"In Khazaria, sheep, honey, and Jews exist in large quantities." Muqaddasi, Descriptio Imperii Moslemici (tenth century).
ABOUT the time when Charlemagne was crowned Emperor of the West, the eastern confines of Europe between the Caucasus and the Volga were ruled by a Jewish state, known as the Khazar Empire. At the peak of its power, from the seventh to the tenth centuries AD, it played a significant part in shaping the destinies of mediaeval, and consequently of modern, Europe. The Byzantine Emperor and historian, Constantine Porphyrogenitus (913-959), must have been well aware of this when he recorded in his treatise on court protocol .1 that letters addressed to the Pope in Rome, and similarly those to the Emperor of the West, had a gold seal worth two solidi attached to them, whereas messages to the King of the Khazars displayed a seal worth three solidi. This was not flattery, but Realpolitik. "In the period with which we are concerned," wrote Bury, "it is probable that the Khan of the Khazars was of little less importance in view of the imperial foreign policy than Charles the Great and his successors." .2 The country of the Khazars, a people of Turkish stock, occupied a strategic key position at the vital gateway between the Black Sea and the Caspian, where the great eastern powers of the period confronted each other. It acted as a buffer protecting Byzantium against invasions by the lusty barbarian tribesmen of the northern steppes - Bulgars, Magyars, Pechenegs, etc. - and, later, the Vikings and the Russians. But equally, or even more important both from the point of view of Byzantine diplomacy and of European history, is the fact that the Khazar armies effectively blocked the Arab avalanche in its most devastating early stages, and thus prevented the Muslim conquest of Eastern Europe. Professor Dunlop of Columbia University, a leading authority on the history of the Khazars, has given a concise summary of this decisive yet virtually unknown episode:
It is perhaps not surprising, given these circumstances, that in 732 - after a resounding Khazar victory over the Arabs - the future Emperor Constantine V married a Khazar princess. In due time their son became the Emperor Leo IV, known as Leo the Khazar. .Ironically, the last battle in the war, AD 737, ended in a Khazar defeat. But by that time the impetus of the Muslim Holy War was spent, the Caliphate was rocked by internal dissensions, and the Arab invaders retraced their steps across the Caucasus without having gained a permanent foothold in the north, whereas the Khazars became more powerful than they had previously been. .A few years later, probably AD 740, the King, his court and the military ruling class embraced the Jewish faith, and Judaism became the state religion of the Khazars. No doubt their contemporaries were as astonished by this decision as modern scholars were when they came across the evidence in the Arab, Byzantine, Russian and Hebrew sources. One of the most recent comments is to be found in a work by the Hungarian Marxist historian, Dr Antal Bartha. His book on The Magyar Society in the Eighth and Ninth Centuries4 has several chapters on the Khazars, as during most of that period the Hungarians were ruled by them. Yet their conversion to Judaism is discussed in a single paragraph, with obvious embarrassment. It reads:
Which leaves us only slightly more bewildered than before. Yet whereas the sources differ in minor detail, the major facts are beyond dispute. .What is in dispute is the fate of the Jewish Khazars after the destruction of their empire, in the twelfth or thirteenth century. On this problem the sources are scant, but various late mediaeval Khazar settlements are mentioned in the Crimea, in the Ukraine, in Hungary, Poland and Lithuania. The general picture that emerges from these fragmentary pieces of information is that of a migration of Khazar tribes and communities into those regions of Eastern Europe - mainly Russia and Poland - where, at the dawn of the Modern Age, the greatest concentrations of Jews were found. This has lead several historians to conjecture that a substantial part, and perhaps the majority of eastern Jews - and hence of world Jewry - might be of Khazar, and not of Semitic Origin. .The far-reaching implications of this hypothesis may explain the great caution exercised by historians in approaching this subject - if they do not avoid it altogether. Thus in the 1973 edition of the Encyclopaedia Judaica the article "Khazars" is signed by Dunlop, but there is a separate section dealing with "Khazar Jews after the Fall of the Kingdom", signed by the editors, and written with the obvious intent to avoid upsetting believers in the dogma of the Chosen Race:
How important, in quantitative terms, is that "presence" of the Caucasian sons of Japheth in the tents of Shem? One of the most radical propounders of the hypothesis concerning the Khazar origins of Jewry is the Professor of Mediaeval Jewish History at Tel Aviv University, A. N. Poliak. His book Khazaria (in Hebrew) was published in 1944 in Tel Aviv, and a second edition in 1951.5 In his introduction he writes that the facts demand -
This was written before the full extent of the holocaust was known, but that does not alter the fact that the large majority of surviving Jews in the world is of Eastern European - and thus perhaps mainly of Khazar - origin. If so, this would mean that their ancestors came not from the Jordan but from the Volga, not from Canaan but from the Caucasus, once believed to be the cradle of the Aryan race; and that genetically they are more closely related to the Hun, Uigur and Magyar tribes than to the seed of Ab raham, Isaac and Jacob. Should this turn out to be the case, then the term "anti-Semitism" would become void of meaning, based on a misapprehension shared by both the killers and their victims. The story of the Khazar Empire, as it slowly emerges from the past, begins to look like the most cruel hoax which history has ever perpetrated.
"Attila was, after all, merely the king of a kingdom of tents. His state passed away - whereas the despised city of Constantinople remained a power. The tents vanished, the towns remained. The Hun state was a whirlwind. ...".Thus Cassel,6 a nineteenth-century orientalist, implying that the Khazars shared, for similar reasons, a similar fate. Yet the Hun presence on the European scene lasted a mere eighty years,*[From circa 372, when the Huns first started to move westward from the steppes north of the Caspian, to the death of Attila in 453.] whereas the kingdom of the Khazars held its own for the best part of four centuries. They too lived chiefly in tents, but they also had large urban settlements, and were in the process of transformation from a tribe of nomadic warriors into a nation of farmers, cattle-breeders, fishermen, vine-growers, traders and skilled craftsmen. Soviet archaeologists have unearthed evidence for a relatively advanced civilization which was altogether different from the "Hun whirlwind". They found the traces of villages extending over several miles,7 with houses connected by galleries to huge cattlesheds, sheep-pens and stables (these measured 3-31/2 x 10-14 metres and were supported by columns.8 Some remaining ox-ploughs showed remarkable craftsmanship; so did the preserved artefacts - buckles, clasps, ornamental saddle plates..Of particular interest were the foundations, sunk into the ground, of houses built in a circular shape.9 According to the Soviet archaeologists, these were found all over the territories inhabited by the Khazars, and were of an earlier date than their "normal", rectangular buildings. Obviously the round-houses symbolize the transition from portable, dome- shaped tents to permanent dwellings, from the nomadic to a settled, or rather semi-settled, existence. For the contemporary Arab sources tell us that the Khazars only stayed in their towns - including even their capital, Itil - during the winter; come spring, they packed their tents, left their houses and sallied forth with their sheep or cattle into the steppes, or camped in their cornfields or vineyards..The excavations also showed that the kingdom was, during its later period, surrounded by an elaborate chain of fortifications, dating from the eighth and ninth centuries, which protected its northern frontiers facing the open steppes. These fortresses formed a rough semi-circular arc from the Crimea (which the Khazars ruled for a time) across the lower reaches of the Donetz and the Don to the Volga; while towards the south they were protected by the Caucasus, to the west by the Black Sea, and to the east by the "Khazar Sea", the Caspian.* ["To this day, the Muslims, recalling the Arab terror of the Khazar raids, still call the Caspian, a sea as shifting as the nomads, and washing to their steppe-land parts, Bahr-ul-Khazar - "the Khazar Sea"." (W. E. 0. Allen, A History of the Georgian People, London 1952).] However, the northern chain of fortifications marked merely an inner ring, protecting the stable core of the Khazar country; the actual boundaries of their rule over the tribes of the north fluctuated according to the fortunes of war. At the peak of their power they controlled or exacted tribute from some thirty different nations and tribes inhabiting the vast territories between the Caucasus, the Aral Sea, the Ural Mountains, the town of Kiev and the Ukrainian steppes. The people under Khazar suzerainty included the Bulgars, Burtas, Ghuzz, Magyars (Hungarians), the Gothic and Greek colonies of the Crimea, and the Slavonic tribes in the north-western woodlands. Beyond these extended dominions, Khazar armies also raided Georgia and Armenia and penetrated into the Arab Caliphate as far as Mosul. In the words of the Soviet archaeologist M. I. Artamonov:10
Taking a bird's-eye view of the history of the great nomadic empires of the East, the Khazar kingdom occupies an intermediary position in time, size, and degree of civilization between the Hun and Avar Empires which preceded, and the Mongol Empire that succeeded it.
But who were these remarkable people - remarkable as much by their power and achievements as by their conversion to a religion of outcasts? The descriptions that have come down to us originate in hostile sources, and cannot be taken at face value. "As to the Khazars," an Arab chronicler 11 writes, "they are to the north of the inhabited earth towards the 7th clime, having over their heads the constellation of the Plough. Their land is cold and wet.
Accordingly their complexions are white, their eyes blue, their hair flowing and predominantly reddish, their bodies large and their natures cold. Their general aspect is wild." .After a century of warfare, the Arab writer obviously had no great sympathy for the Khazars. Nor had the Georgian or Armenian scribes, whose countries, of a much older culture, had been repeatedly devastated by Khazar horsemen. A Georgian chronicle, echoing an ancient tradition, identifies them with the hosts of Gog and Magog - "wild men with hideous faces and the manners of wild beasts, eaters of blood".12 An Armenian writer refers to "the horrible multitude of Khazars with insolent, broad, lashless faces and long falling hair, like women".13 Lastly, the Arab geographer Istakhri, one of the main Arab sources, has this to say:14 "The Khazars do not resemble the Turks. They are black-haired, and are of two kinds, one called the Kara-Khazars, [Black Khazars] who are swarthy verging on deep black as if they were a kind of Indian, and a white kind [Ak-Khazars], who are strikingly handsome.".This is more flattering, but only adds to the confusion. For it was customary among Turkish peoples to refer to the ruling classes or clans as "white", to the lower strata as "black". Thus there is no reason to believe that the "White Bulgars" were whiter than the "Black Bulgars", or that the "White Huns" (the Ephtalites) who invaded India and Persia in the fifth and sixth centuries were of fairer skin than the other Hun tribes which invaded Europe. Istakhri's black-skinned Khazars - as much else in his and his colleagues' writings - were based on hearsay and legend; and we are none the wiser regarding the Khazars' physical appearance, or their ethnic Origins..The last question can only be answered in a vague and general way. But it is equally frustrating to inquire into the origins of the Huns, Alans, Avars, Bulgars, Magyars, Bashkirs, Burtas, Sabirs, Uigurs, Saragurs, Onogurs, Utigurs, Kutrigurs, Tarniaks, Kotragars, Khabars, Zabenders, Pechenegs, Ghuzz, Kumans, Kipchaks, and dozens of other tribes or people who at one time or another in the lifetime of the Khazar kingdom passed through the turnstiles of those migratory playgrounds. Even the Huns, of whom we know much more, are of uncertain origin; their name is apparently derived from the Chinese Hiung-nu, which designates warlike nomads in general, while other nations applied the name Hun in a similarly indiscriminate way to nomadic hordes of all kinds, including the "White Huns" mentioned above, the Sabirs, Magyars and Khazars.*[It is amusing to note that while the British in World War I used the term "Hun" in the same pejorative sense, in my native Hungary schoolchildren were taught to look up to "our glorious Hun forefathers" with patriotic pride An exclusive rowing club in Budapest was called "Hunnia", and Attila is still a popular first name.] .In the first century AD, the Chinese drove these disagreeable Hun neighbours westward, and thus started one of those periodic avalanches which swept for many centuries from Asia towards the West. From the fifth century onward, many of these westward-bound tribes were called by the generic name of "Turks". The term is also supposed to be of Chinese origin (apparently derived from the name of a hill) and was subsequently used to refer to all tribes who spoke languages with certain common characteristics - the "Turkic" language group. Thus the term Turk, in the sense in which it was used by mediaeval writers - and often also by modern ethnologists - refers primarily to language and not to race. In this sense the Huns and Khazars were "Turkic" people.*[But not the Magyars, whose language belongs to the Finno-Ugrian language group.] The Khazar language was supposedly a Chuvash dialect of Turkish, which still survives in the Autonomous Chuvash Soviet Republic, between the Volga and the Sura. The Chuvash people are actually believed to be descendants of the Bulgars, who spoke a dialect similar to the Khazars. But all these connections are rather tenuous, based on the more or less speculative deductions of oriental philologists. All we can say with safety is that the Khazars were a "Turkic" tribe, who erupted from the Asian steppes, probably in the fifth century of our era..The origin of the name Khazar, and the modern derivations to which it gave rise, has also been the subject of much ingenious speculation. Most likely the word is derived from the Turkish root gaz, "to wander", and simply means "nomad". Of greater interest to the non-specialist are some alleged modern derivations from it: among them the Russian Cossack and the Hungarian Huszar - both signifying martial horsemen;*[Huszar is probably derived via the Serbo-Croat from Greek references to Khazars.] and also the German Ketzer - heretic, i.e., Jew. If these derivations are correct, they would show that the Khazars had a considerable impact on the imagination of a variety of peoples in the Middle Ages.
Some Persian and Arab chronicles provide an attractive combination of legend and gossip column. They may start with the Creation and end with stop-press titbits. Thus Yakubi, a ninth-century Arab historian, traces the origin of the Khazars back to Japheth, third son of Noah. The Japheth motive recurs frequently in the literature, while other legends connect them with Abraham or Alexander the Great. .One of the earliest factual references to the Khazars occurs in a Syriac chronicle by "Zacharia Rhetor",*[It was actually written by an anonymous compiler and named after an earlier Greek historian whose work is summarized in the compilation.] dating from the middle of the sixth century. It mentions the Khazars in a list of people who inhabit the region of the Caucasus. Other sources indicate that they were already much in evidence a century earlier, and intimately connected with the Huns. In AD 448, the Byzantine Emperor Theodosius II sent an embassy to Attila which included a famed rhetorician by name of Priscus. He kept a minute account not only of the diplomatic negotiations, but also of the court intrigues and goings-on in Attila's sumptuous banqueting hall - he was in fact the perfect gossip columnist, and is still one of the main sources of information about Hun customs and habits. But Priscus also has anecdotes to tell about a people subject to the Huns whom he calls Akatzirs - that is, very likely, the Ak-Khazars, or "White" Khazars (as distinct from the "Black" Kara-Khazars).**[The "Akatzirs" are also mentioned as a nation of warriors by Jordanes, the great Goth historian, a century later, and the so- called "Geographer of Ravenna" expressly identifies them with the Khazars. This is accepted by most modern authorities. (A notable exception was Marquart, but see Dunlop's refutation of his views, op. cit., pp. 7f.) Cassel, for instance, points out that Priscus's pronunciation and spelling follows the Armenian and Georgian: Khazir.] The Byzantine Emperor, Priscus tells us, tried to win this warrior race over to his side, but the greedy Khazar chieftain, named Karidach, considered the bribe offered to him inadequate, and sided with the Huns. Attila defeated Karidach's rival chieftains, installed him as the sole ruler of the Akatzirs, and invited him to visit his court. Karidach thanked him profusely for the invitation, and went on to say that "it would be too hard on a mortal man to look into the face of a god. For, as one cannot stare into the sun's disc, even less could one look into the face of the greatest god without suffering injury." Attila must have been pleased, for he confirmed Karidach in his rule..Priscus's chronicle confirms that the Khazars appeared on the European scene about the middle of the fifth century as a people under Hunnish sovereignty, and may be regarded, together with the Magyars and other tribes, as a later offspring of Attila's horde.
The collapse of the Hun Empire after Attila's death left a power-vacuum in Eastern Europe, through which once more, wave after wave of nomadic hordes swept from east to west, prominent among them the Uigurs and Avars. The Khazars during most of this period seemed to be happily occupied with raiding the rich trans-Caucasian regions of Georgia and Armenia, and collecting precious plunder. During the second half of the sixth century they became the dominant force among the tribes north of the Caucasus. A number of these tribes - the Sabirs, Saragurs, Samandars, Balanjars, etc. - are from this date onward no longer mentioned by name in the sources: they had been subdued or absorbed by the Khazars. The toughest resistance, apparently, was offered by the powerful Bulgars. But they too were crushingly defeated (circa 641), and as a result the nation split into two: some of them migrated westward to the Danube, into the region of modern Bulgaria, others north-eastward to the middle Volga, the latter remaining under Khazar suzerainty. We shall frequently encounter both Danube Bulgars and Volga Bulgars in the course of this narrative. .But before becoming a sovereign state, the Khazars still had to serve their apprenticeship under another short-lived power, the so-called West Turkish Empire, or Turkut kingdom. It was a confederation of tribes, held together by a ruler: the Kagan or Khagan*[Or Kaqan or Khaqan or Chagan, etc. Orientalists have strong Idiosyncrasies about spelling (see Appendix I). I shall stick to Kagan as the least offensive to Western eyes. The h in Khazar, however, is general usage.] - a title which the Khazar rulers too were subsequently to adopt. This first Turkish state - if one may call it that - lasted for a century (circa 550-650) and then fell apart, leaving hardly any trace. However, it was only after the establishment of this kingdom that the name "Turk" was used to apply to a specific nation, as distinct from other Turkic-speaking peoples like the Khazars and Bulgars.*[This, however, did not prevent the name "Turk" still being applied indiscriminately to any nomadic tribe of the steppes as a euphemism for Barbarian, or a synonym for "Hun". It led to much confusion in the interpretation of ancient sources.] .The Khazars had been under Hun tutelage, then under Turkish tutelage. After the eclipse of the Turks in the middle of the seventh century it was their turn to rule the "Kingdom of the North", as the Persians and Byzantines came to call it. According to one tradition,15 the great Persian King Khusraw (Chosroes) Anushirwan (the Blessed) had three golden guest-thrones in his palace, reserved for the Emperors of Byzantium, China and of the Khazars. No state visits from these potentates materialized, and the golden thrones - if they existed - must have served a purely symbolic purpose. But whether fact or legend, the story fits in well with Emperor Constantine's official account of the triple gold seal assigned by the Imperial Chancery to the ruler of the Khazars.
Thus during the first few decades of the seventh century, just before the Muslim hurricane was unleashed from Arabia, the Middle East was dominated by a triangle of powers: Byzantium, Persia, and the West Turkish Empire. The first two of these had been waging intermittent war against each other for a century, and both seemed on the verge of collapse; in the sequel, Byzantium recovered, but the Persian kingdom was soon to meet its doom, and the Khazars were actually in on the kill. .They were still nominally under the suzerainty of the West Turkish kingdom, within which they represented the strongest effective force, and to which they were soon to succeed; accordingly, in 627, the Roman Emperor Heraclius concluded a military alliance with the Khazars - the first of several to follow - in preparing his decisive campaign against Persia. There are several versions of the role played by the Khazars in that campaign which seems to have been somewhat inglorious - but the principal facts are well established. The Khazars provided Heraclius with 40000 horsemen under a chieftain named Ziebel, who participated in the advance into Persia, but then - presumably fed up with the cautious strategy of the Greeks - turned back to lay siege on Tiflis; this was unsuccessful, but the next year they again joined forces with Heraclius, took the Georgian capital, and returned with rich plunder. Gibbon has given a colourful description (based on Theophanes) of the first meeting between the Roman Emperor and the Khazar chieftain.16
Eudocia (or Epiphania) was the only daughter of Heraclius by his first wife. The promise to give her in marriage to the "Turk" indicates once more the high value set by the Byzantine Court on the Khazar alliance. However, the marriage came to naught because Ziebel died while Eudocia and her suite were on their way to him. There is also an ambivalent reference in Theophanes to the effect that Ziebel "presented his son, a beardless boy" to the Emperor - as a quid pro quo? .There is another picturesque passage in an Armenian chronicle, quoting the text of what might be called an Order of Mobilization issued by the Khazar ruler for the second campaign against Persia: it was addressed to "all tribes and peoples [under Khazar authority], inhabitants of the mountains and the plains, living under roofs or the open sky, having their heads shaved or wearing their hair long".17 .This gives us a first intimation of the heterogeneous ethnic mosaic that was to compose the Khazar Empire. The "real Khazars" who ruled it were probably always a minority - as the Austrians were in the Austro-Hungarian monarchy.
The Persian state never recovered from the crushing defeat inflicted on it by Emperor Heraclius in 627. There was a revolution; the King was slain by his own son who, in his turn, died a few months later; a child was elevated to the throne, and after ten years of anarchy and chaos the first Arab armies to erupt on the scene delivered the coup de grace to the Sassanide Empire. At about the same time, the West Turkish confederation dissolved into its tribal components. A new triangle of powers replaced the previous one: the Islamic Caliphate - Christian Byzantium and the newly emerged Khazar Kingdom of the North. It fell to the latter to bear the brunt of the Arab attack in its initial stages, and to protect the plains of Eastern Europe from the invaders. .In the first twenty years of the Hegira - Mohammed's flight to Medina in 622, with which the Arab calendar starts - the Muslims had conquered Persia, Syria, Mesopotamia, Egypt, and surrounded the Byzantine heartland (the present-day Turkey) in a deadly semi-circle, which extended from the Mediterranean to the Caucasus and the southern shores of the Caspian. The Caucasus was a formidable natural obstacle, but no more forbidding than the Pyrenees; and it could be negotiated by the pass of Dariel*[Now called the Kasbek pass.] or bypassed through the defile of Darband, along the Caspian shore. .This fortified defile, called by the Arabs Bab al Abwab, the Gate of Gates, was a kind of historic turnstile through which the Khazars and other marauding tribes had from time immemorial attacked the countries of the south and retreated again. Now it was the turn of the Arabs. Between 642 and 652 they repeatedly broke through the Darband Gate and advanced deep into Khazaria, attempting to capture Balanjar, the nearest town, and thus secure a foothold on the European side of the Caucasus. They were beaten back on every occasion in this first phase of the Arab-Khazar war; the last time in 652, in a great battle in which both sides used artillery (catapults and ballistae). Four thousand Arabs were killed, including their commander, Abdal-Rahman ibn-Rabiah; the rest fled in disorder across the mountains..For the next thirty or forty years the Arabs did not attempt any further incursions into the Khazar stronghold. Their main attacks were now aimed at Byzantium. On several occasions*[AD 669, 673-8, 717-18.] they laid siege to Constantinople by land and by sea; had they been able to outflank the capital across the Caucasus and round the Black Sea, the fate of the Roman Empire would probably have been sealed. The Khazars, in the meantime, having subjugated the Bulgars and Magyars, completed their western expansion into the Ukraine and the Crimea. But these were no longer haphazard raids to amass booty and prisoners; they were wars of conquest, incorporating the conquered people into an empire with a stable administration, ruled by the mighty Kagan, who appointed his provincial governors to administer and levy taxes in the conquered territories. At the beginning of the eighth century their state was sufficiently consolidated for the Khazars to take the offensive against the Arabs..From a distance of more than a thousand years, the period of intermittent warfare that followed (the so-called 'second Arab war", 722-37) looks like a series of tedious episodes on a local scale, following the same, repetitive pattern: the Khazar cavalry in their heavy armour breaking through the pass of Dariel or the Gate of Darband into the Caliph's domains to the south; followed by Arab counter-thrusts through the same pass or the defile, towards the Volga and back again. Looking thus through the wrong end of the telescope, one is reminded of the old jingle about the noble Duke of York who had ten thousand men; "he marched them up to the top of the hill. And he marched them down again." In fact, the Arab sources (though they often exaggerate) speak of armies of 100000, even of 300000, men engaged on either side - probably outnumbering the armies which decided the fate of the Western world at the battle of Tours about the same time..The death-defying fanaticism which characterized these wars is illustrated by episodes such as the suicide by fire of a whole Khazar town as an alternative to surrender; the poisoning of the water supply of Bab al Abwab by an Arab general; or by the traditional exhortation which would halt the rout of a defeated Arab army and make it fight to the last man: "To the Garden, Muslims, not the Fire" - the joys of Paradise being assured to every Muslim soldier killed in the Holy War. .At one stage during these fifteen years of fighting the Khazars overran Georgia and Armenia, inflicted a total defeat on the Arab army in the battle of Ardabil (AD 730) and advanced as far as Mosul and Dyarbakir, more than half-way to Damascus, capital of the Caliphate. But a freshly raised Muslim army stemmed the tide, and the Khazars retreated homewards across the mountains. The next year Maslamah ibn-Abd-al-Malik, most famed Arab general of his time, who had formerly commanded the siege of Constantinople, took Balanjar and even got as far as Samandar, another large Khazar town further north. But once more the invaders were unable to establish a permanent garrison, and once more they were forced to retreat across the Caucasus. The sigh of relief experienced in the Roman Empire assumed a tangible form through another dynastic alliance, when the heir to the throne was married to a Khazar princess, whose son was to rule Byzantium as Leo the Khazar. .The last Arab campaign was led by the future Caliph Marwan II, and ended in a Pyrrhic victory. Marwan made an offer of alliance to the Khazar Kagan, then attacked by surprise through both passes. The Khazar army, unable to recover from the initial shock, retreated as far as the Volga. The Kagan was forced to ask for terms; Marwan, in accordance with the routine followed in other conquered countries, requested the Kagan's conversion to the True Faith. The Kagan complied, but his conversion to Islam must have been an act of lip-service, for no more is heard of the episode in the Arab or Byzantine sources - in contrast to the lasting effects of the establishment of Judaism as the state religion which took place a few years later.*[The probable date for the conversion is around AD 740 - see below.] Content with the results achieved, Marwan bid farewell to Khazaria and marched his army back to Transcaucasia - without leaving any garrison, governor or administrative apparatus behind. On the contrary, a short time later he requested terms for another alliance with the Khazars against the rebellious tribes of the south. .It had been a narrow escape. The reasons which prompted Marwan's apparent magnanimity are a matter of conjecture - as so much else in this bizarre chapter of history. Perhaps the Arabs realized that, unlike the relatively civilized Persians, Armenians or Georgians, these ferocious Barbarians of the North could not be ruled by a Muslim puppet prince and a small garrison. Yet Marwan needed every man of his army to quell major rebellions in Syria and other parts of the Omayad Caliphate, which was in the process of breaking up. Marwan himself was the chief commander in the civil wars that followed, and became in 744 the last of the Omayad Caliphs (only to be assassinated six years later when the Caliphate passed to the Abbasid dynasty). Given this background, Marwan was simply not in a position to exhaust his resources by further wars with the Khazars. He had to content himself with teaching them a lesson which would deter them from further incursions across the Caucasus. .Thus the gigantic Muslim pincer movement across the Pyrenees in the west and across the Caucasus into Eastern Europe was halted at both ends about the same time. As Charles Martel's Franks saved Gaul and Western Europe, so the Khazars saved the eastern approaches to the Volga, the Danube, and the East Roman Empire itself. On this point at least, the Soviet archaeologist and historian, Artamonov, and the American historian, Dunlop, are in full agreement. I have already quoted the latter to the effect that but for the Khazars, "Byzantium, the bulwark of European civilization to the East, would have found itself outflanked by the Arabs", and that history might have taken a different course..Artamonov is of the same opinion:18
Lastly, the Professor of Russian History in the University of Oxford, Dimitry Obolensky:19 "The main contribution of the Khazars to world history was their success in holding the line of the Caucasus against the northward onslaught of the Arabs.".Marwan was not only the last Arab general to attack the Khazars, he was also the last Caliph to pursue an expansionist policy devoted, at least in theory, to the ideal of making Islam triumph all over the world. With the Abbasid caliphs the wars of conquest ceased, the revived influence of the old Persian culture created a mellower climate, and eventually gave rise to the splendours of Baghdad under Harun al Rashid.
During the long lull between the first and second Arab wars, the Khazars became involved in one of the more lurid episodes of Byzantine history, characteristic of the times, and of the role the Khazars played in it. .In AD 685 Justinian II, Rhinotmetus, became East Roman Emperor at the age of sixteen. Gibbon, in his inimitable way, has drawn the youth's portrait:20
After ten years of intolerable misrule there was a revolution, and the new Emperor, Leontius, ordered Justinian's mutilation and banishment:21
From the chronological point of view, the next event to be discussed should be the conversion of the Khazars to Judaism, around AD 740. But to see that remarkable event in its proper perspective, one should have at least some sketchy idea of the habits, customs and everyday life among the Khazars prior to the conversion..Alas, we have no lively eyewitness reports, such as Priscus's description of Attila's court. What we do have are mainly second-hand accounts and compilations by Byzantine and Arab chroniclers, which are rather schematic and fragmentary - with two exceptions. One is a letter, purportedly from a Khazar king, to be discussed in Chapter 2; the other is a travelogue by an observant Arab traveller, Ibn Fadlan, who - like Priscus - was a member of a diplomatic mission from a civilized court to the Barbarians of the North..The court was that of the Caliph al Muktadir, and the diplomatic mission travelled from Baghdad through Persia and Bukhara to the land of the Volga Bulgars. The official pretext for this grandiose expedition was a letter of invitation from the Bulgar king, who asked the Caliph (a) for religious instructors to convert his people to Islam, and (b) to build him a fortress which would enable him to defy his overlord, the King of the Khazars. The invitation - which was no doubt prearranged by earlier diplomatic contacts - also provided an opportunity to create goodwill among the various Turkish tribes inhabiting territories through which the mission had to pass, by preaching the message of the Koran and distributing huge amounts of gold bakhshish. .The opening paragraphs of our traveller's account read:*[The following quotations are based on Zeki Validi Togan's German translation of the Arabic text and the English translation of extracts by Blake and Frye, both slightly paraphrased in the interest of readability.]
The date of the expedition, it will he noted, is much later than the events described in the previous section. But as far as the customs and institutions of the Khazars' pagan neighbours are concerned, this probably makes not much difference; and the glimpses we get of the life of these nomadic tribes convey at least some idea of what life among the Khazars may have been during that earlier period - before the conversion - when they adhered to a form of Shamanism similar to that still practised by their neighbours in Ibn Fadlan's time. .The progress of the mission was slow and apparently uneventful until they reached Khwarizm, the border province of the Caliphate south of the Sea of Aral. Here the governor in charge of the province tried to stop them from proceeding further by arguing that between his country and the kingdom of the Bulgars there were "a thousand tribes of disbelievers" who were sure to kill them. In fact his attempts to disregard the Caliph's instructions to let the mission pass might have been due to other motives: he realized that the mission was indirectly aimed against the Khazars, with whom he maintained a flourishing trade and friendly relations. In the end, however, he had to give in, and the mission was allowed to proceed to Gurganj on the estuary of the Amu-Darya. Here they hibernated for three months, because of the intense cold - a factor which looms large in many Arab travellers' tales:
Around the middle of February the thaw set in. The mission arranged to join a mighty caravan of 5000 men and 3000 pack animals to cross the northern steppes, and bought the necessary supplies: camels, skin boats made of camel hides for crossing rivers, bread, millet and spiced meat for three months. The natives warned them about the even more frightful cold in the north, and advised them what clothes to wear:
Ibn Fadlan, the fastidious Arab, liked neither the climate nor the people of Khwarizm:
They left on March 3 and stopped for the night in a caravanserai called Zamgan - the gateway to the territory of the Ghuzz Turks. From here onward the mission was in foreign land, "entrusting our fate to the all-powerful and exalted God". During one of the frequent snow-storms, Ibn Fadlan rode next to a Turk, who complained: "What does the Ruler want from us? He is killing us with cold. If we knew what he wants we would give it to him." Ibn Fadlan: "All he wants is that you people should say: "There is no God save Allah"." The Turk laughed: "If we knew that it is so, we should say so." .There are many such incidents, which Ibn Fadlan reports without appreciating the independence of mind which they reflect. Nor did the envoy of the Baghdad court appreciate the nomadic tribesmen's fundamental contempt for authority. The following episode also occurred in the country of the powerful Ghuzz Turks, who paid tribute to the Khazars and, according to some sources, were closely related to them:24
The democratic methods of the Ghuzz, practised when a decision had to be taken, were even more bewildering to the representative of an authoritarian theocracy:
The sexual mores of the Ghuzz - and other tribes - were a remarkable mixture of liberalism and savagery:
He does not say whether the same punishment was meted out to the guilty woman. Later on, when talking about the Volga Bulgars, he describes an equally savage method of splitting adulterers into two, applied to both men and women. Yet, he notes with astonishment, Bulgars of both sexes swim naked in their rivers, and have as little bodily shame as the Ghuzz..As for homosexuality - which in Arab countries was taken as a matter of course - Ibn Fadlan says that it is "regarded by the Turks as a terrible sin". But in the only episode he relates to prove his point, the seducer of a "beardless youth" gets away with a fine of 400 sheep. .Accustomed to the splendid baths of Baghdad, our traveller could not get over the dirtiness of the Turks. "The Ghuzz do not wash themselves after defacating or urinating, nor do they bathe after seminal pollution or on other occasions. They refuse to have anything to do with water, particularly in winter....".When the Ghuzz commander-in-chief took off his luxurious coat of brocade to don a new coat the mission had brought him, they saw that his underclothes were "fraying apart from dirt, for it is their custom never to take off the garment they wear close to their bodies until it disintegrates". Another Turkish tribe, the Bashkirs, 'shave their beards and eat their lice. They search the folds of their undergarments and crack the lice with their teeth." When Ibn Fadlan watched a Bashkir do this, the latter remarked to him: "They are delicious." .All in all, it is not an engaging picture. Our fastidious traveller's contempt for the barbarians was profound. But it was only aroused by their uncleanliness and what he considered as indecent exposure of the body; the savagery of their punishments and sacrificial rites leave him quite indifferent. Thus he describes the Bulgars' punishment for manslaughter with detached interest, without his otherwise frequent expressions of indignation: "They make for him [the delinquent] a box of birchwood, put him inside, nail the lid on the box, put three loaves of bread and a can of water beside it, and suspend the box between two tall poles, saying: "We have put him between heaven and earth, that he may be exposed to the sun and the rain, and that the deity may perhaps forgive him." And so he remains suspended until time lets him decay and the winds blow him away.".He also describes, with similar aloofness, the funeral sacrifice of hundreds of horses and herds of other animals, and the gruesome ritual killing of a Rus*[Rus: the Viking founders of the early Russian settlements - see below, Chapter III.] slave girl at her master's bier. About pagan religions he has little to say. But the Bashkirs' phallus cult arouses his interest, for he asks through his interpreter one of the natives the reason for his worshipping a wooden penis, and notes down his reply: "Because I issued from something similar and know of no other creator who made me." He then adds that 'some of them [the Bashkirs] believe in twelve deities, a god for winter, another for summer, one for the rain, one for the wind, one for the trees, one for men, one for the horse, one for water, one for the night, one for the day, a god of death and one for the earth; while that god who dwells in the sky is the greatest among them, but takes counsel with the others and thus all are contented with each other's doings.... We have seen a group among them which worships snakes, and a group which worships fish, and a group which worships cranes....".Among the Volga Bulgars, Ibn Fadlan found a strange custom:
Commenting on this passage, the Turkish orientalist Zeki Validi Togan, undisputed authority on Ibn Fadlan and his times, has this to say:25 "There is nothing mysterious about the cruel treatment meted out by the Bulgars to people who were overly clever. It was based on the simple, sober reasoning of the average citizens who wanted only to lead what they considered to be a normal life, and to avoid any risk or adventure into which the "genius" might lead them." He then quotes a Tartar proverb: "If you know too much, they will hang you, and if you are too modest, they will trample on you." He concludes that the victim 'should not be regarded simply as a learned person, but as an unruly genius, one who is too clever by half". This leads one to believe that the custom should be regarded as a measure of social defence against change, a punishment of non-conformists and potential innovators.*[In support of his argument, the author adduces Turkish and Arabic quotations in the original, without translation - a nasty habit common among modern experts in the field.] But a few lines further down he gives a different interpretation:
Perhaps both types of motivation were mixed together: 'since sacrifice is a necessity, let's sacrifice the trouble-makers". .We shall see that human sacrifice was also practised by the Khazars - including the ritual killing of the king at the end of his reign. We may assume that many other similarities existed between the customs of the tribes described by Ibn Fadlan and those of the Khazars. Unfortunately he was debarred from visiting the Khazar capital and had to rely on information collected in territories under Khazar dominion, and particularly at the Bulgar court.
It took the Caliph's mission nearly a year (from June 21, 921, to May 12, 922) to reach its destination, the land of the Volga Bulgars. The direct route from Baghdad to the Volga leads across the Caucasus and Khazaria - to avoid the latter, they had to make the enormous detour round the eastern shore of the "Khazar Sea", the Caspian. Even so, they were constantly reminded of the proximity of the Khazars and its potential dangers. .A characteristic episode took place during their sojourn with the Ghuzz army chief (the one with the disreputable underwear). They were at first well received, and given a banquet. But later the Ghuzz leaders had second thoughts because of their relations with the Khazars. The chief assembled the leaders to decide what to do:
They argued among themselves for seven days, while Ibn Fadlan and his people feared the worst. In the end the Ghuzz let them go; we are not told why. Probably Ibn Fadlan succeeded in persuading them that his mission was in fact directed against the Khazars. The Ghuzz had earlier on fought with the Khazars against another Turkish tribe, the Pechenegs, but more recently had shown a hostile attitude; hence the hostages the Khazars took..The Khazar menace loomed large on the horizon all along the journey. North of the Caspian they made another huge detour before reaching the Bulgar encampment somewhere near the confluence of the Volga and the Kama. There the King and leaders of the Bulgars were waiting for them in a state of acute anxiety. As soon as the ceremonies and festivities were over, the King sent for Ibn Fadlan to discuss business. He reminded Ibn Fadlan in forceful language ("his voice sounded as if he were speaking from the bottom of a barrel") of the main purpose of the mission to wit, the money to be paid to him 'so that I shall be able to build a fortress to protect me from the Jews who subjugated me". Unfortunately that money - a sum of four thousand dinars - had not been handed over to the mission, owing to some complicated matter of red tape; it was to be sent later on. On learning this, the King - "a personality of impressive appearance, broad and corpulent" - seemed close to despair. He suspected the mission of having defrauded the money: ""What would you think of a group of men who are given a sum of money destined for a people that is weak, besieged, and oppressed, yet these men defraud the money?" I replied: "This is forbidden, those men would be evil." He asked: "Is this a matter of opinion or a matter of general consent?" I replied: "A matter of general consent."" .Gradually Ibn Fadlan succeeded in convincing the King that the money was only delayed,*[Apparently it did arrive at some time, as there is no further mention of the matter.] but not to allay his anxieties. The King kept repeating that the whole point of the invitation was the building of the fortress "because he was afraid of the King of the Khazars". And apparently he had every reason to be afraid, as Ibn Fadlan relates:
It sounds like a refrain. Ibn Fadlan also specifies the annual tribute the Bulgar King had to pay the Khazars: one sable fur from each household in his realm. Since the number of Bulgar households (i.e., tents) is estimated to have been around 50000, and since Bulgar sable fur was highly valued all over the world, the tribute was a handsome one.
What Ibn Fadlan has to tell us about the Khazars is based - as already mentioned - on intelligence collected in the course of his journey, but mainly at the Bulgar court. Unlike the rest of his narrative, derived from vivid personal observations, the pages on the Khazars contain second-hand, potted information, and fall rather flat. Moreover, the sources of his information are biased, in view of the Bulgar King's understandable dislike of his Khazar overlord - while the Caliphate's resentment of a kingdom embracing a rival religion need hardly be stressed. .The narrative switches abruptly from a description of the Rus court to the Khazar court:
Ibn Fadlan then proceeds to give a rather fanciful description of the Kagan's harem, where each of the eighty-five wives and concubines has a "palace of her own", and an attendant or eunuch who, at the King's command, brings her to his alcove "faster than the blinking of an eye..After a few more dubious remarks about the "customs" of the Khazar Kagan (we shall return to them later), Ibn Fadlan at last provides some factual information about the country:
Ibn Fadlan's travel report, as far as it is preserved, ends with the words:
I have quoted Ibn Fadlan's odyssey at some length, not so much because of the scant information he provides about the Khazars themselves, but because of the light it throws on the world which surrounded them, the stark barbarity of the people amidst whom they lived, reflecting their own past, prior to the conversion. For, by the time of Ibn Fadlan's visit to the Bulgars, Khazaria was a surprisingly modern country compared to its neighbours. .The contrast is evidenced by the reports of other Arab historians,*[The following pages are based on the works of lstakhri, al- Masudi, Ibn Rusta and Ibn Hawkal (see Appendix II).] and is present on every level, from housing to the administration of justice. The Bulgars still live exclusively in tents, including the King, although the royal tent is "very large, holding a thousand people or more".26 On the other hand, the Khazar Kagan inhabits a castle built of burnt brick, his ladies are said to inhabit "palaces with roofs of teak",27 and the Muslims have several mosques, among them "one whose minaret rises above the royal castle".28 .In the fertile regions, their farms and cultivat ed areas stretched out continuously over sixty or seventy miles. They also had extensive vineyards. Thus Ibn Hawkal: "In Kozr [Khazaria] there is a certain city called Asmid [Samandar] which has so many orchards and gardens that from Darband to Serir the whole country is covered with gardens and plantations belonging to this city. It is said that there are about forty thousand of them. Many of these produce grapes."29 .The region north of the Caucasus was extremely fertile. In AD 968 Ibn Hawkal met a man who had visited it after a Russian raid: "He said there is not a pittance left for the poor in any vineyard or garden, not a leaf on the bough.... [But] owing to the excellence of their land and the abundance of its produce it will not take three years until it becomes again what it was." Caucasian wine is still a delight, consumed in vast quantities in the Soviet Union. lHowever, the royal treasuries' main source of income was foreign trade. The sheer volume of the trading caravans plying their way between Central Asia and the Volga-Ural region is indicated by Ibn Fadlan: we remember that the caravan his mission joined at Gurganj consisted of "5000 men and 3000 pack animals". Making due allowance for exaggeration, it must still have been a mighty caravan, and we do not know how many of these were at any time on the move. Nor what goods they transported - although textiles, dried fruit, honey, wax and spices seem to have played an important part. A second major trade route led across the Caucasus to Armenia, Georgia, Persia and Byzantium. A third consisted of the increasing traffic of Rus merchant fleets down the Volga to the eastern shores of the Khazar Sea, carrying mainly precious furs much in demand among the Muslim aristocracy, and slaves from the north, sold at the slave market of Itil. On all these transit goods, including the slaves, the Khazar ruler levied a tax of ten per cent. Adding to this the tribute paid by Bulgars, Magyars, Burtas and so on, one realizes that Khazaria was a prosperous country - but also that its prosperity depended to a large extent on its military power, and the prestige it conveyed on its tax collectors and customs officials. .Apart from the fertile regions of the south, with their vineyards and orchards, the country was poor in natural resources. One Arab historian (Istakhri) says that the only native product they exported was isinglass. This again is certainly an exaggeration, yet the fact remains that their main commercial activity seems to have consisted in re-exporting goods brought in from abroad. Among these goods, honey and candle-wax particularly caught the Arab chroniclers' imagination. Thus Muqaddasi: "In Khazaria, sheep, honey and Jews exist in large quantities."30 It is true that one source - the Darband Namah - mentions gold or silver mines in Khazar territory, but their location has not been ascertained. On the other hand, several of the sources mention Khazar merchandise seen in Baghdad, and the presence of Khazar merchants in Constantinople, Alexandria and as far afield as Samara and Fergana. .Thus Khazaria was by no means isolated from the civilized world; compared to its tribal neighbours in the north it was a cosmopolitan country, open to all sorts of cultural and religious influences, yet jealously defending its independence against the two ecclesiastical world powers. We shall see that this attitude prepared the ground for the coup de theatre - or coup d'tat - which established Judaism as the state religion. .The arts and crafts seem to have flourished, including haute couture. When the future Emperor Constantine V married the Khazar Kagan's daughter (see above, section 1), she brought with her dowry a splendid dress which so impressed the Byzantine court that it was adopted as a male ceremonial robe; they called it tzitzakion, derived from the Khazar-Turkish pet- name of the Princess, which was Chichak or "flower" (until she was baptized Eirene). "Here," Toynbee comments, "we have an illuminating fragment of cultural history."31 When another Khazar princess married the Muslim governor of Armenia, her cavalcade contained, apart from attendants and slaves, ten tents mounted on wheels, "made of the finest silk, with gold-and silver-plated doors, the floors covered with sable furs. Twenty others carried the gold and silver vessels and other treasures which were her dowry".32 The Kagan himself travelled in a mobile tent even more luxuriously equipped, carrying on its top a pomegranate of gold.
Khazar art, like that of the Bulgars and Magyars, was mainly imitative, modelled on Persian-Sassanide patterns. The Soviet archaeologist Bader33 emphasized the role of the Khazars in the spreading of Persian-style silver-ware towards the north. Some of these finds may have been re-exported by the Khazars, true to their role as middlemen; others were imitations made in Khazar workshops - the ruins of which have been traced near the ancient Khazar fortress of Sarkel.*[Unfortunately, Sarkel, the most important Khazar archaeological site has been flooded by the reservoir of a newly built hydro-electric station.] The jewellery unearthed within the confines of the fortress was of local manufacture.34 The Swedish archaeologist T. J. Arne mentions ornamental plates, clasps and buckles found as far as Sweden, of Sassanide and Byzantine inspiration, manufactured in Khazaria or territories under their influence.35 .Thus the Khazars were the principal intermediaries in the spreading of Persian and Byzantine art among the semi-barbaric tribes of Eastern Europe. After his exhaustive survey of the archaeological and documentary evidence (mostly from Soviet sources), Bartha concludes:
The last remark of the Hungarian scholar refers to the spectacular archaeological finds known as the "Treasure of Nagyszentmiklos" (see frontispiece). The treasure, consisting of twentythree gold vessels, dating from the tenth century, was found in 1791 in the vicinity of the village of that name.*[It now belongs to Rumania and is called Sinnicolaul Mare.] Bartha points out that the figure of the "victorious Prince" dragging a prisoner along by his hair, and the mythological scene at the back of the golden jar, as well as the design of other ornamental objects, show close affinities with the finds in Novi Pazar in Bulgaria and in Khazar Sarkel. As both Magyars and Bulgars were under Khazar suzerainty for protracted periods, this is not very surprising, and the warrior, together with the rest of the treasure, gives us at least some idea of the arts practised within the Khazar Empire (the Persian and Byzantine influence is predominant, as one would expect).*[The interested reader will find an excellent collection of photographs in Gyula L�szl�'s The Art of the Migration Period (although his historical comments have to be treated with caution).] .One school of Hungarian archaeologists maintains that the tenth century gold-and silversmiths working in Hungary were actually Khazars.37 As we shall see later on (see III, 7, 8), when the Magyars migrated to Hungary in 896 they were led by a dissident Khazar tribe, known as the Kabars, who settled with them in their new home. The Kabar-Khazars were known as skilled gold and silversmiths; the (originally more primitive) Magyars only acquired these skills in their new country. Thus the theory of the Khazar origin of at least some of the archaeological finds in Hungary is not implausible - as will become clearer in the light of the Magyar-Khazar nexus discussed later on.
Whether the warrior on the golden jar is of Magyar or Khazar origin, he helps us to visualise the appearance of a cavalryman of that period, perhaps belonging to an elite regiment. Masudi says that in the Khazar army 'seven thousand of them*[Istakhri has 12000.] ride with the King, archers with breast plates, helmets, and coats of mail. Some are lancers, equipped and armed like the Muslims.... None of the kings in this part of the world has a regular standing army except the King of the Khazars." And Ibn Hawkal: "This king has twelve thousand soldiers in his service, of whom when one dies, another person is immediately chosen in his place." .Here we have another important clue to the Khazar dominance: a permanent professional army, with a Praetorian Guard which, in peacetime, effectively controlled the ethnic patchwork, and in times of war served as a hard core for the armed horde, which, as we have seen, may have swollen at times to a hundred thousand or more.*[According to Masudi, the "Royal Army" consisted of Muslims who "immigrated from the neighbourhood of Kwarizm. Long ago, after the appearance of Islam, there was war and pestilence in their territory, and they repaired to the Khazar king.... When the king of the Khazars is at war with the Muslims, they have a separate place in his army and do not fight the people of their own faith" [Quoted by Dunlop (1954), p. 206] That the army "consisted" of Muslims is of course an exaggeration, contradicted by Masudi himself a few lines later, where he speaks of the Muslim contingent having a "separate place" in the Khazar army. Also, lbn Hawkal says that "the king has in his train 4000 Muslims and this king has 2000 soldiers in his service". The Kwarizmians probably formed a kind of Swiss Guard within the army, and their compatriots" talk of "hostages" (see above, section 10) may refer to them. Vice versa, the Byzantine Emperor Constantine Porphyrogenitus had a corps d'�lite of Khazar guardsmen stationed at the gates of his palace. This was a privilege dearly bought: "These guards were so well remunerated that they had to purchase their posts for considerable sums, on which their salaries represented an annuity varying from about 2.25 to 4 per cent." (Constantine, De Ceremoniis, pp. 692-3). For example, "a Khazar who received 7.4s. had paid for enrolment 302.8s." (Bury, p. 228n).]
The capital of this motley empire was at first probably the fortress of Balanjar in the northern foothills of the Caucasus; after the Arab raids in the eighth century it was transferred to Samandar, on the western shore of the Caspian; and lastly to Itil in the estuary of the Volga. .We have several descriptions of Itil, which are fairly consistent with each other. It was a twin city, built on both sides of the river. The eastern half was called Khazaran, the western half Itil;*[The town was in different periods also mentioned under different names, e.g., al-Bayada, "The White City".] the two were connected by a pontoon bridge. The western half was surrounded by a fortified wall, built of brick; it contained the palaces and courts of the Kagan and the Bek, the habitations of their attendants*[Masudi places these buildings on an island, close to the west bank, or a peninsula.] and of the "pure-bred Khazars". The wall had four gates, one of them facing the river. Across the river, on the east bank, lived "the Muslims and idol worshippers";38 this part also housed the mosques, markets, baths and other public amenities. Several Arab writers were impressed by the number of mosques in the Muslim quarter and the height of the principal minaret. They also kept stressing the autonomy enjoyed by the Muslim courts and clergy. Here is what al-Masudi, known as "the Herodotus among the Arabs", has to say on this subject in his oft-quoted work Meadows of Gold Mines and Precious Stones:
In reading these lines by the foremost Arab historian, written in the first half of the tenth century,*[Supposedly between AD 943 and 947.] one is tempted to take a perhaps too idyllic view of life in the Khazar kingdom. Thus we read in the article "Khazars" in the Jewish Encyclopaedia: "In a time when fanaticism, ignorance and anarchy reigned in Western Europe, the Kingdom of the Khazars could boast of its just and broad-minded administration."*[Jewish Encyclopaedia, published 1901-6. In the Encyclopaedia Judaica, 1971, the article on the Khazars by Dunlop is of exemplary objectivity.] .This, as we have seen, is partly true; but only partly. There is no evidence of the Khazars engaging in religious persecution, either before or after the conversion to Judaism. In this respect they may be called more tolerant and enlightened than the East Roman Empire, or Islam in its early stages. On the other hand, they seem to have preserved some barbaric rituals from their tribal past. We have heard Ibn Fadlan on the killings of the royal gravediggers. He also has something to say about another archaic custom regicide: "The period of the king's rule is forty years. If he exceeds this time by a single day, his subjects and attendants kill him, saying "His reasoning is already dimmed, and his insight confused"." .Istakhri has a different version of it:
Bury39 is doubtful whether to believe this kind of Arab traveller's lore, and one would indeed be inclined to dismiss it, if ritual regicide had not been such a widespread phenomenon among primitive (and not-so-primitive) people. Frazer laid great emphasis on the connection between the concept of the King's divinity, and the sacred obligation to kill him after a fixed period, or when his vitality is on the wane, so that the divine power may find a more youthful and vigorous incarnation.*[Frazer wrote a special treatise on these lines on "The Killing of the Khazar Kings" (Folklore, XXVIII, 1917).] .It speaks in Istakhri's favour that the bizarre ceremony of "choking" the future King has been reported in existence apparently not so long ago among another people, the Kok-Turks. Zeki Validi quotes a French anthropologist, St Julien, writing in 1864:
When the new Chief has been elected, his officers and attendants ... make him mount his horse. They tighten a ribbon of silk round his neck, without quite strangling him; then they loosen the ribbon and ask him with great insistence: "For how many years canst thou be our Khan?" The king, in his troubled mind, being unable to name a figure, his subjects decide, on the strength of the words that have escaped him, whether his rule will be long or brief.40
We do not know whether the Khazar rite of slaying the King (if it ever existed) fell into abeyance when they adopted Judaism, in which case the Arab writers were confusing past with present practices as they did all the time, compiling earlier travellers' reports, and attributing them to contemporaries. However that may be, the point to be retained, and which seems beyond dispute, is the divine role attributed to the Kagan, regardless whether or not it implied his ultimate sacrifice. We have heard before that he was venerated, but virtually kept in seclusion, cut off from the people, until he was buried with enormous ceremony. The affairs of state, including leadership of the army, were managed by the Bek (sometimes also called the Kagan Bek), who wielded all effective power. On this point Arab sources and modern historians are in agreement, and the latter usually describe the Khazar system of government as a "double kingship", the Kagan representing divine, the Bek secular, power. .The Khazar double kingship has been compared - quite mistakenly, it Seems - with the Spartan dyarchy and with the superficially similar dual leadership among various Turkish tribes. However, the two kings of Sparta, descendants of two leading families, wielded equal power; and as for the dual leadership among nomadic tribes,*[Alf�ldi has suggested that the two leaders were the commanders of the two wings of the horde (quoted by Dunlop, p. 159, n. 123).] there is no evidence of a basic division of functions as among the Khazars. A more valid comparison is the system of government in Japan, from the Middle Ages to 1867, where secular power was concentrated in the hands of the shogun, while the Mikado was worshipped from afar as a divine figurehead. .Cassel41 has suggested an attractive analogy between the Khazar system of government and the game of chess. The double kingship is represented on the chess-board by the King (the Kagan) and the Queen (the Bek). The King is kept in seclusion, protected by his attendants, has little power and can only move one short step at a time. The Queen, by contrast, is the most powerful presence on the board, which she dominates. Yet the Queen may be lost and the game still continued, whereas the fall of the King is the ultimate disaster which instantly brings the contest to an end. .The double kingship thus seems to indicate a categorical distinction between the sacred and the profane in the mentality of the Khazars. The divine attributes of the Kagan are much in evidence in the following passage from Ibn Hawkal: *[Ibn Hawkal, another much-travelled Arab geographer and historian, wrote his Oriental Geography around AD 977. The passage here quoted is virtually a copy of what Istakhri wrote forty years earlier, but contains less obscurities, so I have followed Ouseley's translation (1800) of Ibn Hawkal.]
The passage about the virtuous young man selling bread, or whatever it is, in the bazaar sounds rather like a tale about Harun al Rashid. If he was heir to the golden throne reserved for Jews, why then was he brought up as a poor Muslim? If we are to make any sense at all of the story, we must assume that the Kagan was chosen on the strength of his noble virtues, but chosen among members of the "Imperial Race" or "family of notables". This is in fact the view of Artamonov and Zeki Validi. Artamonov holds that the Khazars and other Turkish people were ruled by descendants of the Turkut dynasty, the erstwhile sovereigns of the defunct Turk Empire (cf. above, section 3). Zeki Validi suggests that the "Imperial Race" or "family of notables", to which the Kagan must belong, refers to the ancient dynasty of the Asena, mentioned in Chinese sources, a kind of desert aristocracy, from which Turkish and Mongol rulers traditionally claimed descent. This sounds fairly plausible and goes some way towards reconciling the contradictory values implied in the narrative just quoted: the noble youth without a dirhem to his name - and the pomp and circumstance surrounding the golden throne. We are witnessing the overlap of two traditions, like the optical interference of two wave-patterns on a screen: the asceticism of a tribe of hard-living desert nomads, and the glitter of a royal court prospering on its commerce and crafts, and striving to outshine its rivals in Baghdad and Constantinople. After all, the creeds professed by those sumptuous courts had also been inspired by ascetic desert- prophets in the past..All this does not explain the startling division of divine and secular power, apparently unique in that period and region. As Bury wrote:43 "We have no information at what time the active authority of the Chagan was exchanged for his divine nullity, or why he was exalted to a position resembling that of the Emperor of Japan, in which his existence, and not his government, was considered essential to the prosperity of the State." .A speculative answer to this question has recently been proposed by Artamonov. He suggests that the acceptance of Judaism as the state religion was the result of a coup d'�tat, which at the same time reduced the Kagan, descendant of a pagan dynasty whose allegiance to Mosaic law could not really be trusted, to a mere figurehead. This is a hypothesis as good as any other - and with as little evidence to support it. Yet it seems probable that the two events - the adoption of Judaism and the establishment of the double kingship - were somehow connected.*[Before the conversion the Kagan was still reported to play an active role - as, for instance, in his dealings with Justinian. To complicate matters further, the Arab sources sometimes refer to the "Kagan" when they clearly mean the "Bek" (as "kagan" was the generic term for "ruler" among many tribes), and they also use different names for the Bek, as the following list shows (after Minorsky, Hudud al Alam, p. 451):
Const. Porphyr. Khaqan Bek Ibn Rusta Khazar Khaqan Aysha Masudi Khaqan Malik Istakhri Malik Khazar Khaqan Khazarl. Ibn Hawkal Khaqan Khazar Malik Khazar or Bek Gardezi Khazar Khaqan Abshad
.The order of the rulers appears to have been switched.]