Zompist Bboard Again

Rounin Ryuuji wrote: ↑Fri Aug 27, 2021 1:58 pm I think I should like an explanation, too. Every language I've ever studied seems to have more than two thousand words of native vocabulary.

If the explanation involved or invokes Piraha, can I have a horse?

Simply. There's no way the +2000 lexical items reconstructed for "PIE" actually came at once from a single protolanguage.

Then where did they come from?

(from two protolanguages? twenty?)

bradrn wrote: ↑Wed Aug 25, 2021 9:00 am The full journal issue also looks fascinating, in that it contains both Forni’s article, several researchers’ rebuttals, then his reply. Unfortunately they have been very careful to paywall the whole thing, so I can’t assess it properly.

https://www.academia.edu/3811354/On_For ... JIES_2013_
https://www.academia.edu/4689618/_Why_B ... _2_203_237
https://www.academia.edu/2020451/Is_Bas ... o_isolates
http://euskararenjatorria.net/wp-conten ... n-JIES.pdf
http://euskararenjatorria.net/wp-conten ... h-JIES.pdf
http://euskararenjatorria.net/wp-conten ... s-JIES.pdf

hwhatting wrote: ↑Fri Aug 27, 2021 6:43 am

Talskubilos wrote: ↑Fri Aug 27, 2021 6:04 am Simply. There's no way the +2000 lexical items reconstructed for "PIE" actually came at once from a single protolanguage.

I don't follow your logic here. Natural languages have tens of thousands of words. The point with most proto-languages is rather the opposite - we are only able to reconstruct a small part of what was actually there, because languages constantly replace words and a large number of words that must have existed in PIE have been replaced in all daughter languages.

It isn't about quantity but internal and external relationships. For most IE-ists, it looks like PIE and its "daughter" languages existed within a bubble.

Talskubilos wrote: ↑Sat Aug 28, 2021 5:49 am

hwhatting wrote: ↑Fri Aug 27, 2021 6:43 am

Talskubilos wrote: ↑Fri Aug 27, 2021 6:04 am Simply. There's no way the +2000 lexical items reconstructed for "PIE" actually came at once from a single protolanguage.

I don't follow your logic here. Natural languages have tens of thousands of words. The point with most proto-languages is rather the opposite - we are only able to reconstruct a small part of what was actually there, because languages constantly replace words and a large number of words that must have existed in PIE have been replaced in all daughter languages.

It isn't about quantity but internal and external relationships. For most IE-ists, it looks like PIE and its "daughter" languages existed within a bubble.

I'm not sure whether to ask "What makes you feel that most IE-ists believe that?" (IE-ist??)...or to ask "Who believes that?"...or to ask "What bubble?"

can i ask them all?

bradrn wrote: ↑Fri Aug 27, 2021 6:36 amSure, no-one is claiming that every single reconstructed word was really PIE. On the other hand, that doesn’t suddenly make PIE disappear in a puff of smoke. There did exist a single language which was the last common ancestor of Hittite, Hindi, French and English, even if we’ve incorrectly reconstructed part of its lexicon.

Said that way, it sounds like a dogma.

bradrn wrote: ↑Fri Aug 27, 2021 6:36 am(And, if you’re inclined to quarrel with me on this point, we do know such a language existed because we can find regular and systematic phonological correspondences between each of these subgroups: Grimm’s law, satemisation and so on. Furthermore, and perhaps more importantly, there are significant morphological similarities between all these languages. The only explanation for such widespread correspondences is a single common ancestor.)

I'd say "several common ancestors" in a series of expansion and replacement processes over several millenia.

keenir wrote: ↑Sat Aug 28, 2021 8:42 amI'm not sure whether to ask "What makes you feel that most IE-ists believe that?" (IE-ist??)...or to ask "Who believes that?"...or to ask "What bubble?"

I think this forum is a good sample of what I referred to. Vertheless, the bubble is that "they" (i.e. these Indo-Europeanists) usualy ignore linguistic data outside the IE family, while IMHO long-range comparisons are mandatory in order to study and understand the complexity of IE.

It looks like IE femenine endings -ā where adapted as -e in Paleo-Basque (namely andere < *anderā) and Iberian (e.g. the toponyms Aŕse < *Arza or Salduie < Salduvia). Athough quite rare, this evolution can be also seen in Romance loanwords, as in e.g. pare 'shovel' < pala.

However, we've also got some former femenines in -o, as e.g. oilo 'hen' < Hispano-Romance polla 'young hen'. In fact, these can coexist with the former as variants of the same etymon, namely saldo 'herd, group' vs. (t)alde id., ultimately from IE *ḱerd-.

Talskubilos wrote: ↑Sat Aug 28, 2021 9:39 am However, we've also got some former femenines in -o, as e.g. oilo 'hen' < Hispano-Romance polla 'young hen'.

Wouldn't it be easier to posit that it was just a borrowing of pollo, with semantic narrowing?

Talskubilos wrote: ↑Sat Aug 28, 2021 5:49 am

hwhatting wrote: ↑Fri Aug 27, 2021 6:43 am

Talskubilos wrote: ↑Fri Aug 27, 2021 6:04 am Simply. There's no way the +2000 lexical items reconstructed for "PIE" actually came at once from a single protolanguage.

I don't follow your logic here. Natural languages have tens of thousands of words. The point with most proto-languages is rather the opposite - we are only able to reconstruct a small part of what was actually there, because languages constantly replace words and a large number of words that must have existed in PIE have been replaced in all daughter languages.

It isn't about quantity but internal and external relationships. For most IE-ists, it looks like PIE and its "daughter" languages existed within a bubble.

I actually find the opposite, that (certain) IE-ists will attribute any innovation found in Proto-Germanic, Proto-Greek, Proto-Celtic, etc, to substrate influence.

Talskubilos wrote: ↑Sat Aug 28, 2021 9:25 am

bradrn wrote: ↑Fri Aug 27, 2021 6:36 am(And, if you’re inclined to quarrel with me on this point, we do know such a language existed because we can find regular and systematic phonological correspondences between each of these subgroups: Grimm’s law, satemisation and so on. Furthermore, and perhaps more importantly, there are significant morphological similarities between all these languages. The only explanation for such widespread correspondences is a single common ancestor.)

I'd say "several common ancestors" in a series of expansion and replacement processes over several millenia.

In that case, how exactly do you explain all the phenomena I mentioned?

(Besides, I’m not even sure how a group of languages can have several common ancestors… by definition there is only one least common ancestor for any given group of languages.)

Rounin Ryuuji wrote: ↑Sat Aug 28, 2021 9:42 am

Talskubilos wrote: ↑Sat Aug 28, 2021 9:39 am However, we've also got some former femenines in -o, as e.g. oilo 'hen' < Hispano-Romance polla 'young hen'.

Wouldn't it be easier to posit that it was just a borrowing of pollo, with semantic narrowing?

A gender transitioning? I don't think so.

Why would Basque speakers care about adjective agreement in the source language when borrowing a word? It ends in -o, the language (presumably) has other feminine nouns ending in -o, erriplen goes into the vegetable gender.

Ketsuban wrote: ↑Sat Aug 28, 2021 9:39 pmWhy would Basque speakers care about adjective agreement in the source language when borrowing a word? It ends in -o, the language (presumably) has other feminine nouns ending in -o, erriplen goes into the vegetable gender.

The thing is Basque has no grammatical gender at all.

Talskubilos wrote: ↑Sat Aug 28, 2021 9:29 am

keenir wrote: ↑Sat Aug 28, 2021 8:42 amI'm not sure whether to ask "What makes you feel that most IE-ists believe that?" (IE-ist??)...or to ask "Who believes that?"...or to ask "What bubble?"

I think this forum is a good sample of what I referred to. Vertheless, the bubble is that "they" (i.e. these Indo-Europeanists) usualy ignore linguistic data outside the IE family,

perhaps you can explain how linguistic data from, say the Piraha or Japanese languages, is relevant to PIE.

while IMHO long-range comparisons are mandatory in order to study and understand the complexity of IE.

long-range?

bradrn wrote: ↑Sat Aug 28, 2021 6:31 pm

Talskubilos wrote: ↑Sat Aug 28, 2021 9:25 am

bradrn wrote: ↑Fri Aug 27, 2021 6:36 am(And, if you’re inclined to quarrel with me on this point, we do know such a language existed because we can find regular and systematic phonological correspondences between each of these subgroups: Grimm’s law, satemisation and so on. Furthermore, and perhaps more importantly, there are significant morphological similarities between all these languages. The only explanation for such widespread correspondences is a single common ancestor.)

I'd say "several common ancestors" in a series of expansion and replacement processes over several millenia.

In that case, how exactly do you explain all the phenomena I mentioned?

(Besides, I’m not even sure how a group of languages can have several common ancestors… by definition there is only one least common ancestor for any given group of languages.)

I'm starting to wonder if Talskubilos is conflating linguistic descent with ancestry...like how I have a grandmother on each side of my family, therefore English (and every other language) has to have at least two grandmother languages leading up to it.

Talskubilos wrote: ↑Sat Aug 28, 2021 10:50 pm

Ketsuban wrote: ↑Sat Aug 28, 2021 9:39 pmWhy would Basque speakers care about adjective agreement in the source language when borrowing a word? It ends in -o, the language (presumably) has other feminine nouns ending in -o, erriplen goes into the vegetable gender.

The thing is Basque has no grammatical gender at all.

this reinforces the question of why would they care about feminine agreement.

keenir wrote: ↑Sat Aug 28, 2021 11:00 pm

bradrn wrote: ↑Sat Aug 28, 2021 6:31 pm

Talskubilos wrote: ↑Sat Aug 28, 2021 9:25 am I'd say "several common ancestors" in a series of expansion and replacement processes over several millenia.

In that case, how exactly do you explain all the phenomena I mentioned?

(Besides, I’m not even sure how a group of languages can have several common ancestors… by definition there is only one least common ancestor for any given group of languages.)

I'm starting to wonder if Talskubilos is conflating linguistic descent with ancestry...like how I have a grandmother on each side of my family, therefore English (and every other language) has to have at least two grandmother languages leading up to it.

Thankfully, even the worst of linguistic crackpots don’t tend to make this mistake. Rather, I suspect he is conflating loaning and stratum influences with ancestry: IE branches have lots of loanwords from substrates, therefore IE cannot exist.

bradrn wrote: ↑Sat Aug 28, 2021 11:16 pmThankfully, even the worst of linguistic crackpots don’t tend to make this mistake. Rather, I suspect he is conflating loaning and stratum influences with ancestry: IE branches have lots of loanwords from substrates, therefore IE cannot exist.

But these "substrates" are still a very large part of the IE lexicon, so IMHO the std PIE model is an oversimplification.

keenir wrote: ↑Sat Aug 28, 2021 11:00 pmthis reinforces the question of why would they care about feminine agreement.

But this doesn't explain why the original femenine suffix gave -o or -e instead of the more widespread -a.

Talskubilos wrote: ↑Sun Aug 29, 2021 4:06 am

bradrn wrote: ↑Sat Aug 28, 2021 11:16 pmThankfully, even the worst of linguistic crackpots don’t tend to make this mistake. Rather, I suspect he is conflating loaning and stratum influences with ancestry: IE branches have lots of loanwords from substrates, therefore IE cannot exist.

But these "substrates" are still a very large part of the IE lexicon, so IMHO the std PIE model is an oversimplification.

So what? No-one is claiming that ‘X is a PIE language’ means that ‘every X word is descended from PIE’, or even that ‘most X words are directly inherited from PIE’ (cf Armenian). Rather, the best definition of language classification I’ve yet seen is that given in the Koch paper linked above:

Koch wrote: A language (or dialect) Y at a given time is said to be descended from language (or dialect) X of an earlier time if and only if X developed into Y by an unbroken sequence of instances of native-language acquisition by children.

According to this definition — which as far as I can see seems to be consensus amongst historical linguists — a language could be PIE even if most of its vocabulary were to be non-PIE, as long as it has inherited PIE vocabulary and morphology through native-language acquisition.

bradrn wrote: ↑Sun Aug 29, 2021 4:17 amSo what? No-one is claiming that ‘X is a PIE language’ means that ‘every X word is descended from PIE’, or even that ‘most X words are directly inherited from PIE’ (cf Armenian). Rather, the best definition of language classification I’ve yet seen is that given in the Koch paper linked above:

Koch wrote: A language (or dialect) Y at a given time is said to be descended from language (or dialect) X of an earlier time if and only if X developed into Y by an unbroken sequence of instances of native-language acquisition by children.

According to this definition — which as far as I can see seems to be consensus amongst historical linguists — a language could be PIE even if most of its vocabulary were to be non-PIE, as long as it has inherited PIE vocabulary and morphology through native-language acquisition.

The thing is the +2000 lexical items commonly reconstructed for PIE can't be from a single language. On the other hand, IE morphology (espceially the verbal ones) doesn't fit very well into the single common ancestor/classical genealogical tree model.

Talskubilos wrote: ↑Sun Aug 29, 2021 4:51 am

bradrn wrote: ↑Sun Aug 29, 2021 4:17 amSo what? No-one is claiming that ‘X is a PIE language’ means that ‘every X word is descended from PIE’, or even that ‘most X words are directly inherited from PIE’ (cf Armenian). Rather, the best definition of language classification I’ve yet seen is that given in the Koch paper linked above:

Koch wrote: A language (or dialect) Y at a given time is said to be descended from language (or dialect) X of an earlier time if and only if X developed into Y by an unbroken sequence of instances of native-language acquisition by children.

According to this definition — which as far as I can see seems to be consensus amongst historical linguists — a language could be PIE even if most of its vocabulary were to be non-PIE, as long as it has inherited PIE vocabulary and morphology through native-language acquisition.

The thing is the +2000 lexical items commonly reconstructed for PIE can't be from a single language.

No-one seriously claims they are. Even the people who made the reconstruction usually admit that it’s an imperfect process. (And if they don’t, they’re a crackpot.)

On the other hand, IE morphology (espceially the verbal ones) doesn't fit very well into the single common ancestor/classical genealogical tree model.

How so? I’d be interested to see a specific example.