|
|
pravda29@xxxxxxxxx wrote:
> ...
> >From the data it seems that there is a rule:
> -stops (plosives) are voiceless when inital consonants
> -otherwise stops are voiced
> -voiceless plosives may occur within a word, but only as a long
> consonant (or double letters)
The last rule contradicts the earlier ones, since the
doubled voiceless stops are neither voiced nor
initial.
A distribution like the one you describe with the
first rules (voiceless in a certain environment and
elsewhere voiced) would be found in a system
with all voiced stop phonemes which have voiceless
allophones (in the certain environment). However,
though phoneme systems with only voiceless
obstruents are common, phoneme systems with
only voiced obstruents are almost unknown. So
there's something wrong here.
>
> To me, this means that each plosive set is in complementary
> distribution and therefore they are allophones (p and b; t and d; k and
> g; etc).
Therefore what are allophones of what? You're not
being clear about just what analysis you're
considering. What are the phonemes?
Greg
|
|