THE IMPERIAL

ROMAN ARMY

Copyrighted Material
Copyrighted Material
THE IMPERIAL
ROMAN ARMY

YANN LE BOHEC

Translated by
RAPHAEL BATE

London and New York

Copyrighted Material
 First published in French 1989 as L'Armee Romaine, sous Ie Haut-Empire
by Les Editions Picard

First published in English 1994

by B. T. Batsford Ltd

First published in paperback 2000

Reprinted 2000,2001
by Routledge
11 New Fetter Lane, London EC4P 4EE

Simultaneously published in the USA and Canada

by Routledge
29 West 35th Street, New York, NY 10001

Routledge is all ill/prillt of tile Taylor & Frallcis Croup

~) llJ8lJ Picard editeur

English translation © 1994 B. T. Batsford Ltd

The right of Yann Le Bohec to be identified as the Author of this Work

has been asserted by him in accordance with the Copyright, Designs and Patents
Act 1988

Typeset by J. & L. Composition Ltd, Filey, North Yorkshire

Printed and bound in Great Britain at the University Press, Cambridge

All rights reserved. No part of this book may be reprinted or reproduced or

utilised in any form or by any electronic, mechanical, or othe r means, now
known or hereafter invented, including photocopying and recording, o r in
any information storage or retrieval system, without permission in writing
from the publishers.

British Library Cataloguillg ill Publieatioll Data

A catalogue record for this book is available from the British Library

Library of Congress Cataloguing ill Publicatioll Data

A catalogue record for this book is available from the Library of Congress

ISBN 0-415-22295-8

Copyrighted Material
 CONTENTS

List of plates 6
Introduction 7

I ORGANIZATION OF THE ARMY

1 The divisions of the army 19
2 The men 36
3 Recruitment 68

II THE ACTIVITIES OF THE ARMY

4 Training 105
5 Tactics 120
6 Strategy 147

III THE ROLE OF THE ARMY IN THE EMPIRE

7 History of the Roman army 181
8 The practical role 207
9 Cultural influence 231
10 General Conclusion 253

Notes 260
Summary bibliography 293
Index of names 298
General index 300

Copyrighted Material
 PLATES

Between pages 160 and 161

1 Funerary monuments 20 Gates of Roman marching camps

2 Funerary reliefs 21 A temporary camp after Polybius
3 Military cemeteries at Ha"idra 22 A temporary camp after Pseudo-
4 The garrison at Rome Hyginus
5 Organization of a legion 23 The groma
6 Tombstone of an auxiliary 24 A tent
cavalryman 25 The siege of Masada
7 The eagle-bearer from the 26 An assault on a besieged city
Antonine Wall 27 Artillery pieces
8 The standard-bearer from a 28 Battle formation
funerary relief 29 Defensive organization: theoretical
9 Cavalry from the column of structure
Antoninus Pius 30 Linear defence: theoretical plan
10 Music and religion 31 Camps
11 The adlocutio 32 Rampart sections
12 Decorations 33 Camps: surface area changes
13 The training ground at Lambaesis 34 Roman walls in Britain
14 Trajan's Column 35 Coin of Clodius Macer
15 Weaponry 36 Coin of Septimius Severus
16 Auxiliaries and legionaries 37 Coins
17 A legionary 38 Rapidum: town and camp
18 The Roman army on the march 39 Coin to honour Disciplina
19 Basic fortifications 40 Triumphs

Copyrighted Material
 INTRODUCTION

In AD 9 three legions with their accompanying auxiliaries, under the orders

of Publius Quinctilius Varus, were annihilated in the Teutoburgwald by
the Germans of Arminius. When he heard the news of the disaster
Augustus went into mourning, and according to Suetonius he was prone
to fits of rage for several months during which he would shout: 'Varus,
give me back my legions! >1 The emperor certainly considered that the
army played an important role within the state, but was his point of view
unbiased? Must we accept it without qualifications?

HISTORIANS AND THE ROMAN ARMY

Historians' views on the subject have evolved. In the nineteenth century
they considered events to be of prime importance, writing 'battle-history'.
From this perspective it would be necessary to relate the Teutoburgwald
affair in minute detail. On the other hand, in the mid-twentieth century
the Annales school of historians focused on the quantitative and social
aspects. From this angle, Varus' army should be described without
virtually any mention of the ambush in which it was destroyed. Today we
consider that the 'structures' (recruitment, tactics, etc) are of fundamental
importance, of course, but evolution is not neglected, while every effort
is made to situate events, major conflicts and even battles exactly.
Moreover, two recent works 2 have shown the importance of war in
ancient times. According to Y. Garlan war is the expression of a society:
thus at Teutoburgwald, senators and equestrians perished, as well as
plebeians, both citizens and foreigners. J. Hamand goes even further: he
thinks that war expresses a civilization in its entirety; it is not related
merely to social history, but also to political, economic, religious and
cultural factors.

Copyrighted Material
 INTRODUCTION

Furthermore, the Roman state is a complex structure, formed by three

major interconnected elements - central administration, provincial
administration and the army - modification to anyone of these three
instruments of power inevitably entailed a reshaping of the other two,
precisely because of the close links that existed between them. Yet no
synthesis about the third of these areas of study has been published
recently; there is therefore a gap to be filled. While the organs of
government, economic and social life, religion and culture in the Early
Empire are becoming more familiar, the army still holds certain mysteries.
There are, of course, many accounts of excavations describing numerous
fortresses; two books 3 have concentrated on the Roman army and the
Roman soldier, but their contents, far from overlapping, complement each
other, and there is a need for an overall picture of the subject. This lacuna
can probably be explained by the risks involved in such a venture (the
fear of narrating 'battle-history' or the history of events) and by the
discredit that has been heaped on military questions. Is it necessary to say
that this lack of esteem seems totally unjustified?

CERTAIN PARADOXES AND PROBLEMS

Rome's military history comprises many points of interest, some of which
are paradoxical. Before discussing them it must be specified that this book
will deal only with the first three centuries AD, that is the Early Empire:
the vast movement of conquests that left its mark on the Republic drew
to a close with Augustus, while a new order was being installed, both in
strategy and organization of the army. On the other hand, Constantine and
Diocletian laid the foundations for a new era in these fields. They modified
significantly the process of recruitment of soldiers and the distribution of
the forces in charge of maintaining security within the Empire. 4
To return to the paradoxes mentioned above, the most important one
is of the greatest interest to historians. Rome certainly built for itself a
vast and long-lasting Empire, and this, of course was thanks to its troops.
But these conquerors suffered disasters like that of the Teutoburgwald;
their weaponry was heterogeneous, the different elements having often
been taken from the losers the previous day,5 and their sense of discipline
would certainly have shocked many a twentieth-century officer. What,
then, is the true value of the Roman army?
But this is not all, and at least four more questions need to be asked.
Firstly, were these soldiers even capable of maintaining law and order? If
certain specialists like P. Petit6 believed in the 'Roman peace', this concept
is largely a myth for others like Y. Garlan. ? The Empire would have been
attacked at the same time by barbarians from the outside and brigands
from the inside.

Copyrighted Material
 INTRODUCTION

Secondly, what was the social composition of this army? This problem,
of fundamental importance to the contemporary historian, is very
complex; many questions have been asked to try to determine both the
social backgrounds and the nationalities of recruits. M. Rostovtzeff wrote
that in 238 the urban civilians were opposed to the peasant soldiers during
particularly violent riots, but this theory has since been criticized. What
is more, we now know that certain values helped to form the collective
mentalities, and P. Veyne 8 has shown how power, prestige and honours
all play their part along with money, in fact, all that goes to make up
'appearances' (though it is true to say that in the case studied by Veyne
the military are not involved).
But there is in addition a technical problem; it is clear that types of
unit, command, strategy and tactics have not been examined thoroughly,
or even at all, for a long time now. Research is often based on erroneous
information: some authors copy others when they are ignorant of the
realities involved, and use in any context - normally, of course, the wrong
one - Latin terms such as vexillatio, castra or the surprising castrum, the
precise meaning of which they are unfamiliar with.
Finally, an effort must be made to answer the last question: 'what was
the precise role of the Roman army in the world at its particular time? '.
What was stated above must be repeated here, namely that it was one of
the component elements of central power, a 'structure' of the state. Its
importance will then be better understood if it is remembered that it was
also linked to civilian society. The army had a certain influence on the
provinces it was situated in, for example when spending salaries, while in
its turn it was subject, through recruitment, to the influence of the
background in which it evolved. There are, then, three areas of concern:
politics, economics and spirituality (i.e. Romanization and religion).
If some new insight is to be brought to a general history of the Roman
army it is essential to stick to a principle which is here called a global
principle. While it would be somewhat pretentious to wish to say
everything there is to be said in one book - and this is not my intention
- nevertheless, it seems evident that many questions will go unanswered
if only one aspect of the subject, one method 9 or one type of source is
adopted; a synthesis must provoke confrontations within each of these
categories. Everything is interconnected and one cannot hope to under-
stand what the Roman army really was if one studies recruitment without
also studying strategy; aerial photography without excavation reports;
inscriptions without literary sources.

Copyrighted Material
 INTRODUCTION

THE SOURCES

Literary sources
The ancient authors have not received their due attention from historians
who have been attracted by the new information to be gleaned from
excavations, while Latinists have often ignored what archaeology and
inscriptions have to offer. Yet the former could have avoided many errors
and the latter many misinterpretations if this were not so.
These authors can be put into two broad categories - one for whom
military matters are not of major concern, but who nevertheless shed
much light on the subject: Polybius and Julius Caesar for the Republican
period, Josephus, Pliny the Younger, Suetonius, Tacitus, Aelius Aristides,
Cassius Dio and the Historia Augusta for the following period. Further
details can be obtained from the Talmud of Jerusalem and that of
Babylonia, treatises on religious matters based on real events, compiled
by Rabbis between the second and fifth centuries. Until now nobody had
thought of reading them with this intention. Unfortunately the facts they
relate are at worst late, at best misdated. The same is true of the legal
documents, the Theodosian Code and the Institutes of Justinian.
However, better material exists, for certain ancient theoreticians wrote
exclusively on the art of war. 10 Essentially they are tacticians and include
specialists on sieges or stratagems: Onesandros, Vitruvius (in Book IX of
De Architectura), Frontinus, Aelian, Pseudo-Hyginus, Arrian, Polyaenus,
Modestus, and especially Vegetius, who interpreted the Early Empire to
the best of his ability from his fourth-century standpoint. In addition, the
possibility that Augustus and Hadrian instituted regulations for the army
has been raised,11 but it is above all Arrius Menander who is to be read
on this subject. Whatever their individual merits, these works often
contribute to a better understanding of inscriptions.

Inscriptions
The Romans often engraved texts on hard materials, and this mania or
fashion, which the army shared, has left us a heritage of several hundreds
of thousands of inscriptions. 12 They can be classified into three groups.
'Military diplomas' are certified true copies of imperial documents
granting citizenship to soldiers at the time of their demobilization or to
their children or to the mothers of these children. Secondly, there are
epitaphs. Finally, there are dedications, which are termed honorific when
written to celebrate the merits of a mortal, religious when addressed to
one or more gods, commemorative when their aim is to immortalize some
event or other (a victory, the construction of a building etc.); moreover,
they are termed individual if requested by one person only and collective
if several people have contributed, as is the case when clubs, called

10

Copyrighted Material
 INTRODUCTION

colleges, are formed at the time of demobilization of soldiers of the same

age group, or in some other circumstances. These collective dedications
are usually composed of two parts: the dedication itself and the list of
sponsors. Very often the two parts have become separated with the
subsequent loss of the first, the second being called 'military lists' (in Latin
latercula which is preferable to laterculi the form universally accepted
since Th. Mommsen). This has given rise to a fairly widespread
misinterpretation since many historians wrongly believe that these lists
are independent - archives established by the authorities to know how
many men were at their disposal or how many were to be demobilized.
In fact, these are private, not official, documents.
One major problem is that these documents rarely carry any clear date.
To obtain some notion of their chronology, both the archaeological
context, when known, and especially the formula used , must be examined.
For example an epitaph found at Mainz' 3 says: '(Here lies) Cneius
Musius, son of Titus, of the tribe Galeria, (born in) Veleia, aged thirty-
two years, having completed fifteen years service, eagle-bearer of the
XIV legion Gemina. His brother, Marcus Musius, centurion, has set up
(this epitaph).' A specialist would date this burial in the first half of the
first century AD, based on the three following criteria: the names of the
individuals; indications of civilian nature (age etc.); indications of military
nature (length of service etc.). What follows is a brief look at those factors
which the specialist bases a judgement on. To be sure, 'there is only local
epigraphY'.14 To study, and above all date, an inscription, only the criteria
established for the region or city of that inscription must be taken into
consideration. Nevertheless, as long as a too precise date is not put
forward, a few major invariables can be established.
The name of a Roman citizen may comprise several elements: the
praenomen (Caius), the gentile nomen (Claudius), at least one cognomen
(Saturninus), the filiation (so n of Lucius), the tribe (Galeria), the
homeland (a city) and the signum (A ntacius):

praenomen gen. nomen filiation tribe cognomen homeland signum

Caius Claudius Caii f. Galeria Saturninus Abella Antacius

Table 1: The names of a Rom an citizen in the second century AD.

It is tempting to translate praenomen as 'first name', cognomen as

'familiar name' and signum as 'nickname', but this would be absurd. The
gentile nomen, common to anyone whose ancestors received citizenship
from the same magistrate or emperor (Julius, Claudius, etc.) presents a
collective aspect, whereas the praenomen, the cognomen and signum

11

Copyrighted Material
 INTRODUCTION

individualize their bearer. The interest of onomastics here lies in its

variation according to date, social class and geographical origin. Thus the
tria nomina (praenomen - gentile nomen - cognomen ) characterize the
Roman citizen of the second century AD; before the Flavians the cognomen
is often missing, whereas in the third century the habit of mentioning the
praenomen is lost. When there is also the filiation, the tribe and the
homeland, then the text dates from the first century AD. The signum,
which appears at the end of the second century, was vulgar at that time
but developed a certain cachet in the Late Empire. The cognomen carries
a lot of information: it ca:1 indicate when the homeland is not mentioned,
and if it is not of Latin origin, a person's province (a Hasdrubal will
necessarily be an African); if it comes from the Greek (Cleitomachus,
Epagathus, etc.) it corresponds to an oriental or servile origin, or possibly
to a fashion - that of Hadrian's era; if it is found without gentile nomen
and particularly if it comes from a barbarian dialect (for instance from
Thracian like Bithus or Phoenician like Hiddibal) it indicates a peregrinus,
a lower-class citizen, even a slave. The presence of several cognomina
indicates nobles and social climbers.
The information of a civilian nature yielded by an inscription can be of
different sorts. The use of 'vulgar' Latin forms (for instance Elius for
Aelius) is considered 'late'. The mention of two emperors ruling together
(e.g. the abbreviation Augg. for Augusti duo) cannot be earlier than
AD161, the year in which Marcus Aurelius took Lucius as his co-ruler.
When it is a question of the price of a monument the word 'sesterces' is
written 'H5' in the second century, '55' in the third and 'I-5' in between.
For epitaphs 15 the use of the nominative case would point to the first
century, but a text beginning with the invocation 'To the 5pirits of the dead'
could not be before the end of the first century, while one beginning
'Memoria of X .. .' could not be earlier than the end of the second century.
Information of a military nature 16 also sheds much light on the subject.
Inscriptions relating to auxiliary troops are considered early (first century
AD) if the number of the corps follows the name of the unit (ala
Pannoniorum J instead of ala J Pannoniorum), if the unit is designated in
the ablative case with or without preposition (miles ala Pannoniorum or
ex or in ala Pannoniorum) or if an officer indicates his rank only by
praefectus equitum without any further precision. On the other hand,
honorific epithets bestowed on wings and cohorts (torquata, felix etc. ) do
not appear before the Flavians.
Soldiers on the whole rarely mention the century to which they belong
except in the first century, and the same is true of their length of service
if they use the verb militavit: 'X of Rufus' century, served Y years'.
However, when their length of service is expressed by the noun
stipendiorum the text is probably from the third century. The use of
aerum was a fashion largely but not exclusively restricted to Spain. The
indication of ranks, and notably successive ranks corresponding to a

12

Copyrighted Material
 INTRODUCTION

career, is largely a third-century fashion. On the other hand, contrary to

what was once received wisdom, the mention of comrades-in-arms as
dedicators of a tomb has no special significance. Finally, it should be noted
that a closely-argued study 17 has shown that the expression centuria Rufi
indicates that the centurion Rufus still occupies his post, whereas the
formula centuria rufiana indicates that he has left his unit and has not yet
been replaced.

Coins and papyri

Numerous mmtmgs of coins 18a also shed light on the history of the
Roman army. Some coin inscriptions glorify legions or even whole armies
whose emperor (or a pretender) sought their support, as did Macer and
his legio I Macriana, Hadrian and the various provincial exercitus. Others
disseminate the military themes of imperial propaganda, the loyalty of
armies (fides exercituum), particularly when the emperor is not very
confident of it. Discipline is another theme thus treated. In the third
century some mints were operating solely to satisfy military needs.
Egypt and especially Dura-Europos have yielded a large number of
papyri, either official archives or private documents, which are of the
greatest interest. 18b

Archaeology
Excavations do not only yield inscriptions. 19 For some time now we have
been aware of coins commemorating victories or else minted in honour
of units that had distinguished themselves in action. But a more immediate
interest can be found in the study of funerary monuments and military
constructions. It is known that burial was practised at a later date than
cremation, although cremation experienced a return to favour from time
to time. With the African army it has been possible to chart an evolution:
in the first century the soldiers' corpses were placed under stelae or flat
stones; in the second century under cube-shaped altars, and in the third
under 'cupules', half-cylinders resting on their flat section (pU,l). Some
of these tombs were decorated with reliefs, especially those of officers and
centurions. There is evidence of busts, some in simple relief, some within
a niche, some in a temple (pU,2a). Sepultures depicting horsemen have
also been found; these may be depicted dismounted and facing the
onlooker or in the act of riding away or killing a fallen enemy (pU,2b).
Another sort represents a standing figure sacrificing, or at a funeral
banquet or facing onlookers (pl.II,2c). These sepultures were generally
grouped together along roads leaving the camp or placed in a semicircle,
at first around the fortresses and later around the civilian agglomeration
that grew up alongside the fortress (pl.II,3).
However, it is military archaeology that provides the most information.

13

Copyrighted Material
 INTRODUCTION

To begin with, several hundred fortresses and 'defensive lines' around the
Mediterranean basin are known, the most famous of the latter is
Hadrian's Wall in Britain. The existence of such ruins has sometimes been
discovered by aerial photography, but these discoveries must always be
checked on the ground. This technique has been used with particularly
good results for the study of the Roman army in Syria by A. Poidebard
and in south Algeria by J. Baradez;20 and there is nothing to exclude the
idea that one day artificial satellites will replace the aeroplaneY
Archaeology also provides some large monuments. Trajan's Column in
Rome represents in fact a volumen , that is a book upon which has been
depicted through sculptures and not words the Roman victories over the
Dacians at the beginning of the second century AD; moreover, it was
erected between two libraries. (It is 29.78m (98ft) high on a 10.05m (33ft)
base). The more damaged reliefs of the Column of Marcus Aurelius, also
in Rome, recount the wars waged by him against the Germans and
Sarmatians (it was sculpted in AD180 on a shaft measuring 29.6m (97ft)).
Finally the Adamklissi monument in Romania has an enormous circular
base which supports a tropaeum, commemorating one of Trajan's
victories. 22

THE DUTIES OF THE ROMAN ARMY

After establishing a list of the sources that can be used to study the Roman
soldiers, and before describing this army and its evolution, it would be
useful to give a brief but precise account of its missions. An initial
reflection of this sort will provide a better understanding of some of the
analyses developed later in the book.

Main function: war abroad

When dealing with military matters historians sometimes have a tendency
to forget a basic truth (for fear of telling 'battle-history'); to quote the
words of an unfortunately little-known specialist, Ch. Ardant du Picq,
'combat is the ultimate aim of armies' Y A soldier's job is to wage war
against a foreign enemy. His main task therefore is to kill without being
killed. Legionaries were also duty-bound to ensure the protection of
Roman citizens, wheatfields and temples, the last being of prime
importance in ancient mentality. As a result they had to undergo
preparation by exercises and manoeuvres; they also had to ensure the
upkeep and guarding of fortresses, while also going on patrol to survey
the enemy.

14

Copyrighted Material
 INTRODUCTION

Secondary function: policing

As they represented a reliable force and as the Roman state had never
thought of organizing the maintenance of law and order within its
boundaries, it was the soldiers who performed this task.
They could act in a preventative manner, but in such a case their role
was limited to one of surveillance of potential trouble-makers. Stationarii
and burgarii patrolled the streets and market-places while the navy
endeavoured to prevent the ever-possible return of pirates. In Judaea
decurions were posted in the towns, centurions in the cities. Other officers
were responsible for monitoring what was said in the schools. 24
Their main task, however, was repressive. Runaway slaves had to be
caught, and Pliny the Younger writes of a stationarius doing just that. 2s
During the persecutions of the third century it was often the soldiers who
arrested, interrogated and executed Christians. In fact, their principal task
during peacetime was to eliminate banditry in general,26 though it must
be remembered that in times of civil war political enemies were also often
called bandits (latrones ). It is true that in any period these could be
physically exterminated by a secret police specially formed for this task P
Finally, it is the army that acted as prison warders 28 and ensured the
security of officials by means of ships and escorts. 29

Subsidiary functions
Soldiers were also called upon to carry out duties that had nothing to do
with the use of the force they represented. Some emperors saw in them
no more than a relatively well-qualified workforce that cost the state
nothing. Thus the army had to carry out administrative tasks,3° carry
official letters,31 perhaps protect the collectors of a tax called the
portorium,32 and even, in some cases, take on civil engineering jobs. 33 On
top of that it will be necessary to look again at what was the indirect role
of the troops in the economy (spending of salaries), religion (cults of
certain gods) and culture (spread of Romanization ).H

15

Copyrighted Material