More thoughts on BrSc orthography & phonology

Going through the Lin material has caused me - as I knew it would - to
rethink some aspects of BrSc.  This is hardly surprising as both R.
Srikanth and I are aiming at compactness; in Srikanth's case it is the main
aim, in mine it has to be tempered with the other (strictly theoretical)
aim of a potential IAL.  I don't think anyone, including Srikanth, would
suggest that Lin could serve as an IAL (except for telepathic beings  :)

My two aims above are the same as Dutton's who invented Speedwords.  I have
added a third aim: that morphemes should be self-segregating (as, e.g., in
the Loglans).  Till recently, I had the following:
(a) all syllables to be simple, open syllables with no consonant clusters;
(b) only stressed vowels will be written; unstressed vowels to be
determined by a rule of vowel harmony.
(c) there will be two classes of morphemes:
   i. functional morphemes (clitics/affixes) written with a single letter.
  ii. lexical morphemes written CVC (pronounced [CVCv])

To get the maximum number of root morphemes I use 7 vowels, that is the
normal 5 plus {w} and {y} as Welsh does (but not with the Welsh phonemic
values).  This leaves 19 consonants, which gives me in theory
2527 possible lexical morphemes (19*7*19).

But there are some problems with this; to give the 19 consonants values
which will be acceptable phonemes in a global IAL is not easy (one would,
for example, not really want to keep /r/ and /l/ - and what to do with {q}
is always a problem!).  Without going into details - which have been argued
over here and elsewhere before - it seems that to get the best solution,
some consonants may be restricted in range (i.e. not used in connexion with
certain vowels).  The number of possible lexical morphemes is likely to be
closer to 2000 than 2500.

One wants to leave a few gaps in the lexical morpheme inventory to allow
for new ideas & concepts in the future.  It means a restricted list of
basic morphemes which must mean that some compounding is neeeded.
Speedwords, admittedly, uses a far more severely restricted inventory of
lexical morphemes - just under 500 - and must compound even more.  Some of
the compounds grow to greater length than the equivalent English word -
hardly desirable in a briefscript!

Also in the scheme I have outlined above, there is no place for two-letter
morphemes - which might be thought odd in a briefscipt!

I've wondered about 'polysemy' before even coming across Lin, but never in
my wildest imaginings had I considered enneasemy!  What I had considered
was using tone, in the manner of Chinese, to distinguish different words
which would otherwise be pronounced alike; but attractive tho it is (to
me), I have dismissed it as inappopriate for an IAL.

But, using the idea of vowel harmony, I am proposing a 'disemy' (two
meanings), e.g. {pt} would mean one thing with front vowels and another
with back vowels.

As well as musing with ideas sparked off by Lin, I've also recalled that:
At 1:40 pm -0700 1/4/02, Dirk Elzinga wrote:
>At 7:57 PM +0100 04/01/02, Raymond Brown wrote:
[snip]
>>Comments?
>
>Well, I still like my vowel harmony scheme, which would enable each
>consonant symbol to represent a syllable :-). But it doesn't seem to
>fit BrSc. That's the way it goes.
I've been intrigued by Dirk's scheme ever since he posted it in 2 years ago
to this very month, and I've gone back to it again.  It means I can keep
the sounds simple and have a very wide range of morphemes (and do not need
to use all 26 letters to do this).

I've dropped Dirk's high, central vowel [1] and kept only front (unrounded)
vowels and back (rounded) vowels.  I've also made one or two other small
modifications; the result is shown below:
( the braces {} enclose a _set_ of syllabic values which are shown
phonemically, but without slash delimiters):

p = {pi, pu}   b = {pE, pO}
t = {ti, tu}   d = {dE, dO}
k = {ki, ku}   g = {kE, kO}

f = {fi, fu}   v = {fE, fO}
s = {si, su}   z = {sE, sO}
x = {Si, Su}   j = {SE, SO}

w = {wi, wu}   o = {wE, wO}  (/wu/ might be pronounced [u])
l = {li, lu}   r = {lE, lO}
y = {ji, ju}   e = {jE, jO}  (/ji/ might be pronounced [i]

m = {mi, mu}   4 = {mE, mO}
9 = {ni, nu}   n = {nE, nO}

This set of 22 letters will allow 484 two-letter lexical morphemes, and
10648 three-letter morphemes - ample!

Not used are: a, c, h, i, q, u.
(I'm quite happy for {c} and {q} not to be used - troublesome letters, tho
very handy for denoting clicks  :)

I'm not over-enamored with the use of {9} and {4}, but don't know what else
to use.  I have toyed with the idea of {h} for instead of {9}, since the
lower-case form is similar to {n}, and the upper-case form is identical
with Cyrillic {H} = /n/.  But that's probably not acceptable for Roman {h};
and in any case it still doesn't help with a symbol for /mE/, /mO/.   Other
suggestions will be welcome :)

Affixes/clitics will be one-letter morphemes with unstressed vowels. In
two-letter & three-letter morphemes stress is on the penultimate vowel,
thus marking the morpheme boundary

If we assume that all affixes/clitics will be suffixes and/or enclitics,
then we need a marker to show the boundary between the lexical morpheme and
the functional morphemes suffixed to it. The marker can then show us
whether we have front vowels or back vowels in the root morpheme and its
suffixes.

We also need a marker to show that we have two lexical morphemes forming a
compound word.  Because the second morpheme will have the marker separating
the root morpheme from the suffixes, we probably need only show whether the
vowels of the first morpheme are front or back.

In Dutton's forerunner of Speedwords called International Symbolic Script,
he used the period/full-stop to separate morphemes.  I had thought of using
this as one of the boundary markers, but probably it's best to keep it for
a sentence delimiter.

At the moment (and this will probably change), I'm thinking:
(a) to separate lexical morpheme from the string of suffixes:
    front vowels:  {-}
    back vowels:   {'}
(b) to separate lexical morphemes in a compound:
    first morpheme has front vowels:   {\}
    first morpheme has back vowels:    {/}

But the percipient reader will see that I've more or less created sets of
three-character & four-character morphemes, rather than two-letter *
three-letter ones!  What would be more in keeing with a briefscript is if
in some way I could do what Srikanth managed to do with his "variable
characters" which he used for "cements", i.e. with a single symbol, show
the vowel qualities of two separate morpheme groups.

Also, I rather think I ought to be using {a}, {i} and {u} as 'cements' or
delimiters in some way.

Comments please.

Ray.

======================
   XRICTOC ANECTH
======================

Replies