Unofficial alphabet lists for Lojban/Latin/English, Greek,

["Mark E. Shoulson" <cbmvax!uunet!cuvmb.cc.columbia.edu!shoulson>]

Cowan:
>I don't know.  Originally "i"/"j" and "u"/"v"/"w" were not distinct letters;
>now they are.  The 1976 Israel Standards Institute computer charset separates
>bet/vet, kaf/khaf, peh/feh, sin/shin.  This is probably reasonable.

I'm feeling sort of like reversing myself.  After all, the alternation for
all the letters like this (except shin/sin, which don't alternate) is
determinable by the grammar of the word.  Then again, that grammar is so
hairy four out of five Israelis would be wrong.  OK.  They deserve separate
lerfu.  The only one missing that significant chunks of population still
distinguish is taf/"saf" (orig. "thaf", but the Ashkenazim pronounce it
"s").  Very few people know that gimel and dalet also sound different
depending on the dot, or how to handle a dagesh in another letter (I
believe Yemenites are careful about this, though).

I'm torn on assigning some of these, though. for kaf/chaf, chet, and qof,
I'd recommend {ky./xy., xy.bu, ky.bu} respectively, on Zipfean and cognate
grounds (the qof is related to the Latin q, I'm told).  peh/feh is easy
because they don't overlap any other letter's sounds.  Bet/vet and vav are
tough, because I'd like bet and vet to be as similar as possible, and since
bet doesn't overlap anyone that'd indicate {by./vy.}, leaving {vy.bu} for
vav.  But vav is a more common letter.  I think {by.bu/vy.bu, vy.} is
better for bet/vet, vav.  Shin/sin, tzadi, and samekh, though, are the real
sticking points, since they step on each other so thoroughly, especially
since {*tsy.bu} is illegal.  Grumble.  I suppose {tsabu} is as well,
because of the cluster, even as {tsa.bu}.  Shin is the most common of
these, probably followed by tzadi, then sin, then samekh.  Not sure of the
positions of tzadi and sin, though.  To make life easier, I guess {cy./sy.}
for shin/sin, but that leaves samekh with something like {sy.bubu}, which
is icky.  bleah.  Maybe I can come up with something better.

Final forms are pretty obvious from context, and you can always break out
of lerfu mode to indicate {my. voi famno} or something.

>Okay.  The character set mentioned above has the following vowels, so they
>at least should be included:  hireq, sereh, segol, qubbus, qamas, pathah,
>shewa, hataph pathah, hataph segol, hataph qamas, holem, shureq.  Needless
>to say, these are just random character strings to me.  Can you map them
>into the Lojban vowel set in any reasonable fashion?

I was splitting overly-fine hair in my complaint about the vowels.  You
need words for the symbols; if you can't tell from the marks on the page
that a qamatz is katan or gadol, why should the lerfu distinguish?  Still
some fine-points left, though.  Let's see.  in order:

.ibu, .eibu, .ebu, .ubu, .abu(?), .a'abu(?), .ybu, [see below for
hataph-*], .obu, .u'ubu.

The trouble with .abu and .a'abu is that the vowels differ not in length
but in quality.  Modern pronunciation has them identical, but Ashkenazic
preserves the more back-sounding qamatz as, well, more back-sounding, while
patach is the true {.a} sound.  I see no problem with collapsing them
phoenitcally to the same sound in Lojban, but whatever we do I can't see
myself being able to remember which is which (having looked away from my
screen, even now I don't remember if I assigned .abu to patach or qamatz).
I almost want .aubu for qamatz, at least that's close to how we might spell
a similar sound in English (au as in Australia), but the Lojban
pronunciation is hopelessly confusing.

The hataph- series are sorta "hyphenated" vowels.  Technically they're
variants of the shewa (where do you think the English word comes from?),
variants of the "moving" (pronounced) shewa, to be precise.  They're
written as the vowel with a shewa to its right, both under the same letter.
Pronounced pretty much the same as a regular vowel, except shorter and
never stressed.  So I can see either listing two vowels under the letter,
or else using {.a'ybu, .e'ybu} and, ugh, {.a'a'y.bu} for those three.
Yuch.  {.a'a'y.bu} has to go.  'Sides, are things like {.a'y.bu} valid?
Didn't think so.  Maybe stick with listing both vowels.

~mark