Old Chinese 6: Chinese Historical Phonology | Guillaume Jacques & Nathan Hill | Leiden 2026

Yeah. So, actually now you can look at the whiteboard. And on it already is well, so because we thought the course was not sufficiently confusing. Uh We decided to switch the way we're writing Middle Chinese halfway through. Yeah. So, so Baxter's way So, so the good news is actually, you know, Guillaume and I do not have our own way of thinking about Middle Chinese. We just slavishly use Baxter's romanization. And Baxter's system is not meant to be He's very He's very sensitive about this. It's not a reconstruction. It's just a romanization. Yeah. So, it captures all of the distinctions that exist in the primary sources trying to be somewhat agnostic about how they're pronounced. So, I think Guillaume and I both think his system is ugly. And I I just mean like ugly to look at. You know, like t s y is ugly, right? And it was mostly motivated by not wanting to use diacritics back in 1992. Yeah. But now it's 2026 and it's What's the problem using diacritics? So, Guillaume has decided that he's a linguist, so he should use the IPA. Yeah. So, that's why on the left you get these equivalences where Guillaume is using IPA. And then the meaning of the the symbols is clearer to you, right? At least you know, and how committed Well, anyhow, now I switch to my own view, which is this the disadvantage of IPA is that it loses some of Baxter's agnosticism, right? Because it it gives a phonetic interpretation to what Baxter meant as a categorical symbol. Yeah. So, so I've decided to use an Indological transcription because also I'm not a linguist, I'm a Indologist. So, so I look at what the stuff on the left, I say what are these crazy symbols whereas I feel comfortable when I look at the stuff on the right. And then also I'll say Tibetan, Burmese, Newar, Metalone, Lepcha, all of the languages of the family that have old written traditions use an Indic script. So, it's incredibly straightforward to write them using Indic Indic transcription. And even actually Chinese of course doesn't use an Indic script, but the Chinese philological tradition of analyzing Chinese pronunciation is heavily influenced by Indic phonological theory. So, I think you know, we should uh have what one ring to rule them all and it should be Indological transcription. So, that's why I'm using this Indological transcription. And then the other thing that I think is a little bit nice about my way of doing it is for Baxter the the the the for Baxter division three you can tell two ways with an I or with a Y, yeah? Uh so, for instance um I don't know. Um if you have like t r y no. No, that one. Never mind. I don't I'm going to do something wrong if I do anything, so uh Uh but anyhow, certainly something like t s a n g you it's supposed to be division three, yeah? And it's being marked with a a Y. Oh, I know. I'll do one which is like um I don't know. Can I think we can have something like pyang in Baxter? So, this is division three and this is division three. So, he has two different ways of writing division three. I think that's a little bit not very elegant. So, I always write division three with an I, yeah? So, that's why I've added the I here and added the I here. Yeah, but otherwise it's, you know, is extremely easy to uh you know, to to switch back and forth between these three systems cuz they're they're exactly the same categorically speaking. >> What's the difference between palatal and central? >> Uh so, uh uh like uh like uh this is what Guillaume is asking, I think, is like uh uh Baxter will have something like this >> Yeah. >> versus this, yeah? >> Yeah. >> Where uh so, I write I write this one like this. You think, yeah? And I write this one like this. >> Oh, okay. >> Yeah. Uh and and this is just uh a mechanical thing. Although, I kind of like how it looks a bit, yeah? Okay. Uh Uh I've written an article actually in 2017 maybe called like an Indological transcription of Middle Chinese, where I Pardon? >> I I said was it that one? I thought it was more recent. >> No, it's it might be more recent, I don't know. My my whole life is a blur. That's fine. Uh But, uh yeah, so and then also I have there's a there's a spreadsheet that Baxter made where he has 9,000 characters in his transcription system and I have added my transcription system to it. So, if you want to use my transcription system and are too lazy to uh do the math in your head or something, I can just give you that uh spreadsheet, although not now because I only have my phone with me, yeah? Um okay. So, now I'm going to start talking about methodology. So, I think it's useful to distinguish two kinds of historical phonology. One we can call comparative phonology, which is not what we're doing, yeah? Uh and uh in comparative phonology we reconstruct unattested stages of languages using the comparative method. And then maybe we can call it the philologico-historical phonology, uh where we're philologically interpreting uh well, sorry, we're phonologically interpreting uh ancient documents. More precisely, uh the postulation of a phonological system that in principle is associated with the speech community that produced or altered a particular document at a particular time, together with an interpretation of how the document instantiates that system. So, I hope that's, you know, reasonably clear. It's that you need to you know, you you need to understand uh the system and its instantiation, right? And it's actually kind of I don't know, like that's how you deal with textual problems, for instance, like you know, for instance, uh Guillaume mentioned, you know, a mistake in the Yunjing, right? But you can you how how can you correct the primary source that's telling you how things are in the first place, right? It's it sort of sounds risky. Well, that's the difference between coming, you know, figuring out the system and then looking at its instantiation, cuz then you can say, according to the system that this document has created, this character can't go here, so I can correct it, yeah? Um and then I would actually say you would also need a kind of a little bit of a theory of how the system uh like like how is it that that mistake happened, you know, historically. Okay. So, uh the tools of comparative phonology, exceptionalist sound change and four-part analogy, And actually, uh I don't know. One thing I find tricky about teaching here in Leiden is some people know historical linguistics very well, but some people don't, yeah. And uh and everyone knows different languages. So, maybe worth going through. So, exceptional sound change. We notice a systematic correspondence like P in Latin, P in Sanskrit, F in English and F in German. Uh and then we say, "Well, maybe the P is kept in Latin and Sanskrit and changes to F in English and German." Uh and we use such patterns to reconstruct the the proto-language, right? Uh and and the point is you can't uh you can't say except in this word, yeah. Except in the word I don't know. Bad or something, yeah. Uh Okay. And then analogy. We we have forgotten I actually think of it as the way I like to think of it is exceptional sound change is where you remember something correctly, yeah? And then basically the sounds your parents make and that you learn from them have like the token type relationship gradually drifts over time, yeah? So, like So, that's why it's exceptionless is that is it like I'm learning the correct category, but the the the precise phonetic instantiation of that category is changing over time because there's because I don't have access to the tokens that my great-grandfather said. I only have access to the tokens that my parents and maybe grandparents had, yeah? Whereas analogy is what happens when you forget things. You say, "I want to use the past tense of the verb help and I just can't remember what it is right now." So, >> [laughter] >> then I say, "Well, the past tense of talk is talked. So, maybe the past tense of helped is helped, yeah?" And sure enough, then that's how I get helped. Uh whereas in fact it used to be holpen, right? Um as a past participle, yeah? Uh so so so you know, when you forget something, you supply the answer from the system synchronically. Whereas if you remember something, you just remember it and and it has to very slowly change over time. Okay. And then we have strive is to strove as dive is to dove, yeah? And that replaced dived in American English, at least. Uh and it shows you analogy can go anyway, yeah? So, whereas whereas uh sound change is exceptionless and directed, yeah? Analogy is is anything goes, yeah? Uh you there will be a tendency to make irregular things regular for obvious cognitive reasons, but it can also make regular things irregular. >> Well, I think in cases where it makes regular things irregular, it's just that you have multiple like paradigmatic systems at the same time. And they both get continued by different uh speakers. >> Yeah, exactly. Yeah, like which is just to say you can form a four-part analogy. Right, I could say I could have said talk is to talked as dive is to and then I would have gotten dived, yeah? Uh but one thing is actually most of the time in a language remembering it and making it up will result in the same thing, yeah? Because that's what happens when when something is regular. Yeah? Um Anyhow, okay. Uh now, what are the tools of So, those are the tools of comparative of comparative phonology are are exceptional sound change and analogy, right? Uh and what are the tools of phi- philologico-historical phonology? I call them imposition and transposition. So, imposition is when you're is a map from from X onto L where L is the language in question and X is something else, yeah? And then transposition is a map inside of our language, yeah? Okay. So, let's look at Egyptian to illustrate what I'm talking about. So, we put proper names in cartouches in in ancient Egyptian and that allows us to associate a Greek personal name with an Egyptian equivalent, yeah? And the Rosetta Stone is famous as the document that, you know, helped us decipher ancient Egyptian, but in fact it was the Philae Obelisk that was decisive. So, we couldn't get we couldn't get there all the way with the Rosetta Stone. We needed the Philae Obelisk. Okay, so here is Ptolemy on the Rosetta Stone. So, and I've I've transcribed it, yeah? So, you can see Ptolemaios, Ptolemaios, yeah? Okay. And then here is Cleopatra, which is which is also on both stones, yeah? So, this one's Cleopatra, right? Yeah. So, um So, Ptolemy occurs on both objects and is spelled the same way. Four signs are common to both names, L E O and P, yeah? So, those ones we we can we can get, basically, right? You know, um and then both instances of A in Cleopatra appear as the eagle, yeah? Um but the T in Ptolemy and the T in Cleopatra are written with different signs, okay? So, that's what I want you to just see. Oops. See the T in Ptolemy and the T in Cleopatra, it's a hand here and it's a it's a kind of I don't know, half here, yeah? Okay. So, Champollion ignored uh this problem, and he assumed that both had T, yeah, cuz he said, "Well, it's a In Greek, they're both T, so forget about it. Both the hand and the half-disk are a T, yeah?" Okay. So, then we find a new Now, we have a theory, yeah, that we've developed a hypothesis about Well, a lion is L, and eagle is ah, and hand is T, and whatnot. Then, he say sees this name, and he says, "This is Al something set something-ter something, yeah?" So, probably Alexander, yeah? So, uh so, we can read these six signs, uh and then he guessed these ones, yeah? Uh so, uh so, only N is new to the hypothetical phonology, right? Cuz we had the sounds K and S already, uh but now we have new transitive relations. So, the transit this So, this is So, let me say the imposition is using the Greek to figure out the the Egyptian, right? And then the transitivity is the the transposition, which is saying uh foot equals whatever bowl, and crook equals whatever this like uh I think it's a kind of knot tying a door shut, okay? Um so, some of these are turn up to to be wrong, right? So, So, one thing that that is I think important in in my philosophy of science is, don't be afraid of making mistakes, because if Champollion had been afraid of making mistakes, he would not have deciphered Egyptian, right? Um so, uh just this just is talking through my um you know, uh kind of formalism. So, we have uh as an imposer, we're mapping from Greek into Egyptian. And we say that this function takes T and turns it into half a circle, yeah. And it takes P and it turns it into a box, etc., yeah. So, that's a function from Greek into Egyptian, yeah. Uh Uh okay. And then here the transposers, they move within Egyptian. So, they take they take a it's a map from a half circle to a hand and from a foot to a bowl and from a crook to a tight thing, yeah. And this is how I think we do historical sort of philological historical phonology. You have to have some external source of phonetic information to allow you to interpret the script. And then you have to have relationships inside the script that give you sort of equivalence classes. So, you might say for instance, in previous class we use Sino-Vietnamese as a transposer as an imposer to give us the in certain words the S final. And then we use the fact that we have a category inside Chinese the chusheng as the transposer to then assign that S value to other words that we didn't necessarily have in Sino-Vietnamese, right? Okay. And then and I just have to polemicize a little bit, of course, yeah. So, what about the hypothetico-deductive method of Karl Popper? He says you propose a hypothesis and how the hypothesis arrived at is only a question for psychologists, yeah. That's something that he says. You can take drugs, you can ask a guy on the street, you whatever, you just come up with your hypothesis. Then you attempt to falsify the hypothesis. And Baxter and Sagart claim to subscribe to this method. Now, I think this method never produces anything. So, I think they have in fact produced some good results. So, that shows that they have not used this method, yeah? Um But, let's just notice that Champollion's work on Egyptian is exactly the opposite. So, I call it the inducto-dogmatic method, yeah? So, what did he do? He noticed a pattern in the data. So, he didn't take drugs or ask a man on the street. He did what we might call uh in in you know, inductive empirical uh generalization. So, he knows the pattern in the data. He formulated the pattern as a hypothesis, right? So, he So, he moved from saying whatever, "An eagle is the letter a in this word." to "The eagle represents the letter a in general." That's moving from an observation to uh hypothesis. And then he stuck to his hypothesis, no matter what happens, yeah? So, uh So, he finds out there are two ways of writing t where prima facie then, you know, there's a problem. Like, right? Why would the Why would Egyptian have two ways to write t? Yeah, or two ways to write s, yeah? So, he said, "Uh whatever. I still think it means t even though I have evidence against that, yeah?" Uh and uh and I would say such a hypothesis, like the hypothesis that, you know, uh what I forget what it was the foot means t and the and the half circle means t uh is never abandoned. Uh it just gets you you might say refined or sublimated. So, actually, uh one of these is written t in ancient Egyptian now and one is written t with the with the line under it, yeah? So, So, they're They are both t even though they're not the same, yeah? Uh and so, uh yeah, anyhow, that's how I think science works is with through through this inducto-dogmatic method. Um and I would say that uh exceptionalness in in sound change an application of the inductive dogmatic method. There it is impossible to prove to me that any sound change is not regular because I can always find a conditioning environment or invent another protoform phoneme or say it's a borrowing or whatnot. And that is not a problem. I actually think that is why it is successful as a research program. Yeah. Okay. So how about in Chinese? Let's talk about Chinese. What are our imposers and our transposers? So here are some imposers. We can have loan words like like the words chariot and the words tiger. I think are quite early loan words. We have transcriptions of Indic terms. So yeah. So so these are both these are ways of assigning a phonetic value to a Chinese character, right? If you know that like now the thing is you might doubt some of these equations, but if you know that comes from some Indo-Aryan word that means chariot something like chakra, then you will want it to have a pronunciation something like that, right? Okay. And then articulatory articulatory phonetic information in the rhyme tables. Yeah. So the rhyme tables say you know, voiceless bilabial plosive or something. And then you think, well, if the rhyme table says it's a voiceless bilabial plosive, maybe it's a voiceless bilabial plosive, right? And that was what we were mostly using as our imposers in the last week. Okay. And how about some transposers? Okay. Well, there's phonetic series. And here you see I've written the let's call it the mother of the phonetic series in a large font. And then we see we have nak nak nak nak nik >> Yeah. Well, I don't know. They They must have something in common. Uh and they and they do. Yeah. >> [clears throat] >> Uh by virtue of having been written with the same phonetic. And then we have paronomastic glosses. So, these are These are actual glosses in Chinese texts, right? Where this one says Du Zichun reads bo as po. Yeah. So, uh you know, I mean, for whatever it's worth, it is a way of identifying the pronunciation of one Chinese character with another Chinese character. So, it's a transposer. We're going from one Chinese character to another like we were, you know, going from one Egyptian sign to another, right? Uh and then we have fanqie uh speller chains. Uh so, you know, it's um let me just explain how how to read this one, yeah? Uh do I have color in this, by the way? >> Uh we see color. >> Yeah, okay. The color is meant to help you sort of understand how the the fanqie chain works. So, we look up lu, and then it says it says uh Uh you you read the Chinese. >> Long qie. Uh but it's a bit strange. This reading it doesn't correspond to lu. Because uh the u the rhyme is not the same, right? >> Yeah. >> So, this >> Yeah, uh >> Is it You sure >> Oh, you're right. Maybe I made a mistake. >> Yeah. It's we It would be the mo qie, maybe. Uh I guess the the one of the you which is in yellow, uh next to the the yellow one, I guess. Yeah. Oh gosh. >> Oh, it must have been luo that I meant to have instead of lu. Yeah. Yeah, yeah, yeah, yeah, yeah. Cuz that's the only way that makes sense. >> yeah. >> So, read it as uh it's long lock chair, yeah. >> Yeah. >> Or no, sorry. So, long is long lock chair, and then long is Anyhow, whatever. You know what fronted chains are. Uh Yeah, okay. And that's my whole methodological preface. So, as as you saw, it didn't actually take that long. And if you want to hear, you know, more about methodology in Chinese historical phonology, you could my inaugural lecture on my YouTube channel. Yeah. Um but now I will ask my attractive assistant to uh >> [laughter] >> to to move on to the next uh >> Oh, okay. Yeah. >> presentation. Yeah. >> let's do that. >> Although if you, you know, you could ask questions if you want. >> But like like like like like when you said that um when you were using loan words, for instance, like by your own schema, isn't that by default application of the comparative method? >> No. Comparative method compares related languages in order to come up with uh sound changes to reconstruct their common ancestor. That's not what I'm doing, right? I'm I'm >> Oh, okay. I see. >> Yes. >> Yeah, but it's the second >> Okay. I do feel like that went a little fast. So >> Yeah. >> So, I don't know. Does anyone have any questions about Egyptian? >> [laughter] >> Well, no, but but like like like like just [clears throat] to clarify, I think part of the confusion like like like like in me, at least, I don't I can't speak for other people, is is is like when we colloquially talk about compare. That that means one thing. But then when you say the comparative method, you're strictly talking about using two languages to deduct their common ancestor. That and only that is what you call an application of the comparative method. >> Yeah, so it's exactly why I distinguish that that that when we say we do comparative linguistics, we should say that means two different things. One is um or no, historical When we say we do historical linguistics, what we're doing two different things. One is comparative linguistics, which you could do without any old documents, right? You can do it with just uh fieldwork on modern languages, yeah? Uh where you find cognates and propose sound changes and reconstruct ancestors. That's one kind of game we play, yeah? In which our tools I really think of this as like uh uh I don't know uh like You know, when you do geometry, you're only allowed two tools, right? A compass and a straight edge. >> Yeah. >> And someone might come along and say, "Actually, maybe you should try using a a a marked straight edge. You'd find it much easier to do your geometry with a marked straight edge." Then you say, "Yes, but then I wouldn't be doing geometry." >> Yeah. >> Okay. So, [clears throat] >> [laughter] >> in the same way I say comparative linguistics is what we can do with uh analogy and uh I should have done them in the other order. Exceptional sound laws and analogy, right? So, then someone comes along and says, "Well, you should really look at frequency or you should uh look at lexical diffusion or what not." And you can say, "You can do what you want, but uh I do something called comparative linguistics where I have two tools, >> Yeah. >> and that's those are the rules of the game, yeah? So, uh in the same way I would say uh for let's let's call it philological historical linguistics, the what you're trying to do is assign uh pronunciation to written documents, yeah? And now >> Oh, I see. >> if if those written documents are recent fieldwork using the IPA, then it's not a very difficult task, yeah? But for uh Sumerian or Chinese, it's a fiendishly difficult task, right? Uh and then like intermediate things are things like Semitic languages where they don't mark the vowels. Uh you know, like And so so that's the let's say it when we reconstruct Old Chinese, all we're doing like ultimately as a task, as a goal, is we're assigning pronunciations to ancient documents. And I think it's easy to lose sight of that. Uh like over the last week, yeah. Uh that's the you know, like they it's it's important to say like the reconstruction of Old Chinese based on Chinese dialects, that would be I should say based on Sinitic languages, that would be uh >> Proto-Sinitic. >> Proto yeah, that would be Proto-Sinitic. Uh and that would be uh something we did as comparative linguists, yeah. But that's not what we're doing in this class, yeah. So uh yeah. >> But wouldn't you agree that uh studying, for instance, the layers of loan words even between unrelated languages is still comparative linguistics? >> Yeah, that that was my original question. >> Like like for instance, I think I understand you you work on Proto-Germanic and you look at the borrowings into Proto-Finnish. >> Yes. Yes, but but But that's but But why is it comparative linguistics? It's because the two tools you're allowing yourself are exceptional sound change and analogy, right? So like you can do let's call it comparative linguistics on Basque where you're looking at loan words, yeah. But the the the function of the loan words is totally different, right? Like when when we're looking at loan words, we're trying to figure out how Chinese characters were pronounced. >> Huh. >> Yeah. Whereas when the Bascologists are looking at loan words, they are trying to figure out what the what sound changes affected Basque in its history, yeah. Okay, so I mean I I don't know. I'm being, let's say, overly stylized and doctrinaire because I think that leads to clarity, right? Yeah. Uh cuz in effect, you know, we are just looking for exceptional sound changes uh in order to study the history of Chinese, but for the purpose of deciding how Chinese was pronounced at a given era, right? Like in order to interpret how this how the script is being used, yeah? That's what I would say. Okay. So, now on to uh phonetic information contained in the Chinese script. So, our goal, yeah. So, this is I should introduce. This is a This is a term of uh of uh uh Nishida Tatsuo. Uh and he has this, I think, very useful idea of the sonus grammatici, which means, like, let's say I write this word in English. Okay, what is its sonus grammatici? What this means is, what is the pronunciation implied by the way of writing it itself, yeah? And I think you will agree it's something like, you know, and I don't know. I would say something like "enough". Yeah? Right? Yeah. Okay. And And let's say, like, you know, let's give a nice Greek example. So, the sheep in in Aeschylus, they say they say this, yeah? Which is, of course, in modern Greek Greek pronounced "v" "v", okay? So, do the sheep go "v" "v"? No, the sheep go "ba" "ba", right? >> [clears throat and laughter] >> And And the the the way you get "ba" "ba" is by, you know, just saying, "What's the pronunciation of B and what's the pronunciation of edit, right? And then you and then you, you know, so this is like you you there you still need to do work, right? Like what why do I think G H would be a a gamma or something like that, right? You have to do a certain amount of work, but the point is that the script itself implies something about the pronunciation. And uh let me give an example that, you know, that like like this one, right? Is probably it was pronounced like knicht or something, right? >> Yeah. >> Uh and you can have that be your, let's say, default hypothesis, your zero your null hypothesis, is that what we say in in science, before doing any philology, right? And I think that's it's very useful conceptually to say like, what is the sonus grama, yeah? Okay, so now we're going to try and figure out the sonus gramae of Chinese. Okay. So, we're going to use Middle Chinese as our imposer. And the phonetic components of the characters, the Shie Sheng series, as our transposers, yeah? So, let's let's give it a try. Yeah, okay. So, we look at uh the character ma and we see that it has a semantic on the left uh nü, and it has a a phonetic on the right, ma. Okay, so uh So, just for this one slide, I've used pinyin, yeah? But now on, I told you I was going to use as my imposer Middle Chinese, so this is the last pinyin you see. Okay, so So, let's look at this Shie Sheng series. Well, there's the head of the Shie Sheng series. Ma and then ma uh ma and then oh, ma and ma. Okay. So, what have we learned from this shieshang series about the you know, what pronunciation is implied by the script itself? Well, I will tell you so you don't have to tell you don't have to actually, you know, answer the my rhetorical question. It doesn't tell us anything about tone, right? Because cuz we you see the tone, we have some we have some shuangshang and we have some chushang. So, the the the phonetic component of the character itself is not telling us anything about tone in this series, yeah? At least, yeah? Uh but maybe it tells us about the rhyme and the initial, yeah? That's what I would say. Cuz so far, everything is uh it has an m initial and a vowel and a rhyme ang, right? So, now my my theory, this is, you know, using this methodology. So, I make an observation, I change it into a hypothesis. So, my observation is these three characters have the same initial and the same rhyme, yeah? So, my hypothesis is all uh what do you call it? The phonetics in uh Chinese characters indicate that the the readings with those phonetics have the same initial and the same rhyme. And uh let's look at this one. Okay, so, bian pie Oh, so far so good. Pie ah, DAMN. YEAH, OKAY. OKAY. And then, pie. Okay, so, uh and then, pa and then, pa pa. Okay. So, uh uh-oh. What do we do, yeah? Well, I don't know. Any any thoughts? What What So, so so it didn't work like like it did with ma, yeah? So, what do all of these readings have in common that is being indicated by the by the phonetic, yeah? And now I am am going to force you to engage with my rhetorical question, yeah? So, anybody? >> Uh somebody says that they're all similar, hopefully. >> Yeah, they are all similar, hopefully. Can we can we be more explicit about what similarity we observe empirically? And someone down there very quietly said, "Same place of articulation." >> Of the initials. >> Of the initial, yeah. So, does everyone does everyone feel like, "Yeah, that's that sounds good, yeah?" Okay. Now, do we notice anything about the rhymes? >> They're all just a single well, or they have no codas. Let's say that, man. >> Yeah, that that's true. They're all open syllables. >> Like this one. >> Yeah, they they're all open syllables. That's good. Uh how many different rhymes do we get? >> Somebody online answers that the there's two different rhymes. There's ya and a. >> Yes, that's correct. And do you notice anything about those two different rhymes? Well, maybe I need to we need to look at more series first. Um but anyhow, um so we weakened our our hypothesis uh to say like maybe they mean hope they have homorganic initials and similar rhymes where where we're not quite sure what similar means yet, yeah? Okay. Uh so yeah, so we have three initials b p and p and two rhymes a and ya. Okay. Now, let's look at a poem from the Shijing. This is ode 18, stanza one. On the lamb furs, five many tresses of white silk. And look, we have a a character from our series b ye that is rhyming with a character not from our series, but that has the rhyme from our series. da Yeah. So, uh at this point can we Well, and then let's look at another one. Here we have Pierre, also a series from our Sorry, a character from our series. Uh and it's rhyming with ha. Oh, I should have put that one up here. Uh I have I write uh I write this where Guillaume writes >> an X >> an X. No, is that right? >> No, no, no. >> No, gamma. >> gamma >> Yeah. >> Also, this is the gamma. >> This is shamu. >> Okay. >> This is like actually how it is. >> Oh, okay. Okay. >> And then she And then Baxter What does he put What does Baxter do for X? >> Uh he does X and for gamma, H. >> Yeah. So, this one's confusing. Yeah. So, Baxter This is Baxter. Guillaume >> I don't like that he uses a capital X and then he uses his X for very different things. >> Yeah. Well, yeah, there you go. That's why you should use an H. And uh I noticed that I use Well, if you're in the my Tibetan class, I use H with this symbol under it for exactly the same thing as in Tibetan. It's also gamma in Tibetan. Uh so, yeah, so you see I'm trying to come up with a romanization that sort of works for Sino-Tibetan linguistics in general. So, because I find it very annoying in my 2019 book, like the same symbol in it occurs in Tibetan, Burmese, and Chinese means different things in each one. I think that's that puts a lot of cognitive pressure on the reader, yeah? Anyhow. So, uh So, let's see. So, yet and ah rhymed in the Odes, yeah? So, uh Okay. And uh yeah, and then the other thing that I'll point out is uh this is maybe being a bit uh a bit uh a bit too strong, but uh uh the Odes ignore tone, yeah? So, here's for instance, a stanza where the chushang is rhyming with the open syllable, yeah? Uh and here's uh another one where the chushang is rhyming with a final K. Uh now, what's going on there? Well, in in any event, it's they're both examples of the chushang not rhyming with the chushang, right? So, it seems that the uh the the ancients uh did not rhyme paying attention to the tone, yeah? So, uh yeah. So, it it Duan Yucai, who it's very irritating we have to talk about him at all, but uh uh >> [laughter] >> Yeah. Yeah, okay. But, he says there's no departing tone in Old Chinese, yeah? Yeah. Uh Now, okay. Having shown all this, let's go back to here and say, "Anyone have a new hypothesis for how we should articulate the similarity in the two rhymes in in here, in this series, yeah?" Uh I I actually am re- wearing my reading glasses, so you're all just a total blur, but uh I think I see >> outlines or that's some possibility like uh Well, I mean I think all you said is there are given consonants and there are given vowels. So cuz like all that would be a morphological hypothesis. Uh so so I'm going to see does anyone else have a suggestion? We're going to say how can I characterize the way that a and yeah are similar? in a uh phonetic series >> They are similar enough that they can rhyme with each other which means that one of the two probably sounds sound different something like that. >> Well, I I think you've gone further than you need to there they are similar enough that they can rhyme together. Yeah? Okay. So uh So So let's go with that. Yeah? And this is what we call the Shixiong hypothesis. Oh and it's also according to Wang Li characters with the same phonetic component write words that would have rhymed in the Shijing. I say would have rhymed because not all characters are used in rhyme position in the Shijing. Right? So you can't actually prove it for for all characters. You can only prove it for some characters which is also you know again we look at we observe an empirical uh you know sorry yeah we observe something empirical like oh look bie rhymes with da and pie rhymes with ha. And then we hypothesize yeah and a always rhymed in the Shijing. Yeah? Okay. So uh And then Li Fang Kuei adds to this hypothesis that phonetic series only have initials with the same place of articulation, yeah? So, let's say I think for for our mythological purposes, uh our we can say, you know, the Shěngmǔ hypothesis is if two characters are in the same phonetic series, they would have rhymed in the Shěngmǔ system. And have phone homo-organic initials, yeah? Um person with hand up, can you please ask your question? >> Yes. Uh so, like I mean, clearly it's, you know, like obvious they're in the same orbit, the same syllable, but I was thinking that yeah, it could be potentially from our ya, because I was thinking of the I was thinking of the syllable in Mandarin, yan. Like y e n. Um and that is I still phonetically write yan with the e because >> Yeah, okay. So, you're I mean, that's that's a very good suggestion, uh but you're sort of uh jumping ahead of us in a in a way where basically, let's say, um so far, we just have uh we just have like, I don't know. So, we have like A and B are both in the same Shěngmǔ series, yeah? Uh ergo, A and B rhyme, yeah? Yeah, and then and then I'll say, I don't know. Big R. Big R, yeah. Yeah. So, uh So, now you Now now you can say, well, ah and ye do not rhyme in my sense of what a possible rhyme is. So, maybe the maybe the ah maybe maybe ah changed to ye under certain circumstances, or maybe ye changed to ah under certain circumstances, or maybe something changed to both ye and ah under certain circumstances, right? Like those are the logical possibilities. If we want to get them to rhyme in our representation of it. Uh but I don't want to move quite so fast, yeah? Okay. So, um Yeah, so I just point out these are like, you know, both aspects of the Shixing hypothesis are instances of the inducto-dogmatic method, right? Where we take an observation and we turn it into uh belief, yeah? And uh and then we stick with it no matter what happens, yeah? And uh this is like how I'm going to end up defending uvulars in uh in Old Chinese is I think that uh those people who don't believe in uvulars have given up on the Shixing hypothesis before they need to, yeah? So, um yeah. Okay. So, uh So, if we look at Shixing hypothesis uh sorry, Shixing series, some of them don't have homo-organic initials and as we've already seen, some of them don't rhyme. Like like the horse one, they do rhyme, right? Like like all all of them are ma, so maybe Old Chinese had ma, right? Um but uh some Shixing series are poorly behaved, yeah? With respect to the Shixing hypothesis. So, we have to uh beat them into shape is uh the way I'll put it. And the first thing I'll point out is that we have uh type A and type B syllables, yeah? Now, uh since you've done a week of uh Chinese historical phonology already, I will just point out type B syllables are uh third uh division, yeah? And type A syllables are not third division. Okay, but what am I even talking about? Uh well, type B I mark with an I and type A doesn't have an I. So, if you just want to take it if you just want to treat Middle Chinese as attested, you can just look at my transcription is what I want to do. My transcription is better than Baxter, yeah. And you can just say if it has an I, then it's type B, yeah. That's all and that's the definition, right? I'm just saying by definition, type B syllable is a Middle Chinese syllable with an I in it, yeah? Okay, so here's an example. Type A, pong. Type B, pyang. And uh now I just want to show that the type A type B distinction is acknowledged by the structure of the Chinese script itself. So, you might say, "Look, Yunjing is where we get uh uh the the divisions from and Yunjing is from 10 01, I think, something like that. So, it's crazy to say that in 1250 BC, there you know, to just back project that uh from from logical distinction. Of course, it's not crazy because of exceptionalness of sound change. Everything in Middle Chinese has to come from somewhere, right? But, I have positive evidence that it was there in the ancient times and here it is. So, uh So, uh at the moment, the only thing I'm asking you to do is to look at this and notice that the arrows are all real, right? Like that the thing at the left actually appears in all of the characters that its arrow is pointing to, yeah? And the thing on the right comes from the thing on the left and everything coming off the thing on the right also contains it, right? So, this is that uh so, we we we shěng series as if it's just flat, but actually they have a structure, right? Because because you see this this guy who uh, who's in the, you know, this this guy. Uh, I don't know. This See, that guy doesn't come off there. He comes off that one, yeah? So, there layers to the shěng series. And in this case, now I write the Middle Chinese and you see everybody on the left is type A and everybody on the right is type B, which means the structure of the Chinese script itself uh, I I wouldn't say encodes type A type B cuz it doesn't very much. But let's say acknowledges type A versus type B, yeah? Uh, and here's just another one to show that, you know, it's like it's it's it's it's it's it's only now and then that it's it basically is in the largest series they acknowledge uh, uh, the type A type B syllable. How are we doing on time? >> Uh, we still have 14 minutes. >> 40 minutes, great. Okay. So, uh, yeah. So, so now I, you know, now I have uh, given myself access to uh, the distinction between type A and type type B syllables because I I know that Old Chinese had this distinction, right? That's the point. Yeah, it's built into the Chinese characters. It's very old. Okay. And then here's another one. Yeah. So, is everyone happy with >> No, this one this is weird. >> What which one's weird? >> The one with the ship. >> Okay. [laughter] The point is that everything on the left is type A and everything on the right is type B. >> That's fine. >> Yeah, that's fine. Yeah. Okay. So, let's improve conformity to the Shang Sheng hypothesis using the type A type B distinction. Yeah? Uh and what do I What do I mean by that? Basically, I mean like uh because I know that the uh distinction between type A and type B existed in Old Chinese, I can use it as a kind of wildcard to clean up Middle Chinese to make it uh behave well. Yeah? So, um So, now I'm going to to present kind of uh a catalog of problems in Shang Sheng series. Yeah? Problems in terms of uh uh well, in terms of violations of the Shang Sheng hypothesis. So, problem number one is the cha ta problem. Yeah? So, we have palatal affricates in the same series as dental stops. So, palatals and palatals and dentals have different places of articulation. So, they shouldn't occur together in a Shang Sheng series. Yeah? And then we have the the ya da problem. Again, ya and da have different places of articulation. They shouldn't occur in a Shang Sheng series. And then let's call the ha na problem, the HN problem. Maybe we should should label them this way, the CT, the YD, and the HN problem. And then the uh glottal stop uh HK. I I know I'll just say I use Actually, I'm not totally consistent about this, but I use this dot, which is called In Unicode, it's called the sinological dot. Yeah? So, I'm a sinologist, so I have to use the sinological dot, right? Uh it's uh uh not very many fonts can make it, yeah. But, I use that for the the the ing mu, right? Yeah. Okay. So, uh, yeah, and I don't know. We'll see what Guillaume thinks, but I like this way of doing it where like I can I can name the problem, yeah? And then I can use the machinery that I've already allowed myself, my you know, my compass and my straight edge to solve the problem. >> So, so in each case the question is the same. Is the situation relation genuine anomalous or does the mismatch point to an earlier phonological contrast? Okay, so here's an example of the D the CD problem. So, we have cho and do do and tu and du all in this same series, yeah? And then here's another example, another series. So, here we have tan and dan and um chan no chan >> Yeah, you you you're doing that. >> No, the chan >> chan Okay, yeah. And then chan and jian, yeah? Okay. So, my proposal is that in Old Chinese T palatalizes in type B syllables. What do you think, guys? Uh, and and Old Chinese D also palatalizes in type B syllables, yeah? So, to goes to cho and do goes to uh, jo. And actually, you know, you can tell yourself, if it makes you feel good, that it's the I that marks the type B syllable that is the conditioning environment for the palatalization, yeah? But Yeah, and then we can also say maybe the cluster TR gives us the retroflex cuz we have to get rid of the retroflex, too. Right? Yeah. We want to make this a pure dental series. And maybe DR gives us retroflex D. Yeah, so these proposals are all motivated by internal reconstruction in Middle Chinese as well, yeah. So, I'm going purely from the Shixiong series, but you know, actually I don't know if you exactly saw these. Yeah, I'm sort of walked you up to the edge, but he didn't quite give you these reconstructions, yeah. Um but yeah, they they they also improved the internal reconstruction in Middle Chinese, yeah. Okay, so now just in terms of notation, this is how I'm going to do it, right? So, I say that these first ones come from type B T and then type B TR and type B DR and type A T and type A D. So, I'm so from uh Yeah, I realized actually it's funny cuz if someone Marika said, she said like that my mind just works a certain way where I like, you know, where where well, anyhow like the point I want to make is no information has been lost, right? That's the point. It's like there I can in in mathematical terms, I can I can create a bijection between my reconstruction of Middle Chinese and and sorry, my transcription of Middle Chinese and my reconstruction of Old Chinese. So, I'm allowed to do it because I'm just moving symbols around, right? I haven't actually done anything. Yeah? Uh is everyone >> They know what to do. >> Okay. >> I just started the chat. There's just There's just people who just did the discussion. Yeah. >> Well, if anyone in the discussion wants to you know tell it to the whole class, then uh then please uh please go ahead, yeah? Okay, and then uh here's the second series, right? So, you know, I don't know, just sort of look and enjoy and say, "Oh, yeah, I see." Now now they obey the Sh Sh hypothesis in terms of they all have their home organic and have the same place of articulation. Okay. Now uh uh it's it's funny cuz my slides are giving you the answers, and maybe I shouldn't do that, yeah? Uh I will I will relabel them after uh after this, yeah? Uh but okay, so this one we have we have >> The it's the ya da problem. >> Yeah, it's the ya da problem, yeah. Yeah. So we So we've solved the Let's actually have our little menu here, you know? So we've solved the the the the CT problem. Uh now we're solving the ya da problem, and then what else do we have? We had HN, and we had the glottal stop K HK, yeah? Uh I don't know. So this one we've done. Yeah, check, yeah? Okay. So this one now. Well, uh we already get D goes back to DR for free, right? Okay, uh that's nice cuz we did that one uh when we were solving the the the previous problem. So then I have the, you know, so then I have the problem of just ya versus da, okay? So what in the world can become both a ya and a da? >> Maybe let >> Yeah. >> some other other >> Whoops. Yeah, okay. >> So we have a voiced palatal stop. >> A ja? >> A ja, yeah. >> Yeah, I actually think a D alone could become a ya, right? Like we have this yeah, yeah, yeah, yeah, yeah, I'm not very happy with with with with a ya turning into a da, but but it yeah >> But but in in so you could imagine becoming the >> Yes, yeah. >> It could be thus so and even like in Vietnamese the year of middle Chinese was the >> Yeah, yeah, so >> You could have that So Karlgren's original solution was to say that the the thing the the da of middle Chinese is da aspirated da and so originally in middle Chinese you didn't have a da and so this ya comes from the >> Oh, I see so so his so Karlgren's answer was >> Yeah. >> This one really needs to be written in Indic >> [laughter] >> Okay, so let's just do it in terms of my alternations, yeah, so we have the we have the the the the da Let's say the da da and uh >> and ya >> and and no the da da and ja >> Oh, yeah, yeah, yeah, yeah, the da da ja >> and and then we have the da da ya >> Yeah. >> Yeah, yeah. >> So Karlgren says this one's the da >> Yeah. >> and this one's da >> Yeah, so it says all the D's are da >> Oh. >> He doesn't distinguish between the two. >> I see. >> Whenever you have a ya in the one series you have a reconstruction of da >> I see and then and then that's how he solves the Shayshan problem because then he just has he has like you know let's just do it like this or something and then and then like probably like that and then he has >> Yeah. >> So that's fine actually that's a fine that's a fine solution. What I would say is that you'd never get a ya Yeah, yeah. Yeah, you never get a ya and a ja in the same situation. >> Yeah, that's So, this is indeed uh something that's So, it's just a coincidence in Calveran's view. >> Yeah, yeah. He sees that as a coincidence. >> Yeah. But so, this shows that that's what Calveran says, that despite his claims to the contrary, in fact, Calveran did have a the mind of a phonologist. >> Yeah. >> [laughter] >> Uh Yeah, and you could, you know, you could say, let's say, let's, you know, we would say, maybe this one's da and this one's uh tha or something like that, you know. Got it. Why not? >> Yeah. >> Uh and I don't actually really have a particularly good reason for doing what I'm about to do. >> [laughter] >> Uh but here [clears throat] comes. Let's go with L. Yeah. >> Yeah. >> And and and you can at least convince yourself that L is not a bad choice. Like, L becomes a D in the word lingua. No, D becomes an L. >> Lots of Bantu languages have have L to D. It's very common. >> Okay, lots of L lots of Bantu languages have L to D. And then And then like Spanish has L to ya. >> So, we're using uh cross-linguistic evidence. So, so is there a general parameter to like like like to to what can change into what or this is purely empirical? We're just observing all these different languages. We just have to say, okay, this is the evidence we have, this can change into that. >> craft sense. You know, like, look, S changes into H, H changes into zero, T changes into S, S doesn't change into T. >> So, it's a meaningless question to say why does S it change into H? Like like that that is an unanswerable question. You just don't ask that. >> That's the kind of question that you should ask Patrick Honeybone. Yeah, he worries about that kind of thing. I don't. Yeah. >> Okay. >> Um but what I was also saying is it would be nice if there was some resource out there that said like you know, I could just look up does does you know ever change into >> There is. >> Yeah. >> It's called Index Diachronica. You can search up whether or not a sound change has happened in a particular language. >> Woah. >> However, their data is sometimes really bad, so you have to check it yourself. >> Well, I didn't know about that. I know that the Clayton Marr is making something like this for the Max Planck Institute now, but he only has 18 months to do it, so who knows >> [laughter] >> who knows how well he'll manage, you know, um uh and then I will also say that that I want to say use the the German word Mangel, yeah, like in the absence of such a resource, a lot of us use uh who's the guy in Jena? A Kummell, yeah. Martin Kummell's Konsonantenwandel, which connects a lot of like or sorry, he's a kind of list of a lot of sound changes among consonants in the Indo-European and Uralic families. Um but, you know I think it's nice to just get a feel for it. >> Yeah, so just accept as is, right? Like in in most of these cases, just accept that this is a fact. Just accept that. >> Well, no, no, no, we're not accepting facts here. Like what accept like like Bantu changes L into D, that yes, accept that fact. >> Yeah, yeah, yeah. Okay, yeah. >> But like the I mean Yeah, I don't know. Like let me let me let me give the the example of uh ask T goes to S goes to H goes to zero, yeah? Like I'm telling you it can't happen the other way. But and that's not just a fact. Like I I that's I like I can feel it in my bones, you know? >> [laughter] >> Okay. >> I mean for for S to H, I guess that that you can't have um uh phoneticians in indeed arguing for the from the point of view of acoustics or of a uh articulatory phonetics, I guess. And we as historical phonologists I I think uh we are poorly equipped actually to really answer why this sound change took place the way it does, you know? Yeah, it's really because this is these are questions for people who do research on the sounds themselves. >> Yeah, yeah, yeah, sure. You know, but but I I but I I I will make a claim which is like in terms of uh uh s- like I don't know, like sonority hierarchy or something like that. Things things tend to get to the lenite. Everything wants to become zero, yeah? And it's a question of how it gets there. >> But zero doesn't want to be to become something else. >> Yeah. Yeah. [laughter] Uh >> Yeah, maybe you could say that once you the you you've lost your articulation within your mouth, like subtle h, then it's difficult to have a um a uh so a a a sound change that gives you back your vocal articulation as a uh um a as a context-free sound change. So, you can have, let's say, he going into she, who [snorts] becoming who, but you wouldn't have a ha becoming fa. >> Yeah, yeah, yeah. >> what I mean? >> Yeah. A- and and and actually I was just about to say uh what Gillon said in another way, which is uh th- there are two types of patterns. Yeah, they're like unconditioned and then highly conditioned. Yeah. In In a highly conditioned context, absolutely anything can happen. Yeah. Like because the context can, I don't know, increase the likelihood of of that thing happening. >> The context can be anything. >> Yeah. But in a in a in a in a you know, which is say like you can certainly have nothing change into a P. Yeah. If it's after an M and before the S. Yeah. Okay. [laughter] But but in terms of unconditioned sound changes, it's always going to be you know, things >> [clears throat] >> Okay. So, I'm going to go with L here in any case. Yeah, and then I and then here I have a note to add some typological comparisons for both L. Oops, I I I I didn't do it. Oh, okay. L to D. Okay. Well, next time if I, you know, this is why I'm recording it is then I can come back to this slide and then look for these typological comparisons. Yeah. So, L is not unreasonable in any case. Yeah, although you know, you can have a something else if you want. Yeah. Okay. So, here are the sound changes. L becomes D in type A syllable. L becomes Y in type B syllable and LR becomes D. In both type A and type B syllables. And now you say, "Do we really want to have a L L L in in Old Chinese?" And I say, "Yeah, you know, whatever." It's what people do. Yeah. And maybe as a Tibetologist, I'm not super happy with it cuz we don't have it in Tibetan. But but that's you know like like how how I how I say everybody who reconstructs clusters reconstructs this one. I think that is fair enough to say yeah. Oh, maybe. Okay. >> [snorts] >> Okay. Well, maybe maybe that would be better. Yeah. Yeah. But I'm also happy to just be a sort of nominalist about these things and say LR means the thing that changes into duh in series that have a ya in them. Yeah. Okay. So let's look at this series now, you know, so this is the series that we start with and then you see I changed it to L in type B syllable and LR in type B syllable even though you know, it's the same in type A and type B. I can tell it's type B because of the I in the middle Chinese. And then uh and then you know, so on. Yeah. And then you'll notice I'm not reconstructing the rhymes because I haven't thought at all about how to do it. Yeah. Okay. Um How are we doing on time now? Another 15 minutes or so 10 minutes? 20 Oh, great. Okay. Okay. So Okay. So now on to uh yeah. So now I guess I'm not well, I'm not going to deal with uh a problem, but I've created a problem for myself, which is uh Middle Chinese has the Laimu initial. Yeah. And so Middle Chinese has an L. But I've just used L for contact between Dingmu and ya you say okay. Yeah. Yeah. So what is going to be the source of the Laimu in in Old Chinese? And once again, I've given you the answer. So, anyhow, here it is. So, we need L to come from from somewhere. And here is a series. Yeah, so we have Liang Liang Lang Lang Lang Lang, yeah. Okay. So, if I had decided to just treat this one first, I would have said, "We should definitely use an L here." Yes. And then I would have not been able to use the L for my DY cases. And indeed, that is what happened in the sort of 1970s or so, where actually I think that yeah, yeah, Pulleyblank actually proposed uh uh Pulleyblank actually proposed this, yeah. Uh and that's yeah, because he wanted to use the L for the Laimu, you know. Uh but, uh you know, I would rather say, "Well, what can change into an L?" And how about R? Yeah, okay, so just R's can change into L's, yeah. Uh and then I'll also say, "In Middle Chinese, there's actually no contrast between R and L, yeah. So, uh you know, so so it doesn't matter, yeah. If if like if we had just written Laimu with an with an R, like a Japanese person would, then it would have [clears throat] been fine, yeah, because Japanese and Middle Chinese have no distinction between L and R. So, uh so this is totally legitimate, right? Might even have been pronounced R in Middle Chinese. Who knows? Okay, so then I get these, yeah. R in type B, R in type B, R in type A, R in type A, R in type A. Yeah? Anyway, I think everyone understands the game now, you know? Yeah. Okay. So, uh-oh. What's going on here? Oh, no. Okay. A- Any Anyone have any ideas? Okay, someone explain the problem to me. >> Well, the problem is we've already used L, so like what is whatever >> Well, well, well, well, no, wait. Just what's the problem with the series sort of synchronically? Why Why is this a violation of the Shishan hypothesis? Yeah, because it has because it has a lateral and a velar initial and laterals and velars do not have the same place of articulation. So, we need to come up with something that can give us both velars and laterals. Well, uh well, I just made the laterals into rotics, so does anyone have a a a hunch about what might do it? Yeah. Huh? Gura. Okay. >> Maybe a lat- a velarized L. >> Maybe a velarized L, yeah. Yeah, why not? Yeah. >> Long becomes tiang. >> Yeah. >> [laughter] >> Okay. Yeah, so let's do gura and kra, yeah? Uh so, here we go. Uh and so, kra becomes ja and uh gra becomes gia in type B syllables, and then uh long is from arang, yeah? Now, I will just point out that I I I I I have made the violation of the Shixiong hypothesis here less bad, yeah? But it's still violating. So, that's something I just want to flag, yeah? So, I'm not I think that's as far as I'm going to go with this for now, but it you should still be a little bit unhappy. You're like, okay, that's okay, but it's it's better, but it's not okay. Like that's the point. >> May I still point out something? You may notice that it's the yang yang bu, that is learned last week. And so, to just >> Yeah, yeah, yeah, yeah, yeah. Go ahead. Yeah. >> So, you remember that we had learned the yang bu, we had um here, we had pan ar here. We had uh yang here, and here we had yang in coronal initials, and yang, so gong san here with grave initials, yeah? So, with labials and velars this is with coronals. And what we see here is exactly this slot here. The rhoticized, so rhotic is zero uh or rhotic and zero palatalized, yeah? So, it's type B and it's with a rhotic. These These are the things that you have seen in this series. >> Yeah, and actually that's let's say that that's a reason why to reconstruct the Lai Mu as a ar actually is because uh because then cuz we've already shown that certain rhymes are rhotic, and in and now I'm I'm giving them rhotic initials, yeah? So, um yeah, so that's nice, isn't it? Okay, now onto the H N problem. So, what to do here? Anyone have an any idea? I need to rename these slides. Yeah. Yeah, yeah. How do you make How do you make that sound? Yeah. >> [laughter] >> Okay. >> [clears throat] >> Yeah. It's not. Yeah. Like I said I In Burmese In Burmese you say na. In Newar you say na. Yeah. But I think But we still think this is the the na one. Yeah, okay. Oh, and then I'll just point out you know, I call it the HN problem. But it's Oh, well, no. No. No, never mind. This is just to show another series. Yeah. Okay. So, yeah. Let's say that in type A syllables na becomes ha. Yeah. And there you go. Right? So so wo was originally na. And then this I don't know what it is. Probably ye Yeah, was originally probably nga. Yeah. nga. Okay. Yeah, okay. Okay. And then I just point out like why stop with the velars? Yeah, okay. So, here it is with labial. So, this is muk and huk. Yeah. Um muk. Yeah. Okay. And then here's another one. And this one is a che You see this is that Chongyin uh syllable. So, I have the Y Yeah, just to to to indicate it. Okay, and then uh So, here it's not going to an H. Yeah, it's going to a T, but it also solves a problem for us. Right? Uh and a lot of these have to do with like you actually turn out not to get uh um you know, in um in type A syllables in series with nasal initials, you don't get H, but you do get TH, yeah? So, let's say that na became ha in type A syllables. And then >> [clears throat] >> I will also just point out that every time we're doing this, we're slowly developing case law about what the type A type B distinction is, you know? Type A is something that can change uh uh voiceless nasals into dental stops, yeah? And type B is something that can change uh dental in stops into palatals and so on, yeah? And then you'll you'll slowly get a sense of like So, I think it's this is the, you know, rather than just starting with like I think type A type B is blah blah blah, you can really sneak up on it and sort of learn all of the things that type B and type A do, and then uh then you can have a conversation about what the actual uh phonetic interpretation in Old Chinese of the type A type B distinction was, yeah? >> Just a quick uh So, so the the the voiceless N to H uh I find articulatorily a little difficult to imagine. I don't know how we can get back to like which sound changes changes into what, but like na to na to na and then to ha, I mean it's almost debuccalized, right? So, you can kind of get there, but it's not so obvious why will become a um right? >> Well, all I can do is talk to your therapist about it. >> Well, one thing I I I have to say that the example was small. And he is a bit odd because because in most cases you have whoa inside. So, it's fa bwa the sound the most common sound choice is >> Oh, yeah, yeah, yeah, yeah, yeah. >> So, the last >> Yeah, and then you can understand the the labialization as being kind of trans phonologized. Yeah. Yeah, yeah, yeah. >> And and that you can actually see it in Pumi because in Pumi you have mi becoming phi with a transfer of nasality on the vowel, actually. >> That makes me think so. >> Yeah, yeah. >> I'm I'm I'm sorry. I'm not I'm not sure if I'm really following this cuz what is type A and B? >> We haven't explained it. >> Yeah. >> I just I just go back a long way and I say type A is what whatever the ancestors whatever made the ancestors break this Shishang series in this way. Yeah? Do you understand? >> But from the point of view of Middle Chinese, maybe. Type A is first, second, and fourth division. >> Okay. >> Type B is third division. >> Okay. >> Yeah, but the but the point is we know Old Chinese had type A had type A type B because because of Shishang series like this. Yeah? Whereas we don't have Shishang series where Well, it's not quite true, actually, but the Shishang series don't generally acknowledge the four dong. But they do acknowledge the the sandong. Yeah? So, so we know that you know, like I I feel like, you know, I feel like there's a Let me just talk this one through and actually Uh Guillaume, you like Oh, actually Here. Then this one. This one. >> New and new. >> Wait, what? Oh, yeah. Okay. It's this one. And then this one. Okay, yeah. So So this is uh Yeah, let me just go to the next slide. [clears throat] New new and new. Okay. So this is new. This is new. And this is >> new. >> Yeah, with an H, yeah? >> With a tilde. >> Yeah. Yeah. Okay, so >> the NA reflex. >> Oh, yeah. Yeah, yeah, yeah. Okay, so When our ancestors, you know, decided to make up new characters, they generally didn't go back to this one. They went to either this one or to this one, right? >> Yeah, it's my job. >> So it means that And and and then if you just look at the slide, you know, like for this Let's take it the example of this sh, yeah? Okay, so it means that this this sh had something that made it more like this one than like this one. >> Yeah, that's >> Right? >> Yeah. >> What was that something? Well, we're calling it type A, type B. >> Yeah. >> Well, but but but like what's preventing us from having a type C? Because like like like we just happen to have type A, type B, but then like like I I can iterate even more examples based on the the ancestor. Like like if there are those. I I don't know if if there are those, but yeah. >> So so so the question is uh put another way, which phonetic distinctions are acknowledged in the structure of net nested Shixing series. Yeah. >> Yeah. >> And that is an empirical question that I have answered by sending you a PDF that you looked at and you said, "What is this garbage?" Yeah. Do you remember it? >> Yeah, yeah. >> Yeah. >> So I I give you all the Shixing series with all their nested structures. >> So we have empirically >> No, it's always binary. There's no >> no. There There's all kinds of stuff that's in Shixing series. Like uh certainly uh the sometimes you get the division two broken out. Yeah. >> Yeah. >> Sometimes you get the the initial you get only voiceless initials or something. There's Like you can just go look at my handout I gave you. >> Yeah. >> Uh and see like what uh the trouble is you can't use the Shixing series kind of systematically uh because the these distinctions are not written systematically. >> Right. >> Some the larger series the larger the series is the more fine-grained the phonetic distinctions are. Uh but it's only by looking at the Middle Chinese transcriptions that you can see what it is that is being transcribed in the Old Chinese uh structure of the Shixing series, right? >> But couldn't we say that uh preliminary fashion that type B had a year medial or something I think? >> Yeah. >> Yeah, I think that's I mean that's how I'm writing it. Yeah. >> Yeah. >> Is like uh Now now then you do get this thing of like do we like a language where half the syllables have Yeah yeah like I mean but uh >> I mean I guess >> I guess you can analyze Russian that way. >> [laughter] >> That's true. You can analyze Russian that way. Yeah. Yeah. And I don't know where it comes from in Russian. >> There's more e So short e >> One of the years. >> Yeah the years yeah that comes from short e. >> Like but was it short e in Proto-Slavic? >> No yeah. >> It was like an e. >> So they had a lot of e's in Proto-Slavic. >> Yeah yeah but yeah. But those those those come from like unstressed unstressed vowels. >> Oh yeah okay okay. Yeah yeah yeah. See? >> There was no joke. >> Everything comes from something right? Okay. Oops I hadn't gone that far yet. We have 6 minutes. Okay. Then uh we have 6 minutes so I'll just do one more. So I did not go to ta in type A syllables. So let's see what happens. Oh hna goes to sha in type B syllables. And this is our is our new Yeah I was just talking about this character right? Yeah. Uh So Yeah. So that's nice isn't it? Yeah. Okay let's do a couple more. Okay. Okay. [laughter] Hna Sna goes to a ta ta Yeah. Uh as you get in this someone say the second word for me. Yeah in opinion why not? Yeah. >> Uh you mean the ch >> Yeah. I mean the ch Yeah. Okay. So cuz it has year which is ni or ar as it's uh as it's uh phonetic. Yeah. >> Yeah sh >> We should say however here that normally the phonetic element is on the right side. So here actually this character shows that maybe the people in a fun dynasty didn't realize it was a phonetic element anymore because they didn't Otherwise we should have type the R. R total R. We should have if they had realized that the phonetic element is the is the ear. Well, they probably would have >> I have never I have never known that the phonetic was usually on the right. I I just always ask the Wiktionary which which side is the phonetic where is the phonetic. And sometimes the phonetic is actually like God there's one character where where the phonetic is split in half vertically with something shoved in between, you know. You can you can really have all sorts of stuff happening, yeah. Um Okay, let's do let's just sneak in a couple more, yeah, cuz I know I know you like voiceless resonance, yeah. Okay, well how about we look at this one this series. So you see it's one of these lateral series cuz you have the Y and you have the D and then you have a sha and a ta. Well, okay, if sha can come from sna in type B syllables maybe it could also come from sla in type B syllables, yeah. And if and and if pa can come from sna in type A syllables maybe sla can also come from sna maybe a sla maybe pa can also come from sla in type A syllables, yeah. So you see I mean the the the the other thing I'll just say about the A B distinction is you see that we can get a ton of work out of this you know you in in terms of, you know, and and and still I would say we're just playing a game here, and the game is take a phonetic series in Middle Chinese and torture it into being uh a homo-organic with the same place of articulation uh for the initial in Old Chinese, where basically the constraints are it has to be plausible that the thing you reconstruct becomes what's there in Middle Chinese. Uh and uh the only tools you have, the only tool you have is the the the distinction between A and B, right? Yeah. >> And you have >> Uh and a medial R. >> Yeah. >> Yeah, so far. Yeah. Okay. Uh and then how about we had these Laimu, which were coming from R. So then if we have a th that intrudes you know, Laimu series, then we can take it back to a uh voiceless R uh uh I should have I meant to get rid of the little pharyngealization thing. Uh cuz I'm cuz cuz that just means type A, but I'm writing them out, yeah? Uh okay. Uh and then uh in this one's different cuz in general like n and sl became sh, but uh sh in type B syllables becomes t t Yeah. Uh as in uh this uh case, right? Okay. And that And now I just uh Yeah. So we're running out of time, but uh I'm coming to the end of voiceless resonance. So uh Dong Tong Hong is the first person who proposed voiceless resonance in 1949, and he uh said it with the M actually was the first one uh that he proposed. And then Pulleyblank extended it to the velars, the labial-velars, uh the dental nasals, and the voiceless lateral. And then Baxter added the voiceless rhotic. So that's the Forschungsgeschichte, yeah. Uh and then here is cuz you were like Sorry, this is going too fast. How does it all work? This is how it works, yeah. So you can, I don't know, take a picture of the slide or something there if you want, yeah. Uh and then you see that actually it's kind of I mean, elegant might be overstating it, but >> [laughter] >> I've I've managed to put the outcomes next to each other in a chart, yeah. And and I feel like that speaks in favor of my uh proposals, yeah. >> It was surprising to me that that that the voiceless r gives you ta but uh between type A and type B. Um something retroflex is fine if if you want it, but why why is it not rhotic in type A? Which it it's supposed to be what Oh, yes, but I I know it's supposed to be >> Yeah, so I totally agree with you. Yeah, and I would say maybe So first of all, I'd say that's why it was discovered last, maybe, yeah. Uh but you know, like So So there's one thing that that I So I'm not a Sinologist, right? I'm a Tibetologist. So my goal here is to just sell you what's for sale, yeah. And if if you want to buy something else, you have to go to someone else, yeah. >> [laughter] >> Or another way of putting that is like I'm a follower of the Baxter and Sagart system because it saves me a lot of time, yeah. I don't have to go and read Starostin in Russian, yeah. So uh but I agree with you. I don't don't like this at all, but uh you know, maybe someone will think about what's going on there. Anyhow, we'll stop here for today, uh but I've gotten you you know, a lot of initials already, right?