c

Alexander and the army

i The invasion army of 334 B.C.

In the spring of 334 Alexander's army was assembled at Amphipolis and a
smaller expeditionary force was already operating in Asia Minor. The total
number of troops is difficult to estimate because of the diversity of figures
given in the sources, ranging from a maximum of 43,000 foot and 5,500 horse
to a minimum of 30,000 foot and 4,000 horse. 1 Some of the discrepancy may
be explained by some authorities including the advance force in their total and
others omitting it; but the inconsistencies run deep and cannot all be resolved
on that hypothesis. Fortunately the size and composition of the Macedonian
contingents is not seriously in doubt. Alexander took with him 12,000 infantry
and left the same number with Antipater, his regent in Macedonia (Diod.
XVII. 17.4, 5). There was also a body of Macedonians already serving in Asia
(Diod. XVII.7.10), several thousands strong. After Alexander crossed the
Hellespont the total of his Macedonian infantry was around 15,000. The
majority were brigaded in six phalanx divisions (which Arrian usually terms
taxeis) and had the collective title of Foot Companions (pezhetairoi). Three
of these divisions at least were recruited from the old principalities of Upper
Macedonia and are termed asthetairoi, a most mysterious appellation which
has yet to be explained satisfactorily.2 The other taxeis do not apparently bear
any distinctive nomenclature, but they may well have been recruited on a
similar regional basis (cf. Arr. in. 16.11). The other major component of the
Macedonian infantry was the corps of hypaspists. This force had evolved from
the old bodyguard of the Macedonian kings and its nucleus, the agema, still
acted as Alexander's guard when he fought on foot. The rest of the hypaspists
were organised in chiliarchies (units of 1,000), perhaps three in number. They
were an elite, selected for their skill and physique and equally expert at
phalanx work in pitched battle and in rapid skirmishing with light infantry and
1
Berve 1926, 1.177-8; Brunt 1963, 32-6; Bosworth 1980a, 98—9.
2
Bosworth 1973, 1980a, 251-3; Milns 1976, 97-101; Hammond and Griffith 1979, 2.709—13.
On the possibility of a cavalry unit named asthippoi see Hammond 1978a, Milns 1981.

259

Cambridge Books
https://doi.org/10.1017/CBO9780511518539.007 Onlineonline
Published © Cambridge University
by Cambridge Press,
University Press 2010
 260 Alexander and the army

cavalry support. In the major battles they stood alongside the taxeis in the
phalanx, and there is no doubt that their armament was the same. 3 It was their
general calibre that was superior.
The equipment of the Macedonian infantryman was a blend of hoplite and
peltast armament. The main offensive weapon was the satisa, a huge pike up
to six metres long with a leaf-shaped blade and a butt spike, both about 50 cm
in length.4 Its total weight would have been nearly 7 kg. As a result the
weapon could only be manipulated with two hands, which allowed only a
small button-shaped shield, slung around the neck to protect the left
shoulder.5 There were subsidiary weapons, a more orthodox short spear and a
slashing sword, but they were of very secondary importance in pitched battle.
The sarisa was primary and not intended to be used in isolation: the story of
the single combat between Corrhagus and Dioxippus is a vivid illustration of
how vulnerable the lone infantryman could be when isolated from the rest of
the phalanx.6 The defensive armour was relatively sparse. 7 There are no
contemporary descriptions of it, but the Roman emperor Caracalla, who
carried Alexander-imitation to absurd lengths, created an allegedly authentic
Macedonian phalanx, armed with the three offensive weapons (sarisa, spear
and sword) together with a helmet of ox-hide, a triply strengthened linen
corselet and tall boots (Dio LXXvin.7.1-2). Greaves are also attested (Polyae-
nus iv.2.10) and are not improbable. The light body armour was logical
enough. It was the line of sarisae that gave the primary protection, making the
phalanx infantry practically invulnerable except to missile attack, and it was
superfluous to equip them with heavy body armour. As it was, if they
temporarily dispensed with the sarisae, they had a mobility which matched
that of the light infantry.
The phalanx was organised in basic units of sixteen (originally ten, as their
technical name dekas shows) which were combined in larger groupings called
lochoi. In its primary formation it seems that the phalanx was sixteen deep, the
dekades deployed side by side with the more expert and highly paid men to the
front. In an actual engagement only the first three or four ranks used their
sarisae in couched position; the rest kept their weapons vertical and used their
body weight to increase the momentum of the front line.8 There was a good
deal of variation possible. The dekades could be doubled to deepen the
phalanx to thirty-two ranks or halved to reduce it to eight; and the change of
front could be gradual, as happened before Issus, when the Macedonian line,
originally thirty-two deep, was gradually expanded as the plain opened out. As
gaps appeared in the front lines, files were continuously transferred from the
3
Tarn 1948, 2.153-4; Hamilton 1955, 218-19; Milns 1971, 186-8; Ellis 1975; Hammond and
Griffith 1979, 2.414-18.
4
Andronikos 1970; Markle 1977, 323-6; 1980.
5
Asclep. Tact.. 5; cf. Markle 1982, 92-4.
6
Diod. XVII. 100.2-8; Curt. ix.7.16-22; Ael. VIIx.22.
7
Moretti, ISE no. 114, B I ; cf. Griffith 1956-7; Markle 1977, 327—8.
8
Polyb. XVIII.29.2-30.4; Ael. Tact. 14.6; Arr. Tact. 12.11.

Cambridge Books
https://doi.org/10.1017/CBO9780511518539.007 Onlineonline
Published © Cambridge University
by Cambridge Press,
University Press 2010
 The invasion army of 334 261

rear until the phalanx depth was finally contracted to eight. 9 The most
impressive performance recorded was the display Alexander organised for the
Illyrians in 335. Then he massed the phalanx 120 deep and carried out a
number of changes of front, turning the direction of march and the thrust of
the sarisae to left and to right. Finally he drew back the entire front line to
create a wedge-shaped apex on the left (Arr. 1.6.1-3). These manoeuvres were
carried out in silence, and the parade ground discipline was obviously
immaculate. The training was geared to produce a flexible, unbroken mass of
infantry. In the period after Alexander the integrity of the phalanx became a
fetish: breaks in the line were fatal and commanders could not conceive taking
their men over broken ground or across watercourses. 10 Alexander's men were
more versatile. They were led in line over highly difficult terrain at Issus and
fought the engagement across a river; and their line was broken both at Issus
and at Gaugamela without catastrophe resulting. 11 The middle-rank men
could obviously deal to some degree with breaches at the front, in a way that
was impossible later, when the sarisa was eight metres or more long. Not
every action will have required the full weaponry. It is most unlikely, for
instance, that the infantry involved in the final pursuit of Darius carried the
heavy sarisa (Arr. in.21.2-7); they presumably marched with spears alone.
But it seems clear that the sarisa was their basic weapon. It was used by the
Macedonian guards at court, in situations where it was cumbersome and
inappropriate (Arr. iv.8.8-9); and we are explicitly informed that when the
Macedonian infantry crossed the Danube in 335 they took their sarisae with
them (Arr. 1.4.1). The individual phalangite, then, was essentially a part of a
corporate mass, intensively trained to a very specialised form of fighting and
enjoying a cohesiveness and weight of offensive armament that was
unmatched in the contemporary world.
The counterpart of the phalanx infantry was the Macedonian cavalry,
collectively termed the Companions (hetairoi). At the Hellespont they were
1,800 strong and divided into eight squadrons (ilai), one of which, the He
basilike, defended the king when he fought on horseback.12 This Royal
Squadron formed an elite, probably comprising most of the courtiers, those of
the Companions proper who had no specific command. Otherwise they were
recruited on a regional basis. Those squadrons whose origins are recorded
came from the Thraceward districts where Philip had established military
settlers: Bottiaea, Amphipolis, Apollonia, Anthemus. 13 The one exception is
a mysterious 'Leugaean He1 (Arr. 11.9.3), which may be a much older unit
created before the institution of territorial recruitment. The Upper Mace-
donian cavalry (Arr. 1.2.5) were another regional group, but there is no direct
evidence that they were taken on campaign in Asia. They may have been part
9
Polyb. XII.19.5-6 = Callisthenes, FGrH 124 F 35; Arr. 11.8.2; Curt. in.9.12.
10
Polyb. xn.22.4-6; cf. Markle 1978, 493-5 (somewhat exaggerated).
11
Arr. 11.10.5; in.14.4-5.
12
Arr. in. 11.8; Diod. xvn.57.1; cf. Hammond and Griffith 1979, 2.411-14.
13
Arr. 1.2.5, I2-7» n -9-3I Qf- Berve 1926, 1.105.

Cambridge Books
https://doi.org/10.1017/CBO9780511518539.007 Onlineonline
Published © Cambridge University
by Cambridge Press,
University Press 2010
 262 Alexander and the army

of the 1,500 cavalry left with the home army under Antipater (Diod.
XVII. 17.5), but it is much more likely that Alexander took troops from all
sectors of his kingdom into Asia, leaving a proportion from each recruiting
area. At first there is no record of any subdivision of the ilai; each squadron
apparently fought as a unit under its local commander. Their armament was
simple, a thrusting lance of cornel wood with a reserve supply of javelins and
perhaps a cavalry shield, together with a minimum of body armour, including
the Macedonian helmet, the broad-brimmed kausia. In pitched battles the
cavalry was Alexander's main striking force. Time and time again it was a
cavalry charge, usually in wedge formation,14 that exploited the vital breach in
the enemy line. Unfortunately the numbers in each squadron are very much a
matter of guesswork. The royal guard may have totalled 300 by the end of the
reign (see below, p. 269), but there is no indication of its strength at the time
of the Hellespont crossing. More seriously, we do not know whether Dio-
dorus' figure of 1,800 covers the Companions alone or includes other units of
Macedonian cavalry.
The matter is complicated by the problem of the prodromoi. These troops
were a division of cavalry regularly associated with the Companions and the
Paeonian light horse; and, as their name implies, they were used on
reconnaissance missions. But they are also termed sarisa-bezrers (sarisopho-
roi)]15 and it is clear that they operated in the vanguard of the assault at the
Granicus armed with the cavalry sarisa > which by all indications was equal in
length to the infantry weapon. 16 They were divided into ilai like the
Companions, and there were at least four of them (Arr. 1.12.7). The use of
sarisae in battle, together with the fact that they are habitually mentioned by
Arrian without any ethnic qualification, suggests that they were native
Macedonians, organised separately from the Companions proper but to be
included in Diodorus' total of 1,800 Macedonian cavalry.17 They seem to have
fulfilled a dual function, advance scouting (clearly without the sarisa) and
anti-cavalry fighting in open order. The sarisa, which projected murderously
before and behind the horse, could not be used in close formation without
mortal danger to one's own troops. Cavalry using it would need to be widely
spaced or to be massed in single extended lines, in which case they would
provide an effective counter to frontal assaults by more lightly armed
opponents (cf. Arr. iv.4.6). Although there is evidence, notably in the
Alexander Mosaic, that the Companions might use the sarisa on occasion, it is
14
Ael. Tact. 18.6; Arr. Tact. 16.6-7; cf. Rahe 1981; against Markle 1977, 338-9.
15
The equation is explicit: Arr. 1.14.1 and 6; m.12.3 with Curt, iv.5.13.
16
Markle 1977, 333-9; 1982, 105-6.
17
Diod. XVII. 17.4 has a note ©QQixec; 6e Jio66QO|i,oi xal nouoveg evvaxooioi, which has been
taken to imply that the prodromoi were Thracians (Tarn 1948, 2.157) or emended to dissociate
the Thracians from the prodromoi (cf. Milns 1966a). Neither is necessary; prodromoi could be
used as a generic term for light cavalry and applied to Thracians and Paeonians (Arr. HI.8.1) as
well as the Macedonian sarisophoroi. But, if the sarisophoroi were Macedonian, as seems almost
certain (Berve 1926, 1.129; Brunt 1963, 27-8), they should be listed by Diodorus with the
main body of Macedonian cavalry, not separated and associated with the cavalry of the north.

Cambridge Books
https://doi.org/10.1017/CBO9780511518539.007 Onlineonline
Published © Cambridge University
by Cambridge Press,
University Press 2010
 The invasion army of 334 263

evident that the weapon could not be used profitably in the wedge-shaped
assault formation. The principal weapon of the Companions must have been
the shorter thrusting lance, and their training was designed for intensive
attacks in close formation and dense column, contrasting with the prodromoi
who were intended to operate in more open conditions. Both were Mace-
donian in extraction, but we have no means of assessing their relative
numbers, and the total of ilai in the prodromoi remains unknown.
The indispensable complement of the Macedonian forces was the light
infantry. Some of these units may indeed have come from Macedon proper,
but there is little explicit evidence. Arrian occasionally mentions divisions
(taxeis) of light-armed troops, but he rarely gives an indication of their
national origins and never designates any of them as Macedonian. On the
other hand he does appear to include Thracians and Agrianians among the
taxeis of javelin-men.18 The only known group of light troops which may have
consisted of Macedonians is the contingent of javelin-men commanded by
Balacrus; it is listed without ethnic in Arrian's description of the line at
Gaugamela, which specifies the national origins of the other units. 19 If there
were Macedonian light-armed they were clearly not numerous. As we have
seen, the phalanx soldiers had a relatively light defensive equipment, and few
Macedonians can have been excluded from service in it on grounds of poverty.
Alexander would have been well advised to concentrate his native infantry in
the phalanx, relying on his neighbours in the north to supply the light
infantry. By and large that was what happened. The most important of
Alexander's light-armed troops were the Agrianian mountaineers, a relatively
small body of javelin-men from the upper Strymon. They are attested some
fifty times in Arrian alone, used on almost every occasion which called for
rapid movement on difficult terrain. From the time of the Danubian campaign
they were employed with the hypaspists and selected phalanx infantry for
particularly arduous marches, and in the formal battles they formed part of a
defensive screen in advance of the main line. Their usual associates were the
archers. Once again there may well have been a corps of Macedonian archers
(Arr. in. 12.2) but its numbers were small. Otherwise the archers largely
comprised Cretans, and two attested commanders of the contingent were
Cretans by birth. 20 The archers were evidently a specialist body, recruited
outside Macedon, but with the Agrianians they were used alongside the
Macedonian troops whenever skirmishing tactics were called for. The Thra-
cians were occasionally used in the same role (Arr. 1.28.4), but the archers and
Agrianians are far more frequently attested, and Diodorus (xvn.17.4) lists
them as a composite body, 1,000 strong at the Hellespont. That was a
minimum figure. Alexander had deployed twice that number in the north

18
Cf. A r r . 1.27.8 with 1.28.4; v i . 8 . 7 .
19
A r r . i n . 1 2 . 3 , 13.5; cf. i v . 4 . 6 ; Berve 1926, 1.131; H a m m o n d and Griffith 1979, 2.430.
20
Arr. 11.9.3; Diod. xvn.57.4; Curt. iv. 13.31. The commanders are Eurybotas (Arr. 1.8.4) a n ( ^
Ombrion (in.5.6). For the Macedonian archers see Arr. in. 12.2 with Bosworth 1980a, 302.

Cambridge Books
https://doi.org/10.1017/CBO9780511518539.007 Onlineonline
Published © Cambridge University
by Cambridge Press,
University Press 2010
 264 Alexander and the army

(Arr. 1.6.6), and once the campaign in Asia began they were swelled by
reinforcements. By the middle years of the reign there were at least 1,000
Agrianians,21 and the archers were brigaded in chiliarchies (Arr. iv.24.10).
They had proved themselves key units and were systematically expanded.
T. ALIAD
The rest of the army consisted of allied troops and mercenaries. Of these the
most important contingent by far was the Thessalian cavalry, probably equal
in numbers to the Macedonian cavalry and almost equal in calibre. Like the
Companions they were divided into ilai, of which the Pharsalian contingent
was the most prestigious and numerous (Arr. 111.11.10), and they performed
much the same functions as the Companions themselves, protecting the left
wing of the phalanx in the first three major battles. The structure of command
seems to have been parallel to that of the Macedonian cavalry, with regionally
based ilai, but at the head was a Macedonian commander. The rest of the
allied cavalry, predominantly from central Greece and the Peloponnese, was
much less important and effective, fewer in number and less prominent in
action. Like the Thessalians they were divided into ilai (Tod, GHI no. 197.3)
under the command of a Macedonian officer. The infantry from the allied
Greek states is more problematic. They formed a contingent numerically
strong, 7,000 of them crossing the Hellespont in 334, and they were
predominantly heavy-armed hoplites. But once in Asia they are mainly
notable for their absence. There is no explicit record of them in any of the
major battles. At Gaugamela we may infer that they provided most of the men
for the reserve phalanx (Arr. 111.12.1), but in the other engagements there is
no room for them. They are only mentioned as participants in subsidiary
campaigns, usually under Parmenion's command (in the Troad, at the
Amanid Gates, in Phrygia and in the march on Persis), 22 and they never
appear in the entourage of Alexander. One contingent, that of Argos, was
detached for garrison duty in Sardes (Arr. 1.17.8) but that is the only case
recorded (the allied cavalry, however, formed the original garrison of Upper
Syria). Part of the reason for the neglect must have been the heterogeneous
nature of the allied infantry, drawn as it was from a plethora of different cities
and virtually impossible for its Macedonian commander to organise as a single
unit. There was also the question of loyalty. Alexander might well have been
reluctant to rely on men recently vanquished at Chaeronea to face the Hellenic
mercenaries in Persian service. It was too much kin against kin, and his Greek
allies naturally had less stomach for the task than his native Macedonians.
The other major infantry group was the 7,000-strong contingent of
Thracians, Triballians and Illyrians. These troops are, if anything, more
elusive than the Hellenic infantry. The Triballians are never mentioned in the
campaign narrative and the Illyrians only rate a passing reference in Curtius'
account of Gaugamela (iv. 13.31), where they are associated with the merce-
nary infantry. The Thracians are a little more prominent. Under the Odrysian
21
A r r . i v . 2 5 . 6 ; cf. C u r t , v . 3 . 6 .
22
A r r . 1.17.8; 1.24.3, H ' S - 1 ; i n . 1 8 . 1 .

Cambridge Books
https://doi.org/10.1017/CBO9780511518539.007 Onlineonline
Published © Cambridge University
by Cambridge Press,
University Press 2010
 The invasion army of 334 265

prince Sitalces they were active before and during the battle of Issus, and at
Sagalassus and Gaugamela they performed exactly the same function on the
left of the line as the Agrianians on the right (Arr. 1.28.4, HI. 12.4). It may be
only chance that we hear nothing more of them in action, but their subsequent
history suggests that Alexander did not find them in any way indispensable. A
large proportion was left behind in 330 to man the satrapal armies of Media
and Parthyaea (Arr. in. 19.7; v.20.7). A few returned to the main army in
326/5, but the entire contingent of Thracians was soon discarded, left in the
unenviable role of garrison army in northern India (Arr. vi.15.2). The
evidence, such as it is, suggests that they were not normally employed as
front-line troops but used on secondary missions or in positions where weight
of numbers rather than expertise was important. It was they who formed the
occupying force on the island of Lade with 4,000 other non*Macedonians,
denying access to the Persian fleet (Arr. 1.18.5), a n c ^ t n e v were also consigned
to road-building in Pamphylia (Arr. 1.26.1). It looks as though Alexander had
no interest in repatriating any of them, and the main raison d'etre of his
Thracian contingent may have been simply to be out of Thrace. Their absence
meant that the territory was easier to control. The same considerations applied
even more forcefully to the Illyrians and Triballians.
The Thracian cavalry comes in the same category. These troops were
placed alongside the Greek allied cavalry at the Granicus and Gaugamela, but
their employment was very sporadic and they were assigned to the Median
garrison together with the infantry. A further contingent of Thracian cavalry
which reached India late in 326 (Curt, ix.3.21) was certainly left with the
satrapal army of northern India. The other cavalry body from the north, the
Paeonians, had a more distinguished career. They are associated with the
prodromoi and were in the vanguard of fighting at the Granicus and Gau-
gamela; and they were lightly enough equipped to be termed scouts in their
own right (Arr. in.8.1). But they are not mentioned in any narrative after 331,
and there is no record of their being detached to any garrison force. Their
numbers must have been small (with the Thracian cavalry they only
amounted to 900), and they could easily have been amalgamated with other
units (see below, p. 271).
Finally Alexander, like his father, made extensive use of mercenaries. Only MERCE
5,000 are recorded at the Hellespont, but there were certainly many thousands
already active in the expeditionary force in Asia Minor, including the whole of
the mercenary cavalry.23 One distinct group, the 'old mercenaries', were
grouped together under the command of the Macedonian Cleander. 24 Other-
wise they seem to have been divided up as the occasion demanded. The
mercenary cavalry at Gaugamela were divided into two groups, under
23
T h e total force in Asia n u m b e r e d over 10,000 (Polyaenus v.44.4), a proportion of which was
Macedonian (Diod. x v n . 7 . 1 0 ) . T h e majority were doubtless mercenaries, as was clearly the
case with t h e cavalry ( D i o d o r u s lists n o mercenary cavalry with t h e army at t h e H e l l e s p o n t ) .
24
A r r . i n . 12.2; cf. Griffith 1935, 17, 2 9 - 3 0 .

Cambridge Books
https://doi.org/10.1017/CBO9780511518539.007 Onlineonline
Published © Cambridge University
by Cambridge Press,
University Press 2010
 266 Alexander and the army

Menidas and Andromachus, again both Macedonians, and the division of

command persisted for some time (Arr. in.25.4). That is the only organi-
sation of any permanence that can be traced. In general the mercenaries
seem to have been organised on a far more fluid basis than the rest of the
army. Their numbers constantly fluctuated, as new units were recruited and
mercenaries already in service were detached to man the satrapal armies or,
in the latter years of the reign, to provide settlers for the new foundations in
the east. They were probably the most dispensable part of the army, in
terms of front-line fighting. In the major battles they were kept in reserve,
except for the cavalry which played an important part at Gaugamela; and,
like the infantry of the Corinthian League, they tended to fall under the
command of Parmenion, to be used in secondary expeditions. In later years
the forces sent out to deal with Satibarzanes in Areia and Spitamenes in
Sogdiana were almost exclusively composed of mercenaries (cf. Curt.
VII.3.2; Arr. iv.3.7). Indeed outside the immediate entourage of Alexander
the fighting troops throughout the Empire tended to be mercenaries. At
first the need to pay them regularly and the fact that the Great King was a
competing paymaster will have kept numbers down. Apart from the 300
RECLU
mercenaries taken in from the garrison at Miletus (Arr. 1.19.6) there are
only attested 4,000 raised in the Peloponnese by Cleander and 400 cavalry
sent to Memphis by Antipater. This is admittedly a partial record, but
it contrasts vividly with the situation after 331, when Alexander's funds
were practically limitless. Then there was a vast and continuous influx;
nearly 60,000 recruits are listed in the extant sources and many more may
have been unrecorded.
TRAS DARÍO AUMENTAN
11 Evolution and reorganisation: 333-323 B.C.
It was the Macedonian nucleus, perhaps 15,000 infantry out of an original
force numbering more than 40,000, which was overwhelmingly important,
the main striking force in the army and the sheet anchor for all the major
battles. On the cavalry side the Macedonians were less dominant, at least
balanced by the Thessalians, but they were still far more frequently used by
Alexander than any other cavalry units. In numerical terms his army was
huge, surpassing the total number of combatants at Nemea in 394, but the
vital Macedonian complement was a relatively small part of it. In practice all
Alexander's victories, except perhaps Gaugamela, were won with a fraction of
the forces at his disposal. Accordingly the first years of the campaign saw a
planned increase in the size of the Macedonian core of the army, as Alexander
prepared to meet the full levy of the Persian empire. Our sources record an
impressive number of reinforcements from Macedon between 333 and 330. At
Gordium 3,000 Macedonian infantry and 500 cavalry, Macedonian and
Thessalian, arrived to swell his forces (Arr. 1.29.4) '•> anc ^ J u s t before he entered
Cilicia later that summer an additional 5,000 infantry and 800 cavalry arrived

Cambridge Books
https://doi.org/10.1017/CBO9780511518539.007 Onlineonline
Published © Cambridge University
by Cambridge Press,
University Press 2010
 Evolution and reorganisation 267

from Macedonia.25 This influx expanded the original phalanx by half, and it
may not represent the entire total. Curtius indicates that other forces had
arrived or were on their way by the time Issus was fought (in. 1.24; 7.8). By
the end of 333 the Macedonian component of the army reached its highest
strength hitherto. Numbers will have fallen the following year as the sieges of
Tyre and Gaza took their toll, and at the end of 322 one of the phalanx
commanders, Amyntas son of Andromenes, was sent on an urgent recruiting
mission to Macedon across the wintry seas of the Mediterranean. 26 Before the
summer of 331 he had amassed a force of 15,000, an army in itself, including 1 AÑADE
6,000 Macedonian infantry and 500 cavalry.27 That is the last that is heard of
specifically Macedonian reinforcements. Though there is ample record of new
arrivals later, no contingent includes native Macedonians; the troops which
came from Antipater were Thracians, Illyrians or mercenaries. Alexander
himself sent for replacements from Macedon in 327 (Arr. rv.18.3), but there
was apparently no response, and in 324, when he demobilised the 10,000
veterans from Opis, he insisted that Antipater bring prime recruits from
Macedon to take their place. 28 But Antipater never left Macedon and the
reinforcements never arrived. The country was already drained of fighting
men by the levies already sent (Diod. xvin.12.2), and the home army could
not be weakened further. As it was, Antipater had trouble enough raising an
army in 331/0 and still more at the outbreak of the Lamian War, when he was
very seriously embarrassed. In effect the Macedonians who were in the army
in 330 had no reinforcement until the end of the reign. Even so their numbers
are impressive. At Opis in the summer of 324 10,000 Macedonians were
demobilised and a strong contingent remained, a minimum of 8,000 at the
time of Alexander's death. 29
The crucial period in terms of reinforcements was 333-331 B.C. Even on the
basis of the defective reports that have come down to us it is clear that both the
Macedonian infantry and cavalry were doubled, and the increase was in all
probability much greater. It is unlikely that the mortality rate was less than
50%, given the constant fighting and the rigour of the physical conditions
encountered. In that case the number of Macedonian troops taken from the
homeland totalled well over 30,000. The actual operational numbers reached a
peak late in 333 and then again at the end of 331, and they declined throughout
the rest of the reign. The effect of the additional numbers is hard to trace.
They did not result in any major change in organisation. As far as we can tell,
the incoming reinforcements were divided up among the existing units
according to their regional origins. The phalanx battalions accordingly
remained six in number between the Granicus and Gaugamela, even though
the complement of each battalion was greatly increased. Even the massive
25
Polyb. X I I . 19.2; cf. Brunt 1963, 3 7 ; Bosworth 1975, 4 2 - 3 ; Milns 1976, 106.
26
D i o d . X V I I . 4 9 . 1 ; C u r t , iv.6.30 (cf. v n . 1.37-40).
27
D i o d . X V I I . 6 5 . 1 ; C u r t . v. 1.40-2; cf. A r r . i n . 16.10.
28
Arr. vn.12.4; Justin xn.12.9.
29
Cf. B r u n t 1963, 19. F o r other estimates see S c h a c h e r m e y r 1973, 4 9 1 ; M i l n s 1976, 112.

Cambridge Books
https://doi.org/10.1017/CBO9780511518539.007 Onlineonline
Published © Cambridge University
by Cambridge Press,
University Press 2010
 268 Alexander and the army

reinforcements brought by Amyntas were simply included in the existing

units according to nationality. It is only at the time of the invasion of India that
there is evidence for a seventh battalion of the phalanx, and the evidence for
that is at best circumstantial: seven commanders appear to be named
simultaneously at the head of phalanx taxeis.30 The reasons for such a change
are totally opaque. One of the battalions may have become disproportionately
large through reinforcements and irregular losses and so became divided into
two commands. That is, however, speculation. We do not know the reason,
and there is no indication when the change took place or for how long it lasted.
There was some reorganisation at the end of 331 when Amyntas' Mace- CAMBIOS
donian reinforcements arrived. Alexander divided the existing cavalry ilai into
two lochoi and appointed new sub-commanders, promoted by merit and not
by regional affiliation.31 It was a break with tradition, and the king must have
had strong reasons for his action. In all probability it was a first step towards
breaking the ties between the original commanders and their men. There was
now an intermediate level of command in which the officers had no necessary
connection with their men but owed their appointment solely to royal favour.
A similar development may well have taken place in the organisation of the
hypaspists. The details are sadly garbled by Curtius (v.2.2-5), 32 but it seems
that another command was introduced in the hierarchy, the new officers again
achieving preferment solely through military merit. From tfyis time onwards
the hypaspist corps had both chiliarchies and pentakosiarchies, units of 1,000
and 500, and there was a new category of officers of relatively humble status.
After 330 the entire Macedonian cavalry appears to have been reorganised.
The basic unit was now not the He but a new formation named 'hipparchy'.
These new units are first recorded in the Ptolemaic part of Arrian's narrative
during the spring of 329,33 and from then onwards 'hipparchy' is the almost
invariable nomenclature of the cavalry units. When we hear of ilai, they
appear as subdivisions, each hipparchy comprising a minimum of two ilai
(Arr. vi.21.3-4, 27.6). The Royal He also disappears as a title and is replaced
by the term agema; Alexander's cavalry and infantry guards now had the same
nomenclature. Unfortunately there is no list of hipparchies and commanders
comparable to the list of ilai at Gaugamela, and we can only infer their
numbers from random hints in the campaign narrative. It looks as though
there were eight hipparchies in addition to the agema throughout the Indian
campaign,34 but whether there was a constant number from the beginning is
an insoluble problem. Ptolemy at least had three hipparchies of Companions
30
Arr. iv.22.7 (Gorgias, Cleitus, Meleager); iv.24.1 (Coenus, Attalus); iv.25.6 (Coenus,
Polyperchon); iv.27.1 (Alcetas). Cf. Milns 1966; Bosworth 1973, 247, 249.
31
Arr. 1n.16.11; cf. Diod. x v n . 6 5 . 3 .
32
A r r . vii.25.6; Plut. Al. 76.6. Cf. Bosworth 1980a, 148-9; Milns 1971, 189-92.
33
A r r . i n . 2 9 . 7 . T h e r e is an earlier anachronistic reference at A r r . 1.24.3. Diodorus x v n . 5 7 . 1
describes the ilai at Gaugamela as hipparchies - a clear instance of substitution of the later
term.
34
Brunt 1963, 2 9 ; Bosworth 1980a, 3 7 5 - 6 .

Cambridge Books
https://doi.org/10.1017/CBO9780511518539.007 Onlineonline
Published © Cambridge University
by Cambridge Press,
University Press 2010
 Evolution and reorganisation 269

for his pursuit of Bessus, and that was clearly only a fraction of the total force.
It is not unlikely that there were eight hipparchies as early as 329.
The other main change at this period is the apparent disappearance of the
prodromoi. The unit is not attested in action under that title after the death of
Darius. There is one fleeting reference to sarisophoroi in 329, when they were
still apparently grouped in ilai (Arr. iv.4.6). They may not as yet have been
divided into hipparchies, but the mercenary cavalry had been so organised and
a fortiori one would expect the Macedonian troops to have undergone the same
transformation. It may be that He is already being used in its later technical
sense and the sarisophoroi in 329 were mobilised in sub-groups against their
Saca enemies. At all events it is their last appearance, and it is a reasonable
assumption that they were amalgamated with the Companion cavalry and
organised with the hipparchies. Each hipparchy for instance may have had a
subdivision of sarisophoroi. They would have added to its versatility when
away from base, and the sub-units could be detached for special service during
an emergency such as the crossing of the Iaxartes.
The relative strength of the hipparchies is unknown. There is some slight
evidence from the period of the Successors that the number in the agema was
300.35 As regards the rest of the Companions there is only Arrian's figure of
1,700 for the troops embarked at the start of the Indus voyage (Arr. vi.14.4),
but there is no indication that they were the whole contingent. Perdiccas at
least was away on other business (Arr. vi.15.1) and he may not have been the
only absentee. 36 All we have is a minimum figure and it is surprisingly high,
almost as large as the entire body of Macedonian cavalry at the Hellespont.
Reinforcements had been numerous until 331/0, but the attrition rate will
have been high. If the Companions numbered a minimum of 1,700 in 326,
there is every reason to believe that they had absorbed the prodromoi.
The causes of the reorganisation are not recorded, but they were pre-
sumably important. Technical terms, especially traditional ones, are not
changed simply for novelty's sake. The execution of Philotas was certainly
related in some way to the move. After his death Alexander refused to appoint
a new commander for the entire cavalry body but divided the Companions
between Cleitus, the veteran commander of the Royal He, and Hephaestion,
his closest personal friend.37 The division was made for reasons of security
(Arr. in.27.4), and it may have been part of a more general reorganisation of
the cavalry, to lessen the ties of personal loyalty. Almost a year before, he had
introduced sub-units (lochoi) with commanders chosen by merit. There is
35
Diod. x i x . 2 8 . 3 , 29.5 {agemata of E u m e n e s and Antigonus in 317).
36
According to Arrian (Ind. 19.2) t h e hypaspists, archers a n d C o m p a n i o n cavalry a m o u n t e d to
8,000 in 325, b u t it is not clear h o w t h e groups are to be divided. Elsewhere (vi.2.2) he states
that only the agema of cavalry was embarked on that occasion.
37
Arr. i n . 2 7 . 4 . Arrian calls these c o m m a n d e r s IJIJICIQX 011 . b u t that does not imply that t h e r e w e r e
only two hipparchies. H e uses t h e t e r m iJtJiciQXTls in a very fluid way a n d can apply it t o m i n o r
officers (cf. V I I . I 1.6) as well as to commanders-in-chief (1.25.2); it is not a technical t e r m in t h e
same sense as h

Cambridge Books
https://doi.org/10.1017/CBO9780511518539.007 Onlineonline
Published © Cambridge University
by Cambridge Press,
University Press 2010
 270 Alexander and the army

other slight evidence for experimentation: his turning force at the Persian
Gates included a tetrarchia of cavalry (Arr. m.18.5), a unit recorded neither
before nor later. Alexander may have been consciously changing the balance
of the cavalry, striking at the regional ties of the ilai and aiming at a more
homogeneous force. There may also have been logistical considerations. The
ilai may have become unbalanced in size, owing to disproportionate losses in
battle and random accretions through reinforcements, which are not likely to
have been divided with mathematical equality between the recruiting areas.
Some levelling off may have been necessary. At all events the old ilai went, as
did their commanders: only Demetrius son of Althaemenes is known to have
continued with a hipparchy. 38 In their place were the new composite units,
whose commanders were the elite of Alexander's court.
No such reorganisation is attested for the phalanx infantry. The division
between hypaspists and phalanx battalions persisted until the end of the reign,
and, except for the addition of a seventh battalion, there is no evidence for any
major change. There is one minor innovation of nomenclature. Just before he
entered India Alexander is said to have introduced silver shields in his army
and coined the title argyraspides (Justin xn.7.5). This new term was reserved
for the hypaspists and was apparently in vogue by the end of the reign; 39 but it
only came into its own after his death. Then Alexander's hypaspists main-
tained their corporate identity and insisted on the title of argyraspides, which
distinguished them from the various corps of hypaspists formed by the
Successors.40 In Alexander's lifetime, when there was no competition,
'hypaspist' seems to have been the term generally used. The unit maintained
its elite status throughout the reign, and it was presumably supplemented
from the rest of the phalanx to keep its complement relatively stable. After
Alexander's death it was still 3,000 strong, its members all battle-hardened
veterans of his campaigns. In 317 B.C. every man of them is said to have been
over sixty, a statement which is no doubt exaggerated but which derived from
the contemporary and eyewitness, Hieronymus of Cardia.41 The hypaspists
contained the most expert of the infantry throughout the reign and there must
have been a constant transfer of picked men from the phalanx battalions,
which suffered as a result. After the demobilisation of the 10,000 veterans at
Opis the nucleus of phalanx troops, including hypaspists, fell below 10,000
and by the end of the reign Alexander was forced to supplement it with Iranian
infantry (Arr. vii.23.3-4). Only four men in each file of sixteen were
Macedonians, stationed to the front and the rear; the mass of the formation
was now Persians, armed with bows and missile javelins. The mixed phalanx
that emerged was designed for frontal attack only and had none of the
flexibility that characterised the infantry manoeuvres at the beginning of the
38
A r r . H I . 11.8; iv. 1 6 . 3 : cf. Berve 1926, 2 n o . 256.
39
A r r . VII. 11.3: it appears anachronistically in t h e vulgate accounts of G a u g a m e l a ( D i o d .
X V I I . 5 7 . 2 ; C u r t . iv.13.27).
40
Anson 1981; contra Lock 1977b.
41
D i o d . x i x . 4 1 . 2 (cf. 3 0 . 6 ) ; Plut. Eum. 1 6 . 7 - 8 ; cf. H o r n b l o w e r 1981, 1 9 0 - 3 .

Cambridge Books
https://doi.org/10.1017/CBO9780511518539.007 Onlineonline
Published © Cambridge University
by Cambridge Press,
University Press 2010
 The use of oriental troops 271

reign. It was an improvisation to make the most of a dearth of sarisa-trained

phalangites and a superabundance of untrained Iranians. This relatively small
number of Macedonian infantry in 323 goes a long way towards explaining
their rapid capitulation to the cavalry in the conflict after Alexander's death.
The vast differential between the two bodies at the beginning of the reign had
been greatly reduced.

in The use of oriental troops

The turning-point in the evolution of Alexander's army appears to have been
the year 330. Until then the Macedonian component was progressively
reinforced, reaching peaks before Issus and after the arrival of Amyntas' great
contingent late in 331. Alexander then thought it safe to divest himself of
non-Macedonian troops. The forces from the Corinthian League, infantry
and cavalry, were demobilised from Ecbatana in the spring of 330; 42 even the
Thessalian cavalry who re-enlisted were dismissed at the Oxus less than a year
later (Arr. 111.29.5). Alexander now relied on the Macedonian nucleus for
front-line work and the mercenaries for support functions. The latter were
important and were recruited in increasing numbers (see above, p. 266); but
they were probably reallocated almost as soon as they arrived. The sources,
partial record though they are, list over 36,000 left in satrapal armies or in the
new foundations, and there are many settlements attested where the number
of mercenaries is not given.43 The wastage was constant, and Alexander
cannot have had a large number with his field army at any given time. With the
Hellenic complement of his army reduced and few, if any, reinforcements
arriving from Macedonia, the king plunged into three years of guerrilla
fighting, first in the Iranian plateau and then in Bactria/Sogdiana. It is not
surprising that he was forced to draw upon oriental troops in increasing
numbers. The date at which the development began is hard to fix. From late
330 Alexander employed a specialist troop of mounted javelin-men, who are
usually assumed to have been Persian in origin. That is possible but nowhere
attested; and the new corps was immediately drawn upon to provide a garrison
force for Areia (Arr. in.25.2), a role in which no oriental unit is attested, even
in the Indian satrapies. It is equally possible that the new unit was composed
of European cavalry.44 The Paeonians, for instance, are not mentioned after
Gaugamela, and as light cavalry they could well have formed the nucleus of
the specialised unit of javelin-men, which is associated in action with the
Agrianians, themselves troops from Paeonia (cf. Arr. iv.26.4, vi.17.4).
The first unambiguous use of oriental troops is in the winter of 328/7. At the
end of the campaign in Sogdiana locally levied cavalry were serving alongside
the Macedonian forces (Arr. iv.17.3). Alexander took more of them with him
42
A r r . i n . 19.6—7; Plut. Al. 4 2 . 5 ; D i o d . x v n . 7 4 . 3 - 4 ; C u r t , v i . 2 . 1 7 ; cf. T o d , GHI n o . 197.
43
F o r the detailed figures see Berve 1926, 1.146-9; Griffith 1935, 2 0 - 3 . Cf. Wirth 1984.
44
Bosworth 1980b, 1 4 - 1 5 ; contra Berve 1926, 1.151; Brunt 1963, 4 2 .

Cambridge Books
https://doi.org/10.1017/CBO9780511518539.007 Onlineonline
Published © Cambridge University
by Cambridge Press,
University Press 2010
 272 Alexander and the army

into India, and at the Hydaspes he employed cavalry from Arachosia,

Parapamisadae, Bactria, Sogdiana and the Saca territories to the north (Arr.
v. 11.3, 12.2). They were supplemented by levies from friendly Indian princes
until the forces under his command reached the almost fantastic total of
120,000 (Arr. Ind. 19.5), only a fraction of which will have been Macedonian.
As yet the oriental cavalry fought in separate national units, the one exception
being the corps of horse archers, recruited in part at least from the nomad
Dahae, who were regularly deployed alongside Macedonian troops. The
majority were in the same category as the Illyrian and Thracian contingents in
334, levied as much to reduce the resistance in their homelands as to increase
the strength of Alexander's army. They apparently kept their separate
identities until the return to the west, when there was some integration with
the Macedonian cavalry. According to Arrian one of the Macedonian griev-
ances at Opis was the admission of selected orientals into the Companion
cavalry: four hipparchies were formed exclusively from the easterners and
there was a fifth which combined Macedonians and Iranians. 45 We should
dearly like to know more about the transition, the numbers of Iranians
involved and their strength relative to the Macedonians, but Arrian's language
is systematically elusive. All we can say is that Alexander gave a certain
number of Iranians, the coveted title of Companions and brigaded them in
hipparchies alongside the Macedonians. Except for the few Iranians in the
fifth mixed hipparchy (and the handful of nobles in the agema) there was little
attempt to integrate the two national groups. Rather Alexander had set two
groups of Companions side by side in a very uneasy relationship. In any case
the selected Iranians who were admitted to the ranks of the Companions were
a small minority of their national contingents, which comprised the great
majority of the cavalry present with Alexander in his last years.
The evidence for oriental infantry is less ambiguous. When Alexander left
Bactria in 327, he gave orders for 30,000 youths to be recruited and trained in
Macedonian arms and discipline. Accordingly the satraps and city comman-
dants in the eastern provinces organised a concerted programme of training, no
doubt using the veterans settled in the new foundations as instructors; and
early in 324 the new phalanx presented itself at Susa and staged an impressive
display before the king himself.46 There was no attempt to amalgamate this
new body with the Macedonian phalanx. It was to be a rival formation, an
antitagma as Diodorus calls it, 47 to be used as a separate entity. The
suggestive name (Epigoni) given to these recruits implies that Alexander
thought of them as the heirs of his Macedonian phalanx, now almost
superannuated. The new unit maintained its integrity while the Macedonian
phalanx proper was adulterated by the admixture of Peucestas' levies from
45
Arr. VII.6.4. For this interpretation see Bosworth 1980a, 15-17, 2 0 - 1 . See also Brunt 1963,
4 3 - 5 ; Griffith 1963, 6 8 - 7 4 ; Badian 1965a; H a m m o n d 1983b.
46
Arr. VII.6.1; Plut./l/. 71.1; Diod. xvn.108.1-3; Curt, vn.5.1.
47
xvii. 108.3: cf. Briant 1982b, 32-41.

Cambridge Books
https://doi.org/10.1017/CBO9780511518539.007 Onlineonline
Published © Cambridge University
by Cambridge Press,
University Press 2010
 The structure of command 273

Persis (see above, p. 170). Until such time as reinforcements arrived from the
homeland, the Macedonians (with the exception perhaps of the hypaspists)
were swamped in a mass of Iranian infantry, indispensable in expertise but
numerically weak. On the other hand selected Iranians were now expert in
Macedonian techniques and formed a reserve which the king might eventually
use as his front-line infantry.
The character of the army had changed irrevocably. The Macedonians no
longer enjoyed supremacy over the other units of the army. Not only were
they outnumbered but there were contingents of Iranians who had almost
equal prestige - the new hipparchies of Companions and the phalanx of
Epigoni. The change in the army reflected Alexander's own transition from
king of Macedon to king of Asia. His Macedonians were in his eyes no longer
a privileged elite but subjects on much the same level as the Iranians. He had
served notice at Opis that, if necessary, he would man his army and officer
corps from Persians, and his new army was a constant reminder of the fact. Its
future moreover was clearly defined. The recruiting grounds had been the
eastern satrapies of the empire, but service was to be in the west, in conditions
as alien as Bactria had been to the Macedonians. Ultimately they would
become deracinated, the only constant being their employer, Alexander. The
process would go even further in the next generation. Alexander had
deliberately retained the offspring of his Macedonian veterans when he
demobilised them, promising to train them in Macedonian style. 48 His
ultimate purpose was to weld them into a military force without attachment of
race or domicile, loyal to himself alone. The transformation of the Mace-
donian national army with its regionally based units could not have been more
complete.

iv The structure of command

Little is known of the command structure of the army at its lower levels,
although there are indications of a complex gradation of ranks. In the phalanx
each file or dekas had four members of superior status, who were paid
accordingly. Apart from the file leader (dekadarches) there were two who
received double pay (dimoiritai) and a third who received pay and a half (Arr.
VII.23.3; Succ. F 24.2). They served in particularly prominent positions and
were paid according to expertise. Above them were the commanders of lochoi.
Nothing is known of their numbers and we do not have a single name on
record, but they were important enough to be included in Alexander's council
before Gaugamela (Arr. in.9.6). The hypaspists will have had a similar
organisation at file level, but the only intermediate rank between file and
chiliarchy appears to have been thepentakosiarchy, the commanders of which
were relatively obscure. 49 The subdivisions of the Companion cavalry are a
48
Arr. VII.12.2; Justin xn.4.2-10.
49
Curt, v.2.5: cf. Berve 1926, 1.127-8; Milns 1971, 192-4; Bosworth 1980a, 149.

Cambridge Books
https://doi.org/10.1017/CBO9780511518539.007 Onlineonline
Published © Cambridge University
by Cambridge Press,
University Press 2010
 274 Alexander and the army

similar mystery. The lowest division recorded is the hekatostys,50 which

should have had a nominal strength of ioo (though the analogy of the dekas in
the infantry is a warning against taking the figure too seriously); and from 331/0
there were the larger subdivisions of lochoi. Nothing is known of these minor
commands, and the only name recorded is that of the 'hipparch' Callines who
intervened briefly at Opis (Arr. vn.11.6).
The positions in the hierarchy which mattered most were of course the
commands of individual contingents, but even at the higher levels there were
differences of rank which are hard to elucidate. Before Issus Alexander
summoned to his council the commanders of the infantry, the cavalry ilarchs
and the commanders of the allies (Arr. 11.7.3); and there was a similar meeting
before the siege of Tyre which included the entire body of hetairoi as well as
the commanders of specific units (Arr. 11.16.8). These episodes merely give us
the totality of the high command, not the gradations of rank within it. It
appears, however, that the commanders of the cavalry ilai were on a slightly
inferior level. Arrian (11.10.2) associates them with the officers of the infantry
lochoi and the more prominent mercenary officers, below the officers of
highest rank. That fits the rest of the evidence. Though the ilarchs are
occasionally mentioned by name in the campaign narrative, they are not of
great distinction (except for Cleitus the Black, commander of the Royal He);
and they never have independent commissions. The same seems to have been
the case with the hypaspist officers: the chiliarchs and pentakosiarchs were on
a markedly lower level than the battalion commanders of the phalanx, the
generals proper (cf. Arr. vn.25.6). It looks as though the first stratum of
command below Parmenion and the king himself comprised the overall
commanders of the Companion cavalry and hypaspists (Parmenion's two
sons, Philotas and Nicanor) and the six generals of the phalanx taxeis. It is
they who are attested at the head of multiple contingents in the absence of the
king and Parmenion: Philotas, for instance, took the cavalry and three phalanx
taxeis to counter the Persian fleet at Mt Mycale (Arr. 1.19.8) and Craterus and
Perdiccas were left in charge of operations at Tyre while Alexander cam-
paigned in the Antilibanus (Curt, iv.3.1). If there was any member of this
group who outstripped the others in the early years, it was probably Craterus,
who had general command of the infantry on the left at both Issus and
Gaugamela and took care of the camp at the Persian Gates while Alexander
made his turning march. 51
A somewhat anomalous position in the hierarchy of command is that of the
Royal bodyguards. This group, the inner circle of the hetairoi, was the
institutionalised relic of the old bodyguard of nobles, and it still provided the
king's immediate entourage. 52 Membership was incompatible with any post
away from court, and both Balacrus and Menes were replaced as soon as they
50
Arr. vi.27.6; vn.24.4.
51
Arr. 11.8.4; Curt. in.9.8; Arr. 111.11.10; 18.4; Curt, v.4.14.
52
Berve 1926, 1.25-30; Hammond and Griffith 1979, 2.403.

Cambridge Books
https://doi.org/10.1017/CBO9780511518539.007 Onlineonline
Published © Cambridge University
by Cambridge Press,
University Press 2010
 The structure of command 275

were assigned to provincial commands. 53 It also seems to have been incompat-

ible with commands in the army. In the early years of the reign there is no
known instance of a bodyguard holding a senior commission; and, when
Ptolemy son of Seleucus took over command of a phalanx battalion he ceased
to hold the title of bodyguard. 54 None the less individual bodyguards are
occasionally attested in command of army groups on an ad hoc basis; another
Ptolemy, also a bodyguard, led a joint force of hypaspists and light-armed
during the siege of Halicarnassus (Arr. 1.22.4-7). I* would appear that the
group as a whole enjoyed the same status as the commanders of the phalanx
battalions, but its members had no place of their own in the command
structure. The same was true of Parmenion. He had no specific command of
his own but was consistently used as Alexander's second-in-command, taking
the left in the major battles and the first choice for subsidiary campaigns. As a
result he was particularly associated with the allied and mercenary forces
(Diodorus xvn.17.3 even gives him a vague general command of the
infantry), which usually came under his leadership, but there was no single
body of troops permanently attached to him.
Various changes took place in the course of the reign. In particular the
cavalry commands became much more important, largely eclipsing the
phalanx positions. This was a definite policy decision on Alexander's part.
After the execution of Philotas and the assassination of Parmenion Alexander
set his face against large single commands and abolished the position which
Philotas had enjoyed.55 The Companions were first divided between Cleitus
the Black and Hephaestion, but the individual hipparchies soon became
important entities in their own right, and by the time of the invasion of India
the hipparchy commanders appear to have been equal in status. Now the more
highly favoured of the phalanx taxiarchs were transferred to hipparchies:
Perdiccas, Craterus and later Cleitus the White. 56 At the same time the
separation of the bodyguard from the command structure was gradually
eroded. Perdiccas, who was elevated to the bodyguard in 330, had command
of a hipparchy by 327 and Hephaestion similarly combined the two functions.
There were possibly other combinations. It is highly probable that Peithon
son of Crateuas, a bodyguard by 325, is the Peithon attested in command of a
phalanx battalion in 326/5.57 But the bodyguard was principally associated
with the cavalry, and the association grew stronger as the infantry declined in
number during Alexander's last years. When Craterus took his contingent of
veterans from Opis in 324, he had with him at least three phalanx taxiarchs,
53
Berve 1926, 2 n o s . 200, 5 0 7 .
54
Arr. 1.24.1; cf. 11.8.4, 10.2, 12.2.
55
A r r . i n . 2 7 . 4 ; cf. 1.25.5.
56
Arr. iv.22.7; v.12.2 (phalanx taxis); v.22.6; vi.6.4 (hipparchy). The fact that Cleitus was
dismissed from Opis with Craterus (Justin xn.12.8) does not entail that he retained his
phalanx command. See, however, Berve 1926, 2 no. 428; Hammond 1980b, 466—7.
57
A r r . v i . 6 . 1 , 7 . 2 - 3 (phalanx taxis); vi.28.4 ( b o d y g u a r d ) . Pace Berve 1926, 2 n o s . 6 2 1 , 6 2 3 ,
there is no good reason to deny the identity of the phalanx leader with Peithon son of Crateuas.
By the end of the reign he m a y have moved on to a cavalry c o m m a n d (Arr. Succ. F 1.2).

Cambridge Books
https://doi.org/10.1017/CBO9780511518539.007 Onlineonline
Published © Cambridge University
by Cambridge Press,
University Press 2010
 276 Alexander and the army

including the veteran Polyperchon.58 Meleager was the only phalanx com-
mander to remain in his post until Alexander's death, and he sustained the
infantry mutiny at Babylon almost single-handed. On the other hand the eight
commanders who are listed with the cavalry were the most distinguished men
remaining at court, and six of them are previously attested as bodyguards. 59 It
is a dramatic illustration of how the balance of power had changed. The
phalanx commands had not been down-graded but the cavalry had become
enormously more important as the cream of the court was assigned commands
in it. Above all, when Alexander installed Hephaestion as vizier (chiliarch), he
did not associate him with his infantry guard, as seems to have been the case
with his Achaemenid predecessors, but left him in charge of his hipparchy,
thereafter known as Hephaestion's chiliarchy.60 The change of emphasis is
certain but the reason for it can only be guessed. There is some evidence for
increasing disillusionment among the infantry from the time of the murder of
Parmenion, culminating in the two mutinies at the Hyphasis and Opis, and it
would not be surprising if Alexander had deliberately aimed at increasing the
prestige and importance of the cavalry.
The other main development in military organisation is an increasing
mobility of command. From late 330, when the guerrilla warfare in eastern
Iran broke out, there was an increasing tendency to divide the army between
several commands, formed for a specific strategic purpose. At the beginning of
328, for instance, Alexander left four phalanx commanders at Bactra to
control the area south of the Oxus and then split the rest of his army into five
separate columns commanded by senior officers, three of whom are known
bodyguards (Arr. iv.16.1-3). These divisions are quite different from the
separate campaigns attested at the beginning of the reign. Then Alexander
tended to detach his allied and mercenary troops while retaining the Mace-
donian forces in their entirety. Now he divided his forces more or less
indiscriminately, Macedonians and mercenaries alike. These separate com-
mands were given to a relatively small number of officers: Craterus, Hephaes-
tion, Coenus and Perdiccas tended to be used in the first instance and Ptolemy
(son of Lagus), Leonnatus and later Peithon if secondary columns were
necessary. On the march into India Hephaestion and Perdiccas were sent
ahead to the Indus with a massive force, comprising almost half the Mace-
donians and all the mercenary foot (Arr. iv.22.7), while Alexander fought an
intensive campaign along the Cophen valley, using a multitude of smaller
columns put together as the occasion demanded. 61 One of the peculiar features
of these missions is the tendency to detach the phalanx commanders from their
troops. In the summer of 327 Craterus was left to pacify the territory around
the city of Andaca together with the rest of the infantry commanders, even
58
Justin XII. 12.8 (Polyperchon, Gorgias and Antigenes (cf. A r r . v.16.3)).
59
A r r . Succ. F 1.2 (Perdiccas, L e o n n a t u s , Ptolemy, Lysimachus, Aristonous, P e i t h o n ) .
60
A r r . vii.14.10; D i o d . x v m . 4 8 . 4 ; cf. Schachermeyr 1970, 3 1 - 7 (see, however, Lewis 1977,
17-19).
61
Arr. iv.23.5-24.1; 24.8-10; 27.1, 5-6; 28.7-8; 30.6.

Cambridge Books
https://doi.org/10.1017/CBO9780511518539.007 Onlineonline
Published © Cambridge University
by Cambridge Press,
University Press 2010
 The structure of command 277

though his force contained at most two taxeis (Arr. iv.23.5-24.1); and at the
Hydaspes three phalanx commanders (Meleager, Attalus and Gorgias) were
apparently separated from their taxeis and employed on diversionary tactics
with the mercenary infantry and cavalry (Arr. v.12.1). Most notably Coenus,
who commanded a phalanx battalion from 334 to his death late in 326, played a
prominent role at the head of the cavalry at the Hydaspes, and he is even
assigned a hipparchy by Arrian. 62 This latter command was probably
temporary, but it is puzzling, a good illustration of the variability of the senior
posts towards the end of the reign. Not only senior posts were affected. When
Nearchus, the hypaspist chiliarch, was sent off on a reconnaissance mission in
327, his troops were confined to light-armed and his own chiliarchy was
assigned elsewhere, maybe to his colleague Antiochus. 63
The reasons for this development were to some degree military. The more
multifarious the operations of the army, the more fluid the command structure
became. But there was also a political factor at work. From the time of
Philotas' execution and even before, Alexander was concerned with the
problem of confiding large bodies of troops to a single commander. He
countered it in various ways, interposing new subordinate ranks in the
hierarchy, transferring commands more frequently and detaching senior
officers for special service away from their units. The result he intended was to
make himself the single focus of the army's loyalty. There was to be no
successor to Parmenion. Craterus is the nearest parallel in the scope of his
commands and the devotion he inspired in his men, but even he was not
allowed to be identified with any single group in the army. Although a cavalry
commander, he is.only once attested with his own hipparchy (Arr. v.11.3),
and the expeditionary forces he led varied widely in composition, usually
including phalanx battalions but a different selection each time. He had no
monopoly over any sector of the army, nor had any other commander.
Regional ties and personal ties had become much less important. The sole
uniting factor was the person of the king.
62
Arr. iv.28.8; v.12.2 (phalanx taxis); v.16.3 (hipparchy); v.21.1 (phalanx taxis).
63
Arr. iv.30.6: see, however, Badian 1975, 150-1.

Cambridge Books
https://doi.org/10.1017/CBO9780511518539.007 Onlineonline
Published © Cambridge University
by Cambridge Press,
University Press 2010