Tai, or whatever

Oh wow, you aren't seriously using the OC reconstructions in Schuessler's ABC Etymological Dictionary of Old Chinese are you? Like, 吠 is OCM *ba(t)s < *bos?, OCB *bjots? That's so 2006, man. Get with the times and download the all-new (September 2014!) reconstructions from Baxter and Sagart's Old Chinese: A New Reconstruction already. They're free! (吠 is now *Cә.bo[t]-s.)

No, but seriously, I also lucked into a very cheap secondhand copy of OCNR, and so far it's been great reading. Like, the very second footnote is entertaining:

We adopt the term "Kra-Dai" proposed by Ostapirat (2000) in place of the traditional "Tai-Kadai," since to Thai speakers, "Tai-Kadai" evidently sounds unintentionally funny, meaning something like "Tai, or whatever" (Montatip Krishnamra, p.c.)

Does anyone reading this know enough Thai to elaborate here?

I do have one complaint about the book, although it doesn't reflect poorly on its authors at all: the printing feels cheap. The text has just enough digital artifacts and jaggies to be obnoxious, and the paper is just thin enough to show through noticeably. None of it's bad enough to harm readability, but it's a shame; a book like this should be a pleasure to look at.

(I honestly thought that I'd ended up with a cheap edition intended for Asian markets or something, which would of course be a whole different story, but I can't find any indication that this is the case, and the title page says "Printed in the United States of America". Î don't know — maybe I got taken by a very specialized gang of counterfeiters?)

Popularity factor: 17


I don't know any Thai, but I imagine it's much like Shm-reduplication and similar phenomena in Turkish ("Twitter-mwitter") and Bengali (which I don't remember where I read about it.)

Matt A:

I agree about the printing—that was also the first thing I noticed about the book, other than its very nice cover, though at least the pixelated printing doesn't seem to bother me when I'm actually reading it. I bought it from Amazon in the US, so if it's a specialized counterfeiting gang, they must be pretty wide-reaching. (& yeah, that note is great.)


Since you're reading this, could you clarify what exactly are the rules of the Baxter-Sargat orthography? People tell me that weird stuff like "trhjowngX" or "qʷʰˤat" are more or less abstract representation of known and proposed phonological relationship, but not an actual phonetic reconstruction of something you should try to pronounce (cf. Bel Matin's comment in this thread: http://paleoglot.blogspot.com.br/2012/01/baxter-sagart-reconstructions-and.html ). But when I browse the book in Google it tells me that they're using IPA, in which case…

Schuessler does use a phonetic orthography, and try to restrict his proposals to the things with better evidence, which I think is much more in line with what most people want from reconstructions (for B–S "trhjowngX" he has "tjwänᴮ").


On pp7-8, they cover the basics: Middle Chinese is given in a "conventional transcription", specifically not a phonetic reconstruction, written in standard ASCII characters, italicized but no asterisk; 華 xwae, 實 zyit, 傳 drjwenH, 奇 kje. Old Chinese is in IPA, with asterisk. "Hyphens indicate morpheme boundaries; angle brackets around prevocalic *-r- indicate that it is an infix." Recognizable synchronic affixes are separated by hyphens, other pre-onset material by periods. Regular parentheses indicate phonemes that may or may not have been there, and square brackets, well, "the notation "*[X] means "either *X, or something else that has the same Middle Chinese reflex." Thus, *qʷʰˤra, *mə.li[t], *N-tron-s, *[k](r)aj, for the four characters above.

That said, if you aren't trying to actually say "drjwenH" and so on, you're missing half the fun of this topic in my opinion.


FM: That's a definite possibility!

Matt A: Thanks for the corroborating data! Glad it wasn't just my copy, at least.


I don't think IPA is limited to things that you're intending to pronounce—we used it in phonology (and even morphology) classes as well. Generally when working with suffixes or prefixes that changed we'd assume a "base" form that changed in particular circumstances….

However, I believe we *did* actually use a symbol for vowels without all of their features expressed (… okay, that's not particularly sensible, but I'm sleep-deprived) when working with vowel harmony, however. Hm.

But anyway: using IPA even for words which may have never existed as pronounced in that way (or, reconstructions) is not all that weird at least in undergraduate linguistics majors. (Actually, we would have probably had to write everything down in IPA if some of the professors would have had their way.)


@Mumei: Yeabut the point is whether or not B-S are actually really trying to convince us that Old Chinese had consonants like (to borrow from the Paleoglot thread) /*qʷʰˤ/, distinct from /*qʷˤ/ & so on; and if they are arguing for that, it will take some pretty darn good convincing!

Matt A:


As I understand it, Baxter and Sagart are really reconstructing the relationship between these sounds, not the sounds themselves.

But the distinction between /*qʷʰˤ/ and /*qʷˤ/ is simply one of aspiration, which doesn't seem unusual at all (they propose a three-way voiced/unvoiced unaspirated/unvoiced aspirated distinction). The labialization is also pretty straightforward. The most speculative part of /*qʷʰˤ/ or /*qʷˤ/ is the pharyngealization, which corresponds to the presence or absence of a circumflex in Schuessler's system (and is related to the Middle Chinese Type A/B distinction). But, yeah, I don't think anyone knows whether their *q was definitively a voiceless uvular stop or not.


@Matt A: None of those features are in themselves strange. What I find hard to accept is that all would be operating at the same time, and distinctively – I wonder how much the spectrograms even change between, say,

*qʷʰˤaj "move, change";
*kʷʰˤaj "great";
*kʷˤaj "dagger-axe; to pass";
*kʰˤaj "may; acceptable";
*qʰˤaj "scold, shout";

and the other possible combinations. Paleoglot said "it's as if a computer programmer created this language", and as an ex-programmer, I feel the same way.

I mean, Nuxálk exists, so I'm not rejecting such a system just because it *seems* weird and unworkable; but I'd want some really good evidence to accept that this is the best guess at what did OC sound like. Then again, I haven't read their book yet, so I'm not allowed to opine. It's just that their (not-quite-phonetic?) OC reconstructions, as well as their totally not-phonetic MC transcriptions, are popping all over as representations of the sounds of OC and MC, by people who are as clueless to their arguments as I am. Of course, this is a consequence of their very laudable Internet openness (which Schuessler et al. should definitely imitate); I just wish that, together with the tables, they also published an easily-accessible introduction to their notation, as well as an objective evaluation of the likelihood of their phonetic hypothesis (and the state of the debate).


It would hardly be a reconstructed ancient language without an extravagant three-way contrast that daughter language speakers can't even recognize as phonemic material, let alone pronounce. (Laryngeals, anyone?)

I'll skip ahead in the book and let you know what they say about plausibility and so on. (It's okay, I've already been spoiled -- OC evolves into a long list of names for orcs used mainly to write rhyme tables and localize Sanskrit jargon, record barbarian poetry in distant lands, etc.)

Also, you could try writing to them and making this suggestion directly -- they say they plan to put more material on the site and a quick summary of orthographic principles and their interpretation would be a helpful resource even for those who have access to the book.


OK, I peeked. Re pharyngealization in particular (pp68-75), B&S definitely believe that there was actual widespread pharyngealization, due to reflexes associated with pharyngealization specifically (lowered vowels, velar + pharyngeal > uvular changes).

They note that their hypothesis "obviously leads to a very large inventory of consonants, in which all plain consonants have a pharyngealized counterpart, regardless of place or manner or articulation. We are aware that such a system is typologically unusual... A possible alternative would be to reconstruct a distinct pharyngeal segment [ʕ] occupying a separate slot at the juncture of the onset and rhyme, a solution we have not adopted but do not wish to exclude."

On the other hand, they don't claim that pharyngealization was a long-term, stable feature of OC; in fact, they specifically raise the possibility that it may have been a short-term thing in the late Hàn period which "actually existed only for a short time, and, being typologically unusual, was rather unstable and soon led to further changes." In other words they frankly admit that their reconstruction might not be OC as it was spoken for centuries, but rather OC at its greatest moment of phonemic crisis -- or even, I suppose, a composite of several such moments.

David Marjanović:

<blockquote>*qʷʰˤaj "move, change";
*kʷʰˤaj "great";
*kʷˤaj "dagger-axe; to pass";
*kʰˤaj "may; acceptable";
*qʰˤaj "scold, shout";</blockquote>

...But that's just bog-standard Average East Caucasian, except for the lack of ejectives. Even I can pronounce these things, and I have no real experience with uvulars <b>or</b> pharyngeals. Try, people! :-)

And [kʷʰˤaj] is a great word for "great".

David Marjanović:

Oh, speaking of ejectives: *lightbulb moment* If a whole bunch of Semitic languages can turn ejectives into pharyngealized consonants, perhaps so can OC?

Alex Fink:

But if the pharyngeals were once ejectives, only stops should appear pharyngealised, or only obstruents at any rate. That's not what we see.

My own opinion is that the distinction might well be pharyngealisation, but tethering it phonemically to the initial consonant doesn't seem very justified. Why not just call it supersegmental?
Also, Ferlus thinks the MC yodlessness developed from some tense and creaky articulation which was a remnant of earlier presyllables, citing Khmer as a partial parallel. To me this seems quite compatible with the [ˤ], either as two different temporal stages or two different aspects of one phonetically complex gesture.

Anyway, phonologies that look like they were designed by programmers are a damn sight better than things like e.g. Karlgren's vowel system.


A very tempting idea I have seen is that Sino-Tibetan languages of old were *not* mono/sesquisyllabic, and that there was a typological shift in root structure later on, with major syllable contractions. Under this scheme, one could probably re-write the various POA modifications into different vowels; e.g. *k : *kʷ : *kˤ : *kˤʷ → *kə : *ku : *ka : *ko.

(Similarly, try to imagine a monosyllabic reconstruction of West Germanic? Could be done, probably, but at the cost of a horrible multitude of vowels, or maybe shunting off some features onto consonants.)

It'd take some serious work to actually sketch out such a system, though. It doesn't sound like a walk in the park, esp. if one wanted to take advantage of this idea to streamline Sinitic dialect developments.


The "Kadai" sounds similar to ก็ได้. The gloss is "also okay" similar to 也行 or 也可以 in Chinese. You say this when you don't fully agree with what someone has said, but you'll go along with it anyway, hence "Tai, or whatever."


Thanks, CE!

Aime la vérité, mais pardonne à l'erreur

LU d'R
Mail d'E

All fields optional. E-mail address will never be displayed, resold, etc. -- it's just a quick way to give me your e-mail address along with your comment, if you should feel the need. URL will be published, though, so don't enter it if it's a secret. You can use <a href>, but most other tags will be filtered out. (I'll fix it in post-production for you if it seems necessary.)