74. I am not sure how to parse this line because (1) I can't find the first or last two tangraphs as a disyllabic surname in Kychanov (2006) or Li Fanwen (2008) and (2) the third and/or fifth tangraphs could be an adjective modifying the surnames that precede them. Some possible translations besides the one in the table below:

Lwyr, De, Lu, the black Va

Lwyr, the beautiful De, the black Va

the beautiful Lwyrde, the black Va

the Lwyrde, the Lu, the black Va


The black Va must be parallel to the white puppy in the previous line. If this line is completely parallel to the last one, then should the last three tangraphs be translated as 'the black Luva'?

Tangraph number 366 367 368 369 370
Li Fanwen number 2755 2463 2488 5156 0176
My reconstructed pronunciation 2lwəʳ 1die 2lɨuu 1va 1nɨaa
Tangraph gloss the surname Lwyr sage; the surname De the surname Lu; first and second syllable of surnames; beautiful the surname Va; first syllable of the surname Valhew; transcription tangraph black
Word the surname Lwyrde? the surname Vana?
Translation Lwyrde, Lu, Vana (?)

366: 2755 (analysis unknown) may be a semantophonetic compound:


2755 2lwəʳ 'the surname Lwyr' (dexbisbescin) =

2888 2mə 'surname' (dexpux; semantic) +

5412 2lwəʳ 'ceremony; to get a haircut' (birbescin; bir = bis; phonetic)

Could the surname Lwyr be like the English surname Barber?

The combination bisbes/birbes is unique to 2755 and 5412. Is bisbescin/birbescin derived from a tangraph with bis/bir plus a tangraph with bescin?

367: 2463 'sage' (bilcukdex) doesn't contain geo 'sage' but does contain dex 'person'. It is the only tangraph with cuk. The function of bil and cuk are unknown and dex may not be 'person' here.

Note how the parts of the analysis don't quite add up. The lower left stroke of cuk is not accounted for:


2463 2die 'sage' (bilcukdex) =

1177 2die 'toil; tired' (baxdilpel; phonetic)

1338 1dzəu 'to love' (bixdiibaedex)

368: 2488 (bilfikdex; analysis unknown) looks like bil (function and source unknown; also in 2463 above): plus its homophone 0542 2lɨuu 'beautiful' (semantic/phonetic?):


369: Were the Va related to the Ru and Bu or families containing those syllables in their names?


5156 1va 'the surname Va' (pikcok) =

5489 2riuʳ 'the surname Ru; first or second syllable in surnames' (pikpux; pik = 'hand') +

1925 1bɨu 'the surname Bu; first or second syllable in surnames' (baedumcok)

370: See this post. THE GOLDEN GUIDE: LINE 73: TANGRAPHS 361-365

73. This line breaks the mold of previous lines. At first glance, the last three tangraphs represent a phrase, not two surnames. What is 'white puppy' doing in a list of names? Is it actually two surnames Kyta 'puppy' and Phon 'white' (which would still not follow the earlier pattern of a monosyllabic surname followed by a disyllabic surname)?

Tangraph number 361 362 363 364 365
Li Fanwen number 5964 5849 3732 1312 1572
My reconstructed pronunciation 2tsha 2zəi 1kə 1ta 1phɔ̃
Tangraph gloss the surname Tsha; transcription tangraph the surname Zi first half of 'puppy' second half of 'puppy' white
Word the surname Tshazi puppy white
Translation Tshazi, white puppy,

361: 5964 (analysis unknown) has the unusual left-hand radical (resembling gux; see below) due found in only three other tangraphs. The remaining components (helcun) form an independent tangraph that must be phonetic in 5964:


5964 2tsha 'Tsha' < 1273 2tsha 'wisdom' (hel 'know' + cun 'language')

362: 5849 has a left-hand radical gux 'small' resembling due in the preceding tangraph:


5849 2zəi 'the surname Zi' =

5036 2zəi 'baby' (biogux; phonetic; gux by itself is 1tsiə)

2132 2ʔiew 'achievement' (dexgoldal; semantic?)

Were the Zi a family of achievers?

363: The first half of 'puppy' has a circular analysis. It is a phonetic in its 'sources' and therefore must predate them:


3732 1kə (first half of 'puppy') (balqur) =

2535 1kə 'a kind of insect' (balquojio) +

4406 1kə 'bramble' (boxbalqur)


2535 1kə 'a kind of insect' (balquojio) =

3732 1kə (first half of 'puppy') (balqur; qur is a non-right-hand variant of quo) +

1887 1kie 'insect' (jio)


4406 1kə 'bramble' (boxbalqur) =

4250 1si 'tree' (boxdexdexcok) +

3732 1kə (first half of 'puppy') (balqur)

364: Did the Tangut really consider puppies to be narrow dogs, or does yir in 1312 come from a different tangraph?


1312 1ta (second half of 'puppy') (boiyir) =

1200 1khwiə 'dog' (boipamfak; cognate to Chn 犬) +

3287 1ruʳ 'narrow' (dexyir)

365: 1572 has a circular analysis. I presume 'white' was created before its 'sources':


1572 1phɔ̃ 'white' (feihon) =

0468 1phɔ̃ 'pair' (feihul) +

1787 1ʃwɪ 'shining white' (hontak)


0468 1phɔ̃ 'pair' (feihul) =

1572 1phɔ̃ 'white' (feihon) +

5855 2lọ 'twin, both' (guxhul)


1787 1ʃwɪ 'shining white' (hontak) =

1572 1phɔ̃ 'white' (feihon) +

5057 1ɣɛ 'true, real' (biohoodim) THE GOLDEN GUIDE: LINE 72: TANGRAPHS 356-360

72. 2335 may be an error for the similar tangraph 4583, since I can find 3429-0789 as a surname in Li Fanwen (2008: 134) but not 2335-0789. Kychanov (2006: 554, 636) lists neither disyllabic sequence.

Tangraph number 356 357 358 359 360
Li Fanwen number 2335 0789 5551 4788 5044
My reconstructed pronunciation 1ʃɨõ 2ɣʊ 1gəəu 2lew 1la
Tangraph gloss second half of the surname Tshenshon; beautfiul second half of the surname Thyghu; fox (borrowed from Chn 狐) the surname Gu the surname Lew the surname La
Word the surname Shon'ghu? (Thyghu?) the surname Lewla
Translation Shonghu? / Dyghu?, Gu, Lewla

356: 2335 is part of a circular analysis. Were the Shon related to the Ghy?


2335 1ʃɨõ 'the surname Shon' (geodal) =

2338 1ɣiəʳ 'the surname Ghy' (geobambilcok) +

0177 1raʳ 'surname' (daidexdal)


0177 1raʳ 'surname' (daidexdal) =

2474 2raʳ (dexdaidex; phonetic) +

2338 1ɣiəʳ 'the surname Ghy' (geobambilcok; not -dal!)

8.11.1:52: 2335 resembles two other tangraphs:

3429 has fak instead of dal on the right:


3429 1sew (geofak; < Chn 羞) =

3435 1ɣʊ 'god' (geogii) +

5921 2zaʳ 'ashamed' (juxfak)

3115 is like 3429 but with on the left:


3115 2thiə (first half of the surname Thyghu) (gerfak) =

2888 2mə 'surname' (dexpux; semantic) +

3437 2ʔõ 'the surname On' (geopor; geo contains bae, the right of ger) +

3150 2thiə (first half of 2thiə-2ɣʊ 'emperor'; see this post) (dexgiifak)

357: Are the Thighu of Xiongnu ancestry and related to the On (see above) and the Tangut imperial family?


0789 2ɣʊ (second half of Thyghu) (belduucin) =

2790 2d (first half of 2diə-1kwiəəʳ 'Xiongnu'; see this post) (dexbeldilcin) +

2813 2ɣʊ (second half of 2thiə-2ɣʊ 'emperor'; see this post) (geoduucin)

Thyghu is homophonous with 'emperor' and could be the Tangut counterpart to European surnames like King, Rey, Kaiser, etc.:


2thiə-2ɣʊ 'the surname Thyghu' = 2thiə-2ɣʊ 'emperor'

358: 5551 has a circular analysis with two homophonous phonetics:


5551 1gəəu 'the surname Gu' (weagii) =

5138 1gəəu 'one of a group of six gods' (weafexgeodex) +

5083 1gəəu 'to patrol, to make a detour' (weagiidal)


5083 1gəəu 'to patrol, to make a detour' (weagiidal) =

5551 1gəəu 'the surname Gu' (weagii) +

3852 1dʒɨe 'to go, parade' (qordal)

359: 4788 (analysis unknown) looks like a horned hat (bio; from 4940? see below) atop

1043 1tõ 'insect' (tuu)

Tangut l- is partly from *t-. Gong's reconstructed reading 1tow for 1043 has a rhyme closer to 4788 1tew than my 1tõ or 1ton in an Arakawa-style reconstruction. If Gong is correct, there is less doubt that 1043 1tow? is phonetic in 4788 1lew. The evidence for nasality in my -õ / Arakawa's -on is ambiguous.

360: 5044 may have a horned hat from the same source as 4788:


5044 1la 'the surname La' (bioguxdex) =

4940 2ʔiə 'the surname Y' (biotumbee) +

5881 1la 'small' (guxdex; phonetic)

Could the name 5044 be derived from 5881? Cf. surnames like Small, Petit, Klein, Maly. THE GOLDEN GUIDE: LINE 71: TANGRAPHS 351-355

71. I can't find 5288-1013 as a disyllabic surname in Kychanov (2006) or Li Fanwen (2008), but I'm guessing that it is one by analogy with previous lines.

Tangraph number 351 352 353 354 355
Li Fanwen number 5288 1013 0533 3168 2838
My reconstructed pronunciation 2khie 1tʃɨu 1gaa 1miəə 1tew
Tangraph gloss happy, cheerful the surname Chu the surname Ga woman; first syllable of surnames the surname Tew
Word Khechu Mytew
Translation Khechu, Ga, Mytew,

351: 5288 (analysis unknown) is a component in 2480 which is in my many birthday wishes on this blog:


2480 1bɛɛ 'amusement, happy' (dexweadex) =

0522 1dziẽ 'quiet' (qisdex; < Chn 靜) +

5288 2khie 'happy, cheerful' (weadex)

352: 1013 is symmetrical but its left and right sides are derived from different tangraphs. I can't make any semantic sense out of the analysis beyond 'the Ba that have men' (but wouldn't any clan have men?).


1013 1tʃɨu 'the surname Chu' (haxbaehax) =

1085 1zi 'man, male' (haxtun) +

2753 2lheew 'to have, exist' (gerdoecin; ger = dexbae) +

1637 2ba 'the surname Ba' (haxyax)

353: 0533 has a variant 0555. Were the Ga a thin family?


0533/0555 1gaa 'the surname Ga' (foodex/fomdex) =

1929 1tew 'thin' (baefomdal; semantic?) +

2888 2mə 'surname' (dexpux; semantic)

354: All of these tangraphs share bix and has in common. Could bix ... has mean 'woman'? The trouble is that no bixhas tangraphs other than 3168 mean 'woman' (though 1214 bumbixhascin 'wonderful' can be translated as Chn 妙 with 女 'woman' on the left).


3168 1miəə 'woman' (dexbixhascin) =

3368 1twị 'young' (dexbixbuohas) +

0960 1miẹ 'woman' (bixbuohascin; cognate to 3168)

355: 2838 and 0533/0555 have very similar analyses. 2838 has the right side of 2888 whereas 0533/0555 has the left side. Could only certain families have names written with pux as opposed to dex, or vice versa? Could those radicals indicate family status?


2838 1tew 'the surname Tew' (baefompux) =

1929 1tew 'thin' (baefomdal; phonetic) +

2888 2mə 'surname' (dexpux; semantic) THE GOLDEN GUIDE: LINE 70: TANGRAPHS 346-350

70. I can't find 3281-1670 as a surname in Li Fanwen (2008) and Kychanov (2006) doesn't even have an entry for 3281, but I'm guessing that it is a disyllabic surname like others in previous lines because 'give birth to' + 'grind' makes no sense.

Similarly, I can't find 2835-1389 as a surname in either of those dictionaries which list a nearly homophonous surname 2835-3201 2vɨu-2di (without a tense vowel in the second syllable) instead. Are 2vɨu-2di and 2vɨu-2dị variants of the same name? Again, I tentatively assume that this is a disyllabic surname by analogy with previous lines of the Golden Guide.

Tangraph number 346 347 348 349 350
Li Fanwen number 3281 1670 0745 2835 1389
My reconstructed pronunciation 2zwị 1swie 2vɨe 2vɨu 2dị
Tangraph gloss to give birth to to grind, pestle the surname Ve the surname Vu the surname Di; second syllable of surnames
Word the surname Zwiswe? the surname Vudi?
Translation Zwiswe, Ve, Vudi

346: 3281 was analyzed as a semantic compound:


3281 2zwị 'to give birth to' (dexqoe) =

2295 2gii 'to give birth to' (dexjei; semantic) +

4057 2dʒɨe 'to give birth to' (buuqoe; semantic)

Why so many apparent synonyms? How did these three words differ?

347: Why add 'move'? Wouldn't 'grind' plus 1swie be sufficient? Is her really a combination of the top of has plus bil?


1670 1swie 'to grind' (herpak) =

0398 2swio 'to grind' (hasdexcok; semantic; obviously cognate) +

4174 2miu 'to move' (boxcuobil) +

3576 1swie 'clear' (dexduupak; phonetic)

348: Were the Ve known for speed?


0745 2vɨe 'the surname Ve' (besbaxbilbaecin) =

0753 2vɨe 'face' (beshercin) +

0744 bõ 'rapid' (duubaecin)

349: 2835 'the surname Vu' (analysis unknown) looks like ㅑ 'house' + the surname Pạ (written as 'grass' + 'surname'). Were the Vu related to the Pa? Did they serve within the house of Pa?


350: The analysis of 1389 is unknown. It can combine with the last two tangraphs from the last line:

2məi-2dị 'the surname Midi'

1khie-2dị 'the surname Khedi'

8.9.0:20: Was -2dị a surname suffix that was once an free morpheme? If it was, it was no longer in use in Tangut, since I can't find any -2dị that would fit the formula 'the ... of the Vu / Mi / Khe / etc.'.

Tangut fonts by Mojikyo.org
Tangut radical font by Andrew West
All other content copyright © 2002-2010 Amritavision