A Personal Encounter with the Submissive Woman/Bound Slave Rorschach Test
« previous post | next post »
[This is a guest post by Michael Broughton.]
I had an interesting Rorschach encounter with the oracle bone graph for woman a couple of years back. Oddly, this experience came in a rather roundabout way through an investigation into the character for interpretation, yi 譯. At the time, I was starting my Chinese translation business and wanted to come up with a meaningful logo for the business. I thought that an investigation into the character yi 譯 might help to inspire some ideas, and so I tried to do a little bit of digging into why it was written the way it was. Of note, the Liji (Book of Rites) has four characters for interpreting officials, as James Legge wrote in his elegant translation:
To make what was in their minds apprehended, and to communicate their likings and desires, (there were officers) – in the east, called transmitters (ji 寄); in the south, representationists (xiang 象); in the west, Di-dis (didi 狄鞮); and in the north, interpreters (yi 譯).
For me, there was something about the yi 譯 character that seemed fundamental to the nature of interpreting/translating. Obviously, the speech radical 言 on the side seemed to suggest that early forms of ‘translation’ were oral interpreting—as would naturally make sense in a predominantly pre-literate Zhou society. But I was more interested in right side of the character, having developed, in my earlier days, a keen interest in what was called youwenshuo 右文說, the idea that a character’s meaning is often conveyed by its phonetic determinative—the part of the character specifying pronunciation.
The right part of the character yi 譯 can be written as its own graph “睪,” and this has quite a few different readings—yi, ze, du and gao. It is used as the phonetic determinative for quite a few characters (the ze in xuanze 選擇 [choose] and Mao Zedong 毛澤東; the shi in jieshi 解釋 [explain]; the yi in yizhan 驛站 [an archaic word for a post station]). In all these words, the character was originally a specifier for pronunciation, and may, for some of them, even be a clue to a possible word family relationship (Axel Schuessler, in the ABC Etymological Dictionary of Old Chinese has yi 譯 [interpret] yi 驛 [relay station] and shi 釋 [explain] in a broad word family). In mainland China today, the character 睪 has been simplified in all these characters to 译*, though in Japan, it has, for reasons unknown to me, been simplified into 尺.
[*VHM: Only the phonophore on the right, which I am unable to type by itself; just imagine that the semantophore on the left has been stripped off.]
Looking up the character 睪 in the Shuowen Jiezi (c. 100 AD), I was interested to see that Xu Shen (c. 58-c. l48 AD) had glossed this character as something like “spy on,” with Xu Shen analysing the graph as a huiyi 會意 character made up of two combined semantic derivatives, mu 目 (eye) and nie 㚔 (an instrument used for punishment). Following the trail further, my oracle bone dictionary glossed nie 㚔 as a form of wooden handcuffs and presented some evocative early graphs that wouldn’t be easily interpreted as manacles today.
That the character for nie 㚔 is, in fact, a representation of some form of manacles is much more evident when we see members of the Shang and Zhou times wearing them, as we do in the oracle bone script for zhi 執 (which meant, in pre-Qin times, something along the lines of “to arrest”). Richard Sears’ painstakingly created (but unfortunately named) website, “Chinese Etymology,” has a great many oracle bone examples of the character zhi 執, which I have screenshotted below:
And thus it was, that as I was looking at these unfortunate pre-Qin captives, a strange thought crossed my mind. The position of the captive looked remarkably familiar, it was almost as though I had seen it before…where was it…ah, that’s right, the character for woman!
Could it be…were those tender hands in front originally bound by shackles. Was today’s lady unshackled, but still bound, her hands tied by the binds of the materially intangible but ever present patriarchy! The moment was a see-it-and-you-can’t-unsee-it one. Rorschach successful.
And that would have been the end of it, a moment of striking similarity and crazy coincidence, another exciting day in arduous challenge of learning Chinese. Except that…except that in 1937, a number of ceramic figurines were discovered in the Yinxu excavation site, some of the earliest figurines ever discovered. Figurines with their hands in shackles, some to the front, and some to the back. The Open Museum hosted by the Academia Sinica Center for Digital Cultures has a great close up of these (link here):
Could it be that the act of distinguishing female slaves from males was determined by the position in which their hands were tied? Perhaps captive ladies may have been seen as less dangerous, less prone to escape—hence the relatively more comfortable shackling position? If such character interpretation revisionism was allowed, imagine the flow-on possibilities.
Thus ends my own Rorschach experience with the lady of the crossed hands. As for the business logo, I took inspiration from the shackles. Perhaps translation and interpretation are, after all, about unlocking confusion and releasing meaning. The manacles alluding not to the binding, but rather the moment of words breaking free. Here is the final logo that came out of it. Thanks go to the captive slaves of the Shang and Zhou who inspired it.
Selected readings
Peter Cyrus said,
February 6, 2026 @ 7:11 am
"has quite a few different readings"
"was originally a specifier for pronunciation"
Don't those seem incompatible?
I've been wondering why Chinese never developed a phonetic script (until the arrival of Europeans). In other places, like Egypt, the phonetic elements became standardized, i.e. they always used the same phonophore for the same sound, and those then became simplified (to mark them as phonetic symbols?), and finally people starting omitting the semantophores. But that seems not to have happened in China.
Maybe multiple readings explains that. But do they arise because the same character is used to write multiple (spoken) words? Or because pronunciations diverge, or the character "captures" other semantically related words? "yi, ze, du and gao" seem pretty divergent.
Victor Mair said,
February 6, 2026 @ 7:45 am
@Peter Cyrus
Absolutely brilliant observations, questions, and suggestions.
Chris Button said,
February 6, 2026 @ 7:46 am
They did:
甲 k
乙 Ɂ
丙 p
丁 t
戊 b
己 ɣ
庚 ƛ
辛 s
壬 n
癸 q
子 ʦ
丑 x
寅 l
卯 ʁ
辰 d
巳 ʣ
午 ŋ
未 m
申 ɬ
酉 r
戌 χ
亥 g
The question is why they didn't use it.
For a (partial) answer to that from a different time/place, I recommend Gari Ledyard's discussion of why Hangul was originally so stubbornly resisted in Korea.
As I mentioned on the previous LLog thread, it is just a generic kneeling figure that "from an inscriptional perspective is undeniably devoid of any feminine features regardless of what was later associated with it".
As someone who worked briefly in the advertising/branding industry, I must say that I love it. It is layered in its representation and subtly evokes the story behind it.
Peter Cyrus said,
February 6, 2026 @ 12:28 pm
Aren't those the Celestial Stems? They're mostly used as ordinals, right? Is the assignment to sounds widely known? Is there a (hidden) relationship between ordinals and sounds, an alphabetic order?
Chris Button said,
February 6, 2026 @ 6:51 pm
@ Peter Cyrus
I should have begun my post accordingly: "In my opinion, which is based on various original proposals by Edwin Pulleyblank, they did."
In terms of controversiality, you might compare it to the proposal that 女 originally depicted a slave rather than a woman.
Lucas Christopoulos said,
February 6, 2026 @ 7:28 pm
For Egypt, unlike in Hellas, traditions were preserved for thousands of years with little change, and those who could read were mainly scribes, priests, high officials, some nobles, and only rarely skilled workers; literacy did generally not extend to the general population.
Chris Button said,
February 7, 2026 @ 4:25 pm
Phonetics aside (which I think are surmountable), the problem with applying a vowel-length distinction along the lines of 荼 *láɰ vs 茶 *láːɰ to 奴 náɰ vs 女 nàːɰˀ is that 女 is a "type b" syllable (symbolized by the grave accent).
Pulleyblank uses the vowel length distinction in "type b" syllables to account for alternations like 者 ~ 諸, which is different from its association with an r-like prosody in "type a" syllables like 茶 *láːɰ. Pulleyblank's (1995) linguistic reasoning makes sense, but it also rules out nàːɰˀ as an option instead of something like nràɰˀ for 女.
The main issue I have with Pulleyblank's vowel length proposal is the implication that the distinction existed across the lexicon, which essentially turns it into a wild card.
I prefer to see the vowel length distinction as something akin to the TRAP-BATH split in British vs American English. Potentially that could allow an aɰ ~ aːɰ alternation in Old Chinese and then a further /aa̯/ [ɑ] ~ /aːa̯/ [a] alternation in Han times. So 者 ~ 諸 would both have been -àɰˀ in OC (i.e., without the aɰ ~ aːɰ alternation in 荼 ~ 茶, etc.), but then in Han times would have alternated as [ɑ] ~ [a] to in turn give their different reflexes in Middle Chinese.
That then opens up nàːɰˀ as a possible reconstruction for 女. All that remains is the phonetic/phonological justification, which I believe is possible but is probably excessive for this post :)
~flow said,
February 7, 2026 @ 5:08 pm
@Lucas Christopoulos I don't think there's anything wrong with your statement, but I do believe that for Ancient Egypt we should reckon with 'literacy' not being a clear-cut binary thing. For one thing, we have differing levels of literacy today, and it is to be expected. Another consideration is that apparently scribes first learned hieratic script, and were only later (if ever?) expected to master hieroglyphs. A third observation is that we have many examples of cursive hieroglyphic writing executed by the (leaders?) of the working gangs that erected the pyramids; one of the earliest to be rediscovered in the early 19th c is written in red ink on the wall of one of the so-called Relieving Chambers that are situated above the King's Chamber inside the Great Pyramid of Giza; it was later deciphered as reading (ḫwfw)|, i.e. Khufu, inside a cartouche. This tells us that, at the least, the group leaders on royal construction sites as early as the 4th Dynasty (~2500BCE) were able to at least write the name of their working gang, the name of the ruling king, and numbers in cursive hieroglyphs. Whether those same people could have or would have read the many inscriptions that Egyptian temples were literally covered in, or, say, a papyrus with a literary content is another question. But we do have private letters from merchants. It's a little bit like the WWW in 1995: everybody knew it was there, and could be used, but most people did not have the means and the knowledge how to, and most of the time has no compelling reason to use it. When the necessity did arise, they might've asked a friend or their nephew to do it for them. This is largely how I imagine Ancient Egyptian literacy to have been in the cultural centers (Thebes, Memphis).
Yves Rehbein said,
February 10, 2026 @ 12:53 pm
@ ~flow, tangential to the previous point of discussion, was Cheop's name (ḫwfw, Khufu) written with consonantal signs already in the Old Kingdom?
@ Chris Button, the question why Chinese did not use consonantal writing is, I think, quite obvious: Where are the vowels? Greek was in the fortunate position amidst sound changes that consonants had become vowel onsets, only later.
By the way, the combination of 子 and 母 in the Oracle Bone Inscription of 乳 seems to be clear as far as feminine features (OC /*noʔ/ "milk; nipple", Baxter–Sagart), which is different from 好 (OC /*qʰˁuʔ/ "good", /*qʰˁuʔ-s/ "love, like", B–S).
Yet another comparison is more difficult to justify, since I mention vowels, compare ěr 耳 (OC /*C.nəʔ/ "ear", Baxter–Sagart).
[longwinding argument omitted] The gist is that hieroglyphic Luwian FEMINA is just an oval. As far as the nie 㚔 sort of shackle is concerned, I have to argue that e.g. the eye of a needle is rather an ear in German (Nadelöhr), though it looks like neither Öse ("eyelet; lug") nor, for that matter, Gewinde-Mutter ("nut (of a bolt)", lierally "mother") are related. Could be, though.
Chris Button said,
February 13, 2026 @ 9:33 am
@ Yves Rehbein
I'm not quite following your comment. However, one point of interest is that the onsets j- and w- were treated as vowels and assigned a glottalic onset as ʔj- and ʔw-, which is why neither w- nor j- appear in the list. They are basically subsumed under 乙 as Ɂ-.