Re: DISC: phonological system of Adain

At 03:09 PM 3/14/2002 -0700, you wrote:
>Thanks to Aidan for letting me put my resolution to the test.
   You're welcome! And thanks for the comments, and allowing me to put the
recent discussion to the test.


>>Basic phonology:
>
>You mean "inventory", right?
    Yes.


>>Length:
>>    vowels (both simple and diphthong) are long or short. The distinction
>>is represented orthographically by geminate consonants or clusters
>>following a short vowel in stressed syllables. Unstressed syllables are
>>always short.
>
>Okay. You realize that this orthographic decision will leave you
>without a clear way of representing long consonants (you mention the
>possibility of long consonants later on).
    Maybe it should be said that length is always present in the stressed
syllable, either on the vowel, or on the consonant, which is thereby
represented in orthography as geminate. Clusters are considered long, which
forces a short length on any given vowel or diphthong.

    son /sown/  vs. sonn /son:/
    cauba /kawb@/  vs.  caubra /kCbr@/


>>          Short     Long
>>a        @          a
>>e        E           e
>>i         I             i
>>o        o            ow
>>u        U           u
>>au      C           aw
>>eu      U           ew
>>ai       E            aj
>>ei       I             ej
>>oe      U            oj
>>ae      &            /aj/ or /ej/ (dialectal variation)
>>ao      E            ew
>>y        @           @ (or wedge /^/ ?)
>>
>>A couple minimal pairs:
>>  son /sown/  vs. sonn /son/
>>  taov /tewv/  vs. taovra /tevr@/
>
>So long /ew/ is matched with both /U/ and /E/ orthographically?
    Yes. The diphthong _eu_ arose through natural processes, whereas _ao_
arose later to deal with the development of the length rule. Originally,
_ao_ represented something more like /aw/, but dissimilational processes
and vowel drift ended up turning it into a fronter diphthong.

>I also wonder why you choose to use <ao> for short /E/ and long /ew/ ...
   The development of /ew/ is described above. Hopefully it makes sense. As
to /E/, it seems to me to be the unstressed variant of /ew/, perhaps by way
of o-umlaut.  Does that explain the back vowel used to represent front
diphthongs/vowels?

>>here's the part I'd like the most comments on (is it realistic? Do I make
>>sense? anything else?)
>
>It seems that there might be some extra-linguistic history behind
>some of the orthographic decisions you've made. Fill us in.
    Well, there may be, but mostly it is what _looks right_ to me, so
unless it's critiqued as being confusing, or unrealistic, etc., I don't
really know what to say about extra-linguistic history. The variation
between f and ph, v and bh, for example, really depends on what looks right
to me. I _think_ that it is f or v, with ph or bh only appearing at the end
of a word (because I really don't like the look of either of them in final
position). _oe_ is the same - I just don't like how _oi_ looks. In cases
like this, I usually create something else ad hoc to explain why.

    For example, the "mythical" transliterator of the alphabet was a woman
named Har Sceillen, and she set things up the way they are. This means that
I will use her whenever handy to say "because Har did it that way!" ;)

    Or, I can explain orthographic choices in terms of the original sounds.
_oe_, for example, was much less of a j-glide diphthong at one point, and
more like a w-glide diphthong, something like /wI/, only with a very much
weaker /w/. Maybe something like a frontish schwa that was eventually
strengthened.

>>Words normally retain the characteristic length of their stressed vowel
>>throughout declension or conjugation. In the case of a long vowel or
>>diphthong preceding a cluster (due to derivational or inflectional
>>processes), this should result in a short vowel (and sometimes it does.
>>However, in general, the long vowel transforms into a diphthong, preserving
>>length visually if not entirely phonemically. For example, the verb
>>_ser_  /ser/ 'flow' becomes _saerra_  /s&rr@/ in the future conjunct (in
>>-ra). This transformation turns front vowels and j-diphthongs (except oe)
>>into ae and all others into ao.
>
>I think I understand what you are doing, but I don't understand why.
>Why is it important that syllables preserve the appearance of length
>if they are not in fact long?Is this another extra-linguistic
>orthographical decision?
   Yes, I think this is an extra-linguistic feature. Again, looks are
important. Let's see if I can come up with a plausible reason for it.
   Perhaps the a- of the ae or ao was originally just there to indicate
length in a situation where length should not exist. Over time, the a- was
incorporated and a diphthong was created, and the short and long
pronunciations developed by analogy with the already extant diphthongs.
This would make these "long long vowels" arise through a complicated
development of orthography and pronunciation, which I think would imply
that writing kind of hit all at once and caused changes on its own. A
preliterate society that was suddenly taught writing, which everyone liked,
and learned to do fairly quickly.
    Oh, I should mention that the ae and ao also arise due to compensatory
lengthening (usually a result of disallowed clusters like -nthr-, where the
n induced nasalization, which became length, which was written with that
extra a-, which became pronounced, and so on..).

>I'd also like to see more examples of the
>"transformations" you mention.
     I'll try to show some long versions of ae/ao as well..

     ura 'water' + -na 'stative verb affix' > urna (by syncope), but -rn-
in disallowed.
         the r was historically always a rhotic, and in disallowed clusters
tended to cause length
         thus the u > aona. Why ao and not au - no one knows (unless you
have a good idea!).
         maybe there was once an au and an ao, but ao took precedence to
disambiguate from the
         standard diphthong, and ai followed suit. or something.
         > aona 'be liquid'
     ser 'flow' > saerra 'about to flow, going to flow' (fut conj)
     vir 'hurt, ache' > vaerra 'going to hurt'
     dom 'eat, chew' > daomra 'going to eat'
     lur 'play a wind instrument, blow' > laorra 'going to play'

     Note that these future conjunct forms can act as stative verbs in
their own right:

    Laorrane              taela    don    paunden         o.
    play:fut-IMPF-3sg all_day when stop-IMPF-1sg 3sg-OS
    He was about to play all day when I stopped him.

verbs with long diphthongs would just revert to short pronunciations. You
know, know that I think about it, its starting to sound like I shouldn't
call this length, but just a function of stress. Excewpt how do I explain
unstressed vowels in stressed syllables (like sonn)? Perhaps this is why
syllables can only end in continuants, because only they can be long? In
stressed syllables, either the vowel or the cons. is long/stressed:

    seph /se(:)f/  vs. seff or sepph /sEf:/
    nar /na(:)r/  vs. narr /n^r:/

   Does this sound like a better explanation of what's going on?

>So if a syllable ending in a consonant precedes a syllable beginning
>with a cluster, won't this produce a cluster of three phonemes? If
>not, what are the rules of cluster simplification?
    It would, technically, produce 3 phonemes, but they're not allowed, so
the cluster changes.

   clusters beginning with n, l, or r > drop the continuant, replace the
vowel with comp. lengthened (or whatever i should call it) ae or ao
    see ura-na > aona above.

   clusters beginning with m, p, b, f, or v > create a w-diphthong. If the
vowel is already a diphthong, a new diphthong is created modeled on the
short pronunciation of the original diph.
    ham 'sit' + thrann 'branch' > hauthran 'perch' (nn>n because not immed.
foll. the stressed syll).
    taum 'occasion' + thrann > (touthran >) tauthran 'puberty'
        au short pron. is /C/, which is enough like o or a that either
could result. However, ou always
        became au, and the pattern is still productive.

    clusters with d, t, g, c,  or their homorganic fricatives > create a
j-diphthong.
     ledh 'separate' (stem led-) > lei- before a cluster-initial word in
compound.

    clusters with s before a stop+cons cause stop > fric.  -str- > -thr-
    clusters with s or h before a fric+cons > s drops, fric is devoiced if
appropriate. -svl- > -fl-
    clusters with h before stop+cons > drop h, devoice stop if appropriate

    I think that's everything for clusters.

    i'm about to address your other post on my euphonic rule...

    Aidan

Reply