Affricates with following r

Josh Lalonde   Sat Feb 24, 2007 3:00 pm GMT
In my accent, /t/ and /d/ with following /r\/ become [t_S] and [d_Z] (though I think they are actually retroflex). Are there any accents that don't do this? Also, when /s/ occurs before these affricates, like in 'strike', it becomes retroflex. What accents do and do not have this?
Stephen W.   Sat Feb 24, 2007 3:10 pm GMT
My "tr" and "dr" are always [tr\] and [dr\] never [t_Sr\] and [d_Zr\]. I certainly don't have affrication in "str" clusters.
Sho   Sat Feb 24, 2007 5:11 pm GMT
I pronounce tr- and dr- clusters as affricates, and str- cluster becomes something like /Str/ when I'm not being careful with my speech.
Liz   Sat Feb 24, 2007 7:34 pm GMT
<<My "tr" and "dr" are always [tr\] and [dr\] never [t_Sr\] and [d_Zr\]. I certainly don't have affrication in "str" clusters.>>

Are you sure? How do you do that?
Lazar   Sat Feb 24, 2007 7:35 pm GMT
<<Are there any accents that don't do this?>>

No, I don't pronounce /tr/ and /dr/ as affricates. When I produce the sequences [t_Sr\] and [d_Zr\] as a phonetic exercise, they sound very different from how I pronounce /tr/ and /dr/.

There is a slight transitional sound between the plosive and the [r\], but I don't think this is significant enough to be transcribed, and I've never perceived it as an affricate.

I've read about some people who, in childhood, intuitively identified /tr/ with the /tS/ phoneme, and considered "true", for example, to be equivalent to "chrue" (which I never did), which I think indicates that they have much more consistent affrication there than I do.

<<Also, when /s/ occurs before these affricates, like in 'strike', it becomes retroflex. What accents do and do not have this?>>

The [s] in my /str/ is always alveolar - I don't think the [s] there undergoes any allophony at all. I have heard some people on TV produce /str/ with [S] or [s`], but this sounds noticeably different from how I say it.
Stephen W.   Sat Feb 24, 2007 7:41 pm GMT
<<My "tr" and "dr" are always [tr\] and [dr\] never [t_Sr\] and [d_Zr\]. I certainly don't have affrication in "str" clusters.>>

<<Are you sure? How do you do that?>>

How? I simply do. If I pronounce a hypothetic word "jraw", it sounds different from my "draw".
Travis   Sat Feb 24, 2007 9:55 pm GMT
I have consistent affrication for /tr/, /dr/, and /str/, them being [tSr\], [dZr\], and [StSr\] (or sometimes [S:r\]), the only exception is if there is an intervening vowel between /t/, /d/, or /st/ and /r/, where sometimes I may not do such affrication (but then, there are cases IMD where historical /t@r/, /d@r/, and /st@r/ *do* have such affrication, but these do not seem to really be predictable).
Josh Lalonde   Sun Feb 25, 2007 1:38 am GMT
Yeah, my /t@r/, /d@r/, st@r/ sequences vary as well. I have 'interesting' [Int_Sr\EstIN], but 'interactive' [Int@`{ktIv]. I think the difference may be that the syllable following historic /@r/ is stressed in 'interactive', blocking the elision of /@/, whereas in 'interesting', it is unstressed. It could also be due to the morpheme boundary between 'inter-' and '-active'.
Sho   Sun Feb 25, 2007 4:47 am GMT
I recorded myself pronouncing the following words.
I think I do have affricate tr- and dr- clusters. interactive is an exception because there's a vowel between t and r.

true
draw
strange
Australia
drink
interactive

How is it different from the way you guys say it?
http://download.yousendit.com/D90C680D0A2D1373
Stephen W.   Sun Feb 25, 2007 1:53 pm GMT
I have affrication of "tr" and "dr" with them merging with the affricates /t_S/ and /d_Z/.

"jaw, draw" /d_Za:/
"jive, drive" /d_ZaIv/
"chew, true" /t_Su:/
MegaBox   Sun Feb 25, 2007 1:56 pm GMT
Sorry. The above post was mine. I was too busy reading Stephen W's posts.
Skippy   Sun Feb 25, 2007 11:25 pm GMT
I don't think anyone's said this yet, I didn't get a chance to read all the way through... but the reason it sounds like "ch" and "j" is because they're retroflex... It may sound about (or even exactly) the same. The difference is that to English speakers they are still perceived to be types of "t" and "d" so they're allphones, not separate phonemes.