“The Turks will take control of Kūfa in southern Iraq and the Khazars the province of al-Jazīra in northern Iraq” runs a popular ḥadīth of the early ‘Abbāsid Caliphate. The patently false attribution to the Prophet notwithstanding, this ḥadīth suggests that the Khazars, a militarised semi-nomadic Turkic people securely established in their Khaganate north of the Caucasus Mountains, loomed menacing and large in the imagination – and fears – of their settled neighbours. This is hardly a surprise. The Khazars, who had emerged as the hegemonic regional power after the fragmentation of the Gök Turk Empire, had grown rich as intermediaries in a north-south trading network, and from the tributes derived from their 25 subject peoples. They employed a centralised fiscal administration which allowed them to maintain a standing army of at least ten thousand, whose ranks could probably swell two-or-three-fold with the retinues of their notables and contributions from their subject tribes, and were ensconced within a network of fortifications that provided them with a degree of internal stability and potential for aggressive campaigning highly unusual within other steppe polities.Footnote 2
Aside from the military threat provided by this most durable of steppe empires, which constituted the most powerful West Eurasian steppe power from the late seventh century to the mid tenth century, there were two peculiarities of the Khazar Khaganate that particularly attracted the interest – and occasional unease – of sedentary observers.Footnote 3 First, Khazaria developed a unique diarchic system of governance in which the office of the Khagan, a traditional Turkic institution that embodied absolute authority deriving from both divine mandate and membership of a charismatic clan, was stripped of its temporal authority which then passed to a Khazar king, who is usually termed išad or beg in our sources. Diminishing earthly power was probably accompanied by increasing heavenly power, and the office of Khagan became entirely sacralised. This was manifested most clearly in the intensification of the intimate link between the health of the Khagan and the prosperity of the state. Perceived weaknesses in the Khaganate suggested weakness in the Khagan, and vice versa, and in times of dire crisis the Khagan was removed from his office so that a more divinely-favoured successor might restore good-fortune to Khazaria.Footnote 4
The fullest reports on the Khazar diarchy are predominantly found in the works of Muslim geographers and travellers who, no doubt because of the political situation in the ‘Abbāsid Caliphate, were fascinated by the exact nuances of the Khagan and king's political and religious authority, and who in any case had superior access to information because of the greater extent of the Islamic world's economic and diplomatic contacts with Khazaria. However, the second of the Khazar peculiarities is found in reports from as far afield as the ‘Abbāsid Caliphate, Byzantium, Umayyad Spain, and Western Europe. This was, of course, the conversion of the Khazar elite to Judaism. While the emergence of the Khazar diarchy and the conversion to Judaism are inextricably entwined, it is the conversion itself – an almost unprecedented event in world history – which provides the main impetus for the present study.Footnote 5
As we shall see, scholars who have contributed to the subject of the Khazars’ conversion, have based their arguments on a limited corpus of textual, and more recently numismatic, evidence, which, even at its most forgiving presents numerous difficulties. Taken together these sources offer a cacophony of distortions, contradictions, vested interests, and anomalies in some areas, and nothing but silence in others. It is probably with these frustrations in mind that Peter Golden, in 1983, published an article underlining exactly how little had been firmly established about the chronology, process, consequences, and motivations for the Khazar conversion. Golden compiled a list of over a dozen questions which required new answers in order to better understand the conversion of the Khazars. A simplified version of this list could be reduced to ‘when’, ‘how’, and ‘why’ the Khazars adopted Judaism. It is with the first two of these questions that this investigation is directly concerned. The ‘why’ is, of course, unavoidable, but owing to its importance in the wider field of steppe conversion it shall receive a more systematic treatment in a subsequent article.Footnote 6
Already formulated before Golden's rallying cry but later refined to incorporate new ideas, Omeljian Pritsak's scenario for the conversion of the Khazars is worthy of attention. Based on a notice in Judah Halevi's theological polemic, the Kuzari (composed ca. 1140), which states that the conversion took place 400 years before Halevi's time, a reference to the conversion of the Khazars in the reign of Hārūn al-Rashīd (r. 786–809) in a work of the prolific tenth-century Baghdādī scholar al-Mas‘ūdī, and some diplomatic correspondence of the late 830s and early 840s which points to the increasing power of the išad/beg, Pritsak posited a three-stage conversion process. The first stage was the private conversion to Judaism by a member of the išad/beg dynasty in about 730. This was followed by an official proclamation in the reign of Hārūn stating that the clan had converted. Finally, sometime in the second quarter of the ninth century, the išad/beg forced the Khaganal dynasty to convert.Footnote 7
In 1995 Constantin Zuckerman published an article which attempted to shed new light on the conversion process by focusing on an entirely different set of textual sources, the three ‘disputation accounts’ found in the Vita Constantinii, the Schechter Document and the Reply of Joseph to Hasdai ibn Shaprut. Zuckerman argued that these accounts, which all depicted theological debates at the Khazar court, described the same event. According to his reading of the Vita Constantinii the Khazars had not yet converted to a monotheistic religion at the time of Constantine's mission – which can be dated by independent evidence to 861 – and so conversion must have occurred after this date. Zuckerman's arguments soon found a willing supporter in Jonathan Shepard.Footnote 8
In 2002, however, an article was published describing that among the many Khazar coins discovered was a type bearing the Jewish legend ‘Mūsā rasūl Allāh’ (Moses is the Messenger of God), which could be securely dated to 837–38. Roman Kovalev, who used this numismatic evidence to formulate a new scenario for Khazar conversion, argued that the minting of this Jewish coinage required that both išad/beg and Khagan profess Judaism. Moreover, the evidence that Pritsak had identified for the growing role of the išad/beg from the late 830s confirmed Kovalev's suspicion that it must have been in 837–38 that the Khazar king forced through the Judaisation of the Khagan and the Khazar elite. Kovalev's arguments have since been strongly supported by Peter Golden.Footnote 9
Each of these three approaches, however, is not entirely sufficient: Pritsak's attempt to form a coherent chronology from highly suspect evidence is laudable for its appreciation of the difficulties that a ruling dynasty would have experienced pushing through conversion but it is untenable; Zuckerman's reconciling of the three ‘disputation accounts’ into one event is noteworthy, but it places too much emphasis on the reliability of the Vita Constantinii and does not account for the new numismatic evidence; finally, Kovalev's case based on the Khazar Jewish coins draws unsubstantiated conclusions from his evidence, fails to reconcile the new numismatic evidence with the full corpus of textual sources, and ignores the lessons that might be drawn from comparative examples of steppe conversion. In the most general of terms then, previous treatments of Khazar conversion have been content to establish a date for conversion based upon a favoured selection of the extant sources, and little attempt has been made either to explore the events surrounding the Khazar conversion, or to see how a particular scenario fits in with the remaining evidence.
It would be easy, however, to erect a straw man by promising a radical reinterpretation of the evidence for the Khazar conversion, but this is simply not possible; a reasonably limited body of evidence coupled with extremely high interest has meant that almost every plausible – and implausible – argument and counter-argument that can be drawn from the sources has already been drawn. What this investigation intends instead is modification and re-elaboration as opposed to wholesale revolution. A re-examination of the evidence for the Khazar conversion combined with the implementation of a source-critical methodology that engages – if only to dismiss – both the textual and documentary sources, will yield new answers for the ‘when’ and ‘how’ of Khazar conversion. In a more general vein, it is hoped that the conclusions put forward here will not just be of interest to Khazarologists, but will prove useful for those interested in the broader areas of steppe historiography and conversion.
The Written Sources for the Khazar Conversion
The evidence for the Khazar conversion forms a veritable kaleidoscope, with slight variations in assemblage resulting in significant distortion in the pattern that emerges. Given that the present study aims first to deconstruct and then to reassemble the various sources in order to build a new scenario for the Khazar conversion from the ground up, we must begin with a brief overview and critique of the available evidence.Footnote 10
The Vita Constantinii Footnote 11 : The Vita Constantinii is the first of our ‘disputation accounts’. The narrative does not directly capture the Khazar conversion but it does provide a great deal of highly useful information on the religious and political practices of the Khazars in the mid ninth century. It offers a hagiographic account of the life and activities of the Byzantine missionary Constantine (d. 869), and was probably composed shortly after his death.
The Vita recounts how envoys from Khazaria arrived at the court of the Byzantine emperor Michael III and announced themselves ready to convert from paganism to monotheism. The envoys informed the Byzantine authorities that representatives from the Jewish and Muslim faiths were already making the case for their own religions, and asked that a learned scholar be dispatched to the Khazar court to make the case for Christianity. Due to the ‘love and friendship’ held between Khazaria and the Byzantium, the Khazars would wait until the Byzantine envoy had arrived and a disputation between the representatives of the three religions had taken place, before converting to the religion which had made the most convincing arguments.
Constantine volunteered to act as a religious ambassador for Byzantium, and travelled to the city of Cherson on the Crimean Peninsula to prepare for the disputation by studying Hebrew and debating with the locals. He also had time to deter both a Khazar and a Hungarian attack on Byzantine cities in the region. External evidence places Constantine in Cherson in early 861 and thus the subsequent journey to the Khazar court must have been made later in that year.Footnote 12
As expected in a hagiographical work, Constantine successfully made the case for the superiority of Christianity in the disputation. The relevant part of the narrative ends with the baptism of 200 Khazars, the recovery of 200 Byzantine prisoners-of-war, and a vague promise by the Khazar Khagan that he would soon convert to Christianity and that all of the Jews and Muslims in Khazaria would be put to death.
Zuckerman has argued that the Vita Constantinii is a ‘first-rate historical document’, in which the only blatant factual distortion in the material on the mission to the Khazars was the claim that the Khagan would convert and that the Jews and Muslims would be killed. This allowed him to conclude that, since the rest of the Vita is supposedly trustworthy, conversion to any new religion could not have occurred until after the disputation had taken place; that is sometime after 861.Footnote 13 There are, however, a number of problems presented both by the account itself and Zuckerman's interpretation of it. Indeed, Pritsak strongly argued that the material on Khazaria could only be a later interpolation based on sources dating from long before the latter ninth century. The Vita, he claimed, presents an anachronistic picture of Khaganal supremacy – in which he is solely responsible for the negotiations with Constantine – which was decades out of date. Pritsak's reasoning for this conclusion was that ‘It is hardly possible that Methodius, an experienced administrator, would have written or edited the chapters of the Vita Constantinii based on unrealistic or inaccurate data’.Footnote 14
While Zuckerman and Pritsak arrived at very different conclusions about the utility of the Vita as a source for the Khazar conversion, they were both far too uncritical in their approach to it. It is not intended as an ‘historical document’ but a hagiographic one. The composers of the Vita claim that their account was based on Constantine's theological writings which had been compiled into ‘eight discourses’ by his brother, Methodius.Footnote 15 This point is emphasised by the fact that Constantine somewhat conveniently refutes in turn the arguments of the iconoclasts, Muslims, Sarmatians and Jews. Indeed, despite the fact that we are informed that there was a Muslim delegation present at the Khagan's court he does not engage them in disputation – presumably this is because he had already outlined the argument to be used against the ‘Hagarites’ during an earlier set-piece debate. Rather than serving as a factual account of Constantine's life, the function of the narrative in the Vita Constantinii was instead two-fold: first, to give Constantine an appropriately interesting setting in which to present his theological arguments; and second, to present him in as ‘holy’ a way as possible in order to cement his legitimacy as the founding father of the Moravian Church.Footnote 16 This is not to say, however, that the description of the saint's adventures in Khazaria is entirely fictitious; yet it is not the ‘first-rate historical document’ which has been postulated.
The Reply of Joseph to Hasdai ibn Shaprut Footnote 17 : This letter purports to be from Joseph, most probably the reigning Khazar king, and was written in the 950s. It was composed in response to the inquiries of the Jewish first minister of the Umayyad Caliphate in Spain, Hasdai ibn Shaprut, who, according to a number of extant texts and fragments from a number of other letters, made several attempts to gather information on the Khazars.Footnote 18 Assuming its veracity, it was almost certainly produced by a scribe employed in the Khazar bureaucracy and thus it represents a product of the Khazar ‘Establishment’.Footnote 19 The letter survives in both a ‘long’ and a ‘short’ version of which it is unclear which is the more authentic. However, the same general conversion narrative is present in both versions.
The Reply indicates that Joseph was the descendant of a pious king named Bulan, who came to power some time after the foundation of the Khazar state. Bulan was repeatedly visited in his dreams by an angel who implored him to accept the Hebrew Laws, but Bulan argued that he needed the consent of the ‘Big Chief’ (the Khagan) before conversion could take place. The angel subsequently appeared in the dreams of the Big Chief and the conversion of Bulan, the Khagan, and all of the Khazars took place the next morning.
The next few years under the new religion were marked by great victories and the accumulation of enormous wealth. However, this provoked the wrath of both the King of the Arabs and the King of Byzantium, who dispatched missionaries to contest the Khazars’ conversion to Judaism. Bulan summoned a Jewish sage and a disputation was arranged so that the merits of the various religions could be debated. The disputation itself appears to have been stage-managed by Bulan. Thus, for example, he sounded out the missionaries in private before manipulating them into admitting the superiority of the Jewish faith in public. Following this victory in the disputation, Judaism was proclaimed with renewed vigour and certain customs – such as circumcision – were enforced. The narrative continues to explain how generations later the Jewish foundations of the Khazar state were strengthened by a certain King Obadiyah.
There are a number of problems with this account. First, Zuckerman has made a compelling case for at least one element of the narrative – the pious King Obadiyah – as marking a later and entirely fabricated interpolation, and this, of course, calls the veracity of the entire source into question.Footnote 20 Second, Devin DeWeese has argued strongly that ‘We will be led into absurdities if we hunt for a “historical” core’ in conversion narratives, and instead we should concentrate our energies on investigating what these accounts tell us about the communities in which they were formed’. For DeWeese, certain elements of the Khazar narrative bear clear signs of symbolic, rather than historical, resonance, and they were introduced into the narrative in order that the converted community might better comprehend the process that had brought them into existence.Footnote 21 Third, of related concern is the structural similarity of this conversion narrative with others, for example the conversion of the Uighur chief, Buqu Khan, to Buddhism in the account of Juwaynī.Footnote 22 It is unclear whether the clear parallels between these accounts are the result of coincidence, direct imitation, or, perhaps, both of the sources drawing on a common narrative prototype for explaining religious conversion. Finally, and perhaps most obviously, the account contains a number of clearly implausible elements, chief among them the angel who brought religion to the King and ‘Big Chief’.
The Schechter Document Footnote 23 : Dated to ca. 949 this letter records a version of the Khazar conversion from the perspective of Khazar Jews, probably tracing their Jewish ancestry back to before the conversion of the Khazar elite.Footnote 24 Similarly to the Reply of Joseph, it seems that this account was composed in response to the wide-ranging inquiries of Hasdai ibn Shaprut. This document can be securely dated to the tenth century, as it has been shown to belong to the same codex as an original letter from Provence which was clearly intended for Hasdai.Footnote 25
The narrative describes how Jews came to Khazaria fleeing persecution and settled there, partially assimilating with the inhabitants. Sometime later a semi-Judaised war-leader named Bulan was granted the office of išad/beg and, influenced by his wife's family, adopted a stricter and more observant form of Judaism. The rise of a Jew to such a prominent position angered the kings of Byzantium and Arabia and they sent scornful messages mocking Israel. It seems that this ignited hitherto suppressed disquiet amongst the Khazar aristocracy and thus a disputation was arranged between representatives of Judaism, Christianity, and Islam.
At the court disputation, the victory of Judaism was ensured when, at the crucial moment, a faction of the Khazar ‘officers’ produced the necessary sacred texts which had been kept in a certain cave. The narrative informs us that after these texts had been explained the people of Khazaria returned to Judaism completely. Bulan was appointed king and renamed Sabriel and the Khagan became a Jewish judge.
Despite offering what might be considered a more plausible account of the Khazar conversion, DeWeese has argued that a number of the elements in the narrative clearly tell us more about the circumstances in which the account was composed rather than the time of their conversion. Chief among these ‘literary’ rather than ‘historical’ elements is the notion of the Khazars having ‘returned’ to Judaism combined with the motif of the religious books in the cave. This was designed to boost the prestige of the Khazar conversion by stressing the notion that rather than a new event, the conversion was actually a return to an heroic Jewish past.Footnote 26
Kitāb al-Khazari Footnote 27 : The Book of the Khazars was composed ca. 1140 by the Spanish Rabbi Judah Halevi. It should be noted from the outset then, that the account is of extremely late provenance. Moreover, Halevi's work is perhaps the most openly distorted of our conversion traditions. The secondary title of the treatise is ‘Book of Argument and Demonstration in Aid of the Despised Faith’. Halevi wanted to show the theological superiority of Judaism at a time of increasing hostility towards Jews from Christian and Muslim populations, and the conversion of the Khazars was explicitly adapted to fulfil this function. At the same time, however, reference to some of the features found in both the Reply of Joseph and the Schechter Document indicates that Halevi had access to these sources, while he may also have had access to another text no longer extant.Footnote 28
Halevi's account begins with an angel visiting the Khazar king and exhorting him to change his ways. Thus the king summoned scholars from the Christian and Muslim faiths, but ignored Judaism because adherents of that religion ‘are of low station, few in number, and generally despised’.
However, finding himself unsatisfied with the cosmological arguments of Christianity and Islam the king decided to summon a representative of Judaism. After a long theological discussion with a Rabbi he pronounced himself pleased with the answers of the Jewish faith. After the disputation the Khazar ruler informed his vizier of his discussions and together they travelled to the mountains of Warsān where they found a Jewish community in a cave. Here they embraced the religion and were circumcised, before they travelled back to the Khazar court and gradually built up a faction in support of conversion to Judaism. When the numbers of their supporters had increased they induced the rest of the Khazar population to adopt the new faith.
Islamic Sources: There are a number of references to the conversion – or at least the Judaization – of Khazaria in the accounts of Muslim geographers and travellers from the beginning of the tenth century. The three accounts which provide the most detailed information on the conversion of the Khazars are:
1. An account by ‘Abd al-Jabbār Ibn Muḥammad al-Hamdānī, who was Chief Qadi in Rayy at the beginning of the eleventh century, describes how in the days of the ‘Abbāsids a man came to the ‘great king’ of the Khazars and managed to convert him. Following the conversion the Khazars took upon themselves ‘the difficult (obligations enjoined by) the law of the Torah’. According to Pines however, while the account may be based on an earlier source, it was clearly designed to fulfil the polemical function of presenting voluntary Jewish conversion as against coerced Christian conversion.Footnote 29
2. An account by the Spanish scholar Abū ‘Ubayd al-Bakrī (d. 1094) describes how the Khazar king at first converted to Christianity before having second thoughts about the decision. A disputation was arranged between a representative of Judaism and Christianity where the former appears to have been victorious. Still unconvinced however, the king invited a Muslim representative to join the disputation but the Jews arranged to have him poisoned en route. Without the Muslim intervention the arguments of the Jewish advocate were successful, and the Khazar ruler converted to Judaism.Footnote 30 It seems most likely that, based in Spain and thus with access to the ‘indigenous’ disputation accounts, al-Bakrī's narrative is an attempt to provide an acceptable explanation for the lack of Muslim success in the court disputation.
3. Al-Mas‘ūdī, writing in the mid tenth century, describes conversion as occurring during the reign of Hārūn al-Rashīd. Moreover al-Dīmashqī (writing ca. 1327), who may have had access to a no longer extant work of al-Mas‘ūdī, narrates how in the days of Hārūn the emperor caused the Jews to emigrate. They came to Khazaria, where they found an intelligent but untutored race, and offered Judaism to the inhabitants who subsequently converted. However, it is equally possible that the al-Dīmashqī was confusing the persecution of Jews and their emigration to Khazaria by the Emperor Romanos, as also described by al-Mas‘ūdī, with the earlier events of the conversion.Footnote 31
There are also a number of Arabic accounts which while not touching on the conversion process directly, do shed light on the political and religious structures in Khazaria and so are useful for our investigation. Islamic geographers and travellers from the early tenth century onwards began to report on Judaism within Khazaria. Many of these authors give slight variations on the characteristics of the Khazar diarchy and the spread of Judaism, but in general these sources agree that by the time that they were writing Judaism was firmly established among the Khazar elite, and that a Jewish Khazar king was the ultimate source of temporal authority in Khazaria.Footnote 32
It should be noted of these Muslim accounts in general, however, that they are of fairly late composition, are highly partisan, and are based on the reports of informants. Moreover, both the geographies and the travellers’ accounts were as much for entertainment as edification; their readers would have been attracted to tales of the unusual customs of northern barbarians, and it is likely that many aspects of the narratives were exaggerated.Footnote 33
Other Textual Sources: The most significant source in addition to main accounts and the Islamic reports is probably the commentary on the Gospel of Matthew by Christian of Stavelot. Here Christian reported that the Khazars ‘profess all dogmata of Judaism. However, the Bulgars, who are from those seven tribes, are now becoming baptised’. Khan Boris of the Bulgars first accepted baptism from Byzantine missionaries in 864 and so this notice seems to place the conversion of the Khazars around this date.Footnote 34
Another important source of information comes from three letters sent by Byzantine Patriarchs to ecclesiastics on the Crimean Peninsula.Footnote 35 The first of these was written by Patriarch Photios to a certain Antonios, Archbishop of Bosporos/Kerch, around 864. The letter does not directly mention Khazaria, but Antonios is praised for ‘having captivated the Jews who live there unto obedience to Christ’.Footnote 36 The other two letters were written by Patriarch Nicholas Mysticus to the Archbishop of Cherson, probably in the early 920s, and describe the Byzantine attempts to re-establish an ecclesiastical presence in Khazaria.Footnote 37
The Numismatic Evidence and its ProblemsFootnote 38
Of relatively recent identification, the ‘Moses’ coinage of the Khazars appears to have circumvented the gaps and contradictions in the textual record by offering undistorted, documentary evidence for the adoption of Judaism. In a flurry of innovative numismatic activity reminiscent of the Islamic coinage reforms of the Umayyad Caliph ‘Abd al-Malik, the Khazars issued three wholly new types of coin in 837/8: one type naming the mint as the ‘Land of the Khazars’ and providing an accurate date, another type containing a tribal or clan marking – the tamgha - the exact meaning of which is unclear, and the final type bearing the legend ‘Moses is the Messenger of God’. These types are all connected by die-chain and so almost certainly originated from the same mint and in the same year.
As we have noted above, Kovalev has argued that these coins could not have been minted without the consent of the Khagan, and it is unthinkable that he would have authorised the issue of Jewish coinage had he not been a convert himself. Thus, the coins point to the Judaisation of the Khazar elite by 837/8. If the minting of this new coinage signalled the fundamental transformation of the Khazar state then the question as to why these coins seem only to have been issued in a single year arises. Kovalev responded to this dilemma by arguing that this numismatic reform ‘was probably a trial run that quickly proved to be a futile venture’.Footnote 39
There seem to be a number of problems with the numismatic case, however, which should lead to some hesitancy in accepting Kovalev's conclusions wholesale. Before looking at the specifics of the argument there are two general points to make. First, it is clear that Kovalev approves of numismatic evidence because it is an ‘unedited’ source. Indeed, with the distorted, contradictory, and suspect textual evidence we have already discussed it is easy to see why unalterable documentary evidence is given pride of place. However, not only is documentary evidence limited in the conclusions it allows the historian to draw, but it has no more inherent authenticity than textual sources.Footnote 40 If a number of independent written accounts seem to suggest that the final stage of conversion did not take place until the 850s or 860s then what right do we have to completely disregard this evidence? Surely a credible scenario for the conversion of the Khazars must draw upon as much of the plausible evidence as possible.
Second, the ‘simple conversion’ offered by Kovalev – in which a complete transition is made from a Turkish Khaganate to a Jewish diarchy in a single year – is not the kind of scenario we find in comparable examples.
Conversion was often a difficult process, taking a number of years, and characterised by numerous setbacks. For example, the conversion of Boris I of Bulgaria led to an aristocratic revolt and the massacre of 52 boyar households, while despite elite and even royal conversion from at least the early ninth century, the ‘official’ conversion of the Rus’ did not take place until 988. Perhaps the most instructive example is to be found in the account of Ibn Faḍlān's journey to the Volga Bulgars in 922. He recounts events soon after his entry into the steppe world: ‘The first of the rulers and chiefs whom we met was the little Yanal. He had been a convert to Islam, but they [his people] told him: “If thou acceptest Islam, then thou canst not be our chieftain”’. It is quite possible then, that the conversion of the Khazars was a much more complex affair than envisaged by the scenario that Kovalev proposes.Footnote 41
This point is further bolstered when we consider that the ‘Moses’ coins were minted only in one year. Kovalev put forward three reasons why this might have been the case: first, that they failed to achieve circulation within Khazaria and thus did not achieve their ideological function; second, that the Rus’ traders could have been uncomfortable with these new coins; and third, that the coins possibly provoked the antagonism of the Islamic royal guard. We shall deal with each of these points in turn.
First, while much of the Khazar coinage was exported it seems implausible that with a centralised taxation system and enterprising merchants there was no monetary circulation within Khazaria.Footnote 42 Coins are normally found in hoards deposited in dangerous times and, as Kovalev has argued elsewhere, Khazaria was a remarkably peaceful and stable polity. This might go some way to explaining why none of the ‘Moses’ coins have turned up within former Khazar territories.Footnote 43 Moreover, if we are to accept that all of the coins were exported then the question arises as to why the Khazars bothered to design three new coins in 837/8 at all. Past experience would have taught the authorities that most of their coinage ended up being exported to the Rus’ and it seems nonsensical that with this in mind they would have introduced ideological reforms on their coinage.
Even in the unlikely event that the Khazars’ new coinage achieved no – or extremely limited – internal circulation this was not a sufficient reason for ending their numismatic experiment. It seems most probable that a group must have objected to the coins. Kovalev first argues that this group might have been the Rus’ who would have been reluctant to accept this modified coinage, but this seems unlikely. There is no indication that these coins were of inferior quality to other Khazar issues, or that the Rus’ would have objected to the new religious content of these coins, and, indeed, the fact that extant examples of the Moses coinage have been found as distant from Khazaria as Gotland, suggests that they were quite possibly traded among Rus’ merchants.Footnote 44 It might be argued that such numismatic discoveries as the ‘Offa dinar’ indicate that Islamic coinage had a great deal of prestige throughout northern Europe, and thus that modification would have been looked upon unfavourably. However, the fact that on this same coin there are a number of mistakes in the script and that the title ‘OFFA REX’ is upside-down in relation to the Arabic suggests that non-Muslims had little idea of what the inscriptions said. It is hard to believe then, that the Rus’ would have noticed or cared that the name of one foreign prophet had been exchanged for another.Footnote 45
The other group Kovalev identifies as possibly objecting to the coins are the royal Muslim guard. According to al-Mas‘ūdī these were Muslim immigrants from Khwārazm who served in the royal army in return for religious freedom and possession of the vizierate. The paucity of our sources means that it is not possible to definitively conclude that these immigrants were not yet present in the early ninth century, but it seems unlikely. The Islamic guard certainly seem to have been established by 913, when they were apparently responsible for a massacre of Rus’ raiders returning from Muslim lands, but apart from a few mentions of a Muslim ‘faction’ in our disputation accounts – which was probably no more than a narrative convention – there is no evidence for a Muslim royal guard in the early ninth century. Moreover, the question might be asked that if this Islamic guard had been so affronted by the release of Jewish coins, then why had they permitted the Judaisation of the Khazar state in the first place?Footnote 46
The forgotten element in Kovalev's scenario is the Khazars themselves. It is argued that the Khagan was both relegated to subordinate status and forced to convert to Judaism in 837/8, possibly as a result of the loss of prestige caused by military defeats throughout the 830s, but this is hardly incontestable. First, if this really was the transition to the fully-fledged Khazar diarchy as described by the Muslim sources, then the Khagan could not have had any input into the minting of coins. Indeed, he would have had no temporal input whatsoever and so his conversion would not have been a necessary prerequisite of the ‘Moses’ coinage.Footnote 47 Second, it seems implausible that the Khazars – many of whom were still adherents of the Turkic Tengri cult even in the tenth century – would have immediately acquiesced to the subjugation of the traditional Khaganal power structures under a new Jewish leadership.Footnote 48 Finally, conclusions drawn from the written sources would suggest that the ‘official’ Judaisation of the Khazar state took place shortly before the time of Constantine's disputation.Footnote 49 It will be argued below then that the numismatic innovations of 837/8, rather than marking the final conversion of Khazaria, were instead the beginnings of a prolonged and difficult process.
Methodological Approach to the Khazar Conversion
It might seem odd that for an investigation which aims at a ‘positivist’ reconstruction of the Khazar conversion, so much time has been spent in highlighting the manifold difficulties presented by every available source of evidence. Even those sources which present little in the way of obvious difficulties – such as the letters from Byzantine patriarchs - are only of limited value in the sense that they shed only the dimmest light on the conversion process. The more informative the source, the less trust can be placed in its historicity. However, it is by adopting an uncritical approach to a particular piece of evidence – whether the Kuzari or the Vita Constantinii or the Khazar coinage – that previous treatments have failed to satisfactorily explain the conversion of the Khazars.
What we intend to do here then, is to reject all of the evidence that is not corroborated – whether in a specific or general sense – by an independent tradition or documentary evidence. Of course, in reality this methodology is not entirely feasible; it is almost impossible to know which of our sources are entirely independent of others, and the paucity of our evidence means that for some events we are left with only a single source. However, by breaking the process down into stages we can mobilise evidence which is thematically related – which corroborates a particular stage in the conversion chronology – even if it is not exact parallel. Moreover, our arguments can be strengthened through source-critical analyses and comparative examples. Despite the conjectural nature of the task at hand, it is hoped that the picture of the Khazar conversion which emerges from a careful assessment of all of the relevant evidence, will seem more plausible than the accounts which have gone before.
The Background to Conversion
The Emergence of Bulan
In the two ‘native’ accounts of conversion the protagonist was a semi-Judaised war-leader named Bulan who orchestrated the adoption of the new religion. In the Schechter Document, Bulan was a descendant of Jewish immigrants who had assimilated with native Khazars. He became chief of the Khazar army after achieving notable victories until, some time later, his heart was stirred to return to Judaism completely. Bulan was instructed in Jewish practices by his wife and her family and, later again, was responsible for the conversion of Khazaria.Footnote 50 Similar features are found in the Reply of Joseph. Here Bulan is again portrayed with pre-conversion monotheistic tendencies. He was wise and God-fearing and ‘He removed wizards and heathens from the land and relied on God alone’. At some point during his rule an angel appeared to him and asked him to convert. Knowing that alone he would be unable to persuade his people of the righteousness of the angel's message he requested that the angel appear before the Khagan. The next morning all of the Khazars converted – albeit it seems only partially – and Bulan subsequently won great victories, accumulated much wealth, and built the symbols of Judaism.Footnote 51
The common feature of the ‘pre-disputation’ section of the two narratives is a semi-Judaised Khazar whose political power was established and/or increased by successful military leadership. There is a significant difference, however, in the chronology of the events. The Schechter Document presents the relatively simple narrative of a semi-Judaised general increasing his power so that he was later in a position to establish Judaism, while the Reply of Joseph portrays a ruler with pronounced monotheistic tendencies who converts his Khagan and his people with the help of an angel, and subsequently achieves divinely-ordained success. However, despite the differences the two accounts are not irreconcilable and it would be impossible to definitively prove that they are independent reports. While the Reply of Joseph describes a ‘shallow’ conversion before the disputation – probably in order to emphasise the early conversion of the Khagan - and the Schechter Document does not mention the conversion of anyone other than Bulan until after the disputation, and while each of the sources contain elements absent in the other which nevertheless find external corroboration, it is difficult to argue that the very notion of Bulan was not a popular oral myth which was subsequently modified to achieve very different purposes by the respective composers of the two accounts.Footnote 52 Therefore we will have to investigate the two elements common to both accounts to ensure the plausibility of the narratives: first, the existence of a military commander wielding increasing political power over Khazaria; and, second, the semi-Judaised nature of this figure.
The Ascendancy of the Khazar Išad/beg
In terms of the other sources which give a background to the conversion of the Khazars, the only one which gives a reasonably similar scenario to the one found in the Schechter Document and the Reply of Joseph is al-Hamdānī's account. However, each of the three main Islamic conversion narratives must be rejected as viable sources for the conversion itself because, composed centuries after the events they describe and suffused with partisan agendas, they act as further distortions of already highly problematic earlier accounts. The other major textual sources are of little use for this early stage of the conversion process.
We do have, however, the fragmentary pieces of evidence used by a number of scholars in their discussion of the development of the Khazar diarchy in the 830s and 840s. Until the beginning of the ninth century our sources suggest that the Khagan was the sole executor of authority among the Khazars. In an eye witness report of Khazaria in the mid-780s, for example, it is the Khagan who is responsible for all negotiations with the renegade Georgian prince Narses and his retinue.Footnote 53 Moreover, it also appears that the Khagan was in command of the Khazar campaigns against Arab-held Derbend in 798–99.Footnote 54 In the late 830s, however, a letter arrived at the Byzantine court addressed to the emperor Theophilos (r. 829–842) from the Khagan and the pech (išad/beg) of the Khazars.Footnote 55
A particularly interesting notice is provided by the al-Masālik wa'l-mamālik (The Book of Itineraries and Kingdoms), the early geographical work by Ibn Khurradādhbih. Containing the supposedly eye-witness account of Sallām the Interpreter, Caliph al-Wāthiq's (r. 842–847) envoy to the north, we are presented with a series of diplomatic exchanges which precede Sallām's journey through the Caucasus and into the steppe beyond.Footnote 56 The final contact is with an individual who can be identified as either ‘Tarkhān, king of the Khazars’ or ‘the Tarkhān of the king of the Khazars’.Footnote 57 Tarkhān is an Altaic term denoting a military commander. While Minorsky suggested that this figure might have been a military governor of the Khazars based in the North Caucasus, in light of what follows it makes more sense to identify him with the išad/beg.Footnote 58 However, even if the Tarkhān is not to be identified with the išad/beg, he was still answerable to the ‘king’ and not to the Khagan. It should also be noted that al-Muqaddasī (wr. ca. 985) suggests that prior to Sallām's mission al-Wāthiq had already dispatched the famous astronomer al-Khwārazmi to the court of this Tarkhān.Footnote 59
At some point during the combined reigns of Theophilos and al-Wāthiq (829–847) then, the extant sources suggest that the Khagan lost his role as the sole wielder of political power in Khazaria. Thus, if we couple this evidence with the numismatic reforms of 838/9, a convincing case begins to emerge for the late 830s and early 840s witnessing the political ascendancy of the Khazar išad/beg.
The Plausibility of a Judaised Išad/beg
While the Reply of Joseph describes the transformation of Bulan's religious convictions from vague monotheism to Judaism, the Schechter Document is far more ambiguous on the issue: Bulan is initially described as a Jew but still needs to be instructed in the Jewish ‘way of life’ by his wife and her family. This situation is paralleled by the respective sources’ description of the transformation of the Khazars’ religious convictions. In the Reply of Joseph they are simply heathens who become Jewish, while in the Schechter Document Khazaria is confusingly described as being inhabited by both the descendants of Jewish-Khazar assimilation, who retained some Jewish practices and became ‘one people’ with the Khazar gentiles, and by distinct Jewish elements such as Bulan's wife and her family. It seems most likely that although there was some Jewish-Khazar intermingling, the ambiguity in the account stems more from the attempt on the part of its composer to reconcile a ‘new’ conversion to Judaism with an ‘ancient’ Jewish past. This was designed to give greater legitimacy to the Khazar conversion by presenting it as a ‘return’ to a natural state of affairs.Footnote 60
Thus while the description of Jewish immigration into Khazaria because of persecution by ‘idol-worshippers’ cannot be entirely discounted, the description of ‘one people’ forming though intermarriage and the assimilation of Jewish customs among the Khazars is highly exaggerated.Footnote 61 However, it is probable that while the Khazars had not been semi-Judaised in the way depicted in the narrative, significant Jewish communities did exist within Khazaria. It seems most likely that the išad/beg who gained increasing power in the 830s emerged from one of these Jewish communities.Footnote 62
The evidence for Jewish elements within pre-conversion Khazaria is not exactly abundant, but there is enough to tentatively suggest that Jewish communities were to be found around the Black Sea area – particularly the Crimean Peninsula – and in the Caucasus. In the first instance, we have a number of archaeological finds from the first century onwards which point to Jewish settlement in a number of Crimean towns. Thus, for example, Greek marble plaques inscribed with names such as ‘Yehudah’ and ‘Enoch’ have been found at a number of places – including Pantikapeum (Bosporos/Kerch) – dating from between the first and third centuries, while a number of accoutrements of Jewish worship have been found in the same region, dated approximately to the sixth century.Footnote 63 Despite a lack of trustworthy material evidence for Jewish settlement after the sixth century there is no reason to think that the Jewish presence disappeared.
Fortunately, a written source is able to supplement the material record by suggesting that Jewish settlement continued into the Khazar period. The Byzantine Chronicle of Theophanes Confessor, which describes events up until the year 813 and was completed soon after this date, describes Jewish settlement to the east of the Sea of Azov.Footnote 64 The information is provided for the year 679–680 in an extended account of the collapse of Magna Bulgaria and the subsequent migrations of Bulgar tribes. While we are not in a position to ascertain the source of Theophanes’ claim or its likely date of composition, it seems plausible to conclude that there was nothing incongruous in the early ninth century about the notion that Jews were living in the Black Sea region.
Furthermore, we learn from the Vita Constantinii that Saint Constantine was able to learn Hebrew and read Jewish scriptures in the Crimean city of Cherson in the early 860s.Footnote 65 Golden has argued that this was a sign of the spread of Khazar Judaisation following the conversion of the elite over two decades beforehand.Footnote 66 However, Golden seems to have the process the wrong way around. It was in fact the Black Sea Jewish communities, with their long-standing settlement in the region, which were responsible for the Khazar conversion. This hypothesis is bolstered by the letter of Photios to Archbishop Antonios of Bosporos/Kerch a few years after Constantine's mission in which the Patriarch congratulated Antonios on ‘having captivated the Jews who live there unto obedience to Christ’.Footnote 67 There is no hint in the letter that these Jews were recent immigrants, nor recent converts to Judaism. Moreover, it is difficult to believe that Antonios was given free reign to convert Jews in one of Khazaria's most important cities more than two decades after Judaism had been officially established as the state religion. What Photios seems to be referring to instead were the communities of Jews who had been settled on the Crimean Peninsula for hundreds of years.
In terms of the evidence for Jewish settlement in the Caucasus, there are a couple of indications. In the Kuzari, for example, it is claimed that the king and his vizier received instruction from Jews living in caves in the mountains of Warsān.Footnote 68 Golden has identified Warsān as Varač‘an, an area both within the Transcaucasian region of the Khazar Khaganate and traditionally associated with Jewish settlement.Footnote 69 The fact that Halevi precisely identified the area suggests that he may have had access to a no longer extant source. Moreover, garbled medieval Georgian traditions may suggest Jewish contact with Khazars. The Kartlis Cxovreba, a composite work which cannot be firmly dated, describes in anachronistic terms how a Jewish community was established in a part of Georgia which witnessed numerous contacts with Khazars.Footnote 70 Despite the paucity of the evidence then, it is reasonable to suggest that there may also have been Jewish communities living inside or near Khazaria's Transcaucasian frontier.
Finally, in more general terms it is clear from the hagiographic account of the martyrdom of Abo that there were other monotheistic communities living in Khazaria in the late eighth century. This eye-witness report describes the ‘many towns and villages in that northern land which by the grace of the Holy Ghost abide securely in the Christian faith’.Footnote 71 Considering that Abo was a refugee from Georgia he is almost certainly describing that part of southern Khazaria north of the Caucasus. While Abo's testimony is not corroborated elsewhere, this is simply down to the fact that we have so few sources for Khazar lands in this period, and it seems extremely unlikely that the report is invented. Khazaria had a long history of military, economic, and diplomatic activity in Christian Transcaucasia and it seems improbable that along such a fluid frontier Christians would not have settled in Khazar lands.Footnote 72 Despite the paucity of our evidence then, there is every reason to accept the Schechter Document's admittedly distorted intimation of Jewish communities within Khazaria.
Therefore, while it is impossible to prove the historical existence of Bulan, it seems entirely plausible that the description given in our conversion narratives, at least in terms of the general contours, is accurate. An išad/beg almost certainly did achieve great political power in the late 830s, and there is nothing in the evidence to suggest, a priori, that he could not have emerged from the Khazarian Jewish milieu.
The 830s
According to the Schechter Document, Bulan achieved power due to his military prowess:
There was (at the time) no king in the land of Qazaria; but rather whoever would achieve victories in war would they appoint over themselves as chief officer of the army. (Thus was it) until the Jews once went out with them to do battle as was their wont. On that day a certain Jew prevailed with his sword and put to flight the enemies come against Qazar. So the men of Qazar appointed him over them as chief officer of the army according to their ancient custom.Footnote 73
Since we have established that it is plausible that a Jewish general could have taken control of Khazaria in the 830s it is time to look more closely at the conversion narratives. It is described how Bulan was initially part of an army of Jewish soldiers – probably as their leader – but following military success he was elevated to the position of ‘chief officer’. Presumably this institution was already well-established but Bulan was able to use his unprecedented ascendency to establish a Jewish monarchy, which was confined to himself and his descendants. The Reply of Joseph also presents Bulan as victorious in war and the founder of a Jewish royal dynasty, while a number of tenth-century Islamic sources describe the Jewish Khazar kingship.Footnote 74 The question which has to be raised then, is what conditions resulted in the fortunes of Bulan waxing so high, and those of the Khagan so low, that a fundamental transformation of Khazaria's power structures could be achieved?
We find agreement among a number of Khazarologists that it was the migration of the Magyars into Khazar territories in the 830s which undermined the prestige of the Khaganal dynasty and allowed for an ambitious general to take over.Footnote 75 Similarly, historians of the early Hungarians, while being less interested in the institutions of the Khazar state, have also argued that the migration of the Magyars to the lands west of Khazaria began around the year 830 and resulted in raiding across the Don.Footnote 76 The debate over the migration of the Magyars is beyond the scope of this investigation, but Kristó, Kovalev and Zuckerman base their arguments on the first mention of the Magyars in the sources.Footnote 77 There are three pieces of evidence which point to a Hungarian invasion of the Pontic Steppes in the 830s.Footnote 78 First, according to Symeon Logothete, in 836 or 837 Byzantines fleeing from Bulgaria discovered Magyars north of the Danube. Second, an entry in the Annals of Bertin for 839 – but probably recording events a few years earlier – reports Rus’ ambassadors unable to return home via the Black Sea, because the route was blocked by ‘primitive tribes that were very fierce and savage’. Third, in 839 the Byzantine Emperor Theophilos received a request for help with the construction of the Khazar fortress at Sarkel, which was probably to serve as a bulwark against marauding Magyars. Moreover, we also have the report from the early tenth-century Muslim geographer Ibn Rusta, that ‘in the past’ the Khazars ‘built trenches around themselves in fear of the Magyars and others from the peoples neighbouring their lands’.Footnote 79 The rise of Bulan then, could conjecturally be explained by his ability to effectively confront these Magyar incursions.
Golden, however, has rejected this hypothesis. He has argued: ‘The dismantling of the western, Pontic zone of Khazaria is conjecture. . . [and] The Pečeneg invasion was far more unsettling than whatever earlier problems the Khazars may have had with the Hungarians. . . Nonetheless, we have no evidence that the Hungarians or Pečenegs were existential threats to Khazaria that would produce, in essence, a restructuring of the modes of governance’.Footnote 80 While it is true that we possess no direct evidence for Magyar conflict with Khazaria in this period, it seems a reasonable hypothesis. Moreover, there was a transformation of political structures in Khazaria in this period, and something must explain this.
It can be suggested that instability caused by the Magyar migration in the 830s was a contributory factor to this political transformation, but that there were perhaps two other forces at work. First, it can be tentatively hypothesised that Byzantium – whether imperially directed or on the initiative of Crimean governors – took advantage of the Magyar invasion to improve or secure its own position in the region. Mark Whittow has argued that the letter addressed to Theophilos should not perhaps be taken at face-value.Footnote 81 The letter asks for Byzantine help in building a fortress and, if we pair it with the relevant passage in the De Administrando Imperio, we are informed how ‘ships of war’ from the Byzantine navy were dispatched under a certain Petronas to help construct Sarkel.Footnote 82
Whittow has argued that this scenario is somewhat incongruous. The Khazars had, for instance, been responsible for constructing numerous and impressive fortifications in the Crimea since the eighth century, and, to take just one example, the citadel at Bosporos had walls around 2.5 meters thick.Footnote 83 Moreover, while Sarkel was an impressive construction, it was only one of at least 24 fortresses guarding the principal river routes into Khazaria in the north and the west, and was of ‘characteristic’ Khazar design.Footnote 84 Finally, James Montgomery has noted that in a similar request for a fortress from the Volga Bulgars to the ‘Abbāsid Caliph, the emphasis was on the Caliph underwriting Bulgar expenses rather than dispatching skilled engineers.Footnote 85 It seems highly likely then, that the technology to construct fortresses already existed among the steppe peoples.
Whittow has suggested that instead of a Khazar appeal for assistance, the Sarkel episode was a cover-story for a Byzantine military expedition to capture the fortress which had gone horribly wrong. Similarly to the spectacularly unsuccessful attempt to seize the Iberian city of Ardanoutzin, as also described in the De Administrando Imperio, a plausible story was concocted to save face.Footnote 86 Of course, it might simply be argued that Byzantium was offering aid to Khazaria following its difficulties throughout the 830s. However, this seems unlikely. As we shall see, Khazaria and Byzantium enjoyed hostile relations throughout the whole of the early ninth century, and if the region really was overrun by ‘primitive tribes that were very fierce and savage’ then it is improbable that Petronas could have ‘made some ovens and baked bricks in them’ and spent time creating ‘mortar out of tiny shells from the river’. Moreover, we are also informed by the De Administrando Imperio that soon afterwards Byzantine control over Cherson was tightened with the installation of a military governor and, presumably, reinforcement by more troops. It seems more plausible then, that the Byzantines launched an expedition to capture the already-established fortress at Sarkel in order to protect their Crimean and mainland possessions from Magyar – and Rus’ - raids. If the notice is fabricated then the dating of the expedition is uncertain but is most likely to have occurred in the late 830s or early 840s.Footnote 87
The final contributory factor for Khazar instability in the 830s is a neglected notice from a problematic Armenian source. The source describes how in the year 838–839 ‘a locust came from the direction of the Khazars, an extraordinary thing larger than a sparrow, and it devoured part of the land of Albania’.Footnote 88 Despite Golden's insistence that there were no ‘existential threats’ to the position of the Khagan in the 830s, if we add together a destabilising Hungarian invasion, Byzantine raids north of the Black Sea, and a devastating famine in the most productive Khazar lands (between the Don and the Volga north of the Caucasus) then a significant cumulative case begins to appear for catastrophe in the Khazar realms.Footnote 89 The state had lost its former prosperity and this had dire consequences for the position of the Khagan.
The Khagan's Loss of Prestige
Al-Mas‘ūdī provides an extended notice on the relationship of the Khagan with Khazaria:
When the land of the Khazars suffers from drought or a misfortune befalls their country or they face a war against other nations or some other matter comes upon them, the common folk and the people of distinction rush to the king of the Khazars and say to him “we have seen a bad omen in this Xāqān and in his reign, we foresee (no good) from him. Kill him or hand him over to us so that we may kill him”. Sometimes, he hands him over to them and they kill him. Sometimes, he takes (the task) of killing him upon himself. Sometimes, he shows him mercy, defends him (saying) he is free of offense, he justifies him (saying) there is no crime which he has committed. This is the customary practice of the Khazars at this time. I have no knowledge whether this was the case in olden times or if it is an innovation.Footnote 90
A similar account of the spiritual link between the prosperity of the Khazars and the health of the Khagan is found in a report by Ibn Faḍlān. Here it is claimed that Khagan is killed after he has reigned for forty years because, according to his subjects: ‘This man's reason has failed, and his mind has become confused’.Footnote 91 Weakness in the Khagan meant weakness in the Khaganate and vice versa. Faced with invasion and famine, it seems plausible to suggest that the anger of the Khazars fell on their Khagan in the 830s – perhaps leading to his deposition but almost most certainly to his marginalisation and the creation of a power vacuum – leaving Bulan with the opportunity to launch a coup d'état.
Of course, al-Mas‘ūdī was unsure whether this was an innovation or a long-standing tradition and thus we must be careful not to back-project this tenth-century account to the 830s. However, it has been convincingly argued that the Khazar Khagan's investiture ceremony reported by Islamic geographers in the tenth century was a direct continuation of Turkic royal rituals reported over three hundred years before in Chinese chronicles. Part of this ceremony involved choking the Khagan with a silk scarf until he was close to death and then asking him to name the length of his rule. This again suggests the existence of a symbiotic relationship between the life of the Khagan and the fortunes of Khazaria. The strangulation ritual gave the Khagan access to the spirit world so he could discover – and inform his courtiers of - the divinely-ordained length of his rule. Reigning beyond this period would bring disaster to the realm. Despite the slight contradiction between the sources on the length of the Khagan's rule their wider implication is clear: the Khagan was the ultimate source for Khazaria's fortunes – and misfortunes. It is impossible to establish if the notion of Khaganal culpability for Khazar catastrophes was as fully-developed in the ninth century as it would become in the tenth, but it seems likely that the prestige of the Khagan was so undermined in the 830s that a powerful military leader could make a bid for power.Footnote 92
Preliminary Conclusions on the Rise of Bulan
Now that this investigation has assessed the credibility of the picture of Bulan's early career that emerges from our sources we are in a position to present an overview of our initial conclusions. Much of what follows is conjectural, but it seems the most likely scenario based on the evidence that we possess.
In the 830s Bulan was a young Jewish military commander in command of a largely Jewish contingent, probably stationed at one of the many Khazar cities on the Crimean Peninsula.Footnote 93 At some point in the 830s the Magyars made their migration into the Pontic Steppes and began raids on the western Khazar territories. Bulan emerged as a successful general able to fend off these incursions. Indeed, he is most likely the ‘Tarkhān’ of Ibn Khurradādhbih's account. In the late 830s Byzantium may also have campaigned in Khazaria but, more significantly, a terrible famine struck the heart of the Khaganate, weakening the position of the Khagan still further. It is also probable that in Khazaria's multi-ethnic confederation, held together mainly by the charisma of the Khagan and self-interest, dissension broke out in the central territories too.Footnote 94 Taking advantage of his growing following and diminishing support for the Khagan, and perhaps frustrated by the paralysis at the heart of Khazaria, Bulan took the opportunity to establish himself as the de facto head of state in a coup d'état. This was marked by the minting of Jewish coinage in 837/8 – most likely 838 if we trust the exact dating of the famine of 838/9 in Movsēs Dasxuranci's chronicle - and was witnessed by Sallām the Interpreter before or in 844.Footnote 95
This scenario however, despite being more thoroughly elaborated than previous investigations, is no different in its essential contours than former findings.Footnote 96 Yet, it is the contention of this investigation that this was not the culmination of Jewish conversion in Khazaria but simply the start. Indeed, it is not clear by this point that any Khazar had converted to Judaism at all. As we shall see, it was not until the late 850s that the ‘official’ Judaisation of the Khazar state began.
The Judaisation of Khazaria
About the events of the years between ca. 840 and 861 in Khazaria virtually nothing is known. It seems probable that Bulan remained in power and spent the time consolidating his position. He is described in the Schechter Document and the Reply of Joseph as being the de facto ruler of the Khazars at the time of the disputation, and while this is far from infallible evidence it seems unlikely that he could have founded a dynasty of Jewish kings simply by becoming the chief military officer of Khazaria in the 830s.Footnote 97 Thus, if the fact that the Jewish coinage was only minted for a single year indicates that not all of Khazaria's inhabitants welcomed a Jewish ruler – as I have argued above – then Bulan had to find a way to legitimise and cement his authority. He eventually decided upon his coronation and the Judaisation of Khazaria.
The Fallacy of Conversion Only After 861
Zuckerman has argued that the conversion of the Khagan – and thus the ‘official’ conversion of Khazaria – did not take place until after the disputation of 861. His conclusions are based upon the letter from the Khagan to the Byzantine Emperor, supposedly preserved within the Vita Constantinii, which describes how, due to the ‘love and friendship’ between Khazaria and Byzantium, the Khagan requested the dispatch of a missionary to make the case for Christianity before deciding upon a religion to replace paganism.Footnote 98 There are, however, a number of reasons why we should reject this part of the Vita Constantinii.
First, and perhaps most obviously, we can return to the issue that the Vita is not a historical source. The disciples of Methodius who composed the account were looking for both an exciting and hagiographically-appropriate framing narrative for the theological disputation, and the self-sacrificing Constantine making the perilous journey to the court of a barbarian ruler in order to secure his conversion, served just this purpose. Thus, the notice cannot be accepted until it finds corroboration elsewhere.
Second, evidence derived from both the Vita Constantinii itself and more generally on Byzantine-Khazar relations in the early ninth century offers a very different picture to the one of ‘love and friendship’ in the report of the letter. For example, the Vita depicts Constantine deterring a Khazar attack on a ‘Christian city’ near Cherson.Footnote 99 While this may simply be part of the hagiographic effort to bolster the missionary's ‘saintly’ reputation by endowing him with near-miraculous powers, it is still noteworthy that the account's composers found no paradox between a journey to convert the Khazar Khagan and a Khazar general attacking a Christian city. The city saved by Constantine was almost certainly one of the Byzantine towns of the province of Gothia which had been, as the Life of John, Bishop of Doros, makes clear, a source of conflict between the two states since the late eighth century.Footnote 100 The Vita also adds that one of the major achievements of the mission was the release of 200 Byzantine prisoners-of-war. This is hardly suggestive of amicable relations between Khazaria and Byzantium. Moreover, as it has been argued above, it is entirely plausible that the Byzantines made a failed bid to capture Sarkel in the late 830s. We must dismiss as a total fabrication, therefore, the notion that the Khazars were patiently waiting for the arrival of their Byzantine allies before deciding upon their future religious affiliation.
Finally, another Byzantine source, the Life of Methodius, gives a very different reason for the Byzantine mission. Here it records that Constantine and Methodius were dispatched to Khazaria ‘because Jews were there who were blaspheming the Christian faith beyond measure’.Footnote 101 Similarly to the saint's earlier mission to the Caliphate then, Constantine was sent to debate against those who were already firm in their convictions, not with any realistic hope of capturing Khazaria for Christianity.Footnote 102
What had the Khazars done to be accused of gross blasphemy in the eyes of the Byzantine Church? The available evidence would suggest that Bulan had enacted the first stage of his three-part plan to achieve the ‘official’ conversion of Khazaria. First, a Jewish monarchy was established and the Khaganal dynasty was converted to Judaism. Second, the disputation of 861 was deliberately rigged to achieve a public success for the new king and his religion. Third, the position of Judaism within Khazaria was strengthened.
The First Stage of Judaisation: The Coronation of Bulan and the Conversion of the Khagan
It seems likely that the conversion of the Khaganal dynasty to Judaism took place soon before 861, and that it was accompanied by the establishment of the hereditary office of the Khazar kings. This chronology is, of course, directly contradicted by the Schechter Document, which describes how it was only after the disputation that the ‘men of the land’ appointed the Khagan as judge and Bulan as king.Footnote 103 However, while the Schechter Document in general is a trustworthy source, it seems to have condensed the complex events of the 850s and 860s into an easy-to-digest chronology. The main problem with its scenario is that it describes how the Byzantine Emperor and the ‘Abbāsid Caliph sent messengers with ‘words of scorn against Israel’, supposedly because Bulan had adopted a more orthodox form of Judaism.Footnote 104 However, it makes little sense that Byzantium or the Caliphate would know or care about the religious beliefs of the Khazar ‘chief officer of the army’.
The Reply of Joseph is equally unreliable on this point because it has a vested interest in associating the Khagan with Judaism from as early a time as possible.Footnote 105 However, here it is reported that both Bulan became king and the Khagan converted to Judaism before the disputation.Footnote 106 Kovalev, as we have noted, thinks that this happened in 838/9 but as we have already argued there is both no evidence for this, and the fact that the ‘Moses’ coins were minted only in one year suggests that Bulan's initially vigorous assertion of Judaism needed to be rethought, probably in light of internal opposition from ethnic Khazars. It may be that Judah Halevi was simply a more politically astute author than the composers of the other narratives, or he may have possessed a no longer extant source, but his description of the conversion process is the most plausible: ‘they [the king and his vizier] kept their conversion secret, however, until they found an opportunity of disclosing the fact gradually to a few of their special friends. When the number had increased, they made the affair public, and induced the rest of the Khazars to embrace the Jewish faith’.Footnote 107 It almost certainly took Bulan time to consolidate his hold over the Khazar state then, and it seems most plausible that it was the elevation of himself to the status of Jewish monarch – and the consequent conversion of the Khagan – which provoked the hostility of Byzantium and the ‘Abbāsid Caliphate shortly before 861, as reported in the Life of Methodius and the Schechter Document.
Tentative evidence for this process comes from the admittedly problematic Vita Constantinii. First, there is a hint that the notion of a Jewish monarchy was a topical subject in Khazaria in 861:
Having boarded a ship, Constantine set out for the land of the Khazars by way of the Meotis Sea and the Caspian Gates of the Caucasus Mountains. The Khazars sent a cunning and resourceful man to meet him, who entered into conversation with him and said to him: “Why do you follow the evil custom of replacing one emperor with another of a different lineage? We do this only according to lineage”.
The philosopher said to him: “Yet in place of Saul, who did nothing to please Him, God chose David, who was pleasing to Him, and David's lineage”.Footnote 108
Second, there are also strong suggestions that the Khagan and the Khazar court, as described by the Vita Constantinii, had already been Judaised. Thus, for example, when Constantine offered a toast in the name of the Trinity, the Khagan replied ‘We say the same but maintain the following difference: you glorify the Trinity, while we, having obtained Scriptures, One God’. Moreover, there were a number of Jews in the entourage of the Khagan.Footnote 109
It seems significant that the first debate Constantine had with an official representative of the Khazar government was over dynastic rule. Whether raised first by the missionary or the envoy, it was clearly a matter of some contention, and the fact that Constantine answered in terms of Old Testament models might indicate that the Khazar representative had first used the example of the Davidic monarchy to support his own arguments for succession within a lineage. Moreover, while Pritsak has argued that the Khazarian episode in the Vita is an eighth-century anachronism, there could hardly be a clearer indication of the Khagan's conversion than his response to Constantine's toast. The question which arises, of course, is whether we can trust these reports. It is impossible to conclusively prove their veracity, but the fact that the account's report of a letter to the Byzantines which portrays the Khazars as both pagans and allies is clearly a fabricated interpolation, and which stands in direct contrast to the Vita's later description of a Judaised Khagan, suggests that this latter part of the narrative was probably drawn from Constantine's own writings, or at least Methodius’ transmission of them.
Overall then, the balance of evidence would suggest that it was shortly before 861 that Bulan implemented the first step in a programme designed to establish his lineage as the Jewish kings of Khazaria. This involved both his coronation as monarch and the foundation of a hereditary dynasty and the transformation of the Khagan's office into that of a Jewish judge.Footnote 110
The Second Stage of Judaisation: The Disputation of 861
Whether Bulan conceived the disputation of 861 as part of a sophisticated strategy to bolster his own authority, or whether he simply capitalised on the arrival of the Byzantine missionary delegation, it appears that he used the debate to help cement the Judaisation of Khazaria. The disputation of 861 was stage-managed by the Khazar court, and the inevitable triumph of Judaism was used to bolster the prestige of the new Jewish political structures created by Bulan.
Evidence for Bulan's manipulation of proceedings abound. In the Reply of Joseph it is described how, because of the king's fame, envoys arrived from the lands of Byzantium and the Caliphate, each accompanied by wise men to invite him to their respective religion. A Jewish sage was summoned and the three ‘refuted the arguments of the other, without agreeing on one point’. At this juncture the king suspended proceedings and spent the next two days sounding out the opinions of the Christian and Muslim envoys on Judaism. On the third day, he resumed the disputation in the presence of ‘all his barons and his people’ and asked the Christian and Muslim representatives to choose between the religion of their rival and Judaism. They both declared that Judaism was the superior religion. Then, the King declared: ‘You have now confessed with your own mouth that the Israelitish religion is the best and the truest. I have chosen it, as it is the religion of Abraham’.Footnote 111
It might be expected that we should immediately dismiss this evidence. The King's sounding-out of the representatives, while a slightly ignoble activity that seems out of place in a conversion narrative, serves the narrative function of allowing the Christian and Muslim envoys to explain, in detail, why Judaism was the superior religion. This is precisely what was required in an ‘official’ conversion narrative, as it reinforced the prestige and legitimacy of the tenth-century Jewish elite.
However, a similar impression of events is given in the Schechter Document. Here it is described how the ‘words of scorn’ from the Byzantine and Arab envoys ‘influenced the hearts of the officers adversely’. In response Bulan convened a disputation of Muslims, Christians, and Jews. In this scenario the Muslims and Christians accepted the veracity of Jewish claims ‘but there also occurred a disagreement between them’. At this point, the ‘officers of Khazaria’ intervened, and had the ‘books of the Torah of Moses’ brought to the disputation. After the Jewish sages explained the books the people of Khazaria embraced Judaism.Footnote 112
Again it might be argued that the agenda in this narrative is too pronounced for the evidence to be considered authentic. The fact that the Jewish position triumphs after the Torah of Moses is brought from a certain cave to the disputation may well be designed to bolster the case for conversion not being a radical departure from paganism, but a ‘return’ to Khazaria's lost Jewish past.Footnote 113 However, could there not be a kernel of historical truth, present in both the Reply of Joseph and the Schechter Document, in the idea that the central authorities ensured the victory of Judaism? After all it is probable that a few years before the disputation a Jewish kingship had been established and the Khaganal dynasty converted. It seems beyond the boundaries of possibility that a Christian missionary could have received a fair hearing in this environment.
In the Islamic world where disputations at a ruler's majlis were popular and formalised, it was possible for some writers to talk of ‘terror sessions’ where a disputant was summoned to a debate that had been rigged against him from the outset. A subcategory of these ‘terror sessions’ were the ‘sessions of high officials’ and, according to one Islamic scholar, ‘One should also beware of discussing theology in the “sessions of high officials”, where the participants seek to be diverted by what they hear, and do not grant the two opponents equal sympathy or equal opportunity to be heard. Participation in such a session is either folly and subjection to humiliation or an incentive to anger and a source of depression’.Footnote 114
Of course, we are not in a position to prove that Constantine was subjected to the Khazar equivalent of a ‘terror session’. The nature of our ‘disputation accounts’ would naturally omit any hint of coercion. Constantine, as a saintly missionary, was under an obligation to describe his encounter as a success, while the two Khazar accounts, despite allowing for Bulan's superior intellect to triumph over the representatives of Christianity and Islam, were hardly going to admit to the envoys enjoying anything other than courteous treatment.Footnote 115
Yet, while the Khazar sources are not forthcoming, tactics of intimidation and humiliation were common strategies employed by the people of the steppe against sedentary envoys. In a couple of highly stimulating articles James Montgomery has explored the idea that some of the disturbing practices recorded in Ibn Faḍlān's account of the Caliphal embassy to the Volga Bulgars in 922, were deliberately designed to scare him. Thus, for example, after making excuses for the absence of 4,000 dinars promised to the Bulgar ruler, Ibn Faḍlān was told the story of a man from Sind who was executed by the Bulgars for being too clever.Footnote 116 Montgomery has also suggested that Ibn Faḍlān's famous description of a Viking burial, which contains a scene in which a slave girl is gang-raped and ritually murdered, might have been the product of Rus’ psychological warfare against the Muslim delegation, possibly because ‘they may not have taken too kindly to the prospect of a Muslim satellite kingdom in Volga Bulgaria’.Footnote 117 We may also note that some centuries later threats and humiliations were often inflicted by the Mongols on participants in religious disputations. Indeed, Peter Jackson has remarked that ‘. . . we cannot discount the possibility that one purpose [of the disputations] was entertainment – that the public religious disputation, in other words, was the intellectual counterpart of the bloody gladiatorial conflicts the Mongols staged between captured enemy soldiers’.Footnote 118
Accepting that, at the very least, the disputation could have been rigged against Constantine might also help to make sense of some ambiguities within the Vita Constantinii. For example, it presents a virtually unrecognisable picture of power relations in Khazaria to the extent that Pritsak thought it was actually based upon an eighth-century source. Thus, the Khagan is presented as the sole arbiter of authority while Bulan, if he is indeed the figure of ‘first counsellor’, is made to attack the Jewish position.Footnote 119 However, there is one passage which might shed light on this problem:
And pointing his finger at the Khagan and his first counsellor, the Philosopher said: “If someone says: ‘The first counsellor cannot entertain the Khagan’, but furthermore says: ‘The latter's lowest slave can entertain the Khagan and render honour unto him’, what are we to call him, tell me, insane or sensible?”Footnote 120
Constantine simply used the first counsellor's relationship with the Khagan as a metaphor to shed greater light on a theological problem, but his choice specifically to utilise the first counsellor in the example – and implicitly compare him to the lowest of slaves – may perhaps suggest that the saint was not on the best of terms with him. It is possible that out of resentment at his unfair treatment in the disputation as rigged by Bulan, he decided to relegate the Jewish king to an inferior position behind the Khagan and, furthermore, to maliciously portray him denouncing Judaism.
Constantine's arguments at the disputation of 861, then, were doomed to failure from the outset. Held in front of the Khazar elite, they were most likely used as an opportunity to publicly undermine Christianity – and possibly Islam - and display the superiority of Khazaria's new religion. The very fact that the disputation scene is a central part of both of the Khazar conversion narratives highlights the ideological advantages for the new Jewish regime which could be derived from such a public triumph. The doubts expressed by the ‘Khazar officers’, as described in the Schechter Document, could be put to rest in an emphatic validation of Khazaria's Judaisation.
The Third Stage of Judaisation: Reinforcing the Position of Judaism
After the establishment of the Jewish kingship, and its ideological reinforcement following the disputation, it is probable that Bulan moved on to the final stage of Khazar Judaisation. This final stage is, however, extremely conjectural as, with the exception of the Reply of Joseph and the Schechter Document, we have absolutely no direct evidence for affairs in Khazaria from 861 to the early tenth century.
According to the Reply of Joseph, following the disputation the Khazars were circumcised and began to observe the laws of the Torah. Later, a certain Obadiah ‘invigorated the kingdom and established religion. He built temples and house of study, gathered many learned Israelites, and gave them much silver and gold. They interpreted for him the Bible, the Mishnah and Talmud and the whole prayer book’.Footnote 121 In the Schechter Document it is described how ‘the Jews began to come from Baghdad and from Khorasan, and from the land of Greece, and they strengthened the men of the land, so that (the latter) held fast to the covenant of the “Father of a Multitude”’.Footnote 122
This is exceedingly fanciful. As Zuckerman has pointed out, Obadiah is a later interpolation and while perhaps another king undertook the reforms credited to him, it seems more likely that this description was added by the Spanish Jews who transmitted the Schechter Document and who must have thought that the initial conversion of the Khazars was not ‘Jewish’ enough. Moreover, it is clear from tenth-century Islamic sources that not all of the Khazars became Jewish and, in fact, the Khaganate allowed its inhabitants to come under the jurisdiction of Christian, Islamic, Pagan and Jewish courts depending on their beliefs.Footnote 123
Nonetheless, a few sources of evidence might suggest increasing Judaisation after 861. First, we have Christian of Stavelot's notice which indicates that around 864 he had got wind of religious change in Khazaria. This could refer to the creation of the Jewish monarchy and the conversion of the Khagan before 861, or it might equally refer to a more emphatic assertion of Judaism following the disputation. Second, in an admittedly ambiguous report on the Khazars, probably dating to the early years of the tenth century, Ibn al-Faqīh states that ‘all of the Khazars are Jews. But, they have been Judaised recently’.Footnote 124 His report could possibly reflect an earlier one that captured the ‘official’ conversion of the Khaganate. Third, and with slightly more credibility, while in ca. 864 Archbishop Antonios was converting Jews in the Khazar city of Bosporos/Kerch, in the early 920s Patriarch Nicholas Mysticus was corresponding with the Bishop of Cherson in an effort to extend Byzantium's ecclesiastical reach into Khazaria.Footnote 125 It is clear that the Byzantine Church's position in Khazaria had been lost in the intervening years and when we combine this with Nicholas’ hope that the salvation of the Khazars ‘may finally be re-established’, it is clear that by the early tenth century the Byzantine ecclesiastical presence north of the Black Sea had contracted to an isolated outpost, and was not operating in Khazar cities as it had in the time of Antonios.Footnote 126
While it is highly unlikely that Khazaria ever attempted a systematic programme of Judaisation - except perhaps at the highest levels of the state - it would make a great deal of sense for the establishment of a Jewish monarchy to have been accompanied by the curtailing of external religious influence. Bulan would not have minded that some of his subjects were Christian, but he would have balked at them offering prayers for the health of the Byzantine emperor.Footnote 127 Similarly, official Islamic missions from the ‘Abbāsid Caliphate seemed to involve recognition, on the part of the convert, of subordination to the Caliph in Baghdad.Footnote 128 Accepting the political or religious claims of foreign rulers over portions of Khazaria's inhabitants would have undermined Bulan's attempts to create a strong Jewish monarchy.
On the balance of the available evidence then, it would appear that shortly after 864 Bulan had taken the project to Judaise Khazaria as far as he intended. A Jewish monarchy had been established, unconvinced aristocrats had been brought onside by the public demonstration of Judaism's theological superiority, and the destabilising influence of Byzantium and the Caliphate's religious activities had been restricted. It was only then that the long and difficult process which had begun in the 830s with a coup d'état, continued in the 850s with a coronation, and was completed – at least in this incarnation – in the 860s with the official conversion of Khazaria, came to an end.
Final Reflections on the Khazar Conversion: The Utility of the Khagan
Much of what has been argued above is simply a modification of what has come before, an attempt to fill in the blanks and to lend slightly more substance to the conjectural arguments which, due to the nature of our source material, are inevitable in Khazar historiography. Through a comprehensive analysis of the evidence that we have described – and, it is hoped, proved plausible – the seizure of power by Bulan in the 830s due to the loss of faith in the divine-favour of the Khagan, his consolidation of power through the 840s and 850s until his coronation as a Jewish monarch and the conversion of the Khagan, and, finally, the bolstering of this new office though the victory of Judaism at the disputation of 861 and the subsequent curtailing of foreign religious influence. This scenario seems possible both in terms of extant evidence and in light of the difficulties faced in other cases of early medieval steppe conversion. When we speak of the conversion of the Khazars, however, we should remember that it was not a ‘mass conversion’ as described for other conversions, but a Judaisation of the Khazar elite designed to establish and reinforce a new Jewish leadership.
With this consideration in mind, we can pose one of the questions raised by the scenario proposed above: if Bulan was responsible for a coup d'état that eventually led to the establishment of a Jewish monarchy, why did he and his successors retain the institution of the Khaganate?
In simple terms it seems that retention of the Khaganate was both necessary and convenient. Despite the nominal conversion of the Khagan to Judaism, it is clear from the tenth-century Muslim accounts that the Khaganal office preserved, and had probably developed to an even greater extent, its traditional Turkic sacral and ritual elements. The Khagan was completely sacralised and tabooised, and he preserved an intimate spiritual bond with the fortunes of Khazaria. In a state whose inhabitants were still mostly adherents to the Tengri cult, the Jewish monarchy had no choice but to keep this unifying institution.Footnote 129 Thus, according to al- Mas'ūdī, even though the Khagan possessed no temporal authority, ‘the sovereignty of the Khazars is not secured by their King save by the Khagan being with him in the capital and his castle’.Footnote 130 This is also why the Reply of Joseph recounts how the Khagan was also visited by the angel that had come to Bulan and was thus converted very soon afterwards; the Khaganal office needed to be tied to the ‘Jewish project’ as tightly as possible, because otherwise the Khagan could offer a dangerous source of autonomous – traditional Turkic – political and religious authority. By accepting Judaism however, the Khagans acknowledged the political supremacy of the Jewish monarchy.
This arrangement, however, was not inconvenient for the Jewish kings. Bulan's successors suffered a variety of military defeats and rebellions, yet by preserving the Khaganate they could insulate the Jewish monarchy from loss of prestige and authority. Decline in Khazaria's fortunes could be blamed squarely on the shoulders of the Khagan, and discontented elements could be appeased through his deposition or execution. This political model would have protected the Jewish monarchs from the very set of circumstances which had allowed them to usurp the Khagan's authority in the first place.
