Remembering Nüshu.

I posted about Nüshu, a form of writing developed by women in Hunan province, back in 2004, but a) that was a long time ago and b) the links are dead, so it might not be amiss to present Lauren Young’s Atlas Obscura piece on the subject:

In 1988, Yi Nianhua, a woman in her 80s, spent many evenings scribbling elegant characters at a table in her kitchen in a small rice-farming village in Shangjiangxu, China. With only a blunt writing brush, the elongated script came out fat and blotchy on the newsprint she used for paper. But Cathy Silber, a professor at Skidmore College in New York, worked alongside Yi in her kitchen, diligently deciphering and studying the written language.

“Out of the thousands of scripts that are gender-specific to men, here we have one that we know is gender-specific to women,” says Silber, who has been researching Nüshu since 1985. Yi was one of the last remaining writers of Nüshu, a fading script that only women knew how to write and read.

Stemming from the southwestern Hunan Province county of Jiangyong, a small group of women in the 19th and 20th centuries practiced this special script that no man could read or write. The writing system allowed these women to keep autobiographies, write poetry and stories, and communicate with “sworn sisters,” bonds between women who were not biologically related. The tradition of Nüshu is slowly vanishing, but at one time gave the women of Shanjiangxu freedom to express themselves.

In the middle of the 20th century, it wasn’t uncommon for Chinese women of higher socioeconomic classes to write songs, ballads, complaints, or stories, as Wilt Idema details in the book Heroines of Jiangyong: Chinese Narrative Ballads in Women’s Script. However, it was extremely rare to find such intimate texts from peasant women. As of 2012, there were approximately 500 known texts written in Nüshu, ranging from four-line poems to long autobiographical narratives. Today, the texts that have survived give researchers such as Silber the opportunity to peer into the daily lives of Chinese women throughout this period of history. […]

It’s an interesting story, and there are some striking images; apparently the last woman who really knew Nüshu died in 2004, the year of my earlier post. Thanks, Jack!


  1. John Cowan says

    Would it not be amiss, sisters, to write in antique script the sad tale of China, of the daughters of China? But no, we must begin by using the language of our time, and not the inventions of Jiāngyǒng.

  2. “Even though it wasn’t a secret, it was for all practical purposes used exclusively by women.”
    This brought to mind hiragana, which was similarly used as women’s writing originally.

  3. Huh, that hadn’t occurred to me.

  4. David Marjanović says

    …and for the same reason: women weren’t taught Chinese characters, but were of course well aware that writing existed and was useful, so they took a few (hundred) characters and expanded them into a syllabary.

  5. Trond Engen says

    So who made an alphabet of hieroglyphs?

  6. David Eddyshaw says

    Clever Canaanite construction workers, idling on the job.

  7. Miners, I thought, scratching their Proto-Sinaitic graffiti on the walls of the levels.

    But what I would like to see is a the specific hieroglyphics which they reinterpreted as ox, house, humper, and so on (as Cousin James had it). There doesn’t seem to be such a table anywhere online.

  8. David Eddyshaw says

    Sorry to say, I was sacrificing sense to alliteration. Mensa-level Moabite miners? Moses?

  9. Definitely miners.

    Even though we know people etched things into clay and painted the sides of pots, the letters that happened to survive inside a mine were definitely the original letters.

    Probably the ones that survived were drawn by the inventor himself, emerging in full flower from his imagination like Venus from her mussel shell.

  10. Stu Clayton says

    which they reinterpreted as ox, house, humper, and so on (as Cousin James had it).

    The internet gives me a few examples of original “and so on”s from Bill, for example in the Pragmatism lectures:

    # It is either a principle or a definition that 1 and 1 make 2, that 2 and 1 make 3, and so on; #

    # Truths emerge from facts; but they dip forward into facts again and add to them; which facts again create or reveal new truth (the word is indifferent) and so on indefinitely. #

    But there are even more “and so on”s in commentaries on him. Of course there’s nothing unusual about the expression “and so on”. Where did you get the idea that this was a particular tic of his (as you seem to suggest) ?

    I guess superior writers don’t do that, I know Luhmann doesn’t. I had never thought about it until now.

  11. What is the meaning of “scripts that are gender-specific to men” in this context?

    Does Young refer to scripts invented and used by men over the centuries, which women contemporaneous to those times could not interpret? The women’s illiteracy being a frequently-recurring effect (either intentional or consequential) of male monopoly on authority in patriarchal societies?

    Offhand, I cannot think of any writing systems that are literally gender-specific, in our world of today. None that are so gender-specific that they defy interpretation by anyone who does not personally share the writer’s gender.

    To endeavor to create an inherently gender-specific script would probably be to pursue an unreachable goal. The brains of male and female humans are structurally identical (although they usually do diverge in their functioning, as they regulate and manage the remainder of the body-at-large.)

  12. Athel Cornish-Bowden says

    Maybe a naive question, but is it just by chance that the character nü means female? What does shu mean?

  13. nü 女 = “female, woman”
    shu 書 = “writing, script”
    (These are the Mandarin pronunciations of the terms, rather than the local southern Hunan language that nüshu was used for.)

    Hangul in premodern Korea was also in part a “women’s script”. In Japan, hiragana was feminine, but katakana was masculine.

  14. Stu Clayton says

    To endeavor to create an inherently gender-specific script would probably be to pursue an unreachable goal.

    Well, gender-specific behavioral scripts are already in place. 🙂 What else is behind “male monopoly on authority in patriarchal societies” ?

    Sorry, your switch from “writing system” to “script” was an instance of Elegant Variation that gave me an opportunity I couldn’t pass up. I’ll just add that the notion of behavioral “script” is chock-a-block with tendentious premises, and so not a notion I have much use for. “Script” is a perfectly acceptable synonym for “writing system”, it’s the psychologists who are oversalting the porridge.

  15. The epic transmission of the early medieval Ogham script of Ireland to Hunan Province in China and its remarkable evolution en route is a story full of lacunae. Much of the credit has to be attributed to Nestorian nuns.

  16. David Marjanović says

    But what I would like to see is a the specific hieroglyphics which they reinterpreted as ox, house, humper, and so on (as Cousin James had it). There doesn’t seem to be such a table anywhere online.

    Tables 2 and 3 in this book chapter that we discussed almost 2 years ago.

    These are the Mandarin pronunciations of the terms

    Let’s add the tones, then: nǚshū.

  17. The internet gives me a few examples of original “and so on”s from Bill

    This is a truly spectacular example of non-communication. When I wrote obscurely “ox, house, humper, and so on (as Cousin James had it)”, I was referring to the use by James Joyce (who I call Cousin James because we are distant relatives) of the phrase “as semper as ox-house-humper” in Finnegans Wake.

    Now ‘ox’ and ‘house’ are the literal meanings of the names of the first two letters of the Hebrew alphabet, and by understanding humper as ‘camel’ we get the meaning of the name of the third letter as well. Joyce probably used humper for its half rhyme with semper, Latin for ‘always’. From these we can interpret the original phrase as meaning “as permanent as aleph-bet-gimel”, which is plainly a parody of “as easy as A-B-C”. Indeed, the Latin letters A and B and C share a common origin with the Hebrew letters א and ב and ג, though the resemblance is not obvious.

    Lastly, I inserted “and so on” (as you say, a commonplace) because there are 19 more letters after these three. Joyce did not need these for his purposes, but I wanted to indicate that I was talking about the whole set. You seized upon this phrase and saw “James” as a reference to William James, quite reasonably, and there we are.

    “Script” is a perfectly acceptable synonym for “writing system”

    They aren’t actually synonyms in technical usage. English is written with the 26 letters of the English alphabet, and this is the usual English writing system, though an alternate writing system with the letters of the International Phonetic Alphabet is used for special purposes. German is written using the 26-letter German alphabet (the same as the English alphabet) plus the umlaut diacritic, and this is the usual German writing system. French is much the same, but with five diacritics. The Icelandic alphabet is A Á B D Ð E É F G H I Í J K L M N O Ó P R S T U Ú V X Y Ý Þ Æ Ö; note the absence of C Q W Z, the presence of Þ, and the fact that (unlike in French) É is a separate letter from E rather than a mere variant of it, and likewise for the other vowels.

    It is plain, however, that all these alphabets (and many others) have something in common that Greek and Russian and Chinese do not share, and we call the union of the sets of letters in such alphabets the Latin script. Similarly, there are the Greek script, the Cyrillic script (used for Russian and many other languages), the Han (Chinese) script, the Cherokee script, about 200 in all. Serbian has two writing systems, one using the Latin script, one using the Cyrillic script. Japanese, on the other hand, has only one writing system, but four scripts are used in it: Han, Hiragana, Katakana, and Latin.

    ObIrrelevant: Just how many letters the Latin script has is a question. There are 1505 Unicode characters with LATIN LETTER in their names, but many of these are a base letter with one or two diacritics, and in principle any number of different diacritics can be added to a Latin base letter. Eliminating also the rotated letters, the subscript and superscript letters, the ligatures and digraphs, the orthographic variants, the Greek letters borrowed for phonetic use, and so on, I make it 64 unique letters, though I have probably made some mistakes.

  18. Stu Clayton says

    This is a truly spectacular example of non-communication.

    On the contrary, it’s an example of communication in which each participant is talking about something different – or rather that is how it may be interpreted (communicated) later, if at all. It is the mechanism responsible for long comment threads.

    Perhaps I responded in that way in order to demonstrate how improbable it is that people should understand each other. Perhaps they don’t, most of the time ? How could one tell ? By communicating the question and counting the responses as votes ? But what if the question is misunderstood ?

    And yet people are satisfied for the most part with the give and take of communication. For that to be so, there must be limits on how much boat-rocking is countenanced. The free play of intellect would create chaos. The stop rules are what make communication possible.

    Thank you for your attention.

  19. Unicode character set is pretty impressive, but unfortunately lacks a symbol that I would like to use. O with dieresis and circumflex on top of it. But it has many other fine characters.

  20. Trond Engen says

    Generally: Miners

    I meant to imply that when nüshu was invented by women without access to the prolonged education needed to master a logogrammatic script, and if there are indications that the same could be the case with katakana and hangeul, shouldn’t we expect also the alphabetic writing of the Levant to have been invented by women? But I should probably take that over to the other thread.

  21. Hangul and the Japanese kana scripts were used by women unfamiliar with Chinese logograms, but as far as we know, they were originally invented/developed mainly by men.

  22. Eli Nelson says

    @D.O.: O with dieresis and circumflex on top of it is writable in in Unicode. It uses combining characters: ö̂ (or ô̈ if you want them the other way around). Unicode is not adding any more precomposed characters (single characters representing a combination of a letter and diacritic(s)), because the multiplicative nature of their construction makes it unwieldy to try to include all of them, but in theory precomposed characters are to be treated equivalently to sequences of base characters and combining characters. It’s up to typographers to make sure that both precomposed characters and sequences containing combining characters are actually displayed correctly.

  23. Thank you! Just what I needed. Don’t know how you pulled it, but no problem, already copied it and can paste whenever I need it. And just checked, today MLK’s head plays a rö̂le of the second o in Google. ö̂ (also ô̈). Thanks again.

  24. @John Cowan: Surely Joyce chose semper there for its similarity to simple, not easy.

  25. David Marjanović says

    Don’t know how you pulled it


    Put one of these, or both, after a base letter.

    Hangul and the Japanese kana scripts were used by women unfamiliar with Chinese logograms, but as far as we know, they were originally invented/developed mainly by men.

    Hangul and katakana, yes (katakana is ascribed to Buddhist monks and apparently had a precursor in Korea); but are you sure about hiragana?

  26. No, not at all sure — that’s why I said “as far as we know.”

    But in any case it may not be sensible to talk about the “invention of hiragana.” In its mature form, hiragana combined two principles: (1) simplification of character forms, based on cursive kanji (and sometimes not visually distinguishable from actual cursive kanji), and (2) phonography (which we casually refer to as “man’yōgana” but whose use was not limited to the Man’yōshū). Some of the earliest surviving examples proposed as “hiragana” writing combined these two principles rather unsystematically, so it can be unclear whether the writer was really thinking of the kana as a “different script” from kanji, or just happened to use cursive forms for some of the phonographic elements.

    The later development of kana literature in the Heian period, of course, owed a great deal to women, but even then it wasn’t an exclusively feminine activity. And there were also women poets whose poems were recorded as kanji (i.e. man’yōgana) in the Man’yōshū. There was definitely a stereotype of hiragana as women’s writing and women’s writing as hiragana (based mainly on literary texts, which were only a small fraction of all writing), but the reality was much messier.

  27. David Marjanović says

    Ah, that makes sense.

Speak Your Mind