 And then actually the first thing I will would like to do is just ask whether there are any burning questions from previous lessons because today we're going to go into the six vowel theory and I don't know it's going to be a little bit of a roller coaster so you know if there's anything is I want to kind of make sure that we have a clean sort of workspace so if there's any lingering questions from the last two times then now is your opportunity to stick your hand up. I think we need to distinguish very sharply here between middle Chinese and old Chinese. Some people would feel like what I'm about to say is an overstatement but I would say the differences between Carlgren's middle Chinese and the middle Chinese is that various people work with today are pretty superficial. So, like, let's say, Leifang Quay's middle Chinese is a kind of orthographically cleaned up version of Carlgren's because Carlgren was was kind of not a big believer in the phoneme. So he tried to capture what he regarded as real phonetic differences, which means that his reconstructions are kind of very ornate orthographically. So, Leifang Quay came along and just tidy those up and removed all distinctions that were not needed from a kind of information theoretical perspective. And I think that Baxter's middle Chinese is in effect, exactly the same as Leifang Quay's, although actually there might be an issue about the Chongyu distinction that someone can tell me about, which is that I think that Baxter would claim that his system contains slightly more information than Leifang Quay's or Carlgren's but in any case just to kind of return to your question, there's a clear continuity there and from my perspective not a lot more progress has been made. Whereas in the case of old Chinese, I think it can be safely said that Carlgren's old Chinese, he kind of posed all the right questions and came up with answers that are almost all considered superseded now. What is true, especially if you read Baxter's 1992 book, is he always starts from Carlgren. And for me, this was a huge obstacle as a reader because because then I need to sort of first understand Carlgren's system and then kind of fix all of the problems that have been seen in it over, you know, 60 years. Yeah, but that's, let's say, which is to say, from a from a pedagogical perspective, Carlgren still has a huge role in in presenting the current understanding of old Chinese phenology, kind of in a way that I was, I was dealing with here with Don Duan-Yu Cai, which is like, well, we start from this person's view and then we fix the problems with it. So I am trying very hard kind of in my presentation to kind of write Carlgren out of history in a sense by way of finding a faster presentation that goes directly from what is the evidence to what is the current thinking. If you read Baxter's 1992 book, you know, he will always say, we start with Carlgren, we go through Li Feng-Qui, and now these are my proposals. So, so I would say that kind of to sum up what I have been saying, Carlgren's mode of thinking and his identification of problems are still the modes of thinking and the identification of problems we live with. But his particular solutions for all Chinese have been basically superseded. The difference between the type A and type B syllables is one of the most controversial areas in Chinese historical phenology and the, let's say, the two, here are two of the most common current perspectives. Zhang Zhengxiang Feng thinks that I hope I get this right, that type A syllables have long vowels and type B syllables had short vowels. Whereas Baxter and cigar think that type A syllables had pharyngealized initials and type B syllables did not. So the main, the main phonetic evidence of what the type A, B distinction is, is that type B syllables systematically palatalized in, let's say, in the early Han dynasty. So there's, there's some kind of, you know, which, so which is to say type A had a non palatalizing phonetic conditioning environment and type B had a palatalizing phonetically conditioned environment. In 1992, Baxter actually projected a medial, yeah, all the way back to old Chinese and said like well yeah is a is a palatalizing conditioning environment. But there's, there's, when you look at like early Sanskrit or early Chinese attempts to write in the language is actually gone Dari more than Sanskrit. There's no evidence that there was this medial yeah, there's no comparative evidence that there was this middle yeah so. So, so he's dropped this medial yeah idea, and instead Baxter and cigar picked up this idea that originally came from Jerry Norman that that that the type A syllables were pharyngealized and the type B syllables are non pharyngealized. And then, you know, you may well ask, well, why, you know, would would non pharyngealization condition palatalization and this is where you should really ask a phonetician. But there's some pretty good evidence from Arabic dialectology that it does. It's that it's reasonable. And actually, one of the students I sent a link to this in email correspondence but Mark Miyake four years ago and so as gave a talk on, you know, his own reflections about the type A type B syllable that you can find on YouTube. And, and he does some very interesting comparisons with with Arabic dialects. So, so the answer is ultimately we don't know. I actually think what is the origin of the type A type B distinction is one of the great. It's kind of one of the one of the, you know, if you if you could tell us the answer in a really convincing definitive way then you'd be famous. So, okay, so now we're going to be talking today about the six vowel theory. And, yeah, so a few sort of preliminary remarks, which is that three investigators. Jong-Jun Chang-Fong, Sergi Starrostin and Bill Baxter all independently came up with the six vowel theory around the same time. I mean kind of let's say roughly around the same time between the late 60s and the early 80s. And kind of word has it that Jong-Jun Chang-Fong was the earliest and that he came up with it. While sort of after he was sent down in the Cultural Revolution and he was out in the countryside with a lot of, you know, time to to think about Chinese historical knowledge just in his by himself. And he kind of saw this theory. And then probably second was Sergi Starrostin. And then third was Baxter. But the, you know, the work of Starrostin and Jong-Jun Chang-Fong was not really available to Baxter so they were all sort of working independently, but starting from some ideas of pulley blanks and and yahontovs. So why am I saying all this? It's because in some sense, if three people at the same time independently came up with the same theory, the theory must be kind of obvious. Right? It must be that like if you have the right background, and you just sit and think for a while, the six vowel theory is kind of obvious. But it's still controversial. So, so the people at in Beijing don't really like it and the people in Taipei, they don't like it either. So it's not so obvious that once it's pointed out any well informed person would agree. And let's say I haven't found it particularly obvious. So one project that I made in my book that now I'm making a sort of second ever that here is how to make the six vowel theory more obvious. And, you know, you'll see what what you think of my presentation. And then maybe yourselves as you think about these things you can think about what kind of presentation of data or what sort of posing of the question would make it even clearer because I firmly believe there is some way that that we can make the six vowel theory just really obvious. Okay, so here we go. So first I'm going to look at type a syllables. And first, you know within that the internal reconstruction of middle Chinese syllables that end with a velar nasal. Then we'll look at the internal reconstruction middle Chinese syllables that end with dentals. And we'll already get the ideas will already get a sort of six vowel theory or or actually we'll start with a sort of five vowel theory. motivated by the first situation which is the internal reconstruction middle Chinese syllables ending with and then we will sort of test that theory against different other situations, and then see that it either works or has to be refined a little bit. So then we'll turn to the type B syllables which are more complicated. And then kind of repeat the, the process so internal reconstruction of middle Chinese syllables, ending with no and internal reconstruction of middle Chinese syllables ending with dentals. And just a notational convention where in Baxter's middle Chinese transcription he writes Oh, I am going to use a schwa. And that's just because I think it simplifies the presentation, because otherwise you get Oh, meaning one thing in all Chinese and a different thing in middle Chinese. And it can just be I think a little bit visually confusing so I've made this sort of orthographic substitution in order to assist with the clarity of the presentation. Okay, so internal reconstruction of middle Chinese type a syllables ending with no. So we'll look at the distribution of these syllables, and then propose the reconstruction of labial dealers. Okay, so this is the, the, the, the, let's say, the presentation of the phonotactics of middle Chinese type. And this syllables ending in dealer nasal, which is to say on the, on the left, you have a list of middle Chinese rhymes. And then in the second column you have the name in, in terms of the, you know, Chinese philological tradition of those rhymes. At the top, you have the position of articulation or, or, yeah, that's the position of articulation, which is to say, you know, capital P this is a normal convention that you're familiar with right. This is pa, pa, and, and, yeah. So then we see which types of syllables occur and which types of syllables don't occur. And let's say the first point to make is just that it's ugly, right like this is a very asymmetrical distribution of curl co occurrences between rhymes and initials. So, you know, the point of internal reconstruction is, which is basically let's say methodologically the same as phonemic analysis in a way right is, how can we look at this mess and make it more beautiful. So the first thing we do is we propose that, I mean actually I'll talk you through the evidence of it. You see that there that the, that those rhymes that have a medial w only occur with the dealer initials. So, then it's tempting to say that in those cases the medial w is actually a feature of the initial rather than a feature of the rhyme. So that's this proposal which is basically that we can reconstruct a class of labial dealer initials. Okay, well then let's do that. And this is what we get right so we've just, we've added a row. Right now we've added a column at the far right. Now we have labial dealer initials, and then we've rearranged the distribution of the cells. And it's much more elegant. I mean, I think you will agree right so we went from this with lots of blanks in lots of places to this where there are fewer blanks. Yeah, so I think we've made some progress already. So now the next hypothesis, which again you can just understand as motivating entirely based on internal reconstruction is that we have these these two Middle Chinese vowels. So the first one is, is, is, you know, it was a written together. And the second one in IPA is is epsilon. So I think. I suppose that the first one came from a wrong with a medial R in China in old Chinese, and the second one came from a ring. So with, with medial R and then a vowel. And then we can now remove from our phonotactic distribution, the rhymes, and, and, yeah. So now I've done that I've removed those. And this is the, the pattern that we get. And, and now we've already reached a five vowel hypothesis for old Chinese, based on the internal reconstruction of type a syllables with a dealer nasal final. And the only remaining gaps now seem quite well motivated phonetically right the labial dealer initial does not co occur with the rounded vowels that that's a, you know, that doesn't bother me that seems like a very normal thing to happen across the world. So, you know, we're done in a sense like oh great we've made a nice elegant internal reconstruction of the old Chinese vowels, but does it work for the dental finals. So we will now take a look at that, and just to sort of preview that we will again make use of the labial dealer hypothesis, but we're going to add another hypothesis which is the rounded vowel hypothesis. And I'm going to just to sort of simplify the presentation, I am going to assume the our hypothesis, which means I can exclude all rhymes with the vowels, and because they're secondary. Yeah, so, then if we start off, you know, which is to say so far I've only assumed the our hypothesis. This is the distribution we get. And once again, you notice that it's a little bit different than last time you notice that medial w occurs. So it's not favorably with the dealer initials, but not only with dealer initials. So, yeah, also just, you know, it's worth the effort is not a very elegant distribution. So we look so we try it first to get rid of this Quinn, and you guessed it we're going to use the labial dealer hypothesis and there we go we got rid of the Quinn by adding this, this column of labial dealers. And, and I'm, you're now seeing how I will be presenting the rest of the argument today, which is basically, I leave the middle Chinese, I make a proposal for for old Chinese, I present it as reconstructed and then we kind of watch how that cell develops through time. So, let's look first at Kwan. So, Kwan goes from having the labial as a part of the initial to a medial. And then we, that's how we get from old Chinese to middle Chinese quite straightforward we leave that one behind. We go to the next one, which, you know, let's I'm not proposing that the old Chinese actually had this syllable. It's just a sort of analytical device right it's it's the initial labial dealer, plus the plus the rhyme in the row. Well, you know, just now changing the labial dealer initial into a velar followed by a labial medial, we get Kwan with two W's in a row, which of course doesn't make any sense. So we reduce that to just one. And there we go. And now you see that these three cells all would have ended up with the same thing in middle Chinese, right. So, Kwan in middle Chinese will sort of occupy three cells of our presumed old Chinese reconstruction. So next we look at this quen, and you kind of get the idea that's how we get quen. And then now we look at a quah before the, sorry, the schwa pretty, you know, you understand so I don't need to talk through all of it. And now that's, we're back to middle Chinese. Okay. So now we look around and see, let's try to fill some of these other holes. Now that we've dealt with quen. So, yeah. So now we're moving on to the rounded vowel hypothesis, which is, we still have some some medial W's right and we kind of want to get rid of them well why do we want to get rid of them. So we can say well we managed to get rid of them with the, the velar finals. So, so it would be quite inelegant if you had a version of all Chinese that had no medial W before velar finals, but there is a medial double before nasal finals. So I sorry, before dental finals. So that's our goal is to get rid of these symbols. Yeah. So we propose the rounded vowel hypothesis. And the rounded vowel hypothesis is that rounded vowels break before dentals. So we get on goes to one and goes to one. Now let's see what happens if we apply the rounded vowel hypothesis. So, you see already that that now we're kind of getting close to that five vowel analysis that we managed to achieve for the dealers. And, and yeah, so now we'll talk through those developments from all Chinese to middle Chinese, like we did already. First looking at tone tone changes into one. Then we look at so on so on changes into so on. Then we look at con con changes into one. And now we look at who on which you know may well not have existed. But in any case, you get again. Con. Okay, now, fun. One. One. Soon. One. And which probably didn't exist goes also to one. So we've made some good progress. And now we just have or now we still want to fill these gaps that are remaining. So we say, let's fill the gaps for one and one. So middle Chinese does not have a phonemic distinction between pa and poor. And this I think it's a good point when we're in child who figured this out in the 1940s. So that's nice for us at the moment. And we can understand it sort of historically speaking as a merger. So there used to be a pa and a plot that were distinguished. And now you just get, you know, you can think of it as either a pa or pa, but we just get one thing that doesn't distinguish the two. So that's a merger. This is maybe a sort of subtle point, but the, the rhyme books. No, sorry, it's actually the rhyme tables will sort of annotate some labial initial syllables as hooker and some as Kyco, which is to say some with a medial w and some without a medial w. There is no phonemic contrast, but in respect to the philological evidence in Baxter's representation of middle Chinese, you still get sometimes, you know, a P and sometimes a PW kind of in accordance to how the rhyme tables presented, but there's no distinction. So I just wanted to emphasize that. And now we can fill in some of these holes by look. So we look at this and and we look at this one, and you see that. Yeah, was that clear? Yeah, you go from. So, so you go from Pawn to Pawn and then Pawn becomes Pawn because there's no distinction between Pawn and Pawn in Middle Chinese. And now we've managed to fill that hole. Now we turn to Pawn, and it's very similar. Pawn goes to Pawn and then Pawn we have. And then the same thing will happen with Pawn. No, Pawn I think we already had. So in any case, this is how we fill these cells and then and then you see that there are a lot of kind of mergers that go on the way from old Chinese to middle Chinese. And now the only in elegance is is done and son are missing. And these ones are sort of harder to motivate. And what would be nice is to find a place in the Shredding where where those syllables rhyme with Kun because then, you know, or where any syllable that starts with a toe or its rhymes with Kun because then you would be able to figure out, oh, okay, probably has the same rhyme as Kun. And then I can see how it develops in Middle Chinese, but there just happened to not be such cases. So we're going to just leave it there and live with the pain of having those two cells missing for a good while. We're sort of done with type a syllables and we'll move on to type B syllables. Like I think you know it kind of you can do that if you want. The question is, how does that interact in in your system with the labial velar hypothesis. That's fine. But then how do you explain. I mean, it's it what it fundamentally comes down to is that there are there. Let's let's kind of put it this way. There's an in elegance in Middle Chinese, which is that some some hooker rhymes are compatible with all initials and some hooker rhymes are only compatible with velar initials. So I think that the combination of the labial velar hypothesis and the rounded vowel hypothesis is very nice in terms of getting rid of that in elegance. Now, there's another consideration to which we sort of haven't gotten to which is, you know, from the perspective of internal reconstruction Middle Chinese kind of who cares, we can have the rounded vowel hypothesis and reconstruct rounded vowels, or we can do what you're proposing and and do the opposite and say those rounded vowels that we do have are kind of kind of merged is not quite the right word but our coalescences of a medial W with a non rounded vowel. That's fine from the perspective of Middle Chinese internal reconstruction. However, they yield different predictions in terms of the behavior of rhyme in old Chinese, where the rounded vowel hypothesis would suggest that that Middle Chinese, let's say to on would never rhyme in all Chinese with Middle Chinese town. Whereas your hypothesis would predict that Middle Chinese that that Middle Chinese to on and Middle Chinese town would enter rhyme in all Chinese. Now, the question is, if you just test those two hypotheses against all Chinese data, what's the result and the answer is, it's controversial. You can think of the whole of Baxter's 1992 book as an attempt to do that. But what he's saying is that his claim is to on and on do not rhyme in all Chinese. So, that's very strong evidence that the labial that they that the rounded vowel hypothesis is correct. Let's say, if I try to take the perspective of a third party and say, there are two camps, there's basically the, the six vowel camp, and the old Chinese had a medial W camp. The question is, has the old Chinese has a medial W camp answered the arguments of Baxter 1992. And my impression is the answer to that question is no. That they're there, like if you read Polly blanks review, or if you read hood on review of the 2014 book, like, there's a lot of venom, you know, against the six vowel hypothesis. I don't see an attempt to offer a superior analysis of the data of old Chinese rhymes that answers the, the arguments in Baxter 1992. And I mean that might, let's say, to give everyone the benefit of the doubt that might be simply for non-organist reasons like that Baxter's argument is extremely sophisticated and involves the use of statistics and whatnot and that, you know, these sort of elderly gentlemen philologists in Taiwan are just not, you know, they just don't have the tools at hand to offer the kind of critique that can be made. Yeah, I'm perfectly willing to believe that. And then the issue is just, well, then someone needs to do it, right? Someone needs to, let's say, rise to the occasion of demonstrating that the belief that old Chinese has a medial W, better explains the Ryan patterns of the Shenzhen than Baxter's system does. And, you know, as far as I can tell, that hasn't been done. It's a hard project because, you know, it's sort of responding to a sort of 1200 page book, right? But I haven't seen it done. And maybe I haven't seen it done just because I don't, you know, know the Chinese language literature well enough, but that's my sense. I would just sort of remind you about the sort of what we covered yesterday in terms of methodology, right, which is kind of the way we try to get at old Chinese is by reconciling three things. One is an internal reconstruction of Middle Chinese. One is Ryan patterns in the churching, and one is the Shesham series. Yeah, so to kind of, you know, sort of attempt to restate your question. It comes down to how well you think this internal reconstruction does against those other two sets of data, right? So like, you know, let's put it this way. And no hypothesis in a way is all Chinese was exactly the same as Middle Chinese. But we can show that that that only gets us so far in making the rhyme patterns of the churching and the Shesham series work. I mean, it does get you a certain distance. Yeah, and I think the fact that it gets you a certain distance shows that Middle Chinese and all Chinese are not unrelated languages. But you sort of when you look at the rhyme patterns of the churching just in Middle Chinese, or you look at the readings of the Shesham series just in Middle Chinese, you sort of say there's something going on here but it's not quite right. Now we're proposing based purely on questions of elegance internal reconstruction Middle Chinese. And then we can see does that, you know, let's say ideologically driven internal reconstruction Middle Chinese actually help us do something like explain the rhymes of the Shesham series or make the Shesham series more regular. Let's say there's the mode of investigation and the mode of presentation, right? And, of course, the this way of internally reconstructing Middle Chinese was not invented in isolation from those two bodies of data. That was a kind of historical dialectic process. But I'm trying to find a mode of presentation that's elegant and I may have failed in doing that, right? So, so, so in doing that I've tried to sort of isolate the act of internal reconstruction and then be able to present it as here's a hypothesis motivated purely on Middle Chinese grounds. Oh, look, it works against the data. The reason why I think I think that is more elegant than doing it the other way around is because otherwise it would just be very anecdotal. It's like, oh, look at this rhyme. It doesn't seem very nice. Let's try to fix it. And that would just be sort of like sort of wandering through the fields and randomly picking flowers. Yeah. Whereas, if you start from the distributional patterns of Middle Chinese, it is motivated the hypothesis to me seem more motivated in terms of, you know, considerations of elegance. Okay. So now we have a theory of all Chinese vowels. And what, yeah, based on type a syllables, and we want to move to type B syllables. So what I do in my book is try to kind of repeat the sort of reasoning that we have just done for the type a syllables with the type B syllables. But I in preparing these lectures I decided that that actually was not that that that was that the type B syllables are sufficiently complicated that that ends up being quite ugly. So instead, what I'm going to do here is more or less assume that our analysis of the type a syllables is correct, and then try to fill in this grid in the type B syllables, which is to say we assume that all Chinese had these five vowels so these five rhymes before and we assume it had this class of initials, and then we can try to fill in each cell with the outcome in Middle Chinese type B syllables of that old Chinese syllable so that's the goal I'm going to set. And we will try and do it, mostly by looking at rhymes and Shesham series. So let's see what happens. This is where you can think of it, I have the grid is all full of question marks, and I'm going to try and fill it in. You will notice that the way where I'm writing the classes of initials have changed where you have all these J's and Y's. That's because that's how type B is indicated in Baxter's transcription of Middle Chinese. So, we look at Middle Chinese type a readings of unambiguous origin. We identify the type B Shesham connections of that reading. We look at rhyme behavior of words with that reading, and we try to fill the slide. Okay, so first we're going to look at Tan, and just to sort of talk you through it. Why am I not looking at Pan because it's not unambiguous in terms of origin, and why am I not looking at Quan it's because it's not unambiguous in terms of origin so I'm starting by looking at Tan in type A syllables and seeing kind of what type B syllables it has connections to. So here we look at a Shesham series where we have Tan in type A syllable, and it has type B connections like Xian and Trian and Xian. So this suggests that all Chinese on change to N in type B syllables, at least after acute initials, which is to say after ta and you know, ta ra. So that's one now sound change we can propose. And this is something that Baxter calls acute fronting. Okay, now let's look at rhyme behavior. So Tan can rhyme with Pian, Dan rhymes with N, Dan rhymes with Drian. So we have further evidence for acute fronting. And we also have evidence for sort of Yan turning into Yan. And this change Baxter calls a raising. Okay. And we also have evidence that acute fronting preceded a raising, which is why we get the schwa only in circumstances, only with those initials that are non acute, or putting another way only with grave initials. So you see in Ode 76.3, we have the schwa after a velar initial. But in Ode 112.1, we have the eval after a retroflex initial. So, so that's my argument that acute fronting preceded a raising. Okay, so this allows us to fill in the first row. Yeah. If if this isn't making sense, then someone, you know, stop me. So now we look at the second row, and we find an unambiguous type a syllable, unambiguous in terms of his origin in the second row. So I'm going to look at one. And then we then look at what are the type B connections in Shesheng series, and in rhymes of this type a syllable one. So we look at the Shesheng evidence, and we see that, for instance, the one is connected with Twen, or they will say Twen, yeah, Twen and Twen. Okay. So this suggests that vowel breaking preceded acute fronting. So what, what, why am I saying this, because imagine that that this one was actually done. So then the first syllable, let's say, here, 0231a would have been sort of type B son, and the second syllable 231c would have been type a don. Well, then you, you change them both to things like Dwan and Chuan, and then you get acute fronting to change Chuan into Twen. Yeah, so vowel breaking has to precede acute fronting. Okay. Now we look at rhyme evidence, and we see that Dwan rhymes with things like Nguyen, and Luan rhymes with things like Shwen. So we, yeah, and then I will just leave to the reader or to the listener that that pawn in type B will turn out as Chuan. Okay, so now we filled in the second row. Now we turn to the third row. No, actually, we skip the third row. Yeah, because the third row. Ends up being tricky, I think. Yeah. So let's go to the fourth row. Chua, N, and we look for an unambiguous syllable in terms of its origins in type A syllables. So we look at Khen, and we look at the Shesong connections of Khen. And the one that's relevant for us is the last one with this barred I, and this gives us the pretty trivial result that, you know, in type B syllables, something like Khen turns into, I mean, I'm not going to try and distinguish but any case, that's one result. Now we look at rhyming evidence, and we see that, okay, the first one just confirms again that type A Chua has connections with type B barred I, but then the second and third examples are more interesting, because here we have, we look at this barred I that we know has a connection to the type A syllable that we started from, and we see that it has two further connections with un and within. So here are our conclusions. So we have something actually like the merger of a W initial before Chua into the round of our Un, yeah, and I do not actually find anywhere that Baxter discusses this change. Maybe you saw it as kind of too, pick I Un to worry about, but it seems like it's necessary to make the whole system work. Okay, so, and yeah, so that's the first thing we noticed from this rhyme evidence was this sort of changes back to just mention. And then the second thing is something we can call Chua fronting, which is that basically between two imputes, Chua fronts to it. And Baxter called this barred I fronting, but that's because in 1992 he used this barred I everywhere where in 2014 Baxter and cigar use a Chua in some of these cases so I think we should call it Chua fronting. Okay. Two remarks. The only evidence I have for the analogous change with with retroflex initials is from this series. And yes, and if if you. We saw this. This sort of pun turns into pure. So then we would expect something like that to happen with a final yeah, and it doesn't. Yeah, no such syllable exists. So I mean, these are my, let's say, just random observations you can understand it that way. And as a third remark, it looks like Chua j type a syllables and Chua j type B syllables, never rhyme in the surging, which just is maybe it's a random gap or maybe it's something that should be looked at more. You know, somehow they were already pronounced differently enough in the surging that they didn't make good rhymes. I don't know. No one I these are observations I don't think I've seen anyone talking about before. But in case, you know, blowing ahead, we have managed to fill in the fourth row of our type B syllables. So now let's go to the last row. And we look for again a type a syllable with an unambiguous origin, and then we look at it's chase on and rhyme connections. So we do that. So we are looking at station connections for one. And here they are we have kind of the stuff you would expect actually you know so twin twin. And this allows us to propose that I think I've maybe done something wrong here. Yeah, it's that they're one of the signs is backwards sorry about that. But so we we start from type B, it turns into one that's vowel breaking, and then we get a twin that's acute fronting. And then similarly turn goes to turn goes to turn. So these are sound changes we've already seen before with the case. Now we're seeing it with the UK's. But it also shows that vowel breaking precedes schwa fronting. Yeah. So, so that's interesting. And we now that was based on a sharing series and now we look at the rhyme evidence. And we see that the one rhymes with mule and then some other rhyme connections that you can look at that there's no need for me to talk you through. And that allows us to fill in the fifth row, like this. Okay, so so far so good. Now. Yeah, so now I'm just showing where the mergers are again. Yeah. Now we need to address this third row, the, the end case, which I've sort of postponed. So we look for an unambiguous syllable pen in type a syllables, and we look at the she's young evidence. And the, the trouble is, yeah, that we, we don't have as much rhyme data as we want. And I don't feel like I've learned very much from looking at pen. So we're going to look at 10, but the trouble with 10 is that because the pun changes into 10, you know, on its way to Chen, I think 10 is actually not unambiguous. Yeah, so, but we look any case, you know, because what choice we have, and we see these things chin and chin. Okay. And, and we also take a look at Ken, although it. Because there's there's no rhyme evidence for pen. So we look at Ken, which is unambiguous in our current theory. And then we see that Ken rhymes with something like Schwann, and on which is covered, let's say, maybe wasn't clear, but a lot of stop initials are covered by my capital K, in terms of cases. And then we get a topic connection with Schwann. So we've more or less been able to fill in this third row, although you can see why I postponed it because it's like there's the evidence starts to be gappy in such a way that I have to kind of reach in this is for different things. But this is what you end up with. And this is our sort of five vowel hypothesis for type B syllables. Okay, so we have one last hole we need to plug. And it's, there's a reason why, you know, it's sort of stuck out, because it's this change. Ken changes to chin. And here's some session evidence for it, which is you have this chin in 377 a in the same series as Ken, which is 0368 C. And this change is the is called the first palatalization, which is the palatization of dealers before front vowels, it happened quite early, whereas the general palatization that you see across all type B syllables is the second palatialization. So there we've done it. We've filled the last hole. It's a little bit counterintuitive because it's an interesting change. But there it is. So and then you see that that's also an interesting kind of merger. So there we did it. Great, hooray. And we've ended up with a five vowel theory of of all Chinese, because you saw, you know, we started from the type A syllables before dealers, we moved to the type A syllables before dentals. So we came up with this five vowel theory. We assumed it worked for type B syllables, looked at Shesong evidence and rhyme evidence and managed to plug in all the holes. So, so far so good and it turns out there's not a six out theory. There's a five vowel theory. Well, not quite, because there are three middle Chinese type B syllables that we have not yet explained so we, you know, so we started with this grid. With of question marks filled it all in, and we still have stuff left over. So we, so we haven't come up with the right analysis. So we might as well presume a sixth the vowel that gave rise to some cases of trim and chin, and then it will give us a place to put our pinkin and to win. So that so that's, you know, my six vowel. And then I put the six vowel into the chart and this is what I get. Yeah. And then you see there's been a lot of mergers, which is why the, the type B syllables are a little bit messy. Okay, so great. Now we've got the six vowel theory. But then you say well Nathan, not quite you've got a five vowel theory for type a syllables and a six vowel theory for type B syllables. So then we need to act kind of return to the type a syllables and say, Well, what happened to the sixth vowel in the type a syllables. So let's look at that. So we're going to look at type B pin and look at Shesha evidence and rhyme evidence so so we're kind of now you see it's a kind of movement back and forth. We've developed a five vowel theory in type a syllables, we saw that basically worked for type B syllables, but then we needed a sixth vowel. So now we're going to see in the same way that we we did on the type B syllables back on the type a syllables, what could fill those slots in the type a situation. Okay, so we look at what are pins connections and we see there's a connection with Ben type a and then in terms of rhyme evidence, we see that pin rhymes with men and me and rhymes with when and mean rhymes with Den and 10. Okay. And now we move on to the next column in the chart and say what are the type a connections to type B syllables like Quinn. Okay, so we see that. Well we have a Shesha connection with a syllable when and we have a rhyme connection with a syllable hen. So this allows us to basically plug these holes in the following way, which is that as you see so we put pen 10 Ken and Quinn into the the in row. And I have not been able to find any direct evidence that allows for it, but it's pretty clear that if you wanted to just speculate what would go into the sin slot it would be send right. So let's just, well we don't have any counter evidence, but you know it's a little bit ex cathedra but there you go. So this is our six vowel theory in type a syllables so far, and you see that there's been a merger a systematic merger of the in vowel and the end vowel in type a syllables and this is a change that Baxter calls mid to high so type a in changes to end. Now there's still two holes left, which we've been leaving aside for a long time. You see them with the question marks here. The two last holes, I see no compelling direct evidence that tells us how type a ton and sun develop. But at this point, we have a lot of theory internal machinery that we can bring to bear. So, so let's do it. The schwa fronting occurred, if we suspect if we if we propose the schwa fronting occurred in type a syllables, as well as in type a B syllables we get that that turn turned into team. And then we already had to propose the high mid change. So then we would get team changes into 10. Some to turn changes into 10. And then that allows us to fill in these last two spots. Okay, so yeah so here's plugging last two holes. I don't see that there's any direct evidence. But here's the internal machinery. We've already proposed schwa fronting. We've already proposed high to mid. So high. So schwa fronting plus high to mid gives us this change. Yeah. And then that tells us how to plug the hole. I just feel like I need to, you know, take a deep breath and say, that's it. We have, we have, we have done it. We've gotten the six vowel theory in both type A and type B syllables. And you say, yeah, this is now you say, I'm not super convinced it seemed a little bit like, you know, you teleological you you you sort of railroaded us through. And then, especially right there at the end for plugging a few holes had no direct evidence to point to. Well, my Achilles heel in this presentation is the distinction between tin ton and 10. So hopefully that will be clear to people. And that distinction was already proposed by Dwan you'd say, which is this, which is to say, you know, I'm trying to sort of step back and say like, is the six vowel theory right or wrong. So, the weakest point in my presentation of the six vowel theory today is something that has been seen as solved done and dusted for, you know, since the night the very early 19th century. So the, the, so, so now I'm just going a little bit into that history. So the, the, yeah, this pioneer in the analysis of old Chinese rhymes. He treated what are called in the philological tradition. The, the, the, the, the, the, the, the the the one rhyme in the wind run, as one category and then this guy, who I think we've met before. John, it came along and he divided out the you on rhyme. And then one You'd say, I distinguished all three. So in one, you use in size numbering or in the traditional numbering that is still, let's say, it used in the kind of one Lee School. There are 30 Ryan classes for old Chinese, and here you see the 20th, the 23rd, and the 26th, the Zheng, the one, and the Yuan. And in Baxter and Cigar's systems, those categories cover all of the reconstructed finals you see, which is to say Baxter and Cigar draw a lot more distinctions than the Wang Li School does or than Duan Yucai did. But the key place where I sort of pulled a rabbit out of a hat has been considered settled since Duan Yucai's time. There is a question of like, well, why did, you know, why is it that this distinction between these three is so non-obvious for me in this presentation, but was so obvious for the Chinese philological tradition? And I think that's a darn good question. And I started to look into it a little bit, but I found it a little, you know, like trying to get inside the head of an 18th century Chinese philologist who was analyzing the patterns that he saw in the poetry that he was looking at, turned out to be more than I could do in preparation for this lecture. So I'll just leave it there and say, it might seem like this is all some kind of grand artifice, but its weakest moment is a moment that there is total unanimity on in the discipline. And there has been for, you know, like 200 years. Yeah, I could do that. I mean, let's say next time, next year I'll do that as well. Yeah, it's a good point, which is like, I've shown that the six vowel theory works for final velars and final dentals, but I haven't shown that it works for final labials. I think it ends up not being interesting, but sort of in Baxter and Cigar's presentation to my mind, they may not think this is fair, they do a really good job with the velar finals, and then they jump immediately to the Qing phynologists. And they basically, like for exactly the same thing that I needed them for, but they jumped to it much earlier. And they say basically, if you combine the Wang Li system, the traditional view with the Qing phynology, with, you know, these observations, then you immediately get the six vowel theory. And to me, it felt just a little bit too magical, you know, like, ooh, look, six vowel theory. And relied too much on the Qing phynologists. So I was trying to kind of a little bit more dig around in the weeds of the Xie Xiang series and the rhymes to kind of build up a plausibility for the six vowel theory based on the primary sources. And then it turns out that I still have to kind of pull the same rabbit out of the same hat in terms of turning to Duan Yuzai at the end, which made me very angry when I realized that my presentation still had that weakness, but I still think it was kind of worth it in a way that I don't know, that I mean, you can read their presentation and I also feel this way about the presentation in my book, which it just ends up being a little bit too fast, a little bit too kind of, poof, see, it works, yeah. Whereas when I thought about actually this presentation, I thought, well, maybe I should also do the labial finals, but I think there will be some people in the audience right now who feel like, no, no, no, that was quite enough. We don't need the labial finals as well. I mean, let's put it this way, there is no, I mean, I haven't actually gone through and done it all, but the labial finals are simpler than the dental finals. So based on, I took a little bit of a look at it and sort of this is where I'm really relying on my memory of very subjective perceptions, but my sense is the velar finals are dead easy. They're the circumstance that motivates the whole theory, right? The labial finals are sort of intermediate, which is that they're not that complicated, but there's some things, there's some labial neutralization that happened very early that kind of confuses the picture a little bit. And then the dental finals are very complicated. So basically what I did was I tried to motivate the theory by the simple case, and then show that the theory does work in the most complicated case, and then hope that sort of those two together means the system works. But I agree with you that sort of a good thorough presentation would go through all of the cases. I mean, one thing that I think people will have suffered through in this presentation is that it's a very sort of interlocking system where you kind of need everything all at once. And as a consequence, if you really systematically presented we're going to do the six-valve theory in all kind of phonotactic circumstances, you would end up having to look at every sound change in the history from all Chinese to middle Chinese, but actually that's probably a good way of presenting it. All of Chinese historical phonology as a motivation for the six-valve theory. That's actually would make a nice little book. Little book, yeah, but I haven't done it.