Amaravati: Abode of Amritas

13.1.12.23:59: A TOXIN FROM WITHIN?: THE ORIGINS OF 'POISON' (PART 2)

Tangut

1do 'poison'

cannot be traced back any further than Pre-Tangut *do on the basis of Tangut-internal evidence. It probably once had a final *-k since it is related to forms like Middle Chinese 毒 *douk and Written Tibetan dug. (See many other -k forms at STEDT.) The question is how. Is the Tangut word inherited or borrowed?

My guess is that Tangut do was borrowed* from Middle Chinese 毒 *douk because it has the mid vowel o which is a Late Old Chinese innovation rather than the high vowel u preserved in Tibetan.

However, things get hazier from here. Although it is certain that Middle Chinese *douk is from a Middle Old Chinese *duk (the underlining represents 'emphasis': i.e., pharyngealization), its Early Old Chinese source is uncertain. Here are four different scenarios involving the combination of two different variables:

	the 'nourish' word family (育毓谷) had an initial *l- (as in my reconstruction and Schuessler's)	the 'nourish' word family (育毓谷) had an initial uvular *q- (as in Baxter and Sagart 2011)
毒 'poison' was Early Old Chinese Cʌ-duk > Middle Old Chinese duk > Late Old Chinese *douk	Scenario 1: The Proto-Sino-Tibetan root for 'poison' was duk. Problem:* The use of a d-graph 毒 'poison' for an l-word for 'nourish' is unusual, though not totally phonetically implausible. Solution (1.13.00:40): 毒 was used for 'nourish' after the l- of 'nourish' had hardened to d-.	Scenarios 3-4: Problem: If 'poison' had d- or l- in Old Chinese, the use of its graph 毒 for a q-word for 'nourish' makes no sense, and 毒 'nourish' would have to e reconstructed as an unrelated word which happened to have a rhyme like those of the qUk-words: 育毓 m-quk (Baxter and Sagart 2011: 159) 谷 m-qok (not in Baxter and Sagart 2011, but reconstructed according to their system) Another problem (1.13.1:43): There is an Late Old Chinese dialectal word 陶 dou 'nourish' written with an l-phonetic 匋 in Fangyan (Schuessler 2007: 515). dou could go back to du or lu. The spelling 陶 'pottery' may have been devised on the spot, so it is not evidence for reconstructing lu. Reconstructing 育毓 luk (< √lwk* with zero grade?) 谷 lok (< √lwk* with -ə-grade?) with l- instead of q- allows me to group them together with 陶 lu (and 毒luk in scenario 4 but not 毒*duk in scenario 3) as members of a word family meaning 'nourish'.
毒 'poison' was Early Old Chinese Cʌ-luk > Middle Old Chinese luk > Late Old Chinese *douk	Scenario 2: Tibeto-Burman words for 'poison' with t- or d- were borrowed from Chinese after EOC l- hardened to MOC d-. Problem: How could a Chinese loanword spread throughout the Sino-Tibetan family, including members that had no known contact with Chinese? Solution (1.13.00:10): This contact occurred between Old Chinese and Proto-Tibeto-Burman, the ancestor of the non-Chinese Sino-Tibetan languages. Problem: Did Proto-Tibeto-Burman really exist, or is 'Tibeto-Burman' merely a convenient category of all ST branches other than Sinitic? 毒 'poison' had an l- at the time it was used to write an l-word for 'nourish'. (1.13.1:13: Or the *l- in both 'poison' and 'nourish' had hardened by the time the latter was written as 毒.)

1.13.00:08: See Sagart (1999: 31) for more on the Old Chinese initial of 毒 'poison'.

*1.13.1:45: Gong (1995: 51) reconstructed Old Chinese 毒 *dəkw and Pre-Tangut *du and regarded them as true cognates (i.e., descendants from a common Sino-Tibetan ancestor).

If Tangut 1do is from *du, then where would his Tangut 1du (my 1dʊ) come from?

I derive 1dʊ from *Cʌ-dru. I cannot explain why my Tangut reconstruction has no 1dəu from *Cʌ-du. Perhaps that is an accidental gap.

13.1.6.23:59: SCORPIONS AND SNAKES: THE ORIGINS OF 'POISON' (PART 1)

The Tangraphic Sea derives 'snake' from 'poison' which it derives from ... 'snake':

=+

1do 'poison' =

top and bottom center of 0la 'scorpion' (tone unknown)
bottom right of 2phɔ 'snake'

I think the tangraph for 'poison' may actually be from a variant of Chinese 毒 with the structure 'top horizontal over left box and right vertical' like

𠟽

𥫗 atop (亯+刂)

𦸕

listed in the ROC's Dictionary of Chinese Character Variants.

The derivation of 'scorpion' is unknown, but I think it is a transparent semantic compound:

=+

0la 'scorpion' (tone unknown) =

all of 1kie 'bug'

all of 1do 'poison'

Perhaps the word that Li Fanwen (2008: 346) glossed as 'snake; scorpion'

=+

2raʳ =

left and center of 2raʳ 'to flow' +

top and bottom right of 2phɔ 'snake'

with the clarifiers

2phɔ 'snake' and 2giuu 'scorpion' (as glossed by Li Fanwen 2008: 346; see below for another definition)

in Homophones was a word for poisonous reptiles and arthropods*.

Next: A Toxin from Within?

*13.1.7.1:07: Is there a word for this nonbiological category in English? Such a concept underlies the use of 虫 as a radical in both 蛇 'snake' and 蝎 'scorpion' in Chinese. These creatures have scales and exoskeletons instead of fur or soft skin.

Maybe no such category is needed. Li Fanwen (2008: 648) also defined 2giuu as 'snake'. Its Homophones clarifiers are

2phɔ 'snake' and 2raʳ '?'

This implies that 2giuu, 2phɔ, and 2raʳ are synonyms for 'snake'.

As far as I know, only 2phɔ is used outside dictionaries. That and the existence of external cognates for 2phɔ may indicate that it is a generic, basic term for 'snake'.

The others

2giuu and 2raʳ

are homophonous with

2giuu 'lucky' and 2raʳ 'to flow'.

and may be terms invented within Tangut for specific kinds of snakes that have not yet been found in texts: 'the lucky one' and the 'flowing one'.

	the 'nourish' word family (育毓谷) had an initial *l- (as in my reconstruction and Schuessler's)	the 'nourish' word family (育毓谷) had an initial uvular *q- (as in Baxter and Sagart 2011)
毒 'poison' was Early Old Chinese Cʌ-duk > Middle Old Chinese duk > Late Old Chinese *douk	Scenario 1: The Proto-Sino-Tibetan root for 'poison' was duk. Problem:* The use of a d-graph 毒 'poison' for an l-word for 'nourish' is unusual, though not totally phonetically implausible. Solution (1.13.00:40): 毒 was used for 'nourish' after the l- of 'nourish' had hardened to d-.	Scenarios 3-4: Problem: If 'poison' had d- or l- in Old Chinese, the use of its graph 毒 for a q-word for 'nourish' makes no sense, and 毒 'nourish' would have to e reconstructed as an unrelated word which happened to have a rhyme like those of the qUk-words: 育毓 m-quk (Baxter and Sagart 2011: 159) 谷 m-qok (not in Baxter and Sagart 2011, but reconstructed according to their system) Another problem (1.13.1:43): There is an Late Old Chinese dialectal word 陶 dou 'nourish' written with an l-phonetic 匋 in Fangyan (Schuessler 2007: 515). dou could go back to du or lu. The spelling 陶 'pottery' may have been devised on the spot, so it is not evidence for reconstructing lu. Reconstructing 育毓 luk (< √lwk* with zero grade?) 谷 lok (< √lwk* with -ə-grade?) with l- instead of q- allows me to group them together with 陶 lu (and 毒luk in scenario 4 but not 毒*duk in scenario 3) as members of a word family meaning 'nourish'.
毒 'poison' was Early Old Chinese Cʌ-luk > Middle Old Chinese luk > Late Old Chinese *douk	Scenario 2: Tibeto-Burman words for 'poison' with t- or d- were borrowed from Chinese after EOC l- hardened to MOC d-. Problem: How could a Chinese loanword spread throughout the Sino-Tibetan family, including members that had no known contact with Chinese? Solution (1.13.00:10): This contact occurred between Old Chinese and Proto-Tibeto-Burman, the ancestor of the non-Chinese Sino-Tibetan languages. Problem: Did Proto-Tibeto-Burman really exist, or is 'Tibeto-Burman' merely a convenient category of all ST branches other than Sinitic? 毒 'poison' had an l- at the time it was used to write an l-word for 'nourish'. (1.13.1:13: Or the *l- in both 'poison' and 'nourish' had hardened by the time the latter was written as 毒.)