Re: sound change question

Yoon Ha Lee wrote:
>> > Thanks!  I feel better.  I keep looking at lists of common sound
changes
>> > trying to find ones that get me to the target phonology (I'm working
>> > bass-ackwards trying to construct a protolang that will result in
>> > something like the conlang I had in mind), but sometimes I can't find
>> > ones that I think might be plausible.  >>
Plausible shmausible.  Natlangs do some veery strange things.  Vaguely
remembered exs.:  in two Turkic languages, the word for 'stone' is 1. [tSaS]
and 2. [tul] (or something like) -- they are cognate.    In one Austronesian
lang. of northern Borneo, a secondary cluster (due to V deletion) of _bh_ >
/s/ IIRC .  Difficult to explain.  Somewhat easier to explain, at least in
generative terms:  Tettum (of Timor) clearly reflects AN *p as /h/ initially
and medially, but /s/ finally.  The development of *p can be traced > **f or
**P > h (and on to 0 in related langs); so at some point Tettum must have
developed a morpheme-structure constraint to the effect that "in final
position, the only permissible voiceless continuant is /s/".
Later on you wrote:

>what I'm struggling with is whether sound changes
come in herds or lone horses >

Both, of course.  It's useful to think in terms of natural classes:  all
stops -- all voiced/voiceless stops -- all nasals --  all
dentals/velars/whatever -- in many languages all stops plus /s/ can be
classed as "obstruents" versus e.g. nasals and resonants.  Changes can
affect an entire class by adding, deleting or changing one or more
features-- this can lead to interesting problems in rule ordering.
Similarly for the "lone horses".
I think it's helpful to learn some generative phonology (no doubt it's
changed from the model I learned and loved); it's very good at describing
historical sound change-- moreso IMHO than Optimality Theory (insofar as I
have that figured out!).

 >used in the creation myth for the phonological relay (what happened to
>that, BTW?), >
    Hmmm, I've been wondering too.........

The
>phonetics/phonology class that I probably will have to drop is actually
>pretty evil in that I keep being tempted to pull out my notebook and do
>conlang things instead of paying full attention, but since I never
>actually took the prereq (Linguistics 101) I'm scared I'll miss something
>crucial.  :-)
We would all agree, no doubt, 'twould be a pity to drop it, as you would
clearly enjoy & profit from it.....but you know your schedule/time
capabilities better than we.  If there were free assignments, papers etc.
you could always present "data from a hypothetical language" without getting
into specifics.  At the least, you could surely just sit in, I assume.
(One can learn a lot by just sitting in; frex, I sat in on Napoleon
Chagnon's course on South American Indians, from which I learned that I did
not want to devote my career to that area.  Chagnon, of Yanomami fame as you
may know, has been in the news lately-- his course  was, in fact,  very
interesting, as was he himself.  Frankly, a hunk (25 years ago), if I may
editorialize. But he wandered, slightly high, into Ann Arbor's only gay bar
one Sat. night and was on the verge of throwing punches before his
companions hustled him out the door.)