The Emergence of Hoplite Warfare, 900–525 BC IV


Although helmets and armour were worn by warriors before and after the introduction of hoplite warfare, it was the hoplite shield that helped to bring about the change. From the beginning of the period until about 700, there was a wide variety of shields in use, judging by both archaeological finds and artistic depictions. At the end of the Mycenaean period we saw that the round shield with central grip was the most common type, perhaps with larger examples being carried by foot soldiers, and smaller examples by chariot-borne infantry. An exception is a vase fragment from Iolkos in Thessaly dating from the thirteenth century, which seems to show a small shield with in-curving sides (Bulletin de Correspondance Hellénique 1961, p. 770, fig. 20). This is similar to contemporary Hittite shields and also to the Dipylon shield.

From 900 to 700, the single-grip round shield remained in use and came in varying sizes (Boardman 1998, p. 118, fig. 213). Much of our evidence comes from bronze shield bosses with which they were sometimes reinforced. These round bosses had a central dome protecting the handgrip, sometimes reinforced with a spike which could be used as an extra weapon at need (Snodgrass 1965a, p. 38). A major problem has been that tombs often contain round bronze objects which are not shield bosses, but cymbals, horse trappings or belt fittings. Snodgrass did admirable work in sorting out criteria to distinguish these items, which seem to show that such shield bosses lasted down to c. 700, but no later. The exception is Cyprus where, as we have seen in the case of helmets and armour, the Greeks there were doing their own thing and may have come to hoplite warfare rather late (Snodgrass 1965a, pp. 40 ff.). The bosses average around 15cm in diameter and are generally reconstructed as fitting onto shields of about 45–50cm in diameter. Snodgrass (1965a, p. 48) points out, however, that a round shield boss does not necessarily mean a round shield. The Kaloriziki shield from Cyprus was probably not round, and Ognenova (1952, pp. 61–81) has reconstructed round shield bosses onto ‘Boeotian’ shields. While that seems unlikely, round bosses could certainly have been attached to the square shields that also appear in art at this time, and perhaps even the Dipylon shields.

Graves are sometimes found with more than one boss, which Snodgrass thinks may have been mounted on the same shield (like Kaloriziki), but I disagree. I think the Kaloriziki reconstruction is incorrect, and that it is far more likely that multiple bosses equals multiple shields, especially when they are of a similar size. A grave at Mouliana had four bosses, each about 19cm across; they are all bosses to protect central handgrips and cannot be seen as decorative extra bosses. The quantity simply points to the status of the warrior. Apart from warrior burials, a large number of shield bosses have been found at Olympia (Fellman 1984). Nearly all of these fall into the 900 to 700 date bracket, although there are some from the seventh century. Fellman sees these as being dedications from further afield in Europe, where single-grip shields with bosses continued to be used. He also points out that there were some areas of Greece that did not adopt hoplite shields and warfare in the seventh century. Achaea, for example, still used rectangular shields (the thureos) a couple of hundred years later, and so small, single-grip shields may still have been used in ‘backward’ parts of Greece.

With the increased availability of bronze, shields with complete bronze facings first made their appearance in Crete in the eighth century. The idea possibly came from the Near East, where similar shields were also in use (Snodgrass 1965a, pp. 52–3). There are three types. The animal-head protome shield has a three-dimensional animal head in the centre of the shield, instead of a simple boss. Examples are known from Olympia, although they may also be Cretan. They date from the eighth century and, although Snodgrass says this may have been a Cretan idea instead of the mainland shield blazon, they did not really overlap chronologically as the animal-head protome shields were no longer in use by 700.

The second type of bronze-faced shield is also Cretan and is called the Omphalos shield, after the conical Omphalos stone at Delphi marking the centre of the world. This shield has a cylindrical or conical bulge in the middle of the bronze facing to protect the handgrip and is decorated with concentric circles. Examples from the Idaean Cave in Crete are 34.5cm and 27cm in diameter. There are no certain examples from Olympia, but fragments are difficult to distinguish from Herzsprung or later hoplite shields. It is probably another mainly Cretan shield.

The last type of bronze-faced shield is called the Herzsprung shield after a find spot in north Germany. Similar to the Omphalos shield, the bronze facing of the Herzsprung type has a V-shaped notch in the centre and no bulge covering the handgrip (Snodgrass 1965a, p. 55 and plate 24; Sekunda 1999, plate A, no. 3 for a reconstruction). This decoration was derived from an original hide construction and, after translation into bronze, the notch became smaller and more stylised, allowing for some typological dating. Examples are known from Crete and Delphi, and there are also votive miniatures from the islands. Snodgrass thinks the Herzsprung shield is Cypriot in origin, and spread from there to the Aegean by 700. It became larger than the Omphalos shield; an example from Idalion is 83cm in diameter, and is thus as large as some hoplite shields, which may have been derived from it. The Herzsprung shield spread northwards and became popular in Europe, but was supplanted by the hoplite shield in Greece by c. 675.

Another type of shield, probably of Cypriot invention, is the small spiked shield, which developed from the small round shields, with or without bosses. Votives of such shields dating from the seventh and sixth centuries have been found on Rhodes and Crete, but it does not seem to have been in use on the mainland (Snodgrass 1965a, p. 56). Rhodes was later famous for its slingers and Crete for its archers, and these small shields may have been used by such troops, where skirmishing warfare rather than hoplite warfare was the norm.

What appears at first to be a very strange shield type is depicted on Geometric vases in the eighth century, particularly those by the Dipylon Master and his followers at Athens from 760 onwards (Coldstream 1977, pp. 109–14). Known as the Dipylon shield, it is oval in shape but with two, often very large cut-outs in each side. In the seventh and sixth centuries this shield design developed into the Boeotian shield, with much smaller cut-outs, depicted on vases showing mythical scenes, and with a hoplite grip that simply would not have worked (see below). Since this is generally considered to be an artistic invention, Snodgrass (following Webster) states that the Dipylon shield is also the result of artistic licence and does not depict a genuine shield (Snodgrass 1965a, pp. 58–9). He is quite correct in his assertion that a shield shaped like those we see on Late Geometric vases (late eighth century) is of no practical use. Such large cut-outs as are generally depicted would render the shield useless. But did the Dipylon shield really look like that? The problem is the style of Geometric art. People are drawn in a stylised way with broad shoulders and a very narrow waist, quite unlike real people. If we assume that the cut-outs are exaggerations, then we come up with a reasonable shield which could be used in combat, and had those cut-outs for a reason.

There are two much more accurate representations of this type of shield which are unaffected by artistic licence. The first is a terracotta miniature shield of the late eighth century in the British Museum. Although Snodgrass (1965a, p. 60) dismisses this as a votive, Connolly (1998, p. 51) shows that it is clearly copied from a real shield. It appears to be made of wickerwork, with supporting braces on the inside. It presumably also had a central handgrip. The second accurate depiction of this shield is on the proto-Corinthian aryballos featuring the hoplite in possible padded linen armour. The figures pitted against him are carrying Dipylon shields and wearing Corinthian helmets. This shows (as does the British Museum terracotta) that the shield was not flat but concave, and that the cut-outs were there to enable a second spear (or more) to be carried by the hand clutching the central grip. It also shows that the Dipylon shield lasted well into the seventh century. The aryballos dates to c. 680. It seems likely that this shield is a direct descendant of the Mycenaean figure-of-eight shield, but there is a dearth of evidence for its use in the thirteenth century, when the round shield dominated. The only exception is the Iolkos sherd mentioned above, which dates from c. 1300 and appears to show a shield with in-curving sides. This is not very similar to the figure-of-eight shield, and may have been separately derived from Hittite shields of the Near East, and the Dipylon shield could then have been descended from that. Then again, it may just have been a fresh Greek invention made to overcome the difficulty of wishing to carry several spears and a handgrip shield at the same time. Apart from the central handgrip, it appears that Dipylon shields usually had a shoulder strap as well, and could be worn on the back when not in use (Snodgrass 1967, plate 16). Greenhalgh (1973, pp. 67 ff.) also suggests that the cut-outs would have helped the shield to keep out of the way of the elbows when worn on the back like this. He is certainly correct when he says that it is the most popular form of shield depicted in late eighth-century art, although this may not reflect reality, especially as so much of this art is Attic. Geometric artists may have just enjoyed depicting this shape in particular.

Depictions of the Dipylon shield also vary in size, from ones almost as big as body shields to those barely reaching below the waist (Ahlberg 1971, p. 59). Something like a metre in height would seem about right to me (Sekunda 1999, plate A gets it about right).

From about 650 onwards, and particularly in the sixth and fifth centuries, a large oval shield with very small cut-outs at the sides features on vases. This is the Boeotian shield, which also features on coins of Thebes and its allies, and later as a shield blazon. It is always featured in heroic scenes, never in those of real life, and it seems to be a poor remembering of the Dipylon shield (Ahlberg 1971, p. 63). It is often shown fitted with a central elbow grip, or porpax, and a handgrip, or antilabe, at the edge, like a hoplite shield. In battle this would have brought the shield up horizontally, which is clearly incorrect. In fact there is no evidence for such a shield being in use at such a time, and it is generally agreed to be an artistic invention (Snodgrass 1965a, p. 58; Greenhalgh 1973, p. 64).

Another shield type which is occasionally represented in Geometric art of the eighth century is the square or slightly rectangular shield. If the Dipylon shield is descended from the figure-of-eight shield, then Snodgrass (1965a, p. 61) suggests this square shield may be a descendant of the Mycenaean tower shield. There is even less evidence for this, however, than for the Dipylon. The square shield seems to appear only for this short period in the eighth century, and might even be considered simply as an artistic device to show two different sides in the depiction of a battle (Boardman 1998, p. 38, no. 50). However, Olympia has occasionally turned up strips of bronze with guilloche decoration which are straight, rather than curved. These may be the edging from straight-sided shields.

All these shield types were gradually ousted by the hoplite shield, which in turn led to the hoplite warfare of the phalanx (Jarva 1995, p. 122). The hoplite shield was round with an offset rim, which could rest on the shoulder to help with the weight. The main distinguishing feature is the way it was carried. It has a central arm-band, called the porpax, through which the left arm was passed up to the elbow. There is then a grip at the edge of the shield called the antilabe, which the left hand gripped. When these features are shown in art, we know we are dealing with a hoplite shield; but when the shield is shown face on, it becomes harder to distinguish it from single-grip round shields like the Herzsprung. This two-handed grip system seems to have been a Greek invention, and may be behind Herodotus’s story of shield grips and blazons being invented by the Carians (Herodotus I, 171). From just before 700, shield blazons are shown on some round shields. These are designs that need to be held the right way up, and it is generally accepted that this works only with the two-handed hoplite grip (Snodgrass 1965a, pp. 61–2; Jarva 1995, pp. 121–2). Far more common than these blazons at this time were circle and spiral designs, which are not subject to a right-way-up rule and which appeared as early as c. 750 on vases with Dipylon and square shields. A series of other non-blazon designs have recently come to light on a late eighth-century vase from Paros. This shows a line of warriors in Kegelhelms carrying large, round shields decorated with a large cross or with chequerboard designs (Zapheiropoulou 2001, p. 289). Such shields are also sometimes shown with a telamon, showing that they are single-grip-type shields.

When the first blazon shields appeared in c. 700, they were not shown on the same vases as spiral-patterned shields; but in the seventh century both shields were shown together and it is clear that the spiral patterns carried on to be used on hoplite shields. Ueda-Sarson (2000; see website listed in References) has compiled illustrations of dozens of shield blazons from c. 700 to the Macedonian period, and spiral patterns continued to be shown until about 550, while other symmetrical abstract designs continued up to the Persian Wars. Only after c. 500 did nearly all shields carry blazons. Spartan bronze figures carry spiral designs down to the third century, but this is probably an artistic convention (Snodgrass 1965a, p. 62).

These designs were mostly painted onto the shield, which was of wood covered with leather and with a bronze rim, the latter usually decorated with a guilloche pattern. The Chigi Vase suggests that the earliest shields may have been of wickerwork or ply, like the Dipylon shield. Many fragments of the bronze rims have been found at Olympia, and some bronze blazons have also been found. Snodgrass illustrates (1967, plate 22) a particularly impressive winged horse and Connolly (1998, p. 54) a gorgon. A cockerel, and a right arm with clenched fist have also been discovered (Olympia Bericht, vol. I, plate 12 and vol. V, plate 32). Connolly (1998, p. 54) thinks that these are too delicate to have been used on shields in battle, and must have been made specially to be dedicated. Most arms and armour dedicated in sanctuaries were, however, precisely that captured from the enemy, and so I am sure these bronze blazons were used in battle (Jackson, in Hanson, 1991, p. 230). Shields completely faced with bronze have also been found at Olympia, and it would have been much easier to paint (and repaint) designs onto these. They also give us the size of hoplite shields, which usually varies from 80cm to 100cm, although one example is 120cm (Snodgrass 1965a, p. 64). Sekunda (2000, p. 10) suggests that bronze-faced shields did not appear until c. 500, judging by the colours used on vase paintings, but the prior existence of the Herzsprung shield shows that the Greeks were capable of producing thin bronze sheet at a much earlier period. Examples of bronze hoplite shield facings from Olympia can be dated to as early as c. 625 (Bol 1989, p. 683). The antilabe handgrip was presumably made of leather, but the porpax was bronze, often highly decorated, and the survival of large numbers of these, more than bronze facings or blazons, shows that most shields were just fronted with leather or some other perishable material. The interiors of the shields also seem sometimes to have been painted (Sekunda 1999, p. 53 and plate C1), and from c. 550 are sometimes decorated with separate tassels (Boardman 1980a, nos 64.2; 73).

By c. 675 the hoplite shield had ousted all other shield forms in most of Greece, and its use lasted right down to the fourth century. Compared to the wickerwork Dipylon, the hoplite shield was much heavier and stronger, and this was because of the new porpax/antilabe grip system. This meant that more weight could be carried by the left arm. The jutting rim of the shield meant that some weight could also be carried on the left shoulder, especially when at rest. Such a heavy shield was hard to swing around in combat and was another reason, along with the heavy bronze armour, for sticking close to your fellow warriors in the heat of combat. The phalanx of hoplite warfare developed from this, with men in close order, usually eight ranks deep. This made throwing spears, which seems to have been the popular form of warfare from before 900 to 700 and later, difficult, and led to the exclusive use of the thrusting spear as outlined below.


Spearheads from the Dark Age and Geometric period are fairly plentiful finds in Greece; Snodgrass devoted many pages (1965a, pp. 116–32) to a listing of different types, and his list tells us that iron replaced bronze for use in spearheads in the eleventh century. An anomaly, however, is that there are many spear finds from Olympia that must be from the eighth century or later, but which are made of bronze. As has been mentioned earlier, iron is much more efficient than bronze for use as a spearhead, and it was also more readily available. It is, however, harder to work and was therefore more expensive as a finished product. Iron might also be considered less aesthetically pleasing. Bronze spearheads seem to have retained their popularity in south Italy and Sicily until the eighth century, and in Crete for even longer; but there is a probability that bronze was reintroduced as a metal for spearheads in mainland Greece at some point in the sixth century (Snodgrass 1965a, p. 134). This may have had something to do with the expansion of hoplite forces. Where spear butts are known, these also tend to be of bronze.

The spearhead of the ninth and eighth centuries was 30cm to 50cm long but was certainly part of a throwing spear, as is depicted in Homer (Lorimer 1950, p. 257). It is difficult to find any artistic evidence of its use as a thrusting weapon (Jarva 1995, p. 123). Spears found in graves are often in pairs of the same size, and it is possible that both were thrown before close combat was joined with the sword. Late eighth-century vases frequently show pairs of spears being carried; most warriors are also shown carrying swords. In the seventh century the sword is depicted less, and one spear was used for thrusting in the new tactics of the hoplite phalanx.

As mentioned above, a second spear for throwing was kept by these early hoplites. A ‘still life’ aryballos of c. 680 shows a thrusting spear paired with a throwing spear, its throwing loop clearly visible. The Chigi Vase seems to show hoplites with more than one spear and, in the arming scene to one side, two spears – one long and one short – are shown waiting to be collected for battle. An anomaly here is that both spears have throwing loops, but this is perhaps an artistic error (Snodgrass 1965a, p. 138; Jarva 1995, p. 123). Both these depictions make it more likely that the second spears were for thrusting, rather than being an extra throwing spear (Snodgrass 1967, pp. 57–8).

Although the use of two spears, one for throwing, seems to have died out by c. 640, there are a few later depictions. These often show warriors on horseback, and date from the middle of the sixth century. They cannot all be explained away as being ‘heroic’ depictions, or cavalry rather than mounted hoplites (Boardman 1979, nos 2.1, 2.2; 1998, nos 375, 464). It seems that on occasion some hoplites still carried a throwing spear into battle, but that it really did go out of use by the 520s and, in most states, probably a lot earlier. The problem with the throwing spear was that, for throwing it, a lot of space and movement was needed, which was entirely incompatible with the close phalanx formation. Once this was discovered, hoplites were encouraged more and more to rely on the single thrusting spear. With the introduction of more solid hoplite shields and bronze armour, it is likely that thrown spears were becoming less effective anyway. Tyrtaeus urged the Spartan hoplites not to stand out of the range of missiles, but to ignore them and close with the enemy (Tyrtaeus: fragment 11, in Sage 1996, p. 28).

The sword retained its importance in the ninth and eighth centuries and was invariably still the cut-and-thrust, Naue II Type sword introduced in the thirteenth century. It continued in use right through the Archaic period to c. 520, although later examples are rare (Snodgrass 1965a, pp. 93–4; Kilian-Dirlmeier 1993, plates 41–54). They vary from 50cm to 70cm in length and have half-moon pommels, clearly depicted in Late Geometric art where the sword is the commonest weapon shown. Since we have seen that spears were thrown in this early period (900 to c. 650), it is natural that all warriors would also have needed a sword for hand-to-hand combat. Greenhalgh (1973, p. 73) suggests that the early phalanx in some states, such as Euboea, may have been composed of men who threw spears or javelins first and then closed with the sword in a phalanx formation. A fragment from Archilochus, who talks of javelin warfare on Naxos and Paros, mentions that Euboeans spurned the bow and sling, and fought close up with the sword, although he also calls them ‘Spear-famed’, indicating that they had this weapon at their disposal as well (Greenhalgh 1973, p. 90). Archilochus was writing in the seventh century, when hoplite warfare was becoming the norm, and he seems to be talking about places that were still fighting in an earlier style.

The introduction of the phalanx certainly undermined the sword’s previous prestige. After 700, weapons were shorter and stubbier, but they were frequently left out of artistic depictions altogether. Some examples have been found at Olympia of a single-edged, straight slashing sword, of a type known from Macedonia, Thrace and Thessaly (Snodgrass 1965a, p. 100). These seem to be mostly from the eighth century, although the type is seen in use later (Sekunda 2002, pp. 16, 29). It was probably from this sword that the kopis or machaira developed. This is a curved sword with the sharp edge on the inside (recurved), like a Gurkha kukri – which indeed is descended from it (Snodgrass 1967, plate 50). An example dating to about 650 is known from Crete (Snodgrass 1965a, p. 100), and one is illustrated on an alabastron from Syracuse of about the same date. This geographical difference has led Sekunda (2000, p. 16) to suggest that they came from the Near East, and Connolly (1998, p. 63) to suggest that they were an Etruscan development. I think, like Snodgrass, that a Thracian origin is the most likely. They do not appear to have become common until the end of the sixth century and were not as popular as the straight sword until perhaps the fourth century.

In the eighth century, the Naue II Type was a cut-and-thrust sword, but Geometric art shows that it was more often used as a cutting weapon (Ahlberg 1971, p. 46). In the later phalanx, a shorter weapon was preferable so as to be more easily wielded, but it seems to have still been used as a cutting or slashing weapon (Connolly 1998, p. 63). The adoption of the machaira shows that the straight sword was seldom used for thrusting, and the machaira’s thicker upper edge would have made it a much stronger cutting weapon. How effective these swords were as a secondary weapon in the warfare of the phalanx is uncertain. A thrusting spear might break in the initial charge so a sword seems essential, but whether it was really effective if your immediate opponent’s spear was still intact is unlikely.


Nearly all the evidence we have discussed refers to infantry developing from lightly armoured skirmishers into the heavily armoured hoplite phalanx. There is, however, evidence for a few other troop types. Although Greeks for the most part abandoned the chariot for the riding horse during the tenth century, the existence of Greek cavalry is not certain before the Peloponnesian War. There are plenty of vase illustrations of armoured warriors on horseback but, as Greenhalgh shows (1973, passim), these are hoplites on horses, who would have dismounted to fight in battle. This is certain because the warriors carry hoplite shields, which cannot be used on a horse; and also they are invariably accompanied by a squire, who would hold the horses while the battle raged and be ready for a quick getaway if necessary. The horses were in fact being used in the same way as chariots were used earlier, as a taxi service to and from battle.

Two questions arise from this. Did all hoplites ride into battle, and did the squires fight as light troops, whether mounted or not? The first question is likely to be answered with a ‘no’. It takes a lot of land and money to look after horses (and to own a squire), and so only a small percentage of hoplites could have afforded to ride to battle. We see a reminder of this when just 300 Spartans were present for the Battle of Thermopylae in 480. These men were the personal bodyguard of King Leonidas, called the Hippeis, or ‘Horsemen’. They suggest the number of hoplites that might have ridden to battle in an earlier period – a small number, that could sometimes be sent on ahead as an advance unit or for small actions. The Theban Sacred Band of 300 may have had a similar origin, or may indeed be copied from the Spartans.

As for squires taking part in the battle, this is certainly what Herodotus suggests at a later period, at the Battle of Plataea in 479, when the Spartan squires (all 35,000 of them!) were supposed to have fought as light troops, along with 34,500 from the other Greek states! However, he fails to then mention any role they played in the battle itself (Herodotus IX, 29). In the sixth century there are certainly signs of light cavalry use, which may have come about from the squires of hoplites joining the fight. Greenhalgh shows many illustrations (1973, pp. 112, 114, 115, etc.), which seem to depict light cavalry fighting with javelins in battle, as opposed to hunting scenes.

Literary evidence is more scant. Athens seems to have had just ninety-six horsemen for coastal patrols organised under the reforms of Solon in 594 (Bugh 1988, p. 5). In the Persian Wars of the early fifth century, no cavalry fought on the Greek side at all, although this may be because it seemed pointless to use a small force that would be so heavily outnumbered by the Persian cavalry (Bugh 1988, pp. 10–11). Finally, we know for a fact that the Spartans did not raise a cavalry force until 424 (Bugh 1988, p. 24). In northern Greece, the situation was somewhat different. The wide plains of Macedonia, Thessaly and, to some extent, Boeotia meant that these states were provided with cavalry. Indeed, Thessalian cavalry was often used by Athens and other states as a mercenary cavalry. They seem to have used javelins and not to have been armoured at all at this period (Snodgrass 1967, p. 86). The Greeks of south Italy and Sicily also developed a true cavalry before 500, some of whom may have been armoured (Snodgrass 1967, p. 87). There are vase illustrations of what appears to be true cavalry from c. 580 (Greenhalgh 1973, pp. 96 ff.) but, as the literary evidence shows, it was very much a small part of the military scene until the Peloponnesian War.

Similarly, the use of light troops is elusive. In the late eighth century they are clearly very common, especially archers, and appear frequently on the Late Geometric vases of the Dipylon Master and his followers (Ahlberg 1971, p. 44). This was the skirmishing warfare which preceded the hoplite phalanx and, for a brief period, the archer had a central role in warfare. Pausanias, writing much later, mentions their use by Spartans in the Messenian Wars of the eighth and seventh centuries (Snodgrass 1967, p. 81), but by the time of the introduction of the hoplite phalanx in c. 650 the archer, slinger and javelin thrower had mostly disappeared. As we have seen, the phalanx did not appear overnight, however, and there are some interesting pieces of evidence showing the transition. Greenhalgh (1973, p. 99) illustrates a vase showing early hoplites possibly fighting in a phalanx formation but, in between them, there is an archer also taking part in the battle. Tyrtaeus, a mid-seventh-century poet, after he talks about the phalanx standing shoulder to shoulder, exhorts the light-armed troops to hide behind the shields of the hoplites and then hurl their stones at the enemy (Sage 1996, p. 29). This last quote seems to refer to ordinary rocks rather than sling stones, for which the evidence is very thin at this period.

There is plenty of other evidence for archery in Greece in the seventh and sixth centuries, but much is concerned with hunting and, once the phalanx was in place, only Crete supplies evidence for the use of the bow in combat. Indeed, just as Thessalian cavalry was used by other states, so Cretan archers were used as mercenary troops.

Evidence for javelin throwers is also thin after the introduction of the phalanx, although they are again mentioned by Tyrtaeus in the mid-seventh century (Snodgrass 1967, p. 79). Otherwise they too did not resurface until the Persian Wars of the early fifth century. The widespread adoption of the phalanx seems to have developed into a semi-ritualised sort of warfare between Greek states. The heavily-armed phalanx would have been very vulnerable to missile troops, and could have been devastated (once defeated) by a cavalry force, but these troops seem to have simply not been used (Hanson 1991, p. 232). It is unfortunate that we have no reliable accounts of hoplite battles in the seventh and sixth centuries, which might have told us more about the use of missile troops.

Only when the threat of the Persian Empire grew, and the Greek states knew that they would have to fight a non-Greek enemy who had thousands of cavalry and missile troops, did they start to adapt the phalanx, and also to introduce or reintroduce cavalry and missile troops.