On Fri, 21 Feb 2003 23:18:10 -0600, "Patrick C. Ryan"
<
proto-language@...> wrote:
>In English and French, a voiced stop either initial or medial usually has voicing bracketing the closure: V-Cl-V-vowel, etc. This is here termed "full" voicing.
>
>A final voiced stop in English has the second voicing only in emphatic speech. Normally, "cad" is [k-h-a-V-t]. Thus, in final position has "partial" voicing. Similarly, the final voiceless stop is unaspirated unless
>
>emphatic, "cat" [k-h-a-t]. The sequence [V-Stop(-V)] is what we interpret as a 'voiced' stop.
Fine, as long as it's understood that we're moving away from acoustic
phonetics to the phonological interpretation of what we see in the
sounswaves and spectrograms.
I don't think there's justification for voicing _before_ the consonant
in initial poition (like _after_ it in final position). In your
notation: "bat" p-V-a-t, "bad" p-V-a-V-t.
Considering that V is also the symbol for "vowel", and that vowels are
the prototypical voiced sounds, the notation would seem to be
redundant. The marked sounds in English are the aspirated plosives,
which are characterized by voicelessness (/h/) intruding into a
following vowel. So an alternative notation, using H (voicelessness)
as the marked feature would be:
bad p-a-t (no pun intended)
cat k-H-a-H-t
bat p-a-H-t
tad t-H-a-d
Or perhaps even better (marking prevocalic voicelesness, and
postvocalic voicedness [as /:/, the length mark]):
bad p-a-:-t
cat k-H-a-t
bat p-a-t
tad t-H-a-:-d
=======================
Miguel Carrasquer Vidal
mcv@...