The Great Proto-Indo-European Thread

Discussion of natural languages, or language in general.
User avatar
Howl
Sanci
Sanci
Posts: 69
Joined: Sat Oct 07, 2017 7:43 pm
Location: Campinia

Re: The Great Proto-Indo-European Thread

Post by Howl »

Vijay wrote:I'm warming up to the idea that Howl just might see how humans might associate increases in temperature with concepts like desire and hope.
Sure, but if one were to propose such an etymology in a long-range comparison, the response of the critics would be scorching.

Still, *kwep is a very interesting root, not just because of 'kw'. The reflexes of this root also show traces of a possible alternation *kwep/*kweb. The Germanic 'hope' requires *kub with a 'b'. And the Lithuanian 'kūpeti' has a strange long 'ū'. But if it were 'kūbeti' from *kub, this would be regular due to Winter's law.

Now, to stay on topic, I would like to ask about another hot PIE root. I have been searching for examples of roots where h3 is next to a stop without involving a zero grade. But I know of only one good example. And that is *bʰeh₃g 'to bake, to roast'. This root has a variant in LIV: bʰerǵ- 'to bake, to roast' (Vedic bhrjjati 'to roast'). Which is nice to know if you want to argue that h3 is ʁ or ʁʷ. But does anybody know any other roots where h3 is adjacent to a stop and where this is not caused by a zero grade?

User avatar
Salmoneus
Sanno
Sanno
Posts: 3197
Joined: Thu Jan 15, 2004 5:00 pm
Location: One of the dark places of the world

Re: The Great Proto-Indo-European Thread

Post by Salmoneus »

Howl wrote:
Sure, but if one were to propose such an etymology in a long-range comparison, the response of the critics would be scorching.
But why? Because there the burden of proof has shifted.

In long-range comparison, the default assumption is that two languages are not closely related (most languages aren't). So you have to actively prove, or at least actively make persuasive, the idea of a connection. Explanations with large but understandable semantic leaps aren't good in that regard, because there are lots of large semantic leaps possible, which means that the chance of a chance resemblance is much higher.

In dealing with a demonstrated family, however, the default assumption is that two languages are related and that most of the vocabulary in inherited where not demonstably uninherited. So now you have to actively make persuasive the idea that a certain formal correspondance has in fact simply arisen by pure coincidence (which is unlikely). To do that you generally have to rule out plausible and understandable semantic leaps, even when they're large, because lots of words do undergo semantic leaps.

The same thing, of course, happens with soundchanges. If your comparison relies on sound changes that knock half the sounds out of the word, that's not great evidence of a relationship, because the chance of a chance resemblance is increased. But if you know two languages are related, there's nothing wrong with proposing those deleting sound changes where the evidence demonstrates their presence.


To give a non-linguistic example: in exactly the same way, what counts as "persuasive evidence" in the trial where X is accused of murdering Y is different from the "persuasive evidence" in the trial where Z is accused of framing X for Y's murder. In the first case, you need to prove beyond a reasonable doubt that X DID kill Y; in the second, it's enough to prove that there is at least a reasonable possibility that X killed Y (so that you can't be sure beyond reasonable doubt that they're innocent and Z framed them). Shifting the burden of proof changes the strength of the evidence.
Blog: [url]http://vacuouswastrel.wordpress.com/[/url]

But the river tripped on her by and by, lapping
as though her heart was brook: Why, why, why! Weh, O weh
I'se so silly to be flowing but I no canna stay!

User avatar
Howl
Sanci
Sanci
Posts: 69
Joined: Sat Oct 07, 2017 7:43 pm
Location: Campinia

Re: The Great Proto-Indo-European Thread

Post by Howl »

Salmoneus wrote:
Howl wrote:
Sure, but if one were to propose such an etymology in a long-range comparison, the response of the critics would be scorching.
So now you have to actively make persuasive the idea that a certain formal correspondance has in fact simply arisen by pure coincidence (which is unlikely). To do that you generally have to rule out plausible and understandable semantic leaps, even when they're large, because lots of words do undergo semantic leaps.
Is it so unlikely?

Take for example English /ɹaɪt/
1. Correct, appropriate
2. Straight, not bent.
3. To form letters, words or symbols on a surface in order to communicate
4. A religious custom.
5. A builder or creator of something

All these meanings can be explained by semantic shifts from the base meaning 'correct, appropriate'.

But we know this is not correct. The English spelling reflects that. 1. and 2. is spelled 'right', 3. is spelled 'write', 4. is spelled 'rite' and 5. is spelled 'wright'.

For PIE we do not have any history. So we can't tell if the same root with a different meaning is a homophone or just a derived meaning.

Vijay
Smeric
Smeric
Posts: 2244
Joined: Sat Feb 06, 2016 3:25 pm
Location: Austin, TX, USA

Re: The Great Proto-Indo-European Thread

Post by Vijay »

How would you get meanings #3-5 from that base meaning?

User avatar
Soap
Smeric
Smeric
Posts: 1228
Joined: Sun Feb 16, 2003 2:57 pm
Location: Scattered disc
Contact:

Re: The Great Proto-Indo-European Thread

Post by Soap »

i think it's at least plausible. Imagine:

3) I had to write my phone number to ensure people got it right.
4) The rite of baptism is so-called because it sets a man right with God.
5) A millwright is the one who rights (erects) the mill.

Meaning #3 is the most common of the latter ones, and the earliest learned. Misspelling of this word does not seem very common. But the less common words rite and wright are probably often mistaken by schoolchildren as semantic variants of #1 and #2.
Sunàqʷa the Sea Lamprey says:
Image

Vijay
Smeric
Smeric
Posts: 2244
Joined: Sat Feb 06, 2016 3:25 pm
Location: Austin, TX, USA

Re: The Great Proto-Indo-European Thread

Post by Vijay »

All three of those words are frequently misspelled, and it's plausible that people can come up with a connection between 'correct' and the other meanings. That doesn't mean it's also plausible that they evolved that way, though, because none of these are analogies that can be found with other semantic shifts whereas "increase in temperature -> desire, hope" is.

User avatar
jal
Sumerul
Sumerul
Posts: 2633
Joined: Tue Feb 06, 2007 12:03 am
Location: Netherlands
Contact:

Re: The Great Proto-Indo-European Thread

Post by jal »

If there's one thing we can be sure of, is that when it comes to semantic drift, everything is possible. I find it silly to argue a certain shift "is not possible", or "highly unlikely".


JAL

User avatar
linguoboy
Sanno
Sanno
Posts: 3681
Joined: Tue Sep 17, 2002 9:00 am
Location: Rogers Park/Evanston

Re: The Great Proto-Indo-European Thread

Post by linguoboy »

jal wrote:If there's one thing we can be sure of, is that when it comes to semantic drift, everything is possible. I find it silly to argue a certain shift "is not possible", or "highly unlikely".
+1.

To give one example from the same semantic area:

PIE *(s)kreybʰ- "to scratch, to tear"
Proto-Italic *skreiβō "to carve"
Latin scrībō "I write"
PGmc *skrībaną "to write"
OE scrifan "to decree, to prescribe"
ME shriven "to impose penance; to hear confession"

So from "scratch" to "perform the Christian rite of penance" in the span of about 3000 years. That's not any more or less implausible than Soap's examples.

User avatar
Raholeun
Sanci
Sanci
Posts: 32
Joined: Mon Jan 26, 2015 7:30 am
Location: sub omnibus canonibus

Re: The Great Proto-Indo-European Thread

Post by Raholeun »

When we're still on the topic of remarkable semantic drift, this is one of my favorites:
PIE: bʰlendʰ - "to blend, to mix up"
P-Balto-Slavic: blenstei - "to muddle, to talk nonsense"
P-Slavic: blędь - "error"
Russian: блядь (blyad') - "whore, bitch"

Edited spelling
language LNK. cover

User avatar
WeepingElf
Smeric
Smeric
Posts: 1630
Joined: Wed Mar 08, 2006 5:00 pm
Location: Braunschweig, Germany
Contact:

Re: The Great Proto-Indo-European Thread

Post by WeepingElf »

Reading The Oxford Introduction to Proto-Indo-European ... by Mallory & Adams made me think again about the PIE laryngeals.

My starting point was the popular idea that the three laryngeals were just the fricative members of the three velar series. There are, though, a few problems with it:

1. If the velar fricatives coloured vowels, why didn't the velar stops?

2. The token frequencies seem to be wrong: one would expect *h1 to be the most common and *h2 the rarest. But *h2 is even more common than *h1!

3. Why is *h2 preserved in Hittite (in most cases), while *h1 and probably also *h3 (there are some unclear cases here) are lost?

The first objection can be met by assuming that the laryngeals originally were velar fricatives but later changed into something else, perhaps pharyngeal approximants, but keeping their palatalization and labialization features.

As for the second objection, I have been suspecting for some while that in *h2, the plain velar fricative merged with something else, perhaps a pharyngeal fricative */ħ/.

Now, Mallory and Adams split *h2 into two: *h2 proper, which colours *e to *a and is preserved in Hittite, and the much rarer *h4, which has the same vowel-colouring effect and is lost in Hittite.

So, now everything seems to fall into place: There were three velar fricatives, one in each of the three velar series, and one other fricative, perhaps pharyngeal. The velar fricatives are all lost in Hittite, and became perhaps pharyngeal approximants in Late PIE. These are Mallory/Adams's *h1 (palatovelar), *h4 (plain velar) and *h3 (labiovelar). Voilà, the frequencies are right, *h1 is the most common, *h4 the rarest. The pharyngeal fricative is Mallory/Adams's *h2, preserved in Hittite. That this one is common as dirt does not hurt as it has nothing to do with the velar series and thus does not upset the token frequencies. That a pharyngeal fricative (which may have become velar or uvular in Hittite, which is why it is written h-rocker rather than h-underdot) survives while velar fricatives are lost, is not that much of a problem. Something similar has happened in English which has lost /x/ but preserved /h/, and if the velar fricatives had become approximants in PIE already, this is even less of a problem. Approximants, of course, get lost easier than fricatives.
...brought to you by the Weeping Elf
Tha cvastam émi cvastam santham amal phelsa. -- Friedrich Schiller
ESTAR-3SG:P human-OBJ only human-OBJ true-OBJ REL-LOC play-3SG:A

User avatar
Howl
Sanci
Sanci
Posts: 69
Joined: Sat Oct 07, 2017 7:43 pm
Location: Campinia

Re: The Great Proto-Indo-European Thread

Post by Howl »

There is one issue with this. Initial h₃e is assumed to preserved as 'ha' in Hittite.
For example: 'hastai' bone from h₃estoi and 'haran' eagle from h₃eron.

User avatar
WeepingElf
Smeric
Smeric
Posts: 1630
Joined: Wed Mar 08, 2006 5:00 pm
Location: Braunschweig, Germany
Contact:

Re: The Great Proto-Indo-European Thread

Post by WeepingElf »

Howl wrote:There is one issue with this. Initial h₃e is assumed to preserved as 'ha' in Hittite.
For example: 'hastai' bone from h₃estoi and 'haran' eagle from h₃eron.
Yes, there are some instances where a *h3 is preserved in Hittite. I don't know yet what to make of that.
...brought to you by the Weeping Elf
Tha cvastam émi cvastam santham amal phelsa. -- Friedrich Schiller
ESTAR-3SG:P human-OBJ only human-OBJ true-OBJ REL-LOC play-3SG:A

User avatar
Howl
Sanci
Sanci
Posts: 69
Joined: Sat Oct 07, 2017 7:43 pm
Location: Campinia

Re: The Great Proto-Indo-European Thread

Post by Howl »

WeepingElf wrote:
Howl wrote:There is one issue with this. Initial h₃e is assumed to preserved as 'ha' in Hittite.
For example: 'hastai' bone from h₃estoi and 'haran' eagle from h₃eron.
Yes, there are some instances where a *h3 is preserved in Hittite. I don't know yet what to make of that.
The theory I'm referring to (By Kloekhorst) says that h₃e was preserved but h₃o/h₃C was not. But how does one determine if the correct reconstruction was h₃e or h₃o? It is probably possible to come up with a theory that has two o-coloring laryngeals. But the o-coloring laryngeal is also more rare in reconstructed PIE than the a-coloring laryngeal(s).

User avatar
KathTheDragon
Smeric
Smeric
Posts: 2139
Joined: Thu Apr 25, 2013 4:48 am
Location: Brittania

Re: The Great Proto-Indo-European Thread

Post by KathTheDragon »

WeepingElf wrote:2. The token frequencies seem to be wrong: one would expect *h1 to be the most common and *h2 the rarest. But *h2 is even more common than *h1!
There's a blindingly simple solution to this: match up *h₂ with the front-velar series instead of the back-velar series.

Of course, that leaves *h₁ dangling, but then I think it was a glottal stop or fricative (or both!) which neatly avoids that issue.
Howl wrote:The theory I'm referring to (By Kloekhorst) says that h₃e was preserved but h₃o/h₃C was not. But how does one determine if the correct reconstruction was h₃e or h₃o?
What Kloekhorst usually does is appeal to theoretic considerations, such as ḫi-conjugation verbs always having o/zero-ablaut. That said, I'm unsure of his theory - certainly I don't understand the phonetic justification for it, and even his supporting evidence I suspect could be interpreted otherwise (though I haven't personally investigated)

User avatar
WeepingElf
Smeric
Smeric
Posts: 1630
Joined: Wed Mar 08, 2006 5:00 pm
Location: Braunschweig, Germany
Contact:

Re: The Great Proto-Indo-European Thread

Post by WeepingElf »

One way out of this conundrum would be to posit a fifth laryngeal, namely a labialized pharyngeal fricative, which colours *e to *o and is preserved in Hittite, while the labiovelar fricative is lost as I said above, but this seems quite ad hoc to me. There are probably better solutions, but I know too little of Anatolian phonology (a difficult matter anyway, with that imprecise writing system) to suggest anything useful.
...brought to you by the Weeping Elf
Tha cvastam émi cvastam santham amal phelsa. -- Friedrich Schiller
ESTAR-3SG:P human-OBJ only human-OBJ true-OBJ REL-LOC play-3SG:A

User avatar
WeepingElf
Smeric
Smeric
Posts: 1630
Joined: Wed Mar 08, 2006 5:00 pm
Location: Braunschweig, Germany
Contact:

Re: The Great Proto-Indo-European Thread

Post by WeepingElf »

KathTheDragon wrote:
WeepingElf wrote:2. The token frequencies seem to be wrong: one would expect *h1 to be the most common and *h2 the rarest. But *h2 is even more common than *h1!
There's a blindingly simple solution to this: match up *h₂ with the front-velar series instead of the back-velar series.

Of course, that leaves *h₁ dangling, but then I think it was a glottal stop or fricative (or both!) which neatly avoids that issue.
And it leaves the labiovelar fricative dangling (your suggestion of a "*h2w" does not really convince me - if it existed at all, it would be far rarer than *h3, while it should be slightly more common). Also, assuming that an unlabialized back velar fricative would o-colour works only under your assumption that PIE *o was not rounded - which is something I am utterly unconvinced of as well. All branches except Tocharian (which has done so many weird things to its vowels that it can't speak for anything) which have kept *o distinct from *a agree on *o being rounded.

I still think my suggestion explains the facts better than yours, but it is not perfect, and I may be wrong.
...brought to you by the Weeping Elf
Tha cvastam émi cvastam santham amal phelsa. -- Friedrich Schiller
ESTAR-3SG:P human-OBJ only human-OBJ true-OBJ REL-LOC play-3SG:A

User avatar
KathTheDragon
Smeric
Smeric
Posts: 2139
Joined: Thu Apr 25, 2013 4:48 am
Location: Brittania

Re: The Great Proto-Indo-European Thread

Post by KathTheDragon »

*h₄ is pretty ad-hoc as well, since it only finds support in Anatolian, but even that support can be explained away.

User avatar
WeepingElf
Smeric
Smeric
Posts: 1630
Joined: Wed Mar 08, 2006 5:00 pm
Location: Braunschweig, Germany
Contact:

Re: The Great Proto-Indo-European Thread

Post by WeepingElf »

KathTheDragon wrote:*h₄ is pretty ad-hoc as well, since it only finds support in Anatolian, but even that support can be explained away.
I thought that way for a long time as well - I have been aware of this proposal for years, but thought the evidence would be too tenuous. Until I realized that it seems to help solving the problem I discussed above. But maybe I am on the wrong track ...
...brought to you by the Weeping Elf
Tha cvastam émi cvastam santham amal phelsa. -- Friedrich Schiller
ESTAR-3SG:P human-OBJ only human-OBJ true-OBJ REL-LOC play-3SG:A

User avatar
Howl
Sanci
Sanci
Posts: 69
Joined: Sat Oct 07, 2017 7:43 pm
Location: Campinia

Re: The Great Proto-Indo-European Thread

Post by Howl »

KathTheDragon wrote:
Howl wrote:The theory I'm referring to (By Kloekhorst) says that h₃e was preserved but h₃o/h₃C was not. But how does one determine if the correct reconstruction was h₃e or h₃o?
What Kloekhorst usually does is appeal to theoretic considerations, such as ḫi-conjugation verbs always having o/zero-ablaut. That said, I'm unsure of his theory - certainly I don't understand the phonetic justification for it, and even his supporting evidence I suspect could be interpreted otherwise (though I haven't personally investigated)
www.academia.edu/9795312/The_Fate_of_PIE_h3_in_Hittite
KathTheDragon wrote:*h₄ is pretty ad-hoc as well, since it only finds support in Anatolian, but even that support can be explained away.
How can that be explained away? The only ways I see are either an 'a' without laryngeal or something much like how the palatal velar series get explained away.

User avatar
KathTheDragon
Smeric
Smeric
Posts: 2139
Joined: Thu Apr 25, 2013 4:48 am
Location: Brittania

Re: The Great Proto-Indo-European Thread

Post by KathTheDragon »

To clarify what I mean by being unsure of his theory - I don't know if it's correct; I know exactly what he claims, having read his papers on it several times.

As I recall, the main piece of evidence for *h₄ is Hittite alpas "cloud", claimed to be cognate with Latin albus "white" from *h₄elbʰos, but Kloekhorst dismisses this etymology.

User avatar
WeepingElf
Smeric
Smeric
Posts: 1630
Joined: Wed Mar 08, 2006 5:00 pm
Location: Braunschweig, Germany
Contact:

Re: The Great Proto-Indo-European Thread

Post by WeepingElf »

Mallory and Adams list a few dozen words with *h4 - not many, but clearly not marginal, and as you surely know, the whole plain velar series is the least common of the three. The punctus saliens in this discussion, I think, is whether PIE *o was rounded or not - and I am not aware of any academic Indo-Europeanist who doubts that PIE *o was rounded.

But I don't think we can seriously expect to crack this riddle which better-equipped academic linguists have shed so much ink on without coming to a definite solution ;) I don't know about your qualifications in this field, Kath, but I am just an amateur and ought not to be too sure of my ideas! So it is perhaps best to leave our models standing side by side, and agree to disagree.
...brought to you by the Weeping Elf
Tha cvastam émi cvastam santham amal phelsa. -- Friedrich Schiller
ESTAR-3SG:P human-OBJ only human-OBJ true-OBJ REL-LOC play-3SG:A

User avatar
Howl
Sanci
Sanci
Posts: 69
Joined: Sat Oct 07, 2017 7:43 pm
Location: Campinia

Re: The Great Proto-Indo-European Thread

Post by Howl »

KathTheDragon wrote:To clarify what I mean by being unsure of his theory - I don't know if it's correct; I know exactly what he claims, having read his papers on it several times.

As I recall, the main piece of evidence for *h₄ is Hittite alpas "cloud", claimed to be cognate with Latin albus "white" from *h₄elbʰos, but Kloekhorst dismisses this etymology.
WeepingElf wrote:Mallory and Adams list a few dozen words with *h4 - not many, but clearly not marginal
I wondered about that, too. But a little googling helped me here. It seems Kloekhorst thinks that both h₂ and h₃ disappear next to o. So h₂o -> Hittite 'a' and h₂e -> Hittite 'ha'. So *h₄elbʰos is only a problem here because its thematic (so no ablaut) and in e-grade in the other branches. And if that is the only problem to Kloekhorst's solution, I would say it is a good solution. And this fits in well with one direction in which I am thinking, which is that some (or now all) of the laryngeals next to apophonic o were already gone in early PIE.

User avatar
KathTheDragon
Smeric
Smeric
Posts: 2139
Joined: Thu Apr 25, 2013 4:48 am
Location: Brittania

Re: The Great Proto-Indo-European Thread

Post by KathTheDragon »

Just an enthusiastic amateur here too, WE! But then, these sorts of discussions are enlightening because they bring up problems you don't think of.

User avatar
WeepingElf
Smeric
Smeric
Posts: 1630
Joined: Wed Mar 08, 2006 5:00 pm
Location: Braunschweig, Germany
Contact:

Re: The Great Proto-Indo-European Thread

Post by WeepingElf »

KathTheDragon wrote:Just an enthusiastic amateur here too, WE! But then, these sorts of discussions are enlightening because they bring up problems you don't think of.
Sure! And being amateurs. we can research what we want to and don't have to worry about our ideas spoiling our careers. This is the reason why I decided against an academic career in linguistics: the money lies not in what interests me but in such drudgeries as developing language teaching courses and translation software.

Of course, we as amateurs are equally bound to using proper methods if we want to avoid ending up as crackpots. And cases like Michael Ventris (an architect who deciphered Linear B) and Ferdinand de Saussure (later to become an eminent linguist, but when he first proposed the laryngeal theory, he was just a student) show that even "outsiders" can achieve meaningful work if they work properly and stay in contact with the scholarly community. Now, I don't seriously expect to be the next Ventris, but I definitely don't want to be another Octaviano!

BTW: I have another idea about the "fifth laryngeal" I tentatively proposed yesterday. Perhaps it was something that nobody would consider calling a "laryngeal" - a labial fricative *f. That one would fill a gap in the system, as all other stop series have a corresponding fricative, plus a lonesome pharyngeal (or glottal) one:

Labial *f ("*h5")
Dental *s
Palatovelar *x́ (*h1)
Plain velar *x (*h4)
Labiovelar *xw (*h3)
Pharyngeal *ħ (*h2)
...brought to you by the Weeping Elf
Tha cvastam émi cvastam santham amal phelsa. -- Friedrich Schiller
ESTAR-3SG:P human-OBJ only human-OBJ true-OBJ REL-LOC play-3SG:A

User avatar
KathTheDragon
Smeric
Smeric
Posts: 2139
Joined: Thu Apr 25, 2013 4:48 am
Location: Brittania

Re: The Great Proto-Indo-European Thread

Post by KathTheDragon »

Well-put.

And if we're enumerating fricatives, here's my set:
Coronal *s
Velar *h₂ (/x/)
Labio-velar *h₂w (/xʷ/) since I think Anatolian preserves the original situation
Post-velar *h₃ (/χ/)
Glottal *h₁ (/h/), which may also contain an older *f if we want symmetry
Last edited by KathTheDragon on Mon Feb 05, 2018 5:07 pm, edited 1 time in total.

Post Reply