 which is the topic, Nathan started today's event by talking about the problem of time with transcription. And I have a system that I use, which is probably unfamiliar to almost all of you, all of me, so I should probably explain how it works here. It incorporates some aspects that are influenced by the system of Professor Al Fowler here. One of those aspects is that I mark target tones using numerals in the beginning of the syllable. So the first number in the target forms in this presentation are tones. And there are two possible tones, one and two, which are conventionally called level and rising, even which are the tone that names for the tones, but whether those are the actual forms of the tones is uncertain as these names are, had become highly conventionalized in the Chinese tradition. The middle of the syllable in my system is fairly simple explanatory. It's the consonants and vowels of the syllable. The one exception is that I use the lecture Y to represent some sort of vowel like a or a. It's a deliberately non-committal symbol. My transcription is not intended as phonetic notation. It's somewhat abstract. So I use numbers for the tones because I am non-committal about how tones sounded. And I want to use Y because all I know about Y for sure is that it's some sort of central vowel. It's not E, it's not A, it's not A. It's not O, it's not O. But exactly what it was, I don't know. The last number in my system at the very end of the syllable is the grade of the syllable. Shungong just talked at great length about grade two. And in my system I have four grades rather than three. But these represent what I used to think Y with the different grades where I don't actually believe the same more. But these are some ideas I used to have for a number of years. You can think of the grades as flavors of the rhymes. So a grade one, two, three, or four Y is they have some kind of common quality, obviously, but they sound slightly different. And exactly how they sound different is something I'm non-committal about. I mean, here I threw in some own suggestions of mine that I don't even believe anymore. So I prefer to just use the abstract numerals instead to speak of a grade one Y or a grade two Y rather than commit to saying that this rhyme is uvularized or whatever. Although I am now strongly tempted to adopt Shungong's system, I prefer this presentation I will be not. Now, when I first started learning Tongue over 20 years ago, one of the first things I did was learn the numerals for one through 10. And I was kind of amused when Professor Alka began his talk on Tongue last week with these very numerals. And so these are the very first 10 characters I learned how to write and the first 10 words I learned. And these are their readings in my system. And one thing I noticed immediately was that some of the numerals are very obviously resemble the written Tibetan numerals, which in part sought mark the similar pairs in green here. One of the, one pair that has bothered me for over 20 years, and which is the topic, subject of today's talk, is the number five. Here you can see in two and three and nine, there's a pretty close resemblance, sort of obvious resemblance. With five, there's something strange going on. There's a W in the Tongue that corresponds to nothing in Tibetan. And these other cases, they have similar vowels like e and e, kind of close, o and o, kind of close. e and o, kind of close, but e and o, kind of different. And then there's this strange W. So this has bothered me for over 20 years. Now, I showed that there were only a few pairs that were really obvious commonance, but if you trace back the history of the Tibetan numerals, which I did with Nathan's help, and tracing back the earlier forms of the Tancret numerals using the system that I published five years ago, you can see that most of these, except for seven and 10, which are completely hopeless and unrelated, most of these are actually commonance. And the resemblances are pretty good in most cases, except five is still not your problem. Even if you go back in time, you still end up with very different vowels, an e and an o. And how can this be reconciled? Assuming these are really commonance, and I'm pretty sure they are, why does a tongue with five end an o? And why does it have a medio w? Using my system, if I take this back in time, I end up with an e, with an e. But throughout some Tibetan, the word for five in the oldest languages, universally has a, old Chinese, Tibetan, Burmese, and the language I've been working on for the last couple of years, pu, which is an extinct, some Tibetan language that was once spoken in Burma. So we have all these votes for a, and Convict has an e here. What's going on? Why is Convict the odd man odd? Let's suppose, for the sake of argument, that all these other languages are conservative in their numerals for five, and that Convict at one point used to have, ah, let's just pretend this is the case. And interestingly, this reconstruction, which is what I proposed about five years ago, looks like pu, which I've never saw until two years ago. So that makes me kind of feel good. So let's just adopt this hypothesis for the time being that, that Tangud used to have ah, but somehow it changed to this. But why? And how? How can we bridge the gap between this w, or whatever it was, and ah? The short answer is you can't do it. It's not possible. Tangud w normally should come from pre-Tangud, u, or ut, but not ah, as Yom Jop demonstrated in his book. So what is going on here? I mean, should I just give up on five and just say that, oh, that, it kind of looks like the other words. It has a nuh like the other words, but the vows just simply make no sense at all. So it's just complete coincidence that Tangud has this nuh word, which is totally unrelated to all these other words. Should we do that? Let's see if we have to. Yom Jop's solution to the five problem was to reconstruct five in Tangud as had an u, which is okay according to his rules, in which w comes from u or ut. And Jop Wug, a library that Shun Gong has referred to, which is a relative of Tangud in, has an u in its word for five. So this is not, this is kind of plausible, but we still have another problem. Okay, so Jop Wug and Tangud agree, but they still don't agree with all these other languages. What's going on? Now, I will present a very different solution to the problem of five. Goh Gwang Chung pointed out that Tangud has a number of pairs of words in which one word has no medial w, so this ties into our first talk today. And the second word, which has a similar meaning, or like for example, in this case, red and to turn red to blush or something, has a w in it. So Goh Gwang found a number of these pairs. In 2012, I published a paper in which I theorized that these medial w's were from earlier p-prefixes. So this turned red is from the root for red plus a prefix p. Pne becomes nueh. The p weakens to a w and kind of switches places. So maybe the medial w in five, which is weird, is also from some kind of p-prefix. So just as pne becomes nueh, pne becomes w. Well, the vowel is still weird, but this could explain it better. And as I've seen before, pue does have a p syllable before the náh. So this p corresponds to the p-prefix. That I propose here for Tanghueh. And I propose this thing for Tanghueh years before I ever looked at pue at all. So this is not some kind of circular thing. So I think that's how the w originated, but we're still stuck at the problem of y. Sorry, I was circling the wrong word in that. Of y, five has e for the vowel, it shouldn't. This part, once again, e cannot be derived from a. Gho yag's book on Tanghueh demonstrates that e can have a whole bunch of different sources, eep, eep, eep, oep and ot, but not a. So we still have a problem here. This is impossible. How can my claim that 5B related to all those other I-words work if 5 has a, it's not possible? You may have noticed that in the program there is an extremely generic title for this talk. That's because I've been working on Pew for the last couple of years and when it came time to write up something for SOAS, Nathan asked me what I wanted to do and I said I don't know so he just made up some generic filler and as the talk approached he and I were on a bus in Burma working on our Pew project and I still had no idea what to do and for some reason I was babbling about Tongut because Pew is my day job currently and Tongut is what I do as a hobby and I was saying something about Tongut and Nathan can correct me if I'm remembering wrong but at some point the numerals came up for some reason maybe because of that the Pew numeral for 5 which has been very much in my mind and I pointed out the vowels in the Pew numerals and Nathan suggested that maybe the vowel for 5 was influenced by the vowel for 4 and that is his contribution to this because everything else after this is all my fault so don't blame him for anything after this. But I thought yeah this is that there might be something to it because look they both have us but on the other hand there are some problems they have different grades for one thing and 4 also has a rough reflex so this is something like ur something and there is a marker here for or a mysterious vowel quality that I have yet to identify so yes they are both romanized and my system was wise but they're not really the same rhyme this is ur and this is ur or something like that so now I'm facing a different problem how do I reconcile these two rhymes and why would this even happen in the first why would you get the vowel like that from in the first place well let's just see if I can try to make this work so once again let's suppose for the sake of argument that the few word for 5 excuse me the time good word for 5 is like all the other verily sounds that words and has an i if we trace back the word for 4 the ur goes back to an e so in the earlier stage they have completely different volumes they're just nothing in common with these two at all now in time good there is a change shift in which e becomes ur so so the e of 4 became became ur and then the original a becomes the new e so punna became so row d became row d and punna became punny so in stage 2 the vowels have changed but they're still distinct from each other on the other hand I should note that a and e are kind of diff are very different but e and e are becoming phonetically more similar they're they're still distinct but they're not as distinct as as high e and low a in stage 3 5 and 4 are jason numerals and Nathan's suggestion was that the vowel 4 kind you know kind of ripped off on the vowel for 5 so at this stage they have now they have the same vowel and the e has been replaced by a ur which is a different vowel but kind of similar so now now these two words rock they are precedents for this sort of thing cases of what I call adjacent neural influence for instance English 4 and 5 have the same initial consonant but a program the European the two words have very different initial constants qua and puh theoretically what to war should have become English whore which is kind of disturbing but we are in fact have four the initial 5 has spread to four this is the case of initials though the time the case involves rhymes so do you have any cases of the ends of numerals becoming adjacent numerals becoming similar yes we do in probably the European nine and ten had different final consonants and to this day English nine still has the final end and in Latin on the other hand we have no them and as in November and December in which the ending of ten has spread to nine theoretically the Latin you will find I should be no then with the end from port of the European but it's not so let's go back to Tom with one warning which is that I'm going to mention a series of changes and the order of those changes is uncertain the order that I present here is arbitrary because I have to present them in some I have to present them in some sort of order I can't dump them all in front of you and say apply the most at the same time it's not going to but I really don't know the order of these things at all so this is my old height my only hypothesis about the grades which are presented in the hamburger which I am pretty skeptical about now is that the grades were partly conditioned by earlier loss of vowels I won't go into huge detail on this I'll just stay here that in this whole theory of mine which is in my 2012 paper low vowels conditioned the lower numbered grades and high vowels conditioned the higher numbered grades so this this little has no low vowels in whatsoever so it gets a high number grade this little has a low vowel so it develops a low grade so at this stage then the the rhymes start to diverge earlier these two words would have rhymed in my hypothesis now but once the grades start to develop they diverge I mean they're still they're still similar vowels which is why you still use the same symbol y for them but now they belong to different grades and have different flavors of some sort the the low vowel that conditioned grade one disappears completely so now you can no longer predict the grades at all they've become completely believing in the next stage I've already mentioned how peak how I think media w develops from prefixed p so the prefix p of five because love and the earlier prefix before conditions a rich flex file so this is so now the rhymes are very very different this is a grade one why and this is a grade four or then they diverge even further have proposed a mystery suffix in Tongut that I call x and this suffix accounts for work work families that go home to discover and that he thought involved short and long vowels so one of his word families was wind with in gong's reconstruction a short vowel and blow so verb based on wind with a long vowel in my system I don't have vowel length and I propose that gong's so-called long vowel derivatives have this mystery suffix x it's a mysterious suffix because I cannot figure out what's phonetic value was using the aforementioned Tibetan Chinese or Sanskrit materials it doesn't seem to correspond to anything all I can say is that syllables with this mysterious x quality tend not to be used to write Sanskrit so whatever it was must have sounded very on the Sanskrit in my Tongut notation I use x to indicate the mysterious excuse me the mysterious source of this distinction and I use an apostrophe to to indicate it in my Tongut transcription so this is the source of this is the outcome phonetically what these things were I'd absolutely no idea I call this apostrophe symbol prime even though it's not really a mathematical prime symbol but it's a lot easier to type and so there it is four so once this average was added to four I don't know what it was doing there really it made them the rhymes extremely different so now we have grade one y on one hand and grade four y with retroflection and the mystery quality I've added the tone numbers here because in my version of Tongut language history Tongut tones develop depending on the presence or absence of a final h and since neither of these words have a final h it developed tone one by default so in my proposed solution I include a number of changes that give you some idea of how my larger Tongut reconstruction system works so this is a review of what we've seen so far there's a change shift from the first and then out of e so it's as if when e goes to a this leaves a blank because there's no more e that's filled by the a the vowel four spread to five I don't really care for this grades for this version of the grades anymore but it's in the park like so we're kind of stuck with it for the next few minutes so in my old system low vowels conditioned low number grades and high vowels conditioned high number of grades with variation depending on the initial these great conditioning vowels in some cases disappeared a prefix p is a source of media w that is found in some word families and in five a prefixed r conditions vowel retroflection that I write as a suffix r but it's not a real r just quality of the vowel there's a process of adding some kind of mysterious affix that I write as x and that this x in turn in my transcription is written as a pseudo prime symbol and finally I propose that Tongut tones reflect the presence or absence of final laryngeals as in Chinese and that if we go back to the number 10 we can see one example of a final h conditioning tone 2 all these other ones have tone 1 because they don't have any final laryngeal they don't have any final laryngeals in my system so my hope here is that not only have my more or less solved the problem of Wi-Fi is a really strange rhyme I say more or less because now that I don't like this great system here this has to be somewhat revised but I think much of this presentation is still solvable but beyond that very very specific problem I thought it might be interesting to give a demonstration of some of the changes and rules in my view of the history of Tongut using five as in five and four as examples with that summary I have come now to the draw or the end