Drinkwater J.F. The Alamanni and Rome 213-496 (Caracalla to Clovis)
Подождите немного. Документ загружается.
presence from the Rhine to Italy. The process began in 382 with
Gratian’s move from Trier to Milan, continued in 387 with Magnus
Maximus’ personal intervention in Italy, and culminated in 402 with
Honorius’ retreat to Ravenna. Overall, therefore, the Alamanni of the
380s and 390s were hardly the bogeymen presented by Ammianus as
he published his ‘History’. A feature of the period is not trouble from
them but from Rome’s earlier bugbear, the Franks.
Around 380, Macrianus of the Bucinobantes was killed by the
Frank Mallobaudes.321 In 388, Franks attacked into Germania II.
According to Gregory, citing Sulpicius Alexander, these were led by
Genobaudes, Marcomeres and Sunno, and did great damage. Some
of them were defeated but the rest, who had already crossed back over
the Rhine, inXicted a heavy defeat on a pursuing Roman force.322 In
389, Arbogast, acting according to the wishes of Valentinian II,
though himself in favour of a much more retributive policy, made
an agreement with Marcomeres and Sunno.323 Arbogast got his way
some years later when, having reduced Valentinian II to a cipher, he
attacked Sunno and Marcomeres in the middle of w inter, crossing
the Rhine and devastating Frankish territory, but failing to make
decisive contact with the enemy. 324 According to Gregory, Eugenius
went to the Rhine border and renewed the old traditional treaties
with the kings of the Alamanni and Franks in order to show oV his
massive army.325 Marcomeres and Sunno were Wnally destroyed by
Stilicho. (Marcomeres was sent into exile in Italy; Sunno, champion-
ing Marcomeres, was killed by his own people.326)
One has to be careful in assessing this apparent resurgence in
Frankish activity because our principal informant, Gregory of
Tours, was interested solely in searching out what he could about
the early history of the Franks of his day. However, the likelihood that
an increase in Frankish raiding is a real historical phenomenon is
strengthened by the fact that Sunno and Marcomeres receive men-
tion in Claudian, a contemporary. The main cause of this was
321 AM 30.3.7. Above 174, 285.
322 Gregory of Tours, Dec. Lib. Hist . 2.9.52–4.
323 Gregory of Tours, Dec. Lib. Hist . 2.9.55.
324 Ibid.
325 Ibid.
326 Claudian, De Cos. Stil. 1.241–5.
318 ConXict 365–94
probably the diverting of imperial attention by civil war. However,
it seems to have been exacerbated by personal hostility between
Arbogast and the Frankish chiefs. The trouble was settled once
control of aVairs had passed into the hands of the capable Stilicho.
At one level, therefore, it appears that despite the change that
aVected the Roman world after Adrianople the relationship between
the Empire and Alamanni and Franks remained much as before,
indeed that it reverted to that of the earlier period. However, this is
not the whole of the story. There is also a hint that the change which
aVected the Roman world also inXuenced what went on between
Romans and Germani. Because this change provoked greater internal
conXict, there was an increased demand for troops among rival
parties. Thus, although according to Gregory of Tours, Eugenius
went to the Rhine in 392 to cow the kings of the Franks and
Alamanni into submission prior to his moving elsewhere, it is likely
that he and Arbogast raised troops from them.327 A line in Orosius
implies that Frankish contingents fought for Eugenius and Arbogast
at the battle of the Frigidus.328 If these were raised at the same time as
Eugenius’ appearance on the Rhine then Alamannic contingents are
also a possibility, although this is not stated explicitly (and Eugenius’
recruitment may have involved only Franks, newly active and with
some, at least, having tribal links to Arbogast). Whatever, after 383
there is no further record of signiWcant Roman conXict with
Alamanni.329
327 i.e. along the lines of Models I and II above (163).
328 Orosius, Hist. 7.35.11–12.
329 Stroheker (1975: 40–1).
ConXict 365–94 319
9
The Fifth Century
Theodosius died in Milan on 17 January 395, leaving the Empire
to be governed by his two sons, the east by Arcadius and the west
by Honorius. Like Valentinian II, however, both were thought
incapable of ruling alone, and so were subordinated to ‘regents’. In
the west was Stilicho, Theodosius’ nephew by marriage. Not long
after Theodosius’ death, probably in 396,1 Stilicho visited the Rhine.
According to Claudian, Stilicho went to calm this river’s Werce
nations and to bring peace to the region. He travelled swiftly, with
no escort, receiving the submission of reges from source to mouth,
and granting them peace without needing to Wght them. He was away
for only a few weeks at most. The Rhine-Germani oV ered military
aid, but this he courteously refused.2
In the western section of the Notitia Dignitatum, where one expects
to Wnd reference to the frontier commander (dux) of Germania I,
there is instead reference to two local commands: a frontier com-
mander at Mainz (dux Mogontiacensis) and, even more oddly, the
commander of a Weld army based in Strasbourg (comes Argentoraten-
sis).3 There has been great debate about the date and circumstances of
the creation of these posts, with suggestions ranging from 369,
and Valentinian I’s work on the frontier, to 413, and the reforms
of Constantius III in Gaul following his defeat of the usurper
Constantine III.4 The issue here is whether they resulted from imper-
ial considerations or were mainly in response to external factors, that
1 Gutmann (1991: 111).
2 Claudian, De Quarto Cos. Hon. 439–59; De Cos. Stil. 1.188–245. Cf. above 171.
3 Not. Dig. Occ. 41; 27.
4 Demougeot (1975: 1123/59); Stroheker (1975: 41 n.45); cf. below 324.
is to some sort of shift in the nature of the ‘threat’ from neighbouring
Alamanni. Demougeot, developing an established line of thinking,
favours the former, seeing the new posts as the work of Stilicho in the
period 396–400. Stilicho created them with a view to internal security,
following other changes he made to the army and administration in
Gaul.5 I therefore exclude the oYces of dux Mogontiacensis and comes
Argentoratensis from discussion here. However, since there is a case for
dating the former somewhat later, to the period following the over-
throw of Constantine III, I will return to them below.6
Stilicho kept himself busy. In particular, he involved himself
closely in the aVairs of the east, attempting to gain dominance over
Arcadius. This drew him into alternating conXict and alliance with
Alaric, leader of a Gothic army, who exploited tensions between the
two halves of the Empire in the hope of gaining a high imperial
position. In 402, for example, Stilicho had to expel Alaric from Italy,
but in 408 sought his aid in actions planned against the administra-
tion of Arcadius and the western usurper, Constantine III. Between
rejection and recruitment of Alaric, in 405 Stilicho had to face
invasion of Italy from the Danube region by another barbarian
general, Radagaisus. And then came the incursion by Vandals,
Alans and Suebi which began on 31 December 406 (or 4057) and
which eventually aVected much of Gaul and Spain.8 Jerome (by
mentioning Vandals, Alans and Alamanni, but not Suebi) hints that
the Suebi were in fact Alamanni; and Gregory of Tours, describing
their crossing into Spain with the Vandals, directly equates Suebi and
Alamanni.9
These events exacerbated trouble in Britain which led, in 407, to
the emergence of the last of the traditional western usurpers,
Constantine III. He, like all his predecessors, crossed to Gaul,
where he renewed old treaties of alliance and aid and agreed new
ones with Franks, Alamanni and Burgundians.10 These and other
5 Demougeot (1975: 1124–6); cf. Bur y (1923: 1.118 n.3).
6 Scharf (1990); below 328.
7 Kulikowski (2000a).
8 Prosper, Chron. ad ann. 406; Orosius, Hist. 7.38.3; Hydatius, Chron. ad ann. 409
(34 [Burgess 42]).
9 Jerome, Ep. 123.16; Gregory of Tours, Dec. Lib. Hist. 2.2: Suebi, id est Alamanni.
10 Drinkwater (1998a: 282).
The Fifth Century 321
disasters led to the fall and execution of Stilicho (22 August 408), and
for a while Constantine seemed to be about to control or even
displace Honorius. However, fortune began to run against him. In
411, under attack in southern Gaul from the Spanish kingmaker,
Gerontius, he sent his Frankish general, Edobichus, north to raise an
army from Franks and Alamanni.11 Gerontius besieged Constantine
in Arles, but withdrew before an imperial army under Honorius’
generals Constantius and Ulphilas. Constantine held on, waiting for
Edobichus, but when Edobichus arrived he was defeated.
What light do these events shed on the relationship between the
Alamanni and Rome? First, why did Stilicho journey along the
Rhine? His refusal of military aid is odd: given his tense relationship
with the east, Stilicho must have sought to build up his military
strength even before Alaric’s Wrst invasion of Italy. Therefore, despite
Claudian, it is usually assumed that Stilicho recruited Germanic
allies.12 Claudian’s remark may be taken as defensive: to protect the
half-Vandal, ‘semi-barbarian’ Stilicho from the (truthful) charge of
hiring barbarians. Embarrassment would have been justiWed.
Because barbarian leaders were able to satisfy such demand, they, as
allies rather than enemies, began to gain the initiative in dealing with
the Empire, a circumstance of which the imperial political nation
began to be increasingly aware. On the other hand, the case for such
recruitment remains speculative. In particular, Stilicho’s formation
of regiments of Brisigavi can no longer be taken as disproving
Claudian’s claim:13 there is a case for such regiments being imperial
in nature.14 If he did recruit, Stilicho, like Eugenius, may have
directed most of his attention towards Franks.
Stilicho is frequently blamed for cannibalizing the western gar-
rison for his eastern adventures and to defend Italy.15 There seems
little doubt that he removed troops.16 However, whether he recklessly
11 Sozomen 9.13.2.
12 So Bury (1923: 1.119), tentatively; Demandt (1989: 141), with assurance.
Cf. above 171.
13 Contra Stroheker (1975: 41 n. 44).
14 Above 170–1.
15 e.g. Bury (1923: 1.169); Demougeot (1975: 1127–8); Stroheker (1975: 41 and
n.43); Demandt (1989: 142).
16 Claudian, Bell. Get. 416–18 (Alaric). Bayard (1993: 227).
322 The Fifth Century
ran down the western defences is debatable, at least in respect
of the Rhine region.17 There had never been a major Germanic
‘threat’. The Empire experienced some trouble in dealing with its
neighbours, but it could handle this through police action. The large
garrison in Gaul had resulted from historical, political reasons, not
local need. It was unnecessary and, in encouraging usurpation,
dangerous.18 The Rhine garrison could, and probably should, be
reduced. The problem for Stilicho was that Gaul was invaded not
by established neighbours, but by warriors from the blue.
This brings us to the Vandals, Alans and Suebi. Given the nature of
their society, it is very likely that some bands of Alamannic warriors
joined the newcomers and were swept with them over the Rhine.
There is, indeed, a case for Gregory of Tours’ king Respendial being
an Alamannic leader operating in southern Gaul and Spain in 410,
rather than an Alanic chief on the Rhine in 407.19 However, given
their customary behaviour, it is equally unlikely that a substantial
number of Alamanni allied themselves with Vandals and Alans and,
as the Suebi of our sources, raided across Spain and into Gaul.
Further, if Alamanni moved en masse against Rome, how may we
explain the loyalty of the Franks, who appear to have resisted them,
and even the relative docility of the Burgundians?20 Jerome’s cata-
logue of invaders must be regarded as fanciful;21 and Gregory of
Tours’ gloss should likewise be seen as speculation on the nature of
the mysterious Suebi.
The origin and identity of all three groups remains a mystery.22
The surprise they caused suggests speed and military skill. This was
17 For what follows see Drinkwater (1996: 29–30).
18 Below 362.
19 Wynn (1997: 81–5).
20 Franks: Orosius, Hist. 7.40.3; Gregory of Tours, Dec. Lib. Hist. 2.9 (though see
Wynn (1997) on the chronological and textual problems involved in the customary
reading of this passage). Zo
¨
llner (1970: 25). Anton (1981: 238–40), following Jerome
and Orosius, has elements of Burgundians caught up with the Vandals, Alans and
Suebi, but clearly demonstrates that further Burgundian movement was then only by
Roman invitation.
21 Freeman (1904: 21).
22 See Kulikowski (2000a: 341–2).
The Fifth Century 323
no ‘Vo
¨
lkerwanderung’.23 We should suppose a group of war-bands.
Perhaps the best explanation for the arrival of mixed forces of
Vandals, Alans and Suebi on the Rhine around 406/7 is that of
Edward Gibbon: they were survivors of Radagaisus’ force, defeated
in northern Italy by Stilicho early in 406.24 As one group may have
moved directly westwards, and reached Gaul over the Cottian Alps,
another could have escaped over the Julian Alps, Wrst north towards
the Danube and then west to cross the Rhine at Mainz.25 The Suebi
were Danubian Suebi.26 The real nature of the Alamannic ‘threat’ is
demonstrated in what happened next: Franks resisted the invaders;
Alamanni did nothing. Apart from Vandals, Alans and Suebi, the
people of the moment were the Burgundians. The events of the early
Wfth century suggest that by 406 they had come down the Main,
reaching the Rhine and displacing local Alamanni.27 They were now
able to cross the river and become a force in Gaul: Jovinus,
Constantine III’s successor, was raised to power by Burgundians
and Alans.
Constantine III used Franks and Alamanni as Eugenius and
Stilicho had: not Wghting them but recruiting them for his army,
the last time through Edobichus. Gregor y of Tours repor ts that
Constantine surrendered having been told that Jovinus had declared
against him in the north and was heading south with a large force,
23 Thus contra Demougeot (1979: 432). The conventional view that the 406/7
invaders comprised a massive folk movement (e.g. Bury (1923: 1.185): ‘vast com-
panies of Vandals, Suevians and Alans’; Thompson (1977: 303): accompanied by
‘their womenfolk, their children and their old people’) appears to derive from the
rhetoric of Jerome (Ep. 123.16: innumerabiles et ferocissimae gentes) and Orosius
(Hist. 7.40.3). Among its most recent proponents is Burns, who (1994: 205–7, 223)
likewise supposes a mixed composition and large numbers, but rejects the idea of
planned ‘invasion’ in favour of ‘mismanaged receptio’ as the starving Vandal nation
sought refuge within the Empire and then overwhelmed Roman resources: in eVect, a
repeat of the Gothic immigration over the Danube under Valens.
24 Gibbon (1901: 268); contra: Freeman (1904: 22) and, for example, Petrikovits
(1980: 272–3: an independent movement of peoples previously settled by the Romans
on the upper Danube). For the likely relatively modest size of Radagaisus’ army,
despite ancient exaggeration of its strength, see Burns (1994: 198, cf. 189 for similar
remarks concerning Alaric’s forces).
25 Paschoud (1989: 22–4, 28–9).
26 Above 48, 168. Cf. Castritius (2005: 202–3): these Danubian Suebi were the
descendants of Quadi.
27 Martin (1997b: 163–4) also citing archaeological evidence.
324 The Fifth Century
including Alamanni. This, however, is a doublet of the tale of
Edobichus’ relief force. Alamanni did not march south twice in
411.28 The old order appeared to continue, but change was on the way.
The Wfth-century history of the Roman west may be said to have
begun with the elevation of Jovinus, probably in Mainz, in 411. The
west had produced many usurpers, but Jovinus was the Wrst to
depend on the support of local barbarian leaders, in his case Goar
of the Alans and Guntiarius of the Burgundians. This brings to mind
the machinations of future Germanic kingmakers inside and outside
the Empire, in particular of king Theoderic II of the Visigoths (who
created the emperor Avitus) and of the Suevo-Visigothic magister
militum Ricimer (who destroyed him and his successors).29 However,
one has to be careful. Once in power, Jovinus was no creature of his
backers, who should therefore be regarded as simply powerful
helpers.30 On the other hand, it is clear that the balance of power
in the west was shifting in favour of Germani.
Jovinus proved unsuccessful, destroyed by Honorius’ Gothic ally,
Athaulf. Peace and order of a sort were re-established in the west. The
Theodosian dynasty endured and preserved stabilit y in Gaul and on
the Rhine for over a generation. This achievement was rooted in two
long reigns—of Honorius (395–423) and of Valentinian III (425–
55)—and the abilities of two Roman western commanders-in-
chief—Flavius Constantius (died 421) and Flavius Ae
¨
tius (murdered
454).31 It was Ae
¨
tius who inXicted a major defeat on the Iuthungi in
430–1, after which we hear no more of them.32 However, this stability
was only relative. The Roman west was changing out of recognition.
The acceptance of autonomous communities of barbarians within
the Empire created internal political and military boundaries. In
Gaul, an early precedent had been set by Franks in Toxiandria.33 In
413 and 418 these were joined by Burgundians on the Rhine (and
later in Savoy) and by Visigoths in Aquitania, respectively. At a
28 Gregory of Tours, Dec. Lib. Hist. 2.9.56; see Drinkwater (1998a: 289 n. 136).
29 See most recently MacGeorge (2002: 178–214).
30 Drinkwater (1998a: 288).
31 On these, see now Lu
¨
tkenhaus (1998) and Stickler (2002).
32 Hydatius, Chron. 93 [Burgess 83]; Sidonius Apollinaris, Carm. 7.234; Stickler
(2002: 189–90). Above 62–3.
33 Above 200.
The Fifth Century 325
slightly later date there was settlement by scattered groups of Alans.34
In addition, the north-western region of Aremorica appears, like
Britain, to have gone its own way; and, again as Bagaudae, disaVected
Roman citizens were quite capable of disrupting nominal imperial
territory.35 In addition, western emperors and generals had to Wnd
time to manage the di Ycult relationship with Constantinople. And
dominating every thing was the problem of North Africa. Geiseric of
the Vandals had invaded the area from Spain in 429 and had quickly
taken it under his control. Its loss was a crippling blow in terms of
morale, rents and taxes. Roman attention was much given to the
questions of how, at best, to reconquer it or, at least, to prevent
Geiseric from using his Xeet against the Italian archipelago and the
peninsula itself.36
The impression given by our scanty sources is of Constantius, and,
particularly, Ae
¨
tius, struggling to maintain the situation in Gaul, not
to improve it. An adventurous policy, aimed at, say, the recovery of
Britain, or even Aremorica, would have endangered the army and the
renown that kept Ae
¨
tius in power and were the west’s only shield
against the eastern Empire and the Vandal kingdom. It is best to
envisage Roman authority in Gaul, especially in nor thern Gaul and
along the Rhine, in the Wrst half of the Wfth century as one force
striving to be the leader among many.37 In this, military police work
against, say, Burgundians and Visigoths is not surprising: these were
new and potentially very dangerous presences in Gaul. More sig-
niWcant is the relative frequency of reports concerning Frankish
aggression, including the multiple sacking of Trier.38 This is in line
with renewed Frankish activity in the later part of the fourth century,39
and, of course, with the success of Clovis at the end of the Wfth.40
34 Burgundians: above 108; Visigoths: Heather (1991: 221); Alans: Thompson
(1982: 25) and, imaginatively, Kovalevskaja (1993: 210).
35 Thompson (1982: 31–7); Drinkwater (1992: 215–16), (1998a: 285–6).
36 See, e.g. Bury (1923: 1.256).
37 Wood (1987: 260); cf. Elton (1992: 170).
38 At least four times before the middle of the century, though the precise dates are
disputed: e.g. Anton (1984: 9–12): 410, 413, 419 or 420, 428 or 435; Runde (1998:
675–6): 411, 413, c.420, c.428.
39 Above 318. Though Stickler (2002: 175–9) assumes that agreements must have
been struck between Romans and Franks in the 430s and 440s, in order to explain the
relative peace of the 450s.
40 Below 344.
326 The Fifth Century
West of the Rhine, therefore, the situation was stable, but such
stability was bound to be only temporary.
What was the place of the Alamanni in all this? Surprising, given
their reputation, but to be expected, given the history of their
relationship with the Empire, is their general passivity. This is not to
say that there was no trouble. The Alamanni still comprised a warrior
society, and despite the absence of any mention of them in the sources
we should envisage raiding-bands attacking out of Alamannia as the
opportunity presented itself. One result of such activity may have
been the destruction of the second fort at Alzey, north-west of Worms,
at some date after 425.41 However, beyond the middle of the Wfth
century there appears to have been no signiWcant Alamannic
movement over the Rhine.42 The conventional explanation is that
the Alamanni were blocked by the Burgundians.43 However, given
what Alamanni had shown themselves capable of in the past when
imperial attention had been diverted, this is unsatisfactory. If they had
wanted to make real trouble they could have done so. A major
discovery of recent years is indeed that, on the contrary, some of
them, in company with other Ger mani, helped to defend the Empire.
The Wrst indications of this came from excavation of burials at the
fort of Krefeld-Gellep, south of Xanten on the Lower Rhine. Finds
indicated the recruitment of right-bank Germani from the late third
into the early fourth century and from the late fourth to the mid-Wfth
century, when they formed the bulk of a regular garrison.44 Other
sites, for example Ju
¨
lich, Alzey and Neuss, subsequently yielded very
similar results.45 The clearest signs of Alamannic involvement were
41 Oldenstein (1993: 126).
42 Stroheker (1975: 43); Keller (1993: 101–2); Koch (1997a: 196); Stickler (2002:
186–7, 203). Contra Fingerlin (1990: 110), that there was movement into Alsace from
the time of Stilicho’s withdrawal of the frontier garrisons.
43 e.g. Thompson (1982: 29, though cf. 32–3 and 36–7, where he shows the
weaknesses of this analysis in respect of Savoy); Stroheker (1975: 46); Runde (1998:
676 n. 105, 677); Stickler (2002: 186) though again with doubts over Savoy (203).
Cf. Pohl (2002a: 173).
44 Pirling (1993); Reichmann (2003: 43–5).
45 Oldenstein (1993: 25–6); Bo
¨
hme (1998: 51–3); Kaiser (1998: 35). For further
likely Wfth-century Roman military sites on the Rhine see: Bayard (1993: 226–7);
Bridger (2003: 25–6); van Enckevort and Thijssen (2003: 94–7).
The Fifth Century 327