In the fifth installment of The Book of the Nisan Shaman there is a word “caise” mentioned as being a kind of food that is prepared for Sergudai Fiyanggo’s funeral.  Pawel looked up the word in P. Schmidt, “Chinesische Elemente im Mandschu. Mit Wörterverzeichnis” in: Asia Major volume 7, 1932, and found a strange Chinese character.  He referred me to the entry in the dictionary:


I couldn’t find that character anywhere, and neither could several people I asked, so I sent the entry to Victor Mair.  After several days of hunting, he finally tracked it down.  He says that when he finally spotted it in Hanyu Da Zidian, 5.3164, he “laughed / sobbed bitterly / mournfully”.  The following is a guest post by Victor Mair:

It turns out that the character 票 + 產 in Schmidt‘s dictionary


must be a corruption of 粟 + 產, i.e., 䊲.  Just so we keep things straight, let me go through all three of the elements in these two characters, although 票 + 產 doesn’t really exist — it’s a non-character, and 䊲 is quite rare:

票 PIAO4   bank note, bill; written certificate, etc.
粟 SU4      millet
產 CHAN4   produce, give birth

All right, 票 + 產 doesn’t exist, so it doesn’t mean anything and it doesn’t even have a sound attached to it.  We can forget about it, except as a joke that history has played on us.

䊲, on the other hand, has both a sound and a meaning, actually two sounds and two meanings:

CHAN4   rough-hulled rice
CHAN3   a. ground millet
b. rehusk (to husk again)

Now, it is generally agreed that 䊲 has been borrowed to represent the morpheme that is normally written with this character:  饊 SAN3 (“deep-fried dough twist”).  This often appears in the word 饊子 SAN3ZI, a word that occurs in Shuihu zhuan and Jin Ping Mei, hence probably early Shandong vernacular, but it is also already to be found in the Song period.  SAN3ZI was originally made of glutinous rice flour and wheat flour with a small amount of salt added.  In modern times, however, just wheat flour is used, and the strips of dough are fine as spaghetti, laid over each other like railings or palings.  Texts from the Yuan on often refer to SAN3ZI as a snack to go with tea.

Since Manchu has both caise and saise for this snack, I suspect that they reflect two different dialectal forms.  Writing 䊲子 seems to reflect caise, whereas writing 饊子 seems to reflect saise.  Obviously, in either case, denazalization has occurred.  The Manchu forms seem to have been based on some pronunciation that had the first syllable as chai rather than chan and as sai rather than san (maybe like Japanese rai, ran for 懶?)

I should note that there are two types of Manchu caise (alternatively, saise), namely, the white (Ma. xanyan caise) and the red one (Ma. fulgiyan caise), corresponding to Chinese bai sanzi 白饊子 and hong sanzi 紅饊子, being fried pastries used for religious offerings.  Zaxarov’s Manchu-Russian dictionary (p. 554) gives the Chinese form in Cyrillic, сань-цзы, which appears to be sanzi as well.

I guess Schmidt‘s informant, though cognizant of the Chinese sound,  was not sure if there may have been a standard Chinese character for it; or, perhaps, he wanted to distinguish it from the sanzi of the southerners or the sanzi that is popular among the Salar, Uyghur and other Muslim minorities; or, perhaps not unlikely, such a “vulgar” form did exist. Anyway, he picked up this form which, I suppose, was originally su 粟 + chan 產. Taking the initial s- of the first and the final –an of the second, we get the sound san.  The sign 票 in Schmidt‘s dictionary may be seen as a corruption of 粟. This may be similar to the trick–the “imperial trick” of Qianlong–whereby the Qing edition of buddhist texts has, e.g.,

Skr. tadyathā

transcribed as

答     達鴉 (combined)     塔阿 (combined).

I think this same kind of transcriptional “trick” was used for “spelling” Mongol, as in the Chinese version of the Secret History of the Mongols.
Now, since I’ve often seen this kind of snack in Central Asia — and even among the mummies thousands of years old — plus the fact that there are a number of variant Sinitic translations and transcriptions that refer to this kind of snack, I suspect that it is a borrowing from some Turkish or Iranian language (or maybe Tocharian!).  Eventually I’ll probably be able to figure that out too, but not right now.  I’m swamped with all sorts of other things that I’ve got to dig out from under.

Thanks to H. T. Toh for help with the Manchu.

One thought on “Caise”

Leave a Reply

Your email address will not be published. Required fields are marked *