ศาลาชุมนุม

[ling]

Thai isn't an easy language to transliterate into the Roman alphabet. But I find some systems better than others, and I understadn that some spelling is based on Thai pronunciation, others on Thai orthography, and still others incorporate Sanskrit romanizaiton conventions. But one of my complaints as a learner of Thai is that just about every textbook, self-instruction book or phrasebook seems to use its own romanization system. (Thank heaven for CDs and .mp3s!)

My personal preference is for systems that use d/t/th over d/dt/t and b/p/ph over b/bp/p. I also like แ represented as "ae" (not the confusing "aa" used in Lonely Planet, or worse, "air" used in Teach Yourself). And Teach Yourself transcribes แถว as "tair-o", which makes me think of "taro". (I would spell it "thaew".)

The last example brings me to my biggest peeve: the use of the letter "r" in syllables like ขอ ("khor"). Why? Because I'm American, and like most Americans I speak a rhotic dialect. When I see "khor", I want to say "core" with a good old American arrrrrr in it.

Another peeve is seeing "krub" instead of "khrap" for the male polite particle: "krub" makes me think it could be pronounced "kroob" (like someone from Northern England eating "groob" at the local "poob"). It's the same reason why Punjab is all too frequently mispronounced as "poonjab": it was romanized with a "u" representing a short "a".

Grrr.

OK, rant over!

[David and Bui]

Simple solution to your psychic pain: ignore Romanizations entirely; stick with the Thai script. Note that this site gives you choices, including "phonemic Thai" which uses Thai orthography to clarify Thai spelling ambiguities.

[ling]

Simple solution to your psychic pain: ignore Romanizations entirely; stick with the Thai script.

Indeed, that's what I'm doing! It's gotten to the point where I pretty much have to look at the Thai script to make sure my pronunciation is correct! (Though those silent letters, like the ร in จริง, sometimes throw me for a loop.)

[Richard Wordingham]

Note that this site gives you choices, including "phonemic Thai" which uses Thai orthography to clarify Thai spelling ambiguities.

Phonemic Thai doesn't always resolve vowel length.

[DonSena]

Simple solution to your psychic pain: ignore Romanizations entirely; stick with the Thai script. Note that this site gives you choices, including "phonemic Thai" which uses Thai orthography to clarify Thai spelling ambiguities.

Dave is right. A phonemic representation of any language will, by definition, contrast all (phonologically) distinctive features of that language --Thai included.

The transcriptions provided in the Forum dictionary, though, are sometimes incorrect for the their corresponding entries. Listening to the native speaker, where audio clips are provided, will occasionally show a discrepancy between the speaker's pronunciation and the transcription provided. For this purpose, get a scholastically-written Thai-grammar text that uses a form of IPA transcription adapted to Thai.

In the case of Thai, the IPA will contrast length, aspiration of voiceless initial stops and affricates, vowel length, (volumetric loudness) stress, rhythm and intonation. These featues are "distinctive" in Thai because altering any one of them results in either a different meaning or no perceived meaning at all.

For the latter three of these features, and why they are relevant, see the paper in the References section named "the Prosodic Features of Thai: Meaningful in Their Own Right."

[Thomas]

Dave is right. A phonemic representation of any language will, by definition, contrast all (phonologically) distinctive features of that language --Thai included.
The transcriptions provided in ...
Listening to the native speaker, where audio clips are provided, will occasionally show a discrepancy between the speaker's pronunciation and the transcription provided....
These featues are "distinctive" in Thai...

Thai isn't an easy language to transliterate into the Roman alphabet. ..

My personal preference is for systems that use d/t/th ...

DonSena ... sorry for this remark, but is yours eventually a sligthly too biased (?) view on the question, or rant, of khun ling?

Trying to improve my listening and speaking facilities I would ask my wife: Could you say มา, ม้า, and หมา? Would listen and, highly likely as always, not hear any difference and ... try it later again. On the other hand, I would be very glad provided that my wife, speaking German, would pronounce at least 2 of the 3 consonants in the final cluster of a German term because these are "distinctive" features in German... Romanization, or vice versa, thap-sap-ing will not help one of us in those situations.

But Romanization is not needed always for means of phonetic studies (would say rather infrequently). In my mind comes a journalist of NYT who needs a headline reporting that somebody (What is his name? Please use Latin letters!!!) doing something in a place in Thailand (Where? Damned, write this in 'normal' letters, not these funny curls and loops you sent me last time by fax!!) --- not so much myself learning Thai from my wife...

Insofar I understand khun ling very well.

My personal preference, for means of ROMANIZATION (i.e. to render it in Latin letters, not necessarily transporting pronunciation - my boss in New York will not pronounce it but will let print it), is RTGS for the single reason that it has a kind of (semi-) official character. Yes, I'm just aware of all incoming critizism of RTGS: It does not discern vowel length (did this Latin, or is vowel length a very specific feature of Romance languages?)! It does not provide information on the tone (how many inhabitants of New York City will know something about the tone system of Thai language?)! What a big mistake by RI using/prescribing ch for จอ จาน, it has to be represented by c, or j, or best by xyz!!!

Let me finish my remark(s) with a story on a Romanzation-accident that happened to me in real life:

I was chatting by Google-chat with a compatriote staying in Thailand. He was using, for means of Romanization, more or less RTGS with the nearly single exception using gk for initial ก (potential argument: German K is no ก so กุ้ง has to be gung ... but in theory ก is a ... compromise: gk).

While chatting I switched from my puristic RTGS to his modified RTGS version for reasons of politeness. Then I asked him:

Me: Where are you?
He: In Bangkok!
Me: Sorry, where? Village of Gkok? Is this in the North-East?

No joke, this happened really in my brain.

[Richard Wordingham]

It does not discern vowel length (did this Latin, or is vowel length a very specific feature of Romance languages?)!

Latin did, but optionally. See for example http://en.wikipedia.org/wiki/Apex_%28diacritic%29 .

The big problem with the RTGS is the difficulty in converting it back to Thai.

Me: Where are you?
He: In Bangkok!
Me: Sorry, where? Village of Gkok? Is this in the North-East?

No joke, this happened really in my brain.

Of course, he should have written Banggkok!

[Thomas]

It does not discern vowel length (did this Latin, or is vowel length a very specific feature of Romance languages?)!

Latin did, but optionally. See for example http://en.wikipedia.org/wiki/Apex_%28diacritic%29 .

I did not say or meant that Latin has no long (or short) vowels. Makron is a sign well known to me but while I learned Latin 35 to 40 years ago the school books we used showed no makron. The apex says to me that sometimes vowel length had to be indicated in a text to avoid misunderstandings. I assume, however, a relatively rare case. Two days ago, in a small talk with a collegue, she was using a (h.Ger) term with short vowel (Sönn-chen) but I percieved it as the same term with long vowel (Söhn-chen). Spelling differs in German already without use of apex. I understood: Little son has gone? while she actually asked: The few sun has disappeared?

The big problem with the RTGS is the difficulty in converting it back to Thai.

To Romanize Thai into unicode would solve, actually solves in our times of computers, the problem. I'm saying it in this extreme since any other transliteration system based on Latin letters, diacrits, and Arabic numerals will not be much more readable for the users than unicode. The big pragmatic advantage I see in RTGS is that, provided that you have a) the Thai term b) the term in RTGS you can deduce the pronunciation, i.e. you can write the Thai term in สัทอักษร (on which you can apply the tone rules).

(((leaving out the few instances in which spoken Thai does not follow สัทอักษร)))

Question:

Seeing สัทอักษร /xyz/sattha-akson

Is it unambigous what 'xyz' is in phonetic alphabet? I guess in 99.5% yes but nearly for sure I forgot some exceptions?!

He: In Bangkok!
Me: Village of Gkok?

Of course, he should have written Banggkok!

Sometimes I thought the same. But Bangkok is in doubt not only a romanized Thai term but also simply a well known German term. Thus, I accuse my brain for this accident, not the usage of the term Bangkok by my compatriot. The overall context was that we were chatting in German (to which some romanized Thai terms were added).

[Thomas]

The big problem with the RTGS is the difficulty in converting it back to Thai.

After reconsidering your remark I want to add:

Yes, of course, as it is a transcription but not a transliteration system.

[Richard Wordingham]

To Romanize Thai into unicode would solve, actually solves in our times of computers, the problem.

Typing Thai in an international encoding does not constitute Romanisation. Indeed, it gives one almost nothing that one does not have in TIS-620. The disadvantage of TIS-620 is that it does not support IPA or indeed most Roman script languages. The ISO 11940 Part 1 and pre-1968 RTGS rely on Unicode for 'easy' computer use.

Seeing สัทอักษร /xyz/sattha-akson

Is it unambigous what 'xyz' is in phonetic alphabet?=

It is ambiguous whenever Thai is ambiguous as to vowel length (and wrong when the Thai spelling misrecords the length). One will also have all the troubles of the tone rules. My favourite example is ตนุ//tanu.

It is also ambiguous when muta cum liquida is spelt with a voiceless stop. I can't remember any of the good examples, but a clumsy example is บาปรับ//baaprap. Most examples will have ก as the consonant.