Tuoba and Xianbei, part 2

« previous post | next post »

Inquiry from Kiraz Perincek Karavit, a Ph.D. candidate in the History Department, Bogazici University, Istanbul, who is interested in the intercultural relations along the Silk Roads, and the mobility of languages, images, tales, and legends. 
I am writing you to ask for your advice on a subject, after reading your blog in Language Log about Tuoba and Xianbei:
"Tuoba and Xianbei: Turkic and Mongolic elements of the medieval and contemporary Sinitic states" (5/16/22)
Vasily Barthold, in his book Orta Asya Türk Tarihi Dersleri, 1927 (a compilation of 12 lessons about Central Asian Turkic history, that he taught at Istanbul University in 1926), that a dictionary of the Xianbei language was found among the Chinese sources. Barthold's reference is Paul Pelliot, who announced the existence of such a dictionary at a conference in Leningrad. He states that Pelliot examined this dictionary and argued it was in the Turkic language. According to Barthold, that is why there is no doubt that the "Sien-pis" are Turkic.

In later years, the book was translated and published in other languages as well, in German, in French, and in Chinese:
Barthold, Zwölf Vorlesungen über die Geschichte der Türken Mittelasiens, 1935, p.25
Barthold, Histoire des Turcs d'Asie Centrale, 1945, p.19
中亚突厥史十二讲【无封面、版权页】 ([苏]巴托尔德; 罗致平译, 中国社会科学出版社1984年, p.20
I think, although very interesting, it is not that important whether the Xianbei language was of Turkic origin or not. Rather, the existence of such a Xianbei language-Chinese dictionary is amazing.
I checked the above versions of Barthold's book for any reference that could have been added in later years by the translators, but no, Pelliot did not publish anything about this dictionary. On the other hand, it does not seem a misunderstanding or misinterpretation for both sides, i.e., for Barthold and Pelliot, since Pelliot has detailed impressions and deductions on scientific matters — such as whether Turkic or proto-Mongol origin — about the Xianbei language and dictionary. Barthold as well, clearly mentions the conference in Leningrad, and the ideas Pelliot expressed, in other parts of the book. He gives a more specific reference for Pelliot's talks in Leningrad in the first lesson (1925 Fall), most probably in the same series of talks.
It is interesting that, in an article dating to 1921, Pelliot states that he is "inclined to consider the Xianbeis as tribes having Mongol language"  (T'oung Pao, "Note sur les Tou-yu-houen et les Sou-p'i"). So, since Barthold's lessons date to 1926, Pelliot must have discovered this document between 1921 and 1926, or most probably before 1925 Fall.
I will check the Russian archives for the event in Leningrad 1925 Fall (maybe it is recorded in the newspapers or university publications) to get the title or hopefully the content of Pelliot's talk. I read some memoirs about Pelliot; thinking about his personality, it is possible that he kept the manuscript for himself, and for some reason he could not accomplish this task. If so, it must be somewhere in the world and if so, I want to find it.

Furthering the research with Chinese colleagues, a possibility is that this manuscript of Pelliot could be one of those mentioned in Sui Shu 隋书, Chapter 32, 经籍志. In this bibliography of books, there are references to such books as 鮮卑語五卷,鮮卑語十卷. According to Chinese sources, these are lost; but maybe, one of these was among the Dunhuang library and brought to Bibliothèque Nationale by Pelliot where he had the chance to encounter and examine it superficially. Then he set it aside planning to go over it in detail in the future. In the meantime he announced it in the Leningrad talks in 1925, provoking curiosity in the field, but the postponement persisted and he did not have the time or opportunity to incline to this material. That is my scenario.

I read in an article referring to Barthold's same book (朱学渊, 鲜卑民族及其语言-下, 2000, p.77) that Pelliot did not finally publish this dictionary. So we can assume with a high probability that a copy of this manuscript does not exist among Chinese sources any more and Chinese academia does not possess the knowledge of it, which can be considered "the first known dictionary of the world." The only extant copy is probably what Pelliot mentioned in Leningrad, 1925.
I admire the tenacity of Kiraz Perincek Karavit in her attempts to track down the document mentioned by Pelliot and Barthold.  If it is ever found, it would be of extraordinary importance for the history of languages in Central Asia.  One thing worries me, however, and that is that I heard from various sources (esp. Professor Francis Cleaves, who studied with the great French Sinologist) that Paul Pelliot's wife "lost" or otherwise failed to preserve many of his manuscripts and research materials.
Selected readings


  1. Ronan said,

    August 15, 2022 @ 11:51 am

    Hartmut Walravens' bibliography of Pelliot's works might have further indications.

  2. Peter B. Golden said,

    August 15, 2022 @ 12:46 pm

    The most recent study of Xianbei (Serbi) is that of Andrew Shimunek, "Languages of Ancient Southern Mongolia and North China. A Historical-Comparative Study of the Serbi or Xianbei Branch of the Serbi-Mongolic Language Family, with an Analysis of Northeastern Frontier Chinese and Old Tibetan Phonology" (Wiesbaden: Harrassowitz Verlag). I do not recall any mention there of a "Xianbei dictionary." Yes, Barthold in his "Twelve Lectures" given in Istanbul in 1926, mentions a lecture that Pelliot gave in Leningrad in which he stated that there was among the Chinese sources a Xianbei dictionary, which "left no doubt that this language was Turkic." There is, indeed, nothing in Pelliot (known to me) that speaks directly to that. Moreover, subsequent work, in particular now, the work of Shimunek, shows that Xianbei (sjen.pjie < s[a]r.pe- Baxter and Sagart, "Old Chinese": 261-262, 367) = särbi/serbi stemmed from a source that was common to Mongolic and "Para-Mongolic" -Shimunek terms it "Serbi-Mongolic."

  3. Lucas Christopoulos said,

    August 15, 2022 @ 3:15 pm

    The Xianbei are described as distinct from the Turks in the Sui books and often fought against them for the Sui dynasty (as in the "heroic" battle of Zhou Pan 周槃 in 582 against Ashina Shetu). Mongolic of Peter B. Golden should be more correct. Their descendants the Monguors (Tuzu) in Gansu-Qinghai are called today “White Mongols” (Tsagaan Mongol).

  4. András Róna-Tas said,

    August 16, 2022 @ 2:40 am

    The enigma of the so-called Tuoba dictionary was treated by Ligeti in his paper "Le Tabghatch un dialect de la langue Sien-pi" in Ligeti, Louis (ed), Mongolian Studies, Budapest, 1970. on p. 277, and note 31.
    Pelliot spoke about a "vocabulaire sino-sien-pi", this was "peu fidélment" interpreted by Barthold, who was present at the conference given by Pelliot in 1924 in Petrograd. This vocabulary is the list of 18 Tuoba titles, which were treated by different scholars, Ligeti wrote:
    "Vers 1934 Pelliot m ̛affirmait à l, occasion d̕ un entretien à propos de sa conference de 1924, que Barthold avait dȗ se méprendre sur ce qu̕̕̕̕ il.avait dit á propos du prétendu lexique sino–sien-pi et de l̕̕ appertance linguistique de Sien-pi, plus exactement des T̕o-pa" (note21

  5. Victor Mair said,

    August 19, 2022 @ 4:25 pm

    From Kiraz Perinçek Karavit:

    I have been thinking about the form of this manuscript as well. I am not sure if I could recognize it if they put it on my desk now. In terms of content, I have more or less an idea: I imagine it not like a real dictionary, but a kind of practical guide in Chinese to contact with Xianbei people, for trade or diplomatic matters. The model I have in mind is Filippo Argenti's Regola del parlare turcho, An educated Florentine, Filippo Argenti arrived in Istanbul in 1524. He held the post of secretary at the Florentine embassy until 1533. His notes, titled Regola del parlare Turcho, meaning "Turkish speaking rules", a kind of handbook, including an extensive Turkish-Italian vocabulary as well as grammar. The order is like this:

    X word in Turkish (as heard by an Italian): meaning of X in Italian, pronunciation of X according to Italian, (sometimes) examples of phrases, proverbs, or frequent uses of X.

    I think this "Xianbei-Chinese dictionary" as well, written for similar practical needs, must have used more or less the same logic, although written at least a millennium ago.

    On the other hand, I can not really visualize what this would look like formally on paper. Argenti's notes look like a dictionary, from left to right; some words are written and there are equivalent words/phrases, one after the other, with a new paragraph. For example "Vocabulario de nomi, fol.244A":

    But what would the Xianbei language-Chinese vocabulary handbook look like? Is there any example of a manuscript produced with a similar logic?

    Hopefully, a title, as it exists in the bibliography of Sui Shu: 鮮卑語.

    Of course vertical; but without space or any other orthography between words?

    How would the words be organized or grouped? I would say thematically, like today's practical tourist/business guides, in the restaurant, in the market, at home…. I think for our case they would undoubtedly have salutation, affirmation/negation, numbers, measure units, colors, animals, animal parts and products (bone, horn, leather, fur), geographical terms, metals, arms, vessels, tribal/kinship/settlement terms, etc. areas determined by the Chinese-Xianbei trade and diplomatic relations' first-hand needs and requirements. I think very few adjectives and verbs, only the fundamental ones embedded maybe inside the examples.

RSS feed for comments on this post