Drinkwater J.F. The Alamanni and Rome 213-496 (Caracalla to Clovis)
Подождите немного. Документ загружается.
was executed en route, near Pola. But this was not the end of
Constantius’ problems. Gallus’ death and the subsequent prosecu-
tion of his supporters intensiWed the atmosphere of fear and distrust
at Constantius’ court. In addition, Constantius now had to Wnd a
way of dealing with Gallus’ half-brother and his own cousin, Julian,
apart from himself the last surviving male in direct descent from
Constantius I. There remained growing tension with Persia. And
Constantius still had to resolve bitter diVerences between eastern
and western bishops which twisted doctrinal disputes, issues involv-
ing Church careers and secular politics, and hunger for imperial
patronage into a maddeningly complex knot.
The religious problem was tackled at the Council of Milan, held
probably in July and August of 355. This must have been long in
the planning, and Constantius kept a close eye on its proceedings.
However, it still proved diYcult, and did not get him the agreement he
had hoped for.193 As the council ended, the emperor suVered a further
reversal in fortune when Silvanus ‘usurped’ in Gaul and had to be
deceived and killed by one of Constantius’ agents, Ammianus’ patron,
the senior general Ursicinus.194 This left Roman military and civil
administration on the Rhine in tatters, and prompted further raids by
Franks and Alamanni.195 After much heart-searching, Constantius,
still bent on returning to the east, decided to make Julian his repre-
sentative in the west. On 6 November 355, he commended his cousin
to his troops at Milan, and with their approval named him Caesar. On
1 December, Julian left for Gaul. As he left Italy, Ammianus tells us, he
received the dreadful news that Cologne had been besieged and taken
by barbarians.196
Establishing the basic framework of Constantius’ 355 campaign is
diYcult. Though Ammianus implies a date fairly late in the year,
with his reference to the emperor’s returning to winter quarters,197 it
is generally agreed that Wghting must have taken place in late spring
or early summer because it is impossible to allocate it reasonably to
193 Barnes (1993: 117).
194 Kienast (1996: 322): Silvanus ‘reigned’ 11(?) August to 8(?) September. Cf.
above 152.
195 Below 213.
196 AM 15.8.1–18.
197 AM 15.4.12.
208 ConXict 285–355
any other period in the year.198 Constantius and Arbetio probably
travelled to Lake Constance over the Splu
¨
gen pass, establishing the
imperial headquarters near the transportation centre and important
naval base of Bregenz.199 It is possible that en route, at Como,
Constantius met Julian making his Wrst visit to the west.200 Proposed
alternative locations of Campi Canini further south are implaus-
ible.201 Arbetio’s advance is likely to have been to the north of
Bregenz, not along the southern shore of the lake which was
Roman territory.202
This location of the campaign raises some diYculties about the
relationship between the Lentienses and the early medieval ‘Linzgau’,
with which they are now commonly associated.203 The Linzgau is
usually taken to be the area just to the north of the western arms of
Lake Constance. The now vanished stream of the Linz, whose name
links the ‘Gau’ and the Lentienses and which survives in that of the
settlement of Linz, near Pfullendorf,204 ran c.60 km (36 statute miles)
as the crow Xies to the north-east of Bregenz.205
Ammianus gives a very negative impression of the aims and
achievements of the campaign. A preliminary protracted council of
war suggests indecision; division of the army implies lack of nerve;
ambush hints at incompetence; and subsequent vacillation at further
lack of leadership. He attributes Wnal success to the bravery of the
troops, not the skill of their generals. Most modern commentators
have taken their tone from him.206 Again, however, what he says has
to be handled carefully. Though the 355 expedition was no enormous
venture, it is a wonder that Constantius bothered with it at all when
he had so many other worries; and, as in 354, he seems to have been
198 Barnes (1993: 221): around June; Lorenz, S. (1997: 28): probably June; Roll-
inger (1998: 188–91): mid-May to mid-June.
199 Rollinger (1998: 180–4).
200 Rollinger (1998: 188).
201 Rollinger (1998: 168–79); cf. Lorenz, S. (1997: 175–6, 179, 183).
202 So Rollinger (1998: 179), contra Lorenz, S. (1997: 30–1), though, certainly, as
Lorenz says, we should not envisage Arbetio as crossing the Rhine at the western end
of the lake.
203 e.g. Rollinger (1998: 178 and n.65), from Staehelin (1948: 291).
204 Zotz (1998: 393).
205 On the Lentienses, see further below 310–15.
206 e.g. Demandt (1989: 84); Lorenz, S. (1997: 31).
ConXict 285–355 209
very pleased with what he achieved.207 It cannot be true that the
355 campaign was ‘probably of no major military signiWcance’, and
that it was of interest to Ammianus only because he happened to be
in Milan.208 The key, as ever, is context; and the context was still that
of 354. Ammianus’ wording suggests that the cause of the 355
expedition, raiding, was recent and perhaps even contemporary.209
However, like that under Gundomadus and Vadomarius, it could
have taken place earlier, occasioned by Constantius’ invitation to
make trouble for Magnentius. Indeed, like that around Kaiseraugst,
the hoard-pattern south of Lake Constance indicates raiding in 351/2
which then more or less ceased.210 Constantius therefore, in his
methodical manner, moved to deal with an important but not critical
problem. Again, he needed to show that he had his allies under
control and that they posed no threat to Italy. Ammianus’ use of
collimitia and the evidence of the coin-hoards suggest that the
Alamanni had not penetrated beyond Raetia. However, since the
aVected area lay in the diocese of Italia Annonaria at the end of a
Wrst-class route leading directly into the Po valley, Constantius must
show that the safety of the peninsula was now assured.211 Success
would also further enhance his military reputation.
In respect of Constantius’ tactics, Ammianus’ detail both illumin-
ates and obscures. Though Ammianus does not say so directly, and
indeed in one place appears to indicate the contrary, Arbetio was sent
ahead with a small but elite cavalry force.212 His orders were to engage
the enemy around Lake Constance. However, Ammianus’ phrasing,
though unclear, does not suggest a direct attack.213 Arbetio’s job was
therefore probably to locate Alamanni. As at Kaiseraugst, this was to
be followed by a show of strength,214 followed in turn by a
peace agreement. Despite Ammianus’ emphatic double mention of
207 AM 15.4.13: ovans.
208 Contra Rollinger (1998: 192), (2001: 150, quotation).
209 AM 15.4.1: collimitia saepe Romana latius irrumpentibus.
210 Wigg (1991: 103–7 Maps 10–14).
211 Cf. Rollinger (1998: 183).
212 Rollinger (1998: 172–4): note especially the alternative translation of validiore
[15.4.1] as ‘very strong’ not ‘the larger part of’.
213 AM 15.4.1: relegens margines lacus Brigantiae pergeret protinus barbaris.I
interpret relegens as ‘going through again’, ‘combing’.
214 Cf. Barcelo
´
(1981: 27); Rollinger (1998: 73).
210 ConXict 285–355
‘war’,215 all-out Wghting was not Constantius’ intention. That combat
occurred may be explained by poor reconnaissance on both sides:
Ammianus reports that Arbetio moved without waiting for news
from scouts.216 One can easily envisage each side running into the
other without warning, and both panicking and coming to blows: a
major skirmish but not a pitched battle.217 It is clear that the violence
was relatively short-lived, and that both sides reached an agreement.
This agreement is likely to have involved treaties.218 These are not
mentioned by Ammianus, but again we have to exercise care; and this
returns us to the question of why Ammianus described these events as
he did. Ammianus was hostile to Constantius because of the emperor’s
treatment of Ursicinus, Ammianus’ patron,219 and it was in 354/5,
following the fall of Gallus, that Ursicinus suVered his Wrst disgrace. He
cleared himself only by removing Silvanus, but then again fell under
suspicion.220 However, even deeper than Ammianus’ regard for
Ursicinus was his regard for Julian, and it is likely that he ignored
the Alamannic treaties of 355 not only because he wanted to bury
anything that would redound to the credit of Constantius but also
because, unlike those of 354, these were not going to Wgure in his
account of Julian’s encounters with Alamanni.221 He also used the
355 campaign to point up the immensity of the threat faced by Julian
in Gaul.
Ammianus includes a relatively long excursus on the upper reaches
of the Rhine and Lake Constance.222 This, however, is entirely
untrustworthy, being highly stereotyped and derivative.223 Alamanni
are depicted as wild people in a wild, ruined land; and wilderness and
ruin begin south of the frontier. Ammianus thus disregards centuries
of development in the original province of Raetia.224 Finds from
215 AM 15.4.1, 7: bellum, aspera ...bellorum.
216 AM 15.4.7.
217 So Rollinger (1998: 173).
218 Benedetti-Martig (1993: 356–7): these were the treaties still operational in 378
(AM 31.10.2); below 310.
219 Above 178.
220 Above 208.
221 Cf. above 155.
222 AM 15.4.2–6.
223 Rollinger (2001: esp. 136–8, 140, 144, 149–50).
224 Rollinger (2001: 132–3).
ConXict 285–355 211
the area to the north and north-east of Lake Constance do not
indicate intensive Alamannic settlement or particularly intimate
contact with the Empire; and it lies outside Stribrny’s regions of
coin circulation.225 On the other hand, the Lentienses, living close
to the Iller-limes, must have had some dealings with Rome.
A Lentiensian could become a member of Gratian’s guard and
could return home on leave; and having fought with Gratian, the
Lentienses were required to supply recruits for the army.226 Thoug h
Lentienses were never formed into full, permanent regiments, this
service was provided, probably under Gratian’s father, by Raetovarii,
who could not have been distant from them.227 In this region lies the
fort at Kellmu
¨
nz, with an unusual apsed structure that may have been
built to facilitate dealings with local Alamanni.228 All this suggests
decades of relatively peaceful interaction between Alamanni and
Romans. The leaders of the Lentienses and their allies would
hardly have recognized themselves in Ammianus’ description of
‘wild’ Alamanni.
There remains one issue concerning Roman military activity on
the Rhine in 355. Ammianus records that in 356, Julian, preparing to
move from Autun to Reims, learned that Silvanus had gone ‘not long
before’ (paulo ante) by a shorter route with 8,000 men.229 This has
been taken to indicate that, at about the same time as Constantius
was campaigning by Lake Constance, Silvanus was active near or on
the Rhine frontier. It has been suggested that both campaigns were
ordered by Constantius in a move that was a protot ype of
pincer attacks of 356 and 357.230 This is unlikely. The signs are that
Constantius’ 355 campaign was short and localized. The theatres
of operations were too distant to allow practical liaison. In early
355, Gallus had fallen and Julian looked well on the way to being
discredited. But Silvanus was trusted, and under his charge Gaul
appeared to be in safe hands.231 He was left to manage things on
225 Germanen (1986: 343 Fig. 69); Stribrny (1989: 401 Fig. 19).
226 Above 160.
227 Above 170–3.
228 Above 99, 157.
229 AM 16.2.4.
230 Jullian (1926: 166 and n.1); Drinkwater (1997: 5 and n.27); Lorenz, S. (1997: 35
and n.73).
231 AM 15.5.2: eYcax ad haec cor rigenda.
212 ConXict 285–355
his own. Ammianus suggests that he drove out the invaders.232
However, an earlier disparaging description of his activities by Julian
has it that he
plundered the cities of their wealth and distributed it among the invading
barbarians, paying it down as a sort of ransom, though he was well able to
take measures to win security by the sword rather than by money. But he
tried to win them over to friendliness by means of money.233
Troop movements, expulsions and subsidies suggest that Silvanus’
policy, like that of Constantius, was to solve ‘the Alamannic problem’
by a blend of aggression and conciliation, not full-scale war. This
should have included treaties with Alamannic communities on the
Upper Rhine and the lower Main.234 It could well have been that
some left-bank settlement by Chnodomarius and his neighbours was
tolerated at this time.235
The situation soon deteriorated.236 Trier, capital of the west since
the third century, had suVered badly during the revolt of Magnentius.
Possibly subject to Alamannic and Frankish attack, it had thrown up
its own leader, Poemenius, in opposition to Decentius, only to be
retaken by the latter before he himself succumbed to Constantius II.
It seems likely that the city had ceased to be a viable base before
Constantius II recovered Gaul.237 This is probably why Silvanus set
up his headquarters in Cologne.238 However, in the summer of 355,
Cologne was closely involved in his usurpation and its suppression
by Ursicinus. Silvanus’ death made Ursicinus de facto imperial v ice-
roy in Gaul, responsible for its defence.239 Thanks to Silvanus and
Constantius, Ursicinus did not have to deal immediately with any
sort of barbarian crisis. Some indication of the restoration of order
may be seen in the fact that he and his small retinue do not appear to
have taken raiding by Alamanni or Franks into account on their
232 AM 15.5.4: barbarosque propellans.
233 Julian, Orat. 2.98C–D.
234 See AM 17.1.3.
235 Cf. below 217–18, 228. Contra Lorenz, S. (1997: 35 n.73), from Gilles: that
Silvanus had to deal with attacks from Chnodomarius at this time.
236 The following linking narrative is a revised version of Drinkwater (1997: 6–8).
237 Drinkwater (1997: 5–6).
238 Above 203.
239 So Fre
´
zouls (1962: 674–5); cf. Matthews (1989: 38).
ConXict 285–355 213
journey to Cologne to meet Silvanus, and that after Silvanus’ fall
Ursicinus was able to arrest his accomplices and send them to Milan
for interrogation.240 Yet Ursicinus was in an unenviable position.
He was physically isolated, w ith very few trustworthy followers in
an area he did not know. He must have been constantly on the
defensive against both the followers of the man he had just betrayed
and an emperor who remained uneasy about his loyalty, who was
soon to complain about the misuse of public funds in Gaul, and who
was never, by promoting him from mag ister equitum to magister
peditum, to make him the full successor of Silvanus in Gaul (the
post went to Barbatio).241 Ursicinus left Cologne. He could not move
to Trier—by late 355 not only badly damaged but also compromised
by the support Poemenius had given Silvanus.242 Ursicinus therefore
transferred his headquarters and, I would argue, most of the forces he
commanded, including the main strength of the Rhine garrisons,
westwards, to winter in the vicinity of Reims. Here, Julian joined him
and his successor as magister, Marcellus, early the following year. 243
Ursicinus’ move is entirely pardonable; but it signalled a military
down-grading of the Rhineland and, worse still, was followed by
complete inaction: the diYculties of his position paralysed Ursicinus’
generalship.244 It was from this time at the latest that Alamanni began
to occupy land on the left bank of the Rhine; and it is arguable that
Cologne, now defended by no more than a token garrison, fell to
purely adventitious raiding by Franks. Contrary to Ammianus, these
240 AM 15.5.24, 6.1f. (cf. 14.5.8: Paul’s job was to take the supposed supporters of
Magnentius to the imperial court for interrogation).
241 On Ursicinus’ duplicity see Drinkwater (1994); for the suggestion of embezzle-
ment see AM 15.5.36 with Fre
´
zouls (1962: 676); and for Ursicinus’ failure to gain
promotion, see Fre
´
zouls (1962: 679–80). Fre
´
zouls comments usefully on Ursicinus’
diYculties in the wake of Silvanus’ disloyalty, but does not develop his arguments,
and makes too much of Silvanus’ ‘Germanness’.
242 AM 15.6.4.
243 AM 16.2.8. Below 219. For Ursicinus’ rank in relationship to Marcellus and
Julian, see Fre
´
zouls (1962: 680–1). Blockley (1980: 475) explains the paralysis noted
below as a result of the somewhat earlier arrival of Marcellus and the failure of
Ursicinus and Marcellus to cooperate in the defence of Gaul.
244 Cf. van Ossel (1992: 73): although in the end there was no total collapse, the
political and military turmoil which accompanied the rise and fall of Magnentius
appears to have led to the abandonment of a large number of rural sites in the
Rhineland.
214 ConXict 285–355
did not take the place by siege. Rather, living up to their name—the
‘bold’/‘impudent’ ones—they just forced their way into a virtually
open city, no doubt astonished by their success.245 By late 355, the
situation in Gaul was very messy but not desperate. Julian and
Ammianus exaggerated the importance of the fall of Cologne to
disparage the achievements of Constantius II and glorify those of
his successor. The Franks posed no real threat. Their capture of
Cologne did not show their overwhelming strength but resulted
from their exploitation of current Roman weakness. As events were
to show, once they faced a serious Roman response they gave up the
city with little resistance.246 Alamanni were the more important
problem, especially now that some were settling on the Roman side
of the Rhine.247 However, they had been brought into line twice
already in recent months, and could presumably be brought into
line again, this time without the commitment of signiWcant add-
itional resources. The manner in which Ursicinus had dealt
with Silvanus, though underhand, had spared the west the horrors
of another civil war. Silvanus’ army was intact and obedient to
Constantius II and his representatives, as must have been the civil
administration in Gaul. The current trouble on the Rhine frontier
had been caused by irresolute leadership. Once Constantius had
decided a course of action, all this could change.248 He decided to
send in Julian. This was uncomfortable for both parties, given the
way that Constantius had beneWted from the massacre of 337, in
which most of Julian’s immediate family had perished,249 and given
the problems that were bound to arise if Julian, quickly married to
Helena, fathered a male heir before Constantius did.250 In the short
term, however, Constantius could take comfort in Julian’s youth and
inexperience; in the fact that he exercised no real power but was
245 Zo
¨
llner (1970: 1) ‘mutig, ungestu
¨
m, frech’; Drinkwater (1996: 23).
246 AM 16.3.2. Below 220. Cf. Elliott (1983: 251 and n.9): ‘so far as one can learn
from Ammianus about this recapture, it was eVected by walking into a deserted
town.’
247 On this, see further below 217–18.
248 As Fre
´
zouls (1962: 683) remarks, Constantius must in the end be credited with
a sensitive treatment of Ursicinus—not punished, but retained in post and eventually
sent back east.
249 Cf. above 199.
250 Cf. Drinkwater (1983b: 357–8).
ConXict 285–355 215
under the control of the new military magister and praetorian prefect
sent to replace Ursicinus and his staV; and in the expectation that, on
past form, it would not take long to restore things to normal. After
this, the ‘Julian problem’ could be considered afresh. Constantius did
not, therefore, as his detractors charged, simply promote Julian to
send him to death in battle.
216 ConXict 285–355
7
Conflict 356–61
Though Julian spent fewer than six years in the west, through his own
writings and those of his supporters we know more about his activ-
ities here than those of any other Roman general since Julius Caesar.
This includes his dealings with Alamanni. Julian’s Wrst Alamannic
campaign took place in 356, but before examining that year, we
should return to the issue of Alamannic occupation of Roman
territory.
When and why this came about is as important as whether
Constantius set Alamanni against Magnentius in the Wrst place.
Alamanni were certainly over the Rhine by the mid-350s. In 356,
Julian found Alamanni living in the territories of cities in the valley of
the Upper Rhine.1 In 357, he again encountered barbarians ‘who had
established their homes on our side of the Rhine’, and conWscated
their crops.2 This move may have begun as early as 352, and cannot
have been later than the end of 355.3 From a Roman point of view,
such a transference of population was neither unprecedented nor in
principle unwelcome. The Ubii had been allowed into the Empire by
Augustus; and, down to the Wfth century, various emperors did not
shrink from increasing the number of their subjects in the same
manner.4 For 359, indeed, Ammianus demonstrates how the idea
of settling barbarians on Roman soil could be presented very posi-
tively to an emperor (albeit, in this case, by court sycophants to
Constantius II): that by taking in Sarmatian Limigantes he would
1 AM 16.2.12. Cf. Julian, Ep. ad Ath. 279A; Libanius, Orat. 12.44, 48; 18.34.
2 AM 16.11.8, 11–12; cf. Libanius, Orat. 18.52. Cf. below 235–6
3 Above 215. Cf. Martin (1997a: 122), (1998: 413).
4 de Ste Croix (1981: 510–16).