Re: OT: What makes a good conlang? (was Re: Super OT: Re: CHAT : JRRT)

Merhaba!

Christophe wrote:

">I agree, these sound changes seem quite unreasonable. I would expect
>something like this instead:
[...]

">h -> ?/t/k/f
>x -> k_h -> k
"Actually, neither your sound changes nor Teoh's are reasonable. Unless
strong influence from the phonetic environment, phones usually don't go
"stronger". They tend rather to weaken (law of least effort). So stops will
naturally fricativise, but fricatives will only rarely become stops (why do
languages still have stops nowadays then you may ask. Simple: there are
things "stronger" than simple stops: aspirated stops, geminate stops - which
can arise from vowel deletion, also a natural sound change going in the way
of the least effort -, etc... Those will naturally weaken into simple stops,
and as I indicated for the geminates, can themselves occur naturally :) ).
Instead, they will tend to become approximants or even disappear completely.

"Also, sound changes rarely go further than one step in the chart at a time.
A change h -> t, besides being already unreasonable for having a fricative
turning into a stop, is even more unreasonable for having a *glottal*
fricative turning into an *alveolar* stop! How do you justify such a leap?!
:) (note that it's not impossible. French has an uvular fricative evolved
from an alveolar trill, through an uvular trill step. Still, that's less
distance than h -> t :) )

"The only way to have such "backwards" sound changes would be to justify
them through their environment, i.e. having a fricative becoming a stop
because of an influence from another sound. But even then it seems quite
difficult.

"Sounds always change for a reason, most often inertia (law of the least
effort). It will tend to assimilate neighbouring sounds (or even less
neighbouring ones, look at the German umlaut), erase the weakest ones
(unstressed vowels for instance), etc... It will *not* strengthen sounds. So
the presented sound changes are indeed unreasonable, at least until they are
given a proper environment."

All of this makes sense. I'm just bad at inventing and applying reasonable
sound-change rules! I have no idea where I got h -> t or x -> k_h!

A revision of my earlier thots:

h -> j
x -> h
And instead of h -> f, I think f -> h would be more likely -- I find /h/
easier to pronounce than /f/ (as in, takes less effort).

As an aside, which is more reasonable (or are they all possible)?
t -> l
t -> h
t -> ?

--Trebor