Drinkwater J.F. The Alamanni and Rome 213-496 (Caracalla to Clovis)

Подождите немного. Документ загружается.

Iuthungi never posed a threat to Italy, it is certain that they were

perceived as having done so. Anti-Germanic prejudice, the terror

teutonicus and terror ger manicus , had been burned into the Roman

soul by the Cimbric invasion and the Varian disaster. Marcus

Aurelius’ Marcomannic Wars will have conWrmed such feelings; and

the incursions of 260, 268/9 and 270/1 must have touched a raw

nerve. Aurelian’s re-walling of the city of Rome, begun immediately

after the second Iuthungian conXict and arguably the greatest monu-

ment to the uncertainty of the third-century ‘Crisis’,180 may be seen

as a political more than a tactical move. It resulted from Aurelian’s

recognition that he needed to be seen to be doing something, espe-

cially in order to distinguish his regime from that of the supposedly

negligent Gallienus, who had also allowed Germanic raiders onto

Italian soil. His Roman walls were imposing but, in military terms,

superXuous and Xawed in design.181 However, they were useful in

projecting the image of Aurelian as a man of action; and they

continued to remind people, wrongly, of the danger from the north.

Although there was, as far as we know, no further Alamannic or

Iuthungian raiding into Italy in the third century, there were still

incursions into Raetia and Gaul. Since 260 Gaul had been protected

from both Alamannic and Frankish depredations by the attention of

Postumus and his successors as Gallic emperors, Victorinus and

Tetricus, all based in the Rhine/Moselle area. This privileged position

was lost in 274 when Aurelian overthrew Tetricus; and the situation

was made worse by the severity with which Aurelian re-imposed

imperial authority and by the uncertainty that followed his assassin-

ation in 275.182 Franks and Alamanni raided over the Rhine.183 Their

direct eVect was probably overestimated by ancient writers.184 How-

ever, there was undoubtedly looting and slaughter throughout the

Gallic provinces. The Wnds at Hagenbach and Neupotz have already

been noted.185 In addition, at least 13 inhabitants of the villa at

180 Watson (1999: 143): ‘the most emblematic and the most enduring monument

of Aurelian’s age.’

181 See Watson (1999: 150–1), on the weaknesses of the original Aurelianic circuit.

182 Okamura (1984: 304); Drinkwater (1987: 44, 202); Witschel (1999: 217–18, 229).

183 Gru

newald and Seibel (2003: xii–xiii).

184 Okamura (1984: 304–9).

185 Above 70.

78 Arrival

Regensburg-Harting were brutally killed and their bodies dismem-

bered, partially scalped and dumped into two wells. Likewise,

remains of a late-middle-aged man and a teenage girl, both of whom

had suVered burning and dog-bites, were found in a well at the villa

of Mundelsheim.186 As earlier in Italy, the perceived danger will have

been very great, causing panic and, in some regions, with the

appearance of indigenous rural rebels, the ‘Bagaudae’, social disrup-

tion.187 After the short reigns of Tacitus and Florianus (275–6), the

situation was taken in hand by Probus (276–82). He campaigned in

the west in 277 and 278, apparently giving his personal attention to

the area over the Upper and High Rhine and the upper Danube, that

is precisely where Alamanni and Iuthungi were to be found, though

Zosimus, in a very muddled section, in this respect mentions only

Vandals and Burgundians.188 SigniWcant raiding into Gaul, and the

disruption that this caused there, ended under Probus’ successors,

Carus and Carinus (282–5) and Diocletian and Maximian (as

‘dyarchs’, 285–96). Though there were still some bar barian incur-

sions, these do not appear to have been serious, and some may even

have been exaggerated to enhance the reputation of the ruler who

dealt with them.189

The absence of any attempt to reconquer the lands lost between

the Rhine and Danube and the old Upper German/Raetian limes at

this time should not be seen as the measure of a prevailing Germanic

‘threat’. The Empire had other concerns, and the new line of demar-

cation worked well enough—to the extent that it was being Wxed in a

line of fortiWcations.190 Inertia had set in. As from the early Wrst

century, the area could serve as a useful source of external ‘danger’

and military glory.

186 Wahl, Wittwer-Backofen, and Kunter (1997: 345–6); Witschel (1999: 104, 106

and n.26); Schro

ter (2000: 180–1); Carroll (2001: 138).

187 For the Bagaudae, see Drinkwater (1984: 363–8).

188 Zosimus 1.67, 1.68.1; Paschoud (1971: 173–5); Okamura (1984: 313).

Cf. below 108.

189 Okamura (1984: 321, 325, 333); below 181.

190 Below 192.

Arrival 79

Settlement

There is very little evidence for Germanic settlement over the Upper

German/Raetian limes before about 300. However, this could be an

artefact of archaeology, perhaps due to frequent changes of simple

settlements, the wooden structures of which are invisible to current

techniques of surveying and excavation.1 Where early settlement

aVected Roman masonry buildings, as at the Wurmlingen villa,

it can be detected.2 We may therefore presume some early activity

of this type, if only to explain what followed.3 But what were its

mechanisms?

In the Wrst instance, war-bands raiding into and through former

provincial territory would have had little inclination to settle. Their

warriors were accustomed to returning home, and are unlikely to

have had the wish or the expertise to turn to farming. As settlers, they

would also have been open to attack by the Empire and by other

bands. On the other hand, continued success and the diYculties

involved in carrying booty and prisoners over long distances

would have encouraged them to put down roots further west. Just

beyond the mouth of the ‘invasion-funnel’ into the Rhine–Danube

re-entrant, to the west of early Elbgermanic settlement around

Bamberg,4 there are signs of later-third-century Elbgermanic settlement

down the Main, around Wu

rzbur g5 (Fig. 4). Further into the fu nnel,

conditionswould have been morediYcult.Here, the‘invasion-corridor’

1 Geuenich (1997b: 76).

2 Fingerlin (1997b: 127–30); and below 86.

3 Cf. Schach-Do

rges (1997: 96): to the early fourth century, Alamannic settlement

was slow, but not non-existent.

4 Above 48–9.

5 Steidl (2002: 108); cf. below 92.

would have to turn into a ‘living-room’.6 But this could only happen

from west to east, since any east to west settlement would have run the

risk of being destroyed by continuing new arrivals.7 Germanic settle-

ment in Alamannia indeed appears to have begun in the north-west,

close to the Roman frontier.8

How many ‘proto-Alamanni’ are we talking about? Alamannia,

however deWned (it is commonly thoug ht of as being bounded by the

old limes, but this is not precisely the case9), was not large, being

about the same size as a small Roman province. As a funnel into the

Empire it could have taken considerable through-traYc, but once

raiders started to settle numbers would have been more restricted. An

average war-party of 600-strong,10 once stationary and acquiring

dependants, would have placed a substantial burden on the local

economy, and the more so if this economy was centred upon a

damaged and degraded agricultural system.11 Furthermore, if, as

seems likely, such bands were prone to quarrel, they could not have

settled in close proximity to each other. The region cannot have

supported very many such groups. Below I propose a maximum

resident population of c.120,000 for the fourth century. The third-

century Wgure must have been considerably lower.

Crucial must have been Roman acceptance and indeed encourage-

ment of a Germanic presence.12 From the imperial viewpoint, with-

drawal from the Upper German/Raetian limes had created a

re-entrant in the frontier which left Italy and central Gaul vulnerable

to Germanic raiding. The corner had to be taken oV and the best way

of achieving this was to persuade some bands to help the Empire

against the rest. Rome demonstrated its overwhelming strength by

restoring the frontiers.13 Hav ing blocked the raiding-bands, what

could the Roman Empire oVer to get some to stay? The obvious

answer is pay or subsidies, in precious metal coin or bullion. Such

payments are attested in the literary sources in respect of the

Iuthungi, and in the archaeological in the case of the Haßleben-

Leuna warriors. Similar ones were probably made to Alamanni,

though it is strange that Roman silver and gold coins are rare in

6 Nuber (1990: 52). 7 Below 106. 8 Below 83; Stribrny (1989: 437).

9 Below 110–11. 10 Above 49. 11 Below 90.

12 Nuber (1993: 103), (1998: 376–7). 13 Below 192.

Settlement 81

elite graves and hill-sites.14 Rome could also oVer toleration of

occupation of former imperial territory. Finally, Rome could give

group-leaders access to imperial goods, opportunities for serv ice

within the Empire, status to help them maintain their positions

and, when necessary, help against rivals: ‘management’, not suppres-

sion.15 Against Nuber, early Roman policy cannot have extended to

the encouragement of the growth of a dependent population

required to provide men and mate

riel suYcient to make up for the

lost tax revenues of the region. How could these lands, their agricul-

ture disrupted and their population diminished, have satisWed such

demands?16 Such a reconstruction of events is heavily ‘tribal’ in

conception, and overinXuenced by details found in the unreliable

Historia Augusta.17 It smacks of anachronism, of fourth-century

conditions under which, as we shall see, there was some restoration

of a Roman presence and Roman exploitation of natural resources

over the Rhine.

Whatever uncertainties there may be about developments in the

second half of the third century, it is now accepted that from around

300 an increasing number of Elbgermanic warrior-bands settled

permanently in Alamannia.18 However, it is very diYcult to perceive

this process. Our best information is from burials, but this is unrep-

resentative and its interpretation controversial.19 The problem is

compounded through diYculty in dating what has been found.

Consequently, the course of permanent settlement, for example

whether it happened at once or in stages, is disputed.20 As far as

geographical distribution is concerned, our understanding is not

helped by a lack of correlation between such groupings as can be

established, such as those of the hill-sites, the coin-Wnds, the ‘sub-

tribes’ of Ammianus and the Notitia Dignitatum and the gold-hilted

sword burials.21

14 Below 105–6, 131. Cf. Reuter (1997: 69): that the earliest ‘Alamanni’ were

mercenaries in the pay of the Gallic Empire.

15 Below 145–6, 178, 189.

16 For the condition of Alamannic agriculture see below 90–1.

17 Historia Augusta, V. Probi 15.2–6.

18 Steuer (1997a: 149), (1998: 281–3, 285).

19 Above 4, 48–9.

20 Cf. Fingerlin (1997b: 125).

21 Steuer (1997a: 152); cf. below 96, 126, 130, 135, 169–74, 341.

82 Settlement

As far as the speed of settlement is concerned, those scholars

prepared to express a view favour a gradual process.22 Prior to around

300, Alamannic settlement (inferred from small clusters of ceremo-

nial burials of group-leaders and their dependants) was to be found

along the lower Main, the lower and middle Neckar, and in the area of

the No

rdlinger Ries at the eastern end of the Swabian Alp23 (Fig. 6). In

the course of the fourth century, settlement was extended to the upper

Rhine and upper Danube as immigrants were drawn to the favourable

conditions they found there24 (Fig. 7). These conditions were

good agricultural land, already opened up by Roman settlement,

and Roman suVerance of newcomers. The availability of land

explains Alamannic settlement in or near former forts, towns, villages

Fig. 6 Elbgermanic burial sites to c.300 [from Schach-Do

rges (1997: 96

Fig. 82); ß Archa

ologisches Landesmuseum Baden-Wu

rttemberg].

22 Fingerlin (1993: 79); Schach-Do

rges (1997: 98); Bu

cker (1999: 18 and n.25);

Steidl (2000: 108); Schmauder (2002: 199).

23 Schach-Do

rges (1997: 96 and Fig. 82). Cf. above 80.

24 Schach-Do

rges (1997: 97 and Fig. 83); cf. (for the Breisgau) Fingerlin (1993:

69), (1997a: 106); Bu

cker (1999: 192–3, 197, 217).

Settlement 83

and, especially, villas.25 Such practice, which was continued when the

Alamanni moved further into imperial territory in the Wfth and

sixth centuries,26 indicates that they did not systematically destroy

or even (contrary to a canard based on a misreading of Ammianus

Marcellinus) sedulously avoid former Roman sites.27 But Alamannic

settlement was not restricted to the neighbourhood of former Roman

Fig. 7 Elbgermanic burial sites c.310–75 [from Schach-Do

rges (1997: 97

Fig. 83); ß Archa

ologisches Landesmuseum Baden-Wu

rttemberg].

25 The pioneering study was Weidemann’s (1972); see now Planck (1990: 71, 82, 84,

92, 94); Fingerlin (1993: 64–5 and nn. 16–17), (1997b: 127–31, 136); Schach-Do

rges

(1997: 96); Bu

cker (1999: 23). (Limes forts/associated vici: Jagsthausen; Osterbu

rken;

Unter bo

bingen; Weltzheim. Hinterland forts/associated vici: Heidenheim; Urspring;

Walheim. Vici:Ko

ngen, Ladenburg. Villas: Hechingen-Stein; LauVen-am-Neckar;

Wurmlingen.)

26 Bu

cker (1997: 320).

27 AM 16.12.2. Accepted, for example, by Stroheker (1975: 33); Martin (1997a: 121

and n.9), (1998: 409–10). Cf. Fingerlin (1997b: 128); Nuber (1998: 375); Sommer

(1999: 191). Zotz (1998: 394) notes that caution would anyway be needed in dealing

with Ammianus’ statement since he says much the same about the Huns: 31.2.4. On

this, see further below 218.

84 Settlement

structures. Examples of free-standing sites are also increasingly com-

ing to light. Some of these were agricultural, but others were devel-

oped to exploit mineral resources, speciWcally iron.28 Access to, and

between, all sites will have been helped by the survival of many of the

Roman roads of the region.29

Though we now know that there was direct continuit y of settle-

ment (between ‘Rhine–Weser’ and ‘Elbgermanic’ artefact users) at

existing Germanic sites beyond the old limes,30 there is less certainty

about what happened within it. The question as to whether there was

a residual ‘Roman’ population will be considered below.31 Much has

been made of Ammianus’ reference to Alamannic villas, built after

the Roman fashion and destroyed by Julian’s troops in his campaign

in the Main valley in 357.32 Matthews, for example, Wnds conWrma-

tion of Ammianus’ observation in the site of Ebel-bei-Praunheim,

west of Nida-Heddernheim near Frankfurt-am-Main (Fig. 4). Here, a

Roman villa abandoned during the later third century underw ent

Germanic reoccupation in the later fourth century, with stretches of

drystone wall and posts being added to the structure. Matthews sees

this as the work of ‘an Alamannic optimate’.33 However, the patching

up of a century-old ruin hardly smacks of careful restoration in the

Roman style; and I argue below that Ammianus’ ‘native’ villas may

anyway have been Roman.34 Further, how or whether the buildings

concerned functioned as the centre of a Roman-style estate remains

unclear; and as far as ‘Alamannic’ nobles are concerned, a nearby

weapon-grave has been interpreted as Eastgermanic (speciWcally,

Burgundian), of a man who achieved success by serving in the

Roman army in Gaul.35 What we have here are signs not of prosperous

28 e.g. Großkuchen (iron); Lauchheim (late seventh century, but very instructive);

Mengen; Schleitheim; Sontheim-im-Stubental (fortiWed): Bu

cker et al. (1997: 318–20);

Planck (1990: 82–4, 87–90); Stork (1997: 304, 306); Wahl, Wittwer-Backofen, and

Kunter (1997: 340). See now Trumm (2002: 218), for the settlement of marginal

agricultural land in eastern Swabia, probably in connection with the extraction and

working of iron ore.

29 Fingerlin (1997b: 131). For the network see Fig. 4. Cf. below 135.

30 Schmauder (2002: 199–200).

31 Below 126–8, 136.

32 AM 17.1.7. Cf. below 226.

33 Matthews (1989: 310), with Werner (1965: 77–80 and Fig. 1).

34 Below 134, 241.

35 Schulze-Do

rlamm (1985: 560–1).

Settlement 85

Alamannic aping of Roman ways, but of tension in the region.36

Likewise, to take a somewhat earlier example, around 250 a smallish

villa at Wurmlingen (Kreis Tuttlingen) on the upper Danube was

Wnally abandoned. About 10 years later it was reoccupied by

Germanic incomers. Their ‘reoccupation’ consisted of no more

than the insertion of timber structures, ‘Grubenha

user’, into the

Roman complex, sometimes in the shells of rooms, which were

re-roofed using Germanic techniques. The precise use of the new

structures is unknown.37 The primitive rooWng of Germanic struc-

tures is particularly signiWcant in distancing these from those ‘built

after the Roman fashion’, because it is precisely the splendour

of Roman roofs that Symmachus evokes in his description of

Valentinian I’s military strongpoints.38 What we Wnd, therefore, is

that Alamannic presence on former villa sites shows association with

but not continuity of these places.

The newcomers ignored the previous Roman economic, social and

administrative structures. Indeed, though we have no clear idea of

the northern and eastern limits of Alamannia, Elbe-Ger mani with

close links to ‘Alamanni’ also settled to the north and east of the old

limes on land that had never been part of the Upper German or

Raetian provinces.39 In the south and west, the survival of Roman

roads did not prevent Alamannic settlement from being broken up

by natural features, in particular the Black Forest.40 Unlike, for

example, what happened in post-Roman Gaul, here there is abso-

lutely no sign of any survival of the civitas-organization or the

juridical framework.41 The evidence is diYcult to Wnd and interpret,

but there is a general consistency and the main concentrations of

archaeological material of the fourth and early Wfth centuries are to

be found in (Fig. 8): (a) the lower Main valley, near the conXuence

with the Upper Rhine; (b) the lower Neckar valley around the

36 Cf. below 112, 114–16, 190, 324.

37 Fingerlin (1997b: 130); Reuter (1997).

38 Symmachus, Orat. 2.20. Cf. below 300.

39 Keller (1993: 110–11); Schach-Do

rges (1997: 97–8 and Figs 83–4). Cf. below

125.

40 Cf. Fingerlin (1990: 103) and (1993: 65–7) on the relative isolation of the

Breisgau.

41 Fingerlin (1997b: 125–6).

86 Settlement

Fig. 8 Alamannia: main concentrations of archaeological remains. Key:

a lower Main; b lower Neckar; c middle Neckar; d Breisgau; e Swabian Alp;

f High Rhine.

Settlement 87