DOWN WITH PALATALIZATION!

A recent post at Anatoly’s blog (now called просто здесь красный, где у всех голубой, a quote from Aquarium‘s song “8200“) shows an absolutely hilarious sign held up by a protester: “Мы за пересмотр итогов палатализации” [We are for a reconsideration of the results of palatalization]. [Photo now available here.] Unfortunately, unless you’re familiar with the Slavic first, second, and third (no Wikipedia article, for shame!) palatalizations, you’re going to have a hard time seeing the humor, but for those who are, Anatoly’s thread is funny enough I think it’s worth posting.
Addendum. Don’t miss Bathrobe’s long comment below (May 15, 2012 09:54 AM) on Mongolian scripts and dialects and the history of the split between Buryat and Mongolian!
Update (2013). John Cowan writes to say: “Anatoly’s blog no longer shows the famous picture, but it is available at
http://img-fotki.yandex.ru/get/6304/24332511.8a/0_6f9ba_76f4ae5d_XXL.” I’ve added a note in the post accordingly.

Comments

  1. a well-known slogan “за пересмотр результатов приватизации” is also probably a prerequisite
    (“for reconsideration of the results of privatization”, referring to the Russian oligarchy’s infamous foundation mechanism of the 1990s)

  2. Third palatalization (which it seems actually took place earliest, or possibly between the first and second). Fourth palatalization (in Polish and the other Lechitic languages only): ke, ky, ge, gy > kie, ki, gie, gi, e.g. bogyni > bogini ‘goddess’.

  3. Of course undoing the English First Consonant Shift and especially the Great Vowel Shift sound like a good idea too.

  4. a well-known slogan “за пересмотр результатов приватизации” is also probably a prerequisite
    Quite right, I should have mentioned that.

  5. marie-lucie says

    Undoing the Great Vowel Shift would be a blessing for children learning to read and write and to foreigners of just about every stripe.

  6. Would the English First Consonant Shift be the consonant shift that’s angry about “press ‘1’ for English”?
    (But seriously — what is that? This page is the only Google-hit for that phrase.)

  7. Ran:
    It’s Grimm’s law, and it applies to all the Germanic languages, not just English. It’s why we have the native words foot, third, hound, what, ten, cold corresponding to Latin-derived pedal, triangle, canine, quodlibet, decimal, gelid.
    The Second Consonant Shift affected German only, and is why we have English path, eat, make and German Pfad, essen, machen.

  8. I think Poles will find this sign offensive.

  9. Having only to deal with two difficult first and second palatalizations in the history of Ancient Greek, I find the very concept of a third or fourth palatalization absolutely frightening.

  10. Marie-Lucie: let’s be fair here, restoring French final consonants and schwa would also ease the burden of native and non-native speakers alike when it comes to reading and writing.
    All: Palatalizations in Slavic certainly created a new layer of morphophonemic complexity, but at least all Slavic languages today represent the new phonemes in writing. Just imagine what Russian (for example) would be like today if, for instance, KTO and CHTO were both spelled with an initial {K] because of etymological considerations…

  11. It’s trying to figure out when French final consonants are pronounced that’s hard, at least for this anglophone.

  12. marie-lucie says

    Etienne, you are right, but I was picking up on John Cowan’s mention of the English GVS.
    In both languages the spelling has largely stood still for several centuries while the pronunciation changed.

  13. Though I suppose if you tried to update English spelling today, you’d end up with different ortographies (e.g. number of vowels) for RP, American, Australian, South African… Written English feels almost logographic to me.
    (Ok, not quite “almost”.)

  14. Bathrobe says

    Different orthographies for different varieties of English. Hmmm. Sounds like a good idea. This is what the Russians liked to do to their minority nationalities (it was a nice way of splitting Buryat from Mongolian, for instance). Applied to English, it would be a good start on the road to destroying Anglo-Saxon dominance of the world…

  15. Leonardo, Bathrobe: Actually, we are at a good moment in the evolution of English to update our spelling. There have been a lot of dialectal phonemic mergers, and a few universal ones like vain/vein, but very few dialectal phonemic splits. The only significant one is bad/lad, which takes different forms in different areas. Consequently, an orthography that continues to maintain most of the 14th-century distinctions of current spelling, but eliminates all the irregularities, will be about as easy to read as French. People who merge cot/caught, when/wen, or meat/meet, for example, will have to learn which spelling to use when, but they will be able to read every word at sight once the system is mastered.

  16. I’m afraid I’m sufficiently interested in palatalisation to want to understand this but dense enough not to get the joke. Why on earth is anyone demonstrating about palatalization anyway?? I know asking to explain a joke usually kills it but could you please oblige?

  17. Bathrobe says

    From my understanding, it would be a bit like a placard saying “We demand the repeal of the Great Vowel Shift”. As to why that would be funny in a Russian political context, well, I’m no wiser than you are.

  18. Cadoro,Bathrobe –
    See primaler’s comment at the begining of the thread. It’s basically an absurdist play on words. Russians tend to be fond of this sort of humor.

  19. Actually, the joke has nothing to do with palatalization, or anything else in the language. It’s politics. The original slogan was “We want the elections results to be reconsidered” (We had elections to the Parliament in autumn, and the results were unfair (Putin’s party Edinaya Rossiya won), in the opinion of lots of people. Now, after Putin’s inauguration, anti-Putin opposition held meetings and demonstrations in many Russian cities. The discussed post tells about one of them,in Krasnoyarsk. Not to be taken by the police for anti-Kremlin prpopaganda, the people pretended to be just “walking” (the word which has aquired a new meaning now!) with ridiculous slogans like the one about palatalization.

  20. Christopher Burd says

    I was surprised that (as far as I can tell) none of those palatalization laws accounts for the hard/soft consonant pairings that are such a striking feature of Russian phonology.

  21. And what’s with the pre-1918 orthography?

  22. Well, if you’re opposing the results of palatalization, it stands to reason you’re not going to like orthographic reform either.

  23. Bring bæc þe olde weieȝes!

  24. Trond Engen says

    Wouldn’t that be weȝes?
    (asked in sincere ignorance)

  25. Well, they didn’t use the ȝ in Old English, so I threw in a fine-looking Middle English form. I don’t think perfect consistency is a strong point of the parodic revivalist movement.

  26. Bathrobe says

    Anna, a belated thanks for that explanation. In the political context it all makes sense. (LH is so focussed on language, he sometimes forgets to fill us in on the political background 🙂 )

  27. So true!

  28. Hi Bathrobe,
    Could you give more information on the story behind splitting Buryat and Mongol?
    I ask because I came across a birthday wish in Buryat: “Турэhэн удэроор – тумэн жаргал хусэнэб!” and while I could recognize what it was, it seems quite far from “Tѳрсѳн ѳдрийн баярын мэнд хvргэe” (that’s correct Mongolian, isn’t it?), and not just orthographically. I can also recognize when I’m wished happy birthday in Dutch when it’s written down, although I couldn’t say it, and English and Dutch are different, not just orthographically.
    I also spoke with a Buryat I knew and he told me on some words that are spelled the same in Buryat and Mongol, Buryats pronounce the final single vowels that are often silent in Mongol. Though he said the same thing as you, that it was one language before the Russians introduced different orthographies, and the languages became more and more separate.
    There is a Buryat working as a Russian teacher at the school I’m working at in Darkhan and she mainly communicates with the other teachers in Russian or poor Mongolian. (Although come to think of it her Buryat may not be strong, since she’s from Ulan-Ude.)
    I don’t know what this all adds up to except I would like more information. What can you tell me?

  29. Bathrobe, you’re welcome)) This link will show you more slogans from that demonstration – have fun!
    http://feelek.livejournal.com/72354.html

  30. Bathrobe says

    @PF: So you’re in Mongolia. Nice! Actually, I think LH could give you much more information about the process by which the Russians decided on the standardisation of the languages of their empire, and I’m also sure it has been covered at LH before.
    At any rate, one of the keys is the script. The old Uighur alphabet is quite difficult because it fails to distinguish between certain sounds that need to be distinguished. For example, ᠦ can represent either ө or ү, which means that you have to guess at the pronunciation if you don’t actually know the word. The script is also very archaic, representing the sounds of Mongolian as it was spoken many centuries ago. While this is a major problem in learning to read Mongolian, it is very good in the way it is able to cover different modern dialects with their varying pronunciations. There are actually a lot of variant Mongolic dialects, and they can be almost mutually intelligible.
    To take a very simple example of the script uniting the language, take the word for ‘father’, which in the traditional script is written ‘abu’ (sorry, can’t reproduce the Mongolian script here). This is pronounced /a:β/ in modern standard (i.e. Khalkha) Mongolian. Since the Cyrillic script is quite phonetic for the Khalkha dialect, this is written аав. This is the same pronunciation as in Chakar dialect of Inner Mongolia. But by way of comparison, the Khorchin dialect pronunciation is /a:b/. If it were written in Cyrillic, it should probably be ааб. So using an alphabet that strictly fits one dialect (and not others) tends to splinter the language. And in fact, the Buryat spelling of this word is аба. As long as Mongols were all using the old script, which wrote ‘abu’, nobody was disadvantaged and everyone had a common written language. But the moment purely ‘phonetic’ spelling was introduced, the old unity was broken. The old script was able to accommodate variation in vowels (which is quite widespread in Mongolian dialects) and at times even the old endings you mention. The modern ‘phonetic’ scripts force just one pronunciation to be chosen as standard.
    The Russians forced the Mongolians, like many other peoples of the Soviet Union, to drop their traditional script and adopt Cyrillic. This happened in Mongolia in the 1940s (after a brief period where the Latin alphabet was used). I’m not sure when they forced the Buryats to adopt the Cyrillic alphabet, but it probably wasn’t that much earlier. What was sneaky about the Russians (or perhaps Stalin) was that they made the decision on what was to be encoded as a ‘standard language’. If they decided that Buryat was a separate language from Mongolian, then a separate script was created for ‘standard Buryat’. If they decided that something was just a dialect of another language, they didn’t develop a separate script. The grounds for making such decisions were, if I understand correctly, just as much related to destroying tradition and ‘divide-and-rule’ as anything cultural or linguistic. Since Buryat was already a distinct dialect from Khalkha and other varieties of Mongolian, and since Buryatia belonged to Russia and Mongolia was a (theoretically) separate country, it wasn’t difficult to set up a different orthography for Buryat from Mongolian. Add to that the massive Russian influence on Buryat vocabulary, and the two standards started to become very different.
    This encoding and enforcing of standards then tends to become entrenched. In Mongolia, the Khalkha variety has been encoded as Standard Mongolian, and all Mongolians are taught this standard, including the Buryats living in Mongolia. (The Buryats living in China are taught the Chahar dialect as standard). This leads to the gradual homogenisation of language within each territory. My understanding is that the original dialects within Mongolia are gradually disappearing and being replaced by standard Mongolian (Khalkha), destroying the variety that once existed — Khalkha Mongolian is a very narrow standard –and also the cross-border similarities among dialects in different countries.
    The example you give is fairly easy to understand, and would be even easier to understand if the script was unified. Турэhэн удэроор – тумэн жаргал хусэнэб. Турэhэн is obviously the same as Tѳрсѳн (с becomes h in Buryat), удэр is the same as өдөр with a difference in vowel and a different case ending (instrumental instead of genitive). You write тумэн, but I suspect that should be түмэн (ten thousand). Жаргал is the same as Mongolian жаргал (fortune, happiness). The last word should be хүсэнэб and is obviously related to the Mongolian word хүсэх (to wish). I’m not sure of the verb ending, but the whole obviously means something like ‘Wishing you ten-thousand happinesses on your birthday’. It’s not the standard Mongolian phrasing for ‘Happy Birthday’, but I think you could be overrating the significance of this difference in wording.
    At any rate, the adoption of different scripts and new ‘standards’ was, I’m pretty sure, deliberate policy designed to split the Mongols into smaller competing groups. Add to this the fractiousness of the Mongols themselves, and it’s easy to split them into competing groups. The Mongolians regularly deny ‘Mongolness’ to Mongols in China or (perhaps to a much lesser extent) in Russia. It helps to reinforce the way that the Chinese and the Russians have already broken them up.

  31. Wow. Thanks, PF, for asking the question and prompting that terrific response! (I’ve written briefly about the general issue of Soviet ethnic/linguistic splitting, for example here and here), but I know nothing about Buryat and Mongol and would have been unable to provide any useful information here.)

  32. Bathrobe says

    Ah, I’ve found some information on Buryat (or Buriad) in The Phonology of Mongolian (Oxford Linguistics) by Jan-Olof Svantesson et al.
    Characteristic phonological features of Buriad are the weaking of *s to h and fricativization of affricates [here I think it refers to things like Mongolian цагаан becoming сагаан in Buriad]. Another feature differentiating Buriad from Halh is the retention of word final short vowels, for example, ʊʊlə ‘mountain’ vs. Halh ʊʊɮ [Cyrillic: уула in Buriad, уул in Mongolian].
    …..
    There is not complete agreement about the dialect division of Buriad, but five main dialects may be recognized:
    Buriad dialects
    (a) Eastern Buriad (Hori, Aga, Muhar-Sheber, Tugnui, Hilok, Buriad of Mongolia)
    (b) Northwestern Buriad (Ehired, Bulagad, Kabansk, North Selenge, Ivalga, Bargazhan, Kachug, Boohon, Ol’hoon, Osa, Baigal-Hudari)
    (c) Southwestern Buriad (Alair, Tünhen, Aha, Ungi, Zakamna)
    (d) Nizhneudinsk
    (e) Bargu
    …..
    The Eastern Buriads traditionally used Classical Written Mongolian as their written language. A special Buriad variant of the Mongolian alphabet was devised in 1905 by Agvan Doržiev, who took the Oirad Clear Script as a model. The Cyrillic alphabet was used for writing Buriad from about 1840 in Western Buriatia, but after 1917 the Mongolian alphabet was introduced there as well. It was officially replaced by a Latin-based script in 1931. At first, Halh Mongolian was used as the dialect basis for this script, but this was almost immediately changed to the South Selenge (Tsongool-Sartuul) dialect, which is phonologically a dialect of Mongolian proper rather than Buriad. The dialect basis was changed to Hori in 1936, and it was decided in 1938 that the Cyrillic alphabet should replace the Latin one. Printing in the old Mongolian alphabet ceased in 1937, and printing in the Cyrillic alphabet started in 1939-40. In China and Mongolia, Buriad is regarded as a dialect of Mongolian, and the speakers use the respective forms of the Mongolian written language.

  33. Bathrobe says

    I have here a Russian-Buryat dictionary from 1954. It’s called Русско=Бурят-Монгольский словарь. Note the use of the compound word Бурят-Монгольский.
    If we look at the naming of Buryatia, Wikipedia says: “In 1923, the Buryat-Mongolian Autonomous Soviet Socialist Republic (Buryat: Буряадай Автономито Совет Социалис Республика; Russian: Бурятская Автономная Советская Социалистическая Республика) was created as a result of the merger of Buryat-Mongol and Mongol-Buryat Oblasts….In 1958, the name “Mongol” was removed from the name of the republic. The Buryat ASSR declared its sovereignty in 1990 and adopted the name Republic of Buryatia in 1992. However, it remained an autonomous republic within the Russian Federation.”
    Since my dictionary was published in 1954, it still retains “Монгольский” in the name. This name no longer appears to be current.

  34. Bathrobe says

    Finally, a quote from Soviet Language Policy in Central Asia by Mark Dickens, which LH links to. It gives the general characteristics of Soviet language policy in those early days:
    “One of the chief linguistic tasks of the new government was to develop a separate literary language for each significant ethnic group in the Soviet Union. “In the USSR, the emergence of a written language is not always the result of a long internal evolution; it is frequently the consequence of a decision by the central authorities who can present a community with a literary language worked out by Russian linguists” (Bennigsen and Quelquejay 1961:16). Each Central Asian Group chosen to constitute a nation was given a literary language which was artificially differentiated from those of neighbouring nations which were often linguistically similar (as, for instance, with the Kazakh and the Kirghiz). Thus, the linguistic unity of the area was broken up while differences between the languages were emphasized. This process of separation was helped further by the National Delimitation of 1924, which fixed the boundaries of the five Central Asian republics, primarily along ethnic and linguistic lines.
    “On the surface, the provision of a territory and a literary language for each Central Asian nationality seems like a generally beneficial development for the groups in question and certainly, the latter was necessary if the people were to become literate in their own language. At the same time, however, definite political motivations can be seen for this move. This divide-and-rule strategy obviously helped to diffuse any potential pan-Turkic tendencies. In addition,
    the early policy-makers harnessed nationalist sentiments and used them to promote unity within the Communist framework. Instead of suppressing the national languages and the expression of cultural heterogeneity, the Soviets encouraged them, thus deflecting any criticisms from nationalist factions. The symbol around which nationalists could rally, namely the language, was openly encouraged by the young Soviet regime and even championed by it, weakening the nationalist cause (Allardyce 1987:4).
    It must be remembered also that it was the central government which initiated and carried out the process, not the local population. the Soviets maintained total control of the situation throughout. ”

  35. Which leads us to korenizatsiya.

  36. As I have said before: Although the conversion to Cyrillic script was surely politically motivated, especially for the Turkic languages, which had perfectly good Latin alphabets from only ten years before, I don’t believe either the separation into novel Dachsprache or the details of the orthography were a matter of deliberate, er, incompatiblizing. It was more a matter of the job being done by separated groups in different places with no coordination between them, according to the best principles they could scrape up in an extreme hurry.
    Note that although Pontic Greek and Yiddish got spelling reforms in Soviet days, neither was cyrillicized. The same is true of Georgian, Armenian, and (after 1940) Latvian, Lithuanian, and Estonian, which retained their native scripts.

  37. “Турэhэн удэроор – тумэн жаргал хусэнэб!” ‘Wishing you ten-thousand happinesses on your birthday’.
    Yes, it says “on your birthday – wishing you thousand joys (happiness)”. I am so very glad that you two are interested in Buriad and Mongolian languages, my father’s family is from hori Buriads, they used to live around the lake Baikal (Baigali dalai) and came to Mongolia in 1916 and were then repressed in 1937, all the men, my grandmother would tell us sometimes stories about that time how all just the women and kids were left, they were considered as if they were counter-revolutionaries who avoided the October revolution, my father’s two uncles then were in the gulags around the White Sea and then served in the Red Army during the WWII, one of them who was a pilot died under Kharkiv in 1944, there is a documentary about them in the Museum of the repressed in UB.
    I think at least in singing our songs, the Inner Mongolian singers now sing in the perfect standard Mongolian, it’s so very moving too, as if Inner Mongolians consider speaking the standard Mongolian cool, isn’t it great and it’s enough only 20 years to open the borders and communications and the language differences start to diminish and the dialects come closer, I am not sure that’s that bad a thing, for our unity and statehood and for people’s perceiving own national and ethnic identity and pride
    otoh, in our countryside the Inner Mongolian Odon or some other TV are so prominent, people watching them daily, seemed as if it is not a very good thing, as if the Chinese state propaganda could reach and wash the people’s minds too easily, and it seemed just pretty boring, what is official is official sounding in any language
    i’ve noticed when i watch some Buriad documentaries they adopt the standard Mongolian words and terms in their speaking Buriad, b/c it seems they forgot their original words and knew only Russian substitutes for those words and had to learn it from our language anew, the speech then sounds pretty strange and as if like patchy, rough, not very polished sounding, “rejet ukho” Russians would say, when the pure Buriad would flow as if like pretty smoothly
    so I mean, if you have any other questions about Mongolian expressions or Buriad, PF, feel free to ask, i’ll try to answer, but Mr.B knows more about the linguistics problems, of course, I have no idea about all those palatalizations etc. would have thought, for example, at first that that was a pun on the Kremlin palaces – palaty in Russian

  38. The 1923 merger was probably done to prevent strife between the Buryat-Mongol People’s Liberation Front and the Mongol-Buryat People’s Liberation Front.

  39. sorry, yeah, tumen – ten thousands, the meaning of the exactly ten thousands is not used much, one would say more like arvan myanga now

  40. Bathrobe says

    Thanks, read, for that follow-up!
    I belatedly realised that Турэhэн удэроор should also be Түрэхэн үдэроор. The same rules about vowel harmony apply (masculine and feminine vowels). (I’m not sure of үдэроор, which, if it were a Mongolian word, would be үдрээр, with vowel harmony in the ending to match the ү in үдэр). At any rate:
    Buryat Түрэхэ = Mongolian Төрөх (to be born)
    Buryat Үдэр = Mongolian Өдөр (day)
    You’ll notice that under the Classical script both words would use the same letter (ᠦ) where Cyrillic uses ө and ү respectively. So the vagueness of the old script actually allowed these two variants to coexist within the one standard, whereas Cyrillic makes this into a difference between two languages.

  41. yes, that is true, the old script writes the both pronunciations the same way, the Buriad or Ordos or other dialects seem like older versions of our language and now they are as if like catching up with us 🙂
    i’ve tried last year to follow a FB page for the Uigarjin, it’s so difficult and without knowing the rules i think i would never start reading it, and the rules seem like too obsolete, so i always ask the admin to put the cyrillic galig together and he always says then i’d never learn it, i even get rebuked for proposing the changes in the writing rules and get labeled unpatriotic and suggested that i should go elsewhere to learn the simplified kanjis if i prefer etc, it’s pretty like that, ironic

  42. I should ask questions more often! It will take me a little while to digest this information. Thanks very much for it!
    (Yes I am in Mongolia, in Darkhan. The Buryat text I took from facebook – its author is an English teacher in Duldurga who is not very familiar with computers, I think, which explains the у for ү. I can only assume that үдэроор should be үдэрѳѳр or үдэрээр or something. From what I recall she speaks Buryat at home and half-Russian, half-Buryat at work and in official/administrative contexts.)

  43. Trond 延元 says

    Support the anti-glottalization movement!

  44. SFReader says
  45. Bathrobe says

    SFReader, I’m much obliged for that information. These ‘predicative personal endings’ look very much like cliticised pronouns (би and бид). What are the endings for second and third persons?
    I wrote my posts above in response to PF’s request for clarification and I was learning as I went along. I would be the first to admit that there is a healthy dose of subjectivism in what I wrote (as JC pointed out), and the issue is obviously less simple than it appears.

  46. SFReader says

    Yes, Buryat personal endings seem to derive from pronouns.
    For example, 2nd person singular would be -sh, -shi, 2nd person plural -t
    yabanab – I go, yabanash – you (singular) go, yabanat – you (plural) go.
    Very close Mongolian cognates can be produced simply by adding pronoun after the word
    yavna bi – yabanab
    yavna chi – yabanash
    yavna ta – yabanat
    3rd person singular – unmarked, 3rd person plural -d
    yabanad – they go

  47. SFReader says

    Generally speaking, once one gets used to the pronunciation and small grammar differences, standard literary Buryat (based on eastern Buryat dialects) is remarkably similar to Mongolian.
    Western Buryat is more divergent, I am told (even eastern Buryats have trouble understanding it)

  48. ura! nashego polku pribylo, so to speak
    i mean i am glad to know there is a reader in SF so knowledgeable about Buriad Mongol language, i would be even more glad if you are yourself of Mongolian descent, but it seems you are not, where have you learnt the language? hiigti, yabagti
    i remember watching on TV an American missionary couple speaking perfect Mongolian without any accent, some people are really good at learning languages, they were talking they lived in Khentii for two years iirc

  49. hiigti, yabagti, i meant the words reminded me my grandma

  50. SFReader says

    I think you mean “хэхэбди” – let’s do, “ябабди” – let’s go.

  51. Bathrobe says

    SFReader, I wasn’t sure if you were in San Francisco or just an avid reader of science fiction.

  52. SFReader says

    Buryat sounds like this:
    http://www.youtube.com/watch?v=P9Zh-GPEY40&feature=player_embedded
    Алиа хүүхэн дуун урин дуудаад
    Аян холохоо ерээб шамдаа
    Анхан түрэхэн талам шэнээр
    Аршаан сүршэжэ, бэеым жэгнээл
    Дуутайхан, суутайхан тоонто нютаг
    Зүрхэнэйм охин, зүдэнэйм эхин
    Дуулим буряадайм тоонто нютаг
    Зүрхэнэйм охин, зүдэнэйм эхин
    Дуулим буряадайм тоонто нютаг
    Сагай хайханда ургааш, хэргээш
    Сагаан эдеэгээ элбэг дэлгэээш
    Сарюун үбэлдөөш, арюун зундааш
    Самряан ажалаар омог зандааш
    Дуутайхан, суутайхан тоонто нютаг
    Зүрхэнэйм охин, зүдэнэйм эхин
    Дуулим буряадайм тоонто нютаг
    Зүрхэнэйм охин, зүдэнэйм эхин
    Дуулим буряадайм тоонто нютаг
    Аяын хангалаар нара золгоош
    Агтын жороое хөгжэм болгоош
    Айдар хүбүүдтээ жэгүүр олгоош
    Абын заншалагаар ерөөл үргөөш
    Дуутайхан, суутайхан тоонто нютаг
    Зүрхэнэйм охин, зүдэнэйм эхин
    Дуулим буряадайм тоонто нютаг
    Зүрхэнэйм охин, зүдэнэйм эхин
    Дуулим буряадайм тоонто нютаг
    While even Mongolian without exposure to Buryat would get the gist of it, this is hardly a level of intelligibility characteristic of one language.

  53. “”хэхэбди” – let’s do, “ябабди” – let’s go.”
    no, hiigti, yabagti are commands, do that, go there
    while hiihebdi, yabhabdi are will do, will go, will be doing, going
    i don’t know whether the forms you wrote are correct, let’s do will be, i am not really sure, in Mongolian xiitsgeye, yavtsgaaya, but in Buriad, I’ll ask my father

  54. What a great discussion! Sometimes the silliest posts produce the most informative comments.

  55. Bathrobe says

    We’re onto the controversial field of what constitutes a separate language. (Mutual) intelligibility is often cited as a test for deciding whether a single language or separate languages are involved, but I don’t think it holds much water. For instance, the German dialects are often so different as to be mutually unintelligible, but thanks to the existence of a common ‘standard language’ from which they are treated as deviations, ‘German’ gets to be regarded as a single language. I think the same could be argued for Buryat. A different historical trajectory (less than 100 years ago) and Buryat would have been a dialect of Mongolian. In fact, it didn’t happen this way, and as a result Buryat is now treated as an independent language. Which was the point of my post.

  56. Well, to me there’s a big difference between being “regarded” or “treated” as a single language and actually being one by linguistic/scientific standards: see Arabic, Chinese, etc. But we’ve gone back and forth on that any number of times.

  57. SFReader says

    Now I understand.
    Yabayabdi – let’s go (Mongolian yav’ya)
    Yababdi – we go (Mongolian yavaa)
    Yabayagty – you go! (Mongolian yavaach!)
    Very, very complicated…

  58. Trond Engen says

    I’ve touted my sociolinguistic definition of language before, haven’t I? I’ve boiled it down to

    A language is a dialect continuum with shared prestige varieties.

  59. SFReader says

    Well, actually in Mongolia and China, Buryat functions as a dialect of Mongolian and in Russia as a separate language.
    Same with Oirat (Western Mongol), which is a dialect of Mongolian in Mongolia and China and a separate Kalmyk language in Russia.
    I think this all depends on existence of schools which teach normative literary standard.
    Every Buryat in Mongolia understands and speaks standard Mongolian, so Buryat is effectively just a dialect of Mongolian.
    The same can not be said for Buryats from Russia which often forces Mongolians and Buryats to speak Russian in order to understand each other.

  60. marie-lucie says

    LH: to me there’s a big difference between being “regarded” or “treated” as a single language and actually being one by linguistic/scientific standards: see Arabic, Chinese, etc.
    Most linguists will agree that there is no hard and fast criteria which will definitely decide on purely linguistic factors whether two speech varieties with many resemblances should be classified as separate languages or separate dialects of a single language. Extra-linguistic factors usually intrude, such as whether speakers agree to ignore their linguistic differences (because they feel they are part of a single ethnic or national group), or emphasize them as markers of separateness.

  61. SFReader says

    More complicated case. Here is a Daguur folk song
    http://www.youtube.com/watch?feature=player_detailpage&v=VtjQP8cH7NM#t=173s
    I am willing to bet that this song would be almost completely unintelligible to any Mongolian (maybe a few words here and there, at most a short sentence can be picked)
    Dagur and Mongolian share a bit over 50% of vocabulary. The languages diverged long ago (over a thousand years ago, perhaps. It is thought that Dagur is related to Khitan language, or maybe even its direct successor)
    Nevertheless, there was actually an attempt by Dagur nationalists to claim that the Dagur is just a dialect of Mongolian and that Daurs are Mongols (curiously, Mongols never regarded them as Mongols and showed very little enthusiasm for this claim)

  62. Bathrobe says

    Interestingly, the few comments that that song attracted are all in Mongolian.

  63. “Yabaya,” i can recognize this form for let’s go, others seem unfamiliar to me, but you, I guess, are referring to a dictionary, so those words can’t be incorrect, right? still sounds a little unnatural to me
    “Interestingly, the few comments that that song attracted are all in Mongolian.” it’s bc the person who posted the song is an Inner Mongolian and he has his own public listening to his posts
    I can understand just a few words from the song, something about berries, jimstei etc. if to listen more carefully it seems very close to how our songs lyrics are constructed, so i’ll try to listen later when will have more time

  64. Most linguists will agree that there is no hard and fast criteria which will definitely decide on purely linguistic factors whether two speech varieties with many resemblances should be classified as separate languages or separate dialects of a single language. Extra-linguistic factors usually intrude, such as whether speakers agree to ignore their linguistic differences (because they feel they are part of a single ethnic or national group), or emphasize them as markers of separateness.
    Of course there are no “hard and fast criteria”; there rarely are when dealing with human affairs. But to say hard and fast lines can’t be drawn and reasonable people can differ on edge cases is very far indeed from saying that whatever people want to be true must be allowed to be true. It is of sociological interest that many speakers of Chinese or Arabic insist that there is only one language involved, but it should be irrelevant to linguists. Frankly, it annoys the hell out of me to see scientists bending over backward to accommodate people who have nothing but their feelings to go on (this applies equally to the issue of bones and other remains that are claimed by people who by any scientific standard have no relationship to the people who left those remains tens of thousands of years ago). If people don’t like the findings of science, the education system needs to be improved rather than having scientists go “Oh well then, if you don’t like it, never mind.”

  65. this horchin song, i can understand almost all the words
    http://www.youtube.com/watch?v=_oCTAVgOcBE&feature=relmfu
    the songs is sung in the khalkha mongolian too, just the melody gets a bit different then http://www.youtube.com/watch?v=vx1_O8H8WDE&feature, the second melody in the compilation, the first one is a uzemchin song

  66. marie-lucie says

    LH, looking for instance at the Scandinavian languages, if they had not already been classified by being spoken in separate nations it could be argued on purely linguistic grounds that there is a single “Scandinavian” language spoken in the Scandinavian peninsula, with several dialects and sub-dialects. There are differences, but Swedes and Norwegians can talk to each other, each using their own “language”, without having had to learn the other one.

  67. Sure, absolutely. That’s what I mean about edge cases. I don’t see that that has much if anything to do with Cantonese and Mandarin, two very different languages, being classified as dialects because it offends people to do otherwise.

  68. marie-lucie says

    Perhaps “dialects” reflects the traditional Chinese way of referring to the non-Mandarin speech varieties? I don’t think linguists think of them as “dialects” rather than separate languages.

  69. Bathrobe says

    I suspect these Mongolic languages are somewhere near the Scandinavian situation, but with a bit of Chinese thrown in. SFReader seems to be suggesting that Buryat is ‘naturally’ different enough that it should be treated as a separate language. And indeed, as SFReader points out, Buryat is rather different from Mongolian in a way that the Mongolian languages of Inner Mongolia (China) and Mongolia are not, although even with the Mongolian dialects there can also be problems of mutual intelligibility.
    The Mongolic languages present various issues of classification. The three well-known languages are Mongolian, Buriad, and Oirat, which The Phonology of Mongolian claims are “more or less mutually comprehensible” and have used Classical Written Mongolian and had contact throughout history. As SFReader points out, Buryat and Oirat are treated as separate languages in Russia (Buryat and Kalmyk respectively) but as dialects of Mongolian in China and Mongolia. In addition there are also many “peripheral” languages (Kamnigan, Dagur, Shira Yugur, Mongguor, Santa, Bonan, Kangjia, and Moghol) which are rather more different. So when a Dagur speaker pops up and claims to be speaking ‘Mongolian’, but this claim is not recognised by a Mongolian, then we have a true ‘edge situation’.
    Underlying all this, as usual, is ethnic and national politics. When SFReader made the comment that “this is hardly a level of intelligibility characteristic of one language”, he/she was pushing back against my rather passionate claim that it was the Russians who decided to make Buryat a separate language from Mongolian. (Russian) Buryat as a separate language is now an established fact. Whether you feel the division is natural or artificial will depend very much on your point of view. At any rate, however you view it, the fact is that decisions relating to script reform and selection of the standard language or Dachsprache led to different outcomes in different places. As SFReader points out, in Russia Buryat is a separate language; in Mongolia and China it is a dialect.

  70. Every Buryat in Mongolia understands and speaks standard Mongolian, so Buryat is effectively just a dialect of Mongolian.
    While this rule of thumb is useful, it’s not universally applicable. Every child over age five in Welsh-speaking Wales understands and speaks English, but that does not make Welsh a dialect of English.
    The Wikipedia article on Ausbausprache, Abstandsprache and Dachsprache may be helpful in clarifying people’s thinking:
    “A dialect [of Welsh, for instance] may be an abstand language (again, with respect to some other dialect [of English, say]) without being an ausbau language. This is often the case with minority languages used within a larger nation state, where the minority language is used only in private, and all official functions are performed in the majority language. [Not quite true in Wales, but near enough.] On the other hand, a language may be an ausbau language even when it has little or no abstand from its neighbours, as noted above for Danish, Swedish, and Norwegian. ”

  71. SFReader says

    I think that people can accommodate wide enough differences within a language if they are inclined to do so. One of the most extreme cases – Mingrelian which is regarded in Georgia as dialect of Georgian, even though the differences are on the level of, say, English and Russian…

  72. Perhaps “dialects” reflects the traditional Chinese way of referring to the non-Mandarin speech varieties? I don’t think linguists think of them as “dialects” rather than separate languages.
    Rather it reflects a traditional but misleading Western translation of the Chinese term 方言 fāngyán, which really means ‘local language’ and in Imperial days applied just as much to Korean or Japanese as to Shanghainese or Cantonese. Victor Mair uses the neologism topolect as a direct translation of fāngyán in order to avoid this confusion. It neatly corresponds to the Tok Pisin term tok ples.
    Unfortunately, the Chinese government insists that all the Han in China speak the same language, and linguists working and living in China can’t safely contradict this.

  73. SFReader says

    re Welsh comparison.
    In Mongolia situation is more complex, since due to closeness of languages, the Mongolians having some exposure to Buryat can understand it without difficulty (and some even speak it).
    I’d say, Mongolian and Buryat most resemble English and Scots(not Gaelic!) in Scotland rather than English and Welsh in Wales.

  74. Bathrobe says

    I was minded to mention Scottish vs English. But for the union of the two thrones, English/Scottish might have become something similar to Mongolian/Buryat.

  75. marie-lucie says

    the Chinese government insists that all the Han in China speak the same language, and linguists working and living in China can’t safely contradict this
    Linguists are people too, and have to live in the real world. Local linguists having to toe the line in China does not mean that foreign linguists cannot have their own ideas.
    Victor Mair’s solution (“topolect”) is a good one, since this word does not have the connotations that “dialect” often has among non-linguists. I often use “language varieties” when I don’t feel it would be wise to make a case on purely linguistic grounds.

  76. SFReader says

    Another useful comparison would be with Russian/Ukrainian/Belarussian.
    In the 19th century Ukrainian and Belarussian functioned and were officially regarded as dialects of Russian.
    A case is often made that evil Russian Communists artificially separated Ukrainian and Belarussian in order to prevent unity of East Slavic peoples… ;-)))

  77. Bathrobe says

    A case is often made that evil Russian Communists artificially separated Ukrainian and Belarussian in order to prevent unity of East Slavic peoples
    You are obviously cynical about the idea of reading motivations into Soviet language policy. I’m curious to hear what you think about the Mark Dickens paper linked to above (Soviet Language Policy in Central Asia).

  78. The problem with dialect/language is that even with hard-and-fast rules it can create too many languages for one to be able to manage. Personally I’m for a Trond-Engen definition of language, which would make Chinese something like 12 or so languages, a picture which is obviously false — Standard Cantonese is as unintelligible with a smaller Cantonese dialect as Standard Mandarin. So practically the most convenient way is to always take the smallest possible unit, like Beijing Mandarin or Guangzhou Cantonese or the Wu Chinese of a certain village, and call it a variety in English (un parler in French and a 土语 in Chinese, I would say), and be precise about the level of classification (whether I’m talking about one variety or a group of varieties etc).

  79. SFReader says

    I think Soviet language policy was dictated by much more mundane practical considerations rather than by any grandiose plans to split, merge or create languages…
    For example, 1920s Soviets attempted eradication of illiteracy for all ethnic groups. Obviously, for average Buryat herder’s kid it was much easier to achieve literacy in Buryat than in Russian, Classical Mongolian or Khalkha Mongolian. Obviously, there was a question of grammars, dictionary and in this regard Buryat was much more extensively studied language in Russia compared to Classical Mongolian (as I know, colloquial Khalkha Mongolian was studied in any detail only in Soviet times). Hence, sheer availability of materials also dictated policy.
    There were of course some political constraints as well. Ukrainian nationalism proved during the Russian Civil War to be a powerful force capable of mobilizing large peasant rebel armies. It was much easier to accommodate it rather than attempt to deny existence of Ukrainian on purely linguistic grounds.
    Then there were personal motives of people in charge of language policy(or nationalities policy in general).
    A prospect of becoming a classic founder of a new literary language, a lifetime fame and employment, well, who could resist that? And on administrative levels, of course there was a great demand from politicians for plausible linguistic explanations why there needs to be a separate republic with separate language and separate republican committee of Communist Party…

  80. marie-lucie says

    minus273: The problem with dialect/language is that even with hard-and-fast rules it can create too many languages for one to be able to manage.
    The Ethnologue catalogue of languages falls into this trap.

  81. I think that people can accommodate wide enough differences within a language if they are inclined to do so. One of the most extreme cases – Mingrelian which is regarded in Georgia as dialect of Georgian, even though the differences are on the level of, say, English and Russian…
    I don’t see what it means to say “people can accommodate wide enough differences within a language”; they can call Mingrelian a dialect if they like, but that doesn’t make it true. If the difference with Georgian is on the level of English and Russian (I’ll have to take your word for it), then they can’t understand each other’s languages without pretty intensive study, and it’s blatantly counterfactual to call the one a dialect of the other. It’s as if for political reasons the US government decided to call Spanish a dialect of English—that wouldn’t make it so.
    I think Soviet language policy was dictated by much more mundane practical considerations rather than by any grandiose plans to split, merge or create languages
    Have you actually read about it in any detail? Because if you read, say, Terry Martin’s The Affirmative Action Empire: Nations and Nationalism in the Soviet Union, 1923-1939, you’ll see that they did in fact have grandiose political plans—it wasn’t at all simply a matter of “mundane practical considerations.”

  82. SFReader says

    Mingrelians understand Georgian because it’s the national language which they are taught in school (but their native Mingrelian is not taught at all). Some Mingrelians speak Georgian badly, because they are refugees from Abhazia (which was mainly Russian-speaking multi-ethnic republic in Soviet times)
    Georgians don’t speak and don’t understand Mingrelian at all.
    Very similar to the Welsh-English comparison above.

  83. SFReader says

    I’ll have to read Martin’s book. What I’ve read about Soviet language policy in 1920-30s gives me an impression of a lack of any common policy (and very abrupt policy reversals).
    First there was an attempt to create a unified written language for Mongolian, Buryat and Kalmyk (disregarding already existing Classical Mongolian which served same role), then reversal of policy and decision to have three different written languages, then change of alphabet in favor of Latin, then sudden change in favor of Cyrillic and all sorts of strange decisions on Cyrillic orthography for every language (the most bewildering one – artificial skipping of short vowels in Kalmyk orthography which produces monstrosities like алдршлгдх (try to pronounce that!))

  84. SFReader and read – I’m a little late to this – I should have checked back earlier. I have a question. My Buryat is non-existent and my Mongolian is very elementary so excuse my mistakes. SFReader wrote:
    Yabayabdi – let’s go (Mongolian yav’ya)
    Yababdi – we go (Mongolian yavaa)
    Yabayagty – you go! (Mongolian yavaach!)
    I don’t know the “yavaa” form – is it possible this is yavna? If not, what form is it?
    Read, in your “xiitsgeye, yavtsgaaya” (let’s do, let’s go), is the tsg the plural infix that makes “sain baina uu” into “sain baitsgana uu”? Is this necessary when speaking in the first person plural? When I am inviting someone to go together with me in the moment I generally say “yavax uu” (shall we go); whereas when I inform someone that I intend to go in the moment I generally say
    “yav’ya” – is this correct? Should I be inviting/exhorting people to go together with me by saying “yavtsgaaya”?
    Also do other Mongolian languages or dialects or topolects retain the personal endings on verbs like Buryat does?
    Thanks!

  85. What I’ve read about Soviet language policy in 1920-30s gives me an impression of a lack of any common policy (and very abrupt policy reversals).
    Yes, there were certainly abrupt policy reversals, but they were for political reasons. Soviet ethnic/national politics changed rapidly in those years!

  86. Bathrobe says

    artificial skipping of short vowels in Kalmyk
    Well, in my understanding this is related to epenthetic vowels, which are basically ‘inserted by rule’. I have always had trouble remembering which vowels to leave out in Mongolian writing (үхэр / үхрийн is easy, but it gets a bit curly with the verbs). At least Kalmyk is consistent.

  87. Reading Bathrobe’s comments: Agvan Dorzhiev seems to turn up everywhere. Last time I read about him he was in Tibet advising the Dalai Lama. I wonder if there’s some more comprehensive resource available about him.

  88. SFReader says

    –I don’t know the “yavaa” form – is it possible this is yavna? If not, what form is it?
    Yes, should be yavna (yavaa is also possible, the meaning is of continuing action – going, walking)
    –Read, in your “xiitsgeye, yavtsgaaya” (let’s do, let’s go), is the tsg the plural infix that makes “sain baina uu” into “sain baitsgana uu”?
    It’s the so called participatory verb. The meaning is let’s do it together, let’s go together.
    –“yav’ya” – is this correct? Should I be inviting/exhorting people to go together with me by saying “yavtsgaaya”?
    This is I think called voluntative suffix -wishing to do something or calling someone to do something.
    —Also do other Mongolian languages or dialects or topolects retain the personal endings on verbs like Buryat does?
    I think it’s unique to Buryat, but need to check

  89. SFReader says

    —Well, in my understanding this is related to epenthetic vowels, which are basically ‘inserted by rule’. I have always had trouble remembering which vowels to leave out in Mongolian writing (үхэр / үхрийн is easy, but it gets a bit curly with the verbs). At least Kalmyk is consistent.
    no, no! They are all pronounced in Kalmyk, so I see no reason why they are skipped. Here is how a Kalmyk dictionary looks like
    http://i38.fastpic.ru/big/2012/0517/9c/90434eb84bb5d9df2e078eff1f4b619c.png

  90. SFReader says

    If I am not mistaken, Welsh orthography also tends to skip vowels, anyone knows why?

  91. SFR:
    The rough estimate of the Mingrelian/Georgian split is about 3000 years ago. So English/Russian is a slight exaggeration, but not much: three millennia or six, you still understand nothing. Laz has only been separated for about 500 years from Mingrelian, the date of the Turkic migration that physically separated them (though there were probably dialect differences before that), and apparently individual words can be mutually understood, but connected sentences cannot.
    In Welsh, “w” and “y” are vowels as well as consonants: is that what you’re referring to? I know of no unwritten vowels.
    Irish (and Hiberno-English) inserts a schwa to break up consonant clusters which is not written, and this could be confused with the fact that most short unstressed vowels go to schwa. But trying to write Irish by ear alone is a hopeless undertaking anyhow, as hopeless as French.

  92. Yabayabdi – let’s go (Mongolian yav’ya)
    Yababdi – we go (Mongolian yavaa)
    Yabayagty – you go! (Mongolian yavaach!)

    so i don’t think Yabayabdi, Yababdi,yabayagty are correct there, yabaya is let’s go, yes, but there is no yabayabdi, that di in the end sounds incorrect, yababdi also sounds not correct, there is yabhab, yabhabdi, which means okay, i’ll go that intonation of okay is in there, maybe bc there is that hidden bi in the end of the word..
    yabagti means you go, if to say yabagti daa, it would be a softer form, as if like asking to go or giving a permit
    in your “xiitsgeye, yavtsgaaya” (let’s do, let’s go), is the tsg the plural infix that makes “sain baina uu” into “sain baitsgana uu”?
    i think it’s plural for do(xiitsgee), go (yavtsgaa), be (baitsgaa) when singular form is xii, yav, bai sounding as commands
    when it gets the suffix ye in the end it becomes let’s do, let’s go, let’s be
    when it has +na attached it’s something like present continuous form i guess, yavtsgaana – are doing, are going, are being
    hope it makes sense
    I have always had trouble remembering which vowels to leave out in Mongolian writing
    there is a lesson on that, so “mongol baavar” mnglbvr always have vowels, before or after, and there are rules of the zarimdag giguulegch which lose the vowels depending on when
    http://myagaabagsh.blogmn.net/74061/egshigt-ba-zarimdag-giiguulegch.html

  93. I don’t know the “yavaa” form – is it possible this is yavna? If not, what form is it?
    Yes, should be yavna (yavaa is also possible, the meaning is of continuing action – going, walking)

    yes, yavaa means going, some continuation of action, just alone it is not used much, it’s like auxiliary to some other action, irj yavaa would mean, for example one is coming, xiij yavaa would mean is doing
    another usage is “yavaa yav” would sound as if one says go go, go away
    “…yavaa daa” would mean someone supposes what action might be is going on

  94. In the 19th century Ukrainian and Belarussian functioned and were officially regarded as dialects of Russian.
    Ukrainian also existed as a separate language in the 19th century across the border in the Habsburg Empire, where it was usually known as “Ruthenian”. The Austro-Hungarians had their own political reasons for promoting Ruthenian/Ukrainian as a separate language from Russian (and apparently no one tried to argue it was a dialect of Polish, at least not prior to 1919). The separate history of Hapsburg Ukraine – Galicia and Bukovina – is why Ukrainian is still a living language today and not as moribund as Belarussian seems to be.

  95. Bathrobe says

    OK, I’m afraid I’m highly dependent on The Phonology of Mongolian, but my understanding goes like this:
    (1) In initial syllables in Mongolian languages, there is contrasting vowel length (i.e., long and short vowels).
    (2) Non-initial syllables can have two kinds of vowels: full and reduced. Historically, these derived from long and short vowels respectively.
    The quality of a reduced vowel can be predicted (centralised version of the vowel in the preceding syllable, except for a reduced vowel following [u], which is a centralised version of [e]).
    The placing of a reduced vowel is also predictable; they can be inserted (epenthesised) by a rule and need not be present in phonological representations.
    This results in the two approaches to writing non-initial vowels in the Cyrillic alphabet:
    a. In Mongolian and Buryat, full non-initial vowels are written with double letters, reduced vowels are written with single letters.
    b. In Kalmyk, full vowels are written with single letters, reduced vowels are not written (because they can be predicted — see above). Thus:
    Mongolian улаан Kalmyk улан (full vowel)
    Mongolian баатар Kalmyk баатр (reduced vowel).
    A Kalmyk orthography which writes schwa vowels was officially adopted in 1999 but not widely accepted and abolished in 2001.
    This makes intellectual sense to me, but I’m not an expert so I’m quite willing to be corrected.
    As for the epenthetic vowels proper, it’s possible to get the hang of them but they can be confusing. For example:
    ойлгох ‘to understand’ is pronounced as written, something like [œlgǝx] (sorry, I know it’s not totally accurate, but I’m having trouble hunting for the right symbols and I have to go to work!). But ойлгосон is not, as it might appear, pronounced [œlgǝsǝŋ], it’s pronounced [œlǝgsǝŋ]! This is, again, rule-based behaviour, but sometimes I feel it would be easier if they just wrote ойлгсн so you could put the vowels in yourself!
    Thus, I tend to sympathise with the Kalmyk approach, even if it’s not quite the same thing.
    (read’s link led to a youtube screen, which is blocked here. I’ll check it later with VPN).

  96. is the tsg the plural infix
    actually it does not go together, -ts is separate and makes a noun : yav (go) – yavts (a noun meaning process) – yavtsgaa (you all go, plural command) – yavtsgaaya (let’s go) – yavtsgaana (are going/will be going)
    the same construction in xar (look)- xarts (a look or looking)- xartsgaa and so on
    there are of course exceptions, there is no a noun baits, for example

  97. Mongolian баатар Kalmyk баатр (reduced vowel).
    Should that be “Kalmyk батр”?

  98. Bathrobe says

    No, because in the initial syllable there is a distinction between long and short vowels, which must be marked. It’s in non-initial syllables that the distinction drops down to ‘full’ and ‘reduced’.

  99. Bathrobe says

    And it was my fault for not making that clear.

  100. Reading Bathrobe’s comments: Agvan Dorzhiev seems to turn up everywhere. Last time I read about him he was in Tibet advising the Dalai Lama. I wonder if there’s some more comprehensive resource available about him.
    i just read the wikipedia page on him and there were some references listed there, i remember reading about him that his remains were exhumated iirc and he was found in a sitting position as he was left according to his own instructions (sounds very doubtful how the nkvd would have allowed that, though could be if the body was given to his relatives) not at all changed, mummified or all just bones, and that was regarded as a sign of a holy spirit or his meditational forces(? after death?) preserving the body, must be just rumors, and if it really was so the Siberian cold weather of course played perhaps a greater role, though still the ordinary remains would have undergone the inevitable changes after death perhaps, so who knows, rumors are rumors
    hope this helps if you know Russian http://www.youtube.com/watch?v=bfin8TFek6U or maybe you already found this, if it is on youtube, seems just a nauchno-popularnaya documentary or this
    http://www.berzinarchives.com/web/en/archives/advanced/kalachakra/shambhala/mistaken_foreign_myths_shambhala.html reads like some science fiction
    this one is a Buriad docu on another medical man buddhist monk Petr Badmaev http://www.youtube.com/watch?v=B660oen1aSc&feature=related
    about the Buriad words, my father adds another different word yabham as yavna, and he also couldn’t recognize the forms i thought were incorrect, must be that is how the western Buriads talk, not the hori Buriads
    apologize for my earlier outburst yesterday not perceived (by me) unprovoked though, deleted in the other thread
    too easy and immediate communications these days, should really try to control self in such cases anyway

  101. Hello,
    found Your interesting page because I’m searching for the translation of a line of the buryat song “Toonto nyutag” – You posted the text here.
    Until I will ask my question still one comment concerning this comparing: (in the hope, this place is still “alive”)

    SFReader says:
    May 16, 2012 at 11:25 am

    Now I understand.
    Yabayabdi – let’s go (Mongolian yav’ya)
    Yababdi – we go (Mongolian yavaa)
    Yabayagty – you go! (Mongolian yavaach!)
    Very, very complicated…

    read says:
    May 16, 2012 at 12:54 pm
    “Yabaya,” i can recognize this form for let’s go, others seem unfamiliar to me, but you, I guess, are referring to a dictionary, so those words can’t be incorrect, right? still sounds a little unnatural

    So far as I know, in the case of “to go” there are no big differences between the Khalkha-Mongol language and the Buryaad -only the typical changing from “v” in Mongolian to “b” in Buryat and the typical vowel in the end of the verb in Buryat. So I would say in Mongolian in the infinitive form “yavakh” (явах) and in Buryat “yabakha” (ябаха). During in Mongol the verb is only changing with times, we have in Buryat also the conjugation (someone already wrote about in this discussion). The form “let’s do” is, so far I know, (I only love music from all over the world and sometimes try to translate lyrics … and Mongolian music concerns to my favourites, but never really studied these languages) very similar. So I would say

    to go …… yavakh ——————– yabakha
    I go ……… yavakh / yavadag ——————– yababa/yabadagbi (differences in time/repeats of acting)

    You go …… ” ” ——————– may be: yabash/yabadagshi
    Let’s go ….. yav’ya! —————————- yabaya!

    Very, very similar.

    These words I constructed myself. May be, I made some mistakes. But let’s have a look at an example of another word, I found in a Buryat grammar book:

    to study – hurakha (in Mongolian surakh) … again the typical changing from s in Mongolian to h in Buryat and the vowel in the end of infinitiv in Buryat

    When I’m speaking about learning in gerneral, not a regular act, not in a form like telling a story etc. I would say in Mongolian suradag – bi suradag – I study, chi suradag – You study …
    In the Buryat grammar book I found the following:

    1.p.sg.: huradagbi 1.p.pl.: huradagbdi
    2.p.sg.: huradagshi 2.p.pl.: huradagta
    3.p.sg.: huradag 3.p.pl.: huradag

    BTW, this system (of kind of conjugation with the help of syllables in the end of the verb, which are similar to the personal pronoun) You will find also in the turk languages, a group of languages concerning together with the mongolian languages to the Altaiic languages.

    And still some very well known Buryat words in the form “let’s do”, which we can hear very often in the traditional Buryat dance “Yokhor”: “khataraya” and “duulaya” – let’s dance and let’s sing. This shows, that they have a similar to the Mongolian language form to express this “let’s do”: with the help of the endings, which in Mongolian are: -‘ya, -‘yo, -‘ye, -ya, -yo and -ye (depending of last letter in the word and the vowels). Examples are orokh – or’yo (go/come in), tegekh – teg’ye (to do (sth. what was spoken about before – read this – teg’ye! – I will do so), khijkh – khijye (make).
    This form is not only used to express “let’s do” but also “let ME do”/”I will do”.

    Well, let’s come to my question 🙂 …

    In the 2nd verse of the song “Toonto nyutag” we have the lines

    Saryuun ubeld88sh, aryuun zundaash
    Samryan azhalaar omog zandaash

    Сарюун үбэлдөөш, арюун зундааш
    Самряан ажалаар омог зандааш.

    Сарюун – fresh
    үбэлдөө – in winter which function could have the letter ш in the end? I found the conjunction-endings already in other parts of the sentences, not always directly in the end of the verb. Could this be such a case? So it could be a sign for “You”/2nd p.sg.? But how to translate in this case the whole sentence? The typical Mongolian word for “coming through the winter” in Buryat should be ubelzhekhe. Cannot imagine this could be a form of this verb.

    арюун – clean, clear, bright
    зундааш – the same construction as үбэлдөөш only concerning summer

    Самряан – ??? could it be Самaряан – echo?
    ажал-аар – work-with – through/with the help of work
    омог – pride
    зандааш – zandakha means sth. like grumble, cry … the double “a” could be past, the “sh” in the end 2nd p.sg.

    Any ideas for translation or help for understanding the grammar? Or some other meanings of single words?

  102. I find the very concept of a third or fourth palatalization absolutely frightening.

    And yet the language you speak every day has four palatalizations in its Latin/Romance-derived words.

    The first palatalization happened in Late Latin, and transformed /tj/ to /sj/, as in enuntiatus > enunciate.

    The second palatalization happened partly in Common Romance and partly in France, and transformed /kI/, /gI/ (where I is a front vowel) > /tsI/, /dzI/ > /sI/, /dʒI/. This is “soft” c, g.

    The third palatalization happened in English, changing /sj/, /zj/ in unstressed syllables to /ʃ/, /ʒ/. The first and third palatalizations together account for the pronunciation of -tion. In cases where the yod became a full /i/, this palatalization was blocked, which is why enunciate is not /iˈnʌnʃeɪt/.

    The fourth palatalization changes /tju/, /dju/, /sju/, /zju/ in unstressed syllables to /tʃu/, /dʒu/, /ʃu/, /ʒu/. In stressed syllables, the change is incomplete, and changes /tju/, /dju/, /sju/, /zju/ to /tʃu/, /dʒu/, /su/, /zu/. In North America, however stressed /tj/, /dj/ have become /t/, /d/, blocking this part of the change.

    There is also a completely separate Anglo-Frisian palatalization changing /k/, /g/ /sk/ to /tʃ/, /j/ (or /w/ or zero), /ʃ/ that affects native words only. It applies when yods or front vowels are adjacent.

  103. David Marjanović says

    changing /k/, /g/ /sk/ to /tʃ/, /j/ (or /w/ or zero), /ʃ/

    The different allophones of /g/ responded differently. Actual [g] became [dʒ] when palatalized; this, together with the rare /gː/ [gː], is the source of /dʒ/ in words of native origin. The trick is that [g] was rare: it only occurred behind /n/. Example for /dʒ/ from [g]: singe; for /dʒ/ from /gː/: hedge, midge, bridge… In all other environments, /g/ was [ɣ] like in Flemish today. When this palatalized, it became [j]. When it was not, it eventually became [g] at the beginnings of words and [w] at the ends; this [w] then merged into the preceding vowel when there was one, as in borrow, follow and many others, and became [ʊ] and then [ʌ ~ ə] when there wasn’t, as in Edinburgh.

    /g/ before front vowels today, as in give, is of foreign origin, Norse in this case. (There is yive somewhere in Chaucer, IIRC.)

  104. /g/ before front vowels also exists in some native words where the vowels were originally back, but became fronted by umlaut and later unrounded, as in “geese” or “gild.”

  105. There are also a bunch of words in j- from lenition of ch-, as jaw, jowl, and some of completely unknown origin, as jack ‘small or large thing’, jolt(-head), jim-jam ‘knick-knack’, jim-jams ‘delirium tremens’, jiffy, jitter, jolly(-boat), jerk, jook/jouk, job, jab, jalopy (unless from Jalapa ‘city in Mexico’; cf jalapeño); the list goes on….

  106. Referring back to the initial discussion in the thread, actually, there is a long tradition of Russian absurdist or, more exactly, wordplay slogans at mock demonstrations. More famous ones ask to “join the Fourier series” (Вступайте в ряды Фурье!) which is a word play for “Join the ranks of <something or other>” or give a shout out to “veterans of Brownian motion” (“привет ветеранам броуновского движения!”), where humor comes from motion=movement like in feminist movement, but of course referring to Soviet realities.

  107. Stephen C. Carlson says

    The fourth palatalization changes /tju/, /dju/, /sju/, /zju/ in unstressed syllables to /tʃu/, /dʒu/, /ʃu/, /ʒu/. In stressed syllables, the change is incomplete, and changes /tju/, /dju/, /sju/, /zju/ to /tʃu/, /dʒu/, /su/, /zu/.
    In Australia, I’ve heard assume pronounced as /ə’ʃu:m/. Really odd to this American’s ear.

  108. January First-of-May says

    Вступайте в ряды Фурье!

    I’ve actually seen a continuation of this one: “Сходимость! Равенство! Гильбертово пространство!” (“Convergence! Equality! Hilbert space!”) – presumably a play on the French motto (of which the second part means “equality”).
    (An example is in this blog post, which I just found on Google.)

    On the topic of palatalization – I recall seeing a comment on some linguistic blog (not sure if here or on Language Log) that Irish orthography could really use some nicer method of denoting palatalization than whatever the ch*rp it has already.
    (I also recall seeing a Russian linguistic puzzle about Irish – it had to do with consonant mutation, I think – that apparently used apostrophes to indicate palatalization, like in transliterations of Russian. For several years I had no idea that Irish wasn’t spelled that way.)

  109. David Marjanović says

    /g/ before front vowels also exists in some native words where the vowels were originally back, but became fronted by umlaut and later unrounded, as in “geese” or “gild.”

    Oh. Of course. Analogously to king, kitchen and a few others.

  110. Вступайте в ряды Фурье!

    Free the Indianapolis 500!

  111. Greg Pandatshang says

    Okay. Question about developments in early Greek, and this seems like about the best place for it, since Greek was mentioned at least once upthread.

    What’s the current thinking on how the infamous labiovelar > dental change (conditioned by front vowels) happened? Are we looking at a third Greek palatalisation affecting velars? e.g. something like kʷe > kɥe > kje > ce or tɕe > te. Or is this contradicted by other evidence?

    The shift to dental is blocked for voiced labiovelars in front of /i/ (as opposed to /e/). Any hypotheses about how that happened?

  112. Good question; I look forward to the responses from those who know more than we.

  113. Eli Nelson says

    I have always assumed it was palatalization, but I don’t know the specifics.

  114. Hat: Could you fix the bad typo in my comment above on the four palatalizations? For enunciatus read enuntiatus. The typo is of course the result of the palatalization!

  115. Done!

  116. David Marjanović says

    Yes, it’s palatalization. The interesting thing is that it can’t have passed through [ts], because other things did and yielded a different outcome (the -tt-/-ss- variation among dialects). So how about: [kʷɛ] > [kɥɛ]* > [cʷɛ] > [tɕʷɛ] > [tʃɛ] > [tɛ].

    One inscription from Arcadia (5th century BC, so not that early) seems to contain a letter for one of the affricates in question.

    This idea isn’t original to me on the whole, but as usual I forgot my source…

    * There doesn’t seem to be a superscript ɥ in Unicode. [kʲʷɛ] = [kʷʲɛ] would be equivalent transcriptions…

  117. Greg Pandatshang says

    I was thinking thinking that the palatalization of labiovelars might have taken place later than the one that produced tt/ss, with the result that it could follow some of the same pathways with different outcomes. The labial articulation might have temporarily insulated the consonant from palatalising tendencies. Come to think, are there examples of KʷjV in Greek, i.e. the conditions for the second palatalisation?

    Any thoughts on the failure of gʷi to follow an analogous path > di?

  118. David Marjanović says

    I should leave that to Piotr 🙂

  119. There was a small discussion of these things on my blog some time ago.

    *Cj palatalisations are older than those before front vowels. Before *j, velars and labiovelars fell together already in Proto-Greek. The stops coalesced with the palatal glide. Word-medially, the voiceless ones (plain and aspirated alike) yielded *ʧː, and the voiced ones *ʤː (long palatoalveolar affricates). The former developed into -ss- in most dialects except Attic, Boeotian and parts of West Ionic, where it became -tt-; the latter developed into -zd- (or whatever ζ stands for) in Attic-Ionic, and mostly -dd- elsewhere.

    Word-initially we have the same developments except that long consonants are shortened. In Myceanean we have the same spelling (transliterated as z) for all these combination; it certainly represents some voiceless and voiced coronal affricates.

  120. Greg Pandatshang says

    Thanks for the link, Piotr. You’re saying, pace Wikipedia, that only the tenuis *kʷ palatalised (became a dental) before -i? Wikipedia says that the aspirate labiovelar behaved like the tenuis, with plain voiced the odd man out.

    Were there other cases of palatalisation in Greek around that time conditioned by front vowels (as opposed to glides)? My impression was that that was much later. If so, that’s a strength of my back-of-envelope model above: it posits an allophonic glide developing in front of labiovelars in this environment. There aren’t any distinctions in Greek between Kʷ vs. Kw, are there?

    It also makes sense, I think, that /ji/ could tend to neutralise to /i/ more readily than /je/ to /e/. Why that would occur only with some stops (plain voiced) and not others (tenuis), I have no idea.

  121. That’s right. Only * was palatalised before *i (at any rate in Attic-Ionic); *, *gʷʰ became Gk. b, in that position. ópʰis ‘snake’ is a good example of what happened to the aspirate. In Aeolic, labial reflexes of all the three series before any front vowel are the norm at least in the basic vocabulary (less likely to be borrowed): pémpe ‘five’, etc. Oddly enough, however, Aeolic has palatalisation in the extremely frequent function words te ‘and’, and tís ‘who’.

    Palatalisations before *j were Proto-Greek. The labiovelar developments (including palatalisation before front vowels) are a much later affair, with a lot of dialectal variation. Of course Mycenaean still had unshifted labiovelar phonemes.

    The contrast between * and *kw/*ḱw is detectable intervocalically in a few Greek words in which the cluster is reflected as -kk- or -pp-. The former outcome is visible in lákkos ‘pond’ and pélekkos ‘axe-handle’; the only obvious example of the latter is the ‘horse’ word híppos.

  122. There is a Doric inscriptional form ενδεδιωκοτα (prob. = ἐμβεβιωκότα) which seems to show palatalization of *gʷ before *i, but then it might be assimilated from ενδεβιωκοτα. I think ὄφις is the only data point for the behavior of the aspirate before *i.

    The asymmetries in these sound changes are certainly odd. It may be useful to think in terms of an ongoing phonemicization process that interacted with a sound change. That is, the old labiovelars had a range of allophones in different environments (and in different dialects), and when the diffused pan-Greek change of labiovelars to labials came along, some of these were perceived as labiovelar and underwent the change, and others were too phonetically different so were missed by the change and struck out on their own as new phonemes (affricates at first, later merging with dental stops except in Arcado-Cypriot). Which still doesn’t explain of course why the allophone of *gʷ before *i fell into the former class and that before *e (except in Aeolic) into the latter.

    There are some other weirdnesses, e.g. Aeolic has ἀδελφεος with a palatalized reflex. That seems to happen with high-frequency words (τις, τε), so maybe we need some kind of Labovian lexical diffusion model.

  123. I think ὄφις is the only data point for the behavior of the aspirate before *i.

    I think you are right. Unfortunately, only the accusative of the root noun for ‘snow’ (*snigʷʰm̥) has left a reflex in Greek (νίφα), and the locative (*snigʷʰi) has been lost. The superlative ἐλάχιστος ‘smallest’ does not directly continue PIE *h₁léngʷʰ-isth₂o- but shows the influence of ἐλαχύς < *h₁ln̥g⁽ʷ⁾ʰú-, with the labial component abducted by the following *u. Still, there is good evidence of *gʷi > βι in Attic-Ionic, and there is not a single example of *gʷʰi > θi.

  124. Are there any good ideas out there about the possible relationship of ἔχις “viper”, which looks similar to ὄφις in sound and sense but can’t be derived from a labiovelar?

  125. Mihaylova (2016), The Pre-Greek Substratum revisited mentions that *kʷ > *k is a typical feature of Pre-Greek substrate / Paleo-Balkan adstrate vocabulary of Indo-European origin, though she doesn’t seem to treat this word in particular. Probably because another similar feature given is *Dʰ > *D, but I guess satemization first, stop shift later would be a natural chronology…

  126. Stephen C. Carlson says

    Beekes, Frisk, and Chantraine all derive ἔχις from PIE *h₁eģʰi- based on their interpretation of ἐχῖνος ‘hedgehog’ as “snake-animal” = “snake-eater”. On the other hand, they derive ὄφις from h₃egʷʰi-.

  127. David Marjanović says

    Interesting paper. That particular substrate looks remarkably Germanic, even more so than it looks Armenian/Phrygian.

  128. Trond Engen says

    Albanian also looks sorta Germanic. Well, at least there are some remarkable isoglosses.

    I think Hamp has argued that the ancestors of Greek and Germanic must have been in close contact.

  129. Greg Pandatshang says

    Ah, but hasn’t somebody reputable argued that the ancestors of Greek and just about anything you can think of must have been in close contact?

  130. I’ll write about the ‘snake’ words in Greek later. It’s an interesting story but one that can’t be summarised in two sentences.

  131. Greg Pandatshang says

    I see the similarities between Germanic and this “Pergamic” substrate that Mihaylova describes. Can someone advise on the highlights of Germanic-Albanian similarities?

    I a long time ago I read, can’t remember where, a detailed hypothesis arguing that the Pre-Proto-Germanic speakers were located in, I believe, the lower Danube area. Close enough, I suppose, to be part of an Old Balkan IE phylum, now largely extinct.

  132. Piotr, that’s what I was hoping to hear 🙂

  133. David Marjanović says

    Albanian also looks sorta Germanic.

    In phylogenetic analyses it tends to come out somewhere around Germanic and Italo-Celtic. However, unlike both and unlike this substrate in Greek, it hasn’t undergone the kentum merger.

    Can someone advise on the highlights of Germanic-Albanian similarities?

    Apparently some word roots are shared; I don’t know which ones. Other than that, the Albanian outcome of *T-T clusters (i.e. things like *dʰ-t across morpheme boundaries) is s in Albanian. In Germanic, it’s *ss; in Celtic and Italic, it’s also *ss, except when it’s *st which may or may not be analogical in every single case; all attested stages of Albanian, which aren’t many, lack consonant length. All other IE branches have quite different developments: in Greek, Iranian and Balto-Slavic, it’s *sT; in Indic it’s *TT; Anatolian kept the original *TsT; I have no idea about Tocharian.

  134. Greg Pandatshang says

    If it everybody handled those clusters like a German would, we’d have Bussism rather than Buddhism!

  135. The ‘snake’ etymologies are so confusing because there must have been two different and deceptively similar terms for snakes (presumably for different species). In the areas where the early IE-speakers may have lived there are several kinds of snakes, and at least some differences between them are important — especially the difference between the venomous ones (various species of Eurasian adders, Vipera) and the harmless ones (especially the grass snake, Natrix natrix, the smooth snake Coronella austriaca, and the Aesculapian snake, Zamenis longissimus, which is impressively big but so harmless that it was kept as a temple pet by the Greek and the Roman priests of Asclepius). Of course the IEs were not zoologists, so any term for a snake could also have been extended to legless lizards (slowworm, Anguis fragilis) and even snake-like fish such as eel (the Latin word for them, anguilla, is evidently some sort of diminutive derived from anguis ‘snake’; cf. also Slavic *ǫgorь ‘eel’ vs. *ǫžь ‘snake’).

    In an article I cited here, Norbert Oettinger (2010) reconstructs (A) *ángʷʰ-i-s, gen.*n̥gʷʰ-éi-s ‘grass snake’ (a.k.a. water snake) and (B) *h₁óǵʰ-i-s, gen. *h₁éǵʰ-is ‘(land) snake’ (let’s call them ‘snake A’ and ‘snake B’). Oettinger makes a solid case and I am inclined to accept these reconstructions, though I would argue that the first word requires an initial *h₂-, even if one has to abandon a potential Hittite cognate (discussed in the article).

    Snake A is the direct ancestor of Lat. anguis, Lith. angìs, Slavic *ǫžь, Arm. awj (all showing the full grade of the PIE strong cases), as well as Ved. áhi-, Avestan aži (with the zero grade reflected as IIr. *a). The Vedic word could equally well reflect *h₁éǵʰ-i-, but the Iranian one couldn’t. Germanic *unka- ~ *unkan- probably also belongs here (though its derivation is more complex). In several languages the word for ‘eel’ is derived from the same root. Snake A seems to have been the “generic snake” (including various other slitherin’ things, from slowworms to eel), perhaps prototypically the grass snake, as suggested by Oettinger. Swims well, eats frogs.

    Snake B is directly reflected in Greek ἔχις ‘adder’ and concealed in the derived words for ‘hedgehog’: Gk. ἐχῖνος < *h₁eǵʰi-h₁n-o-, Germanic *egīla- < *h₁eǵʰi-h₁l-o-, Armenian azni < *h₁oǵʰi-h₁n-ijo-, Balto-Slavic *eźja- < *h₁eǵʰjo-. Hedgehogs usually feed on insects and snails, but they also eat snakes and are to some extent immune to the adder’s venom. They are poor swimmers, however, and don’t like marshy or waterlogged habitats, hence Oettinger’s idea that they were named after snakes inhabiting dry places — and that’s precisely where European adders live. We can therefore prototypically associate snake B with Vipera berus and related species. Dangerous, lives in dry places, gets killed by hedgehogs.

    It seems that at least in Greek and Armenian (possibly also in Tocharian, cf. Toch. B auk) the two kinds of snake hybridised: The hybrid had the head of snake B (*h₁ó- or *h₁é-), the tail of snake A (*-gʷʰ-i), and the meaning of the latter. If the same thing had happened in Indo-Iranian, we wouldn’t be able to tell the difference, since the outcome of *h₁é- and h₂n̥- is exactly the same there. Oettinger thinks that the contamination took place in the common ancestor of Greek and Armenian, but the Graeco-Armenian hypothesis is a shaky construct. If instead we propose an areal innovation of the southeastern periphery of non-Anatolian Indo-European (Greek, Armenian, Tocharian and Indo-Iranian), there is no necessity to reconstruct a zero grade in Indo-Iranian, Oettinger’s reconstruction of *ángʷʰ-i-s (or *h₁ángʷʰ-i-s) is not ruled out, and the connection with the Hittite ‘serpent’ word illuyanka/u- can be saved.

    There are several other terms for snakes or snake-like things, more or less transparently derived from verb roots meaning things that snakes do, e.g. *serp- ‘crawl’ (Lat. serpent-, Alb. gjarpër, Ved. sarpá-) or *(s)neh₁- ‘spin, wind’ (Lat. natrix, PCelt. *natrī, PGmc. *naðra- m. ~ *nēðrōn- f., hence adder from a false analysis of a nadder). Slavic *zmьjь ‘serpent, dragon’ and *zmьja ‘adder’ are derived from the ‘earth’ word.

  136. David,

    PIE *-tt- (phonetically *[tst]) is indeed reflected as /s/ in Modern Albanian. But this development can’t be directly compared to the Italic, Celtic and Germanic change of *-tt- > *-ss-, since Modern Albanian /s/ reflects an older postalveolar affricate (conventionally transcribed *č). To my mind, the most likely scenario seems to be as follows: Early Proto-Albanian had two “colours” of affricates: “light” *[ʨ, ʥ], resulting from the affrication of pre-Albanian “palatal” (front velar) stops, and “dark” (retroflex, labiovelarised) *[tʂʷ, dʐʷ], resulting mostly from the palatalisation of PIE labiovelars before front vowels (but palatals followed by *w merged with the dark set: *[ʨw] > *[tʂʷ]). The inherited cluster *[tst] was phonetically simplified to *[ʦ] at that stage, and the originally allophonic s-insertion was reinterpreted as distinctive affrication. On account of its apical (rather than “domed”) articulation it was auditorily closer to *[tʂʷ] than to *[ʨ], and so became identified with it. The Late Proto-Albanian “hardening” of *[ʨ] to > *c [ʦ] > th [θ] took place after this merger.

    Edited to add: also *[tj, dj] > *[ʦ, ʣ], > *[tʂʷ, dʐʷ] > Alb. [s, z]

  137. David Marjanović says

    That explains why this s hasn’t become gj, as *s did in initial position…

    However, the “West IE” development *[tst] > *[sː] can’t have happened in one stage anyway; we pretty much have to postulate *[ts] as an intermediate stage. So, could this first step still be shared with Albanian?

  138. It could be. The various dialectal developments of *[tst] are often too trivial to mean much but the intermediate stage *[ts] shared by Italic, Celtic, Germanic and Albanian is somewhat unusual. In a cluster like [tst] the final stop belongs to a syllable onset — normally a position where consonants are strengthened, not weakened; so of the three members of the cluster it’s the least likely to be lost.

    By the way, the reflexes of *s in Albanian are really odd. Syllable-initially under stress it became voiced *[ʒ] (which then merged with the reflex of *j, ending up as the modern palatal affricate gj); in other positions, *s generally yielded *[ʃ] (modern sh), but combinations of *[ʃ, ʒ] with other segments gave rise to mind-boggling complications. For example, *sw under stress > *[ʒw] > *[ʥv] (too late to change into *[dʐʷ]), then > *[ʣv] > *[dv] > d (the voiceless development probably leads to th [θ], but the evidence is scarce). The cluster *st yields sht, but *sp, *sk/ks > *[ʃf, ʃx/xʃ] > [f, x] (f, h).

  139. Aramaic עַכְנָא /ʕaxnaː/ ‘snake’ is supposedly from ἐχῖνος, but I don’t know why the ε would be borrowed as /a/. (The ‘ayin doesn’t bother me, but perhaps it should.)

  140. Greg Pandatshang says

    Assuming that Mihaylova’s Pre-Greek stratum was originally a distinct clade from Greek, then one thing I find interesting about it is that its Grassmann’s Law reflexes imply that its so-called voiced aspirate series was actually aspirated. Otherwise, only Greek and Indic attest aspiration/murmur as a distinctive trait of this series, so, if this counts, it’s a 50% increase in attestations.

    P.S. As Mihaylova makes plain in the article, her Pre-Greek stratum is phonologically almost identical to Thracian. Hence, to the extent Germanic resembles the former, it also resembles the latter. Wikipedia has a useful table summarising some similar traits between IE languages of the area (Albanian is not included): https://en.wikipedia.org/wiki/Thracian_language#A_Thracian_or_Thraco-Dacian_branch_of_Indo-European

  141. Many of these traits are suspect. For example, I would agree that the “Thracian” reflex of *sr is str (Ἴστρος for the lower course of the Danube), but if so, Thracian must be distinguished from Proto-Albanian, in which *s was clearly lost in this position (with compensatory lengthening of any preceding vowel). In Germanic (if one wants to compare it with Thracian) the t-insertion can’t be very old (IMO it’s younger than Verner’s Law), and in Baltic it’s less consistent than in Slavic.

    We know very little about Thracian, but what we do know suggests a language with a treatment of syllabic resonants completely different from that found in Albanian, and possibly with a “Grimmoid” shift in the stop system — again something not found in Albanian (or the Albanian-like substrate in Romanian). The linguistic fragmentation of the ancient Balkan/Carpathian area must have been considerable (like that of Italy or Iberia). Macedonian was a close cousin of Greek, Phrygian seems to have been a more outlying member of the same clade; so, possibly, was Messapian, and it’s anybody’s guess where “Illyrian” belonged (if it was a single language in the first place). I personally incline towards the view that “Dacian” ≈ Proto-Albanian. The “Thraco-Phrygian”, “Thraco-Dacian” and “Thraco-Illyrian” hypotheses look equally untenable.

  142. Trond Engen says

    I’ve tried to rediscover a comparative table of vocabulary between the IE branches, which had Albanian grouping surprisingly often with Balto-Slavic and Germanic. With Dacian being Proto-Albanian being the substrate in Romanian, this could have to do with a common origin in the successor cultures of Cucuteni-Tripolye,

  143. David Marjanović says

    Macedonian was a close cousin of Greek, Phrygian seems to have been a more outlying member of the same clade;

    Armenian, too?

  144. David Marjanović says

    Otherwise, only Greek and Indic attest aspiration/murmur as a distinctive trait of this series, so, if this counts, it’s a 50% increase in attestations.

    Aspiration does make it easier to explain the Italic fricatives (especially the voiceless ones), the Germanic ones, and why Winter’s law (Balto-Slavic) didn’t operate before both voiced sets.

  145. David: Armenian, too?

    Phylogenetic algorithms almost always recover a Graeco-Armenian grouping (while Albanian moves around). My impression, however, is that the innovations shared exclusively by Greek and Armenian are mostly lexical (and therefore easily borrowable within the circum-Pontic area). Armenian is not only Satem proper (with an early merger of velars and labiovelars) but may also participate to some entent in the RUKI change (though the jury is still out on that). My personal, slightly heretical hypothesis is that Armenian occupies a basal position with respect to the “Core Satem” group (IIr. + BSl.). The position of Thracian may be similar, while Greek, together with a number of “Paleo-Balkan” satellites, forms a group of its own. Albanian, as always, moves around.

  146. Trond Engen says

    I’ve occasionally enjoyed the idea that Armenian could be the descendant of the Cimmerian migration to Anatolia in the 7th century BCE.

  147. Quite certainly the early Armenians were in contact with Hurro-Urartian speakers somewhere near Lake Van, not later than the Cimmerian invasions. The Thracians were well known to Homer. Both groups may have developed on the northern coast of the Black Sea during the 2nd millennium BC and then migrated south, seeking independendence as the Scythians consolidated their rule over the Pontic steppes.

  148. Trond Engen says

    My scenario was that the Cimmerians settled in Urartu as allies of the Medes after they were driven out of Phrygia by the Lydians. But there were Cimmerian raids on Urartu in the 8th century, even before they are supposed to have settled in Phrygia. So maybe there were Cimmerians in Urartu all through the 7th century, supported by the Medes, and that those driven out of Phrygia found refuge there with their brethren. They then helped the Medes bring down Assyria in 612 BCE and were instrumental in forming the Satrapy of Armenia in the early 6th century.

  149. David Marjanović says

    Interesting.

  150. Trond Engen says

    Yes, at least the chronology seems to work. And with Phrygo-Armenian failing on phonological grounds, it explains how the archeologically attested spread of Phrygian culture from the west ended up bringing Armenian into Urartu.

    Also, with Armenian a founding member of the Legions of Satem, the location of the Urheimat is right. When I first conceived of the idea, I think I imagined Armenian as the stay-behind cousin of Greek, continuing the relationship with Indo-Iranian for longer, but that isn’t really an argument for anything without a good theory of how and when Greek came into Greece.

  151. I would agree that the “Thracian” reflex of *sr is str (Ἴστρος for the lower course of the Danube)

    As well as Strymon — to which compare Rumon, the ancient name of the Tiber according to Servius, which seems to point to a change *sr- > r- in Italic, rather than the *sr- > fr- change for rejecting which we took Vladimir Diakoff to task in the Dziebel thread some time ago. But we don’t know for sure that the river name is Italic, I suppose.

  152. Greg Pandatshang says

    Certainly aspiration could explain the voiceless fricatives in Italic, so that fits the aspiration/murmur model nicely and could be considered an indirect attestation. Is there any reason to think aspiration explains what we see in Germanic or with Winter’s Law? Regarding the former, Greek underwent a similar spirantisation without contrastive aspiration on the mediae, as far as I know.

  153. TR,

    Yes, Στρῡμών is a perfect example of a hydronym with an etymology which is transparent and self-evident even if you know next to nothing about the language it came from. And it’s remarkable how much we can infer from a single river name, which by the way has an exact cognates in Balto-Slavic (*s(t)reumõ/-men-).

  154. Trond Engen says

    (I use too big words up there. It sounds like I think it’s some major breakthrough when it’s just tossing imaginary pieces around on a map and see if you like the pattern. But I do like the pattern…)

    Wikipedia’s list of cognates for attested Thracian words has more and closer matches with Baltic than with any other branch.

  155. Trond Engen says

    So let’s suggest that in the 8th century BCE the expanding Scythians (using newly developed mounted warfare to monopolize trans-continental trade) rammed into the North Pontic semi-nomads with full force and knocked them (or at least their ruling classes) about like billiard balls. The Cimmerians went southeast, Thracians southwest, and maybe Balts up north. And no doubt some into the Pannonian Basin. There is always somebody pouring into the Pannonian Basin. What was to become Slavs stayed behind under Scythian rule.

    Or the Thracians could have moved towards the Aegean riches a little earlier.

  156. David Marjanović says

    the Legions of Satem

    Perfect.

    But we don’t know for sure that the river name is Italic, I suppose.

    Or perhaps a round-trip through Etruscan will do the trick. Or perhaps the name of the city is actually Phoenician, and the name of the river was remodeled after that…

    Is there any reason to think aspiration explains what we see in Germanic or with Winter’s Law? Regarding the former, Greek underwent a similar spirantisation without contrastive aspiration on the mediae, as far as I know.

    I never said anything about “contrastive”. 🙂 I don’t know if phonetically unaspirated voiced plosives can directly become fricatives; it is clear (from many languages in India) that aspirated ones can, and it’s also clear (from Middle Chinese, twice in a row: “first muddy” and “second muddy consonants”) that voiced plosives can spontaneously become aspirated for no discernible reason just as well as voiceless ones.

    Vowel lengthening before voiced plosives seems to be pretty common and comes rather naturally to me: it takes a while to build up enough pressure for a plosive if you’re keeping your vocal cords on. During such lengthening, the pressure needed later for aspiration would escape, so I expect aspiration to block it (and indeed my attempts at pronouncing voiced aspirates result, if anything, in vowel shortening – but of course I’m not familiar with any language that actually has such sounds).

    Wikipedia’s list of cognates for attested Thracian words has more and closer matches with Baltic than with any other branch.

    The Baltic-only matches aren’t many either, though, and the list is generally not well made. That Bulgarian “thunderbolt” word with /f/ is obviously a loan from the Greek word in the same line…

    And no doubt some into the Pannonian Basin. There is always somebody pouring into the Pannonian Basin.

    True.

  157. Trond Engen says

    The Baltic-only matches aren’t many either, though, and the list is generally not well made. That Bulgarian “thunderbolt” word with /f/ is obviously a loan from the Greek word in the same line…

    Yes, I should have added “FWIW”. But it does support the impression that Thracian belongs in the middle of Satem.

  158. That’s why it’s a basin.

  159. David Marjanović says

    I casually mentioned the name of Rome being Phoenician. The full proposal is here.

  160. Very interesting! Here’s the abstract:

    After criticizing attempts to explain the name of Rome without proper regard for the topography of the place and the early linguistic history of Italy, the Carthaginian dimension of early Rome is sketched and a Phoenician origin of the name is proposed: +Rōma, the Phoenician counterpart of Hebrew råmåh ‘elevation’, itself a frequent place name in the Semitic world, may have been the name of a trading post at or on that hill (out of the proverbial seven) standing closest to the ford at Tiber Island where two prehistoric trade routes joined and traversed the Tiber. A possible role of the originally pre-Indo-European names Palātium, Etrusco-Latin designation of the Palatine Hill, and Rūmōn, old name of the Tiber river, for the selection of +Rōma as name for the location is considered.

    What does the + mean?

  161. David Marjanović says

    “Reconstructed”, as opposed to * for “wrong”. Vennemann is pedantic this way (and several other ways).

    Not to be confused with the almost opposite convention of Asterix and Obelix.

  162. What does the + mean?

    Vennemannese for *.

  163. Hah. Thanks!

  164. The Baltic-only matches aren’t many either, though, and the list is generally not well made. That Bulgarian “thunderbolt” word with /f/ is obviously a loan from the Greek word in the same line…

    And the alleged cognates of briza mean ‘rice’ and come ultimately from Dravidian. If anything, one might attempt to compare the Thracian word with the Germanic and Balto-Slavic terms for ‘rye’.

  165. The Venneman article is very unconvincing to me, because he presents no actual evidence that Rome was a Phoenician trading post, and because after rejecting proposals for an IE etymology on the grounds that “The Indo-Europeans are recent intruders into the Peninsula” he then proposes a Phoenician etymology which he himself concedes cannot be older than 700 BC.

  166. Well, that’s Theo Vennemann. He also had a Vasconic etymology of the Thames, supported mainly by the fact that Tamisa might be segmented into Tam-is-a, where *-is- is a hypothetical (Proto-)Basque word for ‘water’ (not a valid reconstruction, according to Michelena and Trask), and -a is of course the Basque definite article. And Tam- (unlikely to be Basque on account of its /m/) is… well… something.

  167. Stephen C. Carlson says

    I don’t know if phonetically unaspirated voiced plosives can directly become fricatives
    Is this what is alleged to have happened with the ancient Greek mediae (β, γ, δ)? Or is the claim that these were aspirated before they became fricatives?

  168. Intervocalically at least, “plain voiced stop > voiced fricative” is a common type of lenition, cf. Spanish and presumably the rest of Romance NW of the La Spezia–Rimini Line. The (unconditioned) fricativisaton of Slavic *g in a number of Slavic languages did not involve any aspiration/murmur either.

  169. David Marjanović says

    There’s no evidence for aspiration/murmur in these cases, but can we rule it out as a short-lived intermediate stage?

    In Spanish, the sounds in question aren’t fricatives, they’re approximants. I think voiced plosive > approximant can go directly.

  170. There’s no evidence for aspiration/murmur in these cases, but can we rule it out as a short-lived intermediate stage?

    We can’t rule it out, but Occam’s razor will slice it right off. In my view, naturalness is at most a statistical property of sound changes; it has only weak predictive power, and for particular cases one must sometimes fall back on the Shit Happens Or Doesn’t theory.

  171. David Marjanović says

    Three proposals for an IE etymology of Rome are considered plausible in this long-winded article in German, from most to least plausible as ranked by the author:

    *h₂arh₃- “to till the soil”
    *h₂r̩h₃-mó- “field”, “furrow” or something
    *h₂roh₃-mó- vr̩ddhi adjective: belonging to the field/furrow
    *h₂roh₃-m-ah₂ collective (“neuter plural”): “everything field-related”, perhaps “the fields” or “arable land”

    There is a place in Germany that is simply called Acker, so people sometimes really are unimaginative like that.

    *werh₁- “say”
    *wroh₁-m-ah₂ as above: perhaps “the places where people speak” > “place of assembly”

    *werh₁/h₃- “be hot, burn (intransitive)”; not otherwise found in Italic
    *wroh₁/h₃-m-ah₂ as above: perhaps “the hearths”, “settlement”

    Several other proposals are discussed at length and deemed too improbable. Vennemann’s is not mentioned.

    The name Rumon (mentioned only once in an ancient source, vowel lengths unknown) for (part of?) the Tiber remains mysterious; there’s no way to make its -n make sense in Italic, and the rest can’t be cognate with stream because *sr- gives fr- in Latin (and probably the rest of Italic).

    (BTW, the presentation of German historical phonology in the first part of the article is… so close and yet so far. However, the context for this is the presentation of a beautiful back-and-forth-and-back-and-forth loan.)

  172. The name Rumon (mentioned only once in an ancient source, vowel lengths unknown) for (part of?) the Tiber remains mysterious

    An old misspelling of rumen ‘throat’, for a narrowing in the river?’

    (Incidentally, cognate with stream.)

  173. Stephen C. Carlson says

    I just thought that Roma was an originally Etruscan name (Ruma). Bichlmeier’s argument against that isn’t very probative in my opinion.

  174. David Marjanović says

    An old misspelling of rumen ‘throat’, for a narrowing in the river?

    That makes sense!

    (Incidentally, cognate with stream.)

    Can’t be, for the same reason.

    I just thought that Roma was an originally Etruscan name (Ruma).

    Other than some of its kings, Rome (and surroundings) never seems to have had an Etruscan population. Bichlmeier mentions that. Is Ruma even a known Etruscan name at all?

    (…Of course, “known” and “Etruscan” in the same sentence…)

  175. Stephen C. Carlson says

    It’s been a long while since I was fascinated by Etruscology, but I can readily find that TLE 300 refers to an Etruscan from Rome, Cneve Tarχunies Rumaχ.

    I realize that an Etruscan etymon for Roma would be obscurum per obscurius, and this guy, https://www.academia.edu/246175/Die_etymologische_Herleitung_des_Namens_R%C5%8Dma also faults the Etruscan hypothesis, which he attributes to Schulze 1904. So I grant that it could be worthwhile to look for possible IE sources for the name. I’m just unimpressed with the results.

  176. the rest can’t be cognate with stream because *sr- gives fr- in Latin (and probably the rest of Italic)

    Is there more than one data point for this (frigor : ῥῖγος)?

  177. David Marjanović says

    this guy, […] also faults the Etruscan hypothesis

    That’s the source Bichlmeier builds his review on.

    I’m just unimpressed with the results.

    Why? Those are the three most parsimonious hypotheses as far as I can see.

    Is there more than one data point for this (frigor : ῥῖγος)?

    I need to check. We discussed this issue in some detail on that really long thread with Dziebel and Diakoff in it…

  178. Yes, I mentioned that upstream; IIRC that was the only example we found.

  179. And that one is no good anyway, because the initial rh may just be automatic, rather than being rh- < hr- < sr- > Latin fr-.

  180. marie-lucie says

    I don’t know if phonetically unaspirated voiced plosives can directly become fricatives

    What about French intervocalic [v] ultimately from Latin [p], most likely with an intermediate [b] as in Spanish, for instance Lat lupa ‘female wolf’, Sp loba (now weakened to approximant), Fr louve (occasional loss of final schwa being relatively recent)?

  181. JC, PIE is thought to have disallowed initial /r/, and there are some possible Slavic cognates that have sr-.

  182. David Marjanović says

    Well, we simply don’t know if there was a short-lived breathy-voiced intermediate, as we happen to know for Middle Chinese because we happen to have Chinese descriptions of the pronunciations of Chinese and Sanskrit from just the right times.

    On the other hand, the intermediate may have been an approximant as found in Spanish today, which then merged with the existing phoneme /v/ (…as also happened in Spanish, though much later and in the opposite direction).

    My dialect more or less always turns intervocalic /b/ into /v/, even though /b/ is voiceless in all environments – no sound between [b̥] and [ʋ ~ ṽ] ever occurs unless people have a cold; you’d expect [b] or [β] to surface on occasion, but they don’t. I suspect the historical intermediate was an approximant that was reinterpreted as /v/ (approximants are after all much more easily voiced than voiceless).

  183. Is there more than one data point for this (frigor : ῥῖγος)?

    Not word-initially, I suppose, but it’s certain that intervocalic *-sr- > Lat. -br- (supported by several exceptionally good etymologies), so at least frīgor is what we would expect.

  184. And that one is no good anyway, because the initial rh may just be automatic…

    But then the otherwise perfect formal and semantic correspondence between ῥῖγος and frīgus (n.) ‘cold, frost; chill, cold shiver’ AND between ῥῑγέω and frīgeō would have to be regarded as pure coincidence despite the fact that both Gk. /hr-/ and Lat. /fr-/ MAY derive from *sr-. The reconstruction *sriHg- (or *sriHǵ-) wins on grounds of parsimony and in order to falsify it one would have to come up with solid etymologies ruling out *sr- > Lat. fr-. It isn’t a question of what we know for sure but where the burden of proof falls.

  185. Stephen C. Carlson says

    Why? Those are the three most parsimonious hypotheses as far as I can see.
    Unless I’ve overlooked something, none of those hypothetical etymons for Roma actually have an attested descendent anywhere else, and the meanings are very vague.

  186. Not necessarily coincidence; could be borrowing, like G rhadix < L radix that we discussed before. German the Dzeibel was arguing for a sound-law operating here too.

  187. Borrowing which way? And why should Gk. /hr/ have been substituted for Lat. /fr/ or vice versa? Actually, it would have to be Gk. –> Lat., considering that ῥῖγος and ῥῑγέω occur already in Homer’s Greek. But I don’t think we have any other example of /hr/ –> /fr/ or a borrowed es-stem producing a whole range of native-looking derivatives in Latin.

  188. The Rumon : Στρύμων connection still looks intriguing. Given that Rumon probably can’t be Latin, if they are related, what are the options? Do we know anything about *sr in the rest of Italic? (I assume any other IE languages that may have been spoken in the region are too obscure to say anything meaningful about.)

  189. TR: So you don’t think much of my Rumon < rumen idea?

  190. Y, the *Srūmōn etymology looks too tempting to me given that we actually have that attested as an IE river name (and *sr- > r- is an easy change also attested in e.g. Greek). Rumen seems to be a rare word, and the deformation to -on would be unmotivated, especially since native Latin words don’t end in -on.

  191. J.W. Brewer says

    Japanese ラーメン (“ramen”) is said by the internet to have a boringly recent and transparent etymology, but if only someone could come up with an alternative etymology tracing it back to a reconstructed Proto-Vasco-Dravidian root, I’m sure we could come up with a semantic just-so story confirming that’s also the non-IE ancestor of “Roma[n].”

  192. The normal derivational cycle in PIE was as follows:

    (1) A proterokinetic deverbal neuter: *sréu-mn̥/*sru-mén-s (Gk. ῥεῦμα).
    (2) An amphikinetic animate noun formed by internal derivation: *sréu-mō(n)/*sru-mn-és (Lith. sr(i)aumuõ, “Thracian” Strūmōn).
    (3) Thematic derivatives, e.g. *srou-m/n-o- (Germanic *sraumaz).

    It is possible that Lat. flūmen is (1) with /fl/ for /fr/ due to the contamination of *frū- < *sreu- with *plū < pleu-. If Rumon has anything to do with (2), the language of the namers would have had to be Greek (or indistinguishable from Greek in this particular case), since the restoration of the final *-n in the nom.sg. is a Greek innovation. One imaginable scenario is that the Italo-Greek colonists gave the River *Tiβeris an alternative name like *Ῥεύμων ‘the Stream’ (unattested but perfectly derivable via animatisation, like τέρμων = τέρμα ‘boundary’). The problem is how we get from something like *Rūmōn to Rōma (a feminine ā-stem). Here Greek isn’t of any help.

  193. Interesting. Of course, Strūmōn itself shows that other languages than Greek also restored the *-n, so we could also be dealing with some other non-Italic IE language (we know such existed in Italy, e.g. Messapic).

  194. Strūmōn is not the original Thracian name but its Greek version (with Greek inflectional morphology). Herodotus thought that all Old Persian personal names ended with -s, which is surprising, given that *-s > -h in Iranian (unless affected by RUKI, as in Dārayava(h)uš). But of course he based his judgement on the Greek versions of Persian male names (Ξέρξης, etc.).

  195. January First-of-May says

    Herodotus thought that all Old Persian personal names ended with -s

    As well as all the Greek holiday names, IIRC (which is what clued me in that it was probably just the Greek nominative ending).

    Similarly, in modern Latvian, all male names (of any origin) end in -s, and the ones that don’t are mangled until they do (as in Džordžs Bušs, Latvian for “George Bush”).
    IIRC something similar is going on in Lithuanian as well (but I don’t recall the details). And yes, as far as I know, the Greek and Latvian endings are cognate to each other.

     
    …This factoid would probably make for a nice Jeopardy-style question: “According to Herodotus, all the names of Persian kings ended with this Greek letter.”
    More obvious in English than in Russian, though (since English kept the Greek nominative endings in those names, while Russian dropped them – Кир, Дарий, Ксеркс).

    I once tried to make up some questions for a linguistic version of What Where When.
    I came up with this one, some Absurdopedia jokes, a question about Middle-Earth, a few others I don’t recall at the moment, and a hilarious one about Old Persian (one of the titles of Achaemenid rulers was “king of kings”; you might not be able to pronounce it in the original Old Persian, but what is it in modern Persian? the answer being shahinshah, the 20th century Iranian title).

  196. Strūmōn is not the original Thracian name but its Greek version (with Greek inflectional morphology)

    Yes, that makes sense.

  197. Džordžs Bušs

    That is absolutely wonderful.

  198. David Marjanović says

    IIRC something similar is going on in Lithuanian as well (but I don’t recall the details).

    I don’t know the details either, but here are two examples: direktorius, Sergejus Tarasovas.

    you might not be able to pronounce it in the original Old Persian

    xšāyaθiya xšāyaθiyānām

  199. Say “xšāyaθiya xšāyaθiyānām Džordžs Bušs” five times fast!

  200. IIRC something similar is going on in Lithuanian as well (but I don’t recall the details).
    Lithuanian has a stem vowel before the /s/ that has been apocoped away in Latvian: Džordžas Volkeris Bušas .

  201. Greg Pandatshang says

    To sum up Bichlmeier a bit, -mā looks like it could be a deverbal collective, so we’re looking for any IE verb root with a relevant meaning and the form XrH (let X be any consonant that would become silent word-initially before r) that could plausibly be in o-grade; or, for that matter, Xrh₃ that could plausibly be in e-grade.

    If we’re using -eh₂ > ā, then I guess we’re assuming that it was shortened analogically after it became a place name?

  202. David Marjanović says

    Old Latin: vocative a, nominative ā, ablative ād
    Classical Latin: vocative a, nominative also a, ablative ā

    When the -d disappeared, the Romans preferred to merge the nominative with the vocative instead of the ablative. Understandable.

  203. Trond Engen says

    Bichlmeier’s derivation of Rome as a -ma collective would have been much stronger if supported with parallels in Italic, and in toponymy, and especially in Italic toponomy.

  204. David Marjanović says

    Interesting paper. That particular substrate looks remarkably Germanic, even more so than it looks Armenian/Phrygian.

    And then I completely forgot about it and will have to read the paper again from scratch, my interest piqued by the facts that 1) Hamp thought his “Prehellenic” was close to Germanic (for reasons at present unknown to me), and 2) “Crotonian” had Verner’s law and a small part of Grimm’s.

    I also didn’t remember the “snake” discussion or the speculations about Armenian, but did remember the Greek labiovelar cacophony and the Rome stuff and most other fragments of this thread…

    It seems that at least in Greek and Armenian (possibly also in Tocharian, cf. Toch. B auk) the two kinds of snake hybridised: The hybrid had the head of snake B (*h₁ó- or *h₁é-), the tail of snake A (*-gʷʰ-i), and the meaning of the latter. If the same thing had happened in Indo-Iranian, we wouldn’t be able to tell the difference, since the outcome of *h₁é- and h₂n̥- is exactly the same there. Oettinger thinks that the contamination took place in the common ancestor of Greek and Armenian, but the Graeco-Armenian hypothesis is a shaky construct. If instead we propose an areal innovation of the southeastern periphery of non-Anatolian Indo-European (Greek, Armenian, Tocharian and Indo-Iranian), there is no necessity to reconstruct a zero grade in Indo-Iranian, Oettinger’s reconstruction of *ángʷʰ-i-s (or *h₁ángʷʰ-i-s) is not ruled out, and the connection with the Hittite ‘serpent’ word illuyanka/u- can be saved.

    Somewhere around “how to kill a dragon”, so probably by Watkins himself, it was proposed that the adder/dragon word was a negation (*n̥-), fitting its mythological role. This could be folk etymology of the zero-grade derived from the mythological role, or taboo deformation, I guess – Polish niedźwiedź “bear” for expected *miedź- comes to mind.

    it’s also clear (from Middle Chinese, twice in a row: “first muddy” and “second muddy consonants”) that voiced plosives can spontaneously become aspirated for no discernible reason just as well as voiceless ones

    Yes, except that bringing up the “first muddy” and “second muddy” consonants was a misunderstanding on my part. The “first muddy” ones are simply the voiced obstruents (which did become aspirated, and are now voiceless aspirated in Mandarin in most tones), the “second muddy” ones the (voiced) sonorants.

  205. Savalonôs says

    David W. Anthony in The Horse, the Wheel, and Language reports that the primordial serpent was called

    *Ngʷhi, the Proto-Indo-European root for negation

    Anthony isn’t a linguist. I just think it’s interesting that he picked up on that ostensible fact.

  206. Commenting on a very old comment but….

    @John Cowan: very few dialectal phonemic splits. The only significant one is bad/lad

    How about BATH/TRAP and LOT/CLOTH?

    Otherwise I agree… As for orthographic reform of English, there is sooooo much low-hanging fruit that worrying about dialects is letting perfect be the enemy of good.

  207. David Marjanović says

    low-hanging fruit

    Obligatory.

  208. BATH/TRAP is a split followed by a merger, but it’s very inconsistent (one of the few regional differences in AusE phonology is whether dance is PALM or TRAP). As for CLOTH, it exists only for old farts in the U.S. Northeast (like me) and for very old RP-speaking farts, I think: everywhere else CLOTH has either re-merged with LOT or been swallowed up in the LOT-THOUGHT merger as if it never existed.

  209. David Marjanović says

    There’s the Scottish TIE-FLY split, as noticeable as it is mysterious.

  210. It’s just an outlier of the Scottish Vowel Length Rule, and of course matches up with Canadian Raising.

  211. David Marjanović says

    Oh.

  212. David Marjanović says

    Me, long ago:

    There doesn’t seem to be a superscript ɥ in Unicode.

    There is: ᶣ – it’s just not in the font Arial Unicode MS.

  213. Lars Mathiesen (he/him/his) says

    When I set the font-family of that p element to Arial, I get the prettiest little ᶣ. Firefox claims that it’s using the Windows 10 system font “Arial”. So maybe modern system fonts have overtaken the ones we used to download to get more of Unicode.

    (The system font Consolas that Firefox uses for <code> also has . Whoda thunk that back when we had 8-bit serial terminals. And matrix printers with fanfold paper for consoles).

    (Also also it turns out that on my fairly vanilla Windows 10, Firefox ends up using the same Arial font for most of Language Hat, since it can’t find matches for “Helvetica Neue” or “Helvetica”. Setting the font family manually did make a difference when I was using a Mac, since œ and ɶ were identical in Apple’s Helvetica).

  214. John Cowan (September 19, 2016):

    The third palatalization happened in English, changing /sj/, /zj/ in unstressed syllables to /ʃ/, /ʒ/. The first and third palatalizations together account for the pronunciation of -tion. In cases where the yod became a full /i/, this palatalization was blocked, which is why enunciate is not /iˈnʌnʃeɪt/.

    That can’t be the reason, since palatalization wasn’t blocked in negotiation, appreciation, etc. (Or at least, *mostly* not blocked.) But it is *mostly* blocked specifically in words with -nunciat-. John Wells wrote:

    (And by the way why does pronunciation have -si- rather than -ʃi-?).

    I’m not sure that I can unravel all the historic and synchronic factors involved here.

    One influence is that the verb is pronounce instead of pronunciate; the same goes for (an-, de-, re-)nunciation, and I guess enunciate, enunciation are carried along in their wake (the opposite verb forms do exist, but they’re much less common). Exactly that reason was given by some 19th-century orthoepists, while others insisted that -tia- and -cia- must always be palatalized, because that’s! the! rule! Here’s Walker’s Critical Pronouncing Dictionary (1791) on the word pronunciation:

    There are few words more frequently mispronounced than this. … But these speakers ought to take notice, that, throughout the whole language, c, s, and t, preceded by the accent, either primary or secondary, and followed by ea, ia, io, or any similar diphthong, always become aspirated, and are pronounced as if written she. Thus the very same reasons that oblige us to pronounce partiality, propitiation, especially, &c. as if written parsheality, propisheation, espesheally, &c. oblige us to pronounce pronunciation as if written pronunsheashun. — See Principles, No, 357, 450, 451.

    And that wasn’t just Walker: it was a controversy throughout the 19th century, with several pronunciation manuals quoting a story of an American traveler in England who was “mortified” by his failure to say “pronun-shee-ashun”. The Century Dictionary endorsed “pronun-shee-ashun”, and well into the 20th century, some dictionaries were still listing it as a standard option; Merriam-Webster didn’t demote it to nonstandard until 1961. Even today, there are /-nʌnʃ-/ pronunciations listed as second options for at least some -nunci- words in some current dictionaries.

    So the moral is: for Latinate words with -tiat-/-ciat-, there’s no simple answer to whether the -tia-/-cia- is palatalized and why. Different analogies are pushing in different directions, and for nouns in -tiation/-ciation, there’s dissimulation as well.

  215. Stu Clayton says

    there’s dissimulation as well

    Nobody is under oath here either.

  216. Damn, too late for the edit window!

  217. Stu Clayton says

    Palliativation is blocked.

Speak Your Mind

*