European languages before Indo-European

[Octavià]

The similarity between the forms used to name 'deer/goat' and 'boar' seen in the IE forms it's also reflected in Arabic, which in addition to ɣafr-, ɣufr- 'young of chamois/goat' also has ʕifr-, ʕufr- 'pig, boar; piglet', which I guess correspond to IE *H₁epero- 'boar'. I guess the origin of this duality was phonosymbolism, with the voiced velar fricative ɣ for 'deer/goat' and the voiced pharyngeal ʕ for 'boar, pig'.

That is, *kapro- ~ *gapro- 'he-goat' corresponds to Arabic ɣafr- 'young of chamois/goat' < PSem *ɣapr- and *H₁epero- 'boar' to Arabic ʕifr- 'pig, boar; piglet' < PSem *ʕipr-, with the addition of the IE male suffix *-o-.

[WeepingElf]

The similarity between the forms used to name 'deer/goat' and 'boar' seen in the IE forms it's also reflected in Arabic, which in addition to ɣafr-, ɣufr- 'young of chamois/goat' also has ʕifr-, ʕufr- 'pig, boar; piglet', which I guess correspond to IE *H₁epero- 'boar'. I guess the origin of this duality was phonosymbolism, with the voiced velar fricative ɣ for 'deer/goat' and the voiced pharyngeal ʕ for 'boar, pig'.

That is, *kapro- ~ *gapro- 'he-goat' corresponds to Arabic ɣafr- 'young of chamois/goat' < PSem *ɣapr- and *H₁epero- 'boar' to Arabic ʕifr- 'pig, boar; piglet' < PSem *ʕipr-, with the addition of the IE male suffix *-o-.

I think we are dealing with Neolithic Wanderwörter here. The Indo-Europeans acquired their goats from the Near East, and the word with them. Same for the pig word. I see no reason to postulate a Semitic substratum language anywhere in Europe.

BTW: The IE thematic vowel is not a "male suffix". It has nothing to do with gender.

[Octavià]

I think we are dealing with Neolithic Wanderwörter here. The Indo-Europeans acquired their goats from the Near East, and the word with them. Same for the pig word. I see no reason to postulate a Semitic substratum language anywhere in Europe.

I disagree, as this would hardly explain the irregularities seen in the Celtic word. Thus these words must have been borrowed somewhere in Europe from speakers of a Semitic-like language whose ancestors ultimately came from the Near East. A likely candidate would be Megalithic culture's people.

[WeepingElf]

I think we are dealing with Neolithic Wanderwörter here. The Indo-Europeans acquired their goats from the Near East, and the word with them. Same for the pig word. I see no reason to postulate a Semitic substratum language anywhere in Europe.

I disagree, as this would hardly explain the irregularities seen in the Celtic word. Thus these words must have been borrowed somehwere in Europe from people speaking a Semitic-like language (e.g. Megalithic culture's people) whose ancestors utlimately came from the Near East.

The Celtic word may have been borrowed from a Central European substratum language that had received the word from Semitic via a Mediterranean pathway, but that doesn't mean that that Central European language itself was Semitic. While I do not definitely rule out Semitic or at least Afro-Asiatic languages somewhere in Europe, I think the evidence is not substantial enough to reach the conclusion that such a language ever existed.

Of course, Theo Vennemann still thinks that the substratum held responsible by some for the typological aberrancy of Insular Celtic was related to Semitic - an old hypothesis, by the way, dating back to the 19th century - but few scholars find such a conclusion compelling nowadays. The typological similarities between Semitic and Insular Celtic have been grossly overstated; there is little other than VSO word order which the two groups have in common.

[Tropylium⁺]

Could the g- in Celtic be an assimilation to the tautosyllabic -b- (continuing the PIE ban on *T-Dʰ roots)? From Matasović's dictionary I can find as possible counterevidence *kāgni "law", and various derivativs of *kred- "hart", but these continue IE *d *ǵ, not *dʰ *ǵʰ; and *kagjo- "pen", but that seems syllabifiable as ka-gjo- (and has limited distribution anyway).

[edit: wrong MatVsF guy]

[Octavià]

The Celtic word may have been borrowed from a Central European substratum language that had received the word from Semitic via a Mediterranean pathway, but that doesn't mean that that Central European language itself was Semitic.

I can't see how your hypothesis would be more plausible than mine.

While I do not definitely rule out Semitic or at least Afro-Asiatic languages somewhere in Europe, I think the evidence is not substantial enough to reach the conclusion that such a language ever existed.

So I'd ask you: which language do you think Megalithic people spoke?

[Octavià]

Could the g- in Celtic be an assimilation to the tautosyllabic -b- (continuing the PIE ban on *T-Dʰ roots)? From Matisoff's dictionary I can find as possible counterevidence *kāgni "law", and various derivativs of *kred- "hart", but these continue IE *d *ǵ, not *dʰ *ǵʰ; and *kagjo- "pen", but that seems syllabifiable as ka-gjo- (and has limited distribution anyway).

I think Celtic *kagjo- 'pen' (< substrate IE *kag^h- 'hedge, enclosure') is actually a loanword from the same substrate: Western Chadic *kag- 'house (with clay walls)', Semitic *kaχ-/*kuχ- 'hut'.

[WeepingElf]

The Celtic word may have been borrowed from a Central European substratum language that had received the word from Semitic via a Mediterranean pathway, but that doesn't mean that that Central European language itself was Semitic.

I can't see how your hypothesis would be more plausible than mine.

It's a draw. In my model of the linguistic landscape of Neolithic Europe, there was the Hesperic family, a sister group of Indo-European, associated with the Linearbandkeramik expansion north of the Alps; in the Mediterranean, in contrast, no single language family spread with agriculture, but as the Neolithic Revolution spread through southern Europe, people took over the words for the new concepts (farm animals, crops, implements, etc.), which therefore spread from language to language throughout southern Europe, in a similar way as words like "automobile" or "telephone" have been borrowed from language to language in the 20th century.

While I do not definitely rule out Semitic or at least Afro-Asiatic languages somewhere in Europe, I think the evidence is not substantial enough to reach the conclusion that such a language ever existed.

So I'd ask you: which language do you think Megalithic people spoke?

I don't know. I would not even say that they all spoke related languages. Maybe they did; maybe they didn't. The whole pre-Indo-European linguistic landscape of Europe is so little known that one cannot really say much about it without indulging in groundless speculation.

[Octavià]

While I do not definitely rule out Semitic or at least Afro-Asiatic languages somewhere in Europe, I think the evidence is not substantial enough to reach the conclusion that such a language ever existed.

So I'd ask you: which language do you think Megalithic people spoke?

I don't know. I would not even say that they all spoke related languages. Maybe they did; maybe they didn't. The whole pre-Indo-European linguistic landscape of Europe is so little known that one cannot really say much about it without indulging in groundless speculation.

I think you're merely repeating what most scholars take as a dogma while at the same you dismiss the available evidence (i.e. substrate loanwords).

[WeepingElf]

While I do not definitely rule out Semitic or at least Afro-Asiatic languages somewhere in Europe, I think the evidence is not substantial enough to reach the conclusion that such a language ever existed.

So I'd ask you: which language do you think Megalithic people spoke?

I don't know. I would not even say that they all spoke related languages. Maybe they did; maybe they didn't. The whole pre-Indo-European linguistic landscape of Europe is so little known that one cannot really say much about it without indulging in groundless speculation.

I think you're merely repeating what most scholars take as a dogma while at the same you dismiss the available evidence (i.e. substrate loanwords).

I don't repeat any dogma and I do not dismiss any evidence! I merely accept the fact that the evidence we have so far is so limited that it is difficult to say anything definite, and that I don't already know what I am trying to find out. That is not a "dogma"; that is proper method.

[Octavià]

I don't repeat any dogma and I do not dismiss any evidence! I merely accept the fact that the evidence we have so far is so limited that it is difficult to say anything definite, and that I don't already know what I am trying to find out. That is not a "dogma"; that is proper method.

I disagree. What it happens is that most linguists aren't capable or don't like to work with small data sets. And when studying substrate languages this is most often the norm rather than the exception.

IMHO, there's enough evidence to state that at least some of the languages spoken in Neolithic Europe were related to Semitic, but unfortunately not much more (e.g. their grammar would remain forever unknown).

[TaylorS]

The Celtic word may have been borrowed from a Central European substratum language that had received the word from Semitic via a Mediterranean pathway, but that doesn't mean that that Central European language itself was Semitic.

I can't see how your hypothesis would be more plausible than mine.

It's a draw. In my model of the linguistic landscape of Neolithic Europe, there was the Hesperic family, a sister group of Indo-European, associated with the Linearbandkeramik expansion north of the Alps; in the Mediterranean, in contrast, no single language family spread with agriculture, but as the Neolithic Revolution spread through southern Europe, people took over the words for the new concepts (farm animals, crops, implements, etc.), which therefore spread from language to language throughout southern Europe, in a similar way as words like "automobile" or "telephone" have been borrowed from language to language in the 20th century.

While I do not definitely rule out Semitic or at least Afro-Asiatic languages somewhere in Europe, I think the evidence is not substantial enough to reach the conclusion that such a language ever existed.

So I'd ask you: which language do you think Megalithic people spoke?

I don't know. I would not even say that they all spoke related languages. Maybe they did; maybe they didn't. The whole pre-Indo-European linguistic landscape of Europe is so little known that one cannot really say much about it without indulging in groundless speculation.

I have wondered sometimes that perhaps Basque is a remnant of the Mediterranean expansion, specifically the Cardium Ware Culture. I remember reading of many place names in the western Med being from a Vasconic substratum.

As for the Megalithic people, they probably spoke languages of a family now extinct, perhaps whose only evidence of existence is non-IE, non-Vasconic inscriptions of Latin script in Spain and a few supposedly "Pictish" Ogham inscriptions in Scotland

[Octavià]

if I had to hazard a guess what the early Anatolian farmers spoke it was probably related to the languages of the Caucasus, most likely the now-extinct Hurro-Urartian Family.

I have wondered sometimes that perhaps Basque is a remnant of the Mediterranean expansion, specifically the Cardium Ware Culture. I remember reading of many place names in the western Med being from a Vasconic substratum.

Yeah, this is actually part of the Vasco-Caucasian hypothesis, of which I'm a supporter.

[WeepingElf]

I don't repeat any dogma and I do not dismiss any evidence! I merely accept the fact that the evidence we have so far is so limited that it is difficult to say anything definite, and that I don't already know what I am trying to find out. That is not a "dogma"; that is proper method.

I disagree. What it happens is that most linguists aren't capable or don't like to work with small data sets. And when studying substrate languages this is most often the norm rather than the exception.

Sure, when studying substrate languages you have to work with small data sets. But that does not mean that you can play fast and loose with the methodology. On the contrary; you must be especially careful. You just have to accept that the knowledge you can gain from the data is limited.

IMHO, there's enough evidence to state that at least some of the languages spoken in Neolithic Europe were related to Semitic, but unfortunately not much more (e.g. their grammar would remain forever unknown).

What rules out the Wanderwort hypothesis? All we can see are some substratum words which look like Semitic words; I have already given you a possible explanation how they could have wound up in Europe. Where's the problem with that? Of course, there could indeed have been languages related to Semitic in Europe; there is no way to prove the opposite. But so far I don't see compelling evidence that such languages were there.

[Octavià]

As for the Megalithic people, they probably spoke languages of a family now extinct, perhaps whose only evidence of existence is non-IE, non-Vasconic inscriptions of Latin script in Spain and a few supposedly "Pictish" Ogham inscriptions in Scotland.

I'd like to know which "Latin script inscriptions in Spain" are you taking about.

[Octavià]

Sure, when studying substrate languages you have to work with small data sets. But that does not mean that you can play fast and loose with the methodology. On the contrary; you must be especially careful. You just have to accept that the knowledge you can gain from the data is limited.

Of course so. But my point is there's enough evidence avaliable little to posit that languages related to Semitic were once spoken in Europe.

What rules out the Wanderwort hypothesis? All we can see are some substratum words which look like Semitic words; I have already given you a possible explanation how they could have wound up in Europe. Where's the problem with that?

Simply that I don't see any reason to think your hypothesis is more probable than mine.

Of course, there could indeed have been languages related to Semitic in Europe; there is no way to prove the opposite.

That's fine.

But so far I don't see compelling evidence that such languages were there.

This is because your interpretation of the linguistic data is different from mine.

[jal]

Sure, when studying substrate languages you have to work with small data sets. But that does not mean that you can play fast and loose with the methodology. On the contrary; you must be especially careful. You just have to accept that the knowledge you can gain from the data is limited.

QFT

Of course, there could indeed have been languages related to Semitic in Europe; there is no way to prove the opposite.

Well, as far as I have been able to Google (e.g here and here, but also Wikipedia), Proto-Semitic is dated to somewhere between 3000 and 4000 BC (with 3500BC more likely), and located in the Levant or Northern Mesopotamia (the first seeming more likely). This is much younger than PIE, and around the same time as the expansion of IE into Europe. So in order for a complete Semitic substrate in Europe for various IE languages, dating of PS needs to be wrong for a few millenia, something I find very doubtful. It seems far more likely that given the PS (or later Semitici languages) and IE languages must have been in close contact in the Middle East and possible Asia Minor, this is the way the Semitic loan words entered the IE languages.


JAL

[WeepingElf]

Well, as far as I have been able to Google (e.g here and here, but also Wikipedia), Proto-Semitic is dated to somewhere between 3000 and 4000 BC (with 3500BC more likely), and located in the Levant or Northern Mesopotamia (the first seeming more likely). This is much younger than PIE, and around the same time as the expansion of IE into Europe. So in order for a complete Semitic substrate in Europe for various IE languages, dating of PS needs to be wrong for a few millenia, something I find very doubtful. It seems far more likely that given the PS (or later Semitici languages) and IE languages must have been in close contact in the Middle East and possible Asia Minor, this is the way the Semitic loan words entered the IE languages.

Yes, Semitic is not any deeper than Indo-European. Proto-Semitic probably was contemporary to Late PIE, i.e. ca. 3500-3000 BC, and therefore it is extremely unlikely that any Semitic language was ever spoken in pre-IE Europe. However, Semitic is part of the larger Afro-Asiatic language family, which reaches much deeper into the past, with Proto-Afro-Asiatic having been spoken about 10,000 years ago, if not earlier. We cannot rule out that any Afro-Asiatic language was spoken in pre-IE Europe, but that is also not very likely, as the origin of Afro-Asiatic lies in Africa, with Semitic being the only branch of which we know that it ever left Africa.

The "goat" word and similar "Indo-Semitic" agricultural worlds may even have originated from an unknown, now extinct language of the Near East which was later displaced by Pre-Proto-Semitic, which borrowed the words itself. Agriculture in the Near East began about 12,000 years ago, at a time when Afro-Asiatic probably still was pretty much confined to its Ethiopian homeland. If that is the case, there is not the slightest shred of reason to assume a Semitic language in Neolithic Europe!

[jal]

[We cannot rule out that any Afro-Asiatic language was spoken in pre-IE Europe, but that is also not very likely, as the origin of Afro-Asiatic lies in Africa, with Semitic being the only branch of which we know that it ever left Africa.

Except, of course, if one accepts the "out of the Levant" theory of Afro-Asiatic origins. Which is, since it's a fringe theory, something I'd expect T. to be a proponent of :).


JAL

[Soap]

Well, I believe that it's been well established that Middle Easterners entered central Europe sometime around 5000 BC, far ahead of the Indo-Europeans, and introduced Neolithic technologies such as cattle farming and likely the practice of drinking cow's milk. We don't know what language they spoke, since it predates proto-Semitic as well, but the theory that they spoke a now-extinct branch of Afro-Asiatic is not out of the question. Indo-Europeans moving west could have borrowed words from them instead of from the Semites, which would explain why many of these words seem to appear only in the western languages (though obviously, that would not explain words that do appear in all branches, such as the numerals six and seven.)

[WeepingElf]

Well, I believe that it's been well established that Middle Easterners entered central Europe sometime around 5000 BC, far ahead of the Indo-Europeans, and introduced Neolithic technologies such as cattle farming and likely the practice of drinking cow's milk. We don't know what language they spoke, since it predates proto-Semitic as well, but the theory that they spoke a now-extinct branch of Afro-Asiatic is not out of the question. Indo-Europeans moving west could have borrowed words from them instead of from the Semites, which would explain why many of these words seem to appear only in the western languages (though obviously, that would not explain words that do appear in all branches, such as the numerals six and seven.)

I wouldn't say that this scenario was impossible, but I don't consider it likely that they spoke an Afro-Asiatic language. In my opinion, no Afro-Asiatic branch other than Semitic (which IMHO moved from Egypt to the Levant about 8,000 years ago) ever left Africa. Of course, it is uncertain where and when Proto-Afro-Asiatic was spoken, but Ethiopia about 10,000 years ago seems to be a good candidate. In my opinion, the Old European Hydronymy points towards the languages of the Central European Neolithic having been languages related to Indo-European, but branching off before the latter developed the ablaut system.

[Octavià]

Yes, Semitic is not any deeper than Indo-European. Proto-Semitic probably was contemporary to Late PIE, i.e. ca. 3500-3000 BC, and therefore it is extremely unlikely that any Semitic language was ever spoken in pre-IE Europe.

These dates are calculated (glottochronology?) from surviving (i.e. historical) branches. But it's also conceivable that the Semitic familiy had older branches now extinct.

However, Semitic is part of the larger Afro-Asiatic language family, which reaches much deeper into the past, with Proto-Afro-Asiatic having been spoken about 10,000 years ago, if not earlier.

I'm affraid even earlier than that. Standard chronologies for large (macro-)families are too low IMHO.

The "goat" word and similar "Indo-Semitic" agricultural worlds may even have originated from an unknown, now extinct language of the Near East which was later displaced by Pre-Proto-Semitic, which borrowed the words itself. Agriculture in the Near East began about 12,000 years ago, at a time when Afro-Asiatic probably still was pretty much confined to its Ethiopian homeland.

But its expansion into Anatolia and Eastern Europe dates several millenia afterwards.

IMHO, one of the these scenarios is possible:

1)This pre-Semitic language was carried to Europe by Neolithic farmers.

2)This pre-Semitic language was replaced by (pre-)Proto-Semitic, a language from which split proto-Semitic proper (i.e. the ancestor of the historical Semitic languages) and the language(s) carried to Europe by Neolithic farmers.

In both scenarios these words entered IE languages of Europe from some of the languages spoken there before IE or its nearest relatives (e.g. "Hesperic").

Well, I believe that it's been well established that Middle Easterners entered central Europe sometime around 5000 BC, far ahead of the Indo-Europeans, and introduced Neolithic technologies such as cattle farming and likely the practice of drinking cow's milk. We don't know what language they spoke, since it predates proto-Semitic as well, but the theory that they spoke a now-extinct branch of Afro-Asiatic is not out of the question.

IMHO, this pre-Semitic language was part of the Vasco-Caucasian macro-family. Notice that PIE has also Vasco-Caucasian loanwords and addition to the Semitic ones.

Indo-Europeans moving west could have borrowed words from them instead of from the Semites, which would explain why many of these words seem to appear only in the western languages (though obviously, that would not explain words that do appear in all branches, such as the numerals six and seven.)

IMHO, some of these loanwords happened before PIE stage while other happened later.

[Octavià]

In my opinion, the Old European Hydronymy points towards the languages of the Central European Neolithic having been languages related to Indo-European, but branching off before the latter developed the ablaut system.

But this theory doesn't exclude other languages being spoken there or in other parts of Europe. It looks like you reject the idea that non-IE languages were once spoken in Europe.

[Shm Jay]

It looks like you reject the idea that non-IE languages were once spoken in Europe.

I think he rejects the idea that anything useful can be said upon which to build theories
about languages which we know nothing about (and will never have data).

[Octavià]

I disagree. Although certainly we've got not much data, this isn't exactly the same thing as nothing. For example, Germanic languages have a noticeable non-IE substrate, but AFAIK very few linguists have afforded to study it. To me, it looks like substrate languages are the lumpenproletariat of historical linguistics.

About the question of a language related to Semitic once spoken in Europe, I could say that PGmc *sibun '7' doesn't come from PIE *septºm- but from the masculine (unsuffixed) variant of PSem *sabʕ-(at-) '7', while the PIE comes from the femenine (suffixed) one.