Drinkwater J.F. The Alamanni and Rome 213-496 (Caracalla to Clovis)
Подождите немного. Документ загружается.
According to our sources, all the Roman prisoners concerned were
taken in the troubles following the revolt of Magnentius. This is
consistent with the general development of Romano-Alamannic
relations during the Wrst half of the fourth century, which were
generally peaceful, with little opportunity for routine Alamannic
slaving in imperial territory.135 Thus, in 358 and 359,136 Julian
purportedly freed a large group of people taken in the short period
from 351 to 355 (when Constantius II began to re-establish order in
Gaul after Magnentius) or to 356 (when Julian Wrst began to
re-establish order in Gaul after the revolt of Silvanus).137 Julian there-
fore liberated a distinct set of people who had been captured in
unusual circumstances and who were, apparently, still available for
mass repatriation very close to Roman territor y—cer tainly in the
territories of Suomarius, Hortarius and Vadomarius (on the Rhine)
and very probably in those of Urius, Ursicinus and Vestralpus (on the
Main artery).138
These had not been dispersed into Alamannia or beyond; they had
not become part of the slave trade. At this point one begins to suspect
that the people involved were not run-of-the-mill enslaved war
captives but rather, as Curta has suggested (albeit via a diVerent
line of reasoning), hostages, acquired for ransom or for bargaining
in future negotiations.139 Those from upper levels of society would
have Wtted this bill better than peasants and slaves, and would be the
sort of people to W gure on relatively hastily prepared lists. Both
Libanius and Eunapius refer to the high birth of the prisoners.140
The taking of hostages in large numbers, though at the most in
thousands, not tens of thousands, would indicate a fair degree of
political acumen on the part of Alamannic war-leaders, but this is
hardly surprising given their long contact with Rome.141 As part of
this cunning, one might adduce the fact that they did not threaten to
135 Below 200.
136 See Lorenz, S. (1997: 56 and nn.210–11), for problems in establishing the
precise chronology of events.
137 Cf. below 204, 212, 219.
138 Below 223 Fig. 18, 250.
139 Curta (2002b).
140 Libanius, Orat. 12.50; Eunapius, Hist. fr. 19.
141 Cf. below 145–6.
138 Society
kill their prisoners before the battle of Strasbourg, or do so after their
defeat. But this leads to another issue.
All these prisoners of the Alamanni would have needed food and
shelter. How was this provided? Surely not in camps, an anachron-
istic concept, well beyond the capacity of the local economy. An
alternative is their being held on rural sites and set to work to feed
themselves. This would have been all the more necessary because
Alamanni occupying land on the left bank were growing crops there
with the help of, one presumes (though Libanius says they worked
‘with their own hands’), labour brought over from the right.142 It
would explain the complaints of the sources that they were treated
like slaves, which well-born hostages would have found hugely
humiliating. Ho wever, the notion of har d work suggests a further group
of potential hostages more used to labour. A signiWcant proportion
of Roman prisoners held by the Alamanni in 358 may have been
people who had been working on the right bank of the Rhine,
supervising production for export into the Empire. These could
have been seized when the political situation deteriorated from 351,
and not released after Magnentius’ fall in 353. They were again
the sort of people whom authorities might well have been able to
list. The possible exceptional nature of the prisoners held by the
Alamanni in the 350s must change our interpretation of sources.
What happened cannot be generalized to yield conclusions about
Alamannic society as a whole: it is not proof that the Alamanni were
peculiarly short of labour.
So what was the nature of day-to-day Alamannic dependent labour?
A simple model of how this may have developed may be constructed
from what has been touched upon.143 In the late third and early fourth
centuries, young warriors overwintering in Alamannia needed a sub-
servient agricultural workforce and drew this from a residual peas-
antry or from prisoners taken in raiding into the Empire, or both. In
the fourth century, as permanent communities developed, Alamanni
continued to use slaves for agriculture, for practical (the advantages
of a subservient workforce) and cultural (the satisfaction derived
from the existence of social inferiors and the belief that the meanest
142 Below 227, 235–6; Libanius, Orat. 18.34.
143 Cf. above 104.
Society 139
tasks were Wt only for the unfree) reasons. However, given the low level
of population, such slaves were not present in massive numbers. Even
less numerous, but equally signiWcant, were the skilled outsiders who
served the elites of the hill-sites. In the later third century, such
outsiders may have been slaves, acquired through raiding and intro-
ducing new skills.144 In the fourth century, such people would have
been less accessible and recourse may have been had to dependent
freemen: we know of a much later Alamannic leader employing free
craftsmen.145
In the fourth centur y, free Roman craftsmen were available over
the Rhine. Indeed, it may have been that a number of these were
voluntary exiles, with no wish to return to Roman jurisdiction. In
other words, the exploitation of transrhenish resources may, in
opening up the Rhine frontier, have encouraged imperial emigration.
From what later writers, such as Salvian, say about deteriorating
social conditions in the west in the Wfth century, which drove the
desperate to Xee to the barbarians, there is the possibility that the
fourth century saw imperial migrants trickling into Alamannia,
escaping problems with landlords, tax-collectors, judges and recruit-
ing oYcers.146 It is signiWcant that legislation of Valentinian I recog-
nized the possibility that, in time of barbarian attack, people might
go over to the enemy of their own accord.147
Overall, we may envisage the population of Alamannia during the
fourth century as being a rich mix of Germanic warriors (Elbgerma-
nic and eastern), Roman provincials of various backgrounds and
144 Cf. above 93. Steidl (2000: 109–10), (2002: 107–10) notes a rise in Germanic
technical expertise in the later third century—visible in the Haarhausen–Eßleben
‘experiments’—and attributes it to a single generation of Roman craftsmen, taken as
slaves in third-century raiding, who were unable to pass on their skills to succeeding
generations.
145 Steuer (1990: 201), citing Eugippius, V. Severini 8.3: Giso, wife of the Wfth-
century Alamannic king, Feletheus, foolishly subjected bar barian goldsmiths brought
to work at court to unlawful arrest.
146 Salvian, De Gub. Dei 5.22. As already noted above (127), according to Tacitus,
Germania 29.4, in the Wrst century the Agri Decumates were colonized in advance of
formal annexation by ‘all the riV-raV of Gaul’: levissimus quisque Gallorum.
147 CT 5.7.1; cf. below 278. On the forces that encouraged desertion (excessive
taxation, extortion by soldiers and civil oYcials, the demands of landowners etc.), see
Gutmann (1991: 20). For the possible presence of Roman or Roman-trained workers
on Germanic soil in the Early Roman Empire see Stupperich (1997: 19).
140 Society
conditions, the products of inter breeding between these, and depen-
dants brought from Free Germany. As far as these last are concerned,
leaders probably had their families with them from the start.148 If
having a Germanic wife was seen as a mark of high status, others
would have striven to bring one in under the easier conditions of
permanent settlement, along the lines of communication that linked
the Vistula, the Oder and the Elbe to the Main and the Rhine.149
This returns us to the thorny problem of population. If Alamanni
were few in the later third century,150 were they more numerous in
the fourth? Much has been made of Aurelius Victor’s description of
the third-century Alamanni as a ‘populous nation who Wght won-
derfully well from horseback’.151 These words have done much to
shape modern popular conceptions of the Alamanni;152 but they are
open to question. It has been suspected that the sources’ projection
of the population of the ‘free’ Germanic territories as massive, as a
potential Xood tide of immigrants, is mistaken, a reXection of
imperial pr opaganda.153 The idea of Alamanni as numerous and excep-
tionally skilled horsemen is likewise explicable as, in modern jargon,
an urban myth: a piece of conventional wisdom which appears
plausible until subjected to scrutiny. Such wisdom may have been
derived, for example, from the ancient belief that cold northern
climes naturally produced an excess of population.154 A related and
perhaps more likely explanation is that Victor used this belief to
point up Julian’s recent victory at Strasbourg.
This famous defeat of the Alamanni took place in summer 357,
just under four years before Victor’s publication of his Caesares in
spring 361.155 Victor had a high opinion of Julian, a feeling that was
reciprocated when Julian, having moved eastward from Gaul, made
him governor of Pannonia II in summer 361.156 Before this, however,
148 Above 49.
149 Cf. above 46 n.20.
150 Above 80–1.
151 Aurelius Victor, Caesares 21.2: gens populosa ex equo miriWce pugnans [trans.
Bird (1994)].
152 e.g. Su
¨
dwestfunk (1997).
153 Wolfram (1990: 28); Drinkwater (1997: 1 and n.2). Cf. below 169.
154 Pohl (2002a: 15).
155 Below 224–5; Bird (1994: xi).
156 Bird (1994: ix).
Society 141
as he was completing the Caesares and dealing with contemporary
matters, Victor had been under the authority of Constantius II
and therefore had to be careful in what he wrote. Thus Julian is
praised, but brieXy and with no mention of Alamanni; and his
successes are credited to the ‘good fortune and planning’ of the
senior emperor.157 But the story of Julian’s victory at Strasbourg
must have spread quickly through the Empire, and have been made
more dramatic by the detail that the day had almost been lost thanks
to the failure of the Roman cavalry, which Julian had to rally, and
later punished.158 I propose that Victor, unable to comment directly
on the battle in its place, made indirect reference to it when describ-
ing Caracalla’s Germanic campaign of 213. Recording this Wrst con-
tact with Alamanni, he enhanced Julian’s victory in two ways. First,
helped by ancient preconceptions of northern bar barian fecundity,
he made the Alamanni generally dangerous by asserting that they
were many in number.159 Second, he made them particularly dan-
gerous, and excused Julian’s lapse as a commander, by transposing
their unexpected cavalry success at Strasbourg into the past and
categorizing them, incorrectly, as horsemen of renown.160
What may safely be said about Alamannic numbers? In truth,
nothing. The archaeological evidence does not suggest a massive
population, but as far as precise Wgures are concerned we can only
speculate. The most direct statements concerning the strength of a
Germanic ‘Volk’ come from Victor Vitalis and Procopius. Combined,
these suggest that in 429 Geiseric led 50–80,000 Vandals and others
across to Africa.161 As Pohl says, citing Wolfram for c.100,000,
including 15–20,000 warriors, as the maximum number for ‘a
major people’, this is not implausible.162 Alamannic numbers at the
battle of Strasbourg may be calculated at around 15,000, made up of
c.1,000 warriors from each of 11 known or likely pagi and 4,000 from
157 Aurelius Victor, Caesares 42.18.
158 Below 239.
159 On Alamannic numbers at Strasbourg, see below 237–9.
160 i.e. contra Springer’s view: that Victor used Julian’s achievements to enhance
Caracalla’s campaign. This is discussed and refuted by Bleckmann (2002: 161–3),
though Bleckmann still believes Victor on the quality of Alamannic cavalry. See
further below 175–6.
161 Pohl (2002a: 76–7).
162 Wolfram (1990: 28): ‘ein großes Volk’.
142 Society
elsewhere.163 Wolfram’s multiplier of about four dependants for each
warrior produces a total associated population of c.75,000. Though
not all Alamanni were involved, this Wgure approaches the proposed
maximum of 80–100,000 and suggests that the population of those
areas which stayed completely aloof from the conXict amounted to
no more than c.5–25,000.
But these sums require adjustment. First, the 4,000 outsiders and
their dependants have to be subtracted from the 75,000 ‘involved’
Alamanni, bringing these down to c.55,000; and some at least of the
former have to be re-allotted to nominally neutral communities. The
4,000 may have included warriors from the wider Elbgermanic tri-
angle. Alamannia was probably still visited by bands raiding west-
wards over long distances from within this region. Though normally
a danger for settled communities, in 357 these could have supple-
mented Alamannic Wghting strength.164 On the other hand, such
recruitment could not have produced the whole 4,000. Given the
rapid turn of events that led to conXict, local Alamanni must have
been driven very much on their own resources.165 A small number—
three or four—roving bands of Elbe-Ger mani could have contrib-
uted, say, half the extras.166 This leaves c.2,000 drawn from Alaman-
nic pagi nominally uninvolved in the conXict. With their c.8,000
dependants, these raise the total population of these communities
to c.15–35,000. But, second, this is not enough. At c.5,000 people per
pagus, these Wgures produce only three to seven additional pagi,
which is too few. There were about Wve north of the Main (those of
the ‘three kings’; and of Macrianus and Hariobaudus);167 a clutch,
say three, north of Lake Constance;168 perhaps a couple along the
Swabian Alp;169 and, say, another three belonging to the Iuthungi.170
This makes 13, or an additional 65,000 people, bringing the total
population of ‘Greater’ Alamannia to around 120,000 (55,000 þ
65,000)—large for a people on the move but plausible for a group
163 Below 238–9; for pagus, see above 119–20.
164 Steuer (1998: 288), his explanation for Alamannic ‘populousness ’. Cf. below 238–9.
165 On the antecedents to Strasbourg, see below 236–7.
166 Above 49.
167 Below 226, 240, 246, 249.
168 Below 207, 209, 211–12.
169 Above 88, 97.
170 Above 97, 99–100.
Society 143
of relatively settled communities. Though these other pagi had a
greater total population than those which found themselves facing
Julian, they were, with the exception of those on the Main, probably
more scattered and larger in area. The main concentrations of popu-
lation were, as the archaeological material and the historical record
suggest, probably along the Rhine, the lower Main and the Neckar.
To set all this in perspective, against c.120,000 Alamanni there was
a population in the neighbouring Three Gauls and Germanies of
around 10,000,000—a ratio of about 80:1 in favour of the Empire.171
171 Drinkwater (1983a: 170–1).
144 Society
5
Service
The Alamanni are best known for Wghting Rome, and conXict will be
the subject of the following three chapters. However, they also served
Rome, and an understanding of this service is essential for recon-
structing the relationship between Alamanni and Romans in war and
at peace.
Alamannic service was essentially military service. As such it was
hardly unusual. The use of outsiders to augment Roman armed
strength went back to the Republic; Julius Caesar hired Germanic
cavalry;1 and earlier emperors, for example Tiberius and Caracalla,
employed Germanic warriors as bodyguards.2 Wider recruitment of
Germani began in the later third century3 and became a matter of
course. Current thinking is not to view such ‘barbarization’ as over-
whelming or as a threat to the security of the Empire. Germani,
oYcers and men, served Rome well.4 Military service was, indeed,
the traditional means by which Germani proWted peacefully from the
later Roman Empire.5 From the later fourth century, the most suc-
cessful were able, through intermarriage, to produce children who
were accepted into the topmost levels of the imperial aristocracy.6
It is helpful to distinguish between oYcers and other ranks. As
Zotz remarks, thanks to the work of Stroheker, Waas and Martin, we
1 Above 14.
2 Tacitus, Annales 1.24.3; Dio 78.6.1; Herodian 4.7.3. Geuenich (1997a: 34).
3 Above 73, 75, 77 (Postumus, Gallie nus, Aurelian).
4 On the ‘normal’ and positive side of Germanic recruitment, see Elton (1996a:
136–52); Witschel (1999: 235). Cf. Jones (1964: 620–2).
5 Martin (1998: 407).
6 e.g. the Vandal father of Stilicho and, indeed, Stilicho himself: PLRE 1.853; and
probably also Bauto the Frank: see below 152–3.
are now much better informed about certain high-ranking Alamanni
in the service of Rome during the fourth century.7 These are now
dealt with in chronological order.
CROCUS8
Following the death of Constantius I in York in 306, prominent among
those who urged his son, Constantine, to replace him was Crocus, a
king of the Alamanni and a military ally of Constantius.9 Caution is
necessary since the Epitome is a relatively late (early Wfth century) and
quirky source, which here is clearly embellishing the account of
Aurelius Victor,10 and since an even later account mentions another
Alamannic king Crocus raiding deep into Gaul in the mid-third
century.11 ‘Crocus’ may therefore have become a legendary ‘bogey-
man’, with the name of the ‘earlier’ being applied to the ‘later’, or even
vice versa. The Epitome’s reference may, indeed, be pejorative, suggest-
ing unhealthy barbarian inXuence at the court of Constantine I.12
However, there is no reason to doubt the basic historicity of what it
says, that is that an Alamannic leader played a role in the elevation of
Constantine I.13
LATINUS, AGILO, AND SCUDILO14
The next three important military Alamanni occur in 354, in
Ammianus Marcellinus’ account of Constantius II’s campaigning
7 Zotz (1998: 388); Stroheker (1961), (1975); Waas (1971); Martin (1997a),
(1998). Cf. Geuenich (1997a: 28–35).
8 Waas (1971: 82); PLRE 1.233.
9 Epitome 41.3: Alamannorum rex; auxilii gratia Constantium comitatus.
10 Aurelius Victor, Caesares 40.4.
11 Gregory of Tours, Dec. Lib. Hist. 1.32, 34; cf. Fredegarius, Chron. 2.40, 60.
12 Festy (1999: 186).
13 Festy, ibid., following PLRE 1.233, even accepts that this Crocus may have been
a descendant of the Crocus of Gregory of Tours.
14 Latinus: PLRE 1.496; Waas (1971: 88). Agilo: PLRE 1.28–9; Waas (1971: 68–9).
Scudilo: PLRE 1.810; Waas (1971: 103–4).
146 Service
near Kaiseraugst. According to Ammianus, at that time they exer-
cised enormous inXuence in the running of the Empire.15 All three
were, presumably, members of the imperial privy council, the sacrum
consistorium.16
In 354, Latinus was comes domesticorum.17 This meant that he
commanded the protectores domestici, elite troops, the cream of the
imperial bodyguard.18 These men attended directly upon the
emperor and were used by him for special errands. They were close
to the centre of power and could expect accelerated promotion. Their
unit has been described as a ‘corps of oYcer-cadets’ and ‘a kind of
staV college’.19 Latinus was, therefore, ‘a senior general’.20 He was also
an imperial politician with important connections, and in his deal-
ings with Constantius had already shown some spirit. In 351, during
the manoeuvring that led up to Magnentius’ defeat at the battle of
Mursa, Constantius II had thrown a great banquet. Latinus, already
comes domesticorum, and Thalassius, an imperial comes, refused to
attend to express their concern about Constantius’ acceptance of
Magnentius’ treatment of an envoy. Thalassius was from a rich
Antiochene family and was related to the famous rhetor, Libanius.
He was a long-standing lieutenant of Constantius II and by the time
of his death in 354 had become praetorian prefect of the east and
played a role in the fall of Gallus.21 Zosimus describes Latinus and
Thalassius in 351 as ‘the emperor’s closest associates’.22 Latinus was,
indeed, mixing with the cream of imperial society.23
In 354, Agilo was t ribunus stabuli,24 in charge of the imperial
stables and remounts.25 He then became commander of a regular
regiment of the imperial guard (schola), but his career only really
15 AM 14.10.8: qui tunc ut dextris suis gestantes rem publicam colebantur.
16 Cf. Elton (1996a: 176).
17 AM 14.10.8.
18 Jones (1964: 1.372, 2.636–40).
19 Jones (1964: 2.638), (1966: 224); Tomlin (1987: 115).
20 Tomlin (1987: 114).
21 Thalassius: PLRE 1.886; Waas (1971: 88).
22 Zosimus 2.48.5: ta pro
¯
ta para basilei pherontes [trans. Ridley]; cf. Paschoud
(1971): ‘qui exerc¸aient les plus hautes fonctions aupre
`
s de l’empereur’. See de Jonge
(1939: 104).
23 Paschoud (1971: 257).
24 AM 4.8.10; de Jonge (1939: 105).
25 Jones (1964: 1.372–3, 2.625), (1966: 141); de Jonge (1939: 104–5).
Service 147