Vujcic ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 Article copyright held by the author(s) and made available under the Creative Commons Attribution License CC-BY https://creativecommons.org/licenses/by/4.0/ The City of Pelion and the Illyrian War of Alexander Nemanja Vujčić N THE SPRING OF 335 B.C. Alexander the Great launched a campaign that led him deep into the interior of the Balkans, up to and beyond the Danube. According to Arrian, our main source for these events, the primary reason for the ex- pedition was the safety of Macedonian borders in the context of the planned Asian campaign (Arr. Anab. 1.1.4):1 When the spring came, he moved into Thrace against the Triballi and the Illyrians, because he was informed that they were restless, and because it seemed unwise to leave these neighboring tribes unconquered, when he was about to leave his home for a campaign far away. We are not given sufficient details but the king was obviously informed that some kind of aggression was brewing. Arrian mentions two groups of enemies, though we are left with the distinct impression that the Triballi, a powerful tribe on the middle Danube, were the more formidable of the two and thus were treated as the primary target of the campaign. It began in early spring and already by harvest time the Triballi and their allies the Getae were defeated (Anab. 1.1–4). After making the necessary arrangements concerning the subjugated peoples, Alexander began the homeward march, at which point the 1 ἅµα δὲ τῷ ἦρι ἐλαύνειν ἐπὶ Θρᾴκης, ἐς Τριβαλλοὺς καὶ Ἰλλυριούς, ὅτι τε νεωτερίζειν ἐπύθετο Ἰλλυριούς τε καὶ Τριβαλλούς, καὶ ἅµα ὁµόρους ὄντας οὐκ ἐδόκει ὑπολείπεσθαι ὅτι µὴ πάντῃ ταπεινωθέντας οὕτω µακρὰν ἀπὸ τῆς οἰκείας στελλόµενον. N. G. L. Hammond and F. W. Walbank, A History of Macedonia III (Oxford 1988) 32, describe the campaign as “a pre-emptive strike.” I NEMANJA VUJČIĆ 497 ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 other threat suddenly manifested itself: three Illyrian tribes made war on him. Initially, it was reported that Cleitus, an Illyrian ruler, revolted, aided by Glaucias, the independent king of the Taulantii. The second news was that the tribe of Autariatae was preparing to ambush the main Macedonian force during the march. King Langarus of the Agriani offered to invade the land of Autariatae, thus preempting their attack, which he im- mediately did. Alexander did not wait for the resolution of this situation but instead moved southward with imposing speed. His goal was a city named Pelion, our main source being un- clear about its location. When he arrived, Cleitus already had taken the city, where he was briefly besieged by the Macedon- ians. On the very next day Glaucias appeared with his forces and took possession of the high ground surrounding the wooded valley in the center of which Pelion was situated. Alex- ander’s army was now essentially trapped, caught between a fortified city and the forces holding the heights. After an un- successful attempt at foraging, Alexander delivered the army from the dangerous situation through a combination of forced march and a bold maneuver. Three days later, after learning that the enemy had relaxed his guard and was recklessly camp- ing in open ground, Alexander quietly returned, crossed the river again and annihilated a multitude of surprised Illyrians. The rest saved themselves by fleeing to the country of Glaucias, burning Pelion in the process. Alexander did not pursue the vanquished monarchs for long, instead going south to deal with the Theban uprising (Anab. 1.5–6).2 2 On these events see N. G. L. Hammond, “Alexander's Campaign in Illyria,” JHS 94 (1974) 66–87; F. Papazoglu, The Central Balkan Tribes in Pre- Roman Times: Triballi, Autariatae, Dardanians, Scordisci and Moesians (Amsterdam 1978) 25–40; A. B. Bosworth, A Historical Commentary on Arrian’s History of Alexander I (Oxford 1980) 68–73, and “The Location of Alexander’s Cam- paign against the Illyrians in 335 B.C.,” in B. Barr-Sharrar et al. (eds.), 498 PELION AND THE ILLYRIAN WAR OF ALEXANDER ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 These, briefly, are the main facts. The sole source that provides them is the account of Arrian. Other sources on Alexander either ignore the Balkan campaigns altogether or mention them in passing.3 Arrian’s description is the product of a drastic condensation of the memoirs of Ptolemy and Aristo- bulus. In this particular case, it is far from perfect and not deserving of the praise it sometimes receives.4 Its geographical context is rather unclear and the continuity of the narrative is fragmented. Essential information is missing, including the very identity of the main attacker. Identity of the attackers According to Arrian, the Macedonian country and army were threatened by three groups. One is that of Cleitus, who is ___ Macedonia and Greece in Late Classical and Early Hellenistic Times (Washington 1982) 75–84; Hammond and Walbank, A History of Macedonia III 39–48; P. Cabanes, Les Illyriens de Bardylis à Genthios (Paris 1988) 132–134; A. B. Bosworth, in CAH VI2 (1994) 795–796; J. Wilkes, The Illyrians (Oxford 1996) 122–124; N. G. L. Hammond, Alexander the Great: King, Commander and States- man3 (London 1996) 45–58; K. Nawotka, Alexander the Great (Newcastle 2010) 98–100. 3 The Illyrian campaign is mentioned by Chron.Oxyrh. (FGrHist 255 T 6); Marm.Par. (FGrHist 239 B 2); Diod. 17.8.1. The first two books of Curtius’ History of Alexander are lost, but he alludes to the conquest of the Triballi and the Illyrians in later books (6.3.2, 9.6.20). Strab. 7.3.8 does not mention the Illyrian campaign, while Plut. Alex. 11.3 only makes clear that Alexander was somewhere in Illyria (the campaign is more directly referred to in De Alex. fort. 342C). Likewise, Just. Epit. 9.2.4–10 is silent on the Illyrian war, and even the conquest of the Triballi receives but a faint allusion; he generally fails to distinguish between the events of 336 and those of 335. A fragment of a narrative history (or a commentary thereon), almost certainly describing parts of Alexander’s campaign in the Balkans, is in a 2nd cent. B.C. papyrus: W. Clarysse and G. Schepens, “A Ptolemaic Fragment of an Alexander History,” ChrÉg 60 (1985) 30–47; except for one important de- tail, discussed below, this badly damaged text is of limited use as a historical source. 4 Cf. Hammond and Walbank, A History of Macedonia III 40; Hammond, Alexander 48–49. NEMANJA VUJČIĆ 499 ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 introduced merely as an Illyrian leader, without specifying the ethnos over which he ruled. The Autariatae are mentioned as a tribe without a king. Only for the third group are we given full information: king Glaucias and the Taulantii, who lived in the area of the modern central Albania and who, unlike the Illyrians of Cleitus, had not accepted Macedonian rule. The Taulantii reached the battlefield from the northwest, a day after the Macedonian army and several days after the troops of Cleitus. In contrast to Cleitus, Glaucias and his people are en- countered later in various episodes of Greek history.5 The Autariatae are mentioned only as an additional threat. We are told that they intended to ambush Alexander’s army on the march (Anab. 1.5.1, οἱ δὲ καὶ τοὺς Αὐταριάτας ἐπιθήσεσθαι αὐτῷ κατὰ τὴν πορείαν ἐξήγγελλον). It is not explicitly stated that the Autariatae and the southern Illyrians were making a joint effort against the Macedonians, though this is usually assumed by historians.6 The entire episode is given in such a circum- scribed manner that it feels as if something is missing. The ambush never came to pass and Alexander moved on un- opposed, with impressive speed. A Macedonian ally, Langarus, king of the Agriani, offered to attack the Autariatae, whom he saw as weak warriors (he describes them as “the least warlike,” ἀπολεµωτάτους), while they were still in their own territory and thus remove the threat (1.5.3).7 With Alexander’s permission 5 Glaucias sheltered the infant Pyrrhus and his supporters during their years of exile (Plut. Pyrrh. 3.1–3); later Pyrrhus maintained cordial relations with the Illyrian king (4.1). Glaucias fought against Cassander (Diod. 19.67.6–7), besieged Apollonia (19.70.7), was given Epidamnus by the Cor- cyraeans (19.78.1), etc. 6 So I. Worthington, Alexander the Great: Man and God (London 2004) 41; Nawotka, Alexander 99; P. Green, Alexander of Macedon 356-323 B.C. A Historical Biography (Berkeley 2013) 131; I. Worthington, By the Spear: Philip II, Alexander the Great, and the Rise and Fall of the Macedonian Empire (Oxford 2014) 129; etc. 7 The statement about the lack of martial prowess of the Autariatae is 500 PELION AND THE ILLYRIAN WAR OF ALEXANDER ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 this was done, the enemy were taken by surprise and their land plundered; Langarus returned to his kingdom with rich spoils. At the moment of Langarus’ raid the Autariatae still did not move to attack, which raises the question whether this was really their intention. How did the messengers even know of plans of this distant tribe which apparently had not even begun to assemble its forces? There is another and simpler way to interpret this episode: Langarus asked for permission to raid and pillage the Autariatae, and this was granted by Alexander, who saw him as a good and dependable ally. If this was the case, the preemptive strike was introduced into (probably Ptol- emy’s) text as a justification for the raid. For a proper under- standing of the events of 335 it would be very useful to know which territories were held by the Autariatae at the time. However, ancient writers do not provide clear information on this. Preserved references are mostly late, often unclear, and sometimes contradictory.8 As for Cleitus, all that we are told is that he led an uprising (which clearly implies that he was a ruler of some Illyrian tribe), and that he was a son of a Bardylis. This probably means that he was a son or grandson of the famous king Bardylis who ___ contradicted by the picture of an expanding tribe of conquerors, which is how Strabo describes them (7.5.11): Αὐταριᾶται µὲν οὖν τὸ µέγιστον καὶ ἄριστον τῶν Ἰλλυριῶν ἔθνος ὑπῆρξεν. 8 Several sources state that the Autariatae where neighbors of the Ardiaei ([Arist.] Mir.ausc. 138; Strab. 7.5.11; App. Ill. 3; cf. Ps.-Scyl. FGrHist 2046, 24.1), as well as that of the Dardani and the Paeonians (Strab. 7.5.1–2). The exact territory of the Ardiaei is also uncertain. The most common opinion among scholars is that the land of Autariatae was somewhere in eastern or southeastern Bosnia. There is a widely-held belief that at some point in history, when they were at the peak of their power, the Autariatae spread eastward into the Morava valley and areas beyond, which they took from the Triballi. As a consequence of this loss, the Triballi ceased to be an im- portant power in the Balkans interior. The power of the Autariatae was in turn terminated by the migrating Celts (Strab. 7.5.11). There are a number of issues with these claims, cf. Papazoglu, The Central Balkan Tribes 90–109 NEMANJA VUJČIĆ 501 ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 invaded Macedonia in 359, though explicit confirmation is lacking.9 There seems to be little support for the hypothesis that Cleitus was the recognized leader of all Illyrians in their struggle against Macedonia,10 or that his kingdom had a special, central position among the Illyrian tribes.11 When ancient writers use a phrase such as “king of Illyrians,” they simply mean an Illyrian king or a king of an Illyrian tribe. It should be noted that Cleitus is not specifically designated as a king by Arrian, whereas Glaucias is. There is still a widespread tendency to identify the Illyrians of Cleitus with the Dardani. This is an old fallacy. At its core, it originated with Droysen who made Monunius, one of the Illyrian kings of the third century B.C., into a Dardanian ruler, using inferred evidence. This served as a basis for a hypothesis, unsupported by any source, that the kingdom of the Dardani in the third century spread far beyond its traditional borders, encompassing a large part of modern Albania.12 However, Droysen never went so far as to claim that the invaders of 9 F. Papazoglou, “Les origines et la destinée de l’état illyrien: Illyrii proprie dicti,” Historia 14 (1965) 159; Hammond, JHS 94 (1974) 79; A. B. Bosworth, Conquest and Empire. The Reign of Alexander the Great (Cambridge 1988) 31; Cabanes, Les Illyriens 132; Nawotka, Alexander 98–99; Worthing- ton, By the Spear 129. 10 As suggested by Cabanes, Les Illyriens 132–133. 11 The hypothesis of the existence of one central and unified Illyrian kingdom, recognized as such by other tribes, which had an unbroken continuity in the 4th to 2nd centuries B.C. was upheld by Fanoula Papazoglu, Historia 14 (1965) 143–179; “Ilirska i dardanska kraljevina. Poreklo i razvoj, struktura, helenizacija i romanizacija,” in M. Garašanin (ed.), Iliri i Albanci (Belgrade 1988) 145–171. Outside of the former Yugoslavia, this idea gained a very modest following and is mostly rejected by scholars: see criti- cism by N. G. L. Hammond, “The Kingdoms in Illyria circa 400–167 B.C.,” BSA 61 (1966) 239–253, and Cabanes, Les Illyriens 124–128. 12 J. G. Droysen, Kleine Schriften zur alten Geschichte I (Leipzig 1893) 87–94. 502 PELION AND THE ILLYRIAN WAR OF ALEXANDER ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 Macedonia in the fourth century were the Dardani.13 He hypothesized that the kingdom of Bardylis and Cleitus was somewhere in northern Albania, and that this is the same king- dom later ruled by Agron and Teuta. Most scholars disagreed and placed the kingdom of Bardylis and Cleitus in southern Illyria.14 The final step in the identification with the Dardani was taken by N. G. L. Hammond, and his opinion was fol- lowed by many, among them A. B. Bosworth. As they were the leading modern scholars on Alexander and his reign, their influence in this regard was immense. To the majority of historians, especially in the English-speaking world, there seems to be little doubt about Cleitus’ origin: he is considered to be the king of the Dardani and his kingdom is sought somewhere in the area of modern Kosovo.15 Hammond held firmly to this opinion,16 whereas Bosworth did express some degree of reser- 13 Hammond, BSA 61 (1966) 252 n.52, wrote that Droysen “regarded Bardylis as Dardanian king.” This is false, Droysen clearly distinguished between the Dardani and the Adriatic Illyrians, cf. Kleine Schriften I 89: “Die Verhältnisse der illyrischen Völker sind besonders dadurch überaus schwierig, weil die Geographie der vielen kleinen Stämme noch unklar und weil die alten Schriftsteller oft von einzelnen Stämmen unter dem Gesamt- namen Illyrier sprechen. Geschichtlich treten in der Zeit von Philipp II. bis Perseus besonders drei Hauptmassen hervor: die Taulantiner in der Nähe von Dyrrachium und Apollonia, ein mit Epiroten gemischtes Volk (Strabo VII 326); dann das illyrische Reich an der Küste zu beiden Seiten des unteren Drilon, welches in Philipp II. zeit von Bardylis gegründet, bald mehr bald minder mächtig unter seinen Nachkommen bis Gentius hin bestand; endlich die Dardaner, deren Sitze oben bezeichnet sind.” 14 K. J. Beloch, Griechische Geschichte III.1 (Berlin 1922) 614; W. W. Tarn, in CAH VI (1927) 355, and Alexander the Great I (Cambridge 1951) 5; U. Wilcken, Alexander der Grosse (Leipzig 1931) 62. 15 So J. R. Ellis, Philip II and Macedonian Imperialism (Princeton 1986) 44, 47–48, 57–58, 60, 62, 67, 90; Worthington, By the Spear 129; C. J. King, Ancient Macedonia (London 2018) 140. 16 Hammond, The Kingdoms 245, 246 n.31, 252; N. G. L. Hammond and G. T. Griffith, A History of Macedonia II (Oxford 1979) 191–192; Hammond, Alexander 49. NEMANJA VUJČIĆ 503 ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 vation.17 However, there is practically no support in our sources for such an identification, while some facts go squarely against it. Hammond brought forward two arguments for this claim: 1) that Bardylis allowed Philip II to annex lands immediately south of lake Ohrid, supposedly indicating that the center of his kingdom must have been far from there, or else such a con- cession would be unacceptable.18 This is a rather hypothetical and unconvincing assertion, for Bardylis’ willingness to give up territory had little to do with political geography and every- thing to do with the crushing defeat suffered at the hands of Phillip’s forces, a defeat that left seven thousand Illyrian war- riors dead on the battlefield. And even if we concede that the center of the kingdom was at some distance, this would hardly prove that it was in the land of the Dardani. 2) The alleged poverty of southern Illyria (i.e. modern southern Albania) and the comparative abundance of Kosovo. Bardylis I in 359 and 358 led an army of over ten thousand men against the Mace- donians; the forces of Cleitus in 335 are not specified, but they had to be at least several thousand strong, since Alexander could not simultaneously besiege them and fight Glaucias. Ac- cording to Hammond, the allegedly destitute Dassaretis could not feed an army of this size, while Kosovo could.19 There are problems with this reasoning. First, we do not actually know the boundaries of Bardylis and Cleitus’ kingdom—it could have included other regions besides Dassaretis. But even the claim that this fairly spacious and productive country could not sup- port an army of around ten thousand men during a summer 17 Bosworth, A Historical Commentary I 69: “Cleitus was in all probability the king of the Dardani.” 18 Hammond, The Kingdoms 252. 19 “It remains to put to the test our suggestion that the kingdom of Bar- dylis was a Dardanian kingdom … As we have seen it cannot have been to the south in Dassaretis, a poor area” (The Kingdoms 252). 504 PELION AND THE ILLYRIAN WAR OF ALEXANDER ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 campaign is dubious, and Hammond himself contradicts it in his other works.20 Our sources clearly differentiate the Dardani on the one hand and the Illyrians of Bardylis and Cleitus on the other. During the reign of Philip II, Macedonians waged three major wars against Illyrians in the west, but no source mentions the Dardani in connection with these events.21 Of course, Alex- ander’s father did subdue Dardania, but this is mentioned separately by Justin, without a specific date: the context implies that this took place probably early in his reign (“After setting the affairs of Macedonia in order, he conquered the Dardani and other neighbors, victims of his deceit”).22 They remained quiet for the rest of Philip’s life, but upon his assassination they were contemplating defection. Again, they are mentioned sep- arately from other Illyrians: “and now there were the Illyrians, Thracians, the Dardani, and other barbarian tribes of dubious loyalty and treacherous intentions.”23 Their unrest did not lead to open war, because it soon became obvious that Alexander had full control of his father’s kingdom and the army. No 20 Hammond, Alexander's Campaign 71: “The great area of plain which extends from the foothills south of Pogradec to the hills south and east of Bilisht is more than 800 m above sea level. It is extremely fertile and very healthy both for men and animals. It has a long tradition of good agri- culture; indeed it is said to be nowadays the best cultivated area in Albania … In the Greco-Roman period this was the heartland of the Dassaretii.” Elsewhere (Hammond and Griffith, A History of Macedonia II 191) he speaks of “the rich agricultural lands and fisheries in the basins of Ochrid, Prespa, and Korcë.” 21 In 358 (Diod. 16.5.3–7), 356 (16.22.3; Plut. Alex. 3.8) and 344 (Diod. 16.69.7); perhaps there was another campaign late in the king’s lifetime (337?) but the reference (Diod. 16.93.7) is to an uncertain date and could easily be to the events of 344. 22 Just. 8.6.3, conpositis ordinatisque Macedoniae rebus, Dardanos ceterosque fini- timos fraude captos expugnat. 23 Just. 11.1.6, nunc Illyrios, Thracas et Dardanos ceterasque barbaras gentes fidei dubiae et mentis infidae. Cf. Diod. 17.3.5. NEMANJA VUJČIĆ 505 ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 source mentions the Dardani in connection with the campaign of the following year, and it should be noted that when Alex- ander received news of the Illyrian uprising, he was either still in the Morava valley or somewhere between the Morava and Struma valleys—in the close vicinity of the kingdom of the Dardani. Would a Dardanian king ride off to invade enemy territory in the south, while the same enemy was present in force at his doorstep? And why would Alexander spare the homeland of his disloyal vassal, when it was there for the taking? Instead, he went around the Dardanian lands, by- passing them first from the east and then from the south, as if he had no grudge against them.24 But the geographical context offers even more serious ob- stacles to the identification proposed by Hammond. If Bardylis and Cleitus are actually rulers of the Dardani, i.e. the country whose heartland is 150 km directly to the north of Macedonia, why do all of their attacks against Macedonia come from the west, via the route that goes south of lakes Ohrid and Prespa?25 We should expect them to use the most direct route, through the Vardar (Axius) valley, as did the Dardanian raids of the later third century.26 A separate question is: how did a king of 24 Regardless of this episode, some sources clearly distinguish the Dardani from the rest of Illyrians, and Dardanian land from Illyria (Polyb. 2.6.4, 28.8.3; Liv. 43.20.1), and some take for granted that they are but one of the Illyrian tribes (Strab. 7.5.6, 12; App. Ill. 1.2). This is not the place to analyze this topic in detail, but, while the Dardani are obviously ethnically and linguistically a part of the Illyrian world, they seem to be separated from other Illyrian peoples by their geography and peculiar socio-political de- velopment. A detailed discussion of the Dardanian language and ethnicity is provided by Papazoglu, The Central Balkan Tribes 210–262. 25 On the territory of the Dardani see Papazoglu, The Central Balkan Tribes 187–209. 26 Bosworth, in Macedonia and Greece 79, was aware of this difficulty, only to use it as a support for the claim that the city and the battle must have been located somewhere in Lyncestis (“To reach Lyncestis Cleitus needed 506 PELION AND THE ILLYRIAN WAR OF ALEXANDER ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 the Dardani manage to reach Pelion many days before the king of the Taulantii whose land is much further to the south and significantly closer to the site of Pelion? Instead of constructing elaborate but unlikely hypotheses to bridge these difficulties27 we should accept the simple solution that Cleitus’ kingdom was in the south of present-day Albania and that his attack on Macedonia went along the shortest and fastest route possible. Once again, there is a question whether Arrian has given us complete information concerning the identity of the attackers. Namely, were there other Illyrian ethnicities participating in the attack, besides the tribes led by Cleitus and Glaucias?28 Given Arrian’s evident contraction of other sources, as well as some serious omissions, this possibility cannot be excluded. That the ethnic and political landscape of southern Illyria in the fourth century was more diverse than this is occasionally implied by some ancient authors, as in the previously quoted statement of Justin. Pelion: its location and character Classical sources do not provide sufficient information to establish the precise location of ancient Pelion. According to Arrian, Alexander was advancing toward the lands of the Agriani and the Paeonians (αὐτὸς δὲ ἐπ’ Ἀγριάνων καὶ Παιόνων προὐχώρει) when he received news of the movements of the Illyrians (Anab. 1.5.1). The next topographical insight we get is that he marched up the river Erigon (Crna Reka, the largest of ___ merely to drive south through the Monastir gap”). Of course, this raises the further question why Cleitus, once in possession of the Vardar route, did not press on directly to the south to take a much greater prize, the heartland of lower Macedonia itself. Did he want to link up with Glaucias as soon as possible? Bosworth constructs a rather convoluted hypothesis to claim that there were sound political reasons for the invasion of Lyncestis. 27 Hammond, JHS 94 (1974) 78–79; Bosworth, in Macedonia and Greece 78–80. 28 As hypothesized by Cabanes, Les Illyriens 134–135. Cf. Just. 11.6. NEMANJA VUJČIĆ 507 ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 the western tributaries of the Vardar) to the city of Pelion (1.5.5, Ἀλέξανδρος δὲ παρὰ τὸν Ἐριγόνα ποταµὸν πορευόµενος ἐς Πέλλιον πόλιν ἐστέλλετο). Obviously, there is a gap in the narrative: in the first instance Alexander was north of Paeonia, about to reach the valley of the Struma (Strymon), in the second he already had crossed the Vardar and was about to leave Paeonia. Another gap follows and then we see Alexander setting up camp on the banks of the river Eordaicus (1.5.5, καταστρατοπεδεύσας πρὸς τῷ Ἐορδαϊκῷ ποταµῷ). This river is mentioned solely by Arrian, and it is usually identified as the Devoll or the Shkumbin, although either is uncertain. If we knew which river is meant, that would give us a firm basis to assess the whereabouts of Pelion. Alexander managed to lead his army out of camp, approach the city, and offer battle (for a moment it seemed that Cleitus would accept the challenge), with plenty of time afterward to encircle the city and build another camp in the neighborhood of the walls. All this took place in a single day (1.5, 7–8). Thus, if not by the river itself, Pelion must have been located in its immediate vicinity, not more than a couple of hours away by foot. Apart from Arrian, only three texts mention Pelion. The information they give is limited, but they mostly confirm one another. In narrating events of the Second Macedonian War, Livy tells us that in 199 the army of the consul Sulpicius Galba entered Upper Macedonia from the north, only to turn west and cross from Orestis into Dassaretis. Soon after entering the area, Romans took Pelion, “a city favorably placed for inroads into Macedonia.”29 From this we can conclude that Pelion was somewhere in the eastern Dassaretis, very close to the historical 29 Liv. 31.40.4–5: ab Celetro in Dassaretios processit urbemque Pelion vi cepit. servitia inde cum cetera praeda abduxit [et], libera capita sine pretio dimisit oppidumque iis reddidit praesidio valido imposito; nam et sita opportune urbs erat ad impetus in Macedoniam faciendos. For the Second Macedonian War Livy relied heavily on Polybius, from whom this information probably comes. 508 PELION AND THE ILLYRIAN WAR OF ALEXANDER ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 border with Macedonia. Since Roman troops came from Orestis, the location should be sought to the south or the west of lake Prespa. This is in line with information given by Stepha- nus of Byzantium who describes Pelion as “a city in Illyria.”30 In the sixth century Procopius (Aed. 4.4.3) mentions a restored fortress of the same name in the province Epirus Nova—again within the traditional confines of Illyria.31 Many opinions have been given about the exact location of Pelion and the nature of the settlement. In his pioneering work on the Illyrians, Gustav Zippel located Pelion on the Devoll river.32 W. W. Tarn held that the city (the river as well) was within the traditional boundaries of Macedonia, which was invaded by Illyrian attackers.33 Fanoula Papazoglu located Pelion deeper in Dassaretis, near modern Korçë, south of lake Maliq (Mališko);34 J. N. Kalleris accepted a similar solution, as did T. Winnifrith in his history of the border region of Epirus and Albania.35 The quest for the location was taken up most thoroughly by Hammond and Bosworth. Their conclusions are certainly stimulating but also highly hypothetical as well as mutually exclusive. Hammond’s well-documented study of the topogra- 30 Steph. Byz. 521 Πήλιον· ἔστι δὲ καὶ Πήλιον Ἰλλυρίας πόλις, ἧς µέµνηται Κουάδρατος. τὸ ἐθνικὸν Πηλῖνος. 31 There are discrepancies in the spelling of the name of the city: Πέλλιον (Arrian), Πήλιον (Stephanus), Πήλεον (Procopius). It is questionable whether this fact is of special importance, as argued by Bosworth. 32 G. Zippel, Die römische Herrschaft in Illyrien bis auf Augustus (Leipzig 1877) 28. 33 Tarn, Alexander 6, and in CAH VI 355–356. 34 Papazoglu, The Central Balkan Tribes: the general map titled “The Area Involved in the History of the Central Balkan Tribes.” 35 J. N. Kalleris, Les anciens Macédoniens I (Athens 1954), “Carte historique de la Macédoine au temps de Philip II”; T. Winnifrith, Badlands, Borderlands: A History of Northern Epirus/Southern Albania (London 2002) 143–148. NEMANJA VUJČIĆ 509 ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 phy of Alexander’s Illyrian campaign claimed that Pelion is on the site of Goricë, west of the Small Prespa Lake.36 Contra Hammond, Bosworth looked for Pelion in Lyncestis, claiming that Livy does not refer to Arrian’s Pelion at all, and that the whole passage is riddled with copyist’s errors. Bosworth ascribed great importance to the fact that the city’s name is written inconsistently in different texts. According to him, this is proof of the corruption of the original text: Arrian’s Πέλλιον is a product of a conflation with the toponym Πέλλα, and is placed somewhere in Upper Macedonia.37 Hammond, already critical of explanations proposed by Tarn and Papazoglu (be- cause “the physical features of Arrian’s account are lacking”), was exceedingly critical of the solution suggested by Bos- worth.38 In general, Hammond’s thesis has had a far wider impact than Bosworth’s, which is accepted by few scholars.39 Newer academic works, as well as many popular books, usually follow Hammond’s interpretation without hesitation.40 One more opinion should be mentioned. The archaeologist Neritan Ceka interpreted the remains of an early Hellenistic settlement west of Lake Ohrid (Selcë e Poshtme), on the route 36 Hammond, JHS 94 (1974) 66–87. He personally inspected all the major sites mentioned in the paper and included detailed though hand- drawn (i.e. imprecise) maps (70 and 72, maps of Goricë and its surround- ings). That on 72 is republished in Hammond and Walbank, A History of Macedonia III 42, fig. 4, and Hammond, Alexander 56, fig. 3 (here the design is somewhat more professional, but all the elements and distances are iden- tical with those in the older map). 37 Bosworth, Commentary I 68–73, and in Macedonia and Greece 75–84; cf. Winnifrith, Badlands 143–144. 38 Hammond and Walbank, A History of Macedonia III 41 n.1. 39 It is accepted by Nawotka, Alexander 99. 40 Cf. J. R. Ashley, The Macedonian Empire: The Era of Warfare under Philip II and Alexander the Great ( Jefferson 2004) 171–173; D. J. Lonsdale, Alexander the Great: Lessons in Strategy (London 2007) 141–144; S. English, Sieges of Alexander the Great (Barnsley 2009) 24–34; Worthington, By the Spear 129–130; etc. 510 PELION AND THE ILLYRIAN WAR OF ALEXANDER ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 of the later Via Egnatia, as the ancient Pelion. The claim is mostly accepted by other Albanian archaeologists and histor- ians. This site was excavated between 1969 and 1972; some monumental (royal?) tombs of the third century B.C. were un- covered. The identification is based on the (highly debatable) assumption that Pelion must have been an important center of the Illyrian culture and a seat of political power, in fact the capital of the Dassaretii. According to this version, the Eordai- cus of Arrian is actually the river Shkumbin and the battle took place deep inside Illyria, far to the northwest of the Mace- donian border.41 Some hypotheses are weaker than others, and can be discarded without much hesitation. First of all, if we follow the sources, Pelion must be somewhere to the west or south of Lake Prespa. To place the city in Lyncestis or Orestis would directly contradict our testimonies. Also, any location west or north of the Cangonj (Tsangon) pass will not match the description of Livy well (“city favorably placed for inroads into Macedonia”). This excludes the solution offered by Papazoglu, but especially the one proposed by Ceka. It is difficult to recognize the site of Selcë e Poshtme in the words of Livy. The location is too far from the border region with Orestis (some 70 km to the north- west, behind the mountain curtain and Lake Ohrid) of which Livy speaks. If the battle really took place at Selcë e Poshtme, then not only did the Illyrians fail to threaten Macedonia, they were in fact nowhere near the Macedonian border. We would have to assume that Alexander was moving so fast that he in fact launched the invasion of their country before they could invade his. There is no reason to think that the Pelion of Arrian 41 N. Ceka, “Qyteti ilir në Selcën e Poshtëme,” Iliria 2 (1972) 177–178; “Les tombes monumentales de la Basse-Selce,” Iliria 4 (1976) 367–379; The Illyrians to the Albanians (Tirana 2013) 136. This interpretation was accepted by A. Stipčević, Iliri: povijest, život, kultura (Zagreb 1989) 70, and (cautiously) by Wilkes, The Illyrians 123–124, 130. NEMANJA VUJČIĆ 511 ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 and Livy was an important center of the Illyrians. Indeed, there is little to suggest that it was a large urban settlement at all (more on this below). Now, Selcë e Poshtme certainly was an important center—the archaeological finds are undeniable— but this fact on its own offers no connection with the historical Pelion. These are two different settlements. The name of the settlement west of Lake Ohrid is simply not preserved (or not recognized) in the works of ancient authors. Bosworth’s reconstruction of the events requires us to accept that Cleitus was a Dardanian king who successfully breached the borders of Upper Macedonia, and that Alexander did not enter Illyria proper with his army, or that he did so only briefly, during the final pursuit. Again, this is hard to accept, given that all sources speak of Alexanders campaign in Illyria.42 Arrian mentions an Athenian embassy that congratulated the king in the name of the people, on account of his safe return from the lands of the Illyrians and the Triballi (Anab. 1.10.3, ὅτι τε σῶος ἐξ Ἰλλυριῶν καὶ Τριβαλλῶν ἐπανῆλθε). According to Plutarch, Demosthenes ridiculed Alexander while he was “among the Illyrians and the Triballi” (Alex. 11.3, ἕως ἦν ἐν Ἰλλυριοῖς καὶ Τριβαλλοῖς). Curtius puts in Alexander’s own mouth the claim that he conquered the Illyrians and the Triballi,43 and that he “subjugated Thrace and the Illyrians.”44 In my opinion, Zippel and Hammond have correctly identi- fied the wider area of ancient Pelion—the country immediately to the west of Small Prespa Lake and south of the Great Prespa. Establishing the precise location is an altogether more difficult task, one that could only be solved by a careful ex- ploration in the field, conducted with the information provided 42 See nn.2 and 3 above. 43 Curt. 6.3.2–3, ut omittam Illyrios, Triballos, Boeotiam, Thraciam, Spartam, Achaeos, Peloponnesum, quorum alia ductu meo, alia imperio auspicioque perdomui. 44 Curt. 9.6.20, orsus a Macedonia imperium Graeciae teneo, Thraciam et Illyrios subegi. 512 PELION AND THE ILLYRIAN WAR OF ALEXANDER ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 by Arrian in mind. So far, only Hammond has made an extensive attempt to do this. But, although his interpretation is widely accepted, it is encumbered with numerous difficulties, relies on unproven hypotheses, and cannot be taken as the solution to this problem. First of all, the match of the topogra- phy of Goricë with the source description is only partial. Arrian’s Pelion is in a forested valley, encircled by cliffs, hills, and mountains. There is elevated terrain to the north and southeast of Goricë; to the west and southwest, however, is the wide and completely open plain of Poloskë through which the Devoll flows. Thus, according to Hammond, even though he had wide-open ground to the west, suitable for both withdrawal and foraging, Alexander opted for a dangerous retreat towards the east, through a narrow gorge that the enemy, otherwise shrewd, left unprotected both then and in the following days, thus making the Macedonians’ retreat and their counterattack possible. This long and narrow gorge (Gryke e Ujkut, “Wolf’s Pass,” so narrow in fact that any army could pass through only in a column of two), a key topographical element in Hammond’s reconstruction, is actually not mentioned by Arrian. According to Arrian, the critical part of the retreat came at a place where the river runs close to a forested ridge (λόφος or γήλοφος), held by hostile forces (Anab. 1.6.5): Alexander’s army fought its way past this point, then rapidly crossed the river, with enemies in pursuit, leaving the unfavorable (broken and wooded) terrain behind (1.6.6–8). In Hammond’s version the army crosses the river first, and then is saved by marching through the gorge. In this reconstruction, the river is but a shallow stream that flows across the battlefield and has little impact on the events; the army could ford it at any point. But the biggest issue is this: the path that Alexander allegedly used to approach Pelion, and then to retreat from it, is a dead end (even Hammond calls it “swampy cul de sac”), that goes down to muddy terrain and ends on the shores of Small Prespa NEMANJA VUJČIĆ 513 ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 Lake (Mikri Prespa, Ventrok in Hammond’s terminology). To solve this cardinal difficulty, Hammond introduced the hypoth- esis that the water levels of the two Prespa lakes were far lower in antiquity than now, and that Small Prespa either did not exist at all or had a much smaller area.45 He also assumed that there was another river (the Eordaicus of Arrian), now gone without a trace, which was a tributary of Devoll and which drew the water out of the lake (and even a sketch of this hypothetical stream is given).46 All this is posited in support of the claim that east of the pass there was a stable belt of dry land. This is followed by yet another hypothesis, that through this belt ran an important road, frequently used in antiquity, and the final hypothesis, that this must have been the path taken by Alexander’s army in 335.47 The first two claims are hydrological and geological hypotheses brought forward by a humanist scholar, but easily disproved by recent hydrological studies. Rather than having a substantially smaller area in an- tiquity, the Small Prespa was actually larger and had higher water levels; in fact, in ancient times the two Prespa lakes likely formed a single body of water. The present size and shape of these lakes is a consequence of recent human interference. Be- cause of agricultural and other uses (and abuses) the water levels of both lakes have been in slow and steady decline since at least the late 1930s,48 a trend interrupted only by short 45 Hammond, JHS 94 (1974) 74. 46 Hammond reached this hypothesis by way of analogy with Lake Ostrovo in northern Greece. 47 JHS 94 (1974) 79, 81–84. 48 Cf. G. E. Hollis and A. C. Stevenson, “The Physical Basis of the Lake Mikri Prespa Systems: Geology, Climate, Hydrology and Water Quality,” Hydrobiologia 351 (1997) 2: “Lake Mikri Prespa has been affected by human action. In 1936 the Agios Germanos Stream was diverted from Mikri Prespa to its present artificial channel leading into Megali Prespa. In 1953 the Albanians linked Mikri Prespa to the River Devoll so as to allow water to enter the lake in winter and to drain from the lake in summer for irri- 514 PELION AND THE ILLYRIAN WAR OF ALEXANDER ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 periods of extensive rainfall. When Hammond surveyed the shores of these lakes in 1972 they were experiencing a con- tinuous decade-long decline that continued for several more years (1963–1975);49 since the 1980s the trend has accelerated, with water loss reaching near catastrophic levels.50 In short, Hammond’s hypotheses on the ancient water levels are in this case untenable and so then is the proposed route of Alexan- der’s army. Before we turn to the reconstruction of the events, one more question of some importance remains: the status and character of Pelion itself. There is no direct evidence on this, but the fact that it is a fortified settlement with a Greek name, situated in- side traditional Illyrian territory conquered by Philip II in 358, is significant and indicative. After defeating Bardylis, the king took possession of the lands immediately south of the Ohrid and Prespa lakes, thus providing Macedonia with a new and strong western border.51 Sources confirm that he invested money and labor in building new fortifications and settlements in these parts, as was indeed the case elsewhere: all his con- quests in the Balkans were followed immediately by significant colonization and construction efforts.52 Thus, the simplest and most straightforward hypothesis explaining the existence of Pelion would be that this fortress-town was founded by Philip, ___ gation. In 1969, the Albanians added a dam and sluice to the canal leading from Mikri Prespa.” 49 C. Popovska, “Hydrology of Lake Prespa,” Vodoprivreda 48 (2016) 23– 24; cf. Hollis and Stevenson, Hydrobiologia 351 (1997) 8. 50 T. van der Schriek and Ch. Giannakopoulos, “Determining the Causes for the Dramatic Recent Fall of Lake Prespa,” Hydrological Sciences Journal 62 (2017) 1131–1148. 51 Hammond and Griffith, A History of Macedonia II 652–657; N. G. L. Hammond, The Macedonian State (Oxford 1989) 107–108. 52 Dem. 4.48; Just. 8.3.7–8. Cf. R. M. Errington, A History of Macedonia (Berkeley 1990) 42–43. NEMANJA VUJČIĆ 515 ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 perhaps through expansion and fortification of an existing settlement of the Dassaretii, to be one of the strongholds (per- haps the stronghold) in the conquered territory. Such defenses served a dual purpose: securing control over the newly ac- quired territory and its population, and acting as serious ob- stacles in the path of any future invader. Though Pelion (and, presumably, other strongholds) fell into Cleitus’ hands before Alexander arrived, the defense system fulfilled its purpose, at least partially, by delaying the Illyrian advance. As a conse- quence, the battle was fought in the vicinity of Pelion, and not in Macedonia proper. Arrian uses the word πόλις for Pelion no less than twelve times, while Livy describes it twice as urbs. No far-reaching conclusions should be made on the basis of these terms.53 Two points are obvious from both texts: it was a settlement of a certain size, and it had fortifications serious enough to give pause to a Macedonian king.54 This is enough to justify the application of the word πόλις in the narrow sense of a fortified settlement: town or citadel, but hardly something more.55 The name itself is significant: all of the proper poleis founded during Philip’s reign carry dynastic names.56 53 J. Wilkes and T. Fischer-Hansen, in An Inventory of Archaic and Classical Poleis (Oxford 2004) 326, quite correctly categorized it among “Pre- Hellenistic settlements not attested as polis.” 54 The claim by Green, Alexander of Macedon 132, that “it was an all but impregnable stronghold” is a significant exaggeration. After all, we do not know what it looked like, and it did change hands twice in the space of several days. 55 See LSJ s.v. πόλις. For walls as one of the defining features of a polis see J. McK. Camp II, “Walls and the Polis,” in P. Flensted-Jensen et al. (eds.), Polis and Politics: Studies in Ancient Greek History presented to Mogens Herman Hansen (Copenhagen 2000) 41–57. 56 Cf. Hammond and Griffith, A History of Macedonia II 354, 358–361, 557–559, 656, 661, 673. 516 PELION AND THE ILLYRIAN WAR OF ALEXANDER ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 Opposing forces and their aims What was Cleitus even doing in Pelion? The passage of Arrian is sometimes interpreted as though the city was already in possession of the Illyrians, and that therefore it functioned as the center of the uprising. This is wrong. The text is clear enough (Anab. 1.5.5): Ἀλέξανδρος δὲ παρὰ τὸν Ἐριγόνα ποταµὸν πορευόµενος ἐς Πέλλιον πόλιν ἐστέλλετο. ταύτην γὰρ κατειλήφει ὁ Κλεῖτος ὡς ὀχυρωτάτην τῆς χώρας· καὶ πρὸς ταύτην ὡς ἧκεν Ἀλέξανδρος, καταστρατοπεδεύσας πρὸς τῷ Ἐορδαϊκῷ ποταµῷ τῇ ὑστεραίᾳ ἐγνώκει προσβάλλειν τῷ τείχει. Alexander, marching along the river Erigon, made for Pelium; this city Cleitus had taken, as being the strongest in the country. When Alexander reached it, he camped by the river Eordaicus and decided to assault next day. (transl. E. I. Robson, Loeb) The basic meaning of the key word, καταλαµβάνω, is to seize, lay hold of, seize for oneself, but also to arrive at (LSJ s.v.) The form κατειλήφει could only mean that Cleitus took possession of Pelion shortly or immediately prior to these events, not that he had held it for a long time57 as is sometimes inferred. Pelion was not Cleitus’ ancestral fortress but one captured during this particular campaign. Arrian mentions messengers who brought the alarming news to Alexander, but not from where they were sent or by whom. The fragmentary historical papyrus could possibly offer some insight. This badly damaged text was once a part of either a Hellenistic history of Alexander58 or a philological commentary on such a work.59 In either case, we can glean some significant 57 This is how the text is understood by most scholars, among others Tarn Alexander 6 and in CAH VI 355; Hammond and Walbank, A History of Macedonia III 41–43 (41 n.1). 58 The opinion held by Clarysse and Schepens, ChrÉg 60 (1985) 30–47. 59 As argued by N. G. L. Hammond, “A Papyrus Commentary on Alex- ander’s Balkan Campaign,” GRBS 28 (1987) 331–347. NEMANJA VUJČIĆ 517 ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 information from it. There are three groups of (mostly dis- connected) fragments, likely parts of three distinct columns of text. Column i contains few recognizable letters, while column iii, although badly damaged, is readable enough to conclude that it deals with the end of the Thracian campaign and/or the beginning of the Illyrian: a river is mentioned, and a stream, the Thracians and the Agriani, Paeonia (twice), and Philotas son of a Parmenio. The most interesting part of the text is the second column.60 There we read that the king appointed one of his friends, a certain Korragos, with a part of the army, to keep watch over the western frontier. Some threat was forming be- yond the borders, and Korragos sent frequent messages to warn the (king’s?) army. After some lacunae comes mention of something (an army?) coming to Eordaia and a “neighboring part of Elimeia,” τῆι πλησίον Ἐλειµία[ι. The preserved text of the column ends at this point. From this little fragment we can infer that troubles on the western border of Macedonia were already considered and, perhaps, expected during or even 60 Reading by Clarysse and Schepens: [– – –] . . / [. . . . . .]ρ̣α̣[. . . . . . . . . . . . . . . . . ]µησασ / [. . . . ]τα̣ν ̣κ̣α̣ὶ̣ .[. . . . . . . . .] πορευθῆι / Κ̣όρρ̣αγον . εν . [. . . . . κατ]α̣στήσας // ἕνα τῶν ̣φίλ̣ων [. . . . .] . ς τῆς δυ/νάµεως τὸ̣ ̣πρὸς τ[ὴν] χρείαν οἰκεῖ/ον παρὰ τὰ τῆς Ἰλλυ[ρίδ]ος ὅρια πυ/κνα̣ς πρ̣οσ̣έ̣ταξε πο[̣ιο]ύµενον / τὰς ἐπὶ τὸ στρατόπ[εδον . ]ς ὑποδει//κνύειν τὸν̣ ἐσόµ[ενον κί]νδυνον / ε̣π ̣. . . . . . . . . . .[. . . . . . . . .]. ἐπερ/χοµ ̣έ[ν]η̣ς ⟦τῆς⟧ [. . . . . . . . . . . χ]ῶ̣ρα̣[.] / —— / Εἰς δὲ τὴν Ἐορδαί[αν . . . . . . . . .] / τῆι πλησίον Ἐλειµία[ι . . . . . . .] Hammond’s proposal: [– – –] . . / [. . . . . . .]ρα[. . . . . . . . . . . . . . . . .]µησασ / [. . . .]ταν καὶ .[. . . . . . . . .] πορευθῆι / Κόρραγον [Μ]εν[οίτου κατ]αστήσας· // ἕνα τῶν φίλων. [µέγα µέρο]ς τῆς δυ/νάµεως· τὸ πρὸς τ[ὴν] χρείαν οἰκεῖ/ον παρὰ τὰ τῆς Ἰλλυ[ρίδ]ος ὅρια. πυ/κνάς· προσέταξε πο[ριο]ύµενον / τὰς ἐπὶ τὸ στρατόπ[εδον. ὡ]ς ὑποδει//κνύειν τὸν ἐσόµ[ενον κί]νδυνον / επ. . . . . . . . . . .[. . . . . . . . .]. ἐπερ/χοµέ[ν]ης ⟦τῆς⟧ [. . . . . . . . . . . ]ωρα. / —— / Εἰς δὲ τὴν Ἐορδαί[αν . . . . . . . . .] / τῆι πλησίον Ἐλειµία[ι . . . . . . .] The text is republished in L. Prandi, Corpus dei papiri storici greci e latini A.2.9 (Pisa 2010); at the time of writing this publication was not avail- able to me. 518 PELION AND THE ILLYRIAN WAR OF ALEXANDER ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 before the Thracian campaign. Korragos, a king’s philos or hetairos, was therefore sent into this region with some forces to observe the situation in Illyria.61 The messengers bringing news of the coming invasion were his. In the later stages of the campaign, Alexander crossed the river Eordaicus twice, once during the retreat from Pelion, and the second time to counterattack and retake the city. It follows that he went along the same route in his initial advance. The river Eordaicus lay between his main encampment and the city; it would present the main obstacle during the retreat. We are not well informed about the size and composition of Alexander’s army and we are told even less about the armies of the two Illyrian kings. There is some evidence about the specific units Alexander had with him: there were the hypaspists (ὑπασπισταί), a number of (mercenary?) archers (τοξόται), light mountaineer infantry of the Agriani (Ἀγρίανες), some cavalry (ἵππεις), and a substantial number of phalanx infantry (φάλαγξ, ὁπλῖται, Anab. 1.5.10, 1.6.1–2.). At one point Alexander ar- ranged the phalanx to form a square 120 rows deep (1.6.1.): obviously, the total number of these soldiers must have been in the thousands. Elsewhere, we are told that there were two thousand of the Agriani and archers combined (1.6.6.). When Philotas took some of the cavalry to protect the foragers, Alexander was able to come to his aid with mere 400 horsemen (1.5.10). Even if he did leave some behind with the phalanx by Pelion (they are not explicitly mentioned), we still may con- clude that his cavalry force was of modest size, probably under a thousand horsemen; when the phalanx advanced, its sides 61 This Korragos is perhaps mentioned at Diod. 17.100.1–101.1: a person of this name is here described as a Macedonian and one of the king’s friends, “often distinguished in battles” (καὶ πολλάκις ἐν ταῖς µάχαις ἠνδρα- γαθηκώς, 17.100.2). He is mentioned only in connection with a duel he fought and lost against an Athenian athlete. The same episode is told by Curt. 9.7.17–26, where his name is given as Corratas. NEMANJA VUJČIĆ 519 ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 were protected by modest detachments of cavalry, two hundred strong each (1.6.1). During the first days of the campaign there is no mention of siege engines of any kind and one can infer that Alexander had none with him. The com- position of the troops the king took with him indicates that he did not expect to conduct a siege: rather, he was preparing himself for a battle in the open against the army of Cleitus. This is supported by the fact that the army lacked supplies for more than a few days. As to the Illyrians, it is probable that their combined armies significantly outnumbered the force that Alexander had on hand, though again no numbers are given. The Illyrian armies were also composed of various units, with different armament. We are told of light infantrymen such as slingers (σφενδονῆται) and soldiers armed with javelins (ἀκοντισταί), as well as num- erous hoplites (ὁπλίταις δὲ οὐκ ὀλίγοις) and, of course, horsemen ( ἵππεις, Anab. 1.5.12). Bowmen are not explicitly mentioned and it is noticeable that Alexander was able to keep Glaucias’ men at a distance using his own archers. All things considered, the Illyrians too seem to have been prepared for a full-on engage- ment. Reconstructing the events Several conclusions reached in the previous discussion will unavoidably affect the proposed reconstruction of the events, which will therefore depart in a number of important aspects from descriptions found in modern literature. First, we will assume that Alexander did not bring forward his whole army, but a smaller and selected force, as was the case later during the final stage of the battle. His main camp and most of the supplies were left on the other side of the river, a decision that had various ramifications for the rest of the campaign. Second, the widespread supposition that Alexander came with the in- tention of conducting a siege will be doubted, because it is not supported either by the text of Arrian, nor does it seem likely from what we are told about the composition of the Mace- 520 PELION AND THE ILLYRIAN WAR OF ALEXANDER ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 donian army. Instead, an argument will be made in favor of a pitched battle. The issue of Macedonian siege machines and their use will be examined separately. Finally, rejecting ques- tionable assumptions about the topography of the battlefield (such as those of Hammond) will enable us to reach a recon- struction that is more in line with the facts provided by Arrian. The whole Pelion campaign can be divided into three distinct stages. The initial one encompasses the first two days. During this stage, Alexander arrived in the area and there was almost no fighting. The forces of Cleitus were the only (visible) enemy. Glaucias had yet to appear, or so claimed Arrian’s source (Ptolemy?). But there is some possibility that this was a ruse, that the second Illyrian army was already in the wider area, ready to strike at Alexander’s rear if the opportunity pre- sented itself. The two allies had local conditions working to their advantage and were in a position to create a tactical de- ception and trap the Macedonian army. When the Mace- donians finally crossed the river and approached Pelion, they encountered prepared defensive positions around the city. The Illyrians were seemingly ready to fight Alexander in the open; according to Arrian, they were expecting him to march straight to the city, in which case they would converge on his troops from the higher ground (Anab. 1.5.6). But either Alexander did not act as expected or Cleitus had a last-minute change of heart. When the Macedonians advanced to engage them, the Illyrians abandoned these positions and withdrew inside the walls of Pelion. They left behind nine sacrificial victims: young boys and girls and (black) rams, three of each (1.5.7). Human sacrifice was still widespread among the Balkan tribes at the time, and would continue to be for centuries.62 The Mace- 62 For an overview of known examples of human sacrifice in the pre- Roman Balkans see F. Papazoglu, “Ljudske žrtve i tragovi kanibalizma kod nekih srednjobalkanskih antičkih plemena,” Zbornik Filozofskog fakulteta 10 (1968) 47–61; cf. Wilkes, The Illyrians 123, 243–244. NEMANJA VUJČIĆ 521 ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 donian king invested the city with his available troops, ending action for the day. Whatever were his intentions for the follow- ing day, they were foiled by the appearance of another Illyrian army. The second stage begins when Glaucias arrived (or revealed himself). The second Illyrian force took possession of the high ground, left vacant by Cleitus’ men (Anab. 1.5.8). The Mace- donians were now in a precarious situation, trapped between the enemies in the fortress and the enemies holding the heights. Since he was cut off from his main camp and the regular food supply, Alexander sent Philotas with some troops to forage (1.5.9). This decision nearly led to disaster. Glaucias grabbed the opportunity to attack the Macedonians who were occupied with searching for food and bringing animals to pasture; Alex- ander had to intervene in haste to save Philotas and his men (1.5.9–11.). After this setback, the only option was to extract the army from the trap. This was anticipated by his enemies who held the elevated points along the route he was going to take (clearly, the same route he used to approach Pelion), with the intention to strike from the rear and on the flanks while the army was on the march (1.5.11–12). The critical part of the retreat, obvious to both Alexander and his enemies, was a narrow path between the river bank and a high forested hill (ὄρος ὑπερύψηλον); the entire area through which the army passed is described as “confined” and “wooded” (1.5.12, τά τε χωρία δι’ ὧν ἡ πάροδος ἦν τῷ Ἀλεξάνδρῳ στενὰ καὶ ὑλώδη ἐφαί- νετο). This is neither a “canyon” nor a “gorge” that we read in some modern texts. Another clear fact is that the river itself presented a formidable obstacle. The Eordaicus of Arrian must have been a watercourse of significant size which, unlike the negligible stream of Hammond’s hypothesis, could only be crossed at a ford. Devoll certainly fits this description. Alexander arranged his phalanx in a formation of extreme depth (120 soldiers), divided into two wings, each flanked by a cavalry detachment. While moving, the phalanx turned its sarisae left then right, signaling that it could beat off an attack 522 PELION AND THE ILLYRIAN WAR OF ALEXANDER ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 from either side. This had a discouraging effect on their adver- saries. At an opportune moment, Alexander sent his left wing against the pursuing enemies. The result was a rout, with very little actual combat; the troops holding the surrounding hills also fled. However, the critical hill above the river was still held by the Taulanti, but now it was a smaller, isolated force. The king sent his hypaspists with some dismounted cavalrymen to take the hilltop; the enemy again withdrew, rather than face the overwhelming force. He left the Agriani and the archers to hold the hill, while sending the rest of the army to the ford that was nearby. Part of the phalanx was used as a rearguard, while the rest of the army crossed the river. Once most of the troops where on the other side, Alexander ordered archers and the Agriani to withdraw from the hill they were holding and move swiftly to the river. The enemy seized the opportunity, charged past the abandoned position and attacked the rear of the col- umn (1.5.12–6.7). At this point in the story, war machines first make their ap- pearance. Alexander “issued the order that machines be set on the bank, to fire the projectiles as far as possible, and also for archers to shoot arrows from the middle of the river.”63 Glaucias did not dare lead his men against this barrage of missiles; soon the entire Macedonian army, cavalry and archers included, was safe on the other side of the river (1.6.8). Here too we are faced with the problems created by the condensa- tion of a much longer text. What were these machines (µηχαναί) and from where did they come? Ancient writers use the word µηχανή to describe a whole range of different contraptions. However, from the context it is obvious that these were missile devices, catapultae or ballistae of some sort, used to fire projectiles 63 Anab. 1.6.8, ἐπιστήσας ἐπὶ τῇ ὄχθῃ τὰς µηχανὰς ἐξακοντίζειν ὡς πορρω- τάτω ἀπ’ αὐτῶν ἐκέλευσεν ὅσα ἀπὸ µηχανῶν βέλη ἐξακοντίζεται, καὶ τοὺς τοξότας δὲ ἐκ µέσου τοῦ ποταµοῦ ἐκτοξεύειν. NEMANJA VUJČIĆ 523 ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 (javelins, sling bullets, perhaps mere rocks) at the enemy.64 It is usually assumed that Alexander had them with him from the beginning, but this seems highly unlikely for several reasons. Ancient war machines were unwieldy contraptions, difficult to set up and move. When not in use, they were disassembled and carried on wagons,65 so it seems impossible that they could be made ready on a whim, to protect the crossing. When as- sembled, they were not particularly mobile, and it strains credulity that Alexander’s men could bring them with them during their rapid and dangerous flight from Pelion.66 The only logical explanation I can see is that the machines remained behind in the main camp and were made ready for use by the troops left there. Perhaps someone brought word in advance that they would be needed. The machines were thus set up on the other side of the river (there is no mention that they were moved across either way: they must have been on the safe side all along), by and around the ford, from where they could throw missiles at the enemy. Alexander used similar tactics in 64 For a general overview of Greek and Macedonian siege warfare see D. B. Campbell, Besieged: Siege Warfare in the Ancient World (Oxford 2006) 30–79; M. Obradović, “Opsadne sprave i razvoj poliorketike u grčkom svetu do vremena Filipa II i Aleksandra Velikog,” Vojnoistorijski glasnik / Military Historical Review (2016.2) 9–30. For detailed discussion of various types of ‘artillery’ available in the fourth century, its use and spread, see E. W. Marsden, Greek and Roman Artillery: Historical Development (Oxford 1969) 5–24, 48–73. 65 Cf. Mardsen, Greek and Roman Artillery 164: “Before the introduction of the carrobalista about A.D. 100, it appears that neither the Greeks nor the Romans possessed any pieces of artillery permanently mounted on mobile carriages. They transported their catapults on ordinary carts in a more or less dismantled state.” 66 Hammond and Walbank, A History of Macedonia III 46, concluded that during the rapid retreat of the army “the baggage train with its men and animals had been lost, though Arrian fails to mention it.” With this in mind, is it possible to assume that the war machines were somehow brought along, with the necessary speed and undamaged? 524 PELION AND THE ILLYRIAN WAR OF ALEXANDER ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 his later campaigns.67 The third stage came after the successful retreat across the river. Alexander went back to his main camp, but kept sending scouts to keep an eye on the enemy. Three days after the retreat, scouts observed that the troops of Cleitus and Glaucias were scattered through the countryside around Pelion; they built no single camp nor any kind of fortifications, and their reconnaissance and watch seemed to be lax and inefficient. Seizing the opportunity, Alexander transferred the immediately available forces (some seven thousand men) across the river during the night (Anab. 1.6.9). These units surprised the sleep- ing Illyrians and annihilated them with savage resolve. Multi- tudes of Illyrian warriors were killed on the spot or during the ensuing confusion and panic. The rest mostly fled (1.6.10). With a fraction of his previous force Cleitus took refuge in Pelion. Seeing that there was no hope of holding the fort and that he could easily be surrounded, he decided to torch the settlement and flee (1.6.11). Two details require further comment. The allied kings did not pursue the Macedonians across the river, and they were seemingly lax about the possibility of their enemies coming back. True, the Illyrians were not effective in inflicting any real casualties on the Macedonian army,68 but they were successful in the sense that they forced the Macedonians to abandon Pelion and its immediate area. Much of the country south of lakes Ohrid and Prespa, the parts that Philip took from Bar- dylis in 358, was now back in their hands, including what was 67 During the campaign in Sogdiana in 329 Alexander deployed war machines on the bank of the Jaxartes (Syr Darya) to force the Scythians on the other bank to flee, removing any opposition to his advancing troops: Anab. 4.4.4, αἵ τε µηχαναὶ ἀπὸ ξυνθήµατος ἐξηκόντιζον ἐς τοὺς Σκύθας παριπ- πεύοντας ἐπὶ τῇ ὄχθῃ. 68 Arrian even claims that “none of them died during their retreat,” ὥστε οὐδεὶς ἀπέθανεν ἐν τῇ ἀποχωρήσει αὐτῶν (1.6.8). NEMANJA VUJČIĆ 525 ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 probably the main fortress. What they achieved thus far could have been interpreted as a victory of a sort. The lack of alert- ness on their part is more difficult to understand; Anab. 1.6.9 suggests that they assumed that Alexander had left per- manently. Aftermath and conclusion Once again, we can ask if Arrian has told us everything we ought to know. And once more, the answer is almost certainly No. For example, we are not told what conditions were im- posed on the defeated kings. We are only informed that Cleitus, after burning Pelion, fled and found refuge in the land of Glaucias.69 That some permanent arrangement was reached is evident from Alexander’s subsequent behavior—he acted as if the problem was fully resolved and rushed south to deal with the Theban uprising—and from the fact that some Illyrians (from Cleitus’ kingdom?) did participate in his Asian wars, albeit in a much less prominent role in comparison with other Balkan tribes.70 Pelion, though torched by the Illyrians, was later rebuilt and continued to exist as a fortified settlement until late antiquity. There is no mention of any territorial changes, but given the general state of the sources for the period this means very little. What we do know of the western borders of Macedonia in the ensuing period, when the country was gov- erned by Antipater, points to the conclusion that they were stable, and remained so at least until Cassander’s Illyrian cam- paign of 314.71 Alexander’s Illyrian war was a success not only in its im- mediate consequences, but also long-term. The restless tribes were pacified in 335 and remained so through Alexander’s 69 Anab. 1.6.11, Κλεῖτος δὲ ἐς τὴν πόλιν τὸ πρῶτον καταφυγὼν ἐµπρήσας τὴν πόλιν ἀπηλλάγη παρὰ Γλαυκίαν ἐς Ταυλαντίους. 70 Cf. Diod. 17.17.4; Arr. Anab. 2.7.5. 71 Diod. 19.67.6–7, 78.1, 89.1–3. 526 PELION AND THE ILLYRIAN WAR OF ALEXANDER ————— Greek, Roman, and Byzantine Studies 61 (2021) 496–526 reign. The following year he crossed into Asia, never to come back to Illyria or Europe. The forces he left with Antipater in Europe appear modest when compared with the army taken into Asia (Diod. 17.17.3–5), but they must have seemed adequate, especially when combined with the levies of the League of Corinth. This judgment was at least partly justified by subsequent events. Antipater experienced serious military challenges twice during Alexander’s lifetime (Memnon’s rebel- lion in Thrace in 332/1 and the war of Agias III in the Pelo- ponnese in 331), and both times managed to overcome them with the troops in hand.72 With the defeat and death of Zo- pyrion, Macedonian control over much of Thrace crumbled,73 but there was no comparable development in Illyria. Only with the onset of the Lamian war in 323 did the aging regent find himself in a situation he could not handle without outside help, with Macedonia exhausted by the need to provide constant re- inforcements for the Asian campaign, one year after another.74 And even then, the Illyrian borderland remained quiet. July, 2021 The University of Belgrade nemanja.vujcic@f.bg.ac.rs 72 Diod. 17.62.1–63.4; Arr. Anab. 2.13.4; Curt. 6.1; Just. 12.1.6–11. 73 Curt. 10.1.44–45; Just. 12.1.4, 12.2.16–17. 74 Cf. Diod. 18.12.1–2.