How should Asian names be transcribed into the Roman alphabet?

Thailand’s island of Phuket might be pronounced “F*ck it” by some Americans. Ambiguity of pronunciation affects other languages as well — should sushi be spelled sooshee to make sure English speakers pronounce it properly? — but I’ll limit this post to Thai examples. And the very word Thai is an example of the problem: Some Americans might read that and think the pronunciation is like English “they” not “tie.”

I’ve not posed my question yet, but “Use the International Phonetic Alphabet” is NOT the answer to my question! We want to map Thai names to the 26 letters of the Roman alphabet, with no other symbols used. (I’d allow hyphen and/or apostrophe as one additional symbol denoting a glottal stop, or syllable separator, as in the Thai name Cha-on.) Thai citizens’ names are written with the Roman alphabet on identity cards, and passports. Place names are rendered in the Roman alphabet on road signs, and so on. Some of these transcriptions become relevant in legal documents.

There is an official Thai government standard (RTGS) in place to address this question. (It has undergone changes and I think some variants allow diacritical marks or other deviations from a pure 26-letter system.) I don’t really think I could improve on it much. Both consonants and vowels pose trouble. For example, there are many Thai names that rhyme, more or less, with English “worry,” “furry” or “jury.” A Thai child nick-named เจ้ย will be taught to write her name in English as “Joei.” Does anyone think she should write her name as “Jurry”? Especially since that English form only approximates the correct sound; the less familiar “Joei” might jog someone into learning a more correct pronunciation.

One reason for RTGS’ rendering ‘u’ instead of ‘oo’ is that many non-English languages use the Roman alphabet. Am I wrong that only English uses ‘oo’ to denote the ‘u’ sound? (However English is increasingly the “official” language of Thailand and its neighbors. Should that affect the transcriptions?)

For me, the transcription issue became a puzzle when writing letters to my U.S. relatives discussing my Thai friends and relatives. Should I use a spelling more likely to induce a correct pronunciation in easy cases? Or a more consistent system? Finally I decided consistency was best, even if it meant trusting the reader to know that ‘u’ was pronounced ‘oo’ and so on.

Here’s another example, a province name shown in Thai, RTGS, IPA, and an easy-to-read phonetic form showing tones:
อุดรธานี … Udonthani … /ʔùʔ dɔːn tʰaː niː/ … oo[sup]L[/sup] daawn[sup]M[/sup] thaa[sup]M[/sup] nee[sup]M[/sup]
Even the “easy to read” form assumes the reader will know ‘th’ is aspirated T, not English ‘th’ as in “this” or “thistle.”

(Thai distinguishes vowels of long and short duration, but RTGS does not. Udonthani has the short ‘u’ but Phuket has a long ‘u’. I might write Phu:ket or Phuuket to denote the duration, though doubling ‘o’ or ‘e’ to show long duration would be far too confusing.)

There are various other quandaries that arise in these transcriptions, but to keep OP short I’ll mention them in follow-ups if at all. One reason this whole question has become topical is that, supposedly, British consulates are now mandating a different transcription system from RTGS, for example insisting that ‘u’ be replaced with ‘oo.’ (The U.K. FCO hasn’t answered my e-mail so I’ve not confirmed this, but “certified translaters” seem to be enforcing the change.) Even slight variations in transcriptions can cause grief with government bureaucracies.

The rule change allegedly mandated by U.K. FCO may be based on a misunderstanding, but meanwhile some schools here are starting to do their own non-RTGS transcriptions. I pointed out to my friend, an English teacher here, that some Thais will no longer be able to write their names properly in English with his system. He didn’t care!

It depends on which language they’re from and which language you’re transcribing to. But this applies whenever you’re trying to work between two languages which don’t use identical writing systems; people with German or Scandinavian names are used to seeing all kinds of strange and unusual things happen to their vowels when they need to take a plane.

Since your question appears to be about transcribing Thai to English (a case which is particularly complicated due to the receiving language not having consistant self-transcription rules in either direction), maybe you could ask to have the thread name changed. Or, since it doesn’t seem to be so much a question as a rant on “how dare country governments not do things the way I think they should!!!?” you could ask for a move to the Pit.

What Nava said. Different cultures that use the Roman alphabet assign different values to the letters of that alphabet. Countries with non-Roman alphabets don’t necessarily want to have a whole series of different Roman transcriptions, to suit the tastes of different users of the Roman alphabet. Any reasonably regular transcription system is as good as any other; all possible transcription system will give some foreigners better clues to pronunciation than other foreigners. Just pick one and stick with it, unless you find a very compelling reason for changing it.

As a librarian, I follow the ALA-LC Romanization Tables. The authoritative reference.

Nitpick: Transcribe means speech-to-text. You seem to be asking about text-to-text, so the word is transliterate. To transliterate into the Roman alphabet is romanize.

Here’s the straight dope on Thai romanization: Romanization of Thai - Wikipedia

There is a conflict between transliterations to a “generic” Roman alphabet, and to English specifically. But a single transliteration is needed! Thais do not get different passports depending on which country they’re traveling to! :slight_smile: Nor do Bangkok street signs show separate renderings for English, French and German!

I did NOT post in BBQ Pit because mine is not a rant (though I’d be peeved by certain transliterations). Thailand’s schools (and the British Embassy) may be embarked on changes to the transliteration system and it would be good to have informed opinions about such changes.

@ Nava — Did you read this part of OP? :-

Nobody in America would ever pronounce a name spelled “Joei” as “Jurry”

And most Americans would pronounce Udon Thani with a “th” but they won’t with “Thailand” since most Americans know it is not pronounced with a “th” (Except in the movie Hangover II) :slight_smile:

I’ve seen a lot of Thai names in English and a lot use the aspirated versions of the consonants, I don’t think many English speaking people will know how to pronounce them no matter how they are translated.

I didn’t want my OP overly long, so only hinted at key points. Whatever system one uses will have trade-offs. If you transliterate to make it easy for Americans to pronounce some words roughly correctly, you’ll end up with some words that can not be rendered at all, and some words that will be mispronounced by non-English foreigners who are more familiar with ‘u’ being the English ‘oo’ sound, ‘i’ being English ‘ee’, ‘ai’ being English ‘ie’ (as in ‘tie’), etc.

OK. But you’d ask how to pronounce it and end up with a closer rendition than ‘Jurry.’ After all, ‘Joei’ doesn’t have an actual R — there’s just a hint of an R from the vowels, as in English ‘adieu.’

I’m not sure why the question is centered around Asian names because it’s a problem for the entire world. Perhaps it’s because problems with other European languages is already well known.

“Sushi” should never be written that way anymore than San Jose should be spelled with an “H” when dealing with English speakers.

The OP asks for something impossible, which is would be to transliterate other languages into an alphabet which can’t even distinguish the separate sounds denoted by the same letters in English.

Yes, did you read the part where you were complaining about the Brits?

I quite like what Wiktionary has done with the romanization of Thai. For example: การเลือกตั้งทั่วไป - Wiktionary

They break it down all the way and then build it back up. First they analyze it letter by letter, to show the underlying orthography, which also helps with understanding its etymology. Then they take it to the phonemic level, illustrating how the alphabet functions in relation to sound. Finally they give both the Paiboon and Royal Institute romanizations. They round it off with accurate IPA pronunciation. That’s the complete package right there for you, Septimus.

Thanks, Johanna ! I like the Paiboon system with /bp/ and /dt/ for unaspirated P and T. Unfortunately almost nobody else uses it.

My relative is nicknamed ต้น (Paiboon /dtôn/; RTGS ‘Ton’). He renders it in the RTGS form ‘Ton.’ I suppose my Brit teacher friend would want to write it as ‘Doan.’ :smack:

Interesting example you picked for your link there, Johanna ! I guess you noticed today was Election Day here.

Why would the UK Foreign Office care? Well, Wikipedeia says: "the ICAO document 9303 part 3 describes specifications common to all Machine Readable Travel Documents. " and “Due to technical limits, characters inside the Machine Readable Zone (MRZ) need to be restricted to the 10 Arabic numerals, the 26 capital Latin letters A through Z, and the filler character <.”

I’m wondering if the passport standard has some specific rules for Thai Romanization, as it does for some other character sets.

That ambiguity affects English words that were never borrowed. Everyone agrees that English spelling is probably the worst at reflecting the actual pronunciation of words. It’s the reason why English is the only language where spelling bees are a significant phenomenon. We live with it by memorizing spellings. The example you give there is one that most people have already memorized. And that will apply to new words we borrow as well as names.

Interestingly enough, Japan itself has an official “Kanji Kentei” spelling test, so it’s not a uniquely English phenomenon. They live with it by memorizing multiple readings for each character.

I have said this type of thing before, but I 100% guarantee that spelling 寿司 “sooshee” will not bring an English speaker any closer to pronouncing it correctly, especially the first vowel.

British spellings of Sub Continental personal and place names are an important guide into how the words were pronounced then and how language has changed. Since the 19th century spellings have persisted even with changes in pronunciation.
Like Lahore. It’s pronounced “La Hoar”, today.
Karachi is a interesting example, since the local pronunciation has changed to Kee Ran Chi, but the official pronunciation (and the rest of the country) remains Ka Ra Chee.

How was it pronounced in the past?

To answer the thread title, very, very carefully! :smiley:

For some names I think it’s due to tradition based on how the original transliteration was done back in the day. ISTM that the people doing the work back then did a poor job with some of the names, and it’s taken time to fix the names as rendered in English. Bombay is now Mumbai. Peking is now Beijing. Those are different enough that I wonder why the original transliteration used the former when the latter are closer to how they actually sound. Couldn’t they have started off with Mumbai, Beijing, and Tailand? If they had, it wouldn’t be as much of an ongoing issue.

I don’t know about Mumbai, but “Peking” is closer to the way the name would be spoken in the southern port cities of China, where most of Western contact took place. In Cantonese, it sounds very much like “puck-ging”, with a hard g. In Mandarin, it sounds more like “bay-jing”, jing as in jingle.