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 (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2021. 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 January 2021

Sex and Salvation in the Vergilian Cento of the Fourth Century

Sex and Salvation in the Vergilian Cento of the Fourth Century

(p.101) 4 Sex and Salvation in the Vergilian Cento of the Fourth Century
The Baptized Muse

Karla Pollmann

Oxford University Press

Abstract and Keywords

In a cento, a writer uses fragments from texts of canonical authors (Genette’s hypotext) taken out of their original context to create a new work (Genette’s hypertext), which one could call ‘literary patchwork’. This chapter argues innovatively to differentiate between the predominantly parodical aim of pagan centos and the predominantly exegetical aim of Christian centos, and illustrates the implications and consequences of this assumption by analysing the centos of the fourth-century writers Ausonius and Proba. While both poets are adhering to a classicism that sees Vergil as indispensable for conveying a culturally convincing message, they manage to exploit the authority and embedded polysemy of the hypotext for diametrically opposed messages. Thus the chapter programmatically demonstrates that the literary genre of the cento, instead of being wrongly accused of epigonality, is a highly original and complex literary form.

Keywords:   Cento, Ausonius, Proba, Vergil, literary patchwork, parody, exegesis, hypotext/hypertext, classicism, polysemy


‘Every book is a quotation; and every house is a quotation out of all forests and mines and stone quarries; and every man is a quotation from all his ancestors.’1 This pre-postmodern claim of a universal intertextuality has gained strong prominence in post-war theoretical discussions concerning the nature of literature. In a less sharply focused way, such a perception of language or literature can already be observed much earlier. Lucretius in the De rerum natura (written before 55 BC) compares the versatility and endless possibility of recombining atoms to create things to the same ability of letters to be combined into different words (1.197–9, 823–7, 912–14; 2.688–9, 1013–22).2 In an analogous way, this is exactly what happens when a poet writes a cento3 by rearranging an original hypotext into a new, equally meaningful but different hypertext. This method can best be called hypertextuality.4

The writing of centos goes back to the Alexandrian period,5 though the first extant centos date to the first6 and second century AD.7 This literary technique remained popular till at least the nineteenth century.8 In a cento, a writer uses (p.102) fragments from texts of canonical authors (Gennette’s hypotext) taken out of their original context to create a new work (Genette’s hypertext), which one could indeed call literary ‘patchwork’ (Greek κέντρων‎)9 or an assemblage (collage) of quotations.10 It was in the truest sense ‘littérature au second degré’ (Genette). It is important that the authors whose texts were used for a cento should be canonical and well known in order to assure the recognition of the technique by the readers. In practice this meant especially, though not solely, Homer and Vergil.11

In a different context,12 I have shown two possible roots for the technique of the cento. First, the method of Alexandrian grammarians of making a difficult text more plausible by changing its word order, which was justified by the rhetorical figure of anastrophe, that is, the inversion of the natural word order. The context of this method is exegetical.13 Second, the Aristophanic mockery of the syntactical monotony of Euripidean verse, which is exposed in Frogs 1206–41 by quoting seven times half lines from Euripidean prologues that can always be completed by the mock ending ‘He has lost his flask’, thus creating a new, ridiculous meaning. The challenge is to maintain the syntactical and metrical coherence of the lines. The context of this method is parodical.

In a cento, the author’s ambition would be to recombine parts of verses from one14 author’s works in order to create a new, unexpected sense, while at the same time moving (almost) exclusively within the metrical and lexical material of this author.15 A computer programme run over the new text would not be able to detect that it was not, in fact, the authentic work of the original author. However, in antiquity the verse cento was normally not intended as forgery. In order to make the cento more practicable, minor alterations of the (p.103) original hypotext were allowed, as, for example, changing a noun in the singular into the plural, or a verb in the indicative into the subjunctive, etc.16 The crucial difference from the original is the semantic change of the material. This is possible on the micro-level of individual words or phrases that are sometimes employed in the new context in a meaning different from that of their original context; and this is necessary on the cento’s macro-level, as the whole of the new text conveys a completely different message.

The possibilities of how to apply the technique of a cento range on the semantic level from sheer parody and caricature17 to serious interpretation with the intention of revealing a hitherto hidden and thus ultimately true meaning of the text. In the former case the noble style of the hypotext is applied to a vulgar subject in the hypertext, implying the parodical transformation of the hypotext; in the latter the noble style of the hypotext is applied to a noble subject in the hypertext, implying the serious transposition of the hypotext.18 It is noteworthy that the pagan tradition from Alexandrian times up to late antiquity stayed more or less on the parodical track, whereas the Christian tradition, with Proba as its first Latin representative,19 pursued the serious intention of revealing the hidden Christian message contained in the pagan canonical poets.20

Though centos had been written well before Ausonius (from around 310 to 395),21 he was the first to write what could be called a ‘theory of the cento’ in the preface to his Cento nuptialis.22 I quote some extracts that are crucial for our context:23

Those who first trifled with this form of compilation (concinnatio)24 call it a ‘cento’. It is a task for the memory (memoria)25 only, which has to gather up scattered tags and fit these mingled scraps together into a whole, and so is more likely to provoke your laughter (ridere) than your praise.…For it is vexing to have Vergil’s majestic verse degraded (dehonestasse) with such a comic theme (p.104) (ioculari…materia).…So take a little work, continuous, though made of disjointed tags; one, though of various scraps; absurd, though of grave content; mine, though the elements are another’s.…A cento is a poem compactly built out of a variety of passages and different meanings, in such a way that either two half-lines are joined together to form one, or one line and the following half with another half. For to place two whole lines side by side is weak (ineptum), and three in succession is mere trifling. But the lines are divided at any of the caesurae which heroic verse admits.…This my little work, the Cento, is handled…so as to harmonize different meanings (sensus diversi ut congruant), to make pieces arbitrarily connected seem naturally related, to let foreign elements show no chink of light between, to prevent the far-fetched from proclaiming the force which united them, the closely packed from bulging unduly, the loosely knit from gaping.

In this passage, Ausonius emphasizes the importance of the formal success of the poem, which is as relevant as the unity of its new content. The convincing combination of formerly heterogeneous parts into a seamless whole has, moreover, the function of giving the impression that the ‘new’ poem is an original creation of the centonist.26 Ausonius’s characterization of the cento is dominated by paradoxes: the dignity of the Vergilian poetry versus the comic theme of his own Cento; the erudition of Valentinian I27 versus the playfulness (ludus, see Green [1991], 132.3, 139.19) of his cento; the continuity of the cento versus the disjunction and disparity of its elements; the seriousness of the hypotext versus the ludicrousness of the hypertext; the tension between the different ownerships of hypotext and hypertext. Defining the genre of the cento as essentially paradoxical corresponds to its particular nature as a text that in itself exposes the paradoxicality of texts or literature in general as potentially or necessarily fluid entities capable of conveying various meanings (parole) as long as the author and the reader (who is also the second author who transforms the hypotext into a cento) operate within the same langue.28

In the prose passages framing the cento, and in the digression (parecbasis) before the detailed description of sexual intercourse, there are several features bearing upon the meaning of the Cento: (a) the apologetic tone of the author; (b) the repeated references Ausonius makes to his Cento as a nullius pretiiopusculum (‘a small work of no worth’), which is, as he alleges, carelessly composed; (c) the description of the Cento as jocular entertainment.

Concerning (a), Ausonius’s poem may first have been composed around 374,29 and so, presumably, after Proba. Yet, although Vergilian centos had been written before Ausonius,30 the need for apology may be explained by the (p.105) fact that here for the first time Vergil was used in an extensively obscene way. Ausonius may have felt he had to give some warning, which at the same time served to incite the reader’s curiosity. Moreover, he combined this with the remark that the emperor Valentinian I himself indulged in such nugae, and that he was really writing at his command. His apology thus turns into confirmation that his literary activity had the backing of the emperor and was, as a conequence, above reproach.31

On (b), his repeated emphasis on the formal carelessness of his Cento could be understood as a topos of modesty, though possibly there existed technically more successful specimens of this genre.32 It seems more likely, however, that this statement of ‘unpolishedness’ is to be understood as hinting at the nature of the literary genre: the theme of the Cento nuptialis is completely different from the high and serious topics of epic and tragedy. At the same time, however, the author’s thorough knowledge of Vergil and his ingenuity are highlighted, both facilitating the speed with which a cento could be created.33 In this, Ausonius follows a tradition as it can be observed in Catullus 1.1 ‘my pretty new booklet’ (lepidum novum libellum). However, the playfulness of Catullus’s work nevertheless claimed to be ‘polished’ (1.2 expolitum), in opposition to Ausonius, who in his letter to Paulus emphasized the roughness of his work (nec labor excudit nec cura limavit, sine ingenii acumine et morae maturitate, ‘neither did labour mould it not diligence polish it, without the sharpness of mental power and the maturity of extended time dedicated to it’; Green [1991] 132.2–3). In a similar way, Statius, Silv. 1 pref. emphasized the improvised state and speedy composition of his Silvae.34 Ausonius uses this kind of terminology several times in his oeuvre.35 That the statement cannot simply be taken at face value is clear from its aim of giving the appearance of authenticity, unity, and originality. The hypertext should no longer bear any traces of the original content or context of the hypotext. This ideal of Ausonius’s aesthetics of production does not necessarily contradict the fact that the readers have to be able to recognize the poet’s model in order to appreciate the cento properly. Moreover, as we shall see in Parts II and III, the contrasting comparison of the old and the new contexts of the reassembled phrases adds a crucial dimension to an enhanced understanding of the cento’s content and message.

(p.106) Concerning (c), the characterization of the cento as something one could rather laugh at than praise is clearly not meant with regard to the semantic surface of the hypertext. Its content as such is not funny at all, with its elements of solemn marriage ritual and, at its end, ‘one of the most detailed descriptions of sexual intercourse in Latin literature, and also one of the most violent’.36 The humour comes in only when the readers recognize that the poetic material used is entirely taken from Vergil (see also on (b)). This mechanism had already been described by Seneca the Elder with regard to Ovid’s technique of Vergilian allusion, in Suas. 3.7: ‘Something he (i.e. Ovid) had done with many other lines of Vergil, with no thought of plagiarism, but meaning that his piece of open borrowing should be noticed (agnosci)’. Though such a cento expresses a certain irreverence towards the poet, this does not necessarily mean that the hypotext as such is not respected. On the contrary, it is the implicit acknowledgement of the familiarity of the reader with Vergil’s works and their canonicity that guarantee that the Cento will work and be recognized for what it is. In a similar way parody in general only works if the text to which it refers is taken entirely seriously.

As already hinted at under (b), it is noteworthy that the theoretical statements made in the preface to the Cento nuptialis are not always followed by the poet (or indeed by other poets, including Proba)37 in practice. First and foremost, Ausonius himself frequently uses whole lines, even two consecutive ones (25–6, 75–6, 97–8), though he describes this as ineptum. Second, not only caesurae are used as cut-off points for the addition of another phrase. Third, though the smoothness of the cento in syntax and content should be as successful as possible, it only achieves its full effect if the reader nevertheless recognizes the underlying original context.38 All in all, Ausonius’s theory of a cento does not present rigid guidelines, but is rather to be understood as an ideal programme.

Sex: Ausonius, Cento nuptialis

Ausonius was not the first to use Vergil’s poetry for vulgar effect in a cento. This was previously undertaken by Petronius in his Satyricon 132.11. In that context the hero Encolpius blames his penis for not having performed in the required way when he was last together with his lover. Petronius describes the reaction of the scolded penis in Vergilian verse combined in a cento:

  • illa solo fixos oculos aversa tenebat, (= Aeneid 6.469)
  • nec magis incepto vultum sermone movetur (= Aeneid 6.470)
  • (p.107) quam lentae salices | lassove papavera collo. (= modified Eclogue 3.83 lenta salix ≈ 5.16 lenta salix quantum + Aeneid 9.436)
  • That one turned aside and gazed fixedly at the ground, nor is moved by the beginning of the speech any more than drooping willows or poppies on their tired necks.39

The first two lines are taken directly from two successive lines in Vergil (which according to Ausonius is inept,40 where he describes Dido’s refusal to speak to Aeneas in the Underworld: illa, referring in the original to Dido, denotes in the new context Encolpius’s penis (ea or illa pars corporis, see Sat. 129.1 and 132.12).41 The crude ‘identification’ is almost macabre, but at the same time has a comic effect, as one imagines the shyly cast-down eyelids of the thus anthropomorphized penis in the first line.42 In the next line the anthropomorphic aspect continues, especially in vultum, but with movetur the physiological state of the penis as such is expressed, making use of the double meaning of movere as metaphorical (of emotional movement, as in the Vergilian original) and literal (the lack of physical movement, i.e. erection of the penis, in Petronius). This is reinforced by the comparison in the third line, where the reader is invited to refer the figurative level of the comparison again to the physiological behaviour and appearance of the flaccid penis: it hangs down like the drooping branches of willows, and its red glans is tilted like red poppy blossoms on their tired necks.43 This interpretation is supported by the fact that Petronius does not continue after Aen. 6.470 with 471 quam si dura silex aut stet Marpesia cautes (‘than if she had stood a hard granite or a Marpesian rock’). Apart from the fact that he might have considered that as too awkward (anticipating Ausonius’s verdict to avoid line clusters),44 stiffness or hardness is the last thing Petronius wants to evoke in this context.

In contrast with the original context, where Dido’s lack of reaction to the entreating Aeneas is a sign of her bitter and unyielding resentment,45 in the case of the penis its lack of response is a sign of its exhaustion. On each occasion the subject in question does not behave as the other side (Aeneas, Encolpius) expects. Similarly, the lentae salices were in the original (where the singular is used) inferior to the pale olive tree, as was, in Menalcas’s (p.108) judgement, Amyntas’s song inferior to that of Mopsus. In Petronius there is no verdict of inferiority connected with the comparison but solely that of physical incapability. Therefore, equally incongruent is Ecl. 3.83, where the lenta salix is as dear to cattle as Amyntas is to Menalcas. Here the epithet lenta is an adornment (epitheton ornans) without much semantic content, whereas in Petronius its significance is vital. The discrepancies become even stronger in the second half of this line, where the romanticized image of the dying Euryalus, compared in Vergil to a flower (in the tradition of Catullus 11.21–4 and Iliad 8.306–7),46 clashes harshly with the crudeness of the new context’s meaning. At the same time, the bitter-sweet pathos of the Vergilian original confers to the new context an emotional depth which, in an analogous way, Encolpius may well have felt facing his predicament of sexual impotency and its negative consequences for him. On the one hand, the context of the original has to be neutralized to make the new meaning possible. At the same time, however, the old context has to stay in the reader’s awareness in order to achieve its telling effect by way of contrasting transformation.47

In the following, I wish to demonstrate with a few select examples how Ausonius in his Cento nuptialis48 employs in principle the same techniques, which are an intentional application of a stylistic fault criticized by the grammarians, namely, an awkward combination (iunctura) of words that involuntarily creates a vulgar sense and is therefore an ill-sounding expression (cacemphaton).49 Martianus Capella, nupt. 5.518, advises that a cacemphaton which is caused ‘by the intrusion or alteration of words’ (vel interpositione vel commutatione verborum) should be avoided and quotes as an example Verg. Aen. 2.413 atque ereptae virginis ira (‘and the wrath of the rescued maiden’), because ereptae can mean ‘rescued’ as well as ‘raped’.50

After the epistolary introduction in prose and a metrical preface to the Cento nuptialis, there follow six roughly symmetrical sections: II. The Marriage Feast (Cena nuptialis), III. The Bridal Portrait (Descriptio egredientis sponsae), IV. The Portrait of the Groom (Descriptio egredientis sponsi), V. The Wedding Gifts (Oblatio munerum), VI. The Epithalamion (Epithalamium utrique), and VII. The Entry into the Bedchamber (Ingressus in cubiculum). Then a prose apology (Parecbasis) separates this from the final VIII. ‘The Bumping of the Uglies’ (Imminutio),51 which is again followed by a prose (p.109) apology. Formally the cento is a prosimetrum in the tradition of the Satura Menippea. Ausonius employs this literary form for various purposes.52 Here it serves to structure the sequence of action in a subtle way, while also evoking the satirical tradition of the prosi-metrical form. Ausonius’s Cento tells the ‘story of a wedding’ (fabula de nuptiis [Green (1991), 139.20]). However, not the whole wedding is described, as, for instance, the actual ceremony of uniting bride and groom as husband and wife and the signing of the wedding contract are missing. After that, a meal would normally take place at the house of the bride. Then, first the bridegroom and after him the bride would leave the house to go in procession to the house of the bridegroom where the marriage would be consummated.53 Sometimes the whole ceremony could take place in a country estate. We have to assume an arrangement like that in this Cento, as the procession is not mentioned. Other details are documented elsewhere, like the offering of gifts to the couple (lines 57–66),54 the throwing of nuts (line 73),55 and the emphasis on the good looks of bride and groom.56 The term epithalamium, literally a ‘song given at the bridal chamber’, did not emerge from the marriage cult, but was created in Hellenistic times as a technical term for the literary genre of songs sung at the bridal chamber. Later the term’s meaning widened to ‘wedding song’ in general, which could be sung at various stages during the wedding.57 Ausonius follows a combination of the two meanings, as the Epithalamium is obviously sung before the bridal chamber, but with the bride and groom still outside and not already inside as would normally have been the case. Another problem is that elements one would normally expect to be addressed in the epithalamium with its intention to praise the bridal couple (with the topoi of the luckiness of the day, the hope for offspring, etc.),58 are dealt with separately in the Cento, especially the description of bride and groom (lines 33–56). Therefore we have a partial overlapping of the genre: the Cento as a whole contains elements of an epithalamium, and the Epithalamium (VI) itself is part of the Cento (lines 70–8).59 The mirroring of the Cento as a whole in the Epithalamium is also visible from the statement that the young men and women ‘sing playfully in unpolished verse’ (line 69 (p.110) versibus incomptis ludunt) echoing in a self-referential way Ausonius’s characterisation of the weakness of his Cento (quoted on pp. 103–4).60

The difference between the first six parts and the final much cruder Imminutio can be described in Genette’s terms as transposition versus transformation.61 In the first six parts it is easier to transpose the original context into a new one, often without harsh contrast. This can, for instance, be seen in the Epithalamium for bride and bridegroom (lines 67–79). First the bride is addressed (lines 70–2) in phrases originally applied to Nysa in the context of her happy but ill-assorted marriage to Mopsus (70a ≈ Ecl. 8.32), to Juno (addressed by Jupiter, 70b ≈ Aen. 10.607), to Venus (addressed by her son Aeneas, who does not yet know her identity, 71a ≈ Aen. 1.330), to the blonde nymph Lycorias (71b ≈ Georg. 4.340), and to the nymph Cyrene (72 ≈ Georg. 4.380, a particularly clever transposition, as in the original it is Cyrene (= mater) who urges her son Aristaeus to pour a libation of wine to Oceanus, whereas in the Cento it is not the mater who utters the rest of line 72). Then the bridegroom is addressed (73–6); the original addressees were Mopsus (73a ≈ Ecl. 8.30), an enchantress (73b ≈ Ecl. 8.64), Aeneas (74a ≈ Aen. 8.500), and Mopsus (74b ≈ Ecl. 8.29). While the bride is associated with nymphs and the divine sphere, the bridegroom belongs to a bucolic-heroic context. In a way, this continues in 75–6, taken from Aen. 1.74–5, where Juno promises Aeolus (the bridegroom in Ausonius) her most beautiful nymph Deiopea (the bride in Ausonius). The following two lines have a more tragic undertone: 77a is taken from Aen. 9.446, where the fallen warrior pair Nisus and Euryalus are promised immortality through Vergil’s poetry; 77b from Aen. 4.382, Dido’s evocation of the divine powers in a curse when Aeneas announces to her that he wants to leave her; 78a from Aen. 3.493, spoken by Aeneas to Andromache and Helenus at Buthrotum as a farewell, adding that their toils are now over. This and the allusion to the Parcae (78b–79 ≈ Ecl. 4.46–7)62 give the epithalamium an almost ominous ending.

In contrast, in the Imminutio, the break with the original context has to be much stronger (Genette’s transformation), and therefore the effects are more intensive, verging on the grotesque. This technique is facilitated by the fact that certain phrases were already considered to be ambiguous in antiquity, potentially obscene or metaphorical. This becomes clear from scattered comments by the grammarians.63 For instance, Diomedes, Ars 2, Keil 1, p. 451.7 (p.111) comments on Aen. 6.406 (≈ Cento 105b) ramum qui veste latebat (‘the branch that was hidden under the garment’), that this is a cacemphaton, caused by ‘a mistake in word arrangement’ (vitio compositionis).64 Servius points out concerning Aen. 1.159 (≈ Cento 110a) that the area described is purely fictitious (topothesia est, id est fictus secundum poeticam licentiam locus, ‘this is a fictitious location, i.e. a place invented according to poetic licence’). Such a comment almost justifies Ausonius’s transformation of this line for the area of the female genitals. Similarly, Servius remarks on Aen. 2.19, that the caverna described there and at 2.53 (≈ Cento 119) is also to be understood metaphorically, which in Ausonius’s case again means the female genitals. Servius’s learned comment on Aen. 11.817 (≈ Cento 121b) that mucro (‘sharp point’) could denote the ‘point’ (acumen) of any weapon makes it easy for Ausonius to transfer it to the penis. Macrobius, Sat. 6.6.17 quotes among other references Aen. 11.804 (≈ Cento 118) as a praiseworthy example for Vergil’s ‘ability to give words or phrases a new meaning’ (excogitatio novorum intellectuum), which Ausonius certainly managed by applying the line to the penetration of the virginal bride. Ti. Donatus’s comment on Aen. 2.52 (≈ Cento 126b) that recusso means omnia illa percussaictu validissimo (‘all that which is struck by a very powerful thrust’) also illustrates the point well when transferred to sexual intercourse.

But there are also instances, where the centonist breaks with the commentary tradition in addition to ignoring the original meaning of a phrase.65 Servius on Ecl. 10.27 (≈ Cento 10b) offers an allegorical interpretation for minium (‘red pigment’) as aether and therefore as the god Pan, whereas Ausonius takes it in its material appearance and applies it to the red glans of the groom. Ti. Donatus comments on Aen. 4.690–691a (≈ Cento 122–123a) that Dido’s final convulsions before her death express her remorse. In the context of the Cento these lines describe the resistance of the bride while being penetrated. Macrobius, Sat. 4.5.2ff. explains that Aen. 6.122 (≈ Cento 126a) expresses pathos and misericordia, which is completely lost in the Cento, where the phrase refers to the coital movements of the groom.66

Salvation: Proba, Cento

At least from the third century AD onwards, another method showed the authority of Vergil and the fractioned usage of his text, but this time for (p.112) serious purposes, namely, the practice of the Sortes Vergilianae.67 Quotations taken from Vergil (or from Homer, or Hesiod) were drawn like lots and taken to offer a prophetical answer to the respective question. For instance, the oracle of Praeneste applied Verg. Aen. 6.882–3 (referring to Marcellus) to Alexander Severus (Hist.Aug. [Lampr.] Alex. 4.6).

Similarly, Proba’s Christian Cento68 serves a serious exegetical purpose. She wrote it presumably between 353 and 37069 and can thus be seen as inaugurating the tradition of the Christian cento. Before Proba, centos were associated by Christians with heretics and condemned as intending to convey dogmatic lies in a manipulative way.70 Her poem comprises 694 Vergilian hexameters,71 and is thus substantially longer than Ausonius’s Cento nuptialis of 131 hexameters. Her Cento, providing a brief survey of the history of salvation, can only be fully appreciated if one has knowledge of the Vergilian text, the biblical text, and the commentary tradition linked with both.72 The cento has two aims:73 first it conserves Vergil’s verse and language for the Christian reader by using him to paraphrase the Bible (and we may remember that Christianity at her time had not produced much poetry of its own), and second it provides an exegesis of Vergil’s Aeneis acceptable to a Christian reader.74 The exegetical figure used here for literary interpretation is typology, i.e. the exploitation of scenes, figures or phrases of the hypotext to illuminate corresponding parts of the hypertext.75 Truly innovative is the idea that Proba (in contrast to many other Christian authors) accepted that Vergil in all his works proclaims Christian truth and that she therefore understood her cento as a kind of Vergilian exegesis, as stated in the programmatic verse 23: ‘That Vergil sang (p.113) of Christ’s sacred duties let me tell’ (Vergilium cecinisse loquar pia munera Christi).76 This leads to the self-confident position that Proba on the one hand refuted pagan poetry including her own previous poetry (which dealt with the traditional pagan themes of panegyric and war), and on the other felt able to use Vergil to convey the Christian message of truth. This, in turn, makes her cento superior to pagan poetry: it is a sacrum…carmen (‘a holy song’; Cento 9, not ludus, as Ausonius’s Cento, see pp. 103–6 in this chapter), and it intends to reveal arcana…cuncta (‘all mysteries’; Cento 12) and the altae res (‘the elevated issues’) of the history of salvation (Cento 50–1).77 This justifies her pride in assuming the title vatis (‘prophetess and poetess’) in line 12.78

Scholarship, apart from criticizing Proba’s poem as absurd and unoriginal,79 has accused her Cento of being not properly structured, with episodes following one after another at random.80 This is not correct. A close analysis shows that both the selection and the sequence of the episodes chosen by Proba follow theological principles to give a brief survey of the crucial points in the history of salvation. The Old Testament section ranges from the creation of the world up to the flood with the sole survival of Noah and his family, with God’s purpose ‘to have the fundamental stock from which a new stem’s roots could be revived’ (Cento 316 ut genus unde novae stirpis revocetur haberet), echoing the traditional view of Noah’s survival as a second creation and a new order (Cento 317–18 diluvio ex illo patribus dat iura vocatis | omnipotens: magnis agitant sub legibus aevum, ‘the Almighty bestows from that flood rights to the fathers that are called: they live under grand laws’). Cento 316 is taken from Georg. 4.282, where it refers to the death of the beehive. Cento 317 is partly based on Georg. 4.154 (after the passing of the Golden Age, in the Age of Jupiter laws are necessary). The original contexts of both these adapted phrases help to mark the change to another era in the cento, that is the age of the patriarchs; it lasts from the second flood till the birth of Christ and can therefore in a theologically stringent way be encapsulated in a praeteritio (319–22). The birth of Christ (p.114) initiates a prophesied new era (Cento 414 regnisque futuris; 447 haut vatum ignarus venturique inscius aevi) as he will rule the earth and save humanity (Cento 340 auxilium; 345 qui viribus occupet orbem; 348 missus in imperium; 409 tu regere imperio populos;81 418 cura salutis; 472 via prima salutis; 666 triumphi). This second era supersedes the first and the poetic task is therefore greater (Cento 334 maius opus moveo).82 It is not surprising that Proba uses verses from Ecl. 4 to illustrate the birth of the Saviour, as this eclogue had been interpreted as a prophecy of Christ’s birth by Christians from Lactantius, Div.Inst. 7.24 (quoting Ecl. 4.21–45) onwards.83 But there are also other instances where she offers a theological interpretation of Vergil. For instance, Jesus as a Christian Aeneas is also a saviour who transcends the past and by his suffering transforms the future, thus representing a truly epic hero.84 This characterization consolidates the Christocentric unity of Proba’s Cento.85

Furthermore, there are typological correspondences within the poem between events from the Old and New Testament, like the Serpent’s successful temptation of Eve, which is undone by the Serpent’s unsuccessful temptation of Christ.86 This is an unusual constellation, because from Romans 5:12 onwards Christ had been seen as the anti-type to Adam, whereas here Eve and Christ form a typological pair. This serves to strengthen her position and to see her as a representative of humanity in general and not as a negative counterpart to Adam.87 Another motif linking the Old and New Testament parts of the poem is the reversal of the role of the Jews according to Proba: the persecuted Hebrews of the Old Testament (317–32) become the persecutors of Christ in the New Testament (600–24).88

In a striking contrast to Ausonius’s profuse apology for his Cento, Proba is concerned with emphasizing the soteriological framework of her Cento which has a strongly didactic aim: in view of Jesus’s mighty deeds for humanity and his power, and in the face of a final judgement, it is important that people keep their faith and lead a good life. This protreptic aim places this Cento between an epic (with narrative action and a hero) and a didactic poem (with personal (p.115) address to the reader and exemplary illustrating episodes), a fairly common transgeneric combination in late antiquity.89

In order to illustrate Proba’s centonic technique more clearly a passage may be analysed in which she paraphrases the event of Jesus walking on water (Cento 531–61), both in relation to her biblical model and to an analogous paraphrase of the same passage in Juvencus’s hexametrical paraphrase of the four gospels from the early fourth century, the Evangeliorum Libri (3.97–109, 124–6).90 As is commonly acknowledged, Juvencus aims at following the biblical text as closely as possible; the changes he makes are very ‘discreet’. Here, he follows the version of Mt 14:22–33 with details missing in Mark 6:45–54 and John 6:16–21.

Proba makes some remarkable alterations. First, she adds a complete section (Cento 531–44) without a biblical equivalent, which can best be described as amplification with the aim of universalizing the biblical message. She puts the sea journey of the disciples in the context of the general human achievement of catching fish from a boat, and later of travelling over water. The latter, especially, risks exposure to the elements, and sailors often have to fear for their lives. This general setting, which echoes pagan ideas about the dangers of technological progress,91 serves to assimilate the biblical narrative to pagan ideas and to make the biblical message more easily absorbable and convincing for both educated unconverted pagans and converted educated ex-pagans. After this, Proba turns in a rather abrupt way to the specific event of the disciples on a boat caught by a storm (545–61; 545 ecce marking the transition from the general to the specific). Whereas in Juvencus it is quite straightforward to determine the exact biblical model, this is not the case in Proba. This is not simply owed to her being restricted by Vergilian verse, but to the fact that she purposefully reshapes the biblical episode.

Jesus’s approach towards the disciples’ boat is richly embellished (Cento 547–9), emulating Juvencus’s amplification (3.102–4). In Cento 550–1 it is stated that the disciples recognize Jesus at once and salute him. This blatantly contradicts Mt 14:26 and Mark 6:49–50, where they mistake Jesus for a ghost and are frightened. In John 6:19 and Juvencus 3.104–6 the disciples are also frightened at the sight of Jesus. Notably, Proba omits the scene between Christ and Peter who wants to walk on water as well, though this scene is related in Mt 14:28–31 and Juvencus 3.110–23,92 but is also missing in Mark 6:48–50 and John 6:18–20. The instantaneous and unquestioned acknowledgement by the disciples of Christ as king is corroborated by his description as a powerful (p.116) giant: 546 cui summa potestas (≈ Aen. 10.100 of Jupiter); as soon as Christ touches the water the storm subsides (552–5)93 without him even wetting his giant loins (556 latera arduaAen. 3.665 of Polyphemus); he is immediately acknowledged as the rector of the boat (558 ≈ Aen. 5.176 of Gyas, having thrown Menoetes overboard), an element that is missing in the biblical pericopes. Finally, the boat groans under the great weight of Christ (559 ≈ Aen. 6.413 about Aeneas climbing into Charon’s boat, with him being heavier than the shadows),94 which is in ancient thought generally a sign of divinity; 560–1 suggests that no rowing is necessary anymore now Christ is on board, again a pagan idea of a god’s ship moving without physical effort. Instead of the merciful and educational attitude of Christ in the biblical model, Proba focuses in her depiction of the scene on Christ’s power and the spontaneous response of his disciples, thereby describing the act of faith as joyful worship instead of the existential doubt and angst of the disciples, especially of Peter, in the biblical narrative.95


According to Genette, the relationship between a hypotext and its hypertext is always that the latter comments on the former.96 As we have seen, this is certainly true for the cento as well. Several differences between Ausonius and Proba are illuminating: Ausonius takes the grammarians’ tradition of finding cacemphata in an author to its extreme, while at the same time claiming that he speaks the truth, because ‘these things really happen in a wedding night’ (aliter haec sacra non constant, Green [1991], 139.21), and by mentioning Vergil and others as his predecessors, even regarding obscenity. However, Ausonius confesses in the Parecbasis that (by manipulation) he makes Vergil shameless (Green [1991] 137.5–6), which means that he, as the ‘second’ poet, devalues his model (dehonestasse, Green [1991] 133.1). The need for ample apology is therefore understandable. In contrast, Proba, while also insisting that she tells the truth, does not do so by claiming to manipulate the poet’s works (as it were, against the grain of their intended content), but by revealing the true and hidden message of the hypotext. Thus, she revalues or enhances the status of her hypotext. She does not have to apologize to lovers of Vergil as Ausonius has to, but rather to lovers of Christ. Hence she offers a lengthy (p.117) renunciation of pagan sources of inspiration and of her own previous (pagan-style) poetry (Cento 1–55).97 As her cento has a missionary aim, and intends to promote the history of salvation, Proba can well describe herself as part of this history and thus, as part, also, of the content of the poem.98 This contrasts sharply with Ausonius, who in his own original way follows, at the very end of his Cento (Green [1991], 139.1ff.), the path of Catullus 16.5–6: ‘For the pious poet ought to be chaste himself, though his poems need not be so’ (nam castum esse decet pium poetam | ipsum, versiculos nihil necesse est), emphasizing the strict separation of the lifestyle of the poet from the content of his poetry.99

For later Christian poets, Proba’s technique of abbreviating and amplifying her biblical model became clearly influential. On a formal level she (even more than Juvencus) encouraged the further development of a Christian poetic language based on classical models, first and foremost Vergil, but also others.100 The intentionally anti-classical (because anti-pagan) poetical enterprise of Commodianus remained the exception. On the level of content, she helped to establish a genre of epic-like poetry with non-martial content.101

In both Ausonius’s and Proba’s centos, part of the ambition, and of the play, is the enigmatization of Vergil’s phrases.102 The necessary and intended aim of speaking in an obscure way urges the reader to try and make sense of the verses with the frame of reference in mind (wedding celebration or biblical text). The collage of verses always requires a hermeneutical frame within which it may be understood.103 In particular, as our analysis of a single passage in Proba has illustrated, more consideration should be given to Proba’s theological intentions that go beyond literary artistry. Despite the Vergilian straitjacket, her considerable modifications of the biblical text aim at making her own theological statement with a protreptic perspective oriented towards her fellow human beings, urging them to remain steadfast in the Christian faith.

Surprisingly (or maybe not, after the preceding argument), if one judges the quality of a cento in terms of the degree of transformation of a given phrase in its new context, then Ausonius’s sexualization of Vergil in his Imminutio and Proba’s Christianisation in her Cento are closely related from a technical point (p.118) of view. The extent of the success of this technique can be seen in particular clarity when one looks at Aen. 7.66 pedibus per mutua nexis (‘with their feet mutually intertwined’, of a swarm of bees clinging together), which is one of the only two Vergilian phrases104 used both in Proba’s Cento and Ausonius’s Imminutio. Whereas in Proba, Cento 618, it describes the fixing of Jesus’s feet when he is mounted on the cross, in Ausonius, Cento 107, it refers to the intertwining of the couple’s limbs during sexual intercourse. If deconstructionism aims at the elimination of the author from the work, and seeks to dismantle the signifier in order to reconstitute what is always already inscribed in a phrase, this would be a prime example to put this theory to the test.

(1) Juvencus, Evangeliorum Libri 3.97–109; 124–6

97 iamque soporata torpebant omnia nocte,

98 cum puppis medio sulcabat in aequore fluctus

99 iactata adverso surgentis flamine venti.

100 ast ubi iam vigilum quarta statione premebat

101 noctis iter rapidos attollens Lucifer ortus,

102 fluctibus in liquidis sicco vestigia gressu

103 suspensus carpebat iter—mirabile visu!

104 iamque propinquabat puppi, sed nescia nautae

105 attoniti tremulo vibrabant corda pavore

106 clamoremque simul confusa mente dederunt.

107 tum pavidis Christus loquitur: ‘timor omnis abesto,

108 credentumque regat vegetans constantia mentem.

109 en ego sum, vestrae doctorem noscite lucis’.


124 ascensaeque rati contraria flamina cedunt.

125 praesentemque Dei subolem stupuere rogantes

126 cuncti, navigio socios quos casus habebat.

(2) Proba, Cento 531–61

531 inde ubi prima fides pelago, tranquilla per alta

532 deducunt socii navis atque arte magistra

533 hic alius latum funda transverberat amnem

534 alta petens, pelagoque alius trahit umida lina.

535 postquam altum tenuere rates nec iam amplius ullae

536 occurrunt terrae, crebris micat ignibus aether,

(p.119) 537 eripiunt subito nubes caelumque diemque,

538 consurgunt venti et fluctus ad sidera tollunt.

539 at sociis subita gelidus formidine sanguis

540 diriguit: cecidere animi cunctique repente

541 pontum adspectabant flentes—vox omnibus una—

542 spemque metumque inter dubii, seu vivere credant

543 sive extrema pati, leti discrimine parvo,

544 qualia multa mari nautae patiuntur in alto.

545 ecce deus magno misceri murmure pontum

546 emissamque hiemem sensit, cui summa potestas.

547 par levibus ventis et fulminis ocior alis

548 prona petit maria et pelago decurrit aperto:

549 nec longo distat cursu praeeunte carina.

550 agnoscunt longe regem dextramque potentem

551 nudati socii et magno clamore salutant.

552 postquam altos tetigit fluctus et ad aequora venit,

553 id vero horrendum ac visu mirabile ferri:

554 subsidunt undae, remo ut luctamen abesset,

555 collectasque fugat nubes graditurque per aequor

556 iam medium necdum fluctu latera ardua tinxit.

557 at media socios incedens nave per ipsos

558 ipse gubernaculo rector subit, ipse magister.

559 intremuit malus, gemuit sub pondere cumba,

560 vela cadunt, puppique deus consedit in alta:

561 et tandem laeti notae advertuntur harenae.


(1) Emerson (1876), 42.

(2) Pollmann (1996a), 115–17.

(3) In all cases known to me words or phrases are recombined. But Ptolemy Philadelphos went a step further and even recombined syllables to mock this method; see Chapter 6, pp. 140–1.

(4) Genette (1982), 11–15. It is not clear to me why Genette does not discuss the cento in his book, though it would illustrate his point in an extreme fashion. His categories and criteria of narratological analysis will prove useful for the following analysis.

(5) Döpp/Geerlings (1998), s.v. Cento; Kunzmann/Hoch (1994), 148.

(6) In Petronius, Satyricon, see Herzog (1975), 13; see also pp. 106–8.

(7) Ermini (1909), 41–55, among them esp. the ‘tragedy’ Medea by Hosidius Geta consisting of Vergilian hexameters.

(8) Slavitt (1998), 43–75 has rendered the Cento nuptialis by Ausonius in a Shakespearean cento, thus extending the history of the genre into the twentieth century. A monograph on the genre of the cento from its beginnings to the Middle Ages and beyond is still a desideratum in literary history. For examples from late antiquity and the Renaissance, see Herzog (1975), 13, 17–18; for the Middle Ages up to the twentieth century, see Kunzmann/Hoch (1994), 152–6.

(9) For the etymology of ‘cento’, see Kunzmann/Hoch (1994), 148. The term cento is not found in the sense of ‘poem composed of odd fragments’ before Ausonius, but Tertullian speaks of Homerocentones in Praescr.Haer. 39.5; it may have been current long before that.

(10) This became again popular in the eighteenth century, where parts of various newspaper headlines were assembled to produce amusement or criticism of society. In the twentieth century, similar effects were achieved by photographic montages, often as a means of political agitation. For all this, see Riha (1971), esp. 7–40.

(11) For the superior status of Homer and Vergil as almost universal poets see, e.g., Seneca, Consol. ad Polyb. 8.2; Quint. Inst. 1.8.9; Ausonius, Epigr. 137.1; Macrobius, Sat. 1.24.5; Augustine, Civ. 1.3; Orosius, Hist. 1.18.

(12) See Chapter 6, pp. 140–4.

(13) This side of the cento is generally underplayed, even by the theologian Sandnes (2011), 17–22, 107–40.

(14) Whereas it is common to combine phrases from several works of one author, it is less frequent to combine bits taken from various authors into a new unit, but see Lucian, Symp. 17 (a certain Histaeus when drunk had quoted a ridiculous poem consisting of verses from Hesiod, Anacreon, and Pindar). A third of the Byzantine cento Christus patiens consists of verses from Euripides, and (to a much smaller degree) from Aeschylus and Lycophron; see Chapter 6, pp. 144–5 with n. 17.

(15) The literary cento had its ‘haptic’ equivalent in various forms of ancient and Chinese puzzles (the tangram), see Evelyn White (1919), 394–6.

(16) For examples in Proba see Schenkl (1887), 556–9 and his apparatus fontium under the text 569–609; for Ausonius, see Green (1991), 519, 522–4.

(17) From the eighteenth century onwards the montages of newspaper fragments or later of contrasting photographs could both be displayed for sheer entertainment or for the satirical criticism of society, see Riha (1971), 7–46. The latter does not seem to me to be the case in Ausonius, but it could be worth pursuing this in other (esp. Christian) centos.

(18) Genette (1982), 30–7.

(19) In the Greek world, the Homerocentones by Eudocia and others follow, according to Zonaras, Ann. 13.23, the tradition of a bishop Patricius, see Mary Whitby (2007), 207–9, 215–17.

(20) See Kunzmann/Hoch (1994), 149, summarizing earlier scholarship.

(21) See for thorough background information Herzog et al. (1989), §554.

(22) Ermini (1909), 31–6. The text of the Cento nuptialis is quoted from Green (1991), 132–9.

(23) Translation with slight modifications taken from Evelyn White (1919).

(24) For concinnatio as a crucial term in Ausonius’s poetics in general see Sánchez Salor (1991), esp. 133–4.

(25) The mnemotechnic artistry links the cento to the rhetorical training at schools, see Herzog (1975), 5; Herzog et al. (1989), 296; Kirsch (1989), 67–8.

(26) Herzog (1975), 4–6.

(27) On Valentinian I and Ausonius, see Sivan (1993), 106–11. Green (1991), 520 emphasizes the careful diplomacy of Ausonius’s statement about Valentinian’s erudition.

(28) Kirsch (1989), 120–2.

(29) Green (1991), 518; Herzog et al. (1989), 296 suggest AD 368/369.

(30) See pp. 106–8, for Petronius.

(31) His emphasis on composing this cento by imperial command is repeated in the poetic preface (line 10 non iniussa cano). For Proba’s completely different strategy of framing her Cento see pp. 116–17.

(32) Herzog (1975), 10: Laelius Capilupus, Centones ex Virgilio (Rome, 1555).

(33) Slavitt (1998), 75 about the Imminutio: ‘a dirty mind is a great comfort, and…if the devil can cite scripture to his purpose, at least he can take credit for having read some’.

(34) subito calore et quadam testinandi voluptate…, gratiam celeritatis….

(35) Sánchez Salor (1991), 114–21.

(36) Green (1991), 519.

(37) Herzog (1975), 39.

(38) Glei (2013) correctly re-emphasizes this point.

(39) The translation is taken from Connors (1998), 32, who analyses this passage 30–3.

(40) See pp. 103–6 in this chapter.

(41) illa can also be interpreted as elliptic for mentula, see Hofmann (193767), 782.52–4.

(42) This is of course made plausible by the preceding personification of the penis in Sat. 132.9–10. For the personification of the penis in Latin, Greek, and other languages see Adams (1981), 205.

(43) The glans or the whole penis are often described as red, see Ausonius (p. 111 in this chapter) and Adams (1981), 204.

(44) See pp. 103–6 in this chapter.

(45) Macrobius, Sat. 5.2.14 speaks of infesta (‘hostile’) Dido. In Sat. 4.1.1 he lists the passage under those describing pathos (unfortunately the text is here incomplete).

(46) Connors (1998), 32.

(47) Term after Gennette; see pp. 101–2 in this chapter.

(48) The edition is Green (1991); Evelyn White (1919), vol. 1, 370–93 offers an English translation with the exception of the Imminutio, hence Slavitt (1998), 67 feels obliged to offer one.

(49) Adams (1981), 201, 203–4; Lausberg (1998), §964.

(50) Ausonius does not use this line in his Imminutio. Servius does not criticize the line in the way of Martianus.

(51) The translations of the subtitles are taken from Slavitt (1998). The noun imminutio (‘diminution, violation’) is used only here in the meaning of ‘defloration’, see Hofmann (1956–70), 463.11; for the verb imminuere in this sense, see Green (1991), 518.

(52) Gruber (1981), 215–19, who emphasizes that in a prosimetrum an author wishes to show his erudition and expects his or her readers to be sufficiently educated to recognize this.

(53) Treggiari (1991), 161–70.

(54) Terence, Phorm. 39–40; Livy 42.12.4.

(55) Cat. 61.121; Verg. Ecl. 8.30 (used in line 73), for which see Coleman (1977), 234; Festus 183 (p. 178 Lindsay).

(56) Generally considered to be of importance, see Treggiari (1991), 101, 260.

(57) Contiades-Tsitsoni (1990), 31–2.

(58) Typical elements of an epithalamium were the praise of the happiness and the beauty of the couple; wishes for a happy consummation of the marriage, for offspring and marital harmony; see Keydell (1962), 928, 932, who emphasizes 938 the purely literary character of Ausonius’s epithalamium.

(59) This differentiation is not made by Green (1991), 518.

(60) This analysis is a bit more differentiated than McGill’s statement (2005), 95 that Ausonius adapted Vergil to a new genre, i.e. the epithalamium.

(61) Genette (1982), 33–40; 237–45.

(62) The Parcae announce the Golden Age which will be inaugurated by a second Achilles, see Coleman (1977), 150–2; in Cat. 64.321–2 the Parcae sing at the wedding of Thetis and Peleus, hinting at the violent deeds and death of Achilles.

(63) The following examples serve as an illustration of the general approach possible to canonical authors in late antiquity. Due to the relative scarceness of the material, I include partly also material that is slightly later than Ausonius.

(64) Adams (1981), 201 with n. 1. This is again equivalent to the technique of a cento, see pp. 103–6 in this chapter.

(65) Herzog (1975), 7 calls this technique ‘neutralization’.

(66) See pp. 106–8 on Petronius, in this chapter.

(67) Ermini (1909), 37.

(68) Schenkl’s edition (1887) is still the standard one. Clark/Hatch (1981) reprint it and offer a helpful English translation. A new one is contained in Schottenius.

(69) See the discussion in Green (1997), 548 n. 5 (against Timothy Barnes), followed by McGill (2007), 186 n. 3. The matter is still controversial, with Bažil (2009), 112 arguing for a date of the poem between 350 and 370 (followed by Schottenius [2015]), Badini/Rizzi (2011), 13–19 before 370, and Sandnes (2011) between 350 and 390.

(70) See Tert. praescr.haer. 39.

(71) The first fifty-five lines of the Cento are mainly non-Vergilian due to their introductory nature. The manuscripts partly also transmit a prefatory poem of fifteen non-Vergilian hexameters, beginning Romulidum ductor, which is not by Proba, see Green (1997), 548–9.

(72) Herzog (1975), 36. For a discussion of Cento 38–42 and 108–9 in this light, see Pollmann (2002), 226–30.

(73) See esp. Herzog (1975), 16–46; Kirsch (1989), 120–3; Jensen (1991a), 47–8; 55; Margoni-Kögler (2001), 143, 151.

(74) It is not clear why Green (1997), 556 denies that Proba reinterprets Vergil whereas 558 he says so himself.

(75) In a literary, but not theological way, this is partly also the case in Ausonius’s Cento, see p. 110, in this chapter. See Margoni-Kögler (2001), 141–2 with further literature. Augustine, De ordine 1.5.12 and 7.24 (written 386 AD) allows a Christian to write philosophical poetry that is wholly allegorized, so that, e.g., the story of the lovers Pyramus and Thisbe would appear as a rarefied allegory of the love of a wise person for wisdom.

(76) Buchheit (1988), 166, 172, 175–6; Jensen (1991b), 87.

(77) Kartschoke (1975), 60–3 with parallels in other Christian writers. Proba shares this ambivalent attitude towards pagan culture and literature with many of her Christian contemporaries, see the survey in Kirsch (1989), 140–50. Nodes (1993), 13–16 rightly emphasizes the theological ambitions of Proba, who does not intend just to offer literary play for an educated audience.

(78) For vatis as nominative see Green (1997), 553 n. 28.

(79) See e.g. again Smolak (1999a), 12–13. This goes eventually back to Jerome’s hostile assessment in Epist. 53.7 puerilia sunt haec [= Homerocentones et Vergiliocentones] et circulatorum ludo similia. In reality, however, Proba’s Cento was very popular, as is already visible from the numerous manuscripts and editions (Ermini (1909), 63–70); see Isidor, vir.ill. 18.22; etym. 1.39.26. Boccacchio also dedicates in his De claris mulieribus a chapter to Proba.

(80) Even Kirsch (1989), 127 claims this, despite his otherwise excellent analysis of Proba’s Cento. Herzog (1975), 14–16 is among the few who emphasize the unity of Proba’s Cento; but he is wrong in stating 19 that Proba does not have any theological aims with her composition, which he nevertheless concedes 41 for parts of the poem.

(81) Clearly a Christian adaptation of the Roman missionary self-awareness as a force civilizing humanity, as expressed in the famous lines Aen. 6.851–3, with the quotation here taken from 851a.

(82) See Clark/Hatch (1981), 191 n. 33.

(83) See esp. Cento 34 iam nova progenies ~ Ecl. 4.7 and the cluster at Cento 377–9 which takes up phrases from Ecl. 4.18–20; 23; 28; see in general Clark/Hatch (1981), 171–81. Benko (1980), 670ff. does not mention Proba at all.

(84) For this and other examples see Kastner/Millin (1981), 39–42.

(85) Margoni-Kögler (2001), 149; see also pp. 115–16 in this chapter.

(86) See in more detail Clark/Hatch (1981), 161–9; Margoni-Kögler (2001), 151.

(87) Jensen (1991a), 51–3, who emphasizes that Proba avoids the more traditional typology of Eve and Mary (54–6); Leisch-Kiesl (1992), 150–7, esp. 155–6; Margoni-Kögler (2001), 149; more critical are Clark/Hatch (1981), 151–9.

(88) Poinsotte (1986), 101, emphasizing 111 that in Proba the Jews are characterized as perpetuating deicide.

(89) Kirsch (1989), 123–5. Genette (1982), 15 emphasizes the partly transgeneric character of hypertextuality in general.

(90) The Latin texts of both passages are reproduced in the Appendix at the end of this chapter.

(91) e.g. Horace, Ode 1.3; for parallels, see Nisbet/Hubbard (1970), 43–4.

(92) Where the lack of understanding is particularly emphasized, see Röttger (1996), 108–9.

(93) This is not explicitly stated in the biblical models for this story. But in Mt 8:26–7 this is said in a different context.

(94) Christ is modelled throughout the Cento after Aeneas, see n. 81.

(95) In Mt 14:33 the disciples recognize Jesus eventually as the son of God; see also Mark 6:54.

(96) Genette (1982), 14–15.

(97) Herzog (1975), xlix–li.

(98) Herzog (1975), 46–51 emphasizes the edifying aspect of the personal passages in Proba’s Cento, which serve to illustrate that the poetess is part of the narrated story.

(99) The lines by Catullus are quoted for a similar purpose by Plin. Ep. 4.14.5, Ovid, Tr. 2.354, and Martial 1.4.8. Ausonius quotes Juvenal, Sat. 2.3 instead.

(100) Kirsch (1989), 139. Ermini (1909), 109–41 offers an almost line-by-line list of phrases in Proba’s Cento which appear also in Juvencus and in later Christian poets.

(101) Kirsch (1989), 137.

(102) Therefore the criticism in Kirsch (1989), 133 that Proba’s Cento is not clear enough, is not quite appropriate. The character of riddle and allusion are an essential part of a cento, see Kartschoke (1975), 35.

(103) Kirsch (1989), 122–3.

(104) The other is Georg. 1.142 alta petens (‘seeking the bottom’ of a river while fishing), which in Proba 534 denotes the same thing, and in Ausonius, Cento 105 is used for the groom seeking to penetrate the bride. These two instances do not provide sufficient evidence to regard Ausonius’s Imminutio as a literary ‘answer’ to Proba’s Cento.