14.9.6.20:39: THE M-ISSING COMMISSAR

Today I read about Genrikh Lyushkov (1900-1945?), whose title brought to mind a question that I've had for a long time: why was German Kommissar borrowed into Russian as комиссар komissar? Why is an m 'missing' from that word and команда komanda 'team' (< French commande)? Was there a rule to simplify sequences of identical consonants at prefix-root boundaries in spellings of loans?

Latin com-missarius > R komissar

Latin com-mendare > R komanda

(The Latin forms are for root identification only and do not necessarily match the later forms' parts of speech, etc.)

But what about

Latin com-mercium > F commerçant > R коммерсант kommersant (not *комерсант komersant) 'merchant'

and Latin com-mutator > R коммутатор kommutator (not *комутатор komutator) 'switchboard'?

That rule obviously does not apply to native words with secondary sequences resulting from syncope: e.g., введение vvedenie 'introduction' < въведеніе vŭ-vedenie 'in-leading' ≠ ведение vedenie 'leadership'.

Sequences of identical consonants within a root word remained intact: e.g., the -ss- of missarius and the -mm- of communis (hence R коммунизм kommunizm 'Communism').

Elsewhere in East Slavic, although both Belarusian and Ukrainian have phonemic gemination, all of the above loanwords (presumably borrowed from Russian) lack geminates:

R комитет komitet / B камітэт kamitet / U комітет komitet does not fall into this category since its French source comité (< English committee) already lacked the double consonants of Latin committere.

Why does Serbo-Croatian комесар komesar have an -e- instead of an -i-?

9.6.20:48: And why did Old Latin comoine(m) become Latin communis 'common' with -mm-?

14.9.5.23:55: LIANMA, LINGMO, LINYIN

Most Chinese character spellings of foreign place names in Japanese can be explained in terms of Chinese and/or Japanese readings.

One baffling exception is 布哇 Hawai 'Hawaii' which makes no obvious sense in either Chinese or Japanese. I have written about it thrice (2008, 2010, 2012). I can't think of a better explanation than what I proposed in 2012.

I discovered what initially appeared to be another exception tonight in Yamamoto (2009: 81): 嗹馬 Denmāku 'Denmark' which would be read as Lianma in Mandarin and as *Renba in normal Sino-Japanese. I didn't think it could have been created by a Japanese speaker because Japanese not only has [d] but also has characters pronounced [den]. Mandarin, on the other hand, has no [d] (what is romanized d is actually an unaspirated [t]). So was voiced l intended to be a substitute for voiced d? Apparently it was, as it and similar Chinese names for Denmark turn up in the 1852 edition of the 海國圖志 Illustrated Treatise on the Maritime Kingdoms by 魏源 Wei Yuan:

嗹國 Lianguo (guo is 'country')

領墨 Lingmo (-ng would be an acceptable substitute for -n to a speaker of a Chinese variety like Shanghai without an -n : -ng distinction)

吝因 Linyin (I have no idea what -yin is doing)

9.6.0:36: The Japanese may have taken the spelling 嗹馬 from the Treatise given its influence in Japan:

Wei's work was also to have a later impact on Japanese foreign policy. In 1862, samurai Takasugi Shinsaku, from the ruling Japanese Tokugawa shogunate, visited Shanghai on board the trade ship Senzaimaru. Japan had been forced open by US Commodore Matthew C. Perry less than a decade earlier and the purpose of the mission was to establish how China had fared following the country's defeat in the Second Opium War (1856–1860). Takasugi was aware of the forward thinking exhibited by those such as Wei on the new threats posed by Western "barbarians" [...] Sinologist Joshua Fogel concludes that when Takasugi found out "that the writings of Wei Yuan were out of print in China and that the Chinese were not forcefully preparing to drive the foreigners out of their country, rather than derive from this a long analysis of the failures of the Chinese people, he extracted lessons for the future of Japan". Similarly, after reading the Treatise, scholar and political reformer Yokoi Shōnan became convinced that Japan should embark on a "cautious, gradual and realistic opening of its borders to the Western world" and thereby avoid the mistake China had made in engaging in the First Opium War. Takasugi would later emerge as a leader of the 1868 Meiji Restoration which presaged the emergence of Japan as a modernised nation at the beginning of the 20th century. Yoshida Shōin, influential Japanese intellectual and Meiji reformer, said Wei's Treatise had "made a big impact in our country".

14.9.4.22:45: *BAKUSHIKO AND *BAKUSHIK(W)A

I have long been puzzled by Chinese character spellings of foreign place names in Japanese. Some seem to be hybrids of Chinese and Japanese readings.

For instance, when I Googled for モスコウ Mosukou and 1939 last night, I found 北京より 莫斯古へ Pekin yori Mosukou e, 高山洋吉 Takayama Yōkichi's 1939 translation of Sven Hedin's Von Peking nach Moskau. Although the Kobe University City Library has it catalogued as Pekin yori Mosukuwa e (with the currently dominant Japanese name of the city - the most likely term to be searched), I assume 莫斯古 was meant to be read as Mosukou since the name appears as モスコウ Mosukou in the title of chapter 11. 莫斯古 Mosukou would be read as *Mosigu [mwɔ sz̩ ku] in Mandarin and *Bakushiko in normal Sino-Japanese. Whoever created that spelling seemed to be thinking of Mandarin 莫斯 [mwɔ sz̩] followed by Sino-Japanese 古 ko. (There is no [mɔ] or [mo] in standard Mandarin.)

I used to think another Japanese spelling 莫斯科 was a direct loan from Mandarin Mosike [mwɔ sz̩ kʰɤ] (ke was once [kʰɔ] and is still [kʰɔ] or the like in many other varieties of Chinese today). But could it be a blend of Mandarin 莫斯 [mwɔ sz̩] followed by Sino-Japanese 科 kwa (pronounced [ka])? Was 莫斯科 first attested in Chinese or Japanese? In any case, it cannot be based on its hypothetical normal Sino-Japanese reading *Bakushikwa.

Tonight I discovered a third Japanese spelling 莫斯哥 in Yamamoto (2009: 78). This looks like a direct loan from the less common Mandarin name Mosige [mwɔ sz̩ kɤ] (ge was once [kɔ] and is still [kɔ] or the like in many other varieties of Chinese today). In normal Sino-Japanese, it would be read *Bakushika which sounds nothing like Moskva or Moscow.

In the Kobe University Library Newspaper Clippings Collection, 莫斯科 appears 815 times between 1912 and 1941, 莫斯哥 appears only once in 1916, and 莫斯古 does not appear at all. The three katakana spellings combined outnumber the kanji spellings by nearly two to one (1564 : 816).

14.9.3.23:59: MOSUKUWA VS. MOSUKŌ (AND MOSUKOU)

While looking up ワルシャワ Warushawa and ワルソー Warusō in various editions of Kenkyusha's New Japanese-English Dictionary, I noticed that their distribution paralleled that of モスクワ Mosukuwa (< Russian Moskva) and　モスコー Mosukō (< English Moscow). Was there a shift toward more Slavic-flavored Japanizations by 1974?

Here is the distribution of both terms and a third term in the Kobe University Library Newspaper Clippings Collection:

Mosukuwa: 866 results, 1912-1942

Mosukō: 568 results, 1912-1942

モスコウ Mosukou: 130 results, 1915-1936

I was expecting Mosukuwa to be less common than Mosukō by analogy with Warushawa and Warusō, but the reverse is true. I wish I had postwar statistics. Here are current Google statistics showing the gaps between the three have widened considerably:

Mosukuwa: 1.36 million

Mosukō: 119,000 results including non-Russian Moscows (cf. the use of Warusō for non-Polish Warsaws)

Mosukou: 10,600 results including モスコウイッツ Mosukowittsu 'Moskowitz'

I've never heard Moscow rhyme with Mexico. Is that pronunciation still current, and if so, where?

14.9.2.23:23: WARUSHAWA VS. WARUSŌ

The influence of English on Japanese has only grown over time, while the influence of other European languages has waned: e.g., German-based dēdētē 'DDT' (in this dictionary of extinct Japanese words) has been replaced by English-based dīdītī. (What was the last major German or French loanword in Japanese?) So when I see a continental European loanword, I assume it is pre-1945: e.g., ワルシャワ Warushawa 'Warsaw' which sounds like Polish Warszawa (though Polish w is [v]).

That was why I was surprised to see an English-like ワルソー Warusō for 'Warsaw' in the September 2, 1939, Asahi shinbun. (Yes, the 75th anniversary of the beginning of WWII is still on my mind.) How far back do Warushawa and Warusō go? I wish Google Ngram Viewer worked with Japanese.

I quickly found various attestations of Warusō from the period:

- the September 18, 1939, entry of the diary of 馬淵良三 Mabuchi Ryōzō

- the October 6, 1939, 大陸日報 Continental Daily News published in Vancouver, BC

- 宮本百合子 Miyamoto Yuriko, "The Flames of the Life of Mrs. Curie" (December 1939)

- the Privy Council's "Abolishing an Imperial Embassy in Poland" (October 1, 1941)

Was Warusō the standard Japanese name for Warsaw at the time? Judging from Wikipedia, today it seems to linger only in a few contexts such as ワルソー条約 Warusō jōyaku 'Warsaw Convention' (1929; cf. ワルシャワ条約 Warushawa jōyaku 'Warsaw Pact' with the same jōyaku) and ワルソー・コンチェルト Warusō koncheruto 'Warsaw Concerto' (1941). The Japanese Wikipedia entry for Warsaw doesn't mention Warusō as an alternative of Warushawa.

14.9.3.21:51: Looking in various editions of Kenkyusha's New Japanese-English Dictionary, I discovered that editions prior to 1974 only listed Warusō. The 1974 edition listed both Warusō and Warushawa for the first time.

The 1975 edition of Sanseido's New Concise Japanese-English Dictionary that I have been using for over thirty years lists only Warushawa in its appendix of place names.

It would be interesting to see when, say, Asahi shinbun shifted from Warusō to Warushawa.

14.9.3.22:36: A couple more data points: I just found Warusō in the December 20, 1919 Ōsaka asahi shinbun (image / HTML) and Richard Austin Freeman's The Case of Oscar Brodski, translated into Japanese by 妹 尾韶夫 Seno Akio in 1957.

Okay, a few more: Warushawa first appears in the Kobe University Library Newspaper Clippings Collection in the May 30, 1913, Jiji shinpō (image / HTML), and appears in papers up through 1939 (image / HTML). Warusō first appears in that collection in 1915 (image / HTML), and last appears in 1941 (image / HTML). Warusō outnumbers Warushawa by a ratio of roughly seven to one (138 : 21). Today in Google, Warusō is vastly outnumbered by Warushawa (20,300 : 515,000). Warusō is the only Japanization of Warsaws outside Poland, but I presume those other Warsaws aren't mentioned enough to give Warushawa serious competition.

14.9.1.22:09: GYDDANYZC

I have long been interested in Slavic partly because it underwent massive vowel loss paralleling the massive vowel losses I reconstruct for Chinese and Tangut.

My favorite example is monosyllabic Gdańsk from *Gŭdanĭskŭ* with four syllables (Comrie 1987: 326). That city has been on my mind lately because today is the 75th anniversary of the German invasion of Poland.

On Saturday I learned that Gdańsk is first attested as Gyddanyzc sometime after 997 AD. Is that spelling evidence for

- the retention of medial *ŭ and *ĭ

- the loss of final *ŭ

circa 1000 AD?

Why were the vowels that were later lost both spelled y? Had they merged into [ɨ] in the version of the name that was transcribed? They could not have merged in the ancestor of the modern name Gdańsk, as ń is from *nĭ and still reflects the palatal quality of the lost vowel *ĭ. Or is ny a transcription of [ɲ]?

Is the doubling of d significant?

Why was the sibilant before c [k] written as z instead of s? The z also appears in the later spellings Kdanzk (1148), Gdanzc (1188), and Danzc (1263). I assume the z of the spelllings Danczk (1311), Danczik (1399), and Danczig (1414) is half of a digraph cz [tʂ] and is not evidence for [z].

*This matches Old Church Slavonic Гъданьскъ <Gŭdanĭskŭ>. Is that form of the name attested in ancient texts, or is it a retroactive creation?

I assume the ISO 639-1 code cu for OCS is from c(h)u(rch). cu makes me think of Cuman which has no ISO 639-1 code; its three-letter ISO 639-3 code is qwm. qum and cum were already taken for Sipakapense in Guatemala and Cumeral in Colombia.

14.8.31.1:31: EAT-YMOLOGY 3: *NZ- > *NDZ-?

I just realized that 'eat' from my last post wasn't the best example of a word that had undergone brightening without lenition. What if lenition were followed by fortition (in bold) after a nasal?

stem 1: *NI-dza > *NI-dzja > *NI-z- > *Nz- > *ndz- > dzi 1.11

stem 2: *NI-dza-w > *NI-dzjaw > *NI-z- > *Nz- > *ndz- > dzio 1.51

Perhaps these are better examples of brightening without lenition:

Tangut 0749 phi 1.11 'to order' (stem 1), 4568 phio 2.44  'to order' (stem 2)  : Japhug kɤ-ɣɤ-xpra 'to order'

also cf. Somang ka-wa-kprá 'to order' preserving Proto-rGyalrong *kpr-

Pre-Tangut *CI-Kpra(-w-H) > *CI-Kprja(w-H) > *Kpr- > phi(o) 1.11/2.44

or Pre-Tangut *KI-pra(-w-H) > *KI-prja(w-H) > *Kpr- > phi(o) 1.11/2.44

Tangut 5449 1tị 'to put' 1.67 (stem 1), 5633 1tiọ 'to put' 1.72 (stem 2)  : Japhug kɤ-ta 'to put'

Pre-Tangut *CI-S-ta(-w) > *CI-Stja(w) > *tt- > ti ~ tiọ̣ 1.67/1.72

or Pre-Tangut *SI-ta(-w) > *SI-tja(w) > *tt- > ti ~ tiọ̣ 1.67/1.72

If lenition had preceded brightening, ph- and t- would have lenited to *v- and *l- in those words.

I do not know if the *I of the brightening presyllable followed *K- that conditioned the aspiration of ph- and/or the *S- that conditioned the tension of rhymes 1.67 and 1.72 (indicated by a subscript dot). *I could have been in a presyllable preceding one or both of those consonants.

Pre-Tangut *K(I-)pr- nicely matches Proto-rGyalrong *kpr-, but pre-Tangut *S(I)-t- does not match Proto-rGyalrong *t-. Perhaps the aspirated th- of Written Burmese thāḥ 'to put' is from *St-.

Old Chinese 置 *trək-s 'to place' may be an unrelated lookalike even if it is from *r-tək-s, as it has a *-k absent in  the other languages.

I cannot explain why stem 2 of 'to order' has a second ('rising') tone from an *-H absent in stem 1.