Re: Artlang-driven English spelling reform

On Second Air of Tenderness, Nik replied to my posting:

> Carlos Thompson wrote:
> >   <ng> into eng, when pronounced /N/, left <ng>
when /Ng/.
>
> Make /Ng/ into eng-g.  /ng/ can exist in at least some
dialects
> in words like "engulf" (/IngVlf/ in my idiolect)
It was something I was thinking about... mainly if we
compare words like _language_ and _conlang_ which in my
foraign accent are /Ng/ and /N/ respectively, then ell-A-
eng-"guage" could become ell-A-eng in compounds, with no
<ng> -> eng conversion.

But.  My point was not in getting a fonetik spelling of
English, but reather in just replacing diagraphs with
extended/accented Latin letters (no IPA).  Then I decided
not to change all /N/-sounding <n>s into eng (most of the
cases predictable), but just the <ng> diagraph.  Well, I
could write test texts of both approaches.

> >   <gh> into yoth, when mude or /f/, <g> when /g/.
>
> The spelling's _yogh_.  And why use it, when dropping or
<f>
> would work?
Maybe because I wanted to use yogh for a _weak_ g.  Still I
don't know about using yogh only when <gh> is mude or in
both positions...

> > Now is time for vowels... volunteers?
>
> *evil grin*, Semetic-style - vowels are unmarked!  ;-)
Hmmgmgmmm...!

Let's see how English vowels are used:

           Stressed
 Ortho-  Usual  Usual
 graphy  Short  Long
  a       &      eI
  e       E      i:
  i       I      aI
  o       A     ow @w
  oo      U      u:
  u       V      ju:
  ai             eI
  ea      E      i:
  ee             i:
 eu ew           ju:
  ie      I      aI
  oi             oI
  ou            ow @w
  ow             Aw

My proposal:
  1) <y> will only be used as a vowel in word final.  All
other cases will be repalced by <i>.  (This rules applies
also for <y> in diagraphs: <ay>, <oy>).

  2) when monograph vowels are in usual* stressed position,
length is predicted, and use the above values, then are
written as above.

    * we have to define the usual stressed position (or
predicted stress position).  See bellow.

  3) when stress nor lenght cannot be predicted, a grave
<`> will mark the short stressed sound and a circumflex <^>
will mark the long one.

  4) unstressed monogrpahs predictively rendered
as /@/, /I/ or droped, according to dialect, will use the
usual orthography.

Now the problems are:
  what to do with diagrpahs?  I suggest leaving diagraphs
for diphthonges and long vowels, and replace for one symbol
when monophtonges.

  what to do with vowels using other values?  I propose
changing them.

  what to do with vocalic consonants like /n=/, /r=/
or /@:/, /l=/, etc.?  I will see if I find some interesting
extended/accented Latin letter, or leave all them as <en>,
<er>, <el>.  If <e> is strong before <n>, <r> or <l>, then
e grave <è> or e circumflex <ê> will be used.

Length:
 when unmarked by grave or circumflex.
 . Open syllables in stress positions use long vowel.
 . Yogh will make previous vowel long.
 . Vowels followed by a geminated consonant are short.
 . Words ending in VCe, if V is stressed is long.

Stress:
 when unmarked:
 . monosyllabes (including those ending in mude "e") are
stressed.
 . If there is a geminated consonant, previous vowel is
stressed (and short).
 . usual unstressed endings like -ing, are unstressed
unless marked.

Alternatives:
  circumflex could be repalced by macron.

-- Carlos Th