Re: II Silindion Returns! (Longish)

En réponse à Elliott Lash <erelion12@...>:

>
> Proto-Nestean Consonant Inventory
> Regular:  Aspirate:  Palatal:  Labial:
> t  d       th  dh     tj dj    tw  dw
> p  b       ph  bh     pj bj    pw  bw
> k  g       kh  gh     kj gj    kw  gw
>
> Glottal:   Pre-Nasal:   Fricative:    Lateral:
> t?         nt  nd        s             l
> p?         mp  mb        S (<shut>)    lj
> k?         Nk  Ng                      l?
>
> Nasal:    Semi-Vowel:
> n          j
> nj         w
> m
> N
>
> Question 1: can one have voiced glottalized
> obstruents?
>
Difficult. Maybe as voiced consonant+glottal stop clusters?

> ------------------------
> Western Developments:
>          Initial   Final (Medial):
> Regular
> t  d       t  D      d  D
> p  b       p  B      b  B
> k  g       k  G      g  G
>
Why the (Medial) in parentheses? Also, I find the change of final voiceless
stops into final voiced stops quite strange. In general, final stops tend to be
*devoiced* rather than voiced. I'd rather expect something like : voiceless
stops stay voiceless, except medially where they are voiced, and voiced stops
stay voiced in initial position (or become fricatives), become fricatives in
medial position and stay as they are or are devoiced in final position.

> (the voiced obstruents become fricativized)
>
> Asipirate:
> th  dh    th  dh     t  dh
> ph  bh    ph  bh     p  bh
> kh  gh    kh  h      k  h
>
I'd expect all the voiced aspirates to become fricatives like gh did, maybe
with devoicing, which would mean dh > T and bh > p\. That would introduce
voiceless equivalents to some of the voiced fricatives you showed above and
would put more balance in the system.

> Palatal:
> tj  dj    tj  dj     tj  d
> pj  bj    tj  bj     tj  b
> kj  gj    tj  Nj     tj  g
>
OK, I agree on that.

> Labial:
> tw  dw    pw  nw     dw  nw
> pw  bw    pw  w      bw  w
> kw  gw    kw  w      gw  w
>
Not sure about those...

> Glottal:
> t?          d         d
> p?          b         b
> k?          g         g
>
Agreed.

> Pre-Nasal:
> nt  nd      t    nd     n   n
> mp  mb      p    b      m   m
> Nk  Ng      Nk   Ng     Nk  Ng
>
Not sure about it here. You prenasals seem to behave a bit independently from
each other, which is quite unlikely. As I already said, sound changes prefer to
apply to groups of sounds rather than differently to each sound. Also, in final
position I'd expect Ng and Nk to become N if all the others lose the stop too.

>
> Initially:      Finally:
> d > l (or d)     d
When does it become l and when does it stay d?

> b > b            b
> g > g            g
>
> Everywhere:
> dh > d
> bh > bh
> h  > -
>
I doubt such a rare type as a voiced aspirate would stay only for *one*
consonant. Voiced aspirates are rare, and no system is known to have voiced
aspirates without the voiceless ones (PIE is an exception but since it's a
reconstruction it doesn't count as we don't know if the reconstructed voiced
aspirates were indeed voiced aspirates or something else), and even less to
have a single voiced aspirate. If you have dh > d, automatically you'd have bh
> b
> tj > tj initially
> tj > s  finally
>
I guess it would do that through a chain tj > tS > ts > s...

>
> Question 2: How does one right voiceless _r_ in Sampa?
>
[r_0].

>
> That's the end of the major developments, comments
> appreciated.
>
Well, you have mine :)) . I have mostly the same critique as I had with the
evolution of the vowel system: each sound is much too independent from the
others. Sounds normally evolve in groups, and separate from a group they belong
to only because they are felt to belong to another group and follow this new
group along. You should try to work sound changes as changes in *features* (you
know, the representation [+voiced] [-continuant] etc...) applying to sounds
showing some features rather than as changes of individual sounds. It would
give more plausibility to your reconstructions and sound changes.

Christophe.

http://rainbow.conlang.free.fr

It takes a straight mind to create a twisted conlang.

Reply