In 590 the fugitive Sasanian king Khusro II (590, 591–628) appealed to
Emperor Maurice (582–602) for military assistance against the usurper
Bahram Chobin, offering generous terms, including substantial territorial
concessions in Armenia. These were accepted by Maurice, and after
the defeat of Bahram in 591 the frontier shifted eastwards.7 The following
decade witnessed unprecedented cooperation between the two ‘great powers’
across Armenia.Maurice and Khusro II set out to strip their respective
Armenian sectors of soldiers for service in distant conflicts. Two rebellions
from the middle of this decade attest the resulting sense of bewilderment
among the Armenian elite.8 Only the uprisings in the 770s and the resistance
to the forces ofMichael IV (1034–41) in1041 outside Ani reveal a similar
desperation. The first of the two rebellions collapsed when threatened by
imperial and Persian forces acting in concert. The second ended in bloodshed.
An army under the general Heraclius and Hamazasp Mamikonean
defeated the rebels, killing the majority and capturing the remainder who
were taken back to Theodosioupolis and executed. The only rebel to escape
fled to Khusro II but was returned, tortured and killed.
The role of Hamazasp Mamikonean challenges the standard picture of
Armenian helplessness in the face of implacable imperial oppression. Here
is an Armenian noble serving imperial interests inside Armenia. The suspicion
must be that there were other Armenian princes prepared to work
with the new regime. When war with Persia broke out after Maurice’s
assassination in late 602, as Khusro II sought to recover those districts previously
ceded, several Armenian princes fought for Byzantium. In 605, the
Byzantine forces defending the district of Bagrevand against Khusro were
led by the local Armenian lord Theodore Khorkhoruni who entered into
negotiations with the Persians only after Byzantine forces had withdrawn.9
Significantly, it took at least five seasons of campaigning for the Persians to
expel the Byzantine forces from Armenia (603–7). Moreover, the fighting
was not restricted to those western districts which had been under imperial
control for generations but was concentrated further east, across the districts
recently acquired by Byzantium. Such a holding strategy would have
been inconceivable without local support.
The decade after 591 also witnessed pressure upon those districts now
under imperial control to conform to imperial orthodoxy. Although
CatholicosMoses II (574–604) refused to attend a council in Constantinople
convened to establish union between the churches and remained in
the Persian sector at Dvin, Maurice ordered the council of Chalcedon to
be preached in all the churches of the land of Armenia, threatening ‘to
unite them in communion through the army’.10 A second catholicos, John
of Bagaran, was established at Avan, provocatively situated just across the
border. John is usually titled ‘anti-catholicos’ and dismissed as little more
than the creature of Maurice with an ephemeral influence upon Armenia.
However, there is good evidence for a sizeable body of support for John,
at and below diocesan level. After the election of Abraham as catholicos
(perhaps in 606, probably in 607), five bishops and nineteen leaders of
religious communities, including those linked to the ‘holy cathedral’ and
the church of St Hrip‘sime in Vagharshapat, acknowledged their error and
returned to the anti-Chalcedonian party.11
Moreover, there were repeated attempts at ecclesiastical reconciliation.
In 604, the Byzantine commander in Armenia, Sormen, wrote to the temporary
head of the monophysite party, Vrt‘anes, noting that they had met
and corresponded on this subject many times. Sormen expressed a hope
that they could meet ‘like fellow brothers, joint heirs in baptism and sons
in the faith of our father St Gregory’, revealing thereby his own Armenian
ancestry.12 This spirit of compromise, which was not reciprocated, seems
to find an echo in the remarkable karshuni version of Agathangelos.13 This
transposes the key events in the original narrative of the conversion of
Armenia to different, contemporary locations. Thus of the seventy-seven
virgins who accompanied St Hrip‘sime, forty are assigned to Dvin and
thirty-seven to Avan, thereby establishing the equal sanctity of both sees.
Gregory the Illuminator baptises in the western district of Ekegheats‘;
he meets King Tiridates fifteen kilometres from Theodosioupolis; and
he dies in Daranaghi. This radical revision represents a rare witness to
the intellectual tradition of the pro-Chalcedonian party in Armenia after
591 and a very subtle development – or rather, subversion – of Armenian
tradition.
Even the Byzantines’ defeat at Persian hands in Basean, probably in 607,
and their subsequent loss of key fortresses, including Theodosioupolis, did
not mark the end of operations in Armenia. The following year, a Byzantine
counter-attack in the district of Theodosioupolis was repulsed, whilst in
610 the city’s inhabitants were transferred to Ecbatana in Persia, suggesting
an ongoing threat. In 613, another Byzantine army marched through these
districts. When Heraclius (610–41) launched a significant campaign in 624
against Theodosioupolis and then Dvin, he was advancing through districts
which had been incorporated into provincial and episcopal structures for
generations. Evidently he was looking to attract additional support. In
autumn 624, Heraclius appealed to the princes and leaders of the lands
of Albania, Iberia and Armenia by letter, urging them to come and serve
him together with their forces but threatening reprisals and subjugation if
they refused.14 It is impossible to gauge the response to his appeal but it
seems that many Armenian princes preferred to support Khusro II.15 Only
one late source refers explicitly to Armenians being attracted into imperial
service beforeHeraclius’ defeat of the Persian army at the battle of Nineveh
on 12 December 627.16
The years between 624 and 628 witnessed a complex series of military
manoeuvres and engagements in the Transcaucasus.17 Three primary
strategic considerations seem to have guided Heraclius. He courted potential
allies across the Transcaucasus and from the steppe world to the north.
The decisive impact of Turkic forces in 627 and 628 cannot be exaggerated.
Secondly, such a strategy drew Persian armies away from Constantinople
and into an environment in which logistical pressures dictated that possession
of the larger army was no guarantee of success. Thirdly, whether or not
instructed by his father, Heraclius had recognised the potential for striking
at the centre of the Sasanian kingdom from the north, using Armenia as a
bridgehead.18 Such considerations go a long way towards explaining why
Armenia continued to command such attention from successive emperors
throughout the seventh century and beyond.
When Byzantine forces were expelled in 607, the monophysite party
in the Armenian church was already in the ascendant and remained so
throughout the reign of Khusro II. The latter began to favour the expanding
monophysite confession across his dominions in preference to the Nestorian
church of the east. In the aftermath of Heraclius’ triumph and the
return of the True Cross to Jerusalem on 21 March 630, the fissures within
the Armenian church were reopened. The recently appointed catholicos
Ezra (630–41) was invited to attend a church council at Theodosioupolis,
probably in early 631, and under threat of the creation of a second catholicos
he accepted union. Statements that Ezra was ‘a humble and gentle man’
and that ‘no indecorous word ever passed from his mouth’ reflect a partisan
opinion.19 In reality his accommodation with Heraclius is likely to have
provoked considerable antagonism, an echo of which may be found in the
exile of John of Mayragom, an ardent monophysite whose own catholical
ambitions had been thwarted by Ezra’s election.20 An inscription commemorating
Ezra, partly in Greek and partly in Armenian cryptograms,
has been unearthed at Avan; evidently Ezra wished to associate himself
with the church founded there by John of Bagaran and the confessional
tradition espoused by him.21
Ezra’s choice of Avan was also dictated by political circumstance, since
Dvin still lay in the Persian sector. The deposition of Khusro II did not
give Heraclius possession of the whole of Armenia. In 628, Khusro II’s successor,
Kavad II, appointed Varaztirots‘ Bagratuni as governor (marzban)
of Armenia. Only under the terms of a subsequent treaty in the summer
of 630, between Heraclius and the latest claimant to the Sasanian
throne, Boran (630–31), were those districts ceded to Maurice returned to
Byzantine control. Even then, Persian influence over eastern and southern
Armenia persisted. In autumn 637, the leading Armenian prince,Mushegh
Mamikonean, responded to a Persian call-to-arms, raising 3,000 troops
whilst Gregory, lord of Siwnik‘, contributed 1,000.22 Both fell at the battle
of al-Qadisiyya on 6 January 638. With the benefit of hindsight, such
loyalty to the Sasanian cause might seem misguided, but the success of the
Arab conquest of Persia was still far from assured at that time.
The loyalty ofVaraztirots‘ Bagratuni andMusheghMamikonean to Sasanian
Persia may also explain the promotion of ‘new men’ to the office
of ‘prince of Armenia’ in the Byzantine sector of Armenia after 630, a
title used to denote the principal client. Mzhezh Gnuni and his successors,
David Saharuni and Theodore Rshtuni, all came from minor
noble houses. Although the narrative sources reveal little beyond this
sequence, epigraphic evidence supports the proposition that this decade
saw an intense Byzantine campaign to attract a broad spectrum of support.
Three inscriptions, recording the foundation of churches at Aghaman
(completed 636/7), Bagavan (August 639) and Mren (between 638 and
mid-640), all give a regnal year of Heraclius and accord him a laudatory
epithet.23 Contemporary regnal formulae and protocols used in imperial
documents and legislation repeat this combination. These inscriptions
therefore attest an otherwise lost body of correspondence between Byzantium
and Armenia.
The inscriptions at Aghaman and Mren also confirm that imperial honours
were distributed and were prized by their recipients. The founder of
the small church at Aghaman chose to define himself as Gregory elustr –
i.e. illustris, no more than a middle-ranking imperial title by this time. This
reveals a considerable down-reach on the part of the imperial authorities
into individual Armenian districts, for Gregory was not the lord of the district
in which he sponsored his church. The founder of the church at Mren,
David Saharuni, is titled patrikios, kouropalat¯es and sparapet of Armenia and
Syria. His remit encompassed all Armenia and must postdate the death of
Mushegh Mamikonean at al-Qadisiyya in 638. The extension of his command
beyond the boundaries of Armenia into Syria is unprecedented and
suggests that Heraclius was prepared to make remarkable concessions in
his efforts to forge an effective opposition to the Arab invasions after the
fall of Syria, one in which Armenian military resources had a leading role
to play.
The contention thatHeraclius invested heavily in a network of Armenian
clients is supported by the numismatic evidence. Seven different issues of
silver hexagrams from the reign of Heraclius and four issues of Constans II
(641–68) have been discovered in hoards or during excavations in Armenia,
the latest issue being struck between 654 and 659.24 This flow of Byzantine
silver into Armenia has traditionally been linked to the presence of Byzantine
forces; however, in light of the epigraphic evidence and the elite’s
prosperity, reflected in the numerous church foundations, one is tempted
to speculate whether this silver was minted for, and paid to, Armenian
clients. Armenia had been integrated into the Sasanian silver-based monetary
system for centuries and silver coins would have been familiar to
Armenians.
This strategy proved effective during the following decade. When an
Arab raiding party advanced from northern Syria through the Bitlis pass in
autumn 640 and sacked Dvin, Theodore Rshtuni ambushed the invaders
during their retreat, albeit without much success.25 A second Arab raid,
attacking from the south-east through Azerbaijan in summer 643, encountered
stiff resistance. One of its divisions, numbering about 3,000, was
heavily defeated by Theodore Rshtuni outside the fortress of Artsap‘k‘.
The major centre of Nakhchawan in the Araxes valley held out. These
operations showed the offensive and defensive potential of Armenia and
may have deterred further attacks.
Armenia was not insulated from the political turmoil engulfing Constantinople
after the death of Heraclius. The failed coup by Valentinus in
645 seems to have prompted widespread changes in the military hierarchy
across Armenia. The new commander, Thomas, was anxious not to damage
the agreement established with Khorokhazat, leader of continuing Persian
resistance against the Arabs in Atrpatakan (Azerbaijan). Thomas visited
him and promised that Theodore Rshtuni would be taken to Constantinople.
26 This episode illustrates how the interests of two clients did not necessarily
coincide. Khorokhazat faced growing recalcitrance from Albania
and was looking for assistance in deterring Armenian support for dissident
elements. In choosing to back Khorokhazat, Byzantium precipitated a crisis
in Armenia.
Theodore Rshtuni was soon restored to his command but the relationship
was clearly strained. In 652 the governor of Syria (and later caliph)
Mu‘awiya (661–80) induced him to switch sides, promising inter alia that
Armenian forces would not be employed in Syria and that Arab forces would
not be stationed in Armenia unless invited to repel a Byzantine attack.27
In response, Constans II travelled to Armenia to shore up his support and
undermine his erstwhile client. He advanced to Theodosioupolis and there
received the submission of a disparate group of Armenian princes and their
armed forces. Evidently they believed that it was in their long-term interests
to return to imperial service. Constans II moved on to Dvin and stayed
with Catholicos Nerses III (641–61). He attended a service with his host
in the cathedral church of St Gregory, during which the liturgy was celebrated
in Greek and the council of Chalcedon was proclaimed. Only one
anonymous bishop refused to participate but this tells us little about the
ongoing confessional tensions within the Armenian church; presumably
anti-Chalcedonians did not attend.
Constans II did not remain in Armenia long, being forced to return and
defend Constantinople in 654. Thereafter Byzantine fortunes fluctuated,
imperial forces being driven out of Armenia twice, but by the first half
of 656, Hamazasp Mamikonean was securely installed as kouropalat¯es and
prince of Armenia.28 At the same time, honours were distributed to the
other princes and treasures to the soldiers, confirming that the benefits of
imperial service were not confined to a few but were spread broadly among
the elite. Nerses III returned from exile in Tao after ‘the lord of Rshtunik‘
had died and the Arab invasion had come to an end’, indicating an earlier
date, perhaps 656, than is generally admitted.
Constans II was determined to exploit the unexpected breathing space
afforded by the outbreak of civil war or fitna across the caliphate. He
sought to establish a broad network of clients across the Transcaucasus.
Juansher, prince of Albania, and the princes of Siwnik‘ quickly submitted.29
In autumn 659, the emperor undertook a second progress eastwards lasting
several months.30 He ventured into Media, meeting and rewarding loyal
True Cross. Constans was also seeking to attract others, including Persians
who wished to fight on against the Arabs. He was still in Armenia in
spring 660, at Vagharshapat, where he rewarded Juansher a second time.
A later text suggests that the emperor was present at the inauguration of
the impressive church of Zvart‘nots‘.31 Whilst this cannot be proved, his
involvement would have done much to bolster the standing of its founder
Nerses III and the pro-Chalcedonian party across Armenia. Intriguingly,
the terse inscription commemorating Nerses’ role is in Greek rather than
Armenian.32
In the event, Constans II’s vision of a chain of clients did not survive
beyond the conclusion of the fitna. As the lynchpin of the network,
Hamazasp was swiftly removed and replaced by his brother Gregory
Mamikonean, previously a hostage of Mu‘awiya. Juansher transferred his
allegiance to the ‘king of the south [Mu‘awiya]’, when ‘the emperor of the
Romans [Constans] took the dregs of his forces and hastened across sea
and land to cross to the . . . distant islands of the west’.33 It seems very
likely that the principal Byzantine clients had been displaced or turned by
late 661 or early 662.
After 661, the limitations of the primary sources make it much harder to
trace the interaction between Byzantium and Armenia. The conventional
approach has been to treat this dearth of information as evidence for the
exclusion of Byzantine influence. Armenian colophons and inscriptions
together with isolated textual references collectively support an alternative
view, of persistent, wide-ranging Byzantine engagement until 730 but a
more limited focus thereafter, concentrated on and operated through those
districts bordering imperial territory.
The second sustained period of civil war across the caliphate after 680
afforded a fresh opportunity for Byzantine intervention. According to
Lewond’s History, Armenia repudiated Arab sovereignty by refusing to pay
tribute, probably in 682, but it is impossible to prove Byzantine influence
lying behind this decision.34 A later Armenian source records how
an Iberian prince, Nerses, massacred the Arab forces in Armenia during
the time of Catholicos Israel I (667–77).35 The Arab blockade of Constantinople
between 674 and 678 supplies an appropriate historical context
for just such a diversionary campaign but a Byzantine connection remains
conjectural (see also pp. 233, 372).
Constantine IV (668–85) was eager to exploit contemporary disorder
across the caliphate. In 685, he invaded Cilicia and threatened northern
Syria, compelling the new caliph, ‘Abd al-Malik (685–705) to sue for peace
on very generous terms on 7 July 685.36 This campaign may have been
coordinated with the devastating Khazar raid into Armenia during which
Gregory Mamikonean and Nerses were killed in battle on 18 August 685.37
According to Theophanes the Confessor, Justinian II (685–95, 705–711) ratified
the truce with ‘Abd al-Malik soon after his accession although its term
was extended to ten years and an additional provision was inserted, requiring
the parties to share the tax revenue of Cyprus, Armenia and Iberia.38
A subsequent passage under the same year entry adds that Justinian II
despatched a strat¯egos, Leontius, into Armenia. He subjugated Armenia,
together with Iberia, Albania, Boukania (probably Vaspurakan) andMedia,
imposed taxes on those countries and remitted a large sum to Justinian.
The changes to the treaty make sense when viewed in the aftermath of
this raid. The revenue arrangements may reflect a more fundamental partition,
of sovereignty.GregoryMamikonean’s successor as prince of Armenia
was Ashot Bagratuni, titled patrikios. Since he also brought an icon of the
incarnation of Christ ‘from the west’ for his church at Daroynk‘, forty
kilometres south of Mount Ararat, it seems likely that he was a Byzantine
client.39
After Ashot’s death – confronting Arab raiders in the Araxes valley
in 689 – a number of Armenian princes switched allegiance. This
prompted Justinian II to travel to Armenia in person, as his grandfather
Constans II had done in similar circumstances. Justinian summoned the
princes to him, taking some of their sons hostage, while rewarding others: he
raisedNerses Kamsarakan, the lord of Shirak, to the rank of prince of Armenia
and the patrikios and exarch Varaz(tr)dat was made prince of Albania.40
He then returned to Constantinople, taking with him Catholicos Sahak III
(677–703) and five bishops. Theophanes likewise reports Justinian’s visit to
Armenia although he places it too early, in his second year, and wrongly
associates it with theMardaites.41 A remarkable, pro-Chalcedonian account
of Armenian ecclesiastical history, which survives only in Greek, records
that Sahak and his bishops accepted Chalcedon at a council convened in
Constantinople in the fifth year of Justinian II, although on their return to
Armenia and under pressure, they reneged.42
This revival in Byzantine fortunes occurred in the context of the second
fitna. Even before his final victory over his main rival in 691, Caliph ‘Abd al-
Malik was turning his attention to Byzantium. Contrary to the traditional
view, it seems very probable that it was ‘Abd al-Malik, not Justinian II, who
broke the ten-year truce.43 The heavy Byzantine defeat in 692 at Sebastopolis
occurred deep inside newly secured Byzantine territory, indicating an
Arab offensive (see below, p. 384). Several Armenian clients promptly transferred
allegiance but the Byzantine position did not collapse overnight. A
colophon confirms that the principal Byzantine client in 689, Nerses Kamsarakan,
was still alive in 696 and in contact with Constantinople.44 The
region of Fourth Armenia also resisted. Although Muhammad bin Marwan,
the governor of al-Jazira, campaigned there in 694/5, evidently it had
not been subjugated in 701/2 when Baanes ‘Heptadaimon’ switched sides.45
Perhaps most surprisingly, in 702 Smbat Bagratuni rebelled and defeated an
Arab force at Vardanakert, being rewarded with the title kouropalat¯es.46 A
parallel account of this uprising, but with a Kamsarakan spin, affords useful
corroboration.47
The aftermath of this rebellion remains confused. Lewond maintains
that Smbat withdrew into Tao and that Catholicos Sahak III negotiated
a three-year peace. According to the History of the Albanians, however,
military operations continued.48 Dvin fell to a joint Byzantine-Armenian
force whilst the Arabs captured a fortress in Sevan only after a three-year
blockade. Both sources agree that a Byzantine force then suffered a heavy
defeat. Lewond adds that this occurred in Vanand in the first year of Caliph
al-Walid I (705–15). The Byzantine troops fled and the Armenian rebels
suffered severe reprisals, with 800 men in Nakhchawan and 400 in Khram
being imprisoned in churches and then burnt alive. Ominously, the lord
of Shirak, Nerses Kamsarakan, was summoned to Syria in 705; his fate
is not recorded. Smbat kouropalat¯es escaped into Byzantine territory and
was settled in the city of Phasis in Lazica. This sequence of events – a
rebellion by Armenian princes, contact with Emperor Tiberius II Apsimar
(698–705), the despatch of Byzantine forces, a successful counter-offensive
by Muhammad bin Marwan followed by the burning alive of Armenian
princes – is corroborated by Theophanes.49 The only significant difference
is chronological. Theophanes records this sequence of events under one
year, AM 6195 (702/3) but it seems more likely that they were spread across
several years (702–5).
Aside from the failed attempt at union in the time of Justinian II outlined
above, relations between the churches after 661 are almost entirely obscure.
In 719, however, Catholicos John III (717–27) stated unequivocally that the
six catholicoi after Komitas (between 628 and 705) were all Chalcedonian,
exempting only his immediate predecessor Elias (703–17) from criticism.50
As outlined previously, Ezra, Nerses III and Sahak III all engaged in discussions
with the imperial church but none of their correspondence or
other writings survives. Indeed the only extant letter between 628 and 705
is a draft Armenian ‘Defence’ of the monophysite position, prepared in
649 for despatch to Constans II.51 Arguably, no records or letters associated
with these catholicoi survive precisely because of their confessional perspective.
An exchange between Patriarch Germanos I (715–30) and Catholicos
John III from the 720s does survive, defining and defending their respective
positions in great detail.52 Conceivably this correspondence marks the
final breach between the churches and was preserved because it articulated
the differences. Confessional tensions at the highest level need not have
deterred other contacts. Colophons reveal that four patristic works were
translated into Armenian in Constantinople between 713 and 717 by David
hypatos and Stephen of Siwnik‘.53
After 730, Byzantine influence persisted but on a more limited scale. An
inscription on a tombstone located in a crypt atNakhchawan in Shirak commemorates
‘the blessed lord Artawazd Kamsarakan apo hypat¯on patrikios and
prince of Armenia, son of Hrahat patrikios lord of Shirak and Asharunik‘’.54
Artawazd was the grandson of Nerses Kamsarakan mentioned previously.
Evidently Byzantine titles continued to be awarded during the eighth century
to Armenian princes. Artawazd does not feature in any other source,
which is surprising given his rank of ‘prince of Armenia’. His omission is
hard to explain unless one views him as a second, rival prince of Armenia
and client of Byzantium.
When the third fitna erupted, two groups of Armenian princes may once
again be discerned. One party, under Ashot Bagratuni, remained loyal to
CaliphMarwan II (744–50); the other underGregoryMamikonean, looked
to Constantine V (741–75). Having taken refuge in Tao, ‘they relied upon
the forces of the king of the Greeks, who were in the regions of Pontos,
because there was a treaty of peace between them, at the command of the
emperor Constantine’.55 After blinding Ashot Bagratuni, perhaps in 748,
Gregory went to Theodosioupolis and broadcast news of his victory. Evidently
Theodosioupolis was under his, or Constantine’s, control and he was
attempting to attract further support. His success or otherwise in this initiative
is not recorded by Lewond, who simply notes that he died in agony
at an unspecified date and was replaced for a short time by his brother.56
Whether Lewond’s hostility stems from a political (anti-Mamikonean) or
confessional (anti-Chalcedonian) perspective is unclear. Again this temporary
Byzantine revival in Armenia was halted by the resolution of the strife
within the caliphate. In 754, Constantine V transferred the population of
Theodosioupolis to Thrace. Lewond adds that many from the surrounding
districts also left and ‘placed themselves on the side of the pious king’, a rare
favourable view of Constantine V.57 This transfer may represent a tactical
withdrawal at the end of a series of initiatives in Armenia rather than the
original goal.
Armenian princes did not risk rebellion against the dominant, controlling
power without support, or expressions of support, from a rival power
other than in exceptional circumstances. At first sight, the complicated
series of rebellions across Armenia in the 770s fall into that category. At
no stage do the narrative sources indicate any Byzantine involvement.58
Two of the rebel leaders, Artawazd and Mushegh Mamikonean, are said
to have begun their uprisings by killing local Arab tax-collectors. New
administrative arrangements and fiscal burdens at district level may have
precipitated their actions. On the other hand, Artawazd moved into Iberia
and later reappears as strat¯egos t¯on Anatolik¯on whilst Mushegh’s rebellion
apparently took the form of a prolonged, and ultimately unsuccessful, siege
of Theodosioupolis. This strategy is hard to fathom unless one accepts that
Byzantine support was anticipated. No Byzantine campaign is recorded
but it may have been planned; in 777 a large Byzantine army, under Armenian
commanders, attacked Germanikeia and devastated the surrounding
region.59
For the following five decades, there is very little evidence for Byzantine
involvement in Armenia. In 788 as many as 12,000 people under the leadership
of Shapuh Amatuni, his son and other Armenian nobles were granted
refuge within the empire by ‘the emperor Constantine’. Lewond portrays
this as a reaction to hardships inflicted by the caliph and his representatives,
specifically the seizure of land.60 It is in the last quarter of the eighth
century that several quasi-independent Arabic emirates emerged, ruling
districts previously under Armenian control.61 At the same time, members
of the Bagratuni princely house exploited their status as preferred Abbasid
clients to secure a dominant position. After 775, Byzantine attention was
concentrated on potential clients in those districts of Iberia which abutted
imperial territory. Ashot Bagratuni, established in neighbouring Klarjet‘i,
was appointed kouropalat¯es before 826.62 Byzantine strategy towards Armenia
came to operate on and through the remote district of Sper which
bordered the theme of Chaldia. The first ninth-century Armenian prince
known to have been accorded an imperial title was another AshotBagratuni,
prince of Sper; he was appointed patrikios and apo hypat¯on by Theophilos
(829–42).63 Intriguingly, his appointment is recorded in the context
of Byzantine operations against Theodosioupolis, Basean and Vanand, all
to the south and east of Sper. Although these operations have been compressed
into a single campaign and linked to a major Byzantine offensive
against Sozopetra, Melitene and Fourth Armenia undertaken in 837, they
could equally comprise separate campaigns spread over a number of years.64
This targeting of Theodosioupolis and its surrounding districts mirrors
the pattern of Byzantine offensives outlined previously, whilst the Khurramite
rebellion under Babek afforded a suitable opportunity (see below,
p. 390).
Caliph al-Mu‘tasim (833–42) responded swiftly to this Byzantine threat.
In 838, his forces inflicted a heavy defeat upon Theophilos at Dazimon
and captured Amorion. Genesios reports that Armenian forces under the
‘Vasparakanites’ (presumably the leading Artsruni prince) and the prince of
princes (probably Bagarat Bagratuni, prince of Taron) participated in these
campaigns.65 This represents a rare instance of active service by Armenian
forces against Byzantium. It illustrates how closely the leading Armenian
princes now identified with caliphal interests and the degree to which
Byzantine influence over them had waned.
Emperor Maurice (582–602) for military assistance against the usurper
Bahram Chobin, offering generous terms, including substantial territorial
concessions in Armenia. These were accepted by Maurice, and after
the defeat of Bahram in 591 the frontier shifted eastwards.7 The following
decade witnessed unprecedented cooperation between the two ‘great powers’
across Armenia.Maurice and Khusro II set out to strip their respective
Armenian sectors of soldiers for service in distant conflicts. Two rebellions
from the middle of this decade attest the resulting sense of bewilderment
among the Armenian elite.8 Only the uprisings in the 770s and the resistance
to the forces ofMichael IV (1034–41) in1041 outside Ani reveal a similar
desperation. The first of the two rebellions collapsed when threatened by
imperial and Persian forces acting in concert. The second ended in bloodshed.
An army under the general Heraclius and Hamazasp Mamikonean
defeated the rebels, killing the majority and capturing the remainder who
were taken back to Theodosioupolis and executed. The only rebel to escape
fled to Khusro II but was returned, tortured and killed.
The role of Hamazasp Mamikonean challenges the standard picture of
Armenian helplessness in the face of implacable imperial oppression. Here
is an Armenian noble serving imperial interests inside Armenia. The suspicion
must be that there were other Armenian princes prepared to work
with the new regime. When war with Persia broke out after Maurice’s
assassination in late 602, as Khusro II sought to recover those districts previously
ceded, several Armenian princes fought for Byzantium. In 605, the
Byzantine forces defending the district of Bagrevand against Khusro were
led by the local Armenian lord Theodore Khorkhoruni who entered into
negotiations with the Persians only after Byzantine forces had withdrawn.9
Significantly, it took at least five seasons of campaigning for the Persians to
expel the Byzantine forces from Armenia (603–7). Moreover, the fighting
was not restricted to those western districts which had been under imperial
control for generations but was concentrated further east, across the districts
recently acquired by Byzantium. Such a holding strategy would have
been inconceivable without local support.
The decade after 591 also witnessed pressure upon those districts now
under imperial control to conform to imperial orthodoxy. Although
CatholicosMoses II (574–604) refused to attend a council in Constantinople
convened to establish union between the churches and remained in
the Persian sector at Dvin, Maurice ordered the council of Chalcedon to
be preached in all the churches of the land of Armenia, threatening ‘to
unite them in communion through the army’.10 A second catholicos, John
of Bagaran, was established at Avan, provocatively situated just across the
border. John is usually titled ‘anti-catholicos’ and dismissed as little more
than the creature of Maurice with an ephemeral influence upon Armenia.
However, there is good evidence for a sizeable body of support for John,
at and below diocesan level. After the election of Abraham as catholicos
(perhaps in 606, probably in 607), five bishops and nineteen leaders of
religious communities, including those linked to the ‘holy cathedral’ and
the church of St Hrip‘sime in Vagharshapat, acknowledged their error and
returned to the anti-Chalcedonian party.11
Moreover, there were repeated attempts at ecclesiastical reconciliation.
In 604, the Byzantine commander in Armenia, Sormen, wrote to the temporary
head of the monophysite party, Vrt‘anes, noting that they had met
and corresponded on this subject many times. Sormen expressed a hope
that they could meet ‘like fellow brothers, joint heirs in baptism and sons
in the faith of our father St Gregory’, revealing thereby his own Armenian
ancestry.12 This spirit of compromise, which was not reciprocated, seems
to find an echo in the remarkable karshuni version of Agathangelos.13 This
transposes the key events in the original narrative of the conversion of
Armenia to different, contemporary locations. Thus of the seventy-seven
virgins who accompanied St Hrip‘sime, forty are assigned to Dvin and
thirty-seven to Avan, thereby establishing the equal sanctity of both sees.
Gregory the Illuminator baptises in the western district of Ekegheats‘;
he meets King Tiridates fifteen kilometres from Theodosioupolis; and
he dies in Daranaghi. This radical revision represents a rare witness to
the intellectual tradition of the pro-Chalcedonian party in Armenia after
591 and a very subtle development – or rather, subversion – of Armenian
tradition.
Even the Byzantines’ defeat at Persian hands in Basean, probably in 607,
and their subsequent loss of key fortresses, including Theodosioupolis, did
not mark the end of operations in Armenia. The following year, a Byzantine
counter-attack in the district of Theodosioupolis was repulsed, whilst in
610 the city’s inhabitants were transferred to Ecbatana in Persia, suggesting
an ongoing threat. In 613, another Byzantine army marched through these
districts. When Heraclius (610–41) launched a significant campaign in 624
against Theodosioupolis and then Dvin, he was advancing through districts
which had been incorporated into provincial and episcopal structures for
generations. Evidently he was looking to attract additional support. In
autumn 624, Heraclius appealed to the princes and leaders of the lands
of Albania, Iberia and Armenia by letter, urging them to come and serve
him together with their forces but threatening reprisals and subjugation if
they refused.14 It is impossible to gauge the response to his appeal but it
seems that many Armenian princes preferred to support Khusro II.15 Only
one late source refers explicitly to Armenians being attracted into imperial
service beforeHeraclius’ defeat of the Persian army at the battle of Nineveh
on 12 December 627.16
The years between 624 and 628 witnessed a complex series of military
manoeuvres and engagements in the Transcaucasus.17 Three primary
strategic considerations seem to have guided Heraclius. He courted potential
allies across the Transcaucasus and from the steppe world to the north.
The decisive impact of Turkic forces in 627 and 628 cannot be exaggerated.
Secondly, such a strategy drew Persian armies away from Constantinople
and into an environment in which logistical pressures dictated that possession
of the larger army was no guarantee of success. Thirdly, whether or not
instructed by his father, Heraclius had recognised the potential for striking
at the centre of the Sasanian kingdom from the north, using Armenia as a
bridgehead.18 Such considerations go a long way towards explaining why
Armenia continued to command such attention from successive emperors
throughout the seventh century and beyond.
When Byzantine forces were expelled in 607, the monophysite party
in the Armenian church was already in the ascendant and remained so
throughout the reign of Khusro II. The latter began to favour the expanding
monophysite confession across his dominions in preference to the Nestorian
church of the east. In the aftermath of Heraclius’ triumph and the
return of the True Cross to Jerusalem on 21 March 630, the fissures within
the Armenian church were reopened. The recently appointed catholicos
Ezra (630–41) was invited to attend a church council at Theodosioupolis,
probably in early 631, and under threat of the creation of a second catholicos
he accepted union. Statements that Ezra was ‘a humble and gentle man’
and that ‘no indecorous word ever passed from his mouth’ reflect a partisan
opinion.19 In reality his accommodation with Heraclius is likely to have
provoked considerable antagonism, an echo of which may be found in the
exile of John of Mayragom, an ardent monophysite whose own catholical
ambitions had been thwarted by Ezra’s election.20 An inscription commemorating
Ezra, partly in Greek and partly in Armenian cryptograms,
has been unearthed at Avan; evidently Ezra wished to associate himself
with the church founded there by John of Bagaran and the confessional
tradition espoused by him.21
Ezra’s choice of Avan was also dictated by political circumstance, since
Dvin still lay in the Persian sector. The deposition of Khusro II did not
give Heraclius possession of the whole of Armenia. In 628, Khusro II’s successor,
Kavad II, appointed Varaztirots‘ Bagratuni as governor (marzban)
of Armenia. Only under the terms of a subsequent treaty in the summer
of 630, between Heraclius and the latest claimant to the Sasanian
throne, Boran (630–31), were those districts ceded to Maurice returned to
Byzantine control. Even then, Persian influence over eastern and southern
Armenia persisted. In autumn 637, the leading Armenian prince,Mushegh
Mamikonean, responded to a Persian call-to-arms, raising 3,000 troops
whilst Gregory, lord of Siwnik‘, contributed 1,000.22 Both fell at the battle
of al-Qadisiyya on 6 January 638. With the benefit of hindsight, such
loyalty to the Sasanian cause might seem misguided, but the success of the
Arab conquest of Persia was still far from assured at that time.
The loyalty ofVaraztirots‘ Bagratuni andMusheghMamikonean to Sasanian
Persia may also explain the promotion of ‘new men’ to the office
of ‘prince of Armenia’ in the Byzantine sector of Armenia after 630, a
title used to denote the principal client. Mzhezh Gnuni and his successors,
David Saharuni and Theodore Rshtuni, all came from minor
noble houses. Although the narrative sources reveal little beyond this
sequence, epigraphic evidence supports the proposition that this decade
saw an intense Byzantine campaign to attract a broad spectrum of support.
Three inscriptions, recording the foundation of churches at Aghaman
(completed 636/7), Bagavan (August 639) and Mren (between 638 and
mid-640), all give a regnal year of Heraclius and accord him a laudatory
epithet.23 Contemporary regnal formulae and protocols used in imperial
documents and legislation repeat this combination. These inscriptions
therefore attest an otherwise lost body of correspondence between Byzantium
and Armenia.
The inscriptions at Aghaman and Mren also confirm that imperial honours
were distributed and were prized by their recipients. The founder of
the small church at Aghaman chose to define himself as Gregory elustr –
i.e. illustris, no more than a middle-ranking imperial title by this time. This
reveals a considerable down-reach on the part of the imperial authorities
into individual Armenian districts, for Gregory was not the lord of the district
in which he sponsored his church. The founder of the church at Mren,
David Saharuni, is titled patrikios, kouropalat¯es and sparapet of Armenia and
Syria. His remit encompassed all Armenia and must postdate the death of
Mushegh Mamikonean at al-Qadisiyya in 638. The extension of his command
beyond the boundaries of Armenia into Syria is unprecedented and
suggests that Heraclius was prepared to make remarkable concessions in
his efforts to forge an effective opposition to the Arab invasions after the
fall of Syria, one in which Armenian military resources had a leading role
to play.
The contention thatHeraclius invested heavily in a network of Armenian
clients is supported by the numismatic evidence. Seven different issues of
silver hexagrams from the reign of Heraclius and four issues of Constans II
(641–68) have been discovered in hoards or during excavations in Armenia,
the latest issue being struck between 654 and 659.24 This flow of Byzantine
silver into Armenia has traditionally been linked to the presence of Byzantine
forces; however, in light of the epigraphic evidence and the elite’s
prosperity, reflected in the numerous church foundations, one is tempted
to speculate whether this silver was minted for, and paid to, Armenian
clients. Armenia had been integrated into the Sasanian silver-based monetary
system for centuries and silver coins would have been familiar to
Armenians.
This strategy proved effective during the following decade. When an
Arab raiding party advanced from northern Syria through the Bitlis pass in
autumn 640 and sacked Dvin, Theodore Rshtuni ambushed the invaders
during their retreat, albeit without much success.25 A second Arab raid,
attacking from the south-east through Azerbaijan in summer 643, encountered
stiff resistance. One of its divisions, numbering about 3,000, was
heavily defeated by Theodore Rshtuni outside the fortress of Artsap‘k‘.
The major centre of Nakhchawan in the Araxes valley held out. These
operations showed the offensive and defensive potential of Armenia and
may have deterred further attacks.
Armenia was not insulated from the political turmoil engulfing Constantinople
after the death of Heraclius. The failed coup by Valentinus in
645 seems to have prompted widespread changes in the military hierarchy
across Armenia. The new commander, Thomas, was anxious not to damage
the agreement established with Khorokhazat, leader of continuing Persian
resistance against the Arabs in Atrpatakan (Azerbaijan). Thomas visited
him and promised that Theodore Rshtuni would be taken to Constantinople.
26 This episode illustrates how the interests of two clients did not necessarily
coincide. Khorokhazat faced growing recalcitrance from Albania
and was looking for assistance in deterring Armenian support for dissident
elements. In choosing to back Khorokhazat, Byzantium precipitated a crisis
in Armenia.
Theodore Rshtuni was soon restored to his command but the relationship
was clearly strained. In 652 the governor of Syria (and later caliph)
Mu‘awiya (661–80) induced him to switch sides, promising inter alia that
Armenian forces would not be employed in Syria and that Arab forces would
not be stationed in Armenia unless invited to repel a Byzantine attack.27
In response, Constans II travelled to Armenia to shore up his support and
undermine his erstwhile client. He advanced to Theodosioupolis and there
received the submission of a disparate group of Armenian princes and their
armed forces. Evidently they believed that it was in their long-term interests
to return to imperial service. Constans II moved on to Dvin and stayed
with Catholicos Nerses III (641–61). He attended a service with his host
in the cathedral church of St Gregory, during which the liturgy was celebrated
in Greek and the council of Chalcedon was proclaimed. Only one
anonymous bishop refused to participate but this tells us little about the
ongoing confessional tensions within the Armenian church; presumably
anti-Chalcedonians did not attend.
Constans II did not remain in Armenia long, being forced to return and
defend Constantinople in 654. Thereafter Byzantine fortunes fluctuated,
imperial forces being driven out of Armenia twice, but by the first half
of 656, Hamazasp Mamikonean was securely installed as kouropalat¯es and
prince of Armenia.28 At the same time, honours were distributed to the
other princes and treasures to the soldiers, confirming that the benefits of
imperial service were not confined to a few but were spread broadly among
the elite. Nerses III returned from exile in Tao after ‘the lord of Rshtunik‘
had died and the Arab invasion had come to an end’, indicating an earlier
date, perhaps 656, than is generally admitted.
Constans II was determined to exploit the unexpected breathing space
afforded by the outbreak of civil war or fitna across the caliphate. He
sought to establish a broad network of clients across the Transcaucasus.
Juansher, prince of Albania, and the princes of Siwnik‘ quickly submitted.29
In autumn 659, the emperor undertook a second progress eastwards lasting
several months.30 He ventured into Media, meeting and rewarding loyal
True Cross. Constans was also seeking to attract others, including Persians
who wished to fight on against the Arabs. He was still in Armenia in
spring 660, at Vagharshapat, where he rewarded Juansher a second time.
A later text suggests that the emperor was present at the inauguration of
the impressive church of Zvart‘nots‘.31 Whilst this cannot be proved, his
involvement would have done much to bolster the standing of its founder
Nerses III and the pro-Chalcedonian party across Armenia. Intriguingly,
the terse inscription commemorating Nerses’ role is in Greek rather than
Armenian.32
In the event, Constans II’s vision of a chain of clients did not survive
beyond the conclusion of the fitna. As the lynchpin of the network,
Hamazasp was swiftly removed and replaced by his brother Gregory
Mamikonean, previously a hostage of Mu‘awiya. Juansher transferred his
allegiance to the ‘king of the south [Mu‘awiya]’, when ‘the emperor of the
Romans [Constans] took the dregs of his forces and hastened across sea
and land to cross to the . . . distant islands of the west’.33 It seems very
likely that the principal Byzantine clients had been displaced or turned by
late 661 or early 662.
After 661, the limitations of the primary sources make it much harder to
trace the interaction between Byzantium and Armenia. The conventional
approach has been to treat this dearth of information as evidence for the
exclusion of Byzantine influence. Armenian colophons and inscriptions
together with isolated textual references collectively support an alternative
view, of persistent, wide-ranging Byzantine engagement until 730 but a
more limited focus thereafter, concentrated on and operated through those
districts bordering imperial territory.
The second sustained period of civil war across the caliphate after 680
afforded a fresh opportunity for Byzantine intervention. According to
Lewond’s History, Armenia repudiated Arab sovereignty by refusing to pay
tribute, probably in 682, but it is impossible to prove Byzantine influence
lying behind this decision.34 A later Armenian source records how
an Iberian prince, Nerses, massacred the Arab forces in Armenia during
the time of Catholicos Israel I (667–77).35 The Arab blockade of Constantinople
between 674 and 678 supplies an appropriate historical context
for just such a diversionary campaign but a Byzantine connection remains
conjectural (see also pp. 233, 372).
Constantine IV (668–85) was eager to exploit contemporary disorder
across the caliphate. In 685, he invaded Cilicia and threatened northern
Syria, compelling the new caliph, ‘Abd al-Malik (685–705) to sue for peace
on very generous terms on 7 July 685.36 This campaign may have been
coordinated with the devastating Khazar raid into Armenia during which
Gregory Mamikonean and Nerses were killed in battle on 18 August 685.37
According to Theophanes the Confessor, Justinian II (685–95, 705–711) ratified
the truce with ‘Abd al-Malik soon after his accession although its term
was extended to ten years and an additional provision was inserted, requiring
the parties to share the tax revenue of Cyprus, Armenia and Iberia.38
A subsequent passage under the same year entry adds that Justinian II
despatched a strat¯egos, Leontius, into Armenia. He subjugated Armenia,
together with Iberia, Albania, Boukania (probably Vaspurakan) andMedia,
imposed taxes on those countries and remitted a large sum to Justinian.
The changes to the treaty make sense when viewed in the aftermath of
this raid. The revenue arrangements may reflect a more fundamental partition,
of sovereignty.GregoryMamikonean’s successor as prince of Armenia
was Ashot Bagratuni, titled patrikios. Since he also brought an icon of the
incarnation of Christ ‘from the west’ for his church at Daroynk‘, forty
kilometres south of Mount Ararat, it seems likely that he was a Byzantine
client.39
After Ashot’s death – confronting Arab raiders in the Araxes valley
in 689 – a number of Armenian princes switched allegiance. This
prompted Justinian II to travel to Armenia in person, as his grandfather
Constans II had done in similar circumstances. Justinian summoned the
princes to him, taking some of their sons hostage, while rewarding others: he
raisedNerses Kamsarakan, the lord of Shirak, to the rank of prince of Armenia
and the patrikios and exarch Varaz(tr)dat was made prince of Albania.40
He then returned to Constantinople, taking with him Catholicos Sahak III
(677–703) and five bishops. Theophanes likewise reports Justinian’s visit to
Armenia although he places it too early, in his second year, and wrongly
associates it with theMardaites.41 A remarkable, pro-Chalcedonian account
of Armenian ecclesiastical history, which survives only in Greek, records
that Sahak and his bishops accepted Chalcedon at a council convened in
Constantinople in the fifth year of Justinian II, although on their return to
Armenia and under pressure, they reneged.42
This revival in Byzantine fortunes occurred in the context of the second
fitna. Even before his final victory over his main rival in 691, Caliph ‘Abd al-
Malik was turning his attention to Byzantium. Contrary to the traditional
view, it seems very probable that it was ‘Abd al-Malik, not Justinian II, who
broke the ten-year truce.43 The heavy Byzantine defeat in 692 at Sebastopolis
occurred deep inside newly secured Byzantine territory, indicating an
Arab offensive (see below, p. 384). Several Armenian clients promptly transferred
allegiance but the Byzantine position did not collapse overnight. A
colophon confirms that the principal Byzantine client in 689, Nerses Kamsarakan,
was still alive in 696 and in contact with Constantinople.44 The
region of Fourth Armenia also resisted. Although Muhammad bin Marwan,
the governor of al-Jazira, campaigned there in 694/5, evidently it had
not been subjugated in 701/2 when Baanes ‘Heptadaimon’ switched sides.45
Perhaps most surprisingly, in 702 Smbat Bagratuni rebelled and defeated an
Arab force at Vardanakert, being rewarded with the title kouropalat¯es.46 A
parallel account of this uprising, but with a Kamsarakan spin, affords useful
corroboration.47
The aftermath of this rebellion remains confused. Lewond maintains
that Smbat withdrew into Tao and that Catholicos Sahak III negotiated
a three-year peace. According to the History of the Albanians, however,
military operations continued.48 Dvin fell to a joint Byzantine-Armenian
force whilst the Arabs captured a fortress in Sevan only after a three-year
blockade. Both sources agree that a Byzantine force then suffered a heavy
defeat. Lewond adds that this occurred in Vanand in the first year of Caliph
al-Walid I (705–15). The Byzantine troops fled and the Armenian rebels
suffered severe reprisals, with 800 men in Nakhchawan and 400 in Khram
being imprisoned in churches and then burnt alive. Ominously, the lord
of Shirak, Nerses Kamsarakan, was summoned to Syria in 705; his fate
is not recorded. Smbat kouropalat¯es escaped into Byzantine territory and
was settled in the city of Phasis in Lazica. This sequence of events – a
rebellion by Armenian princes, contact with Emperor Tiberius II Apsimar
(698–705), the despatch of Byzantine forces, a successful counter-offensive
by Muhammad bin Marwan followed by the burning alive of Armenian
princes – is corroborated by Theophanes.49 The only significant difference
is chronological. Theophanes records this sequence of events under one
year, AM 6195 (702/3) but it seems more likely that they were spread across
several years (702–5).
Aside from the failed attempt at union in the time of Justinian II outlined
above, relations between the churches after 661 are almost entirely obscure.
In 719, however, Catholicos John III (717–27) stated unequivocally that the
six catholicoi after Komitas (between 628 and 705) were all Chalcedonian,
exempting only his immediate predecessor Elias (703–17) from criticism.50
As outlined previously, Ezra, Nerses III and Sahak III all engaged in discussions
with the imperial church but none of their correspondence or
other writings survives. Indeed the only extant letter between 628 and 705
is a draft Armenian ‘Defence’ of the monophysite position, prepared in
649 for despatch to Constans II.51 Arguably, no records or letters associated
with these catholicoi survive precisely because of their confessional perspective.
An exchange between Patriarch Germanos I (715–30) and Catholicos
John III from the 720s does survive, defining and defending their respective
positions in great detail.52 Conceivably this correspondence marks the
final breach between the churches and was preserved because it articulated
the differences. Confessional tensions at the highest level need not have
deterred other contacts. Colophons reveal that four patristic works were
translated into Armenian in Constantinople between 713 and 717 by David
hypatos and Stephen of Siwnik‘.53
After 730, Byzantine influence persisted but on a more limited scale. An
inscription on a tombstone located in a crypt atNakhchawan in Shirak commemorates
‘the blessed lord Artawazd Kamsarakan apo hypat¯on patrikios and
prince of Armenia, son of Hrahat patrikios lord of Shirak and Asharunik‘’.54
Artawazd was the grandson of Nerses Kamsarakan mentioned previously.
Evidently Byzantine titles continued to be awarded during the eighth century
to Armenian princes. Artawazd does not feature in any other source,
which is surprising given his rank of ‘prince of Armenia’. His omission is
hard to explain unless one views him as a second, rival prince of Armenia
and client of Byzantium.
When the third fitna erupted, two groups of Armenian princes may once
again be discerned. One party, under Ashot Bagratuni, remained loyal to
CaliphMarwan II (744–50); the other underGregoryMamikonean, looked
to Constantine V (741–75). Having taken refuge in Tao, ‘they relied upon
the forces of the king of the Greeks, who were in the regions of Pontos,
because there was a treaty of peace between them, at the command of the
emperor Constantine’.55 After blinding Ashot Bagratuni, perhaps in 748,
Gregory went to Theodosioupolis and broadcast news of his victory. Evidently
Theodosioupolis was under his, or Constantine’s, control and he was
attempting to attract further support. His success or otherwise in this initiative
is not recorded by Lewond, who simply notes that he died in agony
at an unspecified date and was replaced for a short time by his brother.56
Whether Lewond’s hostility stems from a political (anti-Mamikonean) or
confessional (anti-Chalcedonian) perspective is unclear. Again this temporary
Byzantine revival in Armenia was halted by the resolution of the strife
within the caliphate. In 754, Constantine V transferred the population of
Theodosioupolis to Thrace. Lewond adds that many from the surrounding
districts also left and ‘placed themselves on the side of the pious king’, a rare
favourable view of Constantine V.57 This transfer may represent a tactical
withdrawal at the end of a series of initiatives in Armenia rather than the
original goal.
Armenian princes did not risk rebellion against the dominant, controlling
power without support, or expressions of support, from a rival power
other than in exceptional circumstances. At first sight, the complicated
series of rebellions across Armenia in the 770s fall into that category. At
no stage do the narrative sources indicate any Byzantine involvement.58
Two of the rebel leaders, Artawazd and Mushegh Mamikonean, are said
to have begun their uprisings by killing local Arab tax-collectors. New
administrative arrangements and fiscal burdens at district level may have
precipitated their actions. On the other hand, Artawazd moved into Iberia
and later reappears as strat¯egos t¯on Anatolik¯on whilst Mushegh’s rebellion
apparently took the form of a prolonged, and ultimately unsuccessful, siege
of Theodosioupolis. This strategy is hard to fathom unless one accepts that
Byzantine support was anticipated. No Byzantine campaign is recorded
but it may have been planned; in 777 a large Byzantine army, under Armenian
commanders, attacked Germanikeia and devastated the surrounding
region.59
For the following five decades, there is very little evidence for Byzantine
involvement in Armenia. In 788 as many as 12,000 people under the leadership
of Shapuh Amatuni, his son and other Armenian nobles were granted
refuge within the empire by ‘the emperor Constantine’. Lewond portrays
this as a reaction to hardships inflicted by the caliph and his representatives,
specifically the seizure of land.60 It is in the last quarter of the eighth
century that several quasi-independent Arabic emirates emerged, ruling
districts previously under Armenian control.61 At the same time, members
of the Bagratuni princely house exploited their status as preferred Abbasid
clients to secure a dominant position. After 775, Byzantine attention was
concentrated on potential clients in those districts of Iberia which abutted
imperial territory. Ashot Bagratuni, established in neighbouring Klarjet‘i,
was appointed kouropalat¯es before 826.62 Byzantine strategy towards Armenia
came to operate on and through the remote district of Sper which
bordered the theme of Chaldia. The first ninth-century Armenian prince
known to have been accorded an imperial title was another AshotBagratuni,
prince of Sper; he was appointed patrikios and apo hypat¯on by Theophilos
(829–42).63 Intriguingly, his appointment is recorded in the context
of Byzantine operations against Theodosioupolis, Basean and Vanand, all
to the south and east of Sper. Although these operations have been compressed
into a single campaign and linked to a major Byzantine offensive
against Sozopetra, Melitene and Fourth Armenia undertaken in 837, they
could equally comprise separate campaigns spread over a number of years.64
This targeting of Theodosioupolis and its surrounding districts mirrors
the pattern of Byzantine offensives outlined previously, whilst the Khurramite
rebellion under Babek afforded a suitable opportunity (see below,
p. 390).
Caliph al-Mu‘tasim (833–42) responded swiftly to this Byzantine threat.
In 838, his forces inflicted a heavy defeat upon Theophilos at Dazimon
and captured Amorion. Genesios reports that Armenian forces under the
‘Vasparakanites’ (presumably the leading Artsruni prince) and the prince of
princes (probably Bagarat Bagratuni, prince of Taron) participated in these
campaigns.65 This represents a rare instance of active service by Armenian
forces against Byzantium. It illustrates how closely the leading Armenian
princes now identified with caliphal interests and the degree to which
Byzantine influence over them had waned.
World History
