When I first started my long series of Tangut posts almost exactly a year ago, I cited the Tangraphic Sea's graphic analysis for the DOG-CALENDRICAL graph for 2006:


TT0571 DOG-CALENDRICAL na 1.11 <

left of TT5353 DOG khywï 1.30 [semantic; cognate to Old Chinese 犬 khw...' 'dog' (rhyme uncertain)?, Written Tibetan khyi 'dog'?] +

left of TT4479 BLACK nyaa 1.21 [phonetic; cognate to Written Tibetan nag 'black']

However, I didn't cite the analysis for its accompanying heavenly stem, so here it is a year late:



< bottom [right] of TT0991 FIRE məə 1.31 (cognate to Written Tibetan me 'fire') +

right of TT5120 BRIGHT swew 1.43

Nishida (1966: 485) identified the left-hand component of HEAVENLY-STEM-III as the 'fire trigram radical' (离部) presumably because it appears on the left of


combined with the bottom right of

TT0881 WOOD syi 1.11

which has no semantic or phonetic connection to FIRE-TRIGRAM. (It would be pushing it to say that WOOD is used to make FIRE.)

The 'fire trigram radical' is semantic in HEAVENLY-STEM-III since that stem is associated with fire.

The right side of BRIGHT in H-S-III, identified by Kychanov (cited in Grinstead 1972: 15) as LIGHT (but by Nishida [1966: 244] as the 'front radical' [前部]!), is presumably also semantic, since BRIGHT swew 1.43 does not sound anything like H-S-III byi 1.11. LIGHT may be a refernce to the association of H-S-III with yang. (But then why isn't its yin counterpart H-S-IV a combination of FIRE and, say, SHADOW? And if Nishida's interpretation of the component as 'front' is correct, H-S-IV could have been a combination of FIRE and BACK, but it isn't.)

The function of the extra two strokes (ソ) over LIGHT in H-S-III is unknown. Why wasn't H-S-III analyzed as left of FIRE-TRIGRAM plus part of some tangraph with the ソ-version of LIGHT? Grinstead (1972: 98) lists 14 such tangraphs including

TT0778 BOWSTAND tsyir 1.79 (with WOOD on top)

TT1293 WHITE? phyiy 1.36 (gloss from Nishida 1966: 396)

TT1638 RED ryïy 2.37 (why is EIGHT 'yar 1.82 on the left side?; EIGHT also appears in TT1648 BRILLIANT byi 1.11 which surprisingly does not contain LIGHT or ソ-LIGHT)

TT2724 CHOOSE tsyiir 1.93

TT4001 LIGHTNING tshywu 1.3 (why is EARTH on the left side?)

TT4568 OTHER tsyiy 1.36

which have no single unifying semantic or phonetic characteristic.

None of these ソ-LIGHT tangraphs are homophonous with H-S-III, so ソ-LIGHT may not be an abbreviated phonetic element, unless it is an abbreviation for TT1293 WHITE? phyiy 1.36 above.

Next: Perceiving faces. THREE PIGS ON THE RIVER STYX

Happy new year! My six week hiatus from blogging - the longest I've ever had in over four years - is over.

Although the Tangut new year has not yet begun, I'm going to discuss the graphs for that year:


equivalent to Chinese 丁亥. As usual, the tangraphs are more complex than their Chinese counterparts. 丁 and 亥 are simple characters which cannot be broken down into smaller parts, but the same cannot be said of their tangraphic translations.

TheTangraphic Sea gives the following analysis for HEAVENLY-STEM-IV:



< left of ? TT0356 ?

+ left of TT0571 HEAVENLY-STEM-III byi 1.11

The third and fourth heavenly stems (= Chn 丙 and 丁; see the complete list here) are associated with fire, so it's not surprising that they both contain the bottom right of

TT0991 FIRE məə 1.31 (cognate to Written Tibetan me 'fire')

Nishida (1966: 392) called the left half of HEAVENLY-STEM-IV the 'arrange radical' (齊部), even though hisglosses for characters with that element don't have much to do with arranging (1966: 392-394):

([Chinese surname 王 Wang], SLAVE, (the syllable ngẽ), OBSTRUCT, MEASURE, WILD-MAN, KICK, TIRED, OPPOSE, BEND, EQUAL (another meaning of 齊), INVADE, HARD, SERVE, DAILY (< equally across a period of time?), STRONG, SPRING, STAIRS (steps are even?), SPEAR/MALLET, KILL, STRIKE, SHAKE, ENEMY, THAN, RUN-OVER, (Sanskrit syllable hi)

The meaning of TT0327, the apparent source of the left half of HEAVENLY-STEM-IV, is uncertain. Nishida (1966: 393) glossed it as REFUSE (拒む), Grinstead (1972: 140) glossed it as CLEANSE, and Shi et al. (2000: 333) glossed it as STOP (止) and PROTECT (护). TT0327 was pronounced something like te. Shi et al. approximated its reading as 丁 - the Chinese fourth heavenly stem! Thus it seems that HEAVENLY-STEM-IV is a compound of

(a) the left side of a tangraph that sounded like the Chinese name of the fourth heavenly stem

(b) the bottom right of TT0991 FIRE

The Tangraphic Sea analysis of PIG-CALENDRICAL is more straightforward:



< left of TT5360 PIG wa 1.17 +

all of TT2191 ? giu 2.3

Nishida (1966: 244, 460) called the left-hand component of PIG-CALENDRICAL and PIG the 'beast radical' (豸部). All of his glosses for tangraphs containing this element involve beasts (mostly dogs).

PIG wa 1.17 is probably cognate to Written Burmesewak 'pig'.

TT2191 ? giu 2.3 is clearly phonetic in TT5355 PIG-CALENDRICAL gyu 1.3. The meaning of TT2191 is not clear. Grinstead did not gloss it at all, but Nishida (1966: 479) glossed it as THREE-ROADS (三途, the Chinese counterpart of the river Styx) whereas Shi et al. (2000: 207) simply glossed it as THREE (三; a mistake for 三途?). In any case, TT2191 is not the normal tangraph for THREE. The components of TT2191 have the following Nishida numbers:

left: 242 (not glossed by Nishida; he lists other tangraphs with it on the left: WITHOUT-EXCEPTION [pron. tạ], DITCH [pron. ngyuH], [Chinese surname 趙 Zhao, obviously not pronounced ngyuH], and [a surname ngyuH] [1966: 478-479]; 242 is phonetic in some but not all of those tangraphs)

(Nishida's ng- and -H are equivalent to g- and -zero in the Gong Hwang-cherng reconstruction of Tangut used elsewhere on this site.)

right: 116 (glossed by Nishida as 'having to do with knowledge' [1966: 243])

Unfortunately, no Tangraphic Sea analysis of TT2191 has survived, so we can't know which tangraphs served as the sources of those components.

Next: Bright fire.

© Copyright 2002-2006 Amritavision