March 09, 2014

The place of the Armenian language in the Indo-European family

A very interesting talk at the Library of Congress making a good case for a Greek-Phrygian-Armenian clade within the Indo-European family.
 


Of course the testimony of Herodotus is -at least for me- the most convincing argument for an ultimately Balkan origin of the Phrygo-Armenians.

Modern Armenians are quite distinct from modern populations of the Balkans but who knows how both they and the populations of the Balkans have changed since the beginning of the Iron Age when the Phrygians established themselves in Asia Minor? The recent study by Hellenthal et al. did not find any good evidence for recent admixture in Armenians but this might be due to (i) Armenians being unmixed descendants of Proto-Armenians, (ii) Armenians being near-unmixed descendants of pre-Armenians, or (iii) the method not having enough power to detect admixture.

My guess is that the relative remoteness of the Armenian highlands coupled with the heterodox position of the Armenian church hindered substantial gene flow into the Armenian population over at least the last 1,500 years.

14 comments:

AgnosticThought said...

Without extensive archaic DNA, this linguistic analysis of Armenian is just an exercise in futility and likely to misguide us. The biggest obstacle here is Turkey's obfuscation of the Armenian Highlands. Granted, the BEAN project is an attempt at this, but without extensive analysis of the Highlands, it will largely remain incomplete. Turkey's lack of interest in the indigenous history of the population, as well as groups that have remained largely conserved, is a huge issue that will not resolve until their revisionist policies are rescinded.

Gary said...

I listened to the talk, and Greek, Armenian, and Indo-Iranian are not the only languages that retains pattern of initial vowel in that particular class of words:

'bone'
ὀστοῦν/ostoun (Ancient Greek)
oskor (Armenian)
óhstien (Mohawk)
ostokhān (Persian)
kost (Czech)

'eye'
ὀφθαλμός/ophthalmos (Ancient Greek)
ačʿkʿ (Armenian)
okà:ra (Mohawk)
oculus (Latin)
akis (Lithuanian)
oko (Czech)

One area where Greek and Armenian differ is in the first person pronoun: Armenian es appears to be related to Kurdish ez , min. Greek and Latin ego on the other hand may be derived from first-person pronomial affixes in the Iroquoian family:

Mohawk (Pronomial affixes) ke- /ka- /ki- /ken- /ko- /wake- /waka- /waki- /waken- /wako- /wak-
Cherokee (Class B affix) aqu- /ag(i),

while Armenian es and Kurdish ez may ultimately be derived from the Cherokee standalone pronoun form for ‘I’, ayv (Note: ‘v’ represents nasal ‘u’) --> Saka azu --> Old Church Slavonic azu --> Lithuanian .

However, words for 'water' in Greek and Armenian, nero and ǰur respectively, appear to be more likely derived from Yeniseian languages: e.g., Kett ul/Kott ur. Hydronymic terms in the Balkans tend to confirm that languages related to Yeniseian may have been spoken there prior to Indo-Europeanization.

andrew said...

A fairly conventional narrative for the origins of the Armenia language assume that the Indo-European Greek language (to distinguish it from pre-Greek languages spoken before Indo-Europeans arrive) appears in the general vicinity of the Aegean ca. 2000 BCE +/- a couple centuries at which point there is rapid language change due to substrate influences and language split differentiation effects.

At about the same time the Hittite language expands with the Hittite empire (mostly displacing non-Indo-European languages present in Anatolia) starting around the same time until the Hittite empire collapses ca. 1200 BCE. Hittite likewise experiences substrate and language differentiation effects involving people linguistically quite different from the pre-Greeks.

The Phyrgians speak a distinctive dialect of Greek on the Hittite empire fringe (with some minor Hittite linguistic influence), and when the Hittite empire collapses there is a folk migration of Phyrgians to "greater Armenia" sometime after the collapse of the Hittite empire ca. 1200 BCE and before classical Greek and Roman civilization start to emerge ca. 800 BCE. Chaos in the Anatolian interregenum of Hittite empire successor states between Bronze Age collapse and the rise of Iron Age civilizations, aided and abetted by a multi-century arid period isolates these Armenians from the Hellenic linguistic cousins and puts them geographically close to the Indo-Iranians of Persian who derived from an entirely different branch of the Indo-European languages from the Greeks, as well as to weakened local Caucasian substrate populations. Isolation and areal influences from the nearby Indo-Iranians causes the Armenian language to change in the direction of old Persian which has similar substrate influences to Armenian upon its arrival in its homeland. Isolation and areal influences simmer for about 1100-1500 years until Alexander the Great comes along and Hellanizes everyone from Anatolia to the fringes of South Asia, reinjecting a new layer of Greek influence into Armenian. After centuries of this second wave Greek influence on Armenian, the language becomes much more isolated until the 19th century CE or so, as Armenians become an island surrounded by Muslims with whom they have little linguistic sharing.

The end result is an Armenian language that is ultimately a genetic descendant of the family of languages that spawn ancient Greek, but with more Indo-Iranian characteristics from areal influences than any of the other Western Indo-European languages, almost giving a false impression that it holds a basal position linguistically.

Dienekes said...

However, words for 'water' in Greek and Armenian, nero and ǰur respectively, appear to be more likely derived from Yeniseian languages: e.g., Kett ul/Kott ur

"Nero" is not an ancient Greek word for water. It comes from an ancient word meaning "fresh" (νηρός).

Unknown said...

Turkey's lack of interest stems because they only settled in modern Turkey from ~600AD. They are foreigners and not the indigenous ancient people.
Of course what they absorbed in the original people became "turks"

Unknown said...

Thanks for sharing Dienekes. I especially enjoyed the transitional pronunciations midway between the Proto-Indo-European forms and the Classical Armenian variations.

The recent study by Hellenthal et al. did not find any good evidence for recent admixture in Armenians but this might be due to (i) Armenians being unmixed descendants of Proto-Armenians, (ii) Armenians being near-unmixed descendants of pre-Armenians, or (iii) the method not having enough power to detect admixture.

All prior studies point quite consistently towards the second scenario. Today, it is essentially safe to consider the Armenians as the linguistically Indo-Europeanized autochthonous population of their historic homeland. The true challenge facing Armenology is determining where, when and how the Proto-Armenian language parted ways with Greek and Old Phrygian, and how it evolved across the Armenian Highland. My hope is that future genetic discoveries can shed further light on this axiomatic issue.


Jim said...

"Today, it is essentially safe to consider the Armenians as the linguistically Indo-Europeanized autochthonous population of their historic homeland."

The same thing is probably true of the Kurds as well.

Gary said...

Dienekes noted: "Nero" is not an ancient Greek word for water. It comes from an ancient word meaning "fresh" (νηρός).

Thanks for the correction. I was looking at the wrong column in my spreadsheet.

There are some issues with the etymology for 'water' (hydor) online in Etymonline.com: ... "from PIE *wod-or, from root *wed- (1) "water, wet" (cf. Hittite watar, Sanskrit udrah, Greek hydor, Old Church Slavonic and Russian voda, Lithuanian vanduo, Old Prussian wundan, Gaelic uisge "water;" Latin unda "wave")"

Most PIE word lists I have seen referenced state that the PIE word for 'water' was *akwa, which is very close to the Proto-Iroquoian equivalent *ahwa. This is consistent with the a- affix we see in words for 'water' in other IE languages such as Italo-Celtic (e.g. aqua) and Indo-Iranian (e.g. Persian ab).

The version of Hittite I have seen shows that it originally written wa-a-tar. It appears that two initial syllables were transposed (i.e. metathesis) from the original IE word and that this is a compound word composed of words for 'water' in two different languages. Later, the a- was dropped entirely and we get the form seen in some other IE languages, such as the Germanic family. The -tar word particle is still found in Persian as the word for 'wet" and as part of Proto-Iroquoian words for 'lake': *-ōtar- and *-nyatar. (While the word for 'lake' in most European IE languages is thought to be derived from PIE *laku- - 'basin', the Iroquoian form apparently survives as 'tarn' for 'mountain lake'.) In turn, -tar is quite likely related to water words in Dene-Yeniseian languages: ta, tʰa, and tu. Elsewhere in these blogs I have described how the Yeniseian forms for 'water' (ul, ur appear to have turned up in Balkan hydronyms, and appears to contribute to Lithuanian word aleti ('flooded').

AdygheChabadi said...

To anyone:

I think DNA has told as much as it can tell.

Linguistics has shed much light on this issue though. Armenian has a substrate of Hurro-Urartian adoptions. This means that Armenian overlay a language that already existed in the area. This strongly indicates that Armenian is intrusive. To further add to this, there is linguistic evidence of Armenian sharing some "Pre-Greek" lexicon with Greek. Armenian also has Kartvelian adoptions, even some that can be reconstructed at the proto-level for Armenian. Armenian shares this with Greek also. Many "Pre-Greek" words are from a Kartvelian-esque language. There are a couple of "Pre-Greek" Hattic adoptions shared by both Armenian and Greek. There is areal morphological substratal influence from the Northwest (Abkhaz[o]-Adyghean), Northeast (Nakh-Dagestanian), and South (Kartvelian) Caucasian languages in Armenian also (glottalic stops). Again, strongly suggesting Armenian intrusiveness. There are other linguistic evidences also.

(I need to be better about proof-reading)

Gary said...

After doing some research, it appears that the creators of the Etymologyonline entry for 'water'/hydor had an agenda: to promote an alternative derivation of PIE words for 'water' that indicate that it is derived from Proto-Uralic weti. Of course their dog doesn't hunt because their derivation ignores words for 'water' in IE families that do not have the '-tar' component.

I did find an intriguing item that might have some bearing on this issue in an unlikely place: an entry in Curtius' discursive and barely readable "Principles of Greek Etymology" referencing words from an unspecified language in which -tar was likely the word for 'water':

"tar-a-wa-s, tar-i-s 'boat', tar-anta-s 'sea'

I have yet to identify this language, but it is a good candidate for the one that contributed the -tar component to Hittite. I also have not been able to find a good etymology for the name of the son of Poseidon, Taras, which I suspect may also be related to water and may be derived from some Balkan or Anatolian substrate language.

By the way, the theory that the Iroquoian language of North America is somehow related to Greek goes back to almost the beginning of the 18th century with Pere Joseph-Francois Lafitau, a Jesuit missionary to the Iroquois. Lafitau believed the Iroquois to be the descendants of the Lycians, based in part of similarities in customs and even proposed etymologies for Greek words based on Iroquoian languages. Rev. J. A. Cuoq, studying Iroquois almost two centuries later, was also struck by the resemblance of Iroquoian languages to Greek.

It is interesting that there appears to be convergent evolution between Modern Greek and North American languages. I ran across this interesting article on polysynthesis in Modern Greek and how it compares to Iroquoian languages.

http://www.linguistik-online.de/34_08/charitonidis.html

Jim said...

"Again, strongly suggesting Armenian intrusiveness. "

Also strongly suggesting a much wider early distribution of all those language families.

Armenian is intrusive, but historically its center of mass was much westward and southward of its present location. The pattern is continual eastward expansion with a loss of territory in the west. That suggests all those early influences from those groups took place when Armenian was much further west.

AdygheChabadi said...

@Jim

I agree with you. Especially those "Pre-Greek" words only shared by Greek and Armenian. They are excellent evidence of Armenian being spoken much further west. Hittite adoptions are present in Armenian as well.

AgnosticThought said...

The westward migration of Armenian from Lesser Armenia, into the East is no different than the Northern and Eastern migration of Kartvelian languages into non-Karvelian Caucasian language groups. Even in recent times, the expansion of Karvelian languages has been documented. Regardless of where the language had origins, ultimately, the population is most likely conserved as it is among Armenians, even within the east.

Unknown said...

In finnish, tar is probably original name for piss. On this basis tar-a-wa-s can be Terävä (sharp flavour) and tar-anta-s (to give the tar). Todays Tarjoilija/waiter and parantaja/healer. Of course the tar (resin) has been used for medicinal purposes, but i think, in this context, this is an older stratum.