Re: Plan B phonology (was Re: Another weird idea!)

Hallo!

R A Brown wrote:

> Jörg Rhiemeier wrote:
>
> >Hallo!
> >
> >R A Brown wrote:
> >
> > [phonetic values of Plan-B letters]
> >
> >Of course, it has its own "logic", in extracting, for most of the
> >letters, both values from the same English words.  The result are
> >vowel/consonant pairings that are entirely arbitrary, and the vowel
> >inventory being patently English.  Bugger.
> >
> >
> Exactly so. And the fact that we have |f| = fOUGHt or Fought, but no
> example of just plain 'short o' in the inventory, leads me to think it
> is patently Merkan English  :)
Yes.  Shows just how phonologically naïve the whole thing is.
But there are some neat ideas in it, e. g. applying Huffman coding
to morphemes such that one can always tell how long the morpheme is
by looking at its initial phonemes.  But the mapping of phonemes
to phones stinks - but then, it is not really our business to
complain (see the "Campaign for rational Klingon romanisation"
thread).  It is much more of our business to do better.

> >>      One might find using
> >>what are clearly not alphabetic symbols, i.e. hex digits, more
> >>acceptable for their dual vowel/ consonant function.
> >>
> >>
> >
> >Yes.  A better solution for the Plan B phonology would be to have
> >16 consonants that are always pronounced as consonants, and a rule
> >that inserts epenthetic vowels to make the whole thing pronouncable
> >
> >
> Exactly the same thought occurred to me yesterday! I began playing
> around with a system whereby the epenthetic vowel was generated by the
> least significant bit of the first consonant & the most significant bit
> of the second one.  Umm - think I'll work on that one   :-)
That's an interesting idea!  One can do so much more than my boring
epenthetic [a]s.

Greetings,

Jörg.

Reply