RAPHAEL SCHWITTER Rheinische Friedrich-Wilhelms-Universität, Bonn A “Roman” Wedding in Vandal Africa Triangular Intertextuality and Generic Appropriation in Luxurius’ Epithalamium Fridi (Anthologia latina 18 Riese2) ABSTRACT The Epithalamium Fridi is a sixth-century Virgilian cento that commemorates the marriage of the Vandal noble Fridus with his unnamed bride. Its author, the African poet Luxurius, engages in versatile poetic play fusing Virgil with multiple epithalamial models such as Statius, Claudian, and Ausonius. Through the dynamics of triangular intertextuality the centonist is able to strengthen the wedding poem’s generic bonds and to connect himself and his work firmly to the classical Roman tradition. At the same time, echoes of distinctive African idiosyncrasies as prefigured by Dracontius highlight the hybrid character of sixth-century Romano-Vandal elite culture and its celebration of what appears to be a distinctive African Romanness. KEYWORDS: Intertextuality, Cento, Virgil, Epithalamium, Anthologia Latina, Romanness, Vandal Africa The evolution and transformation of ethnic, political and cultural identities in the post-imperial West is the subject of a growing field of scholarship dedicated to an era still widely seen as merely interstitial.1 Along with archaeological An early draft of this paper was presented at a workshop in Prague in April 2017. I thank the organizers, Martin Bažil and Daniel Vallat, and the participants, particularly Marie Okáčová, Marcos Carmignani, Maria Teresa Galli, María Luisa La Fico Guzzo and Franca Ela Consolino for their valuable feedback. I am grateful to David van Schoor for his diligent proofreading of the final article, and to the editors and referees at SLA for their extremely helpful responses and advice. . A few notable titles among many: Samuel J. T. Barnish, “Transformation and Survival in the Western Senatorial Aristocracy, C. A. D. –,” in Papers of the British School at Rome  (): –; Geoffrey Greatrex, “Roman Identity in the Sixth Century,” in Ethnicity and Culture in Late Antiquity, ed. S. Mitchell and G. Greatrex (London: Duckworth, ), –; Ralph W. Mathisen and Danuta Shanzer, ed., Romans, Barbarians, and the Transformation of the Roman World: Cultural Interaction and the Creation of Identity in Late Antiquity (Farnham: Ashgate, ); Walter Pohl, Clemens Gantner, Cinzia Grifoni, and Marianne Pollheimer-Mohaupt, ed., Transformations of Romanness: Early Medieval Regions and Identities (Berlin: de Gruyter, ); for Africa see the excellent study of Jonathan Conant, Staying Roman. Conquest and Identity in Africa and the Mediterranean, – (New York: Cambridge University Press, ); for Italy, Patrick Studies in Late Antiquity, Vol. , Number , pps. –. electronic ISSN -. ©  by the Regents of the University of California. All rights reserved. Please direct all requests for permission to photocopy or reproduce article content through the University of California Press’s Reprints and Permissions web page, http://sla.ucpress.edu/content/permissions. DOI: https://doi.org/./sla..... 114 evidence, the literary production of this period is of particular value, helping us to map continuities as well as ruptures and shifts in terms of culture and identity after the Empire’s rule had faded away. While their literary pursuits, especially in poetry and letters, clearly seem to suggest that under barbarian rule local élites in Italy, Southern Gaul, Spain, and North Africa were deliberately negotiating their “Romanness” by upholding the lore and values of the traditional literary canon, the actual extent and impact of this high-status cultural endeavor remains disputed. One way to deepen our understanding of this controversial topic is to scrutinize the cultural implications of reading and writing secular poetry in the Romano-Barbarian age, particularly those works with a supposedly occasional background, which puts them outside both the realms of Church and of mere school exercises. The following article offers a close reading of such a text—a little-known wedding poem from sixth-century North Africa— with attention to its peculiar manner of interweaving Roman traditions with African idiosyncrasies as these are signposted by intertextual references. The Epithalamium Fridi (EF) forms part of a collection of Virgilian centos preserved in the Codex Salmasianus.2 The manuscript’s title identifies the poet as the uir illustrissimus et spectabilis Luxurius, who in all probability lived in the Vandal kingdom in the early sixth century CE. Nothing certain is known about Amory, People and Identity in Ostrogothic Italy, – (Cambridge: Cambridge University Press, ); for Gaul, J. Drinkwater and H. Elton, ed., Fifth-Century Gaul: A Crisis of Identity? (Cambridge: Cambridge University Press, ); Ralph W. Mathisen, Roman Aristocrats in Barbarian Gaul: Strategies for Survival in an Age of Transition (Austin: University of Texas Press, ); Dieter Hägermann, Wolfgang Haubrichs, and Jörg Jarnut, ed., Akkulturation: Probleme einer germanisch-romanischen Kultursynthese in Spätantike und frühem Mittelalter (Berlin: de Gruyter, ). . Anthologia latina (AL) –, ed. A. Riese (Leipzig: Teubner, ), a collection Shackleton Bailey (Stuttgart: Teubner, ) refused to re-edit: “neque is sum qui uati reuerendo denuo haec edendo contumeliam imponere sustineam” (praef. III); for a recent discussion see Marco Carmignani, “La recepción de la épica virgiliana en la tardía: antigüedad Juno y las Furias en los centones de Medea e Hippodamia,” in Varia et diversa. Épica en movimiento: sus contactos con la historia, ed. Rubén Florio (Mar del Plato: U.N. de Mar del Plata, ), –; on the general importance of Virgil in the Codex Salmasianus (now Paris, Bibliothèque nationale de France, Ms. lat. ) see Étienne Wolff, “Virgile dans l’Anthologie latine,” in Virgiliennes. Hommages à Philippe Heuzé, ed. J. Pigeaud (Paris: Belles lettres, ), –. After Riese the EF has been edited by Heinz Happ, Luxurius. Vol. I: Text und Untersuchungen (Stuttgart: Teubner, ), –; for recent discussion of EF, see Scott McGill, Virgil Recomposed. The Mythological and Secular Centos in Antiquity (New York: Oxford University Press, ), –; Sophie Malick-Prunier, Le corps feminin dans la poésie latine tardive (Paris: Belles lettres, ), –; Fabio Dal Corobbo, Per la lettura di Lussorio. Status quaestionis, testi e commento (Bologna: Pàtron, ), – and Sara Ehrling, De Inconexis Continuum - A Study of the Late Antique Latin Wedding Centos (PhD diss., University of Gothenburg, ), http:// hdl.handle.net//. Schwitter | A “Roman” Wedding in Vandal Africa 115 his life.3 He might have been a grammarian, but the arguments are weak.4 In modern scholarship, Luxurius receives attention primarily for his liber epigrammaton, a book of  epigrams, which follows the tradition of Martial.5 Whereas Luxurius enjoyed a high reputation among his contemporaries, modern critics are generally less assured of the literary qualities of his poetry and of the epithalamium not least.6 The -line hexametric poem apparently commemorates the marriage of a certain Fridus and his unnamed bride. The couple probably belonged to the Vandal upper class and may have lived during either the reign of king Hilderichus (– CE) or of his successor Geilimer (– CE), to both of whom Luxurius alludes in his epigrams.7 Critics generally see the EF as a rather poor replication of the Virgilian model text, one replete with logical inconsistencies and semantic infelicities. They thus . See Morris Rosenblum, Luxorius. A Latin Poet among the Vandals (New York: Columbia University Press, ), –; Happ, Luxurius I, –. There is an ongoing debate whether his name was Luxurius or Luxorius; for a synthesis of the arguments, see Dal Corobbo, Per la lettura di Lussorio, –. . The grammarian Coronatus, another Anthologia latina poet, dedicated a grammatical treatise to him; see Lucio Cristante, “Grammatica di poeti e poesia di grammatici: Coronato,” in Grammatica e grammatici latini: teoria ed esegesi, ed. F. Gasti (Como: Ibis, ), –; Luxurius dedicated a poem to the grammarian Felix (AL  Riese). Some critics contend that Luxurius was identical with the grammarian Lisorius; see Rosenblum, Luxorius, –; S. Mariotti, “Luxorius e Lisorius,” Rivista di filologia e di istruzione classica  (): –; Happ, Luxurius I, –. See however Robert Kaster, Guardians of Language. The Grammarian and Society in Late Antiquity (Berkeley: University of California Press, ), –. . AL – Riese (= – Shackleton Bailey); modern editions are by Dal Corobbo, Per la lettura di Lussorio and by Ingrid Bergasa and Étienne Wolff, Épigrammes latines de l’Afrique vandale (Paris: Belles lettres, ), –. For recent discussion, see Marco Giovini, Studi di Lussorio (Genova: Università di Genova, ); Anna Maria Wasyl, Genres Rediscovered. Studies in Latin Miniature Epic, Love Elegy, and Epigram of the Romano-Barbarian Age (Kraków: Jagiellonian University Press, ), –; Étienne Wolff, “Les poems introductifs du recueil épigrammatique de Luxorius,” in La renaissance de l’épigramme dans la latinité tardive, ed. M.-F. Guipponi-Gineste (Paris: Éd. de Boccard, ), – and, in the same volume, Annick Stoehr-Monjou, “La fabrique de l’épigramme en Afrique vandale,” –. . On his high reputation amongst his contemporaries see AL  Riese (=  Shackleton Bailey): Priscos, Luxuri, certum est te uincere uates; / carmen namque tuum duplex Victoria gestat; see in contrast Happ’s judgment of the epithalamium (Luxurius, Vol. II: Kommentar [Stuttgart: Teubner, ], ): “Dichterischen Wert hat unser Gedicht nicht im mindesten; es nimmt unter den Erzeugnissen des L. einen sehr schlechten Platz ein”; similarly C. Morelli, “L’epithalamio nella tarda poesia Latina,” Studi italiani di filologia classica  (): –, here –. . AL  and – Riese; the epigram following Luxurius’ collection (AL  Riese) praises king Thrasamund (– CE). Alessia Fassina offers a possible identification of the spouses (Herminifridus, king of the Thuringians, and Amalaberga, daughter of Thrasamund and Amalafrida): “L’Epithalamium Fridi de Lussorio: una proposta d’identificazione degli sposi,” Bollettino di studi latini  (): –. Fridus is a Western German name, but the Vandal elite was a hybrid and polyglot group. 116 STUDIES IN LATE ANTIQUITY SPRING 2020 consider the text as a failed poetic confection, neither a real cento nor a real epithalamium.8 Microtextual inquiry, however, reveals that the poem’s intertextual relations go far beyond Virgil to include Statius, Claudian, Ausonius, and other late antique poets.9 By focusing on the web of intertextual references underlying the EF, I first demonstrate how the centonist creates a triangular type of intertextual relationship between his own poem, Virgil, and the authors of canonical Latin wedding poetry, in order to connect himself and his work firmly to the Roman epithalamial tradition.10 I then claim that observant readers having this interpretive framework in mind would be able to recognize what appear to be echoes of distinctive Romano-African-Vandal values and wedding practices as prefigured by Dracontius and other fifth- and sixth-century African poets. After disclosing the latent “African” imprint on the EF, I discuss the cultural significance of Luxurius’ hybrid re-imagination of “Roman” marriage customs in the Vandal kingdom and its possible impact on the contemporary audience. As in Hosidius Geta’s Medea and other centos, poetic form and generic appropriation is a major issue in the EF.11 Being a Virgilian pastiche, the text’s function as a wedding poem is—apart from the editor’s title—not immediately apparent. Given the unusual fact that no preface or additional epistle gives a clear indication of the occasional, nuptial background of the text, the exquisiteness of form of a cento composition may have compelled . See for example Filippo Ermini, Il centone di Proba e la poesia centonaria latina (Rome: Ermanno Loescher, ), ; Morelli, “L’epithalamio nella tarda poesia latina,” ; Happ, Luxurius II, –; Sabine Horstmann, Das Epithalamium in der lateinischen Literatur der Spätantike (München: Saur, ),  and . . Morelli, “L’epithalamio nella tarda poesia latina,” –; Zoja Pavlovskis, “Statius and the Late Latin Epithalamia” Classical Philology  (): –, at –; McGill, Virgil Recomposed, –. . See McGill, Virgil Recomposed, , who coined the term “triangular intertextuality”; for similar approaches in the cento Narcissus (AL  Riese), see Marie Okáčová, “Ut imago poesis. A Pastiche of Virgil and Ovid,” Graeco-Latina Brunensia  (): –; Gilles Tronchet, “Corolles pour Narcisse. Une lecture ovidienne au coeur d’un centon virgilien,” Dictynna  () (http://journals. openedition.org/dictynna/); J. Elsner, “Late Narcissus. Classicism and Culture in a Late Roman Cento,” in The Poetics of Late Latin Literature, ed. J. Elsner and J. Hernández Lobato (New York: Oxford University Press, ), –, at  and –. . This observation is true for most extant Latin centos. For Hosidius Geta’s creation of a Virgilian tragedy, see Philip Hardie, “Polyphony or Babel? Hosidius Geta’s Medea and the Poetics of the Cento,” in Severan Culture, ed. S. Swain (Cambridge: Cambridge University Press, ), –; A. Rondholz, The Versatile Needle. Hosidius Geta’s Cento Medea and Its Tradition (Berlin: Walter de Gruyter, ). Schwitter | A “Roman” Wedding in Vandal Africa 117 the author to underscore the epithalamial character of his literary endeavor.12 Since a considerable part of the poem’s quotations derive from generically similar contexts in Virgil, a close connection between the EF and its model text is beyond dispute. However, to further strengthen the generic bonds and, arguably, to engage in an ambitious poetic play, Luxurius employs an innovative poetic technique by which he reduplicates the principal structural feature of the cento, its patchwork, on a generic level. Rather than merely adhering to a series of conventions for wedding poetry, Luxurius consciously imitates preceding epithalamial models by carefully selecting characteristic narrative and structural segments of particular poets, which he integrates into his own wedding poem. Thus, the single epithalamial motifs work not unlike the poetic units in a cento: they are put together piece by piece to form a coherent pastiche composition, even though being fragments out of different contexts. Before moving on to a microtextual analysis to establish my point, I briefly address the concept of triangular intertextuality as essential poetic strategy in Luxurius’ poem. THE CONCEPT OF TRIANGULAR INTERTEXTUALITY Luxurius’ programmatic inclusion of epithalamial models within his own wedding poem follows an established pattern. Critics have noted that later Roman epithalamiasts drew liberally from their respective predecessors, marked by conscious intertextual engagement: Claudian depended chiefly on Statius, Sidonius Apollinaris on Statius and Claudian, Ennodius on Statius, Claudian, and Sidonius Apollinaris, and so forth.13 Thus, by the late fifth and early sixth century CE, intertextual references to preceding models seem to have become a distinctive characteristic of the epithalamial genre. With regard to a Virgilian cento, however, stepping into this tradition is not an easy task. On the level of verbal similarities, no direct borrowing from any other poet except Virgil seems possible. A thematic and structural imitation on the other hand, which, as we will see, works mainly through names, motifs, and plot, faces other problems. Being a cento, Luxurius’ poem is itself a mosaic of incorporated citations, a . All transmitted Roman epithalamia have a preface. In a few cases, centos are also introduced or concluded with a couple of independent lines (for example Ausonius’ CN and Proba’s Cento). My analysis is, however, not concerned with the possibility of a performance context of the poem. Ehrling, De inconexis continuum, treats EF as an epithalamium of occasion. . Morelli, “L’epithalamio nella tarda poesia latina,” –; Pavlowskis, “Statius and the Late Latin Epithalamia,” –. 118 STUDIES IN LATE ANTIQUITY SPRING 2020 pastiche of poetical fragments that does not reproduce the poet’s own voice, but the voice of Virgil. Yet, beneath the linguistic surface, the incorporated voice of Virgil is neither neutral nor homophonic. Although “empty” associations may well exist, nonreferential intertextuality is rather unlikely in a cento:14 each unit inevitably resonates to a certain extent with the semantic context from which it is taken, which in extreme cases may well lead to a cacophonic muddle of competing referential signals, leaving the reader unable to extract meaning in an appropriate way.15 Of course, this is the very nature of a cento, resulting from the author’s willful submission to the authority of the canonical text, and from his complete assimilation of the foreign voice, which intentionally triggers associations that oscillate between the Virgilian original and the new narrative.16 Scholars have sufficiently pointed to the cento’s inherent play with double meanings and its various functions, drawing, for instance, the reader’s attention to the permeability and openness of Virgil’s poetry and its generic limits. In the EF, however, the practice of imitation and allusion reaches a new level, as Luxurius not only coopts Virgil’s voice to sing a nuptial poem but exploits it to interact with traditional models of Latin wedding poetry.17 It is an ingenious strategy to redeploy Virgil while also alluding to other poets’ works, provided the reader is willing to respond to the poet’s hints by . Happ, Luxurius II, –, qualifies the following lines as meaningless or absurd: EF ...–..–; on the concept of “empty” associations in a cento, see Ehrling, De inconexis continuum, –. . However, as Helen Kaufmann has recently argued, in late Latin poetry—unlike Augustan poetry—intertextual allusions do not necessarily make up content: “Intertextuality in Late Latin Poetry,” in The Poetics of Late Latin Literature, ed. J. Elsner and Jesús Hernández Lobato (Oxford: Oxford University Press, ), –; whereas for Aaron Pelttari, The Space that Remains: Reading Latin Poetry in Late Antiquity (Ithaka, NY: Cornell University Press, ), esp. –, Late Latin allusions are in general “nonreferential.” For a general discussion of intertextuality in late Latin poetry, see further Marc Mastrangelo, “Towards a Poetics of Late Latin Reuse,” in Classics Renewed. Reception and Innovation in the Latin Poetry of Late Antiquity, ed. S. McGill and J. Pucci (Heidelberg: Winter, ), –. . Ultimately, this becomes a question of authority and hierarchy; see Hardie, “Polyphony or Babel?,” –. On the cento as poetic form, see for example Marie Okáčová, “Centones: Recycled Art or the Embodiment of Absolute Intertextuality?” in Laetae segetes iterum, ed. I. Radová (Brno: Masarykova, ), –; Marco Formisano and Christiana Sogno, “Petite poésie portable: The Latin Cento in its Late Antique Context,” in Condensing Texts – Condensed Texts, ed. M. Horster and Ch. Reitz (Stuttgart: Steiner, ), – with further bibliographical reference. . The concept of generic mixture originates with Wilhelm Kroll, Studien zum Verständnis der römischen Literatur (Stuttgart: Metzler, ), –; on the aesthetics of hybridization in Late Antiquity, see Jesús Hernández Lobato, Vel Apolline muto. Estética y poetica de la antigüedad tardía (Bern: Peter Lang, ), –. Schwitter | A “Roman” Wedding in Vandal Africa 119 suppressing the original context of a unit—sometimes even against the grain of generic appropriateness. In a triangular constellation, oblique allusions to different authors and different genres can only be acknowledged if the referential context—in this case the Latin epithalamial tradition—is sufficiently signposted by the centonist. As noticed above, constant inclinations towards the nuptial genre in the EF provide sufficient guidance for that reader who is prepared to repress any negative associations potentially evoked by the original context of the Virgilian quotations.18 Once the epithalamial scope of the poem is recognized, the genre’s familiar mode of incorporating preceding models works as a key to unlock further intertextual relations. Bypassing the thematically related passages in Aeneid  and , observant readers will recognize that Luxurius has woven other nuptial poems into his patchwork text. Even single words may possess the power to evoke specific models or passages.19 In some instances, as we will see, a unit may even echo its own intertextual reuse of the poet alluded to, creating a cross-referential play as, for instance, a Virgilian half-line refers to Statius, who himself imitates the same verse. Of course, the reader’s familiarity with the subtexts, rules, and conventions of the genre is a prerequisite for a successful interaction with the non-Virgilian passages and contexts the author has in mind. The self-referential passages of his epigrammata, however, show that Luxurius was speaking to a highly educated elite perfectly capable of appreciating the literary quality of his poetry.20 STITCHING UP THE TEXT If the EF is in fact not only a Virgilian cento, but also a generic pastiche, the poem can be divided into single narrative fragments, all of which allude to one or several models. The following analysis shows that most segments of the cento either refer to an appropriate context in Virgil, to a specific epithalamial model, . For example the tragic outcome of the love story of Dido and Aeneas; the EF is clearly epideictic. On the steady inclining towards the epithalamial genre in this poem, see Ehrling, De inconexis continuum, –. . On the effect of single words in late Latin literature, see Isabella Gualandri, “Words Pregnant with Meaning. The Power of Single Words in Late Latin Literature,” in The Poetics of Late Latin Literature, ed. J. Elsner and J. Hernández Lobato (New York: Oxford University Press, ), –. . In recent years scholars have affirmed the general awareness of model texts and the acknowledgment of the literary tradition in Late Antiquity; see for example Kaufmann, “Intertextuality in Late Latin Poetry,” ; for a discussion of the cento’s model reader, see Ehrling, De inconexis continuum, –; on Luxurius’ self-referential passages, see Wasyl, Genres Rediscovered, –. See below for a discussion of the cultural and literary standards of sixth-century Vandal Africa. 120 STUDIES IN LATE ANTIQUITY SPRING 2020 or to both. It is in the first scene that the centonist demonstrates how he is engaging in an intertextual triangulation: the sun is rising, the streets are filled with noise as the citizens of Carthage, accompanied by a train of nymphs and goddesses under the lead of Venus, make their way to the wedding banquet (EF –). EF 1–10 (H. Happ, Teubner: Stuttgart, 1986) Statius, Silvae 1.2.1–12 (E. Courtney, Oxford University Press: Oxford, 1990) Sol, qui terrarum flammis opera omnia lustrat, | Vnde sacro Latii sonuerunt carmine montes? extulit os sacrum caelo tenebrasque resoluit. | cui, Paean, noua plectra moues umeroque comanti laetitia ludisque uiae plausuque fremebant. | facundum suspendis ebur? procul ecce canoro at Venus aetherios inter dea candida nimbos | demigrant Helicone deae quatiuntque nouena aurea subnectens exertae cingula mammae, | lampade sollemnem thalamis coeuntibus ignem dona ferens, | pacem aeternam pactosque et de Pieriis uocalem fontibus undam. hymenaeos | quas inter uultu petulans Elegea propinquat atque omnem ornatum, | Capitolia celsa tenebat, | celsior adsueto diuasque hortatur et ambit Punica regna uidens, Tyrios et Agenoris urbem. | alternum suffulta pedem, decimamque uideri hinc atque hinc glomerantur Oreades | et bona se cupit et mediis fallit permixta sorores. Iuno; | ipsa manu nuptam genetrix Aeneia duxit incedunt | pariter pariterque | ad limina tendunt | lumine demissam et dulci probitate rubentem The first line—Dido praying to the gods after Aeneas has left her (Aeneid .)—links the poem immediately to the tragic love story of Dido and Aeneas, which henceforth will form the dominant intertextual background of the poem. The following lines, however, distinctively refer—with slight alterations—to the opening scenes in Statius’ epithalamium in honor of Stella and Violentilla: there is festive music (Silvae ..–), divinities assemble from the mountains, and Venus acts as pronuba of the bride (Silvae ..–). This similarity is one of many in Luxurius’ poem. However, the poem’s close assimilation of conventional scenes from Statius comes, with regard to the epithalamial tradition, as no surprise. Like most later epithalamists, Luxurius adheres to the outline of Statius’ poem, with the distinctive feature that he strongly abbreviates the Statian storyline.21 It is nevertheless noteworthy that . On Statius’ influence on EF, see Pavlovskis, “Statius and the Late Latin Epithalamia,” –; McGill, Virgil Recomposed, –, Ehrling, De inconexis continuum, –; on Statius’ general importance for the later epithalamial tradition, see Pavlovskis, “Statius and the Late Latin Epithalamia.” Schwitter | A “Roman” Wedding in Vandal Africa 121 particular echoes of Statius usually occur when the respective Virgilian units are not drawn from the Dido and Aeneas story.22 A prominent example is EF –, where Venus and her companion goddesses take off their divine outfit. EF 14–15 (Happ) Statius, Silvae 1.2.13–15 (Courtney) una omnes, | magna iuuenum stipante caterua, ipsa toros et sacra parat cultuque Latino deueniunt | faciemque deae uestemque reponunt. dissimulata deam crinem uultusque genasque temperat atque noua gestit minor ire marita. While the surrounding units in lines – all originate from the Dido and Aeneas story,23 the slightly altered unit b originally depicts the goddess Iris preparing to incite the Trojan women to rebellion.24 The detail resembles, as Pavlovskis notes, Statius in Silvae ..–, where Venus diminishes her own beauty in order to look like a mortal.25 Further examples can be found in the lengthy dialogue between Venus and Cupid (EF –), which is obviously linked to the similar episode of the Dido and Aeneas story (Aeneid .–).26 While the first lines of the individual speeches of Venus and Cupid are exactly the same as in the Aeneid (EF b–: aligerum dictis affatur Amorem: / “Nate, meae uires, meae magna potentia solus” = Aeneid .–; EF –b: paret Amor dictis carae genetricis et alas / exuit et gressu gaudens = Aeneid .–),27 in the following passages only single units are occasionally taken from the love story of Dido and Aeneas.28 Thus, what is actually said refers chiefly to Statius, and in detail, as we will see, to other epithalamial models, . The first example is the festive sound in EF : laetitia ludisque uiae plausuque fremebant (=Aeneid .) resembling Statius, Silvae ..–: unde sacro Latii sonuerunt carmine montes? / cui, Paean, noua plectra moues umeroque comanti / facundum suspendis ebur? . EF b (=Aeneid .); EF a (=Aeneid .); EF – (=Aeneid .; Aeneid .); EF – (=Aeneid .–). . EF b: faciemque deae uestemque reponunt (=Aeneid .: faciemque deae uestemque reponit). . Statius, Silvae ..–: cultuque Latino / dissimulata deam crinem uultusque genasque / temperat atque noua gestit minor ire marita; see Pavlovskis, “Statius and the Late Latin Epithalamia,” . . The dialogue between Venus and Cupid in the cento is molded upon the dialogue between Venus and Cupid in Aeneid .–. It is, however, a conventional part in all Statian epithalamia: Statius, Silvae ..–; Claudian, Carmina .–; Sidonius Apollinaris, Carmina .–; Venantius Fortunatus, Carmina ..–. The tradition’s nucleus is, of course, Virgil, Aeneid .–. . This is the reason why Luxurius does not follow the conventional motif of Statian epithalamia, namely Cupid instigating his inert mother; see Statius, Silvae ..ff. and Ennodius, Carmina ..ff. . In Venus’ speech: none in EF –, four in EF –, none in EF –, three in EF –; in Cupid’s speech: two in EF –. 122 STUDIES IN LATE ANTIQUITY SPRING 2020 especially Claudian. Similar to Statius (Silvae ..–), Venus first turns Cupid’s attention to the bride’s youthful beauty and lauds her family: EF 29–33 (Happ) Statius, Silvae 1.2.107–110 (Courtney) huc geminas nunc flecte acies, | illam aspice contra, … Mauortia matris quae uocat | insignis facie | uiridique iuuenta, | egregium mirata decus cui gloria patrum iam matura uiro, iam plenis nubilis annis, | et generis certabat honos, tellure cadentem cui genus a proauis ingens clarumque paternae excepi fouique sinu... nomen inest uirtutis | et | nota maior imago. | Being an eminent part of the Statian epithalamial tradition and thus highly conventional, Luxurius’ laus sponsae never loses its connection with Statius, although a direct imitation is hard to substantiate.29 At one point, however, as Scott McGill argues, the linguistic similarity suggests an intentional intertextual echo: EF b quantum egregio decus enitet ore? (= Aeneid .) resembles Statius, Silvae ..–: formae / egregium mirata decus.30 A similar intertextual triangulation occurs in EF , when Luxurius uses a Virgilian line that Statius obviously imitated in Silvae ...31 There are less striking similarities: the resemblance between Venus and the bride,32 the fact that gods and mortals are present at the wedding banquet,33 or the wish for children that marks the ending of the poem.34 The opening scenes as well as the subsequent development of the EF thus sufficiently show that Statius’ poem and the love story of Dido and Aeneas form the two principal pillars around which further models are supposedly grouped. If the plot is mainly Statian and the protagonists mainly Virgilian, the second influential epithalamist, Claudian, still delivers . Compare Statius, Silvae ..–: ... cui gloria patrum / et generis certabat honos with EF –: cui genus a proauis ingens clarumque paternae / nomen inest uirtutis et nota maior imago. . McGill, Virgil Recomposed, . . EF  (=Aeneid .) auspiciis: liceat Frido (Aeneid: Phrygio) seruire marito; Statius, Silvae ..: ... Phrygio si non ego iuncta marito; noted by McGill, Virgil Recomposed, . . EF –a: os humerosque deo similis, cui lactea colla / auro innectuntur; Ehrling, De inconexis continuum, – suggests “a very close identification” with Statius, Silvae ..–: non ideo tibi tale decus uultusque superbos / meque dedi. . EF –: una omnes, magna iuuenum stipante caterua, / deueniunt faciemque deae uestemque reponunt; see Statius, Silvae ..–. . EF b–: geminam dabit Ilia prolem, / laeta deum partu, centum conplexa nepotes; see Statius, Silvae ..–: heia age praeclaros Latio properate nepotes, / qui leges, qui castra regant, qui carmina ludant. A conventional motif with which many later wedding poems end; see for example Claudian, Carmina .–; Claudian, Carmina minora .; Sidonius Apollinaris, Carmina .; Dracontius, Romulea .; Venantius Fortunatus, Carmina ..–. Schwitter | A “Roman” Wedding in Vandal Africa 123 substantial details to Luxurius’ poem. Luxurius’ dependence on Claudian may be less obvious than that on Statius, but the pattern described above is principally the same: intertextual triangulation almost exclusively occurs in instances where Virgilian units do not belong to the love story of Dido and Aeneas. Most similarities occur in the dialogue between Venus and Cupid: Venus’ laus sponsae in EF – owes as much to Claudian as it does to Statius. The laudation of the military virtues of the bride’s ancestors, for example, mirrors Claudian depicting Celerina’s forefathers: EF 29–33 Claudian, Carmina minora 25.70–72 (Happ) (J.B. Hall, Teubner: Leipzig, 1985) huc geminas nunc flecte acies, | illam aspice contra, ... Mauortia matris quae vocat | insignis facie | uiridique iuuenta, | nobilitas spoliis armisque exultat auitis iam matura uiro, iam plenis nubilis annis, | inmensamque trahit Celerini robore lucem cui genus a proauis ingens clarumque paternae nomen inest uirtutis | et | nota maior imago. | And like Celerina, Fridus’ bride has recently come of age.35 Further similarities exist, but they seem rather too vague to suggest imitation with authorial intention. A possible example occurs in EF , in which he states that the bride shines like a jewel, before comparing her to the sea nymphs Doto and Galatea. Luxurius may have thought of Claudian, who depicts both sea nymphs offering precious gems and other gifts to bride and groom.36 Another moment common to Luxurius and Claudian is the bride’s lactea colla (EF b), which again is rather conventional.37 Apart from the laus sponsae, observant readers will detect additional resemblances in the cento, especially in the lines that do . EF –: uiridique iuuenta, / iam matura uiro, iam plenis nubilis annis; Claudian, Carmina minora .–: ... matura tumescit / uirginitas superatque niues ac lilia candor. . Claudian, Carmina .–: cingula Cymothoe, rarum Galatea monile / et grauibus Spatale bacis diadema ferebat / intextum, rubro quas legerat ipsa profundo. / mergit se subito uellitque corallia Doto: / uimen erat dum stagna subit; processerat undis: / gemma fuit. . See Statius, Silvae ..: nec colla genasque; Claudian, Carmina .–: ... non labra rosae, non colla pruinae, / non crines aequant uiolae, non lumina flammae; and more elaborately: Sidonius, Carmina .–: tantus honor geniusque genis; collata rubori / pallida blatta latet depressaque lumine uultus / nigrescunt uincto baccarum fulgura collo; Claudian, Carmina minora appendix .–: lilia ceu niteant rutilis commixta rosetis, / sic rubor et candor pingunt tibi florida uultus. / denique miramur quod colla monilia gestant; / ex umeris frustra phaleras inponis eburnis. / non tibi nam gemmae, sed tu das lumina gemmis. 124 STUDIES IN LATE ANTIQUITY SPRING 2020 not derive from the Dido and Aeneas story. A potential allusion to motifs in Claudian’s epithalamia is the bride’s unawareness of the upcoming wedding or her fear of her future husband in EF .38 Cupid’s anticipation of the wedding night in EF  (omnia praecepi animo mecum ante peregi) may well mirror Hymenaeus informing Venus about the wedding in Claudian, Carmina minora .ff., as Ehling suggests.39 In EF –, Luxurius mentions that the bride’s father has arranged the wedding. Verbal as well as thematic similarities suggest that the passage reflects Claudian, who has Honorius complain about his future father-in-law, because he has not yet given his consent for the marriage.40 Another moment shared by both poets is the ekphrasis of a palace—in Luxurius’ case the location of the cena nuptialis (EF –), in Claudian the palace of Venus (Claudian, Carmina .–). In depicting the location of the cena, Luxurius creates a solemn atmosphere as he consistently adapts phrases that Virgil originally used to picture the mighty palace of Latinus: EF 11–13 (Happ) tectum augustum ingens, centum sublime columnis, | (= Aeneid 7.170) quo sacrae sedes epulis, | atque | ordine longo | (= Aeneid 7.175: hae sacris sedes epulis | Aeneid 1.395) perpetuis soliti patres considere mensis. | (= Aeneid 7.176) The possible allusion to Venus’ divine palace in Claudian’s poem certainly stresses the epideictic moment of the scene. To some extent, the passage may also allude to another influential late Roman wedding poem: the epithalamium Sidonius Apollinaris wrote for his friend Ruricius (Carmina ), which begins with a lengthy and highly artistic ekphrasis of Venus’ mansion in the Aegean sea. While the plot of his poem draws on Statius and Claudian, Sidonius’ originality lies in his descriptive art, which makes him a perfect model for ekphrastic elements in later epithalamial poetry. However, as similarities between the EF and Sidonius are generally rather vague or conventional, one might doubt that . EF : hanc ego nunc ignaram huius quodcumque pericli est may reflect Maria’s unawareness in Claudian, Carmina .–. – and Celerina’s fear in Carmina minora .: quem nunc horrescis amabis. . Ehrling, De inconexis continuum, . . Compare EF –: cui natam egregio genero dignisque hymenaeis / dat pater et pacem hanc aeterno foedere iungit (=Aeneid .–) with Claudian, Carmina .–. Keywords are socer and gener; the thematic nexus lies on the worthiness of the imperial wedding. Schwitter | A “Roman” Wedding in Vandal Africa 125 Luxurius intended to create a closer intertextual relationship between his poem and Sidonius’ epithalamia.41 With Ausonius’ Cento nuptialis (CN), the case of triangular intertextuality is again on more solid ground. Scholars have long seen the inclusion of a sex scene at the end of the cento as a clear sign of Luxurius’ intention to imitate Ausonius’ CN, especially its obscene imminutio (Ausonius, CN –).42 Regarding the whole poem, the EF only shares  out of  Virgilian units with Ausonius’ cento. As to the description of the sexual congress (EF –), however, half of the quotations are identical.43 Ausonius was with all probability the model for this remarkable passage, which finds no parallel in any Statian epithalamia.44 Further analogies confirm the close intertextual relationship between the two poems: EF b– describe the festive music at the cena nuptialis, quoting first Aeneid . (ubi adsuetis biforem dat tibia cantum), then proceeding to Aeneid .– (obloquitur numeris septem discrimina uocum, / iamque eadem digitis, iam pectine pulsat eburno). Ausonius uses exactly the same lines, in reverse order, to describe a singer at the cena (Ausonius, CN –).45 MASTERING THE MODELS: BETWEEN CONVENTION AND DEVIATION Given the close generic relation between the original Virgilian contexts and the EF, it is tempting to qualify Luxurius’ engagement in a triangular intertextuality as an impressive, but rather needless poetic play. The self-imposition of poetic boundaries in the cento composition forms a demonstration of skill advertising its own author. In many cases, however, imitation is barely noticeable and, even when in fact recognized, of only limited relevance for the content of the poem, . For example, the bride shining like a jewel: EF  (qualis gemma micat) and Sidonius, Carmina .–, but also resembling Claudian, Carmina . and Venantius Fortunatus, Carmina ..; the wish for offspring at the end of the poem: EF – and Sidonius, Carmina .–, but also resembling Statius, Silvae ..– and Dracontius, Romulea .–. . Karl Schenkl, Probae cento (Vienna: F. Tempsky,  = CSEL .), ; Ermini, Il centone di Proba, ; Pavlovskis, “Statius and the Late Latin Epithalamia,” –; McGill, Virgil Recomposed, ff. and –; Horstmann, Epithalamium, –; Malick-Prunier, Le corps feminin, –; Ehrling, De inconexis continuum, –. . In both centos, sexual congress is introduced by the same Virgilian unit: Aeneid .: spemque dedit dubiae menti soluitque pudorem (EF  = CN ). The actual intercourse is described almost identically: EF b = CN  (Aeneid .: ramum qui ueste latebat), EF  = CN  (Aeneid .: eripit a femine (CN: femore) et trepidanti (EF: flagranti) feruidus instat (EF: infert)). . Ausonius’ CN is not an epithalamium, but a cento that narrates a wedding; see Horstmann, Epithalamium, –. . McGill, Virgil Recomposed,  note  further points to a verbal echo linking EF : cum dabit amplexus atque oscula dulcia figet to CN –: illum turbat amor figitque in uirgine uultus. 126 STUDIES IN LATE ANTIQUITY SPRING 2020 especially if seen in purely formal features. Through the dynamics of triangular intertextuality, however, Luxurius is able to generate a fusion of epithalamial models within his poem, thus strengthening the generic bond that is threatened by the cento composition and the Virgilian subtext. Generic appropriation by means of constant allusions to the epithalamial tradition strongly underscores the cento’s nuptial character. At the same time, however, by emphasizing the conventionality of his nuptial poem, Luxurius draws the reader’s attention to obvious deviations from classical models. The poem’s conventional plot is abruptly subverted by the inclusion of a graphic sex scene modelled on Ausonius’ notorious immunitio. This ending with a twist forces the readers to re-evaluate their interpretation and, eventually, after a shift of their interpretive horizon, to anticipate further implicit sexual subject matter within the poem. A closer look may then reveal that the final sex scene is just the culminating point of an obscene narrative lurking beneath the panegyric surface of the wedding poem. Implicit references to sexual violence and fear are present throughout the text running counter to its festive tone. Further, the poem’s narrative focus almost exclusively rests upon the bride, while the groom is nearly absent. She is the potential victim of violence and the prime objective of Venus’ and Cupid’s interference within the human realm (which actually consists of preparing and encouraging her for sexual intercourse). When Venus is addressing Cupid, the wedding is already accomplished (EF b: thalamos ne desere pactos!). Thus, as mentioned above, her instructions to Cupid are related solely to the wedding night. The text’s vocabulary is ambivalent and avoids obvious sexual reference: nunc ignaram huius quodcumque pericli est (EF ); occultum inspires ignem (EF a); liceat Frido seruire marito (EF ); sentiet! (EF a). Only in the end does Luxurius finally lift the veil by describing the effects of defloration in an explicit manner: it cruor inque humeros ceruix conlapsa recumbit (EF  = Aeneid .). The disturbing frivolity that covertly undermines the nuptial poem raises questions about the panegyric scope of the EF and its historical meanings. If the cento had indeed an occasional origin, Luxurius’ text might have received an unfavorable response from his initial audience. The poet could then have claimed that the passage is a due reference to Ausonius and thus part of his intertextual strategy to incorporate the major Roman epithalamiasts into his own poem. This scenario, however, seems rather unlikely. Luxurius’ literary goal was not to write a parody of Virgil’s account, much less of the spouses. First, as shown above, in most instances the Virgilian units fit semantically into the new context. Virgil’s words are—in contrast to Ausonius’ CN—not downgraded in Schwitter | A “Roman” Wedding in Vandal Africa 127 any way, or incongruously adapted into a new narrative to generate semantic shifts aiming at a parodistic effect.46 Second, by furnishing his cento with convincing epithalamial details, Luxurius has effectively created a plausible nuptial song that offers credible praise to the bridal couple, even if the text’s subliminal obscenity is perceived. Compared to Ausonius’ account, the sexual intercourse is less aggressive, and the couple are related to one another in closer affective terms.47 A graphic sex scene is, of course, an alien topic in a Roman epithalamium, but sexual suggestion is not. Some liberties, it seems, could be taken without injury to decorum. In fact, the final passage, which provokes readers to recognize the sexual implications hidden within the poem, is a clever punchline put in place by the distinguished epigrammatist in order to dazzle his audience once more with his ludic skill and wit. TRANSCENDING THE TRADITION: THE INCLUSION OF “AFRICAN” INNOVATIONS The preceding discussion has shown that Luxurius engages in versatile intertextual play with different epithalamial models, all of which were written in Italy or Gaul at least a century before his time. However, the wedding’s location is not Rome or Trier but, as in the love story of Dido and Aeneas, Carthage (EF  = Aeneid .). Since Vandal kings reigned undisputedly over the rich and fertile lands, Roman Africa began to foster its own literary community and identity. Recent scholarship has pointed to the “Africanism” that lurks beneath the classicizing surface of many poems by fifth- and sixth-century African authors.48 Given Luxurius’ eagerness to include former models, one might wonder if he was not also alluding to his famous contemporary, the Carthaginian poet Blossius Aemilius Dracontius, whose oeuvre he almost certainly knew. Dracontius wrote two epithalamia (Romulea  and ), both of which deal with the author’s imprisonment under king Gunthamund (– CE).49 Despite . McGill, Virgil Recomposed, – argues that Luxurius seeks to vulgarize Virgil by applying him to sexual subject matter. In my view, parody seems not to have been the poet’s principal aim, yet the notion of a discrete parody within the broader non-parody cannot simply be dismissed. On Ausonius’ approach, see McGill, Virgil Recomposed, –; Raphael Schwitter, “Der obszöne Leser. Vergil-Kritik und apologetische Strategie in Ausonius’ Cento nuptialis, –,” Museum Helveticum . (): –. . Horstmann, Epithalamium, ; Malick-Prunier, Le corps feminin, , and Ehrling, De inconexis continuum,  all agree on this. Certain alterations of the text point into that direction, for example flagranti for the Virgilian (and Ausonian) trepidanti in EF . . See Elsner, “Late Narcissus,” – with further bibliographical reference. . Horstmann, Epithalamium, –; Malick-Prunier, Le corps feminin, –. 128 STUDIES IN LATE ANTIQUITY SPRING 2020 the unusual autobiographical references in his poems, Dracontius was an obvious imitator of Statius, but deliberately transferred the Statian epithalamium to a Carthaginian setting.50 A closer microtextual analysis reveals that there are, in fact, deliberate echoes of Dracontius in the EF. One notable instance is EF a, where Luxurius mentions the Oreades gathering to attend the cena nuptialis, a controversial detail that led troubled critics to question the poems’ qualities.51 It can scarcely be coincidental that the only epithalamial poet who also refers to the Oreades is Dracontius.52 Unlike in other regions—one presumes—the Oreades may more plausibly have been present at a wedding in Carthage. The same could be true for a similarly controversial passage at EF –a, where Luxurius mentions the sound of the military trumpet at the wedding (= Aeneid .–: the Volscians come to battle against the Trojans). The notion is indeed awkward; is it a general reference to Carthage’s martial past in the wars against Rome, or to the Vandal warrior culture? The tuba is a dissonant element in the cento because it ostentatiously opposes Venus’ wish in Claudian, Carmina .–: tibia pro lituis et pro clangore tubarum / molle lyrae festumque canant.53 Yet, in the first scene of Dracontius’ epithalamium for the brothers, references to war seem omnipresent: Pallas is dancing in full armor alongside Apollo, menacingly waving her weapons as if going into battle; Mars is dancing too, trying to impress Venus; Cupid triumphs per castra Veneris; allusion to elegiac love is obvious (Romulea .: cui militat omnis, recalling Ovid, Amores ..: militat omnis amans).54 Again, Luxurius’ reference to a “Carthaginian wedding” as prefigured by Dracontius appears to be a plausible explanation for his deviation from the Statian epithalamial tradition. The trumpet thus signals the presence of Minerva and Mars, as well as of Cupid who has successfully mollified the god of war (EF b: mollitque animos et temperat iras). . Dracontius, Romulea .– (O. Zwierlein, de Gruyter: Berlin/Boston, ): moenia respexit Carthaginis alma Cythere, / “illo”, dixit, “aues, conuertite, mando, uolatus”; for Statius’ influence on Dracontius, see Pavlowskis, “Statius and the Late Latin Epithalamia,” –. . EF a (=Aeneid .): hinc atque hinc glomerantur Oreades; see Morelli, “L’epithalamio nella tarda poesia latina,” ; Happ, Luxurius II, . Ehrling, De inconexis continuum,  tries to explain it with regard to the quotation’s original context. . Dracontius, Romulea .: Oreadas Fauni iungant et Naidas Amnes; also pointed out by Horstmann, Epithalamium, . . Does it refer to the bella amoris, as in Claudian, Carmina .–: ... quae proelia sudas, / inprobe? quis iacuit telis? (Venus talking to Cupid)? . Dracontius, Romulea .–: Pallados armisonae crinem praecingat oliua, / arma ferat Martis, quibus et proludere nostis / et bellum, si tempus eget, perferre ualetis. ... Mars saltat amores / et Venerem placare cupit, cui militat omnis / marcidus et nudis ludit post arma lacertis, / ac furibundus Amor Veneris per castra triumphat. Schwitter | A “Roman” Wedding in Vandal Africa 129 There are more examples. In the EF, Venus is addressing Cupid on the evening of the wedding banquet, and not beforehand as is common in all Statian epithalamia. Thus, her instructions are related to the subsequent wedding night, and not to the wedding ceremony.55 Luxurius shares this peculiar detail with Dracontius’ epithalamium to the brothers: after having addressed the two brides concerning their wedding night, Venus has Cupid shoot his arrows to inflame the spouses with sexual desire.56 Further parallels exist, but they are too conventional to suggest with confidence an authorial intention to imitate Dracontius.57 Dracontius may not have been the only contemporary African poet to whom Luxurius was alluding. There are, for instance, similarities with the Iudicium Paridis, another cento passed down within the Salmasianus collection.58 The poem’s title claims Mavortius as author. Scholars have identified him with the Ostrogothic consul of  CE, Vettius Agorius Basilius Mavortius; yet, from the sixth century, only African poets seem to be included in the Salmasianus collection.59 Even though we do not know for certain the degree to which real-world elements play a role in Luxurius’ or Dracontius’ description of wedding ceremonies, specific Romano-AfricanVandal idiosyncrasies as revealed by a close-reading must have been far more obvious to contemporary readers.60 A “ROMAN” WEDDING IN SIXTH-CENTURY VANDAL AFRICA The creation of a Virgilian cento to be recited as a marriage poem at a sixthcentury Vandal wedding is in many ways remarkable. It draws our attention to . McGill, Virgil Recomposed, . Tunc (EF a) clearly shows that the speech is held after the wedding ceremony. As in Aeneid .– (the “wedding” of Dido and Aeneas), the consummation of the wedding is signaled by lightning and thunder, EF –a: dant signum, fulsere ignes et conscius aether / conubii; EF : ... et consere dextram is not an instruction concerning the ceremony but the wedding night, as the following lines suggest (–: occultum inspires ignem paribusque regamus / auspiciis: liceat Frido seruire marito). . Dracontius, Romulea .–: dixerat et puerum genetrix implere pharetras / imperat. ille libens imbutas melle sagittas / misit et ambarum sensus transfixit arundo. . For example, the wish for offspring at the end of the poem (EF b–) as in Dracontius, Romulea .– (at the end of the poet’s address to Cupido), but see Statius, Silvae ..– and Sidonius, Carmina .–. . AL  Riese; for a recent edition, see Maria Teresa Galli, I Vergiliocentones minores del codice Salmasiano (Florence: Felice le Monnier, ), –. . Galli, I Vergiliocentones,  with further bibliographical reference. . There is, of course, the question of whether the tastes and values shown by Dracontius and Luxurius are particular to them or are more widely shared in Vandal Africa. 130 STUDIES IN LATE ANTIQUITY SPRING 2020 the broader cultural context of Luxurius’ poetic performance, especially if the alleged Germanic noble Fridus is thought to have commissioned the work himself. There is no doubt that both form and sources of the poem require a mastery of Latin and a high-level self affiliation to the classical literary tradition. Still more intriguing are the cultural implications of Luxurius’ re-imagination of a specifically Roman marriage adapted to a contemporary couple. Clearly it would be dangerous to take his poem literally, as a description of an actual wedding ceremony or as secure evidence that marriage poems were still sung at Carthage in the sixth century. However, at least within the realms of literature, it seems, the idea of a culturally defined Romanness untouched by either Christianity or Germans not only persisted among the African-Roman gentry, but was as readily adopted by the Vandal elites.61 Much work has been done in recent years on “Romanization” in North Africa, and its cultural, civic, legal, and economic impact during the Empire and beyond.62 Although being Roman lost its significance in many social and political contexts under Germanic rule, in the Vandal kingdom, as Jonathan Conant has argued, “high culture was Roman culture” and continued to serve both as a mark of identity and of social distinction throughout the sixth century.63 Yet, some core issues still remain unresolved: whether, for instance, Germanic aristocrats who . The omission of any Christian references in EF may be telling. The case of a programmatic secular space created within the poems of the Codex Salmasianus has been made by Robert Miles, “The Anthologia latina and the Creation of Secular Space in Vandal Carthage,” Antiquité tardive  (): –. For the Christian conflict in Vandal Africa see Robin Whelan, Being Christian in Vandal Africa: The Politics of Orthodoxy in the Post-Imperial West (Oakland: University of California Press, ). . For an overview on Roman North Africa, see Jean-Marie Lassère, Africa quasi Roma  av. J.-C. –  apr. J.–C. (Paris: CNRS Éditions, ); on “Romanization” in Africa, see Claude Lepelley, “L’Afrique,” in Rome et l’intégration de l’Empire:  av. J.-C. –  apr. J.-C, Vol. , ed. C. Lepelley (Paris: Presses Universitaires de France, ); on the Vandal kingdom, see Andrew H. Merrills and Richard Miles, The Vandals (Chichester: Wiley-Blackwell, ); Roland Steinacher, Die Vandalen: Aufstieg und Fall eines Barbarenreichs (Stuttgart: Klett-Cotta, ), and from an archaeological perspective Ralf Bockmann, Capital continuous: A Study of Vandal Carthage and Central North Africa (Wiesbaden: Reichert, ); for questions of identity and culture, see Conant, Staying Roman; Roland Steinacher, “When Not in Rome, Still Do as the Romans Do? Africa from  BCE to the th Century,” in Transformations of Romanness: Early Medieval Regions and Identities, ed. W. Pohl, C. Gantner, C. Grifoni and M. Pollheimer-Mohaupt (Berlin: de Gruyter, ), – and the contributions in Andrew H. Merrills, ed., Vandals, Romans and Berbers: New Perspectives on Late Antique North Africa (Aldershot: Ashgate, ). . Conant, Staying Roman, – with further bibliographic reference. Epigraphic evidence suggests that Latin culture still had a political and legitimatizing impact, even in the Berber kingdoms in Mauretania and Tripolitana which evolved as an alternative to Vandal rule: Steinacher, “When Not in Rome,” –. Schwitter | A “Roman” Wedding in Vandal Africa 131 assimilated to elite Roman lifestyle did also regard themselves as “Romans,” in at least some aspect of their multilayered identity.64 In this context, one of the more significant findings of this inquiry is the “Africanism” imprinted on the EF by means of intertextual references. The “Roman” wedding of Fridus, at least in Luxurius’ fictionalized enactment, had in the eyes of the contemporary reader its clear African characteristics, obviously fuelled by a local pride, which seems to have been distinctive of the RomanoVandal high culture that had developed by the sixth century.65 Although references to Dracontius and other African poets were limited by the self-imposed restriction of a cento composition, with the technique of triangular intertextuality and generic appropriation once revealed, the reader could still unlock specific African idiosyncrasies behind the Virgilian units. Thus, while enjoying the versatile intertextual play or the loftily developed sexual implications that pervade the poem, Luxurius’ audience may have been indulging itself as being real Romans. It may have been an imaginary Roman world, but it was one with a distinctive African imprint. . If Fridus and his Germanic upper-class peers were indeed the initial audience of Luxurius’ Epithalamium, the answer may be in the affirmative. There is an ongoing debate for whom exactly the “Vandal” poems collected in the Codex Salmasianus were originally written. Scholarship usually acknowledges the Romano-African elite as the main addressee, but Luca Mondin and Lucio Cristante have recently convincingly argued that Luxurius may well have been the Anthologia latina’s first redactor with Fridus his main dedicatee. See Luca Mondin and Lucio Cristante, “Per la storia antica dell’ Antologia Salmasiana I. La composizione della silloge cartaginese e il suo curator,” Rivista di studi di Anthologia latina  (): –. For a different view see Miles, “The Anthologia latina” who argues that the “‘Vandal’ poems of the Anthologia latina were written by and mostly written for, the Romano-African elite” (at ). Miles stresses the political agenda behind the collection, which in his view ought to promote a secular vision of Vandal North Africa in a period of great religious tensions. . Conant, Staying Roman, – speaks of “a Romanness of a distinctively African stamp” to which the Vandals were exposed; eventually they came to identify themselves with Africa. 132 STUDIES IN LATE ANTIQUITY SPRING 2020