By the 7th-century BC the Greek city-states had adopted the phalanx (pl. phalanxes) as their fighting formation. Armoured hoplites (Greek: ὁπλῖται,
hoplitai; sing. ὁπλίτης
hoplitēs) formed up in close-order, shoulder-to-shoulder, their shields locked together. The first few ranks of men would project their spears beyond the first rank of shields. The phalanxes advanced towards each other usually at walking pace to maintain cohesion. If the formation was lost, the phalanx would be weakened or rendered useless. It is possible, however, that the hoplites picked up the pace, perhaps breaking into a run during the last few metres. The additional speed needed to be sufficient to gain momentum against the enemy in the initial collision. The famous charge of Athenian hoplites at the
Battle of Marathon (490 BC), however, was probably precipitated by a desire to minimize their losses to Persian archery.
Orthodoxy According to the orthodox view of othismos (“pushing”), the opposing phalanxes would collide, possibly breaking many of the front rank’s spears while attempting to maim or kill the opposing front rank. If one side did not collapse because of this clash, then the men in the succeeding ranks pressed their large round shields (Greek: ἀσπίς, aspis; pl. ἀσπίδες, aspides) against the backs of the men in front and pushed them forward. The combined physical onslaught of one densely packed mass of men was opposed by the counterthrust of the enemy phalanx, shield against shield. Eventually, one side was forced back and its formation disrupted, the hoplites perhaps being literally knocked over and trampled. There was little or no actual fighting after the initial, very brief clash of spears and the struggle essentially became a decisive pushing match (Hanson, 1989, 169). Indeed, Thucydides described hoplite warfare as othismos aspidon or “the push of shields”. As a rule, with few recorded exceptions, the deeper phalanx would almost always win. The orthodox model of collision and push (othimos) has been likened to a scrum in rugby or the scrimmage in American football. This is the view championed by V.D. Hanson in his detailed examination of hoplite warfare.
Yet, no ancient Greek historian explicitly tells us that
othismos involved all ranks packing together in a united push to drive their enemy physically backward. Despite this omission, we know hoplite phalanxes were deployed with a series of ranks behind the first. It seems very rare for a phalanx to be less than eight ranks deep from the available evidence, and much deeper formations were not uncommon. Most hoplites in the phalanx, therefore, were unable to reach the enemy with their spears (Greek: δόρυ,
dory or
doru). They might have been able to finish off fallen enemies with their spear’s butt spike (Greek: σαυρωτήρ,
sarouter) and certainly gave moral support to the front rank men doing the actual fighting, but they cannot have inflicted any significant damage upon the enemy. If the
othismos was a massed shove, then it would seem to explain the presence of these, otherwise largely superfluous, men on the battlefield. This is the basic tenet of the traditional view of “massed shoving” to explain the role of the rear ranks of a phalanx. Yet if this was the only purpose of successive ranks, then the larger phalanx would always win.
The literary evidence supporting the collision and pushing match is far from extensive, however. Tyrtaeus’ poetry contains a description of two opposing sides clashing “shield against round shield” as does Aristophanes’ comedic play “Peace”. Neither author names a specific battle but the use of the concept in poetry and theatre suggests audiences were familiar with battles fought this way. Beyond the arts, historical accounts of hoplite engagements refer to a press of shields. At the Battle of Delium in 424 BC, Thucydides describes how the two sides collided at the run and fought with a “pushing of shields”. Likewise, during the Battle of Mantinea (418 BC) he also described how the Spartans “advanced” and “pressed” the opposing Argive and Arcadian hoplites routing them, yet it is unclear whether this was shield-to-shield (Thucydides, “History of the Peloponnesian War”, 5.73). In his account Hellenica, Xenophon states that during the later stages of the Battle of Coronea in 394 BC, the Theban and Spartan phalanxes collided head-on at the run and were similarly pressed “shield against shield” (Matthews, 2009, 397). Yet Xenophon and others also use a variation of the term “shield pressed against shield” to describe a close-order shield wall. In these instances, it is clear the phalanxes were not engaged with an enemy, merely waiting to receive an attack.
Tyrtaeus’ suggestion that hoplites “reach forth and strike the foe” evokes the idea that hoplite combat being conducted at spear’s length from the enemy. Descriptions of battles, for example those of Plataea (479 BC), Sphacteria (425 BC) and Piraeus (404 BC), all record one side being “pushed” or “pressed back” but in a figurative sense. Xenophon employs this sense in his work Hellenica where he describes how the Thebans “pressed” the Spartans at the Battle of Leuctra fought on July 6th, 371 BC (Xenophon, Hellenica, 6.4.14). In these instances, the phalanxes need not have been shield-to-shield, rather the overwhelmed side may simply have been forced backward at the spearpoints of their opponents. Alternatively, as on the island of Sphacteria, when the Athenians engaged the Spartans at range with skirmishers and peltasts [1], the heavily armoured Spartans were pinned in place and could not close to hand-to-hand fighting (Matthews, 2009, 399). The earlier Battle of Thermopylae (480 BC) had ended in a similar manner for the 300 Spartan, 700 Thespian and 400 Theban hoplites defending the famous pass. On the third day, as the Persian Immortals approached, the Greeks withdrew and took a stand on a hill behind the Phocian wall that defended the narrowest part of the pass. Herodotus says:
“In that place they defended themselves with swords, if they still had them, and with hands and teeth. The barbarians buried them with missiles, some attacking from the front and throwing down the defensive wall, others surrounding them on all sides.” (Herodotus, Histories, 7.225.3)
Tearing down part of the wall, Xerxes ordered the hill surrounded before the Persians rained down arrows until every hoplite was dead.
As an aside, the orthodox view that phalanx warfare centred on
othismos seems very akin to the “push of pike” that used to accompany re-enactments of English Civil War battles some years ago (pictured
right). At that time two opposing blocks of pikemen would advance on each other with pikes levelled to threaten the “enemy”. A few metres before contact the pikes would be angled upward, and the two blocks would then crash to together in a pushing contest [2]. While this sounds very much like the description of
othismos, the tactic was not historically accurate but based on health and safety concerns to limit injury to individual re-enactors while still creating a spectacle for audiences. Historically such pike blocks would have advanced to contact whereupon the leading ranks would fence with their pikes to spear their opponents - the equivalent of ancient Greek
doratismos (see
below).
What to do with the spear? The hoplite phalanx dominated warfare among the Greek City States from the 7th- into the 4th-century BC. If we have interpreted phalanx warfare correctly, and it is far from clear that we have, then othismos being a shield against shield pushing match seems only half the story. The hoplite panoply, specifically the degree of protection it provides, is far better suited to fighting with spears. During this period the men equipped themselves with a large round shield (aspis) and a 2.1 m to 2.7 m (7–9 ft) long spear (doru or dory) with a sharpened iron blade (aichme) and bronze butt-spike (sauroter). Practical experimentation based on contemporary representations has determined the most likely manner in which hoplites fought. Within the phalanx each man stood at an angle, their left shoulder braced into the dished bowl of their shield while refusing the other shoulder to protect the vulnerable right side of the body while maintaining the ability to deliver spear thrusts (van Wees, 2000, 128-30). This fighting stance positions the feet and legs in a strong attacking platform but also means the hoplite is braced in defence to resist a clash of shields. The latter fits well with the orthodox view of othismos which sees successive hoplites pushing their shields against the right shoulder or back of the man in front. Yet, with his opponent’s shield pressed against his own, the hoplite would be pinned between his comrades and his enemies, vulnerable and unable to strike with his spear. This lack of manoeuvrability among those in close contact is rather neatly captured in the attempt by modern re-enactors to perform othismos pictured below. The image also shows how the cohesiveness of both front ranks might quickly dissolve and how striking with a spear becomes almost hopeless.
So, if the tactic was for two opposing phalanxes to crash into each other and simply push shield-to-shield, then why carry a long spear? In such close-order fighting, the dory effectively becomes an incumbrance. Far better would be to ditch the spear in favour of a shorter weapon such as the sword (xiphos or kopis) by which the hoplite could strike at his opponent from behind his protective shield. This was clearly the favoured tactic of later Roman armies where it certainly proved decisive against the Macedonian phalanx at the Battle of Cynoscephalae in 197 BC.
The Roman model also raises questions about the density of phalanx formations if hoplites were to use their weapons effectively. Thucydides’ description of the Battle of Mantinea reveals that “fear makes each man do his best to shelter his unarmed side with the shield of the man next to him on the right, thinking that the closer the shields are locked together the better he will be protected” (Thucydides, 5.71). This has been the foundation of nearly all recreations of phalanx warfare but rather unfortunately it ignores the maxim that no plan survives contact with the enemy. It can be shown throughout ancient Greek history, including Homeric period warfare, that each battlefield situation dictated the density of a formation. So, while hoplites might have sought protection from their neighbour, they would still need space both to fight and defend themselves.
Peter Krentz argues that a lack of armour for the
sarissae wielding “phalangites” [3] meant that Phillip II of Macedon’s phalanxes adopted much tighter formations for added morale and physical protection. The Macedonian king also devised the hedge of
sarissae that later so frightened his enemies and, under the right conditions, prevented them from penetrating close enough to do damage with their shorter spears and swords. By contrast, the classical hoplite’s head-to-foot armour instilled confidence in their protection allowing their formations to open the spacing to three feet between men (Krentz, 1985, 52). Both Krentz and van Wees argue that each hoplite thus occupied a six foot space within which they could wield their weapons effectively while limiting an opponent’s opportunities to attack his vulnerable unshielded right side. Any man attempting such an attack would have to be mindful that they were within spear thrusting range from a neighbouring opposing hoplite (van Wees, 2004, 185-186; Krentz, 1985, 51-53).
On SoA Forums, Society member Patrick Waterson described five stages characterising a hoplite battle:
1. Ephodos (the “charge”) as previously mentioned in Thucydides’ description of the Battle of Delium, Xenonphon’s account of the Battle of Coronea, and the famous charge of the Athenians at the Battle of Marathon.2. Doratismos (the “spearing”) evoked by Tyrtaeus’ portrayal of hoplites “reach[ing] forth and [striking] the foe”. Once the phalanxes have closed on one another, then the first contact involved spear fighting rather than a pushing match.
3. En chersi (the “hand-to-hand”) as evidenced by Herodotus’ account of the Battle of Thermopylae in 480 BC. The hoplites defending the Pass fought with spears, until every spear was shattered, and then switched to xiphē (short swords).
4. Othismos (the “pushing”) only appears after all other avenues aimed at defeating or overwhelming the enemy had been exhausted. It is notable that the close-order mêlée (en chersi) precedes the shield-to-shield pushing contest (othismos) suggesting that the latter was a means by which men in the rearward files lent their weight to an otherwise static melee once the front-rank men were exhausted and no longer capable of giving their best.
5. Trope (the “collapse”) could occur very rapidly, even at or before first contact, as indicated by the accounts of battles such as First Mantinea (418 BC).
It would be a mistake to assume all hoplite battles rigidly adhered to these five stages. Circumstance may necessitate a stage being quickly concluded, or long and drawn out, or omitted all together. As Waterson says: “two fairly evenly balanced sides may well find themselves going through all the stages until during othismos it becomes a case of ‘One more push, men, and they will break!’” Regardless, a general pattern emerges whereupon the opposing phalanxes close with each other, perhaps at the run, followed by a mêlée at spearpoint or hand-to-hand with swords. Only when those men in the leading ranks are tired or wounded and unable to effectively continue the hand-to-hand fight might any form of “pushing” contest occur. At this point fatigue might dictate that the first phalanx to yield ground would be unlikely to recover. If, as the orthodox view implies, othismos occurred much earlier when both phalanxes were fresh and invigorated, then a pushing contest might not necessarily produce one side’s collapse.
Undoubtedly this form of combat would have been both physically and mentally exhausting, so it is highly likely that natural pauses would have occurred during a prolonged battle. Such pauses would usefully allow both sides to catch their collective breath, redress their lines, recover the wounded and replace them with men drawn from the succeeding ranks. If the latter is correct, does this interpretation explain the depth of the phalanx? Rather than simply adding their weight to a pushing contest, the additional ranks might be viewed as tactical replacements.
A conclusion? If
doratismos and
en chersi are correct, then having the space to wield their weapons means the pushing contest with opposing hoplites shield-to-shield, and pressed forward by their fellows in succeeding ranks, seems impractical. For
othismos to occur necessitates the phalanx to transition into close-order which, for ordinary hoplites not particularly well drilled in such manoeuvres, might not be so easy to achieve especially in the press of battle. Is it, therefore, more reasonable to conclude that ancient historians used the term
othismos in a more figurative sense. Indeed, Homer uses the word “
othismos” in the Iliad even though most historians generally agree that Homeric warfare was fought in very loose formations (Homer,
Illiad, 569, 655). Compare that style of fighting to the close-quarter, hand-to-hand mêlée of the 4th- and 5th-centuries and it strongly suggests
othismos was a reference to the gradual gaining of ground by forcing - “pushing” - the enemy rearward. This may well have involved pressing forward shield-to-shield, but the primary sources never allude to one mass shove. Rather most describe how armies surged forward, step by measured step, to gain ground from an opposing force.
The evidence we have implies hoplite battles were determined by individual, hand-to-hand combats between hoplites armed with spears and subsequently swords. Once enough hoplites on one side had successfully resolved these individual contests, then they were able to coalesce and advance to “break the enemy’s ranks” before “pushing” them back until the enemy was overwhelmed and routed. Bon appétit!
References:
Hanson, V.D., (1989), “The Western Way of War”, University of California Press.
Godley, A.D. (ed.), “Herodotus: The Histories”, Perseus Digital Library, available online (accessed 9 January 2026).
Krentz, P., (1985), “The Nature of Hoplite Battle”, Classical Antiquity, Vol. 4, No. 1, University of California Press, pp. 50-61.
Matthew, C.A., (2009), “When Push Comes to Shove: What Was the ‘Othismos’ of Hoplite Combat?”, Historia: Zeitschrift Für Alte Geschichte, vol. 58, no. 4, pp. 395–415. available online (accessed 25 June 2025).
van Wees, H., (2000), “Greek warfare : myths and realities”, London: Duckworth.
Waterson, P., (2013), “Othismos - When Push Comes to Shove”, SoA Forums, available online (accessed 11 September 2025).
Endnotes:
1. A peltast (Ancient Greek: πελταστής, peltastēs) was a type of light infantry originating in Thrace and Paeonia and named after the crescent-shaped wicker shield they carried called a “pelte” (Ancient Greek πέλτη, peltē; Latin: pelta). Peltasts often served as skirmishers in Hellenistic armies.
2. This was the author’s first experience of a battle re-enactment where the pike blocks engaged in rugby-like scrummaging. In more recent times, regiments in the English Civil War Society (ECWS) now portray pikemen fencing with their pikes in more historically accurate manner.
3. Philip II of Macedon, equipped his Macedonian “phalangites” with a long spear or pike (Koine Greek: σάρισσα sarissa, pl. sarissae) about 5 m to 7 m (16 to 23 ft) in length. These longer spears improved the strength of the phalanx by extending the rows of overlapping weapons projecting towards the enemy which, in turn, kept the latter at a greater distance. After the conquests of Alexander the Great, sarissae became the mainstay during the Hellenistic era (4th– to 1st-centuries BC) by the armies of the successor states of Alexander's empire, as well as some of their rivals.
