 Today, we'll be looking at Outstanding Problems in Sino-Tibetan Comparison. It's kind of a necessary scientific task, and we'll give you a sense of, you know, where the field is, what we've managed to achieve, and what is still staring us in the face as problems. So, sometimes it's easy, and I won't be discussing these cases. I give paragraph references to my book. Either the three languages have the same thing, then we can reconstruct that thing, like in the case of Final K, or there are straightforward correspondences that point to a merger in one or the other language, like we get in the case of Glottostop or Q. In some cases, the resident files are also pretty easy, including these reconstructions I feel less comfortable with, like the RL reconstruction, but where the correspondence is clear. That one, this RL, is the most common of the resident correspondences, which is maybe a little bit concerning, and the reconstruction is probably wrong. Okay, as we've mentioned before, in some cases, distinguishing Final R and Final RL is hard. In particular, there are two examples. In the first one, there happened to be almost identical words for neck in Tibetan, one that ends with an L and one that ends with an R. And in the second case, it's unclear whether the Chinese word should be compared to the word fare, or the word clear in Tibetan, and which comparison you choose will determine which final year we construct. Of course, maybe this points to some, I don't know, morphological process in Tibetan that needs to be looked into, but with only two examples, the chances of doing that successfully at the moment don't seem super hot. All right, now turning to vowels, we can get back to a six vowel theory pretty easily, more or less saying the proto-language was a lot like Chinese. So according to handle all languages, but Chinese merge ah and schwa, but this is not correct. Burmese changes schwa j into i's, changes i into i, and this is a distinct outcome from its a-j correspondence. Here's should be a-j, I've made a typo, but you can see that it's a-j. And note that Tongut also draws this distinction between ah and schwa in limited context. So here's the general pattern, which is something we call brightening. It seems to have happened late and independently in a bunch of languages related to Tongut. But here's the general pattern, so you have z for i rather than za in Tongut. But schwa changes to u in words like ear and love. So that means not only Burmese but also Tongut has some evidence of the schwa versus ah distinction, which means that Handel's idea is wrong. So one of the most interesting, I think, problems in sign of Tibetan is this nga ka variation. And these, I think, it's worth actually talking through every example so that you kind of have them in your mind. So for this word that means war or conflict, Burmese has let's say proto-Burmish has zik, as Tibetan points to something like zeng in the in the proto language. Similarly for heart, Tibetan and Chinese have a clear ing final, but Burmese has ik and for year, Tibetan and Chinese have ing, Burmese has ik. For wood, Burmese has ik, Tibetan and Chinese have ing, and again for new. So these are the examples I've collected. There aren't very many. And in all of them, Tibetan and Chinese have ing and Burmese has ik. So it looks like a kind of Burmese innovation. But as we talked about a couple of days ago, achong, which is closely related to Burmese, it has this word sang for tree that agrees with Chinese and Tibetan against Burmese, even though it's a language that's clearly closely related to Burmese. So this shows that this this alternation, this ing, ik thing existed at the level of proto-Burmish, or it's a loan word from Jingpo. Those are the two options. And I just want to point out that there's this little language, by long, only preserved in three poems in the Ho Han Chu from the very early part of our era. I think around 50 AD, I would have to check though. And in this language, the word for heart agrees with Tibetan. So you have a Ning for heart, but you have sick for wood like in the Burmese languages. So I think this evidence from By Long is extremely important because it shows that we're not going to be able to solve this problem with sound change. It's not going to be that like, oh, Burmese changed ing into ik in some environments. That's not going to work. The alternation has to be reconstructed at the proto-Sino-Tibetan level because it exists in proto-Burmish possibly. It exists in By Long. And then we can imagine, I think that in all of these words, there was some kind of ing, ik alternation in the proto language, and the Tibetan and Chinese went for the ing, favored the ing versions, and Burmese favored the ik versions for whatever reason. And then other languages still continue to mix maybe. So ik versus ing is not predictive of subgroup affiliation, and it can't be solved with regular phonology. So I would, or like I'm inspired to compare this to heteroclites in Indo-European. So for instance, within Germanic, we get fawn for fire in Gothic, and we get old English fire. We don't need to go into the Indo-European details, but basically they're different case forms, and you get different analogical leveling. So some languages go with the nominative, and they analogically spread it. Some go with the sort of oblique stem and analogically spread it. And I'm not saying that we had case like this in Sino-Tibetan, but I think there was clearly some kind of morphological alternation in the proto language between ing and ik in a core set of basic nouns. So that's the explanation that I currently prefer. Moving just on to other irregularities. Here we have unexpected n in Chinese. Now just keep in mind that any given n in Chinese might be better reconstructed as an r, so this isn't that big a deal. So we have unexpected n in Chinese where we have r in Tibetan, so we sort of want the n to be an r in Chinese. We also have an unexpected n in Chinese where Tibetan has an l, so again we want the Chinese to be an r, but here we would point to an rl reconstruction. And we have unexpected n in Chinese where we have an open syllable in Tibetan, and in these cases, we don't want the n. We don't want the n to be an r. We don't want the n period in Chinese. So who knows what's going on there? Probably some kind of suffix. Maybe there are a lot of n suffixes floating around in Sino-Tibetan, but in any case, it's a problem to be solved. We also have the reverse, if you like. Those are the unexpected n's in Chinese. Here are unexpected r's. And now similarly, you might suspect that maybe Baxter and Cigar have incorrectly reconstructed r here because r and n are hard to distinguish in Chinese. But we would like to have n in these cases because that's what Tibetan has and Burmese when it comes up. More examples of the same. Intriguing rare correspondence to Chinese has a final m where one expects p based on the other languages. So we have a shadow where Tibetan and Burmese point to a final p and Chinese has a final m and needle, which I think we've seen on various occasions for various reasons. And I've ignored this complication because I don't have an answer for it. Definitely these seem similar enough that one expects that they're cognates. Also, sometimes we can have an unexpected T in Chinese. And I'll just talk through these examples. We have rice where you see Chinese has a T final and Tibetan has an S final. And then pitch pipe where Chinese has a T and Tibetan has an S. Cigar makes a very good argument for the semantics of bone being cognate with pitch pipe that has to do with how you made pitch pipes in in ancient China. And then this example of dig. I have suggested in the past, I think following an idea of Simons, although I don't quite remember that Tibetan S sometimes comes from a TS cluster, and that could be invoked here for rice and for bone, but it can't be used for dig. Okay, and then sometimes Chinese has an unexpected final, like in itch, and in this word, ancient times or in the past. Okay, we also have an unexpected D in Tibetan. For instance, we in all Chinese is me, but in in Tibetan is Ned. Okay, now turning to the vowels. We have Tibetan E, where we expect an a so a where we expect all, we just saw this example of we. But there are other examples as well. And some more examples. We sometimes have proto-Burmish, where we expect off, for example, in the word for house in Tibetan, if we if we compare that to roof in Burmese remember that Burmese Oh before dealers comes from proto-Burmish. So in proto-Burmish is something like kung. I don't think it's a coincidence that roof is kung in proto-Burmish and kung in Tibetan, but we don't have any way of getting between an O and an R in that way. And similarly with be scared. Tibetan and Chinese point to an Aq final, but Burmese points to Ok. Okay, and then Chinese sometimes has an ashua where we expect an I. And again, this I think might just be that it's hard to reconstruct Chinese correctly that there are some kinds of mergers in Chinese. But I guess here are the examples. The first one crumbs prohibit in Chinese and crimes law or, or right in Tibetan. And then the second one I think is quite nice to point out because it's a grammatical affects. So, there might well have been exceptional correspondences because of vowel reduction in unstressed syllable something like that. So you have a genitive go in in Chinese and D in Tibet. Yeah, and then in these two cases, Chinese has an evil instead of an I. Sometimes we get shot shot instead of I sometimes we get E instead of I but this is this is a rare one only in these two words. So difficulties with our Tibetan sometimes has a consonant. So, particularly dealers, plus our so in sort of for the tactic terms, an initial dealer and a middle are where Burmese or Chinese have initial are. So just to look at these in a little more detail. In the first one Chinese and Tibetan have crung, but Burmese just has wrong. Similar, you have crop and crop in in Tibetan and Chinese but then you just have a rap with an interesting media y as well. In those two you see agreement between Tibetan and Chinese, but then in the in the in the third one which maybe isn't such a great comparison actually, you have gr in Tibetan but just a raw in Chinese and in the last one cold. These are definitely related grung in Tibetan and an unspecified prefix so maybe it's it's okay prefix and then everything would be fine in a sense in Chinese. Now that's one set of complications. And now another one where Tibetan has our why and Chinese has our and now maybe one solution is to just reconstruct a media or sorry a media why the Chinese words in this case but that would seem a little bit ad hoc. So, so we have flow, which is you in in let's say pre Tibetan, but rule in Chinese. And let's go down to the last one because it's quite famous hundred in pre Tibetan in Burmese and with no sign of a middle yeah in Chinese. So this is a distinction I make or the graphically where our why gives us or yeah or yeah so yeah gives us or yeah this is Lee's law. But then I also have this are superscript why to mean sort of palization, which is gives us Benedict's law and that's how I, I index these two different outcomes of Tibetan and I do think that I don't know that you might find that overly mechanical but let me tell me another solution it's clear it's clear that there were two kinds of our why in pre Tibetan if you like, and in this case, we just have our in Chinese. And rocks for night. But, but we have our why evidence for our why in both Burmese and in Tibetan now, maybe this points to that the backstroke cigar you know should be doing something else here. So they should understand maybe that palatalized are changes to you in middle Chinese and that they should not reconstruct to the end prefix who knows. But in case I think there are some interesting examples. Now PR in Tibetan and are in other languages. So we have, you know, we just mentioned 100 where Burmese is kind of missing the, the P initial for some reason. And then the same thing in in in breast or breastplate but in this case, let's say Chinese and Burmese are compatible with your viewer initial, but Tibetan has a labial. Okay, and then also with tea are where Tibetan has TR DR and the other two languages don't have the cluster. And then I think this one indigo is probably a loan word into Tibetan from Chinese. Maybe at some point where some of the things had disappeared but it was still an R and not an L I'm not sure, but I don't somehow I don't think indigo is a sign or Tibetan word. There are a lot of complicated correspondences involving are that seem to have to do with with prefixes either being maintained or not before are. Okay, we have a lot of the similar problems with L, unsurprisingly. But you can, you can see them here where, you know, generally Tibetan has more going on than the other languages but it can be more complicated in that. Let's look at arrow so an arrow to Ben and Burmese have MLA, but Chinese has some loose initial maybe it was within with an M. Also a slight semantic difference there. And in the in the last one ground. So Tibetan has a seems have a G prefix Burmese and M prefix, and Chinese, no prefix at all. So there are some complications with laterals. Maybe let's look at four, which is the second on this page so you have a be prefix in in Tibetan anchor top, nothing in Burmese and an S prefix in Chinese so some some some mysteries here that need to be solved. And then there are also cases where it's unclear whether are or L is the initial. So in the word for neck. Chinese point to an L but Chinese points to an R. And in joyful Tibetan has an R and Chinese has an L. Now, one question that we've raised earlier is whether or not to reconstruct a yeah. Actually, we're discussing it about Chinese, but now about side of Tibetan. So there are cases where Tibetan and Burmese both have yeah, and Chinese has, you know, say, something not too far from your in middle Chinese but we're back from cigar reconstruct you viewers. And Gion Jacques and Axel Schussler suggest that in these sorts of environments, we should reconstruct a yeah all the way back. Baxter and cigar kind of reply to this criticism in their answer to Schussler's review in diachronica and point out that most of the Tibetan cognates do have a G in them. So, maybe their uvularity is not so bad after all. In any case, I don't have a firm opinion on on this particular question. Now, the origin of Chinese voiceless residents. So we have voiceless residents in in Burmese but we know where they come from. Where do they come from in Chinese. Well, you would be tempted maybe, especially if you were, you know, made to Lynn, you would say these come from s clusters. So the word black, you know, muck, actually started as smuck, something like that. But that's not going to work. Because we have reconstructed s clusters in Baxter and cigar system for other reasons. So where do these voiceless residents come from. One idea that I think I saw somewhere very recently discussed although I don't remember where is it. They did it has to do with the loose tight distinction where something like this tight initial SS became voiceless residents and then loose initial SS became tight initial SS after that. And that might explain what's going on in Chinese. But in any case, it's a problem that needs to be addressed. What is the origin of voiceless residents in Chinese and and can comparisons shed any light on that. There are just a few more examples and a few more examples. Okay, and now one of the big conjectures that would relate to Ben and Burmese so this idea of Shefners that Burmese has. This is how he put it has in some cases an aspirated consonant as an initial where Tibetan uses a superscript s. In kind of more modern terminology in the framework I'm using we can state this conjecture more precisely as where proto Burmese has a pre globalized resonance to Ben has an s resident cluster. So there's a lot of evidence that supports this. There's a lot of evidence, heart knows mucus bamboo. Right. But there are also counter examples. In particular, moon and snake. In the example of moon, Lashi actually does have a pre globalized initial. So we. So there's something going on inside of Burmese. I suspect it's actually that. Yeah, you can't tell it from the, from the orthography in Burmese but this is a creaky tone in Burmese. So it, you can see that with the capital X in Lashi. And so let's assume that Lashi is something like the proto Burmese form. I wonder whether when pre globalized changed into voices resonance in Burmese, there was some kind of conditioning whereby that didn't happen in in the creaky tone, because it's sort of too many, you know, glottal features I'm not sure what we what we can say very confidently is the double moon is a counter example to she's nurse conjecture from the perspective of Burmese, but from the from but it's a kind of it's a proto Burmese to Burmese problem. It's not a sign of Tibetan problem. Whereas snake is is maybe more of a problem. So let's look at counter examples where Tibetan lacks an S as well. So let's look at borrow where Burmese points to pre globalized initial in proto Burmese, but the Tibetan word is Brnya, not business. Yeah. So you can see here paths of investigation like I don't know maybe our prefixes in Tibetan are actually rotor sized S prefixes in some cases, or maybe she's nurse conjecture only addresses some sources of pre globalized initials and in proto Burmese and in proto Burmese, there were other prefixes that they caused it, for instance, in in to where Tibetan has a velar prefix, maybe it's that velar prefix that caused the pre globalization in proto Burmese. In any case, they're counter examples to she's nurse conjecture. Now, there are circumstances where there's disagreement on the manner of the initial obstrant. So places where proto Burmese is unexpectedly voiced and now remember that Burmese doesn't have a voicing distinction so so we take an aspiration distinction Burmese back to a voicing distinction in proto Burmese because of the Burmese split. So if we look at all, we have gun in Burmese or in proto Burmese and con in Tibetan. For slave or servant, we have gun in Burmese coal in Tibetan and corons in Chinese comparisons, you know, it's it's it's tantalizing, but they don't just line up perfectly. Okay. Here are more examples where proto Burmese is unexpectedly voiced. And they include, for instance, the number one. So you have, you know, proto Burmese or, or let's say, Trans Himalayan from the Burmese perspective would give you a deck where you have tech in Chinese and good tech in in Tibetan. Okay, Chinese can be unexpectedly voiceless. So here's this example I gave from elbow. Then there's also nine. And in some of these cases you you might suspect that it's actually Tibetan has voiced for some reason. Maybe there was a little schwa before, before the application of a prefix something like that but it's a, it's a problem to solve. So unexpected voiceless initials in Chinese and a few more examples. And even some more Chinese unexpected voiceless stops. Okay, now there's an idea of Guillaume Jacques that would get rid of some of these, which is that Chinese pre nasalization corresponds to Tibetan voicing. And from the perspective of sign of Tibetan what you would say is you had pre nasalized stops you had like, you had, you had maybe voice voiceless and pre nasalized stops inside of Tibetan, and the pre nasalized stops stayed pre nasalized in Chinese. And they merged with the voice stops in Tibetan very nice idea. I wish it were unambiguously true. So here are some examples that kind of would make you think that so hold in the mouth we get, you know, I don't know, some constant and then come in Chinese, and we get come in Tibetan. And then the G comes from the pre nasalization in the word for shield or contradict, you know, we get a G in proto Burmese and Tibetan and an MK in Chinese so maybe there's maybe the voice the car in Tibetan and Burmese so it's a nice idea, but it only gets as some of the examples. And let's say. Yeah, this is just to give the evidence that actually going back presented in his paper, where he was actually presenting corresponds between pre nasalization in in gyarani clanguages with with voiced in in Tibetan which is of course, you know, not the same idea, but it's the same kind of idea which is we're talking about some of the Tibetan voiced constants as coming from pre nasalization. So, um, the problem is that many of these voiced initials in Tibetan correspond with just plain voiceless in Chinese I've gone through those examples already. They don't correspond with an and pre nasalized initial in Chinese. So, at best, this conjecture of young shocks only gets us a little bit of the way in the direction we want to go. But in his review of my book, where I where he sort of noticed that I had discussed this idea of his, and was a little bit critical of it by saying well it doesn't get us very far and here's a good example. I do let's say I discussed this example which then he has an answer for which I will, which I will present you in a moment. So father is a napa in Chinese so according to his theory, it should be bad in in Tibetan, right. So his, his theory both doesn't explain all the, the pause that need to be explained in Chinese, and it doesn't explain all the buzz need to be explained in Tibet. So, maybe not such a great theory. That's, that's what I said in my book. Yeah. But he points out that in Limbu. Father is usually prefixed in a way that's similar to for now and I've discussed in Garong but in this case we're talking about Limbu, where the prefix is a nasal prefix so you know my father is on Bob, and your father is. I don't know how to say this. Whereas just father in the abstract father as an idea the father is is is pop. So he wonders whether they're there might that this might be similar to what sign of Tibetan had, and then you have a pretty easy explanation, which is that in Chinese, they generalized the prefix form, and in Tibetan they generalized the unprefixed form, which is to put it to put this another way. And that means that we can keep our story about pre nasalization, leading to voicing in Tibetan, where we say well in this case, when that sound change happened, the end prefix was already gone into that and because of morphological And then I'll just point out that the proto-Burmish, the pro-Burmish correspondence is also irregular. So I see has why and not pop. So maybe that's some evidence of some kind of prefixing going on also in proto-Burmish. So I think this is, you know, the goal of this presentation is to sort of show you the cutting edge of sign of Tibetan and and I think this is a good story to tell We've pushed historical phonology a certain distance. And now we know which problems need to be solved. And then in looking to solve those problems, in addition to just finding more and more refined conditioning in our sound changes, we can turn to morphology. But Tibetan and Burmese are not good places to turn to for morphology, whereas Limbu and Giao Rong, they are. So I think that's the next generation of scholarship will be looking at morphology in those languages that have it, and seeing whether that morphology can explain the irregular phonology that we get. So that whole sort of discussion we just had was about where Chinese has a voiceless initial, where we expect a voiced initial, we also have Tibetan having a voiceless initial where we were where Chinese has a voiced initial. So the reverse if you like. So dig, foot, carry, uncle. One Burmese cognate agrees with Tibetan, and one agrees with Chinese. In this one, it's saddle frame where you have Tibetan call Burmese gal and Chinese guy. And then on on here it's a cut or sharp where we're here Burmese agrees with Tibetan. So to Ben has shot Chinese has zot and Burmese has shot. So it's a mess. Yeah. Okay, and then in two cases we have Tibetan and Burmese disagree. And one of them we don't have Chinese. Now, we move on to the origin of Chinese aspirant options. So, just to remind you. So to Ben basically to Ben allows many many clusters but fundamentally, in terms of manners only opposes voiced and voiceless. Whereas Burmese doesn't really allow many clusters and Burmese itself has a voiceless and voiceless aspirant, but that goes back to voice voiceless pregolodized. And that may even go back to a voiceless voiceless pregolodized voiceless pregolodized. Chinese has a three way contrast in manner voiced voiceless and voiceless aspirant. So it just these kind of broad categories don't match very well among the languages. And we would end the aspirant options in Chinese are relatively rare. So I think a lot of people feel like they're secondary and we should get rid of them somehow, but it's not clear how. So in many cases, aspirant in Chinese corresponds to a plain obstrant in the other two languages here are some examples, maybe just look at bitter which is one of the, you know, clearest best Chinese cognize out there. Sign of Tibetan cognize out there. So what why is the Chinese aspirate where the other two were on aspirate, not clear. So Chinese aspirates can correspond to voice stops into bed and Burmese, for instance, in, let's say in Bend, where you have croc in croc in Chinese and gooks into bed. Or maybe scatter divide is even better example. Try it in Chinese where you break. Sorry bra in Tibetan and bra in Burmese which goes back to bra. So the Burmese and then, and the Tibetan tend to agree and something is causing the voicing and aspiration in Chinese. So the big problem in Chinese, let's say summing up our the voiceless residents and these aspirate stuff. So now we turn to the big problems in Tibetan, which generally have to do with palalization. So there are you can understand there is being three, I think, let's say at least conceptually speaking it may not be what happened historically but conceptually we have three kinds of palalization in Tibetan we have Houghton's law which changes. No into me conditioned by some palalizing environment. And we have love changes to show conditioned by some palalizing environment, and then we have lots of other palalization. Now these palalizations we sort of aren't as interesting to look at because they don't lead to mergers they lead to splits. Yeah, so. So they're still, you know, clearly present in the, in the language, and the conditioning environment for for all of these is unclear, although you know as you would expect high front vowels sure help. There are more and more examples with it. So with it is the most examples of the others fewer examples, but there are examples with all vowels including ah, so it doesn't have to do with the vowels. I think you know, in a way the easiest thing to do is just say it's just to reconstruct the palalizing environment, all the way back and then to say that before the value it. So that palalizing feature was was automatic, you know that's that's a way that a lot of languages are analyzed synchronically including for instance, Anton Lustig's analysis of Iowa. So, with that said, we have to distinguish to palalizing environments, because the palalizations happened hundreds of years apart from each other. So Houghton's law is shared with the East Bodish languages. So, it's not the same thing that that triggered these other palalizations right so we have to have at least, you know, if we're reconstructing these mechanically, we can say well we'll just reconstruct a superscript why as whatever the palalizing environment is, we have to do two, one for Houghton's law and one for everything else. And now we're ending up with a quite sort of algebraic proto language which I don't feel super comfortable with and we need to figure out like what actually is the phonetic environment that caused Houghton's law or these other palalizations and I think that's one of the big open questions. And the examples of Houghton's law just to remind you and you see that that Kurtop agrees with Tibetan against Burmese and Chinese. So Kurtop underwent Houghton's law. Now, an idea that's worth discussing to make the correspondences nicer. Nishita points out and particularly based on this word six that Tibetan has DR where Burmese has KR. And actually Nishita thought that Burmese was innovative in these cases, but I think Chinese points points you in the direction that it's probably, it's probably Tibetan that has innovated. Now, the reason I have this proposal here and not in an earlier presentation is it's not the case that you can set this up as a, across the board innovation you know it's not like every place Burmese has KR, Tibetan has TR, DR, it's not that easy. So, there's something going on but we don't let's put it this way don't know the conditioning environment. Okay. And East Bodish does not participate in this innovation. So it's, it's quite late in the history of Tibetan. So you can say, well, for some of Tibetan purposes, we will just remember when you see DR in Tibetan, ask yourself whether it's an old DR or a new DR. And you can check that by looking at East Bodish. And yes, it's a problem but it's a problem that we can solve in the context of Tibetan and not worry about in the context of sign of Tibetan. So just to give, you know, one example will go talk about six, because you know it's the, it's the best example of this whole kettle of fish to look in Tibetan but grok in boom top, and grow in doctor. So yeah, we have a clear velar. So then, so the dental initial is late in Tibet. So now we come back to Schiffner's conjecture we did it first with the resonance. Now we're going to do with the obstrance. And it's more complicated because in in Burmese we have two sources of aspirate obstrance, right, plain voiceless obstrance and preglottalized obstrance. So, I sort of asked myself, I mean, how, how would you phrase Schiffner's conjecture. Now, if knowing more about proto Burmese than he knew and I think it would be the SP where P is any voiceless obstrant SP would correspond in Tibetan to preglottalize P in proto Burmese. So, here are examples. Just look at borrow. As one case so you have preglottalize K I why in in proto Burmese and SK why I in in Tibetan, it's quite a good, you know, seems like a very promising conjecture. Now, a complication, which is, I think quite interesting is, we also have SB corresponding to preglottalized P in in proto Burmese. So, for instance, in the word for hill. In this case we actually can't prove that the proto Burmese was preglottalized because I haven't been able to find it in other Burmese languages but let's give it the benefit of the doubt. And then, most saliently in frog. Yeah. So frog clearly goes back to a preglottalize P in proto Burmese, but an SB in in Tibetan, and this suggests. Try and say it slowly and carefully. Starting from the perspective of Burmese and working backwards. At the Burmese level, all we have is a distinction between P and pH, but at pro pro Burmese, we've been able to rewrite that as a BP distinction and realize that there's a third series preglottalized P, and that the P and the preglottalized P merge on the way to Burmese. This evidence suggests that we, and we can't reconstruct this at the proto Burmese level. It's only with some Tibetan evidence, so we can call it the sort of pre proto Burmese level that we have a fourth series, which you can understand as a preglottalized voiced obstrate so then we would have, you know, P be preglottalized P preglottalized B. So frog I think is the best comparative evidence for that. But there's also some Burmese internal evidence, having to do with causatives. So we have this verb to be afraid, good. And we have the causative to frighten or scare, which reconstructs to preglottalized crook in proto Burmese. And I think purely, you know, for reasons of elegance in terms of internal reconstruction, you can take that one a step further back and say well maybe it was actually, you know, group and preglottalized group where the preglottalization is a sign of some kind of causative prefix. You know, maybe, ultimately an s so you could reconstruct almost entirely on Burmese internal evidence you could do something like group and group, which is just to say that that then this hypothesis that has emerged from frog and the comparison with Tibetan has some Burmese internal evidence that suggests we can we can imagine that kind of pre proto Burmese, let's call it had four way manner contrast in obstruence. Okay. So that shows us that Schieffner's conjecture is is really good, you know, it's really exciting, it really gets us a lot of work done. But there are exceptions. So, we have cases where the P in Burmese is not preglottalized, for example in star. Ah, but curiously, again, like we saw in moon for the residents, Lashi shows it is preglottalized. So, it's not an exception from the from the sign of Tibetan perspective, it's a Burmese internal problem. But it doesn't always help us. So, for instance, the word green Burmese and Lashi agree that it's not preglottalized, whereas Tibetan unambiguously has an S initial. So, what do we do about that. I don't have an answer. It's an exception to Schieffner's conjecture. And then here are two more examples, but I have put them on to this slide by themselves because I'm going to talk more about them. So, the word body looks like a nice cognates right so you have SKU in Tibetan school. Maybe something like goo in Proto Burmese and then call in in Chinese. But it is an exception to Schieffner's conjecture because you would expect to have some kind of preglottalization in Proto Burmese. And then, if we look at knee. You have a spusmo in Tibetan. And then you have. It's actually pronounced in Burmese. Sorry, or actually could be. But let's go with something in terms of the spelling. So, from here again, Burmese, we would expect to have an aspirate in order to confirm conform to Schieffner's conjecture. But a few thoughts. The Burmese word might well be alone from Pali kaya as the final why indicates indigenous. Indigenous Burmese words would not have a final why in this phonotactic position. And the pre aspiration in Lashi suggests that Burmese. It has innovated. It also have lost the final T it's probably a reduction in compound right so you actually would have had something like put charge, and then it becomes a charge in compound. So, maybe these two problems aren't aren't real and I think that that shows you, you know that we can that we're posing the right questions and making some progress. Okay, we also have the reverse so these are examples where we have an S in Tibetan and we don't have a preglottalization in Burmese but we also where we have a preglottalization in Burmese but we don't have an S in Tibetan. For instance, this word for male or armor crop in Tibetan crop in in Chinese for shell, but Burmese and proto Burmese has a preglottalized initial when we don't have an explanation for that. So I won't go through the other examples. And then just to sort of end this discussion of preglottalized initials in proto Burmese and their cognates in sign of Tibetan. We, I remind you of the bizarre word otter where Lashi has a preglottalized shot. So sham, which looks very nice in comparison to serum in Tibetan, but but Burmese has Piam. So there's something there's something really strange going on in the word for otter. But it, it certainly doesn't argue against chief nurse conjecture. Now, to sibilants, Chinese only has one is S. Yeah, so sign with Tibetan from the Chinese perspective you say oh well it had S. And from Tibetans perspective you can say they're to S and show I mean to some extent that reverts to the problem I just talked about which is why does Tibetan paralyzed. Yeah. And then for proto Burmese, we have four sibilants s preglottalized s, paralyzed s and preglottalized paralyzed s. The three languages point to a very different inventory of sibilants, and the correspondences are a total mess. Here we have cognates of proto Burmese s. Here are cognates of proto Burmese preglottalized s where well we have this rock in Tibetan maybe the R has something to do with the preglottalization I don't know. We have to kill where you know maybe the G prefix in Tibetan or the R the apparent are in fix in Chinese or what's cognate with the preglottalization it's really not very clear. And then we have these show examples in proto Burmese where we have drink. Well, I think these this word is pretty clearly cognate to this Chinese word, but here we have SR in in Chinese corresponding to show rather to pre then to preglottalized s. So just to prove the the the correspondences are a mess. Yeah. And then we can't talk about sibilants in proto Burmese without again reminding ourselves about what what a strange little beast author is and here I throw in a Chinese word for which does not look like a particularly good comparison, but it sort of doesn't look particularly worse than the other comparisons so you know might as well throw it up there I thought yeah. So that's order. Just talking about S is an RS. So the book has this conjecture that Burmese show. So a voiceless are from the Burmese perspective and a preglottalized are from the proto Burmese perspective comes from SR. And he points to these two examples. Another sex or shame, which he thinks is cognitive with a shamed and shame in job and a live where you have SR in Chinese and some in Burmese, but I have another I propose another cognate for ashamed, which is that it's, it's cognitive with red in Chinese and blood in in Tibetan. And why do I propose that it seems like proposing that shame is cognitive with shame is much more straightforward right. It's because it works in terms of my proposals for historical phenology which is that you had, you know, a nice uvular onset and the uvulars dropped in Burmese and the uvular becomes a veal or into Ben so the phenology just couldn't be better. And it is extremely normal across the world to have semantic change red goes to ashamed, because people blush when when they're ashamed. Yeah. So, yeah. So, which is not to say that give me Jacques ideas are bad or even wrong. I think maybe these these these other words can be brought into the story somehow, but I feel very fond of my comparison and I'm not ready to give it up. So this second cognate proposal so you only have two examples of this corresponds. The vowels don't work. So, you know, maybe that's a problem. Yeah, you, you, what should you have gotten in in in Burmese maybe that's a good quiz item well I'll just tell you. In should have gone to in and then in should have gone to that's Dempsey's law and then in should have gone to any that's Wolfen's law so you should have gotten shun or would be pronounced. Actually be pronounced she I think in in modern Burmese but it would be RHA and tilde and tilde is what you should have gotten as the vowel correspondence. So, you know, I'm not going to say they're not cognate but it's not a great comparison. And then just to say also in my book I catalog all the one off correspondences as well, all the kind of, oh and then here's, you know, a P corresponding to an end or something like that, where you might well think that they're just bad comparisons but their comparisons exist in the literature. And, and my temptation is to say we don't want to overly clean our cognate sets, because then we won't find new things. We're going to leave some some mess in there, and then we'll have something to clean up, but you know I really didn't think it was a good use of our time or your energy to sort of take take you through individual examples of sweet generous correspondences. Those are all of the problems. So I will just kind of add a word of summary. These are the big questions. So in Chinese, they are. What's the origin of the AB distinction. What's the origin of the voiceless resonance, and what's the origin of the aspirate obstrants in Tibetan. It's basically palatalization. I also think Tibetan is probably the best place to look at oblaut and, and, and more philology should be done on the Tibetan verbal system to figure out exactly what the prefixes mean and where we find the oblauts, not only in terms of you know, naming the stems like you get the present as you get the past but actually find out what those what those things actually mean. I think that's a big desideratum in Tibetan. And then in proto-Burmish, what I talked about had a lot to do with Schieffner's conjecture of sort of linking proto-Burmish and Tibetan, but actually in the case of proto-Burmish we're lucky to have this Burmo-Changik hypothesis, where I think just figuring out the next level up, you know, we did, we did the proto-Burmish. So how does proto-Burmish relate to Lolo-ish, how does Lolo-Burmish relate to Burmo-Changik. And then I think in particular bringing the, the, the velarized, non-velarized vowel distinction that you see across Gaorong and Changik languages into the picture will be very helpful. And I think that we will end up, you know, sorting, we'll end up with, let's say a proto-Burmish Changik that's a much better starting place for comparisons to Tibetan because it will have a more gorgeous set of onsets and the sources of these pre-globalizations will be, will be more clear. And maybe we'll also have a tidier vowel system that doesn't have this imbalance that currently exists with far more correspondences in open syllables than closed syllables. That's a nice idea. Yeah. You know, I wish I'd talked to you before I wrote my book in terms of I think, I think reconstructing B for the MP correspondence, G for the, for where I have K-Schwa and D for where I have RL would lead to an overall more elegant looking Sino-Tibetan