The Great Proto-Indo-European Thread's Sequel

[Talskubilos]

So where's the extraordinary evidence?

What kind of "extraordinary evidence" would convince you? Not a rethorical question.

[Skookum]

There's no reason why we shouldn't expect proto-languages to contain loanwords, and I'm not sure why that would invalidate the existence of a single PIE that all IE languages descend from...

[WeepingElf]

So where's the extraordinary evidence?

What kind of "extraordinary evidence" would convince you? Not a rethorical question.

Surely, you are bluffing!

[Talskubilos]

Surely, you are bluffing!

How many lexical items from Pokorny's or Mallory-Adams' do you think are genuine PIE words?

[WeepingElf]

Surely, you are bluffing!

How many lexical items from Pokorny's or Mallory-Adams' do you think are genuine PIE words?

I can't give you figures, but I think it's the vast majority in Mallory-Adams; Pokorny is now widely considered to have over-reconstructed, many items are questionable because of limited distribution or other problems, but even there, the number of genuine PIE words is quite high, I think.

[Nerulent]

This is why I think we'd need not just one, but several protolanguages for the IE family.

Okay... and what are they?

If you’re going to say the mainstream is wrong, then say why the mainstream is wrong. If you’re going to say there’s an alternative, then say what the alternative is.

How many lexical items from Pokorny's or Mallory-Adams' do you think are genuine PIE words?

Is there a word for evading questions by asking another question?

[Talskubilos]

I can't give you figures, but I think it's the vast majority in Mallory-Adams; Pokorny is now widely considered to have over-reconstructed, many items are questionable because of limited distribution or other problems, but even there, the number of genuine PIE words is quite high, I think.

Just think of the items with /a/ and/or 0 Ablaut, because they're likely candidates to be non-PIE. One example would be summer < *sm-r-o-, which in other branches turns up as *sam-o-.

Okay... and what are they? If you’re going to say the mainstream is wrong, then say why the mainstream is wrong. If you’re going to say there’s an alternative, then say what the alternative is.

I was working on that some years ago, but I didn't get my work done! Since then I've been thinking of 2-3 different proto-languages, although this is still far from being definitive.

[Nortaneous]

There may indeed have been Neolithic Wanderwörter which spread from language to language in the Near East as people adopted agriculture; it makes no sense denying this possibility. There are indeed candidates for such words in IE and Semitic that are seriously discussed by respectable scholars.

Really? I don't think many IE-ists have taken up the matter since Möller.

Bjørn 2017 is an overview recent enough to have an associated blog

[bradrn]

So where's the extraordinary evidence?

What kind of "extraordinary evidence" would convince you? Not a rethorical question.

I noticed KathTheDragon never actually answered this, so here is what would convince me:

• An overview of the modern PIE literature showing that the vast majority of ‘PIE’ words are found only in one branch;
• A demonstration that the various IE branches are not linked through regular sound changes, and thus cannot be descended from a single protolanguage; and
• A plausible theory for why the IE branches have so many words with obvious similarities if they are unrelated.

If you can do that, then consider me convinced.

[Richard W]

Let me ask another one of my questions: does this matter? In English, we have plenty of words like ‘formation’ and ‘question’ and ‘verb’ and ‘matter’: yes, they’re loanwords, but no-one sane would say that they’re not English. In the same way, even if you postulate that most ‘PIE’ words were loaned into it, that doesn’t mean they stop being words of PIE! (EDIT: Oops, just noticed jal had exactly the same critique!)

A very important thing we mustn't forget is PIE isn't a real language like English but a mere theoretical construct.

When did English become a real language? We haven't yet hived off anything south of the Anglo-Scottish border.

Actually, formation has some grammatical behaviour which marks it as a loanword from Latin, and this is not an unusual phenomenon. Loanwords of some description form a grammatically special class (or even classes) in several languages - English, Persian and Thai, and I expect someone to add Japanese to the list. I can’t think of any non-native behaviour for question, and verb has gone completely native. The only non-native behaviour I can think of for matter is a weak connection with material.

[KathTheDragon]

So where's the extraordinary evidence?

What kind of "extraordinary evidence" would convince you? Not a rethorical question.

I noticed KathTheDragon never actually answered this, so here is what would convince me:

• An overview of the modern PIE literature showing that the vast majority of ‘PIE’ words are found only in one branch;
• A demonstration that the various IE branches are not linked through regular sound changes, and thus cannot be descended from a single protolanguage; and
• A plausible theory for why the IE branches have so many words with obvious similarities if they are unrelated.

If you can do that, then consider me convinced.

I never answered because I didn't check the forum again after posting because I got wrapped up in writing code.

So, since your claim is that reconstructed PIE is not the ancestor of all attested IE langs, I'd want to see a substantiated argument about what makes it impossible to derive all attested IE langs from a singular parent language.

[Nortaneous]

But there's evidence of Caucasian loanwords in the language(s) of the Steppe people (aka Kurgans) as well as in individual IE branches, as I stated on FrathWiki.

I must admit that I was curious about your link here, so I had a look at it, and eventually found your statement of your theory on FrathWiki.

hm

In addition to their native lexicon (i.e. the one inherited from its ancestor), all languages have loanwords from other languages, either resulting from language replacement (substrates) or contact (adstrates) processes. Thus they aren't actually monolythic but multi-layer entities (a term which I myself borrowed from the Bulgarian linguist Vladimir Georgiev, who first used it for describing Lycian, an Anatolian language).

Languages are abstractions along multiple dimensions. On the one hand, every language is an abstraction over dialects, every dialect is an abstraction over idiolects, and every idiolect is an abstraction over individual utterances; on the other hand, every language is an abstraction over change in time. There's some sense in which we speak the same language as Shakespeare, but that sense is the culturally and historically contingent product of abstractification. Whatever. Specialists know that. If analytic philosophy blows your mind, you're either doing it wrong or reading about moral realism. (Or thermostat consciousness, but as a dualist I can treat that as a reducto ad absurdum.)

Proto-languages are also abstractions. In principle, every word in every language came from somewhere; in practice, in some cases it's unknown. If English were an obscure tribal language with no written tradition, no amount of etymological and linguistic effort could explain the word "guy"; and if Algonquian were also extinct, they might be able, with exceptional effort and lexicographic thoroughness, to tell that "persimmon" is a loanword from some kind of substrate, but it's unlikely that they could get further than that. (Somebody at some point had a post about the -mon element, but I can't find it now.) "Yeet" has a plausible and even regular Proto-Germanic etymology, but it's a recent sound-symbolic coinage. And so on.

At the same time, though, there was a dialect continuum in pre-expansion England, and this was diachronically "monolithic" in the sense that it was all spoken at the same time. It's been lexically expanded with words like "persimmon" and "yeet", but the innovations, difficulties in reconstruction, interdialectal loans etc. are massively outweighed by the core of inherited vocabulary.

Unfortunately, most comparative linguists have chosen a monolythic approach when reconstructing proto-languages (which to some extent are conlangs), so they implicit assume all the lexicon is inherited from a single source.

Agnostic, not monolithic. Roots are projected back to PIE when it's possible to do so. *meh2lom and *h2ebol- are reconstructed because they can be - their reflexes are regular when they're present - but this doesn't mean they're automatically assumed to have been present in the last common ancestor of all the IE languages. All it means is that the roots are reconstructible in some subset of Indo-European. IIRC, LIV even provides putative IE etymologies for words that only show up in a single branch, on the grounds that it might facilitate future research so they may as well.

Reconstructions of words are not claims that the words were present in the actual LCA!

In the case of the IE (macro)family, the "PIE" reconstructed by specialists doesn't represent a real language spoken by real people but rather a cross-section of the last stages of IE (which IMHO is the result of a complex series of replacement and contact processes). This can be exemplified by the huge gap between Anatolian and the rest of IE languages (cfr. Sturtervant's "Indo-Hittite"), which has lead to scholars such as the Spanish Francisco Rodríguez Adrados to propose a more refined model than the traditional one (coined by Neogrammarians in the 19th century), with several splits and intermediate stages.

Is the gap really so large? Lexical replacement operates quickly under the best of conditions. What percentage of Swadesh list entries are cognate between, say, Tongan and Hawaiian?

IMHO these Paleo-IE layers (in plural) represent a very large amount of the IE lexicon

What percentage?

In the Chalcolithic and the early Bronze Age, they underwent a rapid ("explosive" in Villar's words) expansion, imposing his language to other peoples in a series of elite dominance processes and contributing to the shaping of the later emerging historical IE languages, except Anatolian and possibly also Tocharian. As pointed out by Sergei Starostin, Kurganic has Vasco-Caucasian loanwords such as 'horse' and 'wheel'.

Starostin needs to be taken with an ocean's worth of salt, if at all. What grounds justify the rejection of the IE-internal etymology for *kʷekʷlos - or do you mean a different word for "wheel"? And explosive growth doesn't necessarily imply massive substrate influence - Algonquian left American English a few words and not much else.

However, here and there traces of the languages spoken prior to Kurganic (i.e. pre-Kurganic) survived, especially in Germanic, whose stop system is different from the rest of IE (except Armenian) and comparable to the one of Kartvelian, upon which the so-called glottalic theory was modelled.

How is the Germanic stop system similar to the Kartvelian one? External comparisons generally go to Uralic, which seems more reasonable. The Kortlandtian arguments for the archaic status of the final stop glottalization that occasionally shows up in Germanic are weak IMO - there are solid explanations that don't rely on adstrate influence.

For example, the word *H₂arH₃-tr-o- 'plough' found in some IE languages is cognate to Semitic *ħVruθ- 'to till, to plough; arable land', where IE H₂ ~ Semitic ħ.

So *-tr-o- isn't internal? Or the word was reshaped by analogy? And where would the *h3 have come from in this case?

This Gravettian Paleo-IE would be part of an Eurasiatic phylum in which I'd include Altaic and possibly other Eurasian language (macro)families.

starostinism: not even once

[Richard W]

What grounds justify the rejection of the IE-internal etymology for *kʷekʷlos - or do you mean a different word for "wheel"?

Several:

• Striking Sumerian and Semitic look-alikes make one suspect that we have at best the (P)IE representative of a wanderwort - and then we have timing issues at projecting it back to PIE.
• It is a very odd formation from the root *kʷel. Could it be be a folk-etymology?
• It's not attested in Hittite (though I've a feeling we can't declare it as highly unlikely) and the Tocharian form feels quite different - though it would work as a borrowing parallel to core-IE.
• Numerous irregularities within core-IE - they were discussed here two boards ago.
• Is PIE *(s)ker, or its possibly reduced reduplicated apparent derivatives, a doublet? (Through different routes, this gives English ring and circus.)

And explosive growth doesn't necessarily imply massive substrate influence - Algonquian left American English a few words and not much else.

Whereas pre-IE populations gave rise to a very significant proportion of modern IE populations.

[WeepingElf]

I can't give you figures, but I think it's the vast majority in Mallory-Adams; Pokorny is now widely considered to have over-reconstructed, many items are questionable because of limited distribution or other problems, but even there, the number of genuine PIE words is quite high, I think.

Just think of the items with /a/ and/or 0 Ablaut, because they're likely candidates to be non-PIE. One example would be summer < *sm-r-o-, which in other branches turns up as *sam-o-.

Yes, there are a few items in reconstructed PIE that look bizarre, and are likely to be loanwords. Words with an *a that cannot be explained by preceding or following laryngeals, and words with *b are prime examples. After all, PIE did not exist in a vacuum, but was in contact with other languages, such as NWC (the Maykop culture is IMHO a good candidate for Proto-NWC) and perhaps NEC (though that may have come from south of the Caucasus later, especially if Hurrian-Urartian is related to it) in the south, Uralic in the north, whatever the Cucuteni-Trypillia culture spoke in the west, and whatever was spoken in what is now Kazakhstan in the east. (Assuming that PIE was spoken by the Yamnaya culture, which is very likely, considering the results of archaeology and genetics, which show that there was a massive migration out of the Yamnaya homeland into much of Europe and Central Asia around 3000 BC.) Yet, the majority of the words reconstructed for PIE do not show such problems, and the most likely explanation is that they are indeed inherited from PIE. Of course, chance resemblances are possible, especially if a word is found in only a small number of languages far away from each other; also, when a word occurs only in a few closely adjacent languages (e.g., Italic, Celtic and Germanic), it is probably a dialectal formation or a loanword.

Okay... and what are they? If you’re going to say the mainstream is wrong, then say why the mainstream is wrong. If you’re going to say there’s an alternative, then say what the alternative is.

I was working on that some years ago, but I didn't get my work done! Since then I've been thinking of 2-3 different proto-languages, although this is still far from being definitive.

Show us that you are not bluffing, and that your model explains the facts better than the standard model.

And what regards the *kʷekʷlos word, this seems to be formed the same way as *h₁meh₁msos 'flesh'. This is admittedly a rare pattern, but at least it is not unique.

[Ares Land]

Hey, would any of you IEists have enough patience for a completely noobish question?

This talk of /a/ made me think... could anyone explain to me what exactly was going with the IE vowel system?

What I don't quite understand is that whenever I look up an etymology, the only vowel that comes up in the reconstructed form is /e/, with perhaps /o/ as an exceptional treat. (I'm joking. But only a little.)

What I understand is that for some reason what we reconstruct are abstract roots that didn't actually exist (in the sense of being uttered, or even conceived of, by PIE speakers), and would have been realized with short a, e, o or long a,e,i,o,u by actual speakers.
Oh, and besides, is there any reason but convention to reconstruct the short vowels as /a e o/ and not /a i u/?

[Pabappa]

Hey, would any of you IEists have enough patience for a completely noobish question?

well, it's complicated, so i'll just answer one tiny part of the question, and i pick this one because i suspect it's easy for beginners to misunderstand.

While we're pretty sure that PIE had [a e o] as surface vowels, we're not sure of what caused those realizations to appear. Almost all scholars acknowledge laryngeals that colored underlying /e/ into one of [a e o], and almost all scholars acknowledge vocalic ablaut that colored underlying /e/ into one of [e o ē ō].

These processes seem to be entirely unrelated .... very few scholars have put forth an argument that ablaut arose from the insertion of laryngeals, or that laryngeals arose from older ablaut prcoesses, as far as I know. Yet, it seems that ablaut and laryngeals never interfere with each other, either.

That's about all I know.

[Richard W]

Oh, and besides, is there any reason but convention to reconstruct the short vowels as /a e o/ and not /a i u/?

Yes. [ i] and [ u] were the vocalic allophones of the resonants /y/ and /w/. Thus one may say that that PIE had 5 short oral vowels. Phonetically, there were also the vocalic allophones of the resonants /l/, /r/, /m/ and /n/ and of the three laryngeals. Oddly, there doesn't seem to have been a vocalic allophone of /s/.

[WeepingElf]

Hey, would any of you IEists have enough patience for a completely noobish question?

This talk of /a/ made me think... could anyone explain to me what exactly was going with the IE vowel system?

What I don't quite understand is that whenever I look up an etymology, the only vowel that comes up in the reconstructed form is /e/, with perhaps /o/ as an exceptional treat. (I'm joking. But only a little.)

What I understand is that for some reason what we reconstruct are abstract roots that didn't actually exist (in the sense of being uttered, or even conceived of, by PIE speakers), and would have been realized with short a, e, o or long a,e,i,o,u by actual speakers.
Oh, and besides, is there any reason but convention to reconstruct the short vowels as /a e o/ and not /a i u/?

See here for an introduction. The vowels *e and *o take part in ablaut, alternating with each other and with zero according to rules that are not only quite complex but not even fully understood; *a is mostly the result of the influence of a consonant usually transcribed *h₂ of uncertain phonetic value (probably a uvular or pharyngeal continuant of some sort; such consonants tend to have a backing effect on adjacent front vowels), that changes a preceding or following *e into *a (while *o is not affected). The high vowels *i and *u are essentially vocalic allophones of *y and *w, and considered the zero grades of the diphthongs *ei and *eu. However, there are quite a few irregularities which aren't explained easily, and the scholars' opinions still are divided on these questions. Nobody really knows what exactly was going on with the IE vowel system!

[Ares Land]

Ah, thanks!

What bothers me a little is how odd the system looks.
Is *a attested at all, aside from laryngeal influence? Or is it assumed to have existed, but beyond any possible reconstruction?

For that matter, why are *i *u syllabic *y *u and not *y *w allophones of the vowel? (For that matter, is there any language described that way?)
Wouldn't it make sense to see zero grade as base forms, say *i, *u with *e added? (Or with *i *u lengthened or something and then breaking into diphthongs?)

I'm not trying to come up with my own crackpot theory, BTW (there seems to be no shortage of those), just curious if any "alternate" theories, based on something more typologically common, exist? Something akin to the glottalic theory, but for vowels, in other words...

[KathTheDragon]

For that matter, why are *i *u syllabic *y *u and not *y *w allophones of the vowel?

It's completely arbitrary. PIE "doesn't have" *i *u in the same way that my lect of English "doesn't have" n̩ m̩ - it's by-and-large convention that dictates that the vocalic allophones "don't count" and that the consonantal allophones are the "primary" ones. A much better way to look at it is to allow that there are three disjoint sets of phonemes in PIE: those that're purely vocalic, those that're purely consonantal, and those that're both.