<so nggiyan uliya aniya juwa juwe biya ice ilan inenggi>

'yellow pig year, ten one month, new three day'

(0. 12.29.0:15: I keep thinking the version of <ilan> above looks like Chinese 斗 <DIPPER>, but it is of course in fact cognate to Chinese 三 <THREE>.)

1. Via Andrew West: Abraham Gross' proposal to encode the missing kana <YI> and <WU> in Unicode. That reminds me to upload my August post about <YI> and <WU>.

2. I first heard the song "Year of the Cat" as a child in 1976, and only years later¹ did I learn that it was a reference to the Vietnamese zodiac which is close to the Chinese one with two exceptions:

• Sino-Vietnamese 丑 sửu for the water buffalo rather than the ox

• Sino-Vietnamese 卯 mão (mẹo in an earlier stratum of borrowing)  for the cat rather than the rabbit

The terms for the Vietnamese zodiac are not the normal terms for animals: e.g., in Vietnamese, 'water buffalo' is 𤛠 trâu and 'ox' is 𤙭 ~ 𤞨  bò.

I've long assumed that the reinterpretation of 丑 sửu as water buffalo incorporated a local animal, but water buffalo also exist in China too. Duh. In fact, China has seven times more water buffalo than Vietnam. Shows you what I know about farming: nothing. So I can't explain how sửu came to refer to water buffalo.

As for 卯 mão/mẹo, was its reinterpretation as 'cat' due to a folk etymological association with 貓 ~ 猫 mèo 'cat'?

¹In an interview with Al Stewart that I heard on the radio in 1989?

3. I never heard of screeves until today. The word sounds like it could be a native English word, but in this context it's actually a loan from Georgian მწკრივი cʼkʼrivi 'row, series'. I wonder why it's so Anglicized. It's not as if Japanologists speak of 行 gyō 'rows (of kana sharing the same vowel: e.g., a, ka, sa)' as gheow or however an English speaker might spell it. (It would be fun to ask English speakers unfamiliar with Japanese to write gyō phonetically.)

There turns out to be another screeve which isn't  native or from Georgian.

19.12.27.23:59: YELLOW PIG 12/2

<so nggiyan uliya aniya juwa juwe biya ice juwe inenggi>

'yellow pig year, ten one month, new two day'

1. Dept. of Ideas I Wish I Had: Alexander Zapryagaev's proposal for writing Old Japanese in hentaigana, a logical extension of the common practice of writing the extinct Japanese syllable ye (now [e]) in hiragana as the hentaigana 𛀁 to differentiate it from え e and ゑ we (also now [e]). (More in this thread by Sven Osterkamp.)

2. The reading ritsu for 立 <STAND> is in that stratum of Japanese that I feel as if I've 'always' known. I suspect I learned the reading in the early 80s when I started to read Japanese books with furigana.

When I started learning Korean in 1987, I immediately picked up on the correspondences between Sino-Korean and Sino-Japanese¹. For instance, I noticed that Sino-Korean -l regularly corresponded to Sino-Japanese -tsu or -chi and vice versa. So I should have expected ritsu to correspond to Sino-Korean 릴 ril. But of course, the actual Sino-Korean reading of 立 is actually 립 rip. I learned that reading so early in my studies that I didn't even know the correspondence patterns yet. Hence the mismatch of -p and -tsu didn't bother me at all.

Not long afterward I learned Sino-Korean 잡 chap corresponding to Sino-Japanese zatsu for 雜 <MIXED>.

And then I learned the Cantonese readings of those characters: lap6 and zaap6.

The next step was learning about Chinese reconstruction. Of course all agree that 立 and 雜 originally ended in *-p in Chinese, and that Cantonese preserves that *-p.

So how did the Sino-Japanese readings of 立 and 雜 come to end in -tsu? Alexander Zapryagaev has a thread on the mystery of 立 ritsu.

¹And Mandarin, but that's not relevant here, since Mandarin lacks final stops. Without knowledge of Mandarin, I would have had a much harder time remembering which Sino-Korean words ended in -ng.

12.29.20:35: How I guessed final consonants in Sino-Korean in 1987 (before I knew anything about Cantonese or Vietnamese):

At the time I just memorized which Sino-Korean readings ended in -p, since there was no way to guess Sino-Korean -p on the basis of Sino-Japanese or Mandarin even in regular cases such as

十 <TEN> SJ jū : Md shi : SK 십 ship

In that particular case, *-ip was borrowed into Japanese as *-ipu which became *-iu and then -ū.

Once I learned which Sino-Korean readings ended in -p and -m, I could use that knowledge to guess which Cantonese and Vietnamese readings ended in -p and -m.

19.12.26.23:59: YELLOW PIG 12/1

(I completed this post but lost it before I could upload it, so I reconstructed it on 12.30.16:13.)

<so nggiyan uliya aniya juwa juwe biya ice inenggi>

'yellow pig year, ten one month, new day'

1. The first ten days of the month are ice 'new' in the Ming Jurchen calendar. (In Jin Jurchen, the first day was 一日 emu inenggi 'one day'. Note how the early graphs are identical to Chinese 一日 <ONE DAY>.) Jin (1984: 105) derives the graph for ice from the left side 亲 of Chinese 新 <NEW>. But I think the Jurchen graph may be more directly connected to Chinese 𢀝 <NEW>, a variant of attested in the Jin dynasty dictionary 四聲篇海 Sisheng pianhai (The Four-Tone Text Sea).

2. In 1998 I reviewed William C. Hannas' Asia's Orthographic Dilemma for Korean Studies. I finally got around to reading a Kindle sample of the 2003 sequel The Writing on the Wall: How Asian Orthography Curbs Creativity.

A. East Asia has a "creativity deficit" (Kindle location 146)

B. Writing "affects thought" (Kindle location 245)

C. B causes A - in other words, East Asia writing systems cause a "creativity deficit"

A and/or B could be true. But I am skeptical of C.

19.12.25.23:59: YELLOW PIG 11/30

<so nggiyan uliya aniya juwa emu biya gūsin inenggi>

'yellow pig year, ten one month, thirty day'

gūsin 'thirty' looks like Janhunen's (2003: 397) Proto-Tungusic *gutïn from para-Mongolic or pre-Proto-Mongolic *gutïn. (The Proto-Tungusic form cannot be from Proto-Mongolic *gucin which underwent two changes: *ï > *i and *ti > *ci.) However, Proto-Tungusic *gutïn should become Jurchen gutin, not gusin.

I propose that Jurchen gūsin may be a borrowing that replaced an earlier *gūtïn inherited from Proto-Tungusic. (The macron in Jurchen does not symbolize length; it indicates that u is [ʊ].) The source of Jurchen gūsin may be a para-Mongolic (Khitan?) dialect that shifted *c to sh (unlike the prestigious Khitan dialect preserved in the small script that retains c).

<THIRTY>

is a graphic cognate of Jurchen

<THIRTY>

(12.26.13:19: Left to right: the earliest form from Nüzhen zishu [Book of Jurchen Characters, c. early 12th c.?], variant in 慶源 Kyŏngwon inscription, 1138-1153,  進士 jinshi candidate list, 1224, Berlin copy of the Bureau of Translators vocabulary, 15th c. It is interesting that the early and late forms are more similar to each other than to the forms between them.)

and sounded something like Jurchen gūsin, though there is no evidence for its pronunciation.

2. When I was studying Russian in the late 90s, I was surprised that 'Kremlin' was Кремль <Kreml'> without an n. I asked my professor why and ... I can't remember his answer. Today I learned from Wiktionary that there is an Old East Slavic кремлинъ <kremlinŭ> with -n-. But how did that n-form enter English? Not directly, I assume.

etymonline says:

1660s, Cremelena, from Old Russian kremlinu, later kremlin (1796), from kreml' "citadel, fortress," a word perhaps of Tartar origin. Originally the citadel of any Russian town or city, now especially the one in Moscow (which enclosed the imperial palace, churches, etc.). Used metonymically for "government of the U.S.S.R." from 1933. The modern form of the word in English might be via French.

The un-Turkic initial cluster kr- makes a Tatar (not 'Tartar') origin improbable. The Russian Wiktionary derives kreml' from Proto-Indo-European *kʷrom 'fence'.

12.26.10:09: Merriam-Webster says:

1662 [...] obsolete German Kremelien the citadel of Moscow, ultimately from Old Russian kremlĭ

That gives the impression that German added the -n (but why?).

19.12.24.23:23: YELLOW PIG 11/29

<so nggiyan uliya aniya juwa emu biya orin uyewun inenggi>

'yellow pig year, ten one month, twenty nine day'

orin uyewun 'twenty nine' is a para-Mongolian (Khitan?)-Jurchen hybrid. Compare with Written Mongolian qorin yisün 'twenty nine' containing an unrelated Mongolian word for 'nine'.

Jurchen uyewun is trisyllabic unlike any other Tungusic word for 'nine' at starling other than Negidal ijeɣin with different first and third vowels. Neghidal i can correspond to Jurchen/Manchu u: e.g., N edin : J/M edun 'wind'. I have long assumed that Manchu uyun is a contraction of uyewun. That contraction already existed before Manchu got that name since the Ming dynasty Bureau of Interpreters vocabulary has disyllabic uyun (transcribed 兀容). The roughly contemporaneous trisyllabic uyewun (transcribed 兀也溫) in the Ming dynasty Bureau of Translators vocabulary may be more carefully pronounced and/or from a different dialect.

It's already Christmas in most of the world as I write this, so as a 'gift' to my readers, I'm uploading all the posts I wrote over the last month but had kept on my computer until now:

• 19.11.23: The Etymology of Cantonese 'Tongue'

• Posts from 19.11.24-11.30

• Posts from 19.12.1-12.7

• Posts from 19.12.8-12.14

• Posts from 19.12.15-12.21

• 19.12.22: Yellow Pig 11/27

• 19.12.23: Yellow Pig 11/28

I've been too tired and busy to upload posts late at night.

19.12.23.23:57: YELLOW PIG 11/28

<so nggiyan uliya aniya juwa emu biya orin jakūn inenggi>

'yellow pig year, ten one month, twenty eight day'

1. orin jakūn 'twenty eight' is a para-Mongolian (Khitan?)-Jurchen hybrid. Compare with Written Mongolian qorin naiman 'twenty eight' containing an unrelated Mongolian word for 'eight'.

Jurchen jakūn 'eight' has not changed much from Proto-Tungusic *japkun whose first syllable *ja looks like Proto-Japonic ya 'eight'. Coincidence? How many other instances of Proto-Tungusic intervocalic *j- correspond to Proto-Japonic *y-?

If one wants to link the Tungusic and Japonic words for 'eight' via borrowing, one must deal with the complication of working out a scenario of Tungusic-Japonic contact (see yesterday's post) and with the question of why Tungusic has *-pkun and Japonic doesn't. Proposing a genetic relationship eliminates the contact problem but still doesn't resolve the *-pkun problem.

It may be tempting to link early Korean *yʌtʌrp (Lee and Ramsey 2011: 160) to the Tungusic and Japonic words, but that raises even more problems: e.g., what is *tʌrp?

2. The current state of Korea-Japan relations in a slogan:

• 안가 an ka '[I] don't/won't go [to Japan]'

• 안사 an sa '[I] don't/won't buy [from Japan]!'

(1.2.15:51: Corrections by Kongduino.)

The verbs appear to be bare stems but are actually a-stems that have absorbed an -a ending that Martin (1992: 466) calls the 'infinitive'. But I would rather not use the term 'infinitive' for the ending of a finite verb.

The -a ending is more obvious in forms like 봐! pwa! 'look!' (< po-a) and 팔아! phar-a 'sell!'

3. I was surprised to learn from Martin et al. (1967: 870) that sa- 'buy' is also an "old-fashioned" term for 'sell (grain)', so ssar-ŭl sa-da 'rice-ACC X-STATEMENT' can be either 'buy rice' or 'sell rice'.

19.12.22.23:50: YELLOW PIG 11/27

<so nggiyan uliya aniya juwa emu biya orin nadan inenggi>

'yellow pig year, ten one month, twenty seven day'

1. orin nadan 'twenty seven' is a para-Mongolian (Khitan?)-Jurchen hybrid. Compare with Written Mongolian qorin dologhan 'twenty seven' containing an unrelated Mongolian word for 'seven' with the numeral suffix last seen in jirghughan 'six'.

Jurchen nadan 'seven' can be projected intact all the way back to Proto-Tungusic. Proto-Tungusic *nadan looks like Proto-Japonic *nana 'seven'. Coincidence? How many other instances of Proto-Tungusic intervocalic *-d- correspond to Proto-Japonic *-n-?

What complicates a loan scenario is uncertainty over whether the two proto-languages were in contact. I think Tungusic and para-Japonic languages might have been in contact in Parhae, but that's centuries after the ancestor of Japonic spread from the Korean peninsula to the Japanese islands.


2. I just heard Muir pronounced as [mjʊɚ] which is what I'd expect for a theoretical Miur. Wiktionary lists a General American /mɪɚ/. I have never heard the name pronounced before. I thought it was homophonous with Moore in English. Wiktionary lists five (!) pronunciations for Scots muir 'moor': [møːr], [myːr], [meːr], [miːr], [mjuːr].

3. I also heard Buttigieg pronounced for the first time as [ˈbuːtɪdʒɪdʒ]. I had been mispronouncing it as [ˈbuːtɪdʒɛg], thinking gi was like Italian [dʒ]. Turns out both g's are Maltese ġ [dʒ] and ie is [ɨː] (according to Wikipedia's IPA for Maltese page) or [ɪː], [iɛ], or [iː] (according to Wikipedia's Maltese language page). In any case, ie is from ā, and so I'm not surprised to learn that Wiktionary says Buttiġieġ is from Arabic أبو الدجاج <ʔˀbw ʔldjʔj> ʔabū ad-dajāj, lit. 'father [of] the-poultry' with ā.

The bending of ā to ie in Maltese reminded me of the raising of Old Chinese *a to *ie and various high vowels and convinced me that Norman's pharyngeal hypothesis for Chinese was right. In my take on his hypothesis, pharygealization pushed vowels down, whereas vowels raised in its absence. But David Boxenhorn made me think  pharyngealization might not be a factor; vowel harmony alone might trigger vowel lowering and raising. And vowel harmony is a well-attested phenomenon in north Asian languages.

19.12.15-19.12.21