Goths and Huns 557(although it also translates to the Latin “populus,” which was used for the Roman people
and underlines political, not ethnic aggregation; see Geary 2002). Traditional scholar-
ship had not questioned the homogeneity and durability of these peoples. Ethnicity was
seen as defined by common ancestry, common language, and common culture, and this
is how the history of the migration period has traditionally been told. Methodologically,
this opened many possibilities to extend the range of the patchy evidence we have. If a
written source seemed to attest the presence of Goths in a certain region, the archae-
ological evidence of that region had to be Gothic (and allowed to circumscribe their
settlement area and/or the extension of their rule), and one could safely assume that this
population must have spoken Gothic. One could even go further: as theirs was a Ger-
manic language, their culture and social structure (and, for many, their Nordic race) also
had to be Germanic. Therefore, in spite of the dearth of other clues, one could safely use
the evidence from twelfth/thirteenth-century Scandinavia or Iceland to find out more
about (for instance) Gothic law or religion. This resulted in the lively and colorful rep-
resentations of “the Germans” (or “the Teutons,” as they were also called in English)
that scholars took for granted in the nineteenth and for much of the twentieth cen-
tury (Pohl 2002a). Similar approaches prevailed for the Huns and other steppe-peoples,
with an important difference: it was difficult to judge the language group to which they
belonged, and thus they were variously regarded as Turkic, Mongolian, Iranian, or other,
with all the baggage of interpretation that came with such classifications.
In the second half of the twentieth century, these seemingly natural views of the peoples
of Late Antiquity ran into two problems. One was the general development of theories
of ethnicity in the social sciences and the humanities (e.g., Barth 1969 or Smith 1986),
which (although somewhat patchily and eclectically) gradually permeated historical stud-
ies. The other was the erosion of the paradigm of common blood in the study of Late
Antiquity and the Early Middle Ages itself. It turned out that the evidence had to be
stretched considerably to support anything such as the natural ethnic categorizations
that had dominated older research. To give just one example: after the third centuryAD
(when the Goths moved to Eastern Europe), no contemporary observer regarded them
as Germanic. Although most Goths still spoke a Germanic language (a most valuable
piece of evidence is the Gothic translation of the Bible produced by Bishop Wulfila in
the fourth century), their material culture had become strongly influenced by eastern
European steppe culture. Consequently, contemporaries identified the Goths with the
ancient Scythians orGetae, steppe peoples of the first millenniumBC. What is more,
the Roman ethnographic categoryGermanialmost completely ceased to be used for
any contemporary barbarians after the fourth centuryAD. These (in our scholarly ter-
minology) “Germans” simply did not have enough in common to suggest the use of a
general label.
These, and many other observations, have led to a change of paradigm, but also to
heated debates, many of which are still going on. A few basic points now seem to be
shared by most scholars working in the field. In line with the recent theory of ethnic-
ity, it has become clear that ethnicity is not necessarily based on real shared ancestry,
but only on the subjective notion of common origins. It is not biologically or geneti-
cally but culturally defined. There are numerous instances in the sources for significant
changes in the composition of ethnic groups. For instance, the Germanic-speaking rul-
ing minority of theFranci(Franks) in sixth-century Gaul became, over several centuries,