13 March 2013

Italian cembro, cembra 'Swiss pine' (updated)

Swiss pine


















Italian cembro, cembra, cirmolo 'Swiss pine (Pinus cembra)' is cognate to Friulan cirmul, Ladin cirum, zirm, Raetho-Romance ğèmber (Engadin), žember (Bergün/Bravuogn), Romanian zâmbru, and High German Zirbe, Zirbel, Zirbelkiefer. This would point to a substrate loanword *kemro- ~ *kirmo- which my colleague Marco Moretti considers to be of Tyrrhenian (Rhaetic) origin. 

Another name for the Swiss pine used in Switzerland is Arbe, Arve < *arbo-, with loss of the initial postvelar and denasalization.

For these words I'd propose a Vasco-Caucasian etymology from Caucasian *qq’wǝmV 'grain, fruit stone', Sino-Tibetan *kuam 'fruit (with) kernel', Yeniseian *ʔem- (~ x-) 'cone (of a coniferous tree)'.

Pine cone
 
This etymology would be also the source of Greek kônos 'cone', Altaic *tʃhumu 'seed; cone' (Mongolian, Tungusic)1 and Paleo-Basque *sini > Basque zi (S, Z), (R) 'oak acorn', with palatalization of the initial velar. 
 

_________________________________
1 Conflated in the EDAL with Turkic *tʃɨm 'turf, meadow; various kinds of grass (with seeds)'.

39 comments:

  1. You cannot have a matchup of a root with a velar occlusive k- with one with a fricative or affricate, without proving evidence. You are postulating random sound correspondences, and in any case I see your methodology is blurred, not transparent, vain. I'm sorry to see this trend.

    ReplyDelete
  2. Hi, Marco. Is this you?

    Surely I haven't presented the evidence here, but in my Yahoo list I've shown several examples of this "centumization" of sibilants in non-IE loanwords into IE languages.

    For example, I link the word *k´ormon- 'weasel; ermine, stout' found in several IE languages to NEC *tsa:rggwɨ: 'weasel, marten'. Also Latin cīmex 'bedbug' can be linked to ddzimidzA 'a k. of stinging insect'.

    My guess is this centumization happened in a prehistoric language spoken in the Alpine/Danubian area, from which these words seem to have originated.

    In my work there's no place for what you call "random sound correspondences", Marco.

    ReplyDelete
  3. Dear Octavià, I don't believe in this so-called "centumization". Do you have a single irrefutable occurrence of this phenomenon in a real language? I use to read every post in the Yahoo list, but even there I see something very similar to random sound correspondences, plus several questionable semantic shifts.
    I'm sorry, but it makes no more sense your review of my work, because yours is an entirely different stuff, founded on entirely different assumptions. There's almost no single item not suffering from severe difficulties. I can't approve all this confuse playing with supposed outputs of extremely complicated phonological systems. I hope that it will not offend you, but it's quite better for me to continue my work following my advice.
    Marco

    ReplyDelete
  4. Marco, please don't drown in a glass of water! I think there's no reason to split our "society" before we discuss the matter in deep.

    The idea of "centumization" was actually suggested to me by Starostin himself in an old Russian article where he studies NEC (i.e. Vasco-Caucasian) loanwords into PIE. He discovered sibilant affricates were mapped into PIE palatal(ized) stops, with different outputs in "centum" and "satem" languages. A good example would be (late) PIE *H1ek´u-o- 'horse' from NEC *ɦɨ[n]tʃwi (~ -e) id.

    I think that besides obvious cases of assimilation of a velar stop to a following front vowel, the native core of PIE didn't have palatal(ized) stops at all, as the former post-dental affricates collapsed to DENTAL stops (an idea suggested by Allan Bomhard and Jörg Rheimeier. This also includes the instances of *h1 corresponding to a former voiced affricate, such as the 'elm' word.

    ReplyDelete
  5. I also don't understand your accusations regarding some of my etymologies while your proposal for Italian cembro is IMHO phonetically implausible.

    If you accept for example a Basque word meaning 'oak acorn' is ultimately related to a NEC root designating some k. of fruit or berry, then why do you object the meanings 'pine' or 'elm' couldn't be related at all?

    Also the substrate loanword *k´armon- 'weasel, ermine' found in several European IE languages (Rhaeto-Romanic, Germanic, Balto-Slavic), as well as (as a compound member) in Basque armindderi, erminori , can be straightforwardly linked to NEC *tsa:rggwɨ: 'weasel, marten', assuming a)centumization of the initial sibilant and b)a labial output of the labiovelar cluster (the nasal /m/ is due to the assimilation to the nasal sufix).

    By contrast, within the Vasco-Caucasian phylum thisd root has a broader semantic latitude: Burushaski *ɕargé 'flying squirrel', native Basque sagu 'mouse'. Would you call them "questionable semantic shifts"?

    ReplyDelete
  6. Whit such a methodology you could mix together very different lost strata. Very dangerous. Imagine a future world in which Yiddish and Maltese are the only known survivors of all human languages. What kind of protolanguage could be reconstructed?

    ReplyDelete
  7. The root *k´armon- 'weasel, ermine' is not native in IE, and seems excedingly far from NEC *tsa:rggwɨ: 'weasel, marten'. If a common origin exists, it is more remote than proto-Tyrrhenian. We lack information. What was the real Etruscan word? Probably it was KARMU, that is attested in anthroponymy. But what is the source of this KARMU? It is directly from proto-Tyrrhenian? It was borrowed? From what language? I don't believe at all that NEC *tsa:rggwɨ: 'weasel, marten' could become KARMU in Etruscan in a direct way. Even if the two items were outputs from the same remote root, they are too diverse to be useful in order to understand their history. First it is necessary to have a good reconstruction of Etruscan, then a good reconstruction of proto-Tyrrhenian. If we lack these reconstructions, all we can do is to play with chance resemblance.

    ReplyDelete
  8. This is why predictable sound correspondences are needed to differentiate between them.

    From sibilant affricates we can get up to 3 different kind of outputs: a) a sibilant /s/, b) a dental stop due to the collapse I described before, c) a velar stop from centumization. IMHO each of these outputs belong to a different prehistorical layer.

    Please notice your NEC-Tyrrhenian correspondences also include (b). This should tell us something about the Neolithic languages spoken in the Danubian area. As I said in Yahoo, even the PIE numeral *Hok´te-h3(u) '8' can be reconstructed as a dual of PTyrrhenian *hemq'y-da '4'.

    ReplyDelete
  9. The root *k´armon- 'weasel, ermine' is not native in IE,
    Yes, I agree. This why I labelled it as a "substrate loanword".

    If a common origin exists, it is more remote than proto-Tyrrhenian.
    I don't think Tyrrhenian was the source language, as it had different sound correspondences.

    IMHO the corresponding native Etruscan (i.e. inherited from PTyrrhenian) form is tlapu 'mole' (> Latin talpa). We've also got PAltaic *tHàrba 'a k. of small animal.

    ReplyDelete
  10. Probably some languages, lacking sounds like /ts/, replaced them with /k'/. This could be a rational formulation of the phenomenon you call "centumization". I hope you will agree with me that it is not easy to analyze remnants of languages more ancient than Tyrrhenian and classifying them, understanding their precise origin in time and space. What is native? What is borrowed? E.g., Proto-Altaic *tHarba could be a loanword.

    ReplyDelete
  11. Probably some languages, lacking sounds like /ts/, replaced them with /k'/. This could be a rational formulation of the phenomenon you call "centumization".
    Yes, this is precisely what I think it happened in IE.

    I hope you will agree with me that it is not easy to analyze remnants of languages more ancient than Tyrrhenian and classifying them, understanding their precise origin in time and space.
    Sometimes extra-linguistic data can help us. For example, according to this, Etruscan θra 'milk' can be linked to the Neolithic LBK culture. Then from the NEC etymon is *=a:mdzU 'to milk; to drink' we can tell the voiced affricate *dz collapsed to a dental stop in the language(s) spoken there.

    What is native? What is borrowed?
    As the collapse of post-dental affricates into dental stops can be also seen in loanwords into IE and Altaic, I guess this didn't happened in PTyrrhenian (unless of course you want to resurrect somebody's Indo-Tyrrhenian theory).

    ReplyDelete
  12. Then from the NEC etymon is *=a:mdzU 'to milk; to drink' we can tell the voiced affricate *dz collapsed to a dental stop in the language(s) spoken there.

    Of course! This was an etymology of mine. The collapse of affricate to dental stops is a quite common phenomenon in NEC and in Tyrrhenian, and I always included it in my work - as I find some regular correspondeces.

    However, the collapse of an affricate to a dental stop is quite different form the collapse of an affricate to a velar stop.
    The first one is typical of Tyrrhenian, despite all you are trying to prove, while the second one (if really existing) is not typical of Tyrrhenian, but of some unknown pre-Tyrrhenian language. If an IE or Altaic item has a dental stop from an affricate, I deduce immediately it is a loanword from Tyrrhenian.

    ReplyDelete
  13. You are trying to invalidate the whole set of regular correspondences I found, replacing it with a huge number of invasive random etymologies don't belonging to Tyrrhenian, without any apparent logical order.
    I don't believe in brainstorming, and you have to accept this fact. One single good idea is quite better than a neverending flux of bad ideas. Blogs and Yahoo lists are not appropriate for Science.

    ReplyDelete
  14. Of course! This was an etymology of mine. The collapse of affricate to dental stops is a quite common phenomenon in NEC and in Tyrrhenian, and I always included it in my work - as I find some regular correspondeces.
    Then we must consider those Tyrrhenian items with a different output (there's pently of them in your file) as loanwords from other languages such as Hurrian.

    If an IE or Altaic item has a dental stop from an affricate, I deduce immediately it is a loanword from Tyrrhenian.
    As the mass of loanwords into IE with this sound shift are usually relative to Neolithic technology, I must suppose it's a characteristic of the languages spoken by Neolithic farmers, among them (but not exclusively) Tyrrhenian.

    However, the collapse of an affricate to a dental stop is quite different form the collapse of an affricate to a velar stop.
    The first one is typical of Tyrrhenian, despite all you are trying to prove, while the second one (if really existing) is not typical of Tyrrhenian, but of some unknown pre-Tyrrhenian language.

    I'd say a non-Tyrrhenian language such as native IE, for example.

    You are trying to invalidate the whole set of regular correspondences I found, replacing it with a huge number of invasive random etymologies don't belonging to Tyrrhenian, without any apparent logical order.
    Not at all. I've just pointed the inaccuracy of some of your etymologies and corrected them. I'm afraid you're drowing in a glass of water.

    You must also accept the fact that Tyrrhenian-speaking people are relative late comers (about 1,200 BCE) in Italy, so it's likely they had borrowings from the pre-existing languages.

    I don't believe in brainstorming, and you have to accept this fact. One single good idea is quite better than a neverending flux of bad ideas. Blogs and Yahoo lists are not appropriate for Science.
    Then you're a much more dogmatic rigid person than the Marco I used to know, as he didn't object discussing these things in Yahoo. :-)

    ReplyDelete
  15. Perhaps you don't want to realize that SC and NEC phonological systems are very complicated, so correspondences are not necessarily plain and simple. Sergej Starostin wrote about this matter:
    "I shall start — as is universally required in such cases — with outlining the system of phonetic correspondences between the languages in question. This requires perhaps one preliminary remark. I cannot get rid of a feeling that most of the criticism — including the criticism of such distinguished scholars as S. Yakhontov — is due to the complexity of phonological correspondences between the languages in question. Unfortunately, this is not something that can be amended through any amount of additional research. Correspondences between very complex phonological systems — and SC, especially NC systems probably belong to the most complex in the world — are bound to be complex. So if the reader wants to see a plain and simple system of correspondences between SC families, he
    might as well stop reading this text right now and join the camp of critics."

    Did you ask me for a table of correspondences between SC and proto-Tyrrhenian? No, you did'nt. You simply began to "correct" items in a very random and erratic way. When I uploaded a table of regular correspondences between Vasconic and NEC, this was largely ignored, and my work was so alterated by you to become almost unuseful. Do you have tables of correspondences to prove your proposals? Apparently you don't, and you are not even trying to prepare a single one. What you have is a very long series of posts in the Yahoo list and in this blog, without any attempt of systematic thinking. You have no overview, but only isolated ideas without mutual connection. Your neverending flux is overwhelming but not coherent.

    ReplyDelete
  16. You say: "I've just pointed the inaccuracy of some of your etymologies and corrected them. I'm afraid you're drowing in a glass of water."

    After all you have not pointed the inaccuracy of my etymologies: you have just changed them without providing any evidence of your choice, only inventing absurdities such as "Fournet's Law" or similar gibberish.

    You are acting in this way:
    1) you change something
    2) if I complain about this, you pretend to restore my original proposal, after many pressions
    3) then you change a dozen of other items, just for retaliation, making any further discussion impossible.

    Your "corrections" mainly consist in this:
    1) You wake up with a sudden statement: "Now I think this Tyrrhenian item comes from this NEC root".
    2) If I try to discuss it, not approving your proposal, after my comments, you say: "Now I think that the Tyrrhenian item comes from this other NEC root (different from the former)".
    3) After several iterations like this, you finally say: "Ops! I returned to my primitive idea".

    In the meanwhile, I'm filled with anger and tiredness, so I stop to discuss - having many other things to do in my life.
    All this is a strategy of tension and psychological war. All this must come to an end.

    Your proposals quite often lack of semantic reliability. You apparently don't care of semantics in any way. So a strawberry becomes beer, a sheep becomes a pig, a born becomes a dog, a mole becomes a bear, and so on. I'm not drowning in a glass of water. I decide how I have to work with my own ideas, and I'm very tired of your methods.

    ReplyDelete
  17. You have a strange idea of what a "society" should be. Years ago I deleted my Tyrrhenian files in order to work on them. Unfortunately, I hadn't the opportunity of carry on these studies in the meanwhile. Nethertheless you took those files from your hard disk uploading them in your list (without my permission, but this is only a detail). I told you I would have uploaded new versions soon, but you didn't wait: you began your terrible flux of bad ideas, harassing me, making very difficult to me to have the time and the necessary quiet in order to work.

    "You must also accept the fact that Tyrrhenian-speaking people are relative late comers (about 1,200 BCE) in Italy, so it's likely they had borrowings from the pre-existing languages."
    And so what? I never told Tirrhenian originated in Italy. Every language has borrowing from pre-existing languages, even that of isolated people like the Saami and the Nenets have many substratum and adstratum items. In any way, I wrote a document with many etymologies of loanwords in Etruscan, even from unknown IE languages, so I don't understand your objection.

    "Then you're a much more dogmatic rigid person than the Marco I used to know, as he didn't object discussing these things in Yahoo. :-)"
    Yes, if someone doesn't accept your way of erratic discussion (pardon, it is "brainstorming"), then he is a "dogmatic" guy. Also Glen Gordon uses the very same word to label his oppositors: as for him, they are all "dogmatic". Pokorny and Pedersen wrote books, not posts. We have to do something similar.

    ReplyDelete
  18. I'm very disappointed for your deletion of an old post because of a comment of mine. I repete here my idea. I'm a splitter, not a lumper. Lumpers think that English "much" is related with Spanish "mucho". As for them, English "day" is identical with Spanish "día". And much worser: they would try a connection between English "bad" and Persian "bad". If the well known languages of the present days give such occasions to lumpers, I don't imagine what could be deduced from remnants of lost substrata.

    ReplyDelete
  19. Did you ask me for a table of correspondences between SC and proto-Tyrrhenian? No, you did'nt. You simply began to "correct" items in a very random and erratic way.
    Well, I've pointed out some inconsistences in your etymologies, for example as regarding as the output of affricate sibilants. At least now I know whose ones you regard as Tyrrhenian-native, so everything else must be borrowed.

    You should also notice the number of etymologies I've actually fixed (e.g. cembro, spur, span-) is quite small.

    In any way, I wrote a document with many etymologies of loanwords in Etruscan, even from unknown IE languages, so I don't understand your objection.
    That's OK, but I'd put all the non-native etymologies (even those NEC-related) in that file.

    What you have is a very long series of posts in the Yahoo list and in this blog, without any attempt of systematic thinking. You have no overview, but only isolated ideas without mutual connection. Your neverending flux is overwhelming but not coherent.
    Although at first sight they might seem unconnected, in the long run my ideas tend to form a coherent picture. My aim is directed towards these long-term results.

    I'm sorry if you dislike my methods, but this is the way I think. After all, you won't tell a painter how he/she has to paint, will you?

    Yes, if someone doesn't accept your way of erratic discussion (pardon, it is "brainstorming"),
    Yes, this is the right word.

    Pokorny and Pedersen wrote books, not posts. We have to do something similar.
    I agree with you. However, a book like that might take years to write (maybe a decade or even more), so meanwhile shorter chunks of work are useful.

    ReplyDelete
  20. I'm very disappointed for your deletion of an old post because of a comment of mine.
    Actually I had forgot that post until you made that comment. Now I've put in back the draft list because I didn't want to draw readers' attention to its outdated content.

    After all you have not pointed the inaccuracy of my etymologies: you have just changed them without providing any evidence of your choice, only inventing absurdities such as "Fournet's Law" or similar gibberish.
    I think I've provided enough evidence for what you call "gibberish", although it doesn't look like I've convinced you.

    You have a strange idea of what a "society" should be. Years ago I deleted my Tyrrhenian files in order to work on them. Unfortunately, I hadn't the opportunity of carry on these studies in the meanwhile. Nethertheless you took those files from your hard disk uploading them in your list (without my permission, but this is only a detail).
    A little correction: I didn't uploaded your original files but my own revision of them.

    In the meanwhile, I'm filled with anger and tiredness, so I stop to discuss - having many other things to do in my life.
    Marco, I'm sorry if unintentionally I've done or said things that hurted you, as I truly regard you as a friend of mine.

    But sincerely, I don't think I deserve your nasty judgements. Not of all my ideas are as bad as you think.

    Your proposals quite often lack of semantic reliability. You apparently don't care of semantics in any way.
    By no means.

    So a strawberry becomes beer,
    I never said that! This was other's people idea which I've refuted anyway.

    a sheep becomes a pig, a born becomes a dog, a mole becomes a bear, and so on.
    Although yours is a clumsy generalization, within large time depths like the ones we found in long-range comparison (apparently) puzzling semantic shifts are possible.

    For example, the Sinitic word for 'dog' is derived from a Vasco-Caucasian root which in NEC means 'mouse, rat' and in Basque and Iberian is 'rabbit'. The singularity of the Sinitic word (unrelated to Tibeto-Burman) is a reflex of the early domestication of the animal in East Asia.

    Perhaps you don't want to realize that SC and NEC phonological systems are very complicated, so correspondences are not necessarily plain and simple.
    I'm aware of that, and I've also pointed you some of the inconsistences in Starostin's work. For example, it looks like PNEC is actually very close to the real PSC, with Sino-Tibetan being its offspring rather than a sibling.

    Unfortunately, other long-range constructs such as "Nostratic" are also flawed, and this is why I don't use them. Don't forget also Starostin had a bad reputation (IMHO unjustly deserved) among linguists for his "sloppy" reconstructions.

    Lumpers think that English "much" is related with Spanish "mucho". As for them, English "day" is identical with Spanish "día". And much worser: they would try a connection between English "bad" and Persian "bad".
    I think Starostin was a "lumper" himself, as he lumped together the Tibeto-Burman and Sinitic words for 'dog', which are completely apart, because the TB one is derived from NEC but the Sinitic is not (see above).

    ReplyDelete
  21. a born becomes a dog
    Yes, I confused 'young animal' with 'small animal'. A terrible mistake! :-)

    You know, Hergé's Milou (Tintin's dog) comes from Gaulish *milo- 'small animal' (H. was an erudite himself). Of course, Latin canis must fall into the same category.

    a mole becomes a bear
    You should see these items:
    NEC *χHVr[tɕ’]V 'marten; otter'
    Altaic *karsi 'fox, marten'
    IE *H2ºrtk´o- 'bear'
    Yeniseian *χa(ʔ)s (˜ k-) 'badger'

    Don't forget we're dealing with enormous time depths, in fact much deeper than most long-range comparativists are willing to admit.

    ReplyDelete
  22. I never told Tirrhenian originated in Italy. Every language has borrowing from pre-existing languages, even that of isolated people like the Saami and the Nenets have many substratum and adstratum items.
    Then you must be aware that the Southern Italian word pentuma 'ravine' can't be Etruscan-native, as the root *pent-/*pend-/*penn- is extensively found in toponymy (e.g. Appennines, Pennines) as well as in Basque mendi 'mountain').

    ReplyDelete
  23. "Well, I've pointed out some inconsistences in your etymologies, for example as regarding as the output of affricate sibilants. At least now I know whose ones you regard as Tyrrhenian-native, so everything
    else must be borrowed."
    Methinks you didn't point up anything like this: you simply deleted some proposals of mine replacing with something else, arbitrarily, without providing any logical convincing explanation. I'm still waiting for a table of sound correspondences, explaining in detail why my proposals should be inconsistent. But that table simply doesn't exist.

    "You should also notice the number of etymologies I've actually fixed (e.g. cembro, spur, span-) is quite small."
    I think "spur-" is a loanword from Afro-Asiatic, but I refuse your
    explanations for "cembro" (the protoform of this item had a velar stop, not an affricate) and particularly the one for "span-". The root "span-" cannot contain the output of a lateral, as you suppose. There is not a single occurrence of an output like that you suppose. You fabricated it as the fruit of some unknown and unexplainable twistedness, as a tool in order to demolish my work. I think Etruscan has a lateral as the output of a NEC lateral. I don't throw in waste several important etyma because of a foolish "fixing" of yours. Shit not! There's nothing personal: your attempts of invalidate my work are for me sufficiently annoying.

    ReplyDelete
  24. About "pentuma", Etruscan is not the only Tyrrhenian language. Several other languages from proto-Tyrrhenian did exist, and many of them disappeared without leaving any written record.

    ReplyDelete
  25. "For example, the Sinitic word for 'dog' is derived from a Vasco-Caucasian root which in NEC means 'mouse, rat' and in Basque and Iberian is 'rabbit'. The singularity of the Sinitic word
    (unrelated to Tibeto-Burman) is a reflex of the early domestication of the animal in East Asia."

    As for you, ancient human beings would have domesticated rats, calling them dogs. Ridiculous.
    That's look what's Starostin's db has got.

    Proto-Sino-Caucasian: *x|Hwi>/je
    Meaning: dog

    North Caucasian: *x|_Hwe>je

    Sino-Tibetan: *qhw|i:j ( / *qhw|i:-n)
    Meaning: dog
    Chinese: *khw|i:n dog.
    Tibetan: khji dog.
    Burmese: khwijh dog, LB *khujx.
    Kachin: gui2 dog (cf. also c^@>khjon1 a fox, wolf or wild dog).
    Lushai: ui dog, KC *g|ui.
    Kiranti: *khle\

    Yenisseian: *?y?y-n (~x-,-G-,-x|-)
    Burushaski: *hu-k (? with an old dimin. suffix?)

    Is it all this so absurd? Please provide me a convincing explanation, if you are able to do it. Why the Devil ST should be the output of something else than the NC protoform?

    For more details see SINO-CAUCASIAN by S.A. Starostin, page 93

    http://starling.rinet.ru/Texts/scc.pdf

    ReplyDelete
  26. About "pentuma", Etruscan is not the only Tyrrhenian language. Several other languages from proto-Tyrrhenian did exist, and many of them disappeared without leaving any written record.
    I'm affraid you're confusing Vasco-Caucasian (i.e. NEC-related) with Tyrrhenian. Of course, Tyrrhenian wasn't the only VC language ever spoke in Europe, and Basque is still here to remind us.

    My point is that the root *pent-/*penn- is so widespread in Western Europe, as a substratum item, either fossilized in toponymy or living in Basque, Romance and Celtic, that is nearly impossible it was imported by speakers of a language which came to Italy as late as 1,200 BC.

    However, it's still possible for this root to be cognate to Etruscan penθu-na 'stone cippus' or even Greek pétra 'stone', all deriving from the same VC etymon.

    ReplyDelete
  27. As for you, ancient human beings would have domesticated rats, calling them dogs. Ridiculous.
    I'm not saying this, of course. But in a very deep chronological scale (say, 40,000 years or more), a word designating some kind of carnivore is likely to fan out into various meanings such as 'weasel', 'mouse', 'fox' and even 'dog', the latter only in the language spoken by people who domesticated the animal.

    Like other domestic animals, the words designating the 'dog' spread to other languages in the form of Wanderwörter. Unfortunately, this is forgotten too often by long-range comparitivists, which wrongly assume everything is inherited rather than borrowed.

    This is why I regard long-range constructs such as "Nostratic" as intrisincally flawed.

    Is it all this so absurd? Please provide me a convincing explanation, if you are able to do it. Why the Devil ST should be the output of something else than the NC protoform?
    I'm afraid Starostin's ST is actually a grouping of two different families: Tibeto-Burman and Sinitic (this is somehwat similar to having NEC and NWC lumped together as NC).

    If you look carefully, you'll see the Sinitic word has an extra /n/ not found in Tibeto-Burman. In other words, Starostin is cheating with linguistic data.

    Thus we've got two different words for 'dog' in ST, one in TB and another in Sinitic. While the first is related to the Caucasian one, the second one is almost identical to the PIE one, up to the point Gamkrelidze & Ivanov suggested it was an IE loanword. But archaeology tell us just the contrary, as East Asia was one of the places where dogs were first domesticated.

    My point is the Sinitic word is a distant cognate of Yeniseian *ku:n´ (~ g-) 'wolverine' and NEC *ɦnǝ:q'q'wǝ: (~ *ɦq'q'wǝ:nǝ:) 'mouse, rat'.

    ReplyDelete
  28. I'm still waiting for a table of sound correspondences, explaining in detail why my proposals should be inconsistent. But that table simply doesn't exist.
    I hope we'll come to that in the future, when we've got enough good etymologies.

    but I refuse your explanations for "cembro" (the protoform of this item had a velar stop, not an affricate)
    I regard your etymology as inadequate, among other things because we would expect *n instead of m. This why I've replaced it with PSC *xq'wémV 'nut; kernel', which I think is the source of Greek kônos 'cone'. Your etymology would give Greek kôna 'pitch' instead.

    As regarding the other words with k- (corme, kómaros), I see them as unavoidable linked to sorbus, so either we've got to posit centumization from *ts or satemization from *k. As the latter looks to be more plausible, I've changed my previous etymology and linked these words to Basque gurbitz and Italian corbezzolo.

    The root "span-" cannot contain the output of a lateral, as you suppose. There is not a single occurrence of an output like that you suppose.
    I'm afraid this can be easily disprovable. As you might already know, in Semitic languages the output of PSem voiceless lateral fricative is a sibilant s. This is precisely what we've got in spur.

    There's nothing personal: your attempts of invalidate my work are for me sufficiently annoying.
    Nothing personal, but as the New Testament says, it's always easier to pinpoint other people's mistakes than oneself's.

    ReplyDelete
  29. Methinks your posts and comments are changing, as every time I enter this blog I find something different. Am I suffering from Alzheimer's disease? I don't think that.
    You say: "I regard your etymology as inadequate, among other things because we would expect *n instead of m. This why I've replaced it with PSC *xq'wémV 'nut; kernel', which I think is the source of Greek kônos 'cone'. Your etymology would give Greek kôna 'pitch' instead."
    And more: "Interestingly enough, we've got cognates in dialectal (High Navarrese) Basque gurbi (HN) 'service tree' and gurbe 'wild apple (Malus sylvestris)'. These words are strinkingly similar to gurbitz 'strawberry tree (Arbutus unedo)', found in Western dialects and cognate to Italian corbezzolo 'strawberry tree', and for which I propose a Vasco-Caucasian etymology from NEC *kkurmæɕV / *kkurmæʒV 'a k. of fruit'3."

    But the previous version of this post had something very different, reconstructing a protoform with an affricate, something like *dzumV-rV. I agree with you that PSC *xq'wémV 'nut; kernel' is better than my former etymology of "cembro", but the reason of my anger was the supposed centumization of *dzumV-rV. I didn't see SC *xq'wémV in the list and nor I saw it here. Perhaps you have understood I'm not prone to accept centumization and satemization, so you have changed the text of the post. It's very difficult to discuss in similar conditions.

    You say about "span": "I'm afraid this can be easily disprovable. As you might already know, in Semitic languages the output of PSem voiceless lateral fricative *ɬ is a sibilant s. This is precisely what we've got in spur."
    Etruscan is not a Semitic language, and the shift you imagine is not found in native items.

    I make several mistakes, of course, but brainstorming is not the cure. I don't attribute my mistakes to you. Instead, you are changing even the past of this website. This instability troubles me very much: where it's impossible to find something stable, no real knowledge could be developed.

    New Testament says also: "A tree is recognized by its own fruits".

    I repete it to you: it's better that you feel free to paint your own picture, as I want to be free to paint mine.

    ReplyDelete
  30. "If you look carefully, you'll see the Sinitic word has an extra /n/ not found in Tibeto-Burman. In other words, Starostin is cheating with linguistic data."
    It is reasonable to suppose the extra /n/ disappeared. If this was not the case, it would be easier to suppose the item is a loanword from some unknown language (as you admit, these words are often Wanderwoerter). Can you provide more statistical data about those languages? Before supposing a relationship of a word for "dog" in some languages with a protoform meaning "rat", I would be more careful. Is semantics so irrelevant?

    ReplyDelete
  31. This instability troubles me very much: where it's impossible to find something stable, no real knowledge could be developed.
    When an idea isn't mature enough in my mined, it's prone to be revised (sometimes several times), although this shouldn't confused with brainstorming.

    But the previous version of this post had something very different, reconstructing a protoform with an affricate, something like *dzumV-rV.
    Yes, and also the metathesized form *dzurVmV.

    I agree with you that PSC *xq'wémV 'nut; kernel' is better than my former etymology of "cembro"
    From a theoretical point of view, perhaps NEC *dʒɦumV could be the result of satemization of an older root such as *xq'wémV, as we can't simply assume all the items of a proto-language are inherited and equally old.

    Perhaps you have understood I'm not prone to accept centumization and satemization, so you have changed the text of the post. It's very difficult to discuss in similar conditions.
    Actually, one of your own etymologies is a good example of the latter: Etruscan zip(an)- 'comb' ~ Altaic *tʃhipV 'sharp edge, peg' ~ Latin cippus. Here Altaic has undergone satemization from *k´i (PAltaic has no dorso-palatals whatsoever).

    Of course, I agree these sound shifts never happened in native Tyrrhenian words, but I don't understand your insistence that *kimra must be so.

    Etruscan is not a Semitic language, and the shift you imagine is not found in native items.
    Of course not, but your native etymology doesn't look very good either. This is why I prefer to consider it a loanword in the same way than spur-.

    New Testament says also: "A tree is recognized by its own fruits".
    But trees can be grafted, you know. :-)

    I'm going to continue the discussion about dogs and domestic animals here.

    ReplyDelete
  32. "When an idea isn't mature enough in my mined, it's prone to be revised (sometimes several times), although this shouldn't confused with brainstorming."
    Ok, I can understand this, as several ideas of mine are still immature and others were dismissed as well. But the turnover should not be excessive, or any kind of control would be lost.

    About satemization:
    "Actually, one of your own etymologies is a good example of the latter: Etruscan zip(an)- 'comb' ~ Altaic *tʃhipV 'sharp edge, peg' ~ Latin cippus. Here Altaic has undergone satemization from *k´i (PAltaic has no dorso- palatals whatsoever)."
    This was one of the etymologies I found inadequated. The Nostratic
    forms are very weak. I think the SC protoform meaning 'tooth' would be better: I found it has a *s/t- that would give *z- in Etruscan, and a plural fossilized suffix *-bV is rather common in many NEC languages. When I knew only NEC data, I considered the Starostin's NEC reconstruction not fitting with the Etruscan data, as it has *c- (that would give Etruscan s-).
    http://starling.rinet.ru/cgi-bin/response.cgi?single=1&basename=/data/sinocauc/sccet&text_number=+522&root=config

    "Of course, I agree these sound shifts never happened in native
    Tyrrhenian words, but I don't understand your insistence that *kimra must be so.


    This *kimra (or better *khimra) is a Rhaetian substrate item. This is the reason I consider it a genuine Tyrrhenian word. Several Alpine dialects point to a protoform *gimra.

    Etruscan s'pan- is a very difficult item. The most convincing proposal, as for me is still the SC root whose meaning was reconstructed by Starostin as 'span'. I prefer to attribute it a more general meaning: 'flat, plain'.
    In the Tabula Cortonensis we see s'pan-thi 'in the plain', a quite general meaning fitting the Latin adjective pla:nus and its derivatives. It is rather improbable to consider 'flat stone' as the central meaning of the Tyrrhenian protoform.

    ReplyDelete
  33. About the word meaning 'beer', we should not forget the gloss CEREA, quite probably a Lusitanian word, as it lost -w- (like OILA 'sheep' < *owi-la).

    ReplyDelete
  34. This was one of the etymologies I found inadequated. The Nostratic forms are very weak.
    On the contrary, it's "Nostratic" itself which is dead. IMHO this entity is nothing but a bunch of more or less distantly related languages with a shared stock of Wanderwörter. And some of its proposed members (e.g. Dravidian) would even belong to the Vasco-Caucasian phylum.

    I think the SC protoform meaning 'tooth' would be better: I found it has a *s/t- that would give *z- in Etruscan, and a plural fossilized suffix *-bV is rather common in many NEC languages.
    But semantics is more problematic, not to speak of phonetics (a plural is unjustified IMHO). Therefore I prefer the former non-native etymology.

    This *kimra (or better *khimra) is a Rhaetian substrate item. This is the reason I consider it a genuine Tyrrhenian word. Several Alpine dialects point to a protoform *gimra.
    I still consider as likely a relationship with Greek kómaros and other words designating fruits similar to little apples. This is why I no longer see as adequate my last proposal from SC *xq'wémV and I've replaced it with NEC *ʕæmtɕo: 'apple; medlar' with rhotacism of the sibilant.

    Etruscan s'pan- is a very difficult item. The most convincing proposal, as for me is still the SC root whose meaning was reconstructed by Starostin as 'span'. I prefer to attribute it a more general meaning: 'flat, plain'.
    But 'span' refers to a length mesure, which is quite different from 'broad' > 'flat'. The only solution to this semantical gap is to assume a semantic shift from 'span between the thumb and the small finger' > 'palm of the hand' > 'flat'. But IMHO there're still phonetic difficulties, as (unless clustered to a stop) NEC *m should give Etruscan m.

    It is rather improbable to consider 'flat stone' as the central meaning of the Tyrrhenian protoform.
    Surely this is the meaning in NEC, but we've got 'flat, slice' in Sino-Tibetan and 'flat, broad' in Altaic, for example. Sound correspondences suggest Latin tabula could reflect the native Tyrrhenian item corresponding to this etymology.

    About the word meaning 'beer', we should not forget the gloss CEREA, quite probably a Lusitanian word, as it lost -w- (like OILA 'sheep' < *owi-la).
    I'll deal with that in a forthcoming post.

    ReplyDelete
  35. From other etymologies, I assume the NEC fricative would be reflected in Tyrrhenian as a velar stop, while in another language it would give a sibilant as predicted by Fournet's Law, a special case of satemization of post-velar fricatives.

    ReplyDelete
  36. "On the contrary, it's "Nostratic" itself which is dead. IMHO this entity is nothing but a bunch of more or less distantly related languages with a shared stock of Wanderwörter. And some of its proposed members (e.g. Dravidian) would even belong to the Vasco-Caucasian phylum."
    I agree, Nostratic is not well founded, including not only Wanderwörter, but also a huge plenty of substratum item. E.g. the root *k.arV- 'rock' is reconstructed from non-IE substratum items and has no reliability at all. Almost the half of Starostin's Nostratic db is composed of items found only in Altaic and in Uralic, sometimes also in Dravidian. This roots often seems to be derived from SC (e.g. ).
    In any case I'm skeptikal about your theory of Dravidian as a VC language. There's no much similarity. If you really think Dravidian and VC are related, you are invited to provide some further prove, in order to discuss evidences, or I'll consider this an arbitrary and erratic theory.

    About 'comb' and 'tooth':
    "But semantics is more problematic, not to speak of phonetics (a plural is unjustified IMHO). Therefore I prefer the former non-native etymology."
    The plural is not unjustified, as it is found in several NEC languages. There is some evidence that a root *g'ombho- 'comb' found in Germanic and in Slavic once meant 'tooth'.
    http://starling.rinet.ru/cgi-bin/response.cgi?single=1&basename=/data/ie/piet&text_number=+++315&root=config

    About *khimra:
    "I still consider as likely a relationship with Greek kómaros and other words designating fruits similar to little apples. This is why I no longer see as adequate my last proposal from SC *xq'wémV and
    I've replaced it with NEC *ʕæmtɕo: 'apple; medlar' with rhotacism of the sibilant."
    There is no rhotacism of sibilants in Tyrrhenian. The NEC root you quote is instead the ancestor of Greek 'mespilon'.

    "But 'span' refers to a length mesure, which is quite different from 'broad' > 'flat'. The only solution to this semantical gap is to assume a semantic shift from 'span between the thumb and the small finger' > 'palm of the hand' > 'flat'. But IMHO there're still phonetic difficulties, as (unless clustered to a stop) NEC *m should
    give Etruscan m."

    NEC *-m- cannot give Etruscan -p-, of course, but here we have not NEC *-m-. We have NEC *-mh_-, whose reflexes in several NEC
    languages show a loss of nasality and a labial stop. Any single output of NEC is to be carefully analyzed, not only the protoform. This etymon is not so problematic, after all. Whe have also some sparse words in several western IE languages, such as English 'spoon', whose reconstructed root is *sp(h)e:-, that may also belong
    here. The original meaning coud be something like 'long piece of wood', 'large and flat piece of wood', probably from some Substratum item related to SC and once meaning 'flat'.

    "Surely this is the meaning in NEC, but we've got 'flat, slice' in Sino-Tibetan and 'flat, broad' in Altaic, for example. Sound correspondences suggest Latin tabula could reflect the native Tyrrhenian item corresponding to this etymology."
    Tyrrhenian items don't develop a dental stop for a lateral NEC sound in initial position.

    "From other etymologies, I assume the NEC fricative *? would be reflected in Tyrrhenian as a velar stop, while in another language it would give a sibilant as predicted by Fournet's Law, a special case of satemization of post-velar fricatives."
    This hypothetical Fournet's Law you are speaking about, is not found at all in Tyrrhenian, and it is very dubious it even existed elsewhere. You are continuing to speak about this item? Well, you have to provide a complete list of roots. The few I saw in the Yahoo
    list seem to be all but convincing.

    ReplyDelete
  37. I agree, Nostratic is not well founded, including not only Wanderwörter, but also a huge plenty of substratum item.
    Yes, of course. All them are wrongly considered as inherited items.

    In any case I'm skeptikal about your theory of Dravidian as a VC language. There's no much similarity. If you really think Dravidian and VC are related, you are invited to provide some further prove, in order to discuss evidences, or I'll consider this an arbitrary and erratic theory.
    This isn't a theory, but a mere working hypothesis, based on some correspondences I've myself discovered when studying names of carnivores: NEC *gwe:rdV 'a best or bird of prey' ~ South Dravidian *karaɖ-i 'black bear', NEC *tenttɬ’wV 'weasel' ~ Dravidian *to:nɖl- 'wolf, jackal', NEC *ɦq’q’wǝ:nǝ: 'mouse, rat' ~ Dravidian *nakk- 'jackal'. Please notice that semantic shifts of this kind indicate usually indicate a different physical environment.

    However, this impression could be false and the actual relationship more remote.

    There is some evidence that a root *g'ombho- 'comb' found in Germanic and in Slavic once meant 'tooth'.
    I think 'protunding thing' or similar would be more apropriate. Anyway, I can't still see any convincing reason for dropping your former etymology in favour of the new one.

    There is no rhotacism of sibilants in Tyrrhenian. The NEC root you quote is instead the ancestor of Greek 'mespilon'.
    I guess this *-pilon (Etruscan *-pile) means 'fruit'. I leave the matter open to further research.

    We have NEC *-mh_-, whose reflexes in several NEC languages show a loss of nasality and a labial stop.
    This sounds reasonable enough.

    This etymon is not so problematic, after all. Whe have also some sparse words in several western IE languages, such as English 'spoon', whose reconstructed root is *sp(h)e:-, that may also belong here. The original meaning coud be something like 'long piece of wood', 'large and flat piece of wood', probably from some Substratum item related to SC and once meaning 'flat'.
    I'm afraid these 'spoon' words point to a root whose meaning was 'pointed thing' (e.g. spear), whereas 'flat' corresponds to 'spade'. Notice, however, that Starostin quotes the Kartvelian forms lap’ot’- 'large wooden spoon', laper- 'broad flat wooden spoon' as related to the root *lVpV I propose as the source etymon.

    Tyrrhenian items don't develop a dental stop for a lateral NEC sound in initial position.
    Of course not in native items, but loanwords could be different.

    ReplyDelete
  38. NEC *ɦq’q’wǝ:nǝ: 'mouse, rat' ~ Dravidian *nakk- 'jackal'.
    Of course, the Dravidian form would correspond to the metathesized NEC form *ɦnǝ:q’q’wǝ:, already reconstructed by Starostin.

    Just by chance, this afternoon I've found the German names for the Swiss pine (Pinus cembra). This has leaded me to rethink the matter so I've restated my former proposal from PSC *x’qwémV, which IMHO is able to explain ALL the forms.

    Also notice that the palatalization (i.e. satemization) of the initial stop actually happened in Altaic and Kartvelian.

    ReplyDelete
  39. I've reposted here this comment, as it belongs to this thread.

    It is reasonable to suppose the extra /n/ disappeared.
    In that case, the TB and NEC word would have been borrowed from Sinitic, but I'm not sure of that because the former has other possible cognates.

    If this was not the case, it would be easier to suppose the item is a loanword from some unknown language (as you admit, these words are often Wanderwoerter).
    Not really. I regard the Sinitic word as native because archaeology tell us dogs were domesticated in East Asia. Only afterwards it became a Wanderwort adopted by other languages such as PIE *k´(u)wo:n.

    Can you provide more statistical data about those languages?
    I've not studied in detail the reconstructed ST lexicon, but I think linguistic data shouldn't be tailored to fit pre-conceived family trees as Starostin did, but rather on the contrary.

    Before supposing a relationship of a word for "dog" in some languages with a protoform meaning "rat", I would be more careful. Is semantics so irrelevant?
    Of course not, because the type of animal involved is relevant, so for example 'dog' correlates to 'small mammal' or 'carnivore', 'horse' to 'ungulate', and so on.

    My point is the domestication event caused a major semantic shift in the language spoken by domesticators. This is what it's unlikely we could find genuine cognates (i.e. words inherited from a common ancestor) in names of domestic animals among not closely related languages, because in most cases the date of domestication will be more recent than the split of these languages.

    If for example a long-range comparativist links a word meaning 'dog' to another word with the same meaning in a distantly related language, then at least one of them must be a Wanderwort.

    The consequence of mistaking these and other Wanderwörter as inherited items is that the chronologies given for long-range entities such as "Nostratic" go no deeper than the Mesolithic.

    However, as human prehistory goes far beyond than that, the actual chronologies of distant language relationships must be older, sometimes much older. So if 'dog' in a language X is a genuine cognate of 'rat' in a language Y, this means its common ancestor must have been spoken very long ago, perhaps 40,000 years or even earlier.

    ReplyDelete