murawaki の雑記


Manchu: A Textbook for Reading Documents

Gertraude Roth Li, Manchu: A Textbook for Reading Documents, University of Hawaii Press, 2000.


  • p.88, Notes 2: erei da sekiyen mafa gebu Paihanpar. As noted by the arthor, this sentence has a topic-comment structure. Is Manchu a topic-prominent language?
  • p.101, Vocabulary, the 5th line from the bottom: se bahambi. As noted at page 103, se in this context seems a plural suffix.
  • p.130, last line: wang ni jalin faššaha. Li gives the translation "exerted himself on behalf of the emperor" (p.331). Referring to the emperor by wang sounds strange. I suspect this is a literal translation of some Classical Chinese idiom. 勤王?
  • p.134: hoise sei banin kelfišeme kenehunjere mangga. Another topic-comment construction.
  • p.135: mayan tatabumbi is clearly a literal translation of Chinese 掣肘.
  • p.141, 3-4th lines: ser sere be seremšeme badaran be sibuki sere gūnin. Li gives the translation "a policy with protects small leaders and restrains powerful ones" (p.332). I think this phrase is a literal translation of Chinese 防微杜漸. An English translation would be "nip something in the bud."
  • p.141, 7th line: dere de eterakū ombihe. Li gives the obscure translation "would not have overcome the situation" (p.332). I think dere means 面子 in this context. The following would be a better translation: "[Osman] could not hurt [Gaopu's] feelings."
  • p.149, 5th line: olji be jafaha hūlha be waha gungge ilibuha ... Li translates this phrase as "excelled in taking prisoners and killing rebels" (p.334). jafaha and waha form a coordinate structure. I wonder why jafaha but not jafame or jafafi is used. Without contextual support, it would better be read: "excelled in killing rebels who had taken prisoners."
  • p.150, 4th line: beile i jergi gūsai beise. Li translates this phrase as "beile prince of the rank of a beile" (p.334). Either one of them must be beise.
  • p.152, last line: yadahūn banjici ojorakū ofi ... It seems that yadahūn alone forms a clause although no glue marker is attached.
  • p.171, Note 61: Akim ini deo be ujen weile baharahū seme (also in Review 9 at page 175). Li gives the translation "because Akim feared that his younger brother might have committed a serious crime." I do not see why Li choses the conditional perfect. According to (Gorelova, 202:307-308), -rahū can be analyzed as -ra + akū. baharahū refers to an action that might take place in the future. A better translation might be: "because Akim feared that his younger brother might incur a felony conviction."
  • p.176, Review 13, v): Hotien should be either Khotan or Hetian. I wonder why she uses Wade-Giles.
  • p.194, 11th line: ambasa umai sehe akū aise. I do not see why the Kangxi Emperor used the peculiar expression sehe akū instead of the standard sehekū (also, why did Li ignore this?).
  • p.203, lines 4-5: G'aldan mimbe kemuni Bosihi sei sasa elcin unggiki serede (When Galdan wanted to send me as envoy to go along with Bosihi). A possible SOCV construction. Why is elcin left unmarked?
  • p.203, lines 5-6: G'aldan emu yargiyan ba akū bime. I wonder why de is absent. It should be placed right after G'aldan. A topic-comment structure?
  • p.204, 6th line: yalure jeterengge (provisions and transportation). An interesting example of coordination + the nominalizer -ngge.
  • pp.206-7: G'aldan i ya ici genere. jai G'aldan i gūnin arbun. tubai yaya baita be ... narhūšame wesimbure babi. Li's translation (p.338) is as follows: "But which direction he will turn depends on his intentions. ... I will report in more detail on the situation there." Li treated G'aldan i ya ici genere as a subject and G'aldan i gūnin arbun as a predicate. Can arbun be translated as "depends on"? I think gūnin arbun might be a phrase. Also, as these two phrases is connected with jai, I think that they are coordinated, and that the entire phrase is part of the object of wesimbure.
  • p.208, line 1: (also the transliteration at p.210, and Note 9): giyamun deri genebuhe. To me, the Manchu text seems to represent benebumbi (the causative of benembi), not genembi.
  • p.214, lines 1-2: niyalma largin, jugūn de feksire de tookajara de isinambime ... (the number of people would be large, making for delays during travel). It seems that niyalma largin is a subject-complement construction without a copula. largin is syntactically unrelated to jugūn, which immediately follows largin. Really confusing. We can see a tseg after largin but Li ignored it (p.217).
  • p.226, Transliteration, (220) 3rd line Noyan gelung. Looks like Li interprets this phrase as a personal name (Noyan) plus a title (gelung). Both look like common nouns. Noyan means a lord in Mongolian, and gelüng (from Tibetan dge slong) is a title for Buddhist priests. I am not sure if noyan was used as a personal name in Mongolian.


(might be corrected in the second edition)

  • p.37, 7th line: sargūsarkū (typo in the original?).
  • p.42, last example: jjihejihe.
  • p.59, Vocabulary: banjimbi sainbanjire sain.
  • p.62, Vocabulary: jecen toktobure ici ergi aisilara jiyanggiyunjiyanggiyūn.
  • p.83, Transliteration: kūmsokomso.
  • p.87, Vocabulary, (75), 4th line: Tsewang RaptanTsewang Rabtan.
  • p.161, (127), 3rd line: ubui i nonggime. Li often corrects typos without comment, but here the original spelling is preserved. This must be ubui nonggime or ubu i nonggime.
  • p.164, (146), 2nd line: insert i after Ūdui.
  • p.171, Note 61, 3rd line: rahurahū.
  • p.175, Review 9, 6th example: sildirarasilhidara.
  • p.211, Review 2, 4th example: weleheiweilehei.
  • p.242, Note 8. 'Is the the bad person to prevail?' → ???



何荣伟 ed. 满语365句, 辽宁民族出编社, 2009.

This book is accompanied by a CD, and Manchu utterances were recited by a Sibe announcer. I am curious about exactly what this language is. She pronounced Manchu spellings rather faithfully. But judging from recent field studies I have checked, it clearly different from Modern Spoken Sibe, which should, for example, drop final vowels. Is there any tradition about literary pronunciations in Sibe? Or is this just a modern reinvention? Or do Sibe people actually speak like this?

5: goidame sabuhakū bihe.

(Gorelova, 2002:312) states that goidame, which is derived from goidambi (last for a long time, to endure) means "long ago," "long since," and "for a long time."

6: hanciki de sain yabume biheo.

Does hanciki mean temporal proximity? The glosses and examples given by (Hu, 1994:386) imply that it only means spatial proximity. By contrast, (Li et al., 1984:296) only gives temporal proximity for Sibe:

hanqiki /hanchiki/ 近来, 最近。

Also I suspect de is unnecessary here.

biheo sounds something like /biho/.

10: bi yabume oho.

The syntactic pattern V-me oho cannot be found in grammar books on Written Manchu I have. It is clear from context that this refers to an action that is about to start, but I am not sure about its exact function.

(Kim et al., 2008) contains three examples of V-me oho.

  • 076: te bi 就 jeme oho. 那就不客气了 ???
  • 079: bi neneme yabume oho. 我先走了
  • 133: bi amgame oho. 我要睡觉了

As for Sibe, (Li et al., 1984:50) explains -m oⱨui as "现在-将来完成时" and "表示行为现在已经完成." (Kogura, 2007:147) translates -me o-hui into "~しそうだ" and provides one example.

  • bi sade-maqe buce-me o-hui. (私は疲れてもう死にそうだ)

Quote from (Li, 2000:306):

The verb forms -me oho and -me ohobi, or the negative form rkū oho are very common in Sibe. Examples: ice bele jime oho 'they are about to bring the new grain'; singgeri boo gurime ohobi 'Mr. Mouse is about to move' turigen be bajen de afaburkū oho 'they will no longer turn in rent to the landlords'; duleke de adališarkū oho 'it is no longer as in the past'. This type of oho parallels the Chinese word le 了 with the meaning 'to be about to' or in a negative sentence 'is no longer' or 'no more'.

15: ererengge sini embi jai mudan acaki.

Not sure about the syntactic relation of ererengge to acaki. Looks like it is a topic but not the subject of acaki.

18: ai arame ilihabi.

V-me ilihabi means some action or situation is in progress (Li, 2000:359). Maybe a trait of Sibe. However, (Guan et al., 2002:77) refer to V-me ilihabi as a present form (of Written Manchu).

In a grammatical sketch of Spoken Sibe, (Kida, 1999:52) explain [ilam] as "動作が進行・反復・継続している状態を示す。日本語の「てある、ている」に相する。" He provides six examples, all of which use the -m (-mbi) form.

23: bi ere juwe inenggi šahūrakabi.

juwe sounds like /ju/.

24: bi ergeme geneme oho.

Another V-me oho construction.

25: si ainahabi.

The announcer clearly attached ye to the end of the utterance although it is not shown in the book. According to Li, ye and na are "[p]articles/suffixes for emphasis and questions" (Li, 2000:311). They are described in the chapter of Sibe.

34: sikse dobori de sain amgahabio.

I wonder why de is attached to sikse dobori. The Japanese counterpart "ni" is used only for temporal nouns of absolute dates. "Last night" is not followed by "ni" because it depends on the point of time it is mentioned.

36: majige yasai amu isinjiha.

Quote from (Li et al., 1984:277):

yasim /amu/ 困倦。

yasim /amu isinzhimbi/ 瞌睡。

I suspect Sibe yasim actually corresponds to Written Manchu yasai amu, not amu.

38: ombi, erdeken i erge.

It seems that erdeken is always followed by the instrumental -i. erdeken (erde + -kAn) appears more like a noun than an adjective-adverb.

42: minci juwe se amba bihe.

I do not see why bihe is used for the current state.

44: teike sini emgi yabuha asihata weci.

The initial consonant of teike does not sound like "t" at all to me.

It is clear from context that asihan is used as a singular noun. However, (Kawachi et al., 1994:178) interprets asihata as the combination of asihan and the plural marker -ta. (Hayata, 1998) discussed the very problem. A number of Manchu dictionaries explain asihan as a young man (singular) and asihata as the plural form of asihan, but his analysis on the gin ping mei bithe revealed that asihan was an adjective and that asihata was used as a singular noun. In Sibe, asihata refers to a young man. I guess -ta is a derivational suffix used to form nominal words with semantics of quality (Gorelova, 2002:199).

(Gorelova, 2002:135) gives asihata as an example of the plural marker -ta, but (Gorelova, 2002:370) excerpts a singular form example from SIDI KUR. A Sibe-Manchu Version of "the Bewitched Corpse":

tere tasha aniyange ašikta ainci muke de buce-he dere (That guy who was born in the year of the tiger probably died in the water)

(Norman, 2013:23) gives "2. youth (often used as a singular)."

In the 清文啓蒙・兼漢滿洲套話 (28-1), asihata corresponds to Chinese 小夥子們 (i.e., plural).

At the entry of we, (Hu, 1994:807) lists two examples of weci that is placed at the end of utterances to form who-questions. No grammar books explain this syntactic pattern.

According to (Kawachi, 2014:1175), the 大清全書 gives: 是誰, 那幾個. coohai ejete[O ejente]weci 領兵者是那幾個. 兵主是誰.

(Norman, 2013:403) gives "2. who (plural), (plural of we)." However, asihata is a singular noun in this context.

In the 清文啓蒙・兼漢滿洲套話 (44-2), weci corresponds to Chinese 是誰.

47: i kemuni ajige, teni emu se tuleke.

The Chinese translation "才过一岁" suggests that tuleke (present: tulembi) means "to have passed," but I cannot find this meaning in dictionaries. (Hu, 1994:750) only gives "(1) 下网, 张网, 下套。(2) 套, 套上。" for tulembi. Also, Hu does not mention the -ke form. It was pronounced as /tuleke/, but tuleke must be a misspelling of tulike (present: tulimbi).

48: omšon biya ice ninggun de banjiha inenggi be dulebuhebi.

biya was pronounced as /bi.ya/, not /bja/.

According to (Li, 2000:391) and (Hu, 1994:202), the causative of dulembi is dulembumbi, and dulebumbi is not mentioned.

50: sini boode udu anggala bi.

53: sinde udu ahūn deo eyun non bi.

55: sinde udu jui bi.

86: sini boo ya bade bi.

120: gūwa ai jaka gaimbi..

141: ainu tenteke kangkahabi.

155: si ai omimbi.

157: bi, omimbio.

159: ombi. kara omimbio, embici sun be labdukan i suwaliyambi.

165: si aibide weilembi

175: si ya ba de weilen arambi.

182: tere ya ba de weilembi.

184: tere ai weilen arambi.

223: si an i ucuri de ai gisun be gisurembi.

229: ere be manju gisun i gisureci ai sembi.

246: enenggi ai bithe tuwaki sembi.

255: giya de genefi ainahabi.

273: ya ba deri yabumbi.

298: si tacikū de ai kicen tacimbi.

Untranscribed final ye. Looks like ye almost always follows bi.

60: suweni booi gūwa niyalma ubade terakūn.

suweni sounds like /sen.ni/.

I wonder if suweni can be replaced with sini in this context. They are talking about the listener's family and some other members of his/her family have been mentioned, but only the listener is present in the conversation. The Chinese translation of suweni boo is "你家."

76: ini gebu simengge sembi, tunggiya hala.

Typo: replace terei with ini (only the romanization is wrong; the Manchu script is correct). The announcer used ini, not terei.

85: bi beijing hoton de weilen arambi.

Manchu has weile and weilen. The meaning of the latter seems to have been narrowed.

89: yamjishūn de šolo bi nio.

Is de necessary here? See also No. 34.

90: bi, ai baita.

Untranscribed final ye after bi, but in an affirmative. (Li et al., 1984:70) states that ye "表示惊异, 赞叹, 或疑问语气."

91: sinde šolo bici, emu niyalma be takabume buki.

bumbi is used as an auxiliary verb. I cannot find such a usage in Written Manchu. As for Sibe, (Kida, 1999:53) explains it as one of auxiliary verbs: [bum] (くれる、やる、あげる) 日本語の「やる」「くれる」は区別なく [bum] で表される。私が主格である時には「やる」であり、自分以外の人間が主格で、自分に対してある有利な行為を行えば「くれる」と訳せるだけである。

Also, there are two examples in Sanjiazi Manchu. Quote from (Kim et al., 2008:210-1):

  • 316: bi sefu de meni tokso julen be ǰaŋnan (讲-na-me?) buhe. 我给老师讲了关于我们村子的故事.
  • 317: bi eniye de bithe be hūlame buhe. 我给妈妈读书了.

92: ere weci.

Another weci. See also No. 44. Also, ye is attached, but not to bi.

98: ulintai, erei onggolo si simengge be sabume dulekebio.

onggolo takes a genetive form, and even the demonstrative ere changes its form to erei in the transcript. But I cannot recognize the final i in the utterance.

I cannot find an explanation for the syntactic pattern V-me dulekebio in Manchu grammar books. Looks like a literal translation of the Chinese aspect marker "过" (past experience). As for Sibe, (Kogura, 2007:147) provides the following explanation:

-me dulu-xui 「ーしたことがある」(dule-: 過ぎる*)

* 漢語でも経験を表すのに,同様に動詞+「過」(過: すぎる) という構造が用いられる.

100: tuttu oci, bi imbe sinde takabume bihe.

Pronouns are used as if they cannot be dropped. Aren't they obvious from context?

104: bi inenggidari ningguci erin de ilimbi.

I cannot find the 12/24-hour time system in Written Manchu. Quote from (Li, 2000:295):

When telling time, colloquial Sibe usually uses the word erin. not jungken. It may also omit the ordinal ci.

106: bi inu teni ilihabi, buda jetere unde.

The final ye after ilihabi in the utterance. See also No. 90.

107: muse emu bade erdei buda jeki.

I suspect the genitive i in erdei buda is unnecessary.

emu bade means "together" (in one place -> in the same place). The semantic extensions of emu (one) are really confusing. It can refer to an indefinite object (one, not zero). On the other hand, it can mean "the same" (one, not two or three).

108: tuttu oci bi antahararakū oho.

This is the first appearance of V-rakū oho although Nos. 10 and 24 use the syntactic pattern V-me oho. -rakū seems to substitute for -me here as Manchu converbs lack negative counterparts.

antaharambi derives from antaha (guest) and literally means "to be/act as a guest." I am not sure about the exact nuance but it seems to mean "to be polite or modest." I see a parallel with Chinese 客/客气. They can be found in Written Manchu.

113: sain, jetere de umesi sain.

A comma is placed after sain, but no pause is confirmed.

I think this sentence can be translated as "very good to eat." Need to find other examples of V-ra de sain.

114: coko i umgan jembio.

Again, I suspect the genitive i in coko i umgan is unnecessary.

115: bisire oci, emke bureo.

I don't understand how bisire oci differs from bici. The latter alone might be too short to form a clause.

emke is acually realized as emken. (Li et al., 1984:36) and (Kida, 1999:23) suggests that "one" is emken in Sibe.

119: jeme muterakū oho.

V-me mutembi is certainly used in Written Manchu. (Li, 2000:92) reviews V-me mutembi together with V-ci ombi and V-me bahanambi, but does explicitly not explain the difference in nuance. (Gorelova, 2002:306) does not either. Quote from (Kida, 1999:83):

動詞 [om (なる)]、[mutum (できる: 能力がある)][banem (できる: 修得してできる)] などを使って表現する。... [om] と [mutum] はともに可能を表すが、その意味には差があり、[om] は条件として可能であるという場合で、[mutum] は能力として可能であるという場合に使用する。

123: efen, nasan sogi, sun cai, ememu erin de uyan buda jembi.

(Hu, 1994:698) among others gives sun i cai.

124: bi ere hacin i jemengge be cihalambi.

It is surprising that "to like" is expressed with a verb taking an accusative case in Manchu as Mongolian uses the syntactic pattern N-д дуртай/дургүй.

125: si ai buda jeme cihalambi.

149: arki omime cihalambio.

The syntactic pattern V-me cihalambi. (Gorelova, 2002:296) excerpts one such example from SIDI KUR. A Sibe-Manchu Version of "the Bewitched Corpse":

ši ere ba-de banji-me cihala-ci men'i ere abka i ba-de sasa banji-ki. (If you wish to live in this place, live together with us in heaven.)

129: suwe ai jembi.

suwe sounds like /swe/.

131: sini cihalara buda sogi be sonjo.

buda sogi seems an idiom, but does not appear in dictionaries. According to a blog, it is used in the nišan saman i bithe. It might be a literal translation of Chinese 饭菜.

136: sile be majige gaimbio.

Syntactically majige modifies gaimbio but semantically depends on sile.

137: ombi, emu moro sile bureo.

ombi is used as an answer to the question. Manchu grammar books list various question expressions, but do not mention how to say "yes" and "no."

bureo sounds like /buro/. In No. 115, I can recognize a trace of /e/ in bureo.

141: ainu tenteke kangkahabi.

ainu sounds like /aino/. Its Spoken Sibe form is an.

142: tulergi de hontoho inenggi feksihe.

I wonder if there are any expressions corresponding to "because ..."

144: majige pi jio nure omici antaka.

pi jio is clearly a loan word from Chinese. (Li et al., 1984:215) gives "pivo /piwo/" to 啤酒.

I wonder why there is a huge space between omici and antaka.

146: muke omici sain.

The syntactic pattern V-ci sain. (Gorelova, 2002:147) excerpts one example from SIDI KUR. A Sibe-Manchu Version of "the Bewitched Corpse":

ši encu emu bou ara-ci sain (It will be good if you build another house)

147: muke omime wajifi, muse jai arki omiki.

wajimbi as an auxiliary verb seems to have a long history. As for Spoken Sibe, (Kida, 1999:53) notes: [wazhem] (終わる、~してしまった) は失敗、残念な意を表す。Clearly, the utterance in question has no such nuance.

150: bi an i ucuri de majige omimbi.

Again, I suspect de is unnecessary.

152: pi jio nure omici sain, šanggiyan arki omici uju nimembi.

As a native speaker of Japanese, I naturally think of Japanese translation. Now I wonder why the following sentence is unnatural: 酒を飲んだら頭が痛い.

154: arki omime wajifi, tubihe i muke be majige gaiki.

I have not attested the phrase tubihe i muke in other sources.

159: ombi. kara omimbio, embici sun be labdukan i suwaliyambi.

Quote from (Gorelova, 2002:156):

Diminutives are commonly followed by the genitive case marker i (occasionally they appear in a stem form) and normally function as adverbial modifiers. Due to this, they reveal a tendency to shift from diminutives to adverbs

See also No. 38.

170: manju i suduri, šu wen jai gisun hergen be sibkimbi.

The phrase šu wen is not listed in (Hu, 1994). (Li et al., 1984:230) gives an unsegmented form:

shuven /shuwen/ 文化。

This section is interesting as it adapts the traditional lexicon to modern life.

171: tuttu oci, si emu sefu bihe.

Need to explain why bihe is used here. I guess this expression emphasises the fact that the listener had been a teacher even before the speaker found out that.

179: bithe tuwaki seci, mimbe baime jio.

Can baimbi take person(s) as an object (the be case)? Looks perfectly OK.

180: tuttu oci, sinde ambula baniha.

Saying a thank-you after a conditional expression gives a bad impression at least in Japanese. What about Manchu?

185: etuku adu jergi jaka be uncambi.

Quote from (Gorelova, 2002:138-9):

Plurality can be expressed with the help of the words ... jergi "category," "grade," "rank," "sort." Following the marker for the genitive case added to the noun, these words denote both plurality of objects and the belonging of the latter to a certain class, kind or sort. Following the noun directly, they denote plurality with the shade of meaning "of different/various kinds, every kind." The word jergi is normally used with animate objects ...

However, it is not difficult to find examples of jergi with inanimate objects. Quote from (Kida, 1999:37)

*1)bitke deri tuilhun, serkin, yurɣan, zherɣi i zhaqe gele bi. (本のほかに新聞、画報などもある。)

*2)ere zherɣi afesie. (これなど如何でしょう。)

*3)min deri tiulɣen sulfanga, mutenge, giltuqan zherɣi gurun gele zhim. (私以外にスフファンガ、ムテンゲ、ギルトカンなどの者も来る。)

186: ememu erin de efire jaka uncambi.

This is a declarative sentence. The final "?" in the transliteration must be a typo.

188: bi niyamangga gisun be tacibumbi.

The phrase niyamangga gisun seems a part of the modern vocabulary. It is not found in Manchu dictionaries but in (Li et al., 1984:205):

niamang gisun /niyamangga gisun/ (1) 母语。 (2) 语文。

192: fengšentai oci emu usisi.

Quote from (Gorelova, 2002:198):

the suffixes -si, -msi, -ci, -ji, -lji, -mji, -nju are used to form nouns both from verbs and nouns. The resulting nouns designate people according to their skills, occupations, ranks, trades, and business: ... usisi "farmer," ...

195: manju hergen atanggi fukjileme banjibuhabi.

fukjilembi (< fukjin + -le) is not found in (Hu, 1994). (Li et al., 1984:150) gives "fuhjilem /fukzhilembi/ 创造。"

196: emu minggan sunja tanggū uyunju uyuci aniya de, monggo hergen be ten obufi, manju hergen be fukjileme banjibuha.

obumbi fills monggo hergen in its accusative slot. Then what is the case of ten?

205: daicing gurun i werihe manju hergen i bithe cagan jai dangse elekei gemu ice manju hergen i arahabi.

Is ..bithe cagan jai dangse the object of arahabi? It is interesting that the accusative marker be is not attached to such a very long non-adjacent phrase.

210: i gisurere ararade gemu hafu.

An example of verbal coordination. Also interesting is that hafu takes the de case.

217: manju gisun be tacirede ja nio mangga nio.

I am unsure why tacire and ja/mangga are connected with de. tacire is the theme of the predicate ja. "tacirengge ja nio" looks more natural to me.

219: inenggidari emu erin baitalame tacimbi.

The first example of erin not being accompanied by de. erin modified by a cardinal number refers to a space of time.

221: onggolo aniya de, labdu manju niyalma manju gisun gisureme mutembi.

This example is interesting in that the imperfect -mbi is used for an event that contextually clearly occurred in the past.

237: jai mudan gisurecina.

The first example of -cina. Quote from (Gorelova, 2002:301):

The form in -cina/-kina is considered to be an imperative form by most specialists (Zakharov, 1879:180-81; Pashkov, 1963:37). The form in -cina/-kina is used by the speaker when he addresses someone who is lower in social position. The accomplishment of the action is considered by the speaker as desirable: alacina (alakina) "speak out (if you wish);" tuwacina (tuwakina) "have a look (if you wish)," etc.

According to Zakharov, this form is originated from the combination of the conditional converb in -ci and the interrogative particle na: -ci + na > -cina. In the author's opinion, this form expresses a polite request formulated as a question. The meaning rendered by the form in -cina can be translated as the following words and phrases: "please," "very likely," "let it be," "let it be in this way," "isn't it so?," "isn't it true?". The word form bicina (the form in -cina derived from the verb bi- "to be," "to exist") can be translated as "let it be in this way," "isn't it true?". The word form secina (the form in -cina derived from the verb se- "to say") means literally "let me say." It is very often used as a note of exclamation at the end of an utterance in the meaning "this way it is!," "so it goes!" (Zakharov, 1879:180-1).

239: bi jing manju gisun hergen be sibkimbi.

jing indicates that an action is in progress, and has some parallels with the Chinese aspect marker (正)在.

254: sinde šolo bici, mini emgi giya de gengšeme yabu.

I guess giya is a loan from Chinese (街). It has a variant giyai. (Hu, 1994:342) implies that giyai is more common. Quote from (Kim et al., 2008:73)

gja [kja] street 镇,街道,街上/街头 giyai

As for Sibe, (Li et al., 1984:156) gives "gia /giya/ 街道。"

I had a hard time finding gengšembi in a dictionary. It is not found in (Hu, 1994). As for Sibe, (Li et al., 1984:155) gives:

gengsim /gengshembi/ (1) 游玩, 闲逛, (2) 串门。

260: mini emgi sasa genembio.

The Chinese translation contains the modal verb 要, which has no Manchu counterpart.

261: minde majige icihiyara baita bi.

icihiyara sounds like /icihara/.

262: jai mudan de nashūn bahaci sasa geneki.

I do not understand why jai mudan is followed by -de.

267: ildun de emu gin šatan gaisu.

ildun de 'taking advantage of' without a modifier. (Li, 2000:200) cites one such example:

booi baita be ildun de icihiyame gamambi at the same time I'll take the opportunity to take care of some family matters.

271: majige bense fi jergi jaka be udame gaiki sembi.

bense is clearly a loan from Chinese, 本子.

272: tere hoton i julergi de bihe hūdai ba de geneci, ya emu jugūn sejen tembi.

The conditional converb -ci denotes a purpose here.

sejen in this context means a bus. jugūn corresponds to Chinese 路. "N路车" means "bus route N." The entire phrase is influenced by the Chinese convention.

I do not understand the phrase "ya emu."

273: ya ba deri yabumbi.

The first example of the controversial marker -deri. (Gorelova, 2002:193-4) offers some discussion on this matter. The context suggests that the meaning of -deri in this example is a starting point in space.

275: neneme emuci jugūn sejen tembi.

Dictionaries gloss neneme as "formerly, previously, beforehand" (Gorelova, 2002:354). However, it seems that neneme in this context means "first" as in "first ... then ..."

276: sunja tatan tehe manggi, jakūci jugūn sejen halame tefi, jai ilan tatan yabuci uthai isinambi.

sunja tatan and ilan tatan seemingly function as adverbial modifiers. Adverbial numerical phrases without case markers do need an explanation.

tatan in this context seems to refer to a bus stop. It would be part of the "modern" vocabulary. Qute from (Li et al., 1984:241):

taten /tatan/ 驿站, 住宿地。

jakūci jugūn sejen modifies tefi, and halame is used as an adverb.

277: sejen jime ilihabi.

The j in jime sounds unvoiced.

281: yaka sejen teci gemu ombi.

The context suggests that yaka means "which" (inhuman). (Hu, 1994:822) and (Kawachi et al., 2002:205) imply that yaka is primarily used for persons. As for Sibe, (Li et al., 1984:278) glosses it as "哪个."

In the 清文啓蒙・兼漢滿洲套話 (27-9), yaka bade acaha gese. corresponds to Chinese 在那裡會過是的 (i.e., inhuman).

282: aika turibure sejen teci, tofohon fuwen yabume uthai isinambi.

As suggested by the Chinese translation 出租车, turibure sejen (lit. hired car) means a taxi. Need to be attested in other literature.

fuwen was pronounced as /fun/.

I do not see why the imperfect converb yabume is used instead of yabuci or yabufi. The dependent action marked by the imperfect converb must occur simultaneously with the principal action. In my understanding, however, the action of isinambi occurs after the action yabumbi is completed. Also, yabume is immediately followed by uthai, which strengthens the view that yabumbi must be completed.

285: tuttu oci, muse sejen turime yabu.

The imperative form with inclusive we.

292: amba, tacikū de ilan tacibure leose jai duin indebure leose bi.

leose sounds like /louse/.

Modern terminology: tacibure leose, 教学楼; indebure leose 宿舎楼. It is interesting that causative verbs are used in both phrases.

293: an i ucuri de tacire juse tacikū de indebumbi.

The causative verb indebumbi takes tacire juse as an object, and the accusative marker be must be dropped here. The Chinese translation is expressed by a non-causative construction.

Modern terminology: tacire juse (students). (Lee et al., 1984:240) gives

taqixi /tachisi/. 学生。

tacire juse is attested in the 大清全書 (Kawachi, 2014:1030).

294: šolo sindaci teni boode beredebumbi.

šolo sindambi must be a literal translation of Chinese "放假" (go on vacation, have a holiday). I wonder if šolo co-occurs with sindambi in Classical Manchu.

296: erdeken i gene, ume sitabure.

I don't see why the causative verb sitabumbi is used. The agent of the action sitambi must be the listener of the utterance.

300: suweni banse de udu niyalma bi.

Quote from (Li et al., 1984:98):

bans /banse/ (1) 班。(2) 教室。

304: geren nofi gemu tacibure bithe be bahao.

Modern terminology: tacibure bithe (textbook?). (Li et al., 1984:240) gives

taqivchen /tachibchen/ 课本

314: yabure onggolo leolen šu be afabu.

Modern terminology: leolen šu (paper), a literal translation of 论文.

315: enenggi booi kicen oci cik arambi.

Modern terminology: booi kicen (home work, 家庭作业).

According to various sources I checked, cik only means a circle used as a punctuation marker, but the Chinese translation "句子" suggests that it refers to sentence(s) in this context.

325: niyengniyeri, juwari, bolori, tuweri.

The pronunciation of tuweri. For Sanjiazi Manchu, (Kim et al., 2008:100) gives:

tuli [ˈtʰujle] winter 冬天 tuweri

As for Sibe, (Li et al., 1984:253) gives:

türi /tuweri/ 冬天。

326: niyengneri de, birai juhe weniyehe.

It seems that in Written Manchu weniyehe is associated with iron. The Chinese gloss is "熔" (the "fire" radical) instead of "溶" (the "water" radical). As for Sibe, (Li et al., 1984:264) gives

venem /waniyembi/ 溶化

I don't see why the perfect marker -he is used. It is ignored in the Chinese translation.

327: ser sere edun julergi ci fulgiyeme jifi, tumen jaka arsunambi.

Dictionaries suggest that fulgiyembi (to blow) usually takes people as the agents. Not sure if wind fits. The 大清全書 gives 風其吹汝之吹 (Kawachi, 2014:403).

328: ba na de yasai jalu gemu niowanggiyan fiyan.

It is difficult to parse this poetic sentence. yasai jalu?

There is a huge gap between the spelling of niowanggiyan fiyan and its pronunciation.

According to (Gorelova, 2002:104), fiyan is realized as [fĭɛn] in Sibe.

329: juwari jihe. abka halhūn ohobi.

Again, I don't see why -he and -hobi are used in this context. Do they have a poetic effect?

330: hacingga ilha nemšeme ilhanambi.

nemšembi is glossed "飽くことなく求める, むさぼる" in (Kawachi et al., 2002:196), and "争, 争多, 求多, 贪婪" in (Hu, 1994:572). The second and third example are both "nendere be nemšembi," which is translated into Chinese as "争先." As for Sibe, (Li et al,. 1984:204) gives:

nemsim /nemshembi/ 争, 争先, 竟争

Semantic shift might have taken place.

331: bolori isiname, serguwen edun amargi ci dame jifi, juwari i halhūn sukdun be bašame tucibuhe.

juwari i sounds like /jwarigi/.

334: šahūrun tuweri de, nimanggi dame duribuhe.

duribumbi is used as an auxiliary verb. Quote from (Kida, 1999:53):

[diurevem] (始める) →~始める

337: nai mumuhu šun be emu barun torhome yabuci, uthai emu aniya.

Modern terminology: nai mumuhu seems a literal translation of "地球." I cannot find this term in dictionaries.

338: emu aniya de juwan juwe biya, uheri ilan tanggū ninju sunja inenggi bi.

juwan sounds like /jun/.

339: emu inenggi de orin duin erin bi. emu orinde ninju fuwen. emu fuwen de ninju miyoori bi.

I cannot find the spelling miyoori in dictionaries. (Hu, 1994:541) gives miyori. (Li et al., 1984:197) says:

miauri /miyaori/ 秒。

345: hangsi de niyalma sa tulergi de tucifi niowanggiyan be amcame genembi.

hangsi seems a loan from Chinese but I am not sure about the source. 寒食 (hanshi)?

Looks like niowanggiyan de amcame genembi corresponds to 踏青.

353: tulergi de halukan.

halukan sounds like /haluhan/.

365: sukdun usihiyeken.

(Hu, 1994) misses usihiyeken. He does gloss usihiyan as "同usihiyen" but it is not covered by the dictionary. usihin and usihiken are well attested.


A question that has puzzled me for many years is what is the name of "zero" in Manchu. According to (Li et al., 1984:152), the meaning of fuka is (1)零. (2)泡ㄦ,圈ㄦ. The original meaning of fuka is a circle, as opposed to a dot, and I have not attested fuka as a numeral in text.


  • Liliya M. Gorelova. Manchu Grammar, Leiden, 2002.
  • 关嘉禄, 佟永功. 简明满文文法, Shenyang, 2002.
  • 早田輝洋. 満洲語文語における或る単語の単数と複数について−『満文金瓶梅』のasihanとasihata−, 九州大学言語学研究室報告 第9号, 1988.
  • 胡增益 ed. 新满汉大词典, Ürümchi, 1994.
  • 河内良弘, 満洲語辞典, Kyoto, 2014.
  • 河内良弘, 清瀬義三郎則府. 満洲語文語入門, Kyoto, 2002.
  • 木田章義. シボ語を中心としたアルタイ系言語との比較による日本語の接辞・語尾の発生論的研究. 1999.
  • Kim Juwon, Ko Dongho, Chaoke D. O., Han Youfeng, Piao Lianyu, Boldyrev B.V. Materials of Spoken Manchu, Seoul, 2008.
  • 児倉徳和. シベ語, 文法を描く 2, pp. 131-157, Fuchu, 2007.
  • Gertraude Roth Li. Manchu: A Textbook for Reading Documents, Honolulu, 2000.
  • 李树兰, 仲谦, 王庆丰. 锡伯语口语研究, Beijing, 1984.
  • Jerry Norman. A Comprehensive Manchu-English Dictionary, 2013.


A Study of the Tacheng Dialect of the Dagur Language

By YU Wonsoo, KWON Jae-il, CHOI Moon-Jeong, SHIN Yong-kwon, BORJIGIN Bayarmend, Luvsandorj BOLD, Seoul University Press, 2008.

ソウル大学校出版文化院の Altaic Languages Series の 2 冊目。Materials of Spoken Manchu とあわせて購入。久しぶりに本でも紹介しようかというのに、Amazon は扱ってない。アフィれない。

ダウール語についてはほとんど知らない。そもそも新疆イリの塔城にダウール人のコミュニティがあるというところから知らなかった。基本的にモンゴル語っぽいのに満洲語の語彙が散見される。おまけに古風。10 が xarəb で語頭に x を残していたり、動詞 a- があったり、否定の副詞 (?) ul があったり。



  • p.16 (1) /a/ [-LOW, +HIGH, -FRONT +BACK] -> [+LOW, -HIGH, -FRONT +BACK]
  • p.20 (1) [-LOW, +HIGH, -FRONT +BACK +LONG] /aː/ -> [+LOW, -HIGH, -FRONT +BACK +LONG] /aː/
  • p.20 last line: maɣa(n) -> imaɣa(n)
  • p.21 middle: some of seem -> some of them seem
  • p.21 middle: tone -> stone
  • p.32 3rd line: Septemberistəɾ [tab] saɾ -> September [tab] istəɾ saɾ
  • p.32 (10) Occurs in a [-back] vowel words -> word
  • p.32 (11) Occurs in a [+back] vowel words -> word
  • p.33 three -> tree (えっ、と一瞬思った)
  • p.44 gloss: friend-go-ref -> friend-com-ref
  • p.45 Reflexiveness possessiveness -> Reflexive possessiveness
  • p.47 3.2.2 (1) *Its -> Its (?)
  • p.56 (1) uku-səŋb -> uku-səŋ-b
  • p.68 (3) bəːǰiŋd -> bəːǰiŋ-d
  • p.70 (2) aː-səŋ-b1 -> aː-səŋ-b
  • p.75 (4) gloss: learn-2 -> learn-cond (need to add "cond" to the list of abbreviations on page xi)
  • p.77 (2c) gloss: mother-Per -> mother-2
  • p.79 (2b) irsəŋ -> ir-səŋ
  • p.79 (2d) šiːbuː -> šiː [tab] buː
  • p.83 3.3.3 /bu/ -> /buː/ (?)
  • p.85 3.3.4 [-jo/ -> [-jo]
  • p.86 (9) gloss: insert you at the beginning
  • p.89 (1b) gloss: mother-Per -> mother-1


Empires of the Silk Road

Christopher I. Beckwith. 2009. Empires of the Silk Road.


Beckwith という人は言語の人だと思っていた。しかしこの本では歴史をばりばり書いている。どうやら私の思い込みだったらしい。bibliography に挙げられた本人の論文一覧を見ると、昔から歴史をからめて議論してきた様子。

本文では主に歴史を扱っている。言語の議論は appendix と endnotes に押しやられている。私にとっては、むしろこっちが中心。本文は endnote の文脈を知るためにある感じ。内容的には、過去の論文で提示したものだったり、それが微妙に修正されていたり、初めて見るものだったり。初見のものでは、例えば、「月氏」の「月」が *Tokʷar/*Togʷar を写しているとか、「羌」が印欧語の *Klānk- から来ていて、「戦車に乗る」という意味だとか、にわかに信じがたいが面白い。

本文は完成度が低い気がする。(多分読者にとってあまり馴染みのない) 固有名詞が十分な説明なしにどんどん流れていく。これは Amazon (米) の書評でも指摘されている。著者の主張は prologue と epilogue にある。ここで挙げている論点ごとに章を立てた方が良くて、通史の形にしない方が良かったのではないか。通史としては、他人の論文のつまみ食いになっていて、議論が深化していない。自身の主張を証明する論拠という形で資料が配置されていないので、証拠もなしに主張だけしているような印象を受ける。

著者は Central Eurasian Culture Complex という野心的な概念を提案し、それが中央ユーラシアの諸集団に広く当てはまることを述べる。中央ユーラシアというと、トルコ系やモンゴル系を思い浮かべるが、著者は印欧語話者にかなり重点を置いている。また、印欧語話者の中でも、スキタイだけでなくローマなども Central Eurasian Culture Complex の例に挙げる。

これに対する支那などを大胆にも peripheral people と呼んでいる。peripheral people についても、中央ユーラシアとの関係で適宜言及している。しかし、その選択が恣意的に思える。シュメールやアッカドが一切出てこず、いきなりスキタイのような印欧語話者の話から始まる。まるで印欧語話者が文化を創始したかのような錯覚を受ける。

本文の最後の 1/3 ぐらいが "Modernism" 批判。Amazon の書評でも大不評。"Modernism" という言葉の意味は特殊。著者がやたら力を入ているが読み飛ばした。現代史を無理やり要約されても新情報がないのも問題だが、そもそもの不幸の原因は、この表題の本を手にとった読者がそんな議論を求めていないことにある。


  • Chapter 4. p.106. 日本で comitatus にあたるのが「舎人」と言われてもピンとこない。著者の理解では、古墳時代から武士の成立までが連続しているらしい。殉死の習慣も連続的らしい。しかし、間に平安時代が挟まっているし、軍人の構成員も変化している。通俗的な理解と異なるので、強い主張であれば通説への反論を書いて欲しいところ。引いている Farris 1995 に何が書いてあるのか気になる。
  • Chapter 9. p.225. 明政府が反乱鎮圧を求めてドルゴンを招いたという。Ming government とは呉三桂のことを指しているのだろうが、misleading にもほどがある。
  • p.226 からのオイラトの説明が謎すぎる。そもそもオイラトが Noghay Horde の一部だったとか、一部がロシアに服属したとか。この部分の記述は、主にPerdue 2005 という未読の本によっている。表題が China Marches West: The Qing Conquest of Central Eurasia というところから推測するに、オイラトの史料をちゃんと扱っていないのかもしれない。 追記: その後自分で読んでみたら、ちゃんとした本だった。
  • p.227. ジュンガルの Khara Khula "Khan" というが、ハラフラはハーンになっていないはず。同様に、Baibaghas も Oirat khan ではない。宮脇淳子氏の論文を調べたら、ザヤパンディタ伝にはバイバガスをハーンと呼ぶ記述があるらしい。間接的な記述。「Khara Khula は1634年にハーンを名乗ったが、チンギス裔でないことから翌年殺された」と断定している。ハラフラが混乱の中で死んだというのは確認できるけど、ここまで断定的な記述をするからには根拠が知りたい。
  • Chapter 10. p.235. モンゴルの旗の説明をする時に八旗の話をするのは misleading。ニルが 300 人から構成されていたという話はほとんど無関係。
  • p.236. Battle of Jao Modo とか書かれるとガックリくる。モンゴル語の地名はモンゴル語で写して欲しい。
  • p.243. General (Shôgun) Toyotomi Hideyoshi と言われるとずっこける。
  • p.258. footnote 80. チベットがモンゴルの影響下でずっと統一されていたというのは明確な誤り。この部分、論文を引いていない。
  • Epilogue. p.345. Perdue 2005 を長々と引用している。ここが突っ込みどころ満載。Altan Khan を grandson of Batu としている。Batu Möngke のことか。the twelve Tümed (ten thousand-man units) としているのも misleading。Tümed 部があって、それが 12 の集団から構成されていたということ。Tümed の語源が tümen の複数形であっても、tümen が 12 個という意味ではない。


The Empire of the Qara Khitai in Eurasian History

"The Empire of the Qara Khitai in Eurasian History" by Michal Biran

気の利いた感想を書くには新情報が多すぎた。契丹の歴史も詳しくないし、ペルシャ語圏はもっと知らない。Juwaynī とかも名前しか把握していない。

地名については、Oxus, Jaxartes, Transoxiana とかギリシャ語由来の名称を使っている。Āmū Daryā, Sīr Daryā, Mā warā' an-Nahr としたほうがまだなじみがある。


とりあえず漢字を間違えまくっているのに気づいたので errata。念入りに調べたわけではないから他にもあるはず。


  • p.223: 天佑黄帝 should be 天佑皇帝
  • p.225, column 4, row 4: Xi 系 should be Xi 奚
  • p.231: Abaoji 阿保机 should be Abaoji 阿保機 or see Yelü Abaoji (simplified character?)
  • p.231: Caixiang 相 ? (Zaixiang?)
  • p.233: Song Huizong 宋[魚山一魚攵]宗 should be Song Huizong 宋徽宗
  • p.236: Tang Suzong 唐[魚禾]宗 should be Tang Suzong 唐肅宗
  • p.236: Wugusun Zhongduan 吾古孙仲端 should be Wugusun Zhongduan 吾古孫仲端 (simplified character?)
  • p.237: Xi 系 should be Xi 奚