Belisarius: The Last Roman General (7 page)

Whatever the cause, Belisarius now steps from the shadows cast by Justin and Justinian and stands in the full spotlight of history, to be judged on his own merits.

Chapter 3

The New Roman Army

For any student of the earlier Roman Empire, the force that Belisarius took control of for the Persian Wars was a strange entity. The ‘Roman’ army was in the process of change. It was no longer the army of the early Empire, with its c.5,000 man legions forming the backbone of the fighting force. During the reigns of Diocletian, Constantine and their successors, a series of reforms had changed the face of the army. Although the number of legions had increased, they had been reduced in strength to around 1,000 men. There had also been an increase in the number of cavalry and
auxilia
units. Although the cavalry units had been reduced to a strength of approximately 300 men, the
auxilia
probably remained 500 men strong. The result was a much larger, well-balanced army combining foot and horse, close combat and missile weapons, whilst still retaining many of the traditional Roman elements. In contrast, by the reign of Justinian the army would scarcely have been recognisable as Roman by Julius Caesar or Pompey.

The Battle of Adrianople marked a major watershed for the Roman army. Although not the catastrophe often portrayed by historians, it was the beginning of the evolutionary period within the military that would result in the army of Belisarius.

The battle ended with the loss of a huge number of trained troops. Although conscription could rapidly renew the numbers lost, the recruits did not have time to receive the levels of training given to their predecessors. More difficult to replace was the vast amount of lost equipment. It is clear that the official
fabricae
(state factories) would struggle to replace the sheer volume of equipment lost, and it is probable that large numbers of smiths and ironworkers were hired to help to make good the losses of the East.

The new workers had no training in producing the complex Roman helmets that were spun from a single piece of iron. As a consequence, they produced basic styles, copied from steppe or Sasanid examples, as will be shown. The process of change had probably begun before the Battle of Adrianople: the eastern Roman army may have already been coming under steppe influences due to changes in fashion, but the after-effects of the defeat were to hasten the shift over the following decades.

Not long after Adrianople, the Romans finally came into direct contact with the Huns. Fearsome warriors mounted on sturdy ponies, the Hunnic cavalry
were easily capable of riding rings around the Roman infantry, picking them off with their powerful bows whilst the infantry remained virtually powerless to respond. Needless to say, the Huns held Roman infantry – indeed any infantry – in contempt.

Following the collapse of the Hunnic empire after the Battle of the Nedao in 454, many Hunnic bands and individuals were recruited as mercenaries by the eastern empire. Naturally, they brought their strategy, their tactics and their contempt for the infantry with them. The army of the Empire began a rapid change. The spear and javelin-armed cavalry units that had been the backbone of the cavalry since the days of Augustus were retrained in the use of the bow. The new formations – now named
hippo-toxotai
(horse-archers) – retained the armour for the riders, yet achieved speed and manoeuvrability thanks to the strength of their Cappadocian horses. However, they did not adopt the use of barding or other horse armour; although some earlier units had used barding, both the Romans and the Persians appear to have stopped using it by the time of Belisarius. On the subject of horse armour, much has been made of the existence of the Leones Clibanarii unit in Egypt, which according to its name should have been equipped with armour for the front of the horse. Yet the name of the unit was likely to remain unaltered long after the use of such armour had been abandoned. Also, the unit was stationed in Egypt, a peaceful province, where the use of armour had probably fallen out of favour when the units involved had been given the option of purchasing their own. As a consequence, the continuation of the unit named Leones Clibanarii does not necessarily prove that Roman cavalry units continued to use barding for their horses.

Roman, or as they will now be referred to – Byzantine, cavalry do not seem to have employed the asymmetrical bow favoured by the Huns, using instead a more standard symmetrical bow that was much easier to produce in the large numbers necessary to re-equip the entire cavalry arm. However, like the Huns they relied on the power and accuracy of their shooting to cause casualties. This can be compared to the reliance upon the sheer massed volume of fire preferred by the Sasanids and their Parthian predecessors.

Organisation

The Byzantine cavalry units appear to have declined from their traditional strength. Although the actual strength of units is unclear, it is possible that the army had by now begun its transformation into the Byzantine army of the
Strategikon
of Maurikios. If this is the case, cavalry would now have been formed in
banda
of around 300 men. There would be ten men in a
decharchia,
five
decharchia
in an
allaghion,
two
allaghia
in a
hekatontarchia
and three
hekatontarchiai
in a
bandon,
totalling 300 men plus officers, musicians and other supernumeraries, all led by a
tribounos.

To complete the tale, in later practice there were a variable number of
banda
in a
chiliarchy,
commanded by a
chiliarch
(from
chilias,
a thousand, although a
chiliarch
usually commanded between 2,000 and 3,000 men). Two or three
chiliarchs
were gathered in a
moira
(plural:
merē)
led by a merarch, who commanded perhaps 5,000–6,000 men and never more than 7,000; three
merē
formed a
dhoungos.
Since the emphasis at this time was upon the cavalry, it is probable that the cavalry, rather than the infantry, were the first to be reformed upon the new model.. Although actual strength is likely to have been less than 300 due to deaths, desertion and recruitment difficulties, thanks to their perceived importance and improved status it is likely that the cavalry attracted recruits far easier than the infantry and so were nearer to their paper figure in reality.

With the increasing importance of the cavalry in the eyes of the generals came an increase in training, with exercises in all-round shooting, speed of manoeuvre, and hand-to-hand combat. This was mirrored by a comparative decline in the training available for the infantry, which lost much of its elan and combat ability. By the time of Belisarius, the infantry were sometimes expected to turn and run when charged by cavalry – as was to happen to Belisarius at the Battle of Rome.

If the infantry had already begun their change in unit size, there would now in theory be sixteen men in a
lochagias,
four
lochagiai
in an
allaghion,
two
allaghia
in a
hekatontarchia,
and two
hekatontarchiai
in an
arithmos,
leading to a paper strength of 256 men, again led by a
tribounos.
This unit recalls the division used by Alexander the Great in his phalanxes, or pike blocks. It gives a square of infantry with sixteen men per side, enabling the unit to turn to face any direction necessary without a loss of frontage or force. It is eminently suitable to troops wielding a large pike of 4m or more in length, for whom greater depth can be an advantage. It is not necessarily suitable for troops wielding spears and swords.

As a consequence, the spear and sword-armed infantry (it is probably wise not to use the title
legiones
anymore, as this can cause confusion) under Belisarius probably retained their paper strength of approximately 1,000 men. Again, in reality it is unlikely that many of them managed to maintain these numbers in the field. The infantry, previously supported by the
auxilia,
were now supported by an increasing number of foot archers, drawn up either to the side or in the rear of the infantry to give supporting fire.

The size of the specialist archer units is open to question. In the
Notitia Dignitatum,
most
auxilia
units are assumed to be 500 strong. There does appear to have been at least one exception, where such a unit was 1,000 strong, equating to one of the
cohortes milliaria
(one-thousand strong cohorts) of the army of the earlier imperial period. However, the reorganisation may have been used to standardise the size of units, in which case the single large exception
would have been amended. Consequently the archers were probably organised in units of 500 like the earlier
auxilia,
or much less likely 1,000 similar to the spearmen. They used the traditional Roman bows as used by the cavalry. On the other hand, due to their lower height coupled with the relatively poor visibility on dusty battlefields, it is likely that these archers relied on sheer numbers of arrows rather than on the accuracy and power of their mounted counterparts.

It is unknown if there was a pattern to the rearmament, for example with the old
legiones
retained as the close-combat arm and the
auxilia
equipped with bows. If there was, it is likely that the change was governed by seniority and influence rather than troop type, with the
Palatina
retaining their shock role and the units of lower rank becoming archers.

The troops described above appear to be the descendants of the ‘field armies’ described by the
Notitia Dignitatum,
supplemented by some of the frontier troops where necessary. Still stationed in towns and on the frontiers were the
numeri
(‘bands’). Mustering around 200 men in each unit, they were often brigaded in pairs under the command of a
tribounos.
They seldom appear to have been used in the field, being too small to sustain casualties and too poorly trained to be effective in combat. They were thus left to garrison strategic points, forts and cities outside the remit of the regular army.

The formation which the Byzantine cavalry adopted in battle is unclear. Later Byzantine practice was to assume a close formation, which relied on arrow numbers prior to contact and the impact of heavily armed troops in the front ranks during combat to defeat the enemy. Hunnic practice was to adopt a loose formation and fire at the enemy at point blank range. This would force the enemy to either charge to relieve the pressure, or it would cause their formation to be disrupted. If charged, the Hunnic cavalry could evade, leaving the enemy disorganised on blown horses. When the enemy were either disrupted through charging or because of missile fire, any available Hunnic cavalry could close formation and deliver a charge in the hope of breaking the enemy. If the charge failed to rout the enemy, the cavalry would withdraw and repeat the process. If the charge succeeded, the horsemen could either pursue or turn to a fresh target.

Given the employment of large numbers of Huns, and the success that the Huns had had with their tactics, it is far more likely that the Byzantines adopted the latter method, using a loose formation. Only later would the Byzantines adopt closer formations, probably under the influence of their Sasanid adversaries.

When looking at the battles fought by Belisarius it is important to bear one more factor in mind: of the 1,000 men theoretically in an infantry unit, when the troops took the field some would be left behind in their garrison to receive and train new recruits, and act as a permanent base. Of the nominal total it is
more than likely that the absolute maximum fielded by any infantry unit in battle would be 800.

Finally, we can look at the total strength of the army at the time of Justinian. The calculations are complex and have been covered in great detail by Treadgold (Treadgold, 1995, pp. 59–64). His figures are rough approximations based on the numbers given in various sources, primarily Agathias. Despite the difficulties surrounding them, they appear to fit the facts and are given in the following table.

Although the figures given date to 559, it is unlikely that they had been altered to any great extent during the wars in the west and the ongoing wars with Persia. By this time, the eastern armies of the
Notitia Dignitatum
had declined in numbers. The new statistics reveal armies which are divisible by 5,000, which is also the average size of a
moira,
as described above.