564 Walter Pohl
vaguest ones (such as the category “Germani,” imposed on the rather heterogeneous
inhabitants between the Rhine and the Vistula by Julius Caesar, or the antiquarian use
of “Scythians” for the steppe peoples of the northeast). It is therefore doubtful whether
“Germanic” self-identification ever mattered to anyone apart from some proud Roman
officers of barbarian origin along the Rhine frontier; and we can assume that the Huns did
not identify with the long-gone Scythians. On a more specific level, however, it seems
unlikely that Roman perceptions had nothing to do with reality. Of course, ethnogra-
phy could always blur the perceptions of foreigners. For instance, poetry and panegyrics
sometimes offer extensive catalogues of long-gone or fantastic peoples that an emperor
is supposed to have defeated. However, such ethnographic fantasies are easy to detect,
precisely because matter-of-fact military historiography is much more pragmatic about
naming Rome’s enemies and allies. Roman designations (e.g., for its auxiliary units)
surely had their share in promoting certain self-identifications among barbarians at the
expense of others. However, the success of the Goths most likely relied on people proud
to be Goths.
There is a further observation supporting the relative significance of ethnic identities
for the migrant warrior groups of the fifth and sixth centuries. Between the battle of
Adrianople and their settlement in Aquitaine, the Visigoths went through a series of
hardships, defeats, failed negotiations, internal conflicts, and premature deaths of rulers.
Still, they essentially maintained their cohesion, and attracted further Gothic and other
groups on the march. The armies and retinues of Roman commanders, on the other
hand, were disbanded or transferred after the death of their leaders. Frequently, they also
overthrew and killed their commanders at the first sign of adversity, and joined another
competitor. Many usurpers who had counted on the loyalty of their army ended in this
fashion. A possible explanation is found in the speech that Procopius puts into the mouth
of the Ostrogothic king Totila before the fatal battle at the Busta Gallorum in 552. Here,
the Goth belittles the superior size of the Roman army:
The vast number of the enemy is worthy only to be despised, seeing that they present a
collection of men from the greatest possible number of nations. For an alliance which is
patched together from many sources gives no firm assurance of either loyalty (pistis)orpower,
but being split up in nationality, it is naturally divided likewise in purpose (Procopius 1978,
VIII, 17–18, 30, p. 367).In a potentially adverse environment, Gothic armies must have relied on a shared sense of
identity if they wanted to maintain some autonomy, otherwise they would have eroded
quickly, following the lure of Roman money, as many Goths had always done. Memories
of shared adventures surely inspired such a sense of loyalty, but more distant origins may
also have helped.
Traditions, Myths, and Memories
What significance, then, did myths and memories have for Gothic identity? Admittedly,
this is a largely speculative field. Wenskus had based his notion of “tradition” not least
on theorigines gentium, narratives about the origins of the Goths, Longobards, and