August 27, 2013

The place of Armenian in the Indo-European language family

Journal of Language Relationship • Вопросы языкового родства • 10 (2013) • Pp. 85—137

The place of Armenian in the Indo-European language family: the relationship with Greek and Indo-Iranian*

Hrach Martirosyan

The main purpose of this paper is to present lexical correspondences that unite Armenian with Greek and/or Indo-Iranian. They include shared innovations on the one hand, and isolated lexemes on the other. These two lexical corpora — lexical innovations on an inherited basis and isolated words — can be placed within the same temporal and spatial framework. After the Indo-European dispersal Proto-Armenian would have continued to come into contact with genetically related Indo-European dialects. Simultaneously, it would certainly also have been in contact with neighbouring non-Indo-European languages. A word can be of a substrate origin if it is characterized by: (1) limited geographical distribution; (2) unusual phonology and word formation; (3) characteristic semantics. The material presented here, albeit not exhaustive, allows to preliminarily conclude that Armenian, Greek, (Phrygian) and Indo-Iranian were dialectally close to each other. Within this hypothetical dialect group, Proto-Armenian was situated between Proto-Greek (to the west) and Proto-Indo-Iranian (to the east). The Indo-Iranians then moved eastwards, while the Proto-Armenians and Proto-Greeks remained in a common geographical region for a long period and developed numerous shared innovations. At a later stage, together or independently, they borrowed a large number of words from the Mediterranean / Pontic substrate language(s), mostly cultural and agricultural words, as well as animal and plant designations. On the other hand, Armenian shows a considerable number of lexical correspondences with European branches of the Indo-European language family, a large portion of which too should be explained in terms of substrate rather than Indo-European heritage.

Link (pdf)

7 comments:

eurologist said...

"On the other hand, Armenian shows a considerable number of lexical correspondences with European branches of the Indo-European language family, a large portion of which too should be explained in terms of substrate rather than Indo-European heritage."

I am not so sure of that. The paper is pretty dense, and I only spent a couple of hours - but it could have used some editing. I think while most of the examples are well-argued, some don't convince me. For example:

Arm. surb, o-stem 'pure, clean; holy' seems to be related to Lat. sobrius (Germ. sauber = clean).

The author argues that in a late period around the Indo-European dispersal, Proto-Armenian shared the thematic innovation *loukeno- ‘light, luminous (sphere)’ with Indo-Iranian. I would argue that this widened meaning is very deep PIE, including Germ. Licht, leuchten and Latin lux.

Gr. γί(γ)γλυμος ‘hinge’: why not a cognate of PIE *konk- "to hang" instead of a Mediterranean substrate isolate?

*gorio- ‘drain’ Gr. γοργύριον n. 'subterranean channel'. Seems to be related to Latin gurgulio (German: Gurgel) and PIE *gwere- "to swallow." Again, I see no isolate, here.

*(H)olur-: Arm. olorn, an-stem ‘pea, bean’, Gr. oλυραι, might relate to Germ. Hülle, Hülse (that what covers, esp. re fruit and seeds, IE from the verb hehlen).

In summary, an interesting paper - but it needs some editing by people with a wider linguistic background. And I don't find the substrate arguments much convincing. If anything, I would have liked to see comparisons to known non-IE Anatolian languages or to Semitic.

AdygheChabadi said...

So Armenian is not native to the region where it is now spoken after all. I knew as much from the Hurro-Urartian substrate in the language anyway.

AdygheChabadi said...

@ Eurologist

Your comparisons are incorrect because of sound change constraints. Each IE language has a consistent pattern of sound change from the proto-language.

The words you cited, therefore, must be substratal as the paper asserts. They do not follow the patterns of sound change for those languages.

Many times a word can LOOK IE while not at all being related to IE. This happens in Greek. Greek so thoroughly adapted the substratal non-IE words that they have the appearance of IE words while not being IE at all.

For instance, you cite:

"Gr. γί(γ)γλυμος ‘hinge’: why not a cognate of PIE *konk- "to hang"...The Greek word must be pre-Greek because of sound change constraints. Note PIE: *k does not reflex Greek: *g => it reflexes as Greek: *k

PIE: *o does not reflex Greek: *i, but Greek: *o

One also has to account for the loss of the word-medial PIE: *n in Greek. PIE: *n should reflex Greek: *n

Those are just some of the reasons why the word is not Greek. There is a lot more wrong with that comparison, but I think you get the point.

Let's look at this comparison, Eurologist. It looks the like it could be, but the reality is that it is a false comparison.

"*gorio- ‘drain’ Gr. γοργύριον n. 'subterranean channel'. Seems to be related to Latin gurgulio (German: Gurgel) and PIE *gwere- "to swallow."

First, PIE: *gʷ reflexes as Armenian: *k, Greek: *b, *d before a PIE front vowel (i, e), and *g before or after PIE: *u, Latin: *u [w > v], gu [ɡʷ] after an /n/, Germanic: /gʷ/ > /kʷ/...so already we run into problems with your comparison just off of the first phoneme. The Latin word is either borrowed from Greek or from the same substratal language as Armenian and Greek.

"*(H)olur-: Arm. oloṙn, an-stem ‘pea, bean’, Gr. oλυραι, might relate to Germ. Hülle, Hülse (that what covers, esp. re fruit and seeds, IE from the verb hehlen)"

Prevocalic PIE: *h₁ = [h] (Meier-Brügger) would be lost in Armenian and Greek. PIE: *o reflexes Armenian: o, a, Greek: *o, and PIE: */o/, */a/ merge in Germanic as */ɑ/. No match here. Also the word appears to be Semitic (Akkadian) actually, Eurologist.

Lastly, you cite:

"Arm. surb, o-stem 'pure, clean; holy' seems to be related to Lat. sobrius (Germ. sauber = clean)"

This could not be related to Latin. PIE: *s does not reflex Armenian: s, but Armenian: h. It reflexes /s/ before a plosive (p, t, k) and Ø (null) intervocalically (between vowels). PIE: *ḱ reflexes as Armenian: s and Latin: c [k], so one can see already they are not related terms. Also PIE: *sk reflexes Latin: sc [sk] as the paper adduces these forms, *ḱubʰ -ro- or *(s)kubʰ -ró- . Furthermore, the paper does not say this word is substratal from what I read and understood.

Just because words look alike does not make them cognates.

I hope this helps.

eurologist said...

AdygheChabadi,

Thanks for your contribution. I will need some time to digest your comments, but I disagree in principle because every single thing I stated is from generic PIE derivations, including known sound changes - nothing was really debated, at this point, to my knowledge. The assertion that those elements have a substratum element or origin is therefore absolutely strange, to me.

eurologist said...

I do agree that some of the words appear to be of Semitic origin - hence my comment that it would have been more productive to explicitly state connections to non-IE Anatolian or Semitic language origins.

DDeden said...

AdygheChabadi

might it be the case that Hurro/Urartian was dialect Georgian, and Armenian overlaid it?

Eurologist
" *loukeno- ‘light, luminous (sphere)’ with Indo-Iranian. I would argue that this widened meaning is very deep PIE, including Germ. Licht, leuchten and Latin lux."

deeper than the ocean... lu/li as deluge & 'flood of light/water'

mongolu (Mbuti) dome hut
igloo/igdlu (Innuit) dome hut
harigolu (India) dome boat
kullik/qidlic (Innuit) lamp bowl

AdygheChabadi said...

@Eurologist

Well, those words do not conform to any known IE language so they must be substratal. I know for a fact the Greek word γοργύριον is not related to PIE: *gwere. Greek already has a word derived from this root, Old Greek: bibrṓskō 'to eat, consume', aor. ébrōn. pf. bébrōka, hom. pf. opt. bebrṓthoi̯s; borǟ́ 'Wolverine (a predator)', boró- 'greedy' ; brō̂ma n., brṓmǟ f.; brōtǘ-s, brō̂si-s 'feed, food'; brōtḗr 'eater; eating, consuming'; bárathro-n, ep., ion. bérethro-n, ark. dzerethro-n (< *dérethro-n) 'maw, abyss'. The German and Latin reflexes are different for this word/ root also.

So no, your comparisons remain incorrect for known IE languages. As for unknown IE languages...they are just that, unknown.

@DDeden

I don't think Kartvelian and Hurro-Urartian are related. It is proposed that Hurro-Urartian maybe related to Northeast Caucasian (Nakh-Daghestanian) in an "Alarodian" family. Though, I would not be surprised if there are a few Kartvelian words in Armenian.