 Okay, welcome everyone and again and thanks for joining us at this so as linguistics webinar, our presenter today is Richard Dockham, who is currently a visiting assistant professor at Swarthmore. Also relatively recent recipient of a PhD from Yale we worked on the project that we're here about today quantitative historical linguistics and phenology on primarily the Thai languages but I think also some other languages of Southeast Asia. There's also a longtime resident of Bangkok which I think explains some of the connection to these languages and his personal experience in that area. So we're really great to be able to have you here today to hear about your research and the sort of wealth of experience and ideas and techniques that you put into this working to to hear a piece of all of it so thanks for joining us and look forward to your presentation. That's a nice introduction yes nice spent basically a decade living in in Bangkok so it's a is certainly informed a lot of my work okay are you able to see the slides I've put up now. Yes, it's good. Okay, great. So the title of the talk today is the East Asian voicing shift the segmental phenology and total reconstruction in Southeast Asia. So, this is really part of a much larger project, not to, you know, to my horn it's just that there's so many bits and pieces of it that we can all fit into one talk but there are about one interesting bit. So the quantitative, this won't be so much on the quantitative side but you, there is work on that that I presented elsewhere which if folks are interested in in that side of things but yeah so today we're going to focus on this one, one angle. I would also say greetings from Pennsylvania, although I, I have spent much time in Southeast Asia, I have been back in the US for several years now, and I'm at Swarthmore Swarthmore, they're like four or five local pronunciations Swarthmore Swarthmore yeah. Anyway, I'm here at Swarthmore, and we are actually part of a try college linguistics department so it's these. There are many small colleges, these are actually all established by Quakers in the 19th century and so, even though they're not formally affiliated with the Quakers anymore that is the who started them. And so my office is right next to a Quaker preschool on campus there, but we have these, because we're relatively small schools we have built a department across three campuses so these are our lovely little schools here so hello from here. And on the left is also the one behind me. That's my local campus. Okay, so a roadmap for today. We're going to talk about lexical tone Southeast Asia as a convergence area, and then some things about how tone arises, and then sort of the meat of the of things is the East Asian voicing shift and how that has played out in the dye languages. And then I'll also talk about some of the evidence from a pig for that. So just a quick note about pronunciation. You may know that the word tie spelled with an H as in Thailand, which is the national language of that country. When you see it written without an H here, whether I pronounce it as tie or tie. It's actually an unaspirated T but I may. I sometimes am not very conscientious about which one I'm saying typically when I'm sent whenever I'm referring to this. The branch or the family that tie belongs to, as opposed to tie with an H is the specific language in that family so that's one small point of clarification to make. Okay. I would also, of course, like to thank people who I've worked with on this project. So, Ryan German as a PhD student of James Kirby's at the University of Edinburgh and I have the dissertation defense draft I know it's almost at the very end of the progress of that process and looking forward to seeing the work that he continues to do but he's worked, I've worked with Ryan on developing some of these ideas with related to the East Asian voicing shift we had a paper at the LSA in January of this year. If anyone wants to look at that poster for a little more detail. And also the National Science Foundation, Swarthmore College and the International Center at Yale have all funded different parts of this work also. Okay, so what is tone. Well, let's just see one sec. Okay so their tone is used to mean different things. In this case, we're talking about the use of primarily pitch to encode meaning. So we can talk about the difference between lexical tone and grammatical tone, where lexical tone typically distinguishes the core meaning of words right so phonemic tone. And then that is a different thing from grammatical tone which also, you know changes the meaning of things but is typically seen as a part of inflection so it's a morphological thing. And so I'll be talking about lexical tone in this talk. Now let's take a peek at the walls map. There's actually a lot of tone around the world. It dependent depending how you count this is a figure that Larry Hyman likes to cite the basically half or more than half of the world's languages, make use of tone in some way. Now we don't need to argue about the exact percentages but you can see here from this map, where the white dots represent no tone, and the pink and red dots are simple versus complex tone. I don't love the same simple versus complex dichotomy so much but basically that just means two tones versus more than two tones under the walls classification. It's a nice snapshot to see that they're really these three big areas so you have Africa. And then Southeast Asia, and then the Americas where a lot of tonal languages pop up. And of course I want to share with you a fun tyton good twister. So this is not actually a tongue twister for speakers of the language because it's, it's just part of the regular phenology but this would be my my my my my my my my my so sort of a question and answers this part is the question, which means is the new word my my my my my, and then the answer is my my my my my no, the new word is not burning. And so you can see I snuck a little vowel length in there to allow me some extra words but, but this is other than the long vowels and the short, the long versus short ah this is four out of the five tytones are in this one sentence which is pretty fun. So it's a somewhat famous sentence that just is used to snimey language learners I suppose. There's also a little bit about I could say about typology of tone languages more generally to go a little bit beyond the simple versus complex which walls adopts. We can talk about level tones sometimes also called register tone, where it's generally conceived as being multiple tiers multiple pitches of level tones those can be combined in different ways. There's also some talk about contour tones, where you have movement of the pitch across the word, but that also combines with level tones in many languages. And then we also get a more complex combination of things where we have things like breathiness creakiness. So what we could put under the single category phonation sometimes so contour tones combined with phonation to create even more fun. And so we often see that one pattern is is dominant in a particular region so we'll talk a little bit more about how that plays out in Southeast Asia but this is, you know, a general typology that you'll you'll see cited many places. Okay, so mainland Southeast Asia and I put the south in parentheses because our general geographic labels of East Asia, Southeast Asia, South Asia, like they actually don't line up super well with the linguistic typology. And so the area I'm talking about is sometimes called mainland Southeast Asia, sometimes also we talk about greater Southeast Asia which includes typologically convergent parts of Northeast India, of Southern China will include Maritime Southeast Asia, it really depends on what features you're talking about. But we'll talk about some of the features that can define this as a as a convergence area. So I'll stick to the term Schrock wouldn't that's also available but I'm not so confident in my German pronunciation so I'll stick to the, the linguistic convergence area. People have called it the many things whether it's a Schrock went or a diffusion area wouldn't really be linguistics we weren't arguing about what term to use to call a thing but the idea is that of course there are some features that cross family boundaries. It grows in multiple locations multiple situations within a family, it's not due to a single family with that spread the feature across multiple languages. And so mainland Southeast Asia is really one of the classic convergence areas because it's a geographically huge. It has a huge percentage of the world's population in it. And it just has some really, really remarkable cross linguistic convergence. And so some of those things include the syllable shape so monosyllables and also sesquic syllables this is a term coined by Madison, which means a syllable and a half. So it basically means like a sort of I am where you have a minor syllable that's not stressed and then but it's really like treated like one syllable in terms of the language so you'll get a tone, a single tone on the, the stress portion. And so it's not big enough to be called two syllables hence the sesquicillable label, but between monosyllables and sesquicillables that covers a lot of the, the syllable structure in the region. We'll also talk with tone and register sort of famous I'll tell you a little bit more about what register is if that term is not familiar to you. I see a lot of isolating morphology and then syntax I just picked one famous feature in the area is the prevalence of numeral classifiers these are these obligatory counting terms that appear with numbers so you don't, you know, we have a handful of these in English that you might say three loaves of bread, rather than three breads or five years of corn rather than five corns, but it's really just obligatory with essentially every noun in this area so that's another sort of famous convergent feature. Five language families. And of course there are more languages than just these five these are the five major ones and look at the majesty of this this is a, a, a, not a perfect map but a really pretty map. So it splits Sino-Tibetan into Chinese, Tibetan and Korean, but you can see here that it has all of the five families represented on here. This is the language family that Thai belongs to you also known as Thai, Thai could I if you've heard of that term, but we have Austroasiatic, Austronesian and Momin rounding out the five. And so if your geography is a little shaky this is, this is Southern China. This is Burma or Myanmar, this is Thailand, Cambodia down here, Vietnam, and then Laos, kind of in the middle there and this is part of Malaysia down here. And so then this is also this actually India there. Okay. So, what is convergent. Well, this is one of the big puzzles of dealing with an area that is so huge and, and is so has so much convergence across the family. What is inherited from common ancestors and what is just simply parallel evolution. Okay. So the classic puzzle whenever we see something that pop up in a language or in multiple languages right we have to decide. What is the best explanation, can we attribute it to shared innovation meaning it's, there's a single recent ancestor, it's part of one recent lineage that split off. And so we would want to group those together as part of a single language group or subgroup. So there's some more more distant common ancestor that explains why we see it in geographically or typologically disparate languages. So they can have inherited it from some distant common ancestor, or we can just have chance resemblance so it could be something like universal linguistic features or something about the physiology of the human vocal tract and the the acoustics of speech there are all these kind of other pressures that can be attributed. We used to explain common out commonalities across languages we can just call that chance resemblance where chance is not truly chance but it's it's combining I'm just using it as a cover term to mean a set of other things that are not due to shared innovation, or shared retention. Okay. So let's jump back into tone then so originally early on it was. We didn't know that much about that many languages in the world to be honest so the number of total languages that we had data on was really quite few. The earliest use of, I guess the earliest record of tone in Western linguistics is probably only less than 200 years old I don't want to say that no one was talking about it before but we have like an 1820s grammar of tie that talks about it we have like dictionary notes and sort of colonial records that make mention of it in in the area, but I, we know from, we know from Chinese linguistics of course that they've been dealing with tone for much longer but those. Yeah, the, the two worlds kind of combine though Chinese linguistics and Western linguistics in one figure which is leaf on way. With 1902 day in 1987 he was actually Edwards appears first graduate student, originally at University of Chicago and then once appear moved to Yale, Lee went with him. And so in 1928 Lee finishes dissertation on on my toll, my toll a I'm not actually sure exactly how the language is pronounced. I think it's my toll a. This is an Athabasca language. And the interesting thing is Athabasca languages are tonal. And so there was very early on abandoned not to you know fairly quickly this idea that maybe there's a single tonal super stock and Athabasca languages could be connected to So this is sort of how early into American linguistics and Western linguistics. This kind of work goes, but this is not held up of course even though the funny thing is, we know at Athabasca languages are tonal but the tonality of the sign of Tibetan stock is the one that actually ended up more questionable. So we, we learned a lot more sense then which has clarified our understanding so we know that tone is a much more common than we previously thought so that would cast doubt on the idea that there would be a single origin for tone. We also understand the phonetic mechanisms of tonogenesis thanks to the work of folks like Odricor and Lee and others in the mid 20th century. And there's also the generally accepted consensus now that old Chinese was in fact not tonal and we understand by combination of the tonogenesis phonetics, exactly how that came about to. Okay. So it took us a little while to get here, right there must have been multiple origin points for tone. And yet, even though we have sort of carved off everything else into okay other things are going on we know about tone in the Americas tone in, in Africa. But the question of Tony Asia remain and so Lee throughout his professional life continued to argue for a single tonal Asian super stock although this is now generally also been rejected so essentially that sign of Tibetan included the Thai languages included the homian languages. So now looking a little bit about phonetics of of tone in East and Southeast Asia. We have typically isolated so I mentioned these these features before us I'll just skip past them. But I'll just mention this term, the Sinosphere which is sometimes also used introduced by Madison to refer to beyond the geographical borders this sort of typologically convergent area, specifically that has the synodic type, typological features. Okay. We also know a lot more about tone in this region, we know that this tonal versus a tonal dichotomy that's that's really not. It's more complex than that so Brinnell and Kirby did a nice aerial survey and showed that it's, there's more going on than just tonal versus a tonal we have a lot of phonetics involved. Other features including the phonation that I already mentioned, things like vowel length vowel quality. Other things that contribute to how to what essentially what what functional load tone bears in a particular language how important it is to sustaining phonemic contrast and pitch only turns out to be part of the picture. So let's take a quick peek and look at how it arises this is the this is the account that was worked out starting in the in the mid 20th century. The basic process is, we start from intrinsic effects on pitch, we're voicing, for one example, this is voicing, where voicing will cause a dip at the start of a syllable in the pitch, and voicelessness can cause a bump up in the in the pitch and start of a syllable. And so then when you return to your baseline you now have sort of a bump up up up down and so over time. And then you become phonemic. This is a famous study from a bear at all in 1979. The point is you start with something that phonemic that is intrinsic and phonetic, but you have these phonetic cues is actually a bundle of features. And what changes is the weighting of those cues over time. So as pitch becomes a redundant marker, along with something like voicing. Voicing can disappear or whatever the feature is, and it leaves behind only pitch so subsequent speakers reanalyze pitch as being the primary queue and this is what we call tonogenesis. So, originally, so let me just demonstrate this was three stages. You start with a nice voicing contrast in the first stage by and pa. In the second stage you get redundant marking so by and pa but also have a low pitch on the voice and a high pitch on the voices so by pa. So there's two ways you can use to distinguish that contrast. And then over time, the voicing goes away leaving only the pitch so pa pa. And now you have only pitch that's contrasting meaning. So that's a simplified explanation there. So this is an example of other ways that we have, we can point to tonogenesis arising so for instance in the in the moon, me and languages rattles work showing that historically CV syllables in the early stage in the atonal stage. If it was plain CV, it ended up with a level tone in modern language, if it was CV with a glottal final. It became a rising tone CV with a H final became a falling tone and if it had a stop final. And so it kind of is its own thing, although in some languages those eventually developed their own tones as well. Okay. And so from there we get subsequent splits that conditioned additional tones and mergers are splits and mergers. Okay. What this, what we know is that there is this connection between historical segmental contrast whether it's the, the voicing on onsets, or it's the particular consonant at the end of a syllable. These things have led to tone and so I've tried to bring these together under one title and so I introduced this term de segmental phenology, which is to say, it's a subset of super segmental phenology because it's diachronically meaningful. So we're talking about super segmental phonemic contrasts things like pitch. Things like tone, but not just tone. They became trans phenologized. So this is a term meaning they moved the phonological contrast move from the consonants onto the super segments. And this is different from things like intonation other in stress other areas which we might class under super segmental phenology generally, but because there is this unified historical explanation. For a whole set of things going on in, in East and Southeast Asia, we want to have a term to refer specifically to these so this is an inherently diachronic term. It just means from the segment right so final phonology that comes from segments but is no longer segmental. And so we can really call this two sides of the same diachronic coin. And in this way we can actually unify the tonogenesis. But like I've just talked about and also something called register genesis. So let's talk about, I'll talk about more about what register genesis means in a sec, but essentially just means the, the origin of phonemic phonation like breathy voice and prekey voice. And so this is where Ryan's work is sort of picking up the ball and running with it. I'm really pleased to see how nicely this stuff is, is working out so he's working on a typology of desegmentalization and using terms familiar with us to sort of classify languages and the local origins and register origins and prominently Austroasiatic, but also in the area generally to identify things like progressive base segmental change regressive a nuclear depending on whether the tone or what the, the tone or register came from the onset the coda or somewhere in the syllable nucleus. So just to get a little overview of what kinds of segmental cues have led to tone over the years, we have things like final consonants that I just mentioned for the moment languages. That's also perhaps most famously, where tone arose in middle Chinese. Also Vietnamese. We have cues on the vowels so the Thai languages actually make use of a vowel length contrast to distinguish certain tones, and then onset so this is the voicing example I gave so this is this is the focus. And in fact this is the locus of change in East Asian voicing shift is these onsets. Okay, so let's now talk about that. So the East Asian voicing shift is this term that I have introduced to to encompass things that have already been described in different families. It's this massively cross linguistic loss of voicing contrasts. And not all languages but all all language families in the region. And what happened was the stops and fricatives so the obstruents devoiced, and they merged with their voiceless counterparts. Now we see some variations on this in some languages actually we saw only the fricatives devoiced and the stop state boy there's there's some interesting variation there but by and large this is the pattern, the devoicing of obstruents, merging with voiceless counterparts. And then we also see the sonarance that there some languages that had a voiced voiceless sonarant contrast actually saw a merger in the other direction where the devoiced sonarance merged with their voiceless or their voiced counterparts. So you have a loss of voicing contrast but in a different direction. And then finally we can say that this is something that swept across the the Sino sphere to borrow Madison's term again. During the second millennium C so it's actually more recent than we might have thought. And to give you an example. So in this early stage we have car and God with a nice. Nice voicing contrast right, and we have three proto tones, which we don't know exactly what they sounded like so we just call them a B and C. So we get stage where voicing has been lost car and car right. And so we now have a split into a tone, but hi be tone behind. So we don't know is like I say we don't know exactly how the, the phonetics of the pitches played out at this stage. And then we can observe on the surface tones of modern languages that we get often six categories. And so it's hypothesized that this is a stage that most of these would have passed through the six tone stage. And this is referring only to, to the open and sonarant final syllables the, the stop final syllables were their own thing, because some of them stay day tonal, as I already mentioned. The language was already tonal. The East Asian voicing shift had this doubling effect, we went from three tones, a B and C to six tones we now, you know, have to label them with these a one a two B one B two C one. That kind of thing. If the language was not yet tonal if it was. So this is like most of the australianic languages. This actually led to the, the origins of contrastive phonation so this is registered Genesis. So it's really a single change, the loss of voicing across the region that had differential outcomes depending on what the input conditions were so the input was tone is already phonemic, you got a doubling of tones. If the input was no phonemic tone, you got the origin of contrastive phonation and typically the equivalent would have been modal voice breathy voice creaky voice, instead of categories like tone one tone to tone three. And then in some languages that continued to evolve, becoming vowel quality differences so breathy voice creaky voice has gone away in most dialects of Khmer and has conditioned these vowel splits. So here's my bold claim I'll take a drink of water before I make it. Feel free to prove me wrong but I think that you know this. We really underestimate the breath of this change right so the East Asian Poetian shift is perhaps the most sweeping example of sound change that we have yet described in linguistics. Certainly it puts it, you know, when we talk about the great vowel shift in English, it puts the greatness of the vowel shift into sharp perspective when we're talking about one language they suddenly great. We're talking about you know the, the, essentially the linguistic ancestors of billions of people. And, yeah, anyway, it's a it's a large, a large swath of the world and a really impressive sound change event so we should appreciate this the scope of it. Now, onto the timing. There's more going on than just aerial convergence. I've already mentioned it's more recent than we thought. The question is how do we know it's recent how do we actually date when this happened. And so there are really several converging lines of evidence. There's the phonetic evidence, which is sort of what we know about how tonogenesis and registered Genesis happen I've already done over that and some detail. There's historical evidence where we can reconstruct what the segments were that's another part of how tonogenesis was it in fact discovered was by this combination of phonetic and comparative evidence. We also have another interesting line of evidence in the form of textual evidence. And so we can have some anchor points for dating from the epigraphy that we see in languages like Thai and Khmer. Stone inscriptions. And then also there's Chinese records and translation manuals. And so we actually can look at how words. So when they're recording words in other languages, they're telling us something about the historical phonology at the time of borrowing. And so there's a number of really interesting studies that have started to pick apart evidence from historical documents. So very very quickly I just want to give you an example of how these things played out so I'm going to skip past some of these slides which I will make available for folks to go through in more detail. But what I want to show you is these proto tones remember I mentioned a b and c. We don't know exactly what they sounded like but we can reconstruct some minimal triplets. So we saw, we could see actually more went on than just the, the East Asian voicing shift so originally we would have had a split from three into six. But we have subsequent further splits so it's actually more complicated than this. And I'll avoid getting into that now what I just want to say. It didn't stop at the East Asian voicing shift there were subsequent onset changes that conditioned additional things now it's not that there were ever 12 tones. Each language made use of different conditioning environments. And so some languages might have split some off based on these some languages might have split off based on these. And so different things like whether it was glottalized or implosive, whether it was an aspirated whether it was aspirated versus just voice so this would have been the. This is the East Asian voicing shift. All three of these rows versus the bot just the bottom row. And a lot of work that has shown that there are parallel processes for this across really across the region. And so to give an example of how tie looks tie has five modern tones. The each color is a tone so this is one tone, one surface tone, we have a surface tone that cuts across these sort of multiple historical categories. And you can really see this is why it's complicated the modern tones don't actually line up super transparently with historical categories. If you're thinking about the origins of the of the system, you can pick it apart you can tease it, tease apart the differences. And so here are four examples from Bangkok tie from tie of Chiang Mai in Northern Thailand. From an example from China and Vietnam. We can see that each language really carved up the tonal space differently different conditions splits and mergers affected what the modern tones are. And the numbers don't matter all this and the numbers are arbitrary the point is just to say this is a five tone system. This is a six tone six tone and five tones so really the numbers varied. But if you look even here right you have a split with the top two in the bottom two boxes versus one and three versus three and one so quite different outcomes. But I want to also point you to another important piece of evidence part of the reason why we know there were three tones is because the Thai language actually had the earliest use of phonemic tone marking that we are aware of. So there's we know of, we know of about tone in in Chinese going back many many more centuries earlier than this. But as far as we know tie was the first language to actually mark tone phonemically in the script as used by. It wasn't used by the whole population but would have been used by the sort of literate class at the time in the in the mid second millennium. So if you look here, you can see this kind of plus cross looking thing. That's the easiest one to tell so I've circled some of them we can see that on above some of those characters. There was a no marking versus a single vertical line for one for the first tone and then a sort of cross for the second tone. So unmarked versus two different tone marks that indicates three tones. The, this is, you know, a strong evidence that there was, there were three tones at the time of the invention of the of the Thai writing system around the 13th century, although this is a mid 14th century document. So this is yet another piece of evidence that this, this would have predated the East Asian voicing shift. So I'm going to wrap up by talking about a little bit more evidence from epigraphy, because I don't think it gets nearly enough attention. So epigraphy in Southeast Asia crosses all of the families also. So here are some examples of, let's see, I have okay good I think I hope I have a slide for each of the families but here is one for the crowd I languages. So this is the mangrove inscription of Sukhothai from the 14th century. We can so we have basically these are different scripts these are the names of scripts we have the Thai noise script the fuck com script and Thai or Sukhothai script. We also have epigraphy from Tibetan languages so Pew and Burmese starting the 10th century and 11th century and so here is the myosody inscription, which is sort of a Rosetta stone of Burma. It has different languages on different faces of the same text. And so you have the Pew face and the Burmese face which were both Tibetan Burmian. We have Austronesian languages. So starting in the fifth century with John, but we also have these other scripts. Again, these are the names of scripts not languages, but they recorded some of them also coincide with languages. So they were used in areas like the Philippines and Indonesia, prior to so this would have predated the Islamistization of Malaysian Indonesia of maritime Southeast Asia, and predated the colonialization and Christianization of the Philippines. The languages are some of the oldest records also in the region so modern Khmer starting from the sixth century. Okay, so here's one from from geographic Thailand but would have been the anchor, or the, I guess this would have been pre anchor even Empire in Cambodia in Thailand. So the reason I bring this up is because we have evidence from modern spelling which provides us with interesting puzzles that actually end up shedding light on tone, believe it or not. So we have this question of well, why do we get etymological doublets in Thai spelling. We have Sanskrit loan words so Sanskrit is an Indo European language of course, and we have words like uttara meaning north. We have words for father words for things like music and foot. They're borrowed in Thai but they have these voiced consonants. Why is this a puzzle. Not only do we have and we have, we have doublets so we have two forms with quite different phonology so we have udon with a voice and this vowel change, and then uttara which much more closely matches this right. So we have beat to which is a particular inflected form that got fossilized and done three bars don't three and Pada as bad. Okay, but if Thai always had these voiceless P&T, why do we actually see these forms why would we not just see them borrowed with the sounds that they the accurate sounds that they had. And perhaps I'm surprising this is evidence for multiple pathways of transmission of these loans into the language. And so what I've argued in the past is that the Sanskrit loan words that start with these voiced consonants actually came by a Khmer. And they actually provide a snapshot of Khmer phonology at the time and so there's actually Thai Thai and Khmer has its own mini Rosetta stone, the mangrove inscription of the 14th century has Thai language Thai script Khmer script and Bali language in Khmer script. So the same text in three different languages. So we have, you know, very good evidence of this provides like unarguable like a single point in place in time but we also have the entire body of evidence from the inscriptions. Now people have worked out quite detailed sound changes for Khmer. And if you're interested in this I'll just provide a link, a plug here for the, this is based on Philip Jenner's the late Phil Jenner's life's work on old Khmer. So ceiling.net says okay this is a project I worked on prior to my moving back to the US and starting grad school. But there's a lot, a lot there really interesting stuff to pick through. But the point I want to make is just that the epigraphic evidence here is abundant but really still sorely underutilized by my linguists. So to take a peek here. The, what I want to point out to you is that this is the era where the, the prod I people specifically the ties moved into Southeast Asia was during the Angkorian period. And because of the, the size and breadth of the of the Angkor Empire at this time, early Thai polities like Suwa Thai were essentially vassals or something similar to vassal states of the Khmer Empire there's actually argument for like that. Strong or widespread multilingualism in Khmer, lots of Khmer loan words in sort of acrolecto vocabulary and Thai so plenty of evidence that this was the case. And so this was a period of intense contact. Now after the fall of the of the Khmer Empire, and entering into the middle Khmer period there was a sort of swap of aerial prominence. And so that that's a different dynamic which affects things but this middle period is what we're talking about so what's the big deal, why do we care. Well, in this specific case, when Thai borrowed the words Khmer must have already gone and Korean sound changes. And the reason why we know this is because PNT became implosive, but and the. This was not reflected in the spelling so we wouldn't necessarily be able to date it very precisely. But the fact that Thai borrowed them with certain spellings that we know from our work on tone change came from historically glottalized or implosive sounds. And these would have been actually very close phonetic matches so when Thai was borrowing Sanskrit words from the Khmer's in the area. They were borrowing with a close phonetic match. They weren't borrowing it with the etymological match right so they would have just borrowed it with the peas and the tease but they were borrowing with the button and which is how it was pronounced. later, voicing ended up being renovated so the modern words, the doublets have the modern voiced ones, but originally at the time of contact they would have been implosive and voiceless. And this is just one point of evidence that tells us or the early Thai contact with Khmer must have predated the East Asian voicing shift in in both those languages so we have the evidence from phonemic tone marking, the evidence from comparative spelling differences across loanwords and so this is a really nice example of how we can use epigraphic text you know that tie things to specific dates and locales that really provide a nice extra evidence for our reconstructions. And so this also opens the door to even more evidence from things like archaeology once we know things about the particular phenology at a time and place. That's another interesting angle that that needs more work. Okay, so the textual evidence augments are comparative and reconstructive evidence. Okay, so that's where I want to wrap up, since I've been talking at you for a long time. And so I'll just close by saying look we've come a long way in a century. We've gone from well maybe they're all one language family, just a century ago or a little more than a century ago. Tone is all one thing to becoming okay well it's aerial something aerial is going on but we're working out a more nuanced view. We understand a lot more about the underlying phonetics, we can look at this sort of the specific sound changes how they propagated across the whole region and we can make maximal use of multiple lines of evidence to do so. And yet, and yet despite all this we're still in the early days so there are many open questions. These are this is an example of the ones that I look at but this is just a teeny tiny slice of all the questions that we need to ask. So the ones that I've been thinking about recently or just in general how did these days and voice and shift play out in the Thai languages. We know that it must have post dated the Thai diaspora. So it could not have been a single sound change that happened in their common ancestor, it had to have happened multiple times in the same language family, which is an interesting twist. So then we have to ask things like well how many different independent devoicing events were there. And so that that is a middle stage where we need to do reconstruction back to these not not prototype a more recent common ancestors, and try to pause it. These sub stocks that each underwent their own in instance of the East Asian voicing shift. And then from there we need to ask well how much is that generalizes to the family to the region and to the world and so that's kind of where we're at. If you have ideas on things to do with this I'm super happy to talk about it. Here's a public data set that I've made available 300 by language tone tunnel data on 300 by languages. It includes their historical tone categories and for all the ones that we know the also the modern phonetic values in this particular notation. And, yeah, come talk to me if you have ideas, but thanks very much. Here's some references, and I look forward to any questions you might have. That's great thanks for that's super informative and interesting to learn a lot in that presentation. Let me open it up for questions from anybody or comments you can either use the raise hand function, or you can note in the chat that you want to ask a question you can write out your question in the chat or you can just unmute yourself and start talking for a minute to see if anyone has a question. Well while we're waiting for people to format the questions or maybe I can ask you a bit about the thing you spoke about at the end of like how common voices change those kind of I was wondering, I mean so it seems clear that this is not just say five random innovations in these families that happened to be located in the same area and retained. It's clear that it's happening in this area at a much higher frequency, you know that anywhere else in the world basically except maybe a few pockets in Central Africans, other places but. So, is it basically you have to come up with some kind of contact explanation, plus maybe the right kind of phonological context for this to happen in. Another question so, because of the difference of the typological differences in tone right so you have the sort of the level tone or the only two tone system whether the simple versus complex Apple thing. We don't quite know with as much detail what happened in or it's either much simpler or much more complex. So in the with the two tone languages of Africa often it's very simply just okay it's probably something like voicing condition to two tone system. End of story. And then other other places where it interacts more with. With morphology and things it gets more complicated. So, there are questions about, you know, what is the origin where their segmental origins of morphological tone and if so can we reconstruct what those morphine would have been. There it's also what you mentioned is true that it requires a certain phonological context, and contact seems to be one of the ways in which that happens. This is something that I think Madison also called the the tonal milieu. So you need monosyllables in eastern South Asia you needed monosyllables, you needed this particular voicing thing and so you. The fact that they all line up in time means that there was something going on a really in the phonology already that set the stage for everything else for all these language families to undergo parallel changes. But the fact that we can date it to after the, you know, the dispersal of dye languages means that we're still stuck figuring out well if it's not five how many changes was it. Let's go to a question from Stravya. Hello, thank you for this very interesting talk. I was actually. When you mentioned the epigraphic evidence about three tones. I was thinking about the fact that the interpretation of the tone also depends on the class of the continent, the hike low and mid class continent. So just interested in knowing your thoughts about the interaction of that and tone marks and your research. Yeah, that's a great point. Okay, so let me go back to this table to answer that. Let's see somewhere. This one. So the way that the way that tones are analyzed in in modern time consonants are are given classes and the reason for that is exactly because of the outcome of of the, the, the East Asian voicing shift. The reason is, these consonants in modern tie or high class, these two middle rows are mid class, and these rows are low class and the reason is that the middle two rows pattern this together totally if you compare these two you see that they are not different in some languages they are. The writing system was invented there. It was all different continents so everything that is the reason why they had to come up with the notion of classes is because you have a car in row one, and you have a car in row four. And so they have to be different classes of continents that carry a particular tone with them. This would have been God, so I don't think I have. Did I, I don't think I included a slide that shows what the proto sounds were, but this would have been God, and this or sorry this would have been caught and this would have been God. And so the notion of classes is a more recent sort of pedagogical tool for teaching how to combine what we know about modern tones with what we know about historical tone categories. So classes are actually just stand ins for proto onset categories that no longer exist. That's that have merged. But at the time, like this, every, the, I'm going the wrong way. When I show you something like this. At the time of invention, every continent would have been a unique onset. And so that's the key observation. Thank you. Yeah, that's interesting English isn't the only orthography with all of these retentions from former pronunciations. We're glad to have them. Can I ask you about this device thing again because how, how cyclical is this because it seems like you go, you couldn't easily go from voicing to tone but then you get voicing back in because I think a lot of the African languages I have told their stuff voice consonants and they're complex segments there. So, and that relates to also the question of like the languages that went from three to six. What was the original three was that a voicing to tone cycle that then repeated as talking back and went out again. Yeah, this is a this is the question of much debate. What were the original three. What, what did a B and C mean so there is this really striking commonality where we have a B and C plus stop final like you know sometimes called D so a B and C and D right, a three plus one system. We have that in cynic languages, karate languages moment languages and Vietnamese so it's across the families and Chomick languages. So there's, there's something going on there. But what it might have manifested differently in different languages. So, it might have been contrastive phonation it might have simply been as simple as this column a or category a was originally modal voice column B was breathy voice column C was creaky voice. It could be that. And in fact, Graham Thurgood has argued that there's obligatorily a stage in tonogenesis of passing through this contrastive phonation. That's, you know, not accepted by everyone but that's one of the things that's been suggested. Others have said, maybe it really was just an early proto tone system it actually was pitch and so we don't quite know if it was truly pitch based or what but as for the cyclicality. I don't know if we've, if it's gone truly cyclical anywhere. We've certainly seen to voicing contrast re innovated in a bunch of languages. So that's something that happens. There is evidence that the East Asian voice so it's when I say the East Asian version it's simultaneously a historical thing that happened in a bunch of languages over time, but has is also a type of sound change that has continued to happen. So we've seen osteoasiatic languages, which are not historically tonal undergo something like this in the in the last, you know, century or two probably sense extensive contact with tonal languages like law and tie. I don't think we've observed a full like loss and regeneration of tone, but there's something going on with with voicing. It'll be interesting to observe as we move forward. Maybe there'll be some you know missing link language that will discover. Let me just quickly go to two of the messages that came in in the chat for you. When asking if there's any evidence of recent changes in the tone patterns in this language. Do you believe that political upheavals of the 60 70s, people migrating from communist China change the way people pronounce things, or is there any other recent evidence of tone changes that we could track a little better. Yeah, so this is part of why tone has been so tricky is because when I talk about a B and C, you know, these are historical categories right, but the actual phonetics of how each of these, you know, to go back to the one with the colorful pictures. So, you know, each of these colors is a is a single surface tone right, but the phonetics of it the categories are stable but the phonetics are really not. And so when we compare these things, you might see, actually I think I have it in another slide yeah this one. We have mid level low level falling high level and rising and high, but we have level high rising low rising level and falling with global so they're really quite different despite the fact that they have this very recent innovation so it's absolutely the case that all kinds of things are going on in recent decades with the surface level tones. But what's been remarkable about this is that things still pattern together historically. So the difficult the phonetics are the hard part I really admire the folks who are working on phonetic change the phonological change the category level change that's a bit easier so I don't know the specifics about what would have been going on with with with China. But it really is true that in areas where all the segmental phenology is the same right we have the same consonant and vowel inventories are very similar ones. And that often distinguishes dialects in these areas is the tones. So it's really that you'll have just strikingly similar. If you're not looking at the tone it looks like they have the same phenology but it's really just the tone that is the thing that distinguishes them so I wouldn't be surprised both that it's used as a marker of identity to say well we're different from these areas, but also something that could quite easily be influenced by, you know, political political movements and just the spread of different groups into different areas. Yeah. We also have a comment from Marie tell our residents, let the expert on the, an example from the guest the far northwest area of this convergence, the language a home, a home, show up in us that is today extinct and so Tony construction hasn't been done so, but there's neighboring languages that may have been part of this convergence area. So she says the Indo area languages of the Brahmaputra Valley like Saletti and Geregonian are also total languages. And apparently the evidence for Saletti sonogenesis is a loss of aspiration or aspirated consonant so I guess that's another possible contrast or segmental contrast that could be lost. Yeah, yeah, so that was one of the later ones, it didn't, it didn't lead to the loss, but it led to a split. So like specifically the weather something was aspirated or not. The ones that had historically been aspirated stayed with this tone but the ones that only became aspirated after the voicing collapse kept a different tone so we see sort of bits and pieces of this home is part of the same family. And so I actually spent some time looking at our home and working with folks like Steven Mori and Poppy go boy who've, who've done some work on a lot of work on our home. And I don't quite know exactly how it can, how it overlaps with the other Indo area and languages of the area, but I had it's yeah there's a lot of a lot to work out. We can reconstruct the home categories but we don't know anything about the home tone phonetics. And so you mentioned this tone of Genesis but what about when this tone ever disappears is pretty much a one directional change or what happens when we see loss of tone. I'm aware. I, in my head I am aware I, my understanding is that there are tones languages that have lost tone but because I don't focus on this I'm not sure of. Otherwise we all end up with a tonal language at some point there was purely one directional. Yeah, no it's definitely not but it might be more closely tied to language shift so for instance the home people this is another example of this where historically spoke a tonal language, but rather than our home becoming a tonal what happened was it was more of a language shift to Ossamese and other languages that retained a lot of our home vocabulary as long words but they weren't they didn't have tones with them at that point. So it's not so much that all of them lost tone but it's that the language, the whole community shifted languages. And so that's, but I'm sure there are specific cases of tone actually disappearing over time as well. Well, I know you have more meetings to get to today so maybe we can wrap up there's one say thanks again for this presentation it's interesting work for sharing with us and all those are going to end up watching this video on YouTube as well so thanks for putting this together for us today really interesting. Yes, thank you it's been a pleasure. Okay, thank you everybody.