13.5.11.21:59: J_(R)_EN

In "Back on Track" and "The Vowels of the Central Capital", I reconstructed Jurchen

~

as <jū> [tʂʊ] with [ʊ] instead of [u] as in Jin Qizong's reconstruction dʒu.

?<jū.še> 'Jurchen' (Yongning Temple Stele, 1413)

which also has the spelling

?<ju.šen> (Sino-Jurchen vocabulary of the Bureau of Translators; cf. Manchu jušen 'serf, Jurchen')

So should I reinterpret the latter as <jū.šen>? I would rather not because the vowel sequence <ū ... e> violates Manchu vowel harmony (which I project back into Jurchen). Manchu vowels are 'yin' (higher), 'yang' (lower), or neutral:

The Manchu vowel system is similar to the Middle Korean vowel system:

In standard written Manchu, u ... e (a yin-yin sequence) and ū ... a (a yang-yang sequence) are possible, but not ū ... e (a yang-yin sequence). (However, the yin-yang sequence u ... a is possible only if u is after a nonuvular consonant and is hence from an earlier *ū: e.g., juwa < *jūwa 'ten'. 5.12.3:43: Moreover, the yang-yin sequence ū ... e is in early [i.e., nonstandard] written Manchu Jūsen [sic!]. I will discuss it next time.)

If ū ... e wasn't possible in Jurchen, what was the original Jurchen ethnonym?

In Classical Mongolian, the Jurchen are the Jürčid with the yin vowel ü. Mongolian also has yin/yang vowel harmony (though this is traditionally interpreted in terms of palatality rather than height):

If Jurchen yin/yang categories were preserved in Mongolian borrowings, then the Jurchen original was *Jurcen with yin vowels. (-d is a Mongolian plural marker that replaces stem-final -n.) Perhaps

was originally <jur.cen> or <jur.šen>. (It is not known whether *rc shifted to *rš before or after the Jurchen script was developed c. 1120. That certainly must have happened after the name was borrowed into Mongolian.)

At this stage,

<jur> ≠ <jū>

were not homophonous. But by 1413, both -r- and the u/ū distinction after nonuvulars were lost, so the two characters became interchangeable homophones, and the old character for <jū> was used to write a syllable that had once been <jur>:

<jū.še> for Juše [tʂuʂə] (not *Jūše *[tʂʊʂə]!) < Jurše(n) < Jurce(n)

The choice of

for še in 1413 does not necessarily imply that this character was once read rše, rce, or ce; it merely indicates that its initial consonant and vowel were identical to those of

<šen> < <cen>?

by 1413.

Oddly, if the above scenario is correct, the spelling

<jur.cen> or <jur.šen>

is more conservative even though I don't know of any attestations prior to circa 1500 (Pelliot's estimated date for the Sino-Jurchen vocabulary of the Bureau of Translators).

5.12.2:55: The fact that

<jur>

has only been found in the word for 'Jurchen' so far may indicate that it originally represented a syllable less common than <ju> (i.e., <jur>). Perhaps it was initially a logogram <jurcen> or <juršen> for 'Jurchen' and

<šen> < <cen>?

was added later as a phonetic clarifier.

I thought

<ju>

might be the phonogram for ju, but it is attested in isolation (i.e., not as part of a longer word) in the Nüzhen zishu (Jurchen Character Book) which may be a list of logograms. (That's the impression I got from the description of Nüzhen zishu in Kane 1989: 9. I have not yet studied it myself.) I suspect it might have originally been a logogram for juhe 'ice' which was later written with two characters as

<ju.he>

The second character <he> might be a phonetic clarifier.

<jū jur(cen?) ju(he?)>

are the only characters with the reading dʒu (= my ju) in Jin (1984). If the latter two were initially logograms, then perhaps

- the first character <jū> was a phonogram for both ju and jū

- the u/ū-merger after j was already complete by the time the Jurchen script was devised, so that first character was the only phonogram for ju < *ju and *jū

- the <ju>-phonogram could be among the hundred or so Jurchen characters without any known readings; ju/jū merged early, and that character was then replaced by <jū>