Newest Addition to the Indo-European Family: Burushaski

[WeepingElf]

Mitian looks like a plausible grouping, yes. Within Mitian, there appears to be a fairly clear division between a "Euro-Siberian" group (including IE, Uralic, Yukaghir, Chukotko-Kamchatkan, and Eskimo-Aleut) and an "Altaic" group (including Turkic, Mongolic, and Tungusic). The latter one might not be a single node though; I know too little about those languages to say anything substantial here.

Within Euro-Siberian, I consider Fortescue's Uralo-Siberian grouping more likely than Indo-Uralic. A major point in favor of this is that the reconstructed phoneme inventories of the four Uralo-Siberian groups are very similar to each other, especially in the consonants (only one series of plosives, plus a series of voiced fricatives; as opposed to PIE's three plosive series but no voiced fricatives).

I am undecided about Indo-Uralic vs. Uralo-Siberian. The Uralic phonology indeed looks more similar to the other Uralo-Siberian languages than to Indo-European, but I think we cannot yet exclude the possibility that IE has innovated (perhaps under a Caucasian areal influence) and the Uralo-Siberian phonological pattern is a shared retention rather than a shared innovation. The sound correspondences have not yet been sufficiently cleared up. The Indo-Uralic accusative *-m may be a shared innovation, but this too could be a shared retention lost in the eastern languages. The migration patterns would be less complicated if Indo-Uralic was a valid group. One would have one group (Indo-Uralic) moving west and one (Eskimo-Siberian) moving east from the original homeland (IMHO probably somewhere west of Lake Baykal), while the Uralo-Siberian hypothesis would require IE and Uralic moving west independently from each other (not that westward movements were rare in southwestern Siberia, though).

You may want to join the Nostratic-L mailing list at Yahoogroups; Tropylium (John Vertical) and I are already there. Now that Arnaud Fournet has left it, constructive discussion of such matters is possible there again.

What regards Nostratic: I wouldn't say it was "bullsh*t", but I consider it unlikely that Afroasiatic and Dravidian have anything to do with the Mitian group. It is a long way from southwestern Siberia either to the upper Nile or the Indus, and the languages look vastly different.

[Soap]

Now that Arnaud Fournet has left it, constructive discussion of such matters is possible there again.

Lol, I guess he was somewhat of a crackpot? He seems to be a published author, though ...

[WeepingElf]

Now that Arnaud Fournet has left it, constructive discussion of such matters is possible there again.

Lol, I guess he was somewhat of a crackpot? He seems to be a published author, though ...

Yes, I have heard of him being a published author, but his behaviour in Nostratic-L was that of a crackpot, quite similar in fact to Octaviano (whom he calls the "Lord of All Trolls"). He strongly believes that Hurrian-Urartean was closely related to Indo-European (a case that looks very much like the Burushaski case discussed here), and doesn't admit any criticism of that hypothesis. He had written a monograph about that hypothesis together with Allan Bomhard, but Bomhard later distanced himself from that work because he realized that it was wrongheaded. Fournet also is very dismissive of Mitian (which he nicknamed "Mythian") and Indo-Uralic, and believes in something he calls "Ural-Altaic" even though he excludes Turkic and Manchu (but not the rest of Tungusic!) and includes Na-Dene. He called everybody else in the list names and thereby alienated them until he finally left the list.

[Bob Johnson]

Mitian

is anybody even working on the altaic-korean connection? or altaic-IE. (i'm rather pessimistic about a korean-japanese merger, pitchfork-wielding mobs interfere with data collection)

and less seriously, where would Basque fit

[TaylorS]

Mitian looks like a plausible grouping, yes. Within Mitian, there appears to be a fairly clear division between a "Euro-Siberian" group (including IE, Uralic, Yukaghir, Chukotko-Kamchatkan, and Eskimo-Aleut) and an "Altaic" group (including Turkic, Mongolic, and Tungusic). The latter one might not be a single node though; I know too little about those languages to say anything substantial here.

Within Euro-Siberian, I consider Fortescue's Uralo-Siberian grouping more likely than Indo-Uralic. A major point in favor of this is that the reconstructed phoneme inventories of the four Uralo-Siberian groups are very similar to each other, especially in the consonants (only one series of plosives, plus a series of voiced fricatives; as opposed to PIE's three plosive series but no voiced fricatives).

So if Mitian is indeed a valid language family, I would imagine its internal structure to be something like this:

Code: Select all

Mitian
- Core Mitian
-- Euro-Siberian
--- Europic
---- Indo-European
---- (probably other languages in Europe; e.g. the language of the Old European Hydronymy)
--- Uralo-Siberian
---- Uralic
---- Yukaghir
---- Beringian
----- Chukotko-Kamchatkan
----- Eskimo-Aleut
-- Altaic (doubtful as a single node)
--- Turkic
--- Mongolic
--- Tungusic
- (Tyrrhenian [Etruscan & related languages]; doubtful)
- (Kartvelian; doubtful)
- (Korean-Japanese; very doubtful)

I speculate something like this:

Mitian
- Kartvelian
- Core Mitian
-- Euro-Siberian
--- Europic
---- Indo-European
---- Etruscan (see Glen Gordon's blog Paleoglot)
---- (probably other languages in Europe; e.g. the language of the Old European Hydronymy)
--- Boreal
---- Uralo-Siberian
----- Uralic
----- Yukaghir
---- Beringian
----- Chukotko-Kamchatkan
----- Eskimo-Aleut
-- Altaic (doubtful as a single node)
--- Turkic
--- Mongolic
--- Tungusic
- (Korean-Japanese; very doubtful)

[TaylorS]

Now that Arnaud Fournet has left it, constructive discussion of such matters is possible there again.

Lol, I guess he was somewhat of a crackpot? He seems to be a published author, though ...

Yes, I have heard of him being a published author, but his behaviour in Nostratic-L was that of a crackpot, quite similar in fact to Octaviano (whom he calls the "Lord of All Trolls"). He strongly believes that Hurrian-Urartean was closely related to Indo-European (a case that looks very much like the Burushaski case discussed here), and doesn't admit any criticism of that hypothesis. He had written a monograph about that hypothesis together with Allan Bomhard, but Bomhard later distanced himself from that work because he realized that it was wrongheaded. Fournet also is very dismissive of Mitian (which he nicknamed "Mythian") and Indo-Uralic, and believes in something he calls "Ural-Altaic" even though he excludes Turkic and Manchu (but not the rest of Tungusic!) and includes Na-Dene. He called everybody else in the list names and thereby alienated them until he finally left the list.

LOL, I joined Nostratic-L a while back, but I left soon after because of all the crackpots who think Starostin could do no wrong, despite his baby, "North Caucasian", being completely, utterly implausible.

[TaylorS]

Mitian

is anybody even working on the altaic-korean connection? or altaic-IE. (i'm rather pessimistic about a korean-japanese merger, pitchfork-wielding mobs interfere with data collection)

and less seriously, where would Basque fit

No serious people have a clue where Basque fits. Some crackpots put it in a mythical "Dene-Caucasian" grouping.

I don't see anything that would preclude Basque being related to the NW or NE Caucasian languages, it does not contradict the archeological evidence (the spread of the Cardial Ware culture bringing agriculture across the Mediterranean).

One problem is that Basque's Y Chromosome DNA are almost entirely all of the indigenous Western European haplotype R1b, one would expect a substantial number of Basques with haplotypes J2 and Eb3 if Basque were spread by farmers from Anatolia.

[Herr Dunkel]

It is my belief that such a family is far too hard to prove - instead, I've been a proponent of a genetic link between Indo-European and Uralo-Siberian (Uralic, Yukaghir and possibly Eskimo-Aleut) - the time gap between IE and Uralic is a bit deeper than the gap between Uralic and Yukaghir.

I'll join the Nostratic-L group.

[Drydic]

No serious people have a clue where Basque fits. Some crackpots put it in a mythical "Dene-Caucasian" grouping.

Now you're just parroting Trask.

[WeepingElf]

Mitian

is anybody even working on the altaic-korean connection? or altaic-IE. (i'm rather pessimistic about a korean-japanese merger, pitchfork-wielding mobs interfere with data collection)

and less seriously, where would Basque fit

No serious people have a clue where Basque fits. Some crackpots put it in a mythical "Dene-Caucasian" grouping.

I wouldn't call the Dene-Caucasianists "crackpots", even though I believe that the evidence is far too tenuous and the group seems more or less to be defined as "all languages of northern Eurasia that are not claimed to be Nostratic, plus Na-Dene and Sino-Tibetan". When you are venturing into uncharted territory yourself, you ought to be careful with words such as "crackpot" or "bullshit". (Yes, I did say yesterday that Arnaud Fournet behaved like a crackpot, but that was not because of his hypotheses, but because of his social behaviour on Nostratic-L.)

I don't see anything that would preclude Basque being related to the NW or NE Caucasian languages, it does not contradict the archeological evidence (the spread of the Cardial Ware culture bringing agriculture across the Mediterranean).

Maybe it is just a language left over from the first peopling of Europe by Homo sapiens 40,000 years ago, and any relationship to NWC or NEC dates back to that time; maybe it arrived later, perhaps in the Neolithic. There currently is no way knowing, I fear.

One problem is that Basque's Y Chromosome DNA are almost entirely all of the indigenous Western European haplotype R1b, one would expect a substantial number of Basques with haplotypes J2 and Eb3 if Basque were spread by farmers from Anatolia.

One has to be careful with combining genetic data with linguistic matters. Language shift (the adoption of a new, more prestigious language by a population) is common enough to invalidate such equations.

[Salmoneus]

Well if it is related, it surely must hold the record for smallest amount of cognates. I havent seen his essay, and I'm only basing my opinion on one source (Wikipedia), which admittedly tells me very little ,but I can't see how this language could possibly go back to PIE, even to a branch that is itself little known.

To qualify this, I was just reading an excellent etymological dictionary of latin. In terms of layout and whatnot, it's wonderful - for instance, it includes an appendix not only of all the latin words it talks about, not only about all the PIE words it talks about, but of all the cognates in other languages it talks about.

There are 40 pages of Latin words
There are 10 pages of Faliscan, Oscan, Umbrian, South Picene, Marsian, Volscian, Pre-Samnite, Marrucinian, Hernician, Vestinian, and Paelignian (altogether, not each)
There are 6 pages of Celtic words
There are 2 pages of Anatolian words
There are 10 pages of Indo-Iranian words
There are 6 pages of Greek words
There are 5 pages of Baltic words
There are 4 pages of Slavic words
There are 8 pages of Germanic words
There are 2 pages of Tocharian words
There is a page of Armenian words
There is almost a page of Albanian words
There are... 6 phrygian words. [anar (man), ab-baret (he carries), ad-daket (he makes, showing the -k- that has also been seen in Latin and Greek - middle voice is ad-daketor, if you're interested), bratere (brothers), onoman (name), and 'otuvoi vetei' (in the eight year)]

I know it's not a complete list of all recorded words, but it does show both the level of scholarly knowledge and the interestingness (or otherwise) of the language relative to the other IE languages.

[Salmoneus]

It's a predisposition presumed by TaylorS to exist based on the ancestry of the professor. TaylorS's dismissive argument clearly implies that, being Macedonian, the professor has a predisposition toward the improvement/betterment/etc. of Macedonia, even if it means manipulating or selectively ignoring facts in order to show Macedonia's connection to a minority language in Pakistan (why this would be beneficial for Macedonia, I don't know...to prove somehow that Alexander the Great was Macedonian and not Greek?) Not only, however, has this presumed predisposition not been proven to exist in even the slightest degree, even if it did, being a Macedonia-booster does not, as I said, automatically falsify any claim he makes regarding Macedonia/Burushaski/etc.. It is a point to keep in mind, but not by any means a deciding factor.

When you see a theory that looks crackpot, you have to think 'surely this isn't crackpot - most linguists don't derive crackpot theories!'. However, there is a clear tendency for balkan linguists to be more likely to derive crackpot theories, when those theories give an previously unknown importance to their own nation. So, knowing that this theory is by a balkan linguist and that it gives a previously unknown importance to his own nation makes it easier to accept that this is just a crackpot theory, rather than assuming that there must be something to it because why else would an academic linguist come up with such a bizarre theory if it wasn't true.

Of course, it may ALSO be true. But until we've seen evidence of this, I think it's fair to take the suggestion of alternative motivation as prima facie support for the hypothesis that it might not be true.

[clawgrip]

Don't get me wrong, here. I'm not defending the IE theory. I don't know enough to make an informed opinion either way, although from what I can tell, it sounds dubious. I'm only saying that the professor being Macedonian is not grounds to "dismiss it out of hand." It's a warning sign, but not conclusive evidence.

[Herr Dunkel]

Don't get me wrong, here. I'm not defending the IE theory. I don't know enough to make an informed opinion either way, although from what I can tell, it sounds dubious. I'm only saying that the professor being Macedonian is not grounds to "dismiss it out of hand." It's a warning sign, but not conclusive evidence.

I can say that any Dixie linguist, if he had a chance to say that Dixish is a seperate language from English, would grab it and make shit up to prove it.
This, no different. Certainly, no grounds for instant dismissal, but dubious.

[clawgrip]

Exactly what I am saying. I was just correcting a fallacious argument. It's important to remember that basing a conclusion on fallacious reasoning does not immediately invalidate that conclusion.

[Herr Dunkel]

Certain fallacies only apply for certain arguments.
Associative reasoning is not a fallacy by itself, but an associative dismissal is frequently a fallacy (not always, though)

[Tropylium]

From what I've seen so far (which is just the blurbs…) it sounds like the evidence is on a level where a bunch of words *can* be derived from PIE — but perhaps so few that one could equally well pick a slightly different set and derive it from Semitic or Turkic or something (as noted, if there were tons of cognates, it would have been noticed by now). Looking up one of the examples named in that radio show transcript, "brow" is /ˈpʰenas/ ~ /ˈpʰinis/ which doesn't particularly jump out as IE. Obviously *bʰ > pʰ would provide a nice match for the stop system at least.

But perhaps there is indeed a case to be made. We'll see.

Also whether Burushaski descends from Phrygian is a different question from whether it descends from PIE; it's possible that it is Indo-European but Phrygian was only dragged in as a case of Balkan nationalism.

(As for other macro-comparision issues: I agree that IE/Uralic and U/EA are promising, but we either need to be able to formulate Indo-Eskimo too or leave either of these as contact infuence. Uralo-Yukaghir sinks from "likely" to "not that good" once a number of loans are weeded out. Altaic is just not working out as is. All of the bigger multilateral bundles are messes and will not progress until we have a good grasp of the prehistory of at least one constituent family. And it's always possible something unexpected like Kartvelian-Hmong-Mien or Korean-Haida is already in the works, too.)

[Herr Dunkel]

I'd say that fourty cognates between IE and Uralic, and fourty-two between Uralic and Yukaghir is pretty neat for around eight thousand years of divergences. I don't know how many cognates exist between IE and Yukaghir, though, but I'm betting it's under ten or fifteen (and almost every single one should also be shared with Uralic)

[Tropylium]

The latest count I've seen is 30-ish U/Y comparanda for which a loaning argument isn't available; they don't match particularly well match with the IE/U list either. You'll find a discussion about this from the beginning of the month on Nostratic-L…

[Drydic]

Why the hell are you people writing in size whatever font, it hurts my eyes, stop it

[Herr Dunkel]

Because we're plotting.

I imagine it' because Yukaghir and Uralic have a closer relationship than Indoeuropean and Uralic.

[WeepingElf]

Because we're plotting.

I imagine it' because Yukaghir and Uralic have a closer relationship than Indoeuropean and Uralic.

It is too early to tell - the matter still requires more work. From what I have seen of Yukaghir (I have Maslova's grammar of Tundra Y here), it doesn't really look like a divergent branch of Uralic, and Collinder's claim that the morphology was more or less the same as that of Samoyedic is a vast overstatement. Sure, some resemblances are there, but they are IMHO not stronger than those between Uralic and IE. Hence, I consider the Uralic-Yukaghir relationship to be overestimated. Possible, yes, plausible, even that, but not necessarily closer than that between Uralic and IE.

[Herr Dunkel]

Then a Proto-Indo-Uralo-Yukaghir, with the three branches splitting off at relatively the same time?

I admit, this is mostly guesswork as I understand shit Russian whilst using a dictionary to translate - I can't really see what gears're inside Yukaghir's machinery.

Although, if the simmilarity between Yukaghir and Uralic is on the same level as Uralic-IE, what is the level of simmilarity of Yukaghir to IE?

[Pabappa]

Presumably they'd be about the same distance if they separated at the same time

[Herr Dunkel]

Presumably and under perfect conditions.
You are forgetting about language contact, substrates and bilingualism.
I expect there to be less simmilarities between IE and Yukaghir than those between the two and Uralic if they seperated at the same time - Uralic is middle ground for both - they would've acquired Uralic characteristics easily whilst not acquiring each other's characteristics.
The layout is (PIE <==> PU <==>PY) and, looking at this, I expect that most (not all) of the characteristics IE and Yukaghir share they also share with Uralic.
And that's not even considering they borrowed the exact same stuff.