A. Carrière

The Eight Sanctuaries
of Pagan Armenia

according to Agat'angeghos and Movses Xorenats'i

[Les huit sanctuaires de l'Arménie payenne]

Paris, 1899, English Translation by Robert Bedrosian, 2009

This material is in the public domain. It may be copied and distributed freely.

Carrière's Map of the Eight Sanctuaries (opens in a separate window).

Download in .zip format



If Movses Xorenats'i is to be believed, Artashes I was the grandson and second successor of Vagharshak, founder of the Armenian Arsacid dynasty. He was a warrior king who conquered Asia Minor, invaded Greece, and "filled the Ocean with the multitude of his ships." He died in a military coup, during which his innumerable soldiers killed each other (MX II. 11-13). Artashes left the throne to his son Tigran, who was the son-in-law and ally of Mithridates, and the adversary of Lucullus and Pompey, who extended his conquests to Mesopotamia and Syria.

During his campaigns Artashes had found in Asia Minor and Greece a certain number of statues of the gods which he took back to Armenia to install in the native sanctuaries as victory trophies. Tigran, who had realized his father's aspirations, also enriched these sanctuaries with the addition of a statue found in Mesopotamia.

Here are the passages from Movses Xorenats'i describing these events:


Finding in Asia images of Artemis, Heracles, and Apollo that were cast in bronze and gilded, he had them brought to our country to be set up in Armavir. The chief priests, who were of the Vahuni family, took those of Apollo and Artemis and set them up in Armavir; but the statue of Heracles, which had been made by Scyllas and Dipenes of Crete, [12] they supposed to be Vahagn their ancestor and so set it up in Tarawn in their own village of Ashtishat after the death of Artashes. MX II. 12, p. 148.

Movses Xorenats'i, Book II. 12.


He also took from Hellas images of Zeus, Artemis, Athena, Hephaistos, and Aphrodite, and had them brought to Armenia. But before they had arrived in our land the sad news of Artashes' death was heard. [Those bringing them] fled and brought the images to the fortress of Ani (1). The priests followed and stayed with them. MX II. 14, p. 149.

Movses Xorenats'i, Book II. 12.


As his first task he wished to construct the temples. But the priests, who had come from Greece, decided not to penetrate deep into Armenia. For an excuse they feigned omens to the effect that the gods wished to reside at that very spot. Tigran consented and raised the statue of Zeus Olympus in the fortress of Ani, that of Athena in T'il, the second statue of Artemis in Ere'z (2), and that of Hephaistos in Bagayr'inj. But the statue of Aphrodite, as the beloved of Heracles, he ordered to be set up beside the statue of the same Heracles in Ashtishat. And angered at the Vahuni in that they had taken it upon themselves to set up on their private lands the statue of Heracles sent by his own father, he dismissed them from the priesthood and confiscated to the crown the village in which the statues had been erected. MX II. 14, pp. 151-152.

Movses Xorenats'i, Book II. 14.


He himself went down to Mesopotamia, and finding there the statue of Barshamin, he embellished it with ivory, crystal, and silver. He ordered that it should be brought and set up on the town of T'ordan. MX II. 14, p. 152.

Movses Xorenats'i, Book II. 14.

From these passages cited above it follows that according to the chronology of Movses Xoranats'i, toward the beginning of the first century before our era, nine foreign idols were transported to the pagan temples of Armenia by order of the two conquering monarchs, Artashes and Tigran. Three of these statues came from Asia Minor, five from Greece, and one from Mesopotamia:

Found in Asia Minor

1. Artemis (I), whose statue was erected in Armavir.
2. Heracles, whose statue was erected in Ashtishat.
3. Apollo, whose statue was erected in Armavir.

Found in Greece

4. Dios, whose statue was erected in Ani.
5. Artemis (II), whose statue was erected in Erez.
6. Athena, whose statue was erected in T'il.
7. Hephestos, whose statue was erected in Bagayar'inj.
8. Aphrodite, whose statue was erected in Ashtishat.

Found in Mesopotamia

9. Barshame'n, whose statue was erected in T'ordan.

The city of Armavir which Artashes had designated as the recipient of the three statues taken from Asia Minor, but which only received the statues of [14] Artemis and Apollo, was still Armenia's capital. Artashes' grandfather Vagharshak already had placed there, along with statues of his ancestors, those of the Sun and the Moon (MX II. 8). When Eruand, Sanatruk's successor, abandoned Armavir and moved the court to the city of Ervandashat which he had built on the Arax not far from the Axurian River, he transported everything from the old capital to the new one, except for the idols. This suspicious monarch, fearful of having too many worshippers thronging in his vicinity, relocated the idols and statues to a town especially dedicated to the cult which he had built at a certain distance and named Bagaran (MX II. 39, 40). However Eruand's successor, Artashes II, in turn built a capital named after himself, Artashat. There he assembled everything which his predecessor had taken from Armavir and had used to adorn Eruandashat and Bagaran, also moving "the image/statue of Artemis and all of his patrimonial idols." Artashes also moved the statue of Apollo which was erected, however, "outside the city, close to the road" (MX II. 49). But the rest of the ancient idols of Armavir were not fated to join them there. When the first Sasanian king Artashir had the Armenian king Xosrov assassinated and conquered his kingdom, he destroyed the statues of the Sun and the Moon as well as those of Vagharshak's ancestors at Artashat (MX II. 77). Consequently only the statues of Artemis and Apollo were in the capital.

There is no evidence that any of the other statues had been moved. We need only replace the name of Armavir with that of Artashat to get a list of the eight sanctuaries. Two were at Artashat (those of Artemis and Apollo); one was at Ashtishat, with two idols side by side (Heracles and his lover, Aphrodite); and one in each of these locations: Ani, Erez, T'il, Bagayarindj, and T'ordan.

These sanctuaries were most irregularly spread throughout the Armenian territories (1). Artashat was situated in [15] central Armenia (valley of the Araxes), Ashtishat in southern Armenia (district of Taron); the five others were grouped into a narrow region of Upper Armenia formed by the three neighboring districts of Daranali, Ekegheats' (Akilisene), and Derjan (Derxene). The opposition of the Greek priests to the dispersion of their idols in the different provinces of Armenia only partly explains this peculiar state of affairs.


Toward the end of the 3rd century A.D. and the beginning of the 4th century--the date is uncertain--about four hundred years after the diffusion of Greek idols into the temples of Armenia, King Tiridates (Trdat) resided at Vagharshapat, not far from the modern city of Erevan. As a result of quite extraordinary events which are narrated in the marvelous book of Agat'angeghos, and thanks to the preaching of Gregory the Illuminator, Tiridates embraced the Christian faith. Even before being baptized he "spontaneously" commanded that the ancient national gods "who are not gods" should be removed from the soil of Armenia and that even their memory should be effaced.

So commenced a true war whose most salient points we will review here. Trdat was accompanied by his nobles and soldiers; they were needed to fight against armed demons who frequently defended access to the pagan temples. He also took along Gregory and, leaving his residence at Vagharshapat headed first to [16] Artashat to destroy the altars of the goddess Anahit. However, before entering that city he found "on the road" the sanctuary of the god Tir (or Tiur), which he destroyed before destroying that of Anahit (Agat'. p. 584 ff.).

Thereafter saint Gregory began an evangelizing tour of the cities, towns, and villages of Armenia, marking the places for future churches, planting the cross, and teaching Christian doctrine (Agat'. p. 587 ff.).

War against the idols resumed. Trdat and saint Gregory headed for western Armenia where they found first the temple of the god Barshimnia in the village of T'ordan in the district of Daranaghi. The temple was destroyed and the statue of the god smashed to pieces. Saint Gregory stopped in this district to convert the residents and to successfully pursue the demons (p. 588 ff.).

From there the king and the saint went to the fortress of Ani where the tombs of the Armenian kings were located, to destroy the altars of Aramazd, father of all the gods. In the neighboring district of Ekegheats', in the town of Erez, they destroyed the temple of Anahit. Then they crossed the Gayl (Lycus) River and demolished the temple of Nane, Aramazd's daughter, in the town of T'il. Finally with Gregory continuing to evangelize and Trdat to relate the miracles he had experienced, they arrived in the district of Derdjan and, in the town of Bagayar'inj, they razed the temple of the god Mher [Mithra], son of Aramazd (p. 590 ff.).

After destroying the sanctuaries of Upper Armenia, they took a break. The king and his court had converted to Christianity but still had not been "illuminated by baptism," for Gregory, not being a priest, could not confer it. Consequently, a great assembly was held at Vagharshapat which resolved to send him to Caesarea of Cappadocia to be ordained by the bishop there. Saint Gregory departed "in the royal chariot," escorted by many nobles and 10,000 soldiers. At Caesarea bishops Leontius and other bishops performed the laying on of hands and thus he received "authority in heaven and on earth and the keys to the kingdom [17] of heaven." In conformity with the rites he became priest [and bishop] and head of the Church of Armenia. Then he returned to his country, honored everywhere he passed and bringing with him precious relics given to him by saint Levontius (p. 594 ff.).

When he reached the Armenian border, saint Gregory learned that in the land of Taron a sanctuary had been spared and was still standing. This was the Vahevahean (1) temple, sacred to the god Vahagn, the eighth (2) of the famous sanctuaries, the "place of sacrifice of the kings of Greater Armenia"; it was full of riches and was located in the town of Ashtishat. Agat'angeghos provides us with more details about this celebrated temple than on any of the others. Despite this, his description is not that clear. We see that the Vahevahean contained three altars or temple, the first dedicated to the god Vahagn, the second to the Golden Mother, and the third to the goddess Astghik, Vahagn's lover, whom the Greeks called Aphrodite (3). As far as the second, the Golden Mother, this epithet refers to the goddess Anahit whom we have already encountered at Artashat and Erez. However, here the text does not provide her name.

The destruction of the Vahevahean was accompanied by extraordinary portents. The soldiers sent to destroy the temple, led astray by the demons, were unable to find the entrance. Their iron tools could not mar the walls. When saint Gregory observed this, he ascended a hill opposite the temple and called on the aid of the Almighty. From the cross he held in his [18] hand "a strong wind" blew which went and leveled the building such that no "trace" of it remained there (p. 606).

It was only after this that saint Gregory built the first church and began to baptize the Armenians. When King Trdat in his royal residence in Vagharshapat heard about the return of the bishop, he hastened to go before him and to receive baptism on the banks of the Euphrates River with his court and his entire army.

It is not our intention to consider the historical veracity of these remarkable passages which we have summarized here. We accept the facts as they are narrated to us and confirm that eight temples were destroyed, two as Artashat and one each in the towns of T'ordan, Ani, Ere'z, t'il, Bagayar'inj and Ashtishat, the latter having three idols, two of which are named.


It is perhaps unnecessary to observe that the place names cited here are identical with those mentioned previously in the passages from Movses Xorenats'i. Only the gods' names are different: where Agat'angeghos provides Armenian names, Movses gives Greek names (1). This difference, which is very apparent at first [19] completely disappears if we use the Greek (1) rather than the Armenian text of Agat'angeghos, which translates the names exactly as we find in Movses. The table below clearly shows the similarities and dissimilarities of the three manuscripts.

Carrier, p. 19

[20] The only difference is in the form of the names (1). Where the text of Movses Xorenats'i has Barshaminay, the regular genitive of Barshame'n, the text of Agat'angeghos has the rather unusual form Barshimnia. It seems certain that the latter form is a copyist's error for Barshamina as the corresponding place in the Greek manuscript has Barshamenes. Such an error in a proper name will not surprise those who deal with Armenian manuscripts (20).

It is also noteworthy that this same word occupies different places in the two lists. While the sanctuary at T'ordan appears right after that of Artashat in Agat'angeghos, in Movses Xorenats'i it is in last place. We shall explain this later. However, for now let us leave "Barshame'n" to one side and examine the order of the other names in the table.

Since Movses Xorenats'i himself took the precaution of informing us that the idols of Artashat had been transported there from their previous home in Armawir (see I. p. 14 above), and that the altar of Heracles which had been destined for Armawir was set up at Ashtishat (p. 12), we may conclude that both the temples and the idols appear in the same order in Agat'angeghos and Movses Xorenats'i.

A resemblance of this nature cannot be regarded as simple coincidence. [21] We must suppose some dependence between the two texts.

The following remarks support such a conclusion:

A. The name Barshame'n which is of Semitic origin, though certainly corrupted (perhaps for Belshamin) is not found anywhere else in Armenian literature.

B. The same expression "places of sacrifices (yashtits' teghik')" is used to give a folk etymology for the city of Ashtishat (1). However, the full expression "places of sacrifice of the kings of Greater Armenia" which is used in the text of Agat'angeghos (p. 606) regarding the Vahe'vahe'an temple is greatly abridged in the text of Movses Xorenats'i to "places of sacrifice" (p. 14) and employed as a proper noun in place of Ashtishat.

C. When Trdat went from Vagharshapat to Artashat to destroy the sanctuary of Anahit-Artemis, before entering the city and "on the road" he encountered the temple of the god Tir-Apollos (Agat'. p. 584). Here the matter is about the road the king was traveling on, that is "the road which went [from Vagharshapat] to the city of Ashtishat," which Agat'angeghos had previously referred to (p. 151). Now when Movses Xorenats'i is relating how king Artashes II transported the idols at Bagaran to his new capital at Artashat (MX II. 49), he adds "but the king erected the statue of Apollo outside the city, close to the road." this last remark is rather imprecise, since there was more than one road leading to Artashat.

D. According to Movses Xorenats'i (p. 12) the statue of Dios was erected "in the fortress of Ani (yamurn yAni)." The Armenian appears to be an abridgement of Agat'angeghos' (p. 590) more expansive expression "in the fortified place called Ani (yamur tegin yanuaneal yAni)." The word ordinarily used to designate the fortress of Ani is amrots' (MX II. 12, 38).

[22] Now we come to the dissimilarities found in the study of the text of Agat'angeghos and Movses Xorenats'i. Let us try to clarify the three points already mentioned:

1. The place assigned to the god Barshame'n.

2. The change in the destination for Heracles' statue.

3. The Greek names of the divinities replacing the Armenian.

1. Movses was unable to find a Greek equivalent for the Semitic name Barshame'n. Nonetheless he acknowledged the origin of this mythological entity, since the name with minor alteration appears in his history of the Armenian heroes. Barsham was said to be a tyrant of the race of giants who oppressed the country's southern reaches. Aram marched against him, attacked and defeated him and put him to flight from Korduats' to Asorestan. Barsham himself was killed by Aram's soldiers, but as he was renowned for his exploits, he was deified and worshipped for a long time by the Syrians (MX I. 14). Such euhemerisms are not rare in Movses.

Consequently it is toward the south that we should look for his statue, not in Asia Minor or in Greece, the theater of Artashes' exploits. The discovery of the idol of Barshame'n was made by Artashes' son Tigran when he conquered Mesopotamia and Syria. This is why Agat'angeghos' "glitteringly white god" (Barshamnia) in the aforementioned list is moved into last place [by Movses Xorenats'i] but with a statue made of white materials "ivory, crystal, and silver." We have noted elsewhere (1) his less than conscientious and arbitrary use of the sources.

2. In the text of Agat'angeghos we find no reference to the disobedience of the Vahunik' (2) or of their transporting to Ashtishat the statue of Heracles [23] which had been intended for Artashat. Agat'angeghos knows nothing about the Vahunis. Movses Xorenats'i, in his turn, appears to be ignorant of the Vahe'vahe'an temple, although he does give the name of the Vahunik' to the priestly family whose members served in the sanctuary of Ashtishat. This family derived from Vahagn whom Movses made a son of the Armenian king Tigran I (MX I. 31). They were entrusted with service in the temple and invested with the priesthood by Vagharsh (MX II. 8). We see that they lost these privileges and their goods at the beginning of the reign of Tigran the Great (II. 14). From then on, Movses no longer mentions them. Although the family name appears in a list of lords found in the Life of Saint Nerses (p. 34), this is the only reference to them and it is rather late. Beyond this, the ancient priestly family has left no other trace in all of Armenian literature. The Vahunik' are surely a product of Movses Xorenats'i's imagination.

It seems to us that the statue of Heracles prior to being finally erected in the place mentioned in Agat'angeghos's account had been reunited with the statues first sent from Greece. This provided Movses an opportunity to discourse about the Vahuni's disobedience and the end of their religious authority. It is the same with the third shipment of statues from Hellas accompanied by Greek priests whose protests prevented the dispersion of the statues throughout Armenian territory. Thus he explains why Agat'angeghos' text places five of the eight statues in a small corner of Upper Armenia.

Much of this derives from Movses' imagination, but a few other passages also reveal his characteristic method. When he says that the statue of Heracles is the work of Scyllis and Dipoenos of Crete, he had before him a source, probably Syriac, which may be found some day. Pliny speaks of a statue of Heracles among the works of Scyllis and Dipoenos (Pliny, Natural History, xxxvi, 4) and the chronography of Cedrenus (I, 564, ed Bonn) also mentions the two sculptors, perhaps using a source related to the one Movses used. The last part of the sentence is not unique to Movses' source, i.e., "they erected it iin Taro'n in their hereditary/own (sep'hakan) village of Ashtishat." Here is [24] the language that Ghazar P'arpets'i uses when describing how the Mamikoneans took saint Sahak's body for burial in Ashtishat: "They took it to the district of Taro'n to their own hereditary village named Ashtishat" (Ghazar, p. 104, ed. 1873) (1). It is immediately obvious that dependence [Xorenats'i's text has on Ghazar's]. The idea which has Tigran erecting Aphrodite's statue by the side of that of Heracles at the sanctuary in Ashtishat, thereby uniting the two lovers naturally arose from and was justified by the text of Agat'angeghos.

3. We have already spoken above about the names of the idols--with the exception of the god Barshame'n--being presented by Movses Xorenats'i in their Greek forms, forms corresponding exactly to those in the Greek edition of Agat'angeghos. Had Movses been independently translating them he could never have achieved such perfect correspondence. Moreover, excepting the passages more or less concerning our idols, Movses never provides Greek names for them, except when he is copying from other sources. Dios and Apollo do not appear elsewhere in his book. Artemis and Aphrodite appear in only one other passage (MX III. 33) where he is following Malalas (2). Hephestos appears twice in the same chapter (MX I. 7) following Eusebius' Chronicle/Chronicon (vol. I, p. 200); similarly Heracles appears twice in the same chapter (MX II. 8) also following the Chronicle (vol. I, p. 58 ff.). Athena constitutes the sole exception, appearing twice in a letter of the emperor Julian (MX III. 15).

If Movses Xorenats'i himself did not translate these names, there are only two conclusions to be drawn. Either he was familiar with the Greek text of Agat'angeghos directly or indirectly, which is not impossible, or else the Armenian text of Agat'angeghos which Xorenats'i had before him [25] contained the names in both their Armenian and Greek form. We cannot reject out of hand the existence of a such a parallel text. The text of Agat'angeghos which we are familiar with also contains such phrases as "Astghik, who is Aphrodite," and in a passage in P'awstos Buzand describing the destruction of the sanctuary of Ashtishat--which clearly derives from Agat'angeghos--we read "Heracles, that is to say, Vahagn" (ed. 1889, p. 37). These may be traces of an old recension of the text of Agat'angeghos which is now lost.

These diverse observations which we have presented thus far accentuate further the nature of the relationship between the two documents. Moreover they indicate that Movses Xorenats'i's text is not the older one. The precedence of Agat'angeghos is evident in other ways. Without making observations which pertain to literary history, Agat'angeghos's text further demonstrates precedence in the order of the names, which simply follows the itinery of saint Gregory in his three campaigns about the idols:

1. From Vagharshapat to Artashat;

2. From Vagharshapat to Upper Armenia;

3. From Vagharshapat to Caesarea and a return via Ashtishat.

Movses observes the same order, but encountering a reference to a second temple of Anahit-Artemis, one in Artashat and the other in Ere'z, he is forced to have two statues of Artemis brought to Armenia, one from Asia Minor and the other from Greece. This circumstance alone is sufficient to demonstrate the artificial nature of his account.

We may regard it as proven that Movses Xorenats'i, in creating his list of the idols introduced into Armenia, was following the account of the destruction of the temples provided by Agat'angeghos. [26]


In the period in which Movses Xorenats'i lived one could form only an imperfect idea about the ancient religion of Armenia. The tradition had vanished. The books which circulated in great number then spoke very little about ancient faiths as well as about foreign faiths. The exception were some polemical works which countered Sasanian Mazdaism. The old sanctuaries had been consecrated to the new cult, and the old festivals had been Christianized. What happened in Armenia had happened in all countries which had converted to Christianity with the exception of Greece and Rome, where a brilliant and indestructible literature maintained at least the memory of the old worship.

It is not impossible that the account in Agat'angeghos about the destruction of the sanctuaries was the only text available from which Movses Xorenats'i could derive information about Armenia's national pagan faith, just as today it remains the only basis for all research on the Armenian pantheon. In any case, Movses was familiar with it and used it. That document provided him with a sort of official list of the old sanctuaries and the idols worshipped in them. He made extensive use of it (1), but did not introduce any of its information anywhere in his History in the chronologically appropriate places. For example when speaking of Trdat, he does not mention the gods he worshipped or anything about the destruction of the sanctuaries in the relevant passages. This is a [27] noteworthy shortcoming of the section of Movses' book where he must speak of Armenia's conversion to Christianity.

Information drawn from Agat'angeghos' book was used, so to speak, regressively. Material was used to relate the origin and transplantation into Armenia of the idols destroyed by saint Gregory. Thus one could know the provenance of each of them. The idols were true trophies, marvelous booty which the Armenians had a right To be proud of, since they were simultaneously products and proofs of their kings' conquests in Asia Minor, Greece, and Mesopotamia. Perhaps Movses Xorenats'i also wanted thereby to assign a Greek origin to Armenian idolatry. Indeed he never says a word about the statues which were already standing in the sanctuaries and is more interested in how the idols at Ani were distributed to the neighboring towns. However this last issue is outside the scope of our present investigation.

If, as we believe we have established, the introduction of Greek idols into Armenia does not correspond to any actual event but is instead the product of Movses Xorenats'i's imagination, then the worth of many chapters of his book is gravely compromised. Thus everything concerning the relocation of the statues from Armawir to Bagaran and from Bagaran to Artashat becomes nothing other than a development on the initial fiction. Moreover it reinforces the doubts which have arisen on many other points about the reigns of Eruand and Artashes II (Book II, 37-60). [As examples,] Artashes II had designated his son Mazhan as chief-priest of Aramazd at Ani (ch. 53) and sent an envoy to the temple at Ere'z to seek health and long life (ch. 60).

Events from these two reigns are narrated to us on the authority of another priest of Ani, Olympius (Ughiup), who purportedly had written a History of the Temples (mehenakan patmut'iwnk') (MX II. 48) (1), [28] but who is otherwise unknown. The god worshipped at Ani was Aramazd, that is for Movses Xorenats'i Jupiter Olympian which probably explains the priest's name. As for that "book," Movses never says that he held it in his hands, nor does he relate anything specifically deriving from it. He provides better information elsewhere (MX II. 66) where we read that the Syrian Bardesan "entered the fortress of Ani and read the History of the Temples [naturally the one by Olympius] where he also found information about the deeds of the kings. Bardesan, adding to it contemporary events, put the whole thing into Syriac and his work was subsequently translated into Greek" (1). Movses was acquainted with the work of the Syrian heretic and gives and extract in this manner: "Bardesan says, according to the History of the Temples (2) that the last Tigran, king of Armenia, to honor the tomb of his brother the chief-priest Mazhan (II. 51, 53, 55) in the Town of altars (=Baguan) (3) in the district of Bagre'vand constructed an altar over the tomb where travelers could make sacrifices and find shelter for the night. In the same place subsequently Vagharsh instituted a solemn feast which was celebrated at the beginning of the New Year at the start of the month of Navasard."

[29] It is impossible not to see in this feast, supposedly instituted by Vagharsh the one saint Gregory displaced with another one after Trdat's baptism in honor of the martyrs whose relics he had brought from Caesarea. It is true that the passage in Agat'angeghos (p. 623) presents difficulties in interpretation (1). Some, such as Langlois find the god Amanor (nor-tari, "New Year") referenced here; others, such as Dulaurier, Gelzer, tec. find the god Vanatur ("shelter-giver"). Emin sees both Amanor and Vanatur, but equates these two deities. It must be conceded that the text in its present state, permits various interpretations. Movses Xorenats'i and the Greek translator of Agat'angeghos understood the passage in a different manner. Without a doubt they had before them a text somewhat different from ours. In any case, the citation from the History of the Temples by Bardesan derives from this passage in Agat'angeghos. To sum up, we see here a parallel to the introduction of the Greek statues into Armenia which highlights another instance of the way Movses Xorenats'i used the sources in a regressive manner.


Page opens in a separate window.

Armenian Mythology Page
Historical Sources Menu
History Workshop Menu