Kevin:

>Can anyone tell me which pronunciation is closer to pure Pali, Thai or
>Sanskrit?

If the phonetic description given in Kaccaayana's grammar may be accepted
as "pure Pali", then in the Thai pronunciation, the vowels are roughly
correct, but 19 of the 33 consonants are mispronounced. I suspect the
figure is similar in Laos, Cambodia and Burma to judge from listening to
bhikkhus from these countries.

But if an Indian Sanskrit scholar were to read Pali in Devanagari script,
then his pronunciation would probably match closely the description given
in the old grammars. Likewise, when Pali is read by a Sinhalese or by an
Indian brought up to speak a Prakrit-derived language such as Hindi or
Punjabi, there are usually only one or two letters mispronounced.

The reason for the Thai mispronunciation is that in the languages of SE
Asia many Pali sounds, especially consonants, simply don't exist. In bygone
centuries alphabets were devised with special letters to indicate these
sounds. With the passing of time the special letters came to be popularly
pronounced in the same way as some existing consonant in that language,
rather than in the Indian way. Knowledge of their proper pronunciation
was probably confined to scholars.

Much the same thing has happened with English-speaking students of Pali.
Though diacritics have been devised to indicate the Pali sounds clearly in
roman script, not many of us actually take the trouble to aspirate bh, dh
and gh, or to distinguish between the dental ta, tha, da, dha, na and the
retroflex .ta, .tha, .da, .dha, .na, at least not when using Pali words in
conversation or when giving lectures. In casual use one tends to revert to
the phonetic patterns of one's mother tongue. Take for example the three
sibilants in "san.ghaadisesa". According to Kaccaayana's rules they should
all be pronounced as unvoiced dentals, but most British and American people
would naturally tend to turn the second S into a voiced alveolar (like the
S's in 'disease') and the first and third S's into voiceless alveolars
(like the S's in 'sesame)'.

Thankfully bodhi doesn't depend on correct pronunciation of Pali.

>And how much does knowing Pali actually help in learning Thai or Sanskrit?

It will help somewhat with Sanskrit but not very much with Thai. A knowledge
of Pali would be handy for reading the Thai translation of the Tipitaka and
other high literature, as well as legal texts, royal edicts etc. But in
everyday spoken Thai, though plenty of Pali words are used, many are
pronounced in a way that makes them barely recognizable to a Pali scholar
who doesn't already know Thai. After cutting off the final vowel (as
mentioned by Piya) Thais then make radical changes to the pronunciation
of some of the final consonants: 'L' and 'R' become 'N'; 'S' and most of
the palatal consonants become 'T'. Pali consonant clusters that don't exist
in Thai are either segmented by inserting the vowels 'A' or 'O', or else
one of the consonants is simply ignored.

Also, the primary meanings of such imported words very often derive from
their usage in legal, political or brahminical ceremonial contexts, rather
than Buddhist ones. As a result these are often only tenuously related to
their primary meanings in Pali.

Take the 5 khandhaa for example:

ruupa (Thai: roop): picture, shape
vedanaa (wethanaa): to pity
sa~n~naa (sanyaa): promise, contract, sign
san.khaara (sangkhaan): the physical body
vi~n~naa.na (winyaan): soul, spirit

Best wishes,

Robert