Paleo-European languages

[Zju]

Yes, this was how I understood that diagram as well (though I hadn’t heard of the NWB before). But none of that suddenly means that Indo–European is no longer a tree!

But the thing is the genealogical tree doesn't represent areal influences between IE branches nor the various pre-IE languages surviving as substrates.

The heavy Scandinavian influence on English might be shown as a hybridisation of Old Danish and Old English, or more precisely, Anglian Old English. I'm not saying Tavi has given this episode any attention, but he may hypothesise similar situations.

That's right. This is reflected in doublets such as shirt/skirt and many more.

The role of the genealogical tree isn't to represent areal influences - that's a fundamental misunderstanding of what it is. It is supposed to represent which languages descend from which other.

Areal influence is a different matter from genealogy. So are lexical borrowings. English has plenty of borrowings from Middle French and Old Norse, but it still is a West Germanic language descending from Old English. No IE linguist denies that there was a great deal of areal influence between various IE languages.

You can argue that the comb model or the wave model describe IE languages better than the tree model, and that's fine - and yet all IE languages descend from a single dialect continuum that we label as PIE.

[WeepingElf]

Andrew & Susan Sherratt: The archaeology of Indo-European. An alternative view

A paper from 1988, and since then a lot has happened - a whole new science has arisen: paleogenetics. That said, this paper is very similar to what I called the "Europic" theory before I abandoned it, the main difference being that the Sheratts have Pre-PIE originate in NW Anatolia while I put it on the northern shore of the Black Sea before the Black Sea Flood. It has since turned out that the early European farmers are genetically and archaeologically unrelated to the Yamnaya culture, who probably spoke PIE; so why should they have spoken related languages? (Also, the Black Sea Flood has turned out to be spurious - Ryan and Pitman apparently just misinterpreted their data.)

But the idea of a whole lost branch of IE in western Europe may have something to it, if such things like the "Old European Hydronymy" are real. I now think that this branch ("Aquan") was associated with the Bell Beaker culture. The idea is that there were two migrations out of the PIE homeland, corresponding to Gimbutas's second and third "Kurgan waves" (the first did not happen at all, but was a wave of endogenic cultural changes caused by the diffusion of metallurgy): 1. An earlier one along the Danube river into the Balkan Peninsula and Central Europe; from this descend "Aquan" and Anatolian. 2. A later one north of the Carpathians, associated with the Corded Ware culture; from this descend all the other IE languages.

Phylogenetic and areal models of Indo-European relatedness: The role of contact in reconstruction

An interesting paper, but I don't see how it supports your position - just the wave model that is gaining popularity among historical linguists, for good reasons. It IMHO does give a better account of what happens in expanding language families than the tree model. Or did I miss something?

[Richard W]

But the thing is the genealogical tree doesn't represent areal influences between IE branches nor the various pre-IE languages surviving as substrates.

The heavy Scandinavian influence on English might be shown as a hybridisation of Old Danish and Old English, or more precisely, Anglian Old English. I'm not saying Tavi has given this episode any attention, but he may hypothesise similar situations.

That's right. This is reflected in doublets such as shirt/skirt and many more.

So are lexical borrowings. English has plenty of borrowings from Middle French and Old Norse, but it still is a West Germanic language descending from Old English.

Some see West Germanic as simply that portion of North-West Germanic that is not North Germanic. It's more an areal group than a genealogical group.

The reason for seeing English as a hybrid language is that the loans extend into the grammatical region. The most notorious example is they, but it also includes them 'those' and retains the preposition at, otherwise restricted to North Germanic. Even deeper is the 3s ending -s (but possibly derived from Northumbrian Old English), which extends to all persons in some dialects. I'm tempted to cite the group genitive (known at least in Swedish), but that looks like a parallel development, so it is the inducing conditions that are shared.

[Nortaneous]

The reason for seeing English as a hybrid language is that the loans extend into the grammatical region. The most notorious example is they, but it also includes them 'those' and retains the preposition at, otherwise restricted to North Germanic.

Norse origin for "they" is disputed - see e.g. doi:10.1075/dia.16026.col

[Moose-tache]

It is uncommon to borrow pronouns, but I don't know if that negates a language's placement on a tree diagram. For one thing, these are clearly borrowings, i.e. words entering a new speech community, not two speech communities being mixed together. English speakers would have learned "skirt" from their Anglo-Saxon mothers the same way they learned the word "shirt." For another, many of these items are disputed. there is no evidence that "the" comes from Norse, so the presence of a voiced initial dental fricative does not automatically mean we are dealing with a Norse borrowing.

As for the "thirteenth tribe" of lost PIE speakers, this is perfectly possible, but evidence is obviously limited. The Basque, genetically, are much more closely related to Yamnaya individuals than the Neolithic farmers who entered southern Europe earlier. This would suggest that the Basque's ancestors moved into western Europe as a "bow wave" ahead of other migrations from the steppe into the continent. Other groups in this bow wave could have spoken dialects of PIE. But to demonstrate that, you'd need a hell of a lot more than some similar words in random classical languages, or similar words for "river."

[Richard W]

The reason for seeing English as a hybrid language is that the loans extend into the grammatical region. The most notorious example is they, but it also includes them 'those' and retains the preposition at, otherwise restricted to North Germanic.

Norse origin for "they" is disputed - see e.g. doi:10.1075/dia.16026.col

Show me.

The most complete account I can find is https://www.quora.com/What-is-the-etymology-of-they. The explanation of the vowel is distinctly hand-wavy, and the history of the vowel is the key point. While it's reassuring to see the clearly Old English demonstrative used where one might expect the pronoun, that's hardly big news, for she appears to originate from the demonstrative, as is typical of modern West Germanic.

Old English þæg and þage do look like the ancestors of SW English demonstrative they, but I don't think they will account for the spelling history of the pronoun they.

[Talskubilos]

You can argue that the comb model or the wave model describe IE languages better than the tree model, and that's fine - and yet all IE languages descend from a single dialect continuum that we label as PIE.

OK, let's put it another way: the "real" PIE would be only a (comparatively small) subset of the reconstructed PIE, whose +2000 lexical items come mostly from the pre-existing languages (the ones labelled with a cross in the diagram).



The good news is that these items can be identified and classified thanks to sound correspondences with external data.

[bradrn]

The reason for seeing English as a hybrid language is that the loans extend into the grammatical region. The most notorious example is they, but it also includes them 'those' and retains the preposition at, otherwise restricted to North Germanic.

Norse origin for "they" is disputed - see e.g. doi:10.1075/dia.16026.col

Show me.

He did: see https://www.jbe-platform.com/content/jo ... .16026.col (that’s where the doi link points).

You can argue that the comb model or the wave model describe IE languages better than the tree model, and that's fine - and yet all IE languages descend from a single dialect continuum that we label as PIE.

OK, let's put it another way: the "real" PIE would be only a (comparatively small) subset of the reconstructed PIE, whose +2000 lexical items come mostly from the pre-existing languages (the ones labelled with a cross in the diagram).

OK, this view makes a lot more sense to me. You’re not claiming that PIE never existed, or that IE isn’t a coherent family; you’re simply claiming that PIE is over-reconstructed. And, as far as I can see, that’s a perfectly legitimate opinion to have! (Possibly a wrong one, but to me it certainly seems like a reasonable opinion.) I’d note, though, that this doesn’t suddenly disqualify the tree model — in your model, IE still fits on a classical tree, it just has a lot of loanwords.

[Ares Land]

You can argue that the comb model or the wave model describe IE languages better than the tree model, and that's fine - and yet all IE languages descend from a single dialect continuum that we label as PIE.

OK, let's put it another way: the "real" PIE would be only a (comparatively small) subset of the reconstructed PIE, whose +2000 lexical items come mostly from the pre-existing languages (the ones labelled with a cross in the diagram).



The good news is that these items can be identified and classified thanks to sound correspondences with external data.

One problem with that is that it implies an extremely high level of borrowing, not unheard of, but still quite rare.
Another problem is that as a theory, I don't really see that it has more explanatory power than more standard ones.

[Richard W]

Norse origin for "they" is disputed - see e.g. doi:10.1075/dia.16026.col

Show me.

He did: see https://www.jbe-platform.com/content/jo ... .16026.col (that’s where the doi link points).

That leads to a paywall.

[Richard W]

It is uncommon to borrow pronouns, but I don't know if that negates a language's placement on a tree diagram. For one thing, these are clearly borrowings, i.e. words entering a new speech community, not two speech communities being mixed together. English speakers would have learned "skirt" from their Anglo-Saxon mothers the same way they learned the word "shirt." For another, many of these items are disputed. there is no evidence that "the" comes from Norse, so the presence of a voiced initial dental fricative does not automatically mean we are dealing with a Norse borrowing.

The word with disputed origin is the pronoun they, not the.

I don't think you should overlook how some native English speakers learnt their English from Danish mothers! Where do you get your information on speech community structure?

[Moose-tache]

It is uncommon to borrow pronouns, but I don't know if that negates a language's placement on a tree diagram. For one thing, these are clearly borrowings, i.e. words entering a new speech community, not two speech communities being mixed together. English speakers would have learned "skirt" from their Anglo-Saxon mothers the same way they learned the word "shirt." For another, many of these items are disputed. there is no evidence that "the" comes from Norse, so the presence of a voiced initial dental fricative does not automatically mean we are dealing with a Norse borrowing.

The word with disputed origin is the pronoun they, not the.

I don't think you should overlook how some native English speakers learnt their English from Danish mothers! Where do you get your information on speech community structure?

What I said was "There is no evidence that "the" comes from Norse, so the presence of a voiced initial dental fricative does not automatically mean we are dealing with a Norse borrowing." To break that down, I am saying that because "the" is not Norse, we do not have to assume Norse origin to explain the initial voicing of "they." There are already examples of initial voicing of the dental fricative in native words, so "they" does not cry out for any special explanation.

[Richard W]

There are already examples of initial voicing of the dental fricative in native words, so "they" does not cry out for any special explanation.

It is not the voicing, but the vowel that calls out for explanation. Though the reflexes of Old English þæg(e)/þāg (and of þā in the north) and Old Danish þei (in English) eventually merge, they should have different Middle English reflexes, should they not?

[Talskubilos]

One problem with that is that it implies an extremely high level of borrowing, not unheard of, but still quite rare. Another problem is that as a theory, I don't really see that it has more explanatory power than more standard ones.

On the contrary, I think the study of pre-IE languages and the relationship between IE and other language families are topics neglected by most Indo-Europeanists.

Getting back to my former example, the correspondence between IE *don- 'reed' ~ *ned-o- 'reed, rush' and East Caucasian *ttsˀwǝ̄nHē/*Hnǝ̄ttsˀwē 'reed, cane' (§3.15 in Starostin's article) would correspond to PIE, whereas *yoini- 'reed' would be a loanword, with ttsˀw ~ yo. Different sound correspondences => different lexicon layers.

[WeepingElf]

One problem with that is that it implies an extremely high level of borrowing, not unheard of, but still quite rare. Another problem is that as a theory, I don't really see that it has more explanatory power than more standard ones.

On the contrary, I think the study of pre-IE languages and the relationship between IE and other language families are topics neglected by most Indo-Europeanists.

True. Some Indo-Europeanists treat PIE as if it had fallen from the sky fully formed, though none except a few creationists or other crackpots actually believe so. They just say, "We don't know" - but don't try to change that. The pre-IE languages of Europe are seriously underexplored. But this is not an easy task; it is hard because there is so little that is certain here. This requires the full methodological rigour of conventional historical linguistics; just fishing dictionaries for wild lexical "correspondences" won't do the job.

Getting back to my former example, the correspondence between IE *don- 'reed' ~ *ned-o- 'reed, rush' and East Caucasian *ttsˀwǝ̄nHē/*Hnǝ̄ttsˀwē 'reed, cane' (§3.15 in Starostin's article) would correspond to PIE, whereas *yoini- 'reed' would be a loanword, with ttsˀw ~ yo. Different sound correspondences => different lexicon layers.

Nonsense.

[Talskubilos]

True. Some Indo-Europeanists treat PIE as if it had fallen from the sky fully formed, though none except a few creationists or other crackpots actually believe so. They just say, "We don't know" - but don't try to change that. The pre-IE languages of Europe are seriously underexplored. But this is not an easy task; it is hard because there is so little that is certain here. This requires the full methodological rigour of conventional historical linguistics; just fishing dictionaries for wild lexical "correspondences" won't do the job.

I agree, but I do nothing of the kind.

[Zju]

But the thing is the genealogical tree doesn't represent areal influences between IE branches nor the various pre-IE languages surviving as substrates.


That's right. This is reflected in doublets such as shirt/skirt and many more.

So are lexical borrowings. English has plenty of borrowings from Middle French and Old Norse, but it still is a West Germanic language descending from Old English.

Some see West Germanic as simply that portion of North-West Germanic that is not North Germanic. It's more an areal group than a genealogical group.

The reason for seeing English as a hybrid language is that the loans extend into the grammatical region. The most notorious example is they, but it also includes them 'those' and retains the preposition at, otherwise restricted to North Germanic. Even deeper is the 3s ending -s (but possibly derived from Northumbrian Old English), which extends to all persons in some dialects. I'm tempted to cite the group genitive (known at least in Swedish), but that looks like a parallel development, so it is the inducing conditions that are shared.

I don't know that loaning of a most of dozen pronouns and suffixes can be called hybridisation. If anything, borrowing a major part of the lexicon would be more of a reason. But I haven't seen any linguist claiming that something other than pidginisation, creolisation and maybe koineisation changes the language's genealogy.

If West Germanic turns out to not be a valid monophyletic node, that's fair, but beside the point.

[Talskubilos]

Getting back to my former example, the correspondence between IE *don- 'reed' ~ *ned-o- 'reed, rush' and East Caucasian *ttsˀwǝ̄nHē/*Hnǝ̄ttsˀwē 'reed, cane' (§3.15 in Starostin's article) would correspond to PIE, whereas *yoini- 'reed' would be a loanword, with ttsˀw ~ yo. Different sound correspondences => different lexicon layers.

Nonsense.

Mallory & Adams (2006) list *yoini- as a regional NW word, while *nedo- would be a native PIE word. By contrast, *don- is listed as a "Western-Central" one, because it's only attested in Greek and Baltic. However, comparison with the Caucasian protoform show us it derives from a metathesized variant of the former.

[Znex]

Getting back to my former example, the correspondence between IE *don- 'reed' ~ *ned-o- 'reed, rush' and East Caucasian *ttsˀwǝ̄nHē/*Hnǝ̄ttsˀwē 'reed, cane' (§3.15 in Starostin's article) would correspond to PIE, whereas *yoini- 'reed' would be a loanword, with ttsˀw ~ yo. Different sound correspondences => different lexicon layers.

Nonsense.

Mallory & Adams (2006) list *yoini- as a regional NW word, while *nedo- would be a native PIE word. By contrast, *don- is listed as a "Western-Central" one, because it's only attested in Greek and Baltic. However, comparison with the Caucasian protoform show us it derives from a metathesized variant of the former.

I think WeepingElf is referring to how there's no obvious correspondence between these IE and East Caucasian words at all. For a lot of the correspondences you have proposed, it is rather easy to find way better matches, whether in East Caucasian or in any other outlier languages.

If there is any obvious correspondence in your eyes, you'll have to do better than giving us only two or four words to compare. Show us regular phonetic correspondences.

[Nortaneous]

There are already examples of initial voicing of the dental fricative in native words, so "they" does not cry out for any special explanation.

It is not the voicing, but the vowel that calls out for explanation. Though the reflexes of Old English þæg(e)/þāg (and of þā in the north) and Old Danish þei (in English) eventually merge, they should have different Middle English reflexes, should they not?

From the paper:

For the period 1150–1325, LAEME records a ‘thay’ type for they in -ai(e), -ay(e) in the North and in a few texts in the Northeast and Northwest Midlands and in the Southwest. A ‘þa/ta’ type is attested in the North and West Midlands. A ‘they’ type in -ei-, -ey- (including þeȝȝ found in the Ormulum) is used mainly in the East Midlands, and to a lesser extent in the West Midlands, but only occurs in one northern text. A ‘þe, ye, ȝe’ type is also used in the East Midlands and in one West Midlands text. A ‘þi’ type is attested in one northern text.

For their, a ‘their’ type with medial -ei- occurs only in the East Midlands and includes þeȝȝre found in the Ormulum. A ‘thair’ type with medial -ai/ay- (including yaier) predominates in the North but also occurs in two East Midland texts and in one West Midlands text. Monophthongal possessive forms are attested, including a ‘thar’ type with medial -a- attested only in the North and in one Northeast Midlands text and a ‘ther’ type found in the North and East Midlands.

For them, the North has a ‘thaim’ type with medial -ay, -ai- (and yaem) and ‘tham’ with monophthongal medial a. A ‘theim’ type with medial -ey, -ei- (including the þeȝȝm forms of the Ormulum) occurs only in the East Midlands, and a ‘them’ type involving yem, yem and yeem occurs in one northern text and in two texts in the East Midlands.