Fordham University


Home | Ancient History Sourcebook | Medieval SourcebookModern History Sourcebook | Byzantine Studies Page
Other History Sourcebooks: African | East Asian | Global | Indian | IslamicJewishLesbian and Gay | Science | Women's



Subsidiary SourcebooksAfricanEastern AsianGlobalIndianJewishIslamicLesbian/GayScienceWomen

Special ResourcesByzantiumMedieval WebMedieval NYC
Medieval MusicSaints' Lives
Ancient Law
Medieval Law
Film: Ancient
Film: Medieval
Film: Modern
Film: Saints

About IHSPIJSP Credits

Cyril Mango. Byzantium: The Empire of New Rome. Scribner's, 1980.

[beginning with pg. 13]



All empires have ruled over a diversity of peoples and in this respect the Byzantine Empire was no exception. Had its constituent population been reasonably well fused, had it been united in accepting the Empire's dominant civilization, it would hardly have been necessary to devote a chapter to this topic. It so happens, however, that even before the beginning of the Byzantine period - indeed, when the grand edifice of Rome started to show its first cracks towards the end of the second century AD- the various nations under Roman sway tended to move apart and assert their individuality. The rise of the Christian religion, far from healing this rift by the introduction of a universal allegiance, only accentuated it. We must, therefore, begin with the question: Who were the 'Byzantines'? In an attempt to answer it we shall undertake a rapid tour of the Empire, noting as we proceed the populations of the various provinces and the languages spoken by them. The time I have chosen is about 560 AD, shortly after the recovery by the Emperor Justinian of large parts of Italy and North Africa and several decades before the major ethnographic changes that were to accompany the disintegration of the Early Byzantine State.

It will have been sufficient for our imaginary traveller, provided he did not intend to stray far from the cities, to know only two languages, namely Greek and Latin. The boundaries of their respective diffusion were not in all places sharply drawn. It may be said, however, as a rough approximation that the linguistic frontier ran through the Balkan peninsula along an east-west line from Odessos (Varna) on the Black Sea to Dyrrachium (Durres) on the Adriatic; while south of the Mediterranean it divided Libya from Tripolitania. With the exception of the Balkan lands, where there was a fair amount of mingling, the western half of the Empire was solidly Latin and the eastern half solidly Greek in the sense that those were the languages of administration and culture. Nearly all educated persons in the East could speak Greek, just as all educated persons in the West spoke Latin, but a great proportion of ordinary people spoke neither.

Our traveller would have had considerable difficulty in supplying himself with an up-to-date guidebook. He could have laid his hands on a bare enumeration of provinces and cities called the Synecdemus of Hierocles as well as on a few itineraries of earlier date that gave distances between staging posts along the main roads. He might have drawn some useful but antiquated information from a little book known to us as the Expositio totius mundi et gentium which was composed in the middle of the fourth century; but if he wanted a systematic treatise combining geography with ethnography, he would have had to pack a copy of Strabo in his luggage. If he had been able to find the geographical treatise (now lost) by the Alexandrian merchant Cosmas Indicopleustes, he would probably have derived little practical benefit from it. Let us imagine that our traveller was content with such imperfect documentation and that, starting from Constantinople, he intended to travel clockwise round the Empire.

Constantinople, like all great capitals, was a melting-pot of heterogeneous elements: all seventy-two tongues known to man were represented in it, according to a contemporary source. Provincials of all kinds had either settled there or would drift in and out on commercial or official business. The servile class included many barbarians. Another foreign element was provided by military units which in the sixth century consisted either of barbarians (Germans, Huns, and others) or some of the sturdier provincials like Isaurians, Illyrians and Thracians. It is said that seventy thousand soldiers were billeted on the householders of Constantinople in Justinian's reign. Syrian, Mesopotamian and Egyptian monks, who spoke little or no Greek, thronged to the capital to enjoy the protection of the Empress Theodora and impress the natives with their bizarre feats of asceticism. The ubiquitous Jew earned his living as a craftsman or a merchant. Constantinople had been founded as a centre of latinity in the east and still numbered among its residents many Illyrians, Italians and Africans whose native tongue was Latin as was that of the EmperorJustinian himself. Furthermore, several works of Latin literature were produced at Constantinople, like Priscian's famous Grammar, the Chronicle of Marcellinus and the panegyric of Justin 1l by the African Corippus. Necessary as Latin still was for the legal profession and certain branches of the administration, the balance was inexorably tilting in favour of Greek. By the end of the sixth century, as Pope Gregory the Great avers, it was no easy matter to find a competent translator from Latin into Greek in the imperial capital.

Facing Constantinople lies the huge land mass of Asia Minor which has been compared to a jetty attached to Asia and pointing towards Europe. Its most developed parts have always been the coastal edges, especially the gently shelving west face, favoured by a temperate climate and studded with famous cities. The Black Sea coastal strip is much narrower and discontinuous, while the southern shore has, with the exception of the Pamphylian plain, no low-lying edge at all. The coastal areas, save for the mountainous part of Cilicia (Isauria), where the Taurus range advances to the very edge of the sea, had been hellenized for a good thousand years and more before Justinian's reign. Along the Black Sea the limit of Greek speech corresponded to the present frontier between Turkey and the Soviet Union. To the east of Trebizond and Rizaion (Rize) dwelt various Caucasian peoples, such as the Iberians (Georgians) as well as the Laz and the Abasgians (Abkhazians), the latter two barely touched by Christian missions. The Empire also possessed a Hellenized foothold on the southern shore of the Crimea, while the high tableland of the Crimean peninsula was inhabited by Goths.

Quite different from the coastal areas of Asia Minor is the high inland plateau, where the climate is rough and much of the land unfit for agriculture. In antiquity as in the Middle Ages the plateau was sparsely populated and urban life was relatively undeveloped there. The more important cities were situated along the major highways, such as the so-called Royal Road that ran from Smyrna and Sardis, by way of Ancyra and Caesarea, to Melitene; the road connecting Constantinople to Ancyra by way of Dorylaeum; and the southern road that extended from Ephesus to Laodicea, Antioch in Pisidia, Iconium, Tyana and, through the Cilician Gates, to Tarsus and Antioch in Syria. The ethnic composition of the plateau had not undergone any notable change for some seven hundred years beforeJustinian's reign. It was a bewildering mosaic of native peoples as well as immigrant enclaves of long standing, such as the Celts of Galatia, theJews who had been planted in Phrygia and elsewhere during the Hellenistic period and Persian groups of even more ancient origin. It appears that many of the indigenous languages were still spoken in the Early Byzantine period: Phrygian was probably still extant, since it appears in inscriptions as late as the third century AD, Celtic in Galatia, Cappadocian farther east. The unruly Isaurians, who had to be pacified by force of arms in about 500 AD and many of whom drifted all over the Empire as professional soldiers and itinerant masons, were a distinct people speaking their own dialect, often to the exclusion of Greek. Next to them, however, in the Cilician plain, Greek had solidly taken root, except, perhaps, among the tribes of the interior.

Lying to the east of Cappadocia and straddling a series of high mountain chains were a number of Armenian provinces that had been annexed to the Empire as late as 387 AD when the Armenian kingdom was partitioned between Persia and Rome. These were strategically very important, but practically untouched by Graeco-Roman civilization, and they continued to be ruled by native satraps until Justinian imposed on them a new form of military administration. In the fifth century the Armenians acquired their own alphabet and began building up a literature of translations from the Greek and the Syriac which strengthened their feelings of national identity. Indeed, the Armenians, who were to play a crucial role in later Byzantine history, proved very resistant to assimilation as did the other Caucasian peoples.

The boundary between Armenia and Mesopotamia corresponded approximately to the river Tigris. Three centuries of Parthian occupation (from the middle of the second century BC until the Roman conquest in about 165 AD) had obliterated in Mesopotamia practically all traces of the Hellenization which the Macedonian kings had tried so hard to impose. In the period that concerns us Mesopotamia spoke and wrote Syriac. The literary form of Syriac represented the dialect of Edessa (Urfa), and it was in that 'blessed city' as well as at Amida (Diyarbakir), Nisibis (Nusaybin`, and in the Tur 'Abdin that a vigorous monastic movement of Monophysite persuasion fuelled the cultivation of that language. Mesopotamia was a frontier district: the boundary between Rome and Persia lay a short distance south-east of the garrison town of Dara, while Nisibis had been ingloriously ceded to the Persians by the EmperorJovian in 363. The cultural apartness of Mesopotamia was certainly no help to the imperial government in so sensitive an area.

The dominance of Aramaic dialects, of which Syriac is a member, extended throughout Syria and Palestine to the confines of Egypt. Here we witness a phenomenon of considerable interest. When the Hellenistic kingdoms were established following the death of Alexander the Great, Syria was divided between the Ptolemies and the Seleucids. The Ptolemies, who obtained the southern half of the country, did rather little to plant Greek colonies there. The Seleucids, on the other hand, for whom northern Syria was of crucial importance, carried out intensive colonization. They established a number of new cities, such as Antioch on the Orontes, Apamea, Seleucia and Laodicea, and injected a Greek element into existing cities, such as Aleppo. From that time onward all of Syria remained continguously under a Greek-speaking administration. Yet, some nine centuries later, we find Greek speech confined not only to cities, but largely to those very cities that had been founded by the Hellenistic kings. The countryside generally and the towns of non-Greek origin, like Emesa (Homs), clung to their native Aramaic.

It is unlikely that the use of Greek should have been more widespread in Palestine than it was in northern Syria, except for an artificial phenomenon, namely the development of the 'holy places'. Starting in the reign of Constantine the Great, practically every site of biblical fame became, as we would say today, a tourist attraction. From every corner of the Christian world people poured into Palestine: some as transient pilgrims, others on a longer-term basis. Monasteries of every nationality sprang up like mushrooms in the desert next to the Dead Sea. Palestine was thus a babel of tongues, but the native population - and we must remember that it included two distinct ethnic groups, namely theJews and the Samaritans - spoke Aramaic as it had always done. The pilgrim Egeria, who witnessed the Easter services at Jerusalem about the year 400, has this to say:

"Seeing that in that country part of the people know both Greek and Syriac, another part only Greek and yet another part only Syriac, given also that the bishop, although he knows Syriac, always speaks in Greek and neer in Syriac, there is always by his side a priest who, while the bishop is speaking in Greek, translates his comments into Syriac so that everyone may understand them. Similarly for the lections that are read in church: since these must be read in Greek, there is always somebody there to translate them into Syriac for the benefit of the people, that they may receive instruction. As for the Latins who are there, i.e. those who know neither Syriac nor Greek, to them also is an interpretation given lest they be displeased; for there are some brethren and sisters, proficient in both Greek and Latin, who give explanations in Latin."

Another element of the population of-both Syria and Palestine consisted of Arabs who had spread as far north as Mesopotamia. Some of them, like the Nabataeans of Petra and the Palmyrenes, had become sedentary and lost their native language. Others roamed the deserts either as brigands or as vassals of the Empire whose duty it was to protect the settled areas and oversee the transhumance of the nomads. We should not, in any case, imagine that the Arab conquest of the seventh century introduced a foreign element into those provinces: the Arabs had been there all along, their numbers were increasing and, in Justinian's reign, they assumed more and more the role of keepers of the emperor's peace. When, for example, the Samaritans staged a bloody revolt in 529, it was an Arab chieftain, Abukarib, who put them down.

Closely linked with Syria by virtue of its situation was the island of Cyprus. Here Greek had been spoken since prehistoric times, but there was also a sizeable colony of Syrians as may be deduced from the prevalence of the Monophysite heresy . St Epiphanius, the most famous bishop of Salamis (d. 403), was a Palestinian and is said to have known five languages - Greek, Syriac, Hebrew, Egyptian and Latin. An exaggeration perhaps, but even so an indication of the multilingualism that characterized, as it still does, the more enterprising among the Levantines.

Separated from Palestine by an area of desert lay the rich and ancient land of Egypt. Here, too, the distribution of Greek was a direct legacy of the Hellenistic age. The capital, Alexandria, was a predominantly Greek city, but it was officially described as being ad Aegyptum, not in Aegypto, an intrusion into an alien country; and the farther one travelled from Alexandria, the less Greek was spoken. Apart from the capital, only two cities had been founded by the Greeks, Naukratis in the Delta and Ptolemais in the Thebaid; nor did Hellenization make much progress under Roman administration. Setting aside theJewish colony, which in the first century AD is said to have numbered about one million, the bulk of the population, even though they were administered in Greek, continued to speak Egyptian (Coptic), and there are signs that in the Early Byzantine period Coptic was gaining ground so that, by the sixth century, even some official acts were published in the native tongue. Above all, Coptic was the language of Egyptian Christianity, while Greek was identified with the alien hierarchy that was imposed by the imperial government.

The settled part of Egypt, which was practically limited to the Nile valley and the Delta, was threatened on all sides by barbarian tribes. From the east came raiding Saracens; in the south the black Nobadae and Blemmyes were particularly troublesome, while the west was open to Berber incursions, as was also Libya, a province that was administratively joined to Egypt. St Daniel, who was a monk at Scetis, no great distance from Alexandria, was three times kidnapped by barbarians and managed to escape only by killing his captor - a sin for which he did penance for the rest of his life. When, in the second half of the sixth century, the itinerant monk John Moschus visited the Egyptian monasteries, he picked up many tales of depredations both by barbarians and by native brigands. Some monasteries had become practically deserted.

With Libya we come to the limit of the Greek-speaking provinces. Farther west lay Tripolitania, a narrow coastal strip, then the important regions of Byzacena, Proconsularis and Numidia, and finally the two Mauretanias extending as far as the straits of Gibraltar. These had all been extensively romanized, and the richer areas, corresponding to modern Tunisia, had counted in better days among the most developed and prosperous parts of the Empire. How far the native population had been assimilated is a matter of uncertainty; nor it is entirely clear whether the vernacular language of the cities, which St Augustine calls Punic, was a legacy from ancient Phoenician (as appears more probable) or whether it was Berber. Our traveller in 560 would have found in any case a situation somewhat different from that which the Bishop of Hippo had known a century and a half earlier: for Africa had barely been recovered from the Vandals (in 533) who had held it for a century as an independent power. The Vandals had not been sufficiently numerous to have made a significant impact on the ethnography of the population, but their intrusion had led to the upsurge of the various Berber tribes who now seriously threatened the settled areas.

We need not concern ourselves with Spain, although part of its southern coast was recovered by Justinian from the Visigoths and remained in Byzantine hands for about seventy years. And so we may lead our traveller to Italy, whereJustinian's rule had just been established on a somewhat shaky basis after a great deal of bloodshed. The whole country was then in a dreadful state. Continuous warfare between Byzantium and the Ostrogoths, lasting from 535 until 562, resulted in the destruction of Milan with a reputed loss of three hundred thousand males, the virtual depopulation of Rome which sufered three sieges, and widespread starvation in the countryside. 'Italy has become everywhere even more destitute of men than Libya,' wrote Procopius, perhaps without great exaggeration. As to the composition of the population, there can be little doubt that the ltalitai, as Procopius called them, were basically Latin; even in the imperial capital of Ravenna, which had close ties with the East and numerous oriental settlers, Latin was the normal medium of communication. Some tiny pockets of Greek may have survived in the southern part of the peninsula and Greek certainly continued to be spoken on the east coast of Sicily. There were other minority groups, such as the Jews and the recently arrived Ostrogoths, but the latter could hardly have numbered more than a hundred thousand. Many more waves of invaders and settlers were to come, without, however, altering the fundamentally Latin character of the population.

Crossing the Adriatic, our traveller may have disembarked at Dyrrachium and followed the Via Egnatia all the way back to Constantinople. The regions he would have to traverse were then about as desolate as Italy. To quote Procopius once again,

"Illyricum and all of Thrace, i.e. the whole country from the Ionian Gulf [the Adriatic to the outskirts of Byzantium, including Greece and the Chersonese, was overrun almost every year by Huns, Slavs and Antae, from the time when Justinian became Roman emperor, and they wrought untold damage among the inhabitants of those parts. For I believe that in each invasion more than two hundred thousand Romans were killed or captured, so that a veritable 'Scythian wilderness' came to exist everywhere in this land."

Procopius omits to mention here that some of the most destructive invasions of the Balkan peninsulahad occurred beforeJustinian's time, in particular by the Goths in 378, by the Huns in 441-7, by the Ostrogoths in 479-82, by the Bulgars starting in 493. There can be little doubt concerning the immense amount of havoc caused by these and later incursions, but their effect on the ethnography of the regions in question is difficult to assess. The native populations were the Illyrians to the west, the Thracians and Daco-Mysians to the east and, of course, the Greeks to the south, but it would take a brave historian to state who was living where and in what numbers in the middle of the sixth century. The Slavs had already begun to settle, especially in the area between Nis and Sofia, as proved by the place names listed by Procopius, and we may imagine that the prolonged presence of Gothic and other barbarian troops had left some trace. As to languages, we have already commented on the boundary between Latin and Greek. Of Illyrian (whose relation to modern Albanian is disputed) very little is known, but Thracian, in particular Bessic, was still very much alive in the sixth century.

Such, in brief outline, were the peoples and languages of Justinian's Empire; and if I have laid any stress on the native elements, it was in order to correct the bias of our literary and narrative sources. To take but one instance, the fourth-century rhetorician Libanius, who was born at Antioch and lived most of-his life in that city, whose writings fill eleven printed volumes and are a mine of useful information, mentions only once the existence of the Syriac language. Yet it is an indisputable fact that Greek-speaking Antioch was an island in a sea of Syriac. Cultivated authors simply took no notice of such 'uncivilized' phenomena. Nor are inscriptions much more illuminating. Whoever set up an inscription, be it even on a tombstone, naturally used the 'prestige' language of the area. Besides, many of the vernacular dialects were not written. It is largely in the milieu of monks that we are occasionally brought face to face with ordinary illiterate folk and gain some inkling of what they spoke. Predictably, it was their native patois. Hence the custom of setting up 'national' monasteries. Others, however, were multinational: that of the Sleepless Ones (Akoimetoi) was divided by language into four groups - Latin, Greek, Syriac and Coptic. The monastery founded by St Theodosius the Coenobiarch in Palestine catered for Greek, Bessic and Armenian. On Mount Sinai in the sixth century one could hear Latin, Greek, Syriac, Coptic and Bessic. In 518 the abbot of a monastery at Constantinople could not sign his name to a petition because he did not know Greek. Similar examples could easily be multiplied.

Our survey would have been much more instructive had we been able to express in figures the relative importance of the various ethnic groups. Unfortunately, we have no reliable figures at our disposal, as has already been indicated in the Introduction. One eminent scholar has nevertheless ventured the view that Justinian's Empire, including the reconquered western provinces, had no more than 30 million inhabitants. Not taking into account the losses caused by the great plague of 542, this appears to be too low an estimate: we may be nearer the truth in postulating 30 million for the eastern half of the Empire. In very approximate terms, the distribution would have been the following: 8 million in Egypt, 9 million in Syria, Palestine and Mesopotamia combined, 10 million in Asia Minor, and 3 to 4 million in the Balkans. If these figures are anywhere near the truth, it would follow that the native Greek speakers represented less than a third of the total population, say 8 million, making allowance for the unassimilated peoples of Asia Minor and for the Latin and Thracian speakers of the Balkans. The Greek, Coptic and Aramaic elements would thus have been on a footing of near parity. Compared to the spread of Latin in Gaul and Spain, it must be admitted that the Greek language had made very limited progress between the third century BC and the sixth century AD. This was no doubt due to the fact that Hellenization was largely centred on cities. About a century after the Arab conquest Greek had become practically extinct in both Syria and Egypt, which can only mean that it had not grown deep roots.

One further observation may be made on the basis of our survey, namely that in spite of mounting insecurity in nearly all parts of the Empire, most of Justinian's subjects still lived in their traditional homelands. The diaspora of the Greeks, of the Jews and, to a lesser extent, of the Syrians had occurred several centuries earlier. From the viewpoint of ethnography, as in so many other respects, Justinian's age represents, therefore, the tail end of Antiquity.

It would be wearisome to describe here all the ethnographic changes that the Empire witnessed after the sixth century, but we must say a few words about the greatest mutation of all, which started happening a few decades after Justinian's death. Its first sign was the massive installation of the Slavs in the Balkan peninsula. The Slavs came in several waves and, unlike earlier invaders, they came to stay. In an oft-quoted passage John of Amida (also known asJohn of Ephesus) records that in 581

"an accursed people, called Slaonians, overran the whole of Greece, and the country of the Thessalonians, and all TXrace, and captured the cities, and took numerous forts, and devastated and burnt, and reduced the people to slavery, and made themselves masters of the whole country, and settled in it by main force, and dwelt in it as though it had been their own.... And even to this day [584 AD], they stili encamp and dwell there, and live in peace in the Roman territories, free from anxiety and fear, and lead captive and slay and burn."

Another source, the so-called Chronicle of Monembasia, states that in the year 587-8 the Turkic Avars (with whom the Slavs were usually allied)

"captured all of Thessaly and all of Greece, Old Epirus, Attica and Euboea. Indeed, they attacked the Peloponnese and took it by war; and after expelling and destroying the native Hellenic peoples, they dwelt there. Those who were able to escape their murderous hands were scattered here and there. Thus, the citizens of Patras moved to the district of Reggio in Calabria, the Argives to the island called Orobe, the Corinthians to the island of Aegina.... Only the eastern part of the Peloponnese, from Corinth to Cape Maleas, was untouched by the Slavonians because of the rough and inaccessible nature of the country."

There is some doubt concerning the exact date of these events, but it is undeniable that at the end of the sixth century and the beginning of the seventh, when the Danubian frontier completely collapsed, practicallv the entire Balkan peninsula passed out of imperial control. Only a few coastal outposts, such as Mesembria on the Black Sea, Thessalonica, Athens and Corinth, held out. Elsewhere the old population sought refuge on off-shore islands, as it did on Monembasia, or emigrated to Italy. The domain of barbarism extended as far as the outer defences of Constantinople - the so-called Anastasian Long Walls which described a wide arc from the Black Sea to Selymbria (Siliv on the Sea of Marmora - but even these had soon to be abandoned.

The last important Slavonic settlement was that of the Serbs and Croats who in the reign of Heraclius occupied the lands where they still dwell. Then, in 680, came the Turkic Bulgars and conquered the country that bears their name, where they were eventually assimilated by the sitting Slavonic population. The barbarization of the Balkans began to be reversed only towards the ed of the eighth century, but by that time its effects had become permanent.

Simultaneously with the loss of the Balkans the Empire suffered a more serious amputation by being deprived of its eastern and southern provinces. This happened in two stages. First, between the years 609 and 619, the Persians conquered all of Syria, Palestine and Egypt. They were then defeated by the emperor Heraclius and withdrew to their own country; but a few years later the same provinces were overrun by the Arabs and, this time, lost for good. The whole of the north African coast also succumbed to the invader. The Mediterranean empire of Rome simply ceased to exist, while the Byzantine State found itself limited to Asia Minor, the Aegean islands, a bit of the Crimea and Sicily.

The Persians also initiated another development that was to have important demographic consequences by striking at Constantinople through Asia Minor. In so doing they caused immense havoc. When the Arabs had succeeded to the Persians and made themselves masters of all the territories up to the Taurus mountains, they, too, struck into Asia Minor- not once or twice, but practically every year- and this went on for nearly two centuries. Many of the raids did not penetratc far from the frontier, but several of them extended as far as the Black Sea and the Aegean, and a few reached Constantinople itself. As it turned out, the Arabs never managed to gain a foothold on the Anatolian plateau. What happened instead was that every time they marched in the local population would take refuge in the inaccessible forts with which Asia Minor is so liberally proviced. The Arabs would pass between the forts, taking prisoners and booty, while the Byzantines would burn the crops to deprive the enemy of supplies and keep him on the move. The consequences of this prolonged process are easy to imagine: much of Asia Minor was devastated and depopulated almost beyond repair.

In this way an enormous demographic gap was created. The Empire urgently needed farmers as it also needed soldiers. To achieve this end it had to resort to massive transfers of population. The Emperor Justinian Il, in particular, applied this policy on a wide scale. He moved a good part of the population of Cyprus to the region of Cyzicus on the southern shore of the Sea of Marmora. It was, apparently, a failure: many of the immigrants perished en route, and those who reached their destination later asked to be repatriated. Justinian II also moved 'a great multitude' of Slavs to Bithynia. Once again, he had little luck: the thirty thousand soldiers he raised from among this group to fight against the Arabs defected to the enemy, whereupon the emperor inflicted cruel reprisals on their families. In the 760s, however, we are told that 208,000 Slavs came to live in Bithynia of their own accord. In the eighth century we repeatedly hear of the organized settlement of Syrians in Thrace.

Among the new immigrants the most prominent, however, were the Armenians, many of whom arrived without being forced to do so. The Armenians were excellent soldiers, and the Empire, deprived of its Illyrian recruiting ground, needed them badly. In fact, the immigration of Armenians had started in the sixth century, and from the reign of Maurice onwards they formed the backbone of the Byzantine army. The trickle of Armenians into the Empire was spread over many centuries. Many settled in Cappadocia and other parts of eastern Asia Minor close to their original homeland, others in Thrace, others in the region of Pergamon. It is impossible to give even a rough approximation of their numbers. Unlike the Slavs, however, the Armenians quickly rose to prominent positions, even to the imperial throne, and dominated the military establishment throughout the Middle Byzantine period.

Thus, if we place ourselves at about the time when the Empire started on the slow course of its recovery, say towards the end of the eighth century, we find a population that had been so thoroughly churned up that it is difficult to tell what ethnic groups were living where and in what numbers. It is often stated that by shedding, however painfully, its principal non-Greek-speaking elements, such as the Syrians, the Egyptians and the Illyrians, the Empire had become more homogeneous. It is also asserted that the non-Greeks were gradually assimilated or Hellenized through the agency of the Church and the army, and that this happened in particular to the indigenous populations of Asia Minor as well as to the Slavs in the Peloponnese and elsewhere in Greece. The critical reader may be advised to treat such generalizations with a measure of caution. It is true, of course, that following the eclipse of Latin, Greek became the only official language of the Empire, so that a knowledge of it was mandatory for pursuing a career or transacting business. Neither Armenian nor Slavonic ever supplanted it as a general medium of communication. It is also true that in the long run Slavonic died out in Greece and in Bithynia, and if any Armenian has been spoken in Thrace within living memory, it was not on the part of descendants of the colonists planted there in the eighth century. But then it is also known that Greek survived in Asia Minor on a continuous basis only in Pontus and a small part of Cappadocia, whereas it had become practically extinct in the western part of the subcontinent until its reintroduction there by immigrants in the eighteenth and nineteenth centuries. We would not argue from the last observation that western Asia Minor was not predominantly Greek-speaking in the Middle Ages. However illuminating it may be in some respects, the long view does not help the historian of Byzantium to solve the specific problems that confront him. Was Hellenization, for example, a conscious aim of the imperial government, and if so, how was it implemented and with what success? And if it succeeded in the Middle Ages, why had it not done so in Antiquity under conditions of a more settled life and a higher civilization?

When we look at our scanty sources; we realize that the formulation of the above questions does not correspond to the Byzantine way of thinking. First of all, the very designation 'Greek', which we use so freely today to describe those Byzantines who did not belong to any alien group, is entirely absent from tlie literature of the period. An inhabitant of Greece south of Thessaly would have referred to himself as a Helladikos (a name already current in the sixth century AD), but he could have been a Slav as well as a 'Greek'. The same holds true of other regions whose dwellers called themselves by the names of their respective provinces, for example Paphlagonians or Thraksians (after the Thraksian 'theme' in western Asia Minor). Since, therefore, there was no notion of'Greekness', it is hard to see how there could have been one of 'hellenization'. The only passage, to my knowledge, that may imply something of the kind says that the Emperor Basil I converted the Slavonic tribes from their old religion and, 'having grecized them (graikosas), subjected them to governors according to Roman custom, honoured them with baptism, and delivered them from the oppression of their own rulers'. It has long been, however, a matter of dispute what the term 'grecized' may mean in the present context. What we do hear about, again and again, is the conversion of various peoples to Orthodox Christianity, be they pagan Slavs or Muslim Cretans, and the setting up of an ecclesiastical organization. Here is how the Chronicle of Monembasia describes the activity of the Emperor Nicephorus I in the Peloponnese: 'He built de novo the town of Lacedaemon and settled in it a mixed population, namely Kafirs, Thraksians, Armenians and others, gathered from different places and towns, and made it into a bishopric.' Surely, neither the Kafirs (possibly a generic term for converts from Islam) nor the Armenians would have contributed to the hellenization of Laconia. The emperor's purpose was simply to implant a Christian population and set up a bishopric.

There can be little doubt that the evangelization of non-Christian peoples settled in the Empire was carried out in Greek. This may cause some surprise in the case of the Slavs since the Slavonic alphabet was itself devised by a Byzantine, St Cyril, presumably in the 860s. Its invention, however, and the consequent translation of the essential Christian texts were intended for a far-away Slavonic country, Moravia; and it was entirely a matter of chance that the Cyrillo- Methodian mission, after its initial failure, should have found a fertile soil in a country for which it was not intended, namely the Bulgarian kingdom. As far as our knowledge goes, no attempt was ever made to evangelize the Slavs in Greece in their own language, just as the liturgical use of Greek was imposed on conquered Bulgaria after 1018. Clearly, this must have contributed to the spread of Greek. But was it due to deliberate policy? Is it not more likely that the absence of a linguistically qualified clergy, the relative inaccessibility of the Slavonic Scriptures, and the mixed nature of the population should have combined to make the use of Greek the easier option?

However efficacious the liturgical imposition of Greek may have proved, it has to be admitted that the assimilation of barbarian enclaves was a very slow process. In the Peloponnese the presence of pagan Slavs a short distance south of Sparta is attested in the latter part of the tenth century, that is nearly two hundred years after the first attempts to bring about their conversion. Equally telling is the case of the Slavs in Bithynia. We have seen that these were transplanted in v ery considerable numbers at the end of the seventh century and towards the middle of the eighth. Some two hundred years later, the Byzantine armament assembled in an effort to conquer Crete in 949 included a contingent of 'Slavonians who are established in Opsikion' (this being the administrative name of a part of Bithynia) placed under their own commanders. Clearly, these Slavonians still formed a distinct group. In the next century Anna Comnena refers to a village in Bithvnia 'locally called Sagoudaous', presumably after the tribe of the Sagoudatai, attested in Macedonia in the seventh century. A little later the Slavonic element in Bithynia was augmented by the EmperorJohn II Comnenus who settled near Nicomedia a throng of Serbian captives. Serbian villages are still mentioned in those parts in the thirteenth century. In other words, it is quite possible that the Slavs of Bithynia, or at any rate part of them, were assimilated by the (Ottoman Turks without having ever become 'Greek'.

The obvious conclusion to be drawn from these and many other cases is that the Middle Byzantine Empire was by no means a solidly Greek state. In addition to the Armenians and the Slavs, there were many other foreign elements, such as the Georgians and the Balkan Vlachs. A massive influx of Syrians and other Christian orientals followed the eastward expansion of the Empire at the end of the tenth century; and when, in 1018, the imperial frontier was once more extended to the Danube, it comprised vast areas where Greek had never been spoken or had been extinguished a long time previously. Whether Greek speakers formed at the time the majority or a minority of the inhabitants of the Empire is a guess I should not like to hazard.

It is not altogether easy to define the feelings of solidarity, if any, that bound together the multinational inhabitants of the Empire. In the sixth century the slogan Gloria Romanorum still appeared from time to time on the imperial coinage, but it is not likely that there was much devotion in the eastern provinces to the idea of Romanias. Besides, loyalty to Rome and admiration for her ancient greatness had been a regular theme of pagan polemic, whereas the Church maintained the position that Christians were, above all, citizens of the Heavenly Jerusalem and in so doing probably weakened the cohesion of the Empire. That is not to say that instances of loyalty to the State are absent from Byzantine history: quite the reverse is true. It is enough to recall the despair of the population of Nisibis when their city had been ceded to the Persians in 6, the demonstrations of Dro-Roman sentiments at Edessa in 449 in the context of sectarian strife, and a multitude of similar cases. But then we must remember that at the time the only alternative to living under Roman rule was living under Persian rule (which was usually worse). People crushed by the burden of taxation were often tempted to desert to the enemy, even to join some barbarian tribe that levied no taxes, but that was not an option for those who enjoyed a reasonable standard of living. A feeling of Romanitas was hardly the determining factor.

As far as we can judge, the main links of solidarity were two: regional and religious. People identified themselves with their village, their city or their province much more than they did with the Empire. When a person was away from home he was a stranger and was often treated with suspicion. A monk from western Asia Minor who joined a monastery in Pontus was 'disparaged and mistreated by everyone as a stranger'. The corollary to regional solidarity was regional hostility. We encounter many derogatory statements concerning 'the cunning Syrian' who spoke with a thick accent, the uncouth Paphlagonian, the mendacious Cretan. Alexandrians excited ridicule at Constantinople. Armenians were nearly always described in terms of abuse. Even demons had strong feelings of local affiliation and did not want to consort with their fellows from the next province.

Religious identity was often more strongly felt than regional identity. Had the Church been less intolerant, it may well be that different religious groups could have lived peaceably side by side, but there was usually some zealous bishop or monk who incited a pogrom, and then the fight was on. It is not surprising that Jews and the few remaining pagans should have proved the most consistently disloyal elements in the Empire. Within the Church, however, religion and regionalism overlapped to a considerable extent. And herein, perhaps, lies the key to the 'heretical' groupings. For what seems to have motivated the Syrian or the Egyptian Monophysite was not so much his belief in some abstruse point of doctrine as his loyalty to his own Church, his own bishop and the holy men of his neighbourhood. Whenever a Christian splinter group had a solidly established territorial base, all attempts to impose on it a uniform, imperial orthodoxy ended in failure.

If in the Early Byzantine period the idea of Romanitas held little potency, the same was even truer of the Middle period when the old imerial caDital had receded into some 'Scthian wilderness' and the Latin tongue had been forgotten. Even in contexts of international confrontation the emotive concept became that of Christian rather than that of Roman identity. When, in 922, Romanus I Lecapenus urged his army officers to put up a spirited defence against Symeon of Bulgaria, they vowed to die on behalf of the Christians, and this although the Bulgarians were by this time, at any rate nominally, Christian themselves. Significantly, however, no new term emerged to describe the identity of the Empire as a whole. Nor was it much needed on the level of everyday life. When, in the early ninth century, St Gregory the Decapolite, a native of southern Asia Minor, landed at the port of Ainos in Thrace, he was promptly arrested by the imperial police and subjected to a bastinado. We are not told why; perhaps he looked like an Arab. He was then asked: 'Who are you, and what is your religion?' His answer was: 'I am a Christian, my parents are such and such, and I am of the Orthodox persuasion.' Religion and local origin constituted his passport. It did not occur to him to describe himself as a Roman.