This is the third part of our four-part series (I, II) discussing the debates surrounding ancient Greek hoplites and the formation in which they (mostly?) fought, the phalanx. Last week, we looked at how the equipment which defined the hoplite – hoplite (ὁπλίτης), after all, means ‘equipped man’) – and how it weighs on on the debate.

And what I expressed last time is that I found the ‘strong’ versions of both the orthodox and heterodox arguments uncompelling. The notion that the hoplite was effectively an ultra-encumbered turtle who couldn’t fight outside of a close huddle simply doesn’t stand up when comparing hoplite equipment – heavy, but not extremely so, somewhat constrained, but not particularly so – to other historical heavy infantry equipment. At the same time, the heterodox vision, where hoplites are as at home in open-order or fluid skirmishing as they are in the confines of a shield wall doesn’t hold up either.** You can fight that way with hoplite equipment, but the panoply is terribly adapted for it while being very well adapted for the context of a shield wall**, suggesting to me that this was always its primary intended purpose (albeit with a meaningful amount of flexibility built in).

We’re now going to carry those observations forward to discuss tactics. To the degree that the board public understands the hoplite debates, they understand it as a debate over tactics and often reduce it to the question, “did they shove?” But there are quite a few more tactical questions here than simply the question of the nature of the othismos. As with some of the previous questions, a lot of these questions are linked but weakly so, meaning it is possible to a degree to ‘mix and match’ without adopting a position that is incoherent. So we’ll begin by outlining what I view as the main differences here and also some of the significant elements of those positions I see as meaningfully unsatisfactory.

As we’ll see chronology also matters here: while the orthodox school generally imagines hoplite warfare to have emerged all at once (a position we’ve already seen can no longer be sustained given the archaeological evidence), reached tactical maturity in the phalanx relatively quickly and then remained rigid and relatively unchanged until the end of the fifth century, the heterodox school instead argues for a lot more chronological change.

Now, I wanted to do the discussion of tactics in a single post so that we could get into some of the interesting implications for polis society more quickly, but there really are too many moving parts and I realized – at the point where I had run out of most of the week, written 7,000 words and barely gotten through the Archaic – that this post needed to be split. The split is, as a result, horribly awkward.

This week, we’re going to look at the ‘strong’ orthodox hoplite model (and dismiss it) and then at parts of the ‘strong’ heterodox model (which we’ll also find unsatisfying, but not entirely without value), before finally working through what a ‘proto-phalanx‘ of the late 600s or 500s might have looked like, thinking in terms of comparative models and what little evidence we have.

Then next week we’ll turn to the ‘mature’ phalanx of the classical period, looking at how we might imagine it functions – tactics, ‘standard’ depth, role of supporting arms, etc. – along with the broader question of defining what exactly the phalanx is (and why I think a more flexible definition is more useful).

Since we’re leaving the definitional work to next week we’re going to avoid calling much of anything a ‘phalanx’ this week, even though these two posts are fundamnetally about the phalanx. One of the things I view as a real problem in this debate are the hard definitional boundaries imposed by both sides, which derive from an overly rigid vision – Konijnendijk’s ‘Prussians’ again – of how the phalanx functioned. The problem is that while the orthodox insist that anything called a phalanx must fit that rigid (and as we’ll see, quite implausible) model, heterodox scholars often insist that anything that does not fit the model is not a phalanx in order to push the date for ‘the phalanx’ back. In my view it is well past time to let the evidence lead the definition rather than the other way around – the phalanx is what the phalanx does, not how we define it – so we’ll lead with the evidence and revisit the definitional scrum only at the end.

As always, if you like what you are reading, please share it as I rely on word-of-mouth to find readers! And if you really like it, you can support this project over at Patreon; I don’t promise not to use the money to buy a full hoplite panoply, but I also don’t not promise to do that.1 And if you want updates whenever a new post appears, you can click below for email updates or follow me on Twitter and Bluesky for updates when posts go live and my general musings; I have largely shifted over to Bluesky (I maintain some de minimis presence on Twitter), given that it has become a much better place for historical discussion than Twitter.

Let Us Shove Off

As we’ve noted – nearly ad nauseam at this point – the orthodox and heterodox ‘camps’ differ both in their understanding of the chronology by which something called ‘the phalanx’ developed, but also their sense of the mechanics of what something called ‘the phalanx’ was and how it functioned. I think both tactical models are substantially flawed. I should note while putting this together Paul Bardunais linked his own synthesis (presented here in video form) which I hadn’t seen developed in full. It is not exactly my synthesis, but it is actually pretty close (I think it is a perfectly good, defensible, plausible model, which is more than I can say for the ‘strong’ models we’re about to discuss) as we’ll see and it is good to see someone working on a synthesis position.

One crucial difference between the **orthodox and heterodox models of hoplite warfare is that orthodoxy generally imagines a tactically stable **(or stagnant) phalanx: it doesn’t change after emerging and rapidly reaching ‘mature’ form. By contrast, the heterodox model assumes significant development over time. Now I do want to treat the evidence for tactics in the Archaic and Classical periods separately, because as we’ve already seen, I think the heterodox school is fundamentally correct in assuming meaningful change over time, but first I think it is worthwhile to dispense with the orthodox tactical vision, at least in its narrowest form. We ought to do that in the beginning because – since the orthodox view is that the phalanx is tactically stagnant – this model is supposed to be valid in every period. So rather than repeat myself, we can deal with it once here.

The **modern version of orthodox hoplite tactics comes directly from The Western Way of War **and so that is the ‘strong’ version of the model I will focus on here. The orthodox vision is that in a phalanx formation, hoplites were densely spaced (file widths of 45-60cm, shoulder-to-shoulder), they advanced at a run and then collided at speed with the two formations smashing together at full tilt. Then, the orthodox suppose the othismos was a kind of rugby-scrum style shoving match where the formations tried to push through each other (while also striking over and beneath shields) and as gaps and tears formed in the line from this pushing action, one phalanx would fall apart. Such fighting naturally fully excluded light infantry and cavalry. Moreover, as we’ve seen chronologically, the orthodox camp argues this form of warfare developed swiftly in the 8th and early 7th century and remained pure and unchanged from then to the late fifth century, a long period of relatively static hoplite warfare.

That vision exists within a sort of assumed framework, particularly among earlier scholars, as Roel Konijnendijk notes in his book,1 that derives more from early modern gunpowder warfare than from ancient warfare: there is an assumption of rigid command and control, supported by both training in arms (that is practice with weapons as opposed to just fitness training) and drill (that is, practice moving in unison) of a sort that is, bluntly put, not really attested in our sources until the late Classical period (if even then). Victor Davis Hansen’s work, coming later out of the Face of Battle school instead emphasizes the amateur citizen-soldier nature of hoplites (and thus doesn’t really assume lots of drill or practice) but keeps the rigid tactical system.

This vision is, frankly, nuts. No other shield wall behaves this way, shoving in a mass rugby scrum. It is physically possible – these presses have been demonstrated, it will not necessarily crush the men in the middle – but it cuts against human psychology in combat (humans tend not to want to stay in the ‘danger zone’ of enemy weapons – called ‘measure’ – for very long) and more important against the sort of casualty figures we get, which suggest losses for victors in hoplite battles could be relatively low and thus most casualties occurred after the rout.2 If this kind of shoving were normal, we’d expect *knives *and daggers, not spears, to be the weapon of choice (and I should note that while Greek swords are generally on the short side, a xiphos is not a knife or a dagger) and one man with a knife pressed at the front could make a terrible mess very quickly as he can easily stab over the shields of his enemies into the neck from the side where even the Corinthian helmet offers less than perfect protection. Indeed, notably, something like a combat dagger isn’t even a standard element of the hoplite’s kit (rare to see them in artwork) and won’t be a standard piece of equipment in the Eastern Mediterranean until the early Roman imperial period (by which point the Romans have fallen in love with a devilish dagger from Spain they call a pugio).3

Crucially, as heterodox scholars have been pointing out for decades now,** nothing in the source tradition requires us to interpret othismos** (a term that is not used in every or even most hoplite battles!) this literally: plenty of cultures describe ‘presses’ and ‘pushes’ of infantry that are not literal shoving.** At no point does any source clearly describe the othismos as literal shoving**; instead it is used to mean what we might term ‘coming into contact’ or ‘shock’ (e.g. Hdt. 7.225.1, 9.62.2, Xen. Anab. 5.2.17, etc.etc.), that is, two formations moving into melee range, or in the sense of a given ‘push’ of effort to achieve victory – we use the same phrase metaphorically of infantry assaults with guns that don’t involve anyone getting within 50 yards of a shoving match. While we start to see lines of men in Greek artwork, seemingly in close-order, as early as the 650s, we never see obvious scenes of mass shoving or even a lot of ‘combat grappling’ (it is hard to grapple with one hand secure in a two-point grip on a shield).4 It is striking that the orthodox school in its modern incarnation is thus arguing that the primary mode of high-status Greek hoplite warfare – the supposed shoving othismos – is both the core of experience of battle in the late Archaic and Classical Greek world and also never depicted in artwork, not even once. That is simply, to me, an unsustainable reading of the evidence.

I am struck that early modern European artwork furnishes more examples of nearly-scrum-like engagements (see below) involved in the push-of-pike, but even in the most chaotic push-of-pike scenes, soldiers are not shoving but instead have recourse to draw their swords (generally the katzbalger, which at 70-80cm is not very much larger than a xiphos or kopis) and cut with them.

Via Wikipedia, the classic Hans Holbein the Younger scene of a push of pike (early 16th cent.). I should note not every artist depicts these clashes this way – often they do seem to have been ‘poking matches’ at the edge of pike’s reach, but evidently could produce melees of this sort. That said, while we do see some men grappling at very close range with daggers, many still use their pikes or else draw their swords, suggesting there is still enough space, even in this mass, to use such weapons.

One may well imagine that two shield walls coming together may have created a temporary press similar to crowd collapses or rushes that happen sometimes at overcrowded concerts and similar crowded spaces, but there’s no sign this was the intended goal. As we’ll see in a moment, I suspect rival hoplite formations probably did often collide at some speed (though not perhaps intentionally), but if they did, I would expect them to ‘accordion’ back out rather than for the men in the rear to press their friends into the points of enemy spears. Crowd crushes happen because the psychological pressure is urging people in the back to push forward but in combat the psychological pressure is urging everyone to move away from the enemy.

Given how speculative and awkward the ‘shoving’ othismos is (as opposed, as we’ll see, to othismos-as-pulse) it is a bit frustrating that it persists in many reenactment circles, presumably because – as Roel Konijnendijk once suggested to me – it is a reasonably ‘safe’ way to do a hoplite reenactment as opposed to, you know, jabbing sharp weapons at people.

Problems pile up for the orthodox model from there. The very tight shoulder-to-shoulder spacing seems quite clearly to be a product of reasoning from modern musket formations; no shock formation I know of was ever this dense (including early modern pike formations). As we’ll see in a moment, I don’t think the spacing was loose generally (> 100cm file width), but I also do not think it was ultra-tight generally (< 60cm). Since we’re not shoving, after all, we need some space to actually use our shield and weapon (though nowhere near as much space as some heterodox scholars imagine, more on that next week).

Meanwhile, the developmental timeline does not work either: hoplite equipment didn’t emerge suddenly and so the ‘mature’ all-hoplite phalanx couldn’t have done so either. Moreover, as the heterodox will frequently note, light troops and cavalry continue to appear frequently in Archaic artwork and battle scenes, often intermingled with hoplites, suggesting they still have a battlefield role. Tyrtaeus, writing in the mid-7th century describes “You light-armed men, wherever you can aim/from the shield-cover, pelt them with great rocks/and hurl at them your smooth-shaped javelins” (Fr. 11 West, trans. West), which sure implies that the light-armed have a job to do even c. 650 or so and that it involves being at least in the same zip-code as the shield wall of hoplites (since they are aiming “from the shield-cover”). And of course throwing javelins and rocks would hardly be feasible if the two opposing lines were locked in contact in a shoving match, as you’d end up hitting your own fellows as often as the enemy. So this orthodox vision will not do, especially for the Archaic.

So what will work?

The Archaic Phalanx Did Not Pine For the Fjords

Having beaten up quite a lot on the orthodox vision, I think we must now turn and beat up a bit on the heterodox vision, particularly the version developed by Hans van Wees. Now here I want to note that while the orthodox school has effectively a single vision of hoplite combat, the heterodox school can sometimes contain multitudes and so not every ‘heterodox’ scholar shares Hans van Wees’ combat model. However it is also the case that Hans van Wees is also pretty much the only scholar in print to lay out a complete model, so we have to deal with it.

And I want to begin with a fairly big reasoning problem involving some dead birds. Hans van Wees, it must be noted, is coming at the question of Greek warfare chronologically from the ‘other side’ in that his work before Greek Warfare: Myths and Realities (2004) was focused on war and violence in Homer, so he is advancing forward from the early archaic towards the classical rather than reasoning backwards from the classical towards the archaic.

Van Wees presents in Greek Warfare and again in his chapter in Men of Bronze (2013) warfare among the Dani people of the highlands of Western Papua New Guinea as a kind of ‘key’ to understand Homeric warfare and thus early hoplite warfare. He cites for this Gardner and Heider, Gardens of War: life and death in the New Guinea Stone Age (1968), the print publication of this research, but most people, if they are aware of this work will be aware of it through the famous and foundational documentary film made during that research, Dead Birds (1963), also made by Robert Gardner. The film presents an idealized vision of a single battle among the Dani people, a people living with stone-age technology (no metal working) in the highlands of Papua New Guinea, though the footage is actually a pastiche of several battles fitted together. That said, Dead Birds is essentially the only footage we have of a society waging a real life-and-death battle with contact weapons.

This is an important piece of scholarship and a crucial tool in our understanding of warfare in the past and I have been on and on so far about how I think the study of hoplite warfare would benefit from comparative evidence so you may be expecting me to praise the use of this material as a tool for understanding Greek warfare, but I cannot.

Van Wees clearly reads this warfare – and perhaps, though he does not cite it, watches the film – and sees in it things Homer is describing (remember, he is coming at this originally as a Homerist): initially massed ranks that break up into no-order open skirmishes, spear-throwing, front line fighters advancing and retreating and so on.5 The failure here is not the effort to use comparative evidence (that’s a good instinct) but the failure to ask if the comparandum – the thing being compared6 – is a good match for warfare in the Greek archaic?

Via Wikipedia, warriors of the Dani people from the central highlands. Now we need to suspend our cultureal assumptions for a moment and avoid focusing on if these fellows look ‘strange’ (we probably look strange to them and all of us would look strange to the Greeks). Instead, we want to ask are these fellows equipped to fight similarly to hoplites or other iron-age Greeks. And the answer just has to be ‘no, obviously not.’ They don’t have helmets, or shields, or armor, or shields, or clothing, or shields, or iron-tipped spears, or shields, or swords of any kind OR SHIELDS.

Because it pretty clearly isn’t. In this documented last phase of Dani warfare (they don’t do these battles anymore), the Dani still had an effectively stone-age level of technology, compared to iron-age Greeks. I cannot stress this enough: that is a very big difference, an enormous gap in weapons and armor capabilities which in turn comes with enormous implications for tactics. Metal – be it bronze or iron (much less steel) – is so much better a material for weapons that it significantly alters battlefield dynamics.

**The Dani fight not only unarmored, but almost entirely nude and do not generally use shields in contrast to armored Greeks and Homeric heroes whose armor ‘clatters’ **(ἀρᾰβεῖν, ‘to rattle, clang, clatter’ (of armor)) to the ground when slain and who regularly bear shields. In part, this is because Dani weapons are much less lethal than iron-age weapons, a point that jumps out if one actually watches Dead Birds. These men are trying to kill each other (and to not be killed) but fighting at distance it takes a lot of luck for their weapons to actually inflict lethal harm (and indeed, the casualties for these battles are very low). An arrow with a bone tip, or a spear that is merely a sharpened wooden stake can only be so sharp. Multiple individuals in Dead Birds are hit by arrows or javelins which simply do not penetrate to lethal depth (though one man does eventually die of a wound) despite striking the target. Remember these are *unarmored, *nude combatants who have been hit directly with a weapon. The contrast with what a sharp, iron-tipped broadhead arrow launched from a war bow can do against an unarmored target is quite stark; ancient and medieval artwork regularly show combatants with arrows transfixed in their bodies – all the way through and out the other side. As is typical with ‘first system‘ warfare, the high casualty bursts in Dani warfare come not from battles, which are generally symbolic affairs, but from ambushes and raids.

But even Homer’s heroes are clearly practicing ‘second system‘ warfare: they are laying siege to a large fortified city, with an army that Homer clearly understands to includes tens of thousands of warriors (Homer’s Catalog of Ships, 2.494-756 describes the Greeks as bringing a total of 1,186 ships; if taken literally it might imply an army of c. 150,000 though of course this is all subject to heroic exaggeration). Those warriors wield weapons – typically described by Homer as bronze, though iron is known to him – and wear body armor, helmets and carry large shields. As van Wees notes (op. cit., 166), the most prominent weapon in early Archaic artwork is actually the sword (spears are very common too), a weapon which the Dani did not have and were not capable of manufacturing with any material available to them. Homer’s own world is part of a broader military system that by 750 BC includes large, sophisticated professional armies in the Middle East (the Neo-Assyrians), employing complex siege craft (indeed, more complex than what the Greeks will have for centuries) and increasingly true cavalry. Homer seems to be blending a vague memory of late bronze age warfare (chariots! bronze weapons!) with early iron age warfare on the edge of ‘civilization.’7

So while in absolute chronology the Dani are c. 2,700 years in Homer’s future, in a kind of relative developmental chronology, their warfare is at least two thousand years in Homer’s past (taking the Greek bronze age to start very roughly at c. 3200). We might as well be trying to use footage of Roman warfare as the key to understanding the World Wars. Sure, humans and human psychology doesn’t change, so there may be some valuable insights (and indeed there are some about human psychology in combat which are useful in pushing back against the orthodox model) but we would need to be alert to everything that is different, which is a lot.

Approaching Archaic warfare through the lens of Homer, the Dani and Dead Birds sets van Wees’ entire foundation askew. That doesn’t mean everything in his model is wrong, but it throws a lot of things off.

In particular, the van Wees model of archaic hoplite warfare runs thusly: hoplites emerge in the context of a kind of warfare that looks a lot like the way the Dani fight: extended skirmishes with missiles, with individual warriors occasionally running forward to take more risk (and be more lethal) doing battle at closer range, sometimes with javelins, sometimes with contact weapons (swords and spears). This is, for van Wees, the environment in which the hoplite emerges. Hoplites initially show up carrying two spears (one for throwing), which to van Wees suggests continued participation in the skirmish (see my doubt below) rather than being pure ‘shock’ specialists. For much of the archaic, in van Wees’ model, hoplites continue to fight in open order or even no order at all, with unarmored skirmishers – poorer Greeks – mixed in with them, taking cover behind the shields of hoplites in an intermixed and largely unorganized formation.

Over time, the hoplite grows gradually in importance, with other warriors not vanishing from artwork or literature (Tyrtaeus, importantly) but being less prominent, but those lights remain scattered ‘here and there’ amidst the hoplites even well into the sixth century, with light infantry prominent on the battlefield even to the Persian Wars at the end of the archaic. **Van Wees admits no regular formation for hoplites prior to the first explicit mention of such in text in 426 (Aristophanes, Babylonians, F. 72) and contends that intervals less than six feet (180cm!) would have been unworkable even in the classical period (op. cit. 185).

For van Wees, these formations do not rush into a collision and then the ‘shoving-match’ othismos, but rather charge to release the psychological pressure of the fear of battle (thus the Spartans, better disciplined, walking into contact)8 but then slow down to a stop eis doru (‘into spear’s reach’) to then jab with spears at each other with overhead strikes. Formation collapse is thus not a result of shoving, but rather the line of hoplites collapses due to psychological pressure and casualties (more the former than the latter).

And I should be clear at the outset: some of this is workable. But a lot of it is not.

As we’ve already seen, I think the idea that the hoplite panoply emerged for open-order skirmishing is simply not tenable: no one commits to open order or no-order skirmishing wearing heavy armor and using a large round shield (instead, globally, the most common ‘kit’ for this kind of fighting in metal-working societies is little or no armor, but relatively large oblong shields that can provide full coverage for the body from missiles). Van Wees insists that a hoplite could advance and retreat just as well wearing their heavier equipment as a light infantryman (op. cit., 171) and that is just…obviously not true. The man in 4-8kg of equipment (a ‘light’) is obviously going to be able to run down the man in 18kg of equipment (the hoplite). That is a real liability in a ‘Dead Birds‘ combat scenario because the ‘front’ moves so far forward and so far back: either side often mounts sudden advances which send the other side scurrying backwards – but if you are wearing 2-3 times as much kit as your mates, when your line scurries backwards to get out of range (and those lights aren’t sticking around for you, they’re unarmored and so in real danger of being instantly killed by close range javelin or arrow shots) you are going to fall behind and those enemy lights are going to catch you and all of the armor in the world isn’t going to save you in a fight outnumbered four-to-one.

And I think here is a good time to stop and talk about how hard it can be to interpret artwork and we can take for our example one of the most important pieces of evidence in all of this, the hoplite artwork on the Chigi Vase (c. 645 BC).

ViaWikimedia Commons, three images of the Chigi Vase’s hoplite scene (there is a second scene below), c. 645 BC. Use the flutist to keep your bearings as to how these images come together – there is only the one guy playing the flute (an aulos, technically). So from (our) left to right, we have a shield and some weapons on the ground and men looking like they’re gearing and running to join a battle line (bottom left), then we have the flutist, then a battle line (top) meeting another, with men in lines, spears raised and then (bottom right) we have a better view of the second battle line, with shields presented as overlapping and a second line of men coming behind it.

And the thing is almost every aspect of that evidence – which seems clear at first glance – is open to multiple interpretations, especially in the context of a two-decade old fight where no one wants to admit they might have been wrong. We can begin with the weapons: while orthodox scholars will point to a dense formation of hoplite-armed heavy infantry (with no light infantry in sight!) Hans van Wees and other heterodox scholars point to the fact that each hoplite here carries two spears, potentially with throwing loops and suggest that this two-spear configuration (which fades out by the end of the 600s) is indicative of hoplites still skirmishing.

And I want to stop for a minute and examine that point because I think it is suggestive of one of the problems I keep coming back to in these debates, because “having a throwing spear alongside a thrusting spear means you probably skirmish” is a position that cannot survive a working knowledge of ancient Mediterranean warfare much less warfare generally. After all, Roman heavy infantry famously carry two javelins (the pilum) and yet are very clearly shock heavy infantry.9 Likewise, in Spain among both Iberians and Celtiberians, a javelin (frequently of the soliferreum type, sometimes of other types) was a standard weapon to pair with the ubiquitous thrusting spear; we very frequently find them in pairs in grave deposits suggesting they were basically always carried one-and-one, yet Fernando Quesada Sanz has spent the last two decades arguing – persuasively – that Iberian and Celtiberian warriors fought frequently as ‘line infantry’ in a sort of shield wall.10 Likewise, we know that in certain periods, Gallic infantry carried javelins and no one would accuse the Gauls of generally operating like skirmish infantry. More broadly, history is full of examples of shock infantry that expected to shoot a single volley at close range right before closing into combat, be that Roman volley-and-charge with pila in the third century BC or post-gunpowder shock tactics like with the 17th century Highland Charge or the contemporary Swedish Gå–På (“go on”). It is significant that these hoplites still carry a throwing spear, but it absolutely does not make them skirmishers.

But the heterodox folks are right that there is a lot of interpretive difficulty here. Van Wees (op. cit.) wants to read the image as representing a single moment of combat, with some men fighting in the front, others holding back and still more gearing up in the ‘everyone do their own things’ Dead Birds style of battle, but of course one could just as easily read the image as chronological, showing the battle line forming up, then marching into battle (it’s a pity we don’t have more of the other side). On the other hand is the question of what to do with the fact that each battle line is shown in two ranks, one separated by a flutist, the other just by an open interval. The orthodox reading is that this is an indication of formation depth, a crucial component in their definition of the phalanx, whereas the heterodox note that there’s a separation here, no sign of shoving and so perhaps the second rank is well behind the first, a distant reserve. Everett Wheeler, in exasperation, pointed out once that contact infantry basically never fight without depth in just a single thin line and I tend to think he is right about that objection, but there is certainly no shoving othismos here.11 In terms of spacing, I read these soldiers are tightly spaced, indicating a close-order formation, but the heterodox will dismiss such closeness as artistic license, noting that soldiers are often drawn more tightly packed in artwork than they would have been in reality.

We might note that what we see here looks somewhat similar to something like the Bayeux Tapestry, which we know to depict a shield wall, but of course a chasm of time and art style separates the two, so this is hardly decisive.

Via Wikipedia (though I have cropped) the English shield-wall at Hastings (1066) as depicted on the Bayeux Tapestry (c. 1070s). Note the one little archer fellow, drawn smaller than the heavy infantry around him (because they’re more important) expressing the idea of some English archers being present, although to go by our sources for the battle, not many (far more Norman archers).

For my own part, my reading of the Chigi Vase is closer to the orthodox one: those men are in close order and the second rank of each formation does imply depth even if the artist has created some space for us to see the flutist. I think what is being expressed here is a chronological sequence, showing the formation forming up, then advancing and finally coming into contact, likely showing us the moment of volley before the charge. In this sense it is actually similar to the chronological scroll of the Bayeux Tapestry, where many scenes ‘blend’ into each other. The fact that the opposing formation is also shown at least two depth suggests to me that depth – not a sequence of two widely separated lines – is intended. We’ll come back to definitions next week, but I *would *call the thing on the Chigi Vase a ‘phalanx’ of a sort (we’re going to see my definition of ‘phalanx’ is a bit broader than some). But as you can see everyone has their own interpretation and the chances of convincing anyone of anything – something that seems promising when you first look at it – are slim.

At the same time van Wees is fundamentally right about some things. Light infantry with bows and javelins do not go away in Archaic artwork, though they do diminish over time, from being perhaps half of all depicted figures in the early Archaic to only showing up infrequently in ones and twos by the end. That might indicate an actual reduction in their numbers, but a even a fairly casual reading of Herodotus suggests otherwise: they’re still there, but they’ve become less politically and socially important and so are less frequently depicted or described. So we need a model of archaic battle which allows for both hoplites and light infantry with ranged weapons to share the battlefield; the ‘all hoplite’ Archaic phalanx of the orthodox school will not do with the evidence.

Towards Better Models

Instead, we need to think with iron-age comparanda about how heavy infantry work in concert with lighter ranged infantry. One possible comparison, contemporary to the Greek archaic, is the warfare system dominant in the Near East at the time: Neo-Assyrian infantry working in matched pairs of shield-bearing contact infantry (with spears) and foot archers. As best we can tell (our evidence is not fantastic) these fellows were expected to set up relatively static battlefield formations, with the shield-bearers providing both protection from ranged attack (with their large but thin shields) and also from sudden cavalry or contact infantry attack (with their spears). The archers could then safely develop ‘fire.’12 This has the advantage of being contemporary and there are lines in Tyrtaeus and artwork that support the idea of light infantry sheltering behind the shields of hoplites (van Wees, op. cit., 166-77 assembles the relevant examples). But that Neo-Assyrian paired infantry was also, from what we know, a quite well organized, professional standing infantry force which is not very much like our hoplites and the status distinction ran the other way (it was archery, not contact warfare, which seems to have been the higher status way to fight) and nothing gives us the sense that hoplites are fighting with lights in something like assigned pairs save perhaps some hint for the Spartans towards the end of the Persian Wars (op. cit. 182) and even then it is hardly strong evidence. I think we need to be aware that this combat model was, certainly by the late archaic if not earlier, available to the Greeks (at least some of them), but I do not think it was how they organized.

Another potential comparandum here is the early medieval shield walls I’ve alluded to before. I thought I would have to write a whole big paragraph about this, but actually Paul Bardunais walked through exactly this comparison and reconstruction, using a lot of knowledge gleaned from reenactment and safe combat sparring experiments and I don’t think I can improve very much on it. He presents this ‘hybrid’ shield wall as having a few ranks of heavy infantry, in relatively close order (we’ll get to intervals blow) at the front forming a protective wall, with light infantry skirmishers deployed behind. They might equally be able – with some difficulty – to filter through the ranks (since ‘close order’ does not mean ‘shoulder-to-shoulder’) so your skirmishers could move out in advance to screen the shield wall or drop back behind it if pressed. In this system, the shield wall becomes a kind of ‘base’ from which skirmishers can operate and since, as noted, hoplites are still often carrying a throwing spear of their own, it can also project some amount of ranged threat.

I think this is a workable mental model, though it seems like it may need a bit of modification to fully fit the evidence. I want to be clear that isn’t me saying it is wrong. Greek artists in the archaic tend to show skirmishers intermixed with hoplites when the show them, but it is really tricky to know how to gauge that. As you are presumably seeing from the artwork I’m showing here, going from a stylized 2D representation of a formation to understanding the actual formation is tricky and artists often have to distort, compressing intervals (very frequent in medieval artwork where formations we know were not shoulder-to-shoulder get compressed until they look it, cf. also the Columns of Trajan and Marcus Aurelius for the same effect) removing depth (so showing only a single rank) and so on. Likewise, my reading of Tyrtaeus’ description of hoplites in battle suggests that while there are certainly light infantrymen running about, there is an offensiveness to the ideal hoplite, who doesn’t just stand under ranged fire but gets in close to the enemy that speaks to me of something closer to what Bardunais terms a ‘bludgeon’ shield wall (which he associated with the classical period).

By fierce deeds let him teach himself to fight, and not stand out of fire – he has a shield – but get in close, engage and stab with lance or sword, and strike his adversary down. Plant foot by foeman’s foot, press shield on shield, thrust helm at helm and tangle plume with plume, opposing breast to breast: that’s how you fight, with the long lance or sword-grip in your hand. – Tyrtaeus fr. 11 West (trans. M.L. West)

I might suggest a third comparative model: warfare in pre-gunpowder coastal West Africa, within the range of the tsetse fly. While north of this region, in the Sahel (too try for the tsetse fly), warfare was dominated by cavalry, the tsetse fly’s sleeping sickness is lethal to horses and so warfare further south along the coast (along the Gulf of Guinea, down through to the Congo River) was an infantry affair. Armies here consisted of two kinds of troops, a broad (lower status) militia force which composed the bulk of the army and were armed as relatively light skirmishers and then a ‘core’ of better trained professional warriors maintained by local kings who formed the backbone of the army and were better equipped (notably including large shields, although not much body armor). A battle between two armies might begin with the engagement of skirmishers, intended to soften up the enemy force (and perhaps screen the higher status warriors). But at the right moment those higher status warriors with their large shields and contact weapons would charge forward in a dense mass, ideally scattering the enemy (who would have their own ‘base’ of heavier warriors too), thus winning the victory. Here the battlefield is open enough for the skirmishing troops to work in and around the ‘heavies’ who initially function as a defensive bulwark to the army but then at the right moment are deployed offensively.13

Via Wikimedia Commons, an African warrior with weapons, including a several iron-tipped javelins and a large shield, c. 1641. This warrior was painted fighting in Brazil, but was likely originally form the Kongo people.

Now I want to immediately caveat this model (I’ve spent so much time harrying van Wees for not doing so, I can hardly not do so myself),** there are some major differences**. The first is armor: this West African system had large shields (generally oblong, more useful against missiles, rather than round) but not much body armor and that’s a really big difference. They do have iron weapons, so those shields are necessary to limit the lethality of the skirmish and that professional core of contact infantry might wield deadly iron swords and iron-tipped spears (just like early hoplites). However, whereas warfare in Greece (and much of Eurasia) was about control of land, warfare in this part of West Africa was frequently about control of people (really, control of laborers) and as a result there is an emphasis in the local kit on capture weapons like clubs, not because these guys are primitive, but because they want to take enemies alive as captives. Those are some pretty meaningful differences and so I am by no means suggesting sub-Sahelian West African pre-gunpowder warfare as a 1-to-1 of early Archaic hoplite warfare: instead it is just another tool we can use to think about how people might combine light infantry and something like a shield wall.

But you can see how this model might work, especially if we work in elements of Bardunais’ model as well. Towards the close of the 8th century, the wealthier Greeks begin to start equipping themselves as ‘specialist’ contact infantry (albeit still carrying a perhaps a single throwing weapon), probably suggesting that ‘contact infantry’ (as distinct from skirmisher) was a role that had already existed and was generally the higher status role (as, frankly, Homer clearly seems to think). Fairly quickly these fellows end up grouped together rather than mixed up indiscriminately with the skirmishers, either in a single block as the core of the army (the ‘West African’ model) or as a line in the front of it (the ‘Early Medieval’ model), but still working hand in glove with the skirmishers. As these fellows group up, the equipment that makes the most sense in that context – what will eventually be the hoplite kit – begins to predominate.

By the late 600s, we see the last of the throwing spears carried by hoplites in artwork drop away, w