NEAR and SQUARE

I'm having trouble figuring out the exact phonetic values of these two sets. In 3M-merged accents it's often hard to assign the vowel of SQUARE to either /e/ or /E/, since it's usually realized somewhere in between. I've been transcribing SQUARE as [skwE_rr/] but I don't think that's accurate. I sounds to me almost like a centring diphthong [E@`], but I've never seen any transcriptions of Canadian English that showed this, or of any dialect without a non-rhotic heritage. I also think there is some alternation between pre-vocalic and word-final positions. This is my tentative transcription of these two sets:

care [kE_r@`]
caring [kE_r@`r/IN]
heritage [hE_rr/I4IdZ]

hear [hi@`]
hearing (v.) [hi@`IN]
hearing (n.) [hIr/IN]

I'm not sure if this distinction is consistent though; are there any minimal pairs of a word ending in NEAR or SQUARE plus -ing with a single-morpheme word? Could you guys take a listen to my comma passage here http://media.putfile.com/Comma-Passage and see if my transcriptions seem accurate? I don't have access to a microphone right now, but I can record some of these words when I get home.

<<I sounds to me almost like a centring diphthong [E@`], but I've never seen any transcriptions of Canadian English that showed this, or of any dialect without a non-rhotic heritage.>>

I actually have seen transcriptions of North American English that use [E@`] and [I@`], and in fact that's the transcription that I prefer. But note that I tend to use these as a systematic convention, even if a specific realization may be monophthongal. As I've mentioned elsewhere, it may really be any of a combination of diphthongization, height, and length that distinguishes these rhotic vowels from basic [E] and [I]. In my own case, I think that the nucleus of my [E@`] is closer than my basic [E], and I think that the nucleus of my [I@`] is fronter than my basic [I].)

For example, on the Merriam-Webster's Learner's Word of the Day Archive ( http://www.learnersdictionary.com/wod_archive.htm ), they use rhotic centering diphthongs in their IPA transcription. (I'll admit that they may not be the greatest authority on phonetics.) And I know that I've seen other transcriptions that use them.

<<hearing (v.) [hi@`IN]
hearing (n.) [hIr/IN]>>

So you actually distinguish those words? That's interesting. For me, the noun and the participial form are homophonous as ["hI@`IN].

<<Could you guys take a listen to my comma passage here http://media.putfile.com/Comma-Passage and see if my transcriptions seem accurate?>>

It's hard for me to tell. In fact, when I try to produce sequences like [E:r\] and [E@`], they sound nearly identical to me.

The vowels I have for NEAR and SQUARE are [I_r] (note that this is *not* centralized, unlike my usual [I], which is really [I_"]) and [e_o], even though I will in more careful or emphatic speak occasionally break the /Ir/ in NEAR words to being more like [I_r.R=]. However, many here have tenser vowels in NEAR words than I have, actually having [i] rather than [I_r].

>>care [kE_r@`]
caring [kE_r@`r/IN]
heritage [hE_rr/I4IdZ]

hear [hi@`]
hearing (v.) [hi@`IN]
hearing (n.) [hIr/IN]

I'm not sure if this distinction is consistent though; are there any minimal pairs of a word ending in NEAR or SQUARE plus -ing with a single-morpheme word? Could you guys take a listen to my comma passage here http://media.putfile.com/Comma-Passage and see if my transcriptions seem accurate? I don't have access to a microphone right now, but I can record some of these words when I get home. <<

I do not have any such alternations myself in my dialect, and would normally just use [e_oR] in the first set of words and [I_rR] in the send set of words.