Battle of Gaugamela 331 BCE

Battle of Gaugamela (October 1, 331 BCE): decisive battle in the war between Macedonia and the Achaemenid Empire, fought in northern Iraq. The outcome was influenced by a celestial omen that announced the imminent downfall of the Persian king Darius III Codomannus and the succession by Alexander the Great.

Continue reading Battle of Gaugamela 331 BCE

Herodotus of Halicarnassus

The Greek researcher and storyteller Herodotus of Halicarnassus(fifth century BCE) was the world’s first historian. In The Histories, he describes the expansion of the Achaemenid empire under its kingsCyrus the Great, Cambyses and Darius I the Great, culminating in kingXerxes’ expedition in 480 BCE against the Greeks, which met with disaster in the naval engagement at Salamis and the battles at Plataeaand Mycale. Herodotus’ remarkable book also contains excellent ethnographic descriptions of the peoples that the Persians have conquered, fairy tales, gossip, legends, and a very humanitarian morale. (A summary with some historical comments can be found here.)

This is the first part of an article in eight pieces.


Herodotus of Halicarnassus hereby publishes the results of his inquiries, hoping to do two things: to preserve the memory of the past by putting on record the astonishing achievements both of the Greek and the non-Greek peoples; and more particularly, to show how the two races came into conflict.

These are the confident opening lines of Herodotus’ Histories, and the Greeks who heard them must have been surprised. Preserving the memory of the past by putting on record certain astonishing achievements was not unusual, but the bards who had been singing legendary tales had been less pretentious. Even the great poet Homerhad started his Iliad in a more modest way:

Sing, goddess, the wrath of Peleus’ son Achilles, that brought endless harm upon the Greeks. Many brave men did it send down to the Underworld, and many heroes did it yield a prey to dogs and vultures. In this way, the counsels of Zeus were fulfilled, from the day on which Agamemnon -king of men- and great Achilles first fell out with one another. And which of the gods was it that set them on to quarrel?

The similarity between these two prologues is obvious: we are about to hear a tale about a terrible conflict and the speaker wants us to understand how the two sides came into conflict. The difference is striking, too: Homer invites a goddess to relate the story; Herodotus does not need divine aid. Who was this man, who so proudly gave his personal opinion about the past?

Herodotus’ life

Not much is known about Herodotus’ life. The only reliable source we have is the book he wrote, known as The Histories, and this remarkable text gives us some clues that enable us to sketch the outlines of its writer’s life. As its prologue shows, Herodotus was born in a town called Halicarnassus: modern Bodrum in southwestern Turkey. Not far from Herodotus’ native city is the island Samos, which figures so prominently in The Histories, that it has been argued that Herodotus spent several years on it. The same argument applies to Athens: Herodotus may have spent some time in the leading Greek city of his age., the online home of Ancient Warfare magazine
The theater and acropolis of Halicarnassus. Photo Marco Prins.
Halicarnassus / Bodrum

It is unknown when or why he left his home town. Two or three centuries after Herodotus’ death, scholars from Alexandria assumed that the historian was banished because he had been involved in an abortive coup attempt. Unfortunately, there are many ancient historians who were forced to spend part of their lives abroad after a political failure (e.g., Thucydides, Theopompus of Chios, Timaeus, and Polybius of Megalopolis). Probably, it is safer to ignore this piece of scholarly speculation.
The famous Macedonian philosopher Aristotle of Stagira (384-322 BCE) must have heard or read The Histories. In his book on Rhetorics, he quotes its first line:

Herodotus of Thurii hereby publishes the results of his inquiries…
[Rhetorics 1409a27]
An easy way to explain this variant reading of Herodotus’ opening line is that Aristotle was simply mistaken. However, the philosopher’s infallibility has been axiomatic for centuries, and many scholars -ancient and modern- have tended to believe that Herodotus was one of the settlers in the South-Italian city Thurii, which was founded in 444 BCE. A medieval dictionary, the Suda, mentions Herodotus’ tomb on the market of Thurii (Suda H536); this was a high honor, only attributed to the (often legendary) founders of new cities. Of course it is possible that Herodotus was the founder of Thurii, but probably we are better advised to take the Suda’s statement with a grain of salt, especially since Athens and Pella (in Macedonia) also claimed his tomb. It is imaginable that the Thurians have invented theirs after reading Aristotle.

The year of Herodotus’ death is unknown, but we have two clues. In section 137 of Book Seven of The Histories the execution of two Spartans in Athens is mentioned. From another source, The history of the Peloponnesian War by the Athenian historian Thucydides (2.67), it is known that the two were killed in the winter of 430/429 BCE. Therefore, Herodotus was still alive and writing in 429. Since it is also known that in the summer of 429 many Athenians were killed by the plague, it may be conjectured that Herodotus was one of the victims of this disease. However this may be, he must have died before 413, because he tells (Book Nine, section 73) that a certain village in the neighborhood of Athens, Decelea, was never plundered by the Spartans, something that did in fact happen in 413, as Thucydides tells us (6.93)
Assuming that Herodotus died between 429 and 413, it is reasonable to infer that he was born between 500 and 470. Perhaps we can be a little bit more precise: nowhere in The Histories does he claim to have witnessed the great Persian War (480-479 BC) that he is describing. Therefore, his date of birth can be estimated in the eighties of the fifth century BC.

The author of The Histories seems to have been a real globetrotter. If we are to believe him, he was no stranger in Babylon, where he interviewed the priests; he claims to have gone north to the Crimea and south along the Nile; he visited Sicily and knows the details of North-African topography. However, some doubts are possible: e.g., his description of Babylon is contradicted by archaeological evidence (see below). On the other hand, in his description of the Crimea, he mentions a king known to have lived around 460, which makes it likely that he really visited that part of the world.

That he was able to write, is a fact easily ignored. However, it tells us that his parents could afford a teacher and were well to do. Herodotus must have been a rich man, possibly a member of the old aristocracy. We may speculate that he fought as a heavy armored infantryman (a hoplite), like all Greek men of his class and age. This would explain why his descriptions of battles are always from a soldier’s point of view and sometimes confused. He was a soldier, not a general.

This is all we know about the Father of History: frustratingly little. Yet, there are only a few ancient writers that we know as well as Herodotus. Other authors wrote longer texts, were greater historians, or reached greater intellectual heights, but none of them is able to convey the same feeling of intimate friendship that we experience when we read Herodotus. We meet him when he is in a dark mood, share his surprise, know his religious opinions, hear him chattering, joking and babbling. There is no ancient author whose character we know so well as the man about whose life we know so little. The solution to this paradox lies in The Histories.

Herodotus’ originality

Today, The Histories are usually edited in one volume. In Antiquity, nine scrolls were needed to contain the entire text, and it is still usual to divide The Histories into nine ‘books’. As the Italian classicist Silvana Cagnazzi has pointed out, it is possible to subdivide every ‘book’ into three units, the logoi (overview). When a person reads one of these logoi to an audience, he or she needs about four hours, and it is likely that this is how Herodotus first ‘published’ the results of his inquiries: as a lecture. This idea corroborates an ancient story that he used to recite his work. (On one occasion, a boy started to cry: the future historian Thucydides, who was deeply moved by Herodotus’ narrative.)
It is likely that at one point Herodotus decided to collect his logoi in one continuous text. But now he faced a serious problem. His logoi were about very dissimilar subjects -e.g., a description of Egypt, a logos about Scythian customs, and a narrative about Persian diplomacy in the winter of 480/479- and it was likely that this collection of logoi would become a messy whole. Herodotus has recognized this problem, and decided to group everything around one single theme: the expansion of the Achaemenid (or Persian) empire between 550 and 479. Lectures on topography and ethnography now became integrated chapters of a historical chronicle., the online home of Ancient Warfare magazine
Homer. Glyptothek, M�nchen (Germany). Photo Marco Prins.
Homer (Glyptothek, Munich)
Stories about the past were something that the Greeks primarily knew from the beautiful epic poems of Homer, who had sung about the valiant deeds of past heroes in the Iliad and the Odyssey. Herodotus was heavily influenced by this example. Sometimes he quotes the legendary bard; or he uses words that any Greek would have recognized as homeric. The Iliad contains a catalogue of nations that took part in the Trojan War; in Book Three, Herodotus sums up all Persian provinces, and in Book Seven, he inserts a list of troops that took part in Xerxes’ expedition to Greece. Sometimes, Herodotus copies scenes from Homer. In his description of the Battle of Thermopylae, he tells how the Spartans and Persians fought about the body of Leonidas. This is impossible in a hoplite-battle (the type of warfare Herodotus is describes) but echoes a scene from the Iliad in which the Greeks and Trojans fight about the body of the hero Patroclus.

A very important borrowing from Homer is the circular composition. More than a hundred times, Herodotus interrupts his narrative to digress on a subject. The longest digression is Book Two: Herodotus announces that the Persian king Cambyses wanted to conquer Egypt, and then begins to talk about the geography, the customs and the history of the ancient country along the Nile. Finally, at the beginning of Book Three, Herodotus resumes his narrative and describes the Persian invasion.
The digressions belong to the most entertaining parts of the Histories. For example, we read an interview with an employee of an Egyptian mummy factory, an astonishing anecdote about the first circumnavigation of Africa, a hilarious tale about Indian goldmining, a report about the sources of the Nile and the Danube (see below), a reconstruction of the language of the prehistoric Greeks, a cautionary tale about deposits, and lots more.

Modern bust of Herodotus, Bodrum (Turkey). Photo Jona Lendering.
Modern bust of Herodotus, near the Museum of Bodrum

A final point of similarity between Herodotus and Homer is the impartiality of the narrative: Homer’s heroes are the Greeks, but his Trojans are no villains, and in the same way Herodotus portrays his Greeks and Persians – he treats both parties without partiality or hatred, but with genuine sympathy. It is interesting to compare this with the historiographical texts from the oriental monarchies: the Persian shah -e.g., the Behistun inscription- and the Egyptian pharaoh leave no doubt about the wickedness of their opponents.
But Herodotus is more than just a pupil of Homer who added geographical and ethnographical bits and pieces to his unbiased epic tale. A first difference is that Homer was a poet using a complex meter, whereas Herodotus composed his logoi in prose. But the greatest difference is the fact that Herodotus was a real researcher, an empiricist. (In fifth century BC Greek, the word historia still meant ‘research’; it was Herodotus’ achievement that the meaning of the word changed.) He traveled a lot in order to investigate the cities and opinions of man. Where Homer claimed to be speaking the truth depended on his inspiration from the muses, Herodotus based his narrative on research. It is a tribute to the quality of Herodotus’ geographical descriptions that the works of his predecessors are now lost.

As a corollary of Herodotus’ empiricist method, he is interested in the recent past. Homer had told about distant, legendary antiquities; Herodotus was interested in events that were in living memory and could be verified. For example, he seems to have interviewed the survivors of the Battle of Marathon. Admittedly, interviews are an unreliable source, but it must be said that Herodotus did a remarkable job: when we can check The Histories, it often turns out to be trustworthy. Even though Herodotus makes some serious mistakes, he managed to give a pretty accurate description of the century before his birth.
As it turned out, Herodotus invented a new literary genre: history. He did so by integrating the results of empiricist ethnographic and topographic research into epic, and writing this in prose. This combination was revolutionary.

Thucydides. Mosaic from Jerash, now in the Altes Museum Berlin (Germany). Photo Jona Lendering. Thucydides; mosaic from Jerash (Altes Museum, Berlin)
It is odd that he was hardly appreciated in Antiquity. People admired his entertaining way of telling stories, but they did not believe them. The first to criticize the Father of History was Thucydides, who rejected Herodotus’ religious explanation of what was happening (below). In later times, nobody dared to believe what Herodotus told about strange customs. For almost two thousand years, people considered him just a teller of (excellent) tales and thought that all these strange customs were merely inventions. His never ending stream of tall, short and winding tales earned him -as Salman Rushdie would say- not one but two nicknames: to some, he was the Father of History, but to others, he was the Father of Lies. Only when, after the discovery of the Americas, the Europeans learned to know the customs of hitherto unknown people, the reappreciation of Herodotus started. But even today, many of his claims are the subject of debate.


The appearance of Armenian literature in the second half of the fifth century CE, in the generation which followed the great revolt of the Armenian nobles in 450 against Yazdgird II’s attempt to re-impose Zoroastrianism on their already Christian country (see EIr. II, pp. 429-30), resulted in its almost total obliteration of Armenia’s ties to the Iranian world. The ideology of its exclusively Christian, ecclesiastical authors reiterating that the Armenian self-image was Christian by definition, simultaneously obscured any memory of the country’s earlier past. Consequently, Armenian sources, particularly in the case of Iran, must often be read as through a distorting mirror. Persian sources are all but non-existent except for brief references in early inscriptions, to which late Sasanian seals add occasional, mostly administrative, details. Chance references in classical sources are often inaccurate or hostile. The minimal archeological evidence for the eleven centuries separating the rise of the Achaemenids from the fall of the Sasanian dynasty, is derived almost exclusively from the territory of the present Republic covering but a scant fifth of historic Armenia. Such evidence has done little to remedy the lacunae of the native written sources and brings with it the risk of distortion through conclusions derived from pars pro toto.

Additional difficulties arise from the still obscure cultural chronology of early Armenian history, which does not coincide with that of the great powers on either side. This is particularly important in the case of Irano-Armenian relations, where the junior, Armenian Arsacid dynasty survived by two centuries the overthrow of its senior, Parthian branch by the Sasanian revolution; the time lag is reflected in institutional discrepancies, since Armenian society preserved, anachronistically, an earlier Iranian pattern. Nevertheless, even in the face of such patent obstacles, no serious study can avoid recording the fundamental elements linking pre-Islamic Armenia to Iran, especially in the crucial, if occasionally subliminal, cultural aspects which were to survive the political vicissitudes of more than a millennium from the 6th century BCE to the mid-7th century of the Christian era.

Persepolis_Armenian_Subjects.JPG (35095 bytes)

 Apadana relief showing Armenian subjects representing their sovereign with their Gifts – Nowruz Ceremony at Persepolis

(Click to enlarge)

The Achaemenid period

The first written evidence for the name of the plateau at the easternmost edge of Anatolia, an area increasingly dominated by the Indo-European speakers whom Herodotus (7.73), would call “Armenioi,” comes from the late 6th century BCE; Darius I’s Bisotun (q.v.) inscription refers to it as “Armina” (DB 1.15, par. 6; Kent, Old Persian, p. 119). According to Herodotus (3.93), Armenia was part of the Achaemenid empire, of which it formed the satrapy XIII; and more than a century later Xenophon would mention a “palace of the satrap” in one of the villages he passed in the tribal, non-urban land crossed by his army in its retreat from Mesopotamia to the Black Sea (Xenophon, Anab. 4.4.2). The actual relationship of Armenia to Iran under the Achaemenids is not altogether clear, since Darius’s inscription at Susa lists Armenia among the countries which “Ahuramazda bestowed upon me” (DSm 5-11; Kent, Old Persian, p. 145; cf. DB 1.14-17). At the same time, it is also given among “the countries which I got into my possession along with this Persian folk, which felt fear of me (and) bore me tribute” (DPe 5-18; Kent, Old Persian, p. 136).

Whatever the precise relationship of the two may have been during this early period in which a clear concept of �r�n�ahr (Ir�n�ahr) had not yet developed (Gnoli, 1989), the position of Armenia was especially privileged. Armenian satraps are singled out by Xenophon as having normally intermarried with the family of the king of kings (Xen., Anab. 2.4.8, 3.4.13), so that the refusal of one of the king’s daughters promised to him was considered sufficient cause for one of them to rebel (Xen., Anab. 4.4.4). According to the same author, the satrap of western Armenia, “a friend of the king … was the only man permitted to help the King mount his horse,” whenever he was present (Xen., Anab. 4.4.4). Perhaps as a result of this privileged status, the Armenians generally remained loyal to their Achaemenid overlords. It may have taken three campaigns to subdue Armenia in the chaotic period which attended Darius I’s accession to the Iranian throne, though even here, an Armenian, Dadarshish, commanded the Persian army. However, Armenian contingents under the leadership of Darius’s son-in-law Artochmes (Herod., 7.93) took part in the great expedition of Xerxes against Greece in 480 BCE, and they were still found 150 years later, in 331, supporting Darius III Codomanus against Alexander the Great at the battle of Gaugamela (Arrian, Anab. 3.8.5).

The clearest evidence for the interrelation of Iran and Armenia has been derived from a comparison of classical and eastern sources juxtaposed and interpreted by Manandian (1966, p. 36-38) and more particularly by Cyril Toumanoff (1963, pp. 277-305). These are: Strabo (11.14.15), the later semi-mythological material preserved in the History of Armenia by the Armenian historian Movs�s Xorenachi (MX, 2.27-46), the fragments of Hellenistic Greek graffiti found at Armawir on the left bank of the middle Araxes, and finally the genealogical inscription of king Antiochus I of Commagene (q.v.) at Nimrud Dagh from the middle of the first century BCE, in which he claimed descent from the Achaemenid kings. The result of their work revealed the presence of a forgotten native dynasty of Iranian origin, called “Eruandid” (cf. Av. auruuant- “mighty, hero;” Mid. Pers. arwand) by the Armenians from the repeated name of its rulers, or Orontid from the various awkward Greek transcriptions of their name�such as “Orontes Aruandes” or “Ardoates”�found in classical sources. The presence of this Iranian native dynasty can now be attested from at least 400 BCE, and it can be shown to have ruled, from the centers of Armawir and subsequently Eruanda�at (q.v.) on the middle Araxes with only a brief hiatus, until the first years of the Christian era.

Very little is known about Armenia’s early tribal society, beyond its agricultural wealth and absence of cities, as noted by Xenophon in the description of his journey across the Armenian plateau; but its ties to Iran are also clearly attested. We learn from Xenophon that the Armenians paid tribute to the Achaemenids. According to Strabo (11.14.9 [C 530]), they were particularly known for the prized horses that they raised and sent to Iran for the celebration of the Mithracina (see MEHRAGA�N). These statements are supported by the presence of male figures usually identified as Armenians, wearing the Persian or rather Median dress of short tunics and trousers and leading a horse, represented in the procession of gift-bearers figured on the great staircase at Persepolis (Ghirshman, 1964, p. 271, fig. 216), as well as by the discovery of silver gilt plates with the central relief of a horse raising his right foot in obeisance to a fire altar, found north of Armenia at Aramazis-xevi, near the early Georgian capital of Mtsxetha (Lang, 1966, p. 89, fig. 20).

Post-Achaemenid Period

The campaigns of Alexander shifted the position of Armenia for centuries from that of an intrinsic component part of the Achaemenid empire to that of a disputed borderland at the limit of the classical and the Iranian worlds. The strategic position of the region lying athwart the east-west military and trade routes, both along the valley of the Araxes leading from Iran to Cappadocia and more particularly through the Mesopotamian plain dominated by the Armenian plateau, made it far too important to permit its concession to a rival power. Alexander himself never entered the country, and the control of the plateau by his Seleucid successors was intermittent. Nevertheless, Armenia came under a powerful Hellenic influence that probably reached its zenith in the last century BCE at Tigran II the Great’s Greek-speaking court. Eponymous cities following the pattern of the ubiquitous Alexandrias, such as Artashat, Eruandashat, Zarishat, Zarehawan, Valar�apat, and Tigranakert, were founded in Armenia. The country’s Iranian base seems to have survived, however, since the names of the eponymous founders are invariably of Iranian origin. The reappearance of an “Orontes” as early as 316 BCE, and of a “Mithranes,” another member of the Eruandid house, even earlier (Toumanoff, 1963, p. 280), clearly demonstrates that the rule of this local dynasty of Iranian origin had survived, with hardly any interruption, the destruction of the Achaemenid empire by the Macedonian conquest.

According to Strabo (11.14.5), new dynasties were established in Araxene Armenia and its southwestern neighbor Sophene by Artaxias (Arm. Arta��s) and Zariadris (Arm. Zareh) at the end of the 2nd century BCE. He identifies both rulers as generals of the Seleucid king Antiochus III who had established themselves as a result of Antiochus’s defeat by the Romans in 188 BCE. However, the recent discovery in Armenia of boundary stones with Aramaic inscriptions, in which the ruler Arta��s proclaims himself “the son of Zareh” and an “Eruandid king” (Perikhanian, 1966), demonstrates that both “generals,” far from being Macedonians, belonged in fact to the earlier native dynasty, albeit probably to collateral branches, and that the Eruandids, or Artaxiad/Arta��sids as they came to be known, with their Iranian antecedents, continued to rule Armenia as before. An unexpected corroboration of this dynastic continuity is also provided by Xenophon’s much earlier choice of the name “Tigranes” for the crown prince of Armenia in his historical romance, the Cyropaedia (Xen., Cyr. 3.1.7).

Our information remains very meager concerning the history of the later Eruandid/Arta��sids, beyond the names of some of the rulers and occasional references, until the accession of Tigran the Great (96/5�55 BCE). Politically, these Armenian rulers were forced to resist the repeated, though short-lived, attempts of the Seleucids to establish their rule over the country, as well as the rising power of the Parthians, to whom Tigran himself had been sent in his youth as a hostage and to whom he had been forced to surrender seventy valleys to free himself at his accession (Strabo, 11.14.15). Even so, during Tigran’s reign in the first half of the last century BCE, Armenia briefly became the leading power of the East in the vacuum created by the decline of the Seleucids, as well as by the rivalry between Rome, temporarily distracted by the Mithridatic wars (see PONTUS), and the growing power of the Parthian empire.

The depth and pervasiveness of the very visible Hellenistic wave, which had broken over Armenia as well as the rest of the Near East in the wake of Alexander and produced problematic hybrid cultures, now appears to have been considerably less overpowering than had been assumed previously. Its unquestionable presence in Armenia and the opening of the country to world trade, evidenced by the presence of coin hoards, did not succeed in obliterating earlier Iranian traditions. Not only did the local Eruandid/Arta��sid dynasty survive the conqueror’s death by nearly three centuries, but Armenia maintained many of its political and cultural ties with the Iranian world. Achaemenid Aramaic remained the official written language of the Armenian chancellery. Intermarriages between the Iranian and Armenian royal houses continued to be celebrated with great pomp, as was that of the sister of the Armenian king Artawazd II to the Parthian prince Pacorus at which the head of Crassus was used during a performace of Euripides’ Bacchae (Plutarch, Crassus 33). In his description of the new capital of Tigranakert, Appian (Mithr. 12.94) noted that for all of its typical Hellenistic features, the new city was also flanked by a royal hunting preserve or “paradise” (MIr. pard�z) of purely Iranian type. Despite Tigran II’s use of Greek in the proclamation of his title, its formula “King of kings” was as Iranian as his own Eruandid name. The reverse of the king’s coins, on the famous silver tetradrachms commemorating his capture of Antioch on the Orontes in 84 BCE, displays a purely classical iconography; but on the obverse the king is represented wearing the tiara, decorated with pearls, of the Parthian rulers and marked with the symbolic star of the semi-divine Oriental monarchy (Der Nersessian, 1969, fig. 24); and there is some evidence for the existence of an epic about Tigran following an Iranian pattern.

Additional evidence for the Armenians’ share in the world of Iran is readily found in the realm of religion. In accordance with the syncretic fashion of the times, later Armenian authors, such as Agathangelos (Aa, secs. 785, 809) or Movs�s Xorenachi (MX, 2.12, 77) gave Greek equivalents for the names of the gods worshipped in Armenia during the Hellenistic period; but there can be little doubt that their identifications of Zeus/Aramazd, Artemis/Anahit, Apollo/Tir, Herakles/Vahagn (Av. Vәrәθragna), or Hephaistos/Mithra merely covered a purely Iranian pantheon, whose presence in Armenia has been minutely studied by James Russell (1987). The very name of the Eruandid holy city of Bagaran (< OIr. baga “god” + -d�na the Iranian suffix of place > Arm. �aran) points to the Iranian antecedents of the holy place. The same derivation is also to be found in numerous other Armenian toponyms such as Bagawan, Bagrewand, Bagayari�, whose meaning was still altogether clear to Christian Armenian writers of the 5th century CE (Aa, secs. 790, 817). Anahit/An�h�d/Ana�tis seems to have been especially reverenced, as she is usual styled “the lady” (Mid. Pers. b�n�g, Arm. tikin) in both Pahlavi and classical Armenian sources. Strabo (11.8.4 [C 512]; 11.14.16 [C532]) further describes a temple of Ana�tis erected by the Persians and the “exceptional honor” paid to the goddess by the Armenians, “who have built temples in her honor in different places, and especially in Acilisene.” These shrines were still known to later Armenian writers (Aa, secs. 48, 50, 53, 59, 127, 786). Strabo (1.2.39 [C46]) also mentions shrines called Iazoneia, which he mistakenly associates with a cult of Jason, but which may in fact have been sacred places whose name derived from OIr. yaz- “sacrifice,” although this interpretation has recently been questioned. The famous journey of the Parthian prince Trdat I for his coronation by Nero at Rome was greatly lengthened because the future king of Armenia, accompanied by Magian priests, insisted on traveling the entire way by land from fear of accidentally polluting the sea. Once in the capital, both adored the emperor “as I do Mithra”; and the prince is said to have initiated him into some of the Magian rites (Pliny, N.H. 30.6.16-17; Dio Cassius, 63. 5. 2). Excavations on the site of the Arta��sid capital of Arta�at, destroyed by the Romans and rebuilt with Nero’s permission after Trdat I’s coronation, have yielded, side by side with a statuette of clearly Hellenistic origin, a series of clay plaques with the representation of an idealized rider who must be Mithra the hunter (Garso�an, 1997, p. 15, figs. 3-4). At a much later date, the Christianizing message to the Armenians in the 5th-century CE work attributed to Agathangelos becomes comprehensible only if addressed to an audience familiar with the Zoroastrian epic tradition rather than classical mythology (Garso�an, 1985).

Parthian Period

With the establishment of Roman dominion in the eastern Mediterranean after the campaigns of Pompey and the end of the Seleucid dynasty in the mid-first century BCE, as well as with the disappearance of the Eruanid/Arta��sids at the very beginning of the Christian era, Armenia became for centuries an apple of discord for the Romans and the Iranians, be they Parthians or subsequently Sasanians; both sides vied with each other to place their candidate on the Armenian throne. According to the Roman historian Tacitus (Ann. 12.1), the Parthian king Vologeses (Arm. Valar�ak) considered the Armenian throne “once the property of his ancestors, now usurped by a foreign monarch by virtue of a crime”; but Armenia continued to see-saw for nearly four centuries in alternate allegiance between the Parthians and Rome. The Parthian Arsacid (Arm. Ar�akuni) dynasty, of Iranian origin, like its Eruandid/Arta��sid predecessor, which was eventually to hold the undisputed rule of the country, made its first appearance in Armenia with the king of kings Vonones, who established himself there in 12 CE after having been driven from the Iranian throne, although he did not succeed in maintaining himself there for more than three years. A period of war closed with the compromise accord of Rhandeia in 63 CE, which stipulated the accession in Armenia of a younger son of the Parthian Arsacid king, as long as he was crowned by the Roman emperor. This ceremony incidentally provides us, albeit in Greek translation, with the first recording of the formulaic attributes of the Iranian ruler: baxt ud xwarrah “good fortune and transcendental glory” (Dio Cassius, 62 [63].5.2). This compromise, whereby a junior line of the Parthan Arsacid house ruled Armenia as Roman clients, was not officially abrogated, although it was breached on several occasions: on the Roman side by the emperor Trajan, who annexed Armenia outright between 115 and 117 CE, by Antoninus Pius, who proclaimed on his coinage that he had “given a king to Armenia” (rex Armeniae datus), and by Marcus Aurelius, who seems to have stationed a garrison in the Armenian capital of Valar�apat in 164 for some twenty years. Similarly, the Sasanian ruler Sh�p�r I profited from his overwhelming defeat of the Roman emperor Valerian in 260 by installing two of his sons, Hormizd-Ardash�r and Narseh, on the Armenian throne, which was consequently held directly by Persia until 293 (�KZ, Parth., ll. 20-21; Paikuli, p. 28, sec. 3).

The Armenian Arsacids/Arshakuni

Insofar as we can judge from the chaotic situation reflected in the inadequate sources, Arsacid rulers held Armenia repeatedly during the first century CE. In fact, “for some one hundred and sixty five years [of the Parthian period]…thirteen sovereigns succeeded one another in Armenia, eight of whom were indeed Arsacid cadets” (Toumanoff, 1969, p. 233), whose blood relationship to the Parthian royal house is constantly stressed (Garso�an, 1981, pp. 36-37 and nn. 33, 35). Although they did not begin to consolidate themselves in the country, until the very end of the 2nd century CE, and even then precariously, the Armenian branch of the Parthian Arsacids was to rule the country almost continuously from the end of the following century to 428, despite the overthrow of their kinsmen in Iran by the Sasanian revolution two centuries earlier. In the eyes of the Armenian sources, the bond between the royal houses of Parthia and Armenia was indissoluble. Inflexibly the native authors use the terms ‘king’ and ‘Ar�akuni’ as interchangeable synonyms to the very end of the dynasty. They deny that any circumstance could deprive them of the crown or that anyone else, no matter how illustrious, might legitimately wear it. They exhort the Armenians to die for their “true Arsacid lords.” So deep was the identification of the Armenians with the Arsacid dynasty, that even in the period of its final decline early in the fifth century, the Sasanians usually conceded to them rulers from this house in order to insure their loyalty (BP, 6.1). Except for the occasional princes imposed by the Romans, none of whom succeeded in consolidating himself on the throne, all the dynasties to rule pre-Islamic Armenia were of Iranian stock.

The extreme political instability marking the history of Armenia in the first centuries of the Christian era does not seem to have affected its cultural identity. We have only glimpses of the situation within the country in this period, but these indices, which coincide with some customs and institutions that are far better attested in the subsequent 4th and 5th centuries, continue to reflect a thoroughly Iranized society persisting despite the political upheavals of war. Tacitus (Ann. 2.56) shrewdly observed that even early in the 1st century CE, long before the consolidation of the Arsacid dynasty in Armenia, the Roman candidate to the Armenian throne, prince Zeno of Pontus, wisely changed his Greek name to the more acceptable one of Artaxias/Arta��s upon his accession; and he endeared himself to his new subjects by his taste for hunting and banquets, the only two pastimes suitable for a nobleman in the Iranian world. The 120 “strategies” into which Armenia was subdivided at the time, according to Pliny the Elder (N.H. 6.10.27) and to which he refers as “kingdoms” (regna) were presumably the dynastic principalities of the great magnate; and as such they reflect the decentralized pattern of the Parthian period in Iran. The four kings who are said to have attended Tigran the Great at all times (Plutarch, Lucul. 21.5) may be the prefigurations of the four great keepers of the marches or bdea�xs of later Arsacid Armenia. The hereditary claim of the Armenian Arsacid monarchy to the throne of its ancestors and the dominant power of the “magnates,” called megisthanes or nobiles by Tacitus, all bespeak an aristocratic society of Iranian type in no way compatible with the theoretically republican Roman world.

The first of the two major events which altered significantly the relations of Armenia to Iran was the overthrow of the last Parthian Arsacid ruler Ardaw�n V by the Sasanian Ardash�r I early in the 3rd century CE; the second one followed nearly one hundred years later�the Christianization of the country. Although the “Sasanian revolution” did not interrupt the Arsacids’ rule over Armenia, its immediate result was to transform its Arsacid rulers from relatives of the Iranian royal house into enemies and avengers, as the Armenian king was said to have sworn “[to] seek vengeance for the blood of Artawan” (Aa, sec. 19). This enmity unquestionably pushed the Armenian rulers in the direction of the Romans, though they continued to vacillate in their allegiance. It led to repeated Armeno-Sasanian wars during the 3rd and 4th centuries, culminating in the disastrous campaign of Sh�p�r II in 363/4. Armenia was overrun; its Arsacid king Ar�ak II was deported to die in the Castle of Oblivion in Khuzistan, and all of the country’s earlier Hellenistic cities were destroyed. Armenia briefly recovered under his son Pap, with some help from the Romans, but soon agreed to collaborate with the S�r�n sent by the Sasanians as marzpan “governor” of the country after Pap’s murder, in cooperation with the powerful commander-in-chief of the realm, Manu�l Mamikonean, who acted as regent for the widowed queen and her minor sons.

The Marzpanate

The final solution for the endemic enmity of Rome and Iran over the control of Armenia finally came with the replacement of the unsatisfactory compromise of Rhandeia by the outright partition of the Armenian kingdom between the two great powers ca. 387, in which the overwhelmingly larger part, some four-fifths of the realm, was conceded to the Sasanians. The abolition of the Arsacid monarchy followed soon thereafter: in ca. 390 on the Roman side and by 428 in the Sasanian portion, which was soon to take the name of Persarmenia. The subsequent period, which was to last until the downfall of the Sasanians in the mid-7th century, is known in Armenian history as that of the Marzpanate [Marzpetuthiwn]. The earlier “ignobile decretum” of Jovian in 363 (Amm. Marc., 25.7.12-13) abandoning his client Ar�ak II to the Persians had already tightened the Sasanians’ hold on Persarmenia. It returned to them the eastern portion of the semi-autonomous, southern Armenian principalities lying along the eastern Euphrates/Arsanias (mod. Murad su), commonly known as the Satrapies or ethn�/gentes, which had been lost to the Romans by the earlier Peace of Nisibis of 299 CE. Following the 3rd-century example of Sh�p�r I, Yazdgird I even went so far as to impose his son as direct ruler of Armenia at the beginning of the 5th century (MX, 3.55-56). But the young prince’s reign was brief, and the Sasanians thereafter ruled Persarmenia through marzpans, a number of whom were local Armenian princes. Sigillography also attests the sporadic presence in Armenia of Sasanian military and finance officials, several of whom belonged to the great Iranian house of the Mihr�n (Gyselen, 2001a, pp. 44-45; eadem, 2002, pp. 110-11, 120-21, 131-32). Persarmenia was not only by far the largest portion of the Armenian lands until 591, when Xusrow II surrendered much of it for a time to the emperor Maurice, it was the political, religious, and intellectual center of the Armenian world. Duin, its administrative center and the normal residence of the marzpan, was also the seat of the Armenian patriarch or katholikos from the end of the 5th century, as well as a major center for international trade (Proc., Pers. 2.25.3). The creation of the Armenian alphabet and the consequent development of its early literature, as well as the elaboration of its ecclesiastical doctrine, likewise took place within the territory of the Marzpanate and not in the portion of the country under Roman control.

The restlessness of the Armenia magnates, who jealously guarded their prerogatives against any hint of encroachment, at first on the part of their own Arsacid dynasty, whose abolition they themselves had requested from the Persian king (MX, 3.63-64), now turned them against the Sasanians’ centralizing attempts threatening their privileges; and they revolted repeatedly through most of the period of the Marzpanate. The major rebellion in 450/1 resulted from Yazdgird II’s attempt to force Zoroastrianism on an already Christian Armenia; but its sequel in 482, and especially that of 571/2, inaugurated by the murder of the Persian marzpan, do not seem to have had the same exclusively religious basis. The first ecclesiastical historians, writing in Armenian some two generations after the end of the Arsacid kingdom and the disastrous defeat of 451, understandably stressed the enmity of the Armenians to the Sasanians and exalted the rebels as Christian Maccabees and martyrs for the faith (BP, 3.11). However, the actual situation does not seem to have been so simple. Even after crushing the great Armenian revolt of 451, the Sasanian court, possibly distracted by the Hephthalite (q.v.; “Kushan” in the Armenian sources) threat on their eastern border, made no attempt to pursue religious repression. Rather, it sent a marzpan, whose name Atrormizd Ar�akan indicates his descent from the former Parthian dynasty, with instructions “not to disturb the Armenian populace but to subdue it peacefully and allow everyone to practice Christianity freely” (�Ph, 2.40), in direct reversal of the very policy that had provoked the rebellion. One generation later, the leader of the rebellion, prince Vahan Mamikonean, was recognized by the Persian court as marzpan of the country; and religious as well as political autonomy was granted to Armenia in 485. At various other times also the Persian marzpans were in fact native Armenian princes, and no attempts seem to have been made to destroy local titles and institutions (Garso�an, in press). In response the Armenians often displayed their loyalty to Persia. The Armenian elite cavalry seems to have served regularly in the Sasanian army during its eastern campaigns (Ps. Seb., 11, 28). According to the 7th-century History of the Pseudo-Seb�os, the Armenian commander Mu�el Mamikonean withstood the blandishments of Bahr�m �owb�n and would not support his struggle for the throne against Xusrow II in 591 (Ps. Seb., 11). The last great Sasanian reign of Xusrow II Parv�z proved to be one of particularly felicitous relations between Iran and Armenia, as the latter flourished under the supervision of the king’s favorite, Prince Smbat Bagratuni, called Xosrov �um “the joy of Xosrov” (Mid. Pers. x�n�m) by contemporary Armenian sources (Ps. Seb., 28).

Most significantly, no cultural break seems to have gone hand in hand with the dynastic change in Iran and the attendant political antagonism in Armenia. The exact status of Armenia vis-a�-vis Iran does not seem to have changed greatly and remained ambiguous, as had been the case much earlier under the Achaemenids. In the great trilingual inscription celebrating his victories over the Romans at Naqsh-e Rostam, Sh�p�r I claimed that “Of the Aryan Empire [MIr. �r�n�ahr] the principalities and provinces (are) these: Pars, Parthia, … Armenia …” (�KZ, Parth., l. 1). An inclusion confirmed by the Letter of Tansar (p. 63), which defined “the land called Persia … from the river of Balkh to the furthermost boundaries of the land of Adarb�ig�n and Persarmenia.” However, the great mowbad Kerd�r (q.v.) included Armenia in “the region of non-�r�n, …where the horses and men of the King of kings penetrated” (KKZ, p. 71, l. 15); and one generation later the king of kings Narseh twice underscored the separation of the two realms in his inscription at Paikuli, when speaking of his move “from Armenia hither to �r�n�ahr” (Paikuli, p. 35, secs. 18, 20).

Despite this ambiguity, however, Armenia even in this period remained closer to Iran than to Rome. The late Roman historian Ammianus Marcellinus might call the Armenian king Ar�ak II, “our constant and faithful friend” (Amm. Marc., 25.7.12-13); but Armenia, lying beyond the Euphrates frontier, remained a foreign land, its annexation by Trajan an act of conquest. On the contrary, even for the Sasanians, as for the Parthians before them, Armenia claimed as an ancestral land required particular care. In justifying the inauguration of his second campaign against the Romans, Sh�p�r I complained that, “Caesar, secondly, lied and did wrong to Armenia” (�KZ, Parth., 1.5). Possibly self-servingly, the late 5th century Armenian historian writing under the Greek pseudonym of Agathangelos affirmed that: “whoever was king of Armenia had second rank in the Persian kingdom” (Aa, sec. 18). This assertion was to be corroborated, for the Sasanian period as well, by the contemporary compiler of the Epic Histories (Buzandaran Patmuthiwnkh), formerly mistakenly attributed to an otherwise unknown Phawstos Buzand or Faustus of Byzantium:

Shapuh king of Persia [Sh�hp�r II] invited Ar�ak king of Armenia, whom he honoured with the greatest deference and glory [phar�kh] … and with full royal pomp. He treated him as a brother, like a son and gave him the second domain in the realm of Atrpatakan. And they reclined together on one and the same banqueting throne in the hour of festivity, and they wore the same garments of the same colour with the same insignia and ornaments. And day after day the Persian king prepared the same crown for himself and for him. Linked together like two indivisible blood-brothers, they enjoyed themselves jointly at festivals. (BP, 4.16)

We have no explicit evidence on the Persian side that the Armenian king ranked immediately after the Iranian king of kings, but the title of wuzurg “great” used by Sh�p�r I for his son Hormizd-Ardashir “wuzurg Armen�n ��h” is not repeated for his other sons ruling elsewhere (�KZ, Parth., ll. 20-21). In 293 CE, Narseh found Armenia the final stepping stone to the throne of the king of kings (Paikuli, p. 32, sec. 13; p. 35, secs. 18-19), as had Darius III Codomanus, long before him, in 336 BCE (Justin, 10.3.3-5; Diod. Sic., 17.6). On their side, the Armenian magnates as well as the royal dynasty and even the family of the Christian patriarch prided themselves on their descent from the great Iranian houses of the royal Arsacids, the Sur�n-Pahlav or the K�ren-Pahlav. In the opinion of the Armenian historian Movs�s Xorenachi, the last representative of the patriarchal Gregorid house, Saint Sahak, “was greatly honored [by the Persian king during his visit to the court at Ctesiphon] first, because of his noble Pahlavik family” and only then because of the reverence due to God’s servants (MX, 3.51; cf. 3.64).

Social structure

So far only minor and disconnected yet tantalizing glimpses have been available for the earlier period of Armenian society, going as far back as Seleucid times, during which its basic institutions seemingly first arose. The curiously anachronistic continuity of a Parthian society in Armenia long after its replacement in Iran by its Sasanian successors was first clarified by the extensive linguistic studies of Emile Benveniste, who demonstrated that the linkage of the overwhelmingly large number of Iranian loanwords in classical Armenian is to Middle Parthian rather than to Sasanian Pahlavi forms. The striking similarities of the two, which have permitted the reconstitution of a considerable number of lost Middle Parthian words from their Armenian derivatives, identify this period as the one of particularly close contact and cultural penetration between the two societies.

Similarly, the iconography and ideology of Iranian and Armenian societies were so close at times, despite the absence of chronological synchronism in the political sphere, that the juxtaposition of examples taken from the two sides can serve to supplement the deficiencies of Sasanian written sources and of Armenian monuments during this period, thus providing what may be called a single illustrated document (Garso�an, 1997, pp. 19-23). Even so, the chronological discrepancies manifest themselves visually. On the numerous surviving steles Armenian noblemen invariably wear the typical Parthian trousers and characteristic tunic with longer pointed sides as late as the 7th century CE (ibid., p. 17, figs. 7-8).

Finally, these similarities and their continuity beyond the point of political association can best be traced in our main source for this study, the almost contemporary Epic Histories, which probably date from the 470s CE, whose reliance on the oral tradition preserving the folk memory reveal a far more accurate picture of late Arsacid and Marzpan Armenia than can be gained from more learned authors. They too reveal the image of an unmistakably aristocratic society of Iranian type, but a largely anachronistic, primarily Parthian one, displaying none of the centralizing elements seemingly introduced by the Sasanians (Garso�an, 1976). As such, their information further permits the partial reconstruction of institutions characteristic of the all but lost Parthian period in Iran, for which we have almost no native sources.

The three main estates in Armenia, that of the “magnates” (mecamec naxarars), that of the lesser nobility (azat), and what may be called the third estate, consisting of the “artisans” (r�amiks) and “peasants” (�inakans), correspond precisely to the Iranian wuzurg�n, �z�d�n, and vastrow��n “farmers.” Only the Sasanian administrative fourth estate of the “clerks” (sing. dab�r, Arm. dpir) is missing in Armenia; its function possibly was taken over by the Christian clergy. The foundations of both societies were the great noble families, of whom some fifty can be identified by name in 4th-5th century Armenia, and whose power, at least in Armenia, and probably in Parthian Iran, did not derive from the authority or will of the king, who was but primus inter pares. At the head of each of these families was its senior member, called in Armenian nahapet or more commonly tanut�r “lord of the house,” a title to which Rika Gyselen has found the precise Pahlavi equivalent kadag-xwad�y on a late Sasanian seal (Gyselen, 2001a, pp. 61-68). The economic basis of these clans lay in their possession of vast, unalienable principalities belonging to the “eternal family,” past, present, and future, of which the tanut�r, who led its “contingent” (gund) in battle, was the temporary administrator, but not the owner, and which he consequently could not transmit or dispose of in any way. Perhaps still more important were the hereditary offices held by the chief houses, which were reserved exclusively to each. Thus, the title and office of “commander-in-chief” of the Armenian army or sparapet (< OIr.*sp�dapati, cf. NPers. sip�hbad) could belong to no one except a member of the great Mamikonean house, even if its only representative was a small child patently incapable of fulfilling its duties and for whom temporary substitutes had to be appointed. For he was the heir of the house and the only one entitled to its insignia and privileges (BP, 3.11). Even the king’s manifest will could not alter the hereditary nature of this transmission, and his interference could end only in failure and tragedy:

… when Manu�l [Mamikonean] reached the land of Armenia … Va�h�, who had previously been nahapet before his return, saw him, he handed over to him the princely diadem that he had received from King Varazdat because [Manu�l] was the senior member of the clan. And so Manu�l held the dignity of nahpet-tanut�r of the clan and Va�h� was in second place. And when Manu�l had attained the glory of his lordship [tanut�ruthiwn], he first seized the office of sparapet and commander-in-chief without an order from King Varazdat. Manu�l took back for himself the authority that his ancestors had held by nature from the beginning and which King Varazdat had granted to his tutor Bat. (BP. 5.37)

The position of the Mamikoneans in Armenia was thus the precise equivalent of that of the Sur�ns in Iran, whose hereditary dignity proved incomprehensible to classical historians such as Ammianus Marcellinus, who could not decide whether “Surena” was a family name or the title of an office. The same confusion is found later in the case of the Bagratuni princes, whose hereditary office in Armenia during the Arsacid period was that of aspet “master-of-the-horse,” and whose Byzantine descendants were known as the “Aspetianoi” to the Byzantine historian Procopius (Proc., Pers. 3.12). The very terms for the various high offices in classical Armenian are identical with those found in Iran: sparapet-sp�hbad, aspet-*aspapati, hazarapet-haz�rbad “second after the king,” and others; these, as is the case with most of the Armenian administrative vocabulary, are unmistakably borrowed from Iranian terminology.

Theoretically equal in status, since they belonged to the same estate, these noble families were in practice ranked according to a rigid hierarchy governing the gah “throne” or “cushion” (< OIr. g�u-), occupied by their representatives at court banquets. These correspond to the differing “entrance- and drinking places, sitting- and standing places … according to the dignity of each man’s rank” set by the king of kings, according to the Letter of Tansar (p. 44; Garso�an, 1997, p. 13, figs. 1-2). The precise order of this hierarchy cannot be reconstructed, since the surviving Armenian “Rank List” (Gahnamak), of which the original was said to have been kept in the Sasanian court archives (MX, 3.51), is a late document of dubious authenticity; but the Epic Histories define the return home to normalcy after a period of crisis as the time when “every magnate [was] on his throne” (BP, 4.2). The Armenian sources stress the rage of the Arsacid king Ar�ak when driven from his share of the royal couch at the Persian court to the lowest place (BP, 4.54) and that of prince Andovk of Siunikh, relegated to the fourteenth cushion far below his dignity (MD, 2.1). It is the same range as that of the Iranian epic hero Rostam humiliated by the inferior place unworthy of his rank assigned to him at the banquet of Esfandi�r (Sh�hnameh xv, vol. IV, pp. 492-93).

The Iranian aspects of Early-Christian Armenia were not merely reserved to these social and official aspects; they pervaded the whole of its culture. The cities built in the Hellenistic period under Greek influence and destroyed by the great Persian invasion of 364 were not rebuilt, as unsuitable for an aristocratic society; its magnates preferred to remain entrenched in the fortresses of their distant domains far from the arm of the king (Garso�an, 1987), thus perpetuating the non-urban, Parthian, centrifugal pattern rather than that of the new “royal cities” used by the Sasanians as means for strengthening the direct authority of the ruler (Gnoli, 1989, p. 157). The Armenian kings themselves, far from residing normally in their capitals, continued to lay out hunting preserves or partez (BP, 3.8), such as Tigran the Great had once created near Tigranakert and as still can be seen in the reliefs at T�q-e Bost�n (Ghirshman, 1962, pp. 194-99, figs. 236-38). They chose to move about the country making use of rich and elaborate, but transportable, tents or pavilions, such as the ones (ma�kapa�en, ma�kawarzan) used by the Persian king in his travels and campaigns (BP, 3.21, 4.15). As in earlier times, the only acceptable diversions for kings and nobles were banquets and the hunt; and like his Iranian counterpart, the Armenian king, when in mourning, refrained from hunting (Aa, sec. 211; Suet., Calig. 5).

Most outward features of the nobility spelled out its Iranian antecedents and counterparts. The entire early Armenian anthroponymy, ecclesiastical as well as secular, is riddled with Iranian names, whether understandably in the Arsacid royal house�Trdat, Tiran, Ar�ak, Pap, Varazdat, Valar�ak, Vram�apuh, Xosrov, Vardanuhi, Xosroviduxt, Banbi�n�or more unexpectedly in that of the Mamikonean, martyrs par excellence for the Christian fait h�Artawazd, Vardan, Vahan, Vasak, Hmayeak, Hamazaspuhi. The same is surprisingly true for the Christian clerics, among whom names drawn from the general Judeo-Christian fund are remarkably rare, both in the patriarchal house of Saint Gregory the Illuminator with its Aristak�s, Vrthtan�s, Yusik, Ners�s, Pap, and still more startingly in the unsuitable names of the co-presidents of the great church council held at the patriarchal residence of Duin in 555: the katholikos Ners�s II and bishop of Mer�apuh/Mihr-Sh�p�r of Tarown (Garso�an, 1996, pp. 229-32). In peacetime the aristocracy apparently continued to wear the Parthian dress depicted on steles, but its military armament was clearly Sasanian. Mounted on horseback, heavily armored (as were their steeds), charging with a long lance, but carrying two swords and a bow as well, the Armenian elite cavalry presented the same aspect as the warriors on the monumental Sasanian reliefs at Naqsh-e Rostam (Ghirshman, 1962, pp. 177-79, figs. 219-20). The menacing presence of the sparapet Manu�l Mamikonean, lance in hand, is described: “in the greatness of his stature, the splendor of his person, the extremely strong and impenetrable iron armor [that covered him] from head to foot, also the robustness of his person and the solidarity of his armor-clad charger also bearing indestructible trappings” (BP, 5.37). This description might as easily fit the formidable royal figure in the lower register of the cave at T�q-e Bost�n (Ghirshman, 1962, p. 192, fig. 235).

Political Theory and Ideology

Reaching beyond outward manifestations, Arsacid Armenia in both Parthian and Sasanian times shared in the political, and especially the religious and epic, ideology of the Iranian world; and it is in this area that the closest and most striking similarities are to be found.

The Iranian rulers carefully traced their hereditary descent. Early on, Darius boasted: “I am Darius the Great King of Kings, King of Persia, King of Countries, son of Hystaspes, grandson of Arsames, an Achaemenian. Saith Darius the King. My father was Hystaspes; Hystaspes’ father was Arsames; Ars�mes’ father was �ri�ramnes; �ri�ramnes’ father was Teispes; Teispes’ father was Achemenes. Saith Darius the King: for this reason we are called Achaemenians. From long ago we have been nobles. From long ago our family have been kings” (DB 1.1-8; Kent, Old Persian, p. 119).

Some seven centuries later, Sh�p�r I wrote: “I [am] the Mazda-worshipping divinity [bag] Shahpuhr. King of Kings of Aryans and non-Aryans, who is of the stock of the gods, son of the Mazda-worshipping divinity Artakh�atr, King of Kings of the Aryans who is of the stock of the gods, grandson of the divinity Papak, King” (�KZ, Parth. l. 1).

So too, the Armenian Arsacids regularly transmitted the crown in hereditary succession from father to son, and on the eve of their conversion to Christianity still sought: “greeting and prosperity by the help of the gods … protection [for us] from our heroic Parthians, from the glory (phar�kh) of [our] kings and from [our] valiant [kha�] ancestors” (Aa, sec. 127). This formula interestingly is omitted in the Greek version of this text by its translator, to whom its Iranian ideology was probably incomprehensible. Even Christian authors writing after the fall of the dynasty would deny that any but the Arsacids could be the ‘true lords’ of Armenia. Thus a rigidly hereditary base for the royal power distinguished both the Iranian and Armenian kings from classical rulers, who were, at least de jure, elected magistrates to the very end of the Roman imperial tradition.

Hailed as Helios “the Sun” (Toumanoff, 1963, pp. 477-78), as his Sasanian counterpart was “the brother of the Sun and Moon” (Amm. Marc., 17.5.3), endowed with supernatural strength, like the elephant-bodied Rostam (Sh�hn�meh vii, vol. I, p. 278); Aa, secs. 42, 123, 767), the Armenian Arsacid king received his legitimacy and power from the Zoroastrian deity Vәrәθragna (Arm. Vahagn) the giver of victory and the companion of Mithra whom he invoked directly: “May… valour come [to you] from valiant Vahagn” (Aa, sec. 127). Through this divine mandate, the Armenian kings, even after their conversion to Christianity, as well as the Iranian rulers, were endowed with “valor” (kha�uthiwn), “good fortune” (baxt), and especially the “transcendental glory” (Mid. Pers. xwarrah, Arm. phar�kh) which rendered the Iranian ruler legitimate and invincible, as is evidenced in the K�rn�mag � Arda��r (ed. Antia, 5.8-14, 12.4). This “kingly glory” could manifest itself even in the king’s absence or after his death and abandoned him only for his sins, as in the case of Yazdgird I “the Sinner” (Sh�hn�meh xxxiv, vol. V, pp. 415-19; cf. Zamy�d Ya�t [Yt. 19] 34). These purely Iranian concepts were so deeply ingrained in Armenian tradition that they were still familiar and understandable to the Christian author of the Epic Histories, who clearly shared these beliefs, although he brands them as “heathen”:

[The Persians] opened the tombs of the former kings of Armenia, of the most valiant [kha�] Ar�akuni and they carried off into captivity the bones of the kings … For they said, according to their heathen beliefs: “This is the reason that we are taking the bones of the Armenian kings to our realm: that the glory [phar�kh] and the fortune [baxt] and valor [kha�uthiwn] of this realm might go from here with the bones of the kings and enter into our realm. (BP, 4.24)

Equally familiar to him was the concept of the protection given by the “kingly glory,” which could manifest itself even in the absence of the ruler, as is evident from his citation of what may have been a lost paean to the Armenian, king Ar�ak:

Shapuh king of Persia… wondered at the valor [kha�uthiwn] the fighting contingents displayed before him and said: “I marvel… at the steadfast devotion of the Armenian contingent in its love for its lord. For so many years have passed since Ar�ak their lord has been lost to them, and yet, they were inspired by him in battle. And whenever they struck down their foe, they ever cried out: ‘To Ar�ak!’, and yet he was not among them. But because of the devotion they bore to their own true lord, they dedicated to him every foe that they slew … And so many years have passed since Ar�ak their lord has been lost to them, for he lies in the castle of Andme� [Ir. Andm�] in the land of Xu�astan, yet they, in their piety, believed that he stood at their head as their king, that he stood with them in the midst of the host [gund], at the head of the fray, and that they performed their service to him in his very presence.” (BP, 5.5)

This passage is all the more constant with epic Iranian beliefs in that it invokes the “glory” of the absent, captive king Ar�ak and not that of his son the reigning king Pap, whose devotion from birth to the power of the devs made him unworthy of this divine attribute.

In this epic world, the hunt and the banquet rose above aristocratic diversion or social settings in which noble rank could most easily be displayed, to a transcendental level. The hunt was the setting par excellence of the hero in the Iranian world both real and legendary (Harper, 1978). The man on horseback was not merely a social superior. The horse of the evildoer stumbled at the critical moment. The first apocalyptic sign of the world turned upside down in the vision of the Zoroastrian J�m�sp N�mag was that “a horseman will become a man on foot, and the man on foot a horseman.” In his official representations on numerous Sasanian silver plates, the hunting king, on horseback, wearing all his supernatural attributes�the halo, the crescent moon set on his crown, and the symbolic floating ribbons of the xwarrah (Garso�an, 1997, p. 23, fig. 14)�unmistakably displayed the full majesty of his “glory.” For the hunt in the Iranian epic tradition was not merely the locus of the “transcendental glory,” it was simultaneously the setting for the heroic apotheosis of the royal hunter (Garso�an, 1997, p. 24, fig. 15). It is consequently of particular significance as a sign of the depth of penetration of Iranian culture into that of Armenia that we find the precise conjunction of (1) the iconography of the Sasanian hunting kings on their own silver cups and (2) the transcendental ideology underlying that iconography, in the Epic Histories’ description of the apotheosis, for his valor and nobility, of the Armenian hero Mushel Mamikonean:

In those days, Mushel possessed a charger, a white steed. And whenever Shapuh, king of Persia took [a cup] of wine in his hand to drink in the hour of festivity, as he entertained his forces, he would say: “Let the rider of the white horse drink!” And he ordered a cup decorated with the portrait of Mu�el with his white steed, and in the hour of festivity he placed the cup in front of himself and constantly he remembered, repeating the same words: “Let the rider of the white horse drink!” (BP, V.ii)

The same setting for the royal apotheosis rather than a scene of secular entertainment was provided by the banquet, where the Iranian ruler was also represented adorned with the crescented crown and undulating ribbons of the xwarrah (Ghirshman, 1962, p 218, fig. 259). Its eschatological implications as a prefiguration of the banquet of eternity appear in both the final banquet concluding Mithra’s terrestrial exploits and the heavenly vision of a golden throne dominating a banquet described in Kerd�r’s inscription at Sar Mashhad (KSM, p. 98, ll. 32-34). The two themes of the hunt and the banquet in which the king and the hero sheds his mortal form to reveal his supernatural attributes are constantly joined on Sasanian silverware and in the epic traditions of both Iran and Armenia. They are explicitly linked on the silver plate from Strelka (Garso�an, 1997, p. 13, fig. 2) and in the twin Parthian frescoes of the heroic hunt and the funerary banquet at Dura Europos (Ghirshman, 1962, pp. 49-50, fig. 62; p. 54, fig. 67). In the contemporary Armenian Epic Histories, all the crucial events of the king’s or the hero’s life and especially his death�suicide or murder�occur either on the hunt or at banquets, coinciding with the moment of apotheosis and thus, in the last case, raising the horror of the scene from the level of a crime to that of sacrilege (Garso�an, 1981, pp. 47-64). As in Iran, the two settings are commonly linked, so that the apotheosis of Mu�el “the rider of the white horse” takes place “in the hour of festivity.”

The Christian period

The event which unquestionably had the greatest effect in the separation of Armenia from the Iranian world to which it was so closely tied was the Christianization of Arsacid Greater Armenia at the beginning of the 4th century. Earliest Christianity had probably reached the southern regions of Armenia and of the semi-autonomous Satrapies of the south in the 3rd century and had come to them by way of Mesopotamia from Edessa and ultimately Antioch. The main current, however, which led to the conversion of the Arsacid monarchy and was ultimately to become the dominant one in the Armenian Church, reached the country from Cappadocia to the west and was associated with the mission of Saint Gregory the Illuminator, who was consecrated patriarch of Armenia by the bishop of Caesarea of Cappadocia, probably in 314 rather than the traditional 301. As a consequence, the Armenian State Church remained at first tied to Caesarea, where its patriarchs continued to be consecrated until the latter part of the 4th century, and its orientation was toward the Greek-speaking Christianity of the Byzantine empire, rather than to the Orient.

The Christianization of Armenia obviously separated it once and for all from the Zoroastrian world of which it had formerly been a part, even though its mythology had sunk so deep in the Armenian popular tradition that early Christian writers were apparently forced to alter Biblical stories in order to make their evangelizing mission comprehensible to their hearers (Garso�an, 1985); and the Armenian patriarch still shared officially the title of Zoroastrian mowbads: “Defender of the dispossessed” (Mid. Pers. drigow��n j�takgoww, Arm. j) as late as the second half of the 4th century (Garso�an, 1981, pp. 21-32). Nevertheless, Christian Armenia gradually drew back from the Persian spiritual tradition; and its opposition to the religion of the Sasanian empire was not limited to its official Zoroastrianism but soon progressed to include the Christian Church of the Orient, usually misnamed the Church of Persia, which was officially recognized by the Sasanian state at the Council of Seleucia-Ctesiphon in 410 (SO, pp. 253-75).

We have no evidence that the Armenian Church was ever considered part of the Church of the Orient, although the Persian katholikos was styled on occasion: “Father, head and director of all the bishops of the East” entrusted with the spiritual and canonical direction and ordinations “for every country and every city of the entire territory of the Persian empire, the rest of the East, and the neighboring countries (SO, pp. 285-87, 319-20), which should naturally have included Persarmenia. Greater Armenia was not mentioned in the hierarchical list of the Church of the Orient promulgated in the twenty-first canon of the council of 410 (SO, p. 272). No Armenian bishop ever attended any of the Persian councils, although the bishops of some of the Satrapies bordering on Mesopotamia were listed at the council of 410 as suffragans of the Persian metropolitan sees of Nisibis and Arbela, and their titulars usually, but not always, participated in the Persian councils.

This administrative separation of the two Churches may at first have resulted from the Armenian orientation toward Caesarea of Cappadocia; but by 410, when the Church of the Orient emerged as a religio licita in the Sasanian empire, the Armenian Church was no longer dependent on Caesarea and considered itself autocephalous. Part of this alienation may have been due to a fundamental difference in ecclesiastical structure. Organized under the direction of the Byzantine bishop of Marutha of Martyropolis/Miyafarqin, the Church of the East shared the western practice of geographical sees identified with a particular city and sharing the hierarchical status of that city, in a pattern prefiguring the one promulgated more than one generation later in the twenty-eighth canon of the Council of Chalcedon of 451. Following the secular hierarchy of the state, the bishop of the capital of Seleucia-Ctesiphon became automatically the head of the Church of the East; he and the other bishops were chosen and their status ratified by the ruler. The anachronistic, Parthian, non-urban, and decentralized pattern of Armenian society could not be adapted to this type of hierarchy, which ran counter to its traditions and would have fostered an unacceptable control by the central authorities. Until the Arab conquest, the early Armenian ecclesiastical hierarchy remained inextricably enmeshed in the para-feudal nexus of its society, and Armenian bishops were wholly identified with the families of the magnates of which they were the members and ecclesiastical representatives, just as the tanut�r was their military leader and administrator. Like any other hereditary office, the patriarchate passed from father to son in the house of Saint Gregory the Illuminator until the death of its last direct male descendant in 438, and no other candidate could be considered as long as a member of the Gregorid house was available (BP, 3.13, 15, 17,19). As late as the seventh century, an Armenian prince could still be cited as referring to “the bishop of his house” (Ps. Seb., 23), and surviving conciliar lists show that these family bishops invariably signed conciliar decisions in the name of their clan (Garso�an, 1999, pp. 439, 476-77).

Even more fundamental was the doctrinal opposition which developed between the two churches. At first, the Armenian Church, especially in its southern portion still strongly under the influence of Antiochene Christianity, as was the Church of the Orient, seems to have shared its Christology, which tended to divide the divine and human natures in the person of the incarnate Logos. Although the Sasanian Church separated itself from Antioch, as well as from all the “Western fathers,” at its council of 424, it maintained and eventually adopted officially the dyophysite Christology of the Antiochene School which became identified with the doctrine of bishop Theodore of Mopsuestia. As the northern hellenophile party came to dominate the Armenian Church and the Ecumenical Council of Ephesus condemned dyophysite Christology as heretical in 431, the Armenians accepted the decision of Ephesus and formally anathematized the dyophysites at its councils of 506 and 555, thus creating an impassable breach between itself and the Church of the Orient.

Bitter as was the separation between the Armenian and Sasanian Churches, however, it did not extend into the secular sphere to the Sasanian state itself. As a resident on Persian territory, even after the Byzantine frontier was shifted eastward in 591 by Xusrow II’s cession of most of his Armenian lands to the Byzantine empire, the Armenian patriarch, as a subject of the king of kings, was in no position to disregard his will. During the entire period of the Marzapanate, even after the grant of religious autonomy to Armenia in 485, the Armenian Church recognized Sasanian secular jurisdiction even in ecclesiastical matters. The invariable dating of the Acts of every Armenian council by the regnal year of the current Persian monarch demonstrates ipso facto the recognition of his ultimate sovereign rights (Garso�an, 1999, pp. 55-57, 412, 514).

More immediately, the organizing Council of Seleucia-Ctesiphon of 410 had granted to the Persian king, even though he was not a Christian, the same rights as the Council of Nicea had conceded to the Orthodox emperor in 325, among them, that of ratifying all episcopal elections. This right apparently extended to Armenia, since we are told that Sh�p�r II was angered when the Armenian king, Xosrov III/IV, appointed as patriarch Saint Sahak the Great in 387 without the authorization of the Sasanian ruler (MX, 3.1), even though according to Armenian tradition, Sahak, the last descendant of the house of Saint Gregory the Illuminator, was the only person then entitled to this office, which was hereditary in his family. Going further, at the time that he put an end to the Armenian Arsacid dynasty in 428, Bahr�m V simultaneously deposed Saint Sahak from most of the functions of his office, exiled him to his own estates, and replace him with an Armenian, followed by two Syrians. The Armenians apparently accepted this high-handed interference in their ecclesiastical affairs, since the native sources, while giving unedifying accounts of the morals of the Syrian antipatriarchs, never question the Persian king’s right to appoint them (MX, 3.64, 66). More particularly, after Sahak’s death, the same sources never refer to his successor Joseph, chosen without the ratification of the Persian court, as katholikos, but only as “a certain priest” (MX, 3.67), who, “although he was by ordination [only] a priest, yet at the time held the throne of the katholikate of Armenia” (�Ph, 1.23, 43)�this despite the fact that they do not hesitate to refer to Joseph as “holy” or “saint.” More than a century after the grant of religious autonomy to Armenia in 485, king Xusrow II had his Armenian favorite, prince Smbat Bagratuni, summon the council which elected the new katholikos Abraham I in 607 after a three years’ hiatus following the death of his predecessor. Despite the total doctrinal break between their two Churches, Sasanian secular justification over the Armenian Church remained unchallenged to the end.

This Sasanian secular control over the Armenian Church was not necessarily damaging to the latter, since it simultaneously protected it from Byzantine doctrinal pressures especially during the reign of Maurice (582-602) and his immediate successor Phokas. So great was the favor shown at that time by Xusrow II to his Armenian subjects, that native Armenian sources went so far as to claim unrealistically that the shah had converted to Christiantiy (Ps. Seb., 151). The change of religious policy which manifested itself in Xusrow II’s reign, as he shifted his goodwill to his monophysite subjects from the usual Sasanian recognition of the State Church with its opposing dyophysite doctrine, was probably not due to the influence of his favorite wife, the Christian Shirin, as has sometimes been suggested. Nor is it likely to have been instigated by his favorite, Smbat Bagratuni. The final official doctrinal break of Armenia with the Byzantine Church at the time provides a far more probable cause, since it ended the perennial Sasanian fear that their Christian subjects might betray them to “Caesar their coreligionist” and transformed Armenia from an untrustworthy region strategically located on the border between the Iranian and Byzantine empires into a loyal supporter (Garso�an, 1999, pp. 20, 383-84). Whatever the reason, Armenia and particularly its Church, weakened by the hostile policy of Maurice seeking to force it into doctrinal union with Constantinople, benefited from the Persian king’s benevolence and protection, so that Smbat Bagratuni was not the only one favored. At Xusrow II’s order, Smbat restored the authority and prestige of the Armenian administrative capital of Duin, overriding the objections of the local Persian marzpan and the military commander of the city (Ps. Seb., 27). Abraham’s successor, Komitas, would similarly reconstruct the martyria of Armenia’s early female saints in the holy city of Valar�apat. The position of the Armenian Church was further reinforced by the Persian capture of the great imperial fortress of Theodosiopolis-Karin and the exile of the anti-patriarch John of Bagaran installed by Maurice, thus ending its twenty years’ internal religious schism. Pressure exerted over the heads of the neighboring Churches of Siwnikh and Caucasian Albania forced them to recognize once more the authority of the Armenian katholikos, from whom they had withdrawn during the period of the schism (Garso�an, 1999, pp. 374-78). The strengthening of the Armenian Church under Sasanian auspices during the early part of the 7th century undoubtedly helped it resist the repeated subsequent attempts of the Byzantine emperors to impose a dogmatic union upon it. The enormous building activity which covered the plateau with churches in this period provides still visible evidence of the prosperity of Armenia and its Church under the Marzapanate in the last years of the Sasanian dynasty. At the same time, the modest dimensions and scattered locations of the majority of these churches identify them as the palatine chapels of local dynasts and give additional proof that the centrifugal pattern of Armenian society had not been obliterated by centralizing developments during the Sasanian period.

In the case of the Church, then, as in the other aspects of its society, the relations of the Armenians to Iran were by no means altogether negative, despite the one-sided, invariably hostile image given by the native Christian sources. At first an integral part of the Achaemenid empire, Armenia’s position was radically altered from the time that Hellenism began to spread through the East in the wake of Alexander’s conquests and all the more with the Roman domination of the Mediterranean world. Armenia now tended to be politically ambivalent between the two world powers of Rome, then Byzantium, and Iran � Parthian or Sasanian, albeit its outward forms, customs, and institutions remained throughout almost exclusively eastern rather than classical. The partition of the Armenian Arsacid kingdom at the end of the 4th century CE, put most of Greater Armenia once again within the Iranian empire, though Armenia’s ambivalence unquestionably grew as its conversion to Christianity transformed its self-image and turned it toward the west. Thereafter, its fervent adoption of the new faith pulled it in the opposite direction from its social structure and former traditions. However, for all of its unquestionable, ultimate dedication to Christianity, Iranian social forms and especially Iranian ideology had sunk so deeply into the substratum of its institutions and beliefs that they long outlived its conversion. Politically and culturally, if not religiously, as well as physically linked to Iran, pre-Islamic Armenia continued to lie significantly beyond the eastern limit that Augustus had presciently set for the classical world.


  • Sources. All references to sources are to the translations. All classical sources are cited according to the Loeb Classical Library. Aa: Agathangelos, History of the Armenians, tr. and comm. R.W. Thomson, Albany, 1976. BP: The Epic Histories attributed to Phawstos Buzand (Buzandaran Patmuthiwnkh), tr. and comm. N. G. Garso�an, Cambridge, Mass., 1989. J�masp n�mag: “To the �amasp N�mak,” ed. and tr. H. W. Bailey, BSOS 6, 1930-32, pp. 55-85, 581-600. KKZ, KSM: Les quatre inscriptions du mage Kird�r (Studia Iranica, cahier 9), tr. and comm. Ph. Gignoux, Paris, 1991. �Ph: The History of �azar Pharpechi, ed. and comm. R.W. Thomson, Atlanta, 1991. Letter of Tansar: tr. M. Boyce, Rome, 1968. MD: The History of the Caucasian Albanians by Movs�s Dascuranc�i, tr. C. J. F. Dowsett, London and New York, 1961. MX: Moses Khorenatshi, History of the Armenians, ed. and tr. R.W. Thomson, Cambridge, Mass. 1978. Paikuli: The Sasanian Inscriptions of Paikuli, ed. and tr. H. Humbach and P. O. Skj�rv�, III/1, Wiesbaden, 1983. Ps. Seb.: The Armenian History attributed to Seb�os, tr. and comm. R.W. Thomson and J. Howard-Johnson, 2 vols., Liverpool, 1999. Sh�hn�meh: Le �āh Nameh ou le Livre des Rois par Abu’l Kasim Firdousi, tr. par J. Mohl, 7 vols. Paris, 1838-78; repr., 1976. �KZ: Third Century Iran: Sapor and Kartir, ed. and tr. M. Sprengling, Chicago, 1953, pp. 14-20. SO = Synodicon Orientale ou Recueil des synodes nestoriens, ed. and tr. J-B. Chabot, Paris, 1902.
  • Literature. N. Adontz, Armenia in the Period of Justinian. Political Conditions Based on the Naxarar System, ed. and tr. by N. G. Garso�an, Louvain and Lisbon, 1970. E. Benveniste, “Titres iraniens en arme�nien,” REArm. 9, 1929, pp. 5-10. Idem, “Mots d’emprunt iraniens en arme�nien,” BSL 53, 1957-58, pp. 55-71. Idem, “Ele�ments parthes en arme�nien,” REArm., N.S. 1, 1964, pp. 1-39. Idem, Titres et noms propres en iranien ancien, Paris, 1966. S. Der Nersessian, The Armenians, London, 1969. N. G. Garso�an, “Prolegomena to a Study of the Iranian Elements in Arsacid Armenia,” Handes Amsorya 90, 1976, pp. 177-234; repr. in idem, Armenia between Byzantium and the Sasanians, London, 1985, sec. x. Idem, “The Iranian Substratum of the ‘Agat’angelos Cycle’,” in N. G. Garso�an et al., eds., East of Byzantium: Syria and Armenia in the Formative Period, Washington, D.C., 1980, pp. 151-89; repr. in eadem, Armenia between Byzantium and the Sasanians, sec. xii. Idem, “Sur le titre de ‘Protecteur des pauvres’,” REArm., N.S. 15, 1981, pp. 21-32; repr. in eadem, Armenia between Byzantium and the Sasanians, sec. vi. Idem, “The Locus of the Death of Kings, Armenia�the Inverted Image,” in R. Hovannisian, ed., The Armenian Image in History and Literature, Malibu, 1981, pp. 27-64; repr. in eadem, Armenia between Byzantium and the Sasanians, sec. xi. Idem, “Secular Jurisdiction over the Armenian Church (IV-VIIth Centuries),” in C. Mango and O. Pritsak, eds., Okeanos. Essays Presented to Ihor Shev�enko on his Sixtieth Birthday (= Harvard Ukrainian Studies 7), Cambridge, Mass., 1984, pp. 220-50; repr. in eadem, Armenia between Byzantium and the Sasanians, sec. ix. Idem, “The Early-Medieval Armenian City�An Alien Element?” in Ancient Studies in Memory of Elias J. Bickerman (= Journal of the Ancient Near Eastern Society 16-17, 1984-85), 1987, pp. 67-83; repr. in eadem, Church and Culture in Early Medieval Armenia, London, 1999, sec. vii. Idem, “‘Thagaworanist kayeankh kam banak arkhuni’: les re�sidences royales des Asacides arme�niens,” REArm., N.S. 21, 1988-89, pp. 151-69; repr. in eadem, Church and Culture, sec. viii. Idem, “L’art iranien comme te�moin de l’armement arme�nien sous les Arsacides,” in Atti del V simposio internazionle di arte armena, Venice, 1992, pp. 385-95; repr. in eadem, Church and Culture, sec. x. Idem, “Reality and Myth in Armenian History,” in The East and the Meaning of History, Rome, 1994, pp. 117-45; repr. in eadem, Church and Culture, sec. xii. Idem, “Notes pre�liminaires sur l’anthroponymie arme�nienne du Moyen-Age,” in M. Bourin et al., eds., L’anthroponymie document de l’histoire sociale des mondes me�diterrane�ens me�die�vaux, Rome, 1996, pp. 227-29; repr. in eadem, Church and Culture, sec. ix. Idem, “Les e�le�ments iraniens dans l’Arme�nie pale�ochre�tienne,” in N. G. Garso�an and J.-P. Mahe�, Des Parthes aux Califat. Quarte le�ons sur la formation de l’identite� arme�nienne, Paris, 1997, pp. 9-37. Idem, L’Eglise arme�nienne et le Grand Schisme d’Orient, CSCO 574/Subsidia 100, Louvain-la-Neuve, 1999. Idem, “Le ‘guerrier des seigneurs’,” Studia Iranica 33, 2003, pp. 177-84. Idem, “Frontier-Frontiers? Transcaucasia and Eastern Anatolia in the Pre-Islamic Period,” Atti dell’Accademia dei Lincei, Rome, in press.
  • R. Ghirshman, The Arts of Ancient Iran, New York, 1964. Idem, Persian Art, New York, 1962. G. Gnoli, The Idea of Iran, Rome, 1989. R. Gyselen, The Four Generals of the Sasanian Empire. Some Sigillographic Evidence, Rome, 2001a. Idem, “Le kadag-xwad�y sassanide. Quelques reflexions a� partir de nouvelles donne�e sigillographiques,” Studia Iranica 31/1, 2001, pp. 61-68. Idem, Nouvaux mate�riaux pour la ge�ographie historique de l’Empire sassanide (Studia Iranica, cahier 21), Paris, 2002. P. Harper, The Royal Hunter, New York, 1978. R. Hovannisian, ed., The Armenian People from Ancient to Modern Times, New York, 1997, I, pp. 19-115. D. M. Lang, The Georgians, London, 1966. H. Manandian, The Trade and Cities of Armenia in Relation to Ancient World Trade, Lisbon, 1965. A. G. Perikhanian, “Une inscription arame�enne du roi Artashes,” REArm., N.S. 3, 1966, pp. 17-29. Idem, Obshchestvo i pravo Irana v parfianski� i sasanidski� periody (The society and law of Iran in the Parthian and Sasanian periods), Erevan, 1983. J. R. Russell, Zoroastrianism in Armenia, Cambridge, Mass., 1987. Idem, “Armeno-Iranian Roots of Mithraism,” in J. R. Hinnells, ed., Studies in Mithraism, Rome, 1994, pp. 183-93. C. Toumanoff, Studies in Christian Caucasian History, Georgetown, 1963. Idem, “The Third-Century Armenian Arsacids. A Chronological and Genealogical Commentary,” REArm., N.S. 6, 1969, pp. 233-81. Idem, review of M.-L. Chaumont, Recherches sur l’histoire d’Arme�nie de l’ave�nement des Sassanides a� la conversion du royaume, REArm., N.S. 7, 1970, pp. 473-78.

Sasanid Architecture

By: Professor Dietrich Huff

1. Building materials

Sasanian architecture is characterized by the widespread use of mortar masonry and the associated vaulting techniques. Although mud brick had been developed long before, and mortar constructions were known in Parthian dynastic eras, both became pre-eminent in the high-standard architecture of the Sasanians. Mud brick remained a most important building material (e.g. Dāmghān, Istakhr/Estakkr, Haĵiābād, Kīš, Ctesiphon, Kūh-ī Khwāja), and only its impermanence shifts our attention to the better preserved stone and brick ruins of Sasanian architecture. Among these, rubble stone masonry with gypsum mortar is predominant. Brickwork was frequently used for vaults and domes, although there are a number of buildings made entirely of brick (e.g. Dastegerd, Ayvan-a Karka, Ctesiphon, Takt-a Solayman). Dressed ashlar appears sporadically, mainly in the early (e.g. B1’sapur, Firūzābād, Nurābād, Pāykūli) and late (e.g. Tāq-a Gerrā, Darband, Takt-a Solaymān, Kangāvar) phases of the empire, and seems to be due to western influence (H. Wulff, Traditional Crafts of Persia, Cambridge, Mass., 1966, p. 102).


2. Construction and structural types

(a) Vaulted constructions

Sasanian vaulting techniques depend largely on the special qualities of gypsum mortar, which allows vaulting without centering because of its short setting time. Barrel vaults with “pitched courses,” the most frequent system, owe their elliptical shape and their significant step out above the impost to this technical procedure, which requires only a back wall or a narrow strip of centering for the first courses, with the following ones successively glued in front (K. A. C. Creswell, Early Muslim Architecture 1/2, Oxford, 1969, p. 544; O. Reuther, “Sasanian Architecture,” in Survey of Persian Art I, p.498). Notwithstanding its practical advantages, vaulting without centering prevented the development of geometrically advanced constructions. Semicircular barrel vaults appear only when built on centering as a voussoir arch with “lying courses.” The cross vault, resulting from the intersection of two barrel vaults at right angles, was not developed. There are no examples of pointed arches built by formal intention, although they occur as a result of building practice in lesser monuments (e.g. Qasr-a Shīrīn) (G. L. Bell, Palace and Mosque at Ukhaidir, London, 1914, p. 5 1). The standard unit of the rectangular barrel-vaulted room was frequently enlarged by vaulted bays. Adjoining semi-domes occur rarely (e.g. Kīš, Bozpār, Negār, Sarvestān), although in vernacular architecture the use of the squinch vault, probably an ancient technique and one widely regarded as the origin of the Iranian dome, results in a hybridization of semi-dome or dome and cloister vault (A. Godard, “Voutes iraniennes,” Athar-e Iran 4, 1949, p.221). With the barrel-vaulted ayvān, a rectangular room with the front side open, the visible shape of the vault became the dominant feature of the facade. Already present in Parthian time, the ayvān became the most conspicuous element of Sasanian and later Iranian architecture.



(b) Domed constructions

The propagation of the dome on squinches above a square hall may be regarded as the most significant Sasanian contribution to the World architecture. This most uncomplicated and solid of all constructive systems already appears fully developed in the buildings of Ardašīr I in Firūzābād (Fig. 1). Its tectonic disposition remained basically unchanged throughout the Sasanian period and had a decisive impact on Islamic architecture; its empirical form clearly distinguished Eastern dome construction from the abstract geometrical concept of Western domes with pendentives (J. Rosinthal, Pendentifs, trompes et stalactites dans l’architecture orientale, Paris, 1928, p.43). The variety of squinch forms demonstrates an increasing effort to find satisfying forms for what was originally a purely constructive element. In its early stage (e.g. Fīrūzābād) the cupola proper does not yet have a perfectly circular base, but rises on a fairly well rounded octagon. Later examples (e.g. Qasr-e Shīrīn) draw nearer to geometric perfection, which is finally achieved in Islamic architecture.


  1. The elevation of the domed hall consists of three horizontal zones:
  2. plain walls, generally with doors or arches at the four axial intercepts;
  3. a zone of transition including the corner squinches and generally windows or decorative niches at the main axes;
  4. the cupola proper. The addition of barrel-vaulted bays to all four sides of the square produced the mature scheme that was to become a standard type for representative architecture in Iran until the present. This cruciform plan, based on the Chahār-tāq, the square with four arches, appears in the earliest examples of Sasanian architecture, (e.g. Takt-a Nešin in Fīrūzābād); it may have been inspired by Parthian architecture, although the central square was generally covered by cross or barrel vaults in those monuments.



(c) Columns and other supporting constructions

With the introduction of far-spanning vaults, the use of columns as constructive elements was widely discarded. There are examples of archaizing slender columns with bases, capitals, and sometimes fluted shafts that maintain Achaemenid traditions (e.g. Bišāpūr, Nurābād, Kīš), while those of later monuments (e.g. Bisotūn, Tāq-a Bostān) reflect a fresh Byzantine influence. But most often the column was transformed into a massive, round or rectangular pillar suitable for vaulted masonry constructions according to Iranian traditional architecture.


Apart from their use in colonnades (e.g. Kangāvar), pillars distinguish a characteristic group of generally three aisled halls covered by longitudinal or transversal barrel vaults (e.g. Čāl Tarkhān, Dāmghān, Ctesiphon, Takt-a Solaymān, Tepe Mil). Nonetheless the typical supporting elements remained the massive wall, and pillars more often appear as relics of a wall pierced by arches than as individual tectonic members.



(d) Constructive and decorative details

Clay remained the chief coating material for flat and vaulted roofs as well as for floors which were frequently covered with gypsum plaster, stone, or in rare cases, with foreign influenced mosaics (e.g. Bišāpūr). Plaster of Paris, frequently painted (Bišāpūr, Ayvān-a Karkha, Kīš), was widely used for building facings and for the dominant mode of architectural ornamentation, the stucco relief (Čāl Tarkhān, Dāmghān, Hajiābād, Kīš, Ctesiphon) (D. Thompson, Stucco from Chal Tarkhan, London, 1976; J. Kroger, Sasanidischer Stuckdekor, Mainz, 1982; M. Azarnoush, “Excavations at Hajidbad, 1977,” Iranica Antiqua 18, 1983, pp. 159ff.). The traditional stepped revetment remained a favorite decorative element, normally with four rectangular stages, which were already becoming dovetail-like at the late Sasanian Tāq-a Gerra.



3. Functional types of buildings

(a) Religious architecture

Frequent reference to sacred fires in Pahlavi texts indicate the important role that sanctuaries of the Zoroastrian state religion played in Sasanian architecture, but their architectural type remains disputed (F. Oehlmann “Persische Tempel,” Archäologischer Anzeiger, 1921, pp. 273ff.; U. Monneret de Villard, “The Fire Temples,” Bulletin of the American Institute for Persian Art and Archaeology 4,1936, pp. 175ff.; K. Schippmann, Die iranischen Feuerheiligtümer, Berlin, 1971; M. Boyce, “On the Zoroastrian Temple Cult of Fire,” JAGS 95, 1975, pp. 454ff.; Y. Yamamoto, “The Zoroastrian Temple Cult of Fire in Archaeology and Literature,” Orient 15, 1979, pp. 19ff.; 17, 1981, pp. 67ff.). The prevailing theory suggests that the main sanctuary structures were a freestanding Čahār-tāq, under which the sacred fire, shining through the four lateral arches, was exposed to worshipers during the religious services, and a small Āteāgāh some distance away, where the fire was kept at other times (A. Godard, “Les monuments du feu,” Athār-a Iran 3, 1938, pp. 7ff.; K. Erdmann, Das iranische Feuerheiligtüm, Leipzig, 1941, pp.46ff.). Apart from religious prescriptions that raise doubts about this kind of cult practice (Dārāb Hormazyār’s Rivāyat, ed. M. R. Unvala, 1, Bombay, 1906, pp. 60, 65ff.), archeological field work suggests another type of sanctuary: a closed chamber, where the fire was permanently maintained and served by priests, with adjoining ambulatories or rooms for worship (E. Herzfeld, Iran in the Ancient East, London, 1941, pp. 301ff.; E. Keall, “Archaeology and the Fire Temple,” in C. J. Adams, Iranian Civilization and Culture, Montreal, 1972, pp. 15ff.; D. Huff, “Das Imamzadeh Sayyid Husain and E. Herzfelds Theorie fiber den sasanidischen Feuertempel,” Stud. Ir. 11, 1982, pp. 197ff.). If the suggested identification of the Takht-e Nešīn in Fīrūzābād with a fire temple of Ardasīr I proves right, the early type was a square, domed room with four interior bays and with ayvans or rooms added to the four facades (Huff in Archäologischer Anzeiger, 1972, pp. 517ff.). A unique, semi-subterranean structure at Bīšāpūīr, convincingly attributed to Shāpūr I, is believed to be an ambulatory type fire temple because of its corridors; these surround a courtyard-like square of uncertain roofing, apparently associated with Anahita, as it was connected with an underground water canal (Ghirshman, RAA 12, 1938, p. 14; see, for a different interpretation, R. N. Frye, “The So-called Fire Temple of Bishapur,” in The Memorial Volume of the Vlth International Congress of Iranian Art and Archaeology, Oxford, September 11-16th, 1972, Tehran, 1976, p.93). The Sasanian phase of the mud brick structure at Kūh-e Khwāja, identified as a fire temple by an altar in its principal building, had a square, domed sanctuary surrounded by corridors and halls, with a vast complex of subsidiary rooms and ayvāns around a central court (Herzfeld, op. cit., pp. 291ff.; G. Gullini, Architettura iranica, Torino, 1964, pp. 87ff.). A similar layout was found at Takht-a Solaymān, tentatively dated to the 6th century, which has been identified, on the basis of historical tradition and the excavation of clay bullae bearing priests’ names and titles, as the shrine of Adhur Gūšnasp (Fig. 2), one of the three most important Adhur Wahrāms; the others, Adhur Farnbag and Adhur Burzēnmihr, have not yet been precisely localized. A second shrine excavated here, beside a dome-ambulatory temple, revealed an altar socle in a small sanctuary, preceded by two successive pillar halls rather than ambulatories (H. H. Von der Osten and R. Naumann, Takht-i Suleiman, Berlin, 1961; R. Nauman, “Takht-i Suleiman,” Archäeologischer Anzeiger, 1975, pp. 109ff.; idem, Die Rumen von Tacht-a Suleiman and Zendan-a Suleiman, Berlin, 1977, pp. 57ff.; D. Huff, “Takht-i Suleiman,” AMI 10, 1977, pp. 211ff.). The Čahār Qāpū at Qasr-e Šīrīn, attributed to Khosrow II, seems to have been another dome-ambulatory type temple within a large architectural compound (Bell, op. cit., pp. 51ff.; Reuther, op. cit., pp. 552ff.; differently J. Schmidt, ” Qasr-e Šīrīn,” Baghdader Mitteilungen 9, 1978, pp. 39ff.).


A great number of Čahār-tāq ruins, surveyed all over Iran and most frequent in Fars and Kerman, are regarded as fire temples. Nearly all of them were closed to the outside by blocking walls in their bays or the surrounding vaulted corridors (L. Vanden Berghe, “Récentes découvertes de monuments sassanides duns le Fars,” Iranica Antigua 1, 1961, pp. 163ff.; idem, “Nouvelle découverte de monuments du feu d’époque sassanide,” ibid., 5, 1965, pp. 128ff.; idem, “Les Chahar Taqs du Pusht-i Kuh, Luristan,” ibid., 12, 1977, pp. 175ff.). See further D. Huff, “Sasanian Čahar Taqs in Fars,” in Proceedings of the 3rd Annual Symposium on Archaeological Research in Iran, Tehran, 1975, pp. 243ff.). The two types are represented by the excavated examples at Tūrang Tepe identified as a sanctuary by an altar socle, and at Qal’a-ye Yazdegerd, respectively (J. Deshayes, “Un temple du feu d’époque islamique a Tureng Tepe,” in Gê feu dans le Proche Orient antique, Leiden, 1973, pp. 31ff.; E. Keall, “Qal`eh-i Yazdigird, an Overview of the Monumental Architecture,” Iran 20, 1982, pp. 51 ff.). Several open air altars including those at Naqš-a Rostam and Tang-a Karam most likely served for some Zoroastrian religious practice (A. Stein, “An Archaeological Tour in the Ancient Persis,” Iraq 3, 1936, pp. 175ff.; K. Erdmann, “Die Altare von Naqsh-i Rustam,” MDOG 81, 1949, pp. 6ff.; D. Stronach, “The Kuh-i Shahrak Fire Altar,” JNES 25, 1966, pp. 217ff.). Christian churches discovered at Hīra, Ctesiphon, and Rahalīya have long prayer halls, mostly with two rows of pillars and tripartite choirs (Reuther, Die Au.sgrabungen der Deutschen Ktesiphon-Expedition, Berlin, 1930, pp. llff.; D. Talbot Rice, “The Oxford Excavations at Hira, 193 1,” Antiquity 6, 1932, pp. 276ff.; B. Finster and J. Schmidt, “Sasanidische and frühislamische Ruinen im Iraq,” Baghdader Mitteilungen 8, 1976, pp. 27, 40ff.).



(b) Palaces

Although palaces provide the best known examples of Sasanian architecture, the number of well defined monuments is smaller than generally assumed. They are characterized by a regular layout along an axis of symmetry and an obligatory ayvān. The two palaces of Ardašīr I at Fīrūzābād, Qal’a-ye Doktar (Fig. 3) and Āteškada, both have as public reception areas a deep ayvān with lateral rooms, followed by a central dome and domed or barrel-vaulted subsidiary halls. A courtyard with ayvāns and large, uniform halls behind or in front of the reception area is generally regarded as the royal living quarters, although it gives the impression of belonging to the official area. Therefore the private lodgings may be assumed in small rooms on the upper floor that are otherwise unexplained (D. Huff, “Qal’a-ye Dukhtar bei Firuzabad,” AMI, N. F. 4, 1971, pp. 127ff.; idem, “Ausgrabungen auf Qal’a-ye Dukhtar bei Firuzabad, 1976,” AMI 11, 1978, pp. 117ff.).


There are few palaces remaining from the middle Sasanian period, during which the characteristic combination of ayvān and domed hall seems to have been abandoned. At the Taq-e Kesrā, now generally attributed to Khosrow I (Reuther, op. cit., pp. 15ff.; O. Kurz, “The Date of the Taq-e Kisrā,” JRAS, 1941, pp. 37ff.; differently Herzfeld, “Damascus: Studies in Architecture II,” Ars Islamica 10,1943, pp. 59ff.), and at the probably contemporary ayvān building at Takt-e Solaymān (Nauman, Die Ruinen von Tacht-a Suleiman, pp. 44), the ayvān appears to be the only dominating element. The inadequately documented `Emārat-e Khosrow in Qasr-a Šīrīin and the nearby ruin of Hawš Kūrī, both attributed to the time of Khosrow II, also seem to lack a dome behind the ayvān, where a transverse structure of uncertain elevation and a square courtyard were located instead (J. de Morgan, Mission scientifique en Perse IV, Paris, 1896, pp. 341ff.; Bell, op. cit., pp. 44ff.; Reuther in Survey of Persian Art I, pp. 533ff.). Regular house-like units added to the rear seem to have been living areas. Both palaces stand on artificial terraces with double ramps like the ruin at Kangāvar, now thought to be a late Sasanian palace (V. Lukonin, “The Temple of Anahita in Kangavar” [in Russian], VDI 2/140, 1977, pp. 105ff., cf. G. Herrmann, The Iranian Revival, Oxford, 1977, p. 107; M. Azarnoush, “Excavations at Kangavar,” AMI 14, 1981, pp. 69ff.). Other terraces such as Tall Dhahab and Haram-a Kesrā at Ctesiphon (Reuther, Ktesiphon-Expedition, pp. 23ff.; E. Kiihnel etal., Die Ausgrabungen der zweiten Ktesiphon-Expedition, Berlin, 1933, pp. Iff.) or Sarmaj (L. Trümpelmann, “Die Terrasse des Hosrow,” Archäologischer Anzeiger, 1968, pp. l lff.) may have carried palace-like super structures as well.


The residential function of a number of monuments generally regarded as palaces has been questioned. The ground plan of the well-preserved building of Sarvestān suggests other than palatial use. Its dating in the mid-Sasanian period has also come into question because of its highly developed vaulting system, closely paralleled by early Islamic constructions such as Qasr al-Kharāna in Jordan (O. Grabar, “Sarvistan. A Note on Sassanian Palaces,” in Forschungen zur Kunst Asiens. Festschrift K. Erdmann, Istanbul, 1968, pp. lff.; M. Siroux, “Le palms de Sarvistan et ses voutes,” Stud. Ir. 2, 1973, pp. 49ff.; L. Bier, The `Sasanian’ Palace near Sarvistan, New York, 1979.). The highly complex layout of the socalled palace of Shāpūr I in Bišāpūr raises similar questions of function (Ghirshman, “Les fouilles de Chapour (Iran),” RAA 12, 1938, pp. l5ff.; idem, Bichapour II, Paris, 1956, pp. llff.; Huff, Archäologischer Anzeiger, 1972, pp. 517ff.). The three-naved buildings of Dāmghān (F. Kimball, apud E. F. Schmidt, Excavations at Tepe Hissar, Philadelphia, 1937, pp. 327ff.), Čāl Tarkhān (Thompson, op. cit., pp. 3ff.), Tepe Mil (Kroger, op. cit., pp. 202ff.), and Kīš (P. R. S. Moorey, Kish Excavations 1923-33, Oxford, 1978, pp. 134ff.) can be reasonably regarded as forerunners of similar, early Islamic palaces such as Kūfa and Tall al-Okhayder but are formally connected with the second fire temple at Takht-a Solaymān and other cult buildings as well. There is little decisive evidence for the purpose of the hall on the city wall of ayvān-e Karkha (M. Dieulafoy, L’art antique de la Perse V, 1889, pp. 79ff.; Ghirshman, MDAFI, Paris, 1952, pp. lOff.) or the buildings at Bozpar (L. Vanden Berghe, “Le tombeau achéménide de Buzpar,” in Vorderasiatische Archäologie. Festschrift A. Moortgat, Berlin, 1964, pp.243ff.), Behešto Dozakh (L. Vanden Berghe, “Les ruines de Bihisht-a Duzakh a Sultanabad,” Iranica Antiqua 8, 1968, pp. 94ff.), and elsewhere.



(c) Cities and houses

The political importance of city foundations in Sasanian Iran is indicated by the almost obligatory component of the sponsor-king’s name in the name of the city. Although many attributions may concern some kind of re-founding or shifting of existing places, a number of original foundations are known, the standard pattern of which is a rectangular system of streets. The exceptional concentric and radiating plan of the circular city of Ardašīr-khorra may reflect an individual decision by Ardašīr I, demonstrating the cosmological and sociopolitical ideas of his emerging empire (D. Huff, “Zur Rekonstruktion des Turmes von Firuzabad,” Istanbuler Mitteilungen 19/20, 1969/70, pp. 319ff.; idem, “Der Takht-i Nishin in Firuzabad,” Archäologischer Anzeiger, 1972, pp. 517ff.; idem, AMI 11, 1978, pp. 117ff.). Archeological evidence for other circular geometric city plans is scanty, although they appear at different periods in the ancient Orient and with different stages of refinement. The round layout of Hatra, the best known Parthian example, lacks a genuine geometrical concept. It is unlikely that the round perimeter of Dārābgerd is a prototype for Ardašīr-khorra, as it probably dates from the 8th century (Creswell, Early Islamic Architecture I/2, 1969, p.21). The circular plan of Ctesiphon and the general topography of the site of al-Madā’en are still under discussion (Reuther, in Survey of Persian Art I, pp. 2ff.; J. M. Fiey, “Topography of al-Madā’in,” Sumer 23, 1967, pp. X.), and the reportedly round city of Sasanian Esfahān is not yet uncovered. Ardašīir-khorra may have influenced the layout of later circular cities such as al-Mansur’s Baghdad and its successors.


Few details are known about the architectural and sociological structure of orthogonal cities such as Jondīšāpūr (R. McC. Adams and D. Hansen, “Archaeological Reconnaissance and Soundings in Jundi Shapur,” Ars Orientalis 7, 1968, pp. 53ff.), Ayvān-e Karkha, and Bišāpūr, the last featuring a commemorative monument at the intersection of its two orthogonal main axes (Ghirshman, Bichdpour I, pp. 21 ff.; II, plan I). The majority of cities certainly continued older settlements with regular or organically grown patterns, as at Estakr (D. Whitcomb, “The City of Istakhr and the Marvdasht Plain,” In Akten des VII. internationalen Kongresses fur iranische Kunst and Archkologie, Munchen, 7.-10. September 1976, Berlin, 1979, pp.363ff.). Some residential areas have been surveyed or excavated in Kīš (S. Langdon, “Excavations at Kish and Barghutiat 1933,” Iraq 1, 1934, p. 113), Ctesiphon (Kuhnel, 2. Ktesiphon-Expedition, pp. lff.; R. Venco Ricciardi, “The Excavations at Choche,” Mesopotamia 3-4, 1968/69, p.57; idem, “Trial Trench at Tell Baruda,” Mesopotamia 12, 1977, pp. Ilff.), Lorestan (Morgan, op. cit., pp. 36111.), Roqbat al-Madā’en (Finster Schmidt, op. cit., pp. 151ff.) and Qasr-a Abū Nasr (W. Hauser and J. M. Upton, “The Persian Expedition 1933-34,” Bulletin of The Metropolitan Museum of Art 29, December 1934/11, pp. 3ff.), but the daily life of the middle and lower classes remains incompletely known.



(d) Fortifications

The main elements include ditches, walls with stepped niches, blind windows and arrow slots with horizontal or triangular covering, stepped battlements, corridors or narrow rooms within the walls, and far-protruding bastions, generally with semicircular headings. Unsophisticated gates were placed between pronounced bastions, and gate chambers were connected with the defence platform above by vertical shafts, probably for acoustic communication.


Few city ramparts have survived later changes. Ardašīr-khorra clearly had an earth wall with bastions, a ditch, and a small fore-wall. The ramparts of Bišāpūr were originally lined with semicircular bastions about 40 cm apart (`A. A. Sarfaraz, “Bišāpūr, the Great City of the Sasanians” [in Persian], Bāstān Chenāssi va Honar-e Iran 2, 1969, pp. 27ff.). The presumed palace section of the ramparts of Ayvān-a Karkha shows an elaborate arrangement of brick constructions (Ghirshman, MDAFI, 1952, pp. IOff.). The brick wall of Dastegerd, an unusual 16.6 m thick, harbored narrow corridors with radiating arrow slots and connecting semicircular tower chambers (F. Sarre and E. Herzfeld, Archkologische Reise im Euphrat-and Tigris-Gebiet II, Berlin, 1920, pp. 76; IV, pl. 127). The exceptional cut stone facing of the wall at Takt-a Solaymān (Osten Naumann, op. cit., p. 39) seems to be identical with that of the Darband walls (S. Han-Magomedov, Derbent, Moscow, 1979.). The standard Sasanian fortification type is represented by the mud brick ramparts of Ctesiphon and Estakhr (M. M. Negroponzi and M. C. Cavallero, “The Excavations at Choche,” Mesopotamia 2, 1967, pp.41ff.; Herzfeld, Iran in the Ancient East, pp.276ff.) and by the rubble stone walls of Qal’a-ye Dokhtar at Firūzābād (Huff, AMI 11, 1976, pp. 138ff.).


Most surviving fortresses served as isolated strongholds or protection for cities; this abundant but scarcely explored military architecture gives some insight into the Sasanian social hierarchy. Examples of the regular, generally square, Roman-type fort with rounded bastions are found in Harsin, Qasr-a Šīrīn (Morgan, op. cit., pp. 354ff.), Sirāf (D. Whitehouse, “Excavations at Siraf,” Iran 10, 1972, pp. 63ff.), and at several Mesopotamian sites (Finster-Schmidt, op. cit., pp. 49ff.). More frequent are irregular fortresses on strategically important heights; these usually have straight curtains between rounded bastions, as at Firūzābād, Bišāpūr, Tūrang Tepe (R. Boucharlat, “La forteresse sassanide de Tureng-Tepe,” in Collogues internationaux du C. N. R. S., No. 567: Le plateau iranien et I’Asie Centrale des origines a la conquête islamique, Paris, 1977, pp. 32911f), and the “Ātašgāh” at Esfahān (fig. 4 ), (M. Siroux, “`Atesh-gāh pres d’ Ispahan,” Iranica Antigua 5, 1965, pp. 39ff.). Territorial defence lines are known from literary tradition and archeological evidence (R. N. Frye, “The Sasanian System of Walls for Defence,” in M. Rosen-Ayalon, ed., Studies in Memory of Gaston Wiet, Jerusalem, 1977, pp. 7ff.), such as the ditch of Šāpūr II west of the Euphrates, the limes of Sistan (A. Stein Innermost Asia II, Oxford, 1928, pp. 972ff.), the walls of Darband from the Caspian into the Caucasus (A. A. Kudryavtsev, “O datirovke pervykh sasanidskikh ukreplenii v Derbente,” Sovetskaya Arkheologiya 3, 1978, pp. 243ff.), the wall of Tammisha (Tamīša) from the bay of Gorgān/Astarābād to the Elburz (A.D. H. Bivar and G. Féhervári, “The Walls of Temisha,” Iran 4, 1966, pp. 3511f), and the wall of Alexander north of the Gorgān river, although the last may date back to Parthian times (D. Huff, “Zur Datierung des Alexanderwalls,” Iranica Antigua 16, 1981, pp. 125fl.; M. Y. Kiani, Parthian Sites in Hyrkania, AMI, Erganzungsband 9, Berlin, 1982, pp. I Iff.).



(e) Funerary, commemorative, and rock architecture

The remarkable lack of monumental funeral architecture “maybe” explained by Zoroastrian religious prescriptions (Vd. 6.44ff.) [by the Western-Iranian-Zoroastrians] restricting burial rites to exposure of the dead and a possible but not necessary preservation of the bones in bone receptacles, or astodans (q.v.). Rock-cut exposure platforms and small cavities for preserving the bones are known mainly from southern Iran, notably around Estakhr and Bīšāpūr, where the huge grotto with the statue of Šāpūr I (fig. 5) is interpreted as his tomb (Vanden Berghe, Archéologie de l’Iran ancien, p.45; A. Stein, Old Routes of Western Iran, London, 1940, pp.311ff.; Ghirshman, Bichapour I, pp. 180ff.). Ritual texts describe astodāns as freestanding buildings, a type possibly represented by a bone burial in a fortification tower in Shahr-a Qūmes (J. Hausman and D. Stronach, “A Sasanian Repository at Shahr-i Qumis,” JRAS, 1970, pp. 142ff.) and by the tower of Nurābād (D. Huff, “Nurabad, Dum-i Mill,” AMI, N.F. 8, 1975, pp. 167ff.). Rock-cut tombs on the island of Khārg seem to belong at least partly to non-Zoroastrian communities (E. Haerinck, “Quelques monuments funéraires de file de Kharg dans le Golfe Persique,” Iranica Antiqua 11, 1975, pp. 134ff.) [or followers of the Eastern Iranian tradition of Zoroastrianism, which they were against the exposing their deaths to revered elements]


Some commemorative or triumphal monuments are identified by inscriptions. The twin-column monument in Bīšāpūr was dedicated to Shāpūr I (G. Salles and R. Ghirshman, “Chapour,” RAA 10, 1936, pp. 117ff.). The tower-like monument of Pākūi celebrates the victory of Narseh over his rivals (E. Herzfeld, Paikuli. Monument and Inscriptions of the Early History of the Sasanian Empire I-Il, Berlin, 1924). There is as yet no definitive explanation for [possibly] the late Sasanian Tāq-e Gerrā a small ayvān building with archivolt (H. V. Gall and W. Kleiss, “Entwicklung and Gestalt des Thrones im vorislamischen Iran,” AMI, N.F. 4, 1971, pp. 2ff.; S. Kambakhsh-Fard, “L’arc de Guirra, monument en pierre,” Traditions architecturales en Iran 4, 1976, pp. 2ff.), or for a freestanding gateway building outside the wall of Bīšāpūr (Sarfaraz, op. cit., pp. 27, 73). The tower in the center of Ardašīr-khorra (fig. ), which possibly carried a hall with the king’s seat or his fire, may symbolize God-given royalty (Huff, Istanbuler Mitteilungen 19-20, 1969/70, pp. 319ff.). The late Sasanian Tāq-a Bostān, an ayvān-like artificial grotto, is linked by its monumentality with official Sasanian architecture, and by its decoration with the tradition of Sasanian rock reliefs (E. Herzfeld, Am Tor von Asien, Berlin, 1920, pp. 57ff.; M. C. Mackintosh, “Taq-i Bustan and Byzantine Art,” Iranica Antigua 13, 1978, pp. 149ff.; S. Fukai et al., Taq-i Bustan I-IV, Tokyo, 1968-84). It may be related to other, partly unfinished rock monuments, such as those at Bisotūn (H. Luschey, “Bisotun, Geschichte and Forschungsgeschichte,” Archäologischer Anzeiger, 1974, pp. 114ff.; W. Salzmann, “Die Felsabarbeitung and Terrasse des Farhad in Bisotun,” ibid., 1976, pp. I l Off.) and Harsin (Godard, Athar-a Iran 3, 1938, pp. 67ff.).



(f) Civil engineering architecture

The centralized Sasanian government enabled the realization of large scale community projects such as road communications, bridges, irrigation, and drainage systems, most of which utilized the technical skill and manpower of prisoners of war. Many bridges (e.g. Kūzestān and Lorestān, Fīrūzābād, Bīšāpūr, and Bisotūn) dressed masonry with iron clamps at their preserved piers, which are generally rectangular with a triangular prism upstream; the arched superstructures are mostly destroyed (Stein, op. cit., pp. 15, 48, 71). Bridges were frequently constructed as weirs for irrigation and constituted the starting point of far reaching canal systems, as at Šūštar and Dezfūl (Dieulafoy, V, pp. lOSff.; G. Van Roggen, “Notices sur les anciens travaux hydrauliques en Susiane,” MDAF 17, 1905, pp. 167ff.; R. J. Wenke, “Imperial Investments and Agricultural Developments in Parthian and Sasanian Khuzistan: 150 B.C. to CE 640,” Mesopotamia 10/I1, 1975/76, pp.31ff). Aqueducts were carried on walls or bridges, and the use of syphon tunnels seems to have been known (Adams-Hansen, op. cit., pp. 59ff.).

Bibliography: Given in the text.