 But I'm going to start with with some preliminaries some discussions of historical linguistics methodology in general And I will review very quickly the major sound Sorry, the major language families of East Asia so that we can know okay Well, we're going to talk about sign of Tibetan, but what is its context in terms of other language families and then go through? Some branches of of sign of Tibetan not all of them, but here we go So language families of Asia I've ordered them in roughly speaking. I think the The history of their dispersal so one thing that that that I'll be looking at is is maps where if a Language is quite isolated in its geography or or broken up in its geography. It probably spread earlier and that's You know, that's this is the order that I came up with of language families in in Asia According roughly to when they probably spread so looking first though at Modern nation states and I'm only going to look at the two Chinas. Let's say we have in Taiwan 20 languages 15 of them for Mosin and for synodic, which is says for Chinese dialects and then in the PRC 300 languages from 11 Primary language families. That's means like, you know, like as different as finish in English or something like that And if we compare that with Europe in Europe, there are 24 official languages used in the context of the EU and There's a total of about 130 languages spoken in in this is not just in the EU, but in all of geographical Europe and They are found in four language families. So I think it's just important contextually to realize that China has twice the linguistic diversity of Europe. So so, you know 300 languages 11 families versus 130 languages and Four language families and I think that's I don't know. I don't want to sort of belabor it necessarily, but I Mean, it's very important contextually because because if you look at how Europe is studied in academia And I don't know how Europe manifests politically. You would think Europe was more multilingual than China Whereas in fact China is much more multilingual than Europe And we can think about, you know Why that contradiction exists, but maybe right not right now So looking at isolates these are, you know A single language that is as far as we can tell unrelated to any other language in the world And we can look at their distribution and get a sense of like, I don't know what the world Would have been like a long time ago. So there's I knew spoken Or tested in three dialects Sakhalin, Kuril and Hokkaido But Sakhalin and Kuril are gone Sakhalin survived into the 1970s. So there is some information about it very little about Kuril and And Hokkaido is kind of just barely around Then Kusunda is spoken in Nepal and there's there were actually At the beginning of the summer there were two speakers, but there's only one speaker now They were hunter-gatherer community that switched to I don't know agriculture since the 1950s and Are, you know, probably indicative of the of the the earliest linguistic situation in that part of the world in Nepal and then Andamanese is spoken just in the Andaman Islands and It's probably the most robust of spoken in the Andaman Islands and One island in particular has is not yet kind of interfered with by modern Civilization, so they are probably quite robust as far as language isolates go in terms of maintenance of their language Then there's a language Nihali that is spoken in West Central India seems not to have been worked on very recently, but was still around in the 1990s and has been highly influenced by Indo-Aryan and and they called Tamil What are those Dravidian Dravidian languages? But seems to be a language isolate and then an interesting case. There's a language Veda Where currently it seems that the there is a population of of let's say an indigenous people who who Traditionally used hunter-gathering as their subsistence strategy and they speak now a Divergent dialect of singhalese, but it seems or or you know, I haven't looked into this at all, but They say this Divergent dialect of singhalese has you know, a substrate of a language isolate that would have been the indigenous language of Sri Lanka So yes, and then of course a burusha ski in northern Pakistan is also considered an isolate So this gives you I think a nice impression of basically In every direction, especially on islands, but to some extent in the mountains There's a language here or there that's unrelated to any other language in the world and these are probably the you know, the last remnants of Diversity that's kind of hard to imagine That as other languages came and and spread to to break up you know these To leave these these few isolates and and none of these are in China as far as I know there are there are no Islets in China, which is kind of interesting in its own right. I'm not quite sure what it means But now turning to of Mengmian I think this Where it seems to me is in some ways the oldest if you like a family of China if you look at the map You see that it's it's it's very sparsely spread throughout China and and Vietnam Laos and Thailand and then there's reason to think both based on known history and on the distribution of the branches that the the the people in Vietnam and Laos and whatnot Have moved south and in general like that's a pattern that we will see again and again that Languages are sort of spreading out of China and Generally speaking which has something to do with you know, the world geography Languages are sort of coming in from the Northwest and then pushing other languages out into the Southeast and About Mengmian there so they're 39 Languages they fall into two groups Hmong and Mian in Chinese. They're called meow and yow They have very complex tonal systems. I wasn't sure, you know, just What to say about these these are going through a lot of languages very quickly So I just to point out, you know something to so it's not just a list of names. Yeah, so you can say oh, okay Hmongmian has very complex tonal systems and Hmongmian has very long-term contact with Chinese such that the the the Hmong in particular Have a special place in Chinese mythology where they're sort of the the sister people to the Chinese and and mongmian languages Are full of very early Chinese long words So they're very the study of mongmian is very important to the study of Chinese historical phonology as well Okay now turning to Austroasiatic Here again looking at the map you can see that probably it started in in China With for instance this branch Pachanik still in China and Is very widely spread out so you see you have Munda and Kasi in India and then you have Nicobarese in the Nicobar Islands in the Southwest But the the famous languages are our Vietnamese and Khmer. So let's just go through that. There's a hundred and sixty-eight Austroasiatic languages the first to tested is Mon Which is spoken in mon state in Burma and in parts of Thailand It was a tested from the 6th century and Vietnamese and Khmer are the two languages with official Status at the national level Okay now on to This family is called both Thai Kadae and Kradae and Again, if you look at the distribution you see you know these little enclaves up in China with Kradae very You know dispersed so this Suggests that the family comes from China and then spread south and You notice that the colored blobs get sort of larger as you go south and are are on Increasingly or let's say increasingly less diverse So you have northern Thai central Thai and then southwestern Thai is the is the largest and Those are all you know varieties of Thai. Yeah, so 95 Kradae languages with the first attested being Thai Which was written down in 1283 and is the only one of National status the most famous Kradae language in China is Zhuang associated with the Zhuang Ethnicity so now on to Austronesian Where I have to zoom way out because these languages are spoken over a huge Geographic area That includes you see in the in the lower left Madagascar all the way to Hawaii or something in the east 1257 languages They all spread from Formosa and this is kind of methodologically quite interesting because it can be hard Sometimes to figure out where a language family spread from you may know that there's a controversy about this for Indo-European languages but in this case it's easy because Basically all of the primary branches of Austronesian are still present in Taiwan but And then as you get sort of further and further from Taiwan you get less Divergence less diversity and and Just point out that was personally I have trouble telling difference between Austro-Asiatic and Austronesian but the key is Nisha is referring to islands like in Indonesia, you know Malaysia Micronesia, so this this Language family is is Very strongly associated with islands it comes from Taiwan got as far as Madagascar And basically every island in the in the Pacific would have been Would have been Austronesian speaking at some point And then in terms of national languages Malay, Basi, Indonesia, Tagalog and Malagasy are all Austronesian languages and Yeah, okay, so then Now you know kind of the basic picture and what I'm not going to get into But you should know is out there is kind of big Theories about Relationships between these groups so like a lot of people in China think that crawdye is related to sign or Tibetan and Laurence cigar thinks that Austronesian is related to sign or Tibetan a lot of people think that crawdye is related to Austronesian and and Here's how I look at these things kind of in From sort of first principles on the one hand Language as a technology was probably not invented a lot of times independently So ultimately all languages on earth are probably related and Languages will tend to be related to languages that are close by So at some point for instance the Austronesian family would have had to have gotten from the mainland to Taiwan So so kind of prima facia it is related to to austral aziatic but This is the kind of thing that it's not going to be very profitable to work on because None of these languages are particularly well studied by European language by European standards and and You know something like Indo-European Let's talk about in terms of big families into Eurealic looks quite promising But is extremely hard To figure out and is very controversial and that despite both Indo-European and Eurealic Having 200 years of research done on them at a very high level with a lot of primary evidence So I think it's premature to worry about relationships among the language families of Asia until The work on the individual language families is more advanced and Now some remarks about historical linguistics so the Key method in historical linguistics is the comparative method and the other Thing to keep in mind is analogy and the way I think about this is so so you have Regular phonology and you have analogy and we'll we'll we'll we'll look at them in turn And then I also want to just mention this. I don't know what to call it is not quite a methodology, but a Way of looking at human prehistory. Let's put it that way, which is called virtue and saw him in German We always say it in English in German Words and things where basically if you can reconstruct words In a proto language, it means that there was some group of people who spoke that language and had those words and that means mentions of Technologies practices animals Are also reconstructible, right? It's a way of accessing the the physical and social world of a prehistoric community through its language But I will get back to that So looking at the comparative method And for some of you this will be a totally old hat, but maybe for some a little bit Less so so we notice systematic correspondences between words in languages and Just to kind of keep it easy and hopefully Languages some of us at least I'm familiar with I look at Latin Sanskrit English in German So you notice that where Latin has a P Sanskrit has a P But English doesn't have a P. It has an F and German also has an F. You can ignore the fact that German uses two letters for the F sound That's that's just an issue of spelling So that's for the word for father it also holds for the word for fish and For the word for foot now when we find a pattern like this we We let's say you might think oh well it either it holds sometimes it doesn't hold sometimes But the methodological principle is that a pattern like this needs to hold all the time and Then either if you find exceptions either they're phonetically conditioned or They're evidence of loanwords So so having seen a pattern like this then you can think These languages at one time in the distant past must have been the same language Which gradually changed into these four languages and in these words in this position Either it had a P or it had an F Now how to decide which it was I'll just point out that Latin and Sanskrit are not particularly close geographically and English and German are closer, especially when we know that you know some historical facts like that English was brought to England from Let's say the German speaking world in the fifth and sixth century so That's just purely that sort of geographical and historical context makes you think probably the P is original and the Germanic language has changed the P to an F but Also, we can have recourse to linguistic typology Which is that we see again and again languages that change P's into F's so like Japanese changed P into F Greek changed P into F So so if you just had to guess which one's older the P or the F you would guess the P Yeah, so so that's what we would say here and we can say that the The language into European that later became all four of these languages Had a P in these words and then that changed to an F in English and German So that's the comparative method you you you bring words together you find the patterns for each pattern You come up with an explanation and that way you can reconstruct how the light the words used to be pronounced I'm not going to go into You know the reconstruction of these particular words because that would be a lecture about Indian European and not about science then Yeah, I've already mentioned that okay So now on to analogy but maybe what I should say is that All of this stuff is regular historical phonology is a process that affects inherited vocabulary where the the the let's say I would say the psychological process in question is memory, right? You you learn something from your parents and then you speak it when you speak the language So you're correctly remembering the word remembering the word and then just natural phonetic tendencies when populations are no longer in contact allows a sound to gradually change into another sound through it through you can think of it as errors of production or Perception so, you know someone someone because of the way the lips works someone occasionally pronounces a p as an F and then Maybe the child learns it as an F. I mean, I'm radically oversimplifying but the the the the the part the point I want to make is Phonological change of this kind is indicative of Correctly remembering the language as it was sort of I mean talks to you is not quite the word but as you as you learn whereas analogy is about What happens when you forget something? so So here and I think this is this notation is something that I would sort of like everyone to Come away with where you this is how you do four-part analogy It's called not all analogy is four-part analogy, but it's always worth trying To see whether you can make an analogy in the form of four-part analogy and why is it called for part? Well, you see there are four things here. Yeah, so the way you read it is talk is to talk as help is to X and I wonder can I have any volunteers for for what X equals? anyone Don't be shy. Let's see. I'll just scroll down here and pick someone then Oh Your names aren't showing Whoops How about oops? Oops. Sorry. How about I forgot your name, but the fellow in Pennsylvania Yeah, hi, my name is Nikita Nikita. Okay, what do you what will you solve X as? So it will be a helped. Yes Okay, yeah So, uh, so yeah, so not so hard but the point is Historically speaking it was Holpen. Yeah, and if you Let's say if you live in this country where I live you occasionally hear Holpen Or here is maybe the wrong word but in coral even song the kind of church service that Anglican's like to do in You know, for instance in Cambridge University. It's a famous thing. There's a line thou has Holpen thy servant Israel and When the Queen writes a letter to members of the House of Lords once a year telling them to show up to meetings The letter also uses the word Holpen even in 2020. Yeah, so the point here Excuse me, but but in Infinitive to help it will also be help. It's not like hope. I mean historically speaking in the wider is this In the infinitive is help So speaking it should be help hope Holpen Okay And the Point is at some point, you know, someone Forgot that they should say Holpen and they say helped and the way They decided to say helped was probably using an equation like talks is to talk as help is to Holpen. So As a historian You try to do these things backwards. So you say for instance, okay I see people say helped But based on for instance comparative evidence It should be hope Can I come up with an explanation for how you would go from one to another and then and then for part analogy is the way to go now to give another one Strive is to strove as Dive is to and then let me see if Maybe sue. Can you tell me the answer here? Yes? Um, actually, I saw this you saw it. Okay. Yeah It's dope. Yeah It's dope and that replaced the inherited dives and I've set up these two examples To make a point which is that in the one case we have Let's say us the English those the word uses a strong verb, but you can just think an irregular verb it's becoming regular and in the other case you have a And a regular verb dive dive. It's becoming irregular dive dough So analogy can make things Can change things from one pattern to another and And it could go either way. That's the point I want to make right I think there's it. There's it. There's a tendency in the way people generally teach linguistics to say that in The first case you have a rule like in English we form the past with ed And and that's a rule Whereas we need analogy for an explanation like the second one And the point I want to make is actually that's not necessary right analogy can get you both So you don't need any rules you can do everything with analogy and then it's just a historical question of which models did people turn to when they didn't remember things correctly and the You know, that's that's a that's a historical question that we can try and figure out and You might say even that strife strove was Not the right model. Maybe it was drive drove or stride Shrive Shrove or something like that, but it's clear that whatever the model was that change dive dived into dove Probably shared something with it in terms of its phonology like it it was an eye of ove verb. Yeah, anyhow, so The the let's just say those are the two The two tools we have available as historians of language regular phonology where you set up, you know correspondences and reconstruct them and Analogy which you can use to explain cases where the correspondence patterns don't work and You might be tempted to use other explanations But I think that's maybe sometimes required But it's it's really like, you know, like in the old days you did the geometry just with a Compass and the straight edge. I think it's good to do as much historical linguistics as you can with just regular phonology and analogy and And then I just you know, as I said, it's to throw away comment I don't think you need any kind of rules or deep structure in linguistics to explain How things got the way they are or in fact how people learn or what structure languages have I think it can all be done through through memory and forgetfulness where memory is is correctly reproducing what you learned and Forgetfulness is analogy where you where you use a part of the system that you do know to guess a part of the system That you've momentarily forgotten Okay now turning to virtual one second The words of of an ancestral language tell us about the objects and ideas of the world of its speakers So let's just look at what the sign of Tibetan, you know, world was like And and how can I say I I think it's important Or one thing to keep in mind here is for the Indo-Europeans. We know a lot about Archaeology and genetics and and basically, you know, the Indo-Europeans are from north of the Black Sea 4000 BC something like that. They were pastoralists. We don't know nearly as much about sign of Tibetan But let's look at what words we can reconstruct they had lies. Yeah, so Tibetan and Chinese Sorry about this capital F. It has to do with it's a it's a tricky Chinese character Which my computer didn't want to make even though I encoded it correctly So anyhow Chinese and Tibetan have inherited words for lice. They have an inherited words for fish they have inherited word for field and now I think this So what are we what are we finding out? Well, I think everyone in the world has lice So it's not particularly indicative, but if they have a word for fish it means that they were living by the sea or by rivers and Not everyone does right there. There are certainly plenty of people in the world who who for whom Fish is not a very important part of their of their diet So we have learned something about possible geographic origins just by knowing that they have a word for fish and in terms of the word for field it suggests that the Sino-Tibetans knew some form of agriculture and let's say for for for Archaeological reasons, it's clear that that wasn't rice agriculture Millet seems to be the thing that goes back the furthest in among Sino-Tibetans speakers But also this word field might maybe we shouldn't imagine like a wheat field like in England or rice Patty, but instead something like swidden agriculture where where you burn down part of a forest to plant in it Which is a form of agriculture that's still practiced in Northeast India And they have a Cognate word for field that refer to those patches But in case it's clear that the Sino-Tibetans speakers probably weren't just hunter-gatherers but had some kind of agriculture as indicated by this word for field and They have a cognate word for steel So that suggests that they had you know some notion of Private property because you can't steal if if you don't have a notion of private property So so we're starting to get a little bit of oppression of you know the Sino-Tibetans. They they had lies They had fish they had fields they stole from each other And what about their view of the afterlife? They had some kind of view of the afterlife So they they believed in in a soul Now I want to look very briefly at at lone words and so I should make some methodological observations You know we all Notice that you know the languages we speak for instance probably have a word for coffee or for tea that is Similar to words for coffee and tea in unrelated languages. That's because the word spread with the technology and These word lone words don't follow the inherited rules So looking at the word for horse And and and here already I'm starting to have a kind of polemical Context because for instance James Mattoff reconstructs a word horse for Sino-Tibetan noticing that Tibetan Burmese and Chinese have similar words for horse But they don't follow the right correspondence patterns And in a forthcoming paper with some colleagues we we propose that actually The word for horse in Chinese is borrowed from Indo-Aryan And if if you're incredulous about the phonology of that we can talk about that later And then I'll also just point out that the horse came along with the chariot And the chariot is you know the Indo-European Technology par excellence. It's probably you know why Western Europe and And much of India speaks Indo-European languages is it was it was the nuclear bomb of its day right this kind of Amazing military innovation the war chariot that led to the Indo-Europeans having a remarkable advantage on the battlefield and According to archaeology, which is what I say here, too It's it's pretty clear that the Chinese learned about the war chariot from Indo-Europeans of some kind around 1250 BC And along with it came the word for horse and chariot so let's say that maybe is interesting as details, but I think I want to to to make the methodological point right that they hear you have things that Befit the civilization Of the time period you have in mind Yeah, I'm almost surprised actually that the sign of Tibetans had some form of agriculture because it's so long ago You would sort of expect maybe they wouldn't But they certainly didn't have chariots, yeah And or horses the the horse You know was only used for eating for a long time And um and spread in Asia, you know under the influence of the Indo-Europeans around 1300 1200 BC So so it's it's it's it would be a bad sign If someone reconstructed the word for horse to sign a Tibetan But if we just reconstructed things based on what was archaeologically plausible That would be poor methodology, right? You have to have the linguistics As linguistics lead to the right answer and that right answer has to be compatible with the findings of archaeology So in this case, I think uh, you know, it can be done and it's that the horse and chariot are borrowed from Indo-Aryan Okay So that's oh, yeah, so that's it for this presentation And then I'll switch to the other presentation with the second half of the lecture Um, but just to sort of sum up we have A series of language families spreading out from southern China around the world and We have Two principles that guide us for studying history regular phonology and analogy and used judiciously they should tell us about the the the kind of physical and social world of uh prehistoric peoples And if everything's going well, uh the findings That we get through historical linguistics should match with the findings of archaeology And now I'm just going to go through Sino-Tibetan subgroups Hopefully it won't be too tedious and I'm not going to do all of them. There's about 30 Uh, I'm not going to cover uh, synidic. I'm not going to cover Tibetan but but uh, let's say partly Based on the other speakers in this um in this, uh, doctoral school Uh, my focus is in kind of southwestern China, right? So So and then and then I'm going to kind of um Do it like I did for the other language families in terms of go from the more isolated more more, um Small smaller groups to the larger groups. So first we'll start with Languages of fragmentary attestation Then Then Then Then Then So for most of you, these are probably just names, but you will get a little bit more Of a sense in just a moment So first we turn to uh these And there's two by long and pew So looking at by long by the by the the language by long is known from three songs That are preserved in the ho han shu And the songs were delivered at the chinese court Around 58 to 75 uh c e So the the the it's it's the language is not chinese but it's written in chinese characters And if we look at a few words, they're they're clearly Signed or Tibetan in some of the slides. You'll see trans Himalayan That means the same thing as sign of Tibetan. Basically, there's a kind of um, let's say a political Battle going on among scholars in terms of which one's the better term, but um, it doesn't need to concern us So we just look at some uh words. We have a word saw that means meet Looks very similar to Tibetan Burmese and Mizo. We have a word uh rui, which means rain We have a word b which means give so it's clearly a sign of Tibetan language and it um And it has relatively simple phonology and I think The the location which is sort of southern china And the relatively simple phonology makes uh w s koblin and christopher beckwith Say that a by-long is a low-lowish language Which would mean it's related to honey, for example Uh, but I don't think there is very good evidence for this so in uh in uh cigar So there's a kind of let's say an important paper an influential paper from last year about sign of Tibetan subgrouping Where they propose lexical innovations that are Kind of pinned on different Subgroups and if we look at Burmocangic, which is the branch above the low-ish The only word that they think is indicative of Burmocangic as a subgroup that occurs in by-long Is this word moose, which means heaven But this word doesn't just occur in Burmocangic. So old Burmese has it, which is which is Burma. So so Oops, I sort of I sort of mess mixed myself up, but First I'm proving that it does occur in pro in Burmocangic Then I'll show that it also occurs elsewhere. So in old Burmese. We have this word Mu'i, which means sky and in Japuk, which is uh A gyaronic language of Sishwan that is studied by Giam Jaak and is in many ways very archaic. We have this word tumum Which means heaven so So, yeah, so this word is in Burmocangic, but it's also found in other branches um, so Tibetan has Mu, which means a sky god And Rahuang, which is a numnish language has this Dumul, which is a celestial spirit now one thing that's actually kind of interesting to me And I'm not going to push here, but uh, you know, see what you think is Maybe it's the meaning heaven specifically that's associated with Burmocangic because you see that outside of Burmocangic It it means sky god So if that's true, then maybe it is evidence that by long is somehow Burmocangic, but it seems like a quite uh, quite a small difference to put a lot of emphasis on So my point here is just that um Yeah, that there's not a lot of evidence for uh by long being low-lowish And then another thing that that I just will flag to you that that I'm curious about that no one's really looked into Is in terms of complicated syllable structure. You see that Jopuk, Tibetan and Rahuang all have some kind of T or D At the beginning of this word which you don't get in Old Burmese or in by long That's interesting. Uh, I don't think the loss of it is particularly meaningful because it's quite easy to lose complicated onsets and may have happened independently, but it's more Uh, this general tendency that that as you get sort of out of the core area The syllables become more simple So both Burmese and by long have have Have simplified their syllable structure And I just want to make the point that because someone might say oh, no, no, no By long might have had complicated syllable structure But it was the the the problem of how to write this language in Chinese characters that meant uh, that that leads us as modern researchers to think it had um Simple syllable structure, but I'll just point out that in the in the in the sign of Tibetan peace treaty of 821 The Tibetan word stuck which means tiger Was written with Chinese characters as something like So when the Chinese wanted to they could write a complicated syllable structure even though they didn't have it themselves. So that's A point that that I think The absence of complicated syllable structure in by long We can think is a true reflection of the by long language So that's it for by long just three poems For Pugh Pugh is an urban civilization that preceded the Burmese in in the upper in upper Burma From the sixth through the 13th century and they are let's say among the earliest Buddhists in Southeast Asia And the oldest text in Polly was found in a reliquary excavated From the Pugh so they knew Polly Uh, Pugh has quite conservative phonology, and I'll just look at that by looking at numbers So, uh Pugh tech meaning one knee meaning two Play for four Banga for five The goal for nine and I've given the the Tibetan is just a kind of um I don't know standard point of comparison And then you notice that in some cases, they're the same and in some cases They're not like the words for nine in Pugh and and Tibetan seem really remarkably similar Right, so the the only thing that's different is really the manner of articulation. So you get voiceless in uh Pugh and you get the goo in Uh Tibetan And then um But there are there are differences like you don't get the G prefix in in the Pugh word for one And you get a P instead of a l in five So I don't you know, I don't have explanations for a lot of this, but I think it's uh It's interesting to look at and you see that uh, Pugh has quite complicated syllable structure as opposed to by long Oh, and then uh, let's look at water as well Do water and chew Uh, Pugh so is is preserved in in let's say around a hundred inscriptions. Most of them quite short And was not really worked on at all Uh between around 1910 and the last couple years But right now there's a kind of a renaissance of Pugh scholarship going on And I've I've put some citations there Uh, but there's plenty more to do. Yeah, so if you want to uh, I don't know look into Pugh, uh, we can talk. Okay, so let's talk about Those are our two Fragmentarily attested languages. Uh, let's talk about what their features are Because in some ways they're kind of have some surprising similarities So, uh, the first person singular pronoun in by long is K And in Pugh is gay. So they're quite similar And most languages of the sign of Tibetan family have a first person pronoun like nah So Tibetan does Chinese does Burmese does Uh, so that's interesting Uh And fun beak, uh, who who worked on proto kooky chin Doesn't didn't have any awareness of Pugh or by long C is Gay as the first person singular in kooky chin A kooky chin are um our languages spoken in in in northern Burma. He sees that as indicative of a kooky chin language Um Which would suggest that by long and pugh are also kooky chin But I just want to point out that that around the family. We do have these kind of similar first person singular, uh Pronouns sneaking around here and there. So oleka is a language Uh of, uh, butan And their first person is and then there's a language, uh, pushy chung Which has a first person So, I mean, I don't know. I don't want to you know, um Over stress this in terms of like pugh gay and And kooky chin gay are very similar But it's just to say I don't I don't that might indicate that pugh is a kooky chin language It might not the whole question of the distribution of let's say not like pronouns and gay like pronouns in the signative and family needs more study But uh, but but it's it is just interesting to notice that that our two tumor tumor spracha Have the sort of less normal first person singular pronoun Okay, so that was it for tumor sprach spracha now we look at mruik some so mruik is uh, is is it only has two members mru and anu Kong so it's spoken in the highlands of berma and bangladesh Uh, and has really almost not been worked on so luffler Has an article from 1966 and then david peterson of dartmouth is working on it now But he hasn't really published anything And uh, luffler does some some comparisons with other signative languages It's rare uh for distinguishing final r final l and final ya Although just and this is just me having looked at some of the words. It doesn't look like the l Words are necessarily inherited So I did someone say something Okay, um I mean, maybe this is just a Uh Kind of the problem with me knowing tibetan, but mru words look to me a lot like tibetan So if we look at pig pack in mru and pack in tibetan The word for house kim in mru and kim in tibetan And the word for bear tom in mru and tom in tibetan and um Let's say these words tend to occur all through outside tibetan, but oftentimes won't Look quite like this. So for instance in bermes the word for home is yim Not kim and the word for bear is warm Not tom or something like that. So You know what is going on here with mru? I just think it's It's it's it's I just wanted to you know say I think there's Mru is juicy and people should work on it more. Okay We'll look at the numbers So we have uh So, uh, you don't you know, probably your memory isn't perfect, but uh, if you look at taku And then let's go back to the pew numbers Uh, nine to call in pew the goo in tibetan. So quite similar, right? Which I think is indicative of of shared inheritance, right? This is to say mru is a very archaic language uh, and then I just uh want to make a proposal here, uh, which is so we have these taz These top prefixes, right? So I think that in six and nine It's etymological Where you get it for instance also in tibetan But probably spread via contamination To four and five and and contamination. I think I'll deal with this later But is is the process whereby languages? Sorry words that are part of a semantic subsystem influence each other The textbook example is that uh, whereas french has mal and femelle for male and female in english We changed femelle to female to make it look more like male So it's it's it's a way that words that are associated in meeting Influence each other's form very common in numeral systems For adjacent numerals And then does the you know, we saw uh, does does the chuh reflect the k prefix of three in in in in the word for three chum in mru gusum in tibetan, I think it doesn't uh, and I compare the word for tree in tibetan with this shing uh To ching in in mru, so it seems like mru just changed their s's to chos Okay, so that's it from mru Now on to nungish There are three nungish languages trung rawon and annong you've already heard about uh, rawon Spoken on both sides of the syno bermi's border So uh, just to remind you about the sort of trajectory of my presentation We're going from you know, tiny fragmentary languages to bigger ones So we're still at a quite small branch that only has three language families and the speakers of these families are switching to low-lowish languages So so I mean one thing that I think um Something is easy to forget in a kind of global context uh, where like english is in charge and then I don't know If you're in germany for instance, you would say okay. Well, we speak german and then in in In some parts of higher education. We speak english whereas if you're a Sorb, you'll say like okay, we speak uh In some in in in some kind of context We speak english in some kind of context We speak german and then in the home. Maybe I speak sorbian with my grandmother So there are those kind of layers all throughout the world and in in this part of The syno-tibetan border region actually lisu is the language that is taking over trung rawon and annong Even though, you know, international lisu might not be very strong. Anyhow, no reconstructions of of proto-nungish, although I think it's a Uh an exciting area to look also wouldn't be that hard because there's only three languages Now just deep breath now karen So karen, uh, is the most southern of all, uh, tibetan-berman languages And it has svo syntax like chinese all other languages have s o v syntax Uh, and it has very simple syllable structure About 20 to 30 languages 16 of them reasonably well documented Uh, just the famous ones sigar, po, kayan, li, weh karen is kind of exciting because it was one of the first, uh sub-branches that was worked on historically by, uh odrikur Who, uh, if his name doesn't mean anything to you, uh, you should look into him He is the person who explained the origin of tones in chinese Uh, he worked on vietnamese and he worked on, uh, proto-karen And identified that karen had a tonal split very similar to middle chinese to tai and to vietnamese And, uh, he reconstructed a series of voiceless nasals, uh, so things like There was later confirmed in field work by gordon luce So i think that's uh, how can i say a great endorsement of a methodology that i think is a cool thing about historical linguistics Is you can, uh, you can have the method lead you to posit something And then have it confirmed Uh later and you know the famous example of that is the the inter european laryngeals, which were reconstructed by desis or And then were found in hit tight But i just want to point out here is a little example in sino-tibetan The voiceless nasals of karen were predicted by odrikur and then were confirmed by luce And then, uh, there seems to be an isogloss characteristic of karen, not a very interesting one to change nga into nia uh One of the most interesting karen languages is is this one pao Because most karen languages have so radically simplified their their syllable structure that it's just There's no finals at all. They have syllables like pa and ma and ma But not pat and bak and bak But pao has nasals. So it has things like, uh, pang versus pang Uh, but unfortunately, it's quite understudied. So so i mean another sort of like motif of my talks will be like There are really low hanging fruit out there. So if you want to become rich and famous, you know, just, uh Pick a sino-tibetan language and and you can make, uh, you know Discoveries that will forever change the field So now that was it for karen Uh, now i'm going to move on to this family sol. So sol is a is and now we're getting to the bigger family This is a big family Uh, it was, uh Proposed by Robbins Verling On the basis of the word for sun. So sol languages have a word for sun that's like sol And uh other languages have a word like, uh, ní or or yeah, so ní is what you have in in tibetan and in in chinese There are three sub-branches uh, bodogaro cognac and uh jingpo luish bodogaro and cognac are both mostly spoken in india and then jingpo luish on the sort of chinese burmese indian border region So we're going to look at jingpo luish a little bit the most, uh I don't know the jingpo is is the more famous of the two pieces And it's being studied these days by a japanese study, uh, japanese young scholar of kurabe It's the majority language of kichin state in burma. So so similarly like if you speak, uh Something like lashi at home you'll speak, uh jingpo in the market in kichin state And it's also spoken in india and china it's one of the five Sino-tibetan languages that was used by, uh benedict in his reconstruction of sino-tibetan from 1972 and i think that's just i mean i don't i that work is So out of date now that it's basically useless But i just think it's maybe helpful to mention to people that jingpo i would even say of all the languages i've mentioned so far is probably one that you'll hear about more Because it has been used a lot in sino-tibetan And it has quite archaic features Including iambic syllables. That's you know, like we've seen with pu and with mru things like taruk is an iambic syllable preservation of final r and final j so that's it for the jingpo side and um Well, I think one thing that's been sort of frustrating at least to me about jingpo in the past Is that it's quite well documented. It's quite archaic. It's very interesting. It's been used in sino-tibetan But it there wasn't anything to compare it to immediately Like comparing jingpo and tibetan jingpo and chinese feels like it's quite far But this has just started to change with with work by this, uh guy whose uh name is pronounced Fujiwara Because he's japanese but he writes it using the old nationalist Transliteration scheme so you will see it in bibliography that who's iwara And he has been uh working on uh, so he's written a grammar of chak done a reconstruction of Luwish. So this is a group of very small languages also spoken in basically the same areas. And I personally think that work on the relationship between Jingpo and Luwish will be one of the most exciting areas of Sino-Tibetan in the coming years. So now comes Kukicin, the last two are going to be Kukicin and Burmachangik, which are things I know a little bit more about, but also are a little bit more complicated. So Kukicin, now you get a tree diagram. Don't need to go through all of this and I'll share all these slides. But anyhow, so Kukicin is spoken in where Burma and Bangladesh meet in the hills. So you have the Chin Hills on the Burmese side and the Chittagon Hills on the Bangladeshi side. And they have an interesting verbal system where they have verb agreement, so person agreement, like let's say like Indo-European languages. And they have an alternation between two verb stems. One, which we call stem one, used mostly in finite contexts and the other one stem two used in subordinate contexts. The central branch is the best studied with Mizzou as another one of Benedict's five reference languages. So his reference languages were Tibetan, Burmese, Mizzou, Jingpul, and Chinese maybe, I'm not sure if I remember the last one. Anyhow, it doesn't matter. And I will just point out, in case you've heard the words mean anything to you, Mizzou is the majority language of Mizzaron, which is a state in India, although most things happen in English in Mizzaron. And the change of R into G is characteristic of the Northeast branch. That's something that has been discussed a lot. Yeah, but David Pearson has shown convincingly that it's a late change that's spread through contact and is not indicative of a subgroup, per se. Proto-kugi-chin has been worked on. So somehow recently it's come back into fashion, let's say. So Toru Ono did a basic reconstruction in the 1960s, but then Chris Button and Ken van Beek both had reconstructions around the same time. And then I've also worked on Proto-kugi-chin a little bit. In general, the Northern languages are conservative in their rhymes and the Southern languages are conservative in their initials. So that's kind of elegant, isn't it? Existing reconstructions rely mostly on the Northern languages because more data has been available on them. One thing that happens, especially in Burma, but in this whole region is the languages where the population has converted to Christianity end up being better studied because people want to translate the Bible into that language and need to compile dictionaries and whatnot. So that's definitely what's happened in this case, which is that speakers of Northern kugi-chin languages converted to Christianity and then have been better studied whereas the speakers of Southern kugi-chin languages have retained their traditional religions and have been less studied, but that is slowly changing. And just last year, there was a kind of a dump of data on surveying a varieties of the Southern chin languages that has been shared on Zenodo, which is a European research council's research data repository. Okay, so that's what I will say about kugi-chin and now comes Burma-changik where things are quite complicated. So first, and here I will kind of really talk through the different bits, the stambam, the family tree of Burma-changik. I should say with all of these family trees, take it with a grain of salt. I have been very, in the past, very against drawing family trees because I feel like you shouldn't draw them until you're sure they're right. Each node, you should be able to say precisely what changed and know that it changed from this into that, but now I've become a little bit more relaxed and I feel like it's a good way of organizing information, organizing our thinking, and reveals the state of scholarship even if we have to keep in mind that most of this stuff could change, hopefully will change as more research is done. But anyhow, here's Burma-changik, which has sort of two halves probably, broadly. Na-changik, which then has three branches, Naik, Ursuik, and Changik, and Changik has four branches, Prinmi, Muya, or Munya, also Minyak. I heard someone here works on rituals among the Minyak, Urma, which is, used to be called Chang-chang, but there's good reason to call them Rama because that's what they call themselves and because they have no relationship as far as we can tell to the historic Chang people of Changik's history. And then the Gyalrongik family that will get its own slide later on. So that's it for the Na-changik family and then looking at Lolo Burmese, there are two branches, Loloish and Burmish. Loloish has quite a lot going on, which we'll look at later, but on Burmish it's quite clean. You have Burmik and Maruik split. And generally speaking, the Na-changik languages are more conservative and Lolo Burmese languages are a little bit more innovative in terms of their phonology, for example. So first of all, just to mention that the idea that Lolo Burmese and Changik have some relationship. So if you like this Burmo Changik hypothesis has been kicking around for a while. So you have some citations. Dempsey 1995 is the earliest that I've sort of run across. But this paper by Guillaume Jacques and Alexi Michaud in 2011, I think it really sort of started to be when this idea became something more widely accepted. And even now, I think there are a lot of Sino-Tibetanists who don't believe in Burmo Changik, but I do, so you'll get to hear about it. And then I just discussed these subgroups, Naik or Suik Changik, including Lolo Burmese, and then let's just mention what are the ancient languages in this family. Most of the, I think, of the families I've been mentioning so far to better Burmese branches have not had ancient languages. So there's Tangut from 1036, which is Gyalranic. There's Burmese, first attested in 1113, which is Burmese. There's Yi, first attested in 1485, which is Lolo-ish. And then there's Nashi, and people like to say that in China that Nashi is as old as the hills, but there are no manuscripts that are obviously from before the 18th century. So let's say a Nashi is from the 18th century and it's a Naik language. So more or less, each branch has its own old language. So some features of Burmese Changik are plain versus velarized vowels. I'm not a phonetician, so I won't be able to do these, but I think, let's say, in a language like Zbu, if you take the vowel ah, this would be a difference between a real ah and an ah. And then a complex system of directional prefixes that double as past tense markers. And this is a very fun thing about these languages where, for example, for a motion verb, you'll have go and then you'll have prefixes that mean downstream, upstream, towards the mountain. It can vary between different languages. And then for verbs like eat or, I don't know, things that don't have it to do with motion, it tends to be that one of those prefixes will be associated with that verb in the past tense. So it might be that you eat downstream. Although, you know, semantically, that's not what's going on. And I personally think that the system is sort of comparable to the perfective prefixes that you get in Russian, where like basically every verb has a perfective equivalent, but the prefix that it uses can be different. And those differences do have some semantics, but are largely just grammaticalized. And they have an inverse agreement on the verb. And I will just say a tiny word about inverse agreement and you'll tell me how good my explanation is, but it's something that took me a long time to wrap my head around. So don't worry about it. But here's what it is. In European languages, we tend to index the, let's say, the subject. So you have like, let me think of a nice Indo-European language like Spanish, you would say, tengo tienes tiene, right? And then you know, oh, I have, you have, he has. And the ending indicates who the subject is. In a language that has inverse agreement, you have a sort of assumed way of things normally happening. And that is a hierarchy that goes first person, second person, third person. So it's much more likely that I will do something to, for example, an inanimate object than that the inanimate object will do something to me. And then the person marking marks the involvement of the person in any way. So like if I, I don't know, if I give you a hug, it marks, it would be first person because I'm involved. And if I'm involved, it's always first person. If I'm not involved and you are involved, then it's second person. And then only if neither you or I are involved, is it marked as third person. And that involvement can be a subject or object. But if the relationship is the reverse of what would be normal. So for example, if I say I ate a cheeseburger, then that's a normal relationship. But if I say the cheeseburger ate me, that would be an inverse relationship where third person is acting on the first person. That would be marked as first person and inverse. So that's my simple sort of attempt at describing inverse marking, which is that in the West, we ignore the object and we mark the subject in Bermatronic languages that have person marking. You mark the person regardless of their role. But if the relationship is the reverse of what would be normal, then you put the inverse marker. So that's quite an interesting system. The more conservative languages show all of these features. And, but in many sub branches, only relics of them survive or they're only indirectly observable. And I don't think that this means much in terms of sub grouping, but because let's say complicated, hard things are easy to lose. So where you see them, they're old, where you don't see them, they've been lost, but they could be lost in different ways at different times by different languages. So in the 1960s and 70s, the focus of research was on low-lowish languages, largely because China and Burma were inaccessible to researchers. So they worked mostly in Thailand. And although Thailand is not a place that has a lot of Burma-Chung languages, even of the low-lowish branch, those Tibetan or Burma languages that are spoken in Thailand tend to be low-lowish. So that's why people in the 60s and 70s worked on low-lowish. But I would say today, it's in the Chungek side where you see the vanguard of neo-Gramarian progress in trans-Malayan reconstruction, which is particularly the Gyal-Rang languages, but more generally Chungek is where exciting things are happening in Sino-Tibetan. And I think both that they're very exciting languages, but I also think it has to do with the people who are working on them tend to be quite good and diligent in their work. So just to talk you through sort of some, to prove my point, this is where progress is happening recently. Yeom Jacques in 2014 did a comparison of Tangut and Gyal-Rang in Phenology and Grammar. And by bringing those examples together in a sort of systematic way, his student was able to notice that the Tangut grades, this is a term from Middle Chinese Phenology, which either if you know about Middle Chinese Phenology, you'll know what it is. And if you don't, consider yourself lucky. But anyhow, Tangut is analyzed like Middle Chinese because that was the phonological tradition that Tanguts had access to. And they analyzed their vowels according to grades. And Gongshun discovered that these grades correspond to the velarized, non-velarized distinction in Gyal-Rangic languages. So that's, I think, a real step forward in sign of Tibetan historical phonology. And then similarly, also in 2020, Nathaniel Sims, who's a PhD student at Santa Barbara, published a paper where he showed that the tonal contrasts of Northern Chang dialects correspond to the two Tangut tones. And this is interesting because Jonathan Evans, who had worked on Chang historical phonology, thought that the non-tonal varieties were conservative and the tonal varieties had evolved tone in some way. So Sims shows that the opposite is the case, that it must be that the non-tonal Changut varieties lost tone because there's a systematic correspondence between Tangut tone and Changut tone. So that means we can reconstruct tonal distinctions and this velarized, non-velarized distinction, maybe up to the Na Changic level, it's a little unclear, but those features are old and the details are getting figured out. So I mean, just 2020, good year for Burma Changic historical phonology. No real work has yet been done at the Burma Changic level itself, which is to say, so stuff has been done on Lolo Burmese, stuff has been done on Burmese, stuff has been done on Gyorong and on Changic, but no one has really brought that all together. And I think that makes sense because there's a lot of work. So here's the Naik branch, the smallest probably. So you have Naik and basically, unless I say otherwise, this is all Southern China we're talking about, from Sichuan Yunnan. So Naik has three branches under it, Namu Yi, which is one language, Shumi, which is one language, and Naish, which is three languages, Nashi, Na, and Laze. Generally not a well-researched branch. So if you're wondering what to do with your life, maybe work on Naik. So there's Namu Yi, just a couple articles about. Shumi is currently being worked on by Katya Chirgava. And then for Naish, Nashi is the best known. It's famous for its representational writing system that is used in liturgical documents by the Dongba priests. And yeah, and Alexei Michelle has a sort of survey article about it in Brill's Encyclopedia of Chinese Language and Logistics, which is a very useful resource for any of you who don't know about it. And then Yongning Na, I think is the best, which is, you know, it's easy to mix up, but Na is one language and Nashi is another language. So Yongning Na is the best studied Naik language where an overall grammar was published in 2010 and then a book specifically about tone in 2017 by Alexei Michelle. And they have a very complicated tone-sandy system. And then Laze is really not very well studied. There's one article by Alexei Michelle and Giong Jacques. But yeah, I was surprised how little I could find out about Laze. So Naik Reconstruction, Giong Jacques and Alexei Michelle made some preliminary efforts in 2011 in an article published in Diachronica. And Lize of Peking University has done some forays into Naish reconstruction since then. I found his work a little paradoxical because he criticizes Giong Jacques and Alexei Michelle for making comparisons with Yaron because, you know, if you're reconstructing Naish, you shouldn't peek at Yaron, which was kind of their methodological point where they were saying it's useful to have an archaic language to interpret more innovative languages. But Lize himself uses Tibetan comparisons in his reconstruction of Naish, which is even further away. So it seems a little bit unfair of him to criticize Alexei and Giong while kind of playing the same trick himself. But in any case, I think that we won't see a lot more progress in proto-Naik until some of the languages are better understood, Laze in particular. Okay, so that's it for Naik. Now we're moving on to Ursuik. And this, I hope, will be a particularly fun one. This is a very small branch. We're in Burma Changiq, right? So Ursuik has three members, Tosu, Ursu, and Lizu. Lizu with a Z, not to be confused with Lizu, which is a low-loish language. So three languages, like I said. Dominic Yu in his Berkeley 2012 PhD surveys previous work and provides an initial reconstruction, but most of the changes he discusses like this a-to-e, which we call brightening, is not very diagnostic. All the languages in this part of China changed a-to-e, basically. There's new data on Ursu and Tosu since Yu's study. And I will now look at that a little bit. And this was my, I don't know, to try and spice things up and not just have it be, oh, long lists of names in minority groups in Southern China will do a little bit, kind of an exercise in historical linguistics of Ursuic reconstruction together and then see some of the methodological principles we discussed earlier. So if we look at the words for son and daughter in Proto-Ursuek in 2012, Yu reconstructs Z for son based on these two forms. And he reconstructs two forms for daughter, Ziyou and Zieyi, based respectively on these two forms and then the next two forms, you see? So the first two support Ziyou and the second two support Zieyi. Okay, well, this is a methodological point that I will devote a whole lecture to basically, but Proto-Ursuek did not have two similar looking words for daughter. It doesn't happen in languages, right? Like either they had different words, one that meant daughter of, I don't know, older daughter and younger daughter or one man daughter and one man daughter-in-law or something like that. Or there's only one reconstruction and he has a formal problem, right? Either there's a formal problem or a semantic problem. That's the point. But if someone reconstructs two different similar looking words, meaning the same thing, it means they've done a bad job. That's my view on the matter. So then we're going to try and solve this problem right now. So that's unsatisfactory. So I think Ziyou, that's his first of the two words for daughter is innovative. Because the beginning of it is the same as the word for son. So you could imagine that it's son plus some suffix. And indeed there is a diminutive suffix yo that comes up in the word for calf and the word for mouse. These are in his reconstructions. So Ziyou is analyzable as a compound of a word for son and a diminutive suffix. So like you would call your daughter, your little son, something like that. Or you could think maybe child and little child or something like that. And the association of feminine meaning with diminutive is found all throughout the world, including in Sanskrit. So I think the semantics works. And that means I can drop Ziyou, right? Like if we have these two words, Ziyou and Zieyi, Ziyou is the newer one and Zieyi is the older one, yeah? And then this also means incidentally that mianning Li Zhu and Kala Li Zhu as documented by one person, by Huang Bufan. And first, like are a subgroup, why? Because they changed what their word for daughter was because they both support the newer word for daughter. And something like changing what your word for daughter is is a pretty big change. So I think that's very strong evidence of subgrouping. So let's say that's one thing that I will, I'm trying to, that I'm trying to make a kind of methodological point of is Dominic, you just said, oh, there are two words for daughter and proto-ursuit. But we can do better. We can squeeze a lot more juice out of this. One is by figuring out that one of the two words has to be younger than the other. And a corollary of that is any languages that support the younger reconstruction must have had a period of shared history. So must be part of a sub-branch. Together, whereas if you've preserved the old word for daughter, doesn't mean anything. Okay, so now we just have to look at this other word for daughter, Zieyi. So let's do that. Well, in a conference paper in 2019, you seems to update this is on the basis of new Tosu data. His reconstruction from Zieyi to Zanyi. So let's see why he did that. He did that because the Tosu words for son and daughter are Zieyi, which we are not surprised by, and Zami. And then he reconstructs the, if you like the roots back to Zieyi and Zan. So I'm not quite so sure on why he does Zan, but by implication, he sees the Yi that we saw in his earlier reconstruction and now in the Zanyi form as unrelated to the Mi in Tosu, right? There's just by implication, he doesn't say anything about it and they're not the same. So they must be unrelated, right? Well, I think that's, again, not very elegant. So let's look at this Z versus Zan, because you wouldn't really expect a language to have, you know, how can I say, similar words for son and daughter that aren't following some pattern at least ultimately historically. You know, you would want it to be like child plus feminine suffix or something like that. But you explains that he sees here what are called alophams, which just means variants in Matassas-Tibetta-Burman reconstruction, Zah versus Zan. So his explanation is that Zah stayed, Zah, no, and then when the ah to e sound change happened, the Zah changed into Z. And then the Zah is some kind of variant word for child that the end suffix kind of protected it from the ah to e change, which is why you then have Zami, that's his explanation. But I think this is not very likely where if we look at the similarity between Tosu Zami and Tangmi Samai for daughter, and Tangmi is spoken way over in the middle of Nepal. So I think it cannot be the case that the similarity between this Tosu form in Sichuan and this form in Nepal, it must be inherited, it must be a sign of being old. So I think that means the ahn reconstruction should be updated to ahm, first of all. And I would explain that Proto-Ursuek first had the Zah son and the Zami daughter, quite similar to Tangmi. And then brightening, that's this change of ah to e only happened in open syllables. So then you get Z and Zami, right? And then just by way of comparison, Tosu has another word, nama, sister, which he, you, reconstructs to nami with two M's. And why does he have two M's there? Well, the first M he's using to close the first syllable in order to block the ah to e change. And the second M he's using to explain the ma suffix. But I don't think it's necessary. I think why not use the one M for both purposes? And then you could just reconstruct namae. And then here I'm just pointing out that then structurally speaking, it's exactly the same as Zami and namae from, you know, nama basically, yes. It's not rocket science. And then if you compare it to Ben has a word namae, which means young lady. So I think that's related to this word for sister. Okay, so that was kind of a mini, I don't know, a lemma, if you like, a distraction, but proves a point methodologically, which is if you see someone reconstructing two variants for the same meaning, they've done something wrong. And it could be that solving the problem is quite easy. So I mean, this, like, I don't know anything about these languages. I came at this totally naive, you know, reading his PhD, like any of you. And if you come to, if you just read a PhD naively and say, okay, the only tools I allow myself are historical phonology and analogy. And I can notice people's mistakes, like when they reconstruct two words that are similar in form and the same meaning, then you too can make rapid progress in improving our understanding of sign of Tibetan linguistics. So it's just a question of commitment to the right principles, I think. So now let's continue on with Changiq, which I haven't made a slide of because the structure is so simple. We have Prinmi, Munya, Ramah, and Yaurangik as our four branches of Changiq. And then Yaurangik itself, so now we're three layers down, right? Burmo Changiq, Changiq, Yaurangik is a little more complicated. And this one, I think, is worth talking through. So we have two branches. This is based on very new research of a group of people, but with the lead author is Lai Yunfan. So we have West Yaurangik, which has Tongut, Horpa, and Khrushchap. And then you have East Yaurangik, which has Situ, Japuk, Sovdun, and Zbu. Zbu is probably the most archaic language, I don't know, in the whole Burmo Changiq family. But Japuk is the most easy one for anyone to get the grips with because it's what Yom Jaq has been writing about very copiously for a long time now. And then you see where Tongut is as well. So Tongut is there next to Horpa and Khrushchap. So looking at West Yaurangik, it's confined to Sichuan except for Tongut. I think Galambos 2015 is a good introduction to Tongut studies in English. The Tongut Empire lasted from 1984 to 1227 in today's Ningxia, and they were destroyed by the Mongols, but also killed Jinx Khan, so at least it wasn't a total loss. And then moving on to Horpa is a nice survey article in this Sino-Tibetan linguistics book published by Rutledge in 2017. Khrushchap is worked on by Lang Yunfong, that's his kind of fieldwork language. And then East Galarannik, there's a nice survey article again by Yom Jaq. It's first attested in a Huayi Yu vocabulary, which are the Bureau of Translators in, I don't know what to say, in pre-modern China, compiled vocabularies of foreign languages, and one of them is a East Galarannik language, although exactly which one isn't totally well understood. And then there's Situ, Jopuk, and Sovdon, as I've mentioned. And now I just want to say, and this, I don't know, I will sound a lot like Yom Jaq, but I think the Galarannik languages are extremely important in Sino-Tibetan linguistics for giving a model for morphology in the way that Sanskrit did for Indo-European. And I'll give you an example, like in English we have these causative sit, set, drink, drench, but you wouldn't know what to do with them in English. You're just like, oh, okay, you change the vowel somehow, and then it makes it a causative. But in Sanskrit, it's very clear you add the aya suffix. And then that syncronic fact in Sanskrit becomes a key to unlock the diachronic history of that causative formation across the language family. And I think that Gyaorong will play a similar role in Sino-Tibetan languages for having a lot of morphology that's still synchronically productive. Okay, so now we're done with the Changik side and we move to Burmese. This is the Burmese family. It has two sub-branches, Burmik and Maruik. And then Burmese just has two sides, all Burmese, and then all the modern Burmese dialects under it, and then Nochan. And then Maruik has a bunch of little languages, Pella, Lachit, Lavo, and Zawa, but in order for these not just to be words, remember Jingpoa was the dominant language of Kachin state in Northern Burma. These languages are the less dominant languages of Kachin state in Northern Burma. So basically Kachin state has an ethnic and linguistic diversity. All of the little languages are Maruik, so Burmish, and the dominant language is Jingpoa, which is a Saul language. So Burmish is circa seven languages in Northern Burma and adjoining regions of the PRC. And then you just saw this list. And so let's look at, and this is an area I've worked on so I can give more detail than other areas, is this split real, the two sides, Burmik versus Maruik? Well, Burmik languages merge the pre-glottalized stops of Proto-Burmish as aspirates. So if we look at the word six in Proto-Burmish, it's Krok, and so in Burmese it's Krok, in, I'm not so great at reading IPA, so please forgive me, but in Longchon, Nochon is something like should draw, but with a high tone. And then in Zhaowakiu, in Pella-Kiao. Now if we look at Frighton, and you notice that it's very similar, but with a pre-glottalized initial, Frighton, Krok, exactly the same, like so which is a Burmese cannot distinguish the word six from the word Frighton and neither can Nochon, Nochon, which is a good evidence of a merger. But you see that in Zhaowakiu and Pella, they're different. So they're not aspirated and they have a Krikiu vowel. And that is why we have to reconstruct there as being a difference at the Proto-Burmish level, and there's a clear Burmik isogloss of merging that distinction. So that proves that what I'm calling Burmik is a real branch, is a real clade, but previously, and where previously is just a year ago, I expressed reservations about the reality of Maruik because Yoshio Nishi, the guy who first proposed this Burmik versus Maruik distinction, he only pointed to evidence that Maruik was, sorry that Burmik had this isogloss I just discussed of losing the pre-glottalized stops. But that doesn't mean that all the other languages group together. But subsequently, I have found together with Gungshun who was working as a postdoc in London, what we're calling the chicken mouse split because it affects the words chicken and mouse. So Burmiz and Nochong, the word for chicken goes back to Grak, but in Lavo and Pella, it goes back to something like Arrakh without any evidence of the G and mouse is very similar where it goes back to Grok in Burmiz and Nochong, but it goes back to Orrok in Lavo and Pella. So you immediately say, wait a second, are you saying that proto-Burmish had two similar words for chicken, one Grok and one Rakh because that's exactly the kind of thing you were criticizing in terms of what Dominic Yu was doing. And I would say, no, that's not what we're saying. We probably would just reconstruct if you like a different initial here like a schwa after the G and say that proto-Burmish had Grok and that the Maruik languages just dropped minor syllables at the beginning of words. Whereas in Burmish, in Burmik languages apocapy meant that you just the schwa and then you have the proper cluster. Something along those lines is what we would propose, but we haven't published this yet. So I don't know. So if you have a bright idea, you can tell me. In any case, the point is it shows that Maruik has innovated together. Both Pella and Lavaux have treated these words in the same way. And that means Maruik is also a sub-branch. And at the beginning when I was saying that I was skeptical of all these nice tree diagrams it's because I think we need really for each spot in a tree diagram an explanation like these where you say these languages all share some innovation that's indicative of that place in the tree. And generally speaking, we don't have them. So, which is why I haven't been giving them. But also it's a way of shortening the presentation. So now about pun, I just want to mention this one particular language which is dead unfortunately. It was worked on in the 1980s and then they were in 2007 when Utum Aung Chaw looked at it. No one spoke it anymore. No one could really say anything except that they knew some words from, oh my grandmother used to use this word. But it's a really interesting looking language because it has these ta and ka prefixes of the type that you've been seeing in Pew and Maru. But here in a, oh no, my Burmese script didn't turn out. That's annoying. So you don't get these prefixes in other Burmese languages. But just take a look at some words. You have kali, wind versus le in Burmese. Kashi, wood versus sach or something like that in Burmese. Tami, fire versus mi, fire, tapah for frog versus pa. And it doesn't seem predictable whether it's the ka or the ka. So I think this is very archaic. Similar things happening in Yaranic languages and is an area that I think, again, progress could really be made in terms of these ka and ta prefixes. So then looking at Burmese, and Burmese came up in my 2019 book. So you can look there for more details. Tested first in 1113, lots of stone inscriptions, basically of kings giving land to monasteries. And there are a bunch of Burmese dialects. And I just think it's important that we keep in mind that it's not like, oh, there's Burmese, the national language of Burma, but that the national language of Burma is only one of these types of Burmese that are quite different from each other. And finally, we're looking at Loloish, which is where honey is. This is Loloish, and Loloish is how can I say, one of the good spreaders. It reminds me of like the Bantu languages in Africa. It's a sub branch that has really spread very broadly and has lots of different members. So here is a Stambaum. I won't talk to you through all of it, but basically, this is based on Bradley. We have Northern, Central, Southern, and Southeastern. But Gerener says the classification of Loloish languages must be reexamined in the future by considering more data sets and also grammatical features. Which is, so basically this is, that Stambaum is what Bradley wrote on the back of the envelope. Needs more work. And I also think it's important to emphasize, I've managed to avoid so far, the classification of people according to official PRC ideology, but the mismatch between languages and so-called nationalities is at its worst with the Loloish languages. With the particular absurdity that speakers of Khadzu are classified as Mongolians. And my understanding of how this happened is, they sent ethnographers out into the field in the 1950s to find how many peoples were in China. And they first just asked people for their auto-nims and came back and said, oh, there's about 300, 400 ethnicities. And the government said, that's too many. Let's get it down to like 50 or something like that. And then, so they tried and the Khadzu basically had the choice of being Yi or not and they didn't want to be Yi. So they said, well, what else do you have on that list? And they had some kind of legend about one of their chieftains being descended from the Mongols. So they said, okay, we're Mongolians. But now that means that in PRC official statistics, they're Mongolians, which is ridiculous. Anyhow, speakers of Honey, Lahu, Lisu and Jino have their own nationality. So that's to say, in the official Chinese list of nationalities, there are Honey, there are Lahu, there are Lisu and there are Jino. But other speakers of Laloish languages are generally grouped under the category Yi, regardless of how close their particular language is or not to each other. And the sort of official Yi language is Nosu, which also has its own script. So and there are a number of languages, Laloish languages that are called Yi in China that used to use a logographic family of scripts. And I'm speaking in that way into the family of scripts because it's not like there is a classical Yi language. There are different Laloish languages that use different syllabaries. It's all quite messy and not very well researched and was never used for widespread administration or education in the way that like Tibetan or Uyghur war, but predominantly for religious texts and kept within families. So now just looking at the branches altogether. In Northern Laloish, the big language is Nosu. In Central Laloish, you have Lahu, Lisu. And in Southern Laloish, Hany and Aka. And these are just to give you a sense of the names of investigators who've worked on these. And then particularly for Hany and Aka, there's this book series that I just got my hands on recently that goes county by county and looks at vocabulary mostly of Hany and Aka varieties in the PRC. And then Bissoid has the Bisu language, which was discovered by Nishi as the most prominent member. And then Southeastern Laloish has Fula, which there's a book about by Pelki. Just as sort of if you like an observation, maybe a criticism of work on Laloish is, people who work on Laloish are really into comparative vocabulary lists. And the languages can be quite similar. So you get these kind of hundreds of pages of like, well, in this town they pronounce this word this way and in this town they pronounce it this way. Where like not a lot of discussion of grammar in a way that would be maybe also useful. In terms of Laloish reconstructions, Bradley reconstructed Proto-Laloish in 1979. It was out of date at the time of his publication as pointed out by Thurgood. The reconstructions failed to predict the tested forms. So I'm not particularly impressed by it. And I think that Tatsu and Nishida has done some important work in Laloish comparative reconstruction that hasn't gotten enough attention, mostly because he writes in Japanese. And I think that's generally true of sign or Tibetan linguistics, is that the contributions of the Japanese, especially let's say before 2000 were overlooked and have not been, still have potential for having a good impact on the field. And then Jacob Dempsey in 2005 suggests a new approach to the Laloish tonal split, which I also think is very interesting and hasn't gotten much attention paid to it. It is worth mentioning that James Mattoff is closely associated with the Laloish tonal split. And that's seen by many as kind of one of his real major contributions to sign or Tibetan linguistics. I've never been able to make much sense of his book about it, so can't really endorse it. I just don't, I'm not sure what he actually thinks and hope that one day he or one of his students states it in a way that's easier.