Jump to ContentJump to Main Navigation
The Baptized MuseEarly Christian Poetry as Cultural Authority$

Karla Pollmann

Print publication date: 2017

Print ISBN-13: 9780198726487

Published to Oxford Scholarship Online: March 2017

DOI: 10.1093/acprof:oso/9780198726487.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2019. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. date: 21 September 2019

Jesus Christ and Dionysus

Jesus Christ and Dionysus

Rewriting Euripides in the Byzantine Cento Christus Patiens

(p.140) 6 Jesus Christ and Dionysus
The Baptized Muse

Karla Pollmann

Oxford University Press

Abstract and Keywords

The cento Christus patiens takes about one-third of its c.2,600 iambic trimeters from tragedies by Euripides, telling in dramatic form the story of Christ’s Passion, death, and resurrection. This chapter argues that for metrical and lexical reasons the cento’s much disputed authorship has to be decided against Gregory of Nazianzus and in favour of a twelfth-century Byzantine author. An analysis of the adaptation of the Euripidean verses demonstrates the cento not to be viewed as mere learned play. The aim of this centonic transformation is to reveal the true, hidden sense of the famous pagan original, by employing various different modes of transferral and alteration, which confirms the findings in Chapter 4. The chapter’s groundbreaking contribution consists in elaborating the cento’s ultimate and highly original goal to contrast the vindictive pagan destroyer Dionysus with the merciful redeemer Jesus Christ in this Christian anti-tragedy with its new and different world-view.

Keywords:   cento, Euripides, Byzantine literature, anti-tragedy, hidden meaning, Christian world-view

In the eleventh book of the Iliad (11.636–7), a drinking vessel is mentioned which is so heavy that ‘another man only with great effort could lift it full from the table, but Nestor, aged as he was, lifted it without strain’ (ἄλλος μὲν μογέων ἀποκινήσασκε τραπέζης‎ | πλεῖον ἐόν‎, Νέστωρ δ‎’ ὁ γέρων ἀμογητὶ ἄειρεν‎).1 As Athenaios tells us in the Deipnosophistai 11.493 f–494 b, this prompted the question among the Alexandrine philologists in the third century BC as to whether or not Homer wished to imply that the aged Nestor was stronger than, for example, the heroes Achilles or Ajax. Sosibios,2 a court philologist of Ptolemaios Philadelphos, proposed that the phrase ‘the aged’ (ὁ γέρων‎) should be moved forward, since the text ought to be interpreted in the following way: ‘Another aged man only with great effort could lift it full from the table, but Nestor lifted it without strain’ (ἄλλος μὲν γέρων μογέων ἀποκινήσασκε τραπέζης πλεῖον ἐόν‎, ὁ δὲ Νέστωρ ἀπονητὶ ἄειρεν‎)—i.e. the aged Nestor as the strongest of all old men provided a more satisfactory meaning.3 As a justification for using this transpositional approach Sosibios presupposes the use of the rhetorical device of anastrophe4 in Homer’s text, i.e. a deliberate alteration (p.141) of the natural word order by the author (Deipnosophistai 11.493 d, e). Sosibios ought therefore to be considered the originator of a method of interpretation that aims to establish the meaning of an ambiguous text through rearranging its constituent parts, a method that, in this way, could solve difficulties in comprehending the Homeric text.

Sosibios experienced at first-hand that this method was not immune to a certain degree of capriciousness and opened the door to interpretative arbitrariness. As Athenaios relates, the next instalment of his royal stipend went unpaid. When he raised his objection to this in an audience with Ptolemaios, he was told that his stipend had been paid and he was shown the receipts for the stipends paid to other scholars, namely those of Soteros, Sosigenes, Bionos, and Apollonios. When the syllables So—si—bi—os were taken from their names and put together, then it could be seen that his stipend had also been paid. Thus Ptolemaios applied Sosibios’s interpretative method ad absurdum.

David Daube5 sees the theoretical justification for the technique of the cento in this Alexandrian method of interpretation. The cento is a poem that is comprised of verses or sections of verses from one or more other works, which have been detached from their original context and woven into a new text—like a ‘patchwork’. During this process some textual alterations or amendments made by the cento poet may also occur. According to Daube, the intention behind this was to determine the true meaning of the original text or texts through rearranging the constituent parts of one or more works, in our case works in verse form. So the cento technique has then an interpretative–exegetical objective. The cento first appears as a distinct poetic genre in Alexandria from the third or second century BC onwards.

It should be noted here that Ptolemaios’s ‘method’ goes beyond the practice as undertaken by Sosibios. Ptolemaios produces what might be termed a particularly rigorous cento, because he even segments words and places them together in new combinations. After Sosibios’s word rearrangement in the Homer verse, the meter is disrupted, and the original hexameter has been lost. His method therefore does not precisely correspond to the method in poetic centos, where the meter must also be preserved in the new text. In addition to Daube, I would therefore like to introduce a further technique that demonstrates the division and metrically correct amendment of verses as they occur in centos.

In his literary critical comedy, The Frogs, in which Aeschylus and Euripides conduct a poetic contest, Aristophanes (fifth century BC) has Aeschylus expose the monotony of the prologues to Euripides’ tragedies. Aeschylus claims he can ruin (1200 διαφϑερῶ‎, ‘I will ruin’) any prologue of a Euripidean tragedy by (p.142) inserting the half verse ‘He lost a little bottle’ (ληκύϑιον ἀπώλεσεν‎), because Euripides always constructs his prologues in such a way that such ‘a little bottle’ (ληκύϑιον‎) will always fit the meter, which he then proceeds to demonstrate seven times (1206–41). Every time Euripides begins to recite one of his prologues, Aeschylus interrupts him after the main caesura and completes the respective verse in a metrically correct manner with ‘He lost a little bottle’. A few examples will serve to illustrate this, taken from Aristophanes, The Frogs (1238–41):

(Euripides quotes, among other things, from the prologue of his tragedy Meleagros, of which only fragments have been preserved, here fragment 516 Kannicht)

Οἰνεύς ποτ‎’ ἐκ γῆς‎      (‘Oineus once from the earth’)

(Aeschylus interrupts)

ληκύϑιον ἀπώλεσεν‎. |      (‘lost a little bottle’.)

(Euripides requests he be permitted at least to complete the first verse)

Οἰνεύς ποτ‎’ ἐκ γῆς πολύμετρον λαβὼν στάχυν‎, |  (‘Oineus once from the earth reaping an abundant harvest,’)

ϑύων ἀπαρχὰς‎      (‘offering the first fruits,’)

(Aeschylus interrupts in the second verse)

ληκύϑιον ἀπώλεσεν‎. |      (‘lost a little bottle’.)

Since the iambic trimeter in tragedy very frequently has the word-end coinciding with the end of the fifth half-foot (i.e. in the main caesura), its appearance in Euripides’ work is nothing unusual or even absurd, and of course it is possible to interrupt the verse after this main caesura and to continue in a different way to that which had been planned. Rather, it is more remarkable that this is always possible with the syntactically stereotypical link ‘He lost a little bottle’. Therefore, already the ancient scholia surmised that the target of Aristophanes’ critical parody was the syntactical monotony of Euripidean verses.6

(p.143) This technique can be seen as the ‘practical’ birth of the cento. Once the divisibility of verses has been recognized as a viable literary possibility, this procedure can naturally be varied at will, i.e. parts of verses can be replaced by others which are either taken from another part of the same or from another original work, or are composed by the cento poet him or herself. In doing so, the successful creation of a poetic cento must ensure metrical, syntactical, and narrative7 seamlessness.

In general, the following features are characteristic of a cento:8

  1. 1. The ostensible arbitrariness of its composition, often pilloried, becomes understandable in the light of an overarching aim or hermeneutic framework to which both the cento’s poet and the cento’s readers must refer, since the latter may otherwise fail to recognize the particular quality and function of this kind of poetry.9

  2. 2. The tension between the original source text (i.e. the hypotext) and the newly created hypertext10 must be recognized by the readers in order to appreciate the specific quality of a cento. Rearranging an existing textual structure by recombining its lexical elements permits a broad range of uses or reuses of (p.144) the original content: the learned game, which neglects the contextual meanings of the original as far as possible in favour of the newly created unit of meaning; a parody with the aim of unmasking perceived weaknesses of a hypotext; serious interpretative engagement with the polysemy of the original content with the aim of revealing a hidden meaning of the text. A cento in this latter sense may be considered a subtype of exegesis.11

It is this exegetical potential that makes the genre of the cento12 interesting for some early Christians as well. Prompted by their desire to prove that even works by pre-Christian authors implicitly contain the Christian message,13 from the fourth and fifth centuries AD there appear Homeric and Vergilian centos that take up biblical stories, especially those concerning the life of Christ, and in which hexametrical verses or verse sections from Homer and Vergil are rearranged and partly modified to a greater or lesser extent in order to tell these Christian stories. It is important to note that only canonical texts that were recognized as authoritative are used for the composition of centos, as only these were considered worthy of the effort to interpret them.14 Moreover, only those canonical texts would have been familiar enough to the readers to allow them to recognize the centonic technique.15

As far as we can ascertain from the extant sources, the cento Christus Patiens (Chr. pat.; or: Χριστὸς πάσχων‎),16 which will be examined in detail in the following, is the only cento to be composed in the verse form of Greek tragedies, i.e. in iambic trimeters, rather than hexameters. Using a series of dramatic roles, this cento describes the passion, death, and resurrection of Jesus Christ. Around one-third of its approximately 2,600 verses are composed of Euripidean verses,17 the rest is the original composition of the cento poet, (p.145) although it is possible that the poet also drew from other Euripidean tragedies that have since been lost and are thus unknown to us. The majority of the verses or verse sections are not taken from the source entirely word for word, and the author inserts biblical proper nouns and implements syntactical and lexical alterations.

The manuscripts unanimously attribute authorship of the Christus patiens to the fourth-century Christian writer Gregory of Nazianzus, whose extant work also includes poems in iambic trimeters. However, the textual tradition before these manuscripts, which originated between the thirteenth and sixteenth centuries, does not provide evidence for Gregory’s authorship, so several scholars since the end of the sixteenth century have cited various arguments that undermine this assertion.18 One school of thought tends to ascribe the Christus patiens to another author of the fourth/fifth century,19 mainly because hexametrical centos were popular at this time and because this cento, albeit in iambic trimeters, could fit within the context of the antagonistic cultural disputes between paganism and Christianity in that period. According to this scholarly point of view, the Christus patiens would therefore provide evidence for the Christian ‘usurpation’ of the pagan genre of tragedy for the purposes of promulgating Christian beliefs.

The other school of thought dates it to the eleventh/twelfth century, referring principally to the metrical construction20 or the dogmatic content21 of the cento. In this context, there have also been some attempts at authorial attribution. It is important to note, however, that the criterion of meter can only indicate very approximate chronological limits as a result of the genre-specific (p.146) classical nature of the cento; by necessity there have been Euripidean verses since Euripides (but see pp. 144–5 in this chapter). The same applies, by and large, to the theological content: central statements of faith were largely fixed from the middle of the fifth century and were always strictly upheld by orthodox authorities, albeit paraphrased and differentiated in various ways. However, there is only limited scope for doctrinal speculations in the genre of tragedy as it is represented by the Christus patiens, which make it difficult to use this as a hard criterion of dating the cento. Moreover, attempts to narrow down the date of the Christus patiens through palaeographic evidence in the manuscripts22 or by determining its priorities in relation to Romanos the Melodist (sixth century)23 must be deemed unsuccessful.

Nevertheless, it will be attempted at this point to bring together several arguments that will allow for a more precise placing of the cento within the history of thought and that will make a later dating more plausible.

Viewed theologically, the work is orthodox. Mary is called ‘God-bearer’ (ϑεοτόκος‎), an epiclesis that was generally popular from the fourth century onwards and that was refuted by Nestorius in the fifth century on dogmatic grounds, arguing that Mary did not bear God, but rather the human Jesus. The cento poet was therefore not a Nestorian. However, there are no indications in (p.147) the cento of a direct engagement with this problem: the epiclesis is used as if it were taken for granted and not a matter of controversy.

Gregory of Nazianzus, Oratio 30.6 (Sources Chrétiennes 250.236) describes the gospel account of the Crucifixion of Jesus Christ as a wonderfully arranged drama, as Jesus displays emotions, among other things: τῆς δὲ αὐτῆς ἔχεται ϑεωρίας καὶ τὸ μαϑεῖν αὐτὸν τὴν ὑπακοὴν ἐξ ὧν ἔπαϑεν‎, ἥ τε κραυγὴ καὶ τὰ δάκρυα καὶ τὸ ἱκετεῦσαι‎, καὶ τὸ εἰσακουσϑῆναι καὶ τὸ εὐλαβές‎,24 ἃ δραματουργεῖται καὶ πλέκεται ϑαυμασίως ὑπὲρ ἡμῶν‎ (‘The same consideration applies to another passage, “He learnt obedience by the things which he suffered”, and to his “strong crying and tears”, and his “entreaties”, and his “being heard”, and his “reverence”, all of which he wonderfully wrought out, like a drama whose plot was devised on our behalf’).25 Yet it is striking that in the Christus patiens the suffering of Jesus Christ is never shown. Jesus only rarely appears with a speaking part (sixty-nine verses), but where he does, his speech mode is that of sacred command.26 In contrast, the other dramatic figures are assigned the task of illustrating his suffering. This different concept of the figure of Jesus is further evidence against viewing Gregory of Nazianzus as the composer of the cento.27

In the prologue, the poet explicitly states that the listener, who has until now listened to poems ‘in a devout manner’ (1 εὐσεβῶϛ‎), will now listen to ‘devout material’ (2 εὐσεβῆ‎) rendered in a poetic manner,28 ‘in the poetic style of Euripides’ (3 κατ‎’ Εὐριπίδην‎). This is not the construction of an antithetical rivalry between a Christianity only just beginning its struggle for a cultural identity in the face of the continued dominance of paganism, as was the case in the fourth and fifth centuries. The general attitude of Christian writers in the (p.148) fourth century towards pagan tragedy tended to be hostile,29 which renders the frank formulation of verse 3 unlikely for this period, although the acceptance of Homer’s unmatched qualities is characteristic of the Greek poetic tradition throughout, and occurs, for instance, in the fifth-century centonist Eudokia.30 This shared cultural inheritance can be appealed to entirely unselfconsciously in later times, especially in comparative discourses relating to poetics.31 The point here in the Christus patiens seems to be that the poet is going to offer something new and desirable, i.e. telling Christian content in a tragic form. This is made pressing and justifiable by the importance of the subject matter, Christi’s ‘passion of cosmic salvation’ (4 κοσμοσωτήριον‎…πάϑοϛ‎).32

Regarding meter it is striking that, in addition to its more or less strict adherence to quantitative meter, the cento only ever permits exactly twelve syllables in its iambic trimeters. Even ‘classical’ verses taken over from Euripides are modified in order to remove any doubling of short syllables.33 In other words, the cento uses the Byzantine twelve-syllable verse in line with mid- and late Byzantine practice. This mixed form is not evident in the work of Gregory of Nazianzus, who also composed poetry in iambic trimeters: he includes double short syllables, making him ‘more classicistic’ than the cento poet and for that reason can hardly be the author.34 The Byzantine twelve-syllable verse becomes more prevalent from the seventh century onward.35 If one does not wish to suggest the Christus patiens as evidence for an earlier use of the twelve-syllable verse, this would provide an argument for a terminus post for the work’s earliest possible date of composition. This and the fact that certain words appear in the Christus patiens that have otherwise not been traced any earlier than the eleventh and twelfth centuries36 seem to me to be (p.149) decisive criteria for not dating the work earlier than into the eleventh or twelfth century.37

Let us now turn to the formal and textual aspects of the work.38

In terms of form, it is not surprising that it is above all verses from Euripides that are used to compose the cento. Besides the aforementioned stereotypicality of at least some of the Euripidean verses, pilloried by Aristophanes (but which of course facilitates their use in the cento), there is also Euripides’ preference for general formulations and judgements to take into account, many of which were also collected separately into anthologies. Moreover, Euripides is historically termed the wisest of all the tragedians, an opinion also shared by the early Christian writers, and was highly popular as an author in school curricula, which are all favourable factors that encouraged the use of his works in a cento.

In terms of its content, Euripides’ tragedy Bacchae in particular represents a challenge for Christians. Its subject is the spread of the cult of Dionysus in Greece in the face of resistance from the family of Cadmus, and Dionysus-Bacchus, the god of theatre, is himself an active figure in the drama.39 However, aspects such as the appearance of a god among humans in human form, his suffering, the failure, or even active resistance by many to recognize his divinity, as well as the eventual revelation of his godhead are reminiscent of the Christian understanding of God as revealed in Jesus Christ.40 Thus, it is not surprising that a Christian discussion of this material was very appealing:41 among other things, the Christus patiens provides a specifically Christian interpretation of the Bacchae. Already in the second century AD, the pagan critic of the Christians, Celsus, cites in Origen, Against Celsus 2.34 a Jew who casts doubt on the godhead of Jesus Christ, pointing out that Jesus was not even able to free himself from the cross (see Mark 15:30–2). In the process, and by way of providing an ironically contrasting alternative, the Jew cites the words of the disguised Dionysus from Euripides, Bacchae 498: λύσει μ‎’ ὁ δ (p.150) αίμων αὐτός‎, ὅταν ἐγὼ ϑέλω‎ (‘The God himself will free me when I so desire’).42 In the same chapter Origen refers to Celsus’s criticism that Jesus Christ could not have been a true god, since he allowed himself to be abused without punishing the perpetrators, in contrast to Dionysus, whose adversary Pentheus met a grisly end. As a counterargument, Origen responds that Jesus’s real enemies—i.e. the Jews—had met a much worse fate, since they had subsequently been scattered across the whole world.43

Some further, formal aspects of classical Greek tragedy in comparison to the Christus patiens merit attention. With the exception of the anapaests in 1461–3, only the iambic trimeter is used. There is a chorus, but it does not use lyric meters.44 The overall composition of the Christus patiens does not follow that of classical tragedy; instead, a tripartite division focusing on Jesus’s Crucifixion (1–1133), death (1134–905), and resurrection (1906–2531) dominates the general structure of the poem.45 Another striking feature are the very long—relative to classical examples—narrative passages that can include reports from messengers, dogmatic observations, or lamenting monologues.46 This shows the influence of the post-classical Byzantine tradition following the poetics of Theophrastus.47 Finally, it is worth noting that more than three individuals may appear on stage at the same time, meaning that the author breaks with the traditional classical triad of actors in a tragedy.

Similarly to these formal and textual points, the nature of the reception of Euripidean verses in the cento is also varied, and sometimes more formal and sometimes more related to content. This will be demonstrated in the following in the light of the way in which verses from the Bacchae have been absorbed. Viewed in terms of technique, this is representative of the verse reception in the Christus patiens as a whole and Christian centos in general.

1. The briefest and most superficial form of verse adaptation is the ‘splinter technique’, i.e. individual verse fragments are used as building blocks to create new verses. However, the verse elements that are taken over are too short and too isolated to allow for the convincing establishment of a conscious (p.151) intertextual reference to the content of the source material. A few examples for this are: Chr. pat. 586 καὶ σπένδε καὶ κλέϊζε καὶ τόνδ‎’ εὐλόγει‎ (‘and pour him offerings and celebrate him and bless him’) from Euripides, Bacchae 313 καὶ σπένδε καὶ βάκχευε καὶ στέφου κάρα‎ (‘pour offerings to him and join the Bacchic dance and crown your head!’); Chr. pat. 1136 ἀτὰρ τόδ‎’ ἄλλο ϑαῦμα καὶ παρ‎’ ἐλπίδα‎ (‘But this is another marvel, and against hope’) from Euripides, Bacchae 248 ἀτὰρ τόδ‎’ ἄλλο ϑαῦμα‎, τὸν τερασκόπον‎ (‘But this is another strange business: the diviner’ [i.e. Teiresias dressed in fawnskin]); Chr. pat. 1712–13 τὸ μὲν σὸν εὐδόκιμον‎, εὐσεβὲς γέρον·‎ | ὁ δ‎’ ἐγγενής σοι λαὸς ἐκτίσει δίκην‎ (‘Your lot is favourable, reverend old man: but the people kin to you will pay a penalty’) from Euripides, Bacchae 1327 τὸ μὲν σὸν ἀλγῶ‎, Κάδμε· σὸς δ‎’ ἔχει δίκην‎ (‘I feel grief at your misfortune, Cadmus; your [i.e. grandson] received justice’).48 The reception of the prayer to God49 also belongs here, which is facilitated by the general tone of its formulation; this also applies, to a certain extent, to the reception of certain verse fragments or whole verses.50 Smaller grammatical or lexical alterations, some of which have purely pragmatic reasons,51 can be made when integrating verse fragments or whole verses.52

(p.152) 2. It is far more interesting and important for the cento’s interpretation if such alterations to the hypotext are made because they are necessary for contextual or ideological reasons, which in particular is the case with the substitutions of proper names. In this regard, it is possible to speak of recasting or contrafacture. In the following a few particularly significant verses and even entire verse clusters in which this phenomenon appears will be presented.

In Chr. pat. 193, Cadmus from Euripides, Bacchae 264 is replaced by Adam, who are both, as the progenitors of a human race, presented as worthy of veneration. In Chr. pat. 580 | ὅδ‎’ (‘he’, i.e. God) replaces the subject | Zεὺς‎ (‘Zeus’) from Euripides, Bacchae 291 in an otherwise identical verse: Zeus thwarts Hera’s plan to destroy the prematurely born Dionysus by hiding him; similarly, the Christian God thwarts the plan of the snake, Satan, to destroy humankind by himself appearing as a human on earth. In Chr. pat. 649, the messenger announces news from the Jews (‘Hebrews’), in Euripides, Bacchae 1029, from the Bacchants.53

In general, there is no fixed pattern to which the substitution of personal names must adhere, i.e. Dionysus does not always have to be equivalent to Jesus, nor Semele to Mary, etc. Hence, for example, the substitutions of the female Bacchants (Euripides, Bacchae 732–33, 1091)≈Galilean women (Chr. pat. 1810–12, 2013–17), (liberated) female Bacchants (Euripides, Bacchae 443–4, 445–7)≈(resurrected) dead (Chr. pat. 1928–9, 2073–5); Cadmus (Euripides, Bacchae 264, 1314–15)≈Adam (Chr. pat. 193, 1342–3), Cadmus (Euripides, Bacchae 1360–2)≈Judas (Chr. pat. 1695–7); Pentheus (Euripides, Bacchae 1111–13, 995)≈Judas (Chr. pat. 1430–2, 1437), Pentheus (Euripides, Bacchae 670, 776)≈Pontius Pilate (Chr. pat. 2222), Pentheus (Euripides, Bacchae 1226)≈Jesus (Chr. pat. 1455); Dionysus (Euripides, Bacchae 1095–7, 27, 22, 45–6, 779, 776–7, 1078)≈Jesus (Chr. pat. 666–8, 1550, 1564, 1570–1, 2228, 2244–5, 2256), Dionysus (Euripides, Bacchae 82)≈God (Chr. pat. 1144).

The same is also true of the use of groups of verses, which are not always—in analogy to the speaker of the original text—uttered by the same individuals. This means that words spoken by ‘bad’ people in the original may be spoken by ‘good’ people in the Christus patiens and recontextualized from a negative to a positive statement. At the root of this is the view that the canonical texts of the pagan tragedians may implicitly reveal truths about salvation, even, as it (p.153) were, contrary to their original intention or contrary to first impressions.54 ‘Revealing’ this through skilful adaptation is precisely the core achievement of the cento poet’s technique. In view of this and bearing in mind the intention of the genre, the objection made by some scholars that it is, for example, tasteless of the poet to clothe Mary in the ‘rags’ of a Medea or Hekabe55 can no longer be upheld.56

In this interpretation, the cento poet takes statements without altering their substance and simply links them to Christian individuals rather than the pagan ones of the original text. One could call this a weaker form of transferral, i.e. a straightforward, direct Christianization of pagan thought, employing the method of ‘usurping’ existing, prefabricated conceptual or poetic elements for Christian statements.57 Depending on the viewpoint of the various readers, the Christian message is thereby given either a secondary, derivative, falsifying character (‘old wine in new skins’), or the function of revealing a long-hidden truth that has not yet been recognized by pagan culture, or, because of the close interaction with pagan culture, is regarded as having been falsified through the influence of that pagan culture, what has been termed ‘Hellenization of Christianity’.58 Here belongs, above all, the notion of the transformation of a god into a human, used as a kind of leitmotif, in which Euripides, Bacchae 4 μορφὴν δ‎’ ἀμείψας ἐκ ϑεοῦ βροτησίαν‎ (‘having changed from God to human shape’) repeatedly taken up (Chr. pat. 1533, 1536, 1543, 1758, 2395, 2405, 2574).59

3. However, a stronger form of transferral can also be discerned: it signifies a qualitative break with the original hypotext, since the Christian message is fundamentally entirely different.

Euripides, Bacchae 1213–15 πηκτῶν πρὸς οἴκους κλιμάκων προσαμβάσεις‎, | ὡς πασσαλεύσῃ κρᾶτα τριγλύφοις τόδε‎ | λέοντος‎, ὃν πάρειμι ϑηράσασ‎’ ἐγώ‎ (‘you will climb on the steps of a well-fastened ladder up to the house so that you can (p.154) nail to the triglyphs this head of a lion I hunted down before coming here’) is taken up in Chr. pat. 1263–5 ἔμβαινε πηκτὰς κλίμακος πρὸς ἐμβάσεις‎, | ἐκπασσαλεύσων διγλύφου δοκοῦ δέμας‎ | λέοντος ὃν γέγηϑε ϑηράσας λαός‎ (‘climb up on the well-fastened steps of the ladder in order to remove the nails from the body of the lion on the diglyphic beam, the lion which the people rejoice to have hunted down’). In Euripides, Agave, the daughter of Cadmus, has killed her own son, whom in her bacchanalian frenzy she mistook for a lion. She triumphantly brings the severed head of this supposed lion to Thebes. She wants her son Pentheus (whose head it really is) to climb up a ladder and nail this ‘lion’s head’ to the ‘three-pronged’ (τρίγλυφος‎) battlements of the palace. In contrast, in Christus patiens, Joseph asks Nicodemus to climb up a ladder and take down the body of the ‘lion’,60 whose capture gives so much pleasure to the Jewish people, from the ‘two-pronged’ (δίγλυφος‎) beam of the cross.

First, the obvious parallels stand out: the captured lion, the insertion or removal of the nails, the three or two prongs of the palace battlements or the cross respectively. Even the delusion concerning the identity of the lion is comparable: Agave does not yet recognize that she has not, in truth, killed a wild animal, but her own son. The Jews believe that they have subjected a treacherous person, who pretends to be a god, to a just punishment. The lifting of the delusion is, however, fundamentally different: for Agave, recognition of her error signifies the revelation of the vengeance of the unjustly treated Dionysus, hence her words in 1296: Διόνυσος ἡμᾶς ὤλεσ‎’, ἄρτι μανϑάνω‎ (‘Dionysus destroyed us: now I realize this’). The consequence is the ruin of the royal family; the event, above all the killing of Pentheus, cannot be reversed anymore. Similarly, the Jews will be punished for the sacrilege of having delivered Jesus onto the cross. Their deed, however, is not ultimately successful, for the captured ‘lion’, Jesus Christ, will rise again. With that, any criminal act, and especially that of violent death, is deprived of its finality; the consequence is the revelation of the saving will of God, not the vengeance of a god—thus, tragedy in the full classical sense of the word has ceased to be possible.61 This will also become clear in the following.

The verses in Chr. pat. 1522, 1524–6: λήψῃ δὲ νεκρούς‎, οὐ σὺ ληφϑήσῃ νεκροῖς‎, | (…). | Μόνος γὰρ ἀνὴρ ταῦτα ϑαρρῶν ἱκάνεις‎, | μόνος σὺ φύσεως ὑπερκάμνεις βροτῶν‎. | Ἔσχον δ‎’ ἀγῶνες‎, οἵ σ‎’ ἔμιμνον‎, νῦν τέλος‎ (‘But you will conquer the dead; you will not be conquered by the dead […] For you are the only man sufficiently strong and courageous for this, and you alone labour on behalf of the nature of mortals. You took upon yourself struggles, they awaited you, now it is their end’) draw on Euripides, Bacchae 960, 962–4: λήψει δ‎’ ἴσως σφᾶς‎, ἢν σὺ μὴ ληφϑῇς πάρος‎. | (…) | Μόνος γάρ εἰμ‎’ αὐτῶν ἀνὴρ τολμῶν τόδε‎. (p.155) | Μόνος σὺ πόλεως τῆσδ‎’ ὑπερκάμνεις‎, μόνος·‎ | τοιγάρ σ‎’ ἀγῶνες ἀναμένουσιν οὓς ἐχρῆν‎ (‘Maybe you will catch them, unless you’re caught first! […] For I alone among them am a man brave enough to do this. You alone labour on behalf of this city, you alone: for struggles await you that were necessary’). In Euripides, Pentheus believes himself to be the only man suitable to take up the fight with the god Dionysus, even if tough struggles await him. In verses 967–70, Dionysus hints to Pentheus with tragic irony (which is incomprehensible to Pentheus as he lacks divine knowledge) that he will be the cause of something ‘of significance’ (ἐπίσημον‎) to all, which in fact indicates his gruesome demise. In Christus patiens, Mary praises her son, saying that he is the only one of all humans capable of suffering and battling for humanity, although here the emphasis lies on the temporal limitation of these struggles and the ultimate victory of Jesus Christ (1526–8).

The problem of human failure—as a result of the limited nature of humans—to recognize the will of the gods or the course of fate, with its catastrophic consequences for humans, is a central idea in Greek classical tragedy. Within this world-view, ‘redemption’ (in so far as this term can be applied here at all) means that humans must improve their powers of perception in order that they may avoid recognizing the true state of affairs too late. Although this kind of recognition implies moral improvement and enhanced self-awareness, it forms nevertheless a tragic dilemma for human beings, as the improvement comes too late and in vain, because they are already caught up in the causality of doom. The implicit conclusion is here that humans must conform as best they can to the existing universal law or coordinate system of reality, as represented by the gods. The coordinate system itself may be challenged—which happens especially in Euripides’ tragedies and their portrayal of the gods—but ultimately remains unchanged. The God of the Christians, by contrast, undertakes in Jesus Christ to conform to this system himself by way of a voluntary sacrifice, and in that way undermines it. The task of humans who have been redeemed in this manner is no longer to recognize reality and then conform to it one way or the other, but rather, to recognize the divine will of salvation and then to ask for forgiveness; instead of conforming to reality or a system of coordinates, humanity is thus liberated from it.

This deconstruction of the pagan concept of tragedy and its underlying world-view finds its final climax in the personal closing prayer at the end of the Christus patiens, which an anonymous speaker first directs to Jesus Christ (2532–71), and then to Mary (2572–2602), as the two guarantors of Christian salvation. In verses 2557–66, the cento poet draws on Euripides, Bacchae 1344–8 and 1118–21. At the end of the Bacchae, faced with ruin, Cadmus asks Dionysus in vain to apply mercy (1344 Διόνυσε‎, λισσόμεσϑά σ‎’, ἠδικήκαμεν‎, ‘Dionysus, we entreat your mercy: we have done injustice’). Dionysus refuses to grant mercy, referring to the unchangeable necessity of the divine order as ordained by Zeus and of fate (1349 πάλαι τάδε Ζεὺς οὑμὸς ἐπένευσεν π (p.156) ατήρ‎, ‘long ago Zeus my father ordained this’, and 1351 τί δῆτα μέλλεϑ‎’ ἅπερ ἀναγκαίως ἔχει‎: ‘Why then do you hesitate to carry out what is ordained?’). In doing so, Dionysus implicitly admits that, in their anger and lust for vengeance, the gods are the same as humans (see Euripides, Bacchae 1348–9), an idea that is diametrically opposed to the Christian notion of God. This is explicitly emphasized at the end of Christus patiens. In Chr. pat. 2557, too, the supplicant asks for forgiveness, using exactly the same verses as Agave (1344), only replacing Διόνυσε‎ with ῥύσιε‎ (‘saviour’). In verse 2560, the supplicant admits that ‘we recognized too late, and when it was necessary, we did not know’ (ὄψ‎’ ἐμάϑομεν‎, ὅτ‎’ ἐχρῆν‎, οὐκ εἰδότεϛ‎). The supplicant himself here admits failure before God, which in Euripides had been a reproach by the god Dionysus directed at the family of Cadmus (1345 ὄψ‎’ | ἐμάϑεϑ‎’ ἡμᾶϛ‎, ὅτε δὲ χρῆν‎, οὐκ ᾔδετε‎, ‘You recognized us too late, and yet when it was necessary, you did not know it’); here human repentance can hardly rely on divine forgiveness.

The starkest contrast to the original hypotext can be found in the two following verses, i.e. Chr. pat. 2562–3 (Γινώσκομεν σφάλματα‎, σὺ δὲ παρόρα·‎ | ἴδμεν δέ σ‎’ ὀργὰν οὐχ ὁμοιοῦσϑαι βροτοῖς‎, ‘We now know our errors, but you disregard them. And we know you are not like mortals in your anger’), which radically transforms Euripides, Bacchae 1346 and 1348. The acknowledgement of guilt is followed in Chr. pat. 2562b by the certainty that God will forgive, in contrast to the statement of Cadmus, who in 1346 (ἐγνώκαμεν ταῦτ‎’· ἀλλ‎’ ἐπεξέρχῃ λίαν‎, ‘we know that now; but your vengeance passes bounds’) denounces the excess and implacability of divine revenge, which is then confirmed by Dionysus in 1349. Accordingly, Cadmus requests in 1348 in vain that the wrath of the gods should be less than that of humans (ὀργὰς πρέπει ϑεοὺς οὐχ ὁμοιοῦσϑαι βροτοῖς‎, ‘Gods ought not to be like mortals in their anger)’, while the Christian supplicant is already assured of this (2563 ἴδμεν δέ‎…, ‘we know…’), which is then developed in the rest of the prayer. In this way, the end of the Bacchae is inverted: instead of revenge, there is the possibility of forgiveness and salvation. So the Christus patiens confirms George Steiner’s claim that Christianity is incompatible with classical tragic drama. But instead of abolishing and thus discontinuing this literary genre, the Christus patiens opens up and manifests possibilities of transformation, deepened readings of a classical original, and the establishment of clear boundaries where Christianity offers something radically different from the old world-view, in what one might call a Christian anti-tragedy. The advantages of a transformational rather than a discontinuing approach are the connection to and exploitation of a powerful cultural and educational tradition that generated identities and its creative opening up for new perspectives.62

(p.157) In conclusion, we can note that it is characteristic of our cento to handle the classical hypotexts in a flexible and diverse way, permitting straightforward borrowings, smaller alterations to the text, and even the transformation of negative statements into positive ones. Beyond simple play with the form, the Christus patiens aims to interpret and correct its source material. This implies that there is no clearly fixed personnel, i.e. for example, Dionysus does not always have to be identified as Jesus Christ, or Agave as Mary. At the basis of this lies the belief that the canonical texts of the pagan tragedians implicitly bear witness to the truth of salvation, even though this is contrary to their original intention and their surface meaning. Revealing this through skilful textual adaptation is precisely the technical and interpretative achievement of the cento poet. The borrowing of existing textual structures for a new, Christian message at a literary level parallels at a soteriological level the action of the Christian God, who adopts the human condition to carry out his act of redemption through suffering and mortality, i.e. he also draws on already existing, ‘human structures’. In analogy to this, the Christian activity of writing centos may be understood from the perspective of the history of salvation as a method by which the history of salvation is continued and turned into reality by means of ‘doing literature’. (p.158)


I would like to thank P. E. Easterling (Cambridge) for our useful discussion, as well as my audiences at the Faculty of Philosophy of the University of Constance (Germany), at the International Patristics Conference in Oxford (1995), and at the Classics Research Seminar at the University of St Andrews (Scotland), for their constructive comments when I presented earlier versions of this chapter.

(1) The English translation is taken, slightly modified, from Lattimore (2011), 270.

(2) He bears the epithet ὁ λυτικός‎ and should not be confused with Sosibios of Lakedaimon; see Laqueur (1927), 1149; Peremans et al. (1968), No. 16886, with reference to Jacoby (192368), No. 595 (commentary, 635).

(3) Hainsworth, (1993), 293 explains the text with the scholia as a conventional compliment to Nestor (where, in contrast to any other uses, it is perhaps not used entirely appropriately) that does not involve a specific comparison with another particular person.

(4) i.e. transmutatio or inversio, see Quintilian, Institutio Oratoria 1.5.39–40; 9.4.89.

(5) This has not yet been taken up by research on the theory and practice of the cento, most likely due to the inaccessibility of the relevant work, i.e. Daube (1953), 27–34, and (1970), 53–9 (reprinted in Cohen/Simon (1991), 1252–62).

(6) See the scholia to Aristophanes, Frogs 1219: διαβάλλει δὲ τὴν ὁμοειδίαν τῶν εἰσβολῶν τῶν δραμάτων‎; likewise also Radermacher (1954), 310; Stanford (1958), 174; and Sicking (1962), 81. The whole passage is the subject of controversial debate in scholarship, for which compare the sensibly balanced overview in Dover (1993), 337–9. The phrase ‘He lost a little bottle’ is interpreted in three ways by scholars: (1) ‘innocently’ as an expression for a mishap, much as somebody today might leave their umbrella somewhere by accident; (2) as a sexual metaphor for ‘to lose erectile function’; (3) as a literary metaphor for ‘to lose the force of tragic expression’. All three variants are difficult to verify. The most convincing approach seems to me to view it as a satirical technique that aims to expose the syntactical stereotype and textual superficiality of Euripidean verse. There is further evidence of this technique in the fifth century BC Hegemon of Thasos, who sometimes appended a stereotypical end clausula (τὸ πέρδικος σκέλος‎) in his parodies—see Bain (1985), 37 n. 22. In the research literature right up to Dover there is unfortunately no detailed discussion of whether such criticism of the grammatical and syntactical stereotypicality of Euripidean language is justified from a modern scholarly perspective (only Stanford (1958), 174 emphasizes that the surviving works by Euripides do otherwise not confirm this Aristophanic criticism) or whether such criticism was ever expressed explicitly elsewhere in ancient times.

(7) For this reason, the claim that Aristophanes, Frogs 1285–95 can be deemed the earliest surviving cento is not tenable—see Mansfeld (1992), 153, n. 1. Viewed purely schematically, it is indeed a cento-like joining of Aeschylean verses (Frogs 1262 ξυντέμνειν τὰ μέλη εἰς ἕν‎), but the satirical target is less textual than it is formal: ‘What is evidently satirized here is not only Aeschylos’s fondness for dactylic rhythm but also his use of refrains, which sometimes consist of only a few words’ (Dover (1993), 345; similarly Radermacher (1954), 315–16). It is more appropriate to speak of parody in relation to this passage—see Dane (1988), 56.

(8) See, on the cento in general, Kunzmann/Hoch (1994), 148–57, although the focus of the section on antiquity is on Latin examples. On the Greek Christian centos of late antiquity, see Hunger (1978), 98–107; Smolak (1979), 29–49; (1984).

(9) See Irenaeus, Adversus Haereses 1.9.4, which reports that students entertain themselves and others by combining lines from the Iliad and the Odyssey, and telling an entirely new story. They were able to fool simpler minds with this, because they believed these poems to be authentically Homeric and only the expert would notice that the verses had been taken from what was originally an entirely different context. In his argument, Irenaeus uses this technique of cento composition as an analogy for the similarly arbitrary practice of the heretics, who use Bible quotations out of context for their ‘new’, i.e. heretical, statements, which no longer have anything to do with the true message of the Bible; on this, see Wilken (1967). In the eyes of Christianity, the cento genre thus became negatively connoted from early on; the arbitrariness of the cento’s composition was linked with the danger of misinterpreting the Christian message. For this reason, Tertullian, praescr. 39.4–6 and Jerome, Epistle 53.7 vehemently opposed Christian centos composed of pagan poems, for fear that this might also provide justification for unorthodox ‘centos’. This hostile attitude amongst cento theoreticians did not diminish the admiration for centos written by Christian authors; e.g. both Christine de Pizan in La citédes dames 1.29, and Boccaccio in De claris mulieribus 67, express enthusiasm for the Cento of the fourth-century poet Proba; for the fascinating reception of Proba as prophetic poet and of her Cento, see Schottenius (2015).

(10) I use the terms hypotext and hypertext after Genette (1982), who, astonishingly, does not himself examine the cento genre. For the theoretical perspective, see Verweyen/Witting (1991).

(11) An extreme example is the Cento nuptialis composed by Ausonius (fourth century) of Vergilian verses. In his foreword to the cento, Ausonius warns the reader of this ‘dishonouring’ of Vergil: piget enim Vergiliani carminis dignitatem tam ioculari dehonestasse materia, on which, see Herzog (1975), 3–13; and in this volume Chapter 4.

(12) On this, see Schelkle (1954), 972 with examples.

(13) See Proba, Cento, praef. 3–4 Maro mutatus in melius and Cento 23 Vergilium cecinisse loquar pia munera Christi.

(14) Müller-Sievers (1989), 230. David Daube sees the method of cento composition as an analogy to the compilation of legal texts commissioned by Justinian, as this also links the desire to preserve canonical material with the process of selecting, cutting, and altering; see Daube (1959), also in Cohen/Simon (1991), 789–897, here 893–7, and his (1971), also in Cohen/Simon (1991), 1263–6.

(15) A striking exception, to my knowledge the only one, are two centos stitched together from verses of a late antique, anti-heretical didactic poem, the so-called Carmen adversus Marcionitas—see Pollmann (1991), 13–14.

(16) This title does not appear in the manuscripts, which use various paraphrasing sentences as titles. Rather, it has become the title in general usage since the Editio princeps by Anton Bladus (Rome, 1542).

(17) In descending order, the tragedies by Euripides used are (see the index in Tuilier (1969), 343–55): Medea, Bacchae, Hippolytus (the Pseudo-Euripidean) Rhesus, Orestes, Troades, Hecuba, Phoenissae, and only one verse respectively from Helena, Alcestis, Andromache, Iphigenia in Aulis, and Iphigenia in Tauris; in addition, a few verses respectively from works by Aeschylus (Agamemnon, Prometheus; see Somers (2010)) and Lycophron (Cassandra). The assumption that Iliad 3.342 δεινὸν δερκόμενοι ϑάμβος δ‎’ ἔχεν εἰσορόωντας‎ is the basis for Chr. pat. 2127 ϑάμβος μ‎’ ἔχει βλέπουσαν ἀστραπὴν ϑέας‎ does not seem to me to be verifiable, as there is insufficient concordance between the two verses. In principle, I believe this is also the case for a considerable part of the other supposed verse attributions. Often it is presumably better instead to speak of a fundamental, classical ‘Euripidization’ in the language of the cento poet, analogous to when we speak of ‘Goethe’s German’ or Shakespearean English today.

(18) On this, see the research overview in Ellissen (1855), xx–li; Tuilier (1969), 118–21, and Trisoglio (1974), 351–423, which covers studies from 1571 to 1972. More recently, Most (2008) has also dismissed the manuscript tradition as valid evidence for the authorship of Gregory of Nazianzus.

(19) In particular, Apollinaris of Laodicaea, since Dräseke (1884), 657–704.

(20) Since Brambs (1884), 29–38; see also Hilberg (1886), 282–314.

(21) So De Aldama (1972), 417–23, which Garzya (1984), 237, n. 2, relativizes by correctly remarking that such dogmatic–systematic analyses of the tragedy genre and the dramatic material are not particularly informative. However, the rather more anthropocentric than hieratic demarcation of Mary in the Christus patiens makes it much more likely to be dated after iconoclasm, see Puchner (1992), 136–43, who, in 128–33, also considers it probable that the Christus patiens was influenced by motifs in post-iconoclastic iconography. On the theological content of the Christus patiens, see also Lanowski/Starowieyski (1995; Polish translation with introduction and commentary), 17–19.

(22) Garzya (1984), 237–40, argues that the Christus patiens manuscripts contain characteristic corruptions which can be traced before the transliterative turn at the beginning of the ninth century. These errors, consisting of majuscule errors, also appear in verses not taken from Euripides or else do not appear in the manuscript tradition of the Euripides verses in question. From this Garzya concludes that the earliest historical limit is pre-ninth century, and with that he considers the work’s chronological location in the fourth/fifth centuries to be confirmed, although he leaves the question of the author open (239–40). Garzya’s argument, at first glance very appealing, ignores or at least overlooks the fact that there were hybrid forms of minuscules and uncials in the Byzantine writing traditions from the ninth to the twelfth centuries, which were termed the mixed or middle minuscules; see Maas (1980), esp. 49–52, and Granstrem (1980), esp. 82–4. This writing style has two types for every letter, see Maas (1980), 50, who denies that this is the case for ϑ‎, ι‎, ο‎, ρ‎, τ‎, and χ‎, while according to Hunger (1980), 60–72 both ϑ‎ and τ‎ can appear as uncials and minuscule as well. This reduces the significance of Garzya’s examples of transcription errors for dating the work prior to the ninth century. For illustrations of these mixed minuscules, see, for example, Barbour (1981), figs 80 (fourteenth century) and 81 (twelfth century), as well as, for general information, xxv–xxvi, where she emphasizes that developments in this field never take place in a linear or regionally uniform manner, which is why criteria such as this are not sufficiently reliable for dating or geographic localization. Garzya (1984), 239, n. 6 also mentions examples in the minuscule for scribal errors. A second article of his also fails to overcome this problem—see Garzya (1989), 110–13.

(23) This concerns Chr. pat. 454–60 in relation to the first strophe of Hymn 19 by Romanos (Lament of Mary before the Cross). Generally, motivic similarities between two poets are unreliable points of reference in determining questions of precedence; see for our problem Momigliano (1933), 47–51, and Grosdidier de Matons (1973), 364, who place Romanos before the Christus patiens, while Tuilier (1969), 39–47 believes Romanos to be influenced by the Christus patiens. Finally, Puchner (1992), 135 with n. 179, argues correctly against the latter stance and, on p. 137, comes to the balanced conclusion that the sections reminiscent of Romanos do not provide sufficient proof for direct dependency. This line is e.g. followed by Eriksen (2013), 52–3.

(24) Referring here to Hebrews 5:7–8: ‘During the days of Jesus’ life on earth, he offered up prayers and petitions with fervent cries and tears to the one who could save him from death, and he was heard because of his reverent submission. Son though he was, he learned obedience from what he suffered…’. Similarly, Gregory of Nyssa, Beatitudes 2 (PG 44.1209 b–d, here c) explains that Christ’s parables are partly arranged like tragedies. He had to make use of this literary tool since the message of God’s kingdom surpassed human understanding.

(25) The English translation is taken from C. G. Browne and J. E. Swallow (1894).

(26) See Trisoglio (1978), 121–4, who also emphasizes the sublime, calm majesty of Christ’s words, which he maintains do not contain traces either of the Passion nor of doubt; despite this, Christ’s person is not portrayed superficially, but rather has even greater weight than any other figure in the drama. Trisoglio does not mention whether theological categories and/or poetic principles lie behind this conception of the figure of Jesus Christ and, if so, which ones. Lacore (2002) offers the thoughtful observation that Christ’s suffering and glory is communicated in a non-verbal fashion through the contemplation of his body. She concludes convincingly that thus both the humanity and the divinity of Christ are implicitly confirmed.

(27) Contrary to Cottas (1931); Tuilier (1969), 27–74; Trisoglio (1979), 13–16; Trisoglio (2002); Salanitro (2003); Centanni (2007).

(28) The Greek in verses 1 ἀκούσας‎, 2 κλύειν‎, and again in 3 ἄκουε‎ suggests a performance context, i.e. either a recitation or a dramatic performance (Puchner (1992), argues against the latter on the basis of too many technical dramatic contradictions). The personal prayer at the end of the drama (2532–602) should also be viewed in this context.

(29) Theocharides (1940). Even if this is correct, the fundamentally ambivalent position of the Cappadocians with regard to classical literature should be highlighted, as in Pelikan (1993), 17–18.

(30) Hose (2004), 30–6.

(31) Thus Michael Psellus (eleventh century) compares the quantitative meter of Georgios Pisides (seventh century) with Euripides, see Dyck (1986), esp. 36–7. On Euripides in gnomologies (e.g. in Codex Marcianus Graecus 507, fol. 112v–41), see Meschini (19734).

(32) Friesen (2015), 253 is subtle, but not quite as nuanced as the above on this point.

(33) e.g. Euripides, Bacchae, 692 αἱ δ‎’ ἀποβαλοῦσαι ϑαλερὸν ὀμμάτων ὕπνον‎ (fourteen syllables) compared to the Chr. pat. 2007 αἱ δ‎’ ἀποβαλοῦσαι τὸν ὕπνον ὀμμάτων‎ (twelve syllables). This was already noted by Brambs (1884), 32.

(34) The prosody and meter of Gregory of Nazianzus have not yet been exhaustively researched. However, compare the concise summaries in Meier (1989), 18–22, and Oberhaus (1991), 26–36, on his iambic tetrameter. On the meter of the Christus patiens, see Krumbacher (1897), 748, on the development of the prosodic mixed forms in the Byzantine Empire (648–9), as well as, in particular, the most significant study that continues to be relevant today, by Maas (1903).

(35) See the good summary in Dölger (1948), 13 and 53–5; Maas (1903).

(36) Hörandner (1988). It is for this reason alone that the argument put forward by MacCoull (1985) that the composer of the Christus patiens was active in Egypt in the fifth/sixth centuries must be rejected. She supports her argument by pointing to the appearance of rare words in the non-Euripidean verses, which can also be found in Egyptian texts of the fifth/sixth centuries.

(37) Vakonakis (2011), 97–103; Alexopoulou (2013); Friesen (2015), 252 with n. 3.

(38) In the following translations from the Christus patiens we will follow, occasionally with slight modifications, Fishbone (2002).

(39) This is unique in Greek classical tragedy, see Hose (1991), 334–5. However, see for gods in disguise on stage (probably) Aeschylus, Edoni fr. 61.

(40) Evans (1988), 145–73 (‘Dionysos and Christ’) presents more analogies between the Dionysian and Christian concepts. However, he discusses what he terms the ‘Christian myth’ (Christianity does not conceive of its message as a myth, but rather as a historical event!) only vaguely and without elucidating it any further, and posits—also without further explanation—that this ‘Christian myth’ took up elements of the Dionysus cult. What is remarkable is that there is evidence of reciprocal influences between the cult of Dionysus and Christian motifs in late antique art—see Daszewski (1985); Hamdorf (1986), 37–9, and Eisler (1925), who, on 239, n. 8, states that Euripides’ Bacchae is deliberately emulated in the Christus patiens.

(41) For a superb overview, see Friesen (2015).

(42) In the translations of the Euripidean passages, we follow, occasionally with slight modifications, Kovacs (2014). This verse is not reproduced in the Christus patiens.

(43) This anti-Jewish bias with its heavy emphasis on God’s ‘revenge’ on the Jews (as well as on Judas Iscariot) is also found in the Christus patiens (e.g. 426–32, 738–43, 790–5, 1049–62, 1407–10, 1712–18); on this, see esp. Puchner (1992), 136–43. However, the Christus patiens ultimately emphasizes the idea of repentance, mercy, and redemption, which represents a central difference to any pagan concepts of the gods (on this, see also pp. 155–7 in this chapter).

(44) On the function of the chorus within the plot of the Christus patiens, and in comparison to classical tragedy—see Trisoglio (1979).

(45) Tuilier (1969), 20, suggests, quite plausibly, that the drama which is in fact rather too long to be staged in a single performance, may have been presented over the three days of Easter, most likely in the form of a recitation (perhaps involving parts for several speakers) rather than a complete dramatic production—see Puchner (1992).

(46) Trisoglio (1980).

(47) Dostálová (1982).

(48) This fragmentary absorption of partial verses also allows elements to be loaned from Euripidean chorus songs in which the lyric meter is broken up. Examples of this are Chr. pat. 1051, 1099, 1599–601, 1614, 1801–3. As already mentioned, the Christus patiens employs only iambic trimeters, with the exception of the anapaests in 1461–3.

(49) Chr. pat. 1535 ὦναξ‎, Ἄναξ ἄφϑιτε‎, σὺ Θεὸς μένων‎ from Euripides, Bacchae 1031 ὦναξ Βρόμιε· ϑεὸς φαίνει μέγας‎, from which Chr. pat. 2100 and Chr. pat. 2542 with variations also draw.

(50) e.g. Chr. pat. 228 κἀγὼ προφήτης τῶν λόγων γενήσομαι‎ from Euripides, Bacchae 211 ἐγὼ προφήτης σοι λόγων γενήσομαι‎, or Chr. pat. 2212–13 ἥκω φράσαι σοι καὶ πόλει πολλὰ ξένα‎, | ὡς καινὰ πάντα ϑαυμάτων τ‎’ ἐπάξια‎ from Euripides, Bacchae 666–7 ἥκω φράσαι σοι καὶ πόλει χρῄζων‎, ἄναξ‎, | ὡς δεινὰ δρῶσι ϑαυμάτων τε κρείσσονα‎ and 716 ὡς δεινὰ δρῶσι ϑαυμάτων τ‎’ ἐπάξια‎. The example of the stereotypical description of the rising sun is somewhat different: it is taken from Euripides, Bacchae 678–9…ἡνίχ‎’ ἥλιος‎ | ἀκτῖνας ἐξίησι ϑερμαίνων χϑόνα‎, demarcating the Bacchants setting off in the early morning, in order to describe the break of dawn on the Easter morning in Chr. pat. 1845–6 γῆς τ‎’ ἀναδραμὼν ἥλιος φαεσφόρος‎, | ἀκτῖνας ἐξίησι ϑερμαίνων χϑόνα‎, thus lending a considerably different theological significance to the much more ornamental phrase in Euripides.

(51) The intention of this alteration is sometimes clearly recognizable: the respective verse has to be synchronized with its new context—see Daube (1971), 1266 with reference to Augustine, Epistle 137.12 where Augustine cites Vergil, Eclogues 4.13–14 and in doing so amends the original te duce to quo duce. In some instances, however, it is not possible to determine with absolute certainty whether the Christus patiens does not in fact testify to another, independent interpretation of the Euripidean text. On this, see Giudice Rizzo (1977), 16, who establishes the four main possibilities for the relationship of the two traditions: (1) the Christus patiens represents the correct reading of the Bacchae; (2) the textual reading represented in the Christus patiens is better than the corrupt version in the Bacchae; (3) the different readings in the Bacchae and in the Christus patiens is owed to the cento technique of partially altering verses and is thus of no textual-critical value for the Bacchae; (4) the Christus patiens offers in fact an inferior reading than the Bacchae manuscript copies.

(52) So, for example, Chr. pat. 1053–6 is identical with Euripides, Bacchae 1259–62, except 1054b εἰ δ‎’ ἕως τέλους‎ instead of 1260b εἰ δὲ διὰ τέλους‎, and 1056a εὐτυχοῦντες‎ (Mary addressing the Jews) instead of 1262a εὐτυχοῦσαι‎ (denoting the young Theban women as Bacchants). Chr. pat. 1685 is identical with Euripides, Bacchae 1340 with the exception of the final word: ἔφη‎ (John telling the words of Jesus) instead of λέγω‎ (Dionysus quoting himself), although both cases concern the fact that they both have an immortal father.

(53) The Bacchants as a group of people must always be renamed, so they cannot be considered metaphors for the followers of Christ, see Chr. pat. 1549 taken from Euripides, Bacchae 472, as well as Chr. pat. 1568 from Euripides, Bacchae 40.

(54) Dostálová (1982), 80 seems to me to express a good argument, referring to Eustathius, De simulatione p. 40 (ed. Tafel), who emphasizes the basic ethical neutrality of words, a thought ultimately going back to the Stoics. This naturally makes an unproblematic reception of verses or statements into an interpretative context that is contrary to their original meaning and context much simpler.

(55) Krumbacher (1897), 747.

(56) In this regard, Kyriakidis (1992) has a more measured approach in his analysis of Proba’s Cento by undertaking an interpretative comparison between the original context and the new context of the cento; see also Chapter 4 in this volume, and Schottenius (2015).

(57) This aspect is heavily emphasized by Evans (1988), 145–73, who speaks of the acquisition of Dionysian elements by Christian mythology. The argument that will be given in the following here will refine this position by making a number of distinctions.

(58) On this term and its implications, see Meijering (1993); Markschies (2011).

(59) Here, the very rare appearance of dogmatic technical terms (for example, ἕνωσις‎, κρᾶσις‎, ὑπόστασις‎) should be noted, which is connected with the work’s genre as well as its classical poetical mode of speech. The emphasis on the virgin birth and the story of the Fall and God’s resulting act of redemption are incorporated into this mode of non-dogmatic speech. The expression ἀμείβω‎ from Euripides, Bacchae, 4 is likewise not a dogmatic technical term; see also Friesen (2015), 256f.

(60) The Lion of Juda is meant here, i.e. Jesus Christ. The Old Testament image from Genesis 9 is transferred to Jesus Christ from Revelation 5:5 onwards.

(61) Puchner (1992), 124; Trisoglio (1995), 356.

(62) See also the fine analysis and observation in Friesen (2015), 253–60, who arrives at similar conclusions but is not as pronounced as to the fundamental opposition of world-views in hypotext and hypertext.