Poetry as "Word Temple" — NOT

« previous post | next post »

Andrew Shields encountered the idea — on Facebook and vigorously promoted on this blog — that the Chinese character for poetry, shī 诗, consists of two parts meaning "word" and "temple".  Furthermore, it is claimed that this is a particularly apt way to represent the notion of poetry, one that is conspicuously missing in Western culture.

Such a facile interpretation commits several fallacies, the chief of which is to misunderstand the history and nature of the character in question.  After a careful examination of the evidence, it seems far more likely that shī 诗 has to do with the ritual performance of the odes by eunuchs, i.e., by reciters or singers who were castrati, than that it means "word" + "temple".

Let me begin by saying that, although they've been studying the origins, construction, and evolution of the graph shī 诗 ("poetry") for centuries, cautious and critical scholars admit that they are uncertain about many aspects of the Chinese word for poetry and the character that is conventionally used to write it.

In the late 60s, two well-known American Sinologists, Chen Shih-hsiang and Chow Tse-tsung, published detailed studies of the word shī 诗 ("poetry") in which they attempted to delve into the etymological roots of the term.  Chen maintained that poetry is "derived from and remained closely associated with the concept of beating rhythm with the foot on the ground" (1969:  378), and he saw that as directly reflected in what he considered to be the archaic root of the character for poetry.  Chow also focuses on the foot, but goes one step further, as it were, by focusing on the movement and the direction toward which the dancer's toes point.  He concludes that poetry involves swift movement toward a goal.  Neither Chen nor Chow mention anything about a temple, and rightly so.


Chen Shih-hsiang, "The Shih-ching:  Its Generic Significance in Chinese Literary History and Poetics," Bulletin of the Institute of History and Philology, Academia Sinica, 39 (1969), 371-413.

__________, "In Search of the Beginnings of Chinese Literary Criticism," University of California Publications in Semitic and Oriental Philology, 11 (1951), 45-63.

Chow Tse-tsung, "The Early History of the Chinese Word Shih (Poetry)", first appeared as a booklet from the University of Wisconsin Press in 1968, and then was republished in a number of other places, including in that same year in Chow Tse-tsung, ed. Wen-lin:  Studies in the Chinese Humanities (University of Wisconsin Press), pp. 195-210.

While Chen and Chow offered interesting speculations about the archaic form and meaning of the forerunners of the traditional Chinese character for "poetry" (shī 诗), their arguments have not been accepted as definitive.  Consequently, in refuting the notion that shī 诗 ("poetry") = "word" + "temple", we must start at the very beginning in our analysis of the character.

First of all, it must be pointed out that the overwhelming majority of Chinese characters — as much as 85-90% of the total — consist of a semantic component and a phonetic component.  That is certainly the case with shī 诗 ("poetry"), which is made up of yán 言 + sì 寺, where yán 言 ("word; speech") is the semantic component and sì 寺 is the phonetic component.  Initially, we need not worry about the meaning of sì 寺 because its primary function in the character is to convey the sound, albeit somewhat imprecisely.  The question, however, is whether it also plays a secondary semantic role in the character.  If it does, then those who assert that the Chinese character for "poetry", shī 诗, is made up of "word" and "temple" might have a bit of a point, since one of the meanings of sì 寺 is indeed "temple".  Before we accept such a proposition, though, there are a mass of obstacles that must be overcome.  Let us go through some of them.

We have seen above that shī 诗 ("poetry") is made up of yán 言 + sì 寺, where yán 言 ("word; speech") is the semantic component and sì 寺 is the phonetic component.  But what is 寺?  Did it really mean "temple" at the time the character shī 诗 ("poetry") was created?  In antiquity, it would appear that 寺 referred to a ritual specialist or protocol officer, and also to a government office building like a yamen where he performed his duties.

It seems that only after Buddhism arrived in China (circa 1st c. AD) did sì 寺 come to mean "temple", though, from the above exegesis, one can see the semantic association between the older meanings of ritual specialist (as well as the building in which he carried out his tasks) and the post-Buddhist meaning of "temple".

Sì 寺 is not found on the oracle bone inscriptions (OBI; latter part of the 2nd millennium BC), but it is on the bronze inscriptions (BI; dated roughly to the latter part of the 2nd millennium BC and the first two thirds of the 1st millennium BC), where it has a hand (shǒu 手) on the bottom as semantophore and zhǐ 止 ("stop", but see below for how it got that meaning) above it as phonophore, with the overall meaning of "grasping", and it was indeed the original form of chí 持 ("grasp; hold") in the bronze inscriptions, the hand radical to the left being redundantly added later to distinguish this meaning from all the other new meanings that sì 寺 had acquired (see list below).   By extension, then, "giving an offering with the hand at a government building" comes to signify "temple", but that latter development didn't occur until the advent of Buddhism, long after the character shī 詩 ("poetry") was created, and couldn't have been operative at that time.

To sum up roughly for sì 寺, we have:

BI 止 + 手 = 寺 ("grasp") > BI 寺 ("grasp [an offering]") > Han (?) "the official who grasps an offering" > "the place / building where an official grasps an offering" > Buddhist and post-Buddhist "temple".

Thus, even if sì 寺 is considered not solely to serve as the phonophore for shī 詩 but also to function as a secondary semantophore, it did not mean "temple" at the time the character shī 詩 was created.

The component 寺 is highly productive in the composition of compound characters.  Here are some of them:

shī 詩 poetry
shí 時 time
shì 侍 help; wait upon
zhì 痔 piles
zhì or shǐ / shì 峙 a hill; to pile up; name of a place in Shanxi
zhì, chóu, shǐ / shì 畤 sacrificial mound; place name; transplant
zhì 庤 to keep stock; storehouse
shì 恃 rely upon; trust to
shǐ [水+寺] islet
shī 邿 a place in Shandong
shì [寺 inside of 門] a eunuch (see below for this meaning of 寺 by itself [the 門 [eunuchs stay behind doors / gates] was added later)

chí 持 hold; grasp (as discussed above, this was probably the original meaning of 寺)
dài 待 wait for; await; to treat
tè 特 special; on purpose
děng 等 wait; class; sort; rank; equal; etc.

This by no means exhausts the total complement of characters containing the component 寺.  It is easy to see how ludicrous it would be to attempt to force each of these characters to somehow convey the notion of X + "temple".

It is clear that, into the Han, the graph 寺 was used freely to write many words.  According to the editors of Mǎ wángduī hànmù bóshū 馬王堆漢墓帛書 (Silk Manuscripts from the Han Tombs at Mawangdui), a single group of texts (the "Shíliù jīng" 十六經 (Sixteen Classics), a group of 16 very short texts which precede the Laozi B manuscript), the graph 寺 writes the words dài 待, zhì 志, chí 持, and shì 恃.  And this all in a short Han dynasty text of roughly 4,000 characters.

Additionally, but a little more obscure, there is an interesting passage in the looted Warring States Shanghai Museum text "Zǐ gāo" 子羔:

“ X(吾)昏(聞)夫X(舜)丌(其)幼也,每㠯(以)…寺丌(其)言”

That's the transcription from book 2 of Shànghǎi bówùguǎn cáng zhànguó Chǔ zhú shū 上海博物館藏戰國楚竹書 (Warring States Bamboo Strips from Chu in the Shanghai Museum), 2.  Although it is too fragmentary to make much sense of, it does interestingly use the graph 寺 and yán 言 in the same phrase ("寺 his words" or "make his words 寺", perhaps relating to the ritual performance of recitation).  On the other hand, Li Xueqin and others have read 寺 here as shí 時 ("temporalize" [?], whatever that would mean).  Still others have read this passage as differently as "měi yǐ xiào sì qí qīn 每以孝寺其亲” ("invariably serve their parents with filiality").  The most parsimonious explanation of 寺丌(其)言 would avoid more elaborate explanations such as the latter two and interpret the phrase as I have suggested in the middle of this paragraph.

Let us return to the arguments of Chen Shih-hsiang and Chow Tse-tsung that the earliest roots of the graph for "poetry" had to do with the idea of "foot".  While I do not believe that their theories ultimately bear up under scrutiny (as the analysis below will demonstrate), they are near to the heart of the matter, and it is possible to see on what assumptions they were proceeding.  To wit, they focus on zhī 之 ("go") and zhǐ 止 ("stop [going]"), which constitute the phonophore of sì 寺, which in turn serves as the phonophore of shī 詩 ("poetry").

Incidentally, zhī 之 ("go" on the OBI had the same exact form as zhǐ 止 ("stop"), viz., a picture of a foot, the only difference being that the foot of the former had a horizontal line drawn beneath it, signifying the place from which one sets out, hence goes.  In contrast, zhǐ 止 showed just the foot without a line, signifying the place where one halts or comes to rest, hence stops.

Both zhī 之 ("go") and zhǐ 止 ("stop [going]") belong to the Old Chinese (OC) 之 rime group; Schuessler's and others' /ə/. These OC words should be identical save whatever one uses to account for the rising (shǎng 上) tone class of 止, generally ʔ, so tə vs. təʔ or what have you.

There is a lengthy discussion of some aspects of these questions by Anne Yue in an article in T’oung Pao. Here is the reference:

1998             "Zhi 之 in pre-Qin Chinese,” T'oung Pao lxxxiv, 239-292.

Zhī 之 ("go") was subsequently borrowed for use as a subordinative particle, and this is its main use in Classical Chinese.

I think that the old adage shī yán zhì 詩言志 ("poetry is the expression of the will / aspiration") lies nearer to the truth than the spurious "word + temple" derivation, with zhì 志 ("will; aspiration") being both phonophore and secondary semantophore.  This formulation is found in many ancient texts, from the Zuo zhuan (Chronicle / Commentary of Zuo) to the Shang shu (Book of Documents; Classic of History),  Zhuang Zi (Master Zhuang) and Xun Zi (Master Xun) — all dating to around the 4th-3rd c. BC.

Nonetheless, overall the hypothesis that shī 詩 ("poetry") derives from the notion of ritual castrati reciters of poetry is more elegant and compelling than the shī yán zhì 詩言志 ("poetry is the expression of the will / aspiration") adage, attractive though the latter is.  In order for shī yán zhì 詩言志 to be considered as a genuine etymology, we would have to accept shī 詩 as composed of yán 言 and zhì 志, which is not as direct as having shī 詩 be composed of yán 言 ("speech") and sì 寺 ("eunuch reciter"), as will be shown below.

It seems that the character shī 詩 ("poetry") appears rather late; it is found neither among the oracle bone inscriptions (latter part of the 2nd millennium BC) nor in the bronze inscriptions (dated roughly to the latter part of the 2nd millennium BC and all but the last couple of centuries of the 1st millennium BC). That's intriguing, since the Shījīng 詩經 (Poetry Classic / Odes) was supposed to have been compiled by Confucius around the 6th c. BC (on the basis of poems that were supposedly composed in the preceding centuries), though extant recensions date from later centuries.

Shī poetry, if not the character 詩 itself, must have existed before the seal script (arose during the Warring States period [ca. 475-221 BC]), at which time 詩 is unmistakably present, so how would shī have been written before that stage?  One wonders, then, what character was being used for the word shī ("poetry") before 詩 was devised.

While 之+言 is well-attested and other forms seem to have existed, it does look like Warring States through Han manuscripts overwhelmingly used 寺 as the phonophore for shī (note that 寺 was a very common phonophore in general [see the incomplete list of characters containing 寺 above]).

As far as I know, shī 詩 never appears on any pre-Qin (221-206 BC) inscriptions of any kind, whether bronze or shell and bone, which is puzzling.  Aside from a few appearances in the Odes itself and a couple in the Chu ci (Elegies of Chu / Songs of the South; dating to the few centuries just before and after the beginning of our era), the overwhelming majority of uses of the word in extant pre-Han texts appear in discussions of (or quotations of) the Odes (i.e., Poetry Classic [Shījīng 詩經]).

That said, there is some evidence for how the graph was written.  A few extant Warring States manuscripts contain the word.  In the " Yǔ yī" 語一 text from Guodian (mid 4th to early 3rd c. BC), it is written with a form structurally identical to its modern form. Matt Anderson found two instances from the looted Shanghai Museum texts (and there surely must be others, but there is not yet an efficient way to search through these texts).  In the "Cáo mò" 曹沫, it is written with a form equivalent to the modern form, but in the "Shī lùn" 詩論, it is written 之+言 (with the 之 above 言).

Various forms combining 之 and 言 are attested second hand through a number of other sources — the Hàn jiǎn 汗簡 includes this form, and the Gǔwén sì shēngyùn 古文四聲韻 contains this form (attributed to the ancient script Xiào jīng 古孝經 [Classic of Filial Piety]) and a form with 言 on the left and 之 on the right (attributed to the ancient script Shǐjì 古史記 [The Grand Scribe's Records]).  The Shuō wén jiě zì 說文解字 also gives the form � (that's 言+it㞢, if  doesn't display correctly for you), with the element on the right equivalent to 之.

The Han manuscripts from Mawangdui (186-168 BC) also contain a number of instances of the word shī; most are equivalent to the modern form, but it is once written with the graph shè 設 ("set up; arrange; establish"; on strip 212 of the " Wǔ xíng" 五行.

So, while 之+言 is well-attested and other forms seem to have existed, it does look like Warring States through Han manuscripts overwhelmingly used 寺 as the phonophore for shī (note, as we have seen above, that 寺 was a very common phonophore in general).

I haven't found any example of shī 詩 written with 言+志, but of the many examples of zhì 志 collected from the Houma texts by Imre Galambos in Orthography of Early Chinese Writing: Evidence from Newly Excavated Manuscripts (490221 BC), Budapest Monographs in East Asian Studies (Budapest: Eötvös Loránd University, 2006), while most are written with 之 over 心 (the ancestor of the current form), 6 are written with a graph structurally the same as sì 寺, and another four are written with a combination of the two forms (之 over 心+寸).  That is, no extant manuscript instances of shī 詩 contain the element 志, but the element 志 is sometimes written with 寺, the phonophore of shī 詩.

We have seen above that the original meaning of 寺 was probably that of chí 持 ("hold; grasp"), but it is very important to point out that it has also meant "eunuch" in received texts from the Poetry Classic to the Zuo zhuan (Chronicle / Commentary of Zuo) on down.  This must be somehow related to the "official who grasps an offering" meaning, but I'm not sure by exactly what mechanism.  It is noteworthy that Schuessler, Dictionary of Early Zhou Chinese, p. 583a gives only the meaning of "eunuch" for sì 寺.  In coming to terms with the origins of the graph shī 詩 ("poetry"), this early emphasis on the meaning of sì 寺 as "eunuch" should not be ignored.

Perhaps the evolution of shī 詩 proceeds something like this:  sì 寺 ("eunuch") > official responsible for the ritual recitation of odes at court > place where the ritual recitation of the odes by eunuchs takes place.  In this scenario, the sì 寺 ("eunuchs") who recited the odes would be like the castrati of medieval and renaissance Europe, part of a tradition that goes back all the way to Sumer.

In this regard, it is essential to consider the last stanza from "Xiàng bó" 巷伯 ("Senior of the Inner Passageway") in the "Xiǎo Yǎ" 小雅 (Lesser Elegantiae) section (#200) of the Odes:


It reads, in Karlgren's translation:

The road of the Willow garden (leans on =) is close to the acred hill;
the eunuch Meng-tsï has made this ode;
all you many lords, carefully listen to it!

It is remarkable that here in the Odes we have a sìrén 寺人 ("eunuch") named Mencius (!) composing a shī 詩 ("poem").  Legge (p. 349) refers to Mengzi in the first person, and on p. 347 provides this wonderful annotation:

The title of this ode — Hëang pih — is not taken from any of the stanzas, but is nearly equivalent to the 寺人, or eunuch, of st. 7.  巷 was the name of a passage in the interior of the palace, of which the writer had the superintendance, — as is denoted by the 伯.  He was perhaps the chief of the eunuchs.

Another ode from the Poetry Classic, "Sōng gāo" 崧高 ("Lofty"), #259 in the "Dà Yǎ" 大雅 (Greater Elegantiae) section ends with this remark:


This is translated by Legge (p. 540) as:

[I], Keih-foo, made this song
An ode of great excellence,
Of influence good,
To present to the chief of Shin.

It is interesting that the piece (or its performance?) is here called sòng 誦 ("ode"), whereas shī 詩 seems to be referring not to the poem as such, but to some quality of the piece or its performance ("its SHI [rendering; execution] is expansive; its FENG [air; style] is superior").

Now, in an attempt to bring a close to this already overly long post, I would like to ask to what extent radicals (semantophores) were used on the oracle bones?  bronze inscriptions?  I pose these questions for two main reasons:

1. So many characters we know of that consist of a phonophore and a radical (semantophore) began on the OBI and the BI as just the phonophore, which then had semantic significance, but later had a radical added to distinguish the original meaning from new meanings that were  assigned to the character.

Case in point:  zhǐ 止 ("stop" [the opposite of "go"]; the OBI form showed a foot, with the toes clearly evident).  As this character became coopted by so many other meanings ("rest; stay; dwell; reach; arrive; impede; detain; capture; await; reduce; endpoint; only; just; till; to", and many others), a new character was created, viz., zhǐ 趾 ("foot"), with a supernumerary zú 足 radical (another word for "foot") added to restore and emphasize the original meaning of zhǐ 止.  This is a story that is repeated countless times in the history of the Chinese script.

2. I'm wondering whether shī 詩 itself might have been like this.  As we have seen, in the late 60s, both Chen Shih-hsiang and Chow Tse-tsung wrote long, elaborate papers arguing that shī 詩 (but heavily emphasizing the 寺 part) was ultimately derived from zhī 之 or zhǐ 止, which are probably cognate, and both of which have to do with the idea of "going".  As sì 寺 acquired other meanings, might someone not have added the 言 radical to distinguish the idea of poetic recitation — if the archaic form of this character did indeed signify that notion — from the other added-on meanings?  Of course, the BI form of 寺 did have a 止 on top and a hand underneath and conveyed the notion of holding or grasping, which later shows up as chí 持 (note the added radical to distinguish this 寺 from all the other 寺-derived characters that were popping up.)  I'm thinking that the person who was engaged in courtly poetic recitation may have been an officiant who held a libation or ritual implement or sign of office of some sort.  After all, we do know that — already before 寺 acquired the notion of Buddhist temple — it already meant "officiant" and the official building where he carried out his duties, and it also meant "eunuch" (later written as shì [寺 inside of 門]), presumably because these duties may have been carried out by eunuchs.

Semantophores definitely existed on oracle bones, but to a much, much lesser extent than later became the case. According to Chu Ki-cheung's 朱歧祥 calculations (in "Jiagu zibiao" 甲骨字表), which are not necessarily entirely accurate but are generally reliable, 19.7% of OBI graphs (out of the number of unique graphs, not out of the total count) are xíngshēng 形聲 characters, made up of a phonophore and a semantophore.  Probably the overwhelming majority of those characters, though, write proper names, meaning that semantophores are much less common than the 20% figure if only terms that are not proper names are included. That said, many graphs such as 星、宅、室、暮、龢、律、祀, to choose some of the most obvious examples, appear in the OBI in structurally-identical or -similar forms writing the same words they would write in classical Chinese, and many others are written with graphs that didn't survive into modern script (or write words that didn't survive at all or haven't been identified).  I believe that a higher percentage of the graphs used on Zhou bronzes have semantophores, though this would need to be checked and quantified.  That said, it is very, very common for words in OBI or BI to be written with only a phonophore, and for the same words to be written with more complex graphs in later periods. The kind of process I describe in (2) definitely frequently happened across the history of the script and language.

All that I've written above is about the origins, evolution, and composition of the character used to write the word for "poetry" in Chinese.  What about the word itself?  Here we are on even less firm ground.  Axel Schuessler (ABC Etymological Dictionary of Old Chinese, p. 462), compares the Old Chinese pronunciation of shī 詩 ("poetry"), *lhə to Lushai hlaaR ("song; poem; poetry", where "R" is a superscript; citing Nicholas Bodman [1980, p. 181]).  Lushai is now known as Mizo, a language belonging to the Kukish branch of the Tibeto-Burman family of languages.  (N.B.:  Notice how finely grained the classification scheme for Tibeto-Burman is in contrast to Sinitic, where many people [including some scholars] think that there is only one language and hundreds of large and small "dialects".)

The late Han (ca. 2nd c. AD) reconstruction of shī 詩 ("poetry") is śə.

Schuessler's ABC Etymological Dictionary of Old Chinese doesn't have an entry for sì 寺, but his Minimal Old Chinese and Later Han Chinese does have one.  In it, we find that sì 寺 is grouped with zhǐ 止 ("stop"), and its Old Chinese pronunciation is s-ləh, which — through shī 詩 ("poetry"; *lhə) — would seem to link it phonologically with the Lushai word for poetry (not identical, but still perhaps cognate).

In conclusion, the character for shī 诗 ("poetry") does not mean or derive from "word" + "temple".  While this kind of folk etymology is very widespread for Chinese characters, both in China and abroad, critical thinkers should pay little heed to such whimsical assertions.  Those who accept such childish / grandmotherish pseudo-etymological explanations actually do a disservice to Chinese notions of poetry and all manner of other concepts by exoticising and distorting them.  The real derivations of Chinese characters are both more interesting and more profound than the superficial ones.

[h.t. Andrew Shields; special gratitude to Matt Anderson for supplying me with textual and manuscript sources, as well as phonological and paleographical evidence; thanks to Jonathan Smith, South Coblin, Axel Schuessler, Wolfgang Behr, Bill Boltz, Jeff Riegel, and Bill Nienhauser; all errors are my own]



  1. Rubrick said,

    October 4, 2013 @ 3:34 pm

    This post brings to mind a general Language Log question I've often had: How do you (or, notably, Mark) decide whether a fairly lengthy analysis requiring a considerable amount of research should be a "mere" LL post or developed into a published academic paper (perhaps by a grad student)?

  2. J. W. Brewer said,

    October 4, 2013 @ 4:07 pm

    Even if true, why would it be relevant? Knowing that the English words for poet/poetry distantly descend from a Greek verb meaning approximately "to make" is a fun bit of etymological trivia, but does it tell you anything meaningful about the nature of poetry or the poet's role in our culture? I wouldn't think so. (It does explain why medieval Scots used the calque "maker" or "makar" to mean poet, which in turn helps explain one poem you might come across in a large enough standard anthology, viz. Dunbar's "Lament for the Makaris," but that's full enough of archaic Scots lexical items that most readers will need an annotated version to understand it anyway.)

  3. the other Mark P said,

    October 4, 2013 @ 4:53 pm

    Even if true, why would it be relevant?

    There is a longstanding pattern of "proving" Chinese culture is superior by virtue of its antiquity and/or subtleness. And not just by the Chinese.

    You see statements about how Chinese culture is "x thousand years old", and "continuous". Apparently invasions by Mongols and Manchus and decades of Maoism don't affect the Chinese, but similar events do affect other parts of the world!

    In this case we are asked to believe that the Chinese culture is more attuned to poetry based on, of all things, the etymology of the word for it.

    The best way to fight nonsense is with facts.

  4. Victor Mair said,

    October 4, 2013 @ 4:58 pm

    @Rubrick, @ J. W. Brewer

    I began this post as a reply to a Language Log reader, who — with good reason — questioned a popular etymology. He was very happy when I told him that I'd look into the claim. In order to refute it, I started to do a bit of research, and it ended up taking most of the last week. I think it was worth the effort, both to set the record straight about the false etymology (which is being bruited about the cybersphere), and also because I was able to propose a completely new approach to the origins and meaning of the term, one that has eluded scholars working on this problem in the past.

    I usually try to keep my LL posts relatively short, but — mindful of the large and sharp audience — I never think of them as "mere". On the other hand, I never set out to write a long post, though sometimes the necessary investigation just makes them end up that way. Believe you me, I'd much rather write short and simple posts than one like this, which became excruciating near the end.

    I'm not sure what J. W. Brewer means when he says "relevant". "Relevant" to what? Scholarship? The truth? Life?

  5. SimonMH said,

    October 4, 2013 @ 6:03 pm

    A poem's not a temple, but has feet
    beating the ground, or perhaps an aspiration
    But now we hear, in little odes discrete,
    the scop's old lyre, which harps upon castration

  6. maidhc said,

    October 4, 2013 @ 6:20 pm

    "Ritual castration" is a particularly apt way to represent the notion of poetry, one that is conspicuously missing in Western culture.

  7. julie lee said,

    October 4, 2013 @ 7:25 pm

    Thank you, Victor Mair, for the detailed discussion of the Chinese character for poetry, shi詩 or 诗 (abbreviated form). I certainly agree, with some linguists, that, as Mair has pointed out, at least 85% of Chinese characters are composed of a meaning-component and a sound-component, and this is true of the character 詩/ 诗. If the non-linguist, non-Sinologist, person would realize this 85% fact, many explanations and books on the hidden meanings of Chinese characters would be abandoned. I have Chinese friends who explain their own fanciful etymologies of Chinese characters to me because they simply don't know that at least 85% of Chinese characters are simply combinations of a semantic component and a phonetic component. A respected old Chinese friend sent me a book written by an American who explained various elements in Chinese characters as coding messages from the Bible. The structure of each of those characters could simply be explained as composing of a semantic component and a phonetic component. My friend, a fervent Christian, felt so happy about this book explaining Chinese characters as coding the Bible that I simply did not have the heart to tell him that it was an elaborate exercise in wishful thinking.
    Victor Mair's post shows that even learned Chinese scholars can give us false etymologies of Chinese characters.

  8. flow said,

    October 4, 2013 @ 8:02 pm

    thanks for the read.

    i seem to remember that Peter A. Boodberg—possibly in his "Proleptical Remarks on the Evolution of Archaic Chinese"—once wrote to the effect that even though many people think so, 東 'east' is not really / scientifically / etymologically depicting 'a 日 sun behind a 木 tree: east', but is rather a phono-semantic compound. i couldn't find an online version of that paper, but i did find http://www.grsampson.net/ARoc.pdf whose authors comment on the formation of the character 明 in the way that express my doubts on reading Boodberg's analysis. to quote:

    "In any case, since for centuries many people have indeed believed that the graph 明 was motivated by the semantic relationship between ‘bright’, ‘sun’, and ‘moon’, how can that belief be self-evidently false?"

    the thinking here is: accepting that many centuries or millennia ago, 東 and 明 did not originate as 會意 semantic compounds but rather as 形聲 phono-semantic compounds, it still seems to be a fact that a vast majority of native users of chinese have learned that 東=木+日 and 明=日+月 graphically (obviously true) and semantically.

    now given that writing is a social construct (invented and permanently re-established by a community of users) and a social contract (i.e. depending on mutual understanding what a given piece of writing should signify), how can it be that a belief—if 'wrong' in the sense of 'it was wrong 2000 years ago'—does play no role in the *synchronic* appreciation?

    i do not want to support any language mysticism or claim that chinese is superior because its speakers have the wisdom to write EAST as (the location of the) SUN BEHIND A TREE (which could also mean 'west', right? well, that's BIRD IN ITS NEST according to some).

    still, when you have a written contract that is contested, the ruling should not be based on what its terms do really / scientifically / etymologically mean, but rather what they meant to society at large at the time the contract was made. so maybe when a contemporary learns 詩 and associates "solemn words spoken in a place of worship: prayer > poetry!" they might be completely off the tracks, etymologically. it might still be a force to be reckoned with if a billion people do the same.

    that said, your insights were very elucidating for me. and yes, relevant, too.

  9. Wentao said,

    October 4, 2013 @ 10:01 pm

    @julie lee

    To me there seems to be some truth in the theory that (at least some) phonetic-components in 形声字 are a secondary semantophores. For example the 右文 theory concerning characters 钱、浅、贱、线 suggests the phonetic-component signifies "few, little".

    Characters with the same phonetic-components must have been pronounced similarly, and I don't believe the similarity is entirely arbitrary and purely coincidental, especially for a highly isolating language like Old Chinese.

  10. Carl said,

    October 4, 2013 @ 10:09 pm

    Prof. Mair, I love your posts, but for non-Sinophiles, perhaps it would be use to get a separate blog. I'm not sure how much a non-reader of Chinese can make of some of your more in-depth posts.

  11. Victor Mair said,

    October 4, 2013 @ 11:07 pm


    Thanks for your kind words.

    One principle I try to adhere to scrupulously is to always include romanization and translation for all characters. By doing so, I hope that readers at various levels will be able to follow the gist of my posts.

  12. julie lee said,

    October 4, 2013 @ 11:57 pm

    I agree that some phonetic components, or phonophores, of Chinese characters can also at the same time serve as secondary semantic components, or semantophores. Victor Mair makes a good case for si寺 "eunuch, castrati cantor" in shi 詩 "poetry" serving both as phonophore and secondary semantophore. My own favorite character with phonophore serving also as secondary semantophore is the character ren 忍 “endure (pain, etc.)", where the character ren刃 “blade, knife" is written over the character xin 心 "heart, mind", and where ren刃 is the sound-component of the character ren 忍 "endure" , and xin 心 "heart, mind" is the meaning-component, or semantic classifier, of the character ren 忍 , because "to endure" is something mental or emotional. But ren刃 "blade, knife" is also a secondary semantic classifier, secondary semantophore, because it shows a knife over the heart/mind, indicating mental or emotional pain, which you suffer when you endure something. Quite a lot of Americans know this character ren 忍 because they have it as a tattoo.

    I agree that characters with the same phonophore were more or less similar in pronunciation at some point in place or time.

    I suspect that si 寺 "eunuch" "officiant" and shi侍 (also pronounced si in some dialects) "to serve" are cognates. A eunuch is a kind of servant or servitor.

  13. michael farris said,

    October 5, 2013 @ 1:39 am

    Your esteemed (and oddly adverse to relatively informed give and take) colleague Geoffrey K. Pullum recently wrote:

    "the Chinese languages share a writing system that is simply not fit for purpose: taking years to learn, and incredibly hard to adapt to many purposes, it is holding China's progress back by many decades"

    This cuts straight to the core of some of your postings (not this one so much but in general). Care to comment?

  14. Rubrick said,

    October 5, 2013 @ 1:51 am


    I'm no Sinologist, nor even a Sinophone, but while many of Prof. Mair's posts exceed my full comprehension, I enjoy reading them nonetheless.

  15. Victor Mair said,

    October 5, 2013 @ 6:32 am

    @Michael Farris

    I think the fact that GKP usually does not open comments has actually made the general level of "give and take" at Language Log more informed and polite than it was before he made the decision to close the comments to most of his posts (he still occasionally does open them). Knowing that, if too many people become overly snarky, trollish, and downright unreasonable, we might all close our comments, generally has made for more civilized discourse than before he made that grave decision. While we still occasionally have a few obstreperous individuals to deal with, in most cases I find the conversation in the Language Log comments to be remarkably mellow, highly edifying, and satisfyingly smart.

    Please remind me where GKP made that amazing statement. I have a vague recollection, but can't place it now.

    The best answer I can give to the question you raise in your third and final paragraph is to recommend that you read these three books by William C. Hannas:

    1. Asia's Orthographic Dilemma

    2. The Writing on the Wall

    3. Chinese Industrial Espionage: Technology Acquisition and Military Modernisation (with James Mulvenon and Anna B. Puglisi)

  16. marie-lucie said,

    October 5, 2013 @ 7:23 am

    @Carl: As a non-Sino-whatever, I agree with Rubrick. Dr Mair's postings teach us a lot about Chinese. We don't have to understand everything in order to learn something.

  17. michael farris said,

    October 5, 2013 @ 8:45 am

    "Please remind me where GKP made that amazing statement. I have a vague recollection, but can't place it now"

    Quite recently, here:


    It's in the next to last paragraph.

    I've read 1. and thought it was overall reasonably convincing though many character enthusiasts seem to disagree.

  18. J.W. Brewer said,

    October 5, 2013 @ 8:52 am

    By "relevant" I was trying to separate out the two issues of: a) what is historical etymology of a given word/phrase; and b) what is the present meaning of a given word/phrase. (I'm not sure if "etymology" is exactly the right concept for how a particular complex kanji/sinogram got formed as a combination of simpler ones, but it's close.) Often people make specious claims about b based on a, suggesting that if you know the word's origin you now have some sort of superspecial privileged insight about its true meaning. That is often coupled with a bogus folk etymology, but even if coupled with a historically accurate etymology is still a fallacy. I did not mean to suggest that etymology (or the history of kanji-coining) is not an independently interesting subject for which accurate scholarship should be preferred to inaccurate.

  19. julie lee said,

    October 5, 2013 @ 9:16 am

    @Victor Mair:

    The quote @michael farris gives about Victor Mair's work comes from Geoffrey Pullum's most recent post, "Food Logistics: a Sign of the Times", which is about English becoming a world language. GKP then gives these views about Chinese (which includes the excerpt given @michael farris):

    "Never imagine that because China is growing powerful and Chinese allegedly has a billion speakers we are going to see Chinese becoming more important to the world than English. First, there is no such thing as the Chinese language: Chinese is a language family, and there are far fewer people who are fluent in the politically dominant member, Mandarin, than the Chinese authorities would like you to think. Second, the Chinese languages share a writing system that is simply not fit for purpose: taking years to learn, and incredibly hard to adapt to many purposes, it is holding China's progress back by many decades. And third, nowhere in the world is there a country outside China where Chinese is used by non-Chinese to communicate with other non-Chinese.

    "English (for all its faults) has that peculiar international lingua franca status, and Chinese does not. And I do not see that changing. Not in fifty years, and perhaps not ever."

    It also struck me that the views on Chinese ("Chinese is a language family…writing system…holding back China's progress by many decades") have been those argued and explained by VHM in his LLogs.

  20. dainichi said,

    October 5, 2013 @ 9:43 am

    @Julie Lee

    "I suspect that si 寺 "eunuch" "officiant" and shi侍 (also pronounced si in some dialects) "to serve" are cognates"

    I've often wondered if it was common for similar-sounding morphemes to be cognates (and have similar meanings) when characters were created. If the character for one of these cognates was used as part of the character for the other, I can see how the semantophore/phonophore line could be hard to draw.

  21. Jon said,

    October 5, 2013 @ 3:56 pm

    JW Brewer's comment struck a chord with me. I personally am very interested in word origins, and the weird paths that words take as they pass through history and between languages. But I get extremely irritated with those who take these histories to imply something deep about the essence of a word. Usually these people take one of the meanings of a word in Latin or Greek, and spin some story about what the descendant of that word really means in English. As if Latin was the first language, and some homeopathic essence of meaning had been transmitted down the generations of changing usage,

  22. Jonathan Mayhew said,

    October 5, 2013 @ 8:03 pm

    "It is important to observe this primary sense of making form when discussing the term poiesis in the context of poetry and poetics. The poet as homo faber (man, the maker) and the poem as made thing are commonplaces that persist throughout the history of Western poetics"

    Princeton Encyclopedia of Poetry and Poetics, 45th edition. p. 1071.

  23. J.W. Brewer said,

    October 5, 2013 @ 8:56 pm

    Yeah, that's a good example of the sort of meaningless etymology-as-destiny blather I was talking about. Novels, plays, turtleneck sweaters, park benches, champagne flutes, etc etc etc.–they're all "made things." Homo faber makes lots of sorts of things. A poem is not the "made thing" par excellence any more than a novel is the "new thing" par excellence.

  24. Victor Mair said,

    October 5, 2013 @ 9:08 pm

    "A poem is not the 'made thing' par excellence any more than a novel is the 'new thing' par excellence."

    And words like "poem" and "novel" don't mean anything in particular.


    Goodness, I don't see what you're getting so worked up about. The main purpose of this post was to show that "word + temple" is a false etymology for Chinese shī 詩 ("poetry"). Secondarily, it was interesting to discover that eunuchs had such an intimate relationship to poetry more than two millennia ago. Nobody is claiming that they do today.

  25. JS said,

    October 5, 2013 @ 10:28 pm

    I've often wondered if it was common for similar-sounding morphemes to be cognates (and have similar meanings) when characters were created. If the character for one of these cognates was used as part of the character for the other, I can see how the semantophore/phonophore line could be hard to draw.

    Rather than one case being more "common," to me it's better to say that cognate pairs were at very early stages more likely, compared to non-cognate (near-)homophones, to be written with a single character and/or two structurally-related characters. The existence within the script of such connections, because cognates would naturally have been similar-sounding, could have served as motivator for graphic redeployment/adaptation in cases of etymologically unrelated homophones — the device so pervasive within the mature script and perhaps instrumental to its growth to maturity.

  26. julie lee said,

    October 6, 2013 @ 4:11 am

    "I've often wondered if it was common for similar-sounding morphemes to be cognates (and have similar meanings) when characters were created."

    Aren't there always cognates with similar sounds and meanings in a language (e.g., English "shirt" "skirt"? Why shouldn't there be many cognates in the Chinese language when Chinese characters were created? Characters are a script, a writing system, to be distinguished from language. A script is a way to represent a language in writing. So Chinese can be written with the Cyrillic script (as in Dungan Chinese) or with characters, or with a romanized script such as pinyin.

    @ dainichi also says: "If the character for one of these cognates was used as part of the character for the other, I can see how the semantophore/phonophore line could be hard to draw."

    Yes, a component of a character can be simultaneously the phonophore and a semantophore.

    For instance, take sì 寺 "eunuch" , shì 侍 "wait upon; attendant" and chi 持 "support" as examples.

    sì 寺 "government center/building; hold; support; aid; rely on "
    shì 寺 (note the difference in sound), anciently interchangeable with shì 侍, " wait upon; attendant, often a eunuch"
    (the meanings are given in this order in Hanyu Da Cidian (Comprehensive Dictionary of Chinese,IIB-1249, comparable to the OED of Chinese)

    shì 侍 "wait upon; attendant, often a eunuch; support; submit a recommendation to a ruler; attending concubine or female attendant; be in charge of" etc. (Hanyu Da Cidian, IB-1312).

    In shì 侍, the component on the left is a semantophore, a semantic classifier meaning "person", while the 寺 component is the phonophore. However, it is was also a semantophore anciently. But as time went on, the layman (non-specialist in classical Chinese) would likely not know the ancient multiple meanings of 寺 and would only see it as phonophore.

    si 寺 "to hold, grasp, support" anciently, but later and now written chi 持 (where the left component, a symbol for "hand", is the semantic classifier and the right component 寺 is the phonophore.

    si 寺 "to rely on, depend on" anciently, but later and now written shi 恃 , where the left component, a symbol for "heart", is the semantic classifier and the right component 寺 is the phonophore.

    A non-Sinologist or layman is not likely to know the many ancient meanings of the symbol 寺 in these characters for "to hold" "to rely on" etc. and will only see it as a phonophore. The added symbols on the left are to disambiguate the meanings of the ancient 寺 symbol.

    Just because the above words have the same phonophore and were once homophones doesn't necessarily mean they were all cognates.

    I believe VHM has explained much of this in his post but more succinctly.

  27. Victor Mair said,

    October 6, 2013 @ 5:43 am

    Jonathan Mayhew made an excellent contribution to this thread by quoting The Princeton Encyclopedia of Poetry and Poetics on the concept of the poet as Homo faber, but J. W. Brewer nonchalantly ("Yeah…") laughs it off with a bunch of non sequiturs ("Novels, plays, turtleneck sweaters, park benches, champagne flutes, etc etc etc…."). Of course, Homo faber has taken on a larger role in Western intellectual history, even in modern times, but this does not negate the fact that poets and literary theoreticians have paid particular attention to this concept. As an undergraduate English major specializing in Chaucer, I am especially sensitive to the idea of the poet as Homo faber. Chaucer's view of himself as such is brought out well by V. A. Kolve in Chaucer and the Imagery of Narrative: The First Five Canterbury Tales, pp. 134-135.

    For those who wish to explore the evidence further, I encourage you to read the examples cited on p. 1071 of The Princeton Encyclopedia of Poetry and Poetics (thanks to Jonathan Mayhew for this good reference).

  28. Jonathan Mayhew said,

    October 6, 2013 @ 11:07 am

    Thanks, Victor. And I meant the 4th edition of the PEPP, not the 45th.

    Poetics has a meaningful etymology, in my view, precisely because it has had meaning for many people historically, as well as being the correct (non-fanciful) origin of the word. Western poetics does take a "poetic" view of etymology, as in other words like techné. Within this system, the origins of words are significant and meaningful. Of course, other notions of human creativity might also invoke poiesis as creative principle.

  29. Janet Williams said,

    October 6, 2013 @ 3:19 pm

    @julie lee:

    Christian explained Chinese characters as coding messages from the Bible. When I was young, my church leaders explained that the character for 'ship' 船 meant 'A ship carrying 8 people.' (Left: ship on the left; Right: 8 口人 = 8 people).

    God in Ancient China 古代中國人的神 – Part 1 (English & Chinese) , on Youtube by Kong Hee shocked me with his assertions of Biblical links with many Chinese characters. He explained, for examples:

    園 garden means the Garden of Eden: 土:earth; 口: breathe; 人:2 people; a woman came out; and a border for garden;

    禁 forbid; it was explained as two trees on top 林:Tree of Life and Tree of Knowledge of good and evil) and 示:God gave them a revelation.

    There were many more similar shocking revelations.

  30. Dan Hemmens said,

    October 6, 2013 @ 4:24 pm

    Often people make specious claims about b based on a, suggesting that if you know the word's origin you now have some sort of superspecial privileged insight about its true meaning.

    Quite so.

    Of course this (the original claim, not Professor Mair's post) is an even more specious brand of the same thing. Suggesting that the (supposed) etymology of a word in a *completely different language* can somehow give a superspecial privileged insight into into not merely the word itself but into the whole concept that the word describes.

    I mean maybe I'm dense, or maybe I have no soul, but I honestly don't understand what the original "word temple" claim is suppose to be saying. That we should see poetry as being like a temple for words? Or like a temple built from words? Or that poetry is uniquely suited to religious functions?

    It occurs to me, incidentally, that "the Chinese word for X is made up from the word for Y and the word for Z" is probably a good candidate for the Snowclone Database.

  31. Janet Williams said,

    October 6, 2013 @ 6:59 pm

    @julie lee:

    My apology. I meant to explain this better:

    When I was young, my church leaders explained that the character for 'ship' 船 meant 'A ship carrying 8 people.' (Left: 舟 ship; Right: 八口人 = 8 people).

    The ship 船 was explained to me to be related to Noah's Ark, as there were exactly eight people, 八口人,on board.

  32. julie lee said,

    October 6, 2013 @ 10:53 pm

    @Janet Williams

    Thanks! I probably threw away the book, so it's good of you to provide some examples of the kind of thing I meant. That book was by an American Caucasian, but Chinese people like Hong Kew can also see the Bible coded in Chinese characters.
    I'm in a friend's house right now and don't know how to type Chinese characters on this laptop, but in the Chinese character for chuan "ship" which you give, the component on the left is the semantic element (semaphore) meaning "boat" and the component on the right is probably the phonetic element (phonophore) indicating the pronunciation of the character (I don't have a dictionary on hand).
    In the character for yuan "garden", the component inside the square is the phonetic element yuan and the outer square frame is the semantic element meaning "enclosure" —a garden being an enclosure.
    In the character for jin "forbid", the bottom component is the semantic element or classifier meaning "indicate" —to forbid is to indicate something. I am not sure about the top element and I don't have a dictionary to look it up. It is the character lin"forest", but may have been used as a phonetic element or as a semantic element. The phonetic element need not be identical to the sound of the character. It sometimes only gives a clue to it.
    Thanks again for giving examples of how some people interpret Chinese characters as coding the Bible.

  33. Stephan Stiller said,

    October 7, 2013 @ 8:21 am

    @michael farris

    Geoffrey Pullum explains his decisions regarding comments in this LL post of his.

  34. Vanya said,

    October 8, 2013 @ 4:52 am

    I suspect the efficiency of any writing system is probably just less important than either the reform-minded or the tradition-beholden want to believe. Blanket condemnation of character based systems seems almost as based on emotion as people who think characters are magical repositories of folk wisdom. The benefits of characters are surely overstated, but proponents of far-reaching reforms should be the ones bearing the burden to prove the necessity for reform. Common sense would tell you that characters are inefficient, archaic and should be abandoned – but no one ever produces any empirical evidence to demonstrate that character systems are harmful. For example GKP says "it is holding China's progress back by many decades"". Compared to what? How do we know what China's "progress" is supposed to be? And Taiwan and Hong Kong seem to have mostly caught up with the West in wealth, happiness and culture despite the fact that their children waste 1000s of their childhood-hours memorizing and forgetting characters. Japan seems to be steadily moving away from characters and towards more romaji and kana, yet their "progress" has been slowing noticeably over the past decades. North Korea abandoned characters entirely and has perhaps the most logical writing system humans have ever devised – not a lot of "progress" though. Vietnam abandoned characters decades ago – are they progressing faster than China? Indonesian is much easier to spell and pronounce than English, why isn't Indonesian the world language? The idea that an inefficient writing system holds cultures back (or makes them better) seems to have as much evidence as the Sapir-Whorf hypothesis, and is probably based in the same idea that we can "fix" cultures by treating symptoms rather than the underlying disease. Or maybe the issue is that most people spend years learning useless trivia during their youth in every industrial and post-industrial cultures – whether that time is wasted memorizing characters, memorizing names of US Presidents, or learnng all the lyrics to every David Bowie song may not matter that much in the end.

  35. JS said,

    October 8, 2013 @ 12:02 pm

    It seems a significant part of reformers' motivations relates to providing a readier path to literacy for China's immense numbers of socioeconomically disadvantaged students. It is true that there is an "underlying disease" here that cries out for treatment, but one that improved literacy levels could, in principle, turn out to remedy directly.

    In addition, as regards foreign learners of Chinese, in our experience the script is a central if not the primary mode of accessing the language and is thus experienced as a very real limiting factor — for us, work with Romanized texts, like regular exposure to and practice with spoken language, is a critical tool in mitigating this difficulty.

    That said — for precisely the above reason, foreign observers tend to conflate the difficulties involved in learning the language with those of learning the script. From working with both native speaker and non-native speaker beginning readers, it seems clear to me that for the former, 99% of the hard work is done; in my experience, proficient speakers of Chinese, given a reasonable degree of motivation, become proficient readers by devoting just a couple of hours every Sunday over three or four years. Not to say learning Chinese characters is at all easy, but, as you suggest, there is more than enough time in the day for it.

  36. Janet Williams said,

    October 8, 2013 @ 2:09 pm

    @julie lee:

    I watched a bit more of God in Ancient China 古代中國人的神 -YouTube on how the Bible stories could all be explained in Chinese characters for you. I recommend sinologists (especially venerable V. Mair) watch it if you all could bear the interpretations by the minister Kong Hee.

    A few more of his interpretations are as follows:

    1) Greed: 婪:There is a woman 女 underneath the wood 林, because it was Eve who was the first to be tempted.
    2)兄/兇 look the same, as they mean Brothers and Murder: relating to the first murder – Cain and Abel's story. And, 凶 means that it is the Mark on Cain's Forehead.
    3) 義 (righteousness): There is a lamb 羊, and 我 means people with hands and they use a spear to pierce something. “The Chinese already prophesised that God will bring His Lamb, and people will take a spear, and with their hands, pierce him.”

    There are a lot more interpretations in Mr Kong Hee's sermon for your enjoyment.

  37. J. W. Brewer said,

    October 8, 2013 @ 3:09 pm

    What Vanya said (except perhaps for his failure to allow for the possibility that some Bowie songs are much more worth memorizing than others). How much more powerful could/would Japan really have gotten without its alleged script handicap? And of course English orthography is notoriously difficult/opaque as Latin-scripted languages go, but if that's a horrible disadvantage vis-a-vis the Finns, we nonetheless seem to be doing ok. (What's the next-worst European language, orthographically speaking? Maybe French? They haven't done too badly for themselves globalinfluencewise either.)

  38. julie lee said,

    October 8, 2013 @ 11:02 pm

    @ Janet Williams:

    Thank you for the additional examples of mis-interpretations of Chinese characters.

    I would interpret lan 婪 "greed, gluttony", for example, as composing of a semaphore nvu (nyu)女 "woman" and a phonophore 林, now pronounced lin (sounds change over time). Nvu 女 "woman" is the semantic classifier because the nvu 女 "woman" symbol is often used as semantic classifier for words denoting evil or vices. Articles have been written about this anti-woman bias in Chinese culture historically. For example, jian 姦, also written jian 奸, means "wickedness". In the first, three "woman" symbols are piled up. In the second, the left symbol, the semaphore, is "woman" , the right symbol is the phonophore. Jian 姦 also means " fornicate, fornication"; jian 奸 also means "treachery, traitor, felon". Lan 婪 "greed, gluttony" is also written lan啉 "greed, gluttony", with a kou 口 “mouth" symbol on the left as semantic classifier, since gluttony can be classified under the mouth. The phonophore lin 林 in both characters lan 婪 and lan啉 suggests these three characters were all homophones at one time. Sounds change over time, and lin 林 is pronounced lum in some Cantonese dialects, for instance.

  39. julie lee said,

    October 9, 2013 @ 9:48 am

    p.s. I should add to my comment above that, in the character lan 婪 "greed, gluttony", lin 林 is a character meaning "trees, forest". But here it functions merely as a phonophore. lin 林 is also used as phonophore in other characters, including

    lan惏 "greedy; gluttonous; cruel"—with phonophore lin林 plus a semaphore on the left meaning "heart/mind". Greed, cruelty is here classified under the heart and mind;

    lan 醂 "soak persimmon to remove raw taste [of not fully ripe persimmon]" –where lin 林 again functions as phonophore, and where the symbol on the left, the character you酉 "jar, amphora" is the semaphore or semantic classifier, since jars are used to soak the persimmons.

  40. Suburbanbanshee said,

    October 9, 2013 @ 12:50 pm

    Re: poets as makers – Sir Philip Sidney's The Defence of Poesy is probably the most extended defense of poetry as making things with the imagination. Sidney basically uses this to defend all fiction and all fantasy as well. So I don't think we want to throw out the etymological symbolism yet!

    Re: made-up etymologies and Just So stories — A lot of these things are really useful as memory aids, or are beautiful as poetic symbolism, or are just impressive as examples of ingenuity. The difficulty comes when people take them as factual historical truth.

  41. JS said,

    October 9, 2013 @ 4:44 pm

    Thanks @julie lee for lin 醂 — I've been looking at OC word(s) of the form *rəm, etc., meaning things in the vicinity of 'soak in liquid'; this must be the same word in different clothes.

  42. julie lee said,

    October 9, 2013 @ 7:17 pm

    nice to know the character for lin "soak persimmon in liquid" was useful.

  43. Janet Williams said,

    October 10, 2013 @ 10:25 am

    @ Suburbanbanshee:

    Re: made-up etymologies and Just So stories:

    Of course it is fine to make up stories relating to characters as memory aids. I ask my students to do whatever that helps them to memorise characters.

    However, when a minister treats his own interpretation of the origin of characters as the absolute truth, and this interpretation is accepted as the authority, I would see it as a problem.

  44. J. W. Brewer said,

    October 10, 2013 @ 12:19 pm

    http://en.wiktionary.org/wiki/%E6%9D%B1 claims that one of the just-so stories I specifically remember from when I learned some rudimentary kanji while living in Tokyo as a boy (character for "east" = sun-behind-a-tree) is apparently a false (but highly traditionally so) etymology. And of course it would separately have the problem a lot of just-so stories do: why should sun-behind-a-tree denote sunrise/east rather than sunset/west? (Oh, and the character for "west" = "bird-settling-into-its-nest-at-sunset"? How can you tell that isn't "bird-preparing-to-leave-its-nest-at-sunrise"?)

  45. JS said,

    October 10, 2013 @ 3:53 pm

    @J.W. Brewer –
    The latter property is not an indicator of a folk or "just-so" character etymology — the glyph writing dan4 旦 'sunrise' shows nothing but sun above horizon; the word indicated is not something you can "tell" from the character but rather a matter of convention.

  46. julie lee said,

    October 11, 2013 @ 11:56 am

    @Janet Williams, @JS:

    As you know, the plot can thicken in analyzing the structure of a Chinese character. I said above:

    "lan 婪 "greed, gluttony", lin 林 is a character meaning "trees, forest", that it's a phonophore here. However, the plot thickens. I think lin 林 also serves as a secondary semaphore here. Mu木 is "tree", lin 林, two trees, means "trees, forest", and lin 林 "two trees" here suggests "many, large quantity" , which suggests "greed, gluttony".

    林 (lin in Mandarin) is often romanized as lum (rhyming with English hum) in Cantonese or lam (rhyming with English hahm).

RSS feed for comments on this post

Leave a Comment