Manchu Literati Names in the 19th Century

Giovanni Stary’s A Dictionary of Manchu Names (Harrassowitz, 2000), based on the Jakūn gūsai Manjusai mukūn hala be uheri ejehe bithe (八旗滿洲氏族通譜, “Comprehensive Genealogies of the Manchu Clans and the Families of the Eight Banners,” prefaces dated 1736 and 1745), is an impressive and handy inventory of Manchu names, but also a list frozen in time. The compilation date of the Genealogies itself makes it clear: it covers, in Stary’s own words, “from Manchu prehistory to the first decades of the 18th century, comprising in some cases up to twelve generations” (p. ix). This chronological span necessarily enables a fair good typology of Manchu personal names (as opposed to hala or mukūn names) up to the Qianlong period: animals, plants, things, birth order, topographical features, “strange and amusing,” parental inspiration, colors, ethnonyms, numbers, etc. (p. xi f.). Another study of the Genealogies by Yingsheng 瀛生 of the Aisin Gioro clan (《談談滿族人的姓名》,滿族研究 1985.2) gives a more basic, morphological typology (p. 59):

  • Names that end with the Mongolian commitative marker -tai or -tu (e.g. Bayantu).
  • Names that end with the Manchu nominalizer -ngga/-ngge/-nggo (e.g. Bayangga).
  • Names that end with the Manchu causative marker -bu (e.g. Tatabu).
  • Manchu nouns or adjectives.

Studies of later Manchu names, however, are somewhat more difficult, not least because both translingual practices and normative regimes come into play. In an article titled 《满族姓名历史演变初探》 by Ma Jingyu 馬競淤 (滿語研究 2011.1) we find the genealogy of the Hitara clan compiled in Guangxu 23 (1897), in particular the names of people in the 11th through 18th generations.

  • 第十一世 :德伯讷、官保 、拐图、麻色、归古里、宁武里 、图力根 、巴哈(章京)、沙哈
  • 第十二世 :永常、永平 、木特布
  • 第十六世 :阿唐阿、海保 、海凌 、海泉、连庆、富隆阿、恩喜 、连喜 、连德、硕成、硕泰、纯庆
  • 第十七世 :广禄 、广祥 、广喜、光珍 、奎珍 、庆珍、富隆布、恒顺、恒龄、恒庆 、恒富、恒庆
  • 第十八世 :宝元 、庆德 、庆余、庆林 、庆贵、庆琛、庆有

From this the author concludes: 1) in the 11th generation all names were in Manchu, and Chinese names begin appearing in the 12th generation; 2) even in the 16th generation bisyllabic names are common, traditional Manchu name endings such as -boo, -tai, -ngga can still be found; and 3) the 17th and 18th generations, which corresponds to the Guangxu reign (r. 1875–1908), show strong features of Chinese names, even adapting a zibei 字輩 system wherein every generation shares a same character in its name, taken sequentially from a poem, which in this case is “宝德毓英魁 ,永成盛世书,隆文多富贵 ,福寿庆双余” (p. 35).

The problem here is that, as always when we deal with Manchu strategies of existing, the line between “Chinese” and “Manchu” is difficult to draw–so difficult, in fact, that they force us to ask instead if drawing such a line is meaningful in the first place. It is often mentioned that the Qianlong Emperor (r. 1735–1796), unsurprisingly, had strong opinions about what a “proper” Manchu name should look like: no trisyllabic personal names because they look Chinese, no trisyllabic names that include a surname (even a Manchu one), and no bisyllabic personal names whose first character is a Chinese surname (ibid). That’s asking for a lot: it’s not enough that a Manchu does not adopt a Chinese name, but that their personal name, when written down in Sinographs, cannot look like a Chinese name–even when you squint. I mean, do the Chinese even have trisyllabic personal names? And since when did transliteration of non-Sinitic names into Chinese not have a tendency to use semantically significant graphs (cf. my previous post on naming)?

These are harsh standards befitting an emperor set to regulate every aspect of Manchu cultural life with an eye toward a “pure” Manchuness that’s supposedly lost, but there is no good reason for modern scholars to follow it to the letter when trying to make sense of Manchu personal names. On the contrary, the relation between Manchu naming practices and Sinography can best be gleaned between the two scripts (instead of focusing only on one or the other as more original or more real), in translingual moments of writing. I draw my examples, again, from the Manchu Rare Book Collection at the University of Chicago Library, where a large number of books contain author, translator, or editor information in both languages, in the interlinear hebi 合璧 format.

A simple way of telling Manchu and Chinese names apart in early Qing books is that, regardless of the Chinese character used or their number, Chinese names have a strong tendency to be transliterated monosyllabically. Take the three postscripts of the Han-i araha sain be huwekiyebure oyonggo gisun (御製勸善要言, preface dated Shunzhi 12 or 1655, M1681/3243), written by three different officials.

Here are the three names: Dang Cung Ya 黨崇雅, Tuhai 圖海, and Fu I Jiyan 傅以漸. It is quite clear, from just looking at the transliteration in Manchu, whose name is supposed to be Chinese and whose is not.

This pattern continues into the many court publications of the Kangxi reign (1661–1722), which tends to include very elaborate list of those involved in each publication’s production (a feature sadly dropped in many Qianlong-era publications). Here is the list of copyists of the Chinese text (Nikan bithe be gingguleme arahangge 謄録漢文:) of the Beye dailame, wargi amargi babe necihiyeme toktobuha bodogon-i bithe (親征平定朔漠方略, preface dated Kangxi 48 or 1710, M2785/3133M):

Li Heng Joo 李恒烑, Dung Je 董哲, Wang Doo Hing 王道烆, Gin Pu 金璞, Wang Fung Sun 王鳳孫, Yan Yung 閻詠, Gu Dzun 顧燇, Ni Pan 倪璠, Jiyoo Šoo Dzu 焦紹祖, Geng Guwe Han 耿國翰, Gin Ki 金𤦺, Li Jen 李湞, Dung Šoo Mei 董紹美, Ju Ting Fung 朱廷鳳, Fang Ceng Yuwan 方承源, Yuwan Poo 袁袍, Hūwang Ting Ioi 黃廷鈺, Jiyang Yan 蔣琰, Wang Hiyo Siyūn 王學㢲, Ju Ts Fang 朱嗣芳, Lin Ši Giyūn 林世俊, Tang Bing I 唐秉彛, Yuwan Guwang Ioi, 袁廣譽 Wei Hūng Doo 魏宏道, Ju Moo Hi 朱懋熹, Wang Guwe Dung 王國棟, Ciowan Ju Jy 全朱芝, Li Kung Hiya 李孔嘉, Ioi Ši Moo 俞時懋, Ioi Pei 于沛, Gu Cūn 顧焞, Gu Bing 顧炳.

Here, on the other hand, are those of the Manchu text (Manju bithe be gingguleme arahangge 謄錄滿文):

Ciši 齊世, Yatu 雅圖, Tungtai 佟泰, Fuhai 福海, Sambooju 三保住, Dunggalai 董阿頼, Cangge 常額, Bešeo 栢壽, Canwen 禪文, Badeboo 巴德保, Cangju 常住, Sucengge 蘇成, Tuktan 圖克坦.

Note, here, that a number of names can be suspicious if you only look at the first character of the Chinese transliteration, for example 董阿頼, 常額, and 蘇成 which all begin with a common Chinese surname. And while it is hard to imagine a Chinese literati would be named 阿頼, a name like 蘇成 could very well be a Chinese name–until you look at the Manchu transliteration, Sucengge. It is intuitive, of course, that copyists of the Chinese text would have distinctively Chinese names and likewise for Manchu, and the list of names become less homogeneous when we turn to other roles, for example the translators (Ubaliyambuhangge 繙譯):

Masai 馬賽, Ajintai 阿金泰, Ulcihai 呉爾齊海, Marsai 馬爾賽, Cangling 常凌, Ciši 齊實, G’u Ioi Lin 郭毓麟.

Again, no shortage of trisyllabic names here, nor transliterations of Manchu names that could easily “pass” as Chinese; but there is little ambiguity if and when both scripts are taken into account. Although I must confess that names such as Cangling 常凌 are so convincing as Chinese ones that I almost suspected the bisyllabic transliteration to be a case of “Manjurification”–of a bilingual Chinese adopting a Manchu manner of transliteration–although this does not seem to be the case. At least for this name. But it is not out of the picture that Manchu naming and transliteration practices have co-evolved, so to speak, so that Manchu names that can yield transliterations that look like more elegant Chinese names enjoyed a slight preference. But they remained Manchu names nonetheless.

Although it is more difficult in later periods of the Qing to find an extensive list of contributors to a publication like this, there is no reason to believe that this distinction was ever changed, at a fundamental level, during the entirety of the dynasty. Instead of more High Qing examples, I will simply give the list of contributors of the Ubaliyambbuha amba tacin-i jurgan be badarambuha bithe (繙譯大學衍義, edict for retranslation dated Xianfeng 6 or 1856, M895/4322M):

Uheri tuwahangge 總校: Wenking 文慶, Muyen 穆蔭, Hilin 熙麟

Kimcime acabuhangge 挍正: Meng Boo 孟保

Narhūšame acabuhangge 詳校: Wenšeng 文盛, Beling 柏齡, Jungcang 中常, Sungliyan 嵩連

Bargiyame asarahangge 收掌: Meng De Fang 孟德芳

Gingguleme arahangge 恭錄: Cungheo 崇厚, Canging 長英, Boogan 寶安, Cengšan 成山.

And here we see the preference for Manchu names that transliterate sophisticatedly into Chinese in full swing. In fact, it is quite likely that some of these names are simply “Chinese names” phonetically transliterated into Manchu (I find no “Wenšeng” in the Genealogies, for example), but they are not transliterated the same way the name of an actual Chinese person would be. In other words, whether the person is Manchu or Chinese remains entirely unambiguous for those who read between the lines, so to speak.

Now, all of this serves as background for a peculiar form of Manchu literati names that can be found in 19th century. Consider the list of contributors to the famous translation of the Strange Tales from a Chinese Studio into Manchu by Jakdan, in particular that written in vol. 1 (since the list can vary from volume to volume) of the 1907 Eryou zhai 二酉齋 edition (prefaces dated Daoguang 28 or 1949, M5748/4242M):


gulu fulgiyan-i manju gūsai Jakdan Sio Fung ubaliyambuha 正紅旗滿洲扎克丹秀峯翻譯

kubuhe šanyan-i manju gūsai Salin Žun Tang acabuha 鑲白旗滿洲薩霖潤塘校正

šandung-ni peng lai hiyan-i Mucihiyan Ioi Fan dasame acabuha 山東蓬萊穆齊賢禹範覆校

gulu fulgiyan-i manju gūsai Deyetai Wei I 正紅旗滿洲德音泰唯一

kubuhe fulgiyan-i manju gūsai Kingsi Hi Cen kimcime toktobuha 鑲紅旗滿洲慶錫熙臣恭訂

gulu fulgiyan-i manju gūsai Canghing Siyang Pu 正紅旗滿洲長興祥圃

The names, quite clearly, follow a more complex pattern. Manchu names no longer appear unadorned, but are accompanied by the style names of the literati. It is quite clear that which part of the compound names are “Manchu” in the sense I have just described–having multisyllabic transliterations–and which parts are “Chinese.” In general, personal names are “Manchu” and style names are “Chinese.” Jakdan himself would be the one to tell us, for example, that 禹範 is the style name of Mucihiyan in his biographical sketch of his friend; but in doing so, he also shows idiosyncratic ways of using Manchu-Chinese hybrid names of this kind. Here is the first sentence of his biography of Mucihiyan:


Mr. Mu is from Penglai in the Deng commandery. His personal name is Qixian, his style name Yufan, and his art name Youlian.

So we have a culturally-specific, if not script-specific, way of using Manchu names. The first syllable of the Sinographic transliteration of Micihiyan is used essentially like a Chinese surname, to create a biographical introduction legible in the Chinese monolingual context despite the character 穆 having little to nothing to do with any hala name Mucihiyan might have. (Mucihiyan returns the favor, of course, and refers to Jakdan as 扎翁 or Elder Zha in his own preface to the translation.)

In another preface (the Manchu Strange Tales has many prefaces, to put it lightly) Jakdan goes further in this way of using Manchu names:

余嘗與穆禹範,長祥圃,德唯一為友[. . .]

I was friends with Mu Yufan, Chang Xianpu, and De Weiyi [. . .]

(Don’t be fooled: Deyetai and Kingsi were Jakdan’s students, but the old master is just very humble.) Now not only is Jakdan taking the first character of the Sinographic transliteration of Manchu names as a de facto surname, he is using it to further make the surname+style name form typical for referring to Chinese literati: think of 李太白 or 胡適之, for example. And indeed, while he prefers to sign his own prefaces (yes, he wrote multiple ones) with his art name 五費居士, he occasionally uses 扎秀峰 as well.

It is often noted that the clan name, or hala, of Manchus were not conventionally used on an everyday basis. Thus the compound name Bujilgen Jakdan, while correct, is a somewhat anachronic expression. It is therefore not hard to imagine why Manchu literati such as Jakdan and Mucihiyan would create these complex forms of names that simultaneously preserve aspects of Manchu naming (i.e. multisyllabic in Manchu script) and afford the full range of pseudonymic display enjoyed by their Chinese counterparts.

And those Manchu names can get quite creative. Let me conclude this post by giving the list of editors (uheri acabuha 仝校) of a Guangxu-era retranslation of the Qianlong Four Books (M856/6156B), with the preface by Siyangheng 祥亨 dated Guanxu 16 (1890):

仁菴金壽 Žin An Giašeo,介堂寶俊 Giyai Tang Boogiyūn,秀濤良續 Sio Too Liyangsioi,鶴門寶壽 He Men Boošeo,馴卿蘇芳阿 Siyūn King Sufangga,文軒盛昌 Wen Hiyan Šengcang,如山有壽 Žu Šan Iošeo, 虛如扎魯布 Hioi Žu Jalubu, 禮臣多貴 Lii Cen Dogui, 筱波德海 Siyoo Bo Dehai, 敷臣志寬 Fu Cen Jikuwan, 容齋培寬 Žung Jai Peikuwan.

Here, the style names are moved up and the personal names, which not only are transliterated multisyllabically but also follow “traditional” Manchu naming conventions surprisingly closely, appear in the second position. The style names can be very self-consciously sophisticated or literary (鶴門, 虛如, and 筱波 are such examples) or appear almost to be official titles (e.g. 馴卿 and 禮臣). 仁菴, 介堂, 文軒, 容齋, and such are, instead, more art name-like, since those conventionally refer to the literatus’ abode. This rather elaborate display of names, interestingly enough, was not followed by the two Chinese patrons of the translation, Žui Boo Cen 瑞寶臣 and De Jing Ting 德靜亭, nor by the team of proofreaders (gingguleme araha 繕正) with only ordinary Manchu names:

榮海 Žonghai, 寶明 Booming, 德培 Depei, 額爾德穆 Erdemu, 文明 Wenming, 富勒鑑 Fulgiyan, 恩芳 Enfang, 海林 Hailin.

It is somewhat unclear why this is the case. There may be a difference in status than trumps personal preference. But regardless, these Manchu-Chinese hybrid names for literati is a fascinating phenomenon worthy of a closer look.

Appendix Jul. 13:

As it often happens, I learn of something that benefits the argument of my blogpost the moment I finish writing it. Today I processed an interesting Manchu grammatical work titled Dasame foloho manju gisun-i untehun hergen-i temgetu jorin bithe (重刻清文虛字指南編, Guide to Manchu “Empty Characters,” The Revised Edition). It was originally published by a Mongol named Heo Tiyan Wanfu 厚田萬福 (preface dated 1884) and was revised and republished by his student Fungšan 鳳山 (preface dated 1894), of the Hanjun Eight Banners (ujen cooha).

When the ethnicity of the names are uncertain, one is left to make conjectures; that was how it was with many names in my original post. It is no longer the case here: we see that a Mongolian literatus taking up the naming convention typical of 19th century Manchu literati, and also a Han Chinese bannerman transliterating his name into Manchu letters in the Manchu fashion. Further confirmation, I think, that instead of substantialist distinctions between “Manchu” versus “Chinese” names as a locus of identity differentiation, we are better off thinking in terms of Manjurisms and Siniticisms in naming practices that are often co-present, and mix-and-match through negotiations and improvisations, and tells us a different kind of story of how bannermen of different ethnicities dealt with the difficulty of names.

Unclean data

To procrastinate I put a bunch of papers I recently wrote into a word cloud generator just to see what would show up. Well, instead of learning about keywords of my research I learned more about the few turns of phrase that I use way too often, and a few verbs I should really know more synonyms of. In any event, after briefly cleaning up the data here’s what I got, which I find interesting–but at least unsurprising–to look at.

The Fieldification of Thinking

A propos of nothing, I confess with pleasure that I’ve always had trouble taking the idea of an academic “field” seriously. That someone genuinely interested in a topic–say, materiality in the 18th-century Chinese novel–could peacefully sleep at night having deliberately chosen not to read–and to lose themselves in the methods of–some dozen of other academic “fields” that would enrich, be it through comparison or parallax, one’s own inquiry because they fall outside one’s “field” is simply baffling. So many disciplines, in praxis if not in conviction, are defined by methodological exclusions: rationalized refusals to historicize or to theorize, to engage with this or that tradition that are either too old or too new, to enter into the archival or technical weeds or to place one’s own method and concepts in philosophical context. To hypostatize those exclusions and to build them into the presuppositions of academic infrastructure misconstrues the very nature of thinking. It is one thing to recognize them as the necessary evils of the modern university qua administered institution, but quite another to internalize those straight-jackets and mistake them for markers of professionalism or rigor.

I was therefore delighted to read the following lines in René Girard’s introduction to Detachment (add your own accent aigu), a book by Michel Serres who, incidentally, went through a similar change of disciplines as I did and was influenced–at a fundamental level of thought even as humanists–by similar paradigm shifts in mathematics (him: the Bourbaki seminars and me: category theory). Here’s what Girard writes:

Michel Serres is not an author for those people whose intellectual life consists in “keeping up with the literature” in one of our constantly shrinking “fields” and in believing that steady progress is being achieved simply because, as the field gets smaller, the objects left in it look larger.

I can’t be the only one wanting to know how the fieldification of thinking happened, not in scholarly thinking per se but in all of its preconditions such as publicity-related decisions made by book and journal publishers, or the design of curriculums in pedagogy.

Here, I can’t help thinking about Michael Barany’s article on how the publication of abstracts in Zentralblatt für Mathematik and Mathematical Reviews fundamentally transformed how mathematicians frame and conceptualize their research in general. It is not hard to notice the proliferation of keywords in the titles of books, articles, talks, panels, and conferences, as if the legibility of thought is increasingly defined by the findability of digital objects in massive online databases like JSTOR. The genre of “article title generators,” which if I recall correctly started as a joke among mathematicians, became something different for humanities papers (here’s one for musicology) because the haphazard throwing-together of keywords reliably churn out titles that we all know to be actually feasible. But surely, thinking with a small set of keywords (even “good ones” like embodiment, affect, or whatever else floats your boat) is not always the most effective way to conduct and present everything that we do; they lock us into too much baggage, and offers only ready-made shortcuts of thought in return.

Am I in a ranty mood today? Not really. It’s just that a small bite of sweets make everything else all the bitterer. Or: I read Feyerabend a little too early in life, and now it’s too late to turn back.

The Natures of Language of Ming China

I’ve been reading Nathan Vedal’s excellent monograph The Cultures of Language in Ming China, whose title I deliberately misconstrue (as a low-commitment nod to recent anthropological theory) in the title of this blog. It’s, in short, a must-read for everyone working on literature, history of ideas, comparative thought, “theory” of various traditions and inclinations, and other topics. Not just specialists of premodern China or East Asia but those who care about the history of the humanities at large–I’m thinking, in particular, of scholars of global modernity who too often only gestures vaguely at pre-1850 East Asia as determined by a few “traditional” characteristics. Vedal digs into the real archival depths to resurface with excellent discoveries, thoughtfully synthesized.

I had, for a more serious occasion, written a longer review of the book, of which I include the last two paragraphs that summarize the gist.

But if the matters of concern at the core of Ming philology can be summed up in one sentiment, it is not the celebration of “shu tong wen” (having the same system of writing) propounded by the Classic of Rites but Southern Song encyclopedist and historian Zheng Qiao’s (1104–1162) poignant answer to his own question: “How is it that the Gautama Buddha’s writings can penetrate the various parts of China but Confucian writings cannot reach the Airavati Rover?” In the face of this asymmetrical history of cultural exchange, Zheng Qiao became the first of many to suggest the close study of language as the path to an answer, stating in his answer that “it is because there are obstacles to the passage of tones and sounds”(1), meaning the resistance Chinese writing faced in its claim to cultural universality had to do with the underdevelopment of its phonographic functions (2). Indeed, if the vestiges and traces of the Sanskrit language seem to surface incessantly like a specter in the pressure points of Ming philology, it is because the “Chinese character culture sphere” paradigm has become so habitual that it overshadowed another highly significant cultural and linguistic “sphere” to which China (as well as the remaining members of the “Sinosphere”) historically belonged, one which China shared not only with familiar kingdoms such as Goguryeo and Heian but also less-discussed ones in East Asian studies such as Sogdia, the Tangut Empire, and Ilkhanate of the Mongol Empire. In the “Mahāyāna Sphere” whose significance for Chinese (secular) intellectual history Vedal’s monograph presses us to better acknowledge, China had always understood itself to be peripheral (3). In a sense, Siddham-like writing systems were to second-millennium Chinese scholars what Sinography and Hieroglyphs were to early modern Europeans: a system of signs as linguistically opaque as they were philosophically and philologically pivotal (4). They posit ready-made forms of linguistic universality that Chinese literati strove to approach through processes of appropriation and negotiation, enacting the very strategies of localization usually ascribed to the border-dwellers of another imagined world in which China considers itself to be the origin and the center.

Such decentering of the “Chinese character culture sphere” is but one of the many insights those outside the field of early modern Chinese intellectual history might learn from Vedal’s monograph, but its significance bespeaks the generativity of his method of taking the Ming’s cosmographic philology on its own terms. The limits of one’s language, Wittgenstein writes, mean the limit of one’s world (5). To which might be added: there are few aspects of one’s worldview (the division of culture and nature, the difference between “us” and “them,” the definition of beauty, the sense of historical existence) that does not in some way inform—and is in turn informed by—how one talks about and does things with language itself. The challenge posed by Ming studies of language as Vedal has presented it (and one notes with regret that he had not included even more vernacular and fascinating registers of metalinguistic discourse in mantic, ludic, and sophic texts) is that they prompt a rethinking, both at the level of historiographical methodology and in the domain of early modern Chinese intellectual history, of the relationship between words and worlds. By patiently excavating the operating forms of thought within early modern Chinese philology (before its closure upon itself in the eighteenth-century), Vedal’s book makes manifest multiplicities, tensions, and complexities in East Asian understandings of language that had too often been put under erasure. But in doing so, he also—deliberately or not—shows what a generalized philology would have as its subject: no longer textual materials as a closed system, but the continuous enfoldment of the cosmos into language into the cosmos.


(1) Zheng Qiao, Tongzhi ershi lüe 通志二十略 (rpt. Beijing: Zhonghua shuju, 1995), 1:354. Translation from Victor H. Mair, “Cheng Ch’iao’s Understanding of Sanskrit: The Concept of Spelling in China,” in A Festschrift in Honour of Professor Jao Tsung-i on the Occasion of His Seventy-Fifth Anniversary (Hong Kong: Xianggang Zhongwen daxue Zhongguo wenhua yanjiusuo, 1993), 339. 

(2) Cf. Shuheng Zhang and Victor Mair, “Between the Eyes and the Ears: Ethnic Perspective on the Development of Philological Traditions, First Millennium AD,” Sino-Platonic Papers 300 (2020).

(3) I borrow the term “Mahāyāna Sphere” from Matthew T. Kapstein, “Other People’s Philology: Uses of Sanskrit in Tibet and China, 14th–19th Centuries,” in The Space of Meaning: Approaches to Indian Philology, Silva D’Intino and Sheldon Pollock, eds. (Paris: Collège de France, Institut de civilisation indienne: Boccard, 2018), 465–494.

(4) Following David Lurie, Vedal uses the term “alegibility” to denote writings that function by being spectated rather than read (chap. 2). See Lurie, Realms of Literacy, chap. 1. The term is effective, although in Vedal more so than in Lurie, an emphasis is placed on the role of alegible writing in the formation of metalinguistic discourse.

(5) Ludwig Wittgenstein, Tractatus Logico-Philosophicus, trans. C. K. Ogden (London: Kegan Paul, Trench, Trubner, 1922), Proposition 5.6.

Pontification, Enumeration, and Retribution

In Bujilgen Jakdan’s (fl. early 19c.) translation of the “Mural” story in the Strange Tales from a Chinese Studio (the Chinese original was the basis of a famous 2011 film) we find the following words describing what the protagonist Zhu Xiaolian 朱孝廉 sees immediately after he is absorbed by the magical painted mural:

Tuwaci, deyen asari jergi jergi bisirengge, fuhali niyalmai jalan waka, emu sakda hūwašan teku de nomun nomulara, garša nerefi torhome tuwahangge umesi geren, Ju Hiyoo Liyan inu terei dolo barabume ilimbi.


Bujilgen Jakdan (trans.), 《擇譯聊齋誌異》, j. 1, p. 12a.

In the Giles translation (which broke up the sentences in somewhat strange ways):

[. . .] where halls and pavilions stretched away one after another, unlike the abodes of mortals. Here an old priest was preaching the Law of Buddha, surrounded by a large crowd of listeners. Mr. Chu mingled with the throng, [. . .]

I want to focus on a particular Manchu expression Jakdan uses here, namely the activity the old monk performs atop his seat: “nomun nomulara.” The noun (we’ll turn to it soon enough) is a somewhat standard one by the nineteenth century, meaning Buddhist scripture. But the verb, nomulambi in its dictionary form, is a somewhat strange one, to my knowledge a Qianlong-era neologism meaning, specifically, to preach Buddhadharma. It is akin to the English term “to gospelize” in that the verb not only determines the form of the activity but also its communicative content.

Indeed, the verb nomulambi does not appear in the various imperially commissioned “Mirrors” (Ma. buleku bithe) compiled in the eighteenth century that became, over time, the standard lexicons of the written Manchu language. Nevertheless, it complies with standard Manchu morphology when it turns a noun into a verb by adding lAmbi (where A is a, e, or o according to vowel harmony), dropping the final nasal of the noun if necessary:

aisi (benefit) -> aisilambi (to help)

baturu (hero) -> baturulambi (to go with courage)

okto (medicine) -> oktolombi (to apply medicine)

erun (punishment) -> erulembi (to punish)

Of course, not all nouns can be “verbed” in this way–alin (mountain) and alimbi (to receive) are completely unrelated as far as I can tell. But like all good neologisms, nomulambi is meant to be understood immediately even by those who have never seen it before.

So, in what context was the word nomulambi coined? My educated guess is that it had to do with the massive sūtra translation project under the Qianlong reign, which culminated in the Manchu Buddhist Canon (catalog), aptly described by Marcus Bingenheimer as follows:

The Manchu Canon (ch. Qingwen fanyi dazangjing 清文繙譯大藏經, mnc. Manju gisun i ubiliyambuga amba kanjur nomun) must be the least used of all canonical editions of the Buddhist canon. Nevertheless, its creation in the 18th century is a fascinating chapter in the translation history of Buddhist scriptures. More than 90 scholars worked for more than 20 years to produce a Manchu version of the Buddhist canon. While sūtra texts were translated from the Qianlong edition of the Chinese canon, the vinaya texts were translated from Tibetan.

Marcus Bingenheimer, “History of the Manchu Buddhist Canon and First Steps towards its Digitization,” p. 203.

The project, as we now know through a number of historical accounts, began in 1772 and concluded around 1790, although one may suggest that its preparation began much earlier.

Undoubtedly enabled by the establishment of the Tibet-Qing patronage or protectorate relationship in early 18th century, a significant step was taken toward the incorporation of Tibet in the Qing’s multilingual Weltanschauung with the publication of Tongwen yuntong 同文韻統 in Qianlong 15 (1750), an apparently-philological work that in effect hypostatized various Buddhist-inspired currents of Manchu language ideology already in circulation. Its more concrete goal, however, is to demonstrate that the magical efficacy of the Manchu script–qua phonetic writing–as a medium for Buddhist spells or dharanis is on par with Tibetan letter and superior to Sinographs, a goal achieved through elaborate charts of phonetic values across writing systems and ingenious uses of Sinographs that systematizes practices of transliterating Sanskrit already seen in medieval manuscripts. Just as the need to phonetically transliterate spoken Chinese resulted (per traditional historiography anyway) in Dahai’s 1632 Manchu script reform adding new letters, so was the encounter with Tibetan civilization the stimulus for another revamping of Manchu (as well as Chinese) writing technology.

《欽定同文韻統》, j. 1, p. 5a.

This technology of phonographic equivalence, it would seem, was quickly put to use. It is known that, accompanying the Manchu Buddhist Canon was a separately published set of spells drawn from the scriptures included in the Canon in four writing systems: Manchu, Chinese, Mongolian, and Tibetan. It is titled

Han-i araha manju nikan monggo tanggūt hergen-i kamciha amba g’anjur nomun-i uheri tarni


Qaγan-u bicigsen manju kitad mongγul töbed kele qabsurugsan büküli kenjür (?)-ün tarni


And dates from Qianlong 38 (1773). What is much less known is that a trilingual (Manchu, Chinese, and Tibetan) collection of spells, similar in aspiration but slightly different in execution, is extant in a highly incomplete form, and its prefatory verse dates from Qianlong 24 (1759), a whole 14 years earlier. I know of eleven extant volumes of this work in sūtra binding, ten of which are in the Berthold Laufer collection at the University of Chicago (call# M1803/4486) and one of which is described in the London catalog.

East Asian Collection, University of Chicago
East Asian Collection, University of Chicago

Here we see some of the earliest uses of nomulambi, in the translations of Chinese and Tibetan sūtra titles. For example, Taishō 1254, traditionally titled 佛說末利支提婆華鬘經, is titled in Tibetan as རྒྱལ་བས་གསུངས་པའི་ལྷ་མོ་འོད་ཟེར་ཅན་མེ་ཏོག་ཕྲེང་བ་ཞེས་བྱ་བའི་མདོ། and in Manchu as Fucihi nomulaha enduringge uldengge eme-i ilha-i erihe sere gebungge nomun. It is clear that 佛說, རྒྱལ་བས་གསུངས་པའི་, and fucihi nomulaha are meant to be equivalent expressions (my Sanskritist friend has yet to find out what Sanskrit expression, if any, this corresponds to). But while 說 and གསུང་བ་ are ordinary verbs in their respective languages meaning “to enuciate,” it is interesting that the translators deemed its obvious Manchu equivalent, hendumbi (used extensively to translate 曰 in Chinese classical texts), to be inadequate. However Confucius was speaking to Zeng Shen 曾參 was determined to be fundamentally a different kind of activity from how the Buddha was speaking to Ānanda, the latter being most appropriately denoted by a verb derived from the speech’s final product, the sūtra.

It would be interesting enough if the story ended here, but let me push the history of nomulambi even further. I had mentioned that the verb is not included in any of the imperial Mirrors compiled in the 18th century, but I should have added the even more significant fact that the noun nomun also did not appear in either the original 1708 monolingual Manchu Mirror or the 1743 bilingual Manchu-Mongolian one. It is, however, attested in the 1771 Expanded Mirror as well as the much later pentaglot. Here is its definition in the Expanded Mirror:

Enduringge niyalmai toktobuha enteheme songkolome halaci ojorakū bithe be nomun sembi.

The writings of sagely persons, which are to be followed eternally and never deviated therefrom, are called nomun.

The notion described here is simply that of the Classic without denominational restrictions; and indeed, whereas the verb nomulambi was to my knowledge used exclusively in Buddhist contexts, nomun had a much wider purchase and effectively replaced the earlier ging (phonetic transliteration of the Chinese 經, which is a more etymologically accurate translation of “sūtra”). We need only take stock of how the Manchu title of the Shujing 書經 (Classic of Documents) changed from the transliteration Šu ging (attested as late as a commercial imprint dated 1738, or Qianlong 3) to Dasan-i nomun in the Qianlong 25 (1960) court translation, the title now literally meaning “The Classic of Government.”

This Qianlong-era change can be detected outside the realm of proper names as well. The first major Manchu dictionary, the Comprehensive Book of the Great Qing (Ch. Da Qing quanshu 大清全書) published by Shen Qiliang 沈啓亮 (fl. 1645–93) in 1683, contains the following entry:

enduringge ging saisai juwan


沈啓亮,《大清全書》, j. ᡝ, p. 13.

Where both 經 and 傳 are given in transliteration. This general use of ging persisted into mid-18th-century, and we may observe the change most directly in the definition given for bithe (book, or writing in general) in the 1743 bilingual Mirror and the 1771 Expanded Mirror respectively:

1743: ging suduri geren dzi tanggū boo-i banjibuha arahangge be gemu bithe sembi, jai araha hergen be inu bithe sembi.

Classics [ging] and histories, as well as the writings of the masters of the Hundred Schools, are all called bithe. Additionally, written words in general are called bithe.

1771: nomun suduri geren dzi tanggū boo-i banjibuha arahangge be gemu bithe sembi.

Classics [nomun] and histories, as well as the writings of the masters of the Hundred Schools, are all called bithe.

It seems clear, then, that the replacement of ging by nomun, somewhat different from the coining of nomulambi, is part of the process of Manchu language “purification” closely associated with the Qianlong reign, in which previous Chinese phonetic loans in Manchu are replaced by words deemed more “authentically,” if artificially, Manchu. (My favorite example of this, which is widely attested in manuscript “corrections” on Manchu books from High and Late Qing, is the replacement of yin and yang by a and e.) In other words, nomun is a newcomer to an old chair, whereas nomulambi is the fresh occupant of an equally-fresh position.

From 《五體清文鑑》

One may well doubt what makes nomun more “Manchu” than ging, if the word did not seem to exist at all in Manchu before mid-18th century. Its origin is quite obviously the Mongolian nom (which is still in use today virtually unchanged, as ном), and perhaps Manchu users, not least of whom the Qianlong emperor, thought they shared their origins more closely with the Monggo than with the Nikan.

Nom itself has a fascinating history according to Wiktionary:

From Proto-Mongolic *nom, derived from Old Uyghur ᠨᠤᠮ (nom, “book; Buddhist scripture”) into pre-Classical Mongolian, from Sogdian 𐫗𐫇𐫖‎ (nwm /nōm/, “law, canon”), from Ancient Greek νόμος (nómos, “law”) from Proto-Indo-European *nem-.

The most surprising part of this very rich etymological history is that even in Proto-Mongolic, the word was never nomun or another bisyllabic word: in adapting an existing Mongolian word, the Manchus probably added another syllable so that it better aligns with Manchu phonology. (The only word ending with -m in the imperial Mirrors is tulum, meaning a kind of float made from animal skin, also a Mongolian loan.) But in doing so, they also created a noun of the familiar -n form that lends itself to the -n -> -mbi derivation which, ultimately, resulted in the verb nomulambi with which this post began.

Finally: According to Wiktionary again, the PIE root *nem- has the meaning “to distribute” as well as “to give, to take.” It is the origin of the German nehmen, perhaps unsurprisingly. But more wonderfully, the Latin noun derived from this is numerus; that is, the Number of everything (to borrow from the Pythagorean phrasing of Nathan Vedal). And last but certainly not least, from the original meaning of *nem- is derived the name of the Greek goddess Νέμεσις–the goddess of retribution, who never fails to give what is due.

And this is all a very ferguwecuke world (to use one of Jakdan’s favorite words) to be absorbed into.


P.S. After writing this post I noticed that the verb nomulambi appears in the lyrics of a Manchu cover of 愛的供養 (Ma. Buyecun-i dobon), the theme song of one of many PRC TV romance dramas taking as its setting the Qing imperial palace.

The verb appears around the two minute mark. Here’s a transcription of the lyrics from 1:52 to 2:07:





goidara be bairakū

damu dalbade

soktofi yumbureo

elhei nomulambi

Note that Wang Shuo 王硕, the very knowledgable scholar who translated the Putonghua into Manchu, chose nomulambi as the word closest to the phrase 梵唱, or the Sanskrit chanting in Chinese Buddhism that is the topic of a recent dissertation by Kelsey Seymour. Which is well worth a read if you enjoyed this blog.

The Butterfly Effect?

George A. Kennedy’s “The Butterfly Case,” first published in 1955, is a follow-up to the same author’s polemical article four years prior against the “monosyllabic myth” in the study of the Chinese language. The essay begins with an imaginary dialogue between the Harvard-trained classicist and the entry for hu 蝴 he finds in Robert Henry Mathews’ Chinese-English dictionary.

George A. Kennedy, “The Butterfly Case” (rpt. 1964)

Kennedy’s point is not hard to grasp–that there are characters that do not represent monosyllabic words but belong exclusively to binomes–even if the position of his antagonists, namely Chinese is an exclusively monosyllabic language, is today very much a thing of the past. Indeed, in the bulk of the article Kennedy goes on to cite a number of examples that most people today would agree to be simple yet strong demonstrations of multisyllabic morphemes in colloquial Chinese such as the 子 and 兒 suffixes (e.g. 桌子,花兒). Kennedy is not unfair to caution against treating binomes that appear to be synonym compounds automatically as one, and thereby assuming every character must have formerly existed as a monosyllabic word.

But given the weight “蝴蝶” specifically received in the structure of this article, which may be partially responsible for the word’s role as an exemplary binome in subsequent discussions about the “nature” of the Chinese language, I think it’s worth the effort to take a closer look at its history. I must first confess that I’m neither a trained paleographer nor a specialist in historical linguistics; what follows is simply the tentative findings of someone who likes words.

Preliminarily, the first distinction that needs to be made is that between synchronic and diachronic investigations: are we speaking of 蝴蝶 in the context of 20c Putonghua, or in the context of the entire corpus of written and spoken Sinitic languages going back to bamboo manuscripts and bronze inscriptions? The second is that of genre and media: how does morphology interact with factors unique to poetic or prose genres, manuscript or print culture, or vernacular or self-consciously literary utterances? (The presupposition of language as a determinate, regulated system separable from material, technological, or practical modulations is for me highly suspicious.) Much of Kennedy’s essay do not make these distinctions, even if his arguments end up being sufficient for the kind of claims he was trying to make. He writes, for example,

A dictionary has really no business defining GO [蝴] as “butterfly” unless the word [. . .] has been found sometime somewhere in some record for written or spoken language. We must therefore find out for ourselves whether its existence is real.

Kennedy, “The Butterfly Case,” p. 290.

This is very uncontroversial, I think–but also slightly misleading. Demonstrating 蝴 has never been used as a monosyllabic word does not imply 蝴蝶 is the original (binome) form of the Chinese word for butterfly, and in fact tells us very little about the history of the word itself. Kennedy’s method in the following pages, which surveys largely literary texts mentioning butterflies over an extraordinary chronological span, does not exactly help. After discussing an early medieval tale in which the word húdié appears (only phonetic transliteration is given in Kennedy’s article, and I omit the sinograph for reasons that’ll become clear later), Kennedy concludes:

The specific value of the graph GO [] is certainly beyond question. It has never been used for any other purpose whatever but to write the first part of go-dhiap [húdié], and if GO is seen alone, as we saw it in Mathews’ dictionary, it suggests immediately by association the word godhiap, or its dialect descendant. But in this relatively early literature we are confronted with the following situation: If there was in the language a word go “butterfly” or a word dhiap “butterfly,” then it is inconceivable why the writers should have recorded the story with GO-DHIAP [. . .] Evidently the Chinese of that time had not grasped the potentialities in their writing system that are so clear to modern scholars of the west.

Kennedy, “The Butterfly Case,” p. 293.

Ouch (on behalf of the Karlgrens of the world)!

But hang on for a second: isn’t it fairly common to refer to butterflies simply as 蝶 in literary texts, from as early as the medieval period? Here’s a “Poem in Praise of Snow” 《詠雪詩》, composed by Pei Ziye 裴子野 (469–530):


Whirling is the thousand leagues of snow / Haste is the crossing of the dessert / Like clouds that gather and disperse / With winds that twirl and tilt / Caressing grass like linked butterflies / Upon trees like fluttering flowers / I proffer as a gift of farewell / This, instead of celestial blossom” (my translation).

Many other poems from the Liang (502–557) attest that 蝶 was used in isolation, and by the Tang (618–907) no one would fail to recognize 蝶 as meaning simply butterflies, with 蝶戀花 (Butterfly in love with flowers) becoming the name of a tune in court music repertoire. All of this persisted well into the early modern period. Even if we assume húdié (or whatever morpheme it was trying to represent) is the original form of butterfly in Old Chinese, as early as medieval period saw its unhesitating abridgment. What is surprising, in comparison, is that its first character was never used to stand for the whole.

If you are a connoisseur of the Chinese lore surrounding butterflies–or, like many people in historical and contemporary China, have simply read your standard serving of Classics–you probably know exactly where I’m going with this. If húdié has any claim to being the binome for butterfly in early Chinese texts, it is simply due to its appearance in the famous anecdote in the second chapter of the Zhuangzi:


Formerly, I, Zhuang Zhou, dreamt that I was a butterfly, a butterfly flying about, feeling that it was enjoying itself. I did not know that it was Zhou. Suddenly I awoke, and was myself again, the veritable Zhou. I did not know whether it had formerly been Zhou dreaming that he was a butterfly, or it was now a butterfly dreaming that it was Zhou. But between Zhou and a butterfly there must be a difference. This is a case of what is called the Transformation of Things (trans. Legge).

But notice that, in this text that is known to us through the editorship of Guo Xiang 郭象 (252?–312), húdié is written differently, consistently without the 虫 radical in nearly all major editions. In the 《續古逸叢書》 edition used on ctext, notably, it is only in the “Outer Chapters” (of more controversial provenance) that we see an appearance of 蝴蝶, to be followed immediately by the 虫-less form:

種有幾,得水則為㡭,得水土之際則為蛙蠙之衣,生於陵屯則為陵舄,陵舄得鬱棲則為烏足,烏足之根為蠐螬,其葉為蝴蝶。胡蝶,胥也化而為蟲,生於灶下,其狀若脫,其名為鴝掇。[. . .]

The seeds (of things) are multitudinous and minute. On the surface of the water they form a membranous texture. When they reach to where the land and water join they become the (lichens which we call the) clothes of frogs and oysters. Coming to life on mounds and heights, they become the plantain; and, receiving manure, appear as crows’ feet. The roots of the crow’s foot become grubs, and its leaves, butterflies. This butterfly, known by the name of xu, is changed into an insect, and comes to life under a furnace. Then it has the form of a moth, and is named the Qu-duo [. . .] (trans. Legge).

Of course, none of this orthographic idiosyncrasy was observed by Kennedy, whose account of his examination of the Zhuangzi concludes with

The graph GO in this sense occurs in Juang-tz just seven times, never alone, and always followed by DHIAP.

The graph DHIAP occurs in Juang-tz just seven times, never alone, and always preceded by GO.

Kennedy, “The Butterfly Case,” pp. 294–295.

Someone has clearly been reading a lot of prose with parallelism! But more to the point, what Kennedy calls a “graph” is not a graph at all, but two somewhat interchangeable graphs in which one (and the one that has a large range of meanings as a monosyllabic word at that) is clearly favored.

Pelliot chinois 2495, 《莊子三袟合卅三卷郭子玄注》(Source: IDP)

From a particular point of view, this doesn’t really matter. If we work backwards from the modern Putonghua lexicon it is easy to see the term in the Zhuangzi as essentially an earlier appearance of what the term is now, namely, the binome that is the standard word for butterflies. Once this is taken as a given, it becomes possible to conjecture why it is so. Consider the etymology given on wiktionary:

From Middle Chinese *ɦuo dep (first syllable unstressed), from Old Chinese *ɡaː l’eːb, derived from a proto-form of *kʰleːp ~ *ɦleːp, a prefixed form of the root *lep (“butterfly”), which was also the source of Tibetan ཕྱེ་མ་ལེབ (phye ma leb, “butterfly”) and Burmese လိပ်ပြာ (lippra, “butterfly”).

From this, a renowned Sinologist remarks,

This is a prime example of a widespread phenomenon in Old Sinitic.  Whenever we have a disyllabic morpheme or word in Old Sinitic, we can expect that it will reflect one or more of the properties inherent in húdié 蝴蝶 (“butterfly”).  That is to say, it will often have a proto-form that is a consonantally complex monosyllable that undergoes disyllabification or dimidiation and other phonological changes.

I am not remotely qualified to judge whether the disyllabification constitutes a “widespread phenomenon” in Old Sinitic, but I do have questions about whether this may be considered a prime example of it. The problem that immediately presents itself here–besides that the prefix *ɡaː does not appear in Burmese or Tibetan and seems to be included entirely because of the 胡蝶 binome–is that húdié, as far as I know, is not the only term for butterflies in Old Sinitic, and possibly is not even the standardized one.

Reconstructions of OC pronunciation of 蝶 (Source:

The reason, quite simply, is that 蝶 does not appear in the Shuowen at all. Instead we find these two entries (I’ve added in the reconstructed OC pronunciation given in Wiktionary):

蛺 (*keːb):蛺蜨也。从虫夾聲。

蜨 (*seːb):蛺蜨也。从虫疌聲。

蛺蜨 (jiádié in Putonghua pronunciation) is not the only binome in the Shuowen used to gloss each of its characters, and neither characters appear in pre-Han or even Han period texts that I can easily find (the binome does appear in 抱朴子, a fourth-century text). It is quite clearly a genuine bisyllabic morpheme, generally interpreted to mean butterflies. Indeed, even a cursory search through databases like ctext and 中國基本古籍庫 shows that for most of Chinese history, 蝴蝶 had these characters as graphic competitors, and the constituents of the two binomes often mixed and matched to create new binomes such as 蛺蝶 (attested in a number of early medieval poems).

If 蝴蝶 has not always enjoyed a de facto monopoly over butterflies in written Chinese, it might help to think of its history in a slightly new way. Commentators of the Shuowen were well aware of this synonymy and explained it away with relative ease. Xu Xuan 徐鉉 (916–991), the editor of the earliest extant full text of the Shuowen, states simply in the gloss of 蜨 that “俗作蝶”–that is, 蝶 is the “vulgar form” of this standardized character. Later commentators further justified this claim by citing, not the famous “Inner Chapter” tale but the “Outer Chapter” story, also cited above, of a “wuzu” whose leaves become butterflies. To refresh our memory:


The roots of the crow’s foot become grubs, and its leaves, butterflies.

This line is cited, for example, in the 《說文句讀》 of Wang Yun 王筠 (1784–1854). While it is unclear what “wuzu” is in this text, its body parts are clearly described in the idiom of roots and leaves, and it is unsurprising that the butterfly of all insects is included in this process of wondrous metamorphosis. Notably, the graphic pun is here quite blatant: the 葉 (OC: *leb) becomes an insect, a 蝶 (OC: *l’eːb). If we imagine with the text that the falling leaves turning into butterflies (undeniably a beautiful imagery), then the surprising yet subtle transformation here is a well-coordinated one, at once vital, sonic, and visual. We might even conjecture that, in the 《續古逸叢書》edition, the unique appearance of the 虫 radical with 胡 in this sentence is deliberate, serving to emphasize the process of metamorphosis it is depicting by drawing attention precisely to the addition of the 虫 radical to its neighboring graph.

《說文解字詁林》, p. 12936

This visual play of characters would have been very recognizable to medieval scholars, familiar as they were with such “folk-etymological structures” playfully deployed. A simple example (which I take from Imre Galambos’ wonderful and detailed article on this topic):

For example, the character 蘇 (su ‘to revive, regain consciousness’) was often written as 甦, which consisted of the components 更 (geng ‘again’) and 生 (sheng ‘live’), which if read together as two words would mean ‘to regenerate, revive,’ thereby matching the meaning of the character 蘇.

Imre Galambos, “Medieval Ways of Character Formation in
Chinese Manuscript Culture,” p. 63.

While one of the supposed triumphs of modern Chinese grapholinguistics is the demonstration that much of what was traditionally interpreted as huìyì 會意 script forms are in fact xíngshēng 形聲 ones, a closer eye to how sinographs were actually used in manuscript culture shows that graph creation had always been an ongoing process–that in fact, to know how to read (and write) such neographisms, many of which were indeed of the huìyì type, were part of the know-how of sinography in general. While educated folks such as Xu Xuan do not always speak kindly of such inelegant uses of writing (see, e.g., David Branner’s article on “portmanteau characters”), it was always understood that replacing a standardized character (蜨) with a folk-etymological one that may or may not be sonically related (蝶 = 虫 + 枼 or 虫 + 葉 – 艹) was something people did with graphs. Xu Xuan’s description of 蝶 as a “vulgar form” of 蜨 intimates this interpretation of the Zhuangzi, which may have entailed an understanding that a visual-poetic gesture in the “Out Chapters” moved backwards to transform all appearances of butterflies in the entire work.

The extent to which these interpretations reflect the intentions of the “original” Zhuangzi is beside the point, not least because received early Chinese texts is often formed, in their very being, by the continued interpretations it received in the hands of later readers. The bottom line is that–and this the premodern Chinese commentators observed better than many modern linguists–胡蝶 as morpheme or graphemes may need to be considered in explicit connection to the Zhuangzi‘s rhetoric, canonicity, dialect, and other idiosyncrasies. Whether it directly reflects a spoken word in Old Sinitic, it should never be forgotten that the Zhuangzi is in the habit of inventing language as it goes along. Is it inconceivable that the simpler truth for 蝴蝶‘s eventually emerging victorious in its rivalry with 蛺蜨 is that, for many in historical and even contemporary China, it is very difficult to think of butterflies without thinking also of the Zhuangzi? (The dream is “suggest[ed] immediately by association” upon the imagined sight of the flapping wings, to use Kennedy’s language.) So much so that the High Qing philologists often glossed 蛺 and 蜨 simply as “that whose vulgar form is 胡蝶,” and provided only the old master’s dream as example sentences to elucidate the Shuowen forms (which have become de facto variants).

As for the eventual victory of 蝴 (which more so than 蝶 can be seen in the Zhuangzi as a character playfully constructed on the spot) over 胡 in the binome, a case is to be made that verbatim citations of the Warring State text was subsequently subjected to the force of “assimilated forms” (also leihuazi 類化字 “analogical character formations”). Galambos again:

The most obvious cases of assimilation are those where one of the characters in a binome or a common phrase changes under the influence of the other, typically by acquiring the same semantic component. For example, the word fenghuang 鳳凰 (‘phoenix’) in medieval manuscripts and inscriptions was at times written as 鳳皇 but the second character gradually became written with the 𠘨 radical around it, undoubtedly under the influence of the character preceding it. [. . .] This phenomenon of matching the radicals in a binome is probably why many of the “true” binomes ended up with the same radical, even though in the manuscript tradition they perhaps even more commonly appear with dissimilar radicals or without radicals at all. Thus the words hudie 蝴蝶 (‘butterfly’) and zhizhu 蜘蛛 (‘spider’) commonly appear in an array of combinations in which either one or both characters are written without the 虫 radical.

Imre Galambos, “Medieval Ways of Character Formation in
Chinese Manuscript Culture,” pp. 67–68.

If the first appearance of 蝴 in the Zhuangzi served a specific visual-poetic function, its persistence had more to do with our tendency to emphasize (and sometimes create) commonalities and patterns between things we deem closely connected–in this case, two characters held together by the Zhuangzian tale.

Pelliot chinois 2524 (Source: IDP)

Much of this, of course, is speculative; but if it has any merit at all, it is because this history of 蝴蝶 recognizes many more factors in the formation of written words than its passive duty to represent speech. Much more is to be gained about the strange 蝴 dancing in Mathews’ dictionary when we situate written texts amid relations specific to their medium–a procedure rarely meaningless even for scholars of the modern, vernacular language. “Ludic writing,” citationality, manuscript variance, and visual rhythm of the 虫 radical, are all operations that contribute, with overwhelming force, to the continuous formation of a straightforward lexical item; yet it is easy to lose sight of these processes of individuation when we focus only on their apparently static end product, and ask only whether it can be called a bisyllabic morpheme.

All this leads us to the strange possibility that today, every speaker of Putonghua may well be unknowingly citing Zhuangzian wordplay when they speak of (or perhaps even perceive) a butterfly. The sheer popularity of a strange dream story altering, unknown to many, our linguistically-mediated experience of the everyday, natural world: now, that’s a real Texan tornado if I’ve ever seen one.

The Melancholy of Not Starving

Noteworthy cynicism toward First World Problems in a reader’s annotation in a passage from a Manchu-Chinese bilingual Caigentan 菜根譚, a collection of aphorisms compiled in the Ming. Again from the Laufer Manchu Collection, call# M1010.3/2659.

East Asian Collection, University of Chicago

Rough translation:

People know the joy of having fame and rank, but they do not know that the joy of not having fame or rank is far more real. People know the melancholy of hunger and cold, but they do not know that the melancholy of not being hungry or cold to be more intense still.

Whichever regime of spiritual cultivation this aphorism is trying to sell, it does so by appealing to the reclusive hermit in the heart of its readers, giving voice to the ennui of grumbling urbanites. Well, our reader, who was more of a realist rather than a romantic, was not having any of that:

Such has been my wish for a long time; but without taking up an official position, where does the food come from?

Good question.

Naming and Impropriety

For something that, according to some very serious philosophers, intimates the primal baptism that is the ἀρχή (beginning, principle) of language, proper names are a surprisingly slippery and contentious bunch. It was not too long ago when Dr. Elise Anderson pointed out the questionable transliteration in U.S. media of the name دىلنىگار ئىلھامجان Dilnigar Ilhamjan, who is an Uyghur athlete competing in the Beijing Winter Olympics (and, quite significantly, also its last torch bearer).

Take a look at James Millward’s resourceful Xinjiang Syllabus, and you’ll quick notice why the innocent-looking “Dinigeer Yilamujiang” is more problematic than it might first seem. This is just a sneak peak, however, of a larger problem experienced by those with non-Han Chinese names within an Aufschreibesystem designed around a single (and rather idiosyncratic) language. The website High Peaks Pure Earth (རི་མཐོ་ས་གཙང་། ri mtho sa gtsang), which publishes wonderful posts and translations about Tibetan contemporary poetry, music notation, and society (e.g. this interview on students’ view of homosexuality), also maintains a list of Tibetan names and their usual transliteration in Simplified Chinese precisely to avoid this problem. The hope, at least, is that those who come across it (or otherwise have knowledge of the language) would know to transliterate the second part of the name སྲོང་བཙན་སྒམ་པོ (srong btsan sgam po, usually rendered in Chinese as 松赞干布) as Gonpo rather than Gan Bu—and thus eliminating any potential confusion between the founder of the Tibetan Empire and the word for medial officials in the Communist bureaucracy.

Some of my posts are written a propos of nothing, but this one is not. The reason I am reminded of all these contemporary politics of proper names in multilingual Asia was an article written by the great Giovanni Stary titled “A Note on the Transliteration of Manchu Names and Its Mistakes” (2011). Stary gets straight to the point in the opening paragraph:

In researches on Qing history Manchu and Mongol names are not seldomly quoted using the transliteration of their Chinese syllabation, creating thus distorted word-forms and changing their original meaning into incomprehensible “hybrid creatures.” One of the best examples is the Imperial hunting ground “Mulan,” which has no specific meaning, instead of “muran”—that is, “a battue held at the time of the deer-breeding season” (Norman: 205).

Stary gives more examples: the name of the famous Oboi (c.1610–1669) gets “reconstructed” from the pinyin “Aobai” into “Ebai,” and trisyllabic Manchu names such as Tulišen gets retroactively sinicized not only in script but also in structure, becoming a Lishen of the Tu family. What takes the cake for him is the translation of 四牛彔—that is, the fourth niru or military company—as the “plateau of four cows.” (I must add that the interpretation of lu is here as incredible as the mistranslation of Manchu.)

East Asian Collection, the University of Chicago

Something, of course, is always gained in the shoving of linguistic “stuff” across various borders, whether under the name of translation or transliteration. In (re)writing language with sinographs, specifically, it is hard to record the sounds of speech without being ambushed by the semantic sedimentations latent in the characters of choice. In an earlier post, for example, I mentioned the Ming derogatory transliteration of śramaṇa (conventionally 沙門, the male monastic order in Buddhism) as 喪門, or the Mourning Sect. This interplay between the sonic and semantic aspects of a phonographically deployed sinograph, which often emerges between at least two nominally disjoint meaning-making economies (e.g. Sanskrit and Chinese), was a particularly active zone of play in the early Japanese poetry collection Man’yōshū—a phenomenon that David Lurie quite nicely termed “poetic inscription.” Closer to home now, the rendering of non-Chinese company names into sinographs (required by law for everyone looking to make the big bucks nowadays) is an art in itself, cautious as everyone has become after Coca Cola’s disastrous transliteration as “tadpole chews wax” 蝌蚪啃臘 at an earlier time. This wikipedia page gives further examples of “graphic pejoratives.”

But semantically generative as the improper transliterations of proper names can often be, carefully curating a single channel as the only one through which a heterogeneous interior may communicate with the Outside is a simple way to effect centripetal homogenization under the guise of informational efficiency. The fruits of dissemination is never quite evenly shared, and the graphical sinicization of China’s cultural others (a sinicization not by the supposed charming radiance of Chinese Civilization but by 20th- and 21st-century historians and bureaucrats with specific agendas) is a monolingualism—that is to say logocentrism—with distinctly political effects. One is wise to be alarmed when the train of difference stops, pretending it has never traveled at all. When it happens, I urge you to join me in stepping off and doing some of the walking ourselves.

The Cruelest Month

The news from Shanghai in the last month have been heartbreaking in many ways, made all the worse by the common knowledge that (and there’s no denying it at this point) it is almost entirely a human-made catastrophe. It was the shattering of many illusions, to risk putting it lightly.

Recently a “public account” on WeChat (something akin to a blog internal to the WeChat ecosystem) named Yongyuan de caomei yuan 永遠的草莓園 (The Eternal Strawberry Garden) created this video that compiles many audio recordings leading up to and from the long and tortuous lockdown. “Content creators” (as we now call them habitually) of this independent type in the Sinocybersphere never have the luxury of platform-sponsored ownership of their own creations, not because pirates are quick to steal their texts and videos to turn a dirty profit, but because without their being copied, digitally manipulated, and creatively disseminated by guerilla-like WeChat users with or without the creator’s consent, such information would never survive the censors in the first place.

This is why I find numerous YouTube channels uploading this video in the last few hours, all of which without even the pretension of originality. Here is one such copy.

[Note from two hours after this blog was written: I’ve replaced the video I initially linked with one that has English subtitles.]

“The Sounds of April,” the video is humbly titled. In lieu of a full translation let me give a summary of the first three minutes of audio files included herein:

[0:00] Press conference on March 15. 5 new symptomatic cases, 95 new asymptomatic ones. “There is no lockdown in Shanghai, and there is no need for a lockdown in Shanghai . . .”

[0:22] Press conference on March 26. 45 new symptomatic cases, 2631 new asymptomatic ones. Shanghai cannot enter into lockdown because the economic impact of such a lockdown (even one of a few days) will be grave not only for the city itself but, more crucially, for most of the country.

[0:52] Title: The Sounds of April

[1:05] April 18. 3084 new symptomatic cases, 17332 new asymptomatic cases.

[1:13] April 1. Phone conversation between a patient and a director in the Shanghai Pudong Disease Prevention and Control Center. They are running out of beds, quarantine centers are running out of rooms, and the medical hotline running out of ambulances. Medical professionals felt hopeless, and comment sections were closed for the city’s official social media accounts.

[1:30] April 1. The cry of a baby being separated from their parents for quarantine.

[1:37] April 2. Volunteer truck driver bringing food to Shanghai, but the produce is rotting in his truck because no one is picking them up and distributing them to residents.

[1:50] April 2. Supplies designated for Baoshan District were sent instead to Pudong District.

[1:59] April 2. Residents thanking volunteers and medical professionals.

[2:07] April 2. Residents demanding that food and supplies be distributed.

[2:14] April 3. Chair of local residence committee in tears. The policies from the higher-ups were terrible, and she cannot face the residents she’s responsible for.

[2:28] April 6. Distraught police to an elderly man breaking lockdown orders: at least you have a home you can return to, what about me?

[2:33] April 6. Corgi beaten to death in the street after the owner was sent to quarantine center. “Are they trying to kill it? . . . Oh my god!”

[2:37] April 6. Neighbors helping each other.

[2:48] April 7. Police providing food to truck drivers delivering supplies, who haven’t eaten in a few days.

[2:56] April 7. Young man’s father is very sick, but no hospital would receive him.

[3:09] April 7. Food delivered for residents being disinfected and disposed as garbage by security personnel enforcing the lockdown.

Let me stop here because surely, someone will soon add subtitles to this video and render my rough translations unnecessary.

Anyone with friends or relatives in Shanghai would know that these are just tips of the iceberg. Chaotic mis-coordinations, recklessly enforced orders, pathetic attempts at creating a positive narrative turn everyone into victims who—just to make everything worse—are always already spoken for. The hopelessness in these voices owes not simply to (in Li Yiyun’s eloquence) their being imprisoned in their homes or isolation cells but also, more deeply, in their imprisonment in the melodrama of another’s imagination. Of all practitioners of torture, the Grand Narrative is the cruelest of them all.

But here lies the caveat, the ambivalence I feel toward videos such as this. Many of us are used to representations of suffering through the presentation of acousmatic sounds: think Wang Xiuchu’s Yangzhou Massacre memoir or the banality of circuitously implied deaths in cinema. The widespread recognizability of this grammar of murder (Karla Oeler) means less is said about the cruelty of acousmatic medium itself (to risk sounding too cynical in a moment of rare enthusiasm) as one of simultaneous confinement and appropriation, replacing the face of the sufferer with the spectacle of the brazen bull. Is this video not, in some strange sense, a distillation of the quintessential social media encounter with the Shanghai lockdown: fragmented, decontextualized, and sensationalized in a particular mode? To what extent is the catharsis (“many on WeChat are brought to tears by this video,” this YouTube uploader writes) brought about by watching it not an effect of digital spectatorship itself? And how does the mode of engagement that it demands of the spectator, which is sharing the video for the sole sake of its circulation, not re-inscribing the only form of activism that social media platforms have conditioned their users to take?

I, for one, am unable to answer these questions. There is perhaps even less I can do than the trucker driver, the local residence committee director, or the doctor on the phone, as much as I know (as I’m sure the others do, too) that something is to be done. One thing’s for sure: we’re all getting tired of waiting for the cabbage to rot.

How to Spot a Connoisseur

I only learned this morning that Kanda Nobuo (1921–2003) was the hero behind the card catalogue (now appears to have been lost) for the Laufer Manchu Collection in the then-Far Eastern Library at the University of Chicago. Everything was done in the span of three days: he was in Chicago from May 16th to 20th, 1963, and was unable work on the collection during the weekend. Nit-picky as I am of all the details left out of this catalogue, that’s an impressive feat that I, for one, probably would not have been able to pull off.

Prof. Kanda wrote a short report to 満族史研究通信 (Communications in Manchu Research) in 1999 recounting this trip to the University of Chicago. A few lines caught my eye specifically:


As Laufer himself writes with pride, in the Laufer collection one finds plenty of beautiful palace impressions from the Daily Lectures on the Four Classics, Daily Lectures on the Classic of Changes, Daily Lectures on the Classic of Documents to Manchu Rituals for Sacrificing for Deities and the Heaven and Explications of the Meaning of Greater Learning. However, in my opinion, I thought such works as the commercially printed Newly Printed Classic of Documents in Manchu and Nikan Characters were the real rare gems.

Have to say, I’m Team Kanda on this one. Here are a few pictures of this quirky Qianlong 3 (1738) commercial translation of the Classic of Documents (M335/5622), with quite a few comparatively unusual features if you know what to look for.

East Asian Collection, University of Chicago
East Asian Collection, University of Chicago
East Asian Collection, University of Chicago

If you would like to see what Prof. Kanda has to say about this document, he wrote an article on it titled 「満文本の四書と書経」, in 『第二届中国域外漢籍国際学術会議論文集』(台北: 国学文献館, 1989).

Create your website with
Get started