The part about the "bizarre movements of goats" instantly brings to mind the most well-known piece on epilepsy in the history of Western medicine. The Hippocratic corpus includes an essay called "On the Sacred Disease," in which the author offers a naturalistic explanation of convulsive disease (something like epilepsy) to counter the common perception that it's a "sacred" condition that can by cured by priests and exorcists. One of the author's key pieces of evidence that the convulsions come from the brain, and not from gods or demons, is bizarrely-moving goats; when the goats' heads are dissected, one sees that their brains are rotting and foul-smelling. So apparently it was not only in China that observers made a connection between human and goat convulsions.

i know my impression of the standard language and dialects is shaped by where i grew up, the south, the hotbed of hundreds of dialects. but that is a reality that must be taken into consideration when we talk about chinese 字 and 词 and written and spoken forms.

just a few points.

china unified the writing system when the first emperor of the qin ruled. the language was largely created and expanded and taken care of by scholars. local dialects never had a chance to get written perfectly. the local dialect i speak does have words and pronunciations that i can not find in any standard chinese language dictionary.

words were created during the first period of time of chinese language. i can't define the period, but it must have been there in history. we chinese no longer create words, generally speaking. but we have kept creating 词 for centuries, all on the basis of existing words, so as to cope with life. i suspect these 词 were all created by scholars in educated ways. local dialects never get a big chance to get their words and phrases to get into the standard chinese language.

it is only in the last twenty or thirty years that the so called mandarin has begun dominating in speaking. that is, when television is everywhere, local dialects begin to die. i have been interested in this fact that boys and girls of people who come to cities from remote rural areas, even boys and girls of city residents, are speaking the standard mandarin. i have asked these parents, i know or i chat with. more often than not, parents speak to each other in a dialect, but parents speak mandarin with their children. that is, dialects are dying to some extent. this is the time when mandarin is unifying the spoken words for the first time in chinese history.

one of my points is that chinese language, as far as i know, is a written language. local dialects, that is, the way ordinary people speak, do not affect the language. it is the written language developed and maintained by scholars that gets into the way people speak. whether you are educated can easily be detected. dialects never seriously shaped the mandarin. it is the standard mandarin as used by confucius, li bai, and other essayists and poets, all these centuries that has found its way into dialects. in the dialect i speak, i use the standard pronunciations when i feel it is quite awkward to pronounce them in the dialect. i have heard other people speak that way frequently.

I'd like to know more about the linguistic history of 词…

If I remember correctly, 词/辞 meant "speech, words (言辞)" in Classical Chinese..

Thanks. Funny — I type in Hangeul and use the Hanja substitution key, so I must have typed 증…

[(myl) That separability is no doubt why English (and German, Dutch, French, Italian, etc.) articles are traditionally written as separate words, while e.g. Arabic articles (where the equivalent form would be the-rock the-big) are written together with nouns and adjectives.

But phonologically, English articles form part of adjacent (following) words. Thus the usual pronunciations of "a tack" and "attack" are phonologically and phonetically identical; and similarly "a just settlement" and "adjust settlement". English articles therefore fall into the category of clitics, i.e. formatives that are syntactically independent words but behave phonologically like affixes.]

You know what troubles me about much of what Joe says? It's that he seems to be unaware of the extent to which the Chinese "language" has a history. He speaks with confidence about "Chinese" as if it were one unchanging, monolithic entity across space and time. And so, for example, he cites that Li Bai poem as if the great poet had written in the same language that he knows from the speech he hears around him, that the syntactic units the characters represented in the poem then were the same as the ones he knows now from modern Mandarin. And you know what? It all boils down to the writing system. It's the writing system that has masked all the radical changes and differences in sounds that have taken place in speech in northern China since Li Bai last chanted that poem. When you think about it, it's downright bizarre that people today read the poem in modern Mandarin. What makes a poem a poem, after all, are its sounds just as much as the thoughts it expresses–in most languages, including all varieties of Chinese, poems are chanted, sung–and it's absurd to imagine even for a moment that Li Bai's poem would have sounded like Mandarin!

It's the characters that create the illusion that Li Bai spoke the same language that Joe now does. And many Chinese think that Confucius also spoke that same language. (A Chinese colleague of mine once told me just that, that he and Confucius share the same language, that, unlike Western languages, the Chinese "language" hadn't changed from Confucius's time down to his.)

We know quite well, at least at an intellectual level, that Chinese sounds have in fact changed greatly, and so have its syntactic and lexical units–its words. Those facts are well documented. (Speaking of which, I believe Victor Mair himself has done quite a bit of work on the colloquialisms, the baihua expressions, that were documented in Dunhuang manuscripts.) Of course, the changes in Chinese lexical structure are exactly what Matt points out most cogently in his penultimate note; that note was impressive. (I hope the readers of the string realize how well stated it was.) But the Chinese writing system has done a remarkable job of masking those changes in how people think of Chinese.

There's also one more, somewhat minor point about history that I want to mention. It seems to me somewhat ironic that Joe would use as his illustration of the structure of Chinese vocabulary "社会主义". As most Sinologists know by now, that particular word, like many of the words of modern Chinese life, was actually coined by Meiji Japanese. (After all, the complete works of Marx were translated into Japanese even before they were into English.) And so, it's no accident the structure of such vocabulary has features of Classical Chinese, since that's what those elitist Meiji intellectuals intended it to be.

But in these quibbles, I shouldn't be excessively critical. That's because, even though he pushes it way, way too far, Joe makes an important point. He is not completely off the mark in talking about the squishiness of the "word". That term does not have an easy definition in English, either. Dwight Bolinger used to write about that point a lot, with some really interesting examples. (Why is "the", for example, a word and not a prefix?) We only think the word is a clearly defined linguistic term because we happen to write spaces between the "words". And so, our writing system has affected our perceptions, too, hasn't it? But coming back to Chinese, I certainly don't know how well documented the origin of this particular character usage is, but the Chinese use of 词 that Joe cites is, I'm pretty sure, an adaptation Chinese intelligentsia came up with to deal with the Western concept of "word". Otherwise, a "word" wasn't something they were culturally prepared to deal with in translating English, for example.

癇症 would be 간증. 간질 is 癎疾.

Regarding the word debate, I also don't have time for a comprehensive answer, but I always use 词 (or 字词) for word and 字 (or 文字) for character in Mandarin Chinese. I also don't see "computer virus" as a counterexample or that wordhood is a concept that wouldn't work in Mandarin Chinese.