of the Gothic Tervingi on the Lower Danube in the early 330s, for in-
stance, the emperor Constantine I opened up the entire spread of their
sector of the Lower Danube for trade. This was done by the emperor
from a position of strength, as something designed to give these Goths,
or their leaders who would benefit from the tolls, real reason to keep
the peace.19 It was also customary to take high-status hostages, usually
the sons of kings and princes, as a further prop to any peace deal. If
things went wrong, these captives could be executed, but there is only
one known fourth-century example of this. More generally, these hos-
tages were usually also young, and bringing them up around the impe-
rial court had long been designed to impress on these possible future
rulers of Rome’s borderlands the power and prestige of the empire,
which could act as a deterrent to future misbehavior should the now
former hostages ever come to power as adults.20
Frontier Defense 233
Less positive measures were also available. If a particular barbarian
leader’s ambitions threatened to destroy or distort the peace arrange-
ments, then imperial commanders were regularly ordered to resort to
kidnap or assassination. In just the twenty-four years covered by the
dense contemporary narrative of the late Roman historian Ammianus
Marcel inus (354–78), these techniques were deployed on no less than
five separate occasions.21 Whether al these tactics amounted to a grand
strategy is contestable, but their existence shows the late empire operat-
ing on far more than a merely defensive footing. Rather, what emerges
with great clarity is that the later empire turned its immediate neighbors
into junior client members of a Roman world system, exerting mili-
tary power to order their affairs in the manner that best suited the em-
pire’s interests. The narrative suggests that each major intervention led
to diplomatic settlements with an average life span of some twenty to
twenty-five years—more or less a political generation. On the Rhine, for
instance, the Tetrarchic emperors mounted one major intervention in
the 290s, Constantine mounted another in the 310s, and there then seems
to have been substantial stability down to the 350s. The Tetrarchs were
again busy on the Middle Danube in the decade after 300 AD, Constan-
tine intervened with a major campaign in the early 330s, and peace then
prevailed again until the later 350s. The pattern on the Lower Danube
was again similar, with the Tetrarchs and Constantine mounting cam-
paigns in the 300s and early 330s, but this time the peace deal—perhaps,
among other reasons, because of the special trading privileges granted
the Gothic Tervingi—lasted until the mid-360s.22 This does not amount
to an unblemished record of frontier security, but, especial y for a pre-
modern state operating at such slow speeds over such vast distances,
getting twenty to twenty-five years of peace from each bout of major
campaigning represents a decent return on its military investments, and
no bad overal record of keeping its possessions secure.
To understand Roman–barbarian relations fully, however, and to
grasp the relationship between Roman frontier policies and the even-
tual processes of imperial collapse, it is necessary to explore one further
dimension of the empire’s approach to client management and fron-
tier security. In the short term, any particular round of campaigning-
followed-by-diplomacy was geared toward generating as much stability
234 Heather
as possible on a particular sector of the frontier. Looked at in the long
term—and by the fourth century, these rhythms of Roman frontier
management had been in operation along the Rhine and Danube for the
best part of 400 years—these techniques had had powerfully transfor-
mative effects on the empire’s neighbors across the border. Diplomatic
subsidies and trading privileges, backed up by imperial diplomatic in-
terference, such as providing political and military support for favored
barbarian rulers, tended to put money and power in the hands of par-
ticular kings. Played out over 400 years, the longer-term effect of this
approach was to help concentrate power more generally in the hands
of an entirely new type of king. The Germanic world of the first cen-
tury AD was populated by a host of small-scale sociopolitical units. Well
over fifty appear in the pages of Tacitus’s Germania covering Central
Europe, for the most part between the Rhine and the Vistula. By the
fourth century, this multiplicity of smaller units had given way to a
much smaller number of larger ones, perhaps no more than a dozen.
These were certainly confederative overkingships, so that estimates of
the degree of political revolution they represent need to be kept within
reasonable bounds. But whereas in the Roman period larger confedera-
tions disappeared with the defeat of their leaders, these fourth-century
counterparts could survive even substantial defeat. The immediate
rulers of the Alamanni of the Upper Rhine frontier were a series of
canton kings and princes. Periodically, however, these banded together
under an overking of particular power, especially when expansionary
warfare (against Rome or a neighbor) was in the offing. Even after mas-
sive military defeats, such as that at Strasbourg in 357, which brought
down the Alamannic overking Chnodomarius, the confederation re-
tained its cohesion and could quickly reform under the leadership of
new overkings, of whom Rome faced a sequence in the course of the
fourth century. The durability of the larger political structures of the
fourth century marks them out their earlier counterparts.23
Equally important, the nature of political power had changed out of
all recognition. A much stronger hereditary element had invaded the
top end of politics. Among the Alamanni the overkingship tended not
to be hereditary, not least because Roman policy was geared toward
eliminating a succession of its holders. But the canton kings do seem
Frontier Defense 235
to have been hereditary, whereas royal status among the early Germani
(not all of whom recognized kings at all) was personal and could not
be easily transmitted to an heir. Among the Gothic Tervingi, further to
the east, even the position of confederation leader seems to have been
hereditary, being passed through three generations of the same fam-
ily.24 Entirely concomitant with this development, a new royal ideology
became current in the Germanic world between the early and later
Roman periods. By the fourth century, all the current terms then in
use for “king” were extensions of words meaning military commander.
In the early Roman period, by contrast, military leadership had often
been separate from kingship.25
That it was precisely this much-enhanced military role that lay at
the heart of these new kings’ hereditary power is suggested by the
range of evidence for the importance of military retinues in the late
Roman period. By the fourth century,
kings maintained personal mili-
tary support in the form of professional retainers. Ammianus Marcel-
linus mentions that Chnodomarius had his own force of 200 men, and
archaeological evidence of a destroyed warband also of some 200 men
has been excavated from Ejsbøl Mose, where the defeated group’s
weaponry was ritually interred. These kinds of professional military
forces were new to the Germanic world from the third century on,
and kings employed them not just to fight wars but also to act as en-
forcers of their policies.26
There is no doubt that this fundamental transformation of politi-
cal power in part reflected the long-term impact of all the wealth that
the empire had directed toward particular princes and kings among its
immediate neighbors in the many centuries of the empire’s existence.
Not only was this new wealth to the Germanic world, it was also not
received merely passively. Its arrival set off struggles for power among
Germanic political elites whose consequences show up particularly in
a body of evidence for internal political disturbance in the Germanic
world from the third century on. In this era, ritual deposits of weap-
ons, such as that found at Ejsbøl Mose, suddenly became reasonably
common, and exterior Germanic groups started to expand their power
toward the Roman frontier, precisely to seize some of the wealth con-
centrated there.27
236 Heather
Other transformations, of course, also played an important role
within this broader revolution. The early centuries AD saw the advent of
new farming regimes in Germanic-dominated Central Europe, which
generated large increases in food production and hence in population.
The overall power of the Germanic world, at least in demographic
terms, clearly increased in relation to its imperial Roman neighbor, and
the new kings presumably used some of this surplus food to support
their retinues. Again, Roman economic demand and transfers of Ro-
man technical know-how appear to have played a significant role in this
agricultural revolution, and in accompanying economic expansions in
some areas of manufacture and trade.28 I also strongly suspect that the
generally aggressive, not to say humiliating, nature of the Roman ap-
proach even to its favored client kings—where, after the fashion mas-
tered by the Sarmatian prince Zizais in the presence of Constantius
II in 358, groveling accompanied by gentle sobbing while begging for
favor was the generally favored protocol for barbarian kings in the im-
perial presence29—likewise played a major role in the ability of the new
class of hereditary military kings to build up their control. If this sce-
nario seems far-fetched, one need only remember that part of Roman
frontier policy was to burn down the villages of its neighbors once per
generation, and that several of the princes taken hostage clearly came
back none too enamored of the Roman way.30 Indeed, one gain that
might be had in return for paying the dues necessary to support the
new kings’ military retinues was surely the hope that belonging to a
powerful confederation of the new kind might help fend off the worst
effects of Roman imperial intrusion. In short, all the different kinds of
relationship—positive and negative, political and economic, diplomatic
and military—that had naturally sprung up between the empire and its
originally very much less developed neighbors combined to accelerate
the transformative processes that turned the large number of small so-
ciopolitical units occupying the imperial hinterland in the first century
AD into the much smaller number of more powerful ones that had re-
placed them by the fourth. And what gave these relationships so much
transformative power was the fact that each stimulated its own autono-
mous response among the Germani. It wasn’t just that imperial Rome
did things that transformed Germanic society—although it certainly
Frontier Defense 237
did—but that elements within Germanic society took full advantage
of the new opportunities that emerged from the various new relation-
ships with the empire to create dynamic new political structures.
By the mid-fourth century, the overal extent of these transforma-
tions had not yet reached obviously dangerous proportions. None of
the new units then in existence was a threat to overal imperial integrity.
At most, even the most ambitious barbarian military king of the fourth
century could hope to make only very limited territorial gains from the
empire, or—more usually—to limit the intrusiveness of Roman eco-
nomic or diplomatic demands. Chnodomarius was laboring in the 350s
to annex a perhaps 50-km-wide strip of Roman territory along part of
the Rhine, while the overkings of the Tervingi were trying to limit the
extent to which Roman emperors could demand recruits from them for
their wars, and to resist demands that Christian missionaries from the
now Christian empire be let loose in their lands. None of this threat-
ened imperial survival.31 Indeed, having a smal er number of diplomatic
partners to deal with perhaps simplified the operation of Roman fron-
tier management techniques, since there were fewer competing political
claims to be balanced out on the other side of the frontier. Where the
new order that Rome itself had inadvertently generated on its European
frontiers final y became a problem, however, was when an outside force
imparted an involuntary unity to a substantial number of these new and
larger Germanic sociopolitical units.
In the late fourth century and early fifth century, the Huns, a group
of Eurasian nomads who were probably attracted westward by the
amount of wealth that could be liberated from Rome’s frontier clients,
revolutionized the overall strategic situation on Rome’s European fron-
tiers. In two discrete phases separated by a generation, they first es-
tablished their dominance north of the Black Sea in the 370s, and then
shifted their center of operations into the Great Hungarian Plain at the
heart of Europe about the year 410. The first effect of each of these
moments of large-scale migration was to throw several of the larger
Germanic groups that had been generated in Rome’s frontier region
across the border onto imperial territory. This is explicitly documented
in the case of the Huns’ first move into the northern Black Sea region,
which pushed two separate large groups of Goths, the Tervingi and
238 Heather
the Greuthungi, and a series of smaller groups across Rome’s Lower
Danube frontier. The Huns’ second move, onto the Great Hungarian
Plain, was preceded by another huge exodus of Roman clients from
precisely that region: another large Gothic group led by a certain Rada-
gaisus, who moved into Italy, a large coalition comprising two separate
groups of Vandals together with Alans and Sueves, who moved across
the Rhine, and a further force of Burgundians, who followed them
in the same direction. The sources are not explicit that Huns caused
this second exodus
. But Huns appear in the region vacated by these
migrants immediately afterward, and the most likely explanation for
this unprecedented demographic upheaval is that it was a repeat of
the scenario of the 370s, but this time played out across Rome’s Middle
Danubian and Upper Rhine frontiers as the Huns moved westward.32
In effect, the Huns imparted a unity of purpose among many tens of
thousands of invaders that is hard to imagine occurring otherwise. And
it was the simultaneous appearance of these politically separate barbar-
ian groupings that prevented the Roman Empire from defeating them:
as it attempted to do, and as it certainly would have been able to do had
not so many come at one time.33
The scale of the resulting strategic disaster was made much worse
from the Roman point of view by the fact that the empire’s attempted
military counterstrikes prompted a further phase of alliance build-
ing among the migrants. Thus, from half a dozen or so separate units
that entered the empire in 376–80 and 405–8, there emerged two much
larger units. The Visigoths, who eventually settled in southern Gaul in
418, were composed of the Tervingi and Greuthungi of the first phase,
united further with surviving Goths from the attack of Radagaisus on
Italy in the second phase. The Vandal force that eventually seized the
economically vital lands of North Africa, breadbasket of the western
empire, in the 430s (after a lengthy interlude in Spain) were likewise cre-
ated by the full unification there of both Vandal groups and the Alans,
who originally outnumbered them.34 The crucial point is that these
further political reconfigurations created groups big enough to resist
even major Roman field armies. Hence the new groups were able to
survive on Roman soil in the long term, not least because the intrusive
Huns united several of the other frontier Germanic groups under their
Frontier Defense 239
control beyond the frontier and began to mount campaigns, which
meant that the maximum Roman force could not be deployed against
the original—now reorganized—migrants.35 And while the Hunnic
Empire proved only a ramshackle and temporary phenomenon, its col-
lapse only increased the problems facing the imperial authorities, as it
led several other Germanic groups comparable in scale to the Visigothic
Makers of Ancient Strategy: From the Persian Wars to the Fall of Rome Page 37