The “correspondence” of Leo III and ‘Umar II: traces of an early Christian Arabic apologetic work

This article compares and revisits the corpus of texts pertaining to the socalled “correspondence” between the Byzantine emperor Leo III and the caliph ‘Umar b. ‘Abd al-‘Azīz. By adducing textual, philological and palaeographic arguments, I suggest that all the extant versions of the “correspondence” ultimately derive from an original Arabic Christian apologetic work, composed probably in mideighth century, in the monastic circles of Syria-Palestine. While acknowledging the importance of previous research on the subject, this article hopes to provide an original explanation that might finally account for both the similarities and the differences between the various extant versions of the “correspondence”. Besides clarifying the origin and transmission history of this text, the results of this study have broader implications for the history of Muslim-Christian relations in the early Islamic period, for the creation of a Christian Arabic culture, and for the circulation of literary texts between the Dār al-Islām and Byzantium in the early Middle Ages. The so-called “correspondence” of Leo III (r. 717–41) and ‘Umar II (r. 717–720 CE/99– 101 H)1 is arguably one of the most interesting texts of the Christian-Muslim debate from the early Islamic period. Because of its singular transmission history, it is also a text that has lent itself to many misinterpretations. The dating of the “correspondence”, its authorship, its audience and function, as well as its original language of composition, are all problematic issues. Several hypotheses have been formulated over the years, but none of them seems conclusive or thoroughly persuasive. This is partly a result of the reference to the emperor Leo, which has long sidetracked researchers, and partly it derives from the fact that very different versions of the “correspondence” exist, written in different languages and in different historical contexts. This linguistic barrier has often led to too specialized, narrow analyses that have prioritized one version over the others. This paper will aim at combining the information provided by the various versions of this source, in order to suggest a new explanation of their origin, which may account for both their similarities and their variations. The most recent contributions to the study of the “correspondence” will be acknowledged, and the main hypotheses advanced by scholars recapitulated. At the same time, it is the hope of this paper to contribute to the discussion by radically shifting perspective and intro The hijrī date (H) will be given along with the year of the Common Era (CE) only with reference to Muslim leaders, or to historical figures who lived under the caliphate. ducing a new interpretation, in a way that might enhance our understanding of this complex source. In general, scholarship has tended to emphasise either the Byzantine or, more recently, the Islamic nature and origin of the “correspondence”; it will be argued that, in both cases, this has led to a downplaying of relevant issues, and that either hypothesis is vulnerable to counter-arguments. Instead, I would like to put forward a new explanation concerning the nature of this source, the environment in which it was created, and its implications for the historical context. 1. Indirect references to the “correspondence” The earliest mention of an epistolary exchange between the emperor Leo III and the caliph ‘Umar b. ‘Abd al-‘Azīz is found in the Chronicle of Theophanes (d. 818),2 under the entry for Anno Mundi 6210 (717/18 CE): In the same year, a violent earthquake having hit Syria, ‘Umar banned [the use of] wine from the cities, and forced Christians to become Magaritai (magarizein).3 Those who became Magaritai, he made them exempt [from taxes?], while those who refused, he killed them. He caused many martyrs [to die]. He also decreed that the testimony of a Christian against a Saracen would not be accepted. And he composed (epoiēse) a letter about doctrine (epistolēn dogmatikēn) addressed to the emperor Leo, supposing to persuade him of Islam (peisein auton tou magarisai).4 ‘Umar’s letter is thus included among a list of measures taken by the new “master of the Arabs” (kratēsas tōn Arabōn) in relation to the Christians.5 It is noteworthy that ‘Umar’s epistle is characterised as dogmatikē, probably hinting at its religious mes Ed. Carl G. De Boor, Theophanis Chronographia (Leipzig: B.G. Teubneri, –). All following quotations of Theophanes’ work will depend on this edition. See also Cyril Mango, Roger Scott, The Chronicle of Theophanes Confessor: Byzantine and Near-Eastern history, A.D. – (Oxford: Clarendon Press, ); and Andreas Kaplony, Konstantinopel und Damaskus. Gesandtschaften und Verträge zwischen Kaisern und Kalifen – (Berlin: K. Schwarz, ), p. .  By magarizein, Theophanes probably means conversions to Islam. It is not easy, however, to assess what “conversion” exactly implied at that time. The term probably referred to formal and informal practices of acculturation and, most importantly, social assimilation that would hardly be rendered by the modern “to convert”. Magaritai, moreover, does not precisely correspond to “Muslim”, as it reflects the author’s polemical stance, as well as, probably, the Eastern origin of his source on these events. This name, in fact, is likely to reproduce the early adaptation in Greek of an original Arabic term (muḥājirūn). It appears in seventhand eighth-century Greek documents, and it is attested also in a Syriac form (mhggrāyē). In other passages, Theophanes prefers the Greek Sarakenoi or Hagarenoi. See Robert G. Hoyland, “New Documentary Texts and the Early Islamic State”, BSOAS  (), – (–); Robert G. Hoyland, Seeing Islam as Others Saw it: a Survey and Evaluation of Christian, Jewish and Zoroastrian Writings on Early Islam (Princeton NJ: Darwin Press, ), pp. –, p.  and n. , pp. – and nn. –; Patricia Crone, “The First-Century Concept of Hiğra”, Arabica / (), – (–).  Theophanes, Chron., p. . Cf. Mango’s translation (cf. fn. ), p. .  Theophanes, Chron. pp. –. 232 Cecilia Palombo