The Four Vowels script.txt

I thought that as I have let the cat out of the bag I might
as well post my notes on the Four Vowels Script in whatever
scattered state they are.

Though this might be unclearly stated below the Four Vowels
Script script works similarly to Canadian Syllabics
<http://en.wikipedia.org/wiki/Canadian_syllabics> except
that there are (in the attested usage of the script) no
special signs for syllable-final consonants. Instead one of
the vowel series can also stand for a consonant with no
following vowel, like in Ethiopic script
<http://en.wikipedia.org/wiki/Ge%27ez_alphabet>.

== The Four Vowels Script ==

In the period when Sohloçan philologists were deciphering
the Kijeb script <http://wiki.frath.net/Kijeb_writing> a
Kidiliçan named Viçpirnijg (Mountain Moon) found that some
inscriptions in Kidil city were in a visually similar, and
clearly derived, but systemically different script.

The theory of Viçpirnijg was that the Four Vowels Script was
created as a conscious reform of the Kijeb script (which he
called the Three Vowels Script) at a time when most fully
unstressed vowels had merged as a schwa vowel which was
probably already in free variation with zero. He called this
stage of the language Old Kidilib.

Notation: _._ -- Kijeb phonemes, later Sohlob graphemes.
/./ -- Old Kidilib phonemes in CXS {.} -- Four Vowels
Script graphemes

N.B. Old Kidilib had no /V/ or /C/ phonemes, only /@/ and
      /s\/. The signs V and C are used for 'any vowel' and
      'any consonant'.

By convention Four Vowels Script stop series are transcribed
with the voiced series {b d z g}. In fact the same signs
stood for voiceless stops, voiced stops and voiced
fricatives.

The reformer had perhaps not realized that the merger or
loss of the unstressed vowels meant that formerly allophonic
variations in the stressed vowels had become distinctive, so
that not even four vowels were enough to write the language
clearly any more. It should however be born in mind that the
Kijeb script had left some phonological distinctions
unexpressed, and apparently consciously so in an effort to
reduce the number of graphemes in the script. The Four Vowel
Script clearly continued this tradition, reducing the number
of basic signs drastically by introducing rotation of the
basic signs as a systematic means of graphemic
differentiation. Indeed graphemic economy seems to have been
a more important goal than phonemic accuracy in all Sohlob
writing systems prior to the philologists' Pointed Script.

Viçpirnijg concluded that originally each of the nine vowel
phonemes of Late Kijeb shared a series of {CV} signs with
the two other phonemes derived from allophones of the same
Old Kijeb vowel, while the fourth alternant from each basic
sign designated a {C} followed by the undifferentiated
unstressed schwa or no {V} at all. This system was upset
when one of the former _i_ allophones came to be identified
with the unstressed schwa merger vowel and written with the
fourth sign orientation. This merger of /e/ and /@/ was in
fact reflected in all extant Old Kidilib inscriptions. The
first of the putative eight stressed vowels of Late Kijeb to
merge were /y U I/: they are never distinguished from one
another in Old Kidilib inscriptions, although the actual
spelling hesitates between A and E.

The Four Vowels Script distinguished the three inherited
Kijeb vowels {a i u} but had an additional series of signs
used in the following contexts:

1. where Kijeb had a word- or syllable-final consonant,
2. where Kijeb had an unstressed vowel which was lost in the
    later language,
3. where Kijeb had an _i_ followed by an _a_ in the next
    syllable. Such an _i_ gave /e/ in the later language,
    i.e. /i\/ in Heleb and Classical Sohlob and /E/ in
    Kidilib. It is assumed to have been phonetically [e] in
    Kijeb and [@] in Old Kidilib.

 From this pattern it was deduced that in Old Kidilib the
Kijeb unstressed vowels had merged as an /@/ which may be
optionally silent, and at the same time stressed a-affected
_i_ had an identical or very similar sound.

In three Old Kidilib inscriptions deemed to be older than
the others these are the only situations where the {e}
series of signs are used, with one notable exception: the
{ge} and {xe} signs were used for velars followed by /i\/ --
i.e. u-affected _i_ and i-affected _u_. In all other
positions /i\/ was written with the {i} series of signs, as
indeed all vowels that had not become [@] were written with
the signs corresponding to their etymologically ancestral
Kijeb phoneme, except that /Q/, the u-affected _a_ was
written with the {u} series before a nasal. The {gi, xi}
signs are used only for what in Old Kidilib was probably
cardinal palatal phonemes.

This brings us to the spellings of consonants. The main
differences are that there are four separate series for what
in the later languages are fricative consonants, denoted {F,
Z, X, H}. The {z} series is indeed used for all such
instances of Kijeb _s_ that would not have become /h/, but
also for the reflexes of Kijeb _ty_ and _dy_, i.e. the /s/
and /z/ of the later languages. Other palatalized consonants
are written as in late Kijeb, i.e. as a {Ci} syllable
followed by a vowel- initial syllable; notably /lV/ was
still distinguished from /rV/ as {riV}.

Somewhat surprisingly this neat orthographical system was
upset in the later Old Kidilib inscriptions. Not only was
there increasing hesitation about how to represent the
actual eight-vowel system with only four series of vowel
signs, so that i-affected _a_ /&/ was increasingly written
with {e}, probably reflecting the actual merger of /e/
(whether it was realized as [e], [E] or [@]) and /&/ in
Kidilib, while /Q/ and /o/, which had apparently also merged
could be written whith either {u} or {a} apparently without
rule, except in as far as writers tended to avoid confusion
with other words with /a/ or /u\/ which would otherwise be
spelled alike. More significant is the fact that the {zi},
{gi} and {xi} signs became specialized to write the dental
and palatal affricates and /s\/, while /tsi\/ and /ts\i\/
etc. which still existed were written with {zei} and {gei}.
Conversely former _kya_ etc. were written {gie} etc. Also
hesitation set in over the voiced fricatives: /v/ was
written with the {w} series, with which it may already have
merged, /D/ was written with the {z} series in spite of its
later merger with /d/ in Kidilib (it merged with /z/ in
Classical Sohlob and Heleb), and /G/ was written with the
{h} series or the signs for vowels without a preceding
connsonant.

The word for 'king' is significant, since it occurs in two
of the older inscriptions, where it was spelled {giure}, and
in several of the younger, where it was spelled {ziure}.
This word BTW was long an etymological puzzle, since it was
never spelled out and the forms -- Kidilib _dir_, Classical
_gifer_, Heleb _güür_ and Linjeb _zyl_ pointed to different
ancestral forms.

--


/BP 8^)>
--
Benct Philip Jonsson -- melroch at melroch dot se

    a shprakh iz a dialekt mit an armey un flot

                                 (Max Weinreich)