Greek, Pre-Verner Germanic, and Proto-Indo-European : Daniel Ross on Quora

Q

What if the Proto Indo European language came from the Proto Greek, considering that Greek is an isolated Indo European language, plus the fact that Greeks had already been to India before Alexander the Great according to the Greek tradition?
https://www.quora.com/What-if-the-Proto-Indo-European-language-came-from-the-Proto-Greek-considering-that-Greek-is-an-isolated-Indo-European-language-plus-the-fact-that-Greeks-had-already-been-to-India-before-Alexander-the-Great/answer/Daniel-Ross-71

Answer requested

by Georgios Antoniou

Daniel Ross

I study Linguistics

Answered Mar 8, 2018*

No.

• The entire basis for reconstructing PIE is that there are similar forms in Greek, Latin, Sanskrit, etc., which descend from a common ancestor. Greek is not that original ancestor, specifically because it also descends from it. There are sound changes and grammatical changes and so forth that simply wouldn’t fit if you put Greek in it “backwards”.
• Alexander the Great is several millennia after the breakup of PIE, so that is entirely irrelevant. We’re talking about distant pre-historic times around 6,000–8,000 years ago.
• “What if” isn’t a scientific argument. What if PIE was really the language of aliens, or Atlantis? What if PIE is really English? etc. Greek in place of PIE simply isn’t a better hypothesis than the currently accepted hypothesis; in fact, it is must worse (see 1 above).
• In a sense, Greek is modern PIE. So is English. So is Spanish. So is Russian. So is Hindi. And so forth. Latin isn’t actually a dead language— we just call it “Spanish”, “Italian”, etc. now (Daniel Ross' answer to Can we say when exactly did Latin die out in the city of Rome?). And the same is true for PIE— it split up into dialects and then languages, which went in their own directions. But you could, if you wish, continue calling any (=all) of the branches “[P]IE” and still be correct. That’s just a name. There is no “true” daughter compared to the others, though, so that argument wouldn’t somehow privilege Greek above the others, just indicate that indeed it represents a continued development from PIE just like all of the other branches do. In fact, even if you were somehow correct about this, the timeline would be so much older than your question implies that it would really end up being effectively the same as this hypothesis: several thousand years removed from our earliest records of Greek, so it would be almost unrecognizable. Perhaps you want to claim that Greece is the homeland of PIE, which is really a “what if” because there is no substantial evidence to back that up, and the migration of PIE groups from there doesn’t really fit with how we understand the language spreading. That’s harder to disprove, though, because we don’t know much about the homeland. But that doesn’t really mean the language was “Greek” in the modern sense any more than it was “English”.

*

2.5k Views, Upvoted by Thomas Wier, Assistant Professor of Linguistics at the Free University of Tbilisi. and Nick Pharris, Ph.D. Linguistics, University of Michigan (2006) · Author has 1.1k answers and 885.1k answer views

Answered twice

by me : A and B each giving a thread:

A

Hans-Georg Lundahl

Mar 8

“There are sound changes and grammatical changes and so forth that simply wouldn’t fit if you put Greek in it “backwards”.”

Not Greek as we have it, true but … Germanic as extant would not fit in it either, but a Germanic consonantism pre-Verner minus the Germanic changes in vocalism would.

That hypothesis, which I consider very weak, would however elucidate one etymology.

“Paphl-agon-ia” would involve a paphl = Germanic and PIE Babl : babl, South IE vavl, phaphl > paphl.

So, Paphlagonia would be “land of those who go babble”. Makes sense since they had a language unrelated to neighbours.

Try and see if something like that could be arranged for Greek?

Daniel Ross

Mar 8

But if you approach it that way, then you’re really just taking Ancient Greek and then backing up all the way to PIE, while still calling it “Greek”. The distinction between the conventional hypothesis and the one of this question becomes one of nomenclature, not substance.

If it’s all that for a single etymology (or even various data points), then you could simply reconstruct PIE differently regarding those features, even propose some new sound shifts and make PIE more Greek-like. But none of that would really make PIE into Greek, and certainly not in anything like the modern (or ‘Ancient’) sense!!

Hans-Georg Lundahl

Mar 8

Not really. Let’s take the pre-Verner Germanic one.

On the conventional road, the original would have been bhabhl-.

On the now conventional road the original for father would be pxtehr (if x is a graph you accept for ach-laut), or for “o father” it would be “ogh pxthr” (using gh for labialised uvular voiced fricative). On “my view” it would have been fathéér / fáther (thorn, not eth).

Possibly precisely “a” as the vowel for a vs Skr i … while some other Germanic a would indeed have been more PIE, as in “o”.

I would say pre-Verner-Germanic as much as Hittite would be perhaps easier than Greek.

One recent Finnish construction is so close to Hittite that the known Hittite would be a spelling simplification of it (as Runes also don’t distinguish between p and b).

Also, a computer simulated cladogram for IE langs placed PIE close to Gothic and Germanic on one branch, other IE langs on the other primeval branching.

It is built on words, vocabulary, rather than order of sound changes.

Daniel Ross

Mar 8

The cladograms you refer to are all but meaningless, given that you can find a published one to represent basically any given subgrouping you’d like. As I said, important area of research, but not one we can cite yet for any sort of consensus because every apparently “reliable” source seems to contradict all the others!

As for the rest, that’s a very different question than the original one about Greek vs. PIE. Sure, PIE might have been different than some reconstructions, but that doesn’t make it Greek. Simple as that. And there’s a large burden of proof required to change the consensus that has developed over the last 200 years, but go for it if you want to take that on.

The odds of Greek in any meaningful sense being PIE? 0%.*

The odds of some of the phonological reconstruction for PIE being incorrect? >99%.

*

If for instance relative in PIE was yos, proposed etymon for Greek hos, and not quis or quei and not tos or te (so unlike the etyma for Latin and Germanic relative pronouns, that would have been one way in which PIE could have been in some meaningful sense Greek, even though not phonetically identic. For interrogative, one must presume quis still had qu, not ti as in Greek tis, but that would have been the case in Mycenean Greek which does count as Greek. If "sea" had been either *dhalakya or *eH3keH1eH2nos rather than *mari, that would also make the PIE closer to Greek than now presumed. So, non-identity of Lautstand does not per se preclude PIE being in some meaningful sense Greek. This is however not what I believe, I rather tend to believe that Iavan's tribe did have the word known as thalassa or as okeanos, while mare (Latin, Germanic, Slavic, Celtic) would be from Gomer's tribe, and common words "over" the branches would be from Sprachbund - from "neighbouring" languages borrowing from each other. Note that languages are always "neighbouring" in bilinguals, whatever the geographic position of most speakers.

Hans-Georg Lundahl

Mar 8

“The cladograms you refer to are all but meaningless, given that you can find a published one to represent basically any given subgrouping you’d like.”

These ones are based on vocabulary. The idea behind them is, more close any clades are, more vocabulary they have in common.

Obviously, this makes Ukrainean and Bielorussian closer to Polish than to Russian, due to Tatarisation of Russian vocabulary.

Daniel Ross

Mar 8

Even more meaningless if they are blind computations based on vocabulary rather than the insights of a trained linguist. See the first video here: Daniel Ross' answer to What are some of the best YouTube lectures in linguistics?

But again, you cannot pick and choose a particular published account that supports your version of things, when there are many other published accounts out there that refute it. There is simply no agreement (let alone consensus) about these cladograms. I’ve done a bit of research about them, and it goes nowhere because each publication contradicts the last, in substantial ways. Nor are they currently converging on any sort of common result.

Links

What are some of the best YouTube lectures in linguistics?
https://www.quora.com/What-are-some-of-the-best-YouTube-lectures-in-linguistics/answer/Daniel-Ross-71

Mismodeling Indo-European Origins: The Assault On Historical Linguistics | GeoCurrents
GeoCurrents | 20.XII.2012
https://www.youtube.com/watch?v=4jHsy4xeuoQ

Hans-Georg Lundahl

Mar 8

I don’t recall if it was the one with the full Swadesh list or a reduced one which gave the result.*

*

It was a case of different cladograms to that in "Mismodeling Indo-European Origins", since that one was based on more words than the full Swadesh list, I think. Disagreeing cladograms depending on how many words are used would be one indication against PIE branching out in linguistic clades.

Hans-Georg Lundahl

Mar 8

“There is simply no agreement (let alone consensus) about these cladograms. I’ve done a bit of research about them, and it goes nowhere because each publication contradicts the last, in substantial ways.”

I’d like links to that.

Notably, I noted there was no unity between Swadesh and reduced Swadesh, but can you find any on either contradicting the one which had Gothic Germanic and PIE close by?

Obviously, if a publication has another criterium than “vocabulary of Swadesh” or “vocabulary of reduced Swadesh” (Swadesh–Yakhontov, I just checked), it could contradict the cladogram.

I’m not supporting Evolutionary Phylogenetics, but I suppose AronRa would have more convergence between diverse criteria, one of them being the cladogram for … whatever it was, cytochrome ? C.

Your admission would be one argument against proto-language hypothesis.

Daniel Ross

Mar 8

Here are some references, which provide very clear (and apparently compelling) analyses that come up with completely different results. These are published, peer-reviewed, and inconsistent.

Blanchard, Ph., F. Petroni, M. Serva, & D. Volchenkov (2011): “Geometric representations of language taxonomies.” Computer Speech and Language, 25: 679–699.

Campbell, Lyle (1999): Historical Linguistics: an Introduction. Cambridge, MA: MIT Press. Dobson, Annette J. (1978): “Evolution Times of Languages.” Journal of the American Statistical Association, 73:361: 58-64.
Dyen, Isidore, Joseph B. Kruskal & Paul Black (1992): “An Indoeuropean Classification: A Lexicostatistical Experiment.” Transactions of the American Philosophical Society, 82:5. Greenhill, Simon J. (2011): “Levenshtein Distances Fail to Identify Language Relationships Accurately.” Computational Linguistics, 37:4: 689-698.

Johnson, Keith (2008): Quantitative Methods in Linguistics. Malden, MA: Blackwell.

Ringe, Don, Tandy Warnow & Ann Taylor (2002): “Indo-European and Computational Cladistics.” Transactions of the Philological Society, 100:1: 59-129.

Serva, M. & F. Petroni (2008): “Indo-European languages tree by Levenshtein distance.” EPL, 81:6: #68005.

Serva, M (Online): http://univaq.it/~serva/languages/languages.html

While not yet

having taken time to consult this, I think I may owe him a grazie.

B

Hans-Georg Lundahl

Mar 8

// “What if” isn’t a scientific argument. //

In that case, PIE isn’t a scientific position either.

There is as much what if to it as to pre-X-Greek and pre-Verner-Germanic or as to Sprachbund.

If “what if” can be logically asked, it is scientifically relevant.

The dismissal may be a very obvious and clear one, but the question is still relevant.

Which means it is scientifically relevant or relevant to science if you consider reconstruction of pre-written states of languages we find “related” but not mutually intelligible as a branch of science.

Btw, h- for previous s-, y-, sw- would be a mutation in Greek, since it involves a merger.

[In other words, PIE cannot have had h- for both hedys and hos, but it can have had *swadus and *yos in those Greek meanings. Again, not what I believe, but I think Daniel Ross dismissed the possibility too easily.]

Daniel Ross

Mar 8

“In that case, PIE isn’t a scientific position either.” — Untrue. PIE is based on two major scientific positions: (1) the similarity between the IE languages is too great to be dismissed as coincidence, thus pointing to a common ancestor; and (2) the comparative method, tested over 200 years(!) and one of the most debated and polished aspects of linguistic science, giving us some idea of what PIE was like. Even if it’s wrong, it’s scientific. Just saying “what if” is not. A scientific hypothesis is much more than that.

The more important point implied in your questions here is what the sub-branching is within IE. Which branches are more closely related to others? Which split first? What is the order of sound changes, not just their networked relationships across the attested daughter languages? And that becomes a very important and fascinating scientific question. It’s one that is not yet solved. And yes, you could have some room to argue that PIE was more Greek-like than currently proposed. But that’s a huge amount of research— literally hundreds of scientific careers over the past 200 years, with almost every minute detail already discussed and debated— that you would need to convincingly refute and adjust. I’m not saying this is wrong. But it would take a lot to convince people that it is right. Regardless, the accepted result would not be to think of Greek as PIE, but rather that PIE was more Greeklike than in earlier proposals, and that the branching of other groups fit in a certain way to result in the changes that derived the modern languages.

And as much as a few changes here or there are facilitated by this revision of current consensus, there would be many more problems that pop up by changing things around. Tricky.

Hans-Georg Lundahl

Mar 8

Whether “what if” is a scientific hypothesis or not, it is scientifically relevant.

Also “is too great to be dismissed as coincidence, thus pointing to a common ancestor” - this is a glaring what if, ignoring the rival what if of Sprachbund.

“the comparative method, tested over 200 years(!) and one of the most debated and polished aspects of linguistic science, giving us some idea of what PIE was like.”

Yes, Schleicher’s Fable started out sounding fairly close to Sanskrit and would now sound fairly close to Klingon, if you substitute h for H1, ach-laut for H2 and labialised uvular voiced fricative for H3.

“But that’s a huge amount of research— literally hundreds of scientific careers over the past 200 years, with almost every minute detail already discussed and debated— that you would need to convincingly refute and adjust.”

As with the rival pre-Verner-Germanic, it suffices to adjust a certain point once and you apply that to every carreer or paper that is relevant.

I do not have to refute it once for this paper, once again for that other paper and so on.

The difficult one, which I am most interested in, is the Sprachbund thesis : documenting PIE root for PIE root how a Sprachbund explanation would suffice.

Like the PIE possible wopswa/wepswa where only Germanic and Italic and Baltic are clear exponents (Slavic osa would be the result of it by Slavic sound laws, but is so unspecific it could also be a coincidence).

Germanic is neighbour to both Italic and to Baltic, hence Sprachbund is sufficient.

Daniel Ross

Mar 8

My goal isn’t to convince you. But it seems your goal may be to convince others, and that will be very hard.

A sprachbund? Not entirely unreasonable, but it goes against our basic understanding of historical linguistics (for which PIE is the primary model).

In the end, none of what you are saying is more convincing than the current hypotheses. And that’s the problem. It’s like reinventing a slightly less efficient wheel. Until you can demonstrate that your wheel is actually better than the others, rather than just remotely plausible (and actually less so than the other popular wheels), you won’t convince anyone.

Overall the problem is that the PIE hypothesis does a very good job of systematically explaining the data. It would be hard to construct an alternative theory that is both as effective and as efficient.

But this isn’t my argument to have. I find the current explanations satisfactory, and specifically more satisfactory than the alternatives. Science is open for debate, though, and I have no need or interest in defending the status quo beyond what I’ve written. Good luck!

Hans-Georg Lundahl

Mar 8

"But it seems your goal may be to convince others, and that will be very hard."

Depends on their background.

"My goal isn’t to convince you."

But perhaps others, right?

"Not entirely unreasonable, but it goes against our basic understanding of historical linguistics (for which PIE is the primary model)."

Actually, what I consider a primary model for historical linguistics is Romance linguistics, Germanic linguistics, Balkan linguistics.

In other words, PIE being a proto-language is as reasonable as PIE being a Sprachbund on that view. Neither is the primary model, both are derived from some such.

Balkan linguistics starting later than Romance would be one valid explanation on why Proto-Language is more popular than Sprachbund.

"In the end, none of what you are saying is more convincing than the current hypotheses."

To you, who presumably give a timeline of 6000 years to IE language family.

And who presumably don't believe that Iavanites and Gomerites spoke different languages just after Babel. If you even credit the labels with any meaning.

I identify Iamnaia culture as by Magogians ... so, if I were to accept proto-language model, it would involve a very drastic view of today's situation, as per Apocalypse 20:7.

I consider it very probable that also the Sunghir man on Grave 1 was Magog himself. And the time scale between "32,050 and 28,550 BC" (Sunghir) and "4000 BC" (Iamnaia) would be only a few centuries with my recalibrated carbon tables. With Babel between (Göbekli Tepe).

"And that’s the problem. It’s like reinventing a slightly less efficient wheel. Until you can demonstrate that your wheel is actually better than the others, rather than just remotely plausible (and actually less so than the other popular wheels), you won’t convince anyone."

In your set circles.

It is more efficient on the account of Paphla- getting an etymology that makes sense in context (note, if pre-Verner Germanic was part of a Sprachbund, it makes sense sound laws were reversible in loans between languages, like a French to English loan today would involve -té > -ty and an English to French loan would involve -ty > -té).

It makes sense because many word fields would be central and still lack a common IE gloss reducible to a PIE one. It makes sense because for father the oldest IE language has the Atta gloss in common with Uralic.

And a few more.

My real problem is, when I get in touch with linguists, they refuse discussion.

I want to know how many of the PIE roots have more than 3 language families?

I want to know how many of them have language families that are distant (like Sanskrit and Gaelic) among the three or more?

I'd need to ask a linguist. Again, I wanted to check what wasp was in Sanskrit, Persian and a few more, I couldn't in Google translate use Farsi, because I don't know the Arabic alphabet well enough - so I'd need a linguist. Who, as usual, are as supportive as Ilya Usoskin on the milliSieverts per year implications of faster carbon 14 productions.

"Overall the problem is that the PIE hypothesis does a very good job of systematically explaining the data."

Except the Biblical ones ... unless we take Magogian as PIE ...

"It would be hard to construct an alternative theory that is both as effective and as efficient."

Or easy, once I get the linguists to discuss with.

By the way, what is the difference between effective and efficient to you?

"But this isn’t my argument to have. I find the current explanations satisfactory, and specifically more satisfactory than the alternatives. Science is open for debate, though, and I have no need or interest in defending the status quo beyond what I’ve written. Good luck!"

Thank you. So, what is wasp in Sanskrit?

Daniel Ross

Mar 8

The reason that linguists refuse to discuss this with you is that the conversation is frustrating. It’s like insisting that a mathematician discuss with you the possibility that the number 5 does not exist. That is also the reason I am uninterested in continuing to debate this.

The approach of science is this: 1) do background research; 2) gather data and test specific hypotheses; 3) suggest conclusions. You’re doing that backwards, asking for the background research after you’ve made conclusions. If you want to be taken seriously, take some time to understand the subject well. That is the only useful advice I can give you. There is a reason that experts in the field do not agree with you. And it isn’t because they haven’t considered the possibilities you’re thinking about, or because they overlooked details in the data. You will note that I have a number of upvotes here from people who do have training in linguistics. Interpret that as you wish. I have nothing more than credentials and the consensus of my peers to offer.

As for the bible, I have absolutely nothing to say about that, because we can’t possibly arrive at a productive conclusion from a book that, interpreted loosely, is metaphorical, or interpreted literally claims the universe is 6,000 years old (so of course PIE cannot exist).

Thanks for the comment. I’ll attempt to point you in the right direction for continued research, but otherwise need to end this discussion.

• See Pokorny’s etymological dictionary (easy to find copies online, in various formats, English or original German). I don’t know if it lists wasp specifically but there’s lots of data there to consider.
• I will also give you some references to published accounts with wildly different subgroupings, as a reply to your other comment, for this now too-spread-out conversation.

Good luck.

Pokorny:
https://indo-european.info/pokorny-etymological-dictionary/whnjs.htm

Thanks for the tip!

I checked Pokorny on waps, and while Sanskrit was lacking, it is as far east as Persian:

u̯obhsā

English meaning wasp

German meaning `Wespe'

Material Av. vawžaka- `Skorpion', aber iran. *vawža- `Wespe' in mpers. vaβz `Wespe', Baluchi gwabz `Biene, Wespe';

lat. vespa f. `Wespe' (aus *vopsā);

acorn. guhi-en gl. vespa, mcymr. gw(y)chi, abr. guohi gl. fucos (*u̯ops-), woraus entlehnt air. foich gl. vespa (auch `eruca'), nir. fotlach und puith `Wespe', daraus spoch `heftiger Angriff' (O'Rahilly Sc. G. St. 3, 63);

ags. wæfs, wæps, wæsp `Wespe', ahd. wefsa, wafsa, waspa, bair. webes, thür. weps-chen und wewetz-chen, die auf germ. *wabi-s und *wabi-t weisen;

lit. vapsvà `Wespe', apr. wobse ds.;

ksl. osa, klr. osá (aus *vopsā, baltoslav. *u̯apsā).

References WP. I 257 f., WH. II 770, Trautmann 342, Vasmer 2, 280, Specht Idg. Dekl. 45 f., Szemerényi Arch. Lingu. 4, 52.

See also deutlich zu u̯ebh- `weben'.

Pages 1179