13.8.18.23:27: A 'WIND'-ING DETOUR

In my last post, I proposed the following Old Chinese (OC) etymology:

鳳凰 *N-pəm-s waŋ 'phoenix' < 風 *pəm 'wind' + 皇 *waŋ 'sovereign'?

- I thought it was the source of Siamese lom 'wind' (a borrowing from a Chinese dialect with *l- < *r-?)

- William Boltz suggested to me that 風 could be the source of Korean 바람 param < Middle Korean pʌrʌm 'wind'

- it shares a phonetic 凡 with

梵 Late OC *bramh, a transcription of Sanskrit Brahmā

嵐 Late OC *ləm (< *rəm?) a transcription of the middle syllable of Sanskrit vairambha- 'a kind of wind'

However, last night I checked Pittayawat Pittayaporn's 2009 dissertation which has Proto-Tai *C̬ .lɯm^A 'wind' with an unknown voiced preinitial *C̬-. The Lakkja cognate of 'wind' is /jɤm^A2/ (< Proto-Lakkja *jom^A2 [L-Thongkum 1992: 81])  without a labial, implying that the preinitial of 'wind' in the common ancestor of Lakkja and Proto-Tai was not labial. Perhaps it was palatal *ɉ- or dental *d- (cf. Li Fang-Kuei's Proto-Tai *dlu̯om^A 'wind'). Unfortunately there is no Proto-Kra cognate; Ostapirat (2000: 231) reconstructed Proto-Kra *gwjən 'wind'.  If 'wind' had an initial *b-, I would expect Lakkja */plɤm^A2/, since Proto-Tai *C̬ .lep^D (< *b.l-?) 'fingernail, toenail' corresponds to Lakkja /pliːp^D1/. (Oddly the Lakkja series 1 tone implies *p- rather than *b- which should condition series 2.) In any case, a Proto-Tai *b- would not match Chinese *p- unless it reflected a voiced prefix: e.g., *N-p- > *mp- > *mb- > *b-.

Moving on to the Sanskrit loans:

梵 could have been borrowed into an OC dialect which had already lost *-r- and broken nonemphatic *a  to *ɨa (which later became *ua before labials). *ɨa or *ua could be approximations of Sanskrit ra.

Schuessler (2007: 343) regarded 風 'wind' in 嵐 as semantic. Perhaps 風 was only partly phonetic: i.e., chosen only for its rhyme.

If 'wind' was *pəm without *-r-, then a Chinese etymology for Korean param is no longer viable.

'Wind' resembles Tibeto-Burman buŋ-type words for 'wind', but it is not clear whether those words are just lookalikes of Old Chinese *pəm. Matisoff's (2003) Proto-Tibeto-Burman reconstruction has *pup and *bup, so perhaps *pum and *bum would also have been possible, and those buŋ-words could be from an earlier *bum with *-m to *-ŋ dissimilation (which also occurred in Chinese).

I don't actually believe in Proto-Tibeto-Burman - a common ancestor of all non-Chinese Sino-Tibetan languages. I only use Matisoff's reconstruction as a stand-in for early Tibeto-Burman languages. I use 'Tibeto-Burman' to designate a paraphletic group: Sino-Tibetan minus Chinese. The Tibeto-Burman of Viet-Muong is Muong which refers to non-Vietnamese Viet-Muong languages (Phan 2010). Compare the following diagrams:

Sino-Tibetan: the old two-branch model

Sino-Tibetan: Chinese as one of many (not just two) branches

Viet-Muong: the old two-branch model

Viet-Muong: Vietnamese as one of many (not just two) branches

In both cases, the most famous branch (Chinese and Vietnamese) was recognized while all others were lumped together into a single group (Tibeto-Burman and Muong). The task ahead is to break up these phantom 'other' branches. How many daughter languages did Proto-Sino-Tibetan and Proto-Viet-Muong have?