JAHRBUCH DER ÖSTERREICHISCHEN BYZANTINISTIK, 54. Band/2004, 39–52 © 2004 by Österreichische Akademie der Wissenschaften, Wien WARREN TREADGOLD / ST. LOUIS, MISSOURI THE HISTORICITY OF IMPERIAL BRIDE-SHOWS Seven independent Byzantine sources record that five times in the eighth and ninth centuries the winner in a competition of beautiful women became the bride of an emperor or future emperor. These sources, four written by contemporaries, include two chronicles, an oration by the em- peror Leo VI (882–912), and four lives of saints. Without explicitly ques- tioning this evidence, historians largely ignored the shows until 1979, when I published an article arguing for their historical importance.1 In 1985 my article provoked a response by the late Swedish philologist Lennart Rydén, who maintained that the shows were not merely unimportant as historical events but entirely fictional.2 Since that time, several historians have sum- marily dismissed Rydén’s arguments as inadequate to discredit such over- whelming evidence.3 Yet several other scholars have accepted Rydén’s ob- jections and conclusions, often after considering only a fraction of that evidence.4 The attack on the historicity of the bride-shows seems therefore to call for a more detailed response. Here I shall again review the sources and the arguments against accepting them, considering the five recorded shows in chronological order. 1 W. TREADGOLD, The Bride-Shows of the Byzantine Emperors. Byz 49 (1979) 395–413. Most of this article was first presented in a lecture to the Oxford Byzantine Society in May 2001 at the kind invitation of Professor Elizabeth Jeffreys. 2 L. RYDÉN, The Bride-Shows at the Byzantine Court – History or Fiction? Eranos 83 (1985) 175–91. 3 The dissenters include L.-M. HANS, Der Kaiser als Märchenprinz: Brautschau und Hei- ratspolitik in Konstantinopel 395–882. JÖB 38 (1988) 46–52, and G. DAGRON, Nés dans la pourpre. TM 12 (1994) 137–40. 4 Those who agree with Rydén include R.-J. LILIE, Byzanz unter Eirene und Konstantin VI. (780–802). Frankfurt am Main 1996, 42–43 Claudia LUDWIG, Sonderformen byzan- tinischer Hagiographie und ihr literarisches Vorbild. Frankfurt am Main 1997, 104–45 (repeating Rydén’s arguments and adding others I find even more farfetched), and Martha VINSON, The Life of Theodora and the Rhetoric of the Byzantine Bride-Show. JÖB 49 (1999) 31–60 (repeating Rydén’s arguments). P. SPECK, Eine Brautschau für Staurakios? JÖB 49 (1999) 26–30, discusses only the show of 807, relying on Rydén’s arguments and some others that I find arbitrary and implausible. 40 Warren Treadgold The source for the first recorded bride-show, which reportedly preceded a wedding in November 788, is the Life of St. Philaretus the Almsgiver by Philaretus’ grandson Nicetas of Amnia. Nicetas dates his own work to 821/22, when he was in exile for a reason he fails to mention. His grand- father St. Philaretus was a provincial landowner in Paphlagonia who had allegedly bankrupted himself by his generosity. According to his Life, his granddaughter Maria was selected to compete for the hand of the emperor Constantine VI (780–97) by a panel of imperial envoys, who visited her village of Amnia on their travels to identify suitable candidates. At a bride- show subsequently held in Constantinople, Maria was chosen to marry Constantine by his mother, the empress Irene (797–802).5 Following arguments that had already been made by Paul Speck, Ry- dén attacked the historicity of this bride-show by attacking the historical value of Nicetas’ whole work.6 According to Rydén, Nicetas had three purposes in writing: to suppress the failure of his cousin’s marriage to Constantine VI, which ended in divorce in 795, to please the emperor Michael II (820–29), who married Maria’s daughter Euphrosyne, and to tell a good story.7 The first two motives, however, would not have required any distortions or omissions in the Life of Philaretus, because the Life ends with Philaretus’ death in 792, before the divorce. Though in my opinion Michael married Euphrosyne in 824, at least two years after Nicetas wrote, in a new edition of the Life of Philaretus Rydén restated his view that the Life was a sort of “wedding gift” for Michael and Euphrosyne, and that Nicetas hoped it would persuade the emperor to recall him from exile.8 Nicetas’ desire to tell a good story may well have distorted the earlier part of the Life of Philaretus, which is loosely modeled on the Book of Job and set in Philaretus’ home village, a place unknown to most of Nicetas’ likely readers. Yet the Life changes dramatically in tone in its latter part, which bears no resemblance to the story of Job and is set at the imperial court.9 Neither Rydén nor those who agree with him have ever explained 5 Life of Philaretus, now ed. L. RYDÉN, The Life of St. Philaretos the Merciful Written by his Grandson Niketas: A Critical Edition with Introduction, Translation, Notes, and Indices. Uppsala 2002. 6 P. SPECK, Kaiser Konstantin VI.: Die Legitimation einer fremden und der Versuch einer eigenen Herrschaft. Munich 1978, 204–7. 7 RYDÉN, op. cit. (supra n. 2), 180–82. 8 RYDÉN, op. cit. (supra n. 5), 49–50; cf. W. TREADGOLD, The Byzantine Revival, 780–842. Stanford 1988, 246–47 and n. 339. 9 RYDÉN, op. cit. (supra n. 5), 16, himself acknowledged that “in this tale there is a gradual transition from fiction to reality.” Cf. A. KAZHDAN, A History of Byzantine Literature (650–850). Athens 1999, 281–91 (though here Kazhdan does not concern himself with the historicity of bride-shows, which he informed me he accepted).  The Historicity of Imperial Bride-shows 41 how Nicetas could expect Michael II (or any informed contemporary) to believe a fabrication about an imperial marriage that had occurred thirty- four years earlier, well within living memory. Nor can they explain how Euphrosyne herself, who had lived with her mother for many years in the same monastery, could have been ignorant of the circumstances of her own parents’ marriage. Nor did Rydén suggest how and why, without gathering candidates in the manner described by the Life, Irene selected the grand- daughter of an obscure and impoverished magnate from the tiny and dis- tant village of Amnia. So far, Rydén’s argument seems not only baseless but incredible. The source for the second bride-show, which reportedly preceded a wed- ding on December 20, 807, is the chronicle of Theophanes Confessor. This source, compiled by Theophanes from material collected by his friend George Syncellus, was completed just five or six years after the show, be- tween September 813 and December 814.10 According to the chronicle, the emperor Nicephorus I (802–11) assembled girls from all over the empire in order to choose a wife for his son Stauracius (811). At the show Nicephorus selected Theophano of Athens, a relative of the late empress Irene. The chronicler, who consistently vilifies Nicephorus, adds that Theophano had already been betrothed to another man with whom she had had intercourse “many times” (ϖολλάκις), and was less beautiful than two other contestants, whom Nicephorus himself seduced.11 Rydén, recognizing the evident fact that Nicephorus wanted to associ- ate his new dynasty with Theophano’s relative Irene, asked, “why arrange a bride-show if Theophano was already engaged and apparently not very beautiful?” He concluded that the chronicler invented the show in order to discredit Nicephorus. Even while conceding that the chronicler ”writes as if this method of finding a bride for the imperial bridegroom was a tradi- tional possibility,” Rydén declared that “there is no record of a predeces- sor” because the Life of Philaretus is “both later … and unreliable ….” According to Rydén, since the show had no historical precedent, the chronicler must have “borrowed the bride-show motif from literature, al- though I cannot point to any particular source.”12 However, even if Theophano was chosen in advance, and neither vir- ginal nor attractive, a bride-show would have allowed Nicephorus to claim that he had demonstrated her virtue and beauty by staging her victory 10 See most recently C. MANGO and R. SCOTT, The Chronicle of Theophanes Confessor. Oxford 1997, lii-lxiii. 11 Theophanes 483. 12 RYDÉN, op. cit. (supra n. 2), 179–80. 42 Warren Treadgold over a selection of the most virtuous and beautiful girls in the empire. Moreover, in all probability the chronicler invented his stories of Theoph- ano’s and Nicephorus’ sexual activities, which he was in no position to have known in such detail and few if any of his readers could have known either. As I have remarked elsewhere, “While such gossip would have been easy to invent about a girl who had been engaged and an emperor who was a wid- ower, Nicephorus would have been unlikely to expose his son to ridicule or to risk having a grandson of doubtful legitimacy ….”13 What the chronicler could not have done is expect readers with any knowledge of the court to believe in a very recent and public event that had never happened. The rest of Rydén’s argument is not merely baseless but circular. He can only maintain that the bride-show of 807 lacked an historical prece- dent on his own assumption that the bride-show of 788 was fictional. Yet the fact that Theophanes’ chronicle mentions the bride-show as a recog- nized custom corroborates the historicity of the show of 788, because the chronicler must have been recalling an actual event, not its mention in the as yet unwritten Life of Philaretus. As Rydén accurately observed later in his article, “in the eighth century the literary output was too small to make the introduction of new motifs [like the bride-show] likely.”14 The third bride-show, which reportedly led to a wedding on June 5, 830, is recorded by two independent sources, the chronicle of Symeon the Log- othete (copied by the Pseudo-Symeon) and the Life of St. Theodora the Empress. Although the Logothete compiled his chronicle soon after 948, he evidently took the portion including the bride-show from a single, quite reliable source compiled between 855 and 865, twenty-five to thirty-five years after the show, again well within living memory.15 The Life of Theo- dora must be later than 867, though perhaps not by much.16 Both sources 13 TREADGOLD, op. cit. (supra n. 8), 153. 14 RYDÉN, op. cit. (supra n. 2), 190. 15 See W. TREADGOLD, The Chronological Accuracy of the Chronicle of Symeon the Log- othete for the Years 813–845. DOP 33 (1979) 171–94, esp. 180, 192, and 193–94. R.-J. LILIE et al., Prosopographie der mittelbyzantinischen Zeit I (641–867): Prolegomena. Berlin 1998, 22 and 289–90, have attacked the reliability of Symeon’s chronicle because it records the death of Manuel the Armenian in 838, though according to them a seal of Manuel’s proves he was still alive after 843; but this seal should be dated before 829 (see infra n. 22). 16 See most recently Martha VINSON, Life of St. Theodora the Empress, in Alice-Mary TALBOT, ed., Byzantine Defenders of Images: Eight Saints’ Lives in Translation. Wash- ington 1998, 355 and n. 5 (for a date between 867 and 912), and D. AFINOGENOV, The Bride-Show of Theophilos: Some Notes on the Sources. Eranos 95 (1997) 10–18 (for a “rather late” date of the Life, but not of all its components). Neither argument is conclusive.  The Historicity of Imperial Bride-shows 43 say that girls were brought from various places to compete in a bride-show held for the young emperor Theophilus (829–42), who chose Theodora. Symeon adds that the girls were selected by Euphrosyne, Michael II’s widow and Theophilus’ stepmother, who held the show in the Triclinium of the Pearl in the Great Palace and gave Theophilus a golden apple to award to his choice. Seeing one beautiful contestant, the poetess Cassia, Theophilus tactlessly remarked that evil had come to man through a woman. Cassia gently responded that better things had also come from a woman. (The references, quoted in verse, are of course to Eve and the Virgin.) Wounded by Cassia’s retort, Theophilus chose Theodora.17 In the Life of Theodora, Theodora is naturally the heroine. When Theophilus gives all the contestants apples (apparently not golden but real ones), she is the only contestant who keeps her apple and produces another, a gift from a monk in Nicomedia who had prophesied to her what would happen.18 Rydén dismissed the Life of Theodora in a footnote, observing that by having Theodora win because of her piety rather than her beauty it makes the bride-show “meaningless, at least from an historical point of view.”19 Rydén however overlooked several indications that character and ortho- doxy were just as important as beauty for bride-show contestants.20 He also failed to notice that these two sources, with their radically different inter- ests in the event, cannot derive from a single literary source and therefore appear to be independent reports of the event itself. While mentioning the late date of Symeon’s chronicle, Rydén ignored the much earlier date of its source, which again was so early that knowledgeable readers could hardly have accepted a falsification of such an important public event. Rydén raised three specific objections to the version in Symeon’s chronicle. First, he assumed that Euphrosyne had from the first wanted her stepson to marry her fellow Paphlagonian Theodora, so that a bride- show was unnecessary.21 Euphrosyne, who had probably never set foot in Paphlagonia, may actually not have cared much about her mother’s native province; but she and Theophilus surely cared that Theodora was the niece of Manuel the Armenian, who had been an important general since 813 and 17 On the text of this passage in Symeon’s chronicle, which is still not properly edited, see W. TREADGOLD, The Problem of the Marriage of the Emperor Theophilus. GRBS 16 (1975) 327–29. 18 A. MARKOPOULOS, Βίος τῆς αὐτοκράτειρας Θεοδώρας. Symmeikta 5 (1983) 249–85, esp. 259–60 on the bride-show. 19 RYDÉN, op. cit. (supra n. 2), 188–89 n. 52. 20 See TREADGOLD, op. cit. (supra n. 1), 409–10. 21 RYDÉN , op. cit. (supra n. 2), 187–88. 44 Warren Treadgold had recently served as preceptor to the young Theophilus.22 Yet it is far from certain that Euphrosyne, who had chosen all the contestants, pre- ferred Theodora to the others, or could have dictated any choice to her strong-willed stepson. As a secret iconophile, Euphrosyne presumably wanted her stepson to choose an iconophile bride; but both Theodora and the rejected Cassia were iconophiles, and the empress could have made sure that the other candidates were iconophiles as well. Even if she did advise Theophilus to choose Theodora, the show would still have served its usual function of certifying the winning contestant’s beauty and virtue. Second, Rydén repeated an old observation of J. B. Bury’s that the show could not have been held in the Triclinium of the Pearl, which was built during Theophilus’ reign.23 While this objection made sense on Bury’s mistaken assumption that Theophilus’ marriage took place in 821, Rydén himself accepted Symeon’s date of June 830.24 This date leaves eight months of Theophilus’ reign for the completion of the triclinium, which the show probably inaugurated. Finally, Rydén assumed that the story of Cassia’s exchange of verses with Theophilus was invented to discredit the iconoclast emperor, especially because Rydén thought Cassia was too old to be a plausible bride in 830.25 Again, this is far from clear. Though Cassia’s age is uncertain, she could easily have been of marriageable age in 830.26 22 On Manuel’s earlier career as a general, see TREADGOLD, op. cit. (supra n. 8), 198 and 222, and on Manuel’s preceptorship for Theophilus, see G. ZACOS and A. VEGLERY, Byz- antine Lead Seals I.2. Basel 1972, 1190–91 (no. 2151A). LILIE et al., PmbZ III. Berlin 2000, Manuel #4707, 136–41, assume that the person holding authority over “the five themes” of Anatolia in 819 was not Manuel but the Metropolitan of Smyrna; but a bishop with jurisdiction over such a wide area delimited by military districts would be unparalleled and nonsensical, while Manuel’s commanding all five themes explains, as the editors cannot, why some sources say Manuel commanded the Anatolics and others say the Armeniacs. They also assume that the seal referring to Manuel as βάγουλος τοῦ δεσπότου means that Manuel was guardian of Michael III after 842; but the evidence seems clear that Manuel died in 838 (see TREADGOLD, op. cit. [supra n. 15], 180–83), and coins give the title δεσπότης to Michael (and Theophilus) only before his reign. Lilie and his collaborators identify this “prince” as Michael only because they accept the errone- ous theory of E. W. Brooks that Theophilus married ca. 821 and was therefore much older than the sources indicate (see infra n. 24). 23 RYDÉN, op. cit. (supra n. 2), 188, citing J. B. BURY, A History of the Eastern Roman Empire. London 1912, 82 n. 1. 24 For a discussion and refutation of Brooks’s theory, see TREADGOLD, op. cit. (supra n. 17), 325–41. 25 RYDÉN, op. cit. (supra n. 2), 188, thinks that Cassia was “born at the beginning of the century,” while in 830 Theophilus was 17 (TREADGOLD, op. cit. [supra n. 17], 335–38). 26 I believe she was about 22 at that date; see TREADGOLD, op. cit. (supra n. 8), 269.  The Historicity of Imperial Bride-shows 45 Her recorded exchange with Theophilus has nothing to do with Iconoclasm, and implies that she was a worthier candidate than Theodora, the iconophile saint who later abolished Iconoclasm. In any case, even if the exchange of verses was a pious elaboration, that would no more discredit the essential historicity of this bride-show than salacious elaborations in Theophanes’ chronicle would discredit the bride-show of 807. More recently, Dimitry Afinogenov has found the source of Theophilus’ and Cassia’s verses in a homily by a not easily identified author of the fifth or early sixth century, and consequently concluded that the exchange was fictional. He acknowledges, however, that Symeon’s chronicle describes the verses it quotes as the general sense but “not the exact wording” of Theophilus’ and Cassia’s remarks; such a candid admission suggests that the chronicle improved upon the words of the exchange but accurately reported its substance. Although Afinogenov argues that the Life of Theo- dora is fairly late, he concedes its complete independence of Symeon‘s ac- count, which again indicates that their common link was an historical event, not a literary source.27 The bride-show of 855, like that of 830, is attested by two sources. The earlier is the funeral oration for Basil I (867–886) delivered by his supposed son Leo VI in 886. Leo mentions that his mother Eudocia Ingerina com- peted in a bride-show held for Michael III (842–67), who, because God had destined her for another husband, chose an inferior bride.28 Rydén pointed out that Michael had already made Eudocia his mistress before 855, when Michael’s mother Theodora, who considered Eudocia shameless, forced him to marry another woman. Ten years later Michael, having packed his mother off to a convent, made a nominal marriage for his mistress with the future emperor Basil I. By this marriage Michael was probably trying to legitimize his child, the future Leo VI, whom the pregnant Eudocia was soon to bear.29 According to Rydén, the bride-show “motif makes it possible for Leo to transform the scandalous background of his parents’ marriage and the delicate question of his own paternity into a story that is flattering to his 27 AFINOGENOV, op. cit. (supra n. 16), 10–18 (quotation from 12). 28 A. VOGT and I. HAUSHERR, Oraison funèbre de Basile I par son fils Léon VI le Sage. OC 26 (1932) 38–79, esp. 54 on the bride-show. 29 See C. MANGO, Eudocia Ingerina, the Normans, and the Macedonian Dynasty. ZRVI 14/15 (1973) 17–27. The rambling and inconclusive discussion of Leo’s paternity in Shaun TOUGHER, The Reign of Leo VI (886–912). Leiden 1997, 23–67, adds nothing of significance. 46 Warren Treadgold mother and implies that God regarded Basil as superior to Michael.”30 First, however, Rydén never established that any such literary motif was current at the time, since even inaccurate historical accounts would not constitute a literary motif if they were meant to be believed as history. Second, the bride-show had nothing to do with Leo’s legitimacy one way or the other. Third, Leo was by no means flattering Eudocia by recalling that she had been rejected in a bride-show, or glorifying Basil by recalling that he had married a rejected contestant. Probably Theodora had used the previously established bride-show requirement of virtue to eliminate Eudocia for not being a virgin; but in any case she had been found inferior to the nonentity who was chosen. Although in his oration Leo apparently felt unable to omit the well-known fact of the show, he mentioned it as briefly as possible. Most people who heard his long and tedious speech prob- ably missed the reference entirely, as he would have preferred. Anyone who noticed the hurried allusion probably thought that Leo had finessed the matter about as well as could be expected. The other source for this show is the Life of St. Irene of Chrysobalantum. The best clue to its date is its author’s remark that the rule of the Macedo- nian dynasty had lasted “to the fourth, or indeed (ἢ καὶ) the fifth, genera- tion.” The author’s uncertainty about the fifth generation, which was that of the emperor Basil II (963–1025), suggests a date for the Life between 963 and 976, when the co-emperors Nicephorus Phocas and John Tzimisces held power and no one could be sure that Basil would ever actually rule.31 The Life records that St. Irene, a member of the noble family of the Gu- berii, came from Cappadocia to compete in Michael III’s show, since she was distinguished by the physical and moral beauty required of contest- ants. She was accompanied by her sister, evidently also a contestant, who later married the emperor’s brother, the Caesar Bardas. This Life’s account of the bride-show of 855 is however chronologi- cally incompatible with two other incidents in the same text. It mentions 30 RYDÉN, op. cit. (supra n. 2), 182–83. 31 The Life of St. Irene Abbess of Chrysobalanton, ed. and trans. J. O. ROSENQVIST. Upp- sala 1986, chap. 12, 52. At 52–53 n. 7, the editor argues for a translation of ἢ καὶ “not as truly disjunctive but as marking a climax,” and so dates the Life between 976 and 1025; but he admits that the words “could possibly be taken as a hint at the conditions during Basil II’s minority” (cf. Rosenqvist’s remarks at xxvii–xxix). Since the former interpretation would make the reference to the fourth generation pointless, the latter seems clearly preferable. (On the other hand, when I wrote in 1979 in TREADGOLD, op. cit. [supra n. 1], 404, that the Life was “probably early,” my youthful intuition was badly at fault.)  The Historicity of Imperial Bride-shows 47 that on her way to the show Irene visited the hermit St. Joannicius, who had however died in 846. Interestingly, the Life reports that Joannicius was often invisible, though he took shape to predict that Irene would be the abbess of Chrysobalantum. The Life adds that, after Irene failed to win the bride-show of 855, the nuns of Chrysobalantum could not persuade her to become their abbess until they referred the matter to the Patriarch Methodius, who had however died in 847. According to the Life, Methodius declared that God had already miraculously revealed their choice to him before he consecrated Irene as abbess.32 Because of these chronological incompatibilities and the vagueness of most of the Life, Rydén rejected it as an historical source, while its editor, Jan Rosenqvist, has pronounced it fictional, like the bride-show and even St. Irene herself.33 Skepticism about this admittedly dubious Life may well have been what first led Rydén to skepticism about bride-shows in general. Yet the Life of Irene, unreliable though it may be, is at worst no evidence for the shows, not evidence that they never happened. Moreover, Rydén and Rosenqvist have probably gone too far in rejecting the whole Life and Irene herself as fantasies. After all, according to the Life Irene’s corpse was at the time of writing still in the Convent of Chrysobalantum working miracles at her tomb, which was tended by a relative of hers whom she had assisted during her life. I would rather conclude that by the time the self- effacing Irene died (at age 97) all that her biographer could gather about her early years was scanty and unreliable hearsay.34 Although the miraculous stories of Irene’s encounters with Sts. Joan- nicius and Methodius look very much like pious fictions made up to glorify Irene, her admirers had no reason to invent her failed candidacy in the bride-show, which showed that she had not so much rejected the world as been rejected by it. Her appearance as a contestant provides the only plau- sible explanation of why she had come from her home in faraway Cappa- docia to her convent in Constantinople, or of how her sister had happened to marry the Caesar. If Irene did compete in the show of 855, she was born around 840 and died around 937, only about thirty years before her Life was written. For all its defects, her Life is certainly independent of Leo VI’s oration, and it therefore corroborates Leo’s mention of the bride-show of 855. 32 Life of St. Irene, chap. 3, 8–12 (for the bride-show and prophecy of Joannicius) and chap. 7, 24–28 (for the prophecy and consecration by Methodius). 33 RYDÉN, op. cit. (supra n. 2), 189–90; ROSENQVIST, op. cit., xxiii–xlviii. 34 Life of St. Irene, chap. 24, 112 (for her tomb) and chap. 23, 110–12 (for her age at her death and for her relative). 48 Warren Treadgold The source for the last known bride-show, that of 882, is the Life of St. Theophano, the wife of Leo VI and reportedly the winning contestant. Ac- cording to this Life, after the emperor Basil I had assembled girls chosen for their virtue and beauty to compete for Leo’s hand in the usual fashion, Leo’s mother Eudocia selected her own relative Theophano to be Leo’s bride. The author of the Life describes himself as a relative of Theophano. Alexander Alexakis has made convincing use of an anonymous scholion to identify the Life’s author with the Magister Slocacas, who is known to have addressed an account of Theophano’s miracles to Leo VI himself.35 Rydén hypothesized that the author of the Life of Theophano invented the bride-show “to exalt Theophano’s beauty and grace,” and observed that, as a relative of the empress Eudocia who chose her, “Theophano was obviously married to Leo for dynastic reasons, and thus the whole bride- show is unlikely to be historical.”36 Once again, however, Rydén overlooked the obvious function of the shows in authenticating beauty and virtue when these otherwise might be doubted, and the improbability that anyone would try to falsify the circumstances of Leo’s marriage for a contempo- rary audience. That improbability is even greater because the audience apparently included Leo himself, who had detested his late wife and had no reason to want his marriage idealized.37 Besides these observations about specific bride-shows, Rydén also cited some general arguments that he thought undermined the historicity of all of them together. He declared, “Scholars have been puzzled” that the shows “appeared so suddenly and disappeared so soon.”38 Yet Rydén never explained why this or any other historical practice should not have been introduced at a specific time, or why or how it might have been expected to evolve gradually. Nor has he or anyone else identified any suitable op- portunities for later shows during the next 175 years, because during that time, by historical accident, no emperors were directly succeeded by their sons.39 Later, contradicting his own implication that the shows should have evolved gradually if they were historical, Rydén argued that if the shows had “belonged to a real custom … the descriptions of them would be uni- 35 A. ALEXAKIS, Leo VI, Theophano, a Magistros Called Slokakas, and the Vita Theophano (BHG 1794), in S. EFTHYMIADIS et al., eds., Bosphorus: Essays in Honor of Cyril Mango. Amsterdam 1995, 45–56. 36 RYDÉN, op. cit. (supra n. 2), 186–87. 37 See TREADGOLD, op. cit. (supra n. 1), 408. 38 RYDÉN, op. cit. (supra n. 2), 177–78. 39 TREADGOLD, op. cit. (supra n. 1), 412–13.  The Historicity of Imperial Bride-shows 49 form, given that in all other cases the Byzantine protocol was so strict.”40 In fact, the differences among the descriptions rather suggest the differing interests of independent observers, whether or not the protocol was the same (and in other ceremonies it was not nearly so rigid as Rydén implies). On the contrary, uniformity is what we would expect of literary influence, because one author would have copied another. Rydén’s failure to find the slightest textual similarity among the seven sources is a devastating weak- ness in his argument. Finally, Rydén alleged that “to a high degree, the texts in which the Byzantine bride-shows are described” are “anonymous,” like “the literary genre of the fairy tales.”41 In fact, among Byzantine texts, the sources for the bride-shows are comparatively early and well attributed. Of the seven, five have identifiable authors, one describing each show: Nicetas of Amnia for 788, Theophanes Confessor with George Syncellus for 807, Symeon the Logothete for 830, the emperor Leo VI for 855, and Slocacas the Magister for 882. At one point Rydén seemingly let slip his real objection, which has nothing to do with his textual arguments: he called bride-shows a “frivolous custom,” apparently too frivolous for real emperors to have adopted.42 Any serious case against the historicity of the bride-shows would need to demonstrate that the obvious connections among them can be explained more easily as literary influence than as historical causation. Rydén did not even hypothesize literary influences that might link the seven texts, and did not identify a single verbal parallel between any two of them. On the other hand, after the first recorded show each subsequent one occurred well within living memory of a previous show or shows, so that many peo- ple at court would surely have remembered the precedent, and if necessary reminded the emperor or empress of it. Usually no reminding would have been needed, because the links among the empresses who held and com- peted in the shows are obvious. The second show selected a relative of the empress Irene, who had held the first show. The third show was held by the empress Euphrosyne, daughter of the winner of the first show. The fourth show was held by the empress Theodora, winner of the third show. The fifth show was held by the empress Eudocia, a competitor in the fourth show. Instead of these self-evident historical connections, Rydén assumed separate falsifications in seven independent and mostly early sources which, apparently by pure accident, produced a consistent and historically coherent myth. 40 RYDÉN, op. cit. (supra n. 2), 190. 41 RYDÉN, op. cit. (supra n. 2), 190–91. 42 RYDÉN, op. cit. (supra n. 2), 178. 50 Warren Treadgold The unjustified skepticism that imperial bride-shows happened seems to be based on a justified skepticism that imperial marriages really de- pended on open competitions to identify the best and most beautiful bride in the empire. The empresses and emperors who chose the winning candi- dates plainly had other motives, most of which can be plausibly conjec- tured. In 788 Irene wanted a daughter-in-law from an obscure and impov- erished family that would be unable to support her son against herself. In 807 Nicephorus wanted a daughter-in-law who could connect his new dy- nasty with the family of his predecessor Irene. In 830 Euphrosyne wanted a daughter-in-law who shared her iconophile sentiments, like either Theo- dora or Cassia; probably she either recommended one of them to her iconoclast stepson or made sure that all the contestants were iconophiles. In 855 Theodora wanted a daughter-in-law who, unlike Eudocia Ingerina, would not encourage her son to assert himself. In 882 Eudocia wanted a daughter-in-law who as her relative would be loyal to her. In most cases the wishes of the bridegrooms were ignored. What made bride-shows attractive to the imperial family was the illu- sion, carefully spread over a wide territory by the inspectors who recruited contestants, that the winner would be the most beautiful and virtuous bride in the empire. This was no doubt Irene’s idea in starting the practice in the first place. Her motives would have been similar whether, as seems likely, she had been humiliated by Charlemagne’s breaking off his previous engagement of his daughter to her son, or Irene had broken the engage- ment herself so as to avoid having too powerful a daughter-in-law.43 The bride-shows obscured the awkward facts that Irene and Theodora chose nonentities to marry their sons, that Eudocia chose her own relative, that Nicephorus was trying to distract attention from his usurpation by a dy- nastic marriage, and that Euphrosyne married her iconoclast stepson to an iconophile. Though the sources mention none of these apparent motives, they report favorably on all the shows but that of Nicephorus. Even in that case, the chronicler’s denunciations of the circumstances betray a fear that the show in itself would make a good impression on his readers. The alternative to believing that the bride-shows were historical events is extremely unlikely, and no one has yet made any systematic attempt to defend it. We would need to assume that the bride-shows were deliberate fictions, separately concocted by at least five people with obscure motives, who somehow succeeded in deceiving many others, including well-informed courtiers and emperors, usually about events in the very recent past. Any- 43 See TREADGOLD, op. cit. (supra n. 8), 89–91.  The Historicity of Imperial Bride-shows 51 one who proposes this hypothesis should be ready to conjecture in detail what sort of person or persons invented the original concept of the bride- show, why this was done, by what means of transmission this person or persons influenced each of our sources, which of the sources were deceivers and which were deceived, and why each of them chose to continue per- petuating or believing these falsehoods. Anyone seriously attempting to answer these questions should also try to produce a full stemma showing the relations among all the hypothetical sources, written or oral, along with their conjectural dates or at least relative chronology. The result, even if not actually self-contradictory, would be a lengthy chain of arbitrary and farfetched conjectures, but it could be called a kind of source criticism, unlike what has been offered so far. In comparison with the extreme improbabilities involved in rejecting the sources for bride-shows, what difficulties lie on the other side? Why should we prefer to think that bride-shows were invented by six writers and one Byzantine ruler (Leo VI) in order to deceive the public, rather than think that the shows were in fact staged by five Byzantine rulers in order to impress the public? The only apparent motive for such a preference is the desire of some contemporary scholars to keep political history separate from literary, cultural, and social history, by distinguishing the sober and rational actions of emperors and empresses from the exotic and irrational imaginings of storytellers and showmen.44 Since this is impossible in the case of the emperor Leo VI, to conjecture that he told a frivolous lie is apparently more acceptable than to believe that the empress Theodora staged a frivolous ceremony. Yet are imperial bride-shows really any more exotic or irrational than an imperial throne room filled with roaring golden lions, twittering golden birds, and a mechanical throne that raised the emperor to the ceiling? Al- though no one has yet suggested that these publicly displayed ornaments 44 Cf. AFINOGENOV, op. cit. (supra n. 16), 10: “Here again, as in many other cases, the ma- jor difficulties arise from the fact that the researchers have to operate in an interdisci- plinary ‘twilight zone’ between history and what in Russian is called literaturovedenie (the German translation would be Literaturwissenschaft). … Now, if we find a positive answer to the question of the relationship between ‘the bride-shows’ as a motif of Byzantine literature … and ‘the Byzantine bride-shows’ as a historical reality, we shall be able to extract the whole chain of literature-turned-history in our sources, thus coming much closer to understanding their mentality and their way of handling the information we are striving to extricate from them.” Note that Afinogenov’s approach assumes without argument that the bride-shows are fictional (“literature-turned-his- tory”). 52 Warren Treadgold were hoaxes, sustained for more than a century by several Byzantine chroniclers and Constantine VII’s De Ceremoniis with the cooperation of Liudprand of Cremona, that suggestion would be no harder to defend than the theory that bride-shows never happened.45 In the absence of a detailed and documented argument, which in both cases has yet to be presented, neither idea should be taken seriously. 45 For these well-attested and (so far) unquestioned decorations in the imperial throne room, see, e.g., TREADGOLD, op. cit. (supra n. 8), 283–85 and n. 390.