This is the third part of the fourth part of our four(ish) part (Ia, Ib, IIa, IIb, IIIa, IIIb, IVa, IVb) look at the how the Roman military system and its manipular legion were able to defeat the Hellenistic military system and its Macedonian sarisa phalanx in the third and second centuries BC. Last time, we looked at the largest of the great Roman wars against the Hellenistic successor states, the Roman-Seleucid War (192-188) and the decisive action of that war, the Battle of Magnesia (190). We concluded that, once again, the Romans seem to have a clear tactical edge, but what makes that so catastrophically decisive against Antiochus III's Seleucid Empire were Roman advantages in naval warfare and allies as well as the "remarkable Seleucid glass jaw."
This week, we're looking at the last of the major wars that Rome fights against the Hellenistic 'great powers,' the Third Macedonian War (171-168), leading to the decisive Battle of Pydna in 168, after which the Romans, for the first time, simply abolish one of the successors of Alexander. At the same time, as we'll see, the Roman Republic is able to immediately put the deterrence power of that victory to work forcing an end to the Sixth Syrian War (170-168) between the Seleucid Empire and the Ptolemaic Kingdom and in the process setting the Seleucids on their long road to decline while dramatically asserting Roman hegemony across the eastern Mediterranean.
This thing is just gonna keep on losing. But the Macedonian light infantry and cavalry actually fare a little better this week, until they don't.
But first, if you like what you are reading here and want to help out this project, please share it – I rely on word-of-mouth for all of my readers. And if you really like it and want to help me waste spend more money on swords research, you can support me, my research and this project on Patreon! If you want updates whenever a new post appears, you can click below for email updates or follow me on twitter (@BretDevereaux) for updates as to new posts as well as my occasional ancient history, foreign policy or military history musings, assuming there is still a Twitter by the time this post goes live. I am also on Bluesky (@bretdevereaux.bsky.social) and (less frequently) Mastodon (@bretdevereaux@historians.social).
The Background
You will recall that after the Second Macedonian War (200-197), the Romans had opted to prune the Antigonid kingdom of Macedon but not fully uproot it: Philip V (r. 221-179) was left in power, but Macedonian control over Greece was ended (in favor of a looser but still quite real Roman hegemony), with Philip's kingdom largely pruned back to Macedon proper. The peace also limited Philip V to a 5,000 man army and a token navy.
That limited army and the need not to anger Rome meant that Antigonid ambitions would have to be curtailed for a time, but it is clear that Philip V was not idle. On the one hand, he supported Roman military operations in the region (particularly in the Roman-Seleucid War) as a means of avoiding further Roman intervention in the region. One may read his support for Rome against Antiochus III in particular as a bid to remove Rome's excuse for remaining heavily involved in his own backyard. Meanwhile, Philip V appears to have turned inward towards what strength Macedon proper could provide. Livy (39.24) reports that Philip V expanded state revenues from agriculture, trade and mining, while both encouraging natural population growth but also moving substantial numbers of Thracians into his kingdom to provide for more manpower. The irony then is that this reduced Antigonid state will meet Rome with greater financial resources and more men than the much larger Antigonid state of the late 200s.
Philip V dies in 179 and is succeeded by his son, Perseus. Perseus himself seems to have been anti-Roman in outlook, in contrast to his brother Demetrius, who had been held as a hostage in Rome and seems to have been much more pro-Roman in outlook (e.g. Livy 39.53.1-6). We're told that Perseus, fearing that when his father died the Romans would supplant him with the more pliant Demetrius, schemed to incriminate him in a plot against Philip V in the last years of his reign and have him killed, thus clearing the way for Perseus to take the throne without opposition.
Via
Wikipedia, a tetradrachma of Perseus, minted early in his reign. The reverse, conveniently, reads, ΒΑΣΙΛΕΥΣ ΠΕΡΣΕΥΣ, "King Perseus." Note the cloth fillet he wears: that is the royal diadem, the Macedonian equivalent to a crown.
Relations between Perseus and Rome seem to have soured fairly quickly. Livy, our main source for this, presents Perseus as implacably anti-Roman, having plotted since before he was king to fight Rome (Livy 42.5.1) and so when he becomes king began preparing to do so. In practice, Perseus, it seems to me, was of a similar mind to the other Hellenistic rulers of this day, especially Antiochus IV (r. 175-164). Remember that by this point the successors of Alexander had been fighting over his empire for a century and a half, with variable fortunes. In that context, it was reasonable to suppose that Rome was just one more contestant in the ring, whose fortunes would rise and then fall. Consequently, a few decades after Rome's great victory, with a renewed kingdom, it made sense to begin attempting to reestablish the Antigonid position, the same way Antiochus IV was, at this very moment, reestablishing the Seleucid position and would soon attempt a renewed conquest of Egypt.
What may have actually set Demetrius apart - though our sources give no indication of this, so take this as the pure speculation that it is - was that, having been in Rome, he might have been a bit more aware of the degree to which the Roman Republic was a different creature than the Hellenistic kingdoms. After all, being in Rome, first as a hostage and later as a diplomat, he couldn't have been unaware of how many other wars Rome was waging, almost continuously in this period. and how substantial the military forces Rome kept continuously deployed. After all, in the last ten years of Philip V's reign (so from 189 to 179), Roman annual deployments never slip below about 110,000 men in any year. By contrast, the Antigonids will never deploy more than about 50,000 at any point in their history. But that massive disconnect might simply not have been obvious for an eastern Mediterranean state which might, understandably, be ill-informed about the seemingly endless Roman wars in Spain, Northern Italy and Gaul.
Meanwhile, agitating for renewed war in the region was Eumenes II of Pergamum - the fellow who won on the Roman wing at Magnesia - who clearly saw a resurgent Antigonid Kingdom as a threat and also stood potentially to gain at the expense of Perseus by greater Roman involvement. Seizing on the fact that Perseus seemed to be wooing the Greek poleis to resistance to Rome (which he was), Eumenes came to Rome in person (or perhaps he sent his brother, though Livy thinks it was Eumenes, Livy 42.11.1) to claim that Perseus was preparing for war. Perseus' own embassy, predictably, denied the charge but got little hearing, in part because they didn't seem particularly afraid to risk a war if it came to it (Livy 42.14.2-4). Interestingly, a number of Greek states, fearing that Eumenes rather than Perseus, might use a war to establish his control of the region, appealed to the Romans in the opposite direction, but succeeded mostly in irritating the Senate for abusing such a stalwart Roman ally in their speeches (Livy 42.14.9-10). As a result, late in 172 the Romans move for war against Perseus and the Third Macedonian War begins.
Armies and Allies
Perseus responds to the Roman declaration of war by assembling his own army at Citium (in Macedon, not the Citium on Cyprus) and Livy provides a report of the forces he assembled there (Livy 42.51). Michael Taylor breaks down the numbers - which require a touch of interpretation due to the way Livy presents them - as consisting of 21,000 Macedonian phalangites, 5,000 Macedonian peltasts (elite heavy infantry, not skirmishers - the 2,000 man royal guard, the agema is drawn from this number), 3,000 Macedonian cavalry, 3,000 Paeonians and Agrianians (probably light infantry), 2,000 Gauls (medium infantry), 3,000 'free Thracians' and 1,000 Odrysian Thracians (medium infantry) along with 1,000 Odrysian cavalry, 3,000 Cretans (missile troops), 500 Aetolians and Boeotians (thureophoroi?) and probably another 500 Greek mercenaries (thureophoroi?) because that is what makes the numbers add up. All told, a force of 39,000 infantry and 4,000 cavalry and the strongest army that Macedon proper had marshaled since Alexander. Perseus' alliance with Cotys IV, king of the Odrysians - the largest polity in Thrace - clearly pays off here, but beyond that most of his troops seem to be mercenaries rather than the product of robust local alliances. Notably, for all of the effort Perseus had spent courting the Greeks in the years before this, he gets very little in the way of useful military force from them and one imagines he probably had to 'overpay' for what meager Greek troops he did get.
Via
Wikimedia Commons, Thracian warriors painted in fresco (c. 300 BC) from the Thracian tombs at Kazanlak. The fellow on the left carries a short, forward curving blade that may be a one-handed
falx (but could also be a
kopis), while we see the top of a polearm above and behind teh infantryman on the right which could be the tip of the Thracian rhomphaia, a distinctive Thracian weapon, a forward-curved blade with a long handle. Thracians had a well-earned reputation as capable fighters.
Clearly, Philip V and Perseus' efforts to strengthen the core Antigonid realm had succeeded, though it is also worth noting that another reason for this relatively large mobilization was actually defeat: the loss of territory in the Second Macedonian War had freed the Antigonid state from the need to garrison those areas. The fact that this increases the effective military power of the state speaks to the limited degree to which the Antigonid kingdom could really have ever drawn military force from many of its conquests. Unlike the Romans in Italy, who had a very effective system for converting conquered peoples into military force, the Antigonids never quite figured out how to turn their control of Greece into a military asset instead of a military liability. To be fair, the Romans also make no attempt to draw meaningful military force out of Greece, leading one to assume that the fractious nature of the Greek poleis and federal leagues, as well as their relative military weakness might have meant that effectively any attempt to actually draw military force out of Greece in this period was likely to be counter-productive.
Such an army was clearly formidable, though as we'll see, Perseus never quite gets all of it on to the field at any one point. On the other hand, the Romans for the year 171 are deployed as follows: Of the consuls, Gaius Cassius Longinus was assigned Italy as his province and spent most of the year bashing up the Gauls in Cisalpine Gaul (the Senate has to stop him from trying to invade Macedonia through Illyria), while Publius Licinius Crassus (Publius, not Marcus) is assigned the province of Macedonia. Of the praetors, C. Caninius Rebilus is in Sicily, L. Canuleius Dives is in Spain, L. Furius Philus is in Sardinia and C. Lucretius Gallus is assigned the fleet to support operations against Macedonia. We also have a few brief pro-magistrate holdovers from the previous year in Spain and Greece, mostly holding until the new magistrates could arrive to release them. All told, the Romans have ten legions under arms in the year, for a rough count of 110,000 soldiers, of whom only 12,600 Romans and 16,800 socii (29,400 total, 1,400 of which were cavalry, the rest infantry) were bound for Macedonia. To stiffen this force, the Senate also permits Crassus to enroll older veteran soldiers and centurions and to appoint (rather than elect) their military tribunes, presumably also to get experienced men in the roles (Livy 42.31.4-5).
From that figure, the immediate strategic problem, which it is not clear if Perseus understood, should be obvious. That Roman army will require Perseus' whole, undivided attention in order to stop (though he does handily outnumber it if he can concentrate his whole army at a point), but it represents less than a third of Rome's total military commitments for the year - and a year that, while it wasn't a low activity year, was also nowhere near a Roman maximum effort either. And the small seeming Roman army could also, of course, count on reinforcements from allies in the region. Chief among these, of course, was Eumenes II of Pergamum, but the Romans also could expect at least some military assistance from both the Achaean and Aetolian Leagues, as well as from the Thessalians (who had been removed from Antigonid control by the Romans in the Second Macedonian War), though it is clear that political sympathies among the Greeks were split.
Consequently, the Romans could pressure Perseus at this level of military engagement effectively forever if they had to. Roman command of the sea - which Perseus was entirely unable to challenge - meant that Roman armies could be brought from Italy into Greece via Apollonia and Perseus could do nothing to stop them. Yet Perseus does not appear anxious to avoid a war with Rome in the run-up to hostilities but is content to adopt a largely defensive strategy, which suggests that he seems to have thought he could win on the defensive, perhaps exhausting Roman efforts, and perhaps in the process reassert the previously dominant Antigonid position in Greece.
Callinicus
Like the Second Macedonian War, this war is one where a lot of the maneuvers are defined by the problems created by the mountainous topography of the region, however the area of operations has shifted and so we are concerned with different mountains and passes than before. Whereas the problem in the Second Macedonian War had been how to break through the mountains separating the Adriatic coast (and the Roman port at Apollonia) from Thessaly, for the Third Macedonian War, the Romans start with a foothold in Thessaly (the cities of which were nominally independent) and the question is how to slip past Mt. Olympus into Macedon-proper from northern Thessaly, though of course before they can do this the Romans will first have to secure northern Thessaly as a jumping-off point.
Via
Wikipedia, a map of ancient Thessaly. Note especially the locations of Larissa on the central-eastern section along the Peneus River as well as the Perrhaebia, the more mountaineous, less densely settled northern spur of the region.
Military operations begin in earnest in 171. Crassus moves his army successfully across the Adriatic to Apollonia, then cuts through the mountains to break out into the open plains of Thessaly, making his camp a short distance off from Larissa, the chief city of the region. Perseus, meanwhile, had moved south out of Macedon into northern Thessaly and proceeded to secure the northern part of the region, negotiating the surrender of some towns and assaulting others in a series of actions probably meant to secure the approaches to the passes that led up to Macedon and perhaps secure his own base of supply for operations in Thessaly, as Perseus seems, at this point, to have intended to try the Romans in pitched battle in Thessaly. This accomplished, Perseus attempts to draw the Romans out with some raiding and foraging, which succeeds in getting Crassus to move his camp up to Larissa in central Thessaly, as previously mentioned. Larissa sits on the Peneus River which cuts across Thessaly west-to-east, and at this point effectively divides the zones of Roman control, south of the river, and Macedonian control north of it (all of these actions are Livy 42.53-55).
That sets the stage for the first significant battle of the war, the Battle of Callinicus (171), effectively an overgrown skirmish. Perseus moves his army in force towards Larissa, evidently knowing the Romans' general position but not fully informed of their strength or precise disposition (Livy 42.57.8) but clearly looking to provoke at least a skirmish if not a general engagement. A minor and brief skirmish ensues on contact, but evidently not the larger battle Perseus hoped for and so Perseus begins deploying his cavalry and light infantry in the same area each day, hoping to draw the Roman cavalry and light infantry into a fight; Crassus in turn refused the engagements, deploying his own light troops and cavalry but not engaging, in a rather un-Roman-like turn (Livy 42.57.9-12).
Perseus now springs a rather clever trap. Having lulled the Romans with the apparent consistency of his deployments, he moves his camp forward much closer to the Roman camp and moves his troops out earlier than usual, occupying a hill near the Roman camp called Callinicus (thus the name of the battle) before the Romans were prepared to meet him. Crassus now responded by deploying his own cavalry and light troops, though both armies kept their heavy infantry back (the Roman infantry deployed on the ramparts to defend the camp, if necessary). The Roman force was a mix, with Roman and socii cavalry and light troops (including the socii cavalry extraordinarii, a picked group of elite cavalry drawn from the socii), along with Greek allies, with Eumenes' Pergamese cavalry in reserve.
After an initial exchange of missiles (Livy 42.59.1), Perseus' Thracian cavalry charged the Roman right wing and disrupted the socii cavalry deployed there, but didn't yet chase them off (Livy 42.59.2-3), and then Perseus himself hammered the center of the Roman position, scattering Rome's Greek allies (this passage is a bit odd, because Livy put the Greek allies not in the center, but on the left, 42.58.12 and one wonders if perhaps he - or his sources - is doing a bit of patriotic cover for the failure of the socii extraordinarii in the center failing to hold their ground). While the southern Greeks evidently fall apart, the Thessalians - long famed for having the best cavalry in Greek - managed a fighting withdrawal, grouping up with Eumenes' reserve (Livy 42.59.5-6). At this point, Perseus' heavy infantry begins to come up and the king apparently considers making a general attack, but is persuaded not to, instead withdrawing having inflicted quite a bit more damage on the Romans than he took in return (Livy 42.59.8-11).
Roman losses were sharp, with 200 cavalry and 2,000 light infantry lost in the skirmish and another 600 captured, to Perseus' losses of only 60 (Livy 42.60.1); Crassus responded by retreating across the Peneus and the morale of his army soured, as you might imagine, with bitter recriminations (mostly at the Aetolians, who fled first) in the council of war. Perseus then attempts to negotiate a peace, which the Romans refuse; that Perseus and his councillors thought the Romans might be moved to peace with such a tiny skirmish, I think, reflects a failure to understand both the Roman strategic position and also the Roman mentality. For Rome to back down at this point would mean largely surrendering their position in the Eastern Mediterranean and giving other rulers, like the restive Antiochus IV (soon to start his own war against Egypt) a green-light. The Romans would hardly accept such strategic losses because of such a mild bloody-nose and so it is strikingly odd that Perseus and his advisors thought such negotiations might work (Livy 42.62.6-7).
What follows this is a bit of frustration in our sources. The Romans engage in foraging operations, while Perseus moves his camp forward to Mopselus to harass them, with Perseus managing to ambush some of the foraging operations with a small force (Livy 42.65.2-6), which then again leads to a potentially larger engagement. What Livy then reports is first, in his own voice, that Perseus gets the advantage in the skirmish but is too late to bring up his phalanx, leading to a brief engagement primarily of light troops in which Perseus takes just 324 losses, after which Perseus, unable to get his disordered phalanx into fighting formation in time, instead withdraws (Livy 42.66.5-8) and Crassus, "content with his modest success" (contentus modico successu; this is not a complement) lets them. Then Livy notes that "there are those who affirm that there was a great battle on that day, that eight thousand enemies were killed, among whom were the king's officers Sopater and Antipater; that around 2,800 were captured alive and 27 military standards were captured. Nor was it a bloodless victory, more than 4,300 of the consul's army were slain and five standards belonging to the left wing were lost."
What Livy is signalling with that, "there are those who say that..." opening is that he doesn't believe them and frankly, I don't either. While Livy doesn't name those sources, one quite immediately suspects the Late Annalistic tradition (e.g. Valerias Antias) and that the situation Livy is in is that his preferred source (Polybius) has reported merely a skirmish, while his other sources (the annalistic ones) have reported a big battle. Livy thus concludes that the late annalistic sources have invented a victory, perhaps as a patriotic cover for the embarrassment at Callinicus, when in fact the fight was much smaller. I am inclined to agree with Livy's judgement - Polybius is unlikely to have left out such a major battle (the Polybius for this period is lost to us) and Livy is unlikely to have expressed such skepticism about Polybius' report. More to the point, Perseus' actions immediately following this don't imply the kind of retreat that a battle with 8,000 casualties would create.
Instead, Perseus stays at his camp at Mopselus for a few days (Livy 42.67.1), then moves back into Macedon, dispersing his army into winter quarters (at home, where supplies would be secure); he also releases Cotys IV and his Thracians to head home to deal with problems being stirred up by some of Eumenes II's forces in Thrace. This is not the action of a king retreating to his kingdom after a sore defeat and potentially facing pursuit, but of one who had tried to pull the Romans into a major battle on favorable grounds, hadn't quite managed it and had exhausted the campaigning season. Crassus, attempting to salvage something, anything, from a dismal year, attempts to seize the town of Gonnus, which guarded the approach to the Vale of Tempe, the main route into Macedonia (to which we will, in a moment return). Crassus is unable to take the town, but secures a few other minor towns in northern Thessaly (Livy 42.67.6-8), but accomplishes little otherwise, aside from pillaging some Greek towns so unnecessarily that the Senate later ordered the captives restored (Livy Per. 43.2-3). It will thus not surprise you to learn that he is not retained in command the following year.
Getting Into Macedon
Instead, the Roman command for 170 goes to Aulus Hostilius Mancinus. We're less well informed about the actions in 170, because book 43 of Livy, which would have told us about it, isn't entirely complete and so Livy's narrative here is mostly missing. In practice, the Roman problem, now that Thessaly was largely secured, was getting into Macedon proper to force Perseus into battle. That effort, in turn, had one particular obstacle, namely that Mount Olympus was in the way, which is a pretty decent obstacle, all things considered.
One again, I am leaning quite hard on J.N. Morton's dissertation, "Shifting Landscapes, Policies and Morals: a Topographically Driven Analysis of the Roman Wars in Greece from 200 BC to 168 BC" (2017), because the problems the Romans face in 170 and 169 are fundamentally topographical in nature. Some maps here are necessary to understand the issue. I am quite bad at maps, but here I will try my best; there are better maps in Morton's dissertation, which you can find with a bit of googling. Let's start with the big picture here:
I've only marked the traditional core kingdom of Macedon here, but the Antigonid state was larger than this. Still, it gives a good sense of where the Romans intend to go.
And you can begin to see the problem, in that the Romans have to get from Thessaly over some fairly formidable mountains to get into Macedon, no matter what direction they go. Moreover, while Thessaly was a agriculturally rich country which could easily supply Roman armies, the moment those armies move out of central Thessaly, they're moving off of their supply bases and into more thinly populated regions, until they reach the open plains of Macedon, where they could resupply by foraging or by bringing in grain via ship through one of the ports there. But, as you may note, the straight line route from Thessaly to Macedon passes directly over Mount Olympus, the tallest mountain in Greece. Moreoever, the goal here is that southern spur of Macedon, called Pieria, which in turn bottlenecks repeatedly on the way up the coastal road, with the northernmost bottleneck, Pydna, marked here.
As we zoom in, Roman options narrow even further. We can dispense first with the options not taken. The most direct route would be via the Vale of Tempe, following today what is the E75 to the coast, from where one could simply follow the coastal plain up into that southern spur of Macedonia (which is called Pieria). The problem is precisely that this route is obvious and the Tempe Pass cut by the Peneus River is both really narrow at points and also well-blocked by Macedonian garrisons. Alternately, you could try to get cute and move through Epirus, bypassing Thessaly and instead trekking through the central Pindus mountains to surprise the Macedonians from the west. But the Illyrians and Epirotes on the route aren't particularly friendly and the terrain for that kind of effort is unfavorable. That leaves essentially two options: you can take the pass on the north side of Olympus, or the pass on the south side of Olympus. Finally, you could head straight north to Elimea (where Lake Polyfyto is now), more or less following the route of the GR-3/E65 to get out of Thessaly, but then you have simply delayed the problem of where and how to descend into the Macedonian plain. This route is considered by the Romans, and Mancinus may have attempted it, to no avail. That leaves just two options:
You can go north around Mt. Olympus, through the Petra pass; this terrain here is the most challenging, but it drops out out cleanly into central Pieria, potentially behind Perseus. The more tempting route is the Lake Ascuris (now drained, but at the time, there was a lake) Pass, which drops you off at the southern end of Pieria, just above the top of the Tempe Vale, but terrain here, while not as bad as the Petra pass, is not good. Both routes now feature paved roads, but at this early point, for a Roman army with limited local knowledge, either route is going to require local guides to avoid getting lost as one moves through the passes over the mountains. Perseus, meanwhile, can base his army between the two passes, post garrisons on the passes and lookouts on the hills and wait to see which one the Romans try to take and then block them, all while the Romans' supply 'timer' is running down because, remember, they've moved out of central Thessaly and its rich supply areas to do this, and don't have easy access to a coast to naval resupply. And this is exactly what Perseus does.
As mentioned, we're not well informed about the campaigns of 170. Aulus Hostilius Mancinus seems to have skirmished unsuccessfully in Thessaly with Perseus before trying that northern route through Elimeia but Perseus seems to have anticipated his movements and the year ends with the Romans back in Thessaly and Perseus back in Macedon. We're better informed about 169. Mancinus' year ends and he is replaced by Quintus Marcius Philippus, a senior Roman politicians (he had been consul before, in 186) but a checkered war record who then has to navigate this problem. He considers three routes in his council of war: the northern route, the Petra pass and the southern route past Lake Ascuris and repeating Mancinus' route through Elimeia, aiming to keep some optionality, loads up on a month's worth of supplies and heads into the Perrhaebia (that northern-most spur of Thessaly), where he could conceivably take any of the routes. Perseus responds by guarding all three of the routes, splitting his army in four (one garrison for each route, plus a core force Perseus keeps with himself to reinforce whichever direction the Romans go).
Philippus ends up attempting the southern route, past the now-drained Lake Ascuris, aiming to drop out north of the Vale of Tempe. Philippus gets himself into some trouble in the passes - while both generals are working without exact knowledge of the other, Perseus seems to have guessed Philippus' route, as he defended it the most heavily - but manages to get himself out and descends into Pieria, having lost a fair bit of time.
Perseus, his army currently divided, pulls back, retreating with his army and much of the food supply to Pydna, the northernmost bottleneck. But Philippus also has a serious problem: supplies. Morton (op. cit.) tallies the days, but in essence it took Philippus a bit over ten days to get to the Lake Ascuris pass, and then another roughly ten days to get through them, with a lot of time wasted in skirmishing and dealing with the Macedonian force there. So when he drops into the Pieria, he has a fairly limited food supply (as, you will recall, he loaded up a month's supply, which is quite a lot). Philippus, for all that effort, ends up with just enough time to secure his route out by taking Heracleum - the southern bottleneck between the Pieria and the Tempe Vale - and then heads back to Thessaly to resupply and wait out the winter. Another campaign that had mostly fizzled out and Perseus reasserts control over the Pieria, rebasing to Phila at its southern end (just south of Heracleum, for those keeping track of places).
In fact, as Livy notes (44.6.3), the Romans had largely avoided disaster because of Perseus' panic when Philippus finally gets through the mountains. Had Perseus simply gathered his army and tried to hold his garrisons, Philippus would have rapidly run out of food still trapped between Perseus (in the central Pieria) and the Macedonian garrisons in Heracleum and the Vale of Tempe. Perseus could have 'bagged' an entire Roman army here, but loses his nerve and misses his chance.
And if you are thinking, "wait, so Perseus loses his nerve and misses his chance for a major victory at Callinicus, and then again skirmishing later that year, and again now - is this going to be a thing with Perseus?" Yes. It is.
In any case, the Romans are not impressed by Philippus spending the season blundering through a pass, nearly running out of food and ending up mostly back where he started, so he too is not retained in command, with the new consul Lucius Aemilius Paullus assigned the job, although it will take him some time to get to the theater from Rome.
Paullus has at least some advantages: the route through Tempe is now clear, so when he arrives and takes up command of the army (reinforced, as we'll see in a moment), he can move through the Vale directly to Phila, meaning we're now concerned not with the passes but rather with the bottlenecks of Pieria.
Via Wikipedia, a map of the region, with the key bottlenecks noted. At the northern end, Pydna blocks access to central Macedon, while Philia at the southern end controls access to the Vale of Tempe. Heracleum (or Heraklion, more faithful to the Greek) and Dion are the central bottlenecks.
As Paullus moves to Philia, Perseus opts to drop back to the Elpeus River, near Dion, where he constructs a fort on the banks that evidently was quite formidable (Livy 44.32.6-10; Plut. Aem. 13.5). And it is worth pausing here for a moment to try to assess Perseus' plan in all of this. I think it is fairly clear at this point that Perseus doesn't have the nerve to try a pitched battle, at least not intentionally. Instead, he keeps opting for the cautious approach of confronting the Romans behind favorable terrain and trying to stall out their campaigns from the logistics difficulty of operating in the region. And so far, that has kind of worked. As the crow flies, the Romans have, at this point, spent three years moving about 20 miles from northern Thessaly to their tenuous foothold on the southern end of Pieria. Perseus presumably hopes that putting the Romans in an endless series of campaigns where they lose light infantry and cavalry skirmishes before running low on supplies and having to head back to Thessaly will either sap the Roman will to continue or prompt some sort of fatal rashness allowing Perseus to engage in battle on extremely favorable circumstances; Plutarch (Aem. 13.4-5) explicitly attributes to him this sort of delaying strategy. I say 'extremely,' because as Livy notes, Perseus has had the chance to engage on merely very favorable circumstances and retreated instead.
I don't think this strategy could have ever worked long-term. The Romans were willing to put up with far greater and more frustrating campaigns in areas of far less strategic significance. The First Punic War (264-241) was, as the date labels there imply, a twenty-three year long high intensity war. Perhaps more apt a comparison, the Romans will be attempting to seize the Celtiberian stronghold of Numantia (and thereby end Celtiberian resistance in Spain) for twenty-one years, from 154 to 133 BC, albeit with some pauses. It is a similar campaign in that the Romans are struggling with the logistical difficulty of actually applying their army to the target, combined with clever opponents who know the local terrain. I do not think Perseus could actually wait the Romans out, but that may not have been clear to him at the time.
In any event, Perseus' luck has run out, because the new Roman consul, Lucius Aemilius Paullus, is neither rash, nor slow, nor timid, but both energetic and quite thorough. Livy's narrative stresses Paullus' careful attention to logistics, scouting, sentries and communications. At the same time, Paullus does not delay; in his own accounting he takes just fifteen days to complete the campaign (Livy 45.41.4; App. Mac. 19). Paullus is in a hurry. And of course, he has to be: he is going to have the same logistics problems operating here that his predecessors did. Every day has to count.
Paullus first sends scouts to figure out where Perseus is, and on getting news of his fortified camp on the Elpeus, Paullus moves up and camps opposite it. Paullus considers both a frontal attack and an attempt to use the Roman fleet to essentially flank Perseus, but dismisses both: the camp was too fortified for a frontal attack and Paullus perhaps correctly guessed that Perseus already had forces in Thessaloniki to forestall any such use of naval power. Instead, Paullus decides on a clever plan: he'll dispatch one of his military tribunes, Publius Cornelius Scipio Nasica Corculum (not that Scipio Nasica, but his father) with a small force to head back, through Heracleum, where they'd be resupplied by the Roman navy, then loop back around through the Perrhaebia (taking a route past Lake Ascuris) and up through the Petra pass, dropping into the plains of Pieria behind Perseus' fortified position. Nasica's march is brisk; he's supplied with ten days of cooked rations (cocta ciberia; probably bucellatum, Roman hard tack). As a result, he can presumably leave all of the cooking supplies behind and march fast and light with just rations and equipment.
The plan works: Nasica drops into the plain behind Perseus, able to threaten his line of supply. Perseus' army, which we'll tally in a moment, was almost certainly too large to supply fully locally. Consequently, Perseus instead falls back along his supply lines back, setting the stage for a decisive engagement at the last bottleneck: Pydna.
Pydna
And at long last, we come to it: the last great clash between a Hellenistic army and a Roman one. Our sources for this battle are somewhat difficult. On the one hand, we would prefer Livy, but we're reaching the ragged edge of the Livy we have; books 43 (171-169), 44 (169-168) and 45 (168-166) of his Ab Urbe Condita start to have gaps, before the whole thing breaks off with finality. Unfortunately, one of those gaps is a chunky lacuna in book 44 that covers the early stages of the battle. On the other hand, we have Plutarch, whose biography of Lucius Aemilius Paullus includes a much shorter narrative of the battle with no breaks. Both authors are relying substantially on Polybius (lost to us), as well as an account written by Scipio Nasica.
Fortunately for me, Paul Johstono and Michael Taylor recently went through the trouble of untangling the sources for the battle in light of the terrain of the battlefield in an article for Greek, Roman and Byzantine Studies (GRBS, a major journal) that is gloriously open-access, "Reconstructing the Battle of Pydna" (2022). So my narrative follows their reconstruction, which I think is sound.
Paullus' army is somewhat large if relatively normal Roman force. He has two legions as normal, but both are larger than usual, 6,000 men rather than the typical 4,200 (Livy 44.21.8). The socii wings (alae) were probably around 5,000 strong. We know Paullus has 600 Roman cavalry and presumably some larger number of socii cavalry; Johstono and Taylor guess around 900, which would be fairly typical). Of the socii, we know that a cohort of Paeligni (a people from what is today Abruzzo in Italy) was part of the army, probably among the extraordinarii and they'll have a key role in the battle. Quake with fear, for a Paelignian officer has brought a flag.
Paullus also has a grab-bag of allies and auxiliaries. Plutarch mentions 700 Ligurian auxiliaries (Aem. 18.2), which are presumably what remained of the 2,000 recruited in 171 (Livy 42.35.6; these guys are skirmishers and the Romans have been taking nasty losses so far in skirmishes, so this isn't shocking). After the embarrassment at Callinicus, the Romans were reinforced with allies from Africa: 1,000 each of Numidian cavalry and infantry, along with 22 elephants; apart from the elephants, it is unclear what of that force remains by Pydna. We also have the troops from Pergamum, initially 5,000 of them (4,000 infantry, 1,000 cavalry) and Attalus is present with the army, so presumably some number of these fellows are still there, but we get no exact count. Likewise, no exact count of Greek allies, but there were almost certainly some. Finally, we are told some 600 Gallic cavalry auxiliaries had been sent earlier (Livy 42.21.6), but we have no idea if they arrived. Johstono and Taylor figure perhaps 35,000 men total and that seems about right. In the event, the cavalry and special troops won't play a huge role in the battle, so a precise account of them isn't strictly necessary.
Plutarch reports Perseus' army as four thousand cavalry and just short of 40,000 infantry, which Plutarch presents as being basically all of the phalanx (Plut. Aem. 13.4), but we know the core Macedonian phalanx wasn't that large and that some of these fellows were rather more lightly equipped. Notably, a report to the Senate earlier in 168 had put Perseus' strength at only 30,000 men (Livy 44.20.4), so it isn't clear if Perseus had reinforced since then or someone is in error. Perseus' army clearly included the core Macedonian phalanx, but also Thracian auxiliaries, Paeonians, Greek mercenaries, and so on. In either case, our sources are clear Perseus has the somewhat larger army (Livy 44.38.5; Plut. Aem. 16.6), though some of Perseus' troops may still have been detached guarding the coast or passes. In the event, the armies were of comparable size.
Perseus fell back to a blocking position at Pydna, encamped on some convenient hills at a point where the plain narrows, framed by marsh on one side and hills on the other, allowing a tight, flat battlespace in the center ideal for his phalanx. Paullus arrives late in the day after marching and forms up a battle line, probably in the wider section of the plain (Plut. Aem. 17.1-6), but when Perseus doesn't come out and Paullus refuses the entreaties of his younger officers to rashly attack on unfavorable ground, Paullus carefully peals his maniples off one by one to construct the standard Roman fortified camp for the night.
And that night, we get an endearing little Roman episode; Paullus evidently knew (Plut. Aem. 17.9) that there was going to be an lunar eclipse that night - the ancients were pretty good at estimating the timed occurrences of stellar phenomenon even when they didn't know why they happened (and Plutarch actually does know that the moon is passing through the shadow of the earth). Paullus has Gaius Sulpicius Gallus, one of his military tribunes, explain to his soldiers what an eclipse is and that these things are predictable, natural phenomenon (Livy 44.37.5-9), but then just to be sure, observes the traditional Roman rituals anyway: the Romans would bang pots and make loud noises to call back the goddess Luna and her favor (Plut. Aem. 17.8). The Romans banged their pots, Paullus sacrificed to the moon (Plut. Aem. 17.10) and Luna graciously returned.
The mood in the Macedonian camp was less positive: lunar eclipses were thought to signal the fall of kings and there was only one king on this battlefield.
Instead of a planned engagement, the battle was triggered by a skirmish between foraging parties going for water which then spiraled into a major battle. These sort of scouting and foraging parties were fairly typical for ancient armies, as we've discussed. In this case, the terrain plays a role, as the primary water source, the Agios Georgios stream, sat between the two armies and so made a natural point of contention for these light scouting forces. Plutarch notes reports that Aemilius Paullus planned to precipitate a battle this way, but given that Paullus is pretty clearly caught somewhat off-guard as the battle begins, one doubts it (Plut. Aem. 18.1, Livy 44.2-3). In any case, the aforementioned 700 Ligurians, covering a watering party ended up in a skirmish with some Thracian troops, possibly over a loose horse and the battle expanded from there, with Scipio Nasica evidently riding forward to try to observe the Macedonian advance and coordinate (Plut. Aem. 18.5).
The battle erupted quite suddenly; neither the consul nor the king had time to get their armor (19.3 and 19.7-10), which leads to a battle where units are coming into combat as quickly as they can be moved out of camp and drawn up. We've actually already discussed the difficulties of command, information and sight-lines in this specific battle before, so I won't dwell on them here. Both armies seem to have tried to deploy in a more or less standard battle array, with units on the right side (from the Roman perspective) coming into contact first. Notably, both sides also lead with infantry, with the cavalry doing effectively nothing in the battle. This is going to be an engagement decided by a direct, head-on clash of heavy infantry.
As Paullus is getting his legions out of his camp, the Macedonian army likewise moves out, with the Thracian mercenaries in the lead, followed by the mixed Greek and Paeonians, and then the Macedonian elite peltasts and the Agema, followed by the cream of the main Macedonian phalanx, the chalkaspides, 'bronze shields,' moving in a line to form the left flank of the Macedonian army. Paullus' first legion - commanded directly by Paullus himself - forms up opposite of the chalkaspides, with his right wing covered by the camp guard (the praesidium) of picked socii cohorts and the rest of the dextra ala (the right wing of the socii) forming up to his right.
This is a very basic diagram of where the major units are in the battle, but do see Johstono and Taylor (2022) - it's open access! - for much more detailed maps that place the units on the actual terrain at their estimated widths and depths.
The action proper then starts with the agema, leading out, slams into the troops of the praesidium, in particular a cohort of the Paeligni (Romans are still mostly in maniples in 168, but the socii have been in cohorts for some time, see M.J. Taylor, "Tactical Reform in the Late Roman Republic: The View from Italy" Historia 68 (2019)). The Paelignian officer, doing what the Paeligni do best, rallies his troops by seizing their cohort's standard and flinging it into the mass of the Macedonians, so that his troops, to avoid the shame of losing their standard, would push forward (Livy 44.40.7-8). Believe it or not, this is not the only flag-throwing Paelignian praefectus either, this happened before (Livy 25.14.4). So let me offer my advice: if a Paelignian officer throws a flag, banner or other kind of standard at you, run and do not look back.
The Paelignians are clearly hard-pressed, but Paullus seems to have judged, correctly, that they'd keep the agema occupied; the Paelignians, outnumbered and facing the best of the Antigonid's troops fall back and the agema evidently follows them, leaving the chalkaspides isolated for a head-on-head clash with Paullus' first legion. Meanwhile, the second half of the phalanx, the less impressive equipped leukaspides, 'white shields' (who are, contra Sekunda, sarisa-phalanx troops, not thureophoroi) form up to the right of the chalkaspides and thus opposite of Paullus' second legion, which is bring driven forward by Lucius Postumius Albinus.
Meanwhile, the Roman dextra ala (right wing of socii), with the Roman war elephants comes into contact with the Macedonian left, composed of the mixed Paeonians and Greeks, along with the Thracians. Perseus apparently had a specialized anti-elephant corps prepared for this moment but these fellows fail with some intensity, leading Livy to dryly remark, "For just as many new devices of mortal men have strength in words, but put to the test, when they must function, not just have their function explained, they vanish without any effect, so to it was with the elephant-fighters: they were a name without a use" (Livy 44.41.4; and who says Livy is a boring prose stylist!). The dextra ala follows this up and pushes back the Macedonian left, but are still fighting it as the battle continues.
Meanwhile Legio I and the chalkaspides seem to have come into contact next and the legion is initially pushed back (Plut. Aem. 20.1-6). We know from a fragment of Polybius that Paullus was terrified at the onset of the chalkaspides, so things were evidently not going well (Polyb. 29.17). What happens next is decisive; Livy in his fragmented narrative presents this as a somewhat natural result: the Romans attack catervatim, "by companies or troops" (that is, in smaller units, probably maniples) and in so doing, disrupt the phalanx, as opposed to the Paeligni, who advanced as a single line and got beaten back for their efforts (Livy 44.41.7-9). Plutarch presents this as a command decision by Paullus, to engage in "not a single battle, but many separate, successive battles" (Plut. Aem. 20.8), especially an order for each maniple to advance or retreat "in their own time" as it were. As a result, the Roman heavy infantry started to get inside the pikes of the chalkaspides.
And here we can imagine what comes next, thinking about the tactical and equipment advantages the Romans have. The Romans have a bigger shield (the scutum) and heavier armor, along with a somewhat longer multi-purpose sword (the gladius Hispaniensis) and a fighting style designed for one-on-one combats (in formation, of course). Their Macedonian opponents, while being (Polybius notes this), some of the best soldiers in the Mediterranean, are trained for pike tactics in groups, relatively densely formed (with less space for swordplay) and while still heavy infantry, they lack the heavier Roman armor, particularly the mail lorica hamata. Worse yet, the close-combat weapons they do have, the xiphos straight-sword and kopis forward-curving sword, are both of limited use against the mail armor the Romans have. Clearly, the Roman is going to have the 'edge' in that fight and indeed, given what we're told about the length of the battle and the casualties, the phalanx crumbles with remarkable speed in a fairly lopsided slaughter.
Meanwhile, on the Roman left, Legio II hits the leukaspides, perhaps before they were fully formed up, leading to its swift collapse, at which point the whole Macedonian army effectively falls apart (after all, both its left wing and both elements of the phalanx are now collapsing by this point). The whole battle evidently occurred quite quickly, despite its fierceness, with Plutarch noting the battle took less than an hour (Plut. Aem. 22.1).
Perseus - wait for it - panics and flees (Plut. Aem. 23.1-2; Livy 44.42.2-3). He seems to have been formed up with his cavalry, in proper Macedonian fashion, waiting for the moment when the powerful Macedonian cavalry could be used to hammer a key gap or weak point. In the event, he never saw his opening. The agema's successful initial engagement ought to have created such a gap, but it is possible that in the rolling terrain, Perseus could not see it. On the other hand, Perseus gets a reputation in the sources (particularly Livy) as being something of a vacillating coward, so it is possible he simply lost his nerve again. I don't think it would have mattered, with the Romans relatively quickly winning over essentially the entire field.
The Romans then pursued and butchered the Macedonian phalanx, to staggering casualties. Plutarch reports some 25,000 Macedonian dead (Plut. Aem. 21.7), whereas Livy more carefully reports 20,000 KIA, with another 11,000 prisoners (Livy 44.42.7); both are probably exaggerations, but it is clear Macedonian losses were devastating. Roman losses, by contrast, were very light; Plutarch says Nasica reported 80 losses, an otherwise unknown Poseidonius 100; Livy reports, "not more than one hundred died of the victors, and of these the major part was the Paeligni; many more were wounded" (Livy 44.42.8). The report that many Romans were wounded but few killed is striking and I suspect a testament to the relative effectiveness of Roman body armor in the close-in fighting as the phalanx collapsed: effective armor doesn't negate wounds, but it turns debilitating or lethal piercing and cutting wounds into survivable blunt trauma wounds or shallow cuts.
Why Roman Victory in the Third Macedonian War?
The lopsided Roman victory at Pydna effectively ended the Third Macedonian War and the Antigonid dynasty and the Macedonian state. Perseus fled, but there was no chance of mounting real resistance anymore; Plutarch reports that even Perseus' cavalry bodyguard took the chance to melt away into the countryside (Plut. Aem. 23.1-9). Perseus was eventually captured and taken back to Rome to feature in Aemilius Paullus' triumph. The Romans seized the Macedonian treasury, which was brought back to Rome and displayed in the same triumph, with Polybius reporting it amounted to some six thousand talents and Livy that it was 120 million sesterces (Polyb. 18.35.3, Livy 45.40.1, note also Plut. Aem. 33.2-24 and Diod. Sic. 33.8; all of the figures given, as Taylor notes in Soldiers and Silver (2020), 149 come out to around the same figure of c. 30 million drachma).
We'll have a seperate post next week, an 'epilogue,' as it were, discussing the broader impacts of Roman victory here, but instead I want to close this post by focusing on why the Romans won.
The first thing that I think is quite clear here, especially as we've now looked at all four of the big second century Legion-v-Phalanx battles, is that the Roman legion simply has a tactical edge against Hellenistic armies. Polybius, of course, famously notes as much, though scholars are often somewhat skeptical of his schematic explanation (Polyb. 18.28-32). While the popular conception focuses a lot on rough ground (often imagining very rough ground like forests and mountains), the three pitched engagements (Cynoscephalae, Magnesia and Pydna) occur on ground that, while certainly not flat in all cases, isn't extremely rough, nor is it clear that the roughness of the ground was purely decisive in any case.
Instead, both Cynoscephalae and Pydna come down to the flexibility of maniples as independent maneuvering units, with the Romans in the first battle able to redeploy maniples to exploit a gap and in the second to advance maniples seperately to create gaps. Meanwhile at Magnesia, the Roman left was able to engage in a fighting withdrawal in situations were we might expect its formation to crumble, seemingly also because of the relatively flexible manueverability of Roman maniples.
That flexibility becomes decisive because of what I'd argue is a real Roman equipment advantage, which isn't really a technological advantage, per se. The only new technology the Romans are using is mail armor, which, to be fair, is very good and I think quite impactful. But if Roman and Hellenistic iron-working or production methods differ, we can't see it. Instead, it isn't a technological advantage, but a mismatch that gives the Romans the edge. Roman heavy infantry are heavily armored for the attritional, close-combat the legion is built around, by contrast, Macedonian phalangites, while still heavy infantry, were less well protected, because they were supposed to have those long sarisae between them and the enemy and in any case were a pinning force, not a killing force.
Consequently, once the Romans disrupt the Macedonian formation and the fighting becomes more fluid, the Roman equipment set, designed to produce casualties and avoid taking casualties, is in its element, whereas the key pinning element of the Macedonian phalanx, the sarisa, is suddenly useless. We cannot know, but I suspect equipment followed training here and that Macedonian soldiers who found themselves in that killing-fight rather than a pinning-fight were aware they had strayed outside of what they had trained for, which probably had catastrophic implications for morale and cohesion.
Beyond this tactical edge, Aemilius Paullus is also just clearly a better general than Perseus. In a surprise battle that neither anticipated, Perseus first vaciliates and then panics (again), while Paullus keeps his head, keeps his legion and wins his battle. And that comes after fifteen days of bold, decisive but well-planned and not-at-all-reckless movement from Paullus, who moves with an urgency his predecessors lacked. Even Paullus' management of the eclipse shows a general more in tune with the morale of his men and more active. That said, we should be careful: it's clear that both Livy and Plutarch understand Perseus as something of an achetype of a bad king, cowardly, greedy, self-serving and narcissistic, and they may be constructing that archetype to moralize a bit. Still, that Perseus was solidly out-generaled is not hard to establish.
Finally, moving up the levels of analysis, Roman command of the sea remains crucial for getting Roman armies to the area of combat and - once you have a general like Paullus who can coordinate it - keeping them supplied there. Moreover, Roman command of the sea is important because it forces Perseus on the defensive, giving him essentially no true offensive options, though the Antigonid kingdom hardly had the force necessary to invade Italy in any event.
More broadly, I think Perseus' defensive, exhaustion-strategy was effectively doomed from the beginning. Rome deployed 10 legions in 176, 7 in each year from 175-173, 6 in 172, 10 in 171 and 170, 8 in 169, and 10 again in 168, with some of them over-strength that year. In short, Rome maintained deployments of about double Perseus' maximum, all-effort deployment at effectively all times. The Romans could wage Perseus' war of logistics and exhaustion forever, rolling the dice every year until they turned up a commander who could outmaneuver Perseus' army into a decisive battle. Perseus had to win all of the battles; the Romans merely needed to win one of them.
The difference in military power wasn't a pure product of demographics. The Antigonid kingdom had plenty of people. But mobilizing armies is more complex than merely finding warm bodies. Often to be soldiers those men need to be of the right ethnic group, the right social class, or have enough wealth to buy equipment. Alternately the king must pay them. In any case, raising fresh armies takes time, which Perseus did not have after Pydna, though I doubt he had many fresh troops to raise. Mobilizing ancient armies was, instead, a complex product of harnessing resources, labor, men and money in the context of societies with extremely low productivity and low labor specialization (when what you need is lots of specialized labor). For reasons that essentially summarize to, "read my book project when it is done" the Romans had built the more effective machine for turning farms into weapons and farmers into soldiers, wildly more effective than any other similar system in the Mediterranean (only Carthage comes remotely close).
And that larger resource pool, combined with the Roman Republic's political system and its abilityto furnish a seemingly endless supply of 'good enough' workmanlike generals and strong strategic direction from the Senate meant that despite Rome not being obviously larger or richer or more urbanized or more sophisticated or more advanced than the Hellenistic states, none of Rome's wars in the East were particularly close. Indeed, as Polybius notes, "the progress of the Romans, from the beginning, was not due to chance, nor was it automatic, as some among the Greeks chose to believe" (Polyb. 1.63.9).
And that is how you beat a Macedonian sarisa-phalanx.
Romans!
No comments:
Post a Comment