 So, welcome back, and today we'll be talking about the pre-initials or consonant clusters, onset consonant clusters of all Chinese, and I'll start with some general considerations of whether or not it's plausible to reconstruct consonant clusters or so-called pre-initials, and then move on to some evidence that makes them plausible. And then in separate lectures, I'll look at the concrete evidence for specific proposals like pre-initial S or pre-initial T. So, this lecture is just more about the general approach and the overall plausibility. So, let's begin. Baxuan cigar reconstruct the following pre-initials in old Chinese, PkTm capital N, which is a prefix that assimilates in terms of place of articulation to the following consonant, and finally S. And I'll just throw in here that I have a sort of terminological objection to pre-initial because, you know, initial should mean the beginning. And so a pre-initial means before the beginning, which is a kind of paradoxical sounding, but clearly it has to do with a phonotactic theory of all Chinese. But I wonder whether this terminology is used in other research traditions if there are languages that have this kind of phonotactic structure. I think we should engage more with those traditions and see how they describe this. But here in Sainte-Tibetan linguistics, this is the conventional term pre-initial. So each of these pre-initials occurs in both a tight form, that's to say, you know, segmentally immediately adjacent, which they annotate as, for example, T dot K, or in a loose form where there's a schwaal vowel that interposes itself between the pre-initial and the initial. So with R, there are actually three options. There's the T-initial followed by the R-medial, the T-pre-initial followed by the R-initial, and the T-loose pre-initial followed by a schwaal followed by the R-initial. So now returning to this discussion of T-R, Baxter and Cigar don't discuss what phonetic distinction they see between T-R and T-dot-R, although they develop differently into Middle Chinese. So there must have been a phonetic difference. Now, one thing I want to draw your attention to is that there have been some changes in the treatment of pre-initials in Cigar's work from his 1999 book to the 2014 book together with Baxter. And the change in approach, I guess, implies a difference in how they're treating certain kinds of evidence. So if we look at the word armpit in 1999, Cigar pointed to the Cantonese word for armpit and the Fuchou word for armpit and said, well, look, both of those have some kind of minor syllable or pre-initial. So he reconstructed that also for old Chinese. But if we look at the 2014 reconstruction, they don't reconstruct it that way anymore. So they're actually quite agnostic. That's what the brackets mean. The brackets mean that they're tentatively suggesting the solution that's in the brackets, but that any solution that yields the same result in Middle Chinese, they wouldn't necessarily be unhappy with. And I think that this is a sort of, let's say, a murky area or something that's quite difficult for users of their reconstruction, because it's all fine and well to know that they are a little ambivalent here, but they didn't reconstruct this word with a K initial. So there must be some reason they think it had a U-view or initial. And that is a little bit out of focus. And another way to put it is, it's not clear what these different levels of their subjective certainty we as users of the reconstruction are meant to do with. So how are we meant to interact with a reconstruction differently if it has these brackets or not? In any case, the point I want to draw attention to here is that apparently the Cantonese and Fucho Evidence is no longer probative in terms of their reconstruction and presumably they arrived at the decision that that wasn't significant anymore to them through some process, but they don't make that explicit. So I mean, that's understandable that you can't write down everything you think in a book, but as a user, I've found it difficult to sort of look at the 1999 book and look at the 2014 book and not really know kind of where you are in terms of similarities or differences there. Okay, so yeah, so where I am still then is saying kind of the kinds of reasons why you might be skeptical about the solution of pre-initials in Baxter's cigar 2014, which include that they seem to have changed their minds without telling us why, yeah? So in 1999, Cigar's working assumption is that the pre-initials are all morphologically meaningful and appear in all of them in two forms, tight and loose, as I described earlier, and that basically you get in terms of root structure, you get plain roots, roots with a tight pre-initial and roots with a loose pre-initial. And he says that these three types of forms perhaps existed side-by-side in all Chinese as stylistic or social variants. And now I just want to point out that this is methodologically a problem for those of us who are committed to Neo-Gramarian principles. And the reason for that is that you can't, you know, it's a total wild card. You can't have a sound change that's unpredictable. So, you know, I mean, some of you will say, well, but come on, these things happen. There are sociolinguistic variants and stylistic variants. Yeah, but what I'm saying is that like there's just something a little bit, I don't know, it really is using a wild card to say, well, this word developed like this and this word developed, the same word developed instead like this for sociolinguistic reasons. Yeah, you would want strong positive evidence of that rather than assuming it. And, you know, let's say altogether the evidence for pre-initials is pretty weak. But Baxter and Cigar's system embraces more types of evidence than other reconstruction systems. And their system is sort of informationally richer, which is to say, like if I want to change Baxter and Cigar's system of old Chinese reconstruction into Schussler's, it's quite easy to do. Now, that's partly because Baxter and Cigar and Schussler are both in the six-vowel camp. It's quite hard to move between reconstructions in the six-vowel camp and those not in the six-vowel camp. But let's say if the disagreement is just about initials and one system has more initials and one system has fewer initials, then I guess what I'm saying is I feel like using the system with more distinctions is better even if it's wrong because it's just providing, if it's wrong, it's just providing redundant information that doesn't really get in the way. Yeah, it's like just adding diacritics for no reason. And if it's right, then it's right. So, and then just to kind of talk through the converse, if a system has two few distinctions, if it's wrong, then there's no way to fix that. So I think it's better to work with a system that has more distinctions even if some of those distinctions will turn out to be incorrect. And that's why I think particularly for comparative linguistics Baxter and Cigar's system is better because maybe some of these distinctions that they make will correspond to phenomena in other languages. And you wouldn't notice that if you didn't allow yourself to draw the distinction. So just for, you know, for reasons having like as a research program, it's better to work with an old Chinese reconstruction system with more distinctions. Okay. Now, mostly so far, I have talked about reasons to be, you know, a little skeptical, some problems that I see, the fact that there's not a lot of evidence for pre initials. But now let's look at the strong reasons, the kind of best evidence for their being for their being pre initials. So, particularly, perhaps uniquely in the charging, there are cases where two characters are used to render a single word. So I'm going to go through these, you know, quite slowly. So in the first type, we have a prefix ma. So let's see it owed 235.5. We have this phrase, which at face value means do not remember your ancestors. Well, if if you know anything about traditional Chinese culture, this will seem like a very surprising sentiment to express. And let's say that interpretation relies on the fact that this ma is a common negation marker. It's one of the top two negation markers. But the Mao commentary, which is the earliest extent commentary that, you know, that is, let's say fundamental to interpreting the oaths, has this remark that says, do not remember means remember. So how do we understand that, you know, it's saying ignore the ma. Yeah. So Baxter and cigar interpret this ma as a loose prefix that makes this verb nim to think of volitional. And this, you know, it's not just for this occasion that they propose this, they believe that in terms of old Chinese morphology, there was this ma volitional prefix. And so they think that this prefix was added to make this remember volitional. And that's because you can't command someone to do something unless it's a volitional verb. Yeah. So that's their interpretation of this passage. Now let's look at another one. This is 0256. This phrase seems to mean not strong is this man, the four realms comply with him. Well, that doesn't make any sense, right? If the four realms are complying with him, then he probably is strong. Yeah. Yeah. And the Mao commentary says not strong means strong. So here Baxter and cigar think that this ma spells out the M prefix in the word strive, compete, which they reconstruct for other reasons. And now I'll just say a word about why does this phenomenon, why is it restricted to predominantly to the to the odds? And I think there's a reason that that could be explained, which is that the odds are poetry. And they're very, very old. So imagine or I think it's useful to think about French and how you sing French. And, you know, I don't know whether everyone will know this, but and, you know, if I had prepared better ahead of time, I would have a French song to hand. But if you listen to basically any French art song, but even in some pop music today, you'll notice that the the schwa vowels that have all disappeared from spoken French are still pronounced in in music. And and partly that has to be because I mean, I'm not saying it's the only factor, but it has to be because the tune wouldn't work. If you if you know if you have a note that you're supposed to hit on one of these syllables, you need to have the syllable there. And I think that that's, you know, you can understand that as partly an explanation for, you know, the the odds are the kind of genre in terms of, you know, folk songs, let's say, folk songs that are kind of ossified as traditional literary monuments are the kind of genre exactly where you would expect these these minor syllables to sometimes be preserved for metrical or musical reasons. Okay, so now just to point out that similar phenomena happened with pa as well as with mo. So again, it looks like negation. And this one is, you know, trickier, which is to say I go straight, you know, so far I've said like, oh, it looks like it says this, but it actually says this. Well, in this case, it's really hard to make sense of it at all without the commentary. So so we have these Joe are greatly illustrious, God appointed them, where there's this, there's this put, which is just quite hard to get any it means like not time or something. It's, it doesn't make any sense. So the mild commentary says, it says that not means, and means this other, which is to say that the mouth commentary is saying, but means this. So, you know, how, how should we interpret this, it looks like, you know, the mouth commentary is saying maybe that this pud is somehow the same word as dead. And that's, you know, evidence for some kind of minor syllable. Okay. Now, this one is my favorite, it's in some ways better if you read the whole poem. And that's in my book. Because from this line of poetry alone, it's not clear that it can't be negation. But certainly in the context of the poem overall, you expect the line to say, attentive and filled, rather than not attentive and not filled. But anyhow, you can look in my book for the details of that. But I just want to point out that in this case, the word for filled then is, it's at the very end of the line, you can see it in the opinion, prolonged. And this perfectly matches the, the Burmese situation. So, fill in Burmese is, I don't know, you know, it would actually be pronounced in modern Burmese, but let's say something like plan. Well, it comes from pling in, in Proto Burmese. And, and actually, Proto Burmese merged eng and eng, we don't need to get into those details. But in any case, it's really a perfect match with the, the old Chinese reconstruction. And, you know, some people would say, and actually, Georgie Orlendi in his review of my book says like, okay, yeah, it's kind of good evidence, but this is maybe a coincidence. You know, my own feeling is, is actually, this is just like, you know, this is good, this is basically as good as something like Veda in Sanskrit and Oida in Greek, you know, you can't make stuff like this up. So, so I think this is very compelling evidence that in some cases, old Chinese had pre initials. Oh, that's just a methodological issue, which is, you know, I mean, let's credit this to me, I don't know if that's fair or not. But let's say, if I'm just looking around the world and saying, oh, what are related languages? If I look at extremely specific morphological idiosyncrasies, that is strong evidence of, you know, not a coincidence. Right. So let's say the fact that the fact that Greek actually, I mean, what's interesting about the Veda Oida case is, is even saying what's surprising already shows you that there's a lot of, you know, shared kind of categorical similarities, but both languages use an unreduplicated perfect for to see meaning to know. Yeah, well, that's, that's actually, I mean, some people I've, I've discussed this with some people who say like, Oh, come on, using some kind of a perfect of to see to mean to know is just dead obvious. Right. Like, that doesn't mean anything about how related languages are. But I would say, well, the fact that they're under it's an unreduplicated perfect. That's the part that's a kind of striking coincidence. So I was maybe I was sort of overstating it, but I think the fact that old Chinese and Burmese both have a word, Phil, that's that has the same morphological structure. And so they both go back to something like playing is, you know, quite good evidence. Let's say on the one hand that they're related languages. And on the other hand, that that P in old Chinese is not some random thing, you know, which is to say, if you like, who's my interlocutor, the people in Beijing will have a story to tell about that Boo character in that poem. And usually they say that all these things are sort of rhetorical questions. And you can use rhetorical questions to mean the opposite. So, we can understand it as a negation marker. You know, they actually meant, like the commentary said, they say, not full to mean full, you know, in a kind of that's, I think, I mean, I don't really know this literature, but I think that's those people who really want to deny that evidence say that. And then I would say, look, I don't think it's a coincidence that this, this P is there in old Chinese and in old Burmese. Okay, so now we're moving outside the oats. There's less of this type of evidence, but there is some. The show, which is a character dictionary from the early Han period, says about this character that means brush. You know, it defines it as that with which one writes. And then it gives some comments about dialectology. So, in Chu, it's called Mrit. In Wu, it's called Prut. In Yan, it's called Put. And in Qin, it's called Prut. Yes, Etienne. Sorry to interrupt again. I was just wondering, so do we, do we think that these were actually, these pre-initials were actually one syllable? Or could we not consider that they just indicate maybe a consonant? And because of the writing system, I mean, they did not have any choice but to write full syllable. So they would write a shva. But maybe, so are there people that think that it was not Prut as this little bit, but like really and like, how could we know whether it was a dyselabic or a monoselabic pronunciation? I mean, maybe you weren't about to go. No, I was not going to, let's say, it's a good question. And we certainly know from, I mean, some of you will know this kind of stuff better than I do. I think in the, in in Acadian or in Sumerian or, oh, actually Mycenaean is, yeah, Mycenaean is a good example. Some of you are taking Mycenaean. They don't have the resources in the Mycenaean script to write consonant clusters. So they write them with, kind of in this way. That's the sort of thing you have in mind. Yeah. So I would say, I don't think anyone has discussed this. And it's clearly, let's say Baxter and Cigar have this clear theory of, you know, their tight pre initials and loose pre initials. And that theory helps them do a certain amount of work. And that that's what we'll see kind of in today's presentation. So if you wanted to come up with an alternative theory, it would have to sort of solve the same problems. And, yeah, and then what I would say is kind of at face value that analysis of like, oh, maybe these, this wasn't this these minor syllables, when written out only indicates consonant clusters and not actual syllables would make particular sense in those cases where, you know, someone is trying to write a foreign language or a dialect form. But which, you know, is the case here. But I think like, for the odes, it wouldn't work at all, right? You know, because the odes would have been probably sung. So each character probably really did correspond to a syllable. Anyhow, that's what I would say. I would, you know, it's one of many topics that it would be nice to have more systematic study of there. There is work in Chinese. But I would say actually that most Chinese scholarship of a philological or paleographic variety uses Wang Li's reconstruction as his pointed departure. So it's hard to make interact with kind of today's historical phonology. Although hopefully this is there's some evidence that this is changing right around now. But anyhow, so, so here on this slide, then we have, we see variation among Put Proth and Porut. And that is pretty good evidence for, you know, constant clusters in in old Chinese, although precisely because it's about different dialects, it's not very good evidence of a tight versus loose pre-initial, which is sort of what you were saying, actually, in any particular variety of Chinese. But here is kind of my favorite kind of bit of evidence along these lines. And for one reason, it's my favorite, which I'll talk you through is it actually works with Baxter and Cigar system. So for example, that playing that I discussed matching Burmese perfectly, that's not actually a case that Baxter and Cigar see any need to reconstruct the P initial in terms of their own system. And in terms of Chinese internal evidence, it's really just the evidence of the Schroding itself that points to that. Or to paraphrase what I'm saying, one might believe that old Chinese had pre-initials and and that those pre-initials distinguish tight and loose, and still not think that Baxter and Cigar reconstruct them in the right places or for the right reasons. And this example is the best case of, let's say, everything coming together nicely in the Baxter and Cigar system in terms of different sources of evidence. So let's just go through it. This is a comment by Mr. Yan Jitue. So he says, I have exhaustively visited the Xu region and they pronounce this word, grain or particle, let's call it lip, as pick. But at the time, they had no way to explain it. And then I said, well, in the San Tsang and the Shouen, that's the Shouen Jitue, which we just looked at, this word is written as this character and this character together. And I show you the character, yeah, and is glossed in both cases with the definition, you know, grain or particle. And in this other book, it's given the pronunciation pick. And they were all delighted to discover this. So what, yeah, so what is this saying? In sort of mainstream Chinese, this word would have been pronounced lip, but in this dialect, it was pronounced pick. And there is philological evidence for pronouncing it as pick. And, you know, I don't know how we should reconstruct it then, but presumably with some kind of ple initial cluster. Okay, now we look at proto-min and we see that it has a softened initial. And softened initials in proto-min are the reason that Baxter and Cigar reconstruct or let me put that differently. When min has, when proto-min has softened initials, they reconstruct tight prees. And I'm not, actually, I'm not quite sure I'm getting this right. So we should check my book to be absolutely sure. But in case the min form supports the reconstruction in Baxter and Cigar system of a pre-initial before erotic. So in this case, we have kind of first hand, you know, ancient dialectical field work that, you know, slots in nicely to Baxter and Cigar system. And that's what I wanted to say, which is to say, I think this is the best case of showing not only that Baxter and Cigar's kind of overall system is right, but that they're in fact right in their concrete proposals of what correspondences to reconstruct in which ways. Let's say from within the Baxter-Cigar system, you would say they are loose pre-initials. I don't feel comfortable calling them constant clusters. But it doesn't seem like at that time you had sort of properly disalabic words. And what do I mean by that? I mean, like the cases of two character words in the Ode's are very rare. And those that do occur either have Wu or have Bu as their first character. So that indicates that, you know, the extent of dye is syllabicity is quite limited. And in Baxter's 1999 book, he refers to this kind of syllable as a iambic syllable, you know, because an iamb has the stress pattern, the th, right? So it's kind of a minor syllable and then a major syllable. James Matasoff refers to them as sesquise syllables, which is like to mean like it's sort of half, like one and a half syllables. So between monosyllable and disyllable. Personally, I think that sesquise syllable is just kind of hideous term. But you know, that's not a criticism as an idea. And this kind of, let's say, this kind of phonotactic pattern is not at all unusual for Asia. Many of the, particularly many of the Vietic languages have that kind of profile. And I think, I think Taikudai languages, this gets into stuff I don't know much about. But let's say that kind of phonotactic profile is not at all unusual in Asia. Yeah. What to say? Just a clarification, the claim that sort of the coexistence of the same morpheme in variants with tight and loose preinitials is explainable by socio-linguistic or stylistic variation is a claim only made by cigar, as far as I know, not Baxter and cigar. And it was made in the 1999 book, not in the 2014 book. And you know, I kind of do they, does cigar still believe it? Does Baxter believe it? You know, we have no way of knowing other than asking them. Yeah. So that's just a kind of a clarification around the history of the discipline. And what I would say is I think the Shouenjie presents quite good evidence that there was dialectical variation of this type. I would really like someone to look at the Fanyan. And I'll just explain a little bit about that for those of you who don't know, you know, all these old Chinese books. But there actually is a work from ancient times that is a work of dialectology. And it includes, we think, both Chinese dialects and non-Chinese languages. And I mean, it's a famous work. It's been extensively studied. But it has not been studied as far as I know by anyone working in the six vowel tradition recently. And the last time it was looked at by in the sort of Anglophon scholarship was in the 1960s. So I would really love it if someone would look at the Fanyan as with an eye on this question is are there dialectical variations between tight and loose pre-initials in the Fanyan? I think there likely will be. But the thing that I want to say right now is my point is showing you those slides actually wasn't about is it sociolinguistic? Is it dialectal? It's just to kind of build a sense that reconstructing pre-initials in old Chinese is not a crazy thing to do. And maybe I should have just kind of contextualized that anyone associated with Beda with Beijing University thinks it's just madness. It's just like absolutely only lunatics think that old Chinese had consulate clusters. So that's why I wanted to just sort of show you that in really strong like really solidly sinological evidence, there is good reason to think that there are pre-initials in in ancient Chinese. Of course to be fair to our colleagues at Beda, they know about this material and they have other ways of analyzing. But let's say that's not what this course is about. I just wanted to start with kind of in the way I've been trying to do with all things, let's look at really what is the evidence in the primary sources that's motivating these theoretical ideas. Oh yeah, okay, I think that's a totally fair characterization. But I would object a little bit to saying that it's unusual because all Chinese, you can think of as, you know, and all Chinese isn't a tested language. So we're not doing, we're not reconstructing a proto-language like proto-Germanic here. We're basically figuring out like, you know, it's more like figuring out how Greek was pronounced. It's just that the Greek alphabet is much more transparent than Chinese characters as a way of recording, you know, phonology. And I think if you looked at, you know, the literature about how Greek was pronounced, they will look at exactly the same kinds of evidence like, you know, what does the organization of the script itself imply? How was the script adapted for other languages? What are sort of meta comments in the philological literature from ancient times about the pronunciation of words? And then you'll also look at, you know, loan words into other languages and very occasionally at maybe, you know, modern Greek or something like that. So I think actually if you think of all of this as a philological project of understanding how an actually a tested language was pronounced, there's nothing unusual about these approaches. It's a difference of degree, not a qualitative difference from the kind of work that's done in interpreting philological sources on any ancient languages in any research tradition. Which is not to say that, like, there's another project that work has been done on but not enough work, which is to just reconstruct the ancestor of the the synodic languages. And that would give you the equivalent of, you know, proto romance, if you like, you know, so but but reconstructing proto romance and interpreting the philological evidence in phonetic terms of Latin epigraphy, for instance, are different. Yeah, I mean, I sort of the, you should maybe just read his book, but I'll tell you the kind of crux of it is if he wants to he wants to look at roots, right, rather than sort of saying like, oh, you know, here's a character and here's how was pronounced middle Chinese and how was it maybe, how was I mean, these things get actually even hard to say rigorously, how was the word used, no, sorry, how was the word, the ancestor of the word that would become this middle Chinese reading of this character, pronounced in all Chinese, that's kind of what Chinese historical phonology does, generally speaking. Instead, he wants to look at like, you know, all Chinese is not just a collection of syllables, yeah, used in reading practice, there, it was a language, which means that, let's say, words that have similar pronunciation and related meanings, probably are exhibiting morphological behavior. And so he prop posits a kind of a theory of root structure, which is basically that, you know, Chinese has roots, and then it has affixes, and this is how you build affixes from roots. And he, and then he does some case studies about, you know, how productive that theory is. It's clear that the 2014 reconstruction is, let's say, very inflected by that period of his work, but the 2014 book does not present a similar kind of root theory. And in fact, there are cases, I won't be able to name any right off the top of my head, but there are cases where they propose stuff in 2014 that would have been strictly forbidden in the 1999 framework. So, you know, so I think, you know, there are two questions that we can say kind of now in 2021. One is like, what do Baxter and Cigar think about root structure? And, you know, for that you can ask them. I think it's not knowable, based on their publications. And then the other one is, you know, where are we in terms of bringing together morphology and phenology in the study of old Chinese? And the answer is, well, we're not much further along than we were in 1999.