Dating the beginning of the invasion
Estimates of the date of Shapur’s first Syrian invasion are mostly based on evidence relating to the fall of Antioch. The SKZ lists the cities captured by Shapur in this campaign, and they run from Anatha on the middle Euphrates to Antioch on the Orontes in a generally traceable order. Those cities listed as being captured after Antioch do not appear in an order that is easy to establish and it is probable that the campaign was actually a series of attacks over a number of years. The first of these was an assault directed up the Euphrates, culminating in the capture of Antioch and continuing with a series of smaller attacks on cities and regions in the vicinity of Antioch. Some more concentrated attacks on certain sections of Syria and Cappadocia then took place in the following three to four years.
An important source for dating the capture of Antioch and the beginning of the invasion is Zosimus. He claimed that at the same time as other events in the reign of Trebonianus Gallus, the Persians invaded Mesopotamia and Asia and advanced as far as Antioch.191 The reign of Trebonianus Gallus began in June 251, following the death of Decius in Moesia, and continued until July/August 253. Using Zosimus alone, however, it is not possible to be more precise regarding the time-frame of the Persian invasion.
Numismatic evidence from Antioch was exploited by Alfoldi to reach a more precise date for the invasion.192 Alfoldi identified a break in the coinage of Trebonianus Gallus from Antioch and concluded that no coins were minted there at all in the name of Aemilius Aemilianus (reigned c. August-October 253) or for Valerian in his first year.193 This interruption to the coinage was interpreted by Alfoldi as an indication of the Persian capture of Antioch. As the break in the coinage could be identified in the reign of Gallus, through the brief reign of Aemilianus and into the reign of Valerian, a date of 253 for the capture of Antioch and the beginning of the campaign was seen as more likely. Baldus challenges Alfoldi’s suggestion regarding the break in the Antiochene coinage claiming instead that it was the result of a reorganization of the Antioch mint identifiable as early as 245/246.194 On this analysis, the numismatic evidence may not be very relevant to dating the invasion.
More recently, a date of 252 for the invasion has been preferred by a number of scholars. Potter prefers the date of 252 for the beginning of the invasion on the basis of other textual evidence that was not considered by the numismatists.195 The Chronicle of Se’ert placed Shapur’s invasion in his eleventh regnal year and Tabari claimed that a Sasanian siege of Nisibis took place in Shapur’s eleventh year.196 Potter claims that the eleventh regnal year of Shapur was now known to be 252, but there is still some debate as to when Shapur’s reign officially began. Potter admits that there are chronological flaws in the Chronicle of Se’ert as it claimed that Shapur remained in Syria following the first invasion and then captured Valerian. The capture of Valerian clearly belongs to Shapur’s third campaign, which took place approximately eight years later. The Chronicle of Se’ert, therefore, demonstrates chronological flaws making it difficult to use as evidence for the exact year of Shapur’s invasion. Tabari’s description of a capture of Nisibis in 252 is reflected nowhere else in the literature or in the SKZ and seems not to fit with Shapur’s strategy.197 Potter also cites the Liber Caliphorum, an eighth-century Syriac account, which specifically places Shapur’s invasion in the 563rd Seleucid year (October 251 to September 252). There is also some archaeological evidence from Apamea, and possibly Dura Europos, that points to a date of 252 for the campaign.198 The date of the beginning of the first Persian invasion of Syria is ultimately a matter for debate and without the discovery of new evidence will continue. Potter’s preference for a date of 252 is perhaps the most attractive, but much of the textual evidence on which it is based is chronologically flawed meaning that a date of 253 cannot be ruled out.
The evidence of the SKZ - the capture of Anatha, Birtha Arupan and Birtha Asporakos
While ongoing Roman interference in Armenia was used to justify the invasion in the SKZ, the attack actually began on the lower Euphrates with the capture of Anatha. The Greek version of the inscription refers first to the defeat of a Roman army numbering 60,000 men at Barbalis-sos199 on the Euphrates before listing the individual cities captured as a part of the campaign in Syria and Cappadocia:
(10) The nation of Syria and whatever nations and plains that were above it, we set on fire and devastated and laid waste. And in that campaign <we took> (the following) fortresses and cities from the nation of the Romans: (11) the city of Anatha with its surrounding territory,200 Birtha Asporakos, Sura, Barbalissos, Hier-apolis, (12) Beroea, Chalcis, Apamea, Rephanea, Zeugma, Ourima,
(13) Gindaros, Larmenaza, Seleucia, Antioch, Cyrrhus, (14) another Seleucia, Alexandria (Alexandretta), Nikopolis, Sinzara (Larissa), Chamath (Epiphania), (15) Aristia (Arethusa), Dichor, Doliche, Doura, Circesium, Germanicia, (16) Batna, Chanar and in Cappadocia, Satala, Domana, (17) Artangil, Souisa, Suid, Phreata, a total of thirty-seven cities with their surrounding territories.
The movement of the Persian forces up the Euphrates to Antioch following the capture of Anatha has been described and debated ever since the discovery of the SKZ in the 1930s. Emphasis has been placed on the order in which the cities appear in the inscription and how this indicates the route taken by the Persian army as it marched through Syria. Considerable work has been done on reconstructing the invasion route and identifying the sites captured.201
Anatha was an island fortress belonging to the Persians at the time of Julian’s invasion in 363 but little is known of its significance as a Roman fortification in the mid-third century. It may have been controlled by Palmyra from as early as the first century AD before becoming part of the complex system of fortifications on the Euphrates under Septimius Severus. The Greek version of the inscription lists Birtha Asporakos as the next city
Figure 5.8 The Euphrates and east wall of the fortress of Birtha Asporakos/ Zenobia.
Captured, but in the Parthian version of the SKZ there are two fortifications known as Birtha (byrt) listed after Anatha. These are byrt ’rwpn (Birtha Arupan) and byrt ’spwrkn (Birtha Asporakos), the latter clearly the same location as that listed in the Greek version. Birtha derives from the Aramaic byrt’ meaning ‘fortress’.202 The various attempts made to identify the modern locations of these fortifications are discussed in Chapter 3.
The capture of Sura, followed by Barbalissos and the defeat of the Roman army there, opened up northern and southern Syria to the Persians and they succeeded in capturing many important Roman cities. The Persian army appears to have split into at least two divisions on the basis of the order in which cities were listed in the SKZ after Barbalissos. One ventured north while the other went south. The northern army captured Hierapolis and moved on to capture Zeugma, Ourima, Gindaros and Lar-menaz while the unit that moved south captured Beroea, Chalcis, Apamea and Rephanea.203 In this section of the inscription the order in which the Roman cities were listed does not follow precisely the order in which it might be expected that the cities were captured.204 With one force heading north and the other south it might be expected that the cities from Hier-apolis to Larmenaz would be listed in order, but Hierapolis appears on its own before the capture of the four southern cities from Beroea to
Figure 5.9 The cardo maximus at Apamea in Syria, one of the many cities captured by the Persians in the first Syrian campaign.
Rephanea. The four southern cities do appear in an order we would expect, but the order of the appearance of Cyrrhus on the inscription also falls outside an expected order as it is listed after Antioch on the Orontes.205 We might expect that the northern division of the Persian army that captured Hierapolis, Zeugma, Ourima, Gindaros and Larmenaz would have taken Cyrrhus before advancing to Antioch as Cyrrhus lay close to the road from Zeugma to Antioch. Potter suggests that Cyrrhus’ isolation in the SKZ could be explained by a column being detailed to besiege the city while the rest of the force made its way to Antioch.206 It is important to bear in mind that the order in which the cities fell to the Persians need not have been strictly adhered to in every minute detail when the inscription was set up over a decade later. The SKZ was primarily a document for public consumption in Iran, and publishing the precise order in which the cities of the Roman East were captured was not necessarily Shapur’s highest priority. Some cities would have fallen quickly while others probably held out for longer, possibly enduring sieges for some time. This probably had implications for the order in which cities were listed.207
Figure 5.10 A third century AD bridge on the road to Cyrrhus (above), and the partially reconstructed theatre at Cyrrhus (below), close to the modern border between Syria and Turkey.
Figure 5.11 The Ataturk dam near the ancient bridge crossing at Zeugma. Zeugma now lies largely underwater due to flooding of the dam. Photo: Ross Burns.
The Persian capture of Antioch
Following the division of the Persian army after Barbalissos and the subsequent capture of cities to the north and south, the attack on Antioch took place and its submission appears to have been one of Shapur’s main aims. Older scholarship held that Antioch fell twice to the Persians - once in the first Syrian campaign and again in the campaign in which Valerian was captured.208 It is now thought that the Antioch listed in the ‘third contest’ was a different Antioch.209 For a successful siege of Antioch both the southern and northern divisions of the army would have been required.210 The Persian capture of cities after Barbalissos eliminated their ability to support Antioch, and the defeat of the Roman army at Barbalissos must have had significant implications for the defence of Antioch. It is proposed that the columns that had captured cities to the north and south after Barbalissos met at the intersection of the roads from Apamea, Chalcis and Zeugma to the north-east of Antioch.211 The knowledge of Mariades and the support he still appears to have enjoyed at Antioch would also have been useful to Shapur at this
Point.212
Map 5.1 Western Syria in the third century AD showing, among other things, sites captured by Shapur I in the first Syrian campaign (Ross Burns).
The Persian capture of Antioch in 252/253 is the most significant event to have survived in the Graeco-Roman literary tradition on the Persian attacks of the 250s, apart from the capture of Valerian. The details are patchy, and it requires some imagination to reconstruct the events in chronological order, but it is possible to hypothesize on the basis of the scraps of remaining evidence. The aspects of the attack of which we know most are preserved in the writings of later residents of Antioch who had an intimate knowledge of the city and its history.
In the best-known story to emerge from the capture of Antioch, Ammi-anus Marcellinus, Libanius and Eunapius of Sardis referred to the surprise of theatregoers when Persian archers appeared on the hilltops and began firing arrows into the crowd.213 This evidence is dismissed by Potter as ‘useless’ and ‘inconceivable’ as the citizens of Antioch would have had ample warning of the Persian approach.214 The dismissal of such evidence as useless assumes much in the context of the complex sequence of events leading up to the siege of Antioch. It is true that the extent of the attacks on nearby cities and fortifications, together with Persian preparations for the attack on Antioch, including the encampment of the army only 20 stades away, would have provided the citizens of Antioch with ample
Figure 5.12 The imposing cliffs behind Antioch which rise to Mt Silpius.
Warning.215 The Anonymous Continuator of Cassius Dio stated, however, that a sizeable proportion of the population of Antioch was well disposed to Mariades and that they had not fled the city in the face of the Persian attack as they welcomed the overthrow of the Antiochene boule.216 It is conceivable that many of the citizens who stayed believed in vain that they would be spared due to their support for Mariades prior to his flight to Persia. It is also worth noting that Malalas claimed that Antioch was captured in the evening, which may explain how the citizens were caught by a surprise attack while attending the theatre.217 The heights around Antioch are extensive and the ancient city was located directly beneath them. One of the theatres at ancient Antioch was located directly below Mount Silpius and could have been attacked easily by archers shooting from the steep cliffs above.218
Antioch was the third largest city in the Roman Empire and the most important city in the Roman Near East. It was naturally a very well defended city as it was surrounded by steep mountains that could only be crossed by way of narrow passes. The defences of Antioch were significant in the third century as the Persian army used a large battering ram in the process of capturing it.219 For the residents of the city, leaving Antioch at this time was potentially more dangerous than staying. It has long been thought that Seleucia Pieria, the sea port of Antioch, was listed in the SKZ before Antioch, indicating that it was captured first so as to cut off the
Figure 5.13 Remains of what were the harbour-front walls at Seleucia-Pieria in antiquity. The modern shoreline now lies approximately 500 metres further west.
Escape route to the sea from Antioch.220 The identification of Seleucia is now challenged by Potter who prefers the identification of Seleucia with Seleucobelos on the road from Apamea to Antioch.221 He believes that the Seleucia listed after Antioch and Cyrrhus was Seleucia Pieria so on this interpretation it is possible that some sought to escape by sea through Seleucia. While the Antiochenes had undoubtedly received warning of the approaching Persian army some probably believed that Mariades would protect them, with others perhaps taking comfort in the inaccessible nature of the city and its strong defences.222 Prior to Antioch’s capture by the Persians it was virtually inconceivable that the city could fall as no serious threat to its security appears to have emerged at any stage in more than 300 years of Roman control.
Other surviving traces of the events surrounding Antioch’s capture allow further analysis of this important event. According to the Anonymous Continuator of Cassius Dio, Shapur and Mariades camped approximately 4 km from the city.223 Some citizens fled, but others who were supporters of Mariades stayed and stirred up rebellion and thus assisted Shapur in his capture of the city.224 Antioch was attacked initially in the evening by Persian archers and the next day by troops who had advanced along the road from Chalcis where the two army divisions had reunited, as discussed above.225 It was in this assault that the walls were probably
Figure 5.14 Mt Casius’ summit from the south-east. Mt Casius is the dominant geographical feature in the vicinity of Seleucia-Pieria.
Attacked using the battering ram referred to by Ammianus. Oracula Sibyl-lina XIII predicted that Antioch would be completely devastated and that it would be the prize of war.226 Malalas and Libanius stated that the city was set on fire, but Libanius noted that the Temple of Apollo at Daphne was spared.227 In another oration, Libanius declared that Antioch was so devastated by the attack of Shapur I that in the fourth century the city no longer had any noble buildings.228 This claim, and that of the oracle, were clearly rhetorical; however, the damage to the city must have been significant. Mariades had been useful to the Persians to this point.229 His knowledge of the region, particularly the environs of Antioch, would have been of assistance. Whatever remained of his support base at Antioch probably assisted in reducing resistance, but with the capture of Antioch the usefulness of Mariades to the Persians was at an end. Shapur seems not to have conceived that he owed Mariades anything and he was killed soon after.230
World History
