Re: Math/Phonological formulae

Eric wrote:
<<
Isn't /l/ also a continuant ([+cont])?
 >>

Roger said "no", but I'm afraid the answer is controversial.  In the
traditional feature chart (believe it or not), /l/ is classified as [-
cont].
I think it's because technically there is a complete closure between
the tongue and the roof of the mouth (even though air is escaping
through the sides).  In languages, though, if you have /l/, and
you have to decide whether it patterns with the stops or the fricatives,
more often than not it will pattern with the stops.  That's another
reason why it may be classified as [-cont] (and this may change
for different languages).  I bet if I (or someone) looks up Ladefodged,
there should be an explanation of some kind.

Roger:
<<
Actually, assuming the language has only 1 lateral /l/, you could write:
[+voice] --> [-voice] / [+lateral ___]  +lateral implies that all its
other
defining features are present and needn't be repeated.
 >>

Isn't that cheating?  I thought you weren't allowed to define
anything about A or B in the C and D part.  Of course, that would
make a more interesting rule, if there were other segments that
got devoiced.  Then you could just have one general devoicing
rule with varying contexts.  That would make it seem like all
devoicing rules should go together, which may or may not be
true on a conceptual level for a given language, but it's interesting,
nevertheless.

Eric:
<<
Does [+-cons] mean consonantal? (If so, I don't see how it would
disambiguate different laterals.)
 >>

Well, you might have a vocalic lateral that, phonologically, *is*
a vowel.  (So, something like [for some dialects] the vowel in
"pull", but widespread.)

Eric:
<<
a. Why are some of these features in caps and some not?
b. Is [+-COR] coronal, [+-ant] anterior, and [+-DOR] dorsal?
c. Does / mean "or"?
d. What do the parentheses mean?
 >>

Regarding the rule: C[+COR, +ant/+DOR] > [(+COR) -ant] /_V[-back, +ATR]

This was my attempt to try to write my rule where velar and
alveolar consonants become palatal before a palatalizing vowel.
First, the vowels are [i] and [e], and they contrast with the non-
palatalizing vowels [1] and [E].  The vowels in Sidaan's vowel
inventory that are [-back, +ATR] are [i] and [e].  So that takes care
of that part.

The features for "coronal" and "dorsal" are in caps, because they're
not binary features.  Technically, I think they're not even supposed
to have +'s or -'s in front of them.  Rather, you simply state (in caps)
CORONAL, DORSAL, etc.  I shortened it, and added the plusses
for clarity/consistency.

The reason the parentheses are in this particular rule is that a +COR
consonant doesn't go to +COR; it just becomes -ant.  The +DOR
consonants become +COR.  Again, to make this maximally clear,
it should be stated in two separate rules, but then the generalization
would be lost.

Roger:
<<
There's another way to deal with palatalization of td/kg at least before
High front /i/-- it so happens that tdkg are also specified [+high],
so it's
a form of assimilation. (Not all phonologists use "high" ahem....but
other
features are available).
 >>

Indeed!  That would be a nice way to do it.  I knew there had to
be some technical way of showing that this kind of change is natural.

Eric:
<<
I suppose you could even say "X is a consonant that's alveolar,
palatal, or velar", which to me seems to better capture the "spirit"
of the change (by avoiding an unnatural separation between alveolar
and velar), but that would mean that it would posit a "change" where
the output equals the input in some cases.
 >>

Yes, but I left out the history.  In reality, there *are* no palatal
consonants in Sidaan: they all came from palatalized velars and
alveolars.  I'm struggling right now with how to deal with this,
because there are currently minimal pairs between palatals and
non-palatals (here I'm using the historical *):

*s1taha > staha
*sitaha (or *xitaha) > Ctaha
*x1taha > xtaha

So there's a rule like this (and I'm never going to be able to write
it, because it has to do with stress):

V[+high, -back] > ø / C_'CV...

There are specific constraints on what the C's can be, but essentially,
high vowels drop out in unstressed syllables between certain types
of consonants (among them, a fricative followed by a stop).  The
result is minimal pairs between velar/alveolar fricatives and palatal
fricatives.  The same goes for stops:

*t1laha > tlaha
*tilaha (or *kilaha) > claha
*k1laha > klaha

The orthography was developed at a time after this was common,
so a series of palatal consonants was developed.  So they have some
sort of psychological status, even if they're not full phonemes in
their own right (or phonemes with a restricted distribution).  The
question, then, is if (or when) speakers will be able to take examples
like that, and produce words like:

taha "X"
caha "Y"
kaha "Z"

And then, of course, I'll have to decide what point in the imagined
history of the language I'm going to be creating vocabulary.  There's
already plenty of words with disappearing high vowels, but I haven't
yet created any where the palatal consonants are full-fledged phonemes
that can appear everywhere other members of their class can appear.
Don't know how or when I'll decide.

For reference, this is a description of the phonology:

http://dedalvs.free.fr/sidaan/phonology.html

And a description of the orthography:

http://dedalvs.free.fr/sidaan/orthography.html

-David
*******************************************************************
"A male love inevivi i'ala'i oku i ue pokulu'ume o heki a."
"No eternal reward will forgive us now for wasting the dawn."

-Jim Morrison

http://dedalvs.free.fr/