ศาลาชุมนุม

[Pawyilee]

I suggest reading Wikipedia's article, Chain shift, and discussing how these occur in Lao and Thai consonants, consonant clusters, vowels and tones.
http://en.wikipedia.org/wiki/Chain_shift

[pensive]

Interesting. What I have been thinking about recently, is that only initial consonants in Thai have any stress. Final consonants tend to fade into the background. I wonder if this is a similar shift; that Thai is easier to parse if final consonants are unstressed. It would appear to create unambiguous utterances.

[Richard Wordingham]

Well, we have a small bit of a chain shift in the change of พ ท from [b] [d] to [p] [t] (Northern) or [ph] [th] (other) and then บ ด from [ʔb] [ʔd] to [b] [d]. I can't think of any examples in vowels or clusters.

I look forward to the examples in tones - I have no handle on how tones change. The big tone change is the tone split that gives rise to tone rules in Thai (all dialects and all three basic scripts), Lao, Tai Lue and points North and East.

[Richard Wordingham]

I can think of three other possible examples, though the evidence for calling them chain shifts is weak.

Proto-Tai had four /j/-like consonants or clusters - /ʔj/, /j/, /hɲ/ and /ɲ/, anciently written อย, ย, หญ and ญ. They are respectively mid, low, high and low consonants. For Lao and Northern Thai we have the apparent chain shift /j/ > /ɲ/ and then /ʔj/ > /j/. However, the original /j/ was a low consonant and the final /j/ is a mid consonant. Therefore they are only really a tone shift when the low and mid consonants yield the same tone. For Lao. that only consistently happened for tone class C (mai tho), the least common of the original three live tones. The merger in tone class A (live, no tone mark), where it has happened, may well be late in Lao dialects. On the other hand, tone class A mid (= A3 only, usng Gedney's notation) and low classes have merged in Northern Thai, so there may have been a proper chain shift here.

An example of a vacuum, though I don't see how to present it as a chain shift, is the distribution of consonants and tones after tone splitting and the great Thai consonant shift. In Central and Southern Thai and Lao, the mid class consonants only occurred in three tones. However, I can't think of any systematic filling of this gap. However, in Northern Thai and Tai Lue, the aspirated stops were only represented in the high consonants (unless you count P/S loans in the old Indic voiced aspirates such as ธ). In these dialects, you get the changes /br/ > /pr/ > /ph/ and /gr/ > /kr/ > /kh/. (The merger of the consonant written ᨤ (= Thai ฅ) and the cluster written ᨣᩕ (= Thai คร) had not occurred in the Tai Lue that Li studied in the first half of the 20th century.) PS loans in /dr/ or similar also show the same development in Northern Thai. However, this may be closely related to the development /r/ > /h/. In Northern Thai (I can't speak for Tai Lue) similar changes have happened in high consonant clusters. Most of these clusters are of recent origin, for SW Tai had eliminated *pr (>t) and *tr (>t), and *kr was rare, and possibly being eliminated by sporadic(?) *kr > *kl. Li presents two examples of inherited SW Tai *kr yielding /kh/ in Lao.

The change /r/ > /h/ might conceivably be treated as an example of a vacuum being filled. In SW Tai, the only evidence of *hr distinct from *h is in Ahom. Elsewhere, after tone splitting, /h/ only occurred as a high consonant and /r/ as a low consonant. There was thus no communication impediment to /r/ > /h/ - though Central and Southern Thai have resisted it.