Skip to main content
Loading Explore guides…
Old Chinese Pronunciation · Acedra
Home
Explore
Languages
Old Chinese Pronunciation
Sign in to save
Chat
If YouTube asks you to sign in, open the video on YouTube.
Watch on YouTube
Reload player
Transcript
32 segments
Copy
Hide video
0:00
Puzzle: The Han dynasty fundamentally changed how Chinese characters were pronounced. Old Chinese seems to have worked very differently, then it entered this era and exited a changed language, with syllable shapes that foreshadow Chinese pronunciation today. How did a character like tuns (稕) turn into something more like tsywinᴴ? Not to mention what the word for character itself went through! This
0:28
dramatic shift has drawn scholars from many fields to try their hand at puzzling out the changes. What happened to Chinese pronunciation during the Han dynasty?
0:42
Imagine we watched a video called "What Old Chinese sounded like, and how we know". It'd engage us with the rhymes of ancient songs, and start to reveal the sounds the famously beautiful characters themselves hold locked within them. Consonant-full syllables with suffixes and prefixes, perhaps multisyllabic. But, surprising to many of us, a complete lack of phonemic tones. Yep, a non-tonal Old Chinese, not yet tonal at all!
1:13
Then we think ahead to "What Middle Chinese sounded like". A video that does exist, unlike the one I just teased. Here we learn of books by poetry fretters who worried rhymes wouldn't rhyme anymore, monosyllables with consonant-vowel-consonant shapes, devoid of prefixes or suffixes. Syllables that sound like they could be the source of Mandarin, Cantonese, Shanghainese. Because they are, in two words, recognizably Chinese. What happened in between to cause this shift?
1:45
What's at the root of this dramatic change? An academic project invited me to take a closer look; understated nod to them at the end, but, as I read papers and watched 65 talks for 470 hours, this not-undisputed sketch of an answer began to come into focus for me: there were once six neat vowels, the Han dynasty happened and syllables came out all messed up. At that, I grabbed my sketchpad and
2:13
imagined a singed sack of syllables that survived the era with only hints of their former sounds. But after reflecting on all these notes, I was compelled to pull back from staring into the cauldron of intense change and to turn to see the faces of the many people who were working on this and the tools they held their hands the whole time. That's when it hit me, what I want to animate today: the diverse people and methods, their faces and voices in my head,
2:43
as I sketch one possible outline of what happened. The people surrounding this story, beholding those singed syllables, aren't just contemporary linguists. This is a particularly long-lived and eclectic academic endeavor. Two historians, one then and one now, step forth to connect language to events. From its rebellious rise in the shadow of China's First Emperor to its fiery end, the drama of
3:15
this period is such a wellspring of culture that its dynastic name still stands for Chinese and the Chinese characters to this day. I read how it's set apart from other ancient imperial projects, yes including whichever one you have in mind right now, in how it implements standardizations, including a standard language and a standard writing system. Characters became fixed, no more fiddling around with their inner components. But they emerge from the
3:44
upheavals of the period with scars that still reveal the intervening linguistic drama. Next, syllable specialists step forward, picking up those scorched syllables as they share a bit of common wisdom: old poems sound only so-so in Mandarin today, but over in Shanghainese initial consonants are better preserved, and further south in Cantonese the ends rhyme better. These just represent three outcomes (out of many!) of
4:16
the same Chinese syllable structure. Of the later singed syllables, not the earlier pristine ones. From at least the night of Lù Fǎyán's dinner party turned rhyme debate a milennium and a half ago, Chinese scholars developed a specialty recipe, their own signature anaylsis of what a syllable even is. You have the front, an initial consonant. And the back, the part that rhymes. Inside the rhyme there's a vowel ("Vowel"). And, after the vowel, an ending consonant ("Coda"). Between
4:48
the initial and the vowel may fall a little glide w or y, the medial. Finally, to the delight of anyone who's learned any Chinese, a syllable has a tōne. Tóne? Tòne! Tǒne. The great changes of the period forged syllables that ended up quite different from how they started. It might've gone something like this. Take our first step back. A, ehr, tonogeneticist? Tone origins expert assures us tone arguably develops last. (Eh, sort of, b-b-it's messy.
5:25
Because, you see, tones initially arose from the loss of specific endings...) Of course, thank you, do come back to teach us more some time. I really would like to learn. Briefly, we start to lose tone as we back up into the Han and head toward the ending sounds that tones came from. Our next step back comes from generations of rhyme analyzers who got lost in books and tables charting each piece of Middle Chinese syllables. After divisions over divisions, many say,
5:55
for our purposes today, there are actually two fundamentally different types of syllables: Type A versus Type B. The difference? Type B syllables have a high /i/ or palatal /j/ medial after the initial. Type A do not. So, in Middle Chinese that might've sounded something like kwan versus kjwən, pen versus pj(i)en. Keep them apart as we continue backwards, because A versus B will have gone through their own separate changes to end up this way.
6:29
So, we ditched tone, divided A from B. Now how do we crack open syllables and watch their internals changing? Well, such a pivotal period was no literary question mark with writings before and after and a mystery during. Enter paleographers, who put hands and eyes on tangible linguistic evidence. They highlight an underappreciated point, that we're often reading standardized texts in reprinted or now digital copies, while they haul out original manuscripts and point
7:03
to the actual ink soaked into the surface of bamboo slips and, by the end of the Han, paper. Some fans of digging directly into old texts specialize in ones that use Chinese characters to transcribe non-Chinese words. Buddhist manuscripts spell Indic words, like dharma (dharma/dhamma) was spelled dəm-mɑ (曇摩). They also shed light on who was borrowing from whom exactly. Words weren't going directly from Sanskrit to Chinese but passing through vernaculars like Gāndhārī and then into
7:40
Late Han Chinese. So Sanskrit samādhi reached Chinese as sɑm-məs (三昧) through the Gāndhārī form samasi. Manuscripts also reveal shifting rhymes. This one connects back to my dictionaries video. Recall how the Instant Primer used one rhyme every chapter as a catchy way to teach characters. An excavated Village Teacher's Edition showed that, at least in this community, final -r was turning into a final -j. And broadly, excavated texts suggest that even simple /a/-like
8:18
syllabes were going through complex waves of changes throughout the early Han dynasty. So, graph theory fans think to themselves, why not take these texts and build networks of what does and doesn't rhyme in them? Maybe we'll catch syllable internals shifting at scale. Do that and, indeed, you prove that rhyme structure is not the same as later in Middle Chinese, nor as earlier in Old Chinese. A quote from one talk about that sums up this whole video:
8:48
old manuscripts prove the Han era is "actually a crux of change between" the two stages. The expansive imperialism of the dynasty brought other languages into contact with Chinese as it underwent these changes, contact that draws in academics of diverse languages studying imprints of words borrowed from and into Chinese. This contact may surprise some. As an outsider to the field, it did me. Tocharian, the northeasternmost Indo-European branch,
9:19
was spoken and written for centuries in lands that lay just beyond the Jade Gate. The Tocharian word for asafoetida... dare I try and pronounce it? I had enough trouble with asafoetida! Aṅkwaṣ. Aṅkwaṣ made its way into Chinese as something like ʔa (阿) plus ŋweis (魏), ʔaŋweis. So it joins sɑmməs in suggesting a timeframe when ending -s was still holding on, at least in cases where it came from a final -/ts/ or -/js/.
9:55
Another cool borrowing. The Han invention of paper made its way south, and along with it this paper character (紙). It reached Mǐn, which is a name that gathers Chinese varieties way down south that have long been distinct, and predate Middle Chinese. Mǐn words for paper start with a ch or ts. And in Vietnamese it can be a word I don't quite know how to say but it starts, it can start with a palatal j-, uhm, /jấj/? Paper also made its way down to the related Rục language where,
10:31
intriguingly, it has two syllables. Another word I don't quite know how to say but the first syllable starts with k- and the second with -ts-, so it's something like /kət͡saj³/. That form is interesting. It points to a more complex Han era initial: a palatalized consonant with a k- before it. This pairs pretty well with manuscript evidence that Han words still maintained p- prefixes. (And if we took those two prefixes back any further,
11:04
we'd summon scholars of Trans-Himalayan because, wow, comparing Chinese to its prefix-full relatives has been an absolute pain.) But so, anyways, for today, here we are. Or had been. No tones. This type A versus B stuff. And now evidence of changing vowels and consonants. Historical linguists built a bunch of theoretical machinery to peel back these layers of changes. Given my channel, they could busy us for the bulk of this video, but, on balance, I want to trace the
11:37
unwinding of just one type A and one type B example that ended up roughly like kwən and tsywen. Middle Chinese consonant bookends could be filled in with many vowelly insides. In the case of kwən (昆), that syllable's start, k plus a medial w, could've arisen from three sources: a simple initial k and a medial u or w in the rhyme, a rounded kʷ in the initial, and both options back to back, at least in theory. I chose an easy one, k- plus -wən. We only need to
12:17
rewind one step from there to get back to the start: this wə on the inside broke up from an earlier -u-, a type of vowel breaking, so put it back together and it was once kun. And still Type A, so I'll put a little A here for a note: ᴬkun. Ok. Now for a Type B: tsywen (嫥). From the outset it's a snaggle, since we'll have to deal with a palatalized initial and a changing rhyme. First, drag the -wen back through vowel fronting to an
12:50
earlier -wan. That wan-wenning was itself fed by a prior change, more vowel breaking. Unwind that to when -wan was once a simple, unbroken -on. As for the initial, tsy rewinds to a palatalized t. (Maybe tjon?) /j/ means, yes, it's Type B, so really I'll write ᴮton with a little B here. Hhh, I promise, I'll return to which sounds such As and Bs might've originally stood for when we approach the end. Finally, an extra piece, not enough to satisfy completionists but necessary to
13:28
grasp the work involved in running this hindcasting machine. We can bring what we learned from undoing Type B changes back over to the Type A syllable tsen, speculating that it fits well if tsen came from three different sources: from ts-in, from ts-en and from ts-ən. You can actually work backwards like this using tables of initials and rhymes for each syllable ending. I'm told it's easiest with -ng like kaeng (羹), a bit harder with final -p like
14:00
kaep (甲), and tougher still with like final -n like kaen (菅). (Those are some attempts at Middle Chinese pronunciation there, heh.) If interested, you can follow the A-and-B back-and-forth machinery, and meet about a dozen sound changes as you do, in lectures I link in my sources. This puts us standing at the doorstep of our period, where the messy changes all rewind back into one single six-vowel theory. To tie this up into a tidy story once more,
14:31
syllables entered the Han era with six neat vowels, these Type A and Type B shifts altered those vowels and in turn the consonants until, as we were warned at the outset, those syllables came out messy, well on their way to Middle Chinese. [That's] too tidy, though. I can reel in one last set of experts to send us off by asking, what really were these A and B syllables we so insisted on keeping apart? Do they go back to
15:00
pharyngeal consonant A versus plain consonant B? Did it involve vowel warping, with A vowels going low while B went high, leaving that kind of high-vowel effect that sometimes results in palatalization? Did that happen as in Maltese, where pharyngealization and warping interacted? Well, in that plotline, an emphatic A syllable like /tʼˤuns/... that's too ejective, more pharyngeal, tˤuns, tˤuns! (扽) broke up and warped down to ᴬtˤwəns, while
15:37
plain B tuns (稕) broke and bent up to ᴮtjwins. Then, after twəns lost its emphasis, the big remaining difference between the twowas the higher j-like sound of tsywins: twəns, tsjwins. And now this ending -s. It wasn't holding on this late, at least not in an -s shape. We know it become a tone, so we get twənᴴ, tsywinᴴ. Step back any earlier, and we're in times when single characters could represent gasp-inducing Old Chinese reconstructions. But the story of those
16:18
wordbeasts is a question about what Old Chinese sounded like and how we know. (This video is, well hasn't yet been but may end up being funded, academic paperwork permitting. These are understated supporters with nothing to plug, so let me just thank everyone involved who put so much work into their piece of the project so I could learn from your findings, expand my sources, and share this today.)
16:43
Speaking of support, thanks to every patron of these last three animations about Chinese linguistics and to all of you for joining me today and throughout. A new year is about to bring new topics, so stick around and subscribe for more about our world full of languages and cultures.
Here to help you learn
Sign in to keep a private conversation with your course tutor.
Ask me anything about the material…
Sign in to chat