Re: Monster phonology in need of romanization help

On Tue, 10 Jun 2008 17:18:09 -0400, ROGER MILLS wrote:
>John Vertical wrote:
>>Without further ado�
>>
>>Initials:
>>/    t_> ts)_> tK)_> tS)_> k_> q_>  /
>>/p   t   ts)   tK)   tS)   k   q   ?/
>>/b_t d_t             dZ)_t g_t G\_t /
>>/b_v d_v                            /
>>/f   T   s     K     S     x   X   h/
>>/v_t D_t z_t   l_t   Z_t   G_t R_t  /
>>/v_v D_v z_v   l_v                  /
>>/m_0 n_0             j_0   w_0 R\_0 /
>>/m_t n_t       l~_t  j_t   w_t R\_t /
>>/m_v n_v       l~_v  j_v   w_v R\_v /
>>
>>_t is "brethy voice", _v is "tense voice"; this distinction has a variety
>>of realizations in dialects.
>Aargh! This IS complex. (First off, i keep thinking "_t" is  _tense_...)
Blame whoever designed X-SAMPA for that!


>>The palatoalveolars are only marginally contrastiv
>>with lhaterals & especially velars.
>
>That may be, but unless one or the other is extremely rare, or conditioned
>(even partially so), IMO you still need a way to distinguish them.
I'm trying to go for a Romance style system here, <ca ce> = /tSa tSe/ and so
on forth. Currently it seems the exceptions to the distribution are /a/ not
counting as front, and palatoalveolars occuring before shwa & sometimes back
vowels. But adding whatever I spell /j/ as should deal with that.

Alternately, I could limit this to just the fricativs & spell /tS dZ/ as
<tch djh> or therearound, since I'm spelling the other affricate series also
by components. That's also unambiguous wrt/ palatalization because there's
no /tx dG/!


>>Okay, onto the issue. Here's what I have for the consonants @TM:
>><t' ts'  tx'   c'  c' k'> (ejective stops/affric.)
>OK t' ts' _tl'_ c' _k' q'_
I've never really liked <tl> for /tK/, mainly because I'd like to have
distinct base letters for /l K/.


>><b d j g>
>bh dh jh gh  (breathy vd.stops/affric)
>
>><ƀ đ>
>b d    (tense vd stops)
This was my first idea, but after I saw how well dashes for the obstruent
series worked, I switched. That also leaves <h> for other functions.
Granted, if I spell /h/ as <h> this is the most logical digraph function…

Your another idea of spelling the tense series as geminate seems useful, at
least in ASCII environment.


>><ph th s  x    ch ch kh h>
>    f  th s   l-bar ç   x  qh  h  (vl.frics)
>><bh dh z l jh gh>
>    vh ðh  zh lh  çh gh  (brethy vd.frics)
Maybe I should clarify; the problem is not in thinking of *an* orthography,
but a *satisfying* one. What I don't like with this sort of "most common
phonetic spelling" system is the erratic usage of <h> sometimes as a
fricativizer, sometimes as a devoicer, sometimes as a brethizer (?). I don't
think thorn & edh really fit into the flavor I'm going for here, either. So
what does that leave for /T D/? I need <z> elsewhere, and apparently also at
least dee slash. Tee slash isn't too bad by itself but in context stands out
just as bad as <th>… unless I spell everything applicable with a h-digraph?

I also notice I did not explicitely specify that similar to <ch>, <jh> is
currently being used for both /Z G/. Somewhat unsatisfying, but I need
something left for the uvulars, too.


>hm  hn  hy  hw  hr (vl.nasals/resonants)
>mh  nh  nl  yh  wh  rh (breathy vd.nas/cont.)
>m  n  l   y  w  r  (tense vd. nas/res)
Not bad... I have not considered distinguishing prefix/postfix <h> before.

BTW, I was originally going for only 2 sets of sonorants, but decided to go
for three after all. This may still change when I examine how different
dialects realize the series & their diachronic origins. Say, if there's one
to have modal voice for the "tense" series, the voiceless/brethy distinction
may be a bit too much to also support. I also realize now my previous
phonological analogy p : bh : b :: hm : mh : m (if you get my drift) is
incorrect; a fully symmetric system would be rather along the lines of p' :
p :: b : bh :: m : mh.

Maybe a third set can still exist, but only as a lexical set similar to BATH
(cross-dialectally, but never phone*ically distinct from both of CALM and TRAP).


>Re vowels etc. I understand the need for no diacritics, since you're going
>to use them for tones-- but have you considered some other way of marking
>tone?  Perhaps: low XX-  rising XX/   high unmarked
>That would free up diacritics for the vowels
You mean using actual dashes and slashes? That would look more like a weird
syllable division than tone marks, IMHO.


>>Syllable structure allows only CV or initial syllabic N.
>Is the language monosyllabic???
No, roots are generally bisyllabic, and a word may have up to two prefixes
(with a highly limited phonetic form) plus a so far indeterminate number of
postfixes (with more freedom).

>Do you mean the initial syllabic N can occur before a consonant?
Not only it can, it must, because an onset consonant is compulsory for a
syllable.


> In that case (if not monosyllabic) how would you distinguish say
> sa.m.ba vs. sam.ba = [sa~.ba] ???
Well, the former doesn't occur; a syllabic nasal may ONLY occur as a
word-initial prefix (one might say they're morphologically conditioned
allophones of N@).


>You might use Vy and Vw for some of the diphthongs.
Always an option, sure.


>>Scheme 1: Straightforward
>>Monofthongs spell'd as per IPA, except /i\ A/ = <y o>. Difthongs spell'd by
>>their components, except a final /6/ as <a>, /ai\ ei\/ as <au eu> and /i@
>>u@/ as <ie uo>. Nasality by syllable-final <m> or <n>.
>
>Actually I like this. You might end up with "yy" as a word-- odd, but clear.
>>
>>Main problem here is not that it's boring, but that it clashes with my
>>current consonant scheme. I would have room to change <y u> to <u w>, but
>>that does not seem enticing.
>??????
I generally don't like <y> or anything else standing for both a vowel and a
consonant, especially not for ones that aren't allophonic.


>>Scheme 2: Digraphy
>>/i i\ u/ = <ei eu ou> -- but why digraphs for /i u/??? eu is good, but I
>>see I want to use it for a diphthong..........
Wellll, I just wanted to use <ou> = /u/ in some project, and I didn't really
have room for it elsewhere. After that <ei eu> just suggested themselves.


>>/e @ a A/ = <ea e a aa>  maybe ee e a aa ???
Well this is supposed to be a somewhat historical orthography; all former
long vowels difthongize so <ee> for /e/ won't do. <e> would be logical but
then I'd need the actual shwa for /@/.

Actually, do I? I need to examine if I could just use eg. <a>. Its
occurrence is rather conditioned by the foot/tone structure. Also, IIRC it
can only occur with lo tone, hence something like <ä> would also work. No,
checking my notes, that's just the [@] allophone, [6] may still receive
other tones.


>>/@i @u i@ u@/ = <ii uu i u>  --> ei eu ie ue
>
>>/ei ei\ ai ai\ Ai Au/ = <eai eau ai au aai aau> --> eei eeu ai aeu aai aau
>>/e6 i6 i\6 u6 o6/ = <ee ie ue uo oo> -->  eea ia eua ua oa  (your i\6=ue is
>>odd)
>>This isn't too confusing, is this?
>No worse than Dutch :-)))))
>I hope I haven't added to the confusion.... and now me brain hurts :-))))
No, thanks for the help. My brain tends to hurt too when thinking too much
about this, which has led to this project going in "delayed" status… %#*:7

John Vertical