APPENDIX III
THE "KHAZAR
CORRESPONDENCE"
1
THE exchange of letters between the Spanish statesman Hasdai ibn Shaprut and King Joseph of Khazaria has for a long time fascinated historians. It is true that, as Dunlop wrote, "the importance of the Khazar Correspondence can be exaggerated. By this time it is possible to reconstruct Khazar history in some detail without recourse to the letters of Hasdai and Joseph."1 Nevertheless, the reader may be interested in a brief outline of what is known of the history of these documents. .Hasdai's Letter was apparently written between 954 and 961, for the embassy from Eastern Europe that he mentions (Chapter III,3-4) is believed to have visited Cordoba in 954, and Caliph Abd-al-Rahman, whom he mentions as his sovereign, ruled till 961. That the Letter was actually penned by Hasdai's secretary, Menahem ben-Sharuk - whose name appears in the acrostic after Hasdai's - has been established by Landau,2 through comparison with Menahem's other surviving work. Thus the authenticity of Hasdai's Letter is no longer in dispute, while the evidence concerning Joseph's Reply is necessarily more indirect and complex. .The earliest known mentions of the Correspondence date from the eleventh and twelfth centuries. Around the year 1100 Rabbi Jehudah ben Barzillai of Barcelona wrote in Hebrew his "Book of the Festivals" - Sefer ha-Ittim - which contains a long reference, including direct quotations, to Joseph's Reply to Hasdai. The passage in question in Barzillai's work starts as follows:
Barzillai goes on to quote or paraphrase further passages from Joseph's Reply, thus leaving no doubt that the Reply was already in existence as early as AD 1100. A particularly convincing touch is added by the Rabbi's scholarly scepticism. Living in provincial Barcelona, he evidently knew little or nothing about the Khazars. .About the time when Rabbi Barzillai wrote, the Arab chronicler, Ibn Hawkal, also heard some rumours about Hasdai's involvement with the Khazars. There survives an enigmatic note, which Ibn Hawkal jotted down on a manuscript map, dated AH 479 - AD 1086. It says:
It seems most unlikely that Hasdai actually visited Khazaria; but we remember that he offered to do so in his Letter, and that Joseph enthusiastically welcomed the prospect in the Reply; perhaps the industrious Hawkal heard some gossip about the Correspondence and extrapolated from there, a practice not unfamiliar among the chroniclers of the time. .Some fifty years later (AD 1140) Jehudah Halevi wrote his philosophical tract "The Khazars" (Kuzri). As already said, it contains little factual information, but his account of the Khazar conversion to Judaism agrees in broad outlines with that given by Joseph in the Reply. Halevi does not explicitly refer to the Correspondence, but his book is mainly concerned with theology, disregarding any historical or factual references. He had probably read a transcript of the Correspondence as the less erudite Barzillai had before him, but the evidence is inconclusive. .It is entirely conclusive, however, in the case of Abraham ben Daud (cf. above, II, 8) whose popular Sefer ha-Kabbalah, written in 1161, contains the following passage:
2
The first printed version of the Khazar Correspondence is contained in a Hebrew pamphlet, Kol Mebasser, "Voice of the Messenger of Good News".*[Two copies of the pamphlet belonging to two different editions are preserved in the Bodleian Library.] It was published in Constantinople in or around 1577 by Isaac Abraham Akrish. In his preface Akrish relates that during his travels in Egypt fifteen years earlier he had heard rumours of an independent Jewish kingdom (these rumours probably referred to the Falashas of Abyssinia); and that subsequently he obtained "a letter which was sent to the king of the Khazars, and the king's reply". He then decided to publish this correspondence in order to raise the spirits of his fellow Jews. Whether or not he thought that Khazaria still existed is not clear. At any rate the preface is followed by the text of the two letters, without further comment. .But the Correspondence did not remain buried in Akrish's obscure little pamphlet. Some sixty years after its publication, a copy of it was sent by a friend to Johannes Buxtorf the Younger, a Calvinist scholar of great erudition. Buxtorf was an expert Hebraist, who published a great amount of studies in biblical exegesis and rabbinical literature. When he read Akrish's pamphlet, he was at first as sceptical regarding the authenticity of the Correspondence as Rabbi Barzillai had been five hundred years before him. But in 1660 Buxtorf finally printed the text of both letters in Hebrew and in a Latin translation as an addendum to Jehudah Halevi's book on the Khazars. It was perhaps an obvious, but not a happy idea, for the inclusion, within the same covers, of Halevi's legendary tale hardly predisposed historians to take the Correspondence seriously. It was only in the nineteenth century that their attitude changed, when more became known, from independent sources, about the Khazars.
3
The only manuscript version which contains both Hasdai's Letter and Joseph's Reply, is in the library of Christ Church in Oxford. According to Dunlop and the Russian expert, Kokovtsov,6 the manuscript "presents a remarkably close similarity to the printed text" and "served directly or indirectly as a source of the printed text".7 It probably dates from the sixteenth century and is believed to have been in the possession of the Dean of Christ Church, John Fell (whom Thomas Brown immortalized with his "I do not love thee, Dr Fell..."). .Another manuscript containing Joseph's Reply but not Hasdai's Letter is preserved in the Leningrad Public Library. It is considerably longer than the printed text of Akrish and the Christ Church manuscript; accordingly it is generally known as the Long Version, as distinct from the Akrish-Christ Church "Short Version", which appears to be an abbreviation of it. The Long Version is also considerably older; it probably dates from the thirteenth century, the Short Version from the sixteenth. The Soviet historian Ribakov8 has plausibly suggested that the Long Version - or an even older text - had been edited and compressed by mediaeval Spanish copyists to produce the Short Version of Joseph's Reply. lAt this point we encounter a red herring across the ancient track. The Long Version is part of the so-called "Firkowich Collection" of Hebrew manuscripts and epitaphs in the Leningrad Public Library. It probably came from the Cairo Geniza, where a major part of the manuscripts in the Collection originated. Abraham Firkowich was a colourful nineteenth-century scholar who would deserve an Appendix all to himself. He was a great authority in his field, but he was also a Karaite zealot who wished to prove to the Tsarist government that the Karaites were different from orthodox Jews and should not be discriminated against by Christians. With this laudable purpose in mind, he doctored some of his authentic old manuscripts and epitaphs, by interpolating or adding a few words to give them a Karaite slant. Thus the Long Version, having passed through the hands of Firkowich, was greeted with a certain mistrust when it was found, after his death, in a bundle of other manuscripts in his collection by the Russian historian Harkavy. Harkavy had no illusions about Firkowich's reliability, for he himself had previously denounced some of Firkowich's spurious interpolations.9 Yet Harkavy had no doubts regarding the antiquity of the manuscript; he published it in the original Hebrew in 1879 and also in Russian and German translation,10 accepting it as an early version of Joseph's letter, from which the Short Version was derived. Harkavy's colleague (and rival) Chwolson concurred that the whole document was written by the same hand and that it contained no additions of any kind.11 Lastly, in 1932, the Russian Academy published Paul Kokovtsov's authoritative book, The Hebrew-Khazar Correspondence in the Tenth Century12 including facsimiles of the Long Version of the Reply in the Leningrad Library, the Short Version in Christ Church and in Akrish's pamphlet. After a critical analysis of the three texts, he came to the conclusion that both the Long and the Short Versions are based on the same original text, which is in general, though not always, more faithfully preserved in the Long Version.
4
Kokovtsov's critical survey, and particularly his publication of the manuscript facsimiles, virtually settled the controversy - which, anyway, affected only the Long Version, but not Hasdai's letter and the Short Version of the Reply. .Yet a voice of dissent was raised from an unexpected quarter. In 1941 Poliak advanced the theory that the Khazar Correspondence was, not exactly a forgery, but a fictional work written in the tenth century with the purpose of spreading information about, or making propaganda for, the Jewish kingdom.13 (It could not have been written later than the eleventh century, for, as we have seen, Rabbi Barzillai read the Correspondence about 1100, and Ibn Daud quoted from it in 1161). But this theory, plausible at first glance, was effectively demolished by Landau and Dunlop. Landau was able to prove that Hasdai's Letter was indeed written by his secretary Menahem ben-Sharuk. And Dunlop pointed out that in the Letter Hasdai asks a number of questions about Khazaria which Joseph fails to answer - which is certainly not the way to write an information pamphlet:
Dunlop goes on to ask a pertinent question:
Why the Letter of Hasdai at all, which, though considerably longer than the Reply of Joseph, has very little indeed about the Khazars, if the purpose of writing it and the Reply was, as Poliak supposes, simply to give a popular account of Khazaria? If the Letter is an introduction to the information about the Khazars in the Reply, it is certainly a very curious one - full of facts about Spain and the Umayyads which have nothing to do with Khazaria.15
Dunlop then clinches the argument by a linguistic test which proves conclusively that the Letter and the Reply were written by different people. The proof concerns one of the marked characteristics of Hebrew grammar, the use of the so-called "waw- conversive", to define tense. I shall not attempt to explain this intricate grammatical quirk,*[The interested reader may consult Weingreen, J., A Practical Grammar for Classical Hebrew, 2nd ed, (Oxford, 1959)] and shall instead simply quote Dunlop's tabulation of the different methods used in the Letter and in the Long Version to designate past action:16
Waw Conversive Simple Waw
with Imperfect with Perfet
Hasdai's Letter 48 14
Reply (Long Version) 1 95
In the Short Version of the Reply, the first method (Hasdai's) is used thirty-seven times, the second fifty times. But the Short Version uses the first method mostly in passages where the wording differs from the Long Version. Dunlop suggests that this is due to later Spanish editors paraphrasing the Long Version. He also points out that Hasdai's Letter, written in Moorish Spain, contains many Arabisms (for instance, al-Khazar for the Khazars), whereas the Reply has none. Lastly, concerning the general tenor of the Correspondence, he says:
To sum up, it is difficult to understand why past historians were so reluctant to believe that the Khazar Kagan was capable of dictating a letter, though it was known that he corresponded with the Byzantine Emperor (we remember the seals of three solidi); or that pious Jews in Spain and Egypt should have diligently copied and preserved a message from the only Jewish king since biblical times.