I don't think anybody would seriously contend that Macedonian units had no fixed establishment. The precision of their manoeuvres, such as exhibited by Alexander's men at Pelion and Issos, would indicate a high degree of regimentation even without the surviving Hellenistic tactical manuals' prescription of fixed infantry and cavalry unit sizes.
Turning first to the infantry, Hellenistic tactical manuals exist by Ailian, Arrian and Asklepiodotos that give us some insight into the organisation of Macedonian-style infantry phalanxes. Each is similar to the other, and they all probably derive from a lost manual of Polybios'. It has been greatly debated to what extent these manuals reflect Hellenistic military reality, and to what extent they are the workings of armchair philosophers; it is also debated to what extent they reflect later Hellenistic organisation, rather than that under Philip or Alexander. Ailian for instance claims (0.6) his work represents Macedonian formations under Alexander, but all the sources he cites are post-Alexandrian (1.2).
Asklepiodotos' infantry phalanx organisation is as follows:
The other manuals give virtually identical presecriptions for their idealised phalanxes. It should go without saying that a 'chiliarchia' would not have had exactly 1024 men in it: this is an ideal establishment strength and its actual numbers would have varied somewhat as attrition occurred. However this has not stopped some scholars attacking these manuals because they deal with parade-ground strengths, despite firm literary evidence that the later Seleucid and Macedonian phalanxes were indeed 16000 men strong.1
How then does this organisational list conform to the historical sources for Alexander's earlier Macedonians? Although it was Philip that first created, organised and equipped the Macedonian phalanx 2, details from his reign are so sketchy that virtually all the evidence comes from after his death. From Polybios (12.9) it is obvious that the Alexandrian file depth was either 8,16, or 32 men. We can probably discount 32, since this was the depth of the Seleucid phalanx at Magnesia, where it was said to be abnormally deep. A depth of 8 is possible, but this would mean it would have been shallower than most hoplite phalanxes: according to Xenophon (Hellenica, 4.2.18), the allies were 16 deep at Nemea (with the Thebans characteristically formed up much deeper still), the Spartans 12 deep at Leuctra (Hellenica, 4.4.12); the depth at Issus would perhaps be influenced by the river crossing. A depth of 16 men seems most likely, and this accords with the most basic unit formation given above; Ailian says that while files of 8 or 12 were recorded (4.2), he himself sticks to 16 (e.g. 8.3) just as Asklepiodotos and Arrian do.
What then of larger units? Traditionally the 12000 Macedonian foot that Diodoros (17.17.4) records crossing over to Asia with Alexander have been grouped as 3000 Hypaspists (the Macedonian foot guards) and 9000 Pezetairoi (the pikemen of the regular phalanx, some are termed Asthetairoi instead).3 This is because later on, post-Gaugamela, the Hypaspists are said by Curtius to have included 2 chiliarchies (units of 1000) plus the royal battalion, which was presumably the same strength, since after Alexander's death the number of the Argyraspids, the renamed Hypaspists, is recorded as 3000 (Diodoros, 18.58.1), and that the Hypaspists could be divided into thirds (Arrian, 4.24.10, who records 2 chiliarchs of the Hypaspists at 4.30.5, but commanding three chiliarchies). This traditional view is proposed for instance by Brunt in his introduction to his Loeb translation of Arrian's Alexander, and is now so ingrained that Bosworth (Conquest and Empire, p259) could say "Fortunately the size and composition of the Macedonian contingents is not seriously in doubt".
According to this analysis, each of the six battalions 4 of Pezetairoi, as given by Diodoros in his Gaugamela order of battle (17.57.2-3), would therefore number 1500 men each. As this doesn't fit in with the organisation of phalanxes in the Hellenistic manuals, it has been necessary to assume that the organisation of the phalanx of Alexander was therefore different from those later on. Some 'support' for this 1500-strong assumption comes from a passage in Arrian where later on (i.e. post-Gaugamela) Alexander takes some Pezetairoi with him on a side-expedition (3.23.3, 3.24.1) including the units of Krateros, Amyntas and Koinos, leaving 6000 behind at Ecbatana (3.19.7, 3.20.1). It is then further assumed that before then, a 7th Pezetairoi unit was created out of the 6000 Macedonian foot reinforcements that arrived at Susa under Amyntas (Arrian 3.16.10, Diodoros 17.65.1, Curtius 5.1.39) to make these 6000 men left behind equally divisible by 1500. This at least is plausible, since by the time of the battle of Hydaspes there would seem to have been more than six Pezetairoi units (see below). Alas, there is no indication in Arrian that the units of Krateros, Amyntas and Koinos were the ONLY Pezetairoi units taken with Alexander (the units not being detailed together, but being mentioned is separate places), so that the theory only works if only 3 units were taken, AND another created in the meantime.
Another problem with this theory is that the numbers only work out if all six Pezetairoi units recorded at Gaugamela came over with Alexander immediately prior to Granikos. If there were any Pezetairoi already in Asia that joined the army before the battle, so that there were more than 12000 Macedonian foot present, the theory falls apart entirely, since then there would be then more than 1500 men per unit. Alas for the traditional theory, there is evidence that there were more Macedonian infantry present than just these 12000.
Plutarch records that ancient estimates (by Aristobolos, Ptolemy and Anaximenes) of Alexander's forces at Granikos ranged from 30000 to 43000 foot and 4000 to 5500 horse (Alex. 15, and Mor. 327D-E). Arrian, who used Ptolemy as a source, gives not much over 30000 foot and over 5000 horse (1.10.3). Justin (11.6.2) records 32000 foot and 4500 horse while Polybios (12.19.1) has the contemporary Kallisthenes recording 40000 foot and 4500 horse. Diodoros, the only source to give a force composition, gives 30000 foot and 4500 horse (though his figures actually total 32000 foot and 5100 horse). The figures for both foot and horse fall into two main groups: either 30000-plus foot or 40000-plus foot, and either 4500-odd horse, or 5000-plus horse.
Some time ago Brunt proposed that the these two groups of figures could be reconciled by assuming that the lower figures represent the number of men crossing into Asia with Alexander, while the higher figures represent the number of men Alexander had actually at the battle. The difference between the two represents the advance force that Philip, Alexander's father, had previously sent to Asia, which having protected the landing bridgehead, presumably joined up with Alexander before the battle. Indeed, Diodoros states that the numbers he quotes are for those that crossed over to Asia with Alexander, rather than for those with Alexander at the battle itself; Arrian similarly gives Alexander over 5000 horse, and a little over 30000 foot at the crossing (1.11.3) without giving numbers involved in the battle itself. Even though Arrian doesn't record the strength of the advance force, a clue may found in his narrative, since he gives Alexander's strength at Gaugamela as 40,000 infantry and 7,000 cavalry (3. 12. 5) although this is complicated both by the various garrison forces taken from the army on the way, and the various reinforcements recieved.
By Brunt's analysis therefore, the advance force would have consisted of 10000 men in very round figures when it joined Alexander, and Polyainos (Stratagems 5.44.4) does indeed record a figure of 10000 such men. According to Diodoros (17.7.10), this advance force was composed of both mercenaries and Macedonians. The mercenaries are likely to have outnumbered Macedonians: Philip had the money to pay them at the time, and he was in need of his own native Macedonian troops at home to help keep the Greeks down. Diodoros (16.89) hints that allies (i.e. Greeks coerced into fighting by the League of Corinth) could have been present, but goes on to say (16.91) that Philip only sent a portion of his 'own' forces.
It is reasonable to assume Macedonian guards, the Hypaspists, were unlikely to have been sent in the advance force, as Philip's domestic situation was far from secure: his guards were needed at home (Alexander was in exile, plotting his return at the time). In this case the advance force troops might be supposed to be mainly Greek and Thracian mercenaries (Illyria having not been secured in the same way it was by the time Alexander left for Asia), plus some (non-Hypaspist) Macedonians. Aside from a small number of archers attested as Macedonian much later on (Arrian 3.12.2), the only non-guard Macedonian foot were Pezetairoi. While I will argue (in part 2) that Macedonian horse would also have been present, most Macedonians soldiers were Pezetairoi, and a Macedonian army without them would have been strange indeed.
The presence of Macedonian soldiers is further confirmed by Polyainos, who records (Excerpts 28.3) Memnon dressing his men up as Macedonians in an attempt to fool the inhabitants of Kyzikos, who believed them to be Macedonians under Chalcas (named as Kalas in Stratagems 5.44.5). This would make no sense if there were no Macedonian soldiers in Asia to impersonate. One could argue that even if there were no actual Macedonian units there, then the generals would still be Macedonian, but dressing men up as generals to infiltrate a town while the rest of the soldiery were not disguised seems somewhat far-fetched.5
We can reasonably assume that the whole advance force was present at the battle of Granikos, since Alexander had yet to capture any major strong-points that would require him to detach men for garrison duties and the like. But what was their composition and strength? The Macedonian advance force had started out well, but is likely to have suffered losses over the campaign since their opponent, Memnon, was getting the upper hand towards the end. Polyainos (Stratagems 5.44.4) records the Macedonians generals Parmenio and Attalos having 10000 men when attacked by Memnon, but it seems that even this might not have been the entire force, given that the men led by Chalcas/Kalas (called Kallas by Diodoros, 17.7.10) were a substantial enough force that they could offer battle in their own right. So even allowing for campaign losses, the differential between Brunt's high and low groups, over 10000 men, could easily be accounted for by the advance force.
Therefore we have the 12000 Macedonian foot with Alexander, plus an unknown number of extra Macedonians from the advance force. To break these numbers down into Hypaspists, Pezetairoi, etc., I needed to know the strengths of the individual Macedonian units. We know that the Alexandrian phalanx was at the time split into six taxeis based on recruiting areas since Diodoros lists some of these areas in the order of battle for Gaugamela giving the six units (17.2-3); the Companion cavalry were similarly recruited on a territorial basis (if not from exactly the same territories).
Since the traditional strength of 1500 men per taxis can't be reconciled with any Macedonian foot additional to those Alexander brought with him, an alternative needs to be proposed. The Hellenistic manuals talk of 2000-man units: how would these fit the requirements I asked myself? Three 2000-man units divide into the 6000 men left behind at Ecbatana just as easily as four 1500-man units do (Arrian, 3.23-24), so this is no problem. Furthermore, the creation of a seventh 2000-man unit would soak up more of the 6000 Macedonian foot reinforcements that arrived at Susa from Macedonia than a 1500-man unit would (Arrian 3.16.11); it may even be that two such units were formed, giving eight taxeis (see below)
A problem then is how would units of 2000 men split equally into the 12000 men crossing over into Asia if the Hypaspists numbered 3000 men? The most sensible answer would be that the Hypaspists also numbered 2000 men like all the other units of the phalanx.6 I would argue that they were only expanded to 3000 men in the army reorganisation at Susa when the above-mentioned foot reinforcements arrived.
While it might be argued that new recruits would be less suitable for posting to a guard unit than experienced veterans, the Hypaspists were guards first and foremost: political reliability would have been a much more important qualification than length of service. Alexander was at this point removing as many of the 'old guards' as he could from the army, because of their loyalty to Parmenio. Such newcomers would be naturally loyal to Alexander, rather than his predecessors, and Diodoros emphasises this motive of enhancing obedience to Alexander (17.65.4). (An alternative scenario could be that politically suitable Pezetairoi veterans were transferred to the Hypaspists, and the reinforcements filled their places in the ranks of the Pezetairoi).
Curtius (5.2.3) says that at this time the Macedonians were reorganised from companies of 500 men (the 'pentakosiarchia' of the Hellenistic manuals) to units of 1000 (a 'chiliarchia'), but he doesn't say which infantry were reorganised, or indeed if any of the infantry were (infantry are assumed as a 'lochoi' was normally if not exclusively used of infantry).7
The traditional view is that just the Hypaspists were, but this is only because the traditional view demands the Pezetairoi be in battalions of 1500 men, which of course 1000 hardly fits into squarely! It is much more natural to read the passage as saying that all the Macedonians (and indeed, the non-Macedonians too) were so reorganised - and of course a 2000-strong Pezetairoi taxis (or a 3000-strong Hypaspist force) fits this just fine. It may be noted that earlier on in the campaign, before Curtius' reorganisation, Parmenio took 2500 Macedonian infantry to take over the cities of Magnesia and Tralles (Arrian, 1.18.1) - perhaps one taxis plus archers. From this it seems that 500-man units were indeed at this stage then the tactical building blocks of the Macedonian infantry.
This would mean that Alexander brought over to Asia five units of Pezetairoi plus the Hypaspists, giving us one unit of Pezetairoi already in Asia, having been sent as part of Philip's advance force, in order to give us the six units present at Gaugamela. Proposing a 2000-man unit already in Asia could be the key to understanding the surprising troop totals of Diodoros' invasion force: he states 30000 foot while the units he lists total 32000 men. Could it be that he has confused two sources, one saying 12000 men crossing (and including the Hypaspists), and another saying 12000 men at the battle (and not including the Hypaspists) not realising the two would be different? Such an error is all too understandable if the infantry taxeis were 2000-strong, but much harder to account for otherwise.
A further piece of evidence for 2000-man units comes from Alexander's manoeuvrings at Pelion. Arrian (1.6.2) says Alexander his phalanx 120 deep. No matter what interpretation of 'embelon' is favoured (embelon, or 'wedge', being the name of the formation these men were drawn up in), this number does not divide well into 1500, but goes a very convenient 16 times into 2000, the standard Macedonian file depth.8
Finally, when Alexander, fighting the Thracians, crosses the Danube (Arrian 1.3.6), he takes 4000 infantry with him. These 4000 men are Pezetairoi (or Hypaspists, but this doesn't alter the argument if the Hypaspists were organised in a similar manner) because each man uses his 'sarissa' (pike) to level crops as he advances (1.3.4). This would easily equate with two 2000-man units of pikemen, but it is hard to reconcile with 1500-man units.
This 2000-man taxis formulation fits in more neatly with later practices under the Hellenistic kingdoms: virtually all accounts have phalangites in multiples of 2000, and those that do not divide are diviseable by 1000, but not 1500. It also not only accords with the Hellenistic tactical manuals but requires less fragile assumptions than the traditional formulation: this assumes that Curtius' reorganisation was restricted to the Hypaspists, mine makes no such assumption. I assume one unit in Asia, the traditional theory only works if none were present; yet both Diodoros and Polyainos clearly state that Macedonians were present, so that the traditional theory requires an Macedonian force with without any Pezetairoi: the most numerous and characteristic type of Macedonian soldier.
When I was first preparing this essay for submission to the journal Slingshot (a must for every ancient wargamer by the way, check out the The Society of Ancients' web-page, Peter Hall's article appeared in Slingshot where he too put forward an organisation based on units of 2000 men, although seeming to indicate this organisation was only put in place with the reinforcements at Susa. I would argue that the units were 2000-strong from the very outset: the 6000 reinforcements would not have been enough to both create a new 2000 man taxis, and add 500 men to each of the six old taxeis as well as replace the casualties sustained since Issos, especially if another 1000 Hypaspists were formed at the same time.
With regard to the Hypaspists, there has been some debate (now available on-line, for links see the references at the end of part 2) between Bosworth and Hammond about their organisation, especially with regard to inner guard units. Bosworth says there was an inner elite, the Agema, and then the rest of the Hypaspists. Hammond says there was an inner elite, the Agema, then the rest of the Hypaspists, plus another elite Agema outside the Hypaspists. It seems to me that the most plausible interpretation is that there were two elite formations within the Hypaspists. (A parallel can perhaps be found amongst the later Seleucid cavalry, that had two elite units within it - one called the Companions, and the other called the Agema).
Bosworth demonstrates to my satisfaction that the elite Macedonian foot guards were all Hypaspists, contra Hammond. However at Hydaspes (Arrian 5.13.4) we are told of the Royal Hypaspists, the Royal Agema, and the rest of the Hypaspists, and he unconvincingly argues that the first is a subgroup of the second.
In his words: "The description of the attack on Thebes, as we have seen, does not give support to his (i.e. Hammond's) thesis. Nor does the equally controversial description of the deployment of the army after the Hydaspes crossing (5.13.4). Here the nomenclature can be understood as the usual variation of 'hypaspists' and 'royal hypaspists'. First Arrian mentions 'the royal hypaspists whom Seleucus commanded', then 'the royal agema' and finally 'the other hypaspists'. First a corps d'elite, those hypaspists defined by Seleucus' command, then the guard division as a whole, and finally the whole remaining body of hypaspists."
But his explanation does not tally with his own translation: it is much more natural to suppose two separate formations within the Hypaspists rather than the first being part of the second, since this is nowhere even implied by the etxt. Likewise at the attack upon Thebes (Arrian 1.8.3-4) it is easier to understand there being two units within the Hypaspists ('the rest took refuge with the Macedonian agema and the royal hypaspists'), rather than as Hammond argues, one within and one without, or, as Bosworth argues, one within, and none without.
This could easily fit into an organisation based on sub-units of 500 being reformed as sub-units of 1000. One could suppose a Hypaspist force of 2000 men, subdivided into two senior units (perhaps with one taking precedence over the other) of 500 men each, plus the others totalling 1000. When the Hypaspists were enlarged to 3000 men at Susa, the 500-strong units could be enlarged to 1000 men each, to give 3 units of 1000, or alternatively the two 500-strong units would be brigaded together and the others increased to 1000 each. However, while the Hypaspists may have numbered 3000 men in 327 BC, I believe that they had been expanded to 4000 men by 326 BC. Much has been made of Diodoros' statement that the Argyraspids, the renamed Hypaspists were 3000 strong in 318 BC (18.54.1), but at 19.28.1, he says they were over 3000 strong. It must be remembered that those troops that were sent west with Craterus to Cilicia, who included these Argyraspids, were only the oldest veterans, the implication being that some younger men remained (Arrian 7.12.1, with Diodoros 18.12.1, 18.6.4 for their numbers), so that there were more than 3000 of them before most of them wesent homewards. Support for this comes from Arrian (5.23.7) where Ptolemy is given command of "3 chiliarchies of Hypaspists, all the Agrianians and a taxis of archers", so implying there were by that stage more than 3 chiliarchies of Hypaspists and therefore more than 3000 men. Quite possibly there were these three chiliarchies, plus the agema, for a total of 4000 men. It should be noted that sometime before the battle of Hydaspes Arrian (5.13.1) has Alexander take command of half the hypaspists, which would be tricky if there were still only 3 units of them.
A final piece of evidence comes from the combination of two sources I believe have not been related together before. The Roman historian Dio records (78.7.1-2, see David Karunanithy's article in Slingshot 213, 33-40) that when the emperor Caracalla recreated an entire Macedonian phalanx, was said to be 16000 strong, just as the Hellenistic manuals claim their (and therefore Alexander's) phalanx was. Although there were only 6 taxeis at Gaugamela, by the battle of Hydaspes there were seemingly more. Arrian in fact names 11 taxiarchs in the Indian campaign (see Brunt's notes to the Loeb Arrian, p487), but some of these were not always commanding infantry at the time - Koenos (5.16.3), Kraterus (5.11.3), and Kleitos (5.22.6) all seemingly commanded hipparchies at the battle of Hydaspes, so there may have been only 8 actual taxeis, and 8 taxeis of 2000 men would fit Dio and the 3 Hellenistic manuals admirably.
Let us assume then that Alexander had 4000 Hypaspists and 16000 Pezetairoi/Asthetairoi in India. Curtius (8.5.4, 9.3.21) and Diodoros (17.95.4) both record that 25000 suits of armour arrived during the Indian campaign which was disributed to the troops. Which troops? Curtius implies the Macedonians (10.2.23), and it would be strange if Alexander provided the mercenaries with such arms, given the Macedonian's complaints about their old equipment (9.3.10). In any case mercenaries normally had to provide their own equipment; similarly allied soldiers would have already had their own arms. In contrast, the Macedonians, being part of a royal army, needed to have their gear provided by their king. Since there were probably at this stage somewhat just over 4000 "Macedonian" horse (see part 2), providing these men with armour, along with 4000 Hypaspists and 16000 other phalangites would perfectly accord with 25000 suits.9
To conclude this section then, the Macedonian infantry under Alexander were organised initially in units of 500 men, and later into units of 1000. In the case of the Pezetairoi, these units were combined into regiments of 2000 men, and not 1500 as is usually stated. Recreations of Alexander's battles with 1500-men taxies have therefore underestimated the number of Alexander's infantry.
1. Eg. the translator of the Loeb version of Asklepiodotos, Oldfather, says of the number "no one would dream of allowing it to interfere with practical considerations"; one would rather have thought the number is a power of two precisely because of such practical considerations as the requirement to be able to double and halve depths, etc. Return
2. Diodoros, 16.3.2. It is often claimed that this passage refers not to Philip inventing the sarissa, but just the compact formation of the Macedonian style phalanx, since Macedon had no hoplites of its own before Philip's time. This however completely ignores the testimony of Thukidydes who reports how both sides in a civil war in 424 BC fielded hoplites, (4.124), although only one sides were ethnically Macedonians (the Lynkestians), the other (the Lowlanders) side's were of Greek extraction. At soem time between 413 BC and about 400 BC, the Lowlanders may have started using hoplites as well (Thukydides 2.100). Return
3. The meaning of Asthetairoi is unsure. In this article, whenever I talk of just Pezetairoi, it is to be understood I mean Pezetairoi and Asthetairoi. Bosworth's theory is that they were highlanders, and the Pezetairoi lowlanders. This theory appeals to me, since when the Asthetairoi are mentioned it is often in conjunction with an operation such as climbing a tower or scaling heights, which might have been thought more suitable for men accustomed to living in the hillier regions of Macedon. They are more frequently mentioned later in the narrative, when Alexander was viewed with increasing distrust by the older (and therefore mostly lowlander) part of the army. Alexander had bad relations with the nobles of one of the highland cantons, the Lynkestians, who had (after initial opposition) ingratiated themselves with his father Philip, and suffered under Alexander because of it, but he had very good relations with the nobles of the other two, Elimiotis, the most consistently loyal to the central throne, and Orestis, from which the assassin (and the silencers of the assassin) of Philip came. Alexander, half Epirot himself, and thus closer to the highlanders than his father (he went into exile in neighbouring Illyria before his assumption of the throne), could count on the loyalty of these highland regions, which is why they might have been used more and more frequently for the most demanding and prestigious duties. The Asthetairoi taxeis of Koinos, the one used in the intital final assault on Tyre, was the one recruited from Elimiotis.Return
4. In Greek, taxeis, or in the singular, taxis. Return
5. One version of this source does in fact claim it was officers that were impersonated, but while it is possible to have Macedonian generals commanding non-Macedonian armies, the presence of lower-ranking officers would require their troops' presence as well. Return
6. The Hypaspists, while distinct from the rest of the Macedonian phalanx, were none the less also a part of it. Eg. Arrian 1.28.3: "Alexander deployed the phalanx as follows: on the right wing, where he had stationed himself, he had the Hypaspists, and next to them the Pezetairoi..." Return
7. Eight new chiliarchs were appointed at this time. There have been some rather contrived attempts to try and squeeze these 8 chiliarchs into the Hypaspists' command structure but in fact there is no evidence any of them were commanders of Hypaspists even if it is likely that some were, since other troops were organised into chiliarchies too (eg Arrian, 4.24.10, where 2 chilarchies of archers are mentioned). Return
8. Certainty is impossible, but it appears that the men in this phalanx were actually the hypaspists; for at 1.5.10 we are told Alexander took with him the Hypaspists, 400 cavalry and the Agrianians and archers, at 1.6.1 we are told the cavalry deployed in two bodies 200 strong on each wing of the phalanx, and at 1.6.2 it was 120 files deep, but the archers and Agrianians are not mentioned, perhaps being kept to the rear, since at1.6.6 we hear that Alexander has to send for the Agrianians and the archers. He also then orders the Hypaspists to cross a river, followed by the taxeis of the Macedonian phalanx who are only now mentioned for the first time. If this is so, it is further confirmation that the Hypaspists were 2000 strong, and not 3000 strong at this time. Return
9. The few Macedonian archers would not require armour, being light troops. Eg. Arrian 1.2.5, 1.28.5. Admittedly Diodoros says the panoplies were for foot troops, but it is harder to believe Alexander had some 25000 Macedonian foot given Antipater still had a resonable number in Macedonia than 20000 (which would still be a much greater number than most commentators would admit to). However, if the suits went to non-Macedonians as well as Macedonians, then their number is entirely too few; I believe it is more likely that Diodoros has conflated something along the lines of 'mostly for the infantry' into 'infantry'. Curtius, whose account is much longer, and therefore usually more detailed than diodoros', makes no mention of them being specifically for infantry, and it is in any case hard to imagine infantry recieving gilt armour if the more prestigious cavalry had to do without! Return
Return to my homepage index, or go to my Battle of Granicus DBM scenario.