Another irrelevance from Matisoff:
"
(6) FIRE/BURN/SHINE/BRIGHT/LIGHT
This `spectacular' word-family was first presented in JAM 1997
(`Laryngeals').
First of all, I am claiming that three sets reconstructed separately
in STC16 (#220; #221; pp.172, 174) are all really allofamically
related:
STC #220: *bar/*par, revised to *bwa^r <> *pwa^r
Nung hwar `burn, kindle'; Jingpho /wàn, Moshang var, Garo
wa/l `fire'; Chairel phal (< *par), id., and Written Tibetan b8ar-
ba `burn, catch fire, sbor-ba `light, kindle', Kanauri bar `burn',
Miri par `light (fire)'
STC #221: *hwa-t
Bahing hwa `light', Lepcha om `shine', om-bo `illuminating', a-
om `light', brightness; WT o8d `light, shine, brightness', nyi-
o*d `sunlight; Written Burmese ne-at `sunlight', Thado wat `shine'
For similar alternation between pw- and hw-, cf. WB phwak <>
hwak `hide'.
STC pp.172, 174: PTB *pwa˘r `white' Lushai va˘r `white'; Proto-Karen
*/(b)wa `white'
To these we may then add a number of new supporting forms with
laryngeal initials which point to slightly different but related
prototypes:
*hwa: Apatani hú-tò `light (n.)'; Chepang ha/-/o `shine'; Kulung ha-
me `shine'; Ntenyi wＵ-ghＵ
*hya: Sangtam a-vi-sa `bright light'
*hwat: Damu (AMD) wat `glimmer'; Limbu o:tt-, o:ts- `burn, give light,
shine'; Manang wE1
*hyat: Dumi hÆt-nÆ `burn', Bahing hOt- `id.'
*hant: Limbu ha:nd- `light (lamp, cigarette)'
*yat: Chairel id `burn, catch fire'
*(h)wan: Tangkhul han `shine'; Lotha and Mao won `id.'; Milang a-
un `bright, light'; Damu wun-pit-dung; Chang wan `fire', Yacham
-Tengsa wa-si `id.'
*hwam: Lepcha om, etc. [STC]; Chepang hyum/-sa `burn, scorch'
*(h)(w)a,: WB wa^n, `yellow'; Tagin ong-ka-nam `shine', hung `id.';
Konyak wang-ngai `bright light'
*yan: Bokar Adi a-jen, Bokar a-en `shine'
*(h)(w)al: Thulung hal <> ul `heat slightly', wal `boil lightly';
Lushai ha^l `light, ignite'; Tiddim ha˘l `burn'; Tangsa (Yogli)
wal `fire'; Dimasa wai
*(h)(w)ar: Tangkhul hor `shine'; Maring war `bright light'; Limbu
haqr- burn, alight'; Limbu o/r-u, o/-ma `fire'; Mikir ar-nu `roast,
bake, grill'; Thakali ur `yellow'; Gurung (Ghachok) ur-gya˘ `id.';
Chepang yar-o `yellow'; Magari or-khe `id.'; Hayu ho `id.' (Lushai
va˘r `white' belongs here)
*yar: Ao (Chungli) yar `shine'; Khaling ehr-nya; Yimchungru
yin `kindle'
We can summarize these relationships by constructing a `pan-allofamic
formula'
(PAF), as follows:
h
w N t
a
y r
p/b
l
On the Chinese side, several promising comparanda are available.
(a) STC (pp.172, 174) already identified PTB *pwa:r `white' with:
皤17 [GSR 195r] *b'wâr <> *pwâr [WHB] *paj > pa > bo: <> *baj > ba >
pó
(b) STC (p.172) identifies set #220 with two Chinese words:
火 [GSR 353a-c] *Xwa˘r `burn; fire' [WHB 1067, 1216, C764.27] *hmïj?
>xwax > huô
燔 [GSR 195i] *b'ywan `burn; roast' [WHB C756.19] *bjan > bjon > fán
Another good candidate for relationship (the two characters are
graphic variants of
each other):
煇 `flame; bright(ness)' GSR 458k] Xiw&r <> g'yw&n
輝 `bright' [GSR 458-l] Xiw&r
[WHB:C764.11] OC *hwjïj > MC xjwïj > Mand. hui
For the semantics, cf. PIE *bhel- `shine; flash; burn; shining white
and various bright colors' > Eng. black, blank, blanch, bleak, bald,
bleach, blue, blaze, blind, blend, blond, blink, etc.18
Finally, I would like to include the Chinese word for `moon', one of
the `bright beings of the night', in this word-family:19
月 `moon' [GSR 306a-f] ngywat
[WHB 970] OC *ngWjat (or *Nwjat ?) > MC ngjwot > Mand. yuè
14 Chin reflexes like this motivate STC's reconstruction of a long
vowel.
15 French 1983 sets up Proto-Northern Naga *C-we˘k as the ancestor of
the Konyak and Chang forms, postulating metanalysis of a compound
like the Mikir form, above.
16 For a similar effort, see my paper, `Universal semantics and
allofamic identification -- two Sino-Tibetan case-studies:
STRAIGHT/FLAT/FULL and PROPERTY/LIVESTOCK/TALENT' (1988), where I
grouped into the same word family two other separate STC roots, both
reconstructed *dyam (#227 `straight' and #226 `full, fill').
17 The Guangyun glosses the meaning of this character as `white-
haired appearance of the elderly'. It has two readings, one the same
as `grandmother' and the other the same as `wave'
18 What have I done! Now Greenberg has more ammunition for his Proto-
World speculations.
19 Cf. Matisoff 1980. Admitting this word into the family would
require adding N- to the initial consonant slot of the PAF. Perhaps
it is a `rhinoglottophiliac' outgrowth of the original PST
initial h-. See Matisoff 1975.
"
Torsten