WALTER POHL/MATHIAS MEHOFER (HG.

ARCHAEOLOGY OF IDENTITY – ARCHÄOLOGIE DER IDENTITÄT

ÖSTERREICHISCHE AKADEMIE DER WISSENSCHAFTEN
P H I L O S O P H I S C H - H I S TO R I S C H E K L A S S E
DENKSCHRIFTEN, 406. BAND

FORSCHUNGEN ZUR GESCHICHTE DES MITTELALTERS

BAND 17

HERAUSGEGEBEN VOM
INSTITUT FÜR MITTELALTERFORSCHUNG
ÖSTERREICHISCHE AKADEMIE DER WISSENSCHAFTEN
P H I L O S O P H I S C H - H I S TO R I S C H E K L A S S E
DENKSCHRIFTEN, 406. BAND

FORSCHUNGEN ZUR GESCHICHTE DES MITTELALTERS

BAND 17

Archaeology of Identity –
Archäologie der Identität
HERAUSGEGEBEN VON

WALTER POHL / MATHIAS MEHOFER

 Vorgelegt von w. M. WALTER POHL in der Sitzung am 20.Juni 2008

Gedruckt mit Unterstützung durch den

Fonds zur Förderung der wissenschaftlichen Forschung

Umschlaggestaltung: Dagmar Giesriegl

British Library Cataloguing in Publication data

A Catalogue record of this book is available from the British Library

Die verwendete Papiersorte ist aus chlorfrei gebleichtem Zellstoff hergestellt,

frei von säurebildenden Bestandteilen und alterungsbeständig.

Alle Rechte vorbehalten

ISBN 978-3-7001-6502-6
Copyright © 2010 by
Österreichische Akademie der Wissenschaften
Wien
Satz und Layout: Gerald Reisenbauer, ÖAW
Druck und Bindung: Ferdinand Berger & Söhne Ges.m.b.H., A-3580 Horn
Printed and bound in the EU
http://hw.oeaw.ac.at/6502-6
http://verlag.oeaw.ac.at
 Inhaltsverzeichnis
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
WALTER POHL, Archaeology of identity: introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
SEBASTIAN BRATHER, Bestattungen und Identitäten – Gruppierungen innerhalb frühmittelalterlicher
Gesellschaften . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
PHILIPP VON RUMMEL, Gotisch, barbarisch oder römisch? Methodologische Überlegungen zur ethni-
schen Interpretation von Kleidung . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
JÖRG KLEEMANN, Frühmittelalterliche Bestattungen als Projektionen kontextueller Identitäten. . . . . 79
DIETER QUAST, Die Langobarden in Mähren und im nördlichen Niederösterreich – ein Diskussions-
beitrag . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
PETER STADLER, Ethnische Gruppen im Awarenreich. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
CSANÁD BÁLINT, A contribution to research on ethnicity: a view from and on the east. . . . . . . . . . . . 145
IRENE BARBIERA, The valorous Barbarian, the migrating Slav and the indigenous peoples of the
mountains. Archeological research and the changing faces of Italian identity in the 20th century . . . . . . 183
PRZEMYSAW URBACZYK, Before the Poles: problems of ethnic identification in Polish
archaeology of the Early Middle Ages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
HUBERT FEHR, Am Anfang war das Volk? Die Entstehung der bajuwarischen Identität als archäolo-
gisches und interdisziplinäres Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
BONNIE EFFROS, Anthropology and ancestry in nineteenth-century France: craniometric profiles of
Merovingian-period populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
PIERO MAJOCCHI, The treasure of Theodelinda: ideological claims and political contingencies in
the construction of a myth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
CRISTINA LA ROCCA – STEFANO GASPARRI, Forging an early medieval royal couple: Agilulf,
Theodelinda and the ‘Lombard Treasure’ (1888–1932) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
Quellenverzeichnis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
Literaturverzeichnis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
 Preface
This volume is based on an international conference held in Vienna in January 2006. The conference was
organised by the Institut für Mittelalterforschung der Österreichischen Akademie der Wissenschaften, with
the cooperation of the VIAS – Vienna Institute for Archaeological Science of the University Vienna, Institut
für Ur- und Frühgeschichte der Universität Wien, and financed to a large extent from the Wittgenstein-Prize
project “Ethnic Identitities in Early Medieval Europe”. This project was made possible by the award of the
Wittgenstein Prize in 2004 by the Fonds zur Förderung der wissenschaftlichen Forschung in Österreich
(FWF). Thanks must therefore go to all the institutions involved. But we would also like to thank the many
young scholars who helped with organising the conference and editing the volume, among them, Richard Cor-
radini, Max Diesenberger, Nicola Edelmann, Clemens Gantner, Gerda Heydemann, Dana Polanichka, Mari-
anne Pollheimer, Helmut Reimitz, Wofgang Breibert, Marion Grossmann and Veronika Wieser. As always,
it has been a pleasure working with them, and hosting many international guests and inspiring discussions in
Vienna. We would like to thank Dagmar Giesriegl, who designed the cover layout, and Hannes Weinberger
and Gerald Reisenbauer (Verlag der Österreichischen Akademie der Wissenschaften), who were responsible
for the printing of the book.

Mathias Mehofer Walter Pohl

 WA L T E R P O H L

Archaeology of identity: introduction

Can there be an “archaeology of identity”? The title of this volume circumscribes a contested field of research.
In particular, it is related to the debate about the ethnic interpretation of archaeological evidence.1 This has
been a topic of controversy in Germany in recent years, following Sebastian Brather’s book “Ethnische Inter-
pretationen in der frühgeschichtlichen Archäologie” and some of his other publications.2 However, the present
volume is not simply intended to continue the debate. Rather, it discusses the material traces of past identities
in a broader sense, of which the question of ethnic interpretation is just one, if important, aspect, and reflects
on how archaeological evidence can be inserted into a general picture of the ‘migration age’. Three related
questions have emerged in recent discussions. Firstly, how can we classify archaeological groups or ‘cultures’,
and to what degree do these correspond to ethnic (or religious, social or other) identities of the past? Secondly,
can traces of past self-identification be found in material remains? And thirdly, in what ways have modern
national or ethnic identifications shaped such interpretations in the history of research? All three questions
are essential, but methodologically, they should be kept apart. They specifically concern the interdisciplinary
communication between archaeologists and historians. Therefore, most of the authors of this volume are to
some degree familiar with both disciplines. Several contributions (Csanád Bálint, Irene Barbiera, Bonnie
Effros, Cristina La Rocca/Stefano Gasparri, Piero Majocchi and Premysav Urbaczyk) deal with the chang-
ing interpretation of archaeological evidence in the context of modern identity constructions, and thus add a
valuable dimension to contemporary debates.3 Others take an exemplary look at methodologically sensitive
research questions mainly concerning the 5th and 6th centuries (Hubert Fehr, Jörg Kleemann, Dieter Quast, and
Peter Stadler). The picture is completed by two methodological contributions from members of the ‘Freiburg
school’ (Sebastian Brather and Philipp von Rummel).4

1
See, for instance, Siân Jones, The Archaeology of Ethnicity. Constructing Identities in the Past and Present (London/New York
1997); Cultural Identity and Archaeology. The Construction of European Communities, ed. Paul Graves-Brown/Siân Jones/Clive
Gamble (London/New York 1996); Lynn Meskell, Archaeologies of identity, in: Archaeological Theory Today, ed. Ian Hodder
(Cambridge/Oxford/Malden 2001) 187–213. I am grateful to Nicola Edelmann and Maya Maskarinec for corrections of this arti-
cle, and to Helmut Reimitz and Pavlina Rychterová for suggestions.
2
Sebastian Brather, Ethnische Identitäten als Konstrukte der frühgeschichtlichen Archäologie, in: Germania 78 (2000) 139–177;
id., Ethnische Interpretationen in der frühgeschichtlichen Archäologie. Geschichte, Grundlagen und Alternativen (RGA Erg. Bd.
42, Berlin/New York 2004); for a different opinion, Volker Bierbrauer, Zur ethnischen Interpretation in der frühgeschichtlichen
Archäologie, in: Die Suche nach den Ursprüngen. Von der Bedeutung des frühen Mittelalters, ed. Walter Pohl (Forschungen zur
Geschichte des Mittelalters 8, Wien 2004) 45–84, and id., Ethnos und Mobilität im 5. Jahrhundert aus archäologischer Sicht: Vom
Kaukasus nach Niederösterreich (Abhandlungen der Bayerischen Akademie der Wissenschaften N.F. 131, München 2008). See
also Claus von Carnap-Bornheim, Hans-Jürgen Eggers und der Weg aus der Sackgasse der ethnischen Deutung, in: Eine hervor-
ragend nationale Wissenschaft. Deutsche Prähistoriker zwischen 1900 und 1995, ed. Heiko Steuer (RGA Erg. Bd. 29, Berlin/
New York 2001) 173–198; Soziale Gruppen – kulturelle Grenzen. Die Interpretation sozialer Identitäten in der Prähistorischen
Archäologie, ed. Stefan Burmeister/Nils Müller-Scheeßel (Tübinger Archäologische Taschenbücher 5, Münster 2006). See also
the contribution by Philipp von Rummel, in this volume.
3
See also the volumes Vergangenheit und Vergegenwärtigung, ed. Helmut Reimitz/Bernhard Zeller (Forschungen zur Geschichte
des Mittelalters 14, Wien 2009); Gebrauch und Missbrauch des Mittelalters, 19.–21. Jahrhundert/Uses an Abuses of the Middle
Ages: 19th–21st Century/Usages et Mésusages du Moyen Age du XIX au XXI siècle, ed. János M. Bak/Jörg Jarnut/Pierre Monnet/
Bernd Schneidmüller (MittelalterStudien 17, Paderborn 2009); Nationalism and Archaeology, ed. John A. Atkinson/Iain Banks/
Jerry O’Sullivan (Glasgow 1996); Heinrich Härke, Archaeologists and migrations: A problem of attitude?, in: Current Anthro-
pology 39 (1998) 19–45; Eine hervorragend nationale Wissenschaft. Deutsche Prähistoriker zwischen 1900 und 1995, ed. Heiko
Steuer (RGA Erg. Bd. 29, Berlin/New York 2001); Inventing the Pasts in North Central Europe. The National Perception of Early
Medieval History and Archaeology, ed. Matthias Hardt/Christian Lübke/Dittmar Schorkowitz (Bern 2003); and themed edition:
Early Medieval Material Culture in the Nineteenth- and Twentieth-Century Imagination, ed. Bonnie Effros/Howard Williams,
in: Early Medieval Europe 16, 1 (2008) 1–88. For the history of “culture-historical archaeology” and its nationalist uses (includ-
ing Asian examples), see Bruce G. Trigger, A History of Archaeological Thought (Cambridge 1989) 148–206.
4
Unfortunately, Volker Bierbrauer and Guy Halsall could not attend the workshop, whereas Florin Curta and Falko Daim chose
not to publish their contributions.
10 Walter Pohl

IDENTITY, ETHNICITY
The word ‘identity’ has recently become common in archaeological debates, and has appeared in the title of
two conferences held in 2005, one in Freiburg entitled “Gräber, Siedlungen und Identitäten. Das 4.–7. Jahr-
hundert im Westen”,5 and the other at Mikulov in Moravia, co-hosted by the Prähistorische Kommission of
the Austrian Academy, “Barbaren im mittleren Donauraum – Zur Veränderung der Identitätsphänomene”.6
Identity is a broader term than ethnicity; we could try to describe the identities of, for instance, Roman offic-
ers, pagan senators, Egyptian hermits, steppe shamans, women in barbarian warrior families along the Elbe
or the urban elites of metropolitan cities in the western Mediterranean. Few of these identifications are as
clearly labeled as ethnic identities, which connect a – real or notional – group with a name. Then and now, this
makes a fundamental difference, because the group thus named can become a topic for political, affective or
scholarly communication, and thus acquire a currency that unnamed social groups, networks or communities
do not have. But, again then and now, the seeming validity of the ethnonym may be misleading. What is be-
ing perceived as a solid group by outside observers need not correspond to any strong self-identification, and
may simply represent an ethnographic or political classification using outdated or misconceived ethnic names
to describe a much more complex reality. This creates a methodological problem for early medieval studies
because most of our written sources represent outside views, not clear self-identifications. We can infer that
people who repeatedly act as a group also identify with it. But we need to be aware that a name does not cor-
respond automatically with a group (as the term ‘ethnicity’ suggests), but that the identification has to be made
plausible in each case (an approach that the term ‘identity’ allows for).
‘Identity’ is perhaps a problematic term.7 It may be understood in rather different ways, as static or dy-
namic, as objective or subjective, as social or individual, as factual or as constructed.8 As a complex category,
it certainly is difficult, if not impossible to define in a satisfactory way (which has often been held against it).
There is in fact a danger that it could be used in a reductive and affirmative manner. Claims of past identities
may be part of deliberate or implicit politics of identity that seek to establish certain sets of cultural values and
social norms across a rather varied population: for instance, the attempts to argue that Europe has a Christian
identity, that the Bosnian Muslims are in fact Serbs, or that the Turks are (or are not) Europeans.9 The term
‘identity’ can thus be used as an ideological tool. That is of course also true of many terms that describe hu-
man groupings, such as race, ethnicity, nation, people, tribe, class, elite and others, most of which are hard to
dispense of when writing history. But thus critical reflection becomes all the more necessary.
The success of ‘identity’ both in the scholarly and the public spheres is after all due to its explanatory po-
tential. If we say that individual X belongs to group A, this is a straightforward statement that can be tested
against the perceptions of X himself, of the ‘ingroup’ A, of the others (the ‘outgroup’), and against their re-
spective behaviour.10 At the same time, we establish a binary logic (does/does not belong) that simplifies the
many ambiguities of social groupings, and makes it very hard to decide who belongs where if our evidence
is patchy. If, however, we assume that X has an ‘A’-identity, we say both more and less than that. We claim,
on the one hand, that being an A is an important part of X’s personality and thus link him/her with a cluster

5
Gräber, Siedlungen und Identitäten. Archäologie des 4.–7. Jahrhunderts im Westen, ed. Sebastian Brather (RGA Erg. Bd. 57,
Berlin/New York 2008).
6
The conference in December 2005 was part of the series “Probleme der Archäologie des mittleren Donauraumes”; the proceed-
ings will be published.
7
For a fundamental critique of the use of the term in the humanities and social sciences, see Lutz Niethammer, Kollektive Iden-
tität. Heimliche Quellen einer unheimlichen Konjunktur (Reinbek 2000); cf. also Francesco Remotti, Contro l’identità (Roma/
Bari 2001), which in spite of its title does employ identity as an analytical category. See also Identitäten, ed. Aleida Assmann/
Heidrun Friese (Frankfurt a. M. 1998); and the useful synthesis by Antonia Davidovic, Identität – ein unscharfer Begriff. Iden-
titätsdiskurse in den gegenwartsbezogenen Humanwissenschaften, in: Soziale Gruppen – kulturelle Grenzen. Die Interpreta-
tion sozialer Identitäten in der Prähistorischen Archäologie, ed. Stefan Burmeister/Nils Müller-Scheeßel (Münster etc. 2006)
39–58.
8
Walter Pohl, Identität und Widerspruch. Gedanken zu einer Sinngeschichte des Frühmittelalters, in: Die Suche nach den Ur-
sprüngen. Von der Bedeutung des frühen Mittelalters, ed. Walter Pohl (Forschungen zur Geschichte des Mittelalters 8, Wien
2004) 23–36; id., Aux origines d’une Europe ethnique: Identités en transformation entre antiquité et moyen âge, in: Annales:
Histoire, Sciences sociales 60, 1 (2005) 183–208.
9
Cf. Mythen der Nationen. Ein europäisches Panorama, ed. Monika Flacke (München/Berlin 1998); Patrick J. Geary, The Myth
of Nations: The Medieval Origins of Europe (Princeton/Oxford 2001).
10
A good introduction into social identity theory in the social sciences is Social Identity Processes, ed. Dora Capozza/Rupert
Brown (London/Thousand Oaks 2000).
 Archaeology of identity: introduction 11
of identity-relevant historical, social, cultural and political phenomena. On the other hand, the term ‘identity’
is much more open and flexible than the logic of belonging. It implies that an individual may be related to
several groups to different degrees (or ‘salience’), and that the identity of these groups in turn may be rather
complex and dynamic.11 In that sense (and not in any essentialist or static sense), ‘identity’ seems to be very
appropriate for the communication between historians and archaeologists, provided that it is used in a critical
and methodologically sound manner. It is possible to detect stronger or weaker signs of (for instance) Gothic
or Christian identity in burials, without having to claim that an individual (or a population) actually is a Goth
or a Christian.
What ‘identity’ describes is the interface between the individual and the social group. Sociologists distin-
guish between individual and collective (or better, social) identities.12 But it is important that we realize the dy-
namic relationship between both. Individuals are hardly ever only idiosyncratic, but are also defined by their
social roles and relationships to others. Certainly this was the case in the pre-modern period when personal
identity was to a considerable extent created with reference to social groups and cultural models (although
not simply imposed by them).13 More than historical research based on written sources, archaeology has the
opportunity to find the traces of a great number of individuals through their funerary remains, and to hypoth-
esize about their ‘personhood’ in the context of exchanges and ‘dividuality’.14 But only shared characteristics
allow for the reconstruction of a more general picture. There are countless ways of showing that one belongs,
but they will only be recognizable if they occur frequently.
On the other hand, it would be too simple to assume that identity is a straightforward expression of peoples’
sense of belonging to a given ethnic group. Richard Jenkins warns against “our tendency to reify ethnicity”:
“Although they are talked about endlessly in these terms, neither ethnicity nor culture is ‘something’ that peo-
ple ‘have’, or, indeed, to which they ‘belong’”.15 Rogers Brubaker has called this reifying approach “groupism”:
“The tendency to take bounded groups as fundamental units of analysis, and basic constituents of the social
world.”16 Identity, and especially ethnic identity, does not necessarily have to refer to a well-delineated social
group with a definite number of members. It also “works through categories, schemes, encounters, identifica-
tions, languages, stories, institutions, organizations, networks and events.”17 It may be that in many cases all
of these (and many more) elements coincide and contribute to shaping a basically coherent and stable ethnic
identity. Perhaps there is such a thing as the typical Englishman or Spaniard. But in the study of past identities,
we can never automatically infer one element from the other. For instance, someone who identified himself as
a Hun in an encounter with Byzantine diplomats may have spoken Gothic as his native language, spent much
of his life serving as a bodyguard for Aetius, then fought for Attila, used a rather eclectic but typical ‘mode
danubienne’ outfit and have been regarded as a Scythian by the Byzantines. We may decide, not unreasonably,
to classify him as a Hun in our scholarly reconstruction of the period, or perhaps, as a Gothic member of the
Hunnic warrior elite. But we should not make too far-reaching assumptions about the coherence of either of
these identities and their impact. The term ‘identity’ has acquired the necessary flexibility to describe such
dynamic identifications. But for that purpose, we have to acknowledge, quite paradoxically, that identities are
never identical. Even our typical Englishman may feel more British than English at times, have an Irish grand-

11
In the social sciences, the ‘self-categorization’ and the ‘comparative identity’ approaches have abandoned binary concepts (in-
group vs. outgroup) to study the “degrees of identification with two” (or more) “categories at different levels of inclusion”, for
instance, regional/national identifications: María Ros/Carmen Huici/Angel Gómez, Comparative identity, category salience and
intergroup relations, in: Social Identity Processes, ed. Dora Capozza/Rupert Brown (London/Thousand Oaks 2000) 81–95; So-
cial Identity and Intergroup Relations, ed. Henri Tajfel (Cambridge 1982); Rediscovering the Social Group: A Self-Categorization
Theory, ed. John C. Turner et al. (Oxford 1987).
12
See, among others, Anthony D. Smith, The Ethnic Origins of Nations (London 1986); Ethnic Identity. Creation, Conflict and Ac-
commodation, ed. Lola Romanucci-Ross/George A. de Vos (Walnut Creek/London 1995); Questions of Cultural Identity, ed.
Stuart Hall/Paul du Gay (London/Thousand Oaks 1996).
13
Walter Pohl, Ego trouble: Introduction, in: Ego Trouble. Authors and their Identities in the Early Middle Ages, ed. Richard Cor-
radini/Matthew Gillis/Rosamond McKitterick/Irene van Renswoude (Forschungen zur Geschichte des Mittelalters 15, Wien
2010) 9–22.
14
Chris Fowler, The Archaeology of Personhood (London 2004); Jos Bazelmans, Beyond power. Cermenonial exchanges in Be-
owulf, in: Rituals of Power from Late Antiquity to the Early Middle Ages, ed. Frans Theuws/Janet Nelson (The Transformation
of the Roman World 6, Leiden/Boston/Köln 1999) 311–375; Howard Williams, Review article: Rethinking early medieval mortu-
ary archaeology, in: Early Medieval Europe 13, 2 (2005) 195–218, at 209.
15
Richard Jenkins, Rethinking Ethnicity. Arguments and Explorations (Los Angeles 22008).
16
Rogers Brubaker, Ethnicity without Groups (Cambridge, Mass./London 2004) 2.
17
Brubaker, Ethnicity 3.
12 Walter Pohl

mother, be born in a colony, prefer Balti to Sunday roast or have a French wife.18 Among the many identifica-
tions that correspond with an ethnic (or national) identity, there are usually some that do not coincide.
In recent scholarly literature, especially in English, the variability and complexity of ethnic identities is
often rendered through the shorthand term ‘socially/culturally constructed’.19 Basically, this refers to the clas-
sical sociological insight that all social reality is constructed by human practice and perception according to
cultural models.20 But the formula has acquired a subtext that goes beyond that. Either it presumes that ethnic-
ity was completely malleable and evanescent, and/or it corresponds to a general sentiment that ethnic identity
did not matter much in history.21 Sometimes this is complemented by a deep-rooted scepticism that early medi-
eval sources cannot tell us much about complex phenomena such as ethnicity anyway.22 To claim that ethnicity
did not matter much is a comprehensible reaction to the misuse of ethnic identities in the history of the 20th
century.23 It also reflects the modern experience of nationalist politics: the conscious and sometimes cynical
attempt to manipulate ethnic identities as a means of buttressing and expanding the power of a ‘national’ elite.
Along these lines, Patrick Amory has represented Ostrogothic identity in Italy as “ethnographic ideology” ma-
nipulated by its rulers to guarantee the privileges of a governing military caste. “Such an evanescent identity
cannot be called ethnic, unless we wish to make ethnicity such a flexible term as to be almost meaningless”,
is what he concluded.24 But if being classed as a Goth in Theoderic’s Italy conferred privilege, why should we
assume that these Gothic soldiers and their families did not ‘really’ think of themselves as Goths, but only
pretended? In the sales document P. Ital. 34 from 551, the clerics of an Arian church in Ravenna (under Byzan-
tine rule since 540) sell off church possessions in an evident hurry; but they subscribe in the Gothic language
and call their church ecclesia gotica, basilica Gothorum, and ecclesia legis Gothorum, although at this point
it was hardly politically convenient to make such statements of Gothic and Arian identity.25 They were about
to lose their privilege, and it would have been more expedient to avoid all hints of ‘gothicness’. Late antique
and early medieval ethnic identities were flexible to differing degrees, but certainly not fluid, purely arbitrary
or meaningless. We have a number of cases where people maintained them when it would have been more
profitable to join the winners.
The fact that ethnic identities were socially constructed does not mean at all that they did not exist. Oth-
erwise, why are our sources filled with them? Why did contemporary historiography consistently attribute
political agency to gentes, peoples?26 The Huns plunder, the Franks send an embassy or the Lombards raise a
king. Ethnic groups act, they engage in politics and go to war. Traditionally, no scholar even felt the need to
explain this, since modern historiography largely agreed that it was nations and peoples that made history. But
perhaps, it would be more accurate to say that history was made in their name. Late antique and early medieval
peoples were in no way the stable and self-conscious ethnic groups that traditional historiography thought

18
Cf. Robert Colls, Identity of England (Oxford 2002).
19
For instance, in Florin Curta, Some remarks on ethnicity in medieval archaeology, in: Early Medieval Europe 15, 2 (2007)
159–185, at 184.
20
Peter L. Berger/Thomas Luckmann, The Social Construction of Reality: A Treatise in the Sociology of Knowledge (Garden City,
New York 1966). See also Pierre Bourdieu, Outline of a Theory of Practice (Cambridge 1977).
21
A strong statement to that effect: Walter Goffart, Barbarian Tides. The Migration Age and the Later Roman Empire (Philadelphia
2006) 1–12, who, however, polemicizes against a rather outdated form of ‘ethnogenesis theory’.
22
An interesting thought-piece along those lines is Timothy Reuter, Whose race, whose ethnicity? Recent medievalists’ discus-
sions of identity, in: id., Medieval Polities and Modern Mentalities, ed. Janet L. Nelson (Cambridge 2006) 100–109: “In the first
millennium, the definition of others does not appear to have much to do with the definition of self” (101). It is of course crucial
to distinguish self-identification and identification by others. But where we have statements of ethnic self-identification they cor-
respond, as a rule, to the perceptions of other sources.
23
Nationalism as a transitory stage in history: Eric J. Hobsbawm, Nations and Nationalism (London 1996); as constructing largely
imagined communities: Benedict Anderson, Imagined Communities. Reflections on the Origin and Spread of Nationalism (Lon-
don/New York 21991); The Invention of Tradition, ed. Eric J. Hobsbawm/Terence Ranger (Cambridge 1983).
24
Patrick Amory, People and Identity in Ostrogothic Italy, 489–554 (Cambridge 1997) 317. For different views, see Herwig Wolf-
ram, Die Goten. Von den Anfängen bis zur Mitte des 6. Jahrhunderts (Wien/München 31990); Peter J. Heather, The Goths (Oxford/
Cambridge, Mass. 1996); Walter Pohl, Die Völkerwanderung. Eroberung und Integration (Stuttgart/Berlin/Köln 22005) 13–30;
126–151.
25
P. Ital. 34, ed. Jan-Olof Tjäder, Die nichtliterarischen lateinischen Papyri Italiens aus der Zeit 445–700. 2 vols. (Lund/Stockholm
1955/82) 2, 102.
26
Walter Pohl, Regnum und gens, in: Der frühmittelalterliche Staat – europäische Perspektiven, ed. Walter Pohl/Veronika Wieser
(Forschungen zur Geschichte des Mittelalters 19, Wien, in print) 439– 450; id., Die ethnische Wende des Frühmittelalters und
ihre Auswirkungen auf Ostmitteleuropa. Oskar-Halecki-Vorlesung – Jahresvorlesung des Geisteswissenschaftlichen Zentrums
Geschichte und Kultur Ostmitteleuropas 2006 (Leipzig, forthcoming).
 Archaeology of identity: introduction 13
them to be. Therefore, it is all the more striking that all relevant post-Roman kingdoms were named after
peoples.27 This was the result of historical change from an empire governed by a ruling elite whose origins
mattered relatively little as long as they shared certain cultural codes.28 The elites of the Gothic or Frankish
kingdoms could also come from quite different backgrounds. But they relied on tribal solidarities, and their
identities were much more distinctive than the imperial Roman identity. Our written sources do not allow for
much to be said about these identities when the barbarians took power, but eventually, ethnic self-identification
and myths of common origin appear more clearly. Ethnic identities were not simply ‘there’; they were the re-
sult of identifications, of concrete action, social knowledge and of ‘strategies of distinction’.29
What we can grasp in the written sources are not the ‘natural’ affiliations of individuals to social groups
and their inborn sense of belonging. Rather, the texts represent the political and cognitive uses made of ethnic
identifications, negotiations for identity and loyalty, and (on the Roman side) fears of the dynamic of barbar-
ian aggregation. Histories project the ethnic groups of the present into a glorious or barbarian past; law-codes
prescribe what members of a people or a kingdom should and should not do; ethnographical passages sketch
a broader ethnic landscape; charters confer privileges in the name of a king or his nobles; Christian liturgy
and church councils help to give providential meaning to the kingdom; and so on. These texts are often more
projective than reflective, they offer identifications and contribute to constructing a meaningful social world.
Generally, the early-medieval texts are much more concerned with defining and delineating Christian identi-
ties than ethnic ones. But at the same time, they frame a political landscape that is formed by peoples which
also have an essential role to play in the history of salvation. What exactly it meant for an individual to be a
Goth in Theoderic the Great’s, or a Frank in Clovis’s, kingdom is difficult to establish. But we know that it
made a difference for many contemporary sources. Most of the texts may not have created affiliations; but they
made sense of them and gave them meanings. In this respect, many texts written from the late fourth century
onwards were part of the effort to give shape and significance to a new order in the Roman provinces in which
ethnic identities played a conspicuous role.
What we can observe in the development of textual witnesses from Late Antiquity to the Early Middle
Ages is a gradual change of perspective. To classical authors, ethnicity had been a quality of ‘the Other’, the
barbarians. According to Aristotle, the Greeks lived in the polis, whereas the barbarians belonged to ethn,
tribes or peoples.30 In the sixth century, Latin authors writing in the post-Roman regna (and often from impor-
tant political positions or in close contact with the courts of the new rulers), or from close acquaintance with
them, began to modify this perspective: Cassiodorus, Jordanes, Gregory of Tours or Isidore of Seville did not
identify directly with the ruling gens, but they legitimized its rule or took it for granted. Fredegar, Bede or the
Carolingian Annals completed the shift towards self-identification with the people, which they considered to
be chosen by God to rule over former Roman provinces.

TRIBES, PEOPLES
Similar to ancient ethnography, nineteenth- and early twentieth-century ethnology again used ethnicity as
an organizing principle for ‘them’, the ‘primitive’ peoples that Europeans met in the course of their colonial
expansion, who were classed in tribes, however adequate that happened to be.31 In this view, a tribe was
distinguished by its specific culture that was the chief object of the discipline of ethnology.32 “Thus, while
on the one hand the difference between ‘them’ and ‘us’ was firmly established, on the other, the basic simi-
larity between different sorts of ‘them’ was proclaimed.”33 The concept of the tribe provided a universally-

27
Andrew Gillett, Was ethnicity politicized in the earliest medieval kingdoms?, in: On Barbarian Identity – Critical Approaches to
Ethnogenesis Theory, ed. Andrew Gillett (Turnhout 2002) 85–121, is correct that official ethnic self-denomination in the regna
was patchy at first. But it did catch on; and in historiography, the identification of regnum and gens is overwhelming.
28
Cf. Andrea Giardina, L’Italia Romana – storie di una identità incompiuta (Roma/Bari 1997).
29
Walter Pohl, Telling the difference – Signs of ethnic identity, in: Strategies of Distinction. The Construction of Ethnic Com-
munities, 300–800, ed. Walter Pohl/Helmut Reimitz (The Transformation of the Roman World 2, Leiden/New York/Köln 1998)
17–69, reprinted in: From Roman Provinces to Medieval Kingdoms, ed. Thomas F.X. Noble (Rewriting History, London/New
York 2006) 120–167; id., Die ethnische Wende.
30
Aristoteles, Politika 1261a; 1276a.
31
Jean-Loup Amselle, Logiche meticce. Antropologia dell’identità in Africa e altrove (Torino 1999).
32
For the 18th-century origin of a definition of non-European peoples by their respective ‘culture’, see Niklas Luhmann, Gesell-
schaftsstruktur und Semantik (Frankfurt a. Main 1999) 145ff.
33
Jenkins, Rethinking Ethnicity 18.
14 Walter Pohl

applicable model for ‘archaic’ social organisation. It did not only allow for distinguishing between different
tribes (and often enough, just as the Romans did with their neighbours, to play them off against each other). It
also allowed for hypothesizing a distinctive ‘primitive’ period in the cultural evolution of mankind, a ‘tribal
phase’.34 Tribes, according to a recent definition, are societies “which create all or most of their social roles
by ascribing social importance to biological characteristics, in other words, societies ordered with reference
to kinship, sex and age.”35 Of course, the common descent of a tribe need not be biological fact, but rather a
genealogical construction. Other anthropologists try to define the tribe by objective characteristics, under-
lining common economic factors or political organisation. It has also been described as a “state of mind”.36
Whatever was common to a tribe, it was as much defined by who did not belong to it.37 Tribal organisation
provided a simple and efficient logic to channel and escalate conflict by involving additional groups accord-
ing to an essentially binary structure of inclusion or exclusion, regardless of their interest in what the original
cause of the dispute may have been. This logic of dispute is usually defined as bloodfeud, but may need a
more complex explanation.38 In short, what emerges from recent research is the great variety of forms of
tribal organisation. Tribes may be pastoral or sedentary, they may be settled in a common territory or scat-
tered among others, they may share a common political organisation or live in autonomous clans or villages,
and they may be subject or related to a state, resist its influence or have no contact at all to more complex
forms of social organisation.
An interesting example of the dialectic of tribe and empire is the early Islamic World, especially as it can be
compared to the parallel, but very different development of the relationship between ethnic groups and states
in the early medieval West.39 In a very general sense, it could be said that most of the Islamic conquerors of
the seventh century were Arabs. But beneath this sweeping ethnic identification, there was a tribal coalition
forged by a dynamic new religion and reinforced by other converts. Unlike the barbarians who built their king-
doms on the territory of the Western Empire, in the east no single ethnic groups established their rule over one
or several former Roman provinces. Still, many of the Islamic conquerors preserved (or recreated) their ethnic
ties. In the ninth century when sources become more detailed, there were some cities and regions in which
tribal solidarities constituted a considerable political potential, like Mossul or the country around Damascus;
and others where they do not seem to have been important at all (such as the city of Damascus itself). In many
of these areas dominated by tribal affiliations, tribes were settled in different villages in a rather dispersed
way.40 The rich ethnic terminology in Arabic seems to reflect a whole hierarchy of tribal affiliations, from the
people of the Arabs, within which there was an opposition between the inclusive tribal groupings of the Yaman
(originally ‘Yemenite’ south Arabians) and the (North Arabic) Qais to single tribes, subtribes and clans. This
hierarchy was, however, far from systematic, and changes of allegiance were regularly reflected in modifica-
tions of genealogy. The Caliphs had to coexist with tribal structures up to a point, but also had a vested interest
in not letting them interfere too much in the rule of the Islamic empire.41
This brief sketch may suffice here to demonstrate that the role of ethnic identities evolved very differently in
the post-Roman West. Even the scholarly terminology has developed differently. Scholars of the Islamic world
regularly use ‘tribe’, and regard ‘people’ (in their context, the Arabs) as a very vague concept. For the West,
Reinhard Wenskus, in the 1960s, still employed the term “Stammesbildung”; but soon after that, ‘Stamm’ or
‘tribe’ ceased to be used by most authors, at least for the migration period.42 Traditionally, the ‘people’ of the
34
See, for instance, Julian H. Steward, Theory of Culture Change (Urbana-Ill. 1955); Isaac Schapera, Government and Politics in
Tribal Societies (London 1956).
35
Patricia Crone, The tribe and the state, in: John A. Hall, States in History (Oxford 1986); reprinted with revisions in ead., From
Arabian Tribes to Islamic Empire. Army, State and Society in the Near East c. 600–850 (Aldershot 2008) n. 1. Adding “sex and
age“ to the definition is interesting in view of the strong cultural accentuation of gender and age in many prehistoric burial com-
plexes.
36
Richard Tapper, Frontier Nomads of Iran: A Political and Social History of the Shahsevan (Cambridge 1997); for various defini-
tions of tribe, see also Eva Orthmann, Stamm und Macht. Die arabischen Stämme im 2. und 3. Jahrhundert der Hira (Wiesbaden
2002).
37
A classic anthropological study of the importance of delimitations of tribes and peoples, and one of the most influential con-
tributions to research on ethnicity altogether, is Frederick Barth, Introduction, in: Ethnic Groups and Boundaries. The Social
Organisation of Cultural Difference, ed. id. (Oslo 1969) 9–38.
38
See, for instance, Violence and Society in the Early Medieval West, ed. Guy Halsall (Woodbridge 1998).
39
See Crone, The tribe and the state; Orthmann, Stamm und Macht.
40
For instance in the Gta around Damascus: Orthmann, Stamm und Macht 81–105.
41
Crone, The tribe and the state.
42
Reinhard Wenskus, Stammesbildung und Verfassung. Das Werden der frühmittelalterlichen gentes (Köln/Wien 21977).
 Archaeology of identity: introduction 15
Germans had been seen as constituted by its ‘tribes’, Saxons, Franks, Bavarians and others; many scholars
also regarded Goths, Vandals or Lombards as ‘deutsche Stämme’, with obvious ideological implications.43
A similar view of the genesis of the Polish people from ‘its’ tribes has recently been refuted by Przemysav
Urbaczyk.44 Ideological overtones (and the notion of primitivity) were one reason why the concept of ‘tribe’
virtually disappeared from early medieval studies. Also, the ethnic groups that formed on Roman territory
seem to have reached a size and a degree of complexity that does not correspond with the notion of ‘tribe’. But
it is worth noting that unlike the Arab language with its rich and differentiated vocabulary for ethnic groups,
the most diffused Latin term, gens, does not allow for groups to be distinguished by their size – it can refer to
the Agilolfing dynasty and to the Franks alike. A tribal model may in any case be quite accurate to describe
the situation in the Germania in the day of Augustus, or at least the Roman perception of it. From the third
century onward, the situation became more complicated as larger and often more improvised and unstable
groupings began to form.
Arguably, the notion of ‘tribe’ has little to contribute to understanding, for instance, the formation of
Franks and Alemanni, although they have often been pictured as tribal confederations.45 Quite generally, the
use of ethnic distinctions by early migration-age Romans rarely reflects a discernible self-definition by barbar-
ian descendance groups. There are some cases of ethnic prejudice, for instance when the Goths explained the
name of the Gepids as gepanta, ‘lazy’.46 If tribe is a ‘state of mind’, we may be surprised by the relative lack
of tribal logic in the often dramatic wars and battles of the migration age. Most of the wandering armies had a
rather loose structure and fluctuating membership. As many conflicts arose between members of the same eth-
nic group (many of whom served in Roman armies) as between different tribes. For instance, Gregory of Tours
pictures King Clovis as a cynic who uses the language of kinship just to kill off all the potential rivals in his
family: “And having killed many other kings and his nearest relatives, of whom he was jealous lest they take
the kingdom from him, he extended his rule over all the Gauls. However he gathered his people together at one
time, it is said, and spoke of the kinsmen whom he had himself destroyed. ‘Woe to me, who have remained as
a stranger among foreigners, and have none of my kinsmen to give me aid if adversity comes.’ But he said this
not because of grief at their death but by way of a ruse, if perchance he should be able to find some one still to
kill.”47 Thus, a clear pattern of ‘tribal’ allegiances and enmities could hardly form. Stable tribal systems often
prevent the rise of supra-regional powers because the most powerful tribes soon have to face opposition from a
broad alliance of those threatened by their expansion. If such a system existed among the northern barbarians
in the earlier imperial age, it had obviously collapsed in the early migration period.
That does not mean that ethnic allegiances did not matter at all; but they mattered on a different level. The
most successful barbarian groups on Roman territory seem to have used ethnic identifications in the attempt
to reinforce the cohesion of their armies. The peoples that formed in the nascent kingdoms seem to have relied
increasingly on ethnic solidarities. But it is necessary to historicize the forms of ethnicity that mattered in the
migration period. Ethnic identities did not necessarily mean the same in all cases, and the variety of ethnic
groups that are attested in the sources between the fourth and the sixth centuries is considerable. Ethnic names
may designate one (or several) of the following types of groups:48
1. Local groups, villages or extended clans; normally the written sources hardly talk about these. Many of the
names in Ptolemaeus or the Bavarian Geographer may be understood in this way.

43
Cf. Matthias Becher, Rex, Dux und Gens: Untersuchungen zur Entstehung des sächsischen Herzogtums im 9. und 10. Jahrhun-
dert (Husum 1996); Bernd Schneidmüller, Völker – Stämme – Herzogtümer. Von der Vielfalt der Ethnogenesen im ostfränk-
ischen Reich, in: MIÖG 108 (2000) 31–47; Joachim Ehlers, Methodische Überlegungen zur Entstehung des deutschen Reiches
im Mittelalter und zur nachwanderzeitlichen Nationenbildung, in: Beiträge zur mittelalterlichen Reichs- und Nationsbildung in
Deutschland und Frankreich, ed. Carlrichard Brühl/Bernd Schneidmüller (München 1997) 1–14.
44
See the contribution by Przemysav Urbaczyk, in this volume.
45
See, for instance, Alexander Demandt, Die westgermanischen Stammesbünde, in: Klio 75 (1993) 387–406. For a more sceptical
view, Walter Pohl, Die Germanen (Oldenbourgs Enzyklopädie der deutschen Geschichte 57, München 22002) 109f.
46
Jordanes, Getica XVII, 95 (ed. Theodor Mommsen, MGH AA 5, 1, Berlin 1882) 82.
47
Gregory of Tours, Histories II, 42 (transl. Earnest Brehaut, New York 1916) http://www.fordham.edu/halsall/basis/gregory-hist.
html (13.2.2009).
48
For a fuller treatment, see Walter Pohl, Die Namen der Barbaren: Fremdbezeichnung und Identität in Spätantike und Frühmit-
telalter, in: Zentrum und Peripherie. Gesellschaftliche Phänomene in der Frühgeschichte, ed. Herwig Friesinger/Alois Stuppner
(Mitteilungen der Prähistorischen Kommission 57, Wien 2004) 95–104; see also id., Spuren, Texte, Identitäten. Methodische
Überlegungen zur interdisziplinären Erforschung frühmittelalterlicher Identitätsbildung, in: Gräber, Siedlungen und Identitäten.
Archäologie des 4.–7. Jahrhunderts im Westen, ed. Sebastian Brather (RGA Erg. Bd. 57, Berlin/New York 2008) 13–26.
16 Walter Pohl

2. Regional sub-tribes or settlement units, such as the Brisigavi, Lentienses etc. as part of the Alamanni. The
names may (as in this example) be derived from toponyms or represent a tribal designation. Evidence on
such regional units is also rather rare, both before and after settlement on Roman soil.
3. A special case are bipartite tribal structures within a people or tribe. They usually correspond to double
genealogies. For instance, Var und Chunni were the two ancestral units within the sixth-century Avars (or
Varchonitai; the name was composed from both elements).49 They are not known to have had any political
influence, but sources are meagre. It may be that the Gothic Tervingi and Greutungi, the Vandal Hasdingi
and Silingi or the Frankish Salii and Ripuarii represented similar sub-units at some stage, although in all
cases this assumption has been criticized as a modern projection.50 A similar double structure underlies the
Arab geneaologies that go back to Adnn (the Qais) and Qahtn (the Yaman); unlike other subgroups, they
constituted a kind of tribal superstructure that incorporated other tribes within the wide and ambiguous
frame of Arab identity.51
4. Ethnic groups under foreign rule. This is a wide field; a well-known example are the Rugi under Ostrogoth-
ic rule who were said to have deliberately maintained their cohesion by strict endogamy, and were able to
raise a Rugian to Gothic kingship half a century after their incorporation in the Gothic kingdom.52 The
Bulgars under Lombard rule are another case, having been settled under their duke Alzeco in the duchy of
Benevento in the second half of the 7th century and still distinguishable in the 10th century; or the Saxons
of Bayeux in the Merovingian kingdom. Whereas these groups seem to have been able to maintain some
regional cohesion, other groups, such as the Gepids under Avar and Lombard rule, preserved their identity
at the village level. Archaeologists often explain divergent evidence within a larger group with the exist-
ence of such ethnic minorities (see the contribution by Peter Stadler, in this volume); but usually, those are
equations with too many variables.
5. Autonomous tribes or peoples on a regional level with relatively coherent settlement areas. In the early
imperial Germania, that seems to have been the predominant mode of social organisation of, for instance,
Cherusci, Chauci, Marsi, Chatti and others. Hardly any of the politically active peoples of the migration
period fit this model; perhaps the peoples settled north of the Danube after Attila’s death offer an example,
but even there, no balanced tribal system emerged.
6. Peoples who lived scattered over several settlement areas, or names that occur in more than one region, for
instance the Goths and Heruls in the 4th to 6th centuries, the Alans and the Suebi of the 5th and 6th centuries,
and later Bulgars or Croats. This may be a result of separations and migrations (and is usually described
that way in the texts); but more likely, it is often the case of a prestigious name adopted by several groups
(as Theophylact attests for the Avars53). A corresponding sense of identity that included some or all of these
groups is rarely attested,54 but there are some examples (for instance, the Heruls and the Vandals sending
messengers into their putative old homelands in Procopius). In some cases, the common name may simply
represent an outside perception or a mistake in the sources.
17. Ethnic confederations, tribal unions, agglomerates and groups of tribes with supra-regional significance
but weak or nonexistent organisation and often doubtful delimitation, such as the Suebi of the early Em-
pire, the Huns after the death of Attila, or the Yamans and Qais of the Ummayad period. The distinction to
ethnographic umbrella terms is fluid, and it is often doubtful whether these names corresponded to strong
self-identifications.

49
Theophylactus Simocatta, Histories VII, 7 (ed. Karl de Boor, Leipzig 1887) 257; Walter Pohl, Die Awaren. Ein Steppenvolk in
Mitteleuropa, 567–822 n. Chr. (München 1988, 22002) 31ff.
50
Wolfram, Die Goten 34f.; Helmut Castritius, Hasdingen, in: RGA 2 Aufl. 14 (1999) 26–28; in general id., Die Vandalen (Stuttgart
2007); Matthias Springer, Gab es ein Volk der Salier?, in: Nomen et gens. Zur historischen Aussagekraft frühmittelalterlicher
Personennamen, ed. Dieter Geuenich/Wolfgang Haubrichs/Jörg Jarnut (RGA Erg. Bd. 16, Berlin/New York 1997) 58–83; Pohl,
Die Völkerwanderung 70f.; 167.
51
Orthmann, Stamm und Macht 210f.; see above.
52
Procopius, History of the Wars VII, 2 (ed./trans. Henry Bronson Dewing, Procopius 4, Loeb Classical Library 173, Cambridge-
Mass./London 1924, reprint 2000) 166–173; Peter Heather, Disappearing and reappearing tribes, in: Strategies of Distinction.
The Construction of Ethnic Communities, 300–800, ed. Walter Pohl/Helmut Reimitz (The Transformation of the Roman World
2, Leiden/Boston/Köln 1998) 95–111.
53
Pohl, Awaren 30f.
54
See the ironical remarks by Reuter, Whose race 102: “We don’t know when or how Avar identity ceased to be meaningful to
Avars, though it clearly must have done at some point, since there aren’t any now. It’s far from clear that Avars thought they were
Avars when they still were Avars, if you see what I mean.”