Datastream

Download as pdf or txt
Download as pdf or txt
You are on page 1of 29

THE THIRD CENTURY CRISIS IN

THE ROMAN EMPIRE1


By A. R. BIRLEY, M.A., D.Phil., F.S.A.
PROFESSOR OF ANCIENT HISTORY IN THE UNIVERSITY OF MANCHESTER

H ISTORY retains its popularity as a subject of study not


least because it offers refreshment in troubled times;
and there is a ready temptation to search the past for parallels
to our own situation. The historian has to be wary of this
practice, remembering a comment of Max Weber : "It helps
the narrator to make an impression on his audience when they
have the feeling that his story is relevant to them (de te narratur
fabula) and when he can end his tale with the words ' You have
been warned!' (discite moniti). But the discussion which follows
does not have this advantage. We can learn little or nothing
about present day social problems from the history of the
ancient world ".2 Whatever historians may say in protest, the
decline of ancient civilization is still frequently held up as an
example from which we should learn. In particular, mention is
often made of the lowering of moral standards, to which that
decline is attributed.3 To be sure, evidence can be found to
support such a view. The fourth-century writer Aurelius
Victor traced the onset of the crisis to the murder of Severus
Alexander in 235. Henceforth civil war became endemic, he
said; the moral barriers were broken down and the direction
of affairs passed to men of the lowest social origins and education :
" good men and bad alike were placed in the imperial position,

1 A lecture delivered in the John Rylands University Library of Manchester


on Wednesday, the 12th of November 1975.
2 M. Weber, " Die sozialen Griinde des Untergangs der antiken Kultur ",
Gesanunelte Aufsdtze zur Sozial- tmd Wirtschaftgeschichte] (Tubingen, 1924),
pp. 289-311, p. 291 (reprint of a paper originally published in Die Wahrheit,
May 1896).
3 A letter to the editor of the Daily Telegraph of 18 October 1975 may serve
as an example : " We learn that the Roman Empire declined into oblivion
through the dissipation and licence of its people. In other words, a preoccupation
with drink, gambling, and sex. These three, with greed and envy, are endangering
our nation today."
253
254 THE JOHN RYLANDS UNIVERSITY LIBRARY
nobles and commoners, and even a good many barbarians "
(immissique in imperium promiscue boni malique, nobiles atque
ignobiles, ac barbariae multi).1 According to Victor, when the
crisis was at its height, in 260, the emperor Gallienus was
frequenting the capital's eating-houses and cook-shops in the
company of pimps and wine-dealers, and conducting a scandalous
liaison with a barbarian princess.2 This conduct, he goes on,
provoked a further spate of civil wars, leading to Gallienus's
overthrow in 268, when, with the accession of Claudius II, the
recovery slowly began. During the ensuing sixteen years
Claudius, Aurelian and Probus were able to restore Rome's
fortunes vis-a-vis her external enemies, even if they could not
solve the empire's political instability. The end of fifty years
of crisis, of " military anarchy", came with the advent of
Diocletian in 284 and the establishment of the tetrarchy, the
rule of four men : " All of them, of course, came from Illyricum,
and while they had little in the way of humanitas [liberal education,
good breeding, refinement], nonetheless their peasant origin
and their service in the army had sufficiently inured them to
hardship and they were very good for the Roman state " (His
sane omnibus Illyricum patria fuit: qui, quamquam humanitatis
parum, run's tamen ac militiae miseriis imbuti satis optimi rei~
publicae fuere).3
The enigmatic author of the Historia Augusta embroidered
Victor's account of a frivolous and decadent Gallienus, con-
trasted with the simple but virtuous and valiant men who

1 Aurelius Victor, de Ca.es., 24.9 ff. (the quotation is from 24.9). Victor, too,
was humbly born, the son of an unlettered African peasant, but he had moved
up the social scale by serious study (de Caes., 20.5); shortly after completing
his brief treatise on the emperors he was made consularis of Pannonia Secunda
by Julian (Ammianus, 21.10.6). It is unfortunate that the occasion for his self-
congratulatory remark the accession of Severus after the overthrow of Didius
Julianus reveals the inadequacy of his research : he confuses Didius with the
jurist Salvius Julianus (20.1).
2 Victor, de Caes., 33.1-7.
3 Ibid. 39.26 (for the sense of satis see A. Mocsy, Pannonia and Upper Moesia
(London, 1975), p. 388 n. 138, citing the view of I. Borzsak). Cf. the eloquent
comments on this passage and others like it by A. Alfoldi, A Conflict of Ideas in
the Late Roman Empire. The Clash between the Senate and Valentinian I (Oxford,
1952),98ff.,esp. 104 f.
THE THIRD CENTURY CRISIS 255
repaired the damage. 1 Edward Gibbon did his best to exercise
restraint when drawing on that corrupt source, but in the
absence of alternative information could not escape contami-
nation : much of what he wrote on the third century is for that
reason alone worthless as history, however delightful as literature.
We have a clear enough picture of the state of the empire in the
early principate, from Augustus to the Antonines ; and the
fourth century is vividly accessible in the pages of Ammianus
and the Theodosian Code. But the Christian empire is a very
different world from the age of the Antonines, and to explain
how the changes came about one has to turn to the third century
where the real problem is to discover what actually happened.
The literary sources are meagre in the extreme,2 and the temp-
tation to exploit the fantasy-ridden biographies in the Historia
Augusta is one to which too many still succumb. 3 Other
sources exist, but they are not easy to control: and it is here
that the patient study of the Roman provinces by archaeologists
who are also good historical scholars becomes so valuable. Nor
will one forget the vital contribution made by numismatists and

1 See for example HA, Gallieni duo, 3.6 (libidini et voluptati se dedidit), 7.4
(novo genere ludorum, nova specie pompartun, exquisito genere voluptatum), 8.3
(obstupefacto voluptatibus corde), 16.1 (qui natus abdomini et voluptatibus dies ac
nodes vino et stupris perdidit), etc. : longtun est eius cuncta in litteras mittere, as
the author himself conceded (18.6). The reasons for the blackening of Gallienus's
memory are clear enough : his treatment of the senate (p. 276, below), and the
wish to enhance the memory of the man who supplanted him, Claudius II,
alleged ancestor of the house of Constantine, not to mention his attitude to the
Christian church : see above all A. Alfoldi, Cambridge Ancient History, xii (1939),
223 ff. = Studien zur Geschichte der Weltkrise des 3. Jahrhunderts nach Christus
(Darmstadt, 1967), 416ff., and Alfb'ldi's other studies reprinted there, esp. 1 ff.,
228 ff.
2 See F- Millar, " P. Herennius Dexippus : the Greek world and the third
century invasions ", J.R.S., lix (1969), 12-29 for a valuable insight into one of
the historians whose work is preserved only in a few fragments ; on pp. 14-16
he gives a useful conspectus of " patterns of Greek historiography in the second
and third centuries ".
3 M. I. Rostovtzeff commented nearly fifty years ago that " In no field of
ancient history is so much animosity displayed in the discussion of scientific
problems as in the investigation of the Scr. Hist. Aug. " (The Social and Economic
History of the Roman Empire (Oxford, 1926), p. 613 n. 2). The animosity is
now perhaps less intense, but it has not completely abated and I do not wish to
add to it unnecessarily by citing recent examples of credulity.
17
256 THE JOHN RYLANDS UNIVERSITY LIBRARY
papyrologists, whose material actually increases in volume during
this period. Historical scholarship has moved a long way
since Gibbon. Indeed, in some respects the situation has
reached another extreme : " Anyone who sets out to solve a
problem in ancient history has to spend more effort in research
to discover what solutions have already been proposed than in
finding a solution of his own 'V This lament by a Swiss scholar
struck me very forcibly when I read it a few weeks ago. It is
printed in a bibliographical work handbook hardly seems the
right word of mammoth proportions, of which seven instal-
ments have so far appeared, each weighing several kilograms
and containing an average of a thousand pages ; three of these
volumes contain a great deal that is highly relevant to my theme
and copious citations of publications both unfamiliar and
inaccessible.2 This ought to have been enough to have deterred
anyone from tackling this rather hackneyed theme again, and
1 must explain why I have not abandoned my attempt. One
could always, of course, find an anniversary as an excuse and
this would be a good time to commemorate the death of the
emperor Aurelian (the greatest figure in the period of " military
anarchy ") and the curious circumstances of the accession of the
emperor Tacitus : for this took place seventeen hundred years
ago (perhaps to the day ; complete precision is not yet possible).3
But adventitious considerations of this kind influenced me less
than the longstanding attention paid to the third century by
the History Department at Manchester University, and my own
interest in the Historia Augusta. Besides, there are two fairly
recent developments which can be viewed with equanimity by
the scholar ; indeed they must be accorded an unmixed welcome.
In 1963 Andreas Alfoldi and Johannes Straub launched a series
1 G. Walser, "Die Severer in der Forschung 1960-1972", ANRW (see
next note) 2.2 (1975), 614-56, p. 615 n.
2 H. Temporini and W. Haase (eds.), Aufstieg und Niedergang der romischen
Welt. Geschichte und Kultur Roms im Spiegel der neueren Forschung. I Republik-
2 Principal (Berlin and New York, 1972 ff.), cited here as ANRW with volume
and part numbers.
3 The alleged interregnum in the Latin sources was a misconception: see
PIR2 C 1036 (Tacitus), D 135 (Aurelian) ; R. Syme, Emperors and Biography
(Oxford, 1971), pp. 137 ff.; L. Polverini, "Da Aureliano a Diocleziano",
(1975), 1013-35, PP. 1018ff.
THE THIRD CENTURY CRISIS 257
of Historia Augusta Colloquia1 to elucidate the problems of that
maddening2 work in which Gibbon immersed himself with such
unfortunate results. Some of the fruits of the Colloquia may be
barren or bitter-sweet, and not all are to the taste of some
scholars. But many have already derived sustenance from the
writings of Sir Ronald Syme, 3 which may be described as
offshoots of the Colloquia. Secondly, there is the series founded
by Graham Webster and the late Donald Dudley, The Provinces
of the Roman Empire. The first volume, on Britain, appeared
in 1967, followed by three splendid works on the Danubian and
Balkan lands.4 This marks a real turning point in the study of
the Roman empire, not least of the third century A.D. But my
purpose on the present occasion is not to review recent scholar-
ship on the period5 ; still less is it to make moral judgements
about the collapse of classical culture. I can only offer some
observations on the causes of the third century crisis, which
may well seem banal enough to the specialist. Inevitably, in
the space available, I must be selective and may thus appear to
be emphasizing one factor unduly at the expense of many others.
The years 235-284 are characterized above all by the frequent
1 Published as Antiquitas, Reihe 4, Beitrdge zur Historia-Augusta-Forschung,
unter Mitwirkung von Johannes Straub herausgegeben von Andreas Alfoldi.
The Colloquia will eventually, it is hoped, lead to an historical commentary on
the Historia Augusta.
2 Note the comment of F. Millar, A Study of Cassius Dio (Oxford, 1964), p.
124 : " The problem of the Historia Augusta is one into which sane men refrain
from entering ".
3 Particularly Ammianus and the Historia Augusta (Oxford, 1968) ; Emperors
and Biography. Studies in the Historia Augusta (Oxford, 1971) ; The Historia
Augusta : a call for clarity (Bonn, 1971).
4 S. Frere, Britannia. A history of Roman Britain (London, 1967); J. J.
Wilkes, Dalmatia (London, 1969); G. Alfoldy, Noricam (London, 1974); A.
Mocsy, Pannonia and Upper Moesia. A history of the Middle Danube Provinces
of the Roman Empire (London, 1974).
5 Note the valuable survey by G. Walser and T. Pekary, Die Krise des
romischen Reiches (Berlin, 1962); there is now of course the ANRW (p. 256,
n. 2 above), with particularly valuable contributions by X. Loriot, " Les premieres
annees de la grande crise du I He siecle : de 1'avenement de Maximin le Thrace
(235) a la mort de Gordien III ", 2.2, 657-787 ; id., " Chronologic du regne de
Philippe 1'Arabe (244-9 apres J.-C.) ", ib. 788-97 ; M. Christol, " Les regnes de
Vale>ien et de Gallien (253-68) : travaux d'ensemble, questions chronologiques ",
b. 803-27 ; and L. Polverini (p. 256 n. 3, above).
258 THE JOHN RYLANDS UNIVERSITY LIBRARY
changes of emperor.1 There were some twenty "legitimate"
rulers, and the longest reign, that of Gallienus, lasted only eight
years, from 260 to 268 (although he had been co-emperor with
Valerian from 253 to 260). But Gallienus never controlled the
whole empire : his reign marked a peak for usurpations ; and
well over a dozen major usurpers may be singled out during the
period, one of whom, the Gallic emperor Postumus, lasted longer
than Gallienus did himself. Two, perhaps three, died of
disease2 ; the remainder met a violent death, mostly at the hands
of their own men, although two succumbed to the enemy,
Decius dying a hero's death against the Goths and Valerian
meeting a cruel and squalid fate in Persian captivity. 3 This
illustrates sufficiently vividly the political instability of the period
and underlines Rome's military weakness. Certainly, her
external enemies were now more dangerous. The long Danubian
wars of Marcus Aurelius in the 170s had revealed that the
northern frontiers were vulnerable.4 After the peace of 180
stability had prevailed for a generation, but the Germans had
clearly learned from their experiences : new peoples and new
groupings appeared, Goths, Franks and Alamanni. All three,
and others, invaded the empire in the third century and were to
remain a constant danger. 5 Meanwhile the eastern frontier

1 Of course, Herodian, writing in the mid-third century, characterizes the


period from 180-240 (or perhaps he meant 180-250, cf. 2.15.7) as one of
tremendous upheavals, above all in the exceptional number of rulers, in con-
trast to the preceding two centuries. Cf. C. R. Whittaker, Loeb edition (2 vols.,
London, 1969), ad loc., and G. Alfoldy, " Zeitgeschichte und Krisenempfindung
bei Herodian ", Hermes, xcix (1971), 429-49.
2 Claudius 11 died of disease (PIR2 A 1626), Tacitus perhaps from fever
(C 1036, but cf. L. Polverini, ANRIV, 2.2, 1024 and n. 38), Carus from disease
or a thunderbolt (A 1475). X. Loriot, ANRIV 2.2, 77Q-4, argues that the
account of the death of Gordian III in Zonaras and several other Byzantine
epitomators, that he died of wounds inflicted by the enemy at the battle of
Mesiche, rather than by Roman treachery, may be correct.
3 See now on the fate of Valerian B. H. Stolte, " The Roman emperor Valerian
and Sapor I, King of Persia ", Rivista storica deH'antichitd, i (1971), 157-62.
4 See now G. Alfoldy, Noricum (1974), pp. 143 ff.; A. Mocsy, Pannonia and
Upper Moesia (1974), pp. 183 ff.
5 See generally A. Alfoldi, CAH, xii. 138 ff. = Studien zur Geschichte der
Weltkrise, pp. 312 ff.; E. Demougeot, La formation d'Europe et les invasions
barbares, 1 : Des origines germaniques d I'avenement de Diocletien (Paris, 1969).
THE THIRD CENTURY CRISIS 259
had also come under much greater pressure. The Parthians1
had never posed a serious threat but at intervals they had been
given a serious pounding by Rome ; in the 190s they had been
deprived of northern Mesopotamia. Doubtless the Romans
had over-reacted : in the 220s the ramshackle Parthian realm
collapsed, to be replaced by a revived Persian monarchy. The
greatest Sassanid ruler of the third century, Sapor, reigned for
over thirty years, from 241 to c. 272 : this alone gave Persia a great
advantage over Rome. Sapor inflicted repeated defeats on suc-
cessive emperors (of which his own triumphant account is now
known).2 The period was also marked by inflation on a colossal
scale : by the 270s the imperial coinage had been debased out of
all recognition. Much work still needs to be done on the details,
but the overall picture is clear enough : far more coinage was being
produced without any increase in the supply of precious metal.3
All sections of society were deeply affected by these circumstances.
In the meantime, the Christian Church, in the face of persecution
engendered by the crisis, was increasing in strength. 4
The symptoms are well enough known. Rome's enemies
had apparently become stronger, 5 but her armed forces were
barely able to cope, particularly since their energies were so
1 See now M. A. R. Colledge, The Parthians (London, 1967).
2 A. Maricq, " Res gestae divi Saporis ", Syria, xxxv (1958), 295-360; R.
N. Frye, " Parthia and Sasanid Persia ", in F. Millar et al., The Roman Empire
and its Neighbours (London, 1967), pp. 249-69, provides a useful survey.
3 See now J.-P. Callu, " Approches numismatiques de 1'histoire du I He
siecle (238 a 311) ", ANRIV 2.2, 594-613 ; M. H. Crawford, " Finance, coinage
and money from the Severans to Constantine ", ib. 560-93, who stresses that
inflation did not really begin until the reign of Severus Alexander.
4 See now especially G. Alfoldy, " The crisis of the third century as seen by
contemporaries", Greek, Roman and Byzantine Studies, xv (1974), 89-111 =
" Historisches Bewusstsein wahrend der Krise des 3. Jahrhunderts" in G.
Alfoldy et al., Krisen in der Antike. Bewusstein und Bewaltigung (Diisseldorf,
1975), pp. 112-32; id., "Der heilige Cyprian und die Krise des romischen
Reiches", Historia, xxii (1973), 479-501; id., Romische Sozialgeschichte
(Wiesbaden, 1975), pp. 139-64.
5 P. A. Brunt, " Reflections on British and Roman imperialism ", Comparative
Studies in Society and History, vii (1964-5), 267-38, comments (p. 276) that the
barbarians' " success is mysterious. They do not seem to have been more
numerous than in the past ". Whether or not they were more numerous
(statistics are naturally lacking), more of them seem to have combined to fight
against Rome than had been the case before. Note HA M.Ant. 22., 1 : gentes
omnes ab Illyrici limite usque in Galliam conspiraverant.
260 THE JOHN RYLANDS UNIVERSITY LIBRARY
frequently diverted by civil war. Nor was the economy of the
empire able to bear the strain imposed by continuous warfare.1
Yet during the period of more than two hundred years from the
time of Caesar to the early third century Rome had not only
managed to keep external enemies at bay but had even extended
the territories of the empire.2 How had things changed?
It may be a truism to observe that the seeds of this crisis
were sown in the second century B.C. But the fundamental
problem of the principate the problem which brought it into
existence was the control of the army. This situation had
been created in 107 B.C.3 and the problem was never really
solved. " Soldiers had begun in the days of Marius and Sulla
to assist generals to carry out political plans. One is reminded
of the role of the New Model Army during the English Civil War
and afterwards. Caesar, like Cromwell, could control his
troops ; for his successors his ' major-generals ', one is almost
tempted to call them it proved, after his death, an impossible
task. Only a centralized despotism could tame Roman armies
and remove them from politics ; and it was as a despot however
he might disguise despotism with republican catchwords that
Augustus, the adopted son of Julius, kept them tamed. "4 But
even Augustus did not achieve a permanent solution.
1 M. H. Crawford, ANRW, 2.2, 591 f., asks the question why, when there
was a standing army, war should be a great deal more expensive than peace. His
answer is that " army units were under strength in time of peace " and that the
extra expenditure was created by the need to pay the extra men recruited. This
may have been one factor at times, but I cannot accept that army units were
regularly kept below strength except during campaigns. On the other hand,
major wars frequently required the creation of additional units, auxilia as well
as legions. But it ought to be obvious that vast sums would be required during
campaigns for equipment (arms, armour, materiel of all kinds), road and bridge
building, repair of enemy damage, remounts etc., cf. e.g. Tacitus, Annals, 1.71.2,
2.5.3. On the effects of the increased costs, see p. 278 f. below.
2 See A. R. Birley, " Roman frontiers and Roman frontier policy : some re-
flections on Roman imperialism ", Transactions of the Architectural and Archaeo-
logical Society of Durham and Northumberland, M.S. 3 (1974), 13-25.
3 By Marius's creation of a client army. See J. Harmand, L'armee et le
soldat a Rome de 107 a 50 avant notre ere (Paris, 1967) ; E. Gabba, Eserdto e
societa della tarda republica romana (Florence, 1973), pp. 1 ff, 47 ff.
4 C. E. Stevens, " Crossing the Rubicon ", History Today, ii (1952), 373-8,
reprinted in A. Birley (ed.), Universal Rome (Edinburgh 1967), pp. 20-30. The
quotation is from p. 378 (=29).
THE THIRD CENTURY CRISIS 261
Cassius Dio, writing in the early third century,1 put into the
mouth of Caesar a speech before the senate after the African
campaign of 46 B.C. Caesar urged the senators not to " fear the
soldiers or regard them as anything but the custodians of my
empire, which is yours as well. That they should be supported
[rpefaaOai, i.e. paid] is necessary for many reasons : but they
will be supported for you, not against you. . . This is why
higher taxes than usual are being levied at the moment: to
restrain the mutinous elements and deter those that have con-
quered from becoming mutinous, by giving them an adequate
living wage . . . We are always in need of arms, since without
them it is impossible for us, the inhabitants of so great a city
and the possessors of such an extensive empire, to live in security ;
and abundant finances are extremely advantageous in this matter
as in others. But none of you should suspect that I will harm
any of the rich or establish new taxes ; I shall be satisfied with
the present revenues and shall be more anxious to contribute
in some way to your prosperity than to wrong any man on
account of his money ".2 Caesar may or may not have pro-
claimed his intention not to soak the rich to pay the troops Dio
was writing for a third century audience and it suited his purpose
to make Caesar say this.3 At all events, Caesar's murder threw
1 F. Millar, A Study of Cassius Dio (1964), pp. 28-31, 193f., adopts an
arbitrary chronology of the composition of Dio's history, which is in conflict
with Dio's own account (72.23.1 -5 is the key passage) : see e.g. G. W. Bowersock,
review of Millar, Gnomon, xxxvii (1965), 470 ff.; A. J. Graham, " The division
of Britain", J.R.S., Ivi (1966), 92-107, p. 92 n. 11 ; A. R. Birley, Septimius
Severus. The African Emperor (London, 1971), pp. 8f., n. 1. If one takes
Dio's own account literally, and assumes (which may be illegitimate) that he wrote
at a constant speed, he would have written book 43 in A.D. 218 (and the crucial
book 52 in 219). The question certainly deserves further study.
2 Dio 43. 18.1-2,4-5.
3 Millar, op. cit. (in n. 1 above), p. 80 f. concludes that " not only the content
but the occasion was fabricated by Dio ", but sees it as little more than an example
of the " political thought of the Hellenistic and Roman age, [which] was largely
reduced to pious exhortation to rulers to preserve the lives, dignity, and property
of the upper classes ... Dio's sentiments in the speech were applicable to any
age, but perhaps particularly to that of Severus and Caracalla " they would
have especial relevance after Caracalla (see p. 272 f. below), whereas on Millar's
chronology book 43 would have been composed in 213. Cf. also E. Gabba,
" Progetti di riforme economiche et fiscali in uno stonco dell'eta dei Severi ",
Studi in onore di Amintore Fanfani, i (Milan, 1962), 39-68, p. 67 n. 80.
262 THE JOHN RYLANDS UNIVERSITY LIBRARY
the issue open again. The events that followed the Ides of
March showed that only a Caesar commanded support from
veterans and legionaries, and Caesar's heir solidified this support
when he marched on Rome in the summer of 43 B.C., with eight
legions.1 A delegation of four hundred soldiers appeared
before the senate. The senators temporized and one soldier
fetched his sword, '* touched it and said : ' If you will not give
Caesar the consulship, this will' ".2 When Octavian arrived,
he duly got his consulship, and the soldiers got their money:
2,500 denarii a man, more than ten times the annual pay.3
The enormous donative of August 43 B.C. was the capital
investment on which the principate was founded. Octavian
was to become Augustus and men were encouraged to forget
the years of anarchy. Even so, he did not succeed easily. Only
two years later the situation in Italy was menacing: the pro-
scriptions had simply raised the stakes and the men were far
from satisfied with their pay and bounties and land-grants.
" The veterans gathered in Rome in great numbers", but
instead of addressing their grievances to the senate and people
held an assembly of their own on the capitol and passed a
resolution that they themselves should adjudicate between the
feuding generals. Octavian's rivals made fun of the veterans,
calling them a senatus caligatus, a " senate in hob-nailed boots ".4
(Dio and his contemporaries will not have found much amuse-
ment in the story : it was too near the bone.)5
When the Civil Wars were over, Augustus eventually managed

1 Dio 46. 42.4-43.4 ; cf. Suetonius, Divus Augustus, 26.1.


2 Appian, B.C., 3.12, 88; he was joined by three more legions outside the
city (ib. 3.13, 92).
3 Dio 46. 46.1-7; Appian, B.C., 3.13, 94. See the comment by G. R.
Watson, The Roman Soldier (London, 1969), p. 113 ; A. Alfoldi, " Der Einmarsch
Octavians in Rom, August 43 v. Chr.", Hermes, Ixxxvi (1958), 480-96; R.
Syme, The Roman Revolution (Oxford, 1939), pp. 97-186 remains the classic
account of the events from March 44 to August 43 B.C.
4 Dio 48.12.1 -3. A century and a half after Dio wrote, the orator Symmachus
spoke approvingly of the manner in which Valentinian I had been chosen emperor :
emeritum bellis virum castrensis senatus adscivit ; the army was digna plane comitia
tanti imperil prindpatu (or. 1.9, p. 320 Seeck). See A. Alfoldi, A Conflict of
Ideas (p. 254 n. 3 above), pp. 105 f.
6 See p. 272 f., below.
THE THIRD CENTURY CRISIS 263
to pay the men off, settling them in twenty eight colonies in
Italy and scores of others all over the empire, providing them
with land and with cash grants.1 Before he died, in spite of
extreme reluctance on the part of the senate, he set up a military
treasury, the aerarium militare, funded principally by a 5 per
cent, estate duty, the vicesima hereditatium. Thus the praemia
militiae were assured for the future.2 Events very nearly
wrecked the new stability. The last eight years of his life were
dominated by a military crisis. First, the Pannonian revolt of
6-9 threatened to undo his major achievement, the consolidation
of the land-link between east and west. Had it not been sup-
pressed, the empire would have split into two. 3 Within days
of the victory came news of disaster : three legions wiped out
in the forests of Germany.4 At his death, the empire was
1 Res Gestae, 3.3, 15.3, 16.1-2.
2 Res Gestae, 17.2: in aerarium militare, quod ex consilio meo constitution
est ex quo praemia darentur militibus qui vicena out plura stipendia emeruissent,
HS milliens et septingentiens ex patrimonio meo detuli. Augustus here mentions
only his own contribution ; kings and certain communities also made donations,
Dio 55.25.3; and Agrippa Postumus's confiscated property was assigned to it
(ib. 55.32.5). Dio's whole account, 55.24.9-25.6, stresses Augustus's difficulty
in securing agreement from the senate. The vicesima, paid only by Roman
citizens, was highly unpopular, and was only accepted when, in A.D. 13, Augustus
threatened to impose a property tax in its place. Suetonius, Divus Augustus,
49.2, perhaps indicates that other taxes were involved (vectigalibus novis). Tacitus,
Annals, 1.78.2, shows that a 1 per cent, sales tax was also used to fund this
treasury; this tax was halved in A.D. 17 when the annexation of Cappadocia
provided extra revenue (ib. 2.42.6), perhaps raised again in 31 (Dio 58.16.2),
and abolished in 38 (id. 59.9.7). The vicesima continued in existence (p. 273
below). Augustus refers only to praemia and Dio is generally regarded as being
mistaken when he says in 55.24.9 that the intention in A.D. 5-6 was to provide a
source for pay as well as bounties (/ecu rr)v rpo<f>r)v /cat TO. ye'pa). But in the
absence of further information it might be preferable to conclude that by the
time that Dio was writing the arrangements had changed. Of course, he admitted
elsewhere that he could not distinguish the different sources of imperial finance
(53.22.3 fT.). But Suetonius's account, for that matter, could be taken to refer
to pay as well as bounties : quidquid autem ubique militum esset, ad certam
stipendiorum praemiorumque formulam adstrinxit ... utque perpetuo ac sine
difficultate sumptus ad tuendos eos prosequendosque suppeteret, aerarium militare
cum novis vectigalibus constituit (49.2).
3 J. J- Wilkes, Dalmatia (1969), pp. 62 ff.; A. Mocsy, Pannonia and Upper
Moesia(1974),pp.37ff.
4 Velleius 2.117 : intra quinque consummate tanti operis dies, funestae ex
Germania epistulae.
264 THE JOHN RYLANDS UNIVERSITY LIBRARY
desperately short of both manpower and money.1 The legions
on the Rhine and in Illyricum mutinied, demanding more
pay2 the elite praetorian guard had recently (it would seem)
had an increase which widened the differentials in their favour. 3
The new emperor was patient and cunning and somehow he
pacified them without granting their request. No pay rise was
granted for seventy years, until the 80s. 4 But before then the
legions had shown their hand again. After Nero's suicide in
68, several armies sought to put their own man in power. The
long year 69,5 the year of the four emperors, was a terrible
reminder of what might happen at any time : soldiers casting
off their discipline and taking the initiative.6 Vespasian, the
victor, was proclaimed by the Syrian and Egyptian legions;
but it was the Danubian legions which won the war for him.7
The upheaval passed, and discipline was restored.8 But the
precariousness of the new order created by Augustus had been
revealed,9 and all round the empire, on both sides of the frontier,
native peoples had risen in rebellion.10 Tacitus takes the
opportunity, describing the suppression of the revolt in the
Rhineland, to put in the mouth of the Roman commander
Petillius Cerialis what is perhaps the most famous justification of
Roman imperialism.11 One sentence in particular is relevant
here : " there can be no peace without arms, no arms without
1 Pliny, N.H., 7.149 : iwicta deinde tot mala : inopia stipendii, rebettio Illyrici,
servitiorum dilectus, iuventutis penuria, etc.
2 Tacitus, Annals, 1.16-49, etc. Pay was only one of several grievances.
3 G. R. Watson, The Roman Soldier (London, 1969), p. 98.
4 By Domitian in 83, Dio 67.3.5 ; Suetonius, Domitian, 7.3.
5 Tacitus, Dialogus, 17.3 ; see now K. Wellesley, The Long Year A.D. 69
(London, 1975).
6 Cf. e.g. Tacitus, Hist., 1.25.1 (suscepere duo manipulares imperil poptdi
Romani transferendtan et transtulerunt), 38.3 (nullo tribtmorum centurionumve
adhortante, sibi quisque dux et instigator), 46.1 (omnia deinde arbitrio milihm
acta : praetorii praefectos sibi ipsi legere), etc.
7 Cf. E. Ritterling, " Legio ", RE, xii (1924), 1266, etc.
8 Suetonius, Divus Vespasianus, 8.1-3.
9 As Tacitus puts it, Hist., 1.4.2: evolgato imperii arcano, posse prindpem
alibi quean Romae fieri. Perhaps this had long been an open secret, but now the
vulgus, and all the soldiers, had had its implications made plain.
10 E.g. the Roxolani (Hist., 1.79), Brigantes (ib. 3.45), Germans and Dacians
(ib. 3.46) and others (ib. 3.47-8), Garamantes (ib. 4.50).
11 Hist., 4.73-4.
THE THIRD CENTURY CRISIS 265
pay and no pay without taxation " (neque quies gentium sine
armis, neque arma sine stipendiis neque stipendia sine tributis
haberi queunt), Cerialis told his audience of Gallic rebels.1
Vespasian increased the tributa, doubling them in some cases2 ;
but the imperial finances needed careful nursing and legionary
pay was not increased for over a decade, and then only by one
third, from 225 to 300 denarii a year.3 If one compares this
with the private income of a senator, derived from rents,4 or with
the salary of a proconsul of Africa,5 the figure may appear
derisory. But ex-legionaries were distinctly among the better
off members of society and had other privileges ; nor did they
have to rely solely on their pay and savings.6 There were the
bounties on discharge and from time to time donatives and
booty when victories were won hence wars of conquest were
popular with the troops.7
But the time came when expansion stopped. Augustus
himself had apparently decided shortly before his death that
extension of the empire had gone on long enough.8 On the

1 Hist., 4.74.1. Cf. Dio, 52.6.1 and 52.28 f. on new ideas for financing the
army discussed in detail by E. Gabba, Studi Fanfani (p. 261 n. 3 above), pp. 39 ff.
2 Suetonius D. Vesp., 16.1. Note that Vespasian's generosity with the
citizenship (on which see A. N. Sherwin-White, The Roman Citizenship (2nd edn.,
Oxford, 1973), pp. 251 ff., 360 ff.) had financial implications, of which the
emperor cannot have been unaware : a steady increase in the numbers of citizens,
e.g. in Spain from the curial class by the grant of Latin rights (Pliny, NH.,
3.30), would have benefited the vicesima and hence the aerarium militare (p. 263
and n. 2 above). 3 P. 264 and n. 4 above.
4 R. P. Duncan-Jones, The Economy of the Roman Empire. Quantitative
Studies (Cambridge, 1974), pp. 17-32.
5 Dio, 78.22.5 (A.D. 217): 250,000 denarii. The proconsul would have
enjoyed this salary for one year only, of course.
6 Cf. R. Macmullen, Soldier and Civilian in the Later Roman Empire
(Cambridge, Mass., 1967), esp. pp. 99 ff.; G. R. Watson, The Roman Soldier
(1969), pp. 89 ff., 147ff. ; P. Garnsey, Social Status and Legal Privilege in the
Roman Empire (Oxford, 1970), pp. 245 ff.
7 Cf. G. Walser, " Der Putsch des Saturninus gegen Domitian ", Provincialia.
Festschrift fur Rudolf Laur-Belart (Basel, 1968), pp 497-507, esp. 498 f., 506 n. 53.
8 Note especially Tacitus, Annals, 1.11.7: addideratque consilium coercendi
intra terminos imperil. Cf. C. M. Wells, The German Policy of Augustus : an
examination of the archaeological evidence (Oxford, 1972) ; A. R. Birley, " Roman
frontiers and Roman frontier policy ", Trans. Archit. & Arch. Soc. Durham &
Northumberland, iii (1974), 13-15.
266 THE JOHN RYLANDS UNIVERSITY LIBRARY
accession of Hadrian this policy was taken to its logical con-
clusion with the erection of permanent frontier barriers, of which
the one between the Solway Firth and Wallsend is only the best
known example.1 What is more, the disposition of the armed
forces in the frontier regions now became permanent.2 During
the period from Augustus to Trajan legions had been moved
from province to province and from fortress to fortress as the
situation demanded, with bewildering rapidity on occasion. 3
This applied particularly to the Rhine and Danube armies, but
even in Britain there were considerable changes in the legionary
garrison in the years 43-1 22.* Thereafter the legions 11 Augusta,
XX Valeria Victrix, and VI Victrix remained in their bases at
Caerleon, Chester and York. Of the legions which survived
Hadrian's reign into the mid-third century only V Macedonica
was transferred to a new base, moving from Lower Moesia to
Dacia in the 160s.5 Legions were moved from the northern to
the eastern front in the 160s and up and down the northern
rivers during the 170s ; but they returned to base when the
wars were over. In the 190s and the third century only detach-
ments were drawn off for campaigns, leaving considerable
elements of the legion at the fortress. 6 Not surprisingly the
1 HA Hadrianus, 11.2 : Ergo conversis regio more militibus Brittaniam petit,
in qua multa correxit murumque per octoginta milia passuum primus duxit, qui
barbaros Romanosque divideret ; 12.6 : per ea tempora et alias frequenter in plurimis
locis, in quibus barbari non fluminibus sed limitibus dividuntur, stipitibus magnis in
modum muralis saepis funditus iactis atque conexis barbaros separavit. Cf. now
generally, J. C. Mann, " The frontiers of the principate ", ANRW 2.1, 508-33.
A. Mocsy, Pannonia and U. Moesia (1974), p. 109, attributes the Hadrianic
frontier policy in the Middle Danube region to Trajan. There may be much
truth in this; but I am convinced that the symbolic significance of Hadrian's
continuous frontier barriers was very marked and represented a major shift of
emphasis; cf. my comments in Trans. Archil, & Arch. Soc. Durham &
Northumberland, iii (1974), 15-17.
2 E. Ritterling, RE, xii (1924), 1293.
3 Cf. A. Mocsy, Pannonia and U. Moesia (1974), p. 85.
4 See now W. Eck, " Zum Ende der legio IX Hispana ", Chiron, ii (1972),
459-62, with new evidence to support E. Birley's arguments, " The fate of the
Ninth legion ", in R. M. Butler (ed.), Soldier and Civilian in Roman Yorkshire
(Leicester, 1971), pp. 71-80, that this legion survived, after transfer from Britain,
until the mid-second century.
5 E. Ritterling, RE, xii (1924), 1293, 1301 ; (1925), 1518-19.
6 E. Ritterling, RE, xii (1924), 1297 ff.
THE THIRD CENTURY CRISIS 267
foundation of veteran colonies also came to an end with Hadrian's
accession.1 The time-expired legionaries preferred to settle in
the communities which grew up close to their base, rather than
be sent elsewhere ; and by the same token the new recruits seem
largely to have been found in the frontier districts.2
For the prosperous classes in the heartland of the empire
this was an admirable situation. As far as most of them were
concerned, the Roman army had become invisible. The defences
of the empire were " many months and years journey " from
the city of Rome, said Aelius Aristides, with some exaggeration. 3
Hadrian's policy may not have endeared him to some sections
of the senatorial order, but the abandonment of wars of conquest
was clearly welcomed and approved in the Greek east. 4
The army was unquestionably the largest item in the imperial
budget,5 however primitive the financial planning may have
been. The soldiers were the only large body of wage-earners,
some 400,000 or more in a population of about fifty million. 6
Attempts have been made to calculate the total annual expenditure
on the army, without any great hope of accuracy since the pay
of some officers is unknown and a varying number of legionaries
were paid above the basic rates.7 But in the period from 9-40,
1 E. Kornemann, " Coloniae ", RE, iv (1900), 511-88, col. 566.; A. Mocsy,
Pcamonia and U. Moesia (1974), pp. 117 ff.; G. R. Watson, The Roman Soldier
(1969), p. 148.
2 This process had already begun, and was a gradual one : G. Forni, //
reclutamento delle legioni da Augusto a Diocleziano (Milan, 1953) ; id., " Estrazione
etnica e sociale delle legioni nei primi tre secoli dell* impero ", ANRW, 2.1,
339-91, esp. 362 ff. 3 Aelius Aristides, or. Rom., 80 Keil.
4 Cf. Birley, Trans. Archit. & Arch. Soc. Durham & Northumberland, iii. 15 ff.
5 See M. H. Crawford, ANRW, 2.2, 561, with n. 4.
6 For the number of auxiliary units in the mid-second century see the figures
of E. Birley, " Septimius Severus and the Roman army ", Epigraphische Studien,
viii (1969), 63-82, p. 72: 270 quingenary cohorts, 39 milliary cohorts, 90
quingenary and 9 milliary alae, a total of 218,000 men (if the quingenary cohorts
are estimated at 500 and the milliary at 1,000 men) ; together with 28 legions
(168,000 men, allowing 6,000 to each) and the urban units (11,000) a total of
397,000 is produced excluding the fleets and the ethnic units (the so-called
numeri), for which see now M. Speidel, " The rise of ethnic units in the Roman
imperial army ". ANRW, 2.3, 202-31.
7 R. Macmullen, Roman Social Relations (New Haven, 1974), p. 94, following
B. Dobson, " Legionary centurion or equestrian officer? A comparison of pay
and prospects ", Ancient Society, iii (1972), 193-207, attempts a calculation.
268 THE JOHN RYLANDS UNIVERSITY LIBRARY
when there were twenty-five of them,1 the legions must have
cost at least 50 million denarii a year. Thereafter the number
of legions increased, to reach thirty by the end of the first century
and thirty-three by the end of the second, while the basic pay
had gone up from 225 to 300 denarii a year c. 83 and from 300
to 450 c. 193.2 By that time the legions must have been costing
at least 110 million denarii a year. These figures of course
exclude the cost of the auxilia, who were paid at a much lower
rate, but for whom the total cost must have been considerable.3
The military treasury established in the year 6 was principally
funded by a tax to which Roman citizens alone were liable. But
our information suggests that the aerarium militare only provided
bounties for the citizen troops, not their regular pay. 4 Pay,
stipendia, depended on tributa. The implication is important.
The bulk of the legions and the auxiliary forces, stationed for
the most part in close proximity to them, were in the remote
frontier areas. Of the late Antonine establishment of thirty
legions, one each was stationed in Spain, Africa and Egypt;
there were three in Palestine and Arabia ; three in Syria; two
in Cappadocia ; and the remaining nineteen were all in the
north three in Britain, four on the Rhine and twelve in the
Danubian lands. 5 In military terms this no doubt made sense.
It was vital, as Augustus recognized, that the empire should
control the Danube and its western tributaries the Drau and
the Save; and Trajan extended Roman control across the
Lower Danube into Transylvania. Only by guaranteeing the
land link between northern Italy and the Balkans could Rome be
sure of maintaining her authority over the east and Asia Minor had
been since the late second century B.C. a key source of revenue.6
1 E. Ritterling, RE, xii (1924), 1244 ff.
2 G. R. Watson, The Roman Soldier (1969), p. 91.
3 M. Speidel, "The pay of the auxilia", JRS, Ixiii (1973), 141-7, argues
that their pay was higher than generally reckoned, e.g. by G. R. Watson, The
Roman Soldier (1969), pp. 99-101. 4 See p. 263 above.
5 Note that the reign of Domitian saw the beginning of a marked change in
the distribution of legions, with the main weight moving from Rhine to Danube,
a process completed by the time of Hadrian : E. Ritterling, RE, xii (1924), 1275 f.
6 The Rhine and Danube were of course vital communications arteries, and
it was not least for this reason that the legions were stationed along them rather
than in the hinterland.
THE THIRD CENTURY CRISIS 269
Hence some two thirds of the Roman army was stationed
in regions that were, by the standards of the Mediterranean
world, poor and unproductive.1 The British, German and
Danubian provinces can only have contributed a small pro-
portion, in tribute, to the cost of their military establishment.
Apart from pay and bounties, there were also the capital expen-
diture and cost of maintenance for the elaborate installations
that these forces required.2 To be sure, the army provided its
own labour and many of the materials were generally at hand.
But what was lacking had to be provided and paid for. In the
meantime, the soldiers and veterans and their families had money
to spend ; not surprisingly, there came into being in the frontier
districts an immense number of flourishing towns and villages :
Carlisle, Chester and York, Vienna, Budapest and Belgrade, are
some well-known examples. 3
Tranquillity reigned on most of the frontiers during the
long reigns of Hadrian and Antoninus Pius. Then serious
warfare began again. In the 160s there was an eastern war
and while it was being fought the Marcomannic war, or war of
many nations, 4 was brewing up on the northern frontiers. The
1 Cf. Dio, 49.36.2, Strabo, 7.5.4, C315, on Pannonia.
2 Cf. A. Mocsy, Pannonia and U. Moesia (1974), p. 109.
3 Aelius Aristides, or. Rom., 81 Keil. Cf. E. Birley, " Hadrianic frontier
policy ", Camuntina, ed. E. Swoboda (= Romische Forschungen in Niederosterreich,
3, 1956), pp. 25-33. See now also G. D. B. Jones and N. F. Higham, " Frontier,
forts and farmers : Cumbrian aerial survey 1974-5 ", Archaeological Journal,
cxxxii (1975), for new light on the density of settlement south of the western
end of Hadrian's Wall. E. Gren, Kleinasien und Ostbalkan in der wirtschaftlichen
Entwicklung der romischen Kaiserzeit (Uppsala-Leipzig, 1941), esp. pp. 89 ff.
stressed the major importance of the army in the economic development of the
Danubian provinces. Cf. p. I47f. : "Das bedeutet, dass das Kapital in die
Grenzprovinzen stromte, ganz besonders an die Donaugrenze, wo die grosste
Verbande standen... Die unmittelbare Folge war ein vermehrter Wohlstand in
den Grenzprovinzen. Die Gebiete waren doch keineswegs die natiirliche
Produktionszentren des Reiches und ihre lage was ausserst ausgesetzt." R.
Macmullen, Soldier and Civilian in the Later Roman Empire (1967), pp. 89 f.
criticizes Gren on the grounds that he fails to show a connection between the
presence of the soldiers and the prosperity, but concedes that " much of Gren's
main thesis, despite its drawbacks, is convincing ".
4 HA, M.Ant., 12.13 : dttm Parthictan helium genitur, natum est Marcoman-
nicum ; 22. 1 : gentes omnes ab Illyrici limite usque in Galliam conspiraverant ;
22.7 : bello Germanico sive Marcomannico immo plurimarum gentium.
270 THE JOHN RYLANDS UNIVERSITY LIBRARY
empire was invaded and throughout the 170s Rome struggled
to evict the Germans and their allies.1 Marcus Aurelius, ill-
fitted to cope with such a situation, was obliged to spend years
at the front. But, as Cassius Dio appreciatively records,
" although he was compelled to make extraordinary expenditures,
he burdened no one by exactions of money ".2 When the first
wave of invasion was driven back, " although a mighty conflict
and a brilliant victory had taken place, nonetheless when the
emperor was asked by the soldiers for a donative he did not
give them one : ' Whatever they might get over and above their
regular pay would be squeezed from the blood of their parents
and kinsmen ' ". 3 It must be noted that although troops from
other sectors of the empire had been sent in4 and two new
legions had been raised in Italy,5 the majority of the men, the
Danubian legionaries, could hardly have been expected to have
kith and kin in the wealthy provinces that provided most of
Rome's revenue. Later in the reign when further funds were
desperately needed, Marcus Aurelius auctioned imperial treasures,
rather than resort to new taxes.6
The reign of his son Commodus, who made peace with the
Germans,7 provided the empire with a breathing space of a
kind, although the armies were in a dangerously restless mood
and the self-confidence and unity of the governing class was to
be shattered.8 The senatorial order experienced a reign of
terror that revived memories of the Julio-Claudian and
X A. R. Birley, Marcus Aurelius (London, 1966), pp. 201 ff.; A. Mocsy,
Pannonia and U. Moesia (1974), pp. 183 ff.
2 Dio, 71.32.3. 3 Dio, 71.3.3.
4 E.g. part of XII Fulminata may have been on the Danube in 173 (Dio,
71.8.10 (Xiphilinus)), cf. A. R. Birley, Marcus Aurelius (1966), pp. 237 ff.; part
of III Augusta was there in the 170s : Dessau, ILS, 2747, cf. H.-G. Pflaum, Les
carrieresprocuratoriennesequestres sousle Haut Empire remain (Paris, 1961), no. 198.
5 II and III Italicae : cf. J. C. Mann, " The raising of new legions during
the principate ", Hermes, xci (1963), 483-9.
6 Zonaras 12.1 = Exc. Salm., 117; HA M.Ant., 17.4-6 ; Eutropius, 8.13.
7 See now G. Alfoldy, " Der Friedenschluss des Kaisers Commodus mit den
Germanen ", Historia, xx (1971), 84^109.
8 See generally F. Grosso, La lotta politico al tempo di Commodo (Turin,
1964); A. R. Birley, Septimius Severus (1971), pp. 131 ff. DJo regarded the
change from Marcus to Commodus as the end of an age of gold and the beginning
of an age of iron and rusty iron at that (71.36.4).
THE THIRD CENTURY CRISIS 271
Domitianic periods.1 Commodus was assassinated and replaced
by Pertmax, one of the heroes of the Danubian wars, a self-
made man of humble origin who was no doubt thought best
fitted to combine respect for the senate's prerogative with a
firm hand in military matters.2 He soon showed his intentions
by suspending payments to the northern tribes, as Dio approvingly
records.3 This is revealing : here was another drain on the
empire's resources, in the same direction, to the north. Com-
plete figures are lacking, but there were clearly lengthy periods
when barbarian peoples beyond Rome's frontiers received
annual grants. Whether these are viewed as Danegeld or as
subsidies to underdeveloped nations, they must have been
expensive.4 Pertinax found a nearly bankrupt treasury ; but his
tighthandedness and firm discipline caused his downfall within
1 For the senate Commodus was hostis deorum carnifex senatus, hostis deorum
parricida senatus : hostis deorum hostis senatus (HA Conan. Ant., 18.4), to quote
the celebrated adclamatio after his murder.
2 A. R. Birley, " The coups d'etat of the year 193 ", Banner Jahrbilcher, Ixix
(1969), 247-80; id., Septimius Severus (1971), pp. 106 fi.; G. Alfoldy, "P.
Helvius Pertinax und M. Valerius Maximianus ", Situla, xiv/xv (1974), 199-215.
3 Dio, 73.6.1.
4 Herodian, 1.6.8-9 refers to Commodus' payments. The practice was
long-standing. Note Tacitus, Germania, 42 : (on the eastern Germans) raw
amis nostris, saepius pecunia iuvantur. The following list does not claim to be
comprehensive. Domitian : Dio 67.6.5, 67.7 A (Dacians); 67.10.5 (Cherusci).
Trajan disapproved of the payments to the Dacians (68.6.1), but must have made
regular payments to the Sarmatian Roxolani, which Hadrian had to continue,
after some trouble 117 (HA Had., 7.8 : cum rege Roxalanorum, qui de imminutis
stipendiis querebatur, cognito negotio pacem conposuit). Marcus Aurehus made
payments to some of the northern peoples : Dio, 71.11.1, 19.1. For Commodus,
see above. Under Severus, the British governor Virius Lupus had to buy off
the Maeatae : Dio, 74.5.4. Caracalla paid large sums to the Cenni: Dio,
77.14.2, and Elbe Germans, ib. 3 and largesses to " barbarians " are mentioned
78.3.5 ; while Macrinus claimed that the amount paid to the barbarians was
equal to the soldiers' pay : 78.17.3. An annual payment to Armenia is mentioned :
78.27.4. Macrinus himself paid an indemnity to Parthia of 50 million denarii:
78.27.1. Severus Alexander: Herodian 6.7.9. Philostratus, a contemporary
of Severus and Caracalla, tells the story of Apollonius' meeting with the Indian
king who preferred peace to war and paid the barbarians to prevent them invading :
V. Apoli, 2.26. See for comment J. Klose, Roms Klientel-Randstaaten am
Rhein und an der Donau. Beitrdge zu ihrer Geschichte und rechtlichen Stellung
im 1. und 2. Jahrhundert n. Chr. (Breslau, 1934), who distinguishes between
payments made to allow individual peoples to maintain their independence vis-a-vis
their neighbours and payments made to buy peace for Rome ; he sees a change
18
272 THE JOHN RYLANDS UNIVERSITY LIBRARY
three months. There followed the notorious auction of the
empire in the barracks of the guard.1 The higher bidder,
Didius Julianus, was unacceptable to the legions and civil war
began. Septimius Severus, who won, was the candidate of the
Danubian armies.2 The armed forces as a whole were rewarded
handsomely: legionary pay was increased, probably by 50
per cent.,3 and various privileges were granted to all soldiers4 ;
but the Danubian legionaries were given something extra. The
guard that had overthrown Pertinax was punished with dismissal
by his avenger and replaced by a force twice the size and no
longer recruited mainly from Italy. The new guard was drawn
principally from the Danubian lands, being manned by Pan-
nonians, Dalmatians, and Thracians. 5 Under the principate the
unity of the widely dispersed army had been maintained not
least by seconding praetorian guardsmen to tours of duty in
the legions : after a spell as legionary centurion they would
return to Rome to be centurions in the guard and the other
Rome units ; then go as chief centurions to a legion ; then, back
at Rome, they became tribunes in the Rome garrison. The
change of manpower in the guard after the year 193 must have
had an effect throughout the imperial army.6
As Severus lay dying at York he advised his sons to be
generous to the soldiers and ignore everyone else.7 Caracalla
needed no second bidding. " He loved spending on the
soldiers ", as Dio bitterly observed, " and he had them around
him in great numbers, alleging excuse after excuse and war
after war ; he made it his task to strip, loot, and wear down the
to the second category from 180. There is doubtless much truth in this comment.
But the financial implications are unaffected by the motives for the payments.
1 Gibbon's account of this transaction in chapter five of the Decline and Fall
of the Roman Empire is justly celebrated.
2 A. R. Birley, Septimius Severus (1971), pp. 155 ff.
3 G. R. Watson, The Roman Soldier (1969), p. 91.
4 E. Birley, Epigr. Stud., viii (1969), 63 f.
5 M. Durry, Les cohortes prdtoriennes (Paris, 1938), pp. 81-87, 247ff.; E.
Birley, Roman Britain and the Roman Army (Kendal, 1953), pp. 118 ff.
6 B. Dobson and D. J. Breeze, " The Rome cohorts and the legionary
centurionate ", Epigr. Stud., viii (1969), 100 ff.; B. Dobson, " The significance
of the centurion and primipilaris in the Roman army and administration",
ANRW, 2. ] , 392-434. 7 Dio, 76.15.2.
THE THIRD CENTURY CRISIS 273
rest of the human race, not least the senators *V Army pay
was substantially increased again, and no doubt to fund the by
now enormously increased praemia militiae, the 5 per cent,
death duties were doubled. At the same time, to widen the
scope of the tax, all free inhabitants of the empire were made
Roman citizens and exemptions were abolished.2 " The emperor
himself kept on spending money on the soldiers (and on wild
beasts and horses) ", Dio goes on. " In fact, he often used to
say : * No one except me ought to have money, so that I can
hand it out to the soldiers *. His mother Julia reproached him
for spending vast sums on them, and said that ' There is no
revenue left for us any more, whether legitimate or illegitimate '.
* Cheer up, mother ', he replied, pointing to his sword, * so long
as we have this, the money will not run out' ".3
But the money did run out. Caracalla was murdered
within a few years and his successor tried in vain to stem the
tide.4 Then, in turn, two Syrian youths, the sons of Caracalla's
cousins, were found to revive the Severan dynasty. But they
and their backers had an uphill task. The great lawyer Ulpian
was murdered by the praetorian guards not long after taking
office as prefect. 5 A few years later Cassius Dio avoided a
similar fate by keeping away from Rome for most of his second
consulship : the guardsmen were after his blood, for he had tried
to impose discipline on their friends and kinsmen in Pannonia.6
In some ways it did not matter very much who the emperor
was.7 But when there was fighting to be done, the armies

1 Dio, 77.9.1.
2 Dio, 77.9.4-5. There is an enormous literature on the Edict of Caracalla.
On the issue relevant here, the financial aspects, note J. F. Gilliam, " The
minimum subject to the vicesima hereditatium ", American Journal of Philology,
Ixxii (1952), 397-405, showing that quite modest estates were liable to the duty.
3 Dio, 77.10.4.
4 Dio, 78.12.2-7 (fragmentary), 17.3, 18.5, 28.1-3, 36 (but note 27.1, 34.2-3,
for large payments by Macrmus).
5 Dio, 80.2.2-4 ; the date of Ulpian's death was 223 or 224, see J. Modrzejewski
and T. Zawadski, " La date de la mort d'Ulpien et la prefecture du pretoire au
debut du regne d'Alexandre Severe ", Revue historique de droit francais et etranger,
xlv (1967), 565-611. 6 DJo, 80.4.1-5.2.
7 R. Syme, Classical Review, liii (1939), 79 commented that " the significance
of the individual ruler declines steadily " during the Antonine period.
274 THE JOHN RYLANDS UNIVERSITY LIBRARY
demanded convincing leadership. The Syrian emperors were
conspicuously unfitted to defend the empire. It is surely in
no way surprising that Severus Alexander was replaced by a man
of Danubian origins who had risen from the ranks, C. Julius
Verus Maximinus.1 The author of the Historic. Augusta, after
letting himself go on the extravagances of the pampered
Elagabalus and expatiating longwindedly on the virtuous conduct
of Alexander, was ready for a contrast when he got to Maximinus.
The man is portrayed as a barbarian monster.2 The truth was
less startling or discreditable.3 The change in the social origins
of the emperors, stressed by Victor, reflected a general process :
the gradual turning away of the senatorial order from its old
role in the command of Roman legions. The Marcomannic
wars had taken a heavy toll; many new men had been taken
into the senate. 4 But before these new elements had had time
to be absorbed into the governing class, the reign of Commodus
and the civil wars of 193-7 had their effects. As a result
senators were doubtless less eager to canvass for office : the
risks of incurring suspicion, or indeed, the risk of being forced
to become the figurehead of a rebellion, were real ones.5 At the
same time, the emperors, after what had happened in 193-7,
were reluctant to put senators in command of large armies.
The subdivision of provinces was one solution, which probably
made high command seem less attractive in any case.6 Whereas
Julius Agricola, for example, had commanded four legions in
1 See above all R. Syme, Emperors and Biography (1971), pp. 179 ff.
2 HA Maximini duo 1.5 (barbaro etiam patre et matre genitus), 2.2 (/eras
moribus), 9.2 (neque enimfuit crudelius animal in fern's), 10.1, 11.6 (ferarum more),
\ 5.6 (tristissima belua), 15.8 (illam beluam), 17.1 (homo natura /eras sic exarsit,
ut non hominem sed beluam putares), etc.
3 R. Syme, Emperors and Biography (1971), esp. pp. 185f., pointing out
that only one Latin source (epitome de Caes., 25A) calls him Thrax, a label
which it would be preferable to abandon.
4 HA M.Ant., 22.7 : et multi nobiles bello Germanico. .. interierunt; 10.3 :
multos ex amicis in senatum adlegit cum aediliciis out praetoriis dignitatibus.
5 Cf. Dio, 72.9.2a, HA Comm., 8.4, Pert., 3.6 (it was bad enough in Rome:
ib. 6.4-5).
6 E. Ritterling, RE, xii (1924), 1309 : " fortan [from Severus and Caracalla]
hat keine Provinz mehr als zwei Legionen standige Besatzung " ; besides this,
the tenure of office was generally short, cf. G. Barbieri, L'Albo senatorio da
Settimio Seve.ro a Carino (Rome, 1952), pp. 554 f.
THE THIRD CENTURY CRISIS 275
Britain in the Flavian period, in the third century there were
two British governors, who shared three legions and the man
who commanded the northern province only had one of these.1
The senators were ever more frequently replaced by equestrian
officials ; and these men themselves, unlike the equites who had
served the emperors in the first two centuries, were increasingly
recruited from the military men who had risen from the ranks.2
Tacitus, writing in the reign of Trajan, describes a mutiny
of the praetorian guard soon after Otho's accession in A.D. 69.
The emperor, conscious that the men were capable of anything
and that the senatorial class was rigid with fear, is portrayed
making a subtle speech, soothing and flattering the guardsmen,
but warning them of the dangers of indiscipline and of con-
tinuing their hostility towards the senate : for, " just as you
can become senators, so senators can become emperors ". 3 By
the year 238 that intermediate stage was no longer necessary.
Guardsmen could become emperors. The interests of the most
powerful body of legionaries and of the guard had been made
one by Septimius Severus. Whatever the details of Mamiminus'
career, he was a true representative of the Danubian soldiery. 4
To be sure, one third-century emperor, Decius, was a man from
the Danube lands who was a senator. 5 A handful of men from
this region had been taken into the amplissimus ordo* But only
a handful were likely to be found, since the economic conditions
favourable to the creation of large estates, and hence to the
emergence of men equipped by their wealth to enter the senate,
did not prevail in the Danubian and Balkan lands.7 But it
1 A. R. Birley, " The Roman governors of Britain ", Epigraphische Studien*
iv (1967), 63-102, p. 78 f.
2 Cf. H.-G. Pflaum, Les carrieres procuratoriennes (1961), no. 317.
3 Histories, 1.84.4 : nam ut ex vobis senatores, ita ex senatoribus principes
nascuntur. The distorted version of this sentiment in the exaggeratedly pro-
senatorial Historia Augusta is worth noting, Clod. Alb., 13.10 : senatus imperet,
provindas dividat, senatus nos consules facial, et quid dico senatus'? vos ipsi et patres
vestri: eritis enim ipsi senatores (speech of Albinus to the British legions).
4 R. Syme, Emperors and Biography (1971), pp. 186 ff.
5 Ibid. pp. 195 ff.; X. Loriot, ANRW, 2.2, 697 f.
6 Notably M. Valerius Maximianus : AE, 1956, 124, Diana Veteranorum.
7 At least until the third century ; A. Mocsy, Pannonia and U. Moesia (1974),
pp. 169 ff., 242 ff., and not in the frontier area (ib. 239).
276 THE JOHN RYLANDS UNIVERSITY LIBRARY
must be reiterated it was now too late in any case. The
community of interest of the senatorial order had been dis-
solved, likewise the carefully nurtured unity of spirit between
emperor and senate. From the beginning of the third century
it had become clear that the road to power and influence was no
longer to be sought by climbing the ladder of senatorial office.
The so-called " Edict" of Galhenus, excluding senators from
military service, was not a revolutionary step.1
Maximinus' entire reign was spent in the north : one season
on the Rhine, then in the Danubian lands, with unremitting
campaigning against Rome's old northern enemies.2 To raise
funds, he resorted to an increasingly tough fiscal policy, which
affected all classes and all regions.3 Even in Thrace, a cele-
brated inscription shows that some villagers had petitioned
the emperor when all other appeals had failed : they were being
ruined by the exactions of the soldiery.4 Significantly, the
people of Scaptopara were able to pass their libellus safely into
the hands of the emperor's a libellis by the agency of a native of
the place, who happened to be a praetorian guardsman.5 Else-
where such useful channels of influence would be lacking. In
Africa the behaviour of a procurator provoked an uprising and
the elevation of a rival emperor, the aged proconsul Gordianus.6
Gordian and his son were speedily swept aside by the governor
of Numidia and proconsular Africa was ravaged by the victorious
legion, III Augusta. But after a series of dramatic events in
Rome and northern Italy Maximinus was overthrown ; and by
midsummer old Gordian's grandson was installed as emperor.7
The reign of Maximinus had been a portent: the real possessors
of power had put one of themselves at their head. By a colossal
effort of will and with a considerable element of good fortune,
an alliance of the lands of ancient culture and prosperity had
1 B. Malcus, " Notes sur la revolution du systeme administratif romain au
Hie siecle ", Opuscula Romana, vii (1969), 213-37.
2 X. Loriot, ANRW, 2.2, 674 ff.
3 Herodian, 7.3.1 ff.; X. Loriot, ANRW. 2.2, 681 ff.
4 IGBulg.. 4. 2236.
5 Ibid.
6 Herodian, 7.4. Iff.
7 X. Loriot, ANRW, 2.2, 688 ff., esp. 720 ff. (for the chronology).
THE THIRD CENTURY CRISIS 277
removed him. For another eleven years men from the Greek
east reigned1 and their principal military activities were in the
east. The advent of Decius in 249 marked the beginning of a
long period during which the Danubian forces struggled to
reimpose their nominee on the throne.2 During the 250s and
260s Italians reigned again, but against repeated attempts to
overthrow them.3 It may not be fanciful to suggest that one
reason why Gallienus was able to retain support for as long as
he did from the Danubian and Balkan lands was the origin of
his wife, Cornelia Salinina Chrysogone, a native of Salonae in
Dalmatia.4 However this may be, Gallienus did succumb :
from 268 onwards the Danubian soldiery were able to control
the empire for over a century.6
Naturally I do not believe that a process such as may be
observed in the third century can ever be explained in terms of
a single cause. Nonetheless, I would argue that the economic
imbalance created by the frontier policy of which Hadrian's
Wall is the most potent symbol, was the major factor in bringing

1 The origin of Philip is undisputed : he was a native of the Roman province


of Arabia (PIR2 J 461). There seems to be general acceptance of the view that
the Gordians originated in Asia Minor, put forward by A. R. Birley, " The
origins of Gordian I ", in M. G. Jarrett and B. Dobson (eds.), Britain and Rome.
Essays presented to Eric Birley (Kendal, 1966), pp. 56-60, cf. the bibliography in
X. Loriot, ANRW, 2.2, 694 fi. However, J. Gage, " Programme d' ' italicite '
et de nostalgic d'hellenisme autour de Gallien et Salonine. Quelques problemes
de la ' paideia ' imp6riale au I He siecle ", ANRW, 2.2, 828-852, is sceptical:
"nous ne croyons pas qu'Jl soil tout a fait raisonnable de ramener ces grands
patrons des bourgeoisies italiennes et africaines a une origine cappadoce " (p.
832 : but no reference is given to any discussion of the question). The attempt
by K. D. Grasby, " The age, ancestry, and career of Gordian I ", Classical
Quarterly, xxv (1975), 123-30, to discard Herodian's evidence (7.5.2) for Gordian's
age is unconvincing not least because he is unaware of AE, 1971. 475 (Caesarea
Maritima) confirming that Gordian II had been consul: see now X. Loriot,
" Un milliaire de Gordien II decouvert pres de Cesaree de Palestine et 1'extension
aux provinces de 1'insurrection de 238 apres J.-C. ", Bull, de la Soc. Nat. des
Antiquaires de France, 1972, 90-93.
2 R. Syme, Emperors and Biography (1971), pp. 193 ff.
3 M. Christol, ANRW, 2.2. 804 ff.
4 See G. Alfoldy, " Senatoren in der romischen Provinz Dalmatien",
Epigraphische Studien, v (1968), 99-144, pp. 127-30, 144.
6 Cf. Mamertinus, Pan. Lot., 2 (10) 2.2 : quis enim dubitat quin multis iam
saeculis... Italia quidem sit gentium domina gloriae vetustate, sed Pannonia virtute?
278 THE JOHN RYLANDS UNIVERSITY LIBRARY
about that process of decline, and that many of the other causes
to which it has been attributed were generated by that frontier
policy.1
The creation of an empire made a standing army inevitable,
and the army had to be paid for. Political considerations and
military requirements caused Augustus to distribute the army
away from the Mediterranean heartland of the empire. The
upper classes realized only too clearly that their own survival
depended on the maintenance of firm discipline, and were
allowed to continue their traditional role as legionary and army
commanders, in the service of the emperors. Although the
mutinies of A.D. 14 and the civil war of A.D. 68-69 showed the
fragility of the system, it appeared to recover. But in the second
century the renunciation of further expansion led to the per-
manent stationing of the empire's forces along the frontiers, and
recruitment was largely from these same regions. As long as
peace was maintained, the burden of taxation required to pay
the troops may have been tolerable. But it is important to
recognize that this system resulted in a transfer of resources
1 This view was stated briefly in Trans. Durham & Northumberland, M.S. 3
(1974), 24f., and owes much to A. Mocsy, "Die Expansionsfrage im I. und
II. Jh. und die Ertragsfahigkeit der Grenzprovinzen ", Annales Univ. Sclent.
Budapestinensis, Sec. Hist. 5 (1963), 3-13, and to E. Gren, Kleinasien und
Ostbalkan (1941), (p. 269 n. 3 above). I have also benefited, through the author's
courtesy, from being able to consult the typescript of Professor Mocsy's paper
to the San Francisco Congress of History, 1975, " Der Limes und die Provinzen ".
My friend Prof. G6za Alfoldy reminds me that the urban upper classes were
already in economic difficulties before the wars of M. Aurelius, e.g. in Spain,
where stagnation or even a decline in the economy may be observed in the first
half of the second century. In my view this may have been unwittingly brought
about by the drain of revenue to the northern frontiers which had already begun.
It is astonishing that A. Deman, " Materiaux et reflexions pour servir a une
6tude du developpement et du sous-developpement dans les provinces de
1'empire romain ", ANRW 2.3, pp. 3-83, and J.-H. Michel, " L'insuffisance des
investissments : Signe ou cause de sous-developpement dans deux provinces
romaines (1'Espagne et 1'Afrique du Nord)', ibid. 84-94, fail to recognize, when
they contrast Spain and N. Africa with Gaul, that the latter's prosperity was
enormously enhanced by the spending power of the Rhine armies ; Spain and
N. Africa, with only one legion apiece, did not benefit from the only kind of
investment that took place (after the colonizing programme was completed),
yet even so F. Millar can justifiably remark (The Roman Empire and its Neighbours
(1967) p. 170 f.) that " The prosperity of Roman Africa was largely the product
of the work of the Roman army ".
THE THIRD CENTURY CRISIS 279
from the productive to the unproductive regions of the empire.1
The armies of Britain and the Rhine were, no doubt, financed
mainly by the wealthy Gallic provinces, which would benefit in
return, because of their favourable river-transport system, from
supplying the troops. Britain itself and the Rhine provinces,
in the meantime prospered but much of the prosperity remained
in the hands of veterans. On the other hand, the hinterland of
the Danubian armies was not rich enough to provide the revenue
needed to maintain the much larger military establishment there.
The richer provinces of the Mediterranean Bithyma, Asia,
Egypt and Africa must, effectively, have paid for the defence
of the northeastern sector of the frontier. When the costs
mounted, it was they that noticed the difference2 ; and the
spending power of the Danubian soldiers could not be pumped
back into the economy. Instead, it stayed in the Danube area,
attracting a considerable immigrant population,3 but largely
remaining in the hands of the soldiers themselves, the veterans,
and their families, who formed a solid and united group.4 The
wealth that was put into the frontier lands naturally made the
empire seem much more attractive to the barbarians across the
frontiers, and determined attempts were made by them to
obtain a share in this prosperity. The evidence shows that
barbarian tribes beyond the frontier were intent on securing
peaceful admission to the empire in the second century, or even
asking Rome to annexe their territory.5 For the most part they
had to be content with money payments.6 These, together
with the more settled way of life imposed on them by Roman
supervision, made Rome's barbarian neighbours more prosperous,

1 M. H. Crawford, ANRW, 2.2, 568; and cf. p. 269 n. 3 above.


2 See above on the events of the year 238. One may compare what happened
in the 80s soon after Domitian had increased legionary pay : " many of the
subject peoples revolted when taxes were forcibly collected, including the
Nasamones " [m Africa], Dio, 67.4.6 ; or ca. 372, when Valentinian's exactions
in Africa to pay for his northern campaigns provoked the rising of Firmus,
Zosimus, 4.16.1 ff.
3 A. Mocsy, Pannonia and U. Moesia (1974), esp. pp. 227 ff.
4 Ibid. pp. 242 f.
6 Appian, proem., 7.
6 P. 271 and n. 4 above.
280 THE JOHN RYLANDS UNIVERSITY LIBRARY
and thus, in their turn, a target for their wilder cousins beyond.
It was these more distant peoples, such as the Lombards and
Vandals, who created the disturbances of the 160s which led to
the Marcomannic wars.1
If the governing class had remained united, and the Roman
army had continued to be heterogeneous, divorced from the
native population among which it was garrisoned, and regularly
liable to be transferred to new bases, the situation that has been
described might have been avoided. Militarily, an army kept
further back from the frontiers might have been more effectively
deployed in emergencies, while the soldiers' pay should have
been spent in the communities whose taxes provided it, keeping
their economy bouyant and better able to respond to a crisis,
both economically and psychologically.2 As it was, the con-
centration of so high a proportion of the army in the Danube
area and the recruitment of the same men into the praetorian
guard from the early third century meant that within a generation
from the new guard's formation these men demanded power
themselves. From the 160s their own homeland had been
under continuous pressure. The Danubian provinces had been
badly afflicted in the 170s and they were determined to ensure
that it did not happen again. It was natural that these men
worked for the unity of the empire : they depended on the
Mediterranean world to finance them.3
In the mid-third century the much maligned Gallienus began
the creation of a new mobile field army.4 If militarily effective,
from the political and economic point of view it had come too
late. The men from the Danube were already dominating the
officer corps and Gallienus could hardly supplement them
much from elsewhere. Meanwhile, the depreciation of the
currency, made necessary by the constant increases in pay and
recurrent donatives, had resulted in the reversion to a non-
1 A. M6csy, Pannonia and U. Moesia (1974), pp. 185 ff.
2 It is difficult to see how F. Millar, The Roman Empire and its Neighbours
(1967), p. 181, can describe the general economic decline of the third century as
" still mysterious ". See pp. 259 n. 3, 260 n. 1, 271 n. 4 above.
3 A. Mocsy, Pannonia and U. Moesia (1974), pp. 210 f.
4 A. Alfoldi, CAM, 12, 208 ff. = Studien zur Geschichte der Weltkrise (1967),
pp. 394 ff.; cf. M. Christol, ANRW, 2.2, 822 n. 96.
THE THIRD CENTURY CRISIS 281
monetary economy.1 Hence the field army of Gallienus and
his successors was not contributing to the economy. The men
were no longer being paid in cash.
1 M. H. Crawford, ANRW, 2.2, 570 f. Note that the annona militaris,
often said to have been instituted by Septimius Severus, did not become regular
till later : see now G. Rickman, Roman Granaries and Store Buildings (Cambridge,
1971), pp. 278 ff.

You might also like