Early Medieval Europe   () 159–185
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd, 
Garsington Road, Oxford OX DQ, UK and  Main Street, Malden, MA , USA
Blackwell Publishing LtdOxford, UKEMEDEarly Medieval Europe0963-9462© 2007 The Author. Journal Compilation © Blackwell Publishing LtdXXXOriginal ArticleSome remarks on ethnicity in medieval archaeologyFlorin Curta
Some remarks on ethnicity in
medieval archaeology
F C
Recent critiques of the culture-historical approach to ethnicity have
denounced the idea that archaeological cultures are ‘actors’ on the historical
stage, playing the role that known individuals or groups have in documentary
history. But the critique has gone as far as to claim that, because
archaeologists supposedly have no access to the meaning of cultural traditions,
medieval ethnicity cannot be studied by archeological means. Ethnicity
should be banned from all discussions, if medieval archaeology is to make
any progress in the future. The paper examines the theoretical malaise at
the root of this scepticism verging on nihilism. The understanding of the
archaeological record not as an imprint, but as a text allows for much
learning about meaning in the past. Symbols, style and power are the key
concepts that currently guide anthropological research on ethnicity as a
‘social construction of primordiality’. As several archaeological examples
show, medieval ethnicity was a form of social mobilization used in order
to reach certain political goals. Ethnic identity was built upon some preexisting
cultural identity, in a prototypic manner.
Eine ethnische Einheit, ein Volk oder ein Stamm, mit dem man in
Mitteleuropa ohne weiteres Kulturen oder kulturellen Gruppen identifizieren
pflegt, ist aber kein gar so homogenes und gleichbleibendes
Gebilde, wie man es in Anlehnung an der romantischen Volksbegriffs
Herders annehmen zu können glaubt, sondern eine ungemein
vielschichtige Gesellungsform, die nicht so sehr durch die Bande
des Blutes, als vielmehr durch das Moment politischer Herrschaft
zusammengehalten wird.1
1
Helmut Preidel, ‘Awaren und Slawen’, Südost-Forschungen 11 (1952), pp. 33–45, at p. 57
(emphasis added): ‘Ethnic groups, peoples or tribes, with which cultures or cultural groups
are indiscriminately identified in Central Europe, are not homogeneous or static formations,
as many believe on the basis of Herder’s Romantic concept of Volk. Instead, they are unevenly
multi-layered social forms, which are kept together not so much by the bond of blood,
as by the circumstances of political power.’
Early Medieval Europe   ()
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd
160 Florin Curta
An archaeologist studying the early Middle Ages, who would be curious
about what historians of that same period have to say about ethnicity,
will quickly experience disappointment if venturing to read some of the
most recent publications in the field. Not too long ago, ethnicity was
treated more as a modern construct than as a medieval category, while
students of the Middle Ages were even warned against examinations of
‘ethnic identity’ that risked anachronism when the origins of contemporary
concerns and antagonisms were sought in the past.2
There is now little
encouragement for the neophyte in search of theoretical guidance and
methodological advice. The current debate between the Vienna and the
Toronto ‘schools’ on the issue of the barbarian (specifically Gothic)
ethnogenesis may be stimulating for the critical reading of written sources,
but has little, if anything, to offer in terms of innovative approaches to
medieval ethnicity. The debate is primarily about whether and to what
degree it is possible to use written sources for explaining the forming
of medieval ethnic groups. In the words of a recent commentator,
members of the ‘Vienna school’ purport to be able ‘[to] explain how
national or ethnic (take your pick) communities arose in the early
Middle Ages’.3
By contrast, Walter Goffart and his students have ‘incessantly
attacked the idea that ancient traditions and orally transmitted
myths had any part in the shaping of early medieval peoples’.4
Underlying
assumptions about the intellectual heritage of the ‘Vienna school’
have increasingly politicized the debate and provoked much irritation.
The unofficial founder of the Vienna school, the German historian
Reinhard Wenskus, is said to have rescued the discipline of Germanic
antiquity from total discredit after World War II, and in turn the
members of the Toronto school are accused of ‘misapprehensions and
defamatory insinuations’.5
At this point in the polemic, the archaeologist with an interest in
ethnicity may be able to recognize some familiar tunes. It is by now
common knowledge that the founder of the culture-historical school
of archaeology was the German archaeologist Gustaf Kossinna. Today,
2
Patrick J. Geary, ‘Ethnic Identity as a Situational Construct in the Early Middle Ages’,
Mitteilungen der anthropologischen Gesellschaft in Wien 113 (1983), pp. 15–26, at p. 16; and The
Myth of Nations: The Medieval Origins of Europe (Princeton and Oxford, 2002), pp. 16–19;
Patrick Amory, ‘Names, Ethnic Identity, and Community in Fifth- and Sixth-Century Burgundy’,
Viator 25 (1994), pp. 1–34, at p. 5; and People and Identity in Ostrogothic Italy, 489–554
(Cambridge, 1997), p. 317.
3
Charles Bowlus, ‘Ethnogenesis: The Tyranny of a Concept’, in A. Gillett (ed.), On Barbarian
Identity: Critical Approaches to Ethnicity in the Early Middle Ages (Turnhout, 2002), pp. 241–56,
at p. 243.
4
Walter Pohl, ‘Ethnicity, Theory, and Tradition: A Response’, in Gillett (ed.), On Barbarian
Identity, pp. 221–40, at p. 222.
5
Walter Goffart, ‘Does the Distant Past Impinge on the Invasion Age Germans?’, in Gillett
(ed.), On Barbarian Identity, pp. 21–38, at p. 31; Pohl, ‘Ethnicity, Theory’, p. 239.
Early Medieval Europe   ()
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd
Ethnicity in medieval archaeology 161
both archaeologists and historians attack Kossinna’s tenets and, whenever
possible, emphasize his association with Nazism and the political
use of archaeology. Like Wenskus for the Vienna school, Kossinna is
now regarded as the archetypal incarnation of all vices associated with
the culture-historical approach to archaeology. Unlike Wenskus, though,
Kossinna’s work is rarely cited, except for his famous statement:
‘Sharply defined archaeological culture areas correspond unquestionably
with the areas of particular peoples or tribes.’6
In post-war Germany,
Kossinna’s followers passed over in silence the fundamental question of
equating Völker and archaeological cultures, but continued to believe
that cultural community equalled biological community. In doing so,
they were simply reproducing views about ethnicity that were until
recently common in both German law and German politics. For more
than fifty years of the post-war period, ethnicity in Germany was
approached in the terms of the 1913 law (abolished only in 2000), which
defined German citizenship by descent, not by residence. In German
archaeology this has led to a deeply rooted confusion between ethnic
identity and biological group affiliation. The idea that a person buried
in ‘Germanic’ dress may not necessarily be of ‘Germanic’ descent has
rarely, if ever been questioned.7
As Hans Jürgen Eggers has long noted,
Kossinna’s Glaubenssatz of interpreting archaeological cultures as ethnic
groups was not his invention. Instead, this was a notion directly inspired
by the Romantic idea of culture as reflecting the national soul (Volksgeist)
in every one of its elements.8
According to Reinhard Wenskus, Kossinna’s
mistake was not so much that he aimed at an ethnic interpretation of
culture, but that he used a dubious concept of ethnicity, rooted in the
6
Gustaf Kossinna, Die Herkunft der Germanen. Zur Methode der Siedlungsarchäologie (Würzburg,
1911), p. 3; and Ursprung und Verbreitung der Germanen in vor- und frühgeschichtlicher Zeit
(Leipzig, 1936), p. 15. For a balanced evaluation of Kossinna’s work, see Leo S. Klejn, ‘Kossinna
im Abstand von vierzig Jahren’, Jahresschrift für mitteldeutsche Vorgeschichte 58 (1974), pp. 7–55
and Ulrich Veit, ‘Gustaf Kossinna and His Concept of a National Archaeology’, in H. Härke
(ed.), Archaeology, Ideology, and Society: The German Experience (Frankfurt a.M., Berlin and
Bern, 2000), pp. 40–59.
7
E.g., Volker Bierbrauer, ‘Zu den Vorkommen ostgotischer Bügelfibeln in Raetia II’, Bayerische
Vorgeschichtsblätter 36 (1971), pp. 133–47. The notion of ‘Germanic’ costume (Tracht) was
first introduced in the 1930s for the study of the early medieval ethnicity through material
culture. For an excellent survey of the issue, see Hubert Fehr, ‘Hans Zeiss, Joachim Werner und
die archäologischen Forschungen zur Merowingerzeit’, in H. Steuer (ed.), Eine hervorragend
nationale Wissenschaft. Deutsche Prähistoriker zwischen 1900 und 1995 (Berlin and New York,
2000), pp. 311–415. For the link between the 1913 law and contemporary archaeology in
Germany, see Heinrich Härke, ‘Archaeologists and Migrations: A Problem of Attitude?’,
Current Anthropology 39 (1998), pp. 19–45, at p. 21.
8
For Kossinna’s Glaubenssatz and an early German critique of his ideas, see Hans Jürgen
Eggers, ‘Das Problem der ethnischen Deutung in der Frühgeschichte’, in H. Kirchner (ed.),
Ur- und Frühgeschichte als historische Wissenschaft. Festschrift zum 60. Geburtstag von Ernst
Wahle (Heidelberg, 1950), pp. 49–59, at p. 49.
Early Medieval Europe   ()
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd
162 Florin Curta
Romantic views of the Volk.9
A new generation of archaeologists now
wants to free German archaeology both from such a misconstrued concept
of ethnicity and from the ghost of Kossinna.
One of the most vocal representatives of this movement is Sebastian
Brather, the author of a recent book dedicated to the study of ethnicity
in early medieval archaeology.10
The abrasiveness of this volume has
already provoked the irritation of the die-hard followers of Kossinna.
In a paper for the most recent publication of the Vienna school, a
Festschrift for Herwig Wolfram, Volker Bierbrauer angrily scolds Brather
for having used too much Anglo-Saxon literature in his critique of
German archaeologists. According to Bierbrauer, any discussion of the
ethnic interpretation should begin with the data and move gradually to
theory, not the other way around. Besides, it is essentially wrong to call
ethnic identity a social construct, Bierbrauer argues, because in German
the word ‘construct’ (Konstrukt) is an abhorrently foreign loan (Fremdwort),
the meaning of which requires explanation. To Bierbrauer, Brather’s
accusations of nationalism and repeated references to Gustav Kossinna
are nothing but slander.11
At a quick glimpse, the incipient scandal may appear to indicate that
Brather’s is the most revisionist book ever written in the archaeology of
ethnicity. But is it really? Taken on its own terms, the argument the
author lays out seems reasonable enough: ‘the nature of archaeology as
a historical discipline does not rest upon and cannot be reduced to the
question of ethnic interpretation, just as history cannot be reduced to
the study of politics’.12
Ethnicity is subjective and the boundaries of ethnic
groups are marked with symbols. As a consequence, ethnic identity in
the past is beyond the reach of archaeology, because the meaning
initially attached to the material culture symbols used for building ethnic
boundaries will forever remain unknown. In a typically empiricist stance,
Brather recommends that archaeologists abandon any research on
ethnicity, as long as no independent, written sources exist out there, to
decipher the meaning of those symbols for them. In the meantime, they
should focus on what they can really do, namely research on economic
and social structures, social rank, religious behaviour, and the like.
9
Reinhard Wenskus, Stammesbildung und Verfassung. Das Werden der frühmittelalterlichen
Gentes (Cologne, 1961), p. 137.
10
Sebastian Brather, Ethnische Interpretationen in der frühgeschichtlichen Archäologie. Geschichte,
Grundlagen und Alternativen (Berlin and New York, 2004).
11
Volker Bierbrauer, ‘Zur ethnischen Interpretation in der frühgeschichtlichen Archäologie’,
in W. Pohl (ed.), Die Suche nach den Ursprüngen. Von der Bedeutung des frühen Mittelalters
(Vienna, 2004), pp. 45–84, at pp. 46, 49 and 74–5.
12
Brather, Ethnische Interpretationen, p. 27.
Early Medieval Europe   ()
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd
Ethnicity in medieval archaeology 163
The issue at stake
To be sure, Brather’s bête noire is the culture-historical approach for
which many blame Kossinna without knowing that what happened in
German archaeology between the two World Wars was ‘only the tip of
a pan-European iceberg’.13
The culture-historical approach was based
on an essentially expanded meaning of the concept of culture, as used
in the nineteenth century for classifying human groups, with all the
underlying assumptions of holism, homogeneity and boundedness.14
Traditionally, archaeological cultures were defined in monothetic terms
on the basis of the presence or absence of a list of traits or types, which
had either been derived from excavated assemblages and type sites or
were intuitively considered to be the most appropriate attributes for the
definition of the culture. In practice, it became rapidly clear that no group
of cultural assemblages from any single culture ever contains all of the
cultural artefacts, a problem first acknowledged by a Marxist archaeologist,
Vere Gordon Childe. Childe’s solution was to discard the untidy information
by demoting types with discontinuous frequency from the rank
of diagnostic types, thus preserving the nineteenth-century ideal of a
univariate cultural block.15
The culture-historical archaeologists against
whom Brather directed his attack typically regard archaeological cultures
as actors on the historical stage, playing the role for prehistory that known
individuals or groups have in documentary history. Archaeological cultures
were thus easily equated to ethnic groups, for they were viewed as legitimizing
claims of modern groups to territory and influence.16
The first criticism against the idea that archaeological cultures represent
ethnic groups came from within the framework of culture history, but
critiques usually consisted of cautionary tales and attributed difficulties
to the complexity and incompleteness of the artefactual record, without
13
Brather, Ethnische Interpretationen, p. 27. Nonetheless, Brather insists that the concept of ‘culture’
(Cultur) in its ethnological and archaeological sense is a German invention, having been first
introduced by Gustav Friedrich Klemm (1802–67). Moreover, the Kulturkreis theory was in
turn the invention of another famous German, Leo Frobenius, before gaining reputation
through the studies of Fritz Graebner and the Vienna school of ethnology. As a consequence,
in Brather’s view, the relation established in Germany between prehistoric archaeology and
ethnology was essentially different from that between archaeology and anthropology in
England or the United States. See Brather, Ethnische Interpretationen, pp. 44 and 61–63.
14
Hans-Peter Wotzka, ‘Zum traditionellen Kulturbegriff in der prähistorischen Archäologie’,
Paideuma 39 (1993), pp. 25–44. For a recent demonstration of how resistant nineteenth-century
concepts can be, even when computers replace field notebooks, see Vera B. Kovalevskaia,
Arkheologicheskaia kul’tura: praktika, teoriia, komp’iuter (Moscow, 1995).
15
Vere Gordon Childe, Piecing Together the Past: The Interpretation of Archaeological Data
(London, 1956), pp. 33 and 124.
16
The most balanced assessment of the culture-historical approach to archaeology remains that
of Bruce G. Trigger, A History of Archaeological Thought (Cambridge, 1989), pp. 148–206. See
also the pertinent, if altogether ignored, remarks of C. Ronald, ‘In Defense of Migration and
Culture History Studies’, Journal of Intermountain Archaeology 3 (1984), pp. 43–52.
Early Medieval Europe   ()
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd
164 Florin Curta
calling into question the assumption of an intrinsic link between artefacts
and groups. The general response to such problems was a retreat
into the study of chronology and typology as ends in themselves, and
the emergence of debates concerning the meaning of archaeological
types, in particular whether such types represent etic categories imposed
by the archaeologist or emic categories of their producers.17
By contrast,
Brather now insists that there will always be a considerable distance
between etic and emic categories. Since ethnicity in the past was defined
with conceptual categories different from ours, ethnicity cannot be studied
by archaeological means. A historian not versed in the vagaries of
archaeological theory, but well informed about the scandal surrounding
Brather’s ideas, sarcastically noted that after abandoning Kossinna’s
approach, ‘we are now told by archaeologists that archaeological cultures
tell us nothing about ethnicity’.18
An archaeologist with an interest in
migrations is similarly puzzled: ‘Exactly at the point where identity has
become the key question of the post-modern world, and migration is fast
becoming the key issue in post-Soviet Europe, academic archaeologists
and historians have lost their own convictions and fail to provide the
historical guidance that the general public is looking for.’19
Whether or not they had Brather in mind, both German-speaking
authors point to discrepancies well illustrated in his latest book. For
Brather does not believe that archaeologists have access to ethnic signs.
Because such signs were not mentioned in written sources, ‘we have no
inside report, no message about the meaning of things’.20
He denies that
we will ever have the ability to learn in detail the meaning of cultural
traditions, their symbolic underpinnings, or their importance for past
ritual or group behaviour. ‘There is no hermeneutic path to a real
17
Lev S. Klein, ‘Regressive Purifizierung und exemplarische Betrachtung. Polemische Bemerkungen
zur Integration der Archäologie mit der schriftlichen Geschichte und Sprachwissenschaft bei
der ethnischen Deutung des Fundgutes’, Ethnographisch-archäologische Zeitschrift 15 (1974),
pp. 223–54, at p. 224; and ‘Die Ethnogenese als Kulturgeschichte, archäologisch betrachtet.
Neue Grundlagen’, in H. Kaufmann and K. Simon (eds), Beiträge zur Ur- und Frühgeschichte,
vol. 1 (Berlin, 1981), pp. 13–25, at p. 18. See also Shaun Hides, ‘The Genealogy of Material
Culture and Cultural Identity’, in P. Graves-Brown, S. Jones and C. Gamble (eds), Cultural
Identity and Archaeology: The Construction of European Communities (London and New York,
1996), pp. 25–47, at p. 26. For the earlier criticism of the idea that archaeological cultures
represent ethnic groups, see Ernest Wahle, Zur ethnischen Deutung frühgeschichtlicher Kulturprovinzen.
Grenzen der frühgeschichtlichen Erkenntnis, vol. 1 (Heidelberg, 1941).
18
Walter Pohl, ‘A Non-Roman Empire in Central Europe: The Avars’, in H.-W. Goetz, J.
Jarnut and W. Pohl (eds), Regna and Gentes: The Relationship Between Late Antique and Early
Medieval Peoples and Kingdoms in the Transformation of the Roman World (Leiden and Boston,
2003), pp. 571–95, at p. 588. Pohl explicitly refers to Sebastian Brather, ‘Ethnische Identitäten
als Konstrukte der frühgeschichtlichen Archäologie’, Germania 78 (2000), pp. 139–77.
19
Heinrich Härke, ‘The Debate on Migration and Identity in Europe’, Antiquity 78 (2004),
pp. 453–6, at p. 456.
20
Sebastian Brather, ‘Ethnic Identities as Constructions of Archaeology: The Case of the
Alamanni’, in Gillett (ed.), On Barbarian Identity, pp. 149–76, at pp. 172–3; Brather, Ethnische
Interpretationen, pp. 369, 570 and 577.
Early Medieval Europe   ()
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd
Ethnicity in medieval archaeology 165
understanding’ of ethnicity. Ethnic identity in the past is beyond the
reach of archaeology, whether or not that identity truly existed and
mattered for people in the past. Looking for ethnic identity is just an
archaeological form of nationalism.21
Volker Bierbrauer may well complain
about Brather’s critique of ethnic(ist) interpretations in Germany.
By now, agnosticism has made converts well beyond the linguistic and
institutional boundaries of Germany. This is particularly true for those
countries in Eastern Europe in which there is still an undercurrent of
culture-historical archaeology that Kossinna would have easily recognized.
In the midst of heated debates about the ethnogenesis of the
Slavs, some Polish archaeologists have begun to entertain doubts that
we would ever be able ‘to explain and describe the birth of any kind of
ethnic group’.22
A Lithuanian archaeologist cites Brather in support of
his firm belief that ‘the archaeological material represents social, but not
ethnic identity’.23
Others maintain that ‘we cannot get to the psyche of
those societies in the past, which we study’.24
Studying ethnicity in the
past is therefore only the study of the past with the ethnocentric (or,
worse, nationalist) concerns of the present. The reductionist approach
on which the archaeological search for ethnicity is based comes dangerously
close to the logic of ethnic cleansing. As a consequence, ethnicity
must be banned from all discussions, if archaeology as an academic
discipline is to make any progress in the future.25
What is ethnicity?
At the root of this scepticism verging on nihilism seems to be a serious
theoretical malaise. To be sure, the premise – that collective identities
are social constructs – is theoretically as well as empirically sound. In a
21
Brather, ‘Ethnic Identities’, pp. 173 and 175.
22
Jerzy Gbssowski, ‘Is Ethnicity Tangible?’, in M. Hardt, C. Lübke and D. Schorkowitz (eds),
Inventing the Pasts in North Central Europe: The National Perception of Early Medieval History
and Archaeology (Bern, 2003), pp. 9–17, at p. 9. See also Stanislaw Tabaczynski, ‘Procesy
etnogenetyczne: doswiadczenia badawcze archeologii i przyszlo1c’, in M. Mi1kiewicz (ed.),
Slowianie w Europie wczesniejszego 3redniowiecza. Katalog wystawy (Warsaw, 1998), pp. 79–99;
Przemyslaw Urbanczyk, ‘Archeologia etniczno1ci – fikcja czy nadzieja?’, in A. Buko and
P. Urbanczyk (eds), Archeologia w teorii i w praktyce (Warsaw, 2000), pp. 137–46. For current
debates about the ethnogenesis of the Slavs, see Magdalena Maczynska, ‘Die Archäologie über
die Anfänge der Slawen. Tagung in Kraków, 19.−21. November 2001’, Ethnographisch-archäologische
Zeitschrift 42 (2001), pp. 417–19; Paul M. Barford, ‘Crisis in the Shadows: Recent Polish
Polemic on the Origin of the Slavs’, Slavia Antiqua 44 (2003), pp. 121–55.
23
Mindaugas Bertasius, ‘The Archaeology of Group: From Situational Construct to Ethnic Group’,
in V. Lang (ed.), Culture and Material Culture: Papers from the First Theoretical Seminar of
the Baltic Archaeologists (BASE) Held at the University of Tartu, Estonia, October 17th−19th,
2003 (Tartu, 2005), pp. 29–38, at p. 32 (citing Brather, ‘Ethnische Identitäten’, p. 168).
24
Henryk Mamzer, ‘Ethnischer Mythus in der Archäologie’, in G. Fusek (ed.), Zbornik na
poçest Dariny Bialekovej (Nitra, 2004), pp. 223–7, at p. 226.
25
Mamzer, ‘Ethnischer Mythus’, p. 225.
Early Medieval Europe   ()
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd
166 Florin Curta
Weberian stance, Brather maintains that social identities (ethnic ones
included) are not a direct reflection of ‘social reality’, even though they
are themselves ‘nothing less than real’.26
This is clearly an echo of recent
studies of ethnicity as a mode of action and of representation. While
twenty years ago, ethnicity could not be defined as ‘either culture or society,
but a specific mixture, in a more or less stable equilibrium, of both
culture and society’, now the tendency is to treat it as a decision people
make to depict themselves or others symbolically as bearers of a certain
cultural identity.27
In short, ethnicity has become the politicization of
culture. Ethnicity may not be innate, but individuals are born with it;
it may not be biologically reproduced, but individuals are linked to it
through cultural constructions of biology; it is certainly not just cultural
difference, but ethnicity cannot be sustained without reference to an
inventory of cultural traits. As Thomas Hylland Eriksen puts it, ethnicity
must be regarded as the ‘collective enaction of socially differentiating signs’.28
Such an approach is largely due to the extraordinary influence of the
Norwegian anthropologist Fredrik Barth.29
He shed a new light on
subjective criteria (ethnic boundaries) around which the feeling of ethnic
identity of the member of a group is framed. He also emphasized the
transactional nature of ethnicity, for in the practical accomplishment of
identity, two mutually interdependent social processes are normally at
work: internal and external definition, the latter also known as catego-
rization.30
Barth’s approach embraced a predominantly social interactionist
perspective, derived from the work of the social psychologist Erving
Goffman.31
Objective cultural difference was thus viewed as epiphenomenal,
26
Brather, Ethnische Interpretationen, pp. 100 and 106.
27
Guy Nicolas, ‘Fait “ethnique” et usages du concept d’ “ethnie”’, Cahiers internationaux de
sociologie 54 (1973), pp. 95–126, at p. 107; Anthony P. Cohen, ‘Culture as Identity: An
Anthropologist’s View’, New Literary History 24 (1993), pp. 195–209, at p. 197.
28
Thomas Hylland Eriksen, ‘The Cultural Contexts of Ethnic Differences’, Man 26 (1991),
pp. 127–44, at p. 141. For similar formulations in German sociology and historiography, see
Veit-Michael Bader, ‘Ethnische Identität und ethnische Kultur. Grenzen des Konstruktivismus
und der Manipulation’, Forschungsjournal Neue Soziale Bewegungen 8 (1995), pp. 32–45; Klaus
Erich Müller, ‘Ethnicity, Ethnozentrismus und Essentialismus’, in W. Essbach (ed.), Wir, ihr,
sie. Identität und Alterität in Theorie und Methode (Würzburg, 2000), pp. 317–43. For ethnicity
and the inventory of ‘cultural traits’, see Brackette F. Williams, ‘Of Straightening Combs,
Sodium Hydroxide, and Potassium Hydroxide in Archaeological and Cultural-Anthropological
Analyses of Ethnogenesis’, American Antiquity 57 (1992), pp. 608–12.
29
Fredrik Barth (ed.), Ethnic Groups and Boundaries: The Social Organisation of Culture Difference
(Bergen and London, 1969). For Barth’s more recent ideas, see his ‘Enduring and Emerging
Issues in the Analysis of Ethnicity’, in H. Vermeulen and C. Govers (eds), The Anthropology
of Ethnicity: Beyond ‘Ethnic Groups and Boundaries’ (The Hague, 1994), pp. 11–32.
30
Richard Jenkins, ‘Rethinking Ethnicity: Identity, Categorization and Power’, Ethnic and Racial
Studies 17 (1994), pp. 189–223, at pp. 198–9.
31
Norman Buchignani, ‘Ethnic Phenomena and Contemporary Social Theory: Their Implications
for Archaeology’, in R. Auger, M.F. Glass, S. MacEachern and P.H. McCartney (eds),
Ethnicity and Culture: Proceedings of the Eighteenth Annual Conference of the Archaeological
Association of the University of Calgary (Calgary, 1987), pp. 15–24, at. p. 16.
Early Medieval Europe   ()
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd
Ethnicity in medieval archaeology 167
subordinate to, and largely to be explained with reference to, social interaction.
Studies of ethnicity inspired by Barth’s approach thus suggest
that ethnic groups are more an idea than a thing. It is not so much the
group that endures as the idea of group. Activation of ethnic identity was
thus used to explain contextual ethnic phenomena, but this very ethnic
identity, since it was not directly observable (because of being an idea),
had to be derived from the actor’s ‘ethnic behaviour’. Barth’s model of
social interaction is so general that there is virtually nothing theoretically
unique about ethnic phenomena explained through reference to it,
for the model could be as well applied to other forms of social identity.
The emphasis of the post-Barthian anthropology of ethnicity has tended
to fall on processes of group identification rather than social categor-
ization.32
However, ethnicity as ascription of basic group identity on the
basis of cognitive categories of cultural differentiation is very difficult
to separate from other forms of group identity, such as gender or class.
Indeed, if social identities are social constructs, then ethnic identities
are not different from any other forms of identity in their subjective,
‘constructed’ nature. Any social identity is a Gemeinsamkeitsglauben, to
employ Max Weber’s bon mot. It has been noted that cultural traits by
which an ethnic group defines itself never comprise the totality of the
observable culture, but are only a combination of some characteristics that
the actors ascribe to themselves and consider relevant. People identifying
themselves as an ethnic group may in fact identify their group in a
primarily prototypic manner, with some recognizable members sharing
some but not all traits, and different traits being weighted differently in
people’s minds.33
How is this specific configuration structured and what
mechanisms are responsible for its reproduction?
Attempts to answer this question resurrected the idea that ethnic
groups are bounded social entities internally generated with reference
more to commonality than to difference.34
Such an approach draws
heavily from the French sociologist Pierre Bourdieu’s theory of habitus,
a system of durable, transposable dispositions, ‘structured structures
predisposed to function as structuring structures’.35
According to Bourdieu,
those durable dispositions are inculcated into an individual’s sense of
self at an early age and can be transposed from one context to another.
Habitus involves a form of socialization whereby the dominant modes
32
D.L. Horowitz, ‘Ethnic Identity’, in N. Glazer and D.P. Moynihan (eds), Ethnicity: Theory
and Experience (Cambridge, MA and London, 1975), pp. 111–40, at p. 114.
33
Eugeen E. Roosens, Creating Ethnicity: The Process of Ethnogenesis (Newbury Park and London,
1989), p. 12; Cynthia K. Mahmood and Sharon L. Armstrong, ‘Do Ethnic Groups Exist? A
Cognitive Perspective on the Concept of Cultures’, Ethnology 31 (1992), pp. 1–14, at p. 8.
34
G. Carter Bentley, ‘Ethnicity and Practice’, Comparative Studies in Society and History 29
(1987), pp. 25–55.
35
Pierre Bourdieu, The Logic of Practice, trans. Richard Nice (Stanford, 1990), p. 53.
Early Medieval Europe   ()
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd
168 Florin Curta
of behaviour and representation are internalized, resulting in certain
dispositions that operate largely at a pre-conscious level. Ethnicity is
constituted at the intersection of the habitual dispositions of the agents
concerned and the social conditions existing in a particular historical
context. Barth and his students built on concepts of the self and social
role behaviour typified by the ‘we vs. them’ perspective. In doing so, they
laid emphasis on ethnic boundaries and on the transactional nature
of ethnicity. By contrast, the tendency now is to view the content of
ethnic identity (the ‘we’ perspective) as important as the boundary
around it (the ‘we vs. them’ perspective). In practice ethnicity results
from multiple transient realizations of ethnic difference in particular
contexts of production and consumption of distinctive styles of material
culture. The very process of ethnic formation is coextensive with and
shaped by the manipulation of material culture. No surprise therefore
that such an understanding of ethnicity coincides in time with an
explosion of studies, primarily inspired by Edmund Husserl, that stress
ethnicity as a phenomenon of every day’s life (Alltagsleben).36
Routine
action, rather than dramatic historical experiences, food ways, rather
than political action, are now under scrutiny. As the idea of ethnicity
turns into a mode of action in the modern world, it becomes more
relevant to study the very process by which the ethnic boundary is
created in a specific social and political configuration.
This line of reasoning has inspired Frank Siegmund’s recent interpretation
of the distribution of weapons, pottery and glass vessels on either
side of the early medieval frontier between Franks and Alamans. Stressing
the importance of daily activities for the construction of ethnic
boundaries, Siegmund has noted that many more swords appear in
burial assemblages of the Alamannic than of the Frankish region, while
axes and spearheads seem to dominate in the Frankish region. Similarly,
within the Frankish zone, most, if not all, vessels associated with burials
were either wheel-made pots or glass beakers, while half of all pots
deposited in graves from the Alamannic zone were handmade. Siegmund’s
conclusion is that despite considerable variation within each category of
artefacts, the ethnic boundary created by such means was maintained
throughout the fifth to seventh centuries.37
But was that truly an ethnic boundary? Sebastian Brather rejects the
idea that anyone could be made a Frank or an Alaman by the simple
36
Ina-Maria Greverus, Kultur und Alltagswelt. Eine Einführung in Fragen der Kulturanthropologie
(Munich, 1978).
37
Frank Siegmund, ‘Alemannen und Franken. Archäologische Überlegungen zu ethnischen
Strukturen in der zweiten Hälfte des 5. Jahrhunderts’, in D. Geuenich (ed.), Die Franken und
die Alemannen bis zur ‘Schlacht bei Zülpich’ (496/97) (Berlin and New York, 1998), pp. 558–80,
at pp. 560–1 and 574.
Early Medieval Europe   ()
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd
Ethnicity in medieval archaeology 169
deposition of a glass beaker or a handmade pot, respectively, into his
or her grave.38
According to him, Siegmund’s exercise demonstrated
nothing else than the existence of a difference of habitus between the
two sides of the frontier. This amounts to no less than an explicit
rejection of the link between habitus and ethnicity. As a consequence,
Brather excludes ethnicity, without noticing that by explaining material
culture patterning as the result of habitus, he had turned the clock back
to the culture-historical approach against which he directs his criticism.
For if the material culture patterning is not the result of deliberate choices
inspired by a desire to mark difference, then the habitus is nothing but
a cultural ‘norm’, whose consequences are always outside the awareness
of the actors and always work ‘behind their backs’. In other words, the
Franks and the Alamans were different without knowing it. Such a
conclusion is in direct contradiction to Brather’s otherwise firm conviction
that ethnicity is a matter of complex representation of the entire culture
and of common origin, and not an objective combination of traits.39
Symbols, ethnicity and style
Sebastian Brather’s placing a premium on representation and symbols
opens the door for a ‘subjective’ approach to ethnicity. Ethnicity is truly
represented through such things as certain dress elements, speech forms,
lifestyles, food ways, and the like. But Brather believes that the selection
of these elements is the result of gambling: should the selection turn
out to be wrong, the very identity and therefore existence of the group
is threatened. Once the symbols are gone, the ethnic group disappears.40
While it is true that ethnicity is concocted out of a few cultural elements
(as opposed to the whole ‘culture’), those elements are not arbitrarily
chosen, to the extent that they are meant to mark the boundaries of the
ethnic group as visibly as possible for outsiders to acknowledge the
existence of that group.41
Moreover, the ethnic group does not originate
in the symbols used to mark it as distinct from others. The symbols as
well as the group are in fact the result of human action. It is precisely
that agency that Brather’s approach ignores. In the words of a historian
of the Vienna school, ‘to make ethnicity happen, it is not enough just
38
Brather, Ethnische Interpretationen, p. 297.
39
Brather, Ethnische Interpretationen, p. 313: ‘Ethnische Identität ist die komplexe Vorstellung
gemeinsamer Kultur und gemeinsamer Abstammung und keine objektive Merkmalskombination’
(original emphasis).
40
Brather, Ethnische Interpretationen, p. 108.
41
Sam Lucy, ‘Ethnic and Cultural Identities’, in M. Díaz-Andreu and S. Lucy (eds), The
Archaeology of Identity: Approaches to Gender, Age, Status, Ethnicity, and Religion (London and
New York, 2005), pp. 86–109, at pp. 96–7.
Early Medieval Europe   ()
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd
170 Florin Curta
to be different’.42
The selection of ethnic symbols is a political strategy
in the same way that choosing a certain dress style is for the construction
of social status.43
The success or failure of such strategies does not depend
upon the quality or number of symbols chosen to represent them. Instead,
that is a matter of how much power is exercised by real people over
equally real people. In other words, if social constructs such as ethnicity
do not reflect ‘social reality’, they can certainly shape it in accordance
with the interests of those in power. Material culture is not a passive
reflection of ethnicity, but an active element in its negotiation.44
The idea that political ethnicity is a goal-oriented ethnicity goes back
to the influential work of Abner Cohen, a member of the Manchester
School of social anthropology known for his work on informal social
organization.45
Manchester School anthropologists were concerned
especially with colonization and urbanization, as well as with other
dramatic changes taking place in African societies after contact with
Europeans and other indigenous groups. Cohen’s approach to ethnicity
was therefore very pragmatic: to him ethnic identity was formed by
internal organization and stimulated by external pressures, and held not
for its own sake but to defend an economic or political interest. Such
an ethnicity needed to be built upon some pre-existing form of cultural
identity rather than be conjured up out of thin air. Ian Hodder’s ethnoarchaeological
fieldwork in East Africa may be viewed as an attempt
to test that conclusion. Hodder’s idea was to study how spatial patterning
of artefacts related to ethnic boundaries, and he chose the district
of Baringo in Kenya. He found that, despite interaction across tribal
boundaries, clear material culture distinctions were maintained in a
wide range of artefact categories. He argued that distinct material culture
boundaries were foci of interaction, not barriers. Hodder showed that
material culture distinctions were in part maintained in order to justify
between-group competition and negative reciprocity, and that such patterning
increased in times of economic stress. However, not all cultural
traits were involved in such differentiation, since, typically, interaction
continued between competing groups. Hodder thus suggested that the
use of material culture in distinguishing between self-conscious ethnic
42
Walter Pohl, ‘Telling the Difference: Signs of Ethnic Identity’, in W. Pohl and H. Reimitz (eds),
Strategies of Distinction: The Construction of Ethnic Communities, 300–800 (Leiden, Boston and
Cologne, 1998), pp. 17–69, here pp. 21–2.
43
For an archaeological example, see Harold Mytum, ‘The Vikings and Ireland: Ethnicity,
Identity, and Cultural Change’, in J.H. Barrett (ed.), Contact, Continuity, and Collapse: The
Norse Colonization of the North Atlantic (Turnhout, 2003), pp. 113–37.
44
Ian Hodder, Reading the Past: Current Approaches to Interpretation in Archaeology (Cambridge,
1986), p. 9.
45
Abner Cohen, Custom and Politics in Urban Africa: A Study of Hausa Migrants in Yoruba
Towns (London, 1969).
Early Medieval Europe   ()
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd
Ethnicity in medieval archaeology 171
groups would lead to discontinuities in material culture distributions
that may enable the archaeologist to identify such groups. The specific
kind of relations between groups is often related to the internal organization
of social relationships within one or the other group. In the case
of the Baringo, between-group differentiation and hostility was linked
to the internal differentiation of age sets and the domination of women
and young men by older men, a conclusion later substantiated by Roy
Larick’s fieldwork among the Loikop (Samburu) of northern Kenya.46
Hodder provides a good example of Cohen’s concept of goal-oriented
ethnicity. The Maasai sometimes ‘become’ Dorobo in order to escape
drought, raiding or government persecution. But, although the Dorobo
had a real separate existence in the conscious thoughts of those calling
themselves by that name, there was no symbolic expression of any
difference between Dorobo and Maasai.47
In other words, different groups
may manipulate material culture boundaries in different ways, depending
upon the social context, the economic strategies at stake, the particular
history of the social and economic relations, and finally, the particular
history of the cultural traits chosen for marking the ethnic boundaries.
The history of the Middle Ages is replete with examples of just that.
To cite one among many relevant cases, in the Latin kingdom of Cyprus,
rulers wanted to project the image of a society with a thoroughly Western,
chivalric look. By 1200, the Latins in the kingdom formed a close-knit
caste from which local Greeks were excluded. Not only were Latins
obsessed with hunting and falconry, but the earliest instance of knights
dressing up as characters from the Arthurian cycle occurred in Cyprus
in the mid-1220s at the jousts held as part of the celebrations surrounding
the knighting of the sons of the Lord of Beirut, John of Ibelin (1197–
1226).48
Whether a marker of ethnic identity or social rank, or perhaps
both, material culture in this case was not a mirror of social identity.
46
Ian Hodder, Symbols in Action: Ethnoarchaeological Studies of Material Culture (Cambridge,
London and New York, 1982), pp. 27, 31, 35, 85, 187 and 205; Roy Larick, ‘Age Grading and
Ethnicity in the Style of Loikop (Samburu) Spears’, World Archaeology 18 (1986), pp. 269–83,
and ‘Warriors and Blacksmiths: Mediating Ethnicity in East African Spears’, Journal of
Anthropological Archaeology 10 (1991), pp. 299–331. Kathleen Deagan, Spanish St. Augustine:
The Archaeology of a Colonial Creole Community (New York, 1983), pp. 99 and 271 offers
similar conclusions. The sixteenth-century Spanish settlers of St Augustine (Florida) strove to
create ethnic boundaries around their community by means of a unique Hispanic-American
cultural tradition, itself the result of mestizaje, or Spanish-Indian intermarriage. Deagan found
that elements of that cultural tradition that were linked to male activities and to public stages
of social interaction (tablewares, items of clothing and personal adornments) tended to retain
their Spanish appearance, while those cultural elements that were female-centred and focused
on the household (internal home furnishings, household pottery) were often of Native
American manufacture.
47
Hodder, Symbols, p. 104.
48
Peter W. Edburg, ‘Latins and Greeks on Crusader Cyprus’, in D. Abulafia and N. Berend
(eds), Medieval Frontiers: Concepts and Practices (Aldershot, 2002), pp. 133–142, at p. 135.
Early Medieval Europe   ()
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd
172 Florin Curta
Instead, that identity was at least in part the result of the manipulation
of material culture. To be a Latin meant, among other things, to have
access to the re-enactment of the Arthurian cycle and be allowed to dress
up as one of its characters.
A key component of ethnic boundary building is therefore style,
which may be viewed as the pattern people make around a particular
event, recalling and creating similarities and differences. This became
particularly clear during the ‘style debate’ of the 1980s, in which several
archaeologists argued over the communicative role of material culture.49
Max Wobst first proposed the idea that style operated as an avenue of
communication.50
James Sackett argued that style was a passive aspect
of artefacts, an intrinsic or adjunct function of material culture, which
he called isochrestic variation. According to Sackett, the isochrestic variation
permeated all aspects of social and cultural life and provided the
means by which members of a group express their mutual identity,
coordinate their actions, and bind themselves together. Style thus became
an idiomatic or diagnostic of ethnicity. No matter what meaning style
may have had for its producers in the past, its consequences were not
completely within their awareness and thus often worked ‘behind their
backs’.51
By contrast, Polly Wiessner argued that as a form of non-verbal
communication through doing something in a certain way that communicates
about relative identity, style is an intentional, structured system
of selecting certain dimensions of form, process or principle, function,
significance, and affect from among known, alternate possibilities to
create variability within a corpus of artefacts. According to Wiessner,
a distinction should be made between ‘emblemic’ and ‘assertive’ styles.
Emblemic styles have distinct referents and transmit clear messages to
defined target populations about conscious affiliation or identity, while
assertive styles are personally based and carry information supporting
individual identity. Because an emblemic style carries a distinct message,
it should undergo strong selection for uniformity and clarity, and because
49
For a review of the debate, see Michelle Hegmon, ‘Archaeological Research on Style’, Annual
Review of Anthropology 21 (1992), pp. 517–36; H.M. Wobst, ‘Style in Archaeology and Archaeologists
in Style’, in E. Chilton (ed.), Material Meanings: Critical Approaches to the Interpretation
of Material Culture (Salt Lake City, 1999), pp. 118–32.
50
H.M. Wobst, ‘Stylistic Behavior and Information Exchange’, in C.E. Cleland (ed.), For the
Director: Research Essays in Honor of James B. Griffin (Ann Arbor, 1977), pp. 317–42.
51
James R. Sackett, ‘Style and Ethnicity in the Kalahari: A Reply to Wiessner’, American
Antiquity 50 (1985), pp. 154–9; ‘Isochrestism and Style: A Clarification’, Journal of Anthropological
Archaeology 5 (1986), pp. 266–77; and ‘Style and Ethnicity in Archaeology: The Case
for Isochrestism’, in M.W. Conkey and C.A. Hastorf (eds), The Uses of Style in Archaeology
(Cambridge, 1990), pp. 32–43. For similar views, see Natalie R. Franklin, ‘Research with
Style: A Case Study from Australian Rock Art’, in S. Shennan (ed.), Archaeological Approaches
to Cultural Identity (London, 1989), pp. 278–90, at p. 278; Esther Pasztory, ‘Identity and
Difference: The Uses and Meanings of Ethnic Styles’, Studies in the History of Art 27 (1989),
pp. 15–38, at p. 17.
Early Medieval Europe   ()
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd
Ethnicity in medieval archaeology 173
it marks and maintains boundaries of group membership, it can be
distinguished archaeologically by uniformity within its realm of function.52
When used as a tool in social strategies, style provides the potential
for the control of the meaning and thus for power. Recent studies
demonstrate that an emblemic style appears at critical junctures in the
regional political economy, when changing social relations would impel
displays of groups identity. With the initial evolution of social stratification
and the rise of chiefdoms, considerable stylistic variability may exist
between communities in clothing and display items. At a regional level,
however, iconography and elite status become important to legitimize
and ‘naturalize’ the inherent inequality that exists in such societies.
Extensive interchiefdom trade and shared political ideology serve to
deliver rare and foreign objects linked symbolically to universal forces.53
An examination of the sixth-century archaeological material from the
Middle Danube region of present-day Hungary may illustrate this
discussion of the active use of material culture in the creation of cultural
identities.54
Most finds in that region concentrate either on the right
bank of the Danube or on the left bank of the Tisza River.55
There are
very few finds between the two rivers, an area that may have served as
a ‘no man’s land’ separating the Lombards from the Gepids. But this
buffer zone was not a barrier, for considerable interaction occurred
across the ‘no man’s land’. For example, in the early 500s the Lombard
king Wacho married a Gepid princess.56
In the late 540s, a candidate to
the Lombard throne named Hildigis is said to have fled to the Sclavenes,
who presumably lived somewhere near the Gepids. When Emperor
Justinian decided to turn the Lombards into his new allies, Hildigis fled
52
Polly Wiessner, ‘Style and Social Information in Kalahari San Projectile Points’, American
Antiquity 48 (1983), pp. 253–76; ‘Style or Isochrestic Variation? A Reply to Sackett’, American
Antiquity 50 (1985), pp. 160–6; and ‘Is There a Unity to Style?’, in Conkey and Hastorf (eds),
Uses of Style, pp. 105–12.
53
Castle McLaughlin, ‘Style as a Social Boundary Marker: A Plains Indian Example’, in Auger
et al. (eds), Ethnicity and Culture, pp. 55–66; Timothy Earle, ‘Style and Iconography as
Legitimation in Complex Chiefdoms’, in Auger et al. (eds), Uses of Style, pp. 73–81; Ian
Hodder, ‘Style as Historical Quality’, in Auger et al. (eds), Uses of Style, pp. 44–51.
54
What follows is a summary of a longer discussion to be found in Florin Curta, The Making
of the Slavs: History and Archaeology of the Lower Danube Region, c.500–700 (Cambridge, 2001),
pp. 197–203.
55
Joachim Werner, Die Langobarden in Pannonien. Beiträge zur Kenntnis der langobardischen
Bodenfunde vor 568 (Munich, 1962), p. 116; Neil Christie, The Lombards: The Ancient Longobards
(Oxford and Cambridge, 1995), p. 55. For the conflict between Lombards and Gepids,
see Walter Pohl, ‘Die Gepiden und die Gentes an der mittleren Donau nach dem Zerfall des
Attillareiches’, in H. Wolfram and F. Daim (eds), Die Völker an der mittleren und unteren
Donau im fünften und sechsten Jahrhundert. Berichte des Symposions der Kommission für Frühmittelalterforschung,
24. bis 27. Oktober 1978, Stift Zwettl, Niederösterreich (Vienna, 1980),
pp. 239–305; and ‘Die Langobarden in Pannonien und Justinians Gotenkrieg’, in D. Bialeková
and J. Zábojník (eds), Ethnische und kulturelle Verhältnisse an der mittleren Donau vom 6. bis
zum 11. Jahrhundert. Symposium Nitra 6. bis 10. November 1994 (Bratislava, 1996), pp. 27–36.
56
Paul the Deacon, History of the Lombards I. 21.
Early Medieval Europe   ()
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd
174 Florin Curta
to the Gepids, followed by a retinue of renegade Lombards and Sclavenes.57
Goods seem to have also moved freely between the Lombards and the
Gepids. At least this may be part of the explanation for the fact that
the construction of male identity on both sides of the ‘no man’s land’
operated with the same categories of artefacts. Judging from grave-goods
associated with male burials, there were no differences between Gepid
and Lombard warriors, at least not in their appearance. A specific combination
of weapons – sword, spear and shield – was typically associated
with high-status male burials on both sides of the ‘no man’s land’. Clear
material culture distinctions were maintained not through male accoutrements,
but through a range of artefacts found in female graves. Most
interesting in this respect is the distribution of bow fibulae, which shows
a sharp contrast between western and eastern Hungary. While disc- and
S-shaped brooches are rare in the area east of the Tisza, they were
particularly popular in western Hungary. Completely different types
of brooches were in fashion at the same time on either side of the ‘no man’s
land’. Brooch classes such as Goethe, Cividale, Ravenna and Castel
Trosino that are represented in western Hungary have no analogies outside
that region, except Italy.58
Conversely, the most popular fibulae in eastern
Hungary were those of the Hahnheim class, many of which have parallels
in Germany and France.59
The distribution thus reveals two different styles
of brooch-use, which may be interpreted as two different ways to convey
information about the relative identity of the brooch owners. The patterns
and contrasts created did not produce ethnically specific artefacts. Very
few, if any, brooch classes were creations ex nihilo and many were either
‘imports’ from other regions or replicas of ‘imported’ specimens. In other
words, no particular class of artefacts could be diagnosed as either
‘Gepid’ or ‘Lombard’. The ethnic boundary emerged from the manipulation
of specific types, without assigning an ‘ethnic value’ to any one
of them. Since most, if not all, brooch classes may be dated to the late
fifth and first half of the sixth century, it is important to note that the
pattern visible in the distribution of all specimens came into existence
at the time of the increasing rivalry between the two groups.60
57
Procopius of Caesarea, Wars VII.35.16 and 19.
58
For detailed descriptions of these types of fibulae, see Herbert Kühn, Die germanischen
Bügelfibeln der Völkerwanderungszeit in Süddeutschland (Graz, 1974), pp. 996–1006, 1187–91,
1217–24, and 1239–48.
59
Herbert Kühn, Die germanischen Bügelfibeln der Völkerwanderungszeit in der Rheinprovinz (Graz,
1965), pp. 151–3; and Die germanischen Bügelfibeln der Völkerwanderungszeit in Süddeutschland,
pp. 799–802. See also Alexander Koch, Bügelfibeln der Merowingerzeit im westlichen Frankenreich
(Bonn, 1998), pp. 200–1.
60
Brather is therefore right in rejecting the idea that brooches per se can be an expression of
ethnic identity. Equally to be rejected, however, is his interpretation of such dress accessories
as ‘the continued existence of older cultural traditions even under new political conditions’
(Brather, ‘Ethnic Identities’, pp. 149–76, at p. 161).
Early Medieval Europe   ()
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd
Ethnicity in medieval archaeology 175
The Lombard–Gepid wars of the mid-sixth century may have indeed
contributed to the consolidation among both Lombards and Gepids of
specific dress styles. That such styles were more prominently displayed
by women than by men is the result of the position that aristocratic
women had in the conflict. It has long been noted that the meaning of
dress is a form of social knowledge, where messages become ‘naturalized’
in appearance.61
Since clothing serves to convey information, dress may
be seen as a symbolic ‘text’ or ‘message’, a visual means of communicating
ideas and values.62
One important aspect of the communicative
symbolism of dress is its capacity for providing locative information,
referring to either the individual’s physical location in space or to his
or her position within the social network. Because dress has a distinct
referent and transmits a clear message to a defined target population
about conscious affiliation and identity, it may be treated as a form of
emblemic style.
That brooches appear exclusively in female burials shows that the real
challenge for archaeologists is not so much that ethnicity in the medieval
past did not have any material culture correlates, but that distinctive
features, such as dress accessories may signal other forms of social identity,
such as gender, age or class.63
Ethnicity was a form of social identity,
often combined with, rather than in opposition to, gender. However,
archaeologists normally treat these two forms of social identity separately
and often favour ethnicity over gender. This is especially true for rich
female burials with bow brooches, in the analysis of which gender has
often been neglected in favour of interpretations overemphasizing
the role of brooches as markers of ethnic identity.64
One reason for this
bias may be that many, if not all archaeologists tend to treat gender as
61
Marie Louise Sørensen, ‘The Construction of Gender through Appearance’, in D. Walde and
N.D. Willows (eds), The Archaeology of Gender: Proceedings of the Twenty-Second Annual
Conference of the Archaeological Association of the University of Calgary (Calgary, 1987), pp. 121–9,
here p. 122. See also Marilyn Revell DeLong, The Way We Look: A Framework for Visual
Analysis of Dress (Ames, 1987); Odile Blanc, ‘Historiographie du vêtement: un bilan’, in Le
vêtement. Histoire, archéologie et symbolique vestimentaires au Moyen Age (Paris, 1989), pp. 7–33.
Brather does not believe that it is at all possible to detect ethnic differences in clothes because
too many social identities can in fact be communicated simultaneously through dress (Brather,
Ethnische Interpretationen, p. 411).
62
Cherri M. Pancake, ‘Communicative Imagery in Guatemala Indian Dress’, in M. Blum
Schevill, J.C. Berlo and E.B. Dwyer (eds), Textile Traditions of Mesoamerica and the Andes:
An Anthology (New York, 1991), pp. 45–62. See also Jean-Thierry Maertens, Dans la peau des
autres: essai d’anthropologie des inscriptions vestimentaires (Paris, 1978).
63
Thomas Wallerström, ‘On Ethnicity as a Methodological Problem in Historical Archaeology’,
in H. Andersson, P. Carelli and L. Ersgård (eds), Visions of the Past: Trends and Traditions in
Swedish Medieval Archaeology (Lund, 1997), pp. 299–352, at p. 303.
64
Bonnie Effros, ‘Dressing Conservatively: Women’s Brooches as Markers of Ethnic Identity?’,
in L. Brubaker and J.M.H. Smith (eds), Gender in the Early Medieval World: East and West,
300–900 (Cambridge and New York, 2004), pp. 165–84.
Early Medieval Europe   ()
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd
176 Florin Curta
a given and not so much as socially constructed as ethnicity. Nonetheless,
much like ethnicity, the representation of gender through burial
assemblages is not a direct reflection of social practice.65
In the east Baltic region, gender distinctions are clearly visible in late
fifth- and early sixth-century cemeteries. For example, in eastern Latvia
and western Lithuania, male and female bodies were buried in opposite
directions. In south-western Latvia, male burials were additionally marked
by the inclusion of horse skulls on top of the body.66
Clear gender
distinctions are also visible in grave-goods. In Pagrybis (western Lithuania),
males were buried with fibulae, bronze torcs and weapons, while female
burials produced tools, as well as headdress garlands with bronze spirals,
bracelets, and dress pins joined by chains. Much has been made of the
typical female burial kit consisting of headdress garlands, bracelets and
dress pins. This has more often than not been viewed as an ethnic
marker, the ‘national costume’ of the earliest Lithuanians.67
However,
at a closer examination this interpretation proves to be wrong.68
In
Pagrybis, the standard set of dress accessories appears only in burials
with skeletons of women who died between twenty and thirty years of
age. Neither girls, nor older women were given such adornments. By
contrast, the presence of bronze dress accessories in male burials does
not seem to have anything to do with age. Silver dress accessories,
which are exclusively found in male burials, appear in graves of both
older men and boys. Unlike girls, boys could be ascribed status upon
death. Grave 55 in Pagrybis contained the skeleton of a ten-year-old
boy buried together with the head and the legs of a sacrificed horse,
a practice that is otherwise attested only with graves of adult men.
Despite his young age, the boy was buried with a bronze torc and a
silver crossbow brooch, both indicators of high status.69
The evidence
thus suggests that distinctions based on social status were the most
important for male burials. Indeed, the majority of such burials excavated
in Pagrybis each had only a spearhead, an axe, or a knife. By contrast,
65
For the difficulties of ‘reading’ burial assemblages as a direct mirror of gender divisions in
society, see Heinrich Härke, ‘Die Darstellung von Geschlechtergrenzen im frühmittelalterlichen
Grabritual: Normalität oder Problem?’, in W. Pohl and H. Reimitz (eds), Grenze und
Differenz im frühen Mittelalter (Vienna, 2000), pp. 180–96, at pp. 193–6.
66
Andris Caune, ‘Die Gräbertypen und Bestattungssitten im Ostbaltikum in der Zeit vom 1.
bis 13. Jahrhundert’, in F. Horst and H. Keiling (eds), Bestattungswesen und Totenkult in ur- und
frühgeschichtlicher Zeit (Berlin, 1991), pp. 257–74, at p. 263.
67
Most typical for this approach is Regina Volkaité-Kulikauskiené, Senoves lietuviu drabuziai ir
ju papuosalai (I–XVI a.) (Vilnius, 1997).
68
The following is based on Laima Vaitkunskiené, ‘Moters ir vyro statusas zemdirbi2 bendruomenéje
V–VI amziais’, Lietuvos TSR Moksl4 Akademijos darbai. A serija 1 (1985), pp. 74–85; and ‘The
Formation of a Warrior Elite during the Middle Iron Age in Lithuania’, Archaeologia Baltica
1 (1995), pp. 94–106.
69
For the role of silver artefacts as markers of social status in early medieval cemeteries in
Lithuania, see Laima Vaitkunskiené, Sidabras senoves Lietuvoje (Vilnius, 1981).
Early Medieval Europe   ()
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd
Ethnicity in medieval archaeology 177
age distinctions seem to have been paramount for female burials, which
suggests the existence of such gender categories as ‘married’ and ‘nonmarried’.
Moreover, most high-status men of older age were buried
together with young females. In Grave 65, a male and a female holding
hands were facing each other. The presence of a one-edged sword and
a silver torc associated with the male skeleton has rightly been interpreted
as an indication of high status. By contrast, the female skeleton
was unusual in many respects. First, unlike single female burials in the
cemetery, it was buried in the same direction as the male skeleton.
Second, although aged twenty to twenty-five, the woman in Grave 65
had no standard kit of dress accessories and no traditional tools, such
as spindle whorls or awls. A spearhead placed between the male and the
female skeletons was interpreted as a marker of social distinction, which
further suggests that the female buried together with the high-status
man in Grave 65 was of inferior social status, arguably a slave.70
Material culture is therefore fundamentally social, and artefacts are
rendered ‘appropriate’ for use only in social context. Decisions about
the use of artefacts are embodied in artefacts themselves in terms of the
conventions of culture. Artefacts are not properties of a society, but part
of the life of that society. They cannot and should not be treated as
‘phenotypic’ expressions of a preformed identity. What should concern
medieval archaeologists is not so much what people do, what kind of
pots or brooches they make, what shape of houses they build, but the
‘way they go about it’.71
The archaeological record, ethnicity and power
Despite their plea for shifting the emphasis of archaeological research
from ethnicity to social issues, Sebastian Brather and a number of other
similarly minded archaeologists in Eastern Europe believe that archaeologists
are not even capable of identifying the boundaries of social groups.
Without independent sources, such as written accounts, archaeologists
cannot entertain any hopes of describing the meaning of symbols or the
particular situations in which they were created and used.72
The ‘floating
gap’ between the communicative and cultural memory of any social group
supposedly prevents archaeologists from reconstructing the meanings
initially attached to symbols manipulated to mark the boundaries of the
70
Whether or not she was sacrificed on the occasion of the man’s death remains unclear. See
Vaitkunskiené, ‘The Formation of a Warrior Elite’, p. 100.
71
Paul Graves-Brown, ‘All Things Bright and Beautiful? Species, Ethnicity, and Cultural
Dynamics,’ in Graves-Brown, S. Jones and C. Gamble (eds), Cultural Identity and Archaeology,
pp. 81–95, at pp. 90–1.
72
Brather, Ethnische Interpretationen, pp. 337 and 369.
Early Medieval Europe   ()
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd
178 Florin Curta
group. One needs written sources to decide which symbols were used
in any particular society for building ethnic boundaries.73
But without
evidence of perfect overlap between written and archaeological sources,
the argument is simply circular. Many sources were written by outsiders
and as such reflect the authors’ idiosyncrasies, not the self-identification
of the ethnic groups described. Even when written by ‘insiders’, such
sources rarely describe the artefacts archaeologists usually find in excavations.
How could then such sources help them understand ethnicity?
What about periods in history for which no written sources exist?
An emphatically functionalist declaration underlines this position:
the archaeological sources are artefacts, which are not specifically made
to carry a certain representation of the past, but to respond to some
immediate economic or social demand. But the idea that written sources
speak ‘naturally’, whereas artefacts are ‘inherently’ dumb has already been
denounced as a presentist fallacy of our logocentric times.74
According
to such views, what separates archaeology from history is the fact that
the former deals with material remains (realia), not with notions fixed
by language. By contrast, the historian works in the sphere of thinking
and language.75
Historical records involve encoding of ideas, because
human actors, not physical processes produced them and established
the language code whereby these documents actually recorded past
events. Historical records are thus viewed as ‘active’, in that they are
thought as actively communicating messages and information that may
transform the reader’s ideas or behaviour. But there are at least two
different meanings that can apply to the archaeological record: ‘things’
found which are supposed to be the result of the causes they record;
and ‘things’ found which are supposed to bear the significance of the
symbols they record. The first meaning refers to what Linda Patrik called
the ‘physical model’ of the archaeological record, which represents the
cardinal paradigm of modern archaeology.76
The basic assumption on
which this model is based is that the physical remains are the causal
effects of physical processes. As a consequence, what the archaeologist
is supposed to do is simply to excavate and analyse the ‘record’ in order
to reconstruct the causal phenomenon. Geologists and palaeontologists
use index fossils to identify strata belonging to a particular geological
epoch, and diagnostic artefacts (Leittypen) are nothing more than the
archaeological equivalent of index fossils.77
The idea behind the index
73
Brather, Ethnische Interpretationen, pp. 570 and 577.
74
Its most elegant refutation remains John Moreland, Archaeology and Text (London, 2001).
75
Lev S. Klein, ‘To Separate a Centaur: On the Relationship of Archaeology and History in
Soviet Tradition’, Antiquity 67 (1993), pp. 339–48, at p. 341.
76
Linda Patrik, ‘Is There an Archaeological Record?’, Advances in Archaeological Method and
Theory 8 (1985), pp. 27–62, at p. 33.
77
For index fossils in archaeology, see Trigger, History, p. 96.
Early Medieval Europe   ()
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd
Ethnicity in medieval archaeology 179
fossil is in fact that the recording connection between the record and
what is recorded is a strictly causal relationship, maintained in a physical
medium. Given sufficient knowledge of the laws governing the
causal connection, the original body can be inferred from the fossil
record on the basis of the uniformitarian principle.78
The fossil record
is therefore ‘passive’, because it records its causes by preserving the static
effects of those causes. In other words, the fossil record is something
like an imprint. The ‘physical model’ takes primarily physical remains
for physical marks, human artefacts and residues for past organic bodies
and traces, which the archaeologist excavates and analyses, assuming
that the archaeological record can never be exactly duplicated. Such a
model implies that the regularities in the features and spatial order of
various components of the record can be expressed as universal or probabilistic
laws. The ‘physical model’ is therefore a deductive-nomological
model in that it attempts to infer deductively the past causes of the
record. Archaeologists operating with this model of archaeological
record do not usually distinguish between three different recording
connections linking physical things in the present to referents in the past:
recording connections of physical remains (e.g., skeletons); recording
connections of what is substantially equivalent to something else (e.g.,
fossils); and recording connections of unique, nomological traces (e.g.,
fingerprints).79
Embracing the ‘physical model’ amounts to rejecting the idea that
physical remains are residues of human actions, ideas and events with
human import, which archaeologists can decode, read and analyse.
There is little room in such views for intentionality, signs and human
agency. By focusing on the intentional codifying process of any cultural
activity, the ‘textual model’ of the archaeological record allows ‘a place
for individuals’ idiosyncratic actions, creativity or protest; just as some
authors break new ground (and perhaps break a few grammatical rules
in the process), so too certain individuals produce, use or discard material
items, they “bring things off ” in unique, unpredictable ways’.80
The ‘textual
model’ implies that the archaeological record is part of a symbolic system
and largely encodes ideas or general social behaviour. This actually
means that material culture is a ‘text’ to be ‘read’. Archaeologists are
therefore supposed to identify and study contexts in order to interpret
78
Formulated in 1785 by the Scottish philosopher James Hutton (1726–97), the uniformitarian
principle states that ‘the present is the key to the past’. When applied to geological processes,
this means that the creation of all ancient geological features may be attributed to the same
physical mechanisms that are seen in the present. See Dennis R. Dean, James Hutton and the
History of Geology (Ithaca, 1992).
79
Patrik, ‘Is There an Archaeological Record?’, pp. 35–55.
80
Patrik, ‘Is There an Archaeological Record?’, p. 37.
Early Medieval Europe   ()
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd
180 Florin Curta
meaning.81
It is particularly in this light that an archaeology of ethnicity
becomes possible.
Cultural practices and representations that become objectified as
symbols of group identity are derived from, and resonate with, the
habitual practices and experiences of the agents involved, but they also
reflect the instrumental contingencies of a particular situation. This was
certainly true for Louis the Pious who took seriously his father’s recommendation
to dress like Gascons when ruling over Aquitaine.82
Similarly,
King Dagobert’s envoy, Sicharius, had to dress up like a Slav in order
to be received by Samo, the king of the Wendish Slavs.83
In both cases,
what was at stake was not marking the difference, as with the Latin
noblemen of Cyprus, but on the contrary, the ability to ‘blend in’, a
strategy whose success depended directly upon knowledge of the habitual
practices of the target group, be that Gascons or Wends. Ethnic differences
are constituted simultaneously in the mundane as well as in the
decorative, and become ‘naturalized’ by continual repetition in both
public and private. It is the pattern created by such repetition that lends
itself for interpretation by archaeologists studying ethnicity. There are
of course different ways to interpret that pattern, and the degree to
which the context of social practice can be reconstructed varies considerably.
But to deny the possibility that ethnicity can be the explanation
for such a pattern is at best an exaggeration and at worst a demonstration
of ignorance.
Brather’s understanding of the archaeological record as imprint may
well be a reaction to the tendency of culture-historical archaeologists to
treat artefacts as epiphenomenal, that is as subsumed beneath something
else than themselves. The task of the traditional archaeologist was
indeed to identify diagnostic traits, as ethnic attribution depended upon
a symptomatic logic. But such an approach is meaningless: what do lists
of supposedly similar brooches tell us? What does it mean to say that
this or that brooch is Frankish or Gothic? The correct approach should
start by acknowledging that the brooch in question is what is left of the
past and not a secondary manifestation of something more primary, or
material, or real.84
The fact that several dress accessories associated with
emblemic styles have been found primarily in burial assemblages is an
indication of the public display of such styles, as burial in the early
Middle Ages was an important ‘beyond-the-household’ arena of social
81
Hodder, Reading the Past, p. 153.
82
Thegan, Vita Hludovici IV.
83
Fredegar, IV.68.
84
Michael Shanks, ‘Style and the Design of a Perfume Jar from an Archaic Greek City State’,
Journal of European Archaeology 1 (1993), pp. 77–106 (reprinted in R.W. Preucel and I.
Hodder (eds), Contemporary Archaeology in Theory: A Reader (Malden and Oxford, 1996),
pp. 364–493, at pp. 368 and 373–4).
Early Medieval Europe   ()
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd
Ethnicity in medieval archaeology 181
competition between groups.85
The praxis of ethnicity resulted in multiple
transient realizations of ethnic difference in public displays of group
identity and aspirations. Inasmuch as archaeologists strive to identify
the material correlates of ethnically specific behaviour rather than the
material symbols of ethnic identity, ethnic styles are not attributes, but
patterns people make around particular events of communal significance,
which recall and create similarities and differences. Between the fourth
and the sixth century, tombs, passageways of churches, synagogues, and
residences of Christians and Jews in towns and villages from the Golan
Heights were sometimes inscribed with powerful statements of limits
maintained and defended by divine powers. Such decorations operated as
‘voices’ in the local religious and ethnic competition.86
The ‘imagined
communities’ of late antiquity and the Middle Ages existed only in the
public commemoration of group boundaries, history and identity.
Ethnicity has often been defined as an essential orientation to the
past, to collective origin, a ‘social construction of primordiality’.87
The
key word is of course ‘construction’, for many inferences made by
culture-historical archaeology about cultural continuity, which was
commonly interpreted as ethnic continuity, have been invalidated by the
use of refined chronologies, which revealed instead a considerable degree
of discontinuity, recycling of old themes, and ‘renaissances’. Ethnic
displays are thus carefully staged invocations of pre-existing cultural
practices, a (re-)invention of norms. For example, we know that the
long, braided hair of the first Avar envoys that came to speak to Emperor
Justinian in 558 made a very strong impression on the inhabitants of
Constantinople. During the subsequent decades, a marker of perhaps
multiple forms of social identity was employed as an ethnic stereotype:
during the late sixth and early seventh century, Avars were known in
Constantinople as ‘the filthy race of long-haired barbarians’. Contemporary
burials in the Avar heartland of present-day Hungary produced
no evidence that braided hair was in any way valued by Avars themselves
as a marker of ethnic identity. But the derogatory label attached
to their appearance was soon turned into a badge of identity. During
the so-called Middle Avar period, between c.620/50 to c.680, warriors
were buried together with pairs of richly adorned hair clips, the position
of which suggests that they were used for the ornamentation of long
85
Bonnie Effros, Caring for the Body and Soul: Burial and the Afterlife in the Merovingian World
(University Park, 2002).
86
Robert C. Gregg, ‘Making Religious and Ethnic Boundaries: Cases from the Ancient Golan
Heights’, Church History 69 (2000), pp. 519–57, at p. 556.
87
Hoyt S. Alverson, ‘The Roots of Time: A Comment on Utilitarian and Primordial Sentiments
in Ethnic Identification’, in R.L. Hall (ed.), Ethnic Autonomy – Comparative Dynamics:
The Americas, Europe and the Developing World (New York, Oxford and Toronto, 1979),
pp. 13–7, at p. 15.
Early Medieval Europe   ()
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd
182 Florin Curta
braids.88
Burials with hair clips cluster within the region between the
Danube and the Tisza rivers, which had served as ‘buffer zone’ in the
500s, but was now the core area of the Avar polity. The mid-600s was
a period of considerable instability. The failure of the siege of Constantinople
in 626 and the ensuing civil war brought the Avar qaganate on
the brink of extinction. It is no doubt under these circumstances of
dramatic social and cultural changes that a tradition of the late sixth
century was revived and turned into a badge of identity. A cultural trait
the Byzantines had employed to identify ‘a filthy race’ of barbarians has
now become a matter of concern for those who wished to be viewed as
Avars in both life and death. It is important to note that this recycling
of an ethnic stereotype is associated with men of high status, who were
often buried close to each other within one and the same cemetery.
Archaeologists may well interpret this phenomenon as the continuation
of a tradition first established at the time of the Avar conquest of the
Middle Danube region. In reality, the emblemic style associated with hair
clips was the invention of the second or third generation of Avar warriors.
More than six hundred years later, similar circumstances led to
another invention of traditions. At the time of their migration into
Hungary in the aftermath of the Mongol invasion of 1241, Cumans of
the first generation accepted baptism and Christian burial, but resented
being forced to let their beards grow and abandon their hairstyle (braids
with one or three tresses and the top of the head shaved).89
The strongest
resistance in that respect came from a few clans that rose to prominence
following the integration of the Cuman nobility into the highest
echelons of the medieval society of Hungary. Moreover, a re-invention
of the traditions of the steppe seems to have taken place during a period
of intense factionalism and political fragmentation that coincided with
the reign of King Ladislas IV (1272–90), himself the grandson of a
Cuman chieftain. The Chertan clan rose to political prominence under
such circumstances and it is no doubt with that family that three isolated
burials of warriors must be associated. These were unusual graves
in that in each one of them a horse was buried together with the human
corpse. No such burial seems to have been given to any Cuman of the
first generation in Hungary, but the practice is well attested on burial
sites in the northern Black Sea area under the control of the Cumans
88
J. Andrási, ‘Avar kori varkocsszorítók’, Archaeologiai Értesitö 123–124 (1996–97), pp. 85–123.
See also Eric Breuer, Byzanz an der Donau. Eine Einführung in Chronologie und Fundmaterial
zur Archäologie im Frühmittelalter im mittleren Donauraum (Tettnang, 2005), p. 58.
89
András Páloczi-Horvath, ‘Le costume coman au Moyen Age’, Acta Archaeologica Academiae
Scientiarum Hungaricae 32 (1980), pp. 403–27. See also Nora Berend, At the Gate of Christendom:
Jews, Muslims, and ‘Pagans’ in Medieval Hungary, c.1000–c.1300 (Cambridge, 2001),
pp. 250 and 261–2.
Early Medieval Europe   ()
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd
Ethnicity in medieval archaeology 183
before 1241.90
The tradition of the steppe was recycled in Christian
Hungary in order to respond to a demand for markers of ethnic differentiation
in times of political turmoil.
The fact that the emergence of emblemic styles often coincides in
time with political strife or military conflict points to crises as ideal
circumstances for the creation of ‘imagined communities’. The key
variable that must therefore be considered in explaining changes in ethnic
boundary maintenance is the differential distribution of power understood
as the probability of persons or groups carrying out their will when
opposed by others. Both dress styles and ‘traditions’ become relevant
particularly in contexts of changing power relations, which impel displays
of group identity. The greater the disparity in power between groups,
the higher the degree of boundary maintenance.91
In thirteenth-century
Ireland, the English government legislated against the Anglo-Norman
subjects who wore Irish clothes, had their heads half shaved and grew
their hair long at the back, ‘making themselves like the Irish in clothing
and appearance’.92
However, the continuous conflict between Irish
chieftains and Anglo-Norman lords led to much greater polarization.
While Anglo-Norman lords lived in motte-and-bailey castles, most
native Irish lords before c.1400 lived in crannogs, that is in residences
located on artificial islands in the middle of lakes. The distributions of
medieval mottes and crannogs in Ireland are mutually exclusive, and
this has been interpreted in terms of different modes of waging war.93
However, general changes in warfare after c.1400 brought about a considerable
alteration of ethnic boundaries. With the Gaelic Resurgence
of the 1400s, many mottes were abandoned and a process of Gaelicization
seems to have taken place in Anglo-Norman Ireland with many lords
of Anglo-Norman descent adopting the language, the customs, and the
social behaviour of their Irish counterparts. Both Irish and Anglo-Norman
lords began building tower houses, which became the hallmark of the
military architecture of fifteenth- to seventeenth-century Ireland. Their
distribution overlaps with that of both mottes and crannogs of the
previous centuries. By 1500, Ireland consisted of a large number of
virtually independent lordships controlled either by Gaelic lords or by
men of Anglo-Norman descent. The colonial government lost control
90
Petre Diaconu, Les Coumans au Bas-Danube aux XIe–XIIe siècles (Bucharest, 1978); Victor
Spinei, The Great Migrations in the East and South East of Europe from the Ninth to the Thirteenth
Century (Cluj-Napoca, 2003), pp. 217–40.
91
Randall H. McGuire, ‘The Study of Ethnicity in Historical Archaeology’, Journal of Anthropological
Archaeology 1 (1982), pp. 159–78, at p. 172.
92
Robert Bartlett, The Making of Europe: Conquest, Colonization, and Cultural Change 950–1350
(Princeton, 1993), p. 198.
93
Kieran Denis O’Conor, The Archaeology of Medieval Rural Settlement in Ireland (Dublin,
1998), pp. 97–101.
Early Medieval Europe   ()
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd
184 Florin Curta
over the island during the reign of Edward I and it was not until the
seventeenth century that Ireland was fully re-conquered. The disappearance
of the power responsible for the polarization of high-status site
distributions during the thirteenth century made room for increased
warfare and raiding between a great number of petty lords, all of relatively
equal power and similar ‘ethnic’ identity.94
Conclusion
Understanding ethnicity in the past presents a particular challenge.
Medieval ethnicity was just as embedded in sociopolitical relations as
modern ethnicity is. Ethnicity was socially and culturally constructed,
a form of social mobilization used in order to reach certain political
goals. Then, just as now, ethnic identity was built upon some preexisting
cultural identity, in a prototypic manner. But ethnicity was also
a matter of daily social practice, and as such it involved manipulation
of material culture. Since material culture embodies practices, emblemic
style was the way of communicating by non-verbal means about relative
identity. Because it carried a distinct message, it is theoretically possible
that it was used to mark and maintain boundaries, including ethnic
ones. But ethnicity was also a function of power relations. Both emblemic
styles and ‘traditions’ became relevant particularly in contexts of changing
power relations, which impelled displays of group identity. In most cases,
the study of both symbols and ‘traditions’ implies a discussion of the
power configuration in any given historical situation, with an emphasis
on the political forces that may have been responsible for the definition
of symbols, their organization and hierarchization.
The archaeologist striving to understand the current debate about
ethnogenesis opposing the Vienna to the Toronto school will undoubtedly
be disappointed, if he or she seeks methodological guidance and
theoretical advice on studying ethnicity. After all, the ‘debate about
ethnogenesis in recent scholarship is not about whether one model of
“ethnic discourse” or identity-formation should be substituted for another;
nor does it revisit discussion in the social sciences on whether or not
ethnic identity is a social construct’.95
In fact, the debate is mostly about
whether or not we can understand the past, about interpretative frameworks
within which to position narrative sources, and about issues of
modern historiography. But the archaeological alternative does not look
any more promising. Taken at face value, Brather’s approach to ethnicity
94
O’Conor, Archaeology, pp. 74 and 102.
95
Andrew Gillett, ‘Ethnogenesis: A Contested Model of Early Medieval Europe’, History Compass
4 (2006), pp. 1–12, at p. 3.
Early Medieval Europe   ()
© 2007 The Author. Journal Compilation © 2007 Blackwell Publishing Ltd
Ethnicity in medieval archaeology 185
accords with instrumentalist views that now predominate in anthropology
and sociology. But he stops short of applying that approach to
the analysis of the archaeological evidence. His intention may have been
to exorcize the demons of culture-history and to free the archaeology of
the Middle Ages from the ghost of Kossinna. But the room left open is
now haunted by another spectre, one that Marc Bloch aptly called ‘the
dangerous modern poison of empiricism parading as common sense’.96
University of Florida
96
Marc Bloch, Apologie pour l’histoire ou métier d’historien (Paris, 1974), p. 26: ‘les poisons,
aujourd’hui plus dangereux, de la routine érudite et de l’empirisme déguisé en sens commun’.