WeepingElf's Europic thread

[Etherman]

The morphological similarities boil down to just one pronoun and one case ending, and that's simply not enough to rule out chance resemblance. The same pronoun and the same case ending are also found in Kartvelian -

So, uh, which ones? Out of mere curiousity.

The pronoun in question is Etr. mi, Geo. me (1sg.); the case marker is Etr. -s (earlier -si), Geo. -(i)s (genitive).

Also there's possibly the Etr. neuter plural -chva and IE collective *eh2 (which later became the feminine).

[TaylorS]

Bump again.

Research on the Europic hypothesis keeps me (somewhat) busy. Currently, I am writing a preliminary grammar sketch of Proto-Europic, based on internal reconstruction from PIE. (There is too little known about Hesperic to be of much help here.) Another project is cataloguing and mapping the Old European Hydronymy - I haven't seen a good map of it anywhere yet, so I'll take Krahe's data and piece one together all by myself.

On the topic of Etruscan, I no longer see any reason to connect it to Europic. The morphological similarities boil down to just one pronoun and one case ending, and that's simply not enough to rule out chance resemblance. The same pronoun and the same case ending are also found in Kartvelian - which, like Etruscan, is otherwise so unlike IE that a relationship (at least, one closer than Uralic) seems very unlikely (there are some vague resemblances in phonology between Kartvelian and IE - but not Etruscan - , but that may be areal, and doesn't really say anything about genealogy). Also, lexical cognates appear to be completely absent.

So I take you you reject Glen Gordon's Indo-Aegean hypothesis?

I couldn't read that file (and many others he has produced) because I don't have the right software on my machine

But what I have seen of his hypothesis is utterly unconvincing. He doesn't establish an IE-Etruscan connection, he assumes it. He often rejects the mainstream scholarly opinion on the meanings of Etruscan words, assigning them different meanings based on similarities with IE words - and then adduces the "cognate pairs" obtained that way as "evidence" for Etruscan being related to IE. That's circular reasoning.

It also doesn't help that Glen Gordon behaves pretty much the same way as Octaviano. I remember him from the Nostratic-L mailing list before I withdrew from that list because it was full of crackpots. He accused his critics of ad hominem attacks, called the entire academic mainstream names, and even invoked comparisons with the Nazis. Eventually, he decided to withdraw from Nostratic-L and similar fora, and instead started ranting away on his blog. That alone does not falsify his hypothesis, but I have pointed out the main problems with it, and if someone acts up that way, you can take it as a sure sign that you are dealing with a crackpot.

BTW, you can see the cognate-ness between the PIE active-stative verbal endings (or objective-subjective as Gordon calls them, something I disagree with since what he is describing is just an active-statve variant) in Alpic.

Here is my current verbal ending system for Alpic:

Code: Select all

AGREEMENT  | Agentive S | Patientive S | Reflexive S |
---------------------------------------------------------
1SG        | -m         | -ch          | -gu         |
---------------------------------------------------------
2SG        | -t         | -n           | -du         |
---------------------------------------------------------
3SG        | -s         | -Ø           | -su         |
---------------------------------------------------------
1PL        | -me        | -he          | -ge         |
---------------------------------------------------------
2PL        | -te        | -ne          | -de         |
---------------------------------------------------------
3PL        | -se        | -je          | -sje        |
---------------------------------------------------------
Impersonal |           -p              | -pa         |
---------------------------------------------------------

The agentive endings look cognate, but the patientive endings show no visible connection to the IE ones (singular *-h2a, *-th2a, *-e). In fact, they look like the stative (or whatever) endings in Glen Gordon's reconstruction of Nostratic, which appear to be based on Dravidian. In my opinion, there is no reason to assume a relationship between IE and Dravidian.

OK, thanks for the feedback!

[TaylorS]

OK, here are Alpic's revised verbal endings:

Code: Select all

AGREEMENT  | Agentive S | Patientive S | Reflexive S |
---------------------------------------------------------
1SG        | -m         | -ch          | -gi         |
---------------------------------------------------------
2SG        | -t         | -ta          | -tji        |
---------------------------------------------------------
3SG        | -s         | -Ø           | -si         |
---------------------------------------------------------
1PL        | -me        | -ha          | -ge         |
---------------------------------------------------------
2PL        | -te        | -ka          | -de         |
---------------------------------------------------------
3PL        | -se        | -da          | -se         |
---------------------------------------------------------
Impersonal |           -p              | -pa         |
---------------------------------------------------------

Just to note, <ch> represents coda /x/, which is <h> elsewhere.


and here is the ending of Proto-Danubian, Alpic's ancestor, in X-SAMPA. P-D was spoken around 6000 BCE

Code: Select all

AGREEMENT  | Agentive S | Patientive S |
----------------------------------------
1          | -m@i       | -ka          |
----------------------------------------
2          | -t@i       | -t@ka        |
----------------------------------------
3          | -sai       | -Ø           |
----------------------------------------

The 3rd Person ending is derived from a demonstrative. The plural forms developed from the addition of the plural morpheme *-ta, cognate with the Uralic and early PIE plural morpheme. Then the reflexive endings developed when the Agentive endings were attached to the Patientive endings. Then syncope in unstressed syllables, analogy, final consonant erosion, and assimilation go to work.

[Etherman]

I'm hope I'm not straying too far off topic by giving my own reconstruction for the Proto-Nostratic (or as I prefer to call it, Proto-Siberian) personal pronouns. I'll use V to stand for any vowel. There appears to have been some kind of ablaut which has obscured the vowels (perhaps something like *a when bound and *i when unbound).

Active:
1. *phV
2. *sV
3. *tV

Stative:
1. *khV
2. *tV
3. *i

Oblique:
1. *NV (N is the velar nasal)
2. *nV
3. *sV

The 3rd person pronouns all function as demonstratives or deictics. I also assume that as a prosodic feature, all unbound pronouns were pre-glottalized.

My 1P reconstrcution of *phV might look strange, but it actually explains a number of otherwise problematic details. I do however assume that when the unbound pronoun was extended with *-n the *ph assimilated to *m. In unassimilated form the following changes took place:

1) *ph > *w in IE, Uralic, Kartvelian (and Yukaghir if it's related)
2) *ph > *b in Altaic
3) *ph > *v in Chukchi-Kamchatkan and Eskimo-Aleut

Uralic has completely eliminated the 1P forms in *w in favor of those in *m, but ever other language preserves traces of both (Altaic is most transparent in this regard). The oblique pronouns are the worst preserved, but traces are found in most, if not all, of the languages.

What's striking is that the oblique set is the same set that's reconstructed for Proto-Sino-Tibetan. I don't know if it's a weird coincidence, an areal effect, or if there's a deep genetic relationship there.

[TaylorS]

I'm hope I'm not straying too far off topic by giving my own reconstruction for the Proto-Nostratic (or as I prefer to call it, Proto-Siberian) personal pronouns. I'll use V to stand for any vowel. There appears to have been some kind of ablaut which has obscured the vowels (perhaps something like *a when bound and *i when unbound).

Active:
1. *phV
2. *sV
3. *tV

Stative:
1. *khV
2. *tV
3. *i

Oblique:
1. *NV (N is the velar nasal)
2. *nV
3. *sV

The 3rd person pronouns all function as demonstratives or deictics. I also assume that as a prosodic feature, all unbound pronouns were pre-glottalized.

My 1P reconstrcution of *phV might look strange, but it actually explains a number of otherwise problematic details. I do however assume that when the unbound pronoun was extended with *-n the *ph assimilated to *m. In unassimilated form the following changes took place:

1) *ph > *w in IE, Uralic, Kartvelian (and Yukaghir if it's related)
2) *ph > *b in Altaic
3) *ph > *v in Chukchi-Kamchatkan and Eskimo-Aleut

Uralic has completely eliminated the 1P forms in *w in favor of those in *m, but ever other language preserves traces of both (Altaic is most transparent in this regard). The oblique pronouns are the worst preserved, but traces are found in most, if not all, of the languages.

What's striking is that the oblique set is the same set that's reconstructed for Proto-Sino-Tibetan. I don't know if it's a weird coincidence, an areal effect, or if there's a deep genetic relationship there.

So the IE 1PL pronoun in *w- would have come from the bare *ph-, while the 1SG *mi would have come from *phn?

[Etherman]

So the IE 1PL pronoun in *w- would have come from the bare *ph-, while the 1SG *mi would have come from *phn?

Yeah, but it was probably *phina > *mina. This *-na suffix was a genitive marker (which implies that the stative pronouns were used in the genitive case). In some languages this suffix lost productivity but remained in fossilized forms. The paradigmatic alternation is seen most clearly in Altaic languages.

Tatar: min "I" bez "we"
Uzbek: men "I" biz "we"
Tuva: men "I", bis "we"
Middle Mongolian: bi "I" mino "mine"
Khalkha: bi "I" minij "mine"
Ordos: bi "I" mini "mine"
Udighe: bi "I" minti "we"

In Tungus-Manchu the -n is a plural marker (which also goes back to Proto-Nostratic).

[WeepingElf]

So the IE 1PL pronoun in *w- would have come from the bare *ph-, while the 1SG *mi would have come from *phn?

Yeah, but it was probably *phina > *mina. This *-na suffix was a genitive marker (which implies that the stative pronouns were used in the genitive case). In some languages this suffix lost productivity but remained in fossilized forms. The paradigmatic alternation is seen most clearly in Altaic languages.

Tatar: min "I" bez "we"
Uzbek: men "I" biz "we"
Tuva: men "I", bis "we"
Middle Mongolian: bi "I" mino "mine"
Khalkha: bi "I" minij "mine"
Ordos: bi "I" mini "mine"
Udighe: bi "I" minti "we"

In Tungus-Manchu the -n is a plural marker (which also goes back to Proto-Nostratic).

This alternation is the main reason why I believe in Altaic; while most of the other similarities between the Altaic languages can easily be ascribed to contact, the /b/~/m/ alternation in the first person pronouns is striking, and not easily explained by borrowing or convergence. The relationship between the three Altaic families, however, is not close; it may be as deep as that between IE and Uralic.

Which languages do you consider related through your "Siberian", and what is your subgrouping like?

In my opinion, Europic is related to Uralo-Siberian (Uralic, Yukaghir, Chukotko-Kamchatkan, Eskimo-Aleut; as proposed by Michael Fortescue); this "Indo-Uralic" group (I should think of a better term for it) is about 10,000 years old and in turn more distantly related to Altaic. I call that "Mitian", after the characteristic shapes of the personal pronouns.

[Cedh]

In my opinion, Europic is related to Uralo-Siberian (Uralic, Yukaghir, Chukotko-Kamchatkan, Eskimo-Aleut; as proposed by Michael Fortescue); this "Indo-Uralic" group (I should think of a better term for it)

What about "Euro-Siberian"?

[TaylorS]

So the IE 1PL pronoun in *w- would have come from the bare *ph-, while the 1SG *mi would have come from *phn?

Yeah, but it was probably *phina > *mina. This *-na suffix was a genitive marker (which implies that the stative pronouns were used in the genitive case). In some languages this suffix lost productivity but remained in fossilized forms. The paradigmatic alternation is seen most clearly in Altaic languages.

Tatar: min "I" bez "we"
Uzbek: men "I" biz "we"
Tuva: men "I", bis "we"
Middle Mongolian: bi "I" mino "mine"
Khalkha: bi "I" minij "mine"
Ordos: bi "I" mini "mine"
Udighe: bi "I" minti "we"

In Tungus-Manchu the -n is a plural marker (which also goes back to Proto-Nostratic).

This alternation is the main reason why I believe in Altaic; while most of the other similarities between the Altaic languages can easily be ascribed to contact, the /b/~/m/ alternation in the first person pronouns is striking, and not easily explained by borrowing or convergence. The relationship between the three Altaic families, however, is not close; it may be as deep as that between IE and Uralic.

Which languages do you consider related through your "Siberian", and what is your subgrouping like?

In my opinion, Europic is related to Uralo-Siberian (Uralic, Yukaghir, Chukotko-Kamchatkan, Eskimo-Aleut; as proposed by Michael Fortescue); this "Indo-Uralic" group (I should think of a better term for it) is about 10,000 years old and in turn more distantly related to Altaic. I call that "Mitian", after the characteristic shapes of the personal pronouns.

I agree with this.

Interestingly, IE and Altaic share an important feature, the lentition of final *t to *s, which then became *r_j (IIRC) in Proto-Altaic. This is why the Early PIE 2SG.ACTIVE ending and the Early PIE plural have an *s and not a *t.

[WeepingElf]

In my opinion, Europic is related to Uralo-Siberian (Uralic, Yukaghir, Chukotko-Kamchatkan, Eskimo-Aleut; as proposed by Michael Fortescue); this "Indo-Uralic" group (I should think of a better term for it)

What about "Euro-Siberian"?

Yes, that's a good, handy name, and actually one I consider.

[WeepingElf]

Interestingly, IE and Altaic share an important feature, the lentition of final *t to *s, which then became *r_j (IIRC) in Proto-Altaic. This is why the Early PIE 2SG.ACTIVE ending and the Early PIE plural have an *s and not a *t.

Yes. There are some instances where an IE *s corresponds to a Uralic *t, such as the 2sg. active ending and the plural marker. A parallel correspondence between IE *H (i.e., any laryngeal) and Uralic *k is also found, e.g. in the 1sg. stative ending.

[Etherman]

This alternation is the main reason why I believe in Altaic; while most of the other similarities between the Altaic languages can easily be ascribed to contact, the /b/~/m/ alternation in the first person pronouns is striking, and not easily explained by borrowing or convergence. The relationship between the three Altaic families, however, is not close; it may be as deep as that between IE and Uralic.

I think Tungus-Mongolian is about as old as Indo-Uralic, which makes Altaic (which includes Japanese and Korean, IMO) very deep.

Which languages do you consider related through your "Siberian", and what is your subgrouping like?

IE, Uralic, Kartvelian, Eskimo-Aleut, Chukchi-Kamchatkan, and Altaic. I'm warming up to the idea of including Yukaghir, but I'm not convinced that it's especially close to Uralic. Etruscan seems likely, but I don't know if they'll ever be enough evidence to demonstrate it.

As for subgroupings Indo-Uralic looks like the most solid candidate. Kartvelian would then be related to Indo-Uralic. I guess you could call this group Western Siberian (which would probably include Yukaghir). Altaic would be it's own node. EA and CK would form Eastern Siberian.

In my opinion, Europic is related to Uralo-Siberian (Uralic, Yukaghir, Chukotko-Kamchatkan, Eskimo-Aleut; as proposed by Michael Fortescue); this "Indo-Uralic" group (I should think of a better term for it) is about 10,000 years old and in turn more distantly related to Altaic. I call that "Mitian", after the characteristic shapes of the personal pronouns.

Dating this is fraught with difficulties. 10-12K is probably a good guesstimate though. It's clearly pre-Neolithic.

[TomHChappell]

Dating this is fraught with difficulties.

QFT.

[Etherman]

Interestingly, IE and Altaic share an important feature, the lentition of final *t to *s, which then became *r_j (IIRC) in Proto-Altaic. This is why the Early PIE 2SG.ACTIVE ending and the Early PIE plural have an *s and not a *t.

I propose *-s_hi as a plural. In almost all languages this *s_h merged with *r, but it remain distinct in Turkic (> *z , except in Chuvash > r) and Eskimo-Aleut (> *R). This has a pretty widespread distribution though it may originally only applied to stative pronouns (later extended in some languages). Possibly had a collective connotation, but that might be secondary.

Kartvelian *-ar

IE *-r but limited to 3rd person plurals in the stative conjugation (but later expanded elsewhere in some languages)

Uralic *-r but limited to Nenets and Mari.

Altaic *-r'

CK *-ri, found only in plural pronouns

Etruscan *-r

Note: It's possible that my *s_h was really /z/ but since I don't reconstruct any voice contrast I interpret it as an aspirated fricative (I do have phonemic aspiration in the stops).

[TaylorS]

Interestingly, IE and Altaic share an important feature, the lentition of final *t to *s, which then became *r_j (IIRC) in Proto-Altaic. This is why the Early PIE 2SG.ACTIVE ending and the Early PIE plural have an *s and not a *t.

I propose *-s_hi as a plural. In almost all languages this *s_h merged with *r, but it remain distinct in Turkic (> *z , except in Chuvash > r) and Eskimo-Aleut (> *R). This has a pretty widespread distribution though it may originally only applied to stative pronouns (later extended in some languages). Possibly had a collective connotation, but that might be secondary.

Kartvelian *-ar

IE *-r but limited to 3rd person plurals in the stative conjugation (but later expanded elsewhere in some languages)

Uralic *-r but limited to Nenets and Mari.

Altaic *-r'

CK *-ri, found only in plural pronouns

Etruscan *-r

Note: It's possible that my *s_h was really /z/ but since I don't reconstruct any voice contrast I interpret it as an aspirated fricative (I do have phonemic aspiration in the stops).

Interesting alternative hypothesis! I'll have to look into it.

[Etherman]

Interesting alternative hypothesis! I'll have to look into it.

I forgot to mention Eskimo-Aleut *-R which is a singulative (from an earlier collective?). This one is ambiguous because in word final position *R and *q were not distinguished. Alternatively the singulative could have been *q which could match with IE *eh2 collective (> feminine in the non-Anatolian languages).

[WeepingElf]

Interesting alternative hypothesis! I'll have to look into it.

I forgot to mention Eskimo-Aleut *-R which is a singulative (from an earlier collective?). This one is ambiguous because in word final position *R and *q were not distinguished. Alternatively the singulative could have been *q which could match with IE *eh2 collective (> feminine in the non-Anatolian languages).

How does a singulative evolve from a collective, which has pretty much exactly the opposite meaning? Also, I feel that EA *R is a development of PES *g, not *r. Also, how do you account for the *t-plurals in Uralo-Siberian?

[Etherman]

How does a singulative evolve from a collective, which has pretty much exactly the opposite meaning? Also, I feel that EA *R is a development of PES *g, not *r. Also, how do you account for the *t-plurals in Uralo-Siberian?

A collective takes a bunch of objects and treats them as a singular whole. So there's a sense of a plurality becoming a singularity.

As for the t-plurals they certainly existed in Proto-Siberian. I don't think they survived into IE.

[WeepingElf]

How does a singulative evolve from a collective, which has pretty much exactly the opposite meaning? Also, I feel that EA *R is a development of PES *g, not *r. Also, how do you account for the *t-plurals in Uralo-Siberian?

A collective takes a bunch of objects and treats them as a singular whole. So there's a sense of a plurality becoming a singularity.

Hmm, perhaps.

As for the t-plurals they certainly existed in Proto-Siberian. I don't think they survived into IE.

I disagree, which, however, doesn't mean that you are wrong; I could just as well be wrong.

[Etherman]

I disagree, which, however, doesn't mean that you are wrong; I could just as well be wrong.

A couple things make me suspicious of the IE s-plural coming from an earlier t-plural. First, it requires word final lenition, but fortition is more common in word final position. Not that word final lenition can't happen, but it's not as common. Second, the difference between the nominative singular and plural is not the presence of *-s (which occurs in both) but the presence of *-e- in the plural. This suggests to me that *-s is a nominative marker but *-e- is the plural marker. In most of the plural cases it's ambiguous whether the plural is caused by *-s or *-e/o-. In the cases where it's not ambiguous some of them suggest *-s while others suggest *-e/o-. Third, word final *-t is found so one has to come up with additional assumptions to explain their existence.

[WeepingElf]

I disagree, which, however, doesn't mean that you are wrong; I could just as well be wrong.

A couple things make me suspicious of the IE s-plural coming from an earlier t-plural. First, it requires word final lenition, but fortition is more common in word final position. Not that word final lenition can't happen, but it's not as common.

Sure, but that doesn't mean it didn't happen. Also, there are good correspondences for IE *H ~ Uralic *k, to which IE *s ~ Uralic *t would be excatly parallel. Of course, the original consonant may have been something else than /t/. Adam Hyllestad reconstructs */x/ for the *H ~ *k correspondence set; I don't remember now what he reconstructs for *s ~ *t.

Second, the difference between the nominative singular and plural is not the presence of *-s (which occurs in both) but the presence of *-e- in the plural. This suggests to me that *-s is a nominative marker but *-e- is the plural marker. In most of the plural cases it's ambiguous whether the plural is caused by *-s or *-e/o-. In the cases where it's not ambiguous some of them suggest *-s while others suggest *-e/o-. Third, word final *-t is found so one has to come up with additional assumptions to explain their existence.

The IE plural case endings are not agglutinative, and most seem to be post-Anatolian formations as they are totally different in Hittite, so they do not reflect a Proto-Indo-Uralic (or whatever) situation. It is indeed hard to say what morpheme expressed the plural in Proto-Europic.

[Etherman]

Sure, but that doesn't mean it didn't happen. Also, there are good correspondences for IE *H ~ Uralic *k, to which IE *s ~ Uralic *t would be excatly parallel. Of course, the original consonant may have been something else than /t/. Adam Hyllestad reconstructs */x/ for the *H ~ *k correspondence set; I don't remember now what he reconstructs for *s ~ *t.

He has a number of good H~k correspondences. I think it's also possible that in inflectional endings *k > *h1 and *kh > *h2 (it's often difficult to distinguish *k from *q and *kh from *qh because they only survive as uvulars in Eskimo-Aleut, Chukchi-Kamchatkan, and possibly Kartvelian but that data is inconsistent suggesting secondary developments). I don't know what he reconstructs for s~t. I agree with Kortlandt that *ti > *si (>*sei in open syllables not followed by a resonant) in IE, at least in initial position. There are almost no IE roots of the form *tei and where they do appear the distribution is severely restricted.

The IE plural case endings are not agglutinative, and most seem to be post-Anatolian formations as they are totally different in Hittite, so they do not reflect a Proto-Indo-Uralic (or whatever) situation. It is indeed hard to say what morpheme expressed the plural in Proto-Europic.

The IE cases are fusional, but fusional inflections generally derive from earlier agglutinating cases. The IE cases appear to be made from recurring elements: *m, *s, *e/o, *i (probably a grammaticalized form of *h1ei "to go" which was related to the pronoun *h1i), *u, *d, *h1, *h2, and *bhi (which was a grammaticalized adverb). There's probably no such thing as the Proto-Europic plural.

[WeepingElf]

I am bumping this thread for two reasons.

1. I am referring to it in the "European languages before Indo-European" thread. Obviously, this matter is relevant to that.

2. I have recently picked up my research into Europic again. My next project is a map of the Old European hydronymy, which will eventually appear on my web site.

[jal]

2. I have recently picked up my research into Europic again. My next project is a map of the Old European hydronymy, which will eventually appear on my web site.

Cool, let us know when it's done.


JAL

[WeepingElf]

Europic is obsolete

I have abandoned the Europic hypothesis in its "classic" form as a family consisting of Indo-European and the (unknown) languages of the Linearbandkeramik, Vinča and Cucuteni-Trypillia cultures, spread by refugees of the Black Sea Flood in the early Neolithic. The latter cultures probably had nothing to do with the Proto-Indo-Europeans, neither archaeologically nor genetically, so why should they speak related languages? The closest modern relatives of the LBK people seem to be the Georgians - which invites the speculation that the LBK people and their southeastern relatives may have spoken a language related to the Kartvelian languages: a rocking cool idea for a lostlang, but I don't know enough about Kartvelian (it's difficult: most of the relevant literature is in Georgian) to build a good para-Kartvelian language.

I still think the Old European Hydronymy preserves traces of a language related to IE (or rather, an early diverging branch of IE) which I now call Aquan, but that has nothing to do with LBK but originated in the first wave of Kurgan expansions ca. 4500 BC (Anatolian would be Kurgan II ca. 3500 BC, Nuclear IE Kurgan III ca. 3000 BC).

The Black Sea Flood, if that happened at all which most geologists doubt, is also out of the game.