I. M. Diakonoff

The Pre-history of the Armenian People*
Predystoriia armianskogo naroda


Erevan, 1968, English Translation by Lori Jennings
(Delmar, New York, 1984)

* This material is presented solely for non-commercial educational/research purposes.
[Page 101]

Chapter 3.
The Formation of the Armenian People

1. The Ethnic Composition of the Population of the Armenian Highlands at the Beginning of the 1st Millennium B.C.

We began with a characterization of the ethnic composition of the population on the Armenian Highlands and the surrounding countries during the 3d and 2d millennia B.C.; now we must try to characterize the ethnic composition of the population there at the beginning of the 1st millennium B.C. Just as before, we will dwell chiefly on the linguistic features of the ethnos, as they are more relevant than the anthropological and less vague than the cultural-historical features.

1.1. The Linguistic Situation at the Beginning of the 1st Millennium B.C.

We encounter some of the same linguistic groups now as before. Thus the mountains of Iranian Azerbaijan and Kurdistan were occupied to a certain extent by the descendents of the same Quti who lived here in the 3d millennium B.C. (1). However Iranian languages, part of the Indo-Iranian linguistic branch, had already also spread widely in Iranian Azerbaijan and in the entire Highlands of Iran. The northernmost of them was the language of the Medes (in the Armenian sources--medac'i or mar) (2). These languages belong to the same branch of Indo-European languages as the so-called "Mesopotamian-Aryan" or "Western Indo-Iranian" of Mitanni which we encountered in the 2d millennium B.C. (actually not in Mesopotamia, at least as a living language, but probably near Lake Urmia). The "Western Indo-Iranian" is usually supposed to have been Indo-Aryan, while the Iranian languages belonged to another subgroup of Indo-Iranian. If in the earliest period they ever penetrated to the Armenian Highlands as well (3), they did not leave any stable linguistic traces here. Their appearance in Iranian Azerbaijan has been dated variously, e.g., to the 11th or 10th centuries [102] B.C. (4); however, still earlier dates are now proposed as, e.g., the middle or even the beginning of the 2d millennium B.C. The present author has suggested that the language attested by the Mitannian sources may not have belonged to the Indo-Aryan subgroup at all, but that the division of the Indo-Iranian languages into Indo-Aryan and Iranian postdates the contact of the undivided language with the Hurrians of Mitanni (*1).

The Urartians occupied the center of the Armenian Highlands and the upper valley of the Upper Zab; their relatives the Hurrians are traceable in some regions along the southern and western periphery of the Highlands, possibly from Lake Urmia to the valley of the river Coroh. In Syria and Mesopotamia the Hurrians as such disappeared between the 11th and 9th centuries. The population here was Aramized in their language as a result of the powerful penetration of nomadic Aramaean tribes into these lands by the beginning of the 11th century, and then subsequently by the Assyrian policy of forced deportation, resettlement, and mixing of ethnic groups. By the 7th century B.C. the Aramaic language had already supplanted the Akkadian to an appreciable degree in the daily life of the population of Mesopotamia (5),

Farther to the west, in the mountains of Cilicia, in the valley of the Upper Euphrates, in the Cilician Taurus, and in some regions of Northern Syria, we may suppose a Luwian ("Luwian Hieroglyphic") population. In Asia Minor the Anatolian languages were also preserved in the valley of the river Gediz, the ancient Hermus (Lydian, belonging to the Hittite subgroup); in the valley of the Menderes (Meander) and to the south of it (Carian); and on the peninsula of Lycia (Lycian dialects, certainly late Luwian). The languages of the Pisidians and the Cappadocians ("White Syrians") are unknown except for proper names. From the town of Side, on the southern coast of Asia Minor, come a few inscriptions in a local tongue; apparently all these languages also belonged to the Luwian group of Anatolian languages.

Eastern Pontus, Colchis, western and part of central Transcaucasia were probably occupied by Georgian-speaking tribes -- the Chaldians, the Colchi, the Saspires, and others, some of whom are mentioned in Urartian (6) and Greek sources (7). Probably most of the languages of the Abkhazo-Adyghian group had already been pushed to the north, into the territory of their present distribution. However in the country of Kasku, which at this time seems to have been situated on the upper reaches of the Halys and/or in the valley of the Lycus(?), the old language of the Kaska may have been preserved, but only if they had not been absorbed by a numerically greater aboriginal element or by one of the newly[103] arrived groups which invaded northern Asia Minor in the 12th century B.C. and later. To the west of Pontus, Paphlagonia was probably still speaking Palaic.

Greek colonies now existed in all the coastal areas of Asia Minor, and the Aegean coast of the peninsula was completely settled by Greeks--Aeolians, Ionians, and Dorians. On the coast of Cilicia there were Greeks and Phoenicians.

Thraco-Phrygian-speaking groups show up as an important new ethnic element. In Asia Minor their most important representatives were the Phrygians. The center of their territory was the valley of the river Sakarya (Sangarius) and the central Anatolian plain, but their inscriptions are also found in the area which was earlier occupied by the speakers of Hittite (8), while objects of their archaeological culture are found even in Pontus (Akatan), in the Cilician Taurus (Elbistan), and on the right shore of the Upper Euphrates (Malatya) (9). The classical sources call the area between the river Sangarius and the Sea of Marmara "Phrygia Minor."

The northwest corner of Asia Minor was occupied by the Mysians, who spoke a Phrygian (or Thracian) dialect strongly influenced by Lydian (10), while the western part of the Black Sea coast of Asia Minor, beginning with the Bosporus, was occupied by the Thracian people of the Bithynians, who had moved from the Balkans later than the others. The Mariandyni were possibly the remnant of the pre-Bithynian population; to the east of Bithynia lived the above-mentioned Paphlagonians.

The questions of who inhabited the country of Tabal (Hebrew Tubal) in the Cilician Taurus, and of how to identify the Mushki (Hebrew Meshech or, better, *Moshak') has been much debated. The latter were recorded by the Assyrian, Urartian, and Hebrew sources of the 8th-6th centuries B.C. as located in Asia Minor to the west of the Cilician Taurus, while the Assyrians of the 12th-9th centuries B.C. also recorded the Mushki in the valley of the Upper Euphrates and in the area between the lower reaches of the Arsanias and the Sasun Mountains. The inhabitants of Tabal and the Mushki are frequently identified with the Moschi and the Tibareni, tribes which lived, according to the data of Greek and Roman authors (11), in Pontus and who, in all probability, must have belonged to Georgian or Abkhazo-Adyghian-speaking peoples. However the problem is not resolved by a simple identification of Tabal and the Mushki with the Tibareni and the Moschi.

If it were not for the superficial similarity between the name of the province of Tabal and the name of the tribe of the Tibareni, hardly anyone would doubt that this province, just as in the 2d millennium B.C., was part of the area of the Luwian [104] ("Hieroglyphic") language, since numerous official inscriptions are written in this language, and also the proper names of kings are Luwian (12). However it cannot be denied that the dominance of this language here might not have been undisputed. Even if some of the localities here certainly preserved the ancient Luwian names (13), there also did appear new names (14)--perhaps as the result of an influx of a new population. As far as the Mushki are concerned, at least their western group can be reliably identified with the Phrygians (15).

Greek authors (16) and some modern scholars have assigned Proto-Armenian to the languages of the Thraco-Phrygian group. In any event it has no place in any other branch of the Indo-European languages (17). It is our aim to establish the original area of where the speakers of this language did settle in Asia Minor.

For a long time it was the common opinion of scholars that Armenians and the Armenian language appeared in the Highlands when the term "Armenia" was attested for the first time, i.e., in the 6th century B.C., and that the history of the Armenian people should begin from this period. This point of view must be regarded as naive and in no way satisfactory.

1.2. Principles of Approach to Ethnonymic Terms

Ethnic names are a completely unreliable source for the restoration of the ethnic and linguistic history of a people. The sources very rarely allow us to establish whether or not a given ethnic designation is a self-appellation (autonym) or a name which was used only by neighboring peoples (heteronym). In the latter case it may be very general, encompassing an entire group of different peoples similar in culture (for example, the "Tatars" of Eastern Europe and Asia, the "Indians" of America, the "Scythians" of the Greek authors) (18). Or, vice versa, the designation may originally be the local name of the inhabitants of a limited region, but used by their neighbors in a broader sense (for example, Latvian krievs, originally the designation of the neighboring Slavonic tribe of the Krivichi, now meaning "Russian"; analogous is the derivation of the French allemand with the meaning of "German," from the tribe of Alamanni; Georgian Somexi "Armenian," which strictly signifies an inhabitant of the region of Suhmu on the Upper Euphrates. Or it can be a traditional name, transferred from the earlier inhabitants of a given locality (Gauls in the sense of "Frenchmen," Sarmatians in the sense of "Slavs," Finnish venalainen "Wend," the name of an Old Slavonic people, in the sense of "Russian") or even from a completely different people, on the basis of some sort of [105] historico-cultural associations (for example, the term tadjik, now meaning the Iranian-speaking population of Soviet Central Asia, originally meant "Arab") (19).

Finally purely accidental sound coincidences are frequently found in ethnonyms--cf. Albanians in the Balkans, Albani (Arm. Aghuank') in ancient Transcaucasia, Albani--the inhabitants of the city of Alba in ancient Italy, Albany in Britain and the ancient name of England itself--Albion, the German tribe of the Alamanni (cf. the French name of Germany--Allemagne), and many others. All of these names have nothing in common as to their ethnic origin; trying to make sense of such coincidences is useless. However among the suggested ancestors of the Armenians some scholars have included, with no more reason, such ethnonyms and toponyms as the Arimi, Arme, Urmie, the Urumeans, etc., and sometimes even the Aramaeans. If these last do not enjoy popularity as candidates for being the ancestors of the Armenian nation, then it is only because they are known to have spoken a Semitic language, in no way related to Armenian. If this were not commonly known to be the case, the Aramaeans would certainly be introduced into discussions of the hypothetical ethnogenesis of the Armenians, the more so since they were their southern neighbors. Obviously similarity of names should be backed up by other, more weighty data; otherwise it is completely unreliable.

But even if it is known to us quite precisely that a certain ethnic term is a self-appellation (autonym), it is not always possible to rely on it in ethnogenetic constructions. An ethnic self-designation may change (for example, in the Middle Ages the Greeks at one time called themselves the Rhomaioi, i.e., actually Romans.) While growing into a nation an ethnic unit may sometimes accept as its self-designation a name of foreign or even accidental derivation. (Thus the French call themselves francais after the name of the Germanic tribe of the Franks, who played a secondary role in the ethnogenesis of the French nation. The self-designation Tadjik, as has already been mentioned, meant "Arab," then "a man of the Arabic Muslim culture," and only later the modern Iranian-speaking nation of Central Asia; but that group did exist before this particular name was adopted. The self-appellation of the Americans is accidental, being derived from the name of the geographer Amerigo Vespucci).

Sometimes the designation of earlier inhabitants of a given country, who have long since disappeared, is preserved as a self-appellation (for example, British is now a self-designation of the Anglo-Saxons, who supplanted and destroyed the ancient Celtic tribes of the Britons).

[106[ It is very important to bear in mind that at the early stages of a society's development there does not, as a general rule, exist a comprehensive self-appellation for the entire ethnic mass--the people usually name themselves only after their local community (thus the Phoenicians called themselves "Sidonians," "Tyrians," etc.) or tribe (the Slavic "Krivichi," "Drevlians"; the Germanic "Vandals," "Franks," etc.). Or take another case: the Slavs, Slav. Slovienie, are "those of (intelligible) speech" in contradistinction to the Niemcy--"mutes," i.e., foreigners (at present it is the Russian designation of the Germans, but earlier it was used in a much broader sense). The majority of the peoples of the Far North of the USSR call themselves simply "people." Of a similar derivation is the self-appellation of the Germans--Deutsche; the term "Germans" was not used by the Germanic tribes themselves. In the Ancient Near East the Sumerians had no common self-appellation (they called themselves "the Black-headed ones," but this term also included the Semitic-speaking inhabitants of Mesopotamia).

This is why searching in antiquity for various similarly sounding ethnonyms is an unreliable way to reconstruct ethnic prehistory, and it leads mostly to great confusion. That the words sound similar not by mere chance can be established only after taking into consideration all the regular developments in the languages under comparison and the historical changes in their phonetic structure over the ages. Unfortunately we are rarely able to get the necessary linguistic data for very remote periods. It is naive to suppose that words or proper names, if related in their origins, will in any event retain a similarity in sound through the course of centuries and millennia (20). A too strong similarity in names, separated by a large interval of time, is more often than not a sign of this similarity being fortuitous.

Therefore, in dealing with the question of when and where the speakers of the Proto-Armenian language first appeared, we shall have to proceed not from a search for ethnonyms, but from other, more objective data, drawing in their support on the data of ethnonyms and toponyms only with caution (21).

2. The Problem of the Original Speakers of Proto-Armenian

2.1. The Historical Composition of Old Armenian

Just as every language with a long history, Old Armenian contains many strata of varying origin. A number of words have been borrowed from Middle Persian, a very large stratum is constituted [107] of words of Parthian origin, and there seem to be a few words from the Old Iranian languages. No traces of contacts with the Scythians appear in the language at all, nor do there seem to be any traces of "Western (Mitannian) Indo-Iranian " (22). But the presence of Iranian words even in great numbers does not indicate that Old Armenian belongs to the Indo-Iranian branch of the Indo-European languages: these words are mainly terms of state administration, of the feudal way of life, of book-learning and various abstract concepts, etc. They do not belong to the basic stock of Old Armenian vocabulary and bear witness only to the fact, well known from the history of Armenia, that the Armenian people had exceedingly long and extensive contacts with the states which successively dominated in Iran and, at times, in the Armenian Highlands and Transcaucasia as well. The most important of these was the Parthian state of the Arshakids (Arsacids) (23).

As Perikhanian has shown, there were at least two strata of words of Semitic, more precisely of Aramaic, derivation in Old Armenian. The more ancient stratum is traceable to one of the old Aramaic dialects of Northern Mesopotamia; these are terms which are chiefly connected with trade and handicrafts (24), and with the state scribal offices. They represent a trace of the existence in Armenia of Aramaic chancelleries, which were created by the Achaemenian Empire, and of trade relations existing between the Armenian Highlands and Mesopotamia in the second half of the 1st millennium B.C. A part of these terms were brought by the Aramaic- and Hebrew-speaking townspeople, who were settled in several cities of Armenia during the rule of Tigran the Great and Artavazd II, in 77-40 B.C. (25). The later stratum represents words of an ecclesiastical-literary nature, which derive from the Syro-Edessan dialect of Aramaic, brought to Armenia with the Christian faith and Christian church (26). There are a few words of Akkadian origin, but all can be shown to have been brought into Old Armenian through the Aramaeans or the Hurro-Urartians (27). There is also a certain stratum of Greek words, which are also mainly of an ecclesiastical-literary nature.

Below these strata, which date from approximately 500 B.C. to A.D. 500 (and partly later), there are other strata. These are words of Hurro-Urartian origin (28). For the time being it is difficult to establish the number of Armenian words belonging to this stratum, since we know very little of the Hurrian and Urartian vocabulary. In all probability many words of Old Armenian (perhaps hundreds) which have not yet been explained will turn out to be words of Hurro-Urartian origin (29). However these words do not belong to the basic vocabulary either, and therefore, of course, Old [108] Armenian cannot be considered to be a Hurro-Urartian language. Hurro-Urartian must be viewed as a substratum of Old Armenian--i.e., the remnant of the language of the local population of the Armenian Highlands preserved during its changeover to Old Armenian.

Kapantsjan has dedicated a series of studies to the discovery of Hittite words in Old Armenian. Not all of his etymologies can bear criticism, but there is no reason to doubt that there actually was also a Hittito-Luwian stratum in Old Armenian (30). Unfortunately scholars have as yet looked only for Hittite words proper, but a long contact between speakers of Hittite (Nesite) and Proto-Armenian is unlikely. At the same time no one has looked for words from other ancient Anatolian (Hittito-Luwian) languages, and in particular for Luwian words, which undoubtedly must have existed in Old Armenian (31). Some of the commonly supposed Hittite words might well turn out to be Luwian or Common Anatolian.

The Anatolian stratum also does not include words of the basic vocabulary (32).

Only by stripping away the above-mentioned strata shall we get to the basic lexical stock of Old Armenian. To this main nucleus belong words that denote objects and notions common to all humanity. Designations of such must have existed even in the very oldest form of the language, and no language can do without them. Therefore there seldom was any valid reason for borrowing their designation from outside. To them usually belong the names of the parts of the body, simple kinship terms, elementary actions and states, a few numerals, etc. (33) The basic lexical stock is not absolutely unchangeable, and new designations for old concepts are occasionally created or borrowed in this field also. It has been estimated that the basic lexical stock of a language is renewed by no more than 15 percent during the course of a thousand years. This ratio is not a hard and fast rule, as was thought for some time, but it gives an approximate idea of the comparative stability of the part of the vocabulary in question. Also the grammatical morphs--prefixes, suffixes, cases, verbal endings, etc.--belong to the basic lexical stock of any language.

An analysis of the Old Armenian basic vocabulary shows without any shadow of doubt that the language is Indo-European. By the same token any question about a possible "dual nature" of Old Armenian is excluded (34). The linguistic ancestor of Old Armenian, Proto-Armenian, certainly was Indo-European and was not related to either the Hurro-Urartian languages, nor to Hattic, nor to the modern Caucasian languages (Abkhazo-Adyghian, Kartvelian, Nakh-Daghestanian), nor to the Semitic languages.

Secondly, it appears that Proto-Armenian did not belong to the [109] Anatolian (Hittito-Luwian), Indo-Iranian, or, e.g., the Slavonic (35) branch of Indo-European, since the phonetic changes in comparison with reconstructed Proto-Indo-European which are peculiar to Old Armenian are different from those in all of these branches (36). Also the basic vocabulary differs substantially from that of the other ancient Indo-European languages (37). Important is the fact that terms relating to cattle- and sheep-breeding, and a few agricultural terms, also belong to the basic lexical stock of Old Armenian.

2.2. The Question of the True Date of the Appearance of the Speakers of Proto-Armenian in the Armenian Highlands

Inasmuch as Old Armenian is not related to the languages of the aborigines of the Armenian Highlands--the Hurrians, Urartians, etc. (although it is precisely these languages which form its substratum), it is clear that it was brought here from the outside. And inasmuch as it does not belong to those branches of the Indo-European family which penetrated into Western Asia in the 3d or first half of the 2d millennium B.C., it must have appeared here either before or after that time.

However there are no traces of the presence in Western Asia of Indo-European languages earlier than the Anatolian and Indo-Iranian languages, such as clearly Indo-European non-Anatolian toponyms, proper names, or borrowed words. According to Djahukyan's theory, Urartian was in collateral kinship with Proto-Indo-European. However we have already noted that the Indo-European features in Urartian, if any, should most probably be explained by influence of an adstratum of the neighboring Anatolian languages. In any event the alleged Indo-European elements, either in the so-called "Haiasan language" or in Urartian, do not seem to share the typical phonetic peculiarities of Old Armenian (38). The remaining solution is that Proto-Armenian arrived in the Armenian Highlands after the middle of the 2d millennium B.C. but of course before the middle of the 1st millennium B.C., the period to which the first strata of borrowings from the Iranian and Semitic languages into Old Armenian are traceable. Thus the only branch of the Indo-European language family to which Old Armenian can be related is Thraco-Phrygian, which in Asia dates from the 12th century B.C. It must be conceded that there is no unanimity among the linguists as to whether Thracian, Phrygian, and Armenian really constitute a single branch inside the Indo-European linguistic family, or whether each of these languages represents by itself a different branch. The question is too complicated to treat it here. We shall use the term "Thraco-Phrygian" conventionally. This is not only [110] because the three languages have in common the negative quality of not belonging to any other known branch of Indo-European. Such usage is justified by the fact that at least from the phonetic point of view, all three constitute a single group with a definite common system of features distinguishing them from all other Indo-European languages. Proto-Armenian is perhaps the more archaic in some respects. Unfortunately the data on possible common features in morphology, vocabulary, and syntax are too scant for safe conclusions. However we do have some positive data which point to a nearer kinship of the Armenian vocabulary to Thracian, Phrygian, and Greek than to any other group of Indo-European languages (39). In addition the Greeks who met speakers of the Armenian language at an extremely early stage of its development--a thousand years before the first written Armenian texts--bear witness that at that time it sounded very "similar to Phrygian" (Eudoxus of Cnidus) and that the Armenians in Asia Minor were considered to have been "settlers having moved away from the Phrygians" (Herodotus). We have no ground to disbelieve this evidence (40). Thus Proto-Armenian could have appeared in the Armenian Highlands no earlier than the 12th and no later than the 6th century B.C.

Beginning with the middle of the 8th to the end of the 7th centuries B.C., the entire territory from the mountains of the Cilician Taurus to the east was engulfed by mighty empires which could hardly have allowed significant ethnic movements through their territories. And in any event such movement could hardly have taken place without information about it being preserved by the numerous chronicles, annals, inscriptions, and royal letters which have come down to us. These sources give us some information about the incursion of the Cimmerians from the northern Black Sea coast--apparently through the Klukhor, the Mamison, and the Darial Passes--in the second half of the 8th century B.C., and about the incursion of the Scythians through Daghestan at the beginning of the 7th century B.C. The Scythians could not have had any connection with the formation of the Old Armenian language, since theirs was a language of the Iranian group (41), and this, from our point of view, is also true of the Cimmerians (42). However, some scholars consider the Cimmerians to have been speakers of Thracian, and therefore we must examine whether or not their language can be considered an ancestor of Old Armenian (43).

To this we must answer in the negative. First of all the Cimmerians were too few in number (44). Their danger consisted only in their great mobility and in the fact that they were the first to introduce the tactics of massive warfare on horseback. Next, [111] according to the existent data, they settled not in the Armenian Highlands, which at that time was firmly held by the Urartian Empire, but to the west of it--in eastern Asia Minor and in Pontus, and possibly to the north of it-- in some regions of Georgia (45).

Thus we must exclude the period from 750 to roughly 634 B.C. as a time possible for the penetration of the speakers of Proto-Armenian into the Armenian Highlands .

We have already made reference to the well-known suggestion according to which the Proto-Armenians penetrated into this area between 635 and 590 B.C., in the chaotic period of the Scythian invasion and the collapse of the Assyrian and Urartian empires, before the firm establishment first of the Median kingdom, and then of the Persian Empire of the Achaemenids. But if that were the case, we must explain where the Proto-Armenians were before 635 B.C. The history of the territories of the Cilician Taurus and the more eastern regions is, as we have seen, quite well attested by the sources from roughly 745 to 635, and we can be sure that during this period only a settled agricultural population lived there. It would be strange if such a population should leave its place and go wandering (46). Therefore, if Proto-Armenian tribes were anywhere else at all other than in Armenia Minor, where we encounter them in the 6th century B.C., they would have to have hidden either in the more northerly mountainous regions (however, it is hard to see how the Proto-Armenians could have gotten there, since, as far as we can judge, these regions were then in the hands of the Georgian-speaking tribes), or in the central Anatolian steppes to the west of the Cilician Taurus. But here the powerful empire of Phrygia (Mushku or Mushki, as it is called in the Assyrian and Urartian sources--we have called it "Western Mushki") existed at least from the end of the 9th or the beginning of the 8th centuries to 676 B.C. (actually also later). It might be suggested that until the period of the Scythian invasion the Armenians were part of the Phrygians, and that they moved away in connection with this invasion, advancing to the east. But even this suggestion is unacceptable, since the linguistic data show that Phrygian and Old Armenian separated from a common language-base a very long time before. Old Armenian is a quite specific language of the Thraco-Phrygian group and not at all a dialect of the Phrygian of the 8th century B.C. (47) Therefore if the Proto-Armenians did participate in the general movement of the Thraco-Phrygian tribes at the end of the 2d millennium B.C., it was not as a part of the Phrygians, but as a separate nation or tribal group.

There remains the assumption that Proto-Armenian penetrated into the Armenian Highlands before the 9th century B.C. Thus the [112] period we are looking for is narrowed to three or four centuries: from the 12th to the 9th centuries B.C. Since the general movement of the Thraco-Phrygian tribes was, as we have seen, from the west to the east, we must think of the Proto-Armenians as a vanguard of this movement.

And in fact the Assyrian sources, as we have seen, relate of an incursion of still-mobile tribes from the west into the valleys of the Upper Euphrates and the Arsanias in the first half of the 12th century B.C., immediately after the collapse of the Hittite Empire. The Mushki, the Apeshlaian Kaska, and the Urumeans are mentioned. Of these at least the Mushki and the Urumeans evidently settled here (48), since they are mentioned in this territory as agricultural tribes as far back as the beginning of the 9th century B.C., and the existence of the "country" of Urumu, or Urmie, also roughly in this region, is evidenced by the Assyrian and Urartian inscriptions in the 9th to 8th centuries B.C.

Thus we should apparently seek the Proto-Armenians either in the Mushki or in the Urumeans who penetrated into the valleys of the Upper Euphrates and the Arsanias around 1165 B.C. We must remember that for the time being we are speaking about the Proto-Armenians as the speakers of an ancestor language of Old and Modern Armenian, and not about the problem of the emergence of the Armenian nation as such, which is much more complicated.

2.3. The Theory of the Haiasan Ancestry of the Armenian Nation

Before dealing with the problem of the Mushki and the Urumeans, we must consider the popular theory which sees the "cradle of the Armenian nation" in Haiasa (49).

We shall begin by clearly stating the question: are we talking about a physical, linguistic, or cultural succession, or of all these types of succession together?

We may speak of a physical succession between Haiasa and the Armenian nation only if we can prove that the Haiasans withdrew from their original dwelling places in Pontus, in the valley of the Choroh River and possibly on the upper reaches of the Euphrates, and scattered their settlements throughout the entire Armenian Highlands, completely or to a significant degree supplanting the former Hurro-Urartian population. This would prove that the entire Armenian nation or a large part of it descended physically from the Haiasans. Otherwise we can only speak of a physical succession from Haiasa with respect to those comparatively small groups of the Armenian nation which lived directly in the territory of former Haiasa.

[113] There are no data about a mass resettling of the Haiasans to the south, southeast, or east. The assertion made in Istorija armjanskogo naroda (50) that with the collapse of the Hittite Empire Haiasa consolidated, spread out to the western regions of the Armenian Highlands, and turned into a strong kingdom rivaling Urartu, is entirely unfounded. On the contrary, from our sources it apparently follows that the federation of Azzi-Haiasa fell apart as far back as the 13th century B.C., long before the collapse of the Hittite Empire, and from that time on it is not mentioned in any of the texts. Subsequently the Hurrian kingdom of Daiene emerged in the territory of what was earlier Haiasa, and later its northern part was occupied by Georgian-speaking tribes. It is true that a piece of former Haiasan land in the upper valley of the Euphrates could have been encompassed in the process of the genesis of the Armenian nation, but only at a time when no more traces of Haiasa as such existed. There are no data in the sources about any Haiasans beyond the confines of the former territory of Haiasa, nor about any direct links between Haiasa and the Armenians of later times. As far as cultural succession is concerned, the Armenians are undoubtedly the successors of the entire ancient population of the Highlands, i.e., of the Hurrians, the Urartians, and the Luwians. There is no indication that Haiasa had any more specifically important cultural influence on the later population of the Highlands than did, say, Isuwa, Alzi, Uruatri, or Qumme. Actually we know very little indeed about the culture of Haiasa, except for its marriage customs and the names of some deities, but no recollection of these has remained at all in the Armenian tradition (51).

There remains the linguistic succession. The assumption of a development from the hypothetical Haiasa language to Old Armenian has no base in any known linguistic fact whatever. It rests entirely on a certain similarity between the name of the country Haiasa (probably /xaiasa-/, with the Arm. sound x) and the self-appellation of the Armenian--hayk' (with the sound h) (52). From what has already been said above about the nature of ethnonyms in general it is evident that this similarity can in no way serve as proof of an organic connection between these terms. Moreover, as the Old Armenian words of analogous structure show, it is difficult to say how the initial form of the word hayk' sounded. The initial consonant might have been either *p-, as in hayr "father," from I.-E. *pe"ter, or the Proto-Indo-European laryngeal *H-, as in haw "grandfather," or *h-, which itself has a different derivation in the Indo-European languages, for example, from *s-. The diphthong -ai- might also be traced to different sound combinations, including -ate-, -ati-. The stem of the word hayk' is hayo- (and not, let [114] us say, *haya- (53). Kapantsjan interprets the suffix -sa in Haiasa as the ancient Luwian toponymical suffix -ssas (54), which was certainly widely in use throughout all of Asia Minor. But there is one area where this suffix is not found at all, and this area is Armenia. Therefore in the word "Haiasa" the element -sa, if it is a suffix, has no relation to Old Armenian.

As concerns the other data on the Haiasa language, they are represented by five personal names: Aissias, Annias, Marias, Muttis, and Hukkannas (as well as Karannis [or Lannis; the signs cannot be read clearly]) (55), and by five divine names: Tarumus, Terittutunus, Unakkastas, Utaktannas(?), and Paltaik(?), not counting a few damaged names or names in heterographic writing (56). We exclude the names of localities since there is no guarantee that they belong to the language of Haiasa and not to some other earlier language (57). All the cited names have been distorted by the Hittite transliteration (it is Hittite, in particular, to which the case endings in -s, -as, -is belong). In the original version of the Haiasa theory, as it was developed by Kapantsjan, these names were etymologized on the principle of external similarity (the "siren of consonance") from various languages, primarily from Hurrian, but in any event not from Indo-European, which eo ipso excludes the possibility of identifying the Haiasa language with Proto-Armenian. Therefore Kapantsjan had to resort to the thesis, earlier proposed by N. Y. Marr, on the "dual nature" of Armenian and to negate its Indo-European origin. However the latest research has completely confirmed its Indo-European affiliation, as first established in the 19th century.

Djahukyan (58) has tried to prove the Indo-European nature of the language of the Haiasans. In spite of his erudition in the field of Indo-European linguistics, his conclusions are not compelling. In a root-list of any linguistic family it is possible to choose roots which sound sufficiently similar to a dozen arbitrarily chosen names. Moreover, as the meanings of the names are unknown, there can be no guarantee that the roots, chosen by virtue of similar sound, fit these names in meaning as well (59). But even if we accept the conclusions of Djahukyan, the "Indo-European Haiasa language" belongs, on his own assertion, to another Indo-European group and has no relation whatsoever to Armenian (60). And how could it be that Armenian phonology (if derived from the language of Haiasa) should turn out to be so close to the phonology of the Phrygians (newcomers who arrived in Asia Minor many centuries later and who had no actual contact with Haiasa)? How could it be the Phrygian and Armenian sounded so strikingly similar to the Greek observers?

[115] Thus the connection of the Armenian nation and language with Haiasa is not provable and in its very essence not probable (*2).

Continued on Next Page


These links open in separate windows:


Hittite, Hurrian, Urartian Page
Classical Antiquity Menu
--   This is a mirror of one of Robert Bedrosian's web pages   --