On Oct 24, 2006, at 8:17 PM, David J. Peterson wrote:

> Katya wrote:
> <<
> I can't think of any examples of stress in natlangs affecting
> pronunciation of consonants in a similar way, so I'm hoping someone
> here will either be able to give me examples or tell me that I should
> drop this idea if I care about realism. =)
> >>
> Really?  Not a one?  Not even a phoneme like /t/ in a language
> like...English?  ~:D
> This doesn't apply to all dialects, but a good number of them
> have the following realizations for /t/:
> [t]
> [t_h]
> [4]

I've always been a little skeptical of the claim that the third one  
is a tap/flap; in my dialect it sounds like /d/ and not very much  
like the Spanish /4/ that I'm familiar with. Besides that, sometimes  
I hear people who definitely *do* use [4], and it sounds odd to me.  
For example, Terri Gross on NPR's "Fresh Air" often says something  
like "Coming up on today's Fresh Air" and the word "today" sounds to  
me almost like [log in to unmask]

> With some minor variations, you get the first after [s] and word-
> finally; the second, word-initially.  The latter is dependent entirely
> upon stress (and vowels).  Take the word "Gattaca", for instance.   
> If you pronounce
> it as in the movie, with stress on the first syllable, the /t/ is
> pronounced [4].  Now let's say you wanted to make up a nonce
> word to describe the movie as "Gattacular!"  Now the /t/ is

The word "Gattacular" is interesting. I really am not sure how I  
should pronounce it; for one, there's a question of whether the first  
syllable should have [@] or [{]. If it's [@], I would use [t_h]; but  
if it's [}], I would say [t:_h] (a geminate aspirated /t/) or maybe  
[t_} t_h] (unreleased /t/ followed by aspirated /t/) (I suspect that  
the two are equivalent, but I'm not sure).

But on the other hand, yet a third pronunciation suggests itself to  
my mind, although it seems like my least favorite one. I'm not  sure  
how to describe or transcribe it; I think it either uses an  
unaspirated /t/, a flap (or whatever my equivalent of a flap is), or  
an unreleased /t/ (can you pronounce an unreleased stop right before  
a vowel?), but definitely not aspirated. It's actually the same sound  
as in the phrase "cat actor", with accent on "actor."

> pronounced [t_h].  This is pretty much the exact same environment
> as you describe, the only difference being where English has [4],
> you have [s].  Now, I know that I've actually seen a language
> where [t] and [s] are in complimentary distribution, but since
> I can't think it up, just think about the process.  What's happening
> is the main emphasis of the word is put before the /t/ (this is
> not a scientific description), so the rest becomes kind of less
> emphatic.  To make it flow more easily (and to make it more
> like the vowels around), the segment becomes less stop-like
> and more vowel-like, while still trying to retain its /t/-ness.  In
> English, it achieves this by keeping the place of articulation,
> and essentially shortening it until it becomes a tap, sacrificing
> voicing*.  In your language, you keep the place of articulation
> and the voicing and sacrifice the manner, making it a fricative.
> I'd think of both sound changes as versions of the same change.
> [*] There's now phonetic evidence that suggests, among other
> things, that there is a voicing difference between the /d/ tap
> and the /t/ tap.  Go fig.

Really? Do tell! If there is a voicing difference, I can't perceive  
it -- although my dialect has some Canadian raising tendencies, so it  
distinguishes "rider" from "writer" by the quality of the /ai/  

> -David
> *******************************************************************
> "A male love inevivi i'ala'i oku i ue pokulu'ume o heki a."
> "No eternal reward will forgive us now for wasting the dawn."
> -Jim Morrison