Richard, et alia:

Danvivatana (1987) states that in both the Sukhothai and King
Lithai scripts, there are both {mai eek} and {mai thoo}; the former
identical to the current form/shape, while the latter looks like the
current {mai cattawaa} as you noted. She notes that the shape of {mai thoo}
changes from the <+> in the Wat Asokaram inscription (1399 CE). These were
maintained in the King Narai script.
It is only until the King Rama VI script that {mai trii} and {mai
cattawaa} came into standardized use. Danvivatana notes that {mai trii} and
{mai cattawaa} were already in use during the Thonburi kingdom in the 18th
century.

There is no comment as to the historical origin/rationale for the
shapes of the tone marks. I would, however, submit that the shape of the
{mai thoo} may be cognate to that of the tone mark in Lanna, where one set
of tone marks contrasts </> (tone 1) versus <//> (tone 2). Given the
relative cursive stylization of many Thai glyphs and the subsequent
'simplification', it's not that far of a jump.
The introduction of {mai trii} and {mai cattawaa} into the script
seems to have come about during the years of 'script reform,' where
whichever King wanted the script to be able to reflect the spoken language.
As for {mai trii} and {mai cattawaa} being found primarily for
loan words... sure... but they're also used for native onomatopoeics or
sandhied forms, as well. Historically, the development of tone marking
would only require three unchecked (open syllable) classes -unmarked, 1,
2- and the fourth checked (obstruent codaed) class... It's only after the
evolution of the various tones' phonetic values that required additional
tone marks to allow for "non-natural" consonant/tone interaction.
(For those who don't know... Thai tonal patterns are dependent
upon historical consonantal values, i.e. voiced, unvoiced unaspirated
(glottalized and voiced continuants, as well), unvoiced aspirated (and
unvoiced continuants), and the tone mark associated with it. Of course
current users are unaware of the historical values of the glyphs they use,
but have memorized correspondences of this glyph and this tone make this
sound...)

If anything, it's interesting to note that there are many cases of
overlap in glyph form in Thai - {mai trii} looks like the digit "7"; the
vowel shortener looks like the digit "8"; the {karan}/virama descendant
looks like a sloppy digit "9" (but we know it's probably related to all of
the other virama descendants in the area); etc...

cheers,
-Patrick

At 05:58 AM 8/17/2004, "Richard Wordingham"
<richard.wordingham@...> wrote:

>I'm curious as to the history of Thai tone marks. Originally Thai has just
>two tone marks, a vertical bar (mai ek, literally 'mark one') and a cross
>(+) (mai tho, literally 'mark 2'). When were the other two, mai tri
>(literally 'mark 3') and mai chattawa (literally 'mark 4'), added?
>(Obviously they won't have been used until the orignal tones had split.)
>Mai tri seems chiefly to be used in English and Chinese loanwords. Why
>does mai tho now look like a European '2'? (It bears little obvious
>resemblance to the Thai digit '2'). It is notable that the present day mai
>chattawa is now a cross (+), like the original mai tho.
>
>Richard.