Visions of Community: Reflections on the role of Ethnicity in the Post-Roman
West, Byzantium and the Early Islamic World
In the last few years, the old debate about the ‘Fall of Rome’
has become rather heated once again. Did Rome fall, or was it transformed? And
did its fall lead to an ‘end of civilisation’, or were the barbarian peoples who
grabbed power in the provinces in reality ‘Rome’s last conquest’? What is so
disappointing about these debates is not only that some authors fall back on
rather simplistic models that already seemed superseded by more complex
interpretations. More importantly, they once again concentrate on the Western
Empire, and on the ‘barbarian’ kingdoms that were founded on its territory. But
Rome did not fall only once. It fell in at least four different ways. Apart from
the Western transition from empire to kingdoms in the course of the fifth
century, there was, in the early seventh, the Slavic rupture in the Balkans, and
only a little later the Islamic conquest of the East. As a fourth element, we
have to take the Byzantine transformation of eastern Rome into account. It is
striking at first glance how differently these four post-Roman worlds developed
(not even taking into account their internal differences, especially in the
West). Chris Wickham’s book
Framing the Early Middle
Ages has been ground-breaking in comparing social and economic
development throughout the post-Roman world. Peter Brown has charted the fates
of the different Christendom(s) in East and West.
In 2009, a conference in Vienna, under the title ‘Visions of Community’,
discussed a different field of comparison, in which there has been little
comparative research so far: the formation of communities and identities.
[1] Late Antiquity and the
early Middle Ages were to a large extent shaped by far-flung and wide-reaching
communities. That is all the more surprising because most of their inhabitants
were accustomed to life in small, more or less face-to-face groups: clans,
tribes, villages, or small cities. Still, the Roman/Byzantine and Sassanian
empires, Christianity and Islam, the Caliphate and even the Frankish kingdom
pursued very inclusive policies over wide distances. Such ambitious designs
required strong visions of community. These states also mobilized a wide variety
of resources: cultural, religious, civic, ethnic, military and, of course,
political. Dark Age identity politics seem to have been remarkably successful,
so that many models of identification shaped by early medieval uses are still
available (and, in some cases, contentious) today.
We still know too little about the transformations of these ‘visions of
community’ in the period between about 400 and 900 CE. The civic vision and the
rich grammar of classical culture that Rome had developed from Greek and
Hellenistic models became infused with Judaeo-Christian ideas of spiritual
community. Christian religion pervaded areas of life that had hitherto remained
aloof from ancient cult practice. This was the synthesis that Byzantium built
upon. Roman and Byzantine imperial identities accommodated ethnic differences,
but provided little room for their political use. Already Aristotle had remarked
that the Greeks lived in their
poleis ,
whereas the barbarians were divided into
ethnē . The Romans, more or less adequately, perceived of their
barbarian neighbours as
gentes , as peoples,
who lived in local communities or in rather loose and unstable confederations. But the
gentes over whom Rome once extended
its rule had long since faded into their respective
civitates and provinces, and into the overarching Roman
populus .
In late Antiquity, it was among soldiers of mostly barbarian origin on Roman
territory that new styles of ethnic identity appeared. Barbarian armies who
formed more or less precarious parts of the Roman military administration seem
to have relied increasingly on their ethnic solidarities as an element of
cohesion in a potentially hostile environment. As they asserted themselves as
partners and competitors of the imperial administration, the Western Empire was
gradually replaced by the kingdoms of the Goths, Vandals, Burgundians and
Franks. Their success was hardly due to the ancient and stable identities that
these peoples had brought from the forests of Germany. They were in fact rather
mixed groups who had frequently been reshuffled in the course of their
migrations. Strong self-affirmations of the Gothic or Frankish character of
these kingdoms were relatively slow to appear in the sources. The ideology of
Frankishness reached its peak in the late eighth century, under Carolingian
rule. And it was not because of their strong and indefatigable identities that
these peoples made an impact on European history. Most of the post-Roman
kingdoms fell within a few centuries; the kingdoms of the Vandals, Burgundians,
Ostrogoths, Visigoths and Lombards disappeared. But the model was successful. In
most parts of Europe, the political landscape continued to be shaped by
Christian kingdoms with ethnic designations. Broken as their histories were,
this model of community provided the basis on which modern nations could
develop.
It is easy enough to recognize the differences from the Islamic conquests. The
post-Roman kingdoms in the West were formed by ethnically-identified armies,
most of them with their families and following, who had already spent at least a
generation on Roman territory. They converted to the Christian religion of the
Romans, and distinguished themselves from the provincial Roman majority over
whom they ruled by their ethnic origin. The Islamic conquerors, on the other
hand, were proud of their distinctive religious identity. They came as invaders
who grabbed power in a spectacular series of victories over both neighbouring
empires. Sure enough, Arabic identity could inspire a sense of pride in many
members of the new ruling elite, and the newly-converted non-Arabic
mawālis were initially despised. But Arabic identity was
too vague to be truly distinctive, and Islam remained more important as a key to
power. The result, therefore, was an Islamic realm in which political power
transcended the tribal and ethnic structure at its basis, and where dynastic or
territorial polities did not build their authority on ethnic identity. It is
hardly conceivable that Iraq could have become a kingdom of the tribe of Bakr,
Syria a Ghassanid state or Egypt a kingdom of the ‘Akk. On the other had, in
Europe it would be difficult to imagine that, for example, Bishop Wulfila, the
‘apostle of the Goths’ in the fourth century, could have become a prophet and
forged a tribal coalition so powerful that the Wulfilites would have extended
their power over much of the Empire as the Muslims did.
Byzantium was a different case again. As a heir to the splendour of classical
civilisation, and to the glory of imperial Rome, it maintained a complex
identity that was Roman in a political sense, Hellenic (although the word was
rather avoided for a while for its connotations of paganism) in a cultural
sense, and, of course, Christian. It was ethnically open, as long as ethnic
identities did not become a key to power. However, in the periphery of the
Orthodox world, more or less ethnic states could develop, such as Armenia,
Bulgaria or Serbia. Their dynamic interplay of ethnicity, Christian religion and
political power was in many ways comparable to the ethnic states of the West. However, they had to compete for legitimacy with the Byzantine Empire, which
influenced their development for many centuries.
At a second glance, of course, things are not so easy. It is not that ethnicity
did not matter in the East, and was all-pervasive in the West. All three worlds
were dominated by strong religious identities that were supposed to have
precedence over any ethnic ties. God was not the God of one people alone;
everyone was called to follow him. Political power was believed to be derived
from God through Christ or Muhammad. No stable states were formed in the West
that were not Christian, and none in the East that were not Islamic. Yet, ethnic
loyalties played an important political role in the Islamic and the Western
world, and to a degree even in Byzantium. Obviously, the comparison between the
differing forms of ethnicity in the post-Roman world is also an important
test-case for general theories of ethnicity and identity. What exactly is
ethnicity? If we take the belief in common origin as a criterium, it is easy to
agree that the Bedouin tribes of the Arabian Peninsula or the Germanic peoples
of the West constitute classic cases of ethnicity. But was the overarching Arab
or Germanic identity ethnic? Did the Romans, and later the Byzantines, have a
‘Roman’ ethnic identity? What about the civic identities of the inhabitants of
Tours, Naples or Damascus, or about the provincial identities of Egypt or
Aquitaine? Were the allegiances of the people (
ahl ) of Syria
and the people of Iraq who clashed during the first
fitna at
the battle of Siffin in 657 ethnic to some extent? Or the Christian ‘Coptic’
communities in Egypt and Syria under Byzantine and then under Muslim rule? And
what about the
populus Christianus and the Muslim
umma as a whole? Perhaps not only the belief in a common
origin, but also ideas about a shared history and a common destiny mattered. The
“multiple identities of the Middle East”, to use Bernard Lewis’s phrase, but
also those of the West, cannot be reduced to a simple pattern. The variety of
forms of ethnicity, in a broader sense, was considerable.
In all these cases, it would make little sense to debate whether we define these
groups as ‘ethnic’ in the narrower sense, and to juxtapose them with ‘imperial’
or ‘religious’ identities as opposed patterns of identification. To distinguish
between ethnic states in the West, an empire in Byzantium and a religious
commonwealth in the Islamic world would be too simplistic, and bar the way to an
understanding of what the three political cultures had in common, after all. Perhaps we can describe the research strategies necessary for successful
comparison in the following way: Any larger ethnic group, religious community or
political realm cannot rely on face-to-face experience alone to make its members
feel that they belong. They all need shared ‘visions of community’, symbols of
identity and experiences of solidarity and common purpose. The foci of
identification—for instance, genealogy and ethnicity, cult and religion, city
and territory, kingdom or empire, language and habitus, myth and history, script
and scripture—were present in most parts of the post-Roman (and post-Sassanian)
world. But their respective balance, and their political uses, could be quite
different.
What made people act together in the field of power and politics? What were the
social spaces for cooperation and conflict on a supra-regional level? Which
‘imagined communities’ formed their frame of reference? I would call these
configurations of meaningful collective action social identities, although I am
aware that identity is a problematic term. To underline its relatively flexible
character, we could also speak of patterns of identification. But the exercise
of power is, as the German sociologist Niklas Luhmann has shown, about reducing
complexity and contingency, and identities could help to make political actions
more predictable. If there was, as is normal in complex societies, more than one
social identity that shaped individual behaviour, tensions could arise, but more
often, they formed relatively stable aggregates. One could easily be a Frank, a
Catholic Christian, a citizen of Tours and a noble warrior at the same time. It
was in the way these different identities interacted that post-Roman societies
seem to have differed most. The Christian
gentes in the West seem to have served the purpose of making the
world appear less contingent. Arab tribes may have been a more disruptive than
stabilizing element in the early Islamic world.
What role did ethnicity play in Western, Byzantine and in Islamic societies? One
difference in recent scholarship on the Islamic and the Western world is
obvious: Whereas ‘tribe’ is a key concept in Islamic studies, much debated but
hardly dispensable, it has largely been abandoned for the
gentes in the West; we tend to call Goths or
Franks ‘peoples’. The reason is not so much in the ideological overtones,
because there is hardly a modern term denoting ethnicity that is without
overtones. But the sources seem to convey a rather different image of the shape
and size of the respective groups in the East and in the West. The rich ethnic
terminology in Arabic, as far as I can see, reflects 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 and the Qais, to single
tribes, subtribes and clans. The Latin word ‘
gens ’, even more than the Greek
ethnos , can cover the whole range from a family to the Franks or
even, in late Antiquity, the Romans. Arabic literature presents and often
idealizes the archaic world of the Bedouin tribes, with their severe logic of
tribal obligations. A similar logic appears in some of the internal conflicts in
the early Islamic period, where support could be mobilized along traditional,
and newly-constructed genealogical lines. It is one of the characteristics of
tribal organisation to provide a simple and efficient logic to channel and
escalate conflict by involving additional groups according to an essentially
binary structure of inclusion or exclusion, regardless of their interest in what
the original cause of the dispute may have been. Of course, in reality tribal
affiliations were constantly modified to fit the expectations of the present. But although the tribal system had been fragmented by the dispersed settlement
in the conquered countries, it could still serve as a frame of reference to
mobilize support or express dissent.
It is striking that a similar tribal logic is relatively absent in conflict
narratives and other sources about the
gentes in the West. Interest in the ‘tribal’ past reflected
itself in a number of heroic stories, but on the whole it was far from
dominating early medieval writing. There is little that resembles the elaborate
and ever-changing Arab tribal genealogies. The closest we get to them are the
genealogies of the Anglo-Saxon royal families and those in the early Irish
sources. But they came from a world of small- to medium-size kingdoms in
constant competition with each other, quite unlike the large-scale kingdoms of
Goths, Franks or Lombards on the continent. We know a lot about the conflicts
between and within these kingdoms, and there is little trace of tribal
motivation in the narratives. There were a number of ethnic minorities who lived
within the Frankish kingdoms: Alamanni, Burgundians, Saxons, Alans, Syrians,
Jews and others; but even there, we have little notice of conflict driven by
ethnic rivalry. Post-Roman Europe can hardly be called a tribal society.
The role of ethnic identities in the political system cannot be understood
without their complex relationship to religious beliefs. Initially, both the
Christian
populus and the Islamic
umma had been recruited individually and from people of
different ethnic or tribal origin. Conversion really meant turning from an old
identity to a radically new one. In late Antiquity, for a time, it could seem
that Christendom was almost co-extensive with the Roman Empire, at least in the
West. Augustine and other Christian intellectuals had to try hard to steer clear
of too close an identification with the Christian Empire. But eventually,
barbarian peoples were also baptized. The Gospel of Matthew (28:19) had called
for proselytizing peoples: “Therefore go and make disciples of all nations”;
consequently, the approach to conversion tended to be a collective one. When the
Frankish king Clovis was baptized around 500 CE, thousands of Franks were also
converted. The pressure for collective conversion was especially strong in the
Carolingian empire, for instance in the Saxon mission. To become a subject of
the Franks meant becoming Christian; there was no way past that. But more often,
Christianisation proceeded parallel to the establishment or the reinforcement of
an independent state, as in the cases of Bohemia, Poland, Hungary or
Bulgaria.
Islamic conversion took a different course. Tabari repeatedly quotes Islamic
conquerors giving the defeated three options: “You may enter our religion, in
which case you will enjoy what we enjoy, and you will bear the obligations we
bear”; alternatively, they could pay the poll-tax, or continue their
resistance.
[2] This is,
of course, well known. The conquered peoples, at least Christians and Jews,
could keep their religion and some autonomy but were excluded from the right to
rule. Those who converted to Islam maintained their ethnic background to some
extent. But when Turks, Persians and Kurds began to rule over Islamic lands,
they did so in the name of Islam, hardly in the name of their ethnic
communities. Islam did not reaffirm their ethnic identities in the same way as
Christianity did in the West. The triumph of the Islamic conquests may have been
due to redirecting the inherent rivalries of tribal societies toward a course of
expansion, and to linking their ambitions to a powerful overarching religious
and political framework. But by its very success, the Caliphate made the
construction of particular powers within the Islamic realm problematic. Some of
them, like the Fatimids in Egypt, the Ummayads in Spain or the Ottomans,
maintained universalist pretences. Others relied on the prestige of a dynasty,
past or present. The early medieval West created the ethnic legends that could
much later serve as resources for the imagined communities of a nationalist
age—the baptism of Clovis, the empire of Charlemagne, the Hungarian land-taking,
the martyrdom of St Wenceslas and many others. A different set of myths
configured the social memories of the Islamic world: the Hejra, the conquests,
the death of the sons of Ali.
As this brief sketch indicates, ethnic and religious identities could interact in
different ways when they became politically salient. All three post-Roman
political cultures, the Latin West, Byzantium and the Islamic world, succeeded
in establishing a complex sense of community that united them. The differences
between them can tell us more about the different ways in which identities can
become ‘good to act with’. In this way, early medieval history can contribute to
a better understanding of the ‘visions of community’ of the present, of their
problems and of their potential.
Footnotes
Note 1
Visions of Community , forthcoming. Ed. W. Pohl,
C. Gantner and R. Payne. Aldershot.
Note 2
The History of al-Tabari vol. XI (Year 12), p.
31.