Drinkwater J.F. The Alamanni and Rome 213-496 (Caracalla to Clovis)
Подождите немного. Документ загружается.
reception would have been the responsibility of the magister
oYciorum, from whom they might expect due deference and gener-
ous hospitality. (Complaints of Alamannic ambassadors against the
magister oYciorum Ursacius in 365 derived from their not having
received the welcome they had come to expect.100) In short, they too,
like their militar y cousins, must have had stirring experiences of
imperial life, and, through their acceptance of the imperial perspec-
tive, have helped to foster local acceptance of the identity of
‘Alamanni’ and ‘Alamannia’.
If they were so highly Romanized, why did leading Alamanni
not do more to Romanize their immediate environments, for
example by building vi llas and ‘proto-to wns’?101 I will return to this in
Chapter 9.102 Of more immediate concern is the question of the mid-
fourth-century ‘exclusion’. A strong body of opinion, voiced most
recently by Martin, holds that the disappearance of high-ranking
Alamanni from imperial service in the later part of that century is
not simply the result of the deterioration in the quantity and quality
of our sources, but is a real historical phenomenon, caused by a
change in policy on the part of Valentinian I. Valentinian, to dem-
onstrate that he was a strong emperor, decided to ‘get tough’ with
Alamanni, provoking them to retaliate and causing them to be seen
as the enemies of Rome. As a result, their leading men were excluded
from high oYce. In the second half of the fourth century, Frankish,
not Alamannic, generals rose to leading positions in the Roman
army.103
We have to be careful. The changing fortunes of regional groups
within the Roman Empire were not new. The ‘Spanish’ emperors of
the late Wrst and early second centuries came and went, as did the
‘Danubians’ of the late third and early fourth. This is unexceptional.
The rise and fall of cliques is normal in any social structure and
does not necessarily indicate the operation of Xagrant favour or
100 AM 26.5.7. Below 267; Jones (1964: 1.369), (1966: 140). Contra Lorenz, S.
(1997: 75 n.15), I see no reason to suppose that attendance at the imperial court was
forced on those concerned.
101 Cf. above 105.
102 Below 358.
103 Geuenich (1997a: 53–6); Martin (1997a: 122–4), (1998: 409–18). Cf. Bo
¨
hme
(1998: 40); Heather (2001: 53–4).
158 Service
prejudice.104 As the career of Silvanus shows, Franks were prominent
before Valentinian; and then, like Alamanni before them, Frankish
generals did well for a couple of generations or so only to be eclipsed
by other Germanic leaders and even Romans. It took almost another
century for Franks to return to play a major role in western aVairs,
but no one suggests that this was due to their being the victims of
exclusion. Though there was a decline in the number of Alamanni in
high positions from the middle years of the fourth century, this
would probably have happened anyway. This development may
have been hastened by events which made Alamannic communities
the enemy of choice for Roman commanders, though arising out of
Julian’s need to save himself from destruction and not from the
accession of Valentinian I.105 Yet, even then, there was no brutal
curtailment in the recruitment of Alamannic oYcers, as the careers
of Hariobaudes, Vadomarius, Fraomarius, Bitheridus and Hortarius
show. It simply faded away, suggesting that instead of exclusion we
should think of ‘accelerated evolution’.
Traces of more extensive Elbgermanic/Alamannic service in the
Roman army, that is that of ‘other ranks’, are sparse but signiWcant.
They occur from the later third century. Archaeological Wnds may be
interpreted as evidence for the Gallic emperors’ having recruited
Thuringians.106 Likewise, around 260, Gallienus may have made an
agreement with Iuthungi which included them undertaking to pro-
vide military help to the Empire.107 The Wrst direct indication of such
activity comes from the early fourth century, when we hear of Crocus
exercising signiWcant inXuence at the court of Constantius I as a
provider of auxiliary military support.108 A. H. M. Jones describes
him as ‘an Alamannic chieftain, commanding a body of his country-
men in Britain’.109
After Crocus there is a long gap until we reach Julian’s report that,
in the course of two battles and a siege, he captured a considerable
104 Cf. Elton (1996a: 144), on the rise and fall of other ‘regional cliques’ into the
sixth century.
105 Below 264–5, 272.
106 Above 73.
107 Above 75.
108 Epitome 41.3: Alamannorum rex, auxilii gratia Constantium comitatus.
109 Jones (1964: 2.612).
Service 159
number of men of Wghting age and his intimation that he drafted
them into his own forces.110 If one of these battles was that of
Strasbourg, these prisoners would have included young Alamanni.111
More explicit is Ammianus’ statement that, following Julian’s raid on
the territory of Vadomarius in 361, he received a certain number of
prisoners of war as dediticii, which must mean that he took them into
his army.112 Next is Ammianus’ account of the career of Fraomarius.
Here it is not the career of the colonel which is of interest but the
existence of his regiment which, we are told, was at that time fully
up-to-strength and doing well.113 Ammianus also relates the story of
an Alamannic soldier who served in Gratian’s guard and went home
on leave to his native Lentienses.114 His loose talk is supposed to have
precipitated the Lentienses’ rising of late 377, and contributed to
their defeat at the battle of Argentovaria in 378. After Argentovaria,
Gratian launched a campaign of vengeance against the Lentienses,
and then made a peace with them by which they provided recruits for
distribution throughout the Roman forces.115 It is probably this
incident that Gratian’s tutor and minister, Ausonius, touched upon
when he wrote poetically of Franks and ‘Suebi’ contending to supply
troops to Rome.116 About a dozen years later the anonymous author
of the so-called Historia Augusta, writing the ‘Life’ of the emperor
Probus, described dealings with Germani—by whom (given his ref-
erences to the Neckar and the Swabian Alp) he must mean
Alamanni—which appear better suited to late-fourth than to late-
third-century conditions.117 (Similarities have been noted between
this account and Symmachus’ second panegyric to Valentinian I,
delivered on 1 January 370.118) Probus, having driven Germanic
110 Julian, Ep. ad Ath. 280C–D.
111 HoVmann (1969: 141).
112 AM 21.4.8: dediticios cepit. See Elton (1996a: 129–30); cf. below 166.
113 AM 29.4.7: potestate tribuni Alamannorum praefecerat numero, multitudine
vir ibusque ea tempestate Xorenti.
114 AM 31.10.3: quidam inter principis armigeros militaris. See further below 310.
115 AM 31.10.17: post deditionem, quam impetravere supplici prece, oblata
(ut praeceptum est) iuventute valida nostris tirociniis permiscenda .
116 Ausonius 20 (Precationes ).2 [Green], 29–30; Green (1991: 535). Cf. HoVmann
(1969: 143), giving it a more general application.
117 Historia Augusta, V. Probi 13.5–14.7.
118 Okamura (1984: 316), citing Norden. Okamura refers to Symmachus’ third
panegyric, but this must be a mistake, perhaps resulting from a confusion between the
second panegyric and Valentinian’s third consulship, which it celebrated. Cf. below 289.
160 Service
invaders from Gaul, forced their leaders to give him hostages and
provisions, and then recruits for his army. These amounted to
16,000: ‘. . . all of whom he scattered through the various provinces,
incorporating bodies of Wfty or sixty in the detachments or among
the soldiers along the frontier; for he said that the aid that the
Romans received from barbarian auxiliaries must be felt but not
seen.’119 In 400, Stilicho held his Wrst consulship. Claudian, in a
poem celebrating this event, refers to Alamannic oVers of military
aid courteously declined.120
The most direct source for Alamannic military service is the
Notitia Dignitatum,anoYcial schedule of senior positions in the
civil service and army around 395. The Notitia is quirky, being in
places lacunose, and generally uneven in its state of revision. Its
eastern section reXects the situation just before 395, its western that
of c.420.121 Among a range of eastern frontier regiments named after
Germanic ‘tribes’, there are units, infantry and cavalry, of Alamanni
and Iuthungi.122 Among the eastern household regiments (auxilia
palatina: each with a nominal strength of c .1,200 men) is a unit of
‘Bucinobantes’.123 The Bucinobantes are known as an Alamannic
‘tribe’, living opposite Mainz, from Ammianus Marcellinus’ account
of Valentinian I’s vendetta against their king, Macrianus.124 On the
basis of this reference, other household regiments, both eastern and
western, with names that appear to derive from Alamannic districts,
have been accepted as having been raised from communities making
up the wider ‘Alamanni’. Thus ‘Raetovarii’ were recruited from
119 Historia Augusta, V. Probi 14.7: Accepit praeterea sedecim milia tironum, quos
omnes per diversas provincias sparsit, ita ut numeris vel limitaneis militibus quinqua-
genos et sexagenos intersereret, dicens sentiendum esse non videndum cum auxiliaribus
barbaris Romanus iuvatur.
120 Claudian, De Cons. Stil. 1. 233–6: Quotiens sociare catervas/oravit iungique tuis
Alamannia signis!/Nec doluit contempta tamen ...; HoVmann (1969: 143).
121 Jones (1964: 1.347, 351, 354, 395–413). Cf. HoVmann (1969: 525) and, most
recently, Kulikowski (2000b), proposing the dates 386 and c.419 respectively.
122 Not. Dig. Or. 28.43 (cohors IV Iuthungorum, Egypt); Or. 31.63 (cohors IX
Alamannorum, Egypt); Or. 32.36 (ala I Alamannorum, Phoenice); Or. 32.41 (cohors
V Pacata Alamannorum, Phoenice); Or. 33.31 (ala I Iuthungorum, Syria).
123 Not. Dig. Or. 6.58; Elton (1996a: 89).
124 AM 29.4.7: Bucinobantes . . . quae contra Mogontiacum gens est Alamannica.
Above 122–3, below 283, 304–10.
Service 161
former northern Raetia, and the ‘Brisigavi’ seniores and juniores from
the neighbourhood of Breisach in the Upper Rhine valley.125
The Codex Theodosianus preserves a constitution of 30 January
400, somewhat peevish in tone, from the emperor Honorius to
Stilicho as senior general (mag ister utriusque militiae), concerning
the attempts of various classes of people to evade military service. It
was probably to do with the contemporary defence of Italy.126 In a
short list of miscreants, mention is made of laetus Alamannus.
If, following the current communis opinio, ‘laetus’ is taken with
‘Alamannus’ (‘Alamannic laetus’), and we are not dealing with two
distinct categories (of ‘laeti’ and ‘Alamanni’), this document prov ides
evidence for Alamanni serving as laeti, that is some sort of tied
troops.127 This topic will be pursued further below.128
The evidence is thin and patchy, and presents no immediately
obvious pattern. One can only propose a set of possibilities and
then assess the extent to which each of these is supported by what
we know, that is to use models. However, we Wrst need to establish a
working terminology. Ancient w riters did not categorize foreigners
in imperial service as precisely as we would wish. It is best to follow
Elton in considering what such people did—how they were recruited
and employed—rather than what they were called.129
The basic distinctions were length of service and identity of lead-
ers. We may envisage two broad classes of recruits. The Wrst com-
prised men serving under their native leaders for speciWc campaigns.
They were not part of the Roman army and therefore not subject to
its regulations. The second consisted of men who, at least nominally,
volunteered for service in the Roman army subject to these regula-
tions. They might serve in ethnic units or be distributed among
ethnically-mixed regiments. If they found themselves under
Germanic oYcers, these would be full members of the Roman
125 Raetovarii: Not. Dig. Or. 5.58; Brisigavi juniores: Occ. 5.201/7.25 (Italy); Brisi-
gavi seniores: Occ. 5.202/7.128 (Spain). Generally see, e.g . Stroheker (1961: 40), citing
Schmidt and Wais; Geuenich (1997a: 29). For the location of the ‘Raetovarii’ ‘um
No
¨
rdlingen’ see HoVmann (1969: 1.165).
126 CT 7.20.12; Jones (1964: 2.620).
127 Lorenz, S. (1997: 15 n.35).
128 Below 166–9.
129 Elton (1996a: 92).
162 Service
oYcer-corps. These distinctions are not wholly exclusive: one can
envisage ways in which men of the Wrst group joined the second.
However, they are generally valid. Both groups could serve as the
result of formal agreements between their communities and Rome:
foedera (‘treaties’). Both might therefore legitimately be termed foe-
derati (‘federates’). This confuses the picture, and I will term the Wrst
group ‘allies’ and the second ‘auxiliaries’.130
The easiest type of general Elbgermanic/Alamannic ser vice to
envisage is that of the ‘allies’. Under Model I, we may imagine
Romans hiring itinerant war-leaders (not yet adjacent territorial
chiefs or kings) and their bands on a short-term basis. These would
have fought as instructed, then returned whence they had come.
Their mercenary activities were an extension of their normal long-
distance raiding. This is presumably how ‘Thuringians’ and Iuthungi
served Postumus and Gallienus and their successors. This category is
likely to have been a very early phenomenon, before the Alamannic
‘corridor’ became a ‘room’.131
The emergence of Alamannic settlements on the imperial frontier
is likely to have generated another type of allied service. While it is
unlikely that Rome encouraged kings to abandon their responsibil-
ities, princes and nobles would have been welcome military associ-
ates. Under Model II, one can imagine sons of kings and their
comites, or ambitious optimates, forming war-bands from the local
youth and serving the Empire as short-term allies under the terms of
treaties.132 The Epitomator’s likely exaggeration of the role of Crocus
in 306 makes it possible that for ‘king’ Crocus we should understand
‘prince’ Crocus.133 And, whatever his status, as Martin says, Crocus
was probably not alone in serving Rome in this fashion.134
As a variant of both Model I and Model II, it is possible to visualize
cases where, for various reasons, war-parties did not return home as
130 For ‘allies’ I follow Elton (1996a: 96). My ‘auxiliaries’ he calls ‘federates’ (92),
which returns us to potential confusion.
131 Above 80–1.
132 Cf. Elton (1996a: 33, 35), on subordinate reguli, subreguli and optimates.
133 Above 146. Cf. Zotz (1998: 387 (Fu
¨
rst), also 388: Crocus’ position may have
helped the rise of other Alamanni at this time). (For a somewhat diVerent view of
Crocus’ position and background, see now Drinkwater (forthcoming).)
134 Martin (1998: 408).
Service 163
planned. One may imagine an isolated group ceasing to be ‘allied’
and, perhaps, becoming an auxiliary unit in the Roman army. This
process would have been helped by the similarity in size and behav-
iour between barbarian raiding-bands and standard Roman army
units of the period.135 Such ‘detached’ groups are not visible in the
sources. It is likely that, like the auxiliary ‘native’ units of early
Roman Gaul,136 they were originally called after their leaders (‘the
manus, cohors or ala of x’), but such names are not found in
the Notitia Dignitatum. They may, however, have formed the basis
of the ‘broadly ethnic’ Alamannic and Iuthungian regiments of the
eastern frontier, dated to the late third and early fourth centuries.137
It must have been possible for individuals to cross into the Empire
and to join the army on their own initiative. This is my Model III.
Auxiliaries raised in this way would have been posted to existing
units.138 One text is conventionally regarded as directly conWrming
such recruitment. In the winter of 359–60, Julian is supposed to have
objected to Constantius II’s removal from Gaul of Germanic troops
who had volunteered for service under special terms:
One thing, however, he could neither overlook nor pass over in silence,
namely, that those men should suVer no inconvenience who had left their
abodes beyond the Rhine and come to him under promise that they would
never be led to regions beyond the Alps; for he declared that it was to be
feared that the barbarian volunteer soldiers, who were often accustomed to
come over to our side under conditions of that kind, might on having
knowledge of this thereafter be kept from so doing.139
These could have been Franks or Saxons, or even Ger mani from
much further aWeld, but they may also have included Alamanni. The
fact that such conditions of service could be bandied about at the
135 Elton (1996a: 72, 79, 99, 94).
136 Drinkwater (1978: 829).
137 Jones (1964: 611 and n. 7); HoVmann (1969: 140); cf. Stroheker (1961: 35–6).
Cf. also Kuntic
´
-Mativic
´
(1996: 187–8), on the imperial military’s taste for ethnic—in
this case ‘internal’ ethnic—labels during the fourth century.
138 Elton (1996a: 140).
139 AM 20.4.4: Illud tamen nec dissimulare potuit nec silere: ut illi nullas paterentur
molestias, qui relictis laribus transrhenanis, sub hoc venerant pacto, ne ducerentur ad
partes umquam transalpinas, verendum esse aYrmans, ne voluntarii barbari militares,
saepe sub eius modi legibus assueti transire ad nostra, hoc cognito deinceps arcerentur.
164 Service
highest level of imperial politics suggests that they were plausible.
Ammianus’ report is usually interpreted as direct proof of the
recruitment of individuals, as envisaged in Model III.140 However,
on practical grounds this is unlikely.
Individual recruits must have been posted to a wide variety of
units, and the presence in the same units of men recruited on
diVerent terms would have been impossible to manage. It is more
likely that what we have here is reference to conditions negotiated
between war-leaders and the Empire, that is that we are still dealing
with the ‘allies’ of Models I and II. According to HoVmann, during the
fourth century the use of Germanic allies was a characteristic of
western usurpers from outside the imperial family, such as Carausius,
Magnentius, Magnus Maximus, Eugenius and Constantine III.
However, all of these were ultimately losers and are treated as such
in the historical tradition: it suited the winners to emphasize their
opponents’ scandalous use of barbari. The winners, on the other
hand, may well themselves have employed such forces—it would
have been madness for them not to—while ensuring that this received
little publicity. Constantius II probably used allied Alamanni against
Magnentius.141 It suited Julian to criticize Constantius’ actions, but
this made it diYcult for him to justify his own use of Germanic allies.
It would have been in his interests to confuse the issue by conXating
potential (and legitimate) complaints of the allies and current (and
quasi-mutinous) objections by certain regular regiments: that by
moving south of the Alps they would expose the families they left
behind to barbarian attack.142
This leaves the most likely example we have of an individual
volunteer as Gratian’s Lentiensian guardsman, loyal enough to be
allowed leave. However, even he may have been promoted into the
guards from a unit raised, say, according to Model V (below).
The youths posted to Roman regiments as a result of Gratian’s
arrangement with the Lentienses cannot have been treated diVerently
140 e.g. HoVmann (1969: 142, 151); Heather (2001: 33).
141 Below 201–2.
142 So AM 20.4.10, 16. On this see further below 254. Cf. Martin (1998: 413): that
Julian’s army would have contained many Germans—though he assumes that most of
these would have been ‘laetic’ in origin.
Service 165
from their new companions, and so must have been regarded as
volunteers. The same was probably true of Julian’s forced recruits
of 20 years earlier. Possible archaeological evidence for this sort of
recruiting into the Rhine and Danube armies from young men from
throughout the Elbgermanic triangle is in the form of Wnds of male
cloak-brooches, after the Roman military style and dated to the later
fourth and early Wfth centuries, throughout the region.143
Models I–III concern nominally voluntary service by ‘free’
Alamanni. DiVerent from this is the recruitment of Alamanni living
on Roman soil as prisoners of war (dediticii) or refugees (laeti), dealt
with here as Model IV. The general issue of Gallic laeti is diYcult and
controversial. In taking them to be Germanic in origin I follow den
Boeft et al.144 In return for being granted their lives and/or perman-
ent residence in the Empire, both dediticii and laeti appear to have
been placed in discrete settlements under the supervision of imperial
‘prefects’. Though the Notitia Dignitatum lists only praefecti laetorum,
a reference by the emperor Julian suggests that both laeti and dediticii
were viewed in much the same manner.145 The Wne line between the
two sorts of communities, especially in later generations, may have
led to both eventually being treated as laetic.146 They were obliged to
maintain themselves and supply recruits to the Roman army. How-
ever, such recruits were not formed into their own regiments but
posted in the usual way to existing units.147
A localized and obscure institution, laeti have caused uncertainty
and confusion among historians, ancient and modern. Zosimus saw
them as a ‘tribe’.148 Ammianus Marcellinus has laeti barbari making
an unsuccessful attack on Lyon in 357.149 In his Loeb edition of
Ammianus, Rolfe followed Zosimus in translating the phrase con-
taining laeti barbari as ‘the Laeti, a savage tribe, skilled in seasonable
raids’. By treating laeti as a proper noun and the ‘Laeti’ as a tribe he
by-passed the issue of the existence of a semi-servile population in
143 Schach-Do
¨
rges (1997: 82); cf. above 160.
144 den Boeft, den Hengst, and Teitler (1987: 202–4).
145 AM 20.8.13.
146 Cf. Lorenz, S. (1997: 14–15, 65): laeti were the oVspring of dediticii.
147 Elton (1996a: 131).
148 Zosimus 2.54.1.
149 AM 16.11.4. Below 225.
166 Service
Gaul. In Rolfe’s version, the Laeti appear to be an Alamannic tribe or
sub-tribe. This interpretation reXects a long-standing misinterpret-
ation in anglophone studies, reinforced by Rolfe’s description of the
recruits whom Julian oVered Constantius in 360 as ‘of the Laeti, a
tribe of barbarians on this side of the Rhine’.150 As is being increas-
ingly recognized, Ammianus’ laeti of 357 cannot be interpreted in
this way.151 They are just ‘barbarian laeti ’: certainly not a tribe,
probably not Alamannic, perhaps not principals in the events of
357, and maybe not hostile to Rome.152 One may imagine the
inhabitants of one or more Germanic laetic settlements west of the
Rhine, having weathered the political storms of the early 350s, now
Wnding themselves caught between Alamanni and Romans. Sus-
pected by both sides, they Xed deeper into Gaul. However, their
Xight was interpreted and dealt with as an attack.
Thus, there is no secure mention of tied Alamannic recruits in the
Roman army in Ammianus. All that remains is a reference to
Alamannic prisoners of war being sent to farm in the Po valley in
370, after count Theodosius’ Raetian campaign.153 This has been
connected with the Alamannic laeti of the Codex Theodosianus.154
However, the latter and their fellow draft-dodgers seem to have been
a relatively sophisticated group of people, evading their responsibil-
ities by obtaining documents exempting them from service, and may
have been relatively few in number. It looks as if the Po-dwellers had
done exceptionally well for themselves as farmers, and were now
hardly down-trodden semi-free dependants of the Empire. (The
fact that neither they, nor any of the other groups of potential recruits
mentioned, appear to have been destined for any particular sort of
unit must be a strong indication against the existence of speciWcally
‘laetic’ units.155) In the end, the best evidence for ‘normal’ laetic
150 AM 20.8.13: adulescentes Laetos quosdam, cis Rhenum editam barbarorum
progeniem. See, e.g. Browning (1975: 84): ‘Another German group, the Laeti . . .’;
Elton (1996a: 131): ‘Laeti was also the name of an Alamannic canton.’ Generally,
Lorenz, S. (1997: 15 and n. 35).
151 Nixon and Rodgers (1994: 142 n. 76).
152 Contra Lorenz, S. (1997: 40): that these laeti were hostile Alamanni.
153 AM 28.5.15: quoscumque cepit ad Italiam . . . misit, ubi fertilibus pagis acceptis,
iam tributarii circumcolunt Padum. Cf. below 283, 304.
154 Above 162; HoVmann (1969: 142 and n. 90).
155 Elton (1996a: 131).
Service 167