This week on the blog we are starting what is a planned four-part ten-part series (Ia, Ib, IIa, IIb, IIIa, IIIb, IVa, IVb, IVc, V) looking at the twilight of Hellenistic warfare and the triumph of the Roman legion. Our core question is a really common one: why was the Roman legion able to decisively defeat the Hellenistic sarisa-phalanx?
It’s a question that crops up on social media frequently, usually in the form of someone tweeting or posting something along the lines of this:
And of course a lot of folks use that as an opportunity to give joke answers (machine guns! cruise missiles!), but what struck me was that on the original tweet and both my response to it (where I posited the answer was ‘Romans’) many of the responses confidently asserted suggestions or solutions which simply do not match up much at all to the actual battle record of these armies. A lot of it was what I am going to term ‘Total War tactics’ – the sort of very simple tactics (flank them, pummel them with ranged fire) that everyone knows but which rarely decide the outcome of battles, in part because everyone knows them and so armies are well-prepared for them.
At the same time, as we’re going to see the answer to this question has some complexity to it. On the one hand, we may note one clear historical fact: from 200 to 148 BC, the Romans win every single major land engagement fought against a Hellenistic power – most of them lopsidedly so. On the other hand, all of those victories have their own quirks. None is quite a perfect model set-piece battle, as we’ll see later in the series. And moreover, this staggeringly lopsided Roman success was relatively new: Rome had fought Pyrrhus of Epirus’ Hellenistic-style army in three major engagements from 280 to 275 and didn’t decisively win any of them, though none of Pyrrhus’ victories were anything like as decisive as the parade of Roman triumphs during the second century, despite Pyrrhus being regarded in our sources as the finest general of his generation (and we have reason to think he is, in fact, being tactically innovative).
So we’re going to go through this question. First, today, we’re going to look at how the Macedonian phalanx functioned, but also how the Hellenistic armies it was embedded into were structured, because the Macedonian sarisa-phalanx never fights alone. Then we’re going to address some of what I consider myths about the weaknesses of this form of the phalanx when applies to the whole tactical system it functioned it, namely that it was weak against flanking, weak against archery (horse or foot), or that it was merely a pale imitation of the supposedly greater version of the system used by Alexander and Philip.
Then we’re going to look at the Roman army of this period, the manipular legion. While I’d normally assume this was a pretty well-understood military, looking at the responses to the tweets above, it seems pretty clear that most folks don’t actually have a good sense of how the Romans fought in any period.1 So we’re going to look at the manipular legion, its weapons and tactics and consider the peculiarities that might give it an edge on the Hellenistic military system.
The having done that, we’re going to look at the battle record itself, particularly what we know of the Pyrrhic Wars (280-275) – which is, alas, not much – as well as the much better attested and much more decisive run of Roman victory in the second century that eventually lead to the slow but seemingly inexorable death of the Hellenistic military system, trying to tease out along the way both why the Macedonian phalanx (and the larger Hellenistic armies) were so successful for so long and also why they failed so suddenly against the legions of Rome.
Now I have to note, those are the units of this series as I have planned them. I am, however, in the final stages of getting a chapter done for an edited collection and quite close to the deadline (writing was badly delayed by a mix of job market stuff and illness in November and December), so I am likely going to have to break up a lot of these parts over multiple weeks. Rather than leave you with nothing, what I’m planning doing is rolling these out each week in segments and then knitting the whole thing together again at the end when it is done. That does mean this is going to be a bit more ‘work in progress’ than usual, so you will have to bear with me a bit as I may have to go back and edit earlier parts and so on.
So this week, we’re going to look at the Macedonian phalanx itself, whereas next week, we’ll get to what I’d argue is actually more important for this question, which is placing the phalanx in the context of the larger Hellenistic army in the field, which included many other kinds of troops.
But first, if you want to take part in the ACOUP military-tributary complex, you can support me and this project on Patreon! Like the Hellenistic kingdoms, I prefer tribute in coin rather than in kind (please do not send me bulk grain) and also like Hellenistic kingdoms, I will most likely use your tribute to finance military activity, in this case to waste use it to buy swords as well as financing my lavish court lifestyle. 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).
(Bibliography Note: There are a number of recent, public-facing works on this topic, perhaps most notably M. Cole, Legion vs. Phalanx (2018) and C. Matthew, An Invincible Beast (2015). I’m afraid I have substantial issues with both titles and in particular find Matthew’s work deeply flawed to the point that I do not feel I can recommend it. Instead, for a popular treatment of both armies, I suggest P. Connolly, Greece and Rome at War (1981), which, while old, is reliable. On Hellenistic arms and armor, T. Everson, Warfare in Ancient Greece (2005) is a good summary, though not so strong as the Bishop and Coulston is for the Romans. On the armies of the successors, the best work on the Antigonids is M. Hatzopoulos, L’organisation de L’armée Macédonienne sous les Antigonides: Problèmes Anciens et Documents Nouveaux (2001), both available only in French and also quite hard to get a hold of even in French. More accessible, but with some problems (some we will discuss) is Sekunda, The Antigonid Army (2013). On the Seleucids, the standard work remains Bar Kochva, The Seleucid Army (1976), but it is increasingly dated and due for replacement. The Ptolemaic army’s situation is much better (we must be at Raphia), with not one but two quite capable recent monographs. I recommend starting with P. Johstono, The Army of Ptolemaic Egypt, 323-204 BC: An Institutional and Operational History (2020), but also note Fischer-Bovet, Army and Society in Ptolemaic Egypt (2014). Further articles are cited below.)
What Sort of Phalangite?
We should start then with the phalanx. Remember the phalanx? This is a post about phalanxes.
However, we need to distinguish what sort of phalanx because this is not the older hoplite phalanx in two very important ways: first, it is equipped and fights differently, but second it has a very different place in the overall tactical system: the Macedonian phalanx may be the ‘backbone’ of a Hellenistic army, but it is not the decisive arm of the system.
So let’s start with the equipment, formation and fighting style. The older hoplite phalanx was a shield wall, using the large, c. 90cm diamter aspis and a one-handed thrusting spear, the dory. Only the front rank in a formation like this engaged the enemy, with the rear ranks providing replacements should the front hoplites fall as well as a morale force of cohesion by their presence which allowed the formation to hold up under the intense mental stress of combat. But while hoplites notionally covered each other with their shields, they were mostly engaged in what were basically a series of individual combats. As we noted with our bit on shield walls, the spacing here seems to have been wide enough that while the aspis of your neighbor is protecting you in that it occupies physical space that enemy weapons cannot pass through, you are not necessarily hunkered down shoulder-to-shoulder hiding behind your neighbors shield.
The Macedonian or sarisa-phalanx evolves out of this type of combat, but ends up quite different indeed. And this is the point where what should be a sentence or two is going to turn into a long section. The easy version of this section goes like this: the standard Macedonian phalangite (that is, the soldier in the phalanx) carried a sarisa, a two-handed, 5.8m long (about 19ft) pike, along with an aspis, a round shield of c. 75cm carried with a arm and neck strap, a sword as a backup weapon, a helmet and a tube-and-yoke cuirass, probably made out of textile. Officers, who stood in the first rank (the hegemones) wore heavier armor, probably consisting of either a muscle cuirass or a metal reinforced (that is, it has metal scales over parts of it) tube-and-yoke cuirass. I am actually quite confident that sentence is basically right, but I’m going to have to explain every part of it, because in popular treatments, many outdated reconstructions of all of this equipment survive which are wrong. Bear witness, for instance, to the Wikipedia article on the sarisa which gets nearly all of this wrong.
1) The wrong wood, the correct wood is probably ash, not cornel – the one thing Connolly gets wrong on this weapon (but Sekunda, op. cit. gets right).
2) The wrong weight, entirely too heavy. The correct weight should be around 4kg, as Connolly shows.
3) Butt-spikes were not exclusively in bronze. The Vergina/Aigai spike is iron, though the Newscastle butt is bronze (but provenance, ????)
4) They could be anchored in the ground to stop cavalry. This pike is 5.8m long, its balance point (c. 1.6m from the back) held at waist height (c. 1m), so it would be angled up at something like 40 degrees, so anchoring the butt in the ground puts the head of the sarisa some 3.7m (12 feet) in the air – a might bit too high, I may suggest. The point could be brought down substantially if the man was kneeling, which might be workable. More to the point, the only source that suggests this is Lucian, a second century AD satirist (Dial Mort. 27), writing two centuries after this weapon and its formation had ceased to exist; skepticism is advised.
5) We’ll get to shield size, but assuming they all used the 60cm shield is wrong.
6) As noted, I don’t think these weapons were ever used in two parts joined by a tube and also the tube at Vergina/Aigai was in iron. Andronikos is really clear here, it is a talon en fer and a douille en fer. Not sure how that gets messed up.
Sigh. So in detail we must go. Let us begin with the sarisa (or sarissa; Greek uses both spellings). This was the primary weapon of the phalanx, a long pike rather than the hoplite’s one-handed spear (the dory). And we must discuss its structure, including length, because this is a case where a lot of the information in public-facing work on this is based on outdated scholarship, compounded by the fact that the initial reconstructions of the weapon, done by Minor Markle and Manolis Andronikos, were both entirely unworkable and, I think, quite clearly wrong. The key works to actually read are the articles by Peter Connolly and Nicholas Sekunda.2 If you are seeing things which are not working from Connolly and Sekunda, you may safely discard them.
Let’s start with length; one sees a very wide range of lengths for the sarisa, based in part on the ancient sources. Theophrastus (early third century BC) says it was 12 cubits long, Polybius (mid-second century) says it was 14 cubits, while Asclepiodotus (first century AD) says the shortest were 10 cubits, while Polyaenus (second century AD) says that the length was 16 cubits in the late fourth century.3 Two concerns come up immediately: the first is that the last two sources wrote long after no one was using this weapon and as a result are deeply suspect, whereas Theophrastus and Polybius saw it in use. However, the general progression of 12 to 14 to 16 – even though Polyaenus’ word on this point is almost worthless – has led to the suggestion that the sarisa got longer over time, often paired to notions that the Macedonian phalanx became less flexible. That naturally leads into the second question, “how much is a cubit?” which you will recall from our shield-wall article. Connolly, I think, has this clearly right: Polybius is using a military double-cubit that is arms-length (c. 417mm for a single cubit, 834mm for the double), while Theophrastus is certainly using the Athenian cubit (487mm), which means Theophrastus’ sarisa is 5.8m long and Polybius’ sarisa is…5.8m long. The sarisa isn’t getting longer, these two fellows have given us the same measurement in slightly different units. This shaft is then tapered, thinner to the tip, thicker to the butt, to handle the weight; Connolly physically reconstructed these, armed a pike troupe with them, and had the weapon perform as described in the sources, which I why I am so definitively confident he is right. The end product is not the horribly heavy 6-8kg reconstructions of older scholars, but a manageable (but still quite heavy) c. 4kg weapon.
Of all of the things, the one thing we know for certain about the sarisa is that it worked.
Next are the metal components. Here the problem is that Manolis Andronikos, the archaeologist who discovered what remains our only complete set of sarisa-components in the Macedonian royal tombs at Vergina/Aigai managed to misidentify almost every single component (and then poor Minor Markle spent ages trying to figure out how to make the weapon work with the wrong bits in the wrong place; poor fellow). The tip of the weapon is actually tiny, an iron tip made with a hollow mid-ridge massing just 100g, because it is at the end of a very long lever and so must be very light, while the butt of the weapon is a large flanged iron butt (0.8-1.1kg) that provides a counter-weight. Finally, Andronikos proposed that a metal sleeve roughly 20cm in length might have been used to join two halves of wood, allowing the sarisa to be broken down for transport or storage; this subsequently gets reported as fact. But no ancient source reports this about the weapon and no ancient artwork shows a sarisa with a metal sleeve in the middle (and we have a decent amount of ancient artwork with sarisae in them), so I think not.4
Polybius is clear how the weapon was used, being held four cubits (c. 1.6m) from the rear (to provide balance), the points of the first five ranks could project beyond the front man, providing a lethal forward hedge of pike-points.5 As Connolly noted in his tests, while raised, you can maneuver quite well with this weapon, but once the tips are leveled down, the formation cannot readily turn, though it can advance. Connolly noted he was able to get a English Civil War re-enactment group, the Sir Thomas Glemham’s Regiment of the Sealed Knot Society, not merely to do basic maneuvers but “after advancing in formation they broke into a run and charged.” This is not necessarily a laboriously slow formation – once the sarisae are leveled, it cannot turn, but it can move forward at speed.
The shield used by these formations is a modified form of the old hoplite aspis, a round, somewhat dished shield with a wooden core, generally faced in bronze.6 Whereas the hoplite aspis was around 90cm in diameter, the shield of the sarisa-phalanx was smaller. Greek tends to use two words for round shields, aspis and pelte, the former being bigger and the latter being smaller, but they shift over time in confusing ways, leading to mistakes like the one in the Wikipedia snippet above. In the classical period, the aspis was the large hoplite shield, while the pelte was the smaller shield of light, skirmishing troops (peltastai, ‘peltast troops’). In the Hellenistic period, it is clear that the shield of the sarisa-phalanx is called an ‘aspis‘ – these troops are leukaspides, chalkaspides, argyraspides (‘white shields’ ‘bronze shields’ ‘silver shields’ – note the aspides, pl. of aspis in there). This aspis is modestly smaller than the Hoplite aspis, around 75cm or so in diameter; that’s still quite big, but not as big.
Then we have some elite units from this period which get called peltastai but have almost nothing to do with classical period peltastai. Those older peltasts were javelin-equipped light infantry skirmishers. But Hellenistic peltastai seem to be elite units within the phalanx who might carry the sarisa (but perhaps a shorter one) and use a smaller shield which gets called the pelte but is not the pelte of the classical period. Instead, it is built exactly like the Hellenistic aspis – complete with a strap-suspension system suspending it from the shoulder – but is smaller, only around 65cm in diameter. These sarisa-armed peltastai are a bit of a puzzle, though Asclepiodotus (1.2) in describing an ideal Hellenistic army notes that these guys are supposed to be heavier than ‘light’ (psiloi) troops, but lighter than the main phalanx, carrying a smaller shield and a shorter sarisa, so we might understand them as an elite force of infantry perhaps intended to have a bit more mobility than the main body, but still be able to fight in a sarisa-phalanx. They may also have had less body-armor, contributing that the role as elite ‘medium’ infantry with more mobility.7
Finally, our phalangites are armored, though how much and with what becomes really tricky, fast. We have an inscription from Amphipolis8 setting out military regulations for the Antigonid army which notes fines for failure to have the right equipment and requires officers (hegemones, these men would stand in the front rank in fighting formation) to wear either a thorax or a hemithorakion, and for regular soldiers where we might expect body armor, it specifies a kottybos. All of these words have tricky interpretations. A thorax is chest armor (literally just ‘a chest’), most often somewhat rigid armor like a muscle cuirass in bronze or a linothorax in textile (which we generally think means the tube-and-yoke cuirass), but the word is sometimes used of mail as well.9 A hemithorakion is clearly a half-thorax, but what that means is unclear; we have no ancient evidence for the kind of front-plate without back-plate configuration we get in the Middle Ages, so it probably isn’t that. And we just straight up don’t know what a kottybos is, although the etymology seems to suggest some sort of leather or textile object.10
In practice there are basically two working reconstructions out of that evidence. The ‘heavy’ reconstruction11 assumes that what is meant by kottybos is a tube-and-yoke cuirass, and thus the thorax and hemithorakion must mean a muscle cuirass and a metal-reinforced tube-and-yoke cuirass respectively. So you have a metal-armored front line (but not entirely muscle cuirasses by any means) and a tube-and-yoke armored back set of ranks. I would argue the representational evidence tends to favor this; we most often see phalangites associated with tube-and-yoke cuirasses, rarely with muscle cuirasses (but sometimes!) and not often at all in situations where they have the rest of their battle kit (helmet, shield, sarisa) as required for the regular infantry by the inscription but no armor.
Then there is the ‘light’ reconstruction12 which instead reads this to mean that only the front rank had any body armor at all and the back ranks only had what amounted to thick travel cloaks. Somewhat ironically, it would be really convenient for the arguments I make in scholarly venues if Sekunda was right about this…but I honestly don’t think he is. My judgement rebels against the notion that these formations were almost entirely unarmored and I think our other evidence cuts against it.13
Still, even if we take the ‘heavy’ reconstruction here, when it comes to armor, we’re a touch less well armored compared to that older hoplite phalanx. The textile tube-and-yoke cuirass, as far as we can tell, was the cost-cutting ‘cheap’ armor option for hoplites (as compared to more expensive bell- and later muscle-cuirasses in bronze). That actually dovetails with helmets: Hellenistic helmets are lighter and offer less coverage than Archaic and Classical helmets do as well. Now that’s by no means a light formation; the tube-and-yoke cuirass still offers good protection (though scholars currently differ on how to reconstruct it in terms of materials). But of course all of this makes sense: we don’t need to be as heavily armored, because we have our formation.
The Formation
Polybius (18.28-30) actually explains how this formation works in practice, which is really useful to us, because he is a contemporary observer of it. Each soldier is holding the sarisa with their leading hand four cubits (1.6m) from the rear, which as Peter Connolly notes, if you reconstruct the weapon correctly, puts the leading hand basically on the point of balance. That causes the rear of the weapon to project more than a meter behind the user, but still gives them 10 cubits (4.17m, 13.6ft) of projection beyond their leading hand. Since each soldier occupies a space 90cm square in the formation, you can get the leading tip of the fifth sarisa about half a meter in front of the leading soldier in each file. That in turn means the first five ranks can all usefully get their sarisae into battle. You can see the basic layout in both the normal ‘compacted’ formation and the ultra-tight synaspismos formation below (note that it is the ‘compacted’ not the synaspismos formation that is the standard fighting formation):
But there’s more to it than this. As I’ve noted before, the hoplite phalanx was not tactically sophisticated; it was in most cases an ‘unguided missile’ – the general hit ‘fire’ and hoped for the best. By contrast, the Macedonian sarisa-phalanx was capable of more – a lot more. It had to be – these men have to fight in concert in order to be effective.
The first step to that is developing some coordination and that means officers. And the Macedonian phalanx has a lot of officers. The basic unit of the Macedonian phalanx is the syntagma; this isn’t a maneuver unit, but an organizational one, but still we can start here. The syntagma in standard battle formation is quite simple: it is a 16 by 16 block of soldiers (the Macedonian phalanx generally deploys 16 soldiers deep, double the most typical depth of the hoplite phalanx). Each file (that is, a line of soldiers, front to back) is led by a file leader (a lochagos, if he isn’t a more senior sort of officer), who has under his command a file closer (an ouragos) a half-file leader (a hemilochites) and two quarter-file leaders (two enomotarches), along with eleven regular soldiers. Every other spot where you’d have a lochagos, you instead have a dilochites, who commands two files (so is senior to the lochagos to his left) and every other spot you’d have a dilochites, you get a tetrarches who commands four files. Finally you have a taxiarchos who commands half the unit and a syntagmatarches who commands the whole unit.
That’s a LOT of officer, but the number and their position makes a lot of sense when you imagine how this creature needs to function in combat. This is a formation designed to function at a range of variable depths, which is made possible by the nested command system. What to halve the depth and double the width? Then the hemilochites move up to the front, bringing the back half of their file with them. Want to go the other way and double your depth? Easy, the lochagoi just move their files in behind the file of their dilochites. I should note I’m not speculating here on the purpose: Asclepiodotus says this is what this is for (Asclep. 2.1). It’s also probably important for the basic task of forming from column into line, because unlike armies at 6 or 8 man depth which can do the ‘turn right’ trick and army that fights with a standard depth of 16 is going to need men to find their places, which is easier when every man is in a sub-unit of four, led by an office: now you can march with a standard width of four or eight and still easily get into formation.
What these officers do not seem to be doing, however, is independently maneuvering; the syntagma is not a maneuver unit. That said, unlike the hoplite phalanx, which often had almost no maneuver capabilities, the Macedonian phalanx does have maneuver units smaller than the whole line. In Alexander’s army (and thus presumably in Philip’s) these were the taxeis (singular, taxis), units of variable size (around 1,500 or so) made by bolting a bunch of syntagmata together. Our narratives of Alexander’s battles, especially in Arrian, keep track of these units so we can see them clearly (e.g. Arr. Anab. 2.8.4; 3.11.9-10); that these units could maneuver independently was actually crucial at the Battle of Gaugamela both for enabling Alexander’s tactic of refusing his left but also in creating a nearly catastrophic opening gap in the phalanx as two of the left-most taxeis were unable to stay in contact with the rest advancing to their right without losing contact with the supporting troops on their left (Arr. Anab. 3.14.4-5).
These sub-units are less visible in our later sources and you will sometimes see it insisted that the Macedonian phalanx becomes less flexible under Alexander’s successors, but I don’t think this is so (indeed, I am going to argue here comprehensively that arguments that the Hellenistic army was somehow a pale shadow of Alexander’s army are, in fact, wrong: this was not a degraded form). As we’re going to see, Pyrrhus had no problem creating an ‘articulated’ phalanx by putting other units in the ‘joints’ between the units of his phalanx (probably taxeis, but we’re not told, Polyb. 18.28.10; we’ll come back to this ‘articulated’ phalanx later – Pyrrhus isn’t the only fellow with this idea), the Seleucid phalanx at Magnesia could still pull fancy maneuvers like forming square (App. Syr. 35), and Perseus’ army at Pydna clearly consists of distinct ‘phalanxes’ which in turn seem like they were in subdivided units as well.14 We hear in our sources of ‘chiliarchies’ (units of c. 1000, actually 1,024, Asclep. 2.10; that’s four syntagmata) and the troop-counts at Pydna make a lot of sense if you assume both of Perseus’ phalanxes (the leukaspides and chalkaspides) each have 10 chiliarchies in them. Those might be maneuver units as well; the leukaspides and chalkaspides clearly maneuver totally independently and the narrative of the breakup of the phalanx (we’ll get to it) implies some independent maneuver within those lines as well.
Consequently this Macedonian phalanx is not so inflexible as the hoplite phalanx. Individual taxeis can advance or hold position independently. They can also clearly wheel in formation, although only with the sarisae up; once they’re leveled to fight, the formation can only move forward, an interesting limitation Peter Connolly’s experiments15 exposed.
We’ll get into the combat role of this formation a bit more next week, but one thing I want to note here is just how much protection these troops have, because it’s going to help to explain why a lot of ‘Total War tactics’ approaches don’t necessarily work. The 75cm diameter aspis is sufficient to cover each soldier’s entire torso, from the upper legs to the shoulders and of course behind that is likely that textile tube-and-yoke cuirass, and each phalangite has a helmet to protect the head. Moreover, the leveled sarisae out front are going to make it hard for enemies to get close enough to use most contact weapons or to be able to get attack angles with those weapons around the aspis.
Light missile attack – regular javelins and arrows – are no solution either. On low trajectories, most missiles will hit the aspides of the front rank and the legs of the front rank are likely to be protected by greaves, making the areas where an arrow-strike might actually produce a disabling result very small indeed, mostly just the face and a small portion of the upper legs. Attempting to solve this by firing over that more heavily armored front ranks runs into new troubles, both in that the rear ranks also have aspides, but also that all of the pikes are in the way, a feature Polybius actually notes (Polyb. 18.30.3). Arrows are not bullets and an arrow that glances off of a pike shaft is going to lose a lot of its momentum and penetrating power. This surely won’t stop all arrows, but between the sarisae and the armor and the aspides, this is a formation that really is very, very resilient so long as it can keep enemies at pike’s reach and in front of it.16
Now if you are thinking, “ok, then we’ll just flank it’ – next week we’ll be exploring why that is so hard to do when we get to the tactical function of the phalanx as part of a larger army. But before we do that, I want to discuss…
Who is in the Phalanx?
You may have noticed a pattern whereby I keep referring to Hellenistic armies but a Macedonian phalanx. This is intentional. The armies the Macedonian phalanx operated in were composite, multi-ethnic forces effectively from the very beginning. Alexander’s army at Issus, for instance, has Thessalian (along with Macedonian) cavalry, Greek hoplites, light infantry from Thrace, Illyria and the Agrianians (a people from what today would be western Bulgaria), along with archers, some of whom are Cretan (Arr. Anab. 2.9.2-3). That, of course, only becomes more true when we get to Alexander’s successors, who find themselves in control of huge populations of non-Macedonians and non-Greek speakers and who, unsurprisingly, begin incorporating those peoples into their armies. We’ll talk about where these folks fit in next week.
But our sources position the phalanx itself as distinctively Macedonian, often contrasting in later Hellenistic armies the Makedones (Μακεδόνες, ‘Macedonians’) of the phalanx with the many other ethnic contingents of other peoples in those same armies. And I think it is worth actually stopping for a moment to ask where the men in the Macedonian phalanx come from because this actually is going to be one of its weaknesses, in an odd sort of round-about kind of way.
For the armies of Philip II and Alexander III (‘the Great’) this was easy: the men of the phalanx, the pezhetairoi (‘foot companions’) came from Macedon, drawn from the peasantry of the country (while the elite served in the hetairoi, the companion cavalry). And for the later Antigonid kingdom, which maintained control over Macedon proper, that largely remained true. Notably in Antigonid armies, allied Greek-speakers show up outside the phalanx, frequently equipped as lighter thureophoroi, a type of troop we’ll get to next week (e.g. Polyb. 2.65.1-5; Livy 33.14.5), but they are separated out of the phalanx even when “equipped in the Macedonian fashion” as with the thousand Megalopolitans at Sellasia in 222 (Polyb. 2.65.3). For the Antigonids, even Greek-speakers showing up with sarisae were not in the phalanx proper, but grouped with allied heavy infantry. ‘The phalanx’ as a unit rather than a formation was for Macedonians.
For the Seleucids, of course, the matter was immediately more complex. Their empire, which stretched across much of what today would be Syria, Turkey, Iraq and Iran (inter alia), included precisely zero indigenous Macedonian communities. What they did have was both a chunk of Alexander’s army as well as a bunch of new Greek-speaking settlements populated mostly by folks we’d call Greeks but also some Macedonians. Early Seleucid rulers intensified this system, creating new poleis and settler-colonies (called katoikiai) in their domains (taking the land from the locals to do so) and offering land grants (kleroi) to entice Greek and Macedonian settlers to live there on the understanding that having one of these kleroi came with an obligation for military service.17 These settlers in turn were legally understood as ‘Macedonians’ (even if they were, in fact, ethnic Greeks) and so when our sources describe Seleucid armies and list the ‘Macedonians’ (before listing all of the other contingents; e.g. Livy 37.40) they are technically correct that these fellows are legally Macedonians. In turn, service in the phalanx itself seems to have remained restricted to these settlers, locking in its ‘Macedonian’ identity. It doesn’t seem like local subject peoples ever really filter into the ‘Macedonian’ legal category in the Seleucid Empire, though to be fair that may just be a lacuna in our evidence; if our sources for Ptolemies were as thin as those for the Seleucids, we’d think the same thing about them.
Which brings us to the rather more complex picture of the Ptolemies in Egypt.18 The bulk of the manpower used for the Ptolemaic phalanx was generally produced the same was as for the Seleucids: the Ptolemies hired Greek-speaking mercenaries and then paid them in part by settling them on kleroi which in turn locked them in to serving for the Ptolemies. The documentary evidence lets us know a lot more about these settlers and they are from a pretty wide range of places, often from well outside areas the Ptolemies actually controlled. What becomes quite clear here is that ‘Macedonian’ in Egypt was a legal category at best only loosely tied to one’s actual origins. While nine of out every ten soldiers on surviving infantry roles has the legal status of ‘Macedonian,’ when we have evidence for the origins of military settlers, we find actual ethnic Macedonians only around 18%, and Greeks around 29% (with another 12% from Cyrenaica, who we might also count as Greeks, the descendants of Greek settlers there), with large numbers of Thracians (15%; probably Hellenized, Greek-speaking Thracians) and smaller portions of other groups.19
Complicating this picture further, we have another ethnic category in Ptolemaic armies, that of ‘Persians.’ The Ptolemaic army has a place for most Egyptians, they fight as machimoi, get paid less than ‘Macedonians’ and don’t (usually) serve in the phalanx. But then we have soldiers noted as being ‘Persians.’ These are not, clearly, actual ethnic Persians or any sort of Iranian peoples – there really aren’t any of those in Egypt. Instead ‘Persian’ too is a fictional legal status. Under Alexander, there was briefly a program of training local youths (mostly elites) in Macedonian warfare and customs and the products of these programs were called ‘Persians,’ which fits into Alexander’s other failed efforts to try to fuse his Macedonian ruling-class with a broader Persian (and related peoples) ruling class. While those programs get shut down basically everywhere after Alexander’s death – as his successors instead set up ethnic hierarchies where Macedonians (and Greek-speakers) rule and all other people are ruled – it seems like it ran a little longer in Egypt, producing a class of ethnic Egyptians who were Hellenized in their customs and fighting style and legally coded as ‘Persians.’ In the second century, some number of Egyptians also earned ‘Persian’ status through military service and it seems like several generations in, some of those original ‘Persians’ end up reclassified as ‘Macedonians.’20
For such a system, I must quote Johstono’s assessment that, “something as nonsensical as an origin marker unrelated to the marked origin almost must come from the military world.”21
That said, compared to the ‘Macedonians,’ these Hellenized Egyptian ‘Persians’ seem to have made up a relatively small slice of the ‘Macedonian’ Ptolemaic phalanx (something sub-10% or so). So on the one hand, the makeup of the phalanx is a mess of origins, but on the other hand for a state that rules perhaps 3-4m indigenous Egyptians but then has a phalanx which is three-quarters drawn from either European Greeks, Greek settlers in Africa or Hellenized Thracians suggests a pretty strong program of ethnic recruiting aimed at excluding all by a select group of Hellenized Egyptian ‘Persians’ from the most prestigious (and best paid) positions in the army. This was still an exclusionary institution.
And then we get to the late second century and the Fourth Syrian War (219-217) fought between the Ptolemies and the Seleucids. The Ptolemies, on the back foot, recruit Egyptians directly into the phalanx in large numbers (20,000 of them, though these may have built on reforms longer in the making), which our main source, Polybius (5.62ff), connects to the subsequent ‘Great Revolt’ (a major Egyptian revolt against Ptolemaic rule) in c. 207/6. As clear from above, the Ptolemies had always had some Egyptians (‘Persians’) in the phalanx, though never this many and many more Egyptiams (machimoi) in the army in lower-prestige and lower-pay positions. Christelle Fischer-Bovet has argued that it is a mistake to see the great revolt as connected as Polybius does, mostly due to the time gap, but I actually think its fairly plausible to suppose that the troops recruited for Raphia may have supposed that being in the phalanx would mean social and economic rewards matching those of their ‘Macedonian’ comrades and it may have taken time for the revolt to fully brew once they realized those rewards were not forthcoming.22 In either case, the Great Revolt marked the real beginning of Ptolemaic decline and the Ptolemaic phalanx, with or without Egyptians in it, would never really get a major ‘face off’ with Rome, as by the time the Romans really arrive, Egypt is too weak to be a real opponent (though Caesar does smash a Ptolemaic army in 47 as part of an intervention into a Ptolemaic civil war).
We’ll come back to this but one thing I want to tag is how the cultural and ethnic chauvinism of these regimes makes this part of their armies brittle. Now we can largely understand why the Antigonid kingdom – small, poor and with fractious, difficult to control Greek neighbors – was relatively weak. But the Seleucid and Ptolemaic problem is more striking precisely because these were big, wealthy empires. But they were empires built on an assumed ethnic hierarchy in which the fruits of empire went to an ethnic elite drawn from Greece and Macedon. Neither empire was ever willing to truly abandon this system; even after Raphia, Egyptian soldiers in the Ptolemaic army get paid less (and revolt more, what a surprise). Unwilling to treat non-‘Macedonians’ as equals in peacetime, it was dangerous to do so on the battlefield. But there were only so many ‘Macedonian’ (the legal category) settlers to go around – a major defeat could deplete the supply for years, especially since holding these empires by force of arms (rather than by gaining local support) required a lot of military force to be diffused into garrisons. Consequently, whereas Rome and Carthage can roll with punches, a major defeat means that a Hellenistic kingdom has to spend a few years recovering. That’s not a catastrophic problem when facing other Hellenistic states who also struggle to capitalize on victory for the same reason, but against Rome the defect suddenly becomes fatal.
(I should note that this brittleness is not restricted to the successors. Alexander invaded the Persian Empire with one army and left another at home with Antipater, but that was effectively all the force he had. Had one of Alexander’s narrow victories at Granicus, Issus or Gaugamela been a defeat instead, he had no real second force to fall back on (in stark contrast to Darius who can incompetently lose army after army and still live long enough to be assassinated by his own furious nobles). Alexander never has to pay the price for a brittle army because he doesn’t lose, but this goes back to what is going to be one of my major contentions here: Alexander’s army wasn’t necessarily better, it just faced weaker opponents.)
In all three cases, the phalanx remains ‘Macedonian’ although the sense in which it is so is increasingly attenuated into largely fictive legal categories. But the phalanx didn’t exist alone nor was it ever supposed to. The Macedonian phalanx was never intended to be the decisive arm of the army in which it served. Instead it was a key component of a larger Hellenistic army, the remaining parts of which were every bit as important as the phalanx, but both far less ‘Macedonian’ and far less discussed.
And that’s where we’re turning to next week.
- Too much Gladiator, the battle scene of which bears very little relation to how the Romans actually fought.
- So to be clear, that means the useful is P. Connolly, “Experiments with the sarissa” JRMES 11 (2000) and N. Sekunda, “The Sarissa” Acta Universitatis Lodziensis 23 (2001). The parade of outdated scholarship is Andronikos, “Sarissa” BCH 94 (1970); M. Markle, “The Macedonian sarissa” AJA 81 (1977) and “Macedonian arms and tactics” in Macedonia and Greece in Late Classical and Early Hellenistic Times, (1982), P.A. Manti, “The sarissa of the Macedonian infantry” Ancient World 23.2 (1992) and “The Macedonian sarissa again” Ancient World 25.2 (1994), J.R. Mixter, “The length of the Macedonian sarissa” Ancient World 23.2 (1992). These weren’t, to be clear, bad articles, but they are stages of development in our understanding, which are now past.
- Theophrastus HP 3.12.2. Polyb. 18.29.2. Asclepiodotus Tact. 5.1; Polyaenus Strat. 2.29.2. Also Leo Tact. 6.39 and Aelian Tact. 14.2 use Polybius’ figure, probably quoting him.
- Also, what very great fool wants his primary weapon, which is – again – a 5.8m long pike that masses around 4kg to be held together in combat entirely by the tension and friction of a c. 20cm metal sleeve?
- Christopher Matthew, op. cit., argues that Polybius must be wrong because if the weapon is gripped four cubits from the rear, it will foul the rank behind. I find this objection unconvincing because, as noted above and below, Peter Connolly did field drills with a pike troupe using the weapon and it worked. Also, we should be slow to doubt Polybius who probably saw the weapon and its fighting system first hand.
- What follows is drawn from K. Liampi, Makedonische Schild (1998), which is the best sustained study of Hellenistic period shields.
- Sekunda reconstructs them this way, without body armor, in Macedonian Armies after Alexander, (2013). I think that’s plausible, but not certain.
- Greek text is in Hatzopoulos, op. cit.
- Polyb. 30.25.2. Also of scale, Hdt. 9.22, Paus. 1.21.6.
- The derivation assumed to be from κοσύμβη or κόσσυμβος, which are a sort of shepherd’s heavy cloak.
- Favored by Hatzopoulos, Everson and Connolly
- Favored by Sekunda and older scholarship, as well as E. Borza, In the Shadow of Olympus (1990), 204-5, 298-9.
- Representational evidence, but also the report that when Alexander got fresh armor for his army, he burned 25,000 sets of old, worn out armor. Curtius 9.3.21; Diodorus 17.95.4. Alexander does not have 25,000 hegemones, this must be the armor of the general soldiery and if he’s burning it, it must be made of organic materials. I think the correct reading here is that Alexander’s soldiers mostly wore textile tube-and-yoke cuirasses.
- On this last one, see Johstono and Taylor, “Reconstructing the Battle of Pydna” GRBS 62.1 (2022).
- op. cit.
- This is actually one of my complaints about the Gaugamela battle scene from Alexander (2004) – the arrow volley is devastatingly lethal in the film, but we get no sense that this was the case in the sources.
- On this, see Taylor, Soldiers and Silver (2020), 101-104 and Bar Kochva, The Seleucid Army (1976), 59ff
- On this, see Johstono, The Army of Ptolemaic Egypt (2020), Fischer-Bovet, Army and Society in Ptolemaic Egypt (2014) and the compact summary in Taylor, op. cit., 89-94.
- Figures from Johstono, op. cit., 76-95, whose argument I follow here
- Once again, I am followed Johstono, op. cit., here.
- Johstono, op. cit., 105-6.
- I should note I am vastly simplifying a pretty complex argument and the difficult evidence it is based on. I strongly recommend taking a read the treatment in Johstono, op. cit., 227-279, which I think is the most capable study. He stands against, in many of his conclusions, Fischer-Bovet, op. cit., which is also a very capable study and worth reading to get the other side of the argument, though I find I side with Johstono on most of their points of contention.
Caltrops (as an answer to the first question at the beginning of this item…)
Which actually raises a question as to how the sarissophore phalanx would deal with caltrops in face of enemy – compared to hoplite phalanx.
A hoplite has omnispear, right? About the length of sports javelin (260 cm), centre of mass in the middle – point 130 cm in front of handle. With 90 cm intervals, it means the second rank´s spearpoint is 40 cm in front of first rank. Little space to bend down in front of the line.
The sarissiphore´s sarissa is 580 cm, with centre of mass 160 cm from rear… and point thus 420 cm in front of handle. With 90 cm intervals, it means the second rank´s spearpoint is 330 cm in front of first rank.
Could the hegemones bend down under cover of aspis picking the caltrops and still move ahead while receiving active defence from the second rank´s sarissae?
Most pre-gunpowder infantry simply doesn’t have a valid way to engage an enemy that isn’t actively trying to engage them back.
Then where do the losses of lost battles come from?
Was the heavy infantry, encumbered by their shields and long spears, capable of pursuing defeated infantry who had thrown away theirs? Or did the losers have a safe way of disengaging without further losses in case the winners lacked sufficient cavalry to run down running infantry?
Light infantry would be able to perform a pursuit role, albeit not as well as cavalry.
I think Hannibal refused battle many times when he was locally outmatched, preferring to instead loot Italy while waiting for the Romans to overextend themselves.
As for outright routs, I assume you had an immediate problem with having left all your stuff in the baggage train. So even if you get out, maybe you get captured the next day when you try to steal food from a nearby village or when you get lost and randomly run into the wrong people.
“Was the heavy infantry, encumbered by their shields and long spears, capable of pursuing defeated infantry who had thrown away theirs?”
Yes. Heavy infantry doesn’t mean each individual was slow, plodding, and so weighed down by his arms and armor that he could barely move. The idea that a knight knocked off his horse couldn’t stand up is pure Hollywood; I’ve seen guys in reproduction Maximilian plate do cartwheels. And if a knight in Gothic plate could do it, any ancient soldier (who necessarily wore less armor) could. I mean, they wouldn’t be winning any sprint heats, but nor were they confined to a snail’s pace.
What makes units slow is the “unit” part. You can only go as fast as your slowest person, or your formation staggers, which results in weak points your enemy can easily exploit. But if your enemy is running away in a route, and especially if they’ve tossed their gear, they’re not in a position to do so, which allows you to be more careless about maintaining formation. And there’s ample reason to do so, as a routed enemy given time to calm down and regroup can easily return to attack again.
Also, remember that running away in panic is not exactly a clean process. You’re going to trip over things (like people, shields, spears, parts of people, the ground, etc), you’re going to run into your fellow soldiers, you’re going to not notice things like tree branches or gullies or the like. Anyone who’s ever had to control a crowd during an emergency knows: Panic makes people stupid. Even flat fields are not smooth; without heavy plows they’d be smoother than modern fields, but there’d be plenty of things to trip over. And once you fall, with a bunch of angry people with weapons coming at you, you’re pretty much dead (or, if you’re valuable, captured).
A rout isn’t “everyone in the unit instantaneously turns and flees, while simultaneously discarding all their heavy weapons and armour”. It’s more like a traffic jam. If you have open space you can run into it, if you don’t (say there’s someone else in the way) you have to slow down or move around – now you’re probably blocking someone else, and so on. Mud, corpses, discarded weapons are all trip hazards that might knock you down, or someone in front of you who you now trip over or have to slow down for.
Throughout that whole messy early stage, heavy infantry in pursuit can easily catch up with their fleeing opponents. Over a short sprint the amount of weight being carried doesn’t make that much difference to raw speed – certainly not enough to overcome this traffic jam problem.
Eventually that’ll all sort itself out, of course, and further pursuit at that point is unlikely to cause many more casualties. That’s where it becomes important to have lighter troops or (ideally) cavalry, who can maintain pursuit once the enemy are in full flight. And the other advantage of having other parts of your force do pursuit is that your heavy infantry can remain in good order, instead of becoming disordered and in turn vulnerable to counterattack.
… which always confuses me when I read this stuff. It makes it seem like there was some kind of rule that these armies always had to find a great big flat open field and have a particular kind of battle on it.
Why didn’t they ever fight in other terrain? Especially if I’m a defender trying to keep an attacker from “delivering the siege”, it seems I ought to be able to choose, or even create, some place where they have to pass through something *other* than a flat field. Most of the world isn’t flat fields. If their whole system only works in wide open spaces, why wouldn’t I deny them that?
In a pre-industrial society, nearly all the valuable stuff worth fighting over *is* flat open agricultural fields. A castle is built to protect the local farmers. A big pre-industrial city simply cannot feed itself without a lot of flat open fields nearby. See our pedantic host’s criticism of the LOTR movie:
https://acoup.blog/2019/05/24/collections-the-siege-of-gondor-part-iii-having-fun-storming-the-city/
Ancient / medieval commanders would deny enemy armies flat ground if they could. Thermopylae was a natural choke point. The city of Tyre gave Alexander the Great enormous trouble because it was an offshore island.
Mostly, though, putting your army in a place that isn’t flat open fields just means your enemy will say “sure, you can keep that” and continue on to the more valuable bits of real estate. And if you pick a really desolate and barren stronghold, how are you going to feed your own citizens and army?
https://acoup.blog/2022/07/29/collections-logistics-how-did-they-do-it-part-ii-foraging/
Because battles require consent. If an army doesn’t want a battle, it camps on tactically advantageous ground and fortifies it additionally; or it doesn’t wait around for your army to ponderously form up to engage it; or it forms up itself when your army attempts to attack without doing so, and you are obliterated.
One army can pressure another, slowly asphyxiate it by restricting its foraging through skirmishing and suply-line disruption and diplomatic intervention; and it can block another from taking certain routes, force it to retreat or engage or find another way. But none of that can reliably force an engagement on grounds that the opposing force finds disfavourable.
They did actively fight on more rough terrain. The myth of flat field battles is a Hollywood invention to make the spectacle of battles and delivering them to a screen an easier viewing experience. Two of the battles cited above, Pydna and Magnesia, were fought in hilly terrain and very much so to the detriment of the phalangites.
Magnesia is reasonably flat, but bounded by water on one side.
You’re thinking of Cynoscephalae (and Pydna).
I think mainly because almost any large army (and any hoplite or horse-dependent army) also works best on open ground, both for movement and for command and control.
There certainly are holding actions in advantageous terrain (eg Ariobarzanes holds Alexander for a month with <2000 men at the Persian Gates) and maneuver to engage the enemies weak elements (picking off siege trains, likely lots of unmentioned patrols going missing) to avoid siege delivery, but in general I think insurgent warfare was just a lot less effective pre-gunpowder given the shorter ranges involved.
Unfortunately the replies to you do not have reply links, so I cannot reply to them.
Most of the world isn´t flat fields. It is true that flat fields are the part of the world worth fighting over; but since they make up a minority of the world, it is important whether they are continuous. They arent´t. Getting from one field to another requires passing through other kinds of land.
The alternative with more total area is hills. Hills, however, may be continuous or discontinuous, various width.
Another alternative is stream valleys, and these because of basic physics of water flow, are normally continuous with a few exceptions like karst.
How common were ford battles? How did hoplites or phalangites set up their force to hold a ford? To force a ford?
In the nature of things, the defender will be trying to prevent the attacker from delivering the siege to a place of some strategic significance i.e. a place X from which you can easily access some important part of the world. That means there must be some easy way of getting to place X.
Put another way, an army following the road network is mostly going to be following flat areas.
Part of it is simply that the way pre-gunpowder (and even early gunpowder) works cohesion is key: In modern warfare a small group of men can carry enough firepower to do significant damage, in an ancient context that’s not really the case the same way. You need the mutually supporting elements of an relatively large group of men to do real damage, and this leads to an escalation (IE: You bringa large group, someone else brings a larger group, etc. until you’ve got thousands of men)
This means that rough terrain becomes a lot less of a tactical advantage, you can’t hide and send out lethal volleys of lead like gunpowder armies can and while choke points or simply breaking up a formation can be important it’s mostly a levelling issue: You can make enemy armies as tactically small as yours.l
So you see some ambushing, etc. but it’s mostly small forces tryingt to pick off even smaller forces. Because you need a big army to defeat another big army, a big army can only really fight *as* a big army on relatively flat ground, and you need that big army to besiege the enemy city (which is why you are here in the first place)
In the context of ford battles, two major battles of Alexander were the battles of Granicus and Hydaspes. How did heavy foot perform there?
They did actively fight on more rough terrain. The myth of flat field battles is a Hollywood invention to make the spectacle of battles and delivering them to a screen an easier viewing experience. Two of the battles cited above, Pydna and Magnesia, were fought in hilly terrain and very much so to the detriment of the phalangites.
Assuming you mean Cynoscephalae, not Magnesia, the terrain was actually to the substantial benefit of the phalangites — they had no problems with its roughness, and were rather able to take advantage of the topography to launch a devastating downhill charge on the Roman left.
@chornedsnorkack
Hydaspes wasn’t a battle to cross a river, as Alexander succeeded in using deception to get part of his army across the river some distance away and then marched to attack Porus.
At the Granicus the Macedonian phalanx did OK. (Using the two of three reports that say the army of Alexander had to fight to cross the river, not unopposed.) They pushed back and broke the Persian infantry defending the riverbank.
One strange aspect of Granicus is that the Persians did not use their mercenary Greek hoplites to defend the riverbank, instead keeping them further back where they were defeated by the combined Macedonian cavalry and infantry after they’d crossed and formed up. Modern historians suggest that the Persian nobility were jealous and didn’t want to share the glory (shades of Agincourt), or that only the Persian cavalry arrived in time to contest the actual crossing with the hoplites arriving later.
“Caltrops” is an answer.
But caltrops are a pretty obscure tactic, and as far as we know, not actually used very much. Why? If you look closer you will see why they are not the answer.
So is that Bar Kochva guy another case of nominative determinism?
If Alexander had lived longer would his integration program have worked producing less brittle armies?
I feel like the fact that his successors were his high-ranking supporters in life yet disbanded the program indicates it didn’t have much social traction, and would likely have fallen apart whenever he died.
Though if he lived to like eighty and kept the program going and was enough of a backer its graduates could have risen to positions of power and gotten socially enmeshed enough to hold their positions, and of course his heir would likely be a son who he might have raised to respect the integration program.
Especially if he promoted Persophiles like Pukestes, or guys like Selukos and Eumenes at the expense of the more extreme ethnic chauvinists.
It seems to me that Alexanders system relied upon himself and his victories as being the sole source of authority. When he died there was no system left to be loyal to, and no heir to obey. If he had lived long enough to found a dynasty, then people might have been loyal to the dynast of the day, so long as the dynasty lasted.
But I don’t see a reason why any such system could outlast the dynasty it was built around.
It’s possible the integration would have succeeded and people would have considered the “Persians” and “Macedonians” to be indistinguishable and successor states would be ruled by a mix, especially if Alexander and his heirs made a point of putting graduates into a position of authority.
It’s also possible that even with it running for centuries ethnic distinctions would live on and we’d get much the same result but with a larger “Persian” population to draw the secondary heavy infantry from.
Dynastic legitimacy was strong enough that the first half of the wars of the Successors revolved round support for/control of Alexanders heirs. A longer-lived dynasty could have cemented something more collaborative (probably Greek-Persian).
Aaah really excited for this one!
For such a system, I must quote Johstono’s assessment that, “something as nonsensical as an origin marker unrelated to the marked origin almost most come from the military world.”
Was that supposed to be a “must” instead of “most”?
My intuition is that OGH meant “although most came from the military world.”
Yup. Looking up the original quotation: “Something as nonsensical as an origin marker unrelated to the marked origin almost must come from the military world.”
I thought Hellenistic phalanxes were more like professional armies than free men levy? I suppose “Macedonian” colonists could be used as professional soldiers at lower cost, since the state only had to pay them during campaigns? Or have I completely misunderstood?
As for individual equipment, how and by whom was it supplied? If it was a levy system, I imagine that soldiers had to present themselves with their regulation weapons and armor, but in that case I find it hard to understand how men who were relatively wealthy thanks to the land they received for their military service in a society where they were considered ethnically superior could settle for “good enough” armor… but if the state is providing, to whom is intended the regulations about what armour soldiers and officiers get to wear?
From what I remember, the army of Philip II and Alexander was a professional force. The Seleucids had an elite standing force (about 10,000 men in the case of the Seleucids), while the rest of the phalanx (divided into two additional units of 10,000) was called upon during war. Many, perhaps most, of these “levies” were men who had already served their time in the standing army. The Ptolemaics also had a standing force, but it was smaller and more akin to a royal guard.
I think the answer is that they were neither? Military settlers tends to kinda be a halfway place in a lot of ways: In that they’re already setup in a military organization (rather than getting inducted at the point of mustering like levies) the quality of course depends very much on the system.
Well, I’m riveted. Can’t wait for the next part. I don’t even have one of my usual battery of questions, I’m just hooked.
Excellent and informative post. Thank you.
This is a somewhat unrelated question, but in the Alexander Mosaic, why is the focus on Darius and not Alexander? Darius is closer to the centre and clearly in the foreground. Alexander looks small in comparison.
Perhaps because he’s the loser, and so has greater pathos.
Makes sense. The Roman mosaic is presumably a copy of a Greek original, so I assumed that the Greeks would emphasize their own leader. There’s a sense of “a worthy opponent defeated” to the artwork.
I can confirm that the battle scene in Gladiator has had a disproportionate effect on popular perception of ancient warfare. So much so that, even if you intellectually know it’s inaccurate, those images tend to color your thought.
Also, I always assumed that the Romans defeated the phalanx by outmaneuvering and outflanking it (maybe I saw that in a tv documentary once). So, I’m interested to see how right or wrong that impression is.
Bret has previously commented that Roman tactics do not rely heavily on flanking, instead trying to produce a decisive breakthrough in the centre where their best troops are concentrated. So flanking the phalanx may well not be the solution the Romans used.
If I had to guess, which I do, I’d suggest some combination of the following, to varying extents of importance:
1) Unusually heavy armour of the Roman infantry allowed them to reach closer quarters with the phalanx in greater numbers, where the Romans are ideally equipped for single combat and the phalanges are not.
2) Flexible line system of the Roman army somehow disrupting the phalanx formation
3) Use of pila being (a) more effective than most missile weapons and (b) creating openings just before the infantry clash
It sounds like Brett is setting up the premise this was a war of attrition. The phalanx was reserved for Macedonians, while the Romans could field legions from any part of the empire – meaning they had a larger group of potential soldiers to recruit.
So to spoil myself, my argument is going to be that the Romans had a complex set of overlapping advantages at the strategic, operational and tactical level.
Also worth pointing out:
“we may note one clear historical fact: from 200 to 148 BC, the Romans win every single major land engagement fought against a Hellenistic power”
Which amounts to a grand total of three engagements (four if we count Corinth). And of those, at Cynoscephalai the Romans lost to the formed phalanx but won the battle because the phalanx’s efforts were disjointed, at Magnesia the Romans didn’t even confront the formed phalanx until it was broken up, and at Pydna the Romans lost to the formed phalanx but won the battle because the phalanx’s efforts were disjointed. Any explanations along the lines of ‘armour!’ or ‘pila!’ or ‘flanks!’ or even ‘virtus!’ are doomed to failure (the many overlapping strategic, operational and tactical advantages answer though is likely to be a good one, if less dramatic). Plus my own preferred explanation – “events, dear boy, events”.
What I wasn’t sure about with that explanation is that we would therefore expect Rome to be fighting at a numerical advantage against its Hellenistic enemies on the battlefield, which it usually doesn’t: sometimes they’re at par, but often the “Greeks” have a significant numerical advantage and only at Heraclea and Thermopylae that I can think of among the major battles did the Romans significantly outnumber their enemies (and they then lost at Heraclea anyway).
Of course, a general advantage in numbers doesn’t necessarily mean that those numbers are made up of phalangites, but Rome’s armies weren’t exclusively legionaries either in these wars (often with large allied contingents).
So far as I’ve been able to casually determine, legionaries significantly outnumbered phalangites in almost every single battle. Total counts are more variable, as you note.
Because of how premodern logistics works, an advantage in numbers of potential recruits doesn’t necessarily cash out as an advantage in army size at the point of contact. The practical size of the army you send to a war zone isn’t governed by how many men you can recruit, it’s governed by how many men you can pay, ship to the theater of war, and feed upon arrival.
If the density of agricultural supplies means that an army of over forty thousand men cannot reliably feed itself by foraging, then you will be sending an army of forty thousand men.
Where the attritional effects start to matter is in what happens after your forty thousand man army encounters the enemy army, which will also be limited to about forty thousand men for similar reasons.
Suppose that you have two hundred thousand men back home whom you can call up to replace losses to your frontline army, and the enemy only has twenty or thirty thousand. They may win the first battle against you, by having an army of comparable size. They may win the second battle. They are less and less likely to win the third, fourth, and fifth battles.
The majority of Rome’s allies were socii, though, so they were legionaries in all but perhaps name.
The phalanx was reserved for Macedonians, while the Romans could field legions from any part of the empire – meaning they had a larger group of potential soldiers to recruit.
Not quite any part — even counting the socii, the Romans only seem to have fielded legions from their Italian population. On the other hand, this still left them with a bigger recruiting pool than even the Macedonian kingdom, let alone the Seleucids and Ptolemies.
What Bret describes for the action of the phalanx–a heavy mass pointed in a single direction and charging at above-walking-speed with spikes in front and no way to turn or easily stop–this is a war elephant, except that it’s composited together out of humans and harder to hit with missiles. I assume the tactics for foiling phalanx charges will be essentially the same as tactics for foiling elephant charges.
As far as I can see, no: elephants seem to have been foiled by harrassing and surrounding them with light infantry; that doesn’t usually work on line infantry outside of exceptional circumstances, and phalangites don’t seem to be an exception.
It’s a lot easier to panic a couple of elephants than to panic a block of 256 trained soldiers.
1, 2, and 3 are all non-starters, judging by the accounts I’ve seen; there are almost no battles we have record of where legionary infantry beat phalangitic infantry in frontal combat — generally the phalanx ranges from a tough opponent to a straight-up immovable object, and while it’s not seemingly incredibly decisive or deadly, you stand almost no chance of actually winning against it with an equivalent or even somewhat superior force of Roman heavy infantry.
There does not seem to be any technodeterministic advantage of loosely arraying a bunch of people in multiple lines and giving them heavy armour and javelins versus densely arraying a bunch of people in one line and giving them pikes; in fact, if you line up two such groups of equivalent size and face them against eachother, you will most likely find that the first strategy is completely impotent in the face of the second. As is logical: we are talking about humans, not robots, and humans do not push through five rows of heavy lances stabbing repeatedly at them with malicious intent. That is suicide, and all the other group have in exchange for the pikes are javelins too sparse to kill significant numbers and armour optimized for close combat that will never occur.
Why, then, the Roman advantage? Well, for starters, it doesn’t seem anywhere near a certain thing that there really is one — Devereaux stresses how the Romans dominated the battles after Pyrrhus, but that doesn’t seem to reflect any really robust sample; at least, I’ve only been able to find a fairly scant few battles to begin with, even before their circumstantial oddities are factored in: three battles with the Epirotes, two battles with the Antigonids, and two battles with the Seleucids. Of the first three, two were won by the Epirotes with the decisive action coming from cavalry and elephants, and the third was lost of the Epirotes by decisive elephant back-fire; the phalanx seems neither to be a point of failure nor one of success. Of the second two, the first was lost of the Antigonids by the bad luck of what was essentially an unwitting Roman ambush, while the second was lost of the same by the failure of their cavalry to engage. Of the third two, the first was lost of the Seleucids by a successful Roman operational envelopment (acutely echoing the other, more famous battle of Thermopylae), and the second and final was won by the Romans in an honest staight-up battle, where they artfully defeated the Seleucid cavalry and auxiliary infantry in detail.
That’s three to five “normal” battles, depending on how you count the fatal case of elelphant and the fatal case of cowardice/mutiny; two of them were won by the Hellenistic participants. The rest were lost by the Hellenistic participants on account of the failure of their cavalry and elephantry, not their phalanges.
We just went over one. The combination of 1 and 2 caused the Phalanx to overextend in places and let the Roman infantry push through at certain places with their armor, which let them fatally crush the entire formation.
Which makes perfect sense when you break down the tactical leadership and strategy of the Phalanx; either the entire line advanced or it didn’t, so if a staggered formation that flexibly engaged and gave, held, or took ground as appropriate met it, it would naturally open up because there was no “brake” command. Especially if there are natural gaps for some of your men to walk into unimpeded, perhaps encouraged by missile fire induced frustration.
Then some men are effectively flanked because they overextended into the gaps, and then Romans are in your formation and everything is chaos.
Hence why articulated Phalanxes or defensive ones faired better. The Phalanx just isn’t drilled to deal with something as simple as a partially open line, it wasn’t made for granular maneuver. Putting medium infantry in between blocks neatly solves this-if a gap opens your Galatians push into it or cover the flanks of the phalangites going in.
From the syntagma diagram, I’m surprised to see all the officers from the lochagoi upwards in the first rank, if I’m interpreting it correctly. Wouldn’t that make them more likely to be casualties in battle, leaving the phalanx without officers?
The front rank has heavier armor, they’re better trained (and have more experience, since they’re officers) and they’re not actually confronting the people directly in front of them – that’s the job of the people in ranks 2 to 5 of the phalanx.
They’re killing the people *behind* the ones in the front rank. Someone in the opposing front rank who is concentrating on the dude in front of them (the guy in the shiny armor) is being filled full of holes from the other phalangites.
In this era, particularly with infantry, nearly the only way to lead is from the front. After all, if the lochagoi is in the rear and decides the phalanx needs to halt its march and form up ranks, how are they going to convey that? It would be impossible to hear them yelling over the noise of so many men marching. Whereas if they’re the first man in each column, you literally just follow them.
I’d imagine the men in back also can’t see where the phalanx is going.
Even in a relatively small shield wall it’s hard to see more than what’s immediately around you. People trying to kill you typically have your more or less undivided attention! Plus, it’s wildly chaotic at the individual level. There’s bits of wood and metal everywhere, people everywhere, screaming and shouting going on; your senses are completely overwhelmed. Even in mock combat you can add a fairly large dose of adrenaline and other fight-or-flight hormones to the mix. Really messes with your head. Your sense of time and space really distorts.
This can be overcome. You can learn to focus this into a useful direction–use that adrenaline to fight more effectively, to block out pain, that sort of thing. Plus, you can learn to see the order in the apparent chaos, and to tune out some of the immediate goings-on to focus on the larger battle. But that takes training an experience. I know nothing of Alexander’s training, but from what our good host has said, the Classical Greeks considered training to be unnecessary. So the average soldier in such a phalanx would have had a fairly narrow focus.
Whether you can lead better from the front or back depends on your style of leadership. If you’re using dumb-fire missile formations, leading from the front is a good tactic. People can just follow you, and you can direct them by changing directions (to a certain extent). In contrast, if you’re trying to manage the battle, it’s better to be in the back and even better to be at a bit of a rise, to give you a better view of the battlefield. At any rate, you want to be away from the enemy’s weapons so you can concentrate on what you’re going to have your men do.
Minor errors where correctly spelled but wrong word:
> even those Polyaenus’ word on this point is almost worthless
though?
> contributing that the role as elite ‘medium’ infantry
> to the?
> and army that fights with a standard depth of 16
and an? an?
> aimed at excluding all by a select group
all but?
Did you fix the Wikipedia article?
lol not worth it. Trying to fix Wikipedia is like bailing out a sinking ship. That’s not to say it can’t be convenient & useful sometimes, it is, but there’s also chronic systemic issues that won’t be resolved without a complete overhaul of policy & staff.
I urge you to try it. There is a policy that generally new scholarship is preferable to older one, so if you cite these newer sources and are prepared to explain your reasoning I’m sure you’ll succeed
The best way to fix wikipedia is to write your own article saying that it’s wrong and explaining why, not to try to edit wikipedia. Someone else then goes and edits wikipedia with your corrections, linking to you as a source and also linking to the sources you quote from.
The best you’ll get by just going in and editing with better sources is a bothsides “there is a dispute; sources X and Y say this, sources A and B say that”.
I have to say, I appreciate the nod to Alice’s Restaurant. 🙂
Thank you.
I’m not proud. Or tired.
Just out of curiosity: What would the different Patreon tribute contributions be if converted to bulk grain (say, wheat)?
2 minutes googling says:
$3.50/month is 6.5 bushels of wheat, paid annually at harvest.
$8.00/month is 15 bushels of wheat, paid annually at harvest.
So just counting the wheat it sounds like about 10 loaves of bread per day for the lower level.
I hope my math is right.
Love me a phalanx!
Two questions.
I’m currently in the middle of Mike Coles’ “The Bronze Lie”, which I feel like was at some point praised by you either here or on Twitter, but a search of the blog yielded no mention of it. Regardless, given your negative opinion on “Phalanx vs Legion” (a book I’ve avoided and didn’t realise he’d also written until I began Bronze Lie because I’m leary of those kinds of books), what do you think of Bronze Lie?
Secondly – Bronze Lie remarks on the evolution of peltasts in the period 400-350BC into steadily more heavily armed forces that eventually became the phalangites discussed here. Would the (surprising, given I was raised on depictions of phalanxes generally being unwieldy) flexibility of the phalangites have its roots in that “peltast tradition”? (whatever on earth that means.)
I haven’t read The Bronze Lie.
Interesting point about Darius’ military being more resilient than Alexander; I’d not thought of it in those terms before. Am very much looking forwards to the remaining parts of this series!p
You can see the same effect at play in the Peloponnesian War. The defeat at Sphacteria sends Sparta in a panic and negotiations. In comparison Athens sustains defeats that would have crippled the Spartan army over and over again.
Is it your opinion that Sparta would have lost the Peloponnesian War if she had not had Persian subsidy? Was it a fault of Athenian leadership not to take into account Persian hatred for their city?
So the composition of Ptolemaic heavy infantry, as distinct from lighter and lower status forces comes out as:
18% actual Macedonians
41% Greeks naturalized as “Macedonians”, of whom the largest group (12%) is Cyreneans
31% barbarians naturalized as “Macedonians”, of whom the largest group (15%) is Thracians.
10% “Persians”… who? What is the grounds of simply stating that there were “hardly any Iranians” in Egypt? It may well have been true in 343 BC; but whatever became of the Iranian men who conquered Egypt from Nectanebo in 342 and who surrendered to Alexander without resistance in December 332…February 331? Repatriated? Anyway, how many native Egyptians got naturalized as the other barbarians (16% beside Thracians) among Macedonians and how many as Persians?
As for the incapability of Macedonians to coopt the ruled, how distinct was it from Romans?
In High Empire, the prestigious heavy infantry, the legions, were also reserved for Roman citizens. Egyptians were also only eligible to the lower paid auxilia. There were people given Roman citizenship, but seems that in Greek East, Roman citizens were a small minority into 2nd century AD. So when was it really harder for Egyptian to get into well-paid heavy infantry – by naturalization as “Macedonian” in Ptolemaic times or by naturalization as “Roman citizen” in Empire?
As for the incapability of Macedonians to coopt the ruled, how distinct was it from Romans?
In High Empire, the prestigious heavy infantry, the legions, were also reserved for Roman citizens. Egyptians were also only eligible to the lower paid auxilia. There were people given Roman citizenship, but seems that in Greek East, Roman citizens were a small minority into 2nd century AD. So when was it really harder for Egyptian to get into well-paid heavy infantry – by naturalization as “Macedonian” in Ptolemaic times or by naturalization as “Roman citizen” in Empire?
In the early Republic the Romans were quite willing to co-opt the ruled, but these latter were all Italians with a similar culture, so arguably the Macedonians already did something analogous in co-opting Greeks. Once the Romans began expanding outside Italy, their willingness to grant citizenship to their new subjects declined quite a bit. From the earlier post on ancient demography, it seems most scholars would put the population of ancient Greece and Macedon at quite a bit lower than that of ancient Italy, and of course the Greek and Macedonian population of the Seleucids and Ptolemies would have been lower still, so perhaps the main reason why the Roman Republic was so much more resilient was simply that the number of Romans + Italians was simply so much higher than the number of Macedonians + Greeks.
More importantly, the Antigonid state was the only one that actually had large local Macedonian and Greek populations. And it was quite small and weak. For the Seleucids and Ptolemies, the Macedonians were a tiny minority even in the imperial core – very different from the status of legion-eligible social groups in Roman Italy.
This suggests that a much larger Antigonid Empire could have been a much more powerful state than any of the three Hellenistic states.
For instance, if Ipsus had gone the other way, then the Antigonids could well have gone on to conquer all of Cassander and Lysimachus’ realms, but the Seleucid Empire and Ptolomeic Egypt would have survived, meaning an Antigonid Empire covering Greece, Anatolia and Syria, facing what would likely be a near-perpetual alliance of Ptolemeic Egypt and Seleucid Babylonia and Persia.
If you get all the wealth of Anatolia to combine with the manpower of Macedonia and Greece, then that’s a much more powerful and resilient state, especially as much of Anatolia is already Greek and more is becoming so. Syria is not, and is likely to spend the next century or two as a battleground between these three Empires, which is not going to incline it towards wealth, deep Hellenization, or large quantities of available manpower.
That’s just the Byzantine Empire renamed.
There just weren’t enough Macedonians for that mega-Antigonia
One key point here, if I recall correctly, is that one of the benefits of being an auxilia is that you became a Roman citizen after your term of service. And this status extended to your children and (under the usual inheritance rules for citizenship) their descendants.
So if you’re a typical native Egyptian male, YOU would only be able to join as an auxilia, but your children could join the legions. If I’m understanding correctly, this sort of path didn’t exist for native Egyptians under the Ptolemies.
Indeed. And the Roman citizenshipmcould be earned also via service on civilian level. If you were among local magistrates, you would become a citizen quite likely, and these magistrate held annual offices, so the upper class of any Roman ally would become citizens quite fast.
Though I must point that probably, you could also lose that citizenship if you went down on the social scale. Legally, it was impossible, but we can see the clash between de facto and de jure citizenship in the Acts: Apostle Paul is repeatedly in situations where people are a bit uncredulous about his citizenship. Paul is a charismatic, well-educated man, so he is able to convince Romans of his citizenship, but if you happened to be, say, reduced to a day-worker on the fields, your claims of citizenship would probably not have much effect, no matter how well-founded. To be a citizen in the provinces, you probably needed the education and social position to go with it. Later, this was formalised with the system of honestiores and humiliores.
As I understand these things, if Britain had had Rome’s citizenship laws Churchill would have been a British citizen (his father was British, though his mother was American), but Disraeli would not have been (both parents Italian Jewish immigrants). OTOH, if Britain had typical Greek citizenship laws, neither would have had British citizenship.
The point is: Rome might not have been perfectly inclusive, but it might still have been more so than its Hellenistic rivals.
(And of course, how inclusive you can be also depends on how much the other guy wants to be included.)
Only if Britain granted the right of intermarriage to Americans.
In this hypothetical, it seems reasonably likely that they would have done so, given the history between the two nations.
The hostility?
Well, did Roman men have the right of marrying Gaulish women?
Not if they wanted their children to be Roman citizens.
OTOH, if Britain had typical Greek citizenship laws, neither would have had British citizenship.
That is, unless the UK and USA had signed a treaty of intermarriage. Something which I am sure all those British aristocrats who want to marry their surplus children to wealthy American heir(esse)s would push for.
Whoops, I somehow succeeded to once again mess up my italics. Why does that keep happening?
Churchill’s parents married before the heiresses were a big thing on the marriage market. So if that were the cause of the treaty, he would have been born before it. Ex post facto law would have been possible, I suppose.
About the Iranian men who surrendered to Alexander, did he not incorporate them into his army? Hope for continued employment with the new rising winner would have been a powerful motive for switching sides. And, do not negotiations preceding such a surrender usually include some sort of promise of employment or settlement for the losing side. I don’t think Alexander would have disdained trained troops when he had a use for them.
I believe that’s what we would call a Pyrrhic victory.
“Pyrrhic victory” doesn’t mean an indecisive victory. It means that the side who technically won ends up worse off.
See, for example, Herodotus’ account of the battle of Alalia, in which 60 Phocaean ships are said to have fought off a combined Carthaginian and Etruscan fleet twice that size. The smaller fleet “won” only in the sense that the city was preserved and noncombatants were protected from enslavement.
… right, his campaigns are where we get the name from
Just guessing for snorts and giggles but did the quality or Roman armor substantially improve between the struggles against Pyrrhus and the later victories? It seems to me like the differences between chainmail and simpler forms of armor would matter a lot when fighting a pike formation. The pikemen in the back aren’t going to have a good view and it’s gonna be hard to keep the point of such a long pick steady while stabbing. So presumably you’d get a lot of weak and glancing blows which would matter a lot if there were gaps in the armor but wouldn’t have the force to break chainmail. So with good enough armor maybe sufficiently brave troops could just push straight through, create a gap in the pikes with his shield and get to stabbing?
Cynoscephalae: The Romans beat the left wing of the Macedonian army, which hadn’t formed up yet, and then outflanked the right wing, which had formed up and was busy pushing the Roman left back.
Magnesia: The Seleucid phalanx became disordered when their own elephants panicked and started trampling them, enabling the Romans to get to close quarters and beat them.
Pydna: The Romans were initially pushed back onto uneven terrain, whereupon the Macedonian phalanx became disordered and the Romans were able to turn the tide.
In none of the main victories, then, do the Romans seem to have won by just pushing straight through the wall of Macedonian pikes.
The impression given is that the phalanx was invincible unless something, anything, went wrong. Hard not to be reminded of a Wellington quote:
‘(The French in Spain) planned their campaigns just as you might make a splendid piece of harness. It looks very well, and answers very well, until it gets broken, and then you are done for. Now I make my campaigns of ropes. If anything went wrong, I tied a knot and went on.’
“Anything” is probably a bit of an exaggeration — the phalanx was the most successful infantry system until the Romans came along, and I don’t think that would have been the case if it couldn’t deal with setbacks — but the phalanx certainly does seem to have been less resilient than the legion.
That’s an incorrect impression, though a tempting one; let’s rephrase those battles:
Cynoscephalae: the two sides unwittingly meet and rush to mobilize themselves, but the Antigonids had sent out half their heavy infantry to forage and don’t remobilize them in time. The Roman left wing faces the remaining Antigonid troops and starts getting one-sidedly ground down, but elements of the right wing manage to wrap around and encircle them, ending the battle. The decisive factor of the battle was operational luck, not tactical prowess; however what it does tactically demonstrate is a phalangitic force having a steep advantage in frontal confrontation with an equivalent legionary force.
Magnesia: Roman cavalry routed their Seleucid counterparts and then, alongside the legionary infantry, defeated the Seleucid non-phalangitic infantry in detail. The isolated phalanx was heavily outnumbered and enveloped by the enemy, but kept cohesion and began making an orderly retreat. Then the elephants intermixed with the formation panicked under the pressure and destroyed it from within. Every possible system that could fail, did so before the phalanx; everything else was, ultimately, more fragile and more failure-prone. Again, phalangitic infantry punched well above its weight in frontal confrontation.
Pydna: The Antigonid phalanx met the Roman heavy infantry in frontal combat and won; the Antigonid cavalry did not pursue for unclear reasons, and the legionaries consequently regrouped and rounded on the phalangites, now disordered from pursuit and rough terrain, routing them and winning the battle. Again, phalangitic infantry punched well above its weight in frontal confrontation.
Cynoscephalae was a gimmick of a battle, Thermopylae was also a pseudo-operational victory; that leaves the three battles of the Pyrrhic War, Pydna, and Magnesia as “normal” battles where tactical factors were significant. The first two Pyrrhic battles were Hellenistic victories, in which cavalry and elephantry were decisive. The third, Beneventum, was a defeat incurred by elephant malfunction.
Thus, the Romans only won three out of five battles through tactical factors, and in all three battles that they won, the phalanx was not the first point of failure in the opposing Hellenistic army.
Hellenistic armies were complex, specialized mechanisms; Roman armies were blunt, ruggedized tools.
Phalangites were the best shock infantry that the Ancient Mediterranean ever saw, but they were expensive, specialised, and irreplaceable, and that meant that they always relied on other forces to pad out their numbers and to aid them in things outside their specialty.
Legionary forces could never compete in direct combat, but they were cheap and monolithic and relied on nothing but themselves — they didn’t need to be the best to have a chance at victory, they just needed to be better than something, anything, which their Hellenic adversaries relied on.
But that doesn’t mean their strategy was better — as Devereaux noted, the Achaemenids erred more on that side of the spectrum and it very much did not benefit them. Ultimately, it came down to motivation and economy and politics and luck; the Hellenistic machine had good odds but all luck runs out, and when it broke instead of breaking just a few times, the consequences were existential, for entropy holds as unnatural and despicable that which is complex.
If you lose due to bad luck in three out of five battles, that suggests you don’t have a great advantage over your opponent.
And if your opponent doesn’t copy your military system when they could easily do so, that suggests they don’t think you have a great advantage either.
Funny how the better prepared you are, and the more flexible you are in the moment, the luckier you get.
If you lose due to bad luck in three out of five battles, that suggests you don’t have a great advantage over your opponent.
No — Hellenistic armies didn’t. But, importantly, that does not in any logically sound fashion imply that Hellenistic phalangites were at any disadvantage against equivalent Roman heavy infantry.
Funny how the better prepared you are, and the more flexible you are in the moment, the luckier you get.
Winning just over 50% of battles does not betray any extraordinary luck; the fragility of the Hellenistic system was rather in the Romans only needing to do so.
“but they were cheap and monolithic and relied on nothing but themselves”
There’s a lot about this subject I can only guess at but I’m pretty sure that calling the legionaries cheap is dead wrong. They are heavy infantry equipped with more metal armor then most of their rivals and with an extensive kit of non battle gear.
It’s good to be lucky.
It’s lucky to be good.
Being good means you get more out of lucky breaks and less out of unlucky ones. And while it’s said that fortune favors the bold, I prefer to word it as fortune favors the ready
Our pedantic host has diagrams of the opposing formations in
https://acoup.blog/2023/12/15/collections-shield-walls-and-spacing-hollywood-mobs-and-ancient-tactics/
Short version before we get to next week: probably not going to push through, too many points in the way. Even medieval dismounted men-at-arms in 15th CE plate armour had a tough time with Swiss pike phalanxes
“Bibliography Note: There are a number of recent, public-facing works on this topic…”
At the risk of sounding like a broken record, there is also Richard Taylor, The Macedonian Phalanx (2020) which is both more up to date and in the opinion of (ahem) its author, but also of some other people whose opinion is IMHO worth listening to, quite good. Read it! You might be pleasantly surprised.
This is a good summary and (if you care) I’m in agreement with most of it. It avoids all the usual errors (as contained in the quoted Wikipedia article for example), so bravo. Though: “it [the pelta] is built exactly like the Hellenistic aspis – complete with a strap-suspension system suspending it from the shoulder”. I am certain that this shoulder strap is yet another of the phalanx-myths that everybody accepts as truth but for which there is in fact zero evidence. Do you have any evidence I might have missed?
A couple of thoughts on the images:
“Detail from the Alexander Mosaic, showing the sarisae in the background”
That seems a common view (that they are sarissae) but I’m certain (for meanings of ‘certain’ that are applicable to ancient history) it is wrong. These are Persian cavalry spears. Reasons: the spears are mostly opposing Alexander, but those on the right are disordered and turning in flight, as the Persians are. Plus, on the right two spears pass in front of what is clearly a Persian standard. Plus, it would be an odd composition to have the phalanx approaching from behind like that.
“a phalangite from the ‘Great Tomb’ of Lefkadia in Lefkadia, Greece … along with an aspis and a sarisa”
I’m reasonably certain (see above) this is a cavalryman. Reason – mostly the cloak, yellow with purple border = Companion (or Hellenistic equivalent). There is no aspis – the yellow thing on his left arm is his cloak.
Is your problem with the titles “Legion Vs Phalanx” and “An Invincible Beast” that they sound like they should be posts on an X-Men fansite?
Just to remind everyone its “Legionary”, not “Legionnaire”.
Like Rock, nothing beats Elephant!
Poor, predictable Bart. (or Barca)
Somewhere in my library there’s an essay (or at least a reference to an essay) looking at whether or not Alexander’s army could have beaten Wellington’s army at Waterloo. I believe the author suggested, yes, it could. I wish I could remember more of the details!
I would like to read that article. It sounds like it’d be a good laugh.
The Alex vs. Wellington write up was in a book called, “The Origins of War,” by Arthur Ferrill and took up the last third of Chapter Six. The author posited that Alex and his army had the art of war down to such a science that he stood a reasonable chance of succeeding in place of Napoleon at Waterloo. Clash of Arms Games also made a wargame expansion kit for their “La Bataille de Mont St. Jean” board game called “Alexander at Waterloo” which allow wargamers to test out the claim.
I seem to recall that by 1815 Wellington had helped his country conquer most of India, and that Alexander did not conquer most of India. Some might think that reason to doubt that Alexander would succeed against him.
Joking aside, there are reasons that the armies of 1815 were not trained and equipped as those of Alexanders time, and it is not that societies on the brink of the Industrial Revolution were incapable of doing so.
Our host has also pointed out on this blogsite that the armies of antiquity could not grow beyond a certain size, as a too-large army would swamp any road. And that European armies of 1815 grew beyond that size by utilising several roads. A technique that Napoleon and Wellington would have been familiar with, but that Alexander could not have been.
Alexander probably could not have concentrated Napoleons army at Waterloo, much less made better use of it.
Fascinating stuff! The guys on the Ancient Warfare Magazine Podcast always seem to think that the Hellenistic Kingdoms’ phalanx was less effective than Alexander’s, so I’m interested to see your arguments next time. They usually seem to imply that Alexander used the phalanx as an “anvil” for his cavalry to smash the enemy against, while the Hellenistic Kingdoms just slugged it out phalanx against phalanx.
While everyone concentrates on the “Big three” engagements of Legion versus Phalanx. Magnesia, Cynoscephalae and Pydna; it is important to remember that they did meet on other occasions.
At the Battle of Chaeronea* (86BC) The Pontic army’s heavy infantry fought in the style of a Macedonian Phalanx. And the Citizen Levy of Carthage may have fought in the same fashion at the Battle of Bagradus (255BC).
Connolly also suggests that at least some of the “African” troops that Hannibal led into Italy were armed and equipped in the same way.
(Hannibal and Mithridates are also said to have re-organised their troops into the fashion of Roman Legionaries)
*Chaeronea is one of those places that gets fought on a lot. (There have been at least six Battles of Chaeronea)
I was going to ask about that. My understanding was that at least part of the Carthaginian army was organized as a phalanx, but I have no idea how accurate that is or for what timeframe.
One of the things that seems to be a thing is that the Macedonian system relies on a lot of specialized troops: Different types of cavalry, the phalanx, the various other types of troops etc. While the roman legion seems to relatively quickly settle in on a single basic soldier (armed with heavy armour, a sword, pilum and scutum) the velites are kept for a while, but my impression is that the socii pretty much fights the same way as well. (this then gets a bit more complicated in the imperial period with auxillaries)
Basically the legion is a very one-size-fits all kind of tool: But that also means each part of it is relatively easily replaceable (and the roman manpower pool is deep, so they can replace it) The macedonian system relies on a lot of different parts working together, and if one of them breaks the entire thing comes apart.
The Macedonians and the successor states were empires. The multiplicity of different troop types arose naturally, as every part of the empire gave the troop types they had traditionally had.
“Connolly, I think, has this clearly right: Polybius is using a military double-cubit that is arms-length (c. 417mm for a single cubit, 834mm for the double)”
Did should this be Polybius using a military single-cubit?
Having read enough of this blog before, I’m guessing that the Roman victory had more to do with their social organization than the tactical moment to moment. Doesn’t matter if your formation is at least as good as theirs if they can mass twice as many troops
So, I have a stupid question. Since the sarissa is a two-handed weapon, and each of these soldiers also has a shield, how did that work exactly? Did the left hand project beyond the shield rim to grip the pike? How was the shield attached to the left arm, such that it didn’t depend on the left hand to hold it steady? I know about the neck strap, and while that would definitely help with the weight, how did they prevent the shield from rotating on the left arm straps?
Not a stupid question, a good one, and one to which nobody has a definite answer. But broadly, yes the left hand projected beyond the rim. On the Alexander Sarcophagus a couple of figures have their left hands free in this way – they are using them to grab prisoners/victims but they could as well be holding a sarissa. Exactly how the shield is held is not depicted. The only ancient depiction of shield and sarissa in use – the ‘Pergamon plaque’ – doesn’t make it clear where the hand, or the sarissa shaft, go, but do at least show that the shield and sarissa were wielded at the same time (despite the lingering survival of a few strange theories of phalangites with shields on their backs).
The only clue to the carrying arrangement is from Plutarch’s account of the conversion of the Spartan army to a Macedonian phalanx, where he says they changed to holding their shields by a handle (ochane) not an armband (porpax) but exactly what that means is not certain. Presumably, something more flexible that could fit over the arm above the elbow. Most modern reconstructions do this well enough eg Connolly or Matthew, with a second handle around the wrist to hold the shield in place.
“I know about the neck strap” – no you don’t! The neck strap is a modern invention, originating so far as I can tell from Minor Markle. There is no evidence for it at all. Yes some shields could have straps for carrying – but the idea of a special strap for the sarissa phalanx is a modern fantasy. One which, for some reason, has become deeply imbedded in modern phalanx folklore and is accepted without question by most (eg OGH).
Thank you for your reply, that answer is illuminating. But now I’m even more confused: try as I might, I can think of no way to attach a shield to an arm with stability that does not involve the hand holding a stiff handle. Two arm straps, it seems to me, will inevitably rotate around the arm. And if the left hand is holding the shield, how could it also be holding the sarissa?
Are you absolutely certain about the lack of an extra strap?
“Are you absolutely certain about the lack of an extra strap?”
No! Nobody should be absolutely certain about anything in ancient history. What I am saying is that there is no evidence for a strap. Markle based his theory on a coin of Philip II, I believe – that the depicted shield had a carrying strap is undeniable, that this means that all phalangite shields had a strap and that such a strap was necessary for their function is, to put it mildly, a bit of a leap.
Regular hoplite shields generally didn’t have a stiff handle for the hand – the antilabe (hand grip) was often a piece of rope. So it could be done.
Strap proponents really need to demonstrate that the strap is necessary, and how it functions – I don’t see how a flexible strap provides stability but would be happily proved wrong by a rigorous experiment that demonstrated a) that the strap did provide stability and b) that stability was impossible without said strap. Until then the strap remains an interesting possibility, for all that it has been accepted by, it seems, everybody. More ‘outdated scholarship’ and ‘confident positions on points of real uncertainty in the evidence’ I suppose!
I’ve never understood how the single central handgrip of a Roman shield could provide any stability against blows to top or bottom of the shield. And yet, that’s what it had. Nobody suggests the scutum had to have been held rigid by a neck strap.
It’s important not to confuse not being able to imagine how a thing was done for evidence that such a thing could not be done.
So, I’m absolutely no kind of expert, or even proponent, but here’s the problem as I see it. I imagine using both hands for holding a pole, and that would seem to leave only the left forearm to hold and steady a shield. Let’s say they used two such arm straps to do this. The arm is a tubular shape, and therefore straps around the arm would also necessarily be tubular to fit around it. This arrangement would seem to allow the shield to rotate around the around the arm along the vertical dimension. This would be bad because if sufficient force is applied to the upper or lower rim of the shield (say, from someone poking at it with a spear) the upper or lower rim would tend to tip toward the user, possibly even striking the person holding the shield.
One solution might have been to provide a third point of stability–a third strap that attaches to the shield above and below the arm, and then around the neck. As the arm applies tension to the neck strap by pushing outward, it would tend to stabilize shield in place along the vertical dimension.
I have not attempted to recreate this arrangement, nor have I seen it attempted in this way, but in theory this could provide one reason why the Macedonian phalangist might have a shield equipped with a neckstrap.
Someone please tell me why this is nonsense.
(This is a reply to Demarquis’ post below, which doesn’t have a ‘Reply’ link)
“Someone please tell me why this is nonsense.”
Can’t help you with that, but I can point out the obvious problem. How does a strap which is presumably not itself rigid, but flexible leather (or similar), provide any rigidity?
Would it help if you pictured the shield not attached solely to the forearm, but to the upper arm above the elbow, and the forearm about the wrist? With the elbow at 90 degrees, that would seem to me to give some degree of rotational stability.
Worth pointing out also that Peter Connolly, who did reconstruct a shield (though he made it too small), and gave it a strap, thought that the point of the strap was to help take the weight of the sarissa, not to make the shield rigid.
Phalangites might have used a strap, who knows? My point is just that there is no argument provided by reconstruction or theory to suppose that they did, and (most importantly) there is no evidence that they did. Markle proposed (though ‘proposed’ is too weak a word) a strap based on a mint mark on one particular issue of coins of Philip II (other mint marks, alongside non-military things like thunderbolts and barleycorns, include Argive shields and Boeotian shields, though Markle didn’t incorporate them into his phalanx theory). Connolly seems to have based his reconstruction on Markle, plus a mistranslation or misunderstanding of Plutarch, but he also attached the shield to the arm in the usual way. And absolutelyeverybloodybody since then has simply repeated Markle and Connolly as if it was established fact. Just look at the Wikipedia article on the Macedonian phalanx – “The shield, called a telamon, was made of bronze plated wood and was worn hung around the neck so as to free up both hands to wield the sarissa”. This is how bad history is written.
> Can’t help you with that, but I can point out the obvious problem. How does a strap which is presumably not itself rigid, but flexible leather (or similar), provide any rigidity?
The strap provides rigidity (or more accurately, support) through tension.
If the expected problem is for the shield to tip “forwards”, i.e. the top part of the shield falls away from the user’s shoulder and upper torso, then putting a strap in place which attaches to that part of the shield can prevent it.
It would also help take the load of a heavy weapon like a pike – that acts to pull the arm down, and would load the strap in tension as well.
> Would it help if you pictured the shield not attached solely to the forearm, but to the upper arm above the elbow, and the forearm about the wrist?
I have never seen a strapped shield design which puts one strap around the upper arm. At least in medieval and later shields, where we do have a number of of surviving originals, it’s an approach which seems unattested. While we do have evidence for the use of straps over the shoulder/neck in conjunction with other handles/straps on a shield: https://manuscriptminiatures.com/3924/10759
Obviously neither of these points prevents a design like you propose having been used, since they’re different shields used at a different time by different people. But I think it is at least a little bit suggestive.
(Reply to Tea Kew)
Suggestive, sure, and anyone would be perfectly within their rights to suggest that given the requirements of handling shield and two-handed pike, and given the existence 1500 years later of the guige to aid in shield handling, then Macedonian pikemen might have adopted a similar solution. But there is no evidence that they did (OK not zero evidence, Markle’s coin exists), and such evidence as we have (Plutarch) suggests they did not.
I’m less convinced by the ‘rigidity’ argument since that would seem to apply to all sorts of shields used in antiquity and beyond, none of which used a strap. But some experiments (which nobody has ever done, to my knowledge) could show how useful a strap is for this purpose (but not, of course, that a strap was ever actually used).
I’m not arguing for or against any particular way of holding the shield. I’m just arguing against, as our host put it: “confident positions on points of real uncertainty in the evidence or debate amongst scholars without any indication to the reader of this uncertain ground.”
Or as he also says: “To offer the reader a confident answer to a question of real uncertainty is historical malpractice”. But the ‘Macedonian strap’ is a case where most people do not even realise there is any uncertainty, and to be fair there is barely any debate, since Markle’s suggestion has been adopted without question by almost everyone (and real academics really don’t care).
“That’s a LOT of officer” – officers.
“led by an office” – officer
“Hellenistic armies but a Macedonian phalanx” – personal preference, but it feel like the italics should be on the adjective, not the noun? Hence the quotations mark around “Macedonian” in the final paragraph. That fits better, IMO.
“excluding all by a” – but a
Millennia later in the Renaissance we see the reemergence of pike formations, and then specialized infantry carrying weapons such as zweihänder swords and polearms, which had their day until the rapid development of firearms made most melee weapons obsolete. Are there any common lessons between the two eras?
I would also like to learn about that.
I find it interesting that all the highest-ranking commanding officers were in the front rank, where they’d presumably be in the most danger. So, what would happen if a lochagos were killed during a battle? I’m assuming that the ouragos would be too far back to even be sure what was going on, let alone take command of the file. Would the nearest enomotarch take over, or would a neighbouring file leader take command of 2 files, or something else?
Was it common for there to be cascades of promotions following battles, where ouragoi were promoted to replace dead lochagoi, then the hemilochites to replace the ouragoi, and so on?
I don’t know if this answers the question entirely, but I get the impression that the legacy of Alexander the Great leading charges of Companion Cavalry personally made a big cultural impression, so Hellenistic generals really did have to be on the front lines or else look bad.
Also, we do have a record of a battle won in precisely this matter:
https://en.wikipedia.org/wiki/Battle_of_Adasa
Our source is pretty biased for the Jewish side, but 1 Maccabees does report other Seleucid victories and rebel setbacks, so there’s no reason not to take it at its word here. Seems Judas Maccabeus’s plan was just to gank the commander, and the Seleucids indeed just retreated after that happened – no backup plan.
(Although on re-reading this, I realize this was more asking about the mid-rank officers not about the overall commander. But there’s no edit button. Oh well.)
Re: the Ptolemies’ Persians, reminds me of the French Foreign Legion and how actually French enlisted are called “Belgians”
Since this article was linked on Twitter yesterday, as a mild update, back in January-February I did update the Wikipedia “sarissa” article. As for the bronze vs. iron for the tube at Vergina, that was not the old WP article editor’s fault, but rather was Robin Lane Fox’s fault – his 1973 book “Alexander the Great” does indeed claim the tube was bronze. Whether Fox was writing from memory and didn’t double-check the then-latest developments, or just invented a fact from the aether, that was indeed sourced. (The WP article has been updated to use more recent sources, of course.)
I also made some articles on associated figures that popped up – Minor Markle, Edmund Lammert, Nick Sekunda, Karl F. Kinch, and a few others. Minor Markle, even if he was wrong, has a bit of an unfortunate story – he was another academic who had trouble getting tenure, a tale familiar to modern academics, although he did find a long-term lecturer position eventually at least. (Even if it involved moving to Australia!) Unfortunately, the articles on Hellenistic armies / Seleucid armies / Ptolemaic armies all need complete rewrites… aside from a few stray Sekunda / Bar-Kochva additions I and others made, the Seleucid army article is largely sourced to public domain pre-1925 stuff like WW Tarn & primary sources like Livy. Yikes! (Any volunteers out there?)
In the realm of general cries in the dark for help, Nick Sekunda edited a 2011 academic monograph collection called, creatively, “Hellenistic Warfare 1”, but this book apparently had like 10 copies printed. Even requests for an Interlibrary loan basically said “Nope nothin'”, and this is a system that has gotten me some extremely obscure books before. Nobody’s even selling it! There’s one chapter available in a PDF online, but that’s it… anyone have access to it?