl/ versus /aI@l/, /aU@l/, and /OI@l/, which act in the same exact fashion IMD except that I can actually think of examples that cover all the cases.