498 Ursula Rothe
way ethnic identities were formulated. Yet, the dynamics of cultural groupings could be
idiosyncratic, and some localities reacted in very different ways to others on the basis of
what seem to us, with our limited source-base, to have been identical preconditions.
It is impossible, within the confines of this chapter, to present a comprehensive account
of ethnic developments in the Roman northwest, nor can I do justice to the enormous
amount of scholarly work that has focused on precisely this question, particularly in the
past 20–30 years. What this chapter aims to do is to present, using selected case stud-
ies, a range of different scenarios that give some insight into the complex and varied
nature of ethnicity and ethnogenesis in this region in the first–third centuriesAD.As
the most personal, and at the same time most public, form of material culture, dress
stands somewhat apart as a particularly valuable source for ethnic identity, providing,
of course, we have the sources for it. It is not a coincidence that Tacitus mentions the
toga in his canonical inventory of Roman cultural traits. As Andrew Wallace-Hadrill has
recently shown (2008: 38–70), the toga played a similar membership-defining role in
Roman consciousness as language did in the Greek, with one important difference: as
the dress of the Roman citizen, the toga expressed more than any other element of cul-
ture the civil and legal nature of Roman identity. It is this characteristic that makes it
so hard to equate “being Roman” with anything amounting to ethnicity. However, it
is also this characteristic which meant that practically anyone in the empire, regardless
of their background, could become a Roman citizen, and as such, a “Roman” (see Far-
ney, Chapter 29). The scenario Tacitus describes is not entirely imaginary: the toga was
indeed worn in wide parts of the Roman northwest; but so was native dress. The very dif-
ferent nature of Roman identity to more localized group identities means that this is no
paradox. Dress, in fact, reflects especially well the varying layers of cultural identity a per-
son living in Rome’s northwest provinces could and did have. As a result, in some of the
case studies that follow, it will play a central role as a constant for identifying similarities
and differences.
The Batavi
The Batavians were a Germanic tribe that broke away from the larger Transrhenane Chatti
group and were allowed by the Romans to settle in the eastern Rhine delta in the mid–late
first centuryBC(Tacitus,Germania29;Histories4.12; Tacitus’ account is supported by
archaeological evidence: Roymans 2004: 67–96; 103–32). From 12BConward, the
Batavi had a specialfoeduswith the Roman state which stipulated that they were freed
from regular taxes, but instead had to supply a large number of auxiliary troops to the
Roman army, and a large proportion of the soldiers in the emperor’s German bodyguard
at Rome. By the mid-first centuryAD,there were nine Batavian auxiliary units, in all over
4,000 soldiers, not including the imperial bodyguard. This meant that every Batavian
family is likely to have had at least one family member in the Roman army (Willems
1984: 395). Since the 1980s, Dutch archaeologists have produced methodologically
groundbreaking work on this group of people, in particular concerning the effects of
the exceptional recruitment arrangements on their social structures and ethnic identity.
The group that came to be known as the Batavi—a Chattian prince and his warrior
retinue—migrated to the lower Rhine some time shortly after the Gallic War, and were