The Great Proto-Indo-European Thread's Sequel

[Moose-tache]

The one time a computer just up and died on me, it gave me ample warning by being an HP.

As for the "this random word looks like this other random word" discourse, I thought we were past all that. If you want to demonstrate that something was loaned from language A to language B, you have to address the null hypothesis, i.e. you need to disprove the person in the back row raising their hand to say "OK, cool, but what if not that?" If you can't shut them down, you have no case. If we are looking at individual words out of thousands, there is no chance of disproving the null hypothesis without some supporting context. As people have pointed out, it would be more convincing if it were systematic. But as it is, it's just gaijin-goyim all over again.

[WeepingElf]

Well, etymology is always guesswork and speculation as there is no sure way knowing where the word actually comes from, though when there is a word in a clearly related language that matches nicely in terms of the known sound correspondences, one can be quite certain that it is cognate. Beyond that, it's anyone's guess. And it is even worse with names where we don't even know the original meaning - though even here, there are cases where an element in some place names correlate with a salient feature of the named sites, such as *hal- in the names of some ancient Central European salt production sites which therefore probably meant 'salt' in whichever language it came from.

[Zju]

I feel like it has to count against loanword hypotheses for 7 that other loanword hypotheses for Indo-European numerals don't seem to hold much water. Blažek reviews them in passing for every numeral and only fails to reject any for 7 because there is no consensus on an internal derivation (the best he can find is the idea that *septḿ̥ is clipped from *septm̥mo-, a superlative derived from the root of Sanskrit सपति "worships" and Greek ἕπω which he reconstructs as 'the most honourable' and gestures to "the prominent position of the numeral 7 among Indo-Europeans").

Wow, thanks for the great source! For quite a while I was curious what ever possibly a native etymology for *ksweks '6' could be. I'll eagerly read up the other suggestions, especially the lesser numerals.
Though I'm not sure I follow your line of thought. Surely it's possible that just a single numeral is a loanword? And no native etymology being proposed by such an extensive source and review of the earlier literature surely gives some weight to the notion that *septm is a loanword.

[abahot]

Though I'm not sure I follow your line of thought. Surely it's possible that just a single numeral is a loanword? And no native etymology being proposed by such an extensive source and review of the earlier literature surely gives some weight to the notion that *septm is a loanword.

I think that the lack of a native etymology for the word for "seven" isn't really that much evidence. Proto-Indo-European is a language with just as much of a complex and layered history as English, and the lack of many (or any?) apparent etymologies for English numerals does not mean they are loanwords. It could well be that, just as for all numerals in English, the ultimate etymology of "seven" in PIE is just very deep and layered.

[WeepingElf]

Proto-Indo-European is a language with just as much of a complex and layered history as English, [...]

Yep. This is often forgotten. Some people treat PIE as a pristine, original language, which it was most emphatically NOT.

[hwhatting]

though even here, there are cases where an element in some place names correlate with a salient feature of the named sites, such as *hal- in the names of some ancient Central European salt production sites which therefore probably meant 'salt' in whichever language it came from.

Actually, no serious IEanist assumes that anymore already for a long time. The current assumption is that it has something to so with salt production. What exactly, and what the exact etymology is, is debated; In this article the Celtologist David Stifter makes his own proposal for a (Germanic) etymology - which you can take or leave, but he gives a good overview over the question and the various proposals made.

[WeepingElf]

though even here, there are cases where an element in some place names correlate with a salient feature of the named sites, such as *hal- in the names of some ancient Central European salt production sites which therefore probably meant 'salt' in whichever language it came from.

Actually, no serious IEanist assumes that anymore already for a long time. The current assumption is that it has something to so with salt production. What exactly, and what the exact etymology is, is debated; In this article the Celtologist David Stifter makes his own proposal for a (Germanic) etymology - which you can take or leave, but he gives a good overview over the question and the various proposals made.

Thank you for sharing. This is a valid objection; but the author knocks at an open door when he points out that the *s > h shift did not happen in Continental Celtic - I don't think it has anything to do with Celtic. What this shows is how much the etymological analysis of names is even more guesswork than etymology in general because we don't know the original meanings of the names. Also, names may spread beyond the range of the original language by migrants, as shown by the numerous names of Arabic origin in Latin America, which were brought along by Spanish and Portuguese settlers naming places in their new homeland after ones of their own homeland. I have coined the term name protraction (German Namenverschleppung) for this.

[abahot]

I think that the lack of a native etymology for the word for "seven" isn't really that much evidence. Proto-Indo-European is a language with just as much of a complex and layered history as English, and the lack of many (or any?) apparent etymologies for English numerals does not mean they are loanwords. It could well be that, just as for all numerals in English, the ultimate etymology of "seven" in PIE is just very deep and layered.

Well, on the one hand, the fact that seven is the only numeral without an etymology might be some better evidence than in English where no numeral has one. On the other hand, I am very skeptical of picking apart roots for numerals into etymologies which may well be pure chance.

[Travis B.]

I have coined the term name protraction (German Namenverschleppung) for this.

I just googled Namenverschleppung and came out with some uses in publications of this very word.

[hwhatting]

Thank you for sharing. This is a valid objection; but the author knocks at an open door when he points out that the *s > h shift did not happen in Continental Celtic - I don't think it has anything to do with Celtic.

Well, it being Celtic was the original idea behind it being a substrate word for "salt".

What this shows is how much the etymological analysis of names is even more guesswork than etymology in general because we don't know the original meanings of the names.

Agreed.
But anyway, the main point is that the Hall- names are limited to, but wide-spread in, the German-speaking area, which make a Germanic etymology much more likely than a substrate origin - why would that show such a distribution?

[WeepingElf]

Thank you for sharing. This is a valid objection; but the author knocks at an open door when he points out that the *s > h shift did not happen in Continental Celtic - I don't think it has anything to do with Celtic.

Well, it being Celtic was the original idea behind it being a substrate word for "salt".

What this shows is how much the etymological analysis of names is even more guesswork than etymology in general because we don't know the original meanings of the names.

Agreed.
But anyway, the main point is that the Hall- names are limited to, but wide-spread in, the German-speaking area, which make a Germanic etymology much more likely than a substrate origin - why would that show such a distribution?

A good point.

[keenir]

What this shows is how much the etymological analysis of names is even more guesswork than etymology in general because we don't know the original meanings of the names.

Agreed.
But anyway, the main point is that the Hall- names are limited to, but wide-spread in, the German-speaking area, which make a Germanic etymology much more likely than a substrate origin - why would that show such a distribution?

hmmm...offhand, I'd guess that a word (proto-Hall, basically) was picked up from the substrate and, modified into Hall- by the Germanic speakers, who spread it throughout what was the Germanic-speaking world.

The analogy that comes to my mind, is if a Basque word was borrowed into Latin, got changed a little so it was easier to make it behave like a proper Latin word, and then spread throughout the empire over the course of history. If we didn't know Basque as well as we do, we might not know where the word came to Latin from (and with how it changed some, there might be some who argue that it can't possibly be a true Basque word)

just a thought.

[WeepingElf]

I have noticed an issue with Stifter's idea. The place names with the element *hal(l) do not all refer to sites where salt was produced by boiling brine, but also to salt mines (e.g. Hallstatt). This doesn't strictly mean that Stifter was wrong (there may have been a sematic shift from 'crust forming in boiling brine' to just 'salt', or something - but why then did the word only survive in place names?), but I think this point has to be addressed and thought about.

[hwhatting]

hmmm...offhand, I'd guess that a word (proto-Hall, basically) was picked up from the substrate and, modified into Hall- by the Germanic speakers, who spread it throughout what was the Germanic-speaking world.

Little nitpick, it isn't even Germanic, it's purely the German (= High & Low German, Dutch / Flemish) dialect continuum. (I will use "German" as a shorthand for that below - excuses aan alle Nederlanders).

Maybe a bit about the history of the substrate idea. The original idea was that Hall- was a Celtic substrate word meaning "salt", as some Celtic varieties (Brythonic) have a development PIE -> Proto-Celtic /*s/ -> Brythonic /*h-/, so Hall- was assumed to be the continuation of PIE /*sal-/ "salt" in this variety of Celtic. When it was established that there's no evidence *#s -> #h ever happened in Continental Celtic, and that the Celtic languages use a different formation from that PIE root, what remained of the substrate theory was instead of "Hall-, which shows up in placenames having to do with salt production in areas that spoke Celtic before they became Germanic and the word looks like the Celtic continuer of the PIE "salt" word, and it has been preserved in those place names because the incoming Germanic tribes adapted the place names from the Celts", instead of that we have Hall- is maybe a substrate word meaning "salt" from some unkown, possibly IE language. So instead of something which, at the time, looked interesting, because it not only said "some substrate", but identified a donor language (family) that was known to have been spoken in the area, and gave an etymology, we now just have something that's vague and meh.

Now, the substrate idea still usually assumes that the word wasn't loaned into German as a concept meaning "salt" (German, like the other Germanic languages, has its own IE word for "salt" - German Salz, Dutch zout, and there is not a single dialect that has a *Hall- "salt"), but that it survived in place names. Which then raises the question, why only in the German-speaking dialect continuum? And the consensus is nowadays that the word means something to do with salt production, and that there's no specific reason to assume a substrate origin. The substrate theory is just a ghost from an earlier stage of the research history.

[keenir]

hmmm...offhand, I'd guess that a word (proto-Hall, basically) was picked up from the substrate and, modified into Hall- by the Germanic speakers, who spread it throughout what was the Germanic-speaking world.

Little nitpick, it isn't even Germanic, it's purely the German (= High & Low German, Dutch / Flemish) dialect continuum. (I will use "German" as a shorthand for that below - excuses aan alle Nederlanders).

The substrate theory is just a ghost from an earlier stage of the research history.

ah, okay. I had presumed more still existing to some degree however small to the substrate theory when I'd offered that. And thanks for clarifying its German and not Germanic; mea culpa.

[WeepingElf]

Indeed, substratum theories that don't even specify the substratum are cases of ignotum per ignotius and thus not particularly useful. What do you gain by saying, "Word X is from a substratum language, but we don't know which language"? Such a statement is not falsifiable and just a fancy way of saying "The origin of word X is unknown". Substratum theories were much en vogue in Celtic and Romance linguistics in the early 20th century, but are now largely out of fashion because few people believe in Semitic languages in the pre-Celtic British Isles anymore, and correlations between Romance isoglosses and the pre-Roman linguistic landscape have turned out to be imaginary. Which doesn't mean that they can't be right, but we just don't know. They may be useful as inspiration for conlangs that represent lost linguistic lineages, but not really for elucidating something real.

[Zju]

Dunno if this has been discussed already here, but PIE *Hebl- 'apple' may not be a loanword at all, and instead just be a metastethised form of *meHlom. Can't track down the paper from academia.edu I originally read that in, but the gist is that an intermediate form *Heml- underwent ml → bl. Also, an s-mobile form of *meHlom allegedly gave rise to proto-Kartvelian *msxali 'pear' via **smxali < **smaxli.

[Travis B.]

Dunno if this has been discussed already here, but PIE *Hebl- 'apple' may not be a loanword at all, and instead just be a metastethised form of *meHlom. Can't track down the paper from academia.edu I originally read that in, but the gist is that an intermediate form *Heml- underwent ml → bl. Also, an s-mobile form of *meHlom allegedly gave rise to proto-Kartvelian *msxali 'pear' via **smxali < **smaxli.

Does the author mention anything about a *m > *b change in other positions that would explain Burushaski báalt?

[Zju]

No, pretty sure it was positted to occur only in front of *l - something with related with the phonetics (and/or phonotactics) of the sonorant cluster [ml] made it likely to change to bl. IIRC there were examples given of ml > (m)bl in other languages. If only I could find the paper, I'd cite them.

Idly speculating, the phonetics of it may not be too different from the English sound change ∅ > p / m_C : emty, hamster → empty, hampster

[Ketsuban]

Might the paper in question be Fenwick (2016)?