Re: Sidestepping Spelling Reform

Well, English syllable structure can be as complex as /st4oUmpt/ (and
probably even more).

Trying to cover all possible sylables would be a monumental task, but
we can reduce still the number of symbols used.  For example we can
render /st4oUmpt/ as <to>_<s>^<r> + <up>^<n> + <(e)d>, six symbols
that are just a little better than "stromped".

The idea is to have both C1V1 and V2C2 syllables, with a different set
of consonants and vowels for C1V1 than for V2C2.  And there would be a
series of diacritic that would allow a greater range.

C1 are those consonants that can usually begin a syllable:
   m     n
 p/b   t/d tS/dZ k/g ?(0)
 f/v T/D s/z S/Z     h
 W/w   l  r    j
for a total of 25 phonemes...

C2 are those that usually end a syllable:

for a total of 24 phonemes, plus one null phoneme.

V1 would represent all possible short vowels:
  I E & a/Q V U @ (A)
for a total of 7 or 8 phonemes.

V2 would represent diphthongh/lenghtening vowels:
  I @ U
for a total of 3 phonemes plus a null phoneme.

Now we have 175 or 200 CV components, plus 99 VC components (excluding
null+null).

Some V1+V2 combinations might be:
  E+I = i: (eh+i = ee)
  &+I = eI (ah+i = ay)
  a+U = oU (oh+u = o)
  V+U = u: (uh+u = oo)
note that this might seem a little arbitrarious...

Now, the 175/200 CV can be affected by some diacritics:
  A presibilanting diacritic: s- (_s)
  A glide diacritic: -l- (^l), -r- (^r), -y- (^j) or -w- (^w).

Now note that applying the diacritics to some of the symbols make
little sense: ?E ^r ... however we can save some symbols this way: we
can use ?E ^r instead of rE.  This reduce the number of CV symbols
from 200 to 168 (or 175 to 144).

We can even add a voicing or devoicing diacritic., so we have now 96
or 84 CV symbols, plus 6 diacritics (for of them mutually exclusive):
 (?@ m@ n@ p@ t@ tS@ k@ f@ T@ s@ S@ h@) times (I E & @ (A) Q V U)
 _s +voiced ^r/^l/^w/^j   m     n         N
 p/b   t/d tS/dZ k/g
 f/v T/D s/z S/Z
   w   l  r    j

A symilar analysis might reduce the 99 VC symbols, with the following
diacritics: +voicing, +nasalization; this gives me 51 symbols:
  (I @ U null) times (null m n N p t tS k f T s S l)

Most syllables are then coposed by two base symbols (from a total of
96 onsets and 51 offsets), with 7 diacritics.

We may add a symbol for the past tense morpheme -ed: /d/ /t/ /Id/.

This looks more rational.

-- Carlos Th