German linguist Möllendorff and the earliest recordings of Chinese

« previous post | next post »

"UCSB Library Acquires Rare Chinese Language Audio Cylinder Recordings", UCSB Library Newsletter (September, 2024)

The UC Santa Barbara Library is excited to announce the recent acquisition of the Paul Georg von Möllendorff Chinese Cylinders, a collection of wax cylinders widely considered to be the first audio recordings from China. The cylinders, recorded in the late 1800s by linguist Möllendorff, contain sixteen recitations of a popular, celebrated poem "Returning Home"' by Tao Yuanming. Möllendorff recorded the poem in various Chinese dialects to document the differences in regional languages at the time. Today, the cylinders provide a rare glimpse into the history of Chinese language and include dialects that are considered critically endangered or extinct.

Like many late 19th-century diplomats and Sinologists, Möllendorff was an extraordinary scholar.  I will spend the remainder of this post describing his aims, abilities, and achievements.

Möllendorff (1847–1901) was a German linguist and renowned diplomat in East Asia. While serving as Commissioner of Customs at Ningbo, China, in the 1890s, Möllendorff undertook an ambitious project to document and classify Chinese languages for the 1900 Universal Exposition in Paris. He asked speakers representing various Chinese languages to recite the same poem into a graphophone and then transcribed the results phonetically. Möllendorff sent the cylinders to Léon Azoulay, a prominent figure in the Société d’Anthropologie de Paris, to feature at the exposition as part of a "phonographic museum" where visitors could take an audio tour of cultures around the world. After the exposition, Azoulay published a full catalog of the contents of the phonographic museum, including the donated Chinese cylinders, which were explicitly identified with Möllendorff.

Möllendorff in official Korean dress

From Wikipedia:

Möllendorff is mostly known for his service as an adviser to the Korean king Gojong in the late nineteenth century and for his contributions to Sinology. In English-language publications, Möllendorff is often credited with having designed a system for romanizing the Manchu language, which was in fact the creation of his compatriot Hans Conon von der Gabelentz.

Since Wikipedia says that a citation is needed for that last sentence, and I've been hearing about this uncertainty for the last half-century, I'm hoping that some of the Manchu specialists who read Language Log will finally clarify the situation regarding the romanization of Manchu.

Half a century ago, I knew Möllendorff primarily as the author of A Manchu Grammar (Shanghai:  Printed at the American Presbyterian mission press, 1892).  In the course of preparation of this post, however, I've come to learn a great deal more about Möllendorff's accomplishments during the course of his life, such as learning Hebrew while still a young man, joining the Imperial Maritime Customs Service in Shanghai, becoming proficient in "Chinese", and joining the German consular service as interpreter and later vice-consul in Tianjin.

Something of great significance for me occurred during Möllendorff's service in the German consulate.  Namely, he became a friend of the Qing scholar, Ma Jianzhong (1845-1900), who was an official in the secretariat of the powerful statesman, governor-general Li Hongzhang (1823-1901).  What brought me to a standstill upon seeing the name of Ma Jianzhong was that he was the first Chinese to write a grammar of "Chinese", published in 1898.

Victor H. Mair (1997), "Ma Jianzhong and the Invention of Chinese Grammar", in Chaofen Sun, ed., Studies on the History of Chinese Syntax. Monograph Series Number 10 of Journal of Chinese Linguistics, 5-26.

I had always wondered why any Chinese — before the introduction of modern linguistics — would undertake the improbable task of writing a grammar of Classical Chinese / Literary Sinitic, even though Westerners had previously written grammars of vernacular Sinitic languages.  Naturally, I had several times attempted to read Ma Jianzhong's Mǎ shì wéntōng 馬氏文通 (Ma's Literary Expositor [ironically rendered as "Basic principles for writing clearly and coherently by Mister Ma" here), but I always found it unintelligible, and I have never met anyone, Chinese or foreigner, who could render it into comprehensible English, French, German, Russian, etc.

When explained rigorously and analytically, as Harold Shadick did for his A first course in literary Chinese, 3 vols. (Cornell University Press, 1968), which I still use for my introduction to Classical Chinese / Literary Sinitic course, it is possible to write a useful "grammar" of the language.

Cf. Aṣṭādhyāyī by Pāṇini पाणिनि (between 7th c. and 4th c. BC), the world's first grammar.  At the University of Washington, all graduate students in linguistics used to be required to take a special course on Paninian grammar.  I don't know if that's still the case.

Knowing that Ma Jianzhong was a close associate of Möllendorff makes me wonder whether it was the latter who put a grammar bee in the bonnet of the former.

We started this post with the earliest recordings of Sinitic dialects / topolects and ended with the first grammar of Literary Sinitic by a Chinese scholar, and Paul Georg von Möllendorff was right there at the heart of both these pathbreaking enterprises, and much else of great consequence besides — such as helping the formidable Li Hongzhang obtain munitions and warships from the German industrial firms Vulkan and Krupp, and then going head to head with Li over his attempts to gain greater independence for Korea from China — a formidable intellect and redoubtable personality that belies the wonky-nerdy appearance we see in some of his photographs.

 

Selected readings

[Thanks to Martin Schwartz]



7 Comments »

  1. Ben said,

    September 5, 2024 @ 10:13 pm

    That's a great Halloween costume idea.

  2. Ben said,

    September 5, 2024 @ 10:14 pm

    Hopefully someone can denoise the recordings. I can't identify anything from the audio linked in the article

  3. John Swindle said,

    September 6, 2024 @ 1:32 am

    I listened to the audio sample they provided and wasn't at all surprised that I couldn't follow it either, but of course it's wonderful that these exist and are being made public. Which Chinese language or topolect does the sample represent?

    Here for reference is a modern recording of the poem in Cantonese pronunciation. The poem starts at 1:09, and the link should open at that point:
    https://www.youtube.com/watch?v=EDVzlYN7IIQ&t=69s

  4. Coby said,

    September 6, 2024 @ 11:03 am

    I had always wondered why any Chinese — before the introduction of modern linguistics — would undertake the improbable task of writing a grammar of Classical Chinese / Literary Sinitic, even though Westerners had previously written grammars of vernacular Sinitic languages.
    Grammar (γραμματική, grammatica) — "before the introduction of modern linguistics" — meant just that: the teaching of the principles of a "classical" language, that is, one that was written but no longer spoken: Sanskrit in India, Attic or Homeric Greek in the Hellenistic era, Latin in medieval Europe etc. Nebrija's Gramática de la lengua castellana was the first grammar of a living language.

  5. Gokul Madhavan said,

    September 7, 2024 @ 10:28 am

    I don’t know enough about the grammatical traditions of Greek and Latin to comment on them, but at least in the case of Sanskrit it is not clear that the assertion that the grammar was of a written and no-longer spoken language holds.

    For one, we still don’t know whether Pāṇini himself used writing when composing his Aṣṭādhyāyī: while he seems to be familiar with the idea of writing, a lot of Pāṇinian terminology is based on phonological distinctions and not orthographic distinctions (the way we use italics or Greek letters today).

    Furthermore, it is also not the case that Pāṇini exclusively focuses on the sort of language that would be regarded as “classical” in his own time. (I use the quotation marks around the word “classical” because of course it is the post-Pāṇinian formalized language that we today call Classical Sanskrit.) Many of Pāṇini’s rules have linguistic register qualifiers, such as the term chandasi (“in metrical speech”, i.e., pertaining to the Vedic poetic corpus), and the term bhāṣāyām (lit., “in speech”, likely pertaining to everyday speech in contrast to the metrically-patterned poetry of the Veda).

    Of course, one could argue that this speech was also a formal, élite register distinct from colloquial speech, but that hardly makes it a “non-living language” at the time that Pāṇini composed his masterpiece.

  6. Gokul Madhavan said,

    September 7, 2024 @ 10:54 am

    One final observation (dare I call it a “Pāṇinian shot”?!): It’s easy to forget that Pāṇini was not a one-off wonder but rather a member of a long tradition of Sanskrit grammarians. He himself refers to a number of his predecessors, although their work has not survived. The pre-Pāṇinian texts thaf have survived are generally phonological in nature and usually focused on specific recensions of different Vedas. (Cory’s point is thus more applicable to them, with the obvious nuance that these texts are all oral.) Pāṇini is in fact celebrated in the tradition precisely for writing a work that spanned across recensions and registers.

    And while it is also true that Sanskrit comes to be defined by Pāṇini’s text in the centuries following his lifetime (so that Cory’s point is also more applicable to the later tradition as well), we also often forget the critical role (in every sense of the word) played by Pāṇini’s successor Kātyāyana. His vārttikas, or notes, “correct” hundreds of cases of both over- and under-extension in Pāṇini’s own system, to the point where Classical Sanskrit grammar as we have it today is genuinely reliant on both Pāṇini and Kātyāyana.

    But Kātyāyana’s contribution is also germane to our discussion because I have seen the claim that many of his “corrections” are not so much fixes to Pāṇini’s grammar, but rather accommodations to describe the language of his own time and place, a few hundred years and a few hundred kilometers removed from Pāṇini. (I wish I could cite this claim accurately, but I alas don’t remember. It may have been something by Paul Kiparsky or by George Cardona.) If this claim is valid, then it is further evidence that Pāṇini and Kātyāyana were interested in the living language of their times, even if limited to an élite register.

  7. Gokul Madhavan said,

    September 7, 2024 @ 10:57 am

    My sincere apologies to Coby: please mentally correct all my misreferences to “Cory” above to “Coby” instead.

RSS feed for comments on this post

Leave a Comment