Proto-Armenian is the reconstructed, unattested ancestral stage of the Armenian language, encompassing its development from Proto-Indo-European to the earliest attestations of Classical Armenian. As Armenian is the sole known member of its branch of the Indo-European languages, its prehistory cannot be reconstructed through direct comparison with closely related languages. Linguists instead combine internal reconstruction with evidence from Proto-Indo-European and other Indo-European branches. Armenian is an independent branch of the Indo-European family and is distinguished by its characteristic phonological changes, including a predominance of satem-like developments, although it is not classified as either a centum or satem language.
Definition and origins Proto-Armenian, as the ancestor of only one living language branch, lacks a rigid, singular definition, but it is generally held by scholars to encompass the complex developmental stages of Armenian between Proto-Indo-European and the earliest written attestations of Classical Armenian in the 5th century AD, beginning with the Bible translation of Mesrop Mashtots. Although it is not a proto-language in the strict comparative sense of giving rise to multiple sub-languages, "Proto-Armenian" remains the standard linguistic term for the dialectal lineage defined by distinctive phonological changes within the Indo-European family, most notably exhibiting more satemization than centumization, though without belonging strictly to either group. Linguists situate Proto-Armenian within an early dialect continuum alongside Proto-Greek, Phrygian, Albanian, and Proto-Indo-Iranian, positioning it conceptually between Western centum dialects and Eastern satem dialects. Proposed phylogenetic models place the divergence of Pre-Armenian as early as approximately 2800 BC, with computational models proposing a close relationship between Pre-Armenian and Pre-Greek after 2500 BC, a grouping reflected in shared phonological, morphological, and lexical features such as the verbal augment, prothetic vowels, and the change *s > *h, alongside unique morphological connections to Balto-Slavic languages. Recent archaeogenetic studies associate the emergence of Proto-Armenian in the Armenian Highlands with Middle Bronze Age populations of the Trialeti-Vanadzor culture carrying western Yamnaya-related Steppe ancestry shared with contemporaneous Aegean groups. Throughout its pre-attested development, Proto-Armenian underwent extensive adstrate and substrate language contact, absorbing multiple lexical layers from Anatolian languages (Luwian, Hittite), Hattic, Hurro-Urartian languages, Semitic languages (Akkadian, Aramaic), and Iranian languages (Persian, Parthian), as well as secondary influences from Greek and Arabic.
Further linguistic evidence comes from Indo-European horse terminology. Rasmus G. Bjørn argues that descendants of Proto-Indo-European *h₁eḱwos spread with Armenian across the Caucasus into Southwest Asia in the late 3rd millennium BC, contributing to a Wanderwort attested in early literary languages from Hurrian through Sumerian and Egyptian. He associates this distribution with the Bronze Age expansion of horses from the Western Steppe Zone, whose dated genetic signal coincides with the proposed spread of Armenian into the Caucasus. Bjørn interprets Armenian ēš, "donkey", as a semantic shift from "horse" to "equid". The alternative Sumerian designation for the horse ("donkey of the mountains") is likewise interpreted as evidence for a montane route of entry into Southwest Asia.
Phonological development of Proto-Armenian The Proto-Armenian sound changes are varied and eccentric (such as *dw- yielding erk-) and, in many cases, uncertain. That prevented Armenian from being immediately recognized as an Indo-European branch in its own right, and it was assumed to be simply a very divergent Iranian language until Heinrich Hübschmann established its independent character in 1874.Karl Brugmann, Grundriss der vergleichenden Grammatik der indogermanischen Sprachen (1897) Das Armenische (II), früher fälschlicherweise für iranisch ausgegeben, von H. Hübschmann KZ. 23, 5 ff. 400 ff. als ein selbständiges Glied der idg. Sprachfamilie erwiesen
The development of voicing contrasts in Armenian is notable in being quite similar to that seen in Germanic, a fact that was significant in the formation of the Glottalic Theory. The Armenian Consonant Shift has often been compared to the famous Grimm's Law in Germanic, because in both cases, Proto-Indo-European voiceless stops became voiceless aspirates (with some complications with regard to Proto-Indo-European *p), the voiced stops became voiceless, and the voiced aspirates became voiced stops.Fortson, Benjamin W. 2004. Indo-European Language and Culture. Page 340-41. Meanwhile, Armenian shares the vocalization of word initial laryngeals before consonants with Greek and Phrygian:Fortson, Benjamin W. 2004. Indo-European Language and Culture. Page 342, 402 Proto-Indo-European ("man", "force") renders Greek anḗr, Armenian ayr from a Proto-Armenian *aynrFortson, Benjamin W. 2004. Indo-European Language and Culture. Page 342. and Phrygian anar ("man"), which may be compared to Latin Nero and neriōsus ("strict"), Albanian njeri, Persian nar, Sanskrit nara, and Welsh nerth.
In certain contexts, the aspirated stops are further reduced to w, h or zero in Armenian: Proto-Indo-European (accusative) *pódm̥ "foot" > Armenian ոտն otn vs. Greek (accusative) póda, Proto-Indo-European *tréyes "three" > Armenian երեք erekʿ vs. Greek treis.
+PIE stop consonants in ArmenianPIE→ ArmenianSpecial developments and exceptions*pհ h∅ (before օ ō); ւ w (after vowel); փ pʿ*tթ tʿտ t (after sibilant); յ y (between vowels); դ d (after ր r); ∅ (before consonant); ք kʿ (PIE *tw → Arm. ք kʿ); ւ w (before ր r and between back-vowels); ն n (PIE *-nt → Arm. ն n)*ḱս sշ š (PIE *ḱw → Arm. շ š); ւ w (before ր r); ∅ (before consonant)*kք kʿխ x (PIE *kH → Arm. խ x); գ g (PIE *[R]k → Arm. գ g); չ čʿ (PIE *ky → Arm. չ čʿ)*kʷչ čʿխ x, գ g, ք kʿ*bպ p*dտ tրկ rk (PIE *dw → Arm. րկ rk)*ǵծ c*g, *gʷկ kծ c*bʰբ bւ w (between vowels)*dʰդ dջ ǰ (PIE *dʰy → Arm. ջ ǰ)*ǵʰձ jզ z*gʰգ gջ ǰ *gʷʰջ ǰ; ժ ž +PIE resonants and fricatives in ArmenianPIEArmenianSpecial developments*mմ mն n (word-finally) ∅ (before *s)*n ն n∅ (before *s)*lլ l, ղ łղ ł (adjacent to consonants and word-finally)*rր r*wգ gվ v, ւ w*yջ ǰ∅ (between vowels)*sհ hս s (after consonant or before obstruent); ∅; ք kʿ (word-finally, also PIE *sw → Arm. ք kʿ)*h₁∅է ē (initially)*h₂, *h₃հ hա a (initially); ∅ +PIE vowels in ArmenianPIEArmenianSpecial developments*aա a*e է ēա a (adjacent to *h₂); ի i (before h₁ or nasals); օ ō (after *h₃); ու u (before *h₃)*iի i*oօ ō*uու u
History The origin of the Proto-Armenian language is subject to scholarly debate. The Armenian hypothesis would postulate the Armenian language as an in situ development of a 3rd millennium BC Proto-Indo-European language, while the Kurgan hypothesis suggests it arrived in the Armenian Highlands either from the Balkans or through the Caucasus. The arrival of such a population who spoke Proto-Armenian in the Armenian Highlands is assumed to have occurred sometime during the Bronze AgeRobert Drews. Militarism and the Indo-Europeanizing of Europe. Routledge. 2017. p. 228.|quote="The vernacular of the Great Kingdom of Biainili was quite certainly Armenian. The Armenian language was obviously the region's vernacular in the fifth century BC, when Persian commanders and Greek writers paired it with Phrygian. That it as brought into the region between the early sixth and the early fifth century BC, and that it immediately obliterated whatever else had been spoken there, can hardly be supposed; [...] Because Proto-Armenian speakers seem to have lived not far from Hurrian speakers our conclusion must be that the Armenian language of Mesrop Mashtots was descended from an Indo-European language that had been spoken in southern Caucasia in the Bronze Age." or at the latest, during the Bronze Age Collapse around 1200 BC, according to this theory.Greppin, John A.C. and Igor Diakonoff Some Effects of the Hurro-Urartian People and Their Languages upon the Earliest Armenians, Oct–Dec 1991, pp. 727.
One of the theories about the emergence of Armenian in the region is that Paleo-Balkan-speaking settlers related to Phrygians (the Mushki or the retroactively named Armeno-Phrygians), who had already settled in the western parts of the region before the Kingdom of Van was established in Urartu, Katvalyan, M. and Karo Ghafadaryan. Ուրարտու [Urartu]. Armenian Soviet Encyclopedia. Yerevan: Armenian Academy of Sciences, 1986, vol. 12, pp. 276–283.Uchicago.edu had become the ruling elite under the Median Empire, followed by the Achaemenid Empire., p. 50 The existence of Urartian words in the Armenian language and Armenian loanwords into UrartianPetrosyan, Armen. The Armenian Elements in the Language and Onomastics of Urartu. Aramazd: Armenian Journal of Near Eastern Studies. 2010. (https://www.academia.edu/2939663/The_Armenian_Elements_in_the_Language_and_Onomastics_of_Urartu) suggests early contact between the two languages and long periods of bilingualism.
According to the Encyclopedia of Indo-European Culture:
Findings in Armenian genetics from 2019 reveal heavy mixing of groups from the 3000s BC until the Bronze Age Collapse. Admixture signals seem to have decreased to insignificant levels after c. 1200 BC, after which Armenian DNA remained stable, which appears to have been caused by Armenians' isolation from their surroundings, and subsequently sustained by the cultural, linguistic and religious distinctiveness that persists until today. The connection between the Mushki and Armenians is unclear as nothing is known of the Mushki language. Most modern scholars have rejected a direct linguistic relationship with Proto-Armenian if the Mushki were Thracians or Phrygians.Kim Ronald I. Greco-Armenian. The persistence of a myth // Indogermanische Forschungen. — 2018. — 123. Band. — S. 247–271. Additionally, recent findings in genetic research does not support significant admixture into the Armenian nation after 1200 BC, making the Mushki, if they indeed migrated from a Balkan or western Anatolian homeland during or after the Bronze Age Collapse, unlikely candidates for the Proto-Armenians.
However, as others have placed (at least the Eastern) Mushki homeland in the Armenian Highlands and South Caucasus region, it is possible that at least some of the Mushki were Armenian-speakers or speakers of a closely related language.The Mushki Problem Reconsidered Some modern studies show that Armenian is as close to Indo-Iranian as it is to Greek and Phrygian.
An alternate theory suggests that speakers of Proto-Armenian were tribes indigenous to the northern Armenian highlands, such as the Hayasans, Diauehi or Etiuni. Although these groups are only known from references left by neighboring peoples (such as Hittites, Urartians, and Assyrians), Armenian etymologies have been proposed for their names. While the Urartian language was used by the royal elite, the population they ruled was likely multi-lingual, and some of these peoples would have spoken Armenian. This can be reconciled with the Phrygian/Mushki theory if those groups originally came from the Caucasus region or Armenian Highlands.