The Problem of the Marriage of the Emperor Theophilus Treadgold, Warren T Greek, Roman and Byzantine Studies; Fall 1975; 16, 3; ProQuest pg. 325 The Problem of the Marriage of the Emperor Theophilus Warren T. Treadgold THE MARRIAGE of the Emperor Theophilus (A.D. 829-42) took place in the year 830 according to Symeon the Logothete, the only primary source to mention the event. In 1901, however, E. W. Brooks redated the marriage to 821/22. Brooks noted that Theophanes Continuatus has the youngest of the emperor's five daughters marry at least by 839, while a passage in the De Ceremoniis seems to assume the daughter was married by 831; Brooks con­ jectured that this daughter was the eldest and married at nine.1 Though Brooks' date of 821/22 for Theophilus' marriage has gone unchallenged, its implications, especially for Theophilus' coinage, have been argued over ever since.2 I believe that a resolution of the controversy is impossible under Brooks' theory, but the date and account of the Logothete, if accepted, can lead not only to solving the chronological problem but also to explaining some peculiar facts about Theophilus' life and reign. A survey of the sources will show the attractiveness of believing the Logothete. Except for the chronicle of George the Monk, which has mostly invective against iconoclasm to contribute, Symeon's chronicle is probably the earliest literary source for the reign of Theophilus. According to Bury, Symeon finished writing in either 944 or 948 and made use of the "Lost Amorian Chronicle" of an adherent of The- 1 E. W. Brooks, "The Marriage of the Emperor Theophilus," BZ 10 (1901) 540-45. I would like to thank Professor Ihor Sevcenko of Harvard, in whose seminar I wrote an earlier version of this article, for many references and suggestions, and Professor Philip Grierson of Gonville and Caius College, Cambridge, for reading the typescript and giving me his views on it. 2 See W. Wroth, Imperial Byzantine Coins in the British Museum I (1908) xlii-xliii; J. B. Bury History of the Eastern Roman Empire (1912) 465-68; G. Ostrogorsky and E. Stein, "Die Kron­ ungsordnungen des Zeremonienbuches," Byzantion 7 (1932)'2.24-33; A. R. Bellinger, "The Emperor Theophilus and the Lagbe Hoard," &rytus 8 (1944) 102-06; A.!, Dikigoropoulos, "The Constantinopolitan Solidi of Theophilus," DOPapers 18 (1964) 353-61; rev. of Dikigo­ ropoulos by D. M. Metcalf, HBN 20 (1966) 680-82; 1. Rochow, Kassia (1967) 6-19; P. Grierson, Catalogue of the Byzantine Coins in the Dumbarton Oaks Collection III.l (1973) 406-23. 325 Copyright (c) 2003 ProQuest Information and Learning Company Copyright (c) Duke University, Department of Classical Studies 326 THE MARRIAGE OF THE EMPEROR THEOPHILUS ophilus' son, Michael III (842-67).3 Symeon relates events in a slightly modified chronological order that ROmilly Jenkins has established to be free from error in a later part of the chronicle.' Unfortunately, no complete edition of the Logothete's chronicle exists. Only three of thirteen known MSS. have been published, of which two are plagia­ rizations of Symeon by Leo the Grammarian and Theodosius of Melitene and the third has been interpolated into the chronicle of George the Monk.5 Of these, Leo's version seems to be the best.6 There are three other chronicles. The 'Pseudo-Symeon' is a para­ phrase of the real Symeon with some additional, often valuable material.7 The compiler has also added dates by regnal years, but these are arbitrary fabrications of his own.S Joseph Genesius, dated by Bury between 944 and 948, is interested mostly in military history, for which he mechanically retells his sources, including some he himself considers unreliable.9 Bury dated Theophanes Continuatus to 949-50 and observed that the compiler probably used both Genesius and some of Genesius' sources.10 Since the Continuer's organization is topical, he is able to avoid chronolOgical questions. In addition to these chronicles, the scholar-Emperor Constantine VII (918-59) gives some incidental information in his De Ceremoniis, compiled in 956 or later.ll But this is a manual of imperial protocol with some historical ex­ amples, not a work of history. These late and sometimes contradictory sources can be checked in two ways. First, the De Ceremoniis includes a catalogue of the imperial tombs in Constantinople, presumably compiled from first-hand 8 Bury, op.cit. (supra n.2) 455-59. 'R. J. H. Jenkins, "The Chronological Accuracy of the 'Logothete' for the Years A.D. 867-913," DOPapers 19 (1965) 91-112. I> For a list of the MSS. and a full edition of a later part of Symeon, see G. Moravcsik, "Sagen und Legenden fiber Kaiser Basileios I," DOPapers 15 (1961) 110-22. 8 G. Moravcsik, Byzantinornrcical I (1958) 516, prefers Leo on the ground that he used both hypothetical redactions A and B of the first part of Symeon, while the other two used only A. 7 Cf. R. J. H. Jenkins, "The Supposed Russian Attack on Constantinople in 807: Evidence of the Pseudo-Symeon," Speculum 24 (1949) 403. 8 For example, according to Ps.-Symeon, Michael IT married Euphrosyne in 820 or later, their son Theophilus married Theodora in 830, and their daughter Maria married in 832 (ed. Bonn 1838 as "Symeon Magister," 620, 625, 630). • Bury, op.cit. (supra n.2) 460. 10 Bury, op.cit. (supra n.2) 460-61. 11 J. B. Bury, "The Ceremonial Book of Constantine Porphyrogennetos," BRR 22 (1907) 222. Copyright (c) 2003 ProQuest Information and Learning Company Copyright (c) Duke University, Department of Classical Studies WARREN T. TREADGOLD 327 observation. This supplies a trustworthy list of the members of Theophilus' family. Second, many of the coins of Theophilus' reign bear images of members of his family, and most of the coins of Theophilus' father, Michael II (820-29), bear the image of Theophilus. If correctly ordered and dated, the coins would supply information about the births, deaths and ages of the people they portray. To win credence, any solution of the problem of the marriage of Theophilus must be compatible with the evidence of the imperial tombs and coins. The first event mentioned by the Logothete and Ps.-Symeon after Theophilus' accession with his stepmother Euphrosyne (2 October 829) is the new emperor's bride-show.12 According to the chroniclers, Euphrosyne brought together beautiful girls from all the themes of the empire so that Theophilus might choose a wife from among them. The empress held the show in the Triclinium of the Pearl, where she gave her stepson a golden apple to present to the girl of his choice. Passing over a beautiful contestant named Cassia because she parried a tactless and unmetrical verse of his with an apposite verse of her own, Theophilus gave the apple to a Paphlagonian named Theodora, whom he married on Pentecost. Apparently this means Pentecost 830 (June 5), so that there would have been eight months during which Michael II was mourned and the contestants were assembled. Cassia, the chroniclers continue, retired to a convent and became a writer; in fact, her poems have survived to the present. Euphrosyne, the chronicles conclude, voluntarily gave up her co-regency after the marriage and retired to a convent of her own called the Gastria.13 In redating the marriage, Brooks relied on the texts of the Logo­ thete supplied by Theodosius ofMelitene and the interpolated George the Monk, which say that Theophilus Hcrowned Theodora in the Oratory of St Stephen, being himself crowned with her by the 11 Other bride-shows were held in 788, 807, 855 and 881. See M.-H. Fourmy and M. Leroy, "La vie de S. Philarere," By~antion 9 (1934) 135-43; Theophanes (ed. C. De Boor) 485; Vita S. Irenes, Acta Sanctorum 33, July VI (1868) 603-04, and A. Vogt and 1. Hausherr, "Oraison funebre de Basile I," Orientalia Christiana 26.1 no.77 (Rome 1932) 54; Eduard Kurtz, "Zwei griechische Texte tiber die HI. Theophano," MemAcadPetersbourg SER. 8, m.2 (1898) 21. 18 Leo Grammaticus (ed. Bonn 1842) 213-14=Theodosius Melitenus (ed. T. Tafel 1859) 147= Georgius Monachus (ed. Bonn 1838) 789-90, Ps.-Symeon 624-25. Not just in this story but throughout, the Logothete and Ps.-Symeon make Euphrosyne Theophilus' mother, though his real mother was Michael II's first wife, Thecla (cj. De Ceremoniis, ed. Bonn 1829, 645). The problem is probably the pro-Amorian nature of Symeon's source. Mter Thecla's Copyright (c) 2003 ProQuest Information and Learning Company Copyright (c) Duke University, Department of Classical Studies 328 THE MARRIAGE OF THE EMPEROR THEOPHILUS Patriarch Antonius with the crown both of marriage and of empire. "14 Since Theophilus' imperial coronation is known to have taken place in 821 or 822, Brooks dated the marriage to the same period.1s But Leo the Grammarian and Ps.-Symeon omit the phrase Kat Tip TOV ycX.I-'OV Kat Tef) TijC fJact>.elac CTECP€t. Without it, Symeon's participle CT€cpfhk would obviously mean only 'crowned with the crown of marriage'.16 Since Leo's is the best of the texts of the Logothete and the Ps.-Symeon was probably compiled early, it is less likely that they omitted the phrase than that the others interpolated it. But present knowledge of the text of Symeon does not permit certainty on such a point without confirmation. This is provided by the De Ceremoniis, which shows that in 821 or 822 the ceremonies described would have violated imperial protocol in four ways. First, since all crowned emperors crowned a new empress (Cer. 209), Michael II should have joined in crowning Theodora, even assuming that the implication that Theophilus was not yet crowned is misleading. Since coronations of emperors were held in St Sophia, Theophilus should not have been crowned emperor in the Oratory of Stephen, which was the proper place only for being crowned with the nuptial wreath and crowning empresses,17 Since co-emperors were crowned by the senior emperor (Cer. 194), Michael again, not the patriarch, should have crowned Theophilus emperor. Finally, the imperial crown was always called the C'rEl-'l-'a or 8,cX.87]l-'a, but this crown is a mere cTEcpoc-evidently the nuptial wreath with which any death, despite the Byzantines' disapproval of second marriages, Michael had married Euphrosyne-who was then a nun. The anti-Amorian Continuer exploits this scandal for all it is worth (Theophanes Continuatus, ed. Bonn 1838, 78-79, 86). By contrast, Symeon's pro-Amorian source seems, understandably, to have omitted the whole episode, for Symeon is ignorant of it. But this means that the Logothete read in his source that only Euphrosyne was Michael's wife, while Theophilus was Michael's son. His natural con­ clusion that Euphrosyne was Theophilus' mother hardly discredits his story. a cT'e~ 8~ Beo8JJpcr:v ~v fl/l(T'TJplcp TOV a"lov ETectvov, cTe8ek Ka~ aVr~c ~p.a aVrjj {m(, 'AVTWV{OV '1TaTp~ctpxov [Ka~ Ttl> TOV yctp.ov Ka~ Ttl> rijc {3ac,).elac cT'e~]. Antonius I Cassimates (ca Jan. 821-Jan. 837) could have performed the marriage at either date. See V. Grumel, La Chronologie (Paris 1958) 435. 15 Brooks, op.cit. (supra n.l) 541-42. 18 Cf Symeon's reference to Michael m's nuptial, but not imperial, coronation (Leo Gram. 230= Theod.Mel. 160= Georg. Monach. Add. 816, Ps.-Sym. 655). 17 Cer. 192, if. 196-97, 208-09. More precisely, the coronation of the empress was held just outside the Oratory of Stephen, which she and the emperors entered immediately afterwards. Copyright (c) 2003 ProQuest Information and Learning Company Copyright (c) Duke University, Department of Classical Studies WARREN T. TREADGOLD 329 common bridegroom was crowned.ls It is almost impossible that such a confused ceremony could have been held at the court of Michael II, or that a knowledgeable imperial official like the Logo­ thete could have reported it so erroneously. There are even more problems with the date 821/22. According to Theophanes Continuatus (143-44), the Triclinium of the Pearl, where the bride-show was held, was not even built by 821/22; it was built during Theophilus' reign, presumably in the months before the show and probably for the occasion. The Continuer also dates Euphrosyne's retirement, which is part of Symeon's story, near the beginning of Theophilus' reign.19 Finally, it is unlikely that Symeon would have dated Theophilus' imperial coronation to 830 about a page after dating the same event to 821. All the versions of the Logothete say that Michael II crowned his son in St Sophia, in accordance with protocol, before the revolt of Thomas the Slav, which began in late 821.20 To my mind, this settles the question of whether Theophilus was crowned in 821 or 822. But Symeon mentions no marriage in 821. Therefore the phrase in question can be demonstrated to be an interpolation by a misunderstanding copyist-no surprise in a text that has been tampered with as much as Symeon's. Consequently there is no more reason to date the story to 821/22 than to any other year. All internal evidence confirms the clearly implied date of 830. But Symeon's bride-show story must still be reconciled with the supposed marriage of the emperor's daughter Maria. The Continuer and apparently the Logothete himself believed that Maria was married to a certain Alexius Musele, whom Theophilus, having no sons living, gave the rank of Caesar and thus made his heir. But Byzantine law forbade women to marry before the age of twelve.21 Since Theophilus died in 842, none of his daughters could have been of marriageable age during his reign if he married in 830. Further­ more, since Maria was either the fourth or the fifth daughter and 18 Cer. 191-216 passim. Bury, op.cit. (supra n.2) 80 n.5, is mistaken that the nuptial crown is a cncpavwILa; that is the ceremony of nuptial coronation. See G. W. H. Lampe, A Patristic Greek Lexicon (1961). For the nuptial coronation of laymen, see Ph. Koukoules, BIJ'avTtvwv fJtoc ,ca~ 'TTO>"'.,.'CIL/'C IV (Athens 1951) 46-47. 18 Theoph.Contin. 86. The Continuer says Theophilus forced Euphrosyne's retirement, but I reject this, following Bury, op.cit. (supra n.2) 125-26. 20 Leo Gram. 211= Theod.Mel. 146=Georg.Monach. Add. 783, Ps.-Sym. 62l. 21 K. Zachariii von Lingenthal, Ecloga Leonis et Constantini, Epanagoge Basilii, Leonis et Alexandri (Leipzig 1852) 24 and 107. Copyright (c) 2003 ProQuest Information and Learning Company Copyright (c) Duke University, Department of Classical Studies 330 THE MARRIAGE OF THE EMPEROR THEOPHILUS died before her father, she could scarcely have lived past early child­ hood, much less have reached twelve. Her marriage aside, there is some independent evidence for Maria's age that scholars seem to have overlooked. Both Symeon and Theoph­ anes Continuatus record that when Maria died, Theophilus buried her in a magnificent coffin.22 In the De Ceremoniis, Constantine Porphyrogenitus describes this coffin as it was in his time, when it lay in the Church of the Holy Apostles next to the coffin of Maria's brother Constantine. Philip Grierson has deduced that Constantine died in infancy from the fact that the Porphyrogenitus calls his tomb a AaPVaK£ov-a little coffin.23 But Maria's tomb is also described as a AapvaK£ov. She seems therefore to have died well before she reached full growth, compatibly with Symeon's date for her father's marriage.24 Either Maria was married as a chil~ or she was only betrothed. Either case can be paralleled. The Continuer and Ps.-Symeon, sup­ ported by Symeon, attest an underage imperial marriage in their own time. Though the law forbade men to marry before fourteen, in 943 the future emperor Romanus II was married at four or five to a Provenc;al princess. Since she died five years later, the marriage could never have been consummated, like Maria's marriage if it was one.25 If Maria was only betrothed, a later imperial engagement provides an even better parallel. In 1163 Manuel I betrothed his eleven-year-old daughter, also named Maria, to Prince Bela of Hungary, later Bela J.26 Manuel, like Theophilus, had no male heirs or daughters of marriageable age at the time, and Bela, like Alexius, was a mature man who was made heir to the empire at the time of his betrothal. Under such circumstances, a marriage and an engagement would amount to much the same thing and could easily be confused, especially by chroniclers writing more than a century after the event. Because Maria would have had to be very young indeed at the time, I am inclined to think that she was only betrothed. But as long as 12 Leo Gram. 216-17= Theod.Mel. 149= Georg.Monach. Add. 794, Ps.-Sym. 630, Theoph. Contino 108. 23 Grierson, op.dt. (supra n.2) 407; if. Cer. 645. 1& Cremation was unknown in Byzantium (Koukoules, op.dt. [supra n.18] IV 193-95). 25 Cf. Theoph.Contin. 431 and 469, Ps.-Sym. 748 and 757, Leo Gram. 325=Theod.Mel. 234= Georg.Monach. Add. 917 (Symeon's chronicle ends before the reign of Romanus 11). For the marriageable age for men, see passages dted in n.21. II Nicetas Choniates (ed. Bonn 1835) 147. Cf. CMHI IV.l (1966) 230-31 and 582-83. Copyright (c) 2003 ProQuest Information and Learning Company Copyright (c) Duke University, Department of Classical Studies WARREN T. TREADGOLD 331 the evidence of her coffin is taken into account, the question of whether Maria was