Tilting vessel

« previous post | next post »

Earlier this year, we had a post about a fascinating new Wikipedia article on "Goblet word" (5/30/20).  That post was about a vessel that served as an analogy for a rhetorical device called zhīyán 卮言 ("goblet word").  Now we have another magisterial Wikipedia article by an anonymous master of Chinese esoterica.  It's about another name for a similar type of vessel called qīqì 欹器, "tilting vessel".

The qīqì (欹器, "tilting vessel" or "tipping vessel") was an ancient Chinese ceremonial utensil that automatically overturned and spilled its contents once it reached capacity, thus symbolizing moderation and caution. Both Confucian and Daoist Chinese classics include a famous anecdote about the first time Confucius saw a tilting vessel. In the Confucian tradition (e.g., Xunzi) it was also named yòuzuò zhī qì (宥座之器, "vessel on the right of one's seat"), with three positions, the vessel tilts to one side when empty, stands upright when filled halfway, and overturns when filled to the brim—illustrating the philosophical value of the golden mean. In the Daoist tradition, the tilting vessel was named yòuzhī (宥卮, "urging goblet" or "warning goblet"), with two positions, staying upright when empty and overturning when full—illustrating the metaphysical value of emptiness, and later associated with the Zhuangzian zhīyán (卮言, "goblet words") rhetorical device.

As with many arcane implements from antiquity, the terminology for the tilting / tipping vessel is challenging:

"Phonetic loan characters", using an existing character to be read as a substitution for another of similar or identical pronunciation, have complicated interpreting several "tilting vessel" names. Prior to the Qin Dynasty (221–206 BCE) when the process of standardizing Chinese characters began, scribes would use a current Chinese word and character to write another (near-)homophonous word without a standard written form, which often subsequently resulted in the creation of a new character to write the original word and meaning. For example, yāo from Old Chinese *ʔew (要, "waist") was borrowed to write yào from *ʔew-s "important; want" and the original "waist" sense was later written yāo (腰) semantically clarified with the addition of the "flesh" radical.

The word (, "implement, utensil, instrument; vessel, utensil used as a container or receptacle", Kroll 2017: 357) is a recurrent linguistic element in Chinese names and descriptions for tilting vessels. What became the most common tipping-vessel name qīqì (欹器) is a linguistic compound with (, "exclamation of admiration"), which was a phonetic loan character for (, "slanting, canted, leaning to the side; inclined; oblique; askew; unstable, unsteady", Kroll 2017: 352). Both these logographs are classified as radical-phonetic characters, combining the same (, "uncommon") phonetic component with two different semantic components or radicals, called the "lack" radical () and "branch" radical ().

I recall from my research on and translation of the Sun Zi 孫子 (Master Sun; Art of War) that ("unconventional; unusual; unorthodox ") played an important role in his strategy.

The article continues with further discussion of other related terms, and then moves on to citations of six early texts referring to the tilting vessel.  That is followed by a survey of the tilting vessel in history, including these two intriguing paragraphs concerning tilting vessels in the Islamic world and in Korea:

During the Abbasid Caliphate, tilting vessels "aroused the keen interest" of Persian scholars, who "greatly developed their possibilities", as seen in Banu Musa's 850 Book of Ingenious Devices, which described many types of automata, such as valves that open and close by themselves (Needham and Wang 1962: 35).

The Chinese qiqi (欹器, "tipping vessel") was adapted for a technologically-sophisticated Korean water clock during the Joseon dynasty (1392–1897). King Sejong the Great (r. 1418–1450) ordered the inventor Jang Yeong-sil to develop two automated water clocks. The 1434 Borugak-nu (報漏閣漏) was the national standard clock in the Gyeongbokgung Palace. The 1438 Heumgyeonggak-nu (欽敬閣漏) was an astronomical clock that displayed the movements of celestial bodies and changes of seasons with both visual and audible time signals. It had an overflow tube that transferred surplus water to a miniature waterfall, pond, and tilting vessel (Kim et al. 2017: 174).

The article concludes with a long section on modern scholarship concerning the tilting vessel that draws on historical sources and archeological evidence.  The debates over the visual symbolism and philosophical significance of the tilting vessel make for especially interesting reading.

As usual with Wikipedia, the article closes with a list of more than a dozen references.

Is your glass half full or half empty?  Be satisfied if it is at the midpoint.  Don't be so greedy that it fills up and runs over, for then you're apt to lose everything.


  1. John from Cincinnati said,

    July 28, 2020 @ 3:43 pm

    anonymous master of Chinese esoterica

    Pseudonymous perhaps. The Wikipedia "view history" tab for the cited article credits the contribution to user id Keahapana, a very active contributor according to "user contributions". Not that that's a recommendation one way or the other. If one has a Wikipedia account and is logged in, then upon visiting a user talk page one can click on "Email this user" and perhaps learn if the user is willing to become known, at least privately, or at least to correspond. As for "Keahapana", it might be a valley on the Hawaiian island Kauai, although Google maps didn't help me to find it.

  2. Peter Maydell said,

    July 28, 2020 @ 3:45 pm

    If, like me, you finished the wikipedia article still not much the wiser about what the vessel might have actually looked like or how it would have behaved as described, the paper by Daniel Fried from its citations list has an illustration of one in its three positions on page 19…

  3. Victor Mair said,

    July 28, 2020 @ 11:24 pm

    Searched for a webpage with Fried’s illustration but this from JSTOR was all I could find.

    But that is essentially like the amphora pictured in the Wikipedia article.

  4. Jens Østergaard Petersen said,

    July 29, 2020 @ 3:53 am

    In late imperial times, the vessel was thought to be suspended. It designed to be unstable when empty, thus apt to lean to one side. When half-full it became stable and upright and when more than half-full, it again became unstable and leaned to one side, pouring out its contents.

    An illustration from 1589 in a work illustrating the Kongzi Jiayu, 孔聖家語圖, show this, in all three positions:


    Another Ming illustration is here,


    The illustration in Fried 2007 derives from a 1958 article interpreting the neolithic amphora as such “tilting vessels.” It is not available to me, but must be inspired by illustrations like these. (I haven’t had time to read in the relevant literature, but Fried does not render any of these later images.)

    More illustrations can be found here,


  5. Victor Mair said,

    July 29, 2020 @ 4:27 am

    "Anonymous" definition: without any name acknowledged, as that of author, contributor, or the like: an anonymous letter to the editor; an anonymous donation.

    Wikipedia articles do not have their authors' names attached. They are anonymous.

  6. Philip Taylor said,

    July 29, 2020 @ 4:50 am

    "Wikipedia articles do not have their authors' names attached. They are anonymous" — is that not a slight mis-representation of the facts ? The whole point of a wiki, as I understand it, is that no one individual is (or can be held) responsible for the authorship of any one article, since all articles can be amended/edited/etc., by anyone, including (for example) those who know absolutely nothing about the subject, and those who are universally acknowledged as the world's leading experts. Thus even if the article in question was originally written by A. Nonymous, any number of persons could subsequently have amended it, and thus the concept of authorship qua authorship simply does not arise. It is a collective work, where the number of contributors may vary literally from second to second.

  7. Victor Mair said,

    July 29, 2020 @ 4:58 am

    "…do not have their authors' names attached."

  8. Andreas Johansson said,

    July 29, 2020 @ 7:22 am

    If you care to dig into the history of a WP article, you can see exactly which username (or IP address) contributed what, so I dunno if I'd consider it anonymous exactly. The author's names may not be "acknowledged" in a manner similar to a book or paper, but they're there to be found.

  9. Victor Mair said,

    July 29, 2020 @ 7:27 am

    "The author's names may not be "acknowledged" in a manner similar to a book or paper…".


  10. Sniffnoy said,

    July 29, 2020 @ 1:21 pm

    Hm, reminds me of the Pythagorean cup.

    (…so I went and added a "see also" to the Wikipedia articles on both of these. :) )

  11. Victor Mair said,

    July 29, 2020 @ 7:13 pm

    Thanks to Jens Østergaard Petersen, two images have been uploaded to Wikimedia Commons and added into the qiqi ("tilting vessel") article.

    This is how Wikipedia goes and grows.

  12. Terry Hunt said,

    July 30, 2020 @ 7:45 pm

    Magisterial the article may certainly be, but it's also at risk of being proposed for deletion, because it currently contains no citations whatsoever.

    Yes, it has a long list of reference works, and doubtless the main contributor (beyond whom only one other has contributed a single 13 byte addition) drew most if not all of the Article's facts from them, but this is not how a Wikipedia article is meant to be verified (at least in the English Language Wikipedia; others may have different rules).

    Every fact (beyond those of "the sky is blue" level) is supposed to have an in-line citation to a reliable source (which would in this case likely be one of the listed sources, plus a page number) so that the relevant text phrase or sentence is followed by a (or several) superscript number(s) in square brackets that hyperlink(s) to a bibliographical entry in a Citation list in the References section at the bottom of the article. Thus any reader can quickly check exactly whence any given assertion comes, and could (in theory) consult the work in question to confirm its accuracy for themself.

    To be sure, I doubt that a deletion proposal would actually be intended to result in such a well-writted article being deleted – it would be made in order to start an effort by editors to create the required citations. This however would (will?) involve an enormous amount of work for anyone who isn't the original main contributor, who presumably already knows exactly what sources were used where, and would have saved a lot of effort by inserting the citations (there are templates to help with this) while drafting the article.

RSS feed for comments on this post

Leave a Comment