 So the first thing is, nobody suspects me of being a phonologist, and therefore I need correction after my talk. So the discussion will take the form of instructing me on my mistakes. Secondly, I would like to suggest that the old Chinese is described in great detail in two books that are missing in the new reconstruction. I believe that Hua Xue Cheng, Zhou Qin, Han Jin, Fang Yan, Yan Jiu Shi, is a uniquely rich semblance of all the material that should have been considered, I think. And this book would be by far the most rich source for it. Even more striking is, but of course not available at the time of writing. The new reconstruction is Wang Zhi Ping. This is an extraordinary thing which I recently received of terrific interest. So it is Chu Tu Wen Xian, Yu Xian Jin, Liang Hai, Fang Yan. That is a very, a thing which then goes systematically beyond Hua Xue Cheng, who is a student of our good friend Liu Guo Yao. And Liu Guo Yao has written a terrifically useful summary of Hua Xue Cheng's findings that I find in itself is a small monographic contribution and an introduction to this work. So Wang Zhi Ping says, This I think puts my point, I don't have more to say, and I think that more of this kind of material might have been recommended to readers, if not used in the book, if you see what I mean. Then we have Fang Yan, which is even by my teacher and friend Liu Guo Yao, who wrote extensively on it, interpreted as a local language and so on, as I understand him. And I think that is a misreading of the ancient Chinese, Fang Yu, does exist also and would be a language, and Fang Yan are actually local words. And Fang Yan is therefore a list of words and not a description of a language. And I say probably, you see, I have discussed it with many of our linguist colleagues, and there is actually a very general suspicion that we have read new Fang Yan into old Fang Yan in two wrong ways. One is not any more current, which is to simply say Fang Yan in the 1920s is dialect, which it is not, is local language. But that local language itself is again a mistake, because it has language which would be Yu and not Yan, so that is an idea which may be wrong. But so we now have some other issues here. I have two indispensable works on morphology, which I think if we look at what we are told about derivation by tone change or by S-addition suffixation, if that is the solution we want to subscribe to that everybody shares. Then Yu Wen, and especially Xie Weiwei, Han Yu, Yin Jian, Goi Zi, Yan Jiu, are a wealth of materials that should have been, I think could have greatly changed the description of S-addition suffixation in classical Chinese, I must say. I have studied this for decades. I learned a furious amount from these books, because these were theses by students. Sun Yu Wen actually got many prizes for this book, and I think that Xie Weiwei richly deserves similar prizes for his thesis. These are two young scholars, I don't think they're much more than 30 or something, and I think they're doing excellent work. And I'm just trying to be useful. I think it would help to at least remind readers of our new reconstruction that there are these works that tell a great deal more about S-addition suffixation than is in the book. Then there is a very nice statement by Zhang Xinhu. You see Zhang Xinhu, Gu Ren Wu Yunshu, Shi Zhiyun, Ge Suiqi, Fang Yin, or Shuyin. They each follow the Fang Yin. And I miss the expression Fang Yin in the new reconstruction, because I think it is correct, as Lin Yu Tang learned from August Konradie in Leipzig when he wrote his thesis there, that the phenomenon of Fang Yin, local variant pronunciation, is crucial for the kind of unneedness that was mentioned a moment ago in the rhyming. And of course, Lin Yu Tang in his Lu Nguan Ji, Yuan Lu Nguan Ji has, I think, five or six important articles on Fang Yin, Shui An. So I won't bother you anymore with this. I think it's just useful to remember. To my mind, it is useful to remember that there is this extensive research on Fang Yin as opposed to Fang Yin, which might have been brought to bear. There is a question I have on the very term reconstruction. Wang Li prefers to be nice, and he uses nice nitsa, literally, a draft guess to Chong Yan. And I just want to, since he's a colleague of mine in Bedain and so on, and I've known him and so on, it is very interesting that he sounds very dogmatic, but in fact I have forgotten to put in the quotation. But he puts very explicitly, he rejects the term Chong Yan and dislikes it and prefers nitsa, and thinks that it is, if you are thinking of your things as hypothesis, it's a very good idea to call them that. You see, to use a terminology that does not have this kind of way, I found Wang Li's comments here really rather moving, because he really felt that nyi and sir, yes, you see, yes, unfortunately in a language like Chinese, with the kind of data we have, we cannot move much beyond this, is his thought. The next point I have is that Middle Chinese is said to be transcribed, and I think that Egerrod already in 55 put it bluntly and correctly, only the modern dialect will provide material for reconstructing actual values for many elements such as initial and so on. This is uncontroversial and moreover, what we reconstruct depends on what dialects we look at, and on many, many other things, especially of course a very long history of reconstruction of Middle Chinese that we can trace. Even Shakespeare's pronunciation was reconstructed. I was going to carry to you, Helge Kökeritz Shakespeare's pronunciation. I Wang Li recommend that if you are going to speak of the philosophy of writing, please read this book of the great professor of Yale, Helge Kökeritz. He became very famous in Shakespeare's studies. This is a classic. It's a very beautiful book. I wish I had taken it. So there you get problematization of, for example, computability of what rhymes and what doesn't rhyme. You see, of course rejected because of the conventional nature of rhyming conventions that are cultural and that can actually link very distant things as was mentioned a moment ago, and as is very common in German and so on. So I just mentioned it. I hope you don't misunderstand me. I'm not saying this should have been in the book, but I am saying that we would be talking about making better sense about the problems of rhyming in literature. If we had in the backs of our minds this fantastically readable book, it is riveting. I'm not a phonologist and I have read large swathes of this book with great pleasure. And so there it is. The very substantial evidence there is for performing his reconstruction has been found radically insufficient for the task. So no magic formula exists by means of which we can single out I rhymes in Shakespeare. That's a harsh statement. And I assure you we have lots and lots of books about the pronunciation at the time and still we cannot in literature be sure. Now I assure you that we could discuss rhyming in Shijin at a much better level if we actually have a commonly reading books of this order. And if we used Wang Xian, Shijin Yunpu, which of course is 2002 or something, and which gives you the 249 rhyme schemes in Shijin and so on and has a remark which I find very helpful. It would take more than his 500 page book to justify a scription of rhymes in each case because it is in fact a literary decision also. It is a philological decision. It is not a computational matter. So that is what he thinks and I must say I just do a permit myself to agree. So the Middle Chinese transcription in the new reconstruction basically the same as in bills explicitly replaces Kagan's phonetic reconstruction with an attempt at scornological reconstruction. I will come back to that. Take Xian first and Xian immortal where Bill, by the way, has my sympathy for not respecting the Kagan's opposition. And this is an explicit refusal in the new reconstruction to transcribe what Lufa Yan tried to register in the spirit of the famous dictum which goes like this. It goes like this. Po, si, ho, li, fun, bia, and then shu, lei. Don't ask me. But it does mean what I hear say apparently. It's a very curious thing. I find it almost a teaser. I don't know. But it does mean that and everybody agrees. And this is the phoneticism of Lufa Yan and it is converted into a phonologism explicitly because in fact it is of course a very honest, honestly done. But you can't at the same time replace a phonetic analysis with a phonemic one and say that you are just transcribing what the other guy is doing. So we have Zhizin, you see, which of course is what Lufa Yan would transcribe the word Zhizin for. And standard Russian orthography is phonemic here and follows a different strategy which is very fine but certainly not a transcription of the early sources. So Lufa Yan would definitely try Zhiz to live as Zhiz because that is how it is pronounced. You have to have some operation in your throat to say I can't do it. If you are down there, you want to say you are very clearly saying Zhiz before you have said good morning. So Ryan then which is taken by Bill to be one of the important arguments for transcribing them as the same. It goes nowhere to prove phonetic identity or closeness. It's very obvious, not very comfortable to remind you of that but it's very clear. And in any case I have to say that Lufa Yan knew this and Bill and Laurent as I see it did not quite realize that this is not an argument. And if they are going to transcribe Lufa Yan's effort then they must transcribe this as Zhiz and not as Zhiz, which is what I do. So that's how I take it. In Middle Chinese we have one phoneme A with three allophones and we have one E with three allophones and one with two allophones and so on. And all this raises this question everywhere. We are told on page, I don't, we have it here. We are told that about the inadequacies of Karlgren's phonological reconstruction, which made it difficult to identify patterns involved. I think I want to go quickly over this. Let me just remind you that Karlgren never used the word term chronology in the Schubertz coin sense except to malign and to in fact reject grotesque with grotesque enmity. All manner of phonology. He was an anti-phonologist of the first order and my dear teacher, Sir Ern Egerwood had his thesis not actually not examined by Karlgren because in his thesis he mentioned the word phoneme. That is the story as Egerwood had it. So there is no such thing and the word phonology used before the Prince Troubet's coin does not refer to Troubet's coin phonology. So it is true enough that Karlgren wrote a book on phonology, if you like, on phonology, but it is a grotesque misunderstanding of one of the main features of Karlgren's work. And here is Troubet's coin. The Prince, whose brother, by the way, I think died in a Ferrari accident in Paris. So this is a really quite a family. I'm sorry. It's not relevant. So then I have something to... I really think this is not very important about Karlgren's phonology, but it is a very major part of the history of the study of old Chinese, certainly. Old Chinese and proto-Chinese indistinguishable. It is difficult at this stage to make a meaningful distinction between old Chinese and proto-Chinese. This, I believe, is a serious category confusion. Unlike Indo-European, a real form of speech current on the Eurasian or Anatolian grassy steppes a few millennia ago, the reconstructed proto-Indo-European parent language is not a concrete historical fact. Rather, it is a very recent hypothetical construct or an abstract scheme, a set of formulae from which properties of the extant daughter languages can hopefully be derived more or less successfully. To say that these are hard to distinguish is unfortunate if this is the right interpretation. But of course it may not be the right interpretation. So I go on, first, to say very clearly that Louis Jemslieu, Omkring's Paul Theorien's Grunlegelsen, also known as Proligomena, and both are equally unreadable. I do not recommend Jemslieu as linguistic prose, neither in Danish nor in English. But he had a very clear vision of this. The Indo-European is, in his terms, a semiotic, and proto-Indo-European is a meta-semiotic. So one thing is a semiotic system, and the other is a description of it. It is about it, hence the word meta. And so not be able to distinguish between these two would be, to be seriously categorically at... Stephen Colvin, my dear friend from across the street, says, this is a basic category mistake. All Chinese is an attested cultural artifact, and proto-Chinese a fairly dodgy construct. He put it more... well, he is more English than I am. He put it more smoothly than I would have been able to put it. So perhaps then the newer instruction intends all Chinese here to be used in the narrow technical sense, referred to the earliest stage of Chinese that we can reconstruct from Chinese evidence. That's a possibility. But in that case we have a plain tautology. The earliest stage of Chinese we can reconstruct for Chinese being defined as being exactly proto-Chinese. They would be telling us essentially that they find it difficult at this stage to distinguish between proto-Chinese and proto-Chinese. If that is the interpretation, we can discuss this. I'm just desperately trying to understand what this can be meant. This cannot be meant. So then thirdly, what they intend is two historical phenomena. One that is attested in the early Chinese texts and the other imagined as a concrete manifestation of the proto-Chinese system. They reconstruct as the common ancestor of all later varieties of Chinese. In this case, now in this case that does make sense and if true it would be important that these two could be shown by compelling evidence to coincide. And that is an open, it is an issue we can discuss. But I think if that is the meaning it could have been expressed a little less misunderstanding. So, but our earliest texts, the oracle bone inscriptions dating from the 14s, according to Chiusi Gui, whatever, I don't care, throw relatively little light on matters of phonology, certainly not enough on their basis alone to reconstruct phonological system underlying them. For this, we badly need the much later oaths. David Keeke puts it in his inevitable spirited way. He was a journalist, you know, he grew up as a journalist. The inscriptions tell us so little about their sound that the problem of the pronunciation of Shang graphs has been declared near insurmountable. And Takashima yesterday concurred. And so there we go to conclude on this reading the new reconstruction statement would at least make sense. Unfortunately, it would be untrue, but at least it makes sense. It is just not right. So I think I've said some outrageous things and displayed enough of my ignorance of phonology for corrections to be in order. We have how many minutes do we have for discussion?