Re[2]: [tied] More nonsense: Is English /d/ truely voiced?

From: Brian M. Scott
Message: 19226
Date: 2003-02-25

At 4:02:07 PM on Monday, February 24, 2003, Patrick C. Ryan
wrote:

>>>A voiceless /d/ is one where the voice does not occur
>>>during the entire time it takes to implement the phoneme.

>> And a voiced /d/ is one where the voice occurs during the
>> entire time it takes to implement the phoneme.

> First off, what language has a "voiceless /d/" and how
> would it contrast with a "voiceless /t/"?

Icelandic. Stefán Einarsson, _Icelandic_, p. 13:

_d_ is always voiceless. Initially it sounds like English
_d_ [d]: _dagur_ [da:qYr.] day. Medially and finally it
sounds more like _t_ (French _t_) and is marked [d.].

E.g., <døkk> 'dark' (nom.sing.fem.) [tøhk]. Icelandic /t/
has several allophones: [tH] initially, in /VtV/, /Vtj/,
/Vtr/, and /Vtv/, medially after voiced consonants, and
finally after a vowel or voiced consonant; [t] after a
voiceless fricative, liquid, or nasal and often after long
vowels. So far as I can discover, this last is
indistinguishable from /d/ after a long vowel, though I
believe that /V:d/ occurs only in compounds (e.g., <hádegi>
'noon, midday').

Brian