It has been brought to my attention that this blog is becoming Russocentric to a degree (I use the phrase in its original sense), so I thought I’d ruminate on a different part of the linguistic universe altogether.

For a long time I was confused by the terms “Thai language” and “Tai languages”: what was up with the h, and what was the relation between the two terms? Eventually (once I got out of the sandbox of Indo-European) I discovered that the Tai family of closely related languages spread from Assam in eastern India (named after invaders who spoke the now extinct Ahom language) to the mountains of northern Vietnam, and from southern China (whence they originated) to the Malay Peninsula. The best-known of them is of course Thai, the official language of Thailand; I’m not sure when and how the h came in (probably the Brit tendency to add it to foreign terms to make them look more exotic, cf. dhal), but it does come in handy to differentiate the terms, which are both from a T(h)ai word meaning ‘free’ (“Thailand” is half translated from prathet tai ‘country of the free’). Eastwards from India we find Shan (in Burma), Northern (or Lanna) Thai (in northern Thailand), Lao (the official language of Laos, but the majority of its speakers are in Thailand, where it is often called Northeastern Thai), and Red Tai, Black Tai, and White Tai (in northern Vietnam—no formal-wear jokes, please). An interesting point is that an alternate ethnic designation is at the base of Assam/Ahom, Shan, and Siam, as well as the Cambodian term Siem (the name of the town Siem Reap, near Angkor Wat, means ‘defeat of the Thai’).


  1. The “h” in Thai reflects the standard pronunciation with [tʰ]; Thai has a three-way distinction of unvoiced unaspirated, unvoiced aspirated, and voiced stops which may go back to Proto-Tai. The terms “Tai” and “Daic” reflect related languages that use an unvoiced unaspirated stop in the cognate word.

  2. In Thailand I was once told that Thai and Tai refers one to the language and the other to the people (the Free).
    Can’t quite remember which was which, though.
    Pronunciation in Thailand is for the initiated only.

  3. David Marjanović says

    Sound files of Thai syllables with /b p pʰ d t tʰ/.

    As expected, /p t/ are voiceless lenes [b̥ d̥].

  4. Hang on, how can you tell it’s [b̥] as opposed to [p]?

  5. David Marjanović says

    Because I natively speak one of apparently two languages worldwide where that difference is phonemic without being backed up by voice, aspiration, glottalization or whatnot.

    Yeah, OK, there’s not supposed to be a difference between these IPA representations. That’s just the closest the IPA can get to representing the distinction.

    The distinction is in the loudness of the release, which depends on the air pressure in the lungs. Unreleased consonants cannot be distinguished that way. Of course that’s a phonetic continuum, but most languages tend to stick close to one extreme.

    For example, the Spanish or Finnish /p t/ and to a lesser extent /k/ don’t sound like the French or Hungarian or Slavic or Japanese ones to me, or like mine; they sound like my /b d g/. That makes sense: all of the latter’s sound systems, except mine, contrast voiced and voiceless plosives, and enhance the distinction by making the voiceless ones fortis. Finnish comes close to having just one series of plosives, in that [b g] are limited to recent loans and [d] to the standard that nobody speaks all that much, so the voice distinction doesn’t carry so much functional load that it would need to be shored up by an extra feature. Spanish basically distinguishes plosives from approximants; the rare voiced plosive allophones of the approximants evidently aren’t common enough to justify the extra effort either. Pure voice distinctions, between voiced and voiceless lenes, that bear a higher functional load are only, to my knowledge, found in more crowded systems, like Thai or East Armenian or Hindi.

    The other language that appears to use both extremes is the kind of Mandarin heard here being shouted at one of China’s vice prime ministers in Wuhan. LLog had a post saying what you can hear there would be jia de, jia de, quanbu shi jia de in Standard Mandarin (except that the conservative form di is used instead of de) and means “fake, fake, it’s all fake”. The q is a strongly aspirated dorso-palatal affricate in northern/Standard Mandarin. Wuhan is far enough south that the local variety doesn’t retroflex, so shi comes out as se (the other vowel change is that -ua- [ɥæ] is contracted to [œ]), and the palatal affricates (j, q) become rather rounded postalveolars. As everywhere in Mandarin, b, d, j are voiceless in the video. Yet, interestingly, the q is not aspirated at all – and still doesn’t merge with j.

    Dschadi! Dschadi! Tschömbusö dschadi! Write it like this, give that to Arnold Schwarzenegger, ask him to read it aloud, and he has a good chance of getting it exactly right.

  6. @David Marjanović: Of course he would.

  7. David Marjanović says

    …Interesting. As famous as that quote is, I had never heard it in the original. get to gets a simple [d̥], perhaps because he doesn’t have time to close his mouth far enough to build up pressure; the pp is [b̥] as if by overapplication of American flapping; but the ch starts and ends with a fortis, so we know he’d get quanbu right.

  8. David Marjanović says

    the Spanish or Finnish /p t/ and to a lesser extent /k/ don’t sound like the French or Hungarian or Slavic or Japanese ones to me

    Oh, I should have mentioned I’ve had a direct comparison: I’ve heard French with a Chinese accent. Putting [d̥] in intérieur jumps out to me.

  9. Thai and Tai are referring to the ethnicity. “Tai” as you say meaning free are two different words, they may sound the same to non speakers. Totally two different words.

Speak Your Mind