Thomas Wier

And in at least two languages, a word meaning 'enemy' has come from words meaning 'destroy':

Greek Ἀπολλύων, avenging spirit of the Abyss < ἀπόλλυμι destroy
Hebrew אֲבַדּוֹן Abaddon < א־ב־ד ʾ-b-d perish, destroy, lose

So the semantic shift in Georgian has good precedents.

What is more clear is that words for 'enemy' across languages frequently come from words of ill-intent:

Greek δυσμενής < PIE *dus- bad *men- thought
Proto-Iranian *dušmánās < same as Greek
Russian враг < PIE *Hwergʰ- commit a crime
Old English fēond < PIE *peh₁- hate

As for where Kartvelian *ṭwr- destroy, shatter comes from, this is much less clear. One possibility is as a loan from Urartian 𒌅𒌨 tur- destroy, which may itself be related to Hurrian 𒌅 𒊒 𒁉 turubi 'enemy'. If so, the Kartvelian word may have been a very ancient calque.

We can show that the /m/ of such verb forms is specific to Georgian b/c the Svan cognate ლიტვრამაუი li-ṭwr-am-aw-i 'destroy, smash, break apart, shatter' lacks it.

If so, why did Georgian lose the *v in *m-ṭver-i? This is a common sound-change when one labial is near another.

So the word seems to be a Georgian-internal development, an agentive formation (m-) of the verb root ṭvr- found in a number of other words:

მტვრევა mṭvreva smash, break to pieces
გამტვრევა gamṭvreva crash down
დამტვრევა damṭvreva shatter, destroy

The word is not found in Georgian's sister lgs except as loans from Georgian:

Meg ნტერი nṭeri
Laz დუშმანი dušmani, < Turkish
Svan ამახუ amaxw

If the Megrelian form were inherited, we would expect **(n)ṭari or **(n)ṭaji because of Zan sound-laws.

The word is first attested in the oldest extended text, the Martyrdom of Queen Shushanik, from the year 476: ნუუკუე მტერმან ვითარცა სრსჳლმან საძოვარი პოოს შენ თანა 'Therefore do not let the enemy like a canker find pasture amongst you'

Weekly Georgian Etymology: მტერი mṭeri 'enemy', from Old Georgian ႫႲႤႰႨ mṭeri, from *m-ṭver-i, agentive participle of Kartvelian *ṭwr- break apart, destroy, possible ancient loan of Urartian 𒌅𒌨 tur- destroy. Many languages' words for enemy come from nouns of ill intent.

Historically many linguists attempted to find systematic links between such cognate sets as families, but these have generally failed. Instead, linguists today take more seriously language contact in classical spread zones like the Eurasian Steppe to explain how such forms arise.

In Altaic we see a similar root *kèra meaning to bind or wind:

Khalkha Mongolian xere- bind, connect
Ulcha kergin to bind
Nanai kergi bunch

Uralic presents many possible points of contact as well:

In fact the word is so old, it may reflect a very ancient layer of language contact with other major Eurasian language families. Indo-European e.g. has various reflexes of *ḱer- plait, weave:

Greek καῖρος row of thrums on a loom
Armenian սարդ sard spider
Albanian thur braid

from Fähnrich's Etymological Dictionary of Kartvelian

That verbal root *ḳr- is found in Georgian and all the other lgs with its historical order in other forms, so it is clearly a very old part of the language:

So it looks like the Georgian nominalization with სა- sa- is unique to its branch, and was borrowed into its sister lgs. In Georgian the earlier form based on the verb *ḳr- bind underwent metathesis from *saḳre to sarḳe, which is common with /r/-sounds in the language.

That means if Georgian სარკე sarḳe came from something like *ɬa-ḳr-e in Kartvelian, we would expect completely different output forms in the sister lgs:

Megrelian/Laz **orḳe/oḳre
Svan **läḳre

Instead we find founds that look like they underwent the Georgian shift *ɬ>s.

The reason is that in these cases they look look suspiciously like loans from Georgian. The Georgian prefix სა- sa- comes from Kartvelian *ɬa-, but this same *ɬ which word-initially became /s/ in Georgian became /l/ in Svan and was lost entirely in Zan lgs.

Most of the modern Kartvelian languages have similar words for mirrors: besides Georgian, Megrelian has სარკე sarḳe and Svan has სა:რკ särḳ. So why don't we reconstruct a word for 'mirror' to Proto-Kartvelian?

The word is attested in Old Georgian texts such as the 9th century translation of Ephrem the Syrian's work on repentence:

ვაჲ მისა, რომლისა გონებაჲ მისი სარკე
ცოდვისა არს და მარადის მას შთახედავს
Woe to him whose mind is a mirror of sin,
and always gazes down into it

"Dali from the Back Painting Gala from the Back", 1972-73 by Salvador Dali

Weekly Georgian Etymology: სარკე sarḳe 'mirror', from Old Georgian ႱႠႰႩႤ sarḳe, metathesis of earlier *saḳre, from Kartvelian *ḳr- bind, connect, possibly an ancient Eurasian Wanderwort: cf. Indo-European *ḱer- plait, weave; Altaic *kèra bind, wind; Uralic *kärV bind, knot.

This is so obviously falsified by a moment's historical reflection it's hard to see why it's taken seriously. Not all your 'Anglophone' countries are Common Law (Quebec, Scotland, Louisiana aren't). Both legal systems have existed for centuries, but your data suggest a recent divergence.

17th century illuminated Georgian manuscript of the Visramiani

Which translates: "When Vis heard this comment, with shame she slowly and beautifully expressed a tender answer and called out to him"

The word is found first in late Old Georgian or early Middle Georgian texts, such as the 12th century Persianate romantic epic the Visramiani, which celebrates the forbidden love between Vis and Ramin: რა ესე საუბარი ვისსა ესმა, სირცხუილითა ნელად და ლამაზად ნაზი პასუხი გამოსცა და აგრე არქუა

Weekly Georgian Etymology: ნაზი nazi 'tender, gentle, delicate', from late Old Georgian ႫႠႦႨ nazi, from Parthian 𐭭𐭠𐭦nāz pleasure, delight, from Proto-Iranian *nāz- take pleasure, exult, be mild. Perhaps a loan from Akkadian 𒈾𒀀𒄭𒍪 naḫsum thin, delicate, fine (of clothing).

This makes it also cognate (through borrowing) with Georgian ფეჩი peči 'cast-iron stove', a recent borrowing from Russian, which also ultimately comes from *pekʷ- cook.

The PIE root is the source of many words in IE languages:

English wind < *h₂wéh₁-n̥ts
Greek ἀτμός < *h₂wéh₁-(t)-mo-
Old Persian 𐎺𐎠 vā⁠ < *h₂wéh₁

In Proto-Iranian, a form like *h₂weh₁-ro- would naturally have become something like *var-, a possible loan source.

There is no obvious noun in Proto-Kartvelian that could fit the meaning and structure here. There are however foreign loans which might fit. One possibility is that Proto-Kartvelian borrowed a form of Indo-European *h₂weh₁-ro- wind, breath.

It seems thus like we're dealing here with an old form of incorporation: *war-tkw-l-i, where the verb incorporates a noun to create a compound verb and is itself nominalized. But where does this incorporated noun come from, if so?

The other issue is the morphology. The final -l suffix is a common way to nominalize verbs. Then there are analogous constructions like სუნთქვა suntkva 'breathe', which appears to contain an incorporated noun სუნი suni 'odor, breath' plus the light verb stem თქვა tkva, say, utter.

There are two possible explanations for this:

(a) Georgian /wa/ became /o/, and Svan underwent umlaut to /e/;
(b) Georgian originally had /we/ like Svan, but lost its original cognate and borrowed one from a Zan, such as Megrelian

The normal expected vowel reflex should be:

Georgian /a/ :: Megrelian /o/ :: Svan /a,ä,e/ (the latter through umlaut)

Normally, Svan /e/ implies either Georgian /a/ or /e/, but here we see /o/ instead.

Normal Kartvelian vocalic sound reflexes. (There is somewhat more complexity than this chart shows: there are a number of sound changes specific to Zan and to Svan which alter or reflect this basic pattern.)

The word is found across the family with somewhat unusual sound-reflexes:

Georgian ორთქლი ortkli
Megrelian ორქი orki
Svan უერთქლ wertkl

This vowel sound-reflex of o~o~e is not what we expect from normal cognate sets.