¹ 13.159–68; 210–15; 240–329.

² Geddes, A. G., ‘Who's Who in “Homeric” Society?’, CQ 34 (1984), 36CrossRefGoogle Scholar.

³ Andrewes, A., ‘Phratries in Homer’, Hermes 89 (1961), 138Google Scholar. As such ‘schemes of organisation’, dropped immediately after their invention, Andrewes lists: the five stikhes of the Myrmidons, the mention of purgoi, and the advice of Nestōr, all of which will be dealt with below; the five contingents formed to attack the wall around the ships (12.86–104) — a scheme which, in fact, is maintained thoughout the storming of the wall, the poet taking care to point out that he is switching to another point of attack when dealing subsequently with the actions of Asios, Hektōr and Sarpedōn (12.110ff., 195ff., 290ff.); the thousand watchfires of the Trojans and the Achaean patrols (8.562–3 and 9.80–8), which obviously must be dropped as soon as the night has passed and battle must begin again; and the armour–exchange and mass–retreat episodes (14.370–7 and 15.294–9), which are not organisational schemes, but temporary ‘manoeuvres’.

⁴ E.g. Snodgrass, A. M., ‘An Historical Homeric Society?’, JHS 94 (1974), 114–25CrossRefGoogle Scholar.

⁵ Kirk, G. S., ‘War and the Warrior in the Homeric Poems’, in: Vernant, J.-P. (ed.), Problèmes de la guerre en Grèce ancienne (Paris, 1968)Google Scholar, see note 8. Compare Couissin, P., Les Institutions militaires et navales (Paris, 1931), 19Google Scholar.

⁶ Finley, M. I., The World of Odysseus ² (London, 1977), 74Google Scholar.

⁷ Albracht, F., Kampf und Kampfschilderung bei Homer (Naumburg a.S; 1886 and 1895)Google Scholar; E., and Lammert, F., ‘Schlachtordnung’ RE ii.a.l (1921), 436–81Google Scholar; Kromayer, J. and Veith, G., Heerwesen und Kriegführung der Griechen und Römer (Munich, 1928)Google Scholar. Other works that deal with aspects of Homeric warfare without doubting its plausibility are: Küsters, A., Cuneus, Phalanx und Legio (Würzburg, 1939)Google Scholar; Jeanmaire, H., Couroiet courètes (Lille, 1939)Google Scholar and Mireaux, E., La Vie quotidienne au temps d'Homère (Paris, 1954)Google Scholar.

⁸ Kirk, op. cit. (n. 5), 113.

⁹ In detail: Latacz, J., Kampfparänese, Kampfdarstellung und Kampfwirklichkeit in der Ilias, bei Kallinos und Tyrtaios (Munich, 1977), 83–5Google Scholar; Albracht, op. cit. (n. 7), 28.

¹⁰ Finley, loc. cit. (n. 6).

¹¹ Boeotians: 50 x 120 (2.509–10), Myrmidons: 50 x 50 (16.168–70). The standard crew seems to consist of 50 men, plus two leaders. However, crews of 20 appear as well. Exceptions to the usual number of leaders per contingent are, apart from the Boeotians, the Epeans (4), and the Dardanians and Argives (3). Twenty-six groups have one leader only, and of the fifteen contingents with two leaders in eight cases these are brothers.

¹² 16.171–97. There are two less explicit parallels to this: see section 2.4.

¹³ 16.164–6. Cf. section 2.4. In the Lycian contingent, too, there are hēgētores other than Sarpēdōn and Glaukos, the ‘official’ leaders (16.495). Finley, 's argument that ‘hēgētores ēde medontes’ is an empty formula meaning little more than ‘the men’ in general (Economy and Society in Ancient Greece [Harmondsworth, 1983], 220)Google Scholar, does not stand up to closer examination. In most cases the phrase clearly refers to leaders, and of the less clear cases there is not a single one in which it has to be translated as ‘men’. It is hard to see how a word with a meaning as obvious as hēgētores could be misapplied, anyway.

¹⁴ The other Trojan leaders are mentioned at 12.196, 13.691, 15.337; in general, it is said that there are Trojan agoi apart from Hektōr, 12.61, 17.335. Athenian leaders emerge at 13.196, 691, 15.337, the Epean leaders at 13.692 and 15.519. Although the poet generally treats Megēs as the leader of the Epeans, the Catalogue of Ships gives him ‘Dulichium and the Echinean Islands’ only (2.625–6). Other ‘uncatalogued’ leaders are Hyrtios of the Mysians (14.512), Mentēs of the Cicones (17.73) and Iphitiōn of the Maeonians (20.383).

¹⁵ 11.301–3.

¹⁶ 16.532–5; 13.91–3.

¹⁷ The presence of many leaders other than the main ones has, to my knowledge, been recognised only by Küsters (op. cit. (n. 7), 4) and Lammert and Lammert (op. cit. (n. 7), col. 438). Further explicit utterances on the subject come from Kromayer (op. cit. (n. 7)) and Couissin (op. cit. (n. 5), 16–17), both of whom deny the existence of any subdivision of leadership, apparently restricting their information to the Catalogue of Ships.

¹⁸ 11.343–4; 12.330; 13.491–2; 13.800–1. There are many more examples.

¹⁹ 13.709–18.

²⁰ 5.25–6,48, 165, 325; 6.52–3; 13.641; 16.506, 665; 17.189; 21.32; and cf. note 26. Hetairoi are also often said to defend a wounded or dead man, or carry one out of the fight. In these cases it is not always clear if the followers of the victim are meant, or his hetairoi in general, including other leaders.

²¹ Alas' men are implicitly said to be heavily armed and explicity to ‘fight hard in front’ (prosthe mamanto, 13.719–20). Akhilleus' therapontes are called ‘close-fighting’ (ankhemakhoi, 16.272, 17.165), and Patroklos in particular is by implication said to have fought hard at Akhilleus' side (16.244–5).

²² Most recently in Greenhalgh, P. A. L., ‘The Homeric therapon and opaon and their Historical Implications’, BICS 29 (1982), 69–80Google Scholar. Compare e.g.: Nilsson, M. P., Homer and Mycenae ² (New York, 1968 [1933 ¹]), 230–3Google Scholar; Jeanmaire, op. cit. (n. 7), 98–104; Mireaux, op. cit. (n. 7), 63 and 103; Finley, op. cit. (n. 6), 103–4; Kirk, op. cit. (n. 3), 113; Garlan, Y., War in the Ancient World (London, 1975), 24Google Scholar; and Murray, O., Early Greece (London, 1980), 54Google Scholar.

²³ 11.336–7. First a dual is used to indicate the resisting heroes, and later, when Diomēdēs has withdrawn, Odysseus is said to be alone (oiōthe), ‘none of the Argives remained with him’ (11.401–2). Shortly afterwards Aias seems to be annihilating the Trojans single-handedly, yet when Hektor rides to the rescue, he ‘avoids Aias’ and attacks ‘the ranks of the other men’ (11.540–2). In other cases a few leaders joining forces turn into ‘strong phalanges’ 13.91–3 with 13.126–35, 145–54; and 16.532–5, 553–5 with 16.563–6. That six Greeks oppose an ethnos of Trojans (13.478–9,491–2), and that an unspecified number of hēgētores is an ethnos (11.587–95), points in the same direction. ‘Falling in heaps’, finally, occurs when leaders fight over the body of Patroklos (17.215–9, 248–61, 360–2).

²⁴ 12.331–72 and 17.702–6. In both cases the context makes it clear that they are not accompanied by followers.

²⁵ However, a group of leader and followers probably does not always stick closely together. The manner of fighting would seem to make for frequent fissioning and fusion of groups and for the occasional separation of individual warriors from them.

²⁶ 13.598–600; 12.372; 15.445–57. Dēiphobos has a charioteer as well: 13.535–7.

²⁷ With the exception of Greenhalgh, whose position, though argued along different lines, comes very close to mine (see n. 73). Of the others Nilsson, Jeanmaire and, presumably, Garlan opt for the first alternative, Mireaux and Murray for the second. Finley would seem to incline towards the latter, since he speaks of retainers in battle and at the same time claims that we do not know how (the rest of?) the army was recruited. Kirk, too, perceives noble bands, plus a mass of commoners, but argues that these are incompatible forms of organisation. Those who have not recognised the existence of followers imagine that the contingents are indivisible units under single leadership, except Küsters, who thinks that the contingents consist of tactical units of about 50 men, each under a leader. These fight in ‘deep column’ formation. The king is supreme leader and does not lead a particular unit. References in n. 22.

²⁸ Indeed we are told that Idomeneus left the field with unspecified hetairoi, though not that they returned with him. Also, it may be more than a conventional epithet when the poet says ‘thus Idomeneus and Mērionēs, leaders of men (agoi andrōn) went into battle’ (13.304–5). The existence of other Cretan agoi is proved by 3.231.

²⁹ 4.457–65.

³⁰ Pairs are e.g. Antilokhos and Menelaos: 5.561–89; Menelaos and Aias: 17.113–24, 11.463–89; Aineias and Pandaros: 5.166–310; Diomēdēs and Odysseus: 11.312–400. Groups of leaders occur at e.g. 13.91ff., 478ff., 16.532ff.

³¹ 17 366–80– 715–54

³² 4.532–3.

³³ 13.685: ‘laones helkekhitōnes’; 689: ‘hoi Athēnaiōn prolelegmenoi’.

³⁴ 13.712–18,721–2.

³⁵ This is my interpretation of the scene described in 16.155–220. Part of the leadership is provided by Patroklos, but normally all would be Akhilleus' responsibility. My reading of the passage involves taking hēgētores (164) literally; assuming that, as usual, their (and later Patroklos', 218–20) followers are present by implication; and further assuming that the hēgemones around whom the men are placed ‘well-divided’ (eu krinas, 198–9), are the leaders mentioned immediately before (173, 179, 193, 196–7).

³⁶ 4.293–6. At 12.139–40 the contingent of Asios, too, is said to advance around him and five others, and at 3.1 all men are said to be drawn up ‘with the hēgēmones’. It is generally recognised that the further details of Nestōr's arrangements (4.297–300) and the content of his speech (301–10) seem inappropriate in the context of Homeric warfare. Although even on my interpretation the passage is awkward, it does become possible to fit Nestōr into the general picture. One might assume that his arrangements — the chariot(s) ahead, the men on foot behind them — apply to each individual band of warriors, rather than to the contingent as a whole. Thus, when he advises the charioteers ‘not to fight alone, ahead of the others, nor to fall back’, ‘the others’ (alloi) would refer to the other members of the band, rather than to the other charioteers of the contingent.

³⁷ Drawing up: 2.525, 558, 704, 727. Inspecting: 3.196; all of the army is ‘inspected’ by Agamemnōn, 4.231, 250. And cf. the figurative use of the expression in 11.264, 540. Encouraging: 2.589, 4.254. Menestheus and Nestōr are commended for their consummate skill in these matters, 2.552–5.

³⁹ 16.210–7.

³⁹ This passage, with two others, has been taken as the ‘hard core’ of so-called hoplite passages. Supposedly, it is an ‘interpolated’ description of a classical phalanx (Webster, T. B. L., From Mycenae to Homer [London, 1958], 214–20Google Scholar; Kirk, G. S., ‘Objective Dating Criteria in Homer’, Museum Helveticum 17 [1960], 194Google Scholar). This, of course, could only be true if there were ordered columns and lines. Another objection is that the Myrmidons are only standing and waiting in this way — not marching or fighting, which is essential to hoplite warfare. When they move, they do so ‘in throngs’ (aollees, 16.276).

⁴⁰ 17.466ff. Asios' five leaders (cf. n. 36) are all mentioned again when they die(12.193;13.506, 545ff., 560ff.). Of the Mymidon leaders, Phoinix plays a part in the non-fighting events, but none of the others ever reappears. Nor do Nestor's men.

⁴¹ Küsters (op. cit. (n. 8), 4–5 and 171 n. 57), Lammert and Lammert (op. cit. (n. 8), col. 438), and Lammert, F., ‘Phalanx’, RE xix.2 (1938)Google Scholar, col. 1626, think of stikhes and phalanges as tactical units. Kromayer (op. cit. (n. 7), 23) equates phalanx with regional contingent. A very different view is taken by Latacz (op. cit. (n. 9), 45–67). He argues that both are single lines of men, not units of leadership at all. This comes close to the classical military use of the words. It should be noted, however, that this usage only emerges with Xenophon. Before him, formations are called ‘taxis’ or ‘lokhos’. From the argument below, it will be evident why I think none of these interpretations is correct.

⁴² The first of the five Myrmidon section-commanders is said to lead' the first stix'; the others lead simply ‘the second, the third’ etc. (16.173ff.). The word stix does not, by the way, actually occur in this nominative singular form. The same meaning of the word is appropriate in other passages: Skhedios and Epistrophos ‘draw up the stikhes of Phocians’ (2.525). Stikhes are present in the contingents of Odysseus (3.196, 4.330), Pandaros (4.90) and Makhaōn (4.202) when these are standing ready for battle. As long as Akhilleus refuses to fight, the Myrmidons are not led ‘into stikhes’ (2.687). Stikhes are inspected by Odysseus and Agamemnōn (cf. n. 37).

⁴³ The references are, in the order cited: 4.90, 201,330; 7.61,65; 4.221; 11.412; 17.107; 5.166, 746; 8.390; 11.188, 203; 17.510; 20.353; 13.680; 15.615; 5.461. Stikhes and leadership are associated once in 17.107: ‘meanwhile the stikhes of the Trojans came closer; Hektōr led (them)’. This only shows that Hektōr led the Trojans in general. As ‘supreme’ leader he would not be leading a particular stix, anyway.

⁴⁴ 20.325–9.

⁴⁵ 20.362. The victims are Iphitiōn (382ff.), Dēmoleōn (395ff.) and Polydōros (407ff.).

⁴⁶ In a broad sense, because ‘any cross-section from front to rear’ does not necessarily imply a section of greater depth than width, let alone a single file, which is the more usual meaning of ‘column’. As far as etymology is concerned, I think it quite probable that stix, related to steikho, ‘to move forward’, could mean something like ‘a group of men, facing and moving towards the enemy’. Cf. Küsters, ' ‘Marschkolonne’ (op. cit. (n. 7), 171 n. 57)Google Scholar.

⁴⁷ Referring to the men as stikhes, then, would have a different nuance from, say, ethnos or laos: it implies a certain structure within the mass. The expression kata stikhas, used when somebody sees or shouts something ‘through the stikhes’, does in fact in two or three cases definitely mean looking or calling to one's left or right, rather than forwards or backwards (11.343,17.84; 15.353 is less clear; 5.590,11.91,16.820 are inconclusive). Likewise, when Aineias passes by ‘many stikhes’ this means he moves a great distance along the front (he ends up ‘at the very edge of the battlefield’, 20.328), rather than away from the front to the rear. Finally, when Akhilleus says he will penetrate ‘a stix’ this means he will not fight in the usual manner, attacking only the men first in line along the front, but intends to press on until he reaches the enemies’ rear.

⁴⁸ 4.427–9. In two other passages (5.591 and 11.344) phalanges are said to ‘follow’ Hektōr. This is as vague as the statement that he leads stikhes (n. 43), or, e.g. that he leads Trojans ‘advancing in throngs’ (13.136).

⁴⁹ There are phalanges ‘of the Athenians’ (2.558), within the Cretan contingent (4.254); and ‘with the Aiantes’ (4.280–1). The other references are 8.279, 11.503, 19.152, 6.6, 7.141, 11.90, 12.718, 15.408, 13.806–7, 4.254, 6.83, 13.90. Furthermore, phalanges can start to fight (4.281, 332,427; 19.158). They can be stopped (3.77, 7.55,11.567), ‘cut through’ (16.394), brought into disorder (5.93, 96; 11.148; 15.448; 16.280), ‘strengthened’ (11.215, 12.415; 16.563), and scattered (17.285). The only lime phalanx (singular) occurs is 6.6, when we hear it is ‘broken’ by Aias — nothing else.

⁵⁰ ‘Prōtas phalangas;’, 16.394; ‘pumatas phalangas’ contrasted with promakhoi, 4.254; ‘epassuterai’, ‘like the waves of the sea’, 4.427–9 (cf. n. 48).

⁵¹ Phalanges taking their place in the battle order: 2.558; withdrawing, 13.83–90; and subsequently moving into the fight again: 13.126–35. The translation of 13.126 and 13.90 may be stretched to give these lines a different meaning from the one adopted here. The verse ‘amphi d' ar' Aiantes doious histantophalanges’ (13.126), which I translate (quite literally) as ‘phalanges went and stood around the two Aiantes’, might be more freely interpreted as ‘they went and stood around the two Aiantes (in the form of) phalanges’. Cf. Michel, C., Erläuterungen zum N der Ilias (Heidelberg, 1971), 43, 45Google Scholar. The phrase ‘(Poseidōn) krateras ōtrune phalangas’ (13.90) — ‘encouraged the strong phalanges’, in my translation — might in the same way be taken to mean that he ‘encouraged the (Greeks, so that they formed) strong phalanges’. Cf. Ameis, K. F. (bearb. C. Hentze), Homers Ilias (Leipzig, 1888)Google Scholar, ad 13.90). While my version implies that phalanges were existing entities in both situations, the alternatives imply that they were not, and had to be formed from a disorganised mass of people. I prefer my version not only because it is more literal, but because I suspect the freer translations to have been inspired by the commentators' preconception that the phalanx must be a formation of some sort, and cannot be anything else.

⁵² See n. 48. In these two passages it is clear from the context that the movement is to the left or right (kata stikhas, cf. n. 47).

⁵³ There are places where ‘most’ (pleislai, 11.148) or ‘by far the most’ (polupleistai, 15.448) phalanges are present.

⁵³ 13.125ff., cf. 13.91ff.; 16.563ff., cf. 532–5 and 553–5.

⁵⁵ The etymology is discussed in Lammert (op. cit. (n. 41), col. 1625), who gives all the relevant passages. The ‘log’ -interpretation is in Küsters (op. cit. (n. 7), 171 n. 57) and Latacz (op. cit. (n. 8), 53).

⁵⁶ Since stikhes and phalanges, on the latter interpretation, are very close in meaning, one might suspect that they are, after all, synonyms, and that the apparent differences between them are merely differences in formulaic usage. I am not able to judge to what extent metric requirements may have led to the choice of one word rather than the other. Only one established formula — ‘kata stikhas’ — is used in the passages that constitute my evidence for the differences between the two. In any case, my argument is not affected even if they are synonyms.

⁵⁷ Yet another etymological argument is slightly in favour of interpreting phalanges as bands of followers. The later change to the singular to indicate the entire army (cf. n. 41) could reflect an organisational change: the epic army with its independent leaders and their phalanges has been replaced by an integrated city-state army, which constitutes a single phalanx under central leadership. The transition is not so easy to explain when one takes phalanx as ‘part of a whole’.

⁵⁸ Kiefer, T., The Tausug (New York, 1972), 71–3Google Scholar; 80. Kiefer also mentions, as things of the past, ‘maximal alliances’, aggregates of ‘medial’ ones (ibid. 74): this would parallel the epic aggregate of contingents, but see below. With respect to feudal armies, it may be noted that Nilsson (op. cit. (n. 33), 229ff.) and Jeanmaire (op. cit. (n. 7), 28–30, 111) go so far as to claim that the Homeric bands are in fact based on feudal obligations towards the leader.

⁵⁹ Agamemnōn deals with those he thinks too slow in moving towards the enemy (Odysseus, Menestheus and Diomēdēs, and their men) in harsh terms, 4.327–422. It is a problem at any time during the battle, because there are always men who stay away from the real fight, but one can see that the problem would be most acute during the advance.

⁶⁰ Cf. Turney-High, H. H., Primitive War, its Practice and Concepts (Columbia S.C., 1971), 43–4Google Scholar.

⁶¹ E.g. Andrewes, op. cit. (n. 3), 132.

⁶² 2.802–6. ‘Those’ is emphasized: hoisi per. See section 3.3.

⁶³ 2.362–8.

⁶⁴ Küsters is, it seems, the first to state that Nestōr's advice is not followed (op. cit. (n. 7), 4). Finley, (Early Greece. The Bronze and Archaic Ages [London, 1970], 84–5)Google Scholar agrees, and so does Andrewes, who finds the advice ‘not… appropriate’ and supposes it to reflect the military organisation of the time of the poet, ‘an intrusion’ in an otherwise more ancient type of organisation (op. cit. (n. 3), 131–2). He considers this a ‘natural guess’, and a guess it is, though ‘natural’ only within a particular tradition of Homeric scholarship. Moreover, as far as I can see, there is no evidence for a tribe-and-phratry army organisation at any time in Greek history (see. n. 74).

⁶⁵ 2.439–46 (krinontes); 474–7 (diekosmeon entha kai entha).

⁶⁶ That ‘purgos’, literally ‘wall’ or ‘tower’, is a word for ‘contingent’, has been assumed by Lammert and Lammert (op. cit. (n. 7)), Kromayer (op. cit. (n. 7), 23), Couissin (op. cit. (n. 5), 17), Küsters (op. cit. (n. 7), 4–5), Latacz (op. cit. (n. 8), 52). There is one scene only in which the word purgos seems to be used in this sense, twice. Odysseus, Menestheus and their men are ‘standing and waiting for the moment that another purgos of the Achaeans would attack the Trojans…’ (4.333–5). Agamemnōn abuses them for holding back: ‘Gladly would you just watch, even if ten purgoi of the Achaeans would be fighting in front of you’ (4.347–8). From these contexts it is not absolutely clear that contingents are meant. I think purgos is used in a metaphorical sense: a leader (and his men) might be called a ‘wall’ or ‘tower’ of the Greeks, in the sense of being their ‘bulwark’. This is in fact what we find in the Odyssey: the Greeks lamented the death of Aias, since he was ‘such a purgos to them’ (11.556). Here, the metaphor is clearly understood; the poet does not even need to say ‘like a purgos’ (cf. Kallinos J, 20). Furthermore, when the phrase ‘purgēdon’ is used, it does not mean ‘by contingents’ as one would expect on the current interpretation, but just ‘like a wall’ (12.43; 13.152; 15.618). Inscriptions of later date (CIG 3064,3081 et al.) attest to the existence on Teos of geographical or social units called purgoi. Here, a purgos is some section of a community, but in the Homeric context, where Odysseus and Menestheus and the others are (still) leading entire contingents, this meaning would be inappropriate.

⁶⁷ In the epic the word sometimes takes on a very broad sense, meaning almost ‘races’, and at other times may mean a descent-group as small as a single family. Cf. Andrewes, op. cit. (n. 3), 132.

⁶⁸ Another explanation of the relation between ‘tribes and phratries’ and ‘contingents’ is to equate each contingent with a single tribe, and to assume that it was subdivided into phratries (Eustathius, ad 2.362; Couissin, op. cit. (n. 5), 17; presumably Kromayer (op. cit. (n. 7), 23), too). Of this subdivision, however, there is no further trace, unless one assumes with Glotz, (La Solidaritè de la famille en Grèce [Paris, 1904], 13, 89–90)Google Scholar, that phratries are the bands of followers. This interpretation would suit my view of army organisation rather well, but unfortunately there is nothing in the epic to support the equation of phrateres and hetairoi.

⁶⁸ 12.86–7 (pentakha kosmēthentes ham' hēgemonessin). In the subsequent description we find that the fourth ‘column’ corresponds to Aineias' contingent in the Catalogue (12.98–100; 2.819–23). The first group, led by Hektōr and others, may correspond to the original Trojan contingent (12.88–92; 2.816–18). The second group, however, contains more Trojans and more Dardanians (12.93); the third group still more Trojans, and the men who in the Catalogue formed Asios' contingent (12.94–7; 2.835–9); and the fifth group consists of the ‘allies’ (12.101–4), which means about a dozen of the Catalogue's contingents! Their massed advance: 12.105–6.

⁷⁰ 11.57–60. Hektōr is said to be giving orders, now in front, now in the rear (61–5), which would be in accordance with the idea that he is the supreme commander, taking his place outside of the main body (cf. 2.4), while the other five lead sections.

⁷¹ Stikhes: 4.231 and 250. The ‘stops’ are: one, Idomeneus and Mērionēs; two, the two Aiantes; three, Nestōr; four, Menestheus and Odysseus; five, Diomēdēs and Sthenelos (4.251–419).

⁷² 2.805–6. On the Trojan side we have only this general statement of the effect of the measures proposed, no individual examples. On the Greek side we have no general statement, but many examples of this effect: see n. 42.

⁷³ Greenhalgh, P. A. L., ‘Patriotism in the Homeric World’, Historia 21 (1972), 528–37Google Scholar, also argues that ‘the Panachaean War is both conceptually and practically alien’ to the world depicted in the epic (531, 533). His arguments differ from mine, and unfortunately they do not seem to me to be valid. He thinks the Panachaean organisation is Mycenaean, and that the poet shows his ignorance of the Mycenaean feudal system by (i) providing a vague (and mistaken) explanation for the obligation of the lesser kings to follow Agamemnōn; (ii) not explaining the letter's right to a very large cut of the spoils; and (iii) disregarding the distinction between wanax and basileus. This does not prove the Panachaean army is a conception alien to the poet, it only proves that he does not think of it as organised in a Mycenaean and/or feudal way. Indeed, Greenhalgh himself states that, actually, Agamemnōn's ‘relation to the Panachaeans in the matter of obligation is, on a grand scale, exactly the same relation as that of the king to his own citizens in the single-state situations’ (533). This form of organisation is evidently not alien to the poet at all. The question, of course, should be whether such obligations ever really existed on a Panachaean scale, and Greenhalgh's only argument against this is that they seem ‘overstrained’ (op. cit. (n. 22), 81). Nevertheless, Greenhalgh's view of the organisation of real-life warfare (which he would date to Geometric times) is, by implication, more or less the same as mine.

⁷⁴ Hesiod, , Theogony 935Google Scholar; Tyrtaeus 12.21; Mimnermos 14.3. Later it appears only once, in a story about Troy told by Bacchylides (15.40–3), where phalanges of the Trojans are called to the ‘army–receiving’ agora. Those who think that Homeric armies are organised in tribes and phratries might see another bit of positive evidence in Tyrtaeus' reference to the three Dorian tribes (1.12). This verse, however, is hardly good evidence for their contemporary role in war (cf. Hammond, N. G. L., ‘The Lycurgean Reform at Sparta’, JHS 70 [1950], 51CrossRefGoogle Scholar). Athenaeus' reference to phratries at the Karneia (4.141ef) is worth even less. Tribes are, of course, army units in Athens, but these are Kleisthenes' ten tribes, and hence they date from 501 B.c. — a very late terminus ante quern for Homeric warfare.

⁷⁵ For the ‘at a stroke’ theory, see Lorimer, H. L., ‘The Hoplite Phalanx’, ABSA 42 (1947), 76–138Google Scholar. For the ‘gradual development’ theory, see Snodgrass, A. M., Early Greek Armour and Weapons (Edinburgh, 1964)Google Scholar; idem, ‘The Hoplite Reform and History’ (JHS 85 [1965], 110–22)CrossRefGoogle Scholar. The suggestion that the development of the ‘true’ phalanx started only in the late fifth century, is George Cawkwell's (Philip of Macedon [London, 1978], 150ff.), labelled ‘heresy’ by Holladay, A. J. in his attempt to refute it (‘Hoplites and Heresies’, JHS 102 [1982], 94–7)CrossRefGoogle Scholar.

⁷⁶ The Lelantine war is conventionally dated to c. 700 B.c. Another early ‘exception’ would be the First Sacred War (c. 600 B.c.) by the Delphic Amphictyony versus the Krisaians. Amphictyonic Leagues were established at early dates, but unlike symmachies, they rarely engaged in warfare. The Peloponnesian League's battle order was well established in 479 B.c. at the latest. In this year the Tegeans were able to claim a traditional place in it, next to the Spartans (Herodotus 9.25–6).

⁷⁷ For convenience, I have mostly referred to a single ‘supreme’ leader. His tasks, however, can be shared by two, and possibly more, men, as is shown by Idomeneus, who leads the men in front, while Mērionēs commands the rear phalanges (4.253–4). The number of ‘section’ -leaders seems to be five, but I suppose that this varied with the size of the army. Though the number of five occurs quite frequently in the epic, judging from Germain's study (Homère et la mystique des nombres [Paris, 1954], 37, 49, 59Google Scholar), it does not have any specific mythical, ritual or social connotations that might explain its appearance here. Maybe this implies that the number of five sections is more or less realistic.

I was enabled to write this article by a British Council scholarship. Earlier drafts were read and criticised by S. C. Humphreys, M. M. Willcock, A. M. Snodgrass, C. J. Ruijgh and the editors. Their comments inspired thorough revisions. They are, of course, not necessarily in agreement with my views. Quotations from the Iliad have been freely borrowed and adapted from E. V. Rieu's translation (Harmondsworth, 1950).