iii. Pre-Islamic History
Like other parts of Iran, the northwestern province of Azerbaijan can look back on a long history. For the earliest periods, however, archeological research has barely begun.
Before the Achaemenids. In 1949 C. Coon discovered a cave of the Paleolithic period at Tamtama, north of Urmia (Reżāʾīya) (Coon, Cave Exploration in Iran 1949, Philadelphia, 1951, pp. 15-20, 36-37, 44, 65). No indications of Paleolithic settlement, however, were found in the first larger-scale surveys in the west of the province made by R. R. B. Kearton (see Iran 7, 1969, pp. 186-87) and R. S. Solecki (ibid., pp. 189-90) in 1968. Opinions differed on the question why no Paleolithic sites had been discovered in this seemingly favorable region. D. Perkins, Jr. (ibid., p. 189) surmised that scarcity of “raw chipping material, such as flint,” in the region might be the explanation, but Solecki considered this improbable because flint was certainly available in the local mountains and spark-flints had been found at later Neolithic sites; in Solecki’s opinion, the lack of Paleolithic settlements was more likely to be due to ecological factors. Investigations by B. G. Campbell in the Tabrīz-Marāḡa-Mīāna triangle in 1974 and 1975 yielded evidence to the contrary for this area, as three caves and seven “open air localities” of the lower Paleolithic period were discovered (H. Sadek-Kooros, “Earliest Hominid Traces in Azerbaijan,” Iran 14, 1976, p. 154; idem, Proceedings of the 4th Annual Symposium on Archaeological Research in Iran 1975, Tehran, 1976, pp. 1ff.). The University of Pennsylvania’s large-scale expedition, called the “Ḥasanlū Project,” to the Soldūz valley under the leadership of R. B. Dyson, Jr., in 1965 produced evidence that the effective human occupation of this area began ca. 6000 B.C. (Bibliotheca Mesopotamica VIII, 1977; also L. Vanden Berghe, Bibliographie analytique de l’archéologie de l’Iran ancien, Leiden, 1979, and idem and E. Haerinek, Supplément l: 1978-80, Leiden, 1981). Other excavations in Azerbaijan, e.g., by C. Burney at Yanik Tepe and later at Haftavān and by A. Lippert at Kordlar Tepe, point to widespread settlement in Azerbaijan in the Neolithic and Chalcolithic periods.
For historical times, which in Azerbaijan begin roughly in the seventh century B.C., new archeological data have come to hand from the investigations of W. Kleiss, who excavated the big Urartian fortress of Besṭām (Basṭām) and from 1967 onward conducted systematic surveys in several parts of the province. Thanks mainly to Kleiss’s surveys, we now know that Azerbaijan, in particular the western region, was densely populated by Urartians—something that would not have been believed twenty years ago. In addition, Kleiss’s surveys brought to light many other large and small settlements of early historical times (see his successive articles in AMI, N.S., 1967 and later; and Vanden Berghe and Haerinck’s bibliographies). Also worthy of mention here is the fact that the long known rock chamber of Karaftū, twenty km west of Takāb, with its inscription “Hercules dwells here, let nothing evil enter,” has now been dated, in the light of its paleographic characteristics, from the late fourth or early third century B.C. (H. van Gall, AMI, N.S. 11, 1978, pp. 91ff.; P. Bernard, Studia Iranica 9, 1980, pp. 301ff.).
Achaemenid period. In the Achaemenid period Azerbaijan was part of the satrapy of Media. When the Achaemenid empire collapsed, Atropates, the Persian satrap of Media, made himself independent in the northwest of this region in 321 B.C. (thus H. H. Schmitt, Untersuchungen zur Geschichte Antiochos’ des Grossen und seiner Zeit, Wiesbaden, 1964, p. 61; in 328 according to V. Minorsky in EI2 I, p. 188, or 328-27 according to Kaerst, in Pauly-Wissowa, II, co1. 2150). Thereafter Greek and Latin writers named the territory Media Atropatene or, less frequently, Media Minor (e.g. Strabo 11.13.1; Justin 23.4.13). The Middle Persian form of the name was (early) Āturpātakān, (later) Ādurbādagān) whence the New Persian Āḏarbāyjān (on the name Atropatene and its derivation, see Minorsky, loc. cit.; Andreas, “Adarbigana,” in Pauly-Wissowa, I, cols. 345ff.; Weissbach, “Atropatene,” in Pauly-Wissowa, II, cols. 2149-50, and Streck, in Pauly-Wissowa, Suppl. I, cols. 223-24; Schwarz, Iran, repr. 1969, pp. 959ff.).
Atropates managed to keep on good terms with Alexander. At the famous mass wedding at Susa in 324, his daughter was married to Perdiccas (Arrian 6.4.5). After Alexander’s death he was left in command of his territory (Diodorus Siculus 18.3.3). He founded a dynasty which was to last long. The exact extent of the state of Media Minor or Media Atropatene is not known; in the opinion of Schwarz (op. cit., p. 61) it probably reached the Caspian, but how much of the coast it embraced is debatable.
Seleucid period. A successor of Atropates known to us from Greek sources is Artabazanes, who was contemporary with the Seleucid ruler Antiochus III (223 or 242-187). On the basis of a statement of Polybius (5.44.8 and 5.55.27) that Atropatene stretched to the Caucasus mountains, E. Herzfeld (AMI 4, 1931-32, p. 56) described Artabazanes as ruler of “Armenia and Atropatene.” Antiochus III, after his successful campaign against Molon, satrap of Media, who had rebelled, decided to march against Artabazanes with the intention to warn all concerned against supporting rebels with troops or arms (Polybius 5.55.1-2). Whether Artabazanes had in fact sent troops to help Molon is doubtful (Schmitt, op. cit., p. 124). Artabazanes, who was growing old, did not put up much resistance and appears to have acquiesced in submission to Seleucid suzerainty, in return for which he was probably confirmed in his rulership of Atropatene (Schmitt, op. cit., p. 149).
Parthian period. The exact date of Atropatene’s incorporation in the Parthian empire is not known. Most probably it occurred in the reign of Mithridates I (ca. 171-139/38 B.C.) when this Parthian great king, taking advantage of the Seleucid empire’s weakness after the defeat of Antiochus III by the Romans at Magnesia in 190 B.C., moved to extend his sway eastward and northward. Presumably Media Atropatene became a vassal state under Parthian suzerainty at the same time as the rest of Media. This must have been after 148 B.C. because the Seleucid rock-inscription at Bīsotūn (Behistun) shows that there was then still a Seleucid governor of the “Upper Satrapies,” which certainly included Media (K. Schippmann, Grundzüge der parthischen Geschichte, Darmstadt, 1980, p. 24). It seems, however, that the small state of Atropatene kept a good measure of autonomy. Descendants of Atropates are said to have “married into the (Arsacid) royal house” (Minorsky, in EI2 I, p. 188).
The next mention of Media Atropatene comes in reports that after the death of Mithridates II in 88-87 B.C., the Armenians succeeded in recovering lands which they had earlier lost to the Parthians. According to Strabo (11.14.15) and Plutarch (Lucullus 26), the Armenians occupied Atropatene at this time.
Atropatene’s history in the following years is confused. Dio Cassius (36.14) states that a certain Mithridates, king of Media and son-in-low of the Armenian king Tiridates, supported the latter when he went to war with the Romans and invaded Cappadocia in 67 B.C. Quite possibly this Mithridates was the future Parthian monarch Mithridates III, who together with his brother Orodes murdered their father Phraates III in 58-57 (Dio Cassius 39.56.2). He has been described, on the strength of Dio Cassius’s statement, as “king of Media Atropatene” in several works by modern scholars (e.g., A. Gutschmid, Geschichte Irans und seiner Nachbarländer, Tübingen, 1888, p. 98; H. Volkmann, in Der Kleine Pauly-Wissowa III, col. 1358; Geyer, with reservations in Pauly-Wissowa, XV, 2, col. 2207), though in fact he is called by Dio Cassius (36.14 and 39.56.2) simply “king of Media” (see also C. Le Rider, Suse sous les séleucides et les parthes, MADFI 38, 1965, p. 400; E. Herzfeld, AMI 4, 1932, p. 72).
In some sources (Appian, Mithridatica 106, 117; Diodorus Siculus 12.40.4), the Romans under Pompey are reported to have attacked a certain Darius, king of Media, in 65 B.C. Here again this person has been described as “ruler of Media Atropatene” by some modern writers (N. C. Debevoise, A Political History of Parthia, New York, 1938, p. 74; Gutschmid, op. cit., p. 98; less explicitly Herzfeld, art. cit., p. 56), whereas in the sources only Media is given. Acceptance of the supposition that he ruled Media Atropatene is also made difficult by the evidence of other sources (Monumentum Ancyranum VI, 11f.; see also Wilcken in Pauly-Wissowa, II, col. 1309) which speak of Artavasdes king of Atropatene, born in 59 B.C. or a little earlier, son of Ariobarzanes, king of Atropatene. This suggests that the father had come to the throne some time before 59 B.C. If so, the time-scale would appear to preclude a reign of this Darius in Media Atropatene. The Greco-Roman writers have left much more detailed and precise accounts of the expedition led by Mark Antony against the Parthians in 36 B.C. Having obtained the support of the Armenian king Artavasdes, Antony made Armenia his base for an invasion of Media Atropatene, whose identically named (and just mentioned) king Artavasdes was an ally of the Parthians. As is well known, the Roman campaign was bungled and ended ignominiously. After a Parthian attack which destroyed his rearguard and siege-train, Antony had to abandon his siege of Atropatene’s capital city Phraata (in some sources Praaspa or Phraaspa) and flee back to Armenia. It has not yet been possible to determine where Phraata lay; the often mooted identification with Taḵt-e Solaymān southeast of Lake Urmia where the German Archeological Institute conducted excavations in 1959 and subsequently, remains unproven (K. Schippmann, Die iranischen Feuerheiligtümer, New York and Berlin, 1971, pp. 309ff.; H. Bengtson, Zum Parther-Feldzug des Antonius, Munich, 1974, pp. 24ff.).
Soon after the defeat of the Romans, so Plutarch (Antonius 52-53) and Dio Cassius (49.33) state, enmity arose between Artavasdes of Media Atropatene and Phraates, the Parthian great king, over the division of the spoils and the fears of Artavasdes concerning his autonomy, with the result that the Median king offered Antony an alliance. The offer was accepted in 33 B.C. (K.-H. Ziegler, Die Beziehungen zwischen Rom und dem Partherreich, Wiesbaden, 1964, p. 36). It was very welcome to Antony who, in the belief that Artavasdes of Armenia had left him down in his campaign, now planned a pincer movement against Armenia while also cherishing hopes of Atropatenian support in his continuing war with the Parthians and impending contest with Octavian (for an assessment of the different motives, see Bengtson, op. cit., pp. 43-44). Troop detachments were exchanged and at the same time some Armenian territory, consisting mainly of the Sambyke district which had earlier belonged to Atropatene, was ceded to the Median ruler. To strengthen the bonds, a son of Antony was betrothed to Iotape, a daughter of Artavasdes. The alliance at first proved advantageous to Artavasdes of Atropatene, who with the help of the Roman reinforcements repulsed an offensive launched jointly by Artaxes, a son of Artavasdes of Armenia, and the Parthians.
These dealings indicate that not only Artavasdes, but also previous rulers of Media Atropatene, were more or less independent of the Parthian great kings. No doubt the geography of this relatively inaccessible mountain region facilitated the maintenance of its autonomy.
Artavasdes, however, could no longer hold out against the Parthians when Antony withdrew the Roman detachment from Media because he needed the men for his war with Octavian. In 30 B.C. Artavasdes was taken prisoner, but later he contrived to escape, probably as a result of the outbreak of civil war between Phraates IV and Tiridates, a rival claimant to the Parthian throne. He took refuge with Octavian, now Augustus, who gave him a friendly reception. He is reported to have died at Rome shortly before 20 B.C. (see Wilcken in Pauly-Wissowa, II, col. 1311).
Soon afterward, probably in 20 B.C., Augustus is said to have nominated Ariobarzanes II, the son of Artavasdes, to be king of Media Atropatene. At some later date, Ariobarzanes was appointed king of Armenia also. (Thus E. Meyer in Pauly-Wissowa, Suppl. I, col. 130; M. L. Chaumont in H. Temporini and W. Haase, eds., Aufstieg und Niedergang der römischen Welt, Berlin, 1976-81, II, Principat 9.1; J. G. L. Anderson in CAH X, pp. 264, 276. Between 20 B.C. and A.D. 2 in the opinion of Gutschmid, op. cit., p. 116. For a different interpretation, see U. Kahrstedt, Artabanos III und seine Erben, Bern, 1950, pp. 15-16). The actual induction of Ariobarzanes took place much later, namely in A.D. 9 following the accession of Vonones to the Parthian throne with Roman support (Ziegler, op. cit., p. 57 n. 81).
Ariobarzanes II was succeeded, on the thrones of both Media Atropatene and Armenia, by his son Artavasdes (Artavasdes II in the reckoning of Herzfeld, art. cit., p. 57; III in that of Chaumont, op. cit., p. 82, with many bibliographic references). Not long afterward, according to M. L. Chaumont in A.D. 19 or 20, this king was murdered. The event marks the virtual end of the rule of the dynasty founded by Atropates over Media Atropatene. It may have been consequent on the negotiation of the peace treaty of A.D. 18-19 between Germanicus, the Roman commander, and Artabanus II, the Parthian monarch since A.D. 10-11 (on whose background see below). Peace with Rome evidently gave Artabanus a free hand to deal with internal issues. Media Atropatene was one of a number of vassal kingdoms where the indigenous dynasts were eliminated and replaced with Arsacid younger sons (Kahrstedt, op. cit., p. 18; Ziegler, op. cit., p. 60 n. 104, basically agrees but points out that the sources for the treaty contain no word of any Roman promise of non-intervention in Media Atropatene). The later princes of the Atropatenian dynasty probably lived in exile in Italy. Two inscriptions bearing the name Artavasdes which were found in Rome are probably epitaphs of the son and grandson of an Atropatenian king Ariobarzanes, whether Ariobarzanes I or II being uncertain (Meyer in Pauly-Wissowa, Suppl. I, col. 130; Kahrstedt. op. cit., pp. 15, 17; Herzfeld, art. cit., p. 57).
It is necessary to comment here on the assertion, which has been frequently made (e.g., by Gutschmid, op. cit., p. 119; Herzfeld, art. cit., p. 74; Anderson in CAH X, p. 278) and is based on a passage in Josephus’s Antiquitates Judaicae (18.48), that Artabanos had been king of Media Atropatene before he became the great king of the Parthians. Kahrstedt (op. cit., pp. 11ff.) has found ample and convincing evidence that this is not so and that Artabanus probably stemmed from eastern Iran.
For the following period few events involving Atropatene are reported in the sources. Josephus (20.74) mentions that the first official act of the Parthian monarch Vologases I (A.D. ca. 51-ca. 76 or 80) was to appoint his brother Pacorus king of Media Atropatene (for a different interpretation, see R. Hanslik, in Pauly-Wissowa, IX, cols. 1839-40; Vonones II may have previously assigned the throne of Media Atropatene to Vologases, see Chaumont, op. cit., p. 97). When the Alans invaded Atropatene A.D. ca. 72, Pacorus had to flee into the trackless mountains (Debevoise, op. cit., p. 200; Chaumont, op. cit., p. 126). Another Alan invasion took place between A.D. 134 and 136.
Information about Atropatene (Azerbaijan) is then lacking until the last years of Parthian rule, when the conflict with Ardašīr, the founder of the Sasanian dynasty, had already begun. Artabanus IV, the last Parthian great king, was simultaneously engaged in a contest for the throne with his brother Vologases VI. His supporters were strongest in Media (where his coins appear to have been minted, probably at Ecbatana, the present-day Hamadān) and in Azerbaijan, Ḵūzestān, and Adiabene (G. Widengren, in La Persia nel Medioevo, Accademia Nazionale dei Lincei, Quaderno 160, Rome, 1971, pp. 711ff., esp. p. 741). Widengren has found evidence, however, that the common people of Media Atropatene were allies of Ardašīr (p. 749). In any case Azerbaijan submitted with little resistance to Ardašīr once he had defeated and killed Artabanus in 226 (the date preferred by Widengren, pp. 748-49). The well-known Sasanian rock relief at Salmās, not far from lake Urmia in which Ardašīr and others are depicted, is in Widengren’s opinion quite possibly a monument to this success (but see Chaumont, Recherches sur l’histoire d’Arménie de l’avènement des sassanides à la conversion du royaume, Paris, 1969, pp. 173ff.); the opinion of W. Hinz (Iranica Antiqua 5, 1965, p. 159) that it commemorates Ardašīr’s conquest of Armenia seems less well grounded.
Sasanian period. The next information given in the sources is that Šāpūr I, in the first year of his reign, i.e., 241-42, conducted two campaigns, first against the Khwarazmians then against the “Medes in the mountains,” which evidently means in Azerbaijan (Christensen, Iran Sass., p. 219). Thereafter Azerbaijan appears to have been pacified, because no more campaigns against its inhabitants are reported in the sources.
Atropatene/Āturpātakān, as the province appears to have been officially named throughout the Sasanian period (M. Streck, in EI1 I, p. 142), was governed on behalf of the Sasanian monarchs by a marzbān (margrave) who had all the authority of a satrap. It was a religious center, the principal temple being at Šīz, now Taḵt-e Solaymān. This was the hearth of Ādur Gušnasp, one of the empire’s three most sacred fires. The name Šīz often appears in linkage with other names, particularly Ganzaca (Ganzak) and Thebarmais, but the supposition that all refer to the same place is questionable (Schippmann, Feuerheiligtümer, pp. 341ff.; D. Huff, Comptes rendus de l’Académie des inscriptions et belles-lettres, 1978, pp. 774ff.). As a result of the existence of this great fire-temple, so revered that every newly crowned Sasanian king had to walk all the way to it on foot, and of the establishment of a royal palace in the province, Azerbaijan became a tightly integrated part of the empire instead of a loosely attached vassal state as in Parthian times. Herzfeld (art. cit., p. 57 n. 2) thought that “personal names incorporating gušnasp, the name of the sacred fire of Ganzak, were distinctively Atropatenian;” if so, this province produced many men of worth who held high office in the four centuries of Sasanian rule (Herzfeld, loc. cit.; Christensen, Iran Sass., pp. 518ff.).
Azerbaijan reenters the historical scene at the end of the Sasanian period. In A.D. 590 the decisive battle in the contest for the throne between the usurper Bahrām Čōbīn and Ḵosrow II was fought at Ganzak in Azerbaijan, ending in victory for Ḵosrow. In 628, on Easter day, the Byzantine emperor Heraclius captured Ganzak. Sasanian authority then began to collapse. Azerbaijan fell to the Arabs between A.D. 639 and 643 (Minorsky, in EI2 I, p. 190), and a new phase of its history began.
Bibliography : Given in the text.
Originally Published: December 15, 1987
Last Updated: August 18, 2011
This article is available in print.
Vol. III, Fasc. 2, pp. 221-224