Vascano-Turkic?

[Shemtov]

So for various reasons (partly because of curiosity, partly because of religious teachings relating to using languages not spoken a lot [at all, in this case, I think] by my coreligionists) I'm teaching myself Basque, and I'm noticing some similarities between it and the Turkic languages. I don't think an actual linguist has proposed a connection (though that may be caused by "Altaic Theory", as the correspondences sort of disappear if you say Old Turkic descended from Proto-Altaic), bu they are still interesting:
- No Basque word begins in a Rhotic. No native Turkic word begin in a rhotic.
-Turkic languages are SOV. Basque has some SOV tendencies.
-Proto-Turkic Directive case ending: -/ʁaru/, -/ɡery/. Basque Allative Case-Ending -(r)a. Both end in /rV/
-Proto-Turkic Locative -/ta/, -/da/, -/te/, -/de/; Basque Locative Genitive Form 2 <etako>. Both contain sequence /ta/.
-Proto-Turkic Genitive -/ŋ/; Basque Genitive Form 2 -en. Both end in a Nasal.
-Proto-Turkic 2P Pronoun /se/; Basque 2P Plural Pronoun <Zu>. Both consist of sibilant-vowel sequence.
-Proto-Turkic 3P pronoun /o/; Basque demonstrative <Hau>.
-Proto-Turkish Numeral one /bir/; Ancient Basque Numeral one *bade. (It seems the Proto-Turkic Rhotic was /ɾ/)
-Proto-Turkish Numer two /iki/; Ancient Basque Numeral two *biga.
-Proto-Turkish Numeral Four tört; Ancient Basque Numeral four *laur (Both end in rounded vowel-rhotic sequence; both begin with coronal.)
-Proto-Turkish Numeral Five besh; Ancient Basque Numeral five *bortz (begin with /b/; end in sibilant.)
-Turkish <kurt> "wolf"; Basque <txakur> "Dog" (Both contain /kur/)
-Turkish <bükme> "Bending" Basque <bihurgune> "Curve"
-Turkish <ağaç>; Basque <zuhaitz> "tree"
-Turkish <ısı> "heat"; Basque <su> "fire"
-Turkish <yem-> "eat"; Basque <jan> "to eat"
Now, I'm not seriously saying they are related, just there's enough "coincidences" in morphology, pronouns, numerals, and certain Swadesh list words, that I think further investigation would be worthwhile, even if it turns out to be trash right away.

[Vijay]

You can find all sorts of linguistic coincidences, even ones that are probably much closer than this, but that's not how we can tell whether two languages are related. There have to be sound correspondences that hold up consistently across these languages, not just random words and affixes.

[Hallow XIII]

sergei, is that you

[zompist]

In general the answer to "Has anyone tried to link Basque and ____" is yes.

[Nortaneous]

i was going to make a joke about vasco-sinitic, but this is what starostin actually believes

[Shemtov]

You can find all sorts of linguistic coincidences, even ones that are probably much closer than this, but that's not how we can tell whether two languages are related. There have to be sound correspondences that hold up consistently across these languages, not just random words and affixes.

I know that, but I'm suggesting that someone (not me, at least right now) look at a broader word list and really see if there is a correspondence. IMO, the coincidences, especially with the morphology, pronouns and numerals, and the initial rhotic fact, warrant a deeper look, even if, as I said in the opening post (please read the whole thing; don't just assume that I'm a crackpot), it turns out to be trash right away, which I think it probably will but the off-chance is worth it.

[Zaarin]

In general the answer to "Has anyone tried to link Basque and ____" is yes.

We all know that all agglutinating languages are Sumerian.

[Salmoneus]

I know that, but I'm suggesting that someone (not me, at least right now) look at a broader word list and really see if there is a correspondence.

But there isn't a correspondence. Because Turkish and Basque are not related.

[Tropylium]

IMO, the coincidences, especially with the morphology, pronouns and numerals, and the initial rhotic fact, warrant a deeper look

The "initial rhotic fact" applies to a bit too many languages in Eurasia (at minimum also Proto-Indo-European, Proto-Uralic, Proto-Mongolian, Proto-Tungusic) to count as more than typological evidence. Numeral-wise then, 1 and 5 seem similar enough but obviously *iki ~ *biga for '2' doesn't work alongside them; we would definitely expect rather e.g. **biki.

But ok, let's have an elementary deeper look and break out Trask's Etymological Dictionary of Basque on the other hand, the Russians' etymological database of Turkic on the other:
– Turk. kurt comes from a root meaning 'worm'; the more original Turkic word for 'wolf' is börü < *bȫrü.
– Bsq. bihurgune is some kind of a compound or derivative. The basic native Basque root for 'twisted, folded' is just /bihur-/. This makes the Turkic comparison look rather less probable (especially if you were aiming for the correspondences g ~ k, m ~ n).
– Turk. bükme is similarly though a derivative from bük- 'to bend', which may partially mitigate the problem (but I don't thnk sufficiently much).
– Bsq. zuhaitz: per Trask this means 'wild tree' and is a derivative of basic zur 'wood'. (Haitz means 'stone', but I'm not sure if this is necessarily the 2nd component.)
– Turk. yemek 'to eat' is segmented as ye-mek, not ˣyem-ek. Cf. e.g. ye-r 's/he eats'.
So that's already four comparisons that break down under scrutiny. My hopes for Vasco-Turkic are not rising at all.

[Vijay]

I know that, but I'm suggesting that someone (not me, at least right now) look at a broader word list and really see if there is a correspondence. IMO, the coincidences, especially with the morphology, pronouns and numerals, and the initial rhotic fact, warrant a deeper look, even if, as I said in the opening post (please read the whole thing; don't just assume that I'm a crackpot), it turns out to be trash right away, which I think it probably will but the off-chance is worth it.

I did read the whole thing, and honestly, I think I know where you're coming from because I used to be surprised by stuff like this myself especially because some coincidences really seem too good to be true at least at first glance (seriously, you should see Hawaiian and Greek!). Since then, though, I've learned that things like this can be disproven pretty easily, however weird that may seem.

First of all, lots of languages don't even have rhotics, plus they can be notoriously difficult to distinguish from lateral approximants anyway. Even for native speakers of languages where there is a phonemic distinction between /l/ and a rhotic, it can be very difficult or even impossible to tell the difference under the right conditions; I've personally seen this demonstrated for English. And feel free to correct me if I'm wrong, but unless I miss my guess, no native Japonic, Korean, or Bantu words seem to begin with a rhotic, either, and if I understood Tropylium correctly, there aren't even any in PIE. So the fact that two languages don't have word-initial rhotics in native words doesn't seem so remarkable.

Second, word order is a typological feature that spreads pretty easily and is hardly evidence of genetic relatedness, and SOV is one of the most common word orders in the world, so that's not really evidence of anything. For that matter, if you take a look at WALS, it seems like almost every language in India but Kashmiri (for some reason) has SOV word order, too, regardless of genetic affiliation, even though probably most of these languages also have relatives spoken outside India that are not SOV.

Third, morphemes in two different languages containing similar sequences is not evidence of a correspondence of any sort. If it were, I could probably connect Dravidian languages to just about anything.

Fourth, zu in Basque is not plural. Zuek is.

Finally, even if you look in your limited sample, there are no correspondences that hold up across the board. For example, you noted that Proto-Turkish has /bir/ 'one' and Ancient Basque has *bade, as well as besh and *bortz for 'five', and <bükme> for 'bending' and <bihurgune> for 'curve', so all of these start with /b/ except that you also have Ancient Basque *biga 'two' vs. Proto-Turkish /iki/! So why is it that the Proto-Turkish word in that case doesn't have /b/ even though it does in the others? There doesn't seem to be an explanation that works or has a phonetic motivation.

[Salmoneus]

On the subject of prohibiting initial rhotics, I'd just add that numerous languages have (re-)introduced this constraint even in recorded history. Some dialects of Sardinian, for example, geminate initial rhotics and then add epenthetic vowels to stop them being initial.

[Nortaneous]

Finally, even if you look in your limited sample, there are no correspondences that hold up across the board. For example, you noted that Proto-Turkish has /bir/ 'one' and Ancient Basque has *bade, as well as besh and *bortz for 'five', and <bükme> for 'bending' and <bihurgune> for 'curve', so all of these start with /b/ except that you also have Ancient Basque *biga 'two' vs. Proto-Turkish /iki/! So why is it that the Proto-Turkish word in that case doesn't have /b/ even though it does in the others? There doesn't seem to be an explanation that works or has a phonetic motivation.

Differences in noun class marker fossilization, of course.

[Vijay]

Oh man, that's right, let's make it even more complicated!

[Zju]

IMO, the coincidences, especially with the morphology, pronouns and numerals, and the initial rhotic fact, warrant a deeper look

The "initial rhotic fact" applies to a bit too many languages in Eurasia (at minimum also Proto-Indo-European, Proto-Uralic, Proto-Mongolian, Proto-Tungusic

Really? Then how come so many words and roots in most, if not all, IE branches begin in a rhotic?

On the subject of prohibiting initial rhotics, I'd just add that numerous languages have (re-)introduced this constraint even in recorded history. Some dialects of Sardinian, for example, geminate initial rhotics and then add epenthetic vowels to stop them being initial.

Rather interesting. Do you happen to have a source?

The only proposal for genetic affiliation of Basque that I have a bit trust in is the Afro-Asiatic hypothesis, on the grounds on explaining the gender marking in 2.SG.

[WeepingElf]

IMO, the coincidences, especially with the morphology, pronouns and numerals, and the initial rhotic fact, warrant a deeper look

The "initial rhotic fact" applies to a bit too many languages in Eurasia (at minimum also Proto-Indo-European, Proto-Uralic, Proto-Mongolian, Proto-Tungusic

Really? Then how come so many words and roots in most, if not all, IE branches begin in a rhotic?

They dropped either an initial */s/ or an initial laryngeal. You can tell by the Greek cognate: the former have rh-; the latter have an initial vowel (e from *h1, a from *h2, o from *h3). For instance, the Greek cognate of English red is erythros; the PIE word was *h1rudhros.

[Vijay]

The only proposal for genetic affiliation of Basque that I have a bit trust in is the Afro-Asiatic hypothesis, on the grounds on explaining the gender marking in 2.SG.

Doesn't hika/hitano just generally do weird stuff in Basque grammar, though?

I've definitely heard people suggesting that Afro-Asiatic is itself kind of shaky because we don't even know how many of the similarities between Semitic, Chadic, etc. languages are due to centuries of close contact rather than inheritance.

[WeepingElf]

I've definitely heard people suggesting that Afro-Asiatic is itself kind of shaky because we don't even know how many of the similarities between Semitic, Chadic, etc. languages are due to centuries of close contact rather than inheritance.

I think that Afro-Asiatic is much like Indo-Uralic: good agreements in inflectional morphology, but few if any lexical cognates. Both may have similar time depths (perhaps about 10,000 years). The difference is mainly that Africanists are much more ready to accept distant relationship proposals than Eurasianists, it seems to me. There are two different reconstructions of Proto-Afro-Asiatic on the market (one by Ehret, the other by Orel and Stolbova), which can't both be right - which perhaps tells us that both are wrong.

[Zaarin]

The only proposal for genetic affiliation of Basque that I have a bit trust in is the Afro-Asiatic hypothesis, on the grounds on explaining the gender marking in 2.SG.

Doesn't hika/hitano just generally do weird stuff in Basque grammar, though?

I've definitely heard people suggesting that Afro-Asiatic is itself kind of shaky because we don't even know how many of the similarities between Semitic, Chadic, etc. languages are due to centuries of close contact rather than inheritance.

From my understanding, the only questionable member of Afro-Asiatic is Omotic, and that's largely from a lack of information.

[KathTheDragon]

IMO, the coincidences, especially with the morphology, pronouns and numerals, and the initial rhotic fact, warrant a deeper look

The "initial rhotic fact" applies to a bit too many languages in Eurasia (at minimum also Proto-Indo-European, Proto-Uralic, Proto-Mongolian, Proto-Tungusic

Really? Then how come so many words and roots in most, if not all, IE branches begin in a rhotic?

They dropped either an initial */s/ or an initial laryngeal. You can tell by the Greek cognate: the former have rh-; the latter have an initial vowel (e from *h1, a from *h2, o from *h3). For instance, the Greek cognate of English red is erythros; the PIE word was *h1rudhros.

Similarly, there are no Hittite words beginning in r.

[Tropylium]

I've definitely heard people suggesting that Afro-Asiatic is itself kind of shaky because we don't even know how many of the similarities between Semitic, Chadic, etc. languages are due to centuries of close contact rather than inheritance.

From my understanding, the only questionable member of Afro-Asiatic is Omotic, and that's largely from a lack of information.

Afro-Asiatic as a family is relatively sound, but if you want to know which similarities exactly are old common inheritance — yes, that part is still very shaky. There isn't really even a clear lexical reconstruction of Proto-Semitic yet, let alone nodes like Proto-Chadic or Proto-Cushitic.

Probably a lot of lexical cognates haven't even been identified yet.

[Vijay]

How does that work, honestly? How is the evidence for a language family sound if you can't identify lexical correspondences, or consistent sound correspondences? I feel like there must be something I'm missing here. Or are you saying we can, and the problem is something else? I'm confused lol.

EDIT: Well, okay, I guess you're saying that figuring out which similarities are due to inheritance and which are due to contact is the problem? But then...that's what I wonder about, how then is the evidence for the family sound?

[Tropylium]

How does that work, honestly? How is the evidence for a language family sound if you can't identify lexical correspondences, or consistent sound correspondences? I feel like there must be something I'm missing here. Or are you saying we can, and the problem is something else? I'm confused lol.

EDIT: Well, okay, I guess you're saying that figuring out which similarities are due to inheritance and which are due to contact is the problem? But then...that's what I wonder about, how then is the evidence for the family sound?

Well, the research situation looks as bare-bones as this: https://en.wikipedia.org/wiki/Proto-Afr ... c_language

There are lexical and sound correspondences that are consistent enough to say that they are onto something (e.g. there does not appear to have been any early major Grimm's-Law-style consonant shifts anywhere; fingerprint phonemes like emphatics and pharyngeals in one branch also correspond to emphatics and pharyngeals in others), but that's just the broad strokes. The Afro-Asiatic equivalents of Verner's Law or *kʷ > p/t/k in Greek or the Slavic palatalizations surely still await discovery. And before those are made, it's very possible that fairly old loans will not be accurately identified, or that some accidental similarities will be treated as significant.

[Nortaneous]

Oh man, that's right, let's make it even more complicated!

IIRC this is what Starostinites actually believe: they project the NEC gender system onto Proto-Vasco-Caucasian (which also includes Burushaski, Yeniseian, Sino-Tibetan, and occasionally Na-Dene, Salishan, Algic, and Sumerian) and think the prefixes fossilized in Basque.

[Sumelic]

I could do this for Turkish and Hungarian:

- Proto-Uralic apparently didn't have word-initial "r," although it exists in many modern languages such as Hungarian.
- Hungarian has some SOV tendencies (specifically, Wikipedia says topics and comments are often in pre-verbal position)
-the Hungarian sublative case -ra, -re ends in /rV/
-the Hungarian genitive suffix -nak,-nek starts with a nasal
-the Hungarian second-person pronoun te is composed of a dental consonant plus the vowel /e/, just like Proto-Turkic /se/
-the Hungarian third-person pronoun ő is just a vowel, just like the Proto-Turkic third person pronoun /o/

I'm sure I could find more.

[Αυτοβοτα]

But weren't proto-Hungarian and proto-Turkish in close contact for long time in the steppe, like all the other Central Asian/Uralic/Altaic languages?