(This posting is best viewed in UTF-8.)

I have been having a discussion at
http://www.thailandguidebook.com/cgi-bin/forum/ikonboard.cgi?act=ST;f=3;t=6817;st=10;r=1
about whether Thai sara a (ะ, U+0E30) is effectively a consonant, albeit one
whose sound (the glottal stop) is often dropped in speech. As it the
discussion has become a duologue between me and 'Alone', I would appreciate
some outside input on the issues it has raised, especially about other
writing systems. This is essentially a cross-posting of posting no. 16 in
that thread. I gave two reasons, A and B. I would like some help with
answering the response to Reason A, but do feel free to make any on-topic
comments on the rest of the discussion.

Richard:
A. Sara a represents a consonant sound, it doesn't combine vertically with
other vowels,
[Whoops! I should have said 'with consonants', not 'with other vowels']

Alone:
Could you give me an example of any languages which have this sound and
consider it as a consonant?

Richard:
Doesn't Thai consider the sound a consonant (อ) word-initially? If you ask
the Thai Transcription Service (http://www.arts.chula.ac.th/~ling/tts/) to
transcribe อ๊ะ, and select IPA mode, you'll get a glottal stop at start and
end - /?á?/.

Anyway, here are a few other examples:

1. Hamza in Arabic.

2. Ayin (initial and final) in Persian. The sound is also represented by
initial alif. A striking example is rob' 'quarter', where the glottal stop
occurs word-finally after a consonant. (Ayin represents a different sound
in Arabic.)

3. Khmer? Not only is it present by the same rules as in Thai (with
yukaleakpintu, not reahmuk), but is also the sound represented by final 'k'
after /a/-like vowels.

4. Tagalog? It isn't expressed in the spelling, but it's there all the
same, both syllable-initially and syllable-finally. I think all Tagalog
syllables begin with consonants.

5. Malay. Final 'k' in native words represents a glottal stop. In educated
use at the very least, this contrasts with final /k/ (also spelt 'k') in
European loan words. I'm not sure about Arabic loans.

[Can Qalamites suggest any good examples? Or demolish any of my examples?]

Finding clear examples is complicated because Latin, Greek, Sanskrit and
Pali don't have glottal stops, and it seems that the original dialect of the
Koran didn't have them after vowels. The Arabic alif is therefore rather
like Thai อ. Classical Arabic is based on more conservative dialects.

The spelling strongly indicates that glottal stops (written with the
consonant aleph) were full-blown consonants in the earliest stages of Hebrew
and Aramaic, but by the time the vowels were recorded, glottal stops after
vowels (but not between vowels) had been absorbed into such vowels,
lengthening them. That's how Latin 'a' derives from the letter for a
glottal stop, aleph.

There is a Devanagari glottal stop for the Limbu language of Nepal, but you
might argue that this is under the influence of Western-based linguists.
It's in the Devanagari part of Unicode since Version 4.1, which came out a
few weeks ago. It looks as though it occurs word-finally, but I'll have to
check. [The proposal for its inclusion was
http://std.dkuug.dk/jtc1/sc2/wg2/docs/n2543.pdf , and one of the contacts
named there is a member of this group.]

Richard:
B. In dictionary order sara a is placed as though it were the last
consonant.

Alone:
I don't think so. In dictionary order it's placed as a vowel as I can say.

Richard:
Can you give me an example of whether treating it as a consonant or as a
vowel would make a difference to alphabetical order? I've convinced myself
that there can't be any.

Alone:
For nikkhahit I can accept that it can be a consonant in P/S , but not in
Thai.

Richard:
Interesting. Can you give an example of an argument that works for Thai but
not for P/S?