Oldest Analects MS Found.

Eiichi Miyashiro writes for the Asahi Shimbun:

A manuscript of commentaries about Confucianism written apparently between the sixth and early seventh centuries in China was confirmed in Japan, a discovery one scholar described as “invaluable.” It is believed to be one of the oldest of any religious teaching written on paper, except for those of Buddhist scriptures, found in the country.

Researchers at Keio University and other institutions say the writing is also most likely the oldest among manuscripts of commentaries on the Confucian Analects that have been handed down at temples, shrines or homes. […] The manuscript is a compilation of commentaries, known as Lunyu Yishu (the elucidation of the meaning of the Confucian Analects), put together by Huang Kan, a Confucian scholar of Liang (502-557), during the Northern and Southern dynasties period. All manuscripts of Lunyu Yishu had been lost in China by around the 12th century, according to experts. […]

Researchers consider the books on the commentaries as part of the Confucian Analects in a broad sense. The manuscript discovered in Japan is expected to provide scholars with significant clues into the history of exchanges between Japan and China in philosophy and other realms. […] Based on the shape of the characters, the team concluded that the manuscript was most likely written between the Northern and Southern dynasties period and the Sui Dynasty (581-618). They also believe it was brought to Japan through Japanese missions sent to the Sui Dynasty and Tang Dynasty (618-907) in China. […]

Until the recent discovery, the world’s oldest manuscript of the Confucian Analects originated from the Song Dynasty between the end of the 12th century and early 13th century. The oldest one in Japan dated back to the latter part of the Kamakura Period (1185-1333).

It’s nice to know such things are still coming to light.


  1. Your headline is a little misleading. Just to be clear: this is purportedly the oldest Analects MS that has ever been found _in Japan_. Manuscripts dating from about 50BC have been found in China and North Korea.

  2. Woops, sorry — thanks for the clarification!

  3. …I guess it wasn’t really wasn’t worth posting then, was it? Oh well.

  4. David Marjanović says

    There was a LLog post about it recently, likewise with commenters pointing out that it’s only the oldest one in Japan and that the original Japanese version of the article says so – all sorts of things were lost in translation.

  5. The article says that it was the oldest copy of the 論語義疏 (Lunyu Yishu), a set of commentaries on the Analects), that had been “passed down” (伝世品), as opposed to being “unearthed” (出土品).

    The article can be found in Japanese here (although headline differs).


    A rough translation (based on Google Translate, I don’t have time to translate it properly so it’s a bit messy) is:

    Newly confirmed “Lunyu Yishu”. Kanji such as “Confucius” can be seen in the text. (provided by Keio University)

    A manuscript of a commentary on “The Analects”, the “Lunyu Yishu”, which summarizes the dialogue between the ancient Chinese thinker Confucius (c. 551 to 479 BC) and his disciples, has been confirmed in Japan. The manuscript appears to have been written in China in the early 6th and 7th centuries. According to a research team led by Keio University, which was investigating the manuscript, there is a high possibility that it is the oldest passed-down copy of the Analects to be brought to Japan and carefully stored at temples and shrines in Japan.

    The “Lunyu Yishu” were lost in China around the 12th century. Apart from Buddhist scriptures, this is considered to be among the oldest surviving paper manuscripts . In China, the Analects are mainly transmitted through annotations (books of commentary), and experts note that they are valuable historical materials for research on the history of thought and exchange between Japan and China.

    What was confirmed was a manuscript of a part corresponding to “volume 5” (in the current document identified as “volume 6”) written by the scholar Huang Kan (Wang Kan) of Liang (502-557) in the Northern and Southern dynasties period when China was divided into north and south. It is formed of twenty sheets of paper glued together to form a scroll with a length of 27.3 cm. It was purchased by Keio University from an antiquarian bookstore in 2017. A research team consisting of experts in bibliography, Chinese literature, Japanese literature, Japanese history, etc. was formed on campus to conduct research from 2018. From the shape of letters it is highly possible that it was written from the end of the Northern and Southern Dynasties to the Sui dynasty. The research team says, “Except for excavated items, it is considered to be the oldest manuscript of the Analects.”

    Note: Even the Japanese is a bit fuzzy about it being the oldest “what of where” because it takes a highly Japan-centric approach, but it appears that this really is the oldest copy of this manuscript anywhere because the last one was lost in China centuries ago. As for how important this particular commentary is among various commentaries on the Analects, I really have no idea.

    (Google Translate really is crap. Twice I got “that the sex is high” “It is highly sexual” for “there is a high likelihood”. Have they been training it on porn?)

  6. @Bathrobe: Your last sentence reminded me of something. I cannot find the clip online, but there is a great scene in the BBC comedy Coupling in which the main male character Steve, and his peculiar Welsh best friend Jeff, argue over whether all-female pornography should be considered a valid form of evidence about what women do when men are not around.

  7. Sinologists I’ve talked to suggest that this is indeed important and significant — this is apparently a major commentary, but was known previously only either from scattered citations, or from relatively late (18th c.+) Japanese recensions that were already (on the basis of comparisons with said citations) strongly suspected of having been significantly reedited. It’s apparently been a major part of Confucian understanding in Japan during the past few centuries. This manuscript, though just one volume of the work, dates from not too long after the composition of the commentary in the first place (a sixth-c. commentary, a sixth- or seventh-c. MS).

    A very rough European comparison might be a third-c. manuscript of Germania turning up somewhere.

  8. Jen in Edinburgh says

    I quite enjoyed the second sentence, although it made my brain hurt slightly too early in the morning.

    This is, I think, the oldest religious teaching…
    – on paper
    – not Buddhist
    – in Japan
    – apart from a small number of unnamed others!

  9. David Marjanović says

    “Unearthed” is meant literally as “bamboo strips found in an archaeological excavation”.

    Have they been training it on porn?

    One of Japan’s major exports…

    when men are not around

    Keyword “camera man”. 🙂

    A very rough European comparison might be a third-c. manuscript of Germania turning up somewhere.

    That would be awesome.

  10. I’m reminded of Leo Wiener’s contention, 100 years ago now, that the Germania was a fifteenth-century (?) forgery and the early Germans, IIRC, pretty much lived like apes till Semites brought them culture. Trying to track that down, I find (a) that he was prone to denounce anything that contradicted his theories as a forgery, and (b) that he was the father of Norbert Wiener. Well.

  11. John Cowan says

    main male character Steve, and his peculiar Welsh best friend Jeff

    “In the first [excerpt,] they are a little crazy, and in the second they are not only crazy but Welsh.” —Le Guin

    early Germans, IIRC, pretty much lived like apes

    That later German H.L. Mencken, who was singularly fond of the 18C (it appealed to his anti-democratism), said that before it “even kings lived like hogs”.

    father of Norbert Wiener

    Fortunately, we bear no responsibility for the sins of our parents, for entropy is directional.

  12. Trond Engen says

    whether all-female pornography should be considered a valid form of evidence about what women do when men are not around.

    Jeff: “How can we discard the only evidence we’ve got?”

    I often think of that scene when I think of scant evidence.

  13. @Trond Engen: That is one of two lines from that scene that really stuck in my memory. The other was (also Jeff’s): “Maybe they’re not cross all the time,” which got a really big laugh from my wife

  14. Trond Engen says

    Me too. Except I think maybe it’s “Why discard the only evidence we’ve got?” I stopped, thought, went back and changed it, and now I regret.

  15. There are a couple of papers on the Internet that help understand the background of the new Japanese discovery. There is a whole world of Confucian scholarship involved in this that could barely be conveyed in a newspaper article, especially if the writer didn’t know much about it.


    “Most of the annotations to the Analects of Confucius 論語 made by scholars from the Han Dynasty and the Wei and Jin periods have lost their original form, with the exception of a few chapters by Zhen Xuan 郑玄(127-200) excavated from Dunhuang and Turfan. Some annotations made by Han Dynasty scholars were included in the Lunyu Jijie 論語集解, compiled by He Yan 何晏 (?-249) in the Wei period, and thus are still extant. After He Yan, compilations generally included annotations made by scholars of the Jin Period and compiled by Jiang Xi江熙 (?), which now have disappeared.

    “The Assistant of the Imperial Academy of Liang, Huang Kan 皇侃 (488-545), made annotations to the text of the Analects of Confucius and the work of He Yan, as well as passages from Jiang Xi and other scholars in his work Lunyu Yishu 論語義疏. The Lunyu Yishu had disappeared in China after the Southern Song Dynasty until a Japanese edition emendated by Sonshi Nemoto 根本遜志 (1699-1764) in 1750 was taken from Japan around 1770.

    “The Lunyu Yishu is valued today not only because it preserves many old annotations which have been lost in full form, but also because it is the only extant “Yishu” 義疏 of the Confucian classics, which is a style of annotation deeply influenced by Buddhism and prevalent in the Six Dynasties.”

    So yes, it is a big deal. It was lost in China and reintroduced from Japan in about 1770. The new discovery is much older than the Nemoto edition and presumably closer to the original source, Huang Kan, who actually taught the Analects to students.

    Much of what this means is made clear in the second, very long article, which looks at a text found at Dunhuang, which gives a picture of the teaching of the Confucian classics to students of a much later time period than that of Confucius who needed to be taught how to understand them — hence the importance of commentaries.


    The coda:

    “The omissions, interpolations, and amendments in the sub‐commentary from Dunhuang do not result from alterations made in Japan but are primarily related to amendments made by the tutor delivering lectures on the Lunyu based on Huang Kan’s readings. It thus follows that manuscript fragment P 3573 [Dunhuang] is the earliest extant textual witness of the Lunyu Yishu and invaluable source material for our understanding of its transmission. It represents what might be described as a working manuscript either taken from or used for lectures based on the Lunyu Yishu at the periphery of the empire. It offers valuable insights into teaching practices and manuscript culture, but it does not qualify as a faithful and trustworthy representation of the original textual condition of the Lunyu Yishu. The manuscript from Dunhuang transmits selected passages from the Lunyu and from Huang Kan’s sub‐commentary that were deemed relevant by either the tutor or the copyist, and it was received and studied with the commentaries assembled in the Lunyu Jijie in mind, and with reference to a full copy of the Lunyu Yishu.

    “The Lunyu Yishu allows for meaningful insights into one man’s teaching at the National University of the Liang, where his lectures attracted considerable numbers of students. It represents what may be described as edited lecture notes that take the Lunyu Jijie as their exegetical starting point and develop an intellectual environment in which the openness of the main text is emphasized by Huang Kan who discusses the exegetical corpus and is open to divergent, sometimes even contradictory, interpretations of the Master’s utterances. The transmitted text shows considerable residual oral influences and remnants of oral teaching practices, and it offers telling insights into traditional pedagogy.

    “The manuscript fragment P 3573 builds on Huang Kan’s elaborations and attests to a second layer of oral teaching performance. The additional elucidations by the tutor in Dunhuang provide us with windows into what was expected of students of the Lunyu in a specific environment toward the end of the Tang Period, and into how students were led to advance their skills and knowledge. The tutor not only provides pointers and focused summaries but also additional references wherever he finds these necessary in view of the students’ perceived lack of familiarity with relevant readings. On the other hand, the manuscript clearly indicates that familiarity with the Lunyu Jijie was a prerequisite and that the main text of the Lunyu was approached with reference to the glosses established by preceding exegetes. By focusing on what the tutor considered the core explanatory material, the manuscript P 3573 reflects historical changes of exegetical parameters, a development that found its culmination in Xing Bing’s Lunyu Zhushu. In this sense P 3573 can be described as a precursor of the Lunyu Zhushu.

    “Although the appearance of text‐cum‐commentary editions is widely associated with developments during the Song, the main text of the Lunyu, the pointers toward the Lunyu Jijie, the elaborations, discussions of and addenda to it in the Lunyu Yishu, as well as the tutor’s treatment of these materials are all written up in a comprehensive manuscript format that reflects substantial components of the oral transmission of knowledge in a particular environment. As an outline of teaching sessions, most likely recorded by a student, manuscript fragment P 3573 was never intended for later generations. It does, nevertheless, provide pictures frozen in time that allow us to gauge the interaction between a tutor and his students at this outpost of the Tang empire, and how readings of the Lunyu were passed on from generation to generation.”

    As a final comment, the newspaper article’s claim that the manuscript “is expected to provide scholars with significant clues into the history of exchanges between Japan and China in philosophy and other realms” is probably true, but its significance for Confucian studies in general is what makes this an important find.

  16. Many thanks — now I’m glad I posted it!

  17. I see that my Google Translate-based translation was a bit sloppy. It should have read:

    A manuscript of the “Lunyu Yishu”, a commentary on “The Analects”, which summarize the dialogue between the ancient Chinese thinker Confucius (c. 551 to 479 BC) and his disciples, has been confirmed in Japan.

    (Actually confusing, but the Analects summarise the dialogue between Confucius and his students; the “Lunyu Yishu” is a commentary on the Analects.)

Speak Your Mind