While a considerable amount of scholarly energy has been devoted to the Latin versions of the Passion and Acts of the African martyrs Perpetua and Felicitas, by comparison rather little serious attention has been devoted to the Greek translation of the narrative of their martyrdom. Such an investigation requires a focus not just on technical problems of the similarities and differences between the Greek translation and a putative Latin original, but also attention to the more strategic problem of its place in the context of translations of Latin Christian texts. Although a Greek translation could have been made soon after the first appearance of the Latin narrative, this essay argues that a more likely context for the translation and for a heightened interest in the cult of Perpetua in Italy and in the East is a much later fifth- and sixth-century one. When we consider the cultural as much as the literary “translation” of Perpetua's martyrdom, we see that the drive to exploit the images and social power of a specific group of African martyrs explains the emphasis placed on them not only in Africa (specifically at Carthage) but also in a cluster of sites at the head of the Adriatic. These particular connections logically suggest concomitant ones with the eastern Mediterranean of the Byzantine state of the fifth and sixth centuries C.E.

Veritable rivers of interpretive scholarship have poured forth on the Latin accounts of the martyrdoms of saints Perpetua and Felicitas—a long descriptive narrative, the Passio, and two short Acta. All three versions have merited comment, although by far most of the attention has been lavished on the long Passio account.1 By comparison, rather little effort has been devoted to the significance of the Greek translation of Perpetua's martyrdom. The single manuscript containing the Greek account was discovered in 1889 by the Cambridge Quaker James Rendel Harris in the Library of the Convent of the Holy Sepulchre in Jerusalem—that is, the library of the Greek Orthodox Patriarchate.2 In the decade immediately following its discovery and publication, the Greek text evoked much interest, but then rather little.3 The first and practically only focus of intense attention that emerged in the aftermath of its discovery was a debate over the relation of the Greek text to the long Latin narrative.4 At first, Rendel Harris, followed by some others, held that the Greek was the original text and that the Latin version was a translation of the Greek. Initially, Harris's claim received support from no less an authority than Harnack who, to his credit, quickly changed his mind.5

In more recent times, no less an eminence than Louis Robert held that the Greek text was the original. As one of the most authoritative of modern-day Hellenists, Robert was able to ferret out “the Hellenic” in sometimes unusual circumstances. It is perhaps not surprising that he was provoked by the presence of Pythian-Apollonian games at Carthage to think of Roman Carthage as a kind of imperial Greek polis.6 This was an excusable overreach by a great scholar who was perhaps exporting too much of what would be sensible in an Anatolian or eastern Mediterranean context to the Latin West. Robert's claim was part of a long stream of arguments that continued unabated from the late nineteenth century about whether the Greek or the Latin was the original text.7 The inconclusive nature of these debates surely created the frustration that provoked Adolf Hilgenfeld, as early as 1890, to propose that the original was composed in Punic, a claim that the Abbé Duchesne described as tending more to excite astonishment than any serious discussion.8 I can only report that I was fortunate enough to happen across the author's copy of Rendel Harris's publication along with his full marginal annotations. From these notes, it is clear that Harris had come to doubt his original assertion and realized that the Latin must be the text from which the Greek translation was made. Indeed, he had made this decision very soon after his ill-conceived assertion concerning the prior status of the Greek.9 In fact, rather early on detailed studies of the language and the technical terms in the Greek version demonstrated beyond reasonable doubt that it was the translation of a Latin text.10 Even if this Latin account was not precisely the Passio that we now have, the one that we do possess is still fairly close to the text that the Greek translator used.

Since the Greek version of the Passion displays clear signs of independence of the long Latin narrative, we are not confronting a translation that straightforwardly links the two texts. Christian texts were frequently subject to additions, deletions, and alterations that suited various parties over time, and there is no reason to believe that these texts were any different. Although there are good grounds to believe that the text of the Passio was fairly stable, the existence of different streams of transmission of slightly different versions virtually guarantees that the eleventh-century Codex Casinensis, the basis of modern editions, was not the same as the original Latin text of the early third century C.E. It is almost certain that the Greek is a translation of a slightly different Latin Passio account than the one that we now possess.11 For example, the Greek text contains critical pieces of information that are not in the Latin narrative but that seem to have been present in an earlier Latin version. One of these details is particularly important. At the beginning of the narrative in the Greek text, the author, who was framing the report of Perpetua's own words, states “The young catechumens were arrested in the lesser polis of the Thoubourbitanoi”—᾿Εν πόλει Θου[βου]ρβιτάνων τῇ μικρότερᾳ συνελήφθησαν νεανίσκοι κατηχούμενοι. The additional clause in Greek containing the name of the home town of Perpetua and her fellow Christians is prefixed to what is the beginning of our Latin narrative.12 That a Greek translator would have happened to select a specific town in the hinterland of Carthage, Thuburbo Minus, by chance seems most improbable.13 The information about Thuburbo might be otiose, since a claim has been proffered that the martyrs and their families did not in fact come from Thuburbo, but from Carthage. The evidence and arguments in support of the hypothesis, however, are not compelling.14 There was a choice between two towns called Thuburbo: the lesser, Thuburbo Minus, a relatively insignificant place, to the northwest of Carthage; and the greater, Thuburbo Maius, the larger and more important center located to the southwest of the metropolis and the more obvious place to pick to an untrained outside eye.15 The less conspicuous town that was Perpetua's home would not be the natural choice based on a happenstance guess about her origins. The way that the translator transcribes the name of the town shows that they probably had a Latin text in front of them referring to the precise Thuburbo.

It is important that an indication of Perpetua's patria appears twice in the ‘A’ recension of the Latin Acta, first in the introduction, where the initial action is said to have taken place apud Africam in civitate Turbitanorum, and again in the conclusion where the words identifying the hometown of the martyrs are repeated verbatim.16 The writer of the second Latin Acta, the ‘B’ version, also knows of the home town of the martyrs. In the introduction, they were arrested in civitate Tuburbita; and in the conclusion the events are likewise said to have happened in civitate Tuburbita.17 The same recollection of the name of the town appears in the titulus of the B family of manuscripts of the Latin Passio: Passio SS. Felicitatis et Perpetuae, quod est nonis Martiis in civitate Turbitana.18 The distinct possibility of a more precise knowledge of her patria in the Latin manuscripts of the Passio is indicated by readings in the tituli of some late and marginal manuscripts that refer specifically to a Tuburbitana minor.19 However the Greek translator accessed the information—surely through one of the alternative texts hinted at in these minor MSS—the Greek text has good information on the precise home town of Perpetua and her fellow martyrs. In this connection, we might also note that there is a lapse that the first of the Latin Acta (‘A’ version) and the Greek translation have in common. The Latin Acta states that Perpetua and Felicitas came from the city of the Turbitanorum and the Greek account that they were from the town of the Θου{κ}ρ<β>ιτάνων. When they wrote the name of the city, both writers dropped out the same middle syllable of the city's name.20 The similarity raises questions about who was copying whom or, at least, about the relationship between the texts that were reproducing a shared error.

The patria of the martyrs is not the only place where the Greek version has preserved better knowledge. The Latin Passio specifies that the imperial procurator Hilarianus had taken over as acting governor from the proconsul Minucius Timinianus, who had perished while in office.21 By comparing imperial careers in epigraphical texts that we possess, it is certain that the cognomen Timinianus is a mistake and that the proconsular governor was in fact named Minicius Opimianus.22 The Greek text reads Μινουκίου ᾿Οππιανοῦ (viz. Minoukios Oppianos) which, although it somewhat misconstrues both the nomen and cognomen, is still close to the actual name of the governor, Minicius Opimianus. The Greek text and not the long Latin narrative is certainly more accurate on this item.23 Just as with the transfer of the name of Thuburbo, the translator has dropped out the middle syllable of the proconsul's name—a lapse that once again points to a link between the Greek and the ‘A’ recension of the Latin Acta.

Other connections between the Acta and the Greek translation are worth noting. First of all, both mark their own externality. The first titulus of the Greek version states that the events happened “in Africa” (ἐν Ἀφρικῇ), surely indicating that its writer was not only assuming a readership that was outside Africa, but revealing that they themselves were writing from Africa. The author of the ‘A’ recension of the Latin Acta also twice states that the events happened apud Africam, which suggests that the writer was also assuming a readership outside Africa and was, similarly, not living in Africa. Notably, for a Latin writer, the writer of the Acta echoes other odd lapses or misunderstandings that are found in the Greek. For example, this text gives Perpetua's age but omits the conditional qualifier circiter. The concatenation of shared differences and similarities suggests that the author of the Latin ‘A’ Acta somehow had information shared by the Greek version that contained accurate data not found in the Latin Passio and that they also shared the same omissions found in it.24

The matter of Perpetua's hometown and the name of the governor hearing her case are just two striking items where the Greek text is technically superior to the long Latin narrative. But there are many other subtler words and passages, though no less important ones, where the access of the Greek translator to a more detailed Latin original is apparent. Whatever its merits or demerits as a reflection of a Latin text, the Greek translation has distinctive elements throughout that in some fashion differ from the Passio narrative. Indeed, it preserves a good number of different readings, major and minor, that are improvements on those in the long Latin narrative that we possess. Consequently, the Greek must be constantly consulted in the restoration of the most accurate text that modern editors can establish.25 For all these reasons, and because it contains important data relevant to Perpetua's life, the Greek text merits more consideration in its own right.26 Despite the utility of analyzing its small different readings for matters of dating and derivation, it must be emphasized that the Greek version is mostly a detailed and faithful reflection of the Latin narrative as we have it. To understand the nature of the relationship between the two versions, however, we must also understand some of the basics of their difference.

In beginning to make a comparison of the two texts at a macro-level, we might note that the long Latin narrative, the Passio, represents the collation of the texts of three different authors. These are the framer whose complex, rebarbative Latin introduces and concludes the whole account; Perpetua's own words, cast in a simple conversational style, placed at the center of the narrative; and, finally, a second first-person report by one of her fellow martyrs named Saturus that is directly appended to her words and is couched in a more standard Latin. The three accounts are distinguished not only by obvious differences of expression, syntax, and diction, but also in more subtle matters such as their rhythmic clausulae.27 The Greek translation includes all three distinctive parts of the Latin original in order, carefully following the narrative sequence sentence by sentence.28 Generally speaking, the translation approximates the close mode of translation followed in the East for translating texts that were considered to have a special sacred or liturgical aspect.29 Divergences from the Latin are often, though not always, due to the choices or options normally found in any translation.30 In some cases, the writer has no idea of the Greek equivalents of technical Latin terms and either mistranslates, glosses over them, or offers reasonable substitutes. In other instances, the translator has simply dropped parts of the Latin text that we have, mostly at the level of omitting this or that clause, but nothing (as far as I can see) of great significance. Most of the alterations could be the result of arbitrary choices natural to any translator who was using what they considered to be more appropriate words or concepts to convey the message to a different readership, access to a slightly different Latin original, or simply lapses of attention. Despite some details that are important for the establishment of a text, most of these divergences do not shed much light on the cultural background of the translation.

The Greek version is surely worthy of consideration on grounds other than its explicit contents or its closer fidelity at some points to a putative original. First of all, it deserves special notice for a history of communications in the Roman empire.31 Being one of the rare cases where we have a Greek translation of a Latin literary text, the translation deserves analysis in the context of cultural transfers in the empire. Important understandings can be gleaned from the comparison by paying close attention to the differences in pitching the message. How a translator wished his or her intended audience to understand the original is often revelatory of the cultural place of the new text. In considering the Greek translation of a Latin original of Perpetua's martyrdom, an obvious but not generally remarked upon fact is evident: from their inception all martyr acta and narratives from Africa were composed in Latin. In itself, this is a significant development that perhaps deserves more emphasis than it has hitherto received.

We might begin by noting that the sacred or technical language of Christianity in the West remained Greek for a long time, certainly into Perpetua's own lifetime. Backstopping this superior standing of Greek in Christian discourse was the parallel elevated status of a secular elite culture in two languages, Greek and Latin. We find praise of this specific language duality to the end of the Severan age of the empire in the West in texts that laud those who were utraque lingua eruditi.32 This elite bilingualism was especially celebrated in African inscriptions, and it is evident in the self-presentation of its highly educated persons, like Apuleius. Like more than a few elite young women of her time, it appears that Perpetua herself had been educated in Greek as well as Latin.33 We are told, albeit in a dream sequence, in the vision of Saturus, that when Perpetua met the bishop Optatus and the priest Aspasius, she was able to address them in Greek.34 The African ecclesiastical elites represented in her narrative, like the bishop Optatus and the priest, were most probably well versed in Greek. Indeed, it might well have been their first language. They had superior access to and control of the sacred texts of the faith. The Greek names of Saturus (Satyros) and of Perpetua's deceased brother, Deinocrates, might reflect some of this bilingual background, but they might signify nothing other than the idiosyncratic proclivities in assigning names.35 In his edition of both texts, Franchi de' Cavalieri thought it possible that Saturus' account might originally have been composed in Greek, an hypothetical that has found some supporters.36 A close study of the Latin text confirms, as the anonymous editor states, that Saturus' account was indeed composed by a writer other than Perpetua.37 But there is no need, apart from the writer's name, to think that he wrote his account in Greek.

As far as the language world of Christians was concerned, biblical books were in Greek and remained so for a long time even in the Latin West, because they were kept in an original sacred language that was regarded as authoritative. Many texts that aim for what we might call a “near canonical” effect, like the Shepherd of Hermas, were similarly composed in Greek. In stark contrast, from the very beginning in Africa the narratives and other court-transcript-like accounts of the deaths of God's witnesses were written in Latin. Insofar as there are some Greek terms found in Perpetua's Latin, it should be noted that almost all of them, as well as those found in the framer's preface and tailpiece, were technical words of Christian Greek which, for that very reason, were maintained in Latin transcriptions: κατηχούμενος, τέκνον, διάστημα, φιάλη, ἀγῶν, ἀγάπη, ἔκστασις, ἅγioς, προφήτης, and ὅραμα among them.38 The martyr acts composed in Latin were therefore a striking departure from the norm for Christian writings. We should remember that in these same years, even at the center of the western Christian world at Rome, the bishops were native Greek speakers and Greek was virtually the “official” ecclesiastical language until Victor became bishop of Rome—significantly himself an African whose native language was Latin. Only in his pontificate, beginning in the last decade of the second century (189–98?), did Latin first become a normal language of formal written communication in the western church.39 It is not surprising that the only martyr texts that pre-date his pontificate, the three accounts of the martyrdom of Justin, executed by the Urban Prefect about the year 170, were based on Latin transcripts of his court proceedings. But there is no probative evidence to indicate the reliance of the Greek texts on a putative original Christian narrative in Latin.40 Every piece of evidence that we have suggests that whatever primary evidence upon which they were based, as literary compositions the originals were written in Greek. But they were the last ones to be so composed in the western empire.

If the Greek text of the martyrdoms of Perpetua and Felicitas had been the original, it would have been a striking exception in the new genre of martyr texts in Africa that from their inception were designed for Latin-speaking audiences.41 It deserves special emphasis that the Greek text as we have it is one that has been translated into Greek from a Latin original at some subsequent stage in the communication of Perpetua's story. This simple fact, drawn to our attention by this double text, indicates how unusual the African martyr accounts in the West were as Christian texts. That they were originally written in Latin indicates that they were composed for direct consumption by local audiences whose primary public language was Latin. In the standard use of Christian texts in Greek, there was a communicative gap that was usually bridged by some type of oral transfer. The person learned in the original Greek text, whether the Bible or ancillary texts, would convey its meaning by an oral mode: discussion, teaching, or public explanation. This type of glossing did not have to be employed with the martyr narratives and acta in Africa. Because of the high status of Christian texts as sacred writings, there was a long-term move to provide Latin versions of them, from canonical biblical texts to other texts of related value, like the works of Josephus. But secular texts of Christian interest, like the Antiquities of Josephus, were only translated later on—in the case of Josephus, probably in the 570s—notably in northern Italy, in Cassiodorus' Vivarium at Squillace.42 But this is the opposite side of this process that is our focus here.

More significant for our investigation is the fact that a Latin text, a piece of Latin literature, was translated into Greek. This simple observation stands out because of the exceptional nature of the translation itself. In fact, very little Latin literature was translated into Greek. The Greek-Latin translation project was almost wholly a one-way street, with a vast range and number of writings in Greek being translated into Latin, but almost nothing going in the opposite direction.43 As for anything at all transiting from Latin in the West to Greek in the East, what little evidence there is indicates that a very small and select number of established classics like Vergil, Cicero, and later Ovid, were accorded a very low-level but sufficient interest in the Greek-speaking East to provoke translation.44 Even in these limited cases, the extension of these translations and their impact—for example, in terms of their apparent readership—appears to have been virtually nil. The same directionality in translation was just as true of Christian writings.45 Large numbers of Christian texts of all generic types were translated from Greek to Latin, but a truly exiguous few from Latin to Greek.46 Much the same applies to the translation, the movement, of the cult of African saints and martyrs to the East. The Carthaginian bishop Cyprian was surely the most elevated and famous of them all, yet his transition to the East was a bubble that quickly burst. In translation, he was almost immediately conflated with Cyprian of Antioch, and in the process his African-ness and western character disappeared.47

An important conclusion that emerges from this contrast is that Perpetua's narrative is one of the very few western texts in Latin that was thought to be worth translating into Greek, the dominant language of early Christianity.48 We must wonder why. As has been pointed out, the number of Christian texts in Latin that received translation into Greek in Perpetua's time and that were known in the Greek East are more or less countable on the fingers of one hand.49 For our purposes, it is significant that one of the few attested cases is Tertullian's Apologeticum.50 What appears elsewhere as a close nexus of Tertullian and Perpetua might not be mere happenstance. Tertullian indicates that he composed earlier versions of his works in Greek or, at least, composed arguments similar to the ones later found in his Latin texts, like the De spectaculis. Others include the De corona, the De baptismo and, probably, the De virginibus velandis.51 These works were written in Africa, almost certainly for consumption by Christian communities in the large coastal cities like Hadrumetum and Hippo Regius, but mainly in Carthage and its environs: communities that would have expected formal Christian writings to be in Greek.52 Unfortunately, we know so little about the place of Greek in normal discourse, or the influence of Greek-speaking communities in these cities, that it is difficult to provide a better context for such writings and translations.53 Our ignorance on these matters, however, should not prevent us from acknowledging their considerable significance for a history of communications in the empire. The subsequent reworking of his early ideas into Latin by Tertullian and his first literary efforts solely in Latin must have been a deliberate decision.54 This background suggests that it is possible that there could have been a translation of Perpetua's martyrdom into Greek in the first decades of the third century C.E.

Of special relevance to our problem is the fact that the only other martyr text in Latin that we know to have been translated into Greek is the trial of the Scillitan martyrs in 180 C.E. Like Perpetua they were also from Africa and from the same first age of celebrating martyrdoms in writing.55 There is no reasonable doubt that the earlier text of the Scillitani, significant parts of which consist of not much more than an abbreviated and literarily edited version of a court transcript, was in Latin.56 The question then arises as to when the translation from Latin to Greek was done. It could have been made at an early time contemporary with the first age of African martyrdoms, but it could have been rather later when the cult of the Scillitani appears to have been transported to southern Italy.57 It was also in a later age when there was sufficient Greek interest, perhaps provoked by numbers of western exegetes and others coming to the East in the late fourth and early fifth century, to incite translations or versions in Greek of the lives of western saints like that of Melania the Younger.58 Even so, Melania's case is a special outrider since she had decamped from Italy and Africa to live in Jerusalem where she remained resident until her death in December 439. In her case, there was a direct on-the-spot local incitement to produce a Greek version of her life. Much the same applies to the Greek translations of Jerome's Latin lives of Malchus and Hilarion. The provocations, again, were local—not only did Jerome spend time in the East and compose these lives when he was there, but the subjects themselves were eastern monastics.

There is an interesting error in the Latin version of the acts of the Scillitani that might indicate a later date and context for the text: the fact that one of the two consuls used in the dating formula at the head of the account is called Claudianus whereas his correct cognomen was Condianus.59 The error made by the copyist is understandable. Condianus was an exceedingly rare cognomen, whereas Claudianus was commonplace. Probably not being able to understand the very odd Condianus in the record that he was transcribing, the copyist translated it as the commonest name close to what he imagined that it must have been. The importance of this lapse for our argument is that the Greek translator translated the Latin text of the martyrdom of the Scillitani with this error in it. If the error been introduced into the Latin text by later copyists, this would indicate that the Greek translation was made at a later date when the error had already crept into the Latin manuscript tradition. This hypothetical scenario is not necessarily the only one. The error might have occurred as soon as the original redactor produced his/her own transcript-like account filled with verbal quotations from the governor and the defendants. It is possible that even at this early stage a copyist had misunderstood the odd name of one of the two consuls. Unlike the governor of Africa, who could have been well known locally, the consuls might have been known only by a dating formula at the head of the original court transcript. By far the more probable scenario, surely, is that the Latin text that we have is a later copy made when the odd cognomen of Condianus no longer made sense. This would seem to indicate a fourth-century context, or perhaps an even later one.

There are also signs in the Greek translation of the martyrdom of the Scillitani that indicate a later date for it as well. The names of the martyrs are prefaced throughout with the title of ὁ ἅγιος/ἡ ἅγια, “the saintly” or “the holy.” The same additions of saintliness also occur in the Greek translation of the Perpetua text. In it, the blood of Saturus is described as “holy,” a quality that does not exist in the original Latin; as are the bodies of the executed: they, too, become “holy.”60 What is more, the Roman dates—the only ones found in the Latin original—are glossed in the Greek by adding the month and day dates of the later Christian calendar. So the dating in the Latin text of xvi [sedecim] kalendas Augustas is translated in the Greek but is then followed by additional words that specify that this is 17 July: ὅπερ ἐστὶν Ἰουλίῷ ιζ.61 This type of dating formula also points to a late date for the Greek version of the text. The use of continuous dates in the month, as opposed to the traditional Roman kalends, nones, and ides system, first appears in the West in the late fourth century and becomes common only later. In the East, the new dating form was adopted from the fifth and sixth century onwards.62 This particular version of the Greek translation was probably made in Late Antiquity, an age when a number of other sacred Latin texts were being translated into Greek. Furthermore, since Greek was the target language, real suspicion must fall on the time of the early Byzantine presence in the western Mediterranean.63 Finally, there is another significant parallel between the Greek translation of the Acta of the Scillitani and the Greek translation of Perpetua's story. The Latin text contains no notice of their place of origin, but the Greek translator somehow had access to this information, presumably in the Latin text available at the time. The earlier parts of the transcript that would contain this information were omitted in the editing of the Latin version that we possess. The Greek text refers to Iskhlē as their place of origin, surely the translator's version of Scil(l)ium.64 Like the Greek translator of the Perpetua text, the Greek translator of the acts of the Scillitan martyrs might have had access to a slightly different Latin version from the one that we possess.

Turning to the Greek translation of Perpetua's martyrdom, we might begin with the two titles at the head of the manuscript. The first titulus reads: “The martyrdom of holy Perpetua and of those who ended their lives with her in Africa. On the fourth day before the nones [i.e. the 2nd] of February. May He [sc. God] bless [us].” The second title, composed independently of the first, reads: “This happened in the persecution of Valerian and Gallienus in which were martyred Satyrus, Saturnilus (viz. Saturninus), Revocatus, Perpetua, Feliciētas (viz. Felicitas) on the nones (i.e. the 5th) of February.” Clearly, the person who wrote the first title was from outside Africa. As has already been noted, no internal Latin document from Africa ever took the trouble to state the obvious: namely, that its events were happening apud Africam. Both the day and the month of the martyrdom are different than the ones, the nones of March, found in the Latin tradition. The second of the prefatory headings is arguably a late fourth- or fifth-century confection, or even later, when more than one revamped martyr account was re-dated to the Valerianic persecution of 258–59. In other cases, admittedly far more numerous, martyr narratives were arbitrarily restaged in the Diocletianic “Great Persecution.” If a martyrdom was pre-Decian, being one of the localized and isolated parochial events that produced martyrs at various places in the empire, the later tendency was to re-embed the story in one of the well-known empire-wide persecutions. It was through this same process that the date of the Valerianic persecution for Perpetua's martyrdom became popularized in the mediaeval Legenda aurea. For purposes of enhancing her power, the re-dating usefully linked Perpetua with Cyprian, the great noble martyr of the African church.65 More important for our purposes is the fact that this is the same dating found in both recensions of the later Latin Acta of Perpetua's martyrdom. Exactly as in the Greek version, the ‘A’ recension of the Acta states that ‘a persecution broke out in the reign of Valerian and Gallienus’ while the ‘B’ Acta states that the events happened “in the consulships of Valerian and Gallienus.”66 The re-datings point decisively to the late antique provenance of the Acta. There is no reasonable doubt that the much reaffirmed date of the nones of March (7 March) found in a line of western sources and in some epigraphical sources, is as accurate a date as we are likely to get.67

The dates and the brevity of the late versions of the Acta of the Scillitani, as well as the shorter Latin Acta of Perpetua's martyrdom, indicate that they were primarily intended for liturgical and catechetical purposes. The Latin Acta of the Scillitani is palpably of late date, probably of the fifth century or perhaps even later. The existence of one of the short recensions earlier than the fifth century has been postulated, but the case must be regarded as undecided.68 In fact, no African writer, neither Augustine nor Quodvultdeus of Carthage, refers to the Acta, which is concordant with the text being produced outside Africa for a non-African audience.69 The later time of the fifth or sixth century is surely also the context of the preservation of Perpetua's Greek narrative in the East. It is not accidental that the account of a martyr who is said to have died in February is bound together with a series of “lives of saints and martyrs” for that month.70 The problem is that both of the headings to the Greek text might be later additions that do not necessarily indicate much about the time or place of the translation that they label. It is within the realm of possibility that the original Greek translation had been done in Africa and that the manuscript titles that we have were affixed to a later and subsequently altered version of the original translation. All that can be said is that the purpose of the dated prefaces, signaling that these events happened in Africa, was to locate the events for persons outside Africa who were expected to be the readership of a Greek text. The original Greek translation was probably more accurate as to details of place and personal names than is the manuscript that we have. Errors seem to have crept into its text in the process of copying. These problems move us forward to a brief consideration of content and form.

Sometimes bluntly, sometimes more subtly the Greek translator shifts individual meanings in the Latin original to make the text more readily understandable for a Greek readership. Roman peculiarities are avoided. For example, Roman exemplarity is reshaped throughout. At the beginning of the text exempla become “teachings” and at the end “exempla no less important than ancient ones” is replaced with a simple reference to the “writings” of the ancients.71 In this and many other small ways, the Roman-ness of the world in which the martyrs lived is lost in translation. I offer a few examples. The backdrop of the Roman law is never once adequately transferred. The framer who continues the narrative of Perpetua's last days to her death says that he gets to do so because he had received a mandatum and a fideicommissum from her. The Greek translator simply bails on the latter and interprets the former with a general verbal form: “she ordered him” to do this.72 The translator did not understand the Roman system of expressing ages and so makes Perpetua 22 years old whereas in the Latin it is precisely stated she was “almost of 22 years.” And not understanding the force of the Latin circiter in age computation, the qualifier is just left out. The Latin states that she was “almost 21” in our terms, that is 20 years old in our computation.73 So the translator makes her two years older than she actually was. Technical matters like this would be ones of relatively minor interest, save for the fact that they indicate a Greek translator whose first and principal language was Greek.

The translator knew Latin well enough to translate most of the text more or less correctly, but their ability to convey value-laden Latin terms was challenged right from the beginning. The picture of Perpetua as she is introduced in the long Latin narrative is the succinct portrait of a very Roman woman: Inter hos [i.e. the apprehensi] et Vibia Perpetua, honeste nata, liberaliter instituta, matronaliter nupta (2.1). The Greek offers instead a depiction of someone of “noble birth, grown in wealth, and brilliantly married,” basically misconceiving the value of being honeste nata and making her “noble”—a suitable for social rank in Late Antiquity, but not in her own time.74 The translator does not understand liberaliter instituta. The translator attaches the liberaliter to a sense of being rich and understands instituta as someone who is “established” in such wealth rather than someone who has been “raised up” or “educated.” They then bail entirely on matronaliter, since they seem not to understand the moral sense of the Roman matrona.75 There is also a manifest tendency to quieten or to make more civil the harshness of the confrontations and punishments. Of her first meeting with her father, Perpetua reports that he was so angry that he wanted to tear out her eyes, sed tantum vexavit (3.3). “But he only beat me” would be a reasonable translation. This harder physical side of the father's assault on his daughter disappears in the Greek, leaving us with: “then, just shouting at me, he left.”76 As with other peculiarities, the diluted Greek interpretation is shared by the ‘A’ recension of the Latin Acta.77 In the Greek version, the violence done to the martyrs is toned down, as for example in the translator's version of the final executions.78 In this flattening process, the Greek translator tends not only to reduce the different voices of the framer, Perpetua, and Saturus, to a more uniform and monotone speech, but also to smooth out other untidy or disjunctive elements in the original. Perpetua's dream of her fight with the Egyptian in the amphitheater is a curious mélange of a gladiatorial fight, physical combat sports, and fist-fights like the pugillationes so popular in Africa, as well as other competitive contests. It is the type of jarring mixture of images that one would expect of a real dream. The Greek translator, on the other hand, makes the contest neat, uniform, and coherent as a Greek pankration.79 Generally speaking, the translator employs a more literary form of language, eliminating or correcting perceived colloquialisms and idiomatic speech in the Latin. The Greek also simplifies its focus: the titulus understands the martyrdom to be that of “the holy Perpetua,” subordinating Felicitas to an anonymous group of “the others who met their end with her.”

Of some interest is that the number and range of misunderstandings that are characteristic of the translator of the Greek version of Perpetua's martyrdom (of which I have offered only a few examples) are more or less the same ones made by the Greek translator of Tertullian's Apologeticum.80 After a thorough investigation of the alternatives, Harnack thought that the following items could be deduced about the translator of Tertullian's treatise: that he was a Greek writing for a Greek audience, that he was someone whose knowledge of Latin was acceptable, but not exceptional—for example, they typically misunderstand standard Latin expressions and legal terms; that he had connections that linked him with Palestine and Judaea in the Severan age; and, finally, that he was somehow connected with circles around the Christian chronographer Sextus Julius Africanus. Most probably, as his cognomen indicates, Africanus was from Africa, although he spent a considerable part of his later life in Palestine in the east. In other words, he was the kind of cultural mediator who would have had both African roots and an eastern home. At the end of his analysis, Harnack teasingly noted that “what meaning these results (i.e. of his investigation) have for the question of the origins of the Greek version of the martyr acts of Perpetua and Felicitas is something that must be left for further research.”81 Although Harnack's attention was drawn to the likelihood of a near contemporary (i.e. Severan) date for the Greek translation, it must be said that its characteristic glitches are typical of Latin-Greek translation projects in general and do not necessarily fix the Perpetua translation to this earlier time. Harnack's observations set a terminus post quem, so the translation could be a near contemporary one, but it could very well come from a much later age.

There remains a fundamental point of importance that bears restating: the number of works translated from Latin into Greek was truly exiguous. Within the body of Christian texts a distinctive group of them included the Acta Scillitanorum, the Passion of Perpetua and Felicitas, and one or more treatises of Tertullian; and, from a generation or so later, possibly some letters of Cyprian.82 This is it. Later writings were of little interest. Even Augustine of Hippo, the greatest of African churchmen, acquired no Greek translator of his works during all of antiquity.83 This sharply confined group of translated texts indicates the existence of parties specifically interested in the foundational martyrs of Africa. Different mises-en-scène present themselves which might in fact be compatible, but all of them point to the significance of Italy as a staging point in the translations. In this light, it is hardly surprising that the African “Big Three” among the martyrs—Cyprian, Perpetua, and Felicitas—made their way to Italy in the early exchanges that were made with Christian churches at Rome. And they were also the specific ones who became especially important from the fourth century onward. It was these African martyrs and no others who took their place in the liturgical celebrations of the church at Rome.

The liturgical function of these martyrs conforms to the place that Perpetua and Felicitas came to have in the prayers for sinners, the nobis quoque peccatoribus, in the canon of the Catholic mass at Rome.84 Felicitas, admittedly in an oddly transmogrified form, became one of the hallmark martyrs whose cult was celebrated in the city.85 Other than biblical figures, Perpetua and Felicitas, along with their fellow African Cyprian, are the only external martyrs to whom appeal is made either in the communicantes or the nobis quoque sections of the mass. Why and when this happened is still perplexing. A time earlier in the fourth century is indicated, perhaps no later than the mid-330s. We can begin by noting that Perpetua, Felicitas, and Cyprian, are already noted in the depositio martyrum that is part of the Chronograph of 354.86 In terms of our interest in language use, we should note that these liturgical documents are late in date and that all of them always existed only in Latin. They were part of a physical and ideological translation of African martyrs and saints to Campania and central Italy that was largely separate from their movement to other regions of the empire.87 The bringing of these particular African martyrs to the imperial metropolis is important, but it was only part of broader flows of information to venues outside Africa.

These indicators lead one to suspect that there was an unusual focus of interest in Perpetua in Late Antiquity, a significant locus of translation that should draw our attention. While it is possible that African refugees from the Vandal incursions in the early fifth century could have brought some of their martyr narratives with them to the East, this does not seem to be the specific connection that we are seeking.88 The deficit of hard evidence in support of such a transfer and any sign that the arrival of the texts in the East at this time had any effect in eliciting translations, tells against this precise nexus as the cause. On the other hand, the early sixth-century armed movements into the western Mediterranean by the eastern Byzantine state that were focused on the recovery and occupation of the lands of Africa and Italy are significant. They created a divisive and contentious situation in which the sanctified power of regional martyrs became an important resource to be exploited. Connected with this overseas adventure, the court at Constantinople engaged in nothing less than a complex ideological program linked to the imposition of its authority in western lands. The program included an extensive remake of large basilica complexes and the construction of new ones both at Carthage and elsewhere in Africa, as well as at Ravenna and the nearby north Adriatic cities in Italy. At Carthage alone, there were at least six, but perhaps as many as eight basilica complexes that were rebuilt on a larger scale or that were substantially altered and refurbished under the aegis of the Byzantine power. Almost all of them were important martyrial churches that depended on imperial patronage for their construction and renovation.89 The building program included the substantial increase in size of sacred areas, as in the construction of the basilica at Bir Messaouda inside the city in the last decades of the sixth century. It was a project that demanded a huge investment in labor and resources that radically changed the previous structures on the site.90 The monumental Byzantine ecclesiastical building projects in the large urban centers of the north Adriatic, including Ravenna, are surely to be connected with the contemporary programs at Carthage.91 The two cities were the opposite poles of the late antique axis of Byzantine control in the western Mediterranean.

The Byzantine construction and renovation of basilicas in Carthage provides critical context for the interment of the superstar martyrs of African Christendom, including Perpetua and Felicitas and their companions, in the massive rebuild of the Basilica Maiorum at Carthage. By the sixth century, it was the largest and most impressive church in all of Africa. The memorial inscriptions in the basilica included not only the famous one celebrating Saturus, Saturninus, Revocatus, Secundulus, Felicitas, and Perpetua (in that order), but also a series of memorial wall plaques that commemorated the individual martyrs.92 Already in the Vandal period, Victor of Vita noted that the basilica was the burial place of the martyred saints Perpetua, Felicitas, and their companions, and that it was associated with other basilicas devoted to Celerina and the Scillitani.93 The basilicas of these particular martyrs were the focal point of seizures, first by the Vandal Arian bishops and then by their Byzantine Nicene peers.94 Moreover, the addition of Maiulus of Hadrumetum as one of the foundational martyrs in the inscription in the Basilica Maiorum probably indicates another Byzantine shaping of the same message. The importance attributed to Hadrumetum is surely to be coordinated with the massive new presence of ARS wares not just at Ravenna but at many north Adriatic sites. Most of these wares derived from production centers directly inland from Lepti Minus and Hadrumetum in the Tunisian Sahel. The new efflorescence of these wares at the head of the Adriatic indicates the importance of the Sahelian coastal sites to a peculiar western Mediterranean connectivity in the fifth and sixth centuries.95 The region and its one big martyr were now very important.

The Vandal seizure of the Basilica Maiorum in particular suggests a struggle to control the power of its martyrs, a struggle that was only to become more intense with the Byzantine military intervention in Africa in the 530s. The drive to assert eastern orthodoxy against the Arian faith of the Germanic states in Africa and northern Italy excited the exploitation of famous martyrs as part of the respective programs. This effort is perhaps most evident at Carthage, but it is manifested elsewhere in Africa, as in the reclaiming of the “Basilica of Melleus” at Ammaedara (modern Haïdra) well to the south in the hinterland.96 Not just any martyrs were reclaimed for orthodoxy at this basilica, but specifically the great father-figure Cyprian, some of whose relics were brought from Carthage and deposited near its altar.97 Although one Victorinus, bishop “of the Vandals,” had once possessed the basilica as his church, its rededication, along with the new deposition of Cyprian's relics, was achieved in the late 560s by Melleus. In reclaiming the basilica for Nicene orthodoxy, he boasted of himself as “the bishop of unity.”98 The cult of the great bishop Cyprian, especially exalted in this period, was united with that of Perpetua and Felicitas to form the great trinity of African saints, not just in Africa but also abroad.

Vitally connected to the reshaping of the Mediterranean at this time was the emergence of newly important communicative links between Africa, from the great port of Carthage, via the Adriatic, to the large imperial centers like Ravenna and Aquileia, at the head of that sea. The connections with Ravenna had been long-standing. It is surely no coincidence that a Lucius Aemilius Sullectinus, deriving from Sullecthum on the Sahelian coast of Byzacium, had been prefect of the Ravenna fleet.99 The great growth in commerce of the fifth and sixth centuries considerably enhanced the presence of Africa at the head of the Adriatic. The flows of trade and tribute incited not only an immense transfer of material artefacts to this specific region, but also of ideas, including Christian beliefs.100 It is at Aquileia that some of the most striking of new African martyr acts were discovered in the 1990s: the accounts of the martyrdoms of Gallonius at Thimida Regia and Uthina, and of Pellegrinus at Vol.101 In long-term context, this hardly seems accidental, since Fortunatianus was one of the most important bishops of Aquileia, a mid-fourth-century founding figure whom Jerome specifically identifies as an African.102 That Fortunatianus at Aquileia and Maurus at Parentium were founding or re-founding bishops in the northern Italian center was not unusual.103 On the other hand, the northern Adriatic cities were also centers for the transfer of commodities and ideas from East to West. Rufinus from Aquileia served as a late fourth-century cultural intermediator “to channel the wisdom of the Greek East into the Latin world.”104 In this labor he served as a translator of a large number of Greek texts including those of Origen. As in this past, Rufinus' monumental translation project was a one-way street: he translated works from Greek to Latin and never the other way around.

The great ecclesiastical construction projects at Ravenna and Parentium stand out as special in this respect.105 It is no mere coincidence that the brilliant mosaic decorations of six female martyrs, including the Africans Felicitas and Perpetua, and the eastern Euphemia, are found on the north side soffits of the vault of the early sixth-century Archepiscopal Chapel at Ravenna, facing six male saints on the south side. They directly illustrate the use of the African martyrs in a context of ecclesiastical conflict.106 The chapel was constructed under the rule of Theoderic and its program was grossly partisan in nature, manifestly intended as an anti-Arian statement. It was not by accident that the female saints included Euphemia of Chalcedon.107 The inclusion of the Africans was surely part of a drive to exploit the best known African martyrs by far in order to claim their authority, no doubt in part because of the enhanced African presence in the north Adriatic. Most famously, Perpetua and Felicitas appear in mosaics on the north wall of the nave of the mid-sixth-century basilica of Sant' Apollinare Nuovo at Ravenna. The mosaics, dating to the third quarter of the sixth century, feature Felicitas and Perpetua in a parade of 22 female saints and martyrs who face 26 male saints and martyrs on the opposite south wall.108 The line of female saints also includes Crispina, another African.109 The line of male martyrs and saints includes the iconic African figure of Cyprian. The selection of the Africans was deliberate: part of the orthodox remaking of the basilica. The choices are correlated with a similar program of activity in Africa that is manifested in the contemporary reshaping of the basilicas at Carthage. The two developments are surely connected and had their parallel in the production of the Greek translations of the texts that I have been discussing.

The sixth-century Euphrasian basilica at Porec, Roman Parentium, on the Istrian peninsula, probably constructed in the 540s and 550s, is another relevant case. The church boasts brilliant mosaics on the front arch of the apse that feature the two African martyrs Perpetua and Felicitas as part of two groups of six women saints and martyrs. The women parallel twelve male counterparts, the apostles, featured on the linear band mosaic above them.110 The founding figure of the Christian church at Parentium, at whose side Euphrasius stands in the mosaic in the apse itself, was a St. Maurus, whose name indicates an African origin.111 On the other hand, the construction of the church features columns and decorative pieces of Proconnesian marbles that were imported from the East. This East/West mix in physical construction is also apparent in the depiction of saints and martyrs. They include a selection mainly of central and north Italian martyrs; a few easterners, among them the women Thekla and Euphemia; and some Africans.112 All these elements strongly suggest a special Byzantine interest in exploiting the image of these particular African martyrs, especially since the Adriatic had become the main north-south seaborne corridor connecting the Exarchate in Africa with the one centered on Ravenna in northern Italy. The collation reflects a constellation of interests that in turn reflects the triangular connections between Carthage and Constantinople via Ravenna and the north Adriatic. A mirror image of this use of martyrs can be found far away in the heart of Africa in the Basilica V at Sufetula (modern Sbeïtla), where the church was reclaimed for orthodoxy by Byzantine authority. It was identified with two martyrs from northern Italy, Ambrose's ersatz saints Protasius and Gervasius, and a saint from the east, Tryphon, a singular import from Asia Minor known in Africa only from this one instance.113 Part of the program in the basilica at Parentium, in fact its central axis, were anti-Arian moves supported by the eastern court. While the inclusion of eastern saints and martyrs, especially Euphemia of Chalcedon, played a core part of this program, the choice of Perpetua and Felicitas served as the sanctified representatives of Africa, saints from the opposite pole in the Byzantine control of the western Mediterranean.

None of these African elements were as prominently selected or displayed in pre-Byzantine Ravenna or other centers in the north Adriatic. Through the fifth century, the volume of African trade with Ravenna and region in the fifth century seems to have increased markedly, especially after the move of the western imperial court to the city. Recent excavations at both Ravenna and its port at Classe have revealed the huge economic influence of the two trade circuits. The first is attested by masses of ARS lamps and transport amphorae shipped from Africa, mainly during the Vandal and Byzantine periods. The lamps derive from production centers in the interior of the Tunisian Sahel and date mainly from the mid-fifth to the mid-sixth century.114 The enormous numbers of transport amphorae reveal a double sourcing. Large numbers come from north Africa; but equally large numbers come from the eastern Mediterranean, divided between Aegean production centers and others in the Gaza.115 Other transport amphorae point to increasingly important imports from other points in the eastern Mediterranean, specifically the Aegean, sources supplying mainly wines to the West.116 If the massive increase in the African trade is especially apparent for the Vandal period, after the imperial court had moved from Milan to Ravenna, this intense phase of Africa-north Italy links gets Perpetua only as far as the north Adriatic. In all likelihood, the Byzantine military presence in the region and the foundation of the Exarchate significantly enhanced the second arm of the West-to-East transfers that connected Ravenna with Constantinople and the East. In this later period the increased importation of trade goods from the eastern Mediterranean can be traced via transport amphorae found at sites all around the northern Adriatic.117 Everything points to the great port cities in the northern Adriatic as important nodal links in the transfer of African martyrs to the Greek East.118

The monumental recollections and affirmations of Perpetua and Felicitas are other indicators of the channels through which their story was perpetuated. Firstly, all the epigraphical and monumental memorials to them are found only in one place in all of Africa: Carthage. They are attested nowhere else in Africa.119 In fact, all the martyrs with whom Perpetua is habitually associated, including Secundulus, Saturninus, and Saturus, are also attested only at Carthage120 Significantly, the same applies to the Scillitan martyrs who had their own memorial basilica, but who were also celebrated in the Basilica Maiorum.121 The impressions imparted by the literary evidence make sense in the light of this other evidence: from its inception through the fifth and sixth centuries, Perpetua's story was so imperial, urban, and metropolitan. Much the same applies to the other western Byzantine pole of her story: the brilliant mosaics at Ravenna and Parentium reflect the same focus in urban centers in the system of imperial communications. Iconic memorializations of Perpetua and Felicitas were found nowhere else in Italy in Late Antiquity. In terms of literary memory, they were also strongly linked to liturgical functions in the churches and basilicas of the region.122 The distribution of the literary and material data suggests two possible scenarios. One is that there was a first stage in which Greek translations were done soon after the production of the Latin texts of the martyrdoms of Perpetua and of the Scillitani. These texts were then taken up for different purposes in Late Antiquity and were slightly re-edited at that time. The second and more plausible scenario is that the Greek translations were done in the early sixth century in a context of a struggle for authority in the Latin West that hailed them forth. If this is so, it suggests that there existed at the later date different Latin versions of both the Scillitani and of Perpetua, some of which were fuller in detail than the ones that we now possess.

A particular triad of martyrs and saints found in the ecclesiastical displays in north Italian centers—north Italian, African, and eastern Mediterranean—indicate the conscious triangulation of elements that were specific to the Byzantine period. The mobilization of the superstars of African martyrdom—Perpetua, Felicitas, and Cyprian—was part of a deliberate program that was encouraged by an anti-Arian eastern state and its church. A basic fact to which I return in explaining “why Perpetua” was part of this mix, and the importance of the translation of her story into Greek, is that she had long been a very popular and powerful figure among the martyrs of Africa. The final phase in the construction of the memorial to her and her fellow martyrs in the Basilica Maiorum at Carthage is of Byzantine date. The late memorials only served to affirm the existing importance and popularity of a specific group of martyrs. These particular martyrs, the proto-martyrs of the faith, made this basilica the greatest in all of Africa because it was the basilica that housed the maiores, the foundational martyrs of the church in Africa. The great ancestral martyrs, the maiores—the Scillitani, Perpetua and her companions—were precisely the ones that were celebrated in the basilica. It seems probable that it was in this sixth-century context that her account was transferred to the East. It is notable that almost the only African saints to have made it to the eastern Mediterranean to be celebrated by the great church in Constantinople included Speratus, one of the Scillitan martyrs, and Perpetua.123 The cult of Felix of Thibiuca also made its way to the East. He is, as far as can be discerned, he is the solitary addition to the core “big three” of African martyrs to make this transition.124 Like the others, it was by means of the Greek translation of a late Latin version of his passion, probably in the sixth century, that he entered into the Synaxarion of the saints and martyrs at Byzantium.

As for the specific collection of martyrs and saints in the churches at the head of the Adriatic of which Felicitas and Perpetua were part, a good argument has been made that the choice of the ones to be represented on the brilliant mosaics was closely linked to liturgical purposes.125 But the main reason why Perpetua and Felicitas were part of that liturgy in the first place was because Perpetua had long been among the most highly honored martyrs in Africa whom travelers, both African and others, brought with them to their main ports of call, as at Ravenna, Aquileia, and Parentium.126 Perpetua's established importance helps to explain why the Byzantine state, and the parties linked to it, the bishops and churchmen who were connected to its interests, featured her not just in their liturgical services but also in the costly decorative arts of their impressive basilicas. The connection at the head of the Adriatic cannot be accidental, since it is demonstrably linked with the Byzantine state's interests and expenditure of resources that are documented at Carthage. That was why Perpetua's basilica, the Basilica Maiorum at Carthage, and the ones of the other martyrs were so worth the wealth and resources invested in them by the eastern state and church. That nexus of interests shows why she was transported to the ports of the Italian exarchate at the head of the Adriatic and from there further afield to the eastern Mediterranean. The Greek translation of her martyrdom was part of this later constellation of power, a peculiar significance that helps to explain why we find it in a synaxarion of Byzantine martyrs and saints and in a manuscript in the Library of the Orthodox Patriarchate in Jerusalem.


Earlier versions of this paper were presented to a round-table in honor of Barbara Gold held at Hamilton College on 19 September 2018; and to the Philadelphia Seminar on Christian Origins held at the University of Pennsylvania on 6 December 2018. I would like to thank all who asked questions and made useful comments on those occasions. I express my gratitude to David Jenkins, Classics librarian at Princeton, who, amidst the lockdowns of the coronavirus pandemic, rescued me more than once. I also thank the two Érics, Fournier and Rebillard, for casting their critical eyes over early drafts of this article. A similar note of thanks is also owed to the Journal's external reviewers for their critical comments. I hope that John Eldevik, who did not receive a satisfactory reply to his good question at Hamilton, might find some of it answered here.


APERP1: Acta Perpetuae et Felicitatis (Bibliotheca Hagiographica Latina [henceforth BHL] 6634–35) (also referred to as the ‘A’ version).

APERP2: Acta Perpetuae et Felicitatis (BHL 6636) (also referred to as the ‘B’ version).

ASCILLG: Acta Scillitanorum, textus Graecus (Bibliotheca Hagiographica Graecum [henceforth BHG] 1645)

Baratteet al. Basiliques chretiennes, 2: F. Baratte, F. Bejaoui, N. Duval, S. Berraho, I. Gui, and H. Jacquest, eds., Basiliques chrétiennes d'Afrique du Nord (inventaire et typologie), II: Inventaire des monuments de la Tunisie (Bordeaux: Ausonius Mémoires no. 38, 2014)

ICKarth. 2: L. Ennabli, Les inscriptions funéraires chrétiennes de Carthage, 2: La basilique de Mcidfa (Rome: Ecole française de Rome, 1982).

PPL: Passio Perpetuae et Felicitatis, textus Latinus (BHL 6633).

PPG: Passio Perpetuae et Felicitatis, textus Graecus (BHG 1482).

In referring to the Perpetua texts, I shall refer to the following standard works by their author(s) only:

Amat: J. Amat, Passion de Perpétue et de Félicité, suivi des Actes: introduction, texte critique, traduction, commentaire et index (Paris: Editions du Cerf, 1996): Sources chrétiennes (henceforth SC) no. 417.

Franchi de' Cavalieri: Pio Franchi de' Cavalieri, La Passio SS. Perpetuae et Felicitatis = Römische Quartalschrift für christliche Alterthumskunde und für Kirchengeschichte, 5, Supplementheft, Rome, 1896 = Scritti agiografici, 1, Studi e Testi 221 (1962): 41–155.

Harris and Gifford: J. Rendel Harris and Seth K. Gifford, The Acts of the Martyrdom of Perpetua and Felicitas: The Original Greek Text, now first edited from a MS. in the Library of the Convent of the Holy Sepulchre at Jerusalem (London: C. J. Clay & Sons, distributed by Cambridge University Press, 1890) = “A New Version of the Acts of Perpetua and Felicitas,” Haverford College Studies 3 (s.d.): 1–73.

Rebillard: E. Rebillard, Greek and Latin Narratives About the Ancient Martyrs (Oxford: Oxford University Press, 2017).

Robinson: J. Armitage Robinson, The Passion of S. Perpetua = Texts and Studies: Contributions to Biblical and Patristic Literature, no. 2 (Cambridge: Cambridge University Press, 1891).

Van Beek: C. I. M. I. Van Beek, Passio Sanctarum Perpetuae et Felicitatis, I: Textum Graecum et Latinum ad fidem codicum MSS (Nijmegen: Dekker & Van de Vegt, 1936 [editio minor with commentary: Bonn, Hanstein, 1938]).


In addition to the text editions noted above, see T. J. Heffernan, The Passion of Perpetua and Felicity (Oxford: Oxford University Press, 2012), 104–24, who provides his edition of the long Latin text equipped with a full apparatus criticus.


A. Falcetta, “James Rendel Harris: A Life on the Quest,” Quaker Studies 8 (2004): 208–25; and, now, in greater detail: A Biography of James Rendel Harris, 1852–1941: The Daily Discoveries of a Bible Scholar and a Manuscript Hunter (London-New York: T. & T. Clark, 2018), esp. 86–88, 99–100, 265 (for the dramatic impact on his own beliefs).


In addition to the information presented in Harris and Gifford, see the earlier report on the library by J. Rendel Harris, “The Library of the Convent of the Holy Sepulchre at Jerusalem,” Haverford College Studies 1 (1889): 1–18.


The main lineaments of these debates are ably reviewed by Clementina Mazzucco, “Il dibattito sulla lingua originaria della Passio,” § 1 in “Il rapporto tra la versione greca e la versione latina della Passio Perpetuae,” in Bilinguismo e scritture agiografiche: Raccolta di studi, ed. V. Milazzo and F. Scorza Barvellona (Rome: Viella, 2018), 17–75, at 17–28.


Among the numerous works on debates on priority of the Greek and Latin versions, only the following will be particularly relevant: L. Duchesne, “En quelle langue ont été écrits les actes des saintes Perpétue et Félicité?” Comptes rendus des séances de l'Académie des inscriptions et Belles-Lettres (1891): 39–54; A. Harnack, Geschichte der altchristlichen Literatur bis Eusebius (Leipzig: J. C. Hinrichs, 1893), 1: 674–819; L. Massebieau, “La langue originale des actes des saintes Perpétue et Félicité,” Revue de l'histoire des religions 24 (1891): 97–101; and V. Reichmann, “Die Passio Perpetuae,” ch. 6 in Römische Literatur in griechischer Übersetzung (Leipzig: Dieterich'sche Verlagsbuchhandlung = Philologus, Supplementband 34, Heft 3, 1943), 101–30, at 103–05. For a full census of almost all published views, see the collation by Heffernan, Passion of Perpetua, 80–81n4.


L. Robert, “Une vision de Perpétue, martyre à Carthage en 203,” Comptes rendus des séances de l'Académie des inscriptions et Belles-Lettres (1982): 228–76 = ch. 130 in Opera minora selecta: épigraphie et antiquités grecques (Amsterdam: Hakkert 1989), 5: 791–839; supported by other eminent scholars, e.g. G. Bowersock, Martyrdom and Rome (Cambridge: Cambridge University Press, 2002), 34. An earlier suggestion made on linguistic grounds by Aarne Salonius, Passio S. Perpetuae: Kritische Bemerkungen mit besonderer Berücksichtigung der griechisch-lateinischen Überlieferung des Textes (Helsingfors: Översikt Finska Vetenskaps-Societetens Förhandlingar, no. 63, 1921), does not convince.


See the full references in Franchi de' Cavalieri, 12n2.


A. Hilgenfeld, Berliner philologische Wochenschrift 10 (1890): 1488–91; see Duchesne, “En quelle langue ont été ecrits,” 53–54.


In a footnote to his edition, Robinson states that he had communicated with Rendel Harris and reports that “he [sc. Harris] has seen reason to change his view on the matter and to regard the Latin as the original” (Robinson, 3n1). Since Robinson published his study in 1891, Harris must have changed his mind rather quickly, indeed within a year of his original publication in 1890. Robinson must have known this because he was in constant contact with Rendel Harris, serving as the Cambridge intermediary in the urgent process of publication: see Falcetta, Biography, 99.


The conclusion was already reached by Franchi de' Cavalieri (99) after an exhaustive analysis of the language of the two texts, published in 1896, not long after the original publication of the Greek text.


Already seen by Franchi de' Cavalieri, 43–44.


PPG 2.1 (Amat, 104). The words are not in the long Latin narrative. It should be indicated at this point that the actual MS reading is Θουκριτάνων which Rendel Harris prints but which all subsequent editors have “corrected” to Θουρβιτάνων (see n19 below). Heffernan (see n14 below) notes that four of the late MSS of the long narrative do contain a reference in their incipits to a “Thuburbo” as the place of arrest of the martyrs, but he makes (to my eyes) a convincing argument (131–32, 134–35) that these late MSS were copying materials from the short Latin Acta.


An objection that not even this much is certain was forcefully made by Robinson, “Were our Martyrs Thuburbitan?” 22–26. He made the argument that the martyrs were from Carthage and that this explains the absence of a named place of their origin in the long narrative where, he argued, the readership and those who were knowledgeable just assumed that Carthage was their patria. The argument deserves to be taken seriously. But I think that the weight of the argument stands in favor of Thuburbo Minus as the patria of the martyrs. Robinson says the fact that Augustine does not mention their place of origin in his sermons is a presumption in favor of Carthage as their home. But Augustine never mentions a place of origin in his sermons on other African martyrs like Castus and Aemilus, Crispina, Marianus and Jacobus, Maxima, Donatilla and Secunda, the Maxulitani, Quadratus and the Massa Candida, the Scillitani, or the Volitani, all of whom are known to come from places outside Carthage: see G. Lapointe, La célébration des martyrs en Afrique d'après les sermons de saint Augustin (Montreal: Communauté chrétienne, 1972), 27–56. Then again, the fact that later martyr calendars, being external sources, specify their martyrdom as in Africa or apud Africam does not weigh on either side of the argument. Finally, Robinson claims that there was likely a “confusion” between the martyrdoms of Perpetua and Felicitas and the later account of the martyrdoms of Maxima and Donatilla, and Secunda, where they are said to have died at a Thuburbo. For the text, see J.-L. Maier, Le dossier de Donatisme, 1, Des origines à la mort de Constance II, 303–361 (Berlin: Akademie Verlag, 1987), no. 5: 92–105, at 95. The notices appear only in the very late titulus of Ado's version which reads: “apud Africam civitate Tuburbo Lucernariae natale sanctarum virginum Maximae Donatillae et Secundae quae persecutione Gallieni sub Anolino iudice passae sunt.” The text itself says even less: the proconsul Anullinus “iussit eas ad civitatem Thuburbitanam perduci” (Passio 3 = Maier, Le dossier de Donatisme, 99). Robinson argued that this Thuburbo has somehow crept into the Greek version. The connection is remotely possible, but it seems dubious since the mere mention of a Tuburbo, much less of an otherwise unattested Tuburbo Lucernaria, would be difficult to transform into the specific Thuburbo Minus. It seems more probable that the mostly fictitious fifth-century confection of the martyrdoms of Maximilla and Donatilla (and Secunda) was heavily influenced by the existing Perpetua story rather than the other way around. Secondly, the appearance of a utilitarian epithet like “Lucernia” for a city seems to be a late antique affectation, as when Boseth is called Boseth “Amphoraria” (Passio Mammarii, 4.5: “in civitate Boseth Anforaria”). Presumably such towns were production centers for late antique ceramics. S. A. Morcelli, Africa cristiana, 3 vols. (Brescia: Bettoniana, 1816–17), 1: s.v. “Tuburbitanorum maiorum,” 332, already guessed as much: civitas Tuburbo Lucernaria “de qua appellatione veteres nihil tradidere… nomen illi haesisse ab officina fictilium lucernarum, quas inde Afri peterent, non inepte credas. Sic enim et Boseth Amphoraria dicta videtur.”


T. J. Heffernan, “The Legacy of Misidentification: Why the Martyrs in the Passio Sanctarum Perpetuae et Felicitatis were not from Thurburbo Minus,” Journal of Early Christian History 6 (2016): 126–51. It is not possible to rehearse the whole of his complex argument here. I note only that Heffernan himself seems to allow that that the martyrs might have been arrested at Thuburbo Minus (e.g. 131–32) and were later tried and executed at Carthage. This is the sequence of the events that I accept. I have performed a systematic census of all of the Vibii known for north Africa as noted in the Corpus inscriptionum Latinarum, Année épigraphique, and Christian epigraphical sources (N = 202). The results suggest two items. The African Vibii were derived from central Italian stock, persons who were first settled primarily on lands in the middle and upper Bagrada-W. Miliana valleys (N = 43). Many of them were (and continued to be) army men, ultimately derived from soldiers of the Legio Octava (specifically noted in Année épigraphique 1919: 29, Thuburnica) settled by Julius Caesar at Uthina and Thuburbo Minus. Indeed, Thuburbo Minus was formally known as the splendidissima colonia VIII (Octavanorum) Thuburbitana (Année épigraphique 1915: 37, Thuburbo Minus). Most of their descendants are found at army bases like Lambaesis and closely allied towns (N = 70). These data suggest that Perpetua's family of the Vibii had a good probability of coming from Thuburbo Minus and not from Carthage, where very few instances of the family name are attested (exactly 5 out of 202 instances, all on items of small commerce).


On the contrast between the two Thuburbo's, see C. Lepelley, Les cités de l'Afrique romaine au Bas-Empire (Paris: Etudes Augustiniennes, 1981), 2: 199–206. Thuburbo Maius (Hr. Qasbat), a larger important regional economic center in the Pont du Fahs plain south of Carthage, rose to the level of a municipium under Hadrian and a colonia under Commodus. Although Thuburbo Minus (Tébourba) was the site of a veteran settlement made by Julius Caesar, its importance faded rather than grew. It remained a small town about which rather little is known. According to the Antonine Itinerary, it was located 46 (Roman) miles by road to the northwest of Carthage: Itin. Anton. 44.1–3 in Itineraria romana, ed. O. Cuntz (Leipzig: Teubner, 1929), 1: 6.


APERP1: 1.1 and 9.5 (Rebillard, 332 and 338): apud Africam in civitate Tuburbitanorum.


APERP2: 1.1 and 9.5 (Rebillard, 342 and 348): in civitate Tuburbitana capti.


Codex Compendiensis: Bibl. Nat. lat. 17626 = Franchi de' Cavalieri, 108, apparatus; the reading is the same as that found in APERP1.


Heffernan, Passion of Perpetua, apparatus criticus ad 104, cf. 370–71: in civitate tuburbitana minore (E); and in civitate Tyburtinarum minore (G); E is a twelfth-century MS from Einsiedeln and is related to G the earliest surviving MS, from St. Gall, dated to the late ninth-early tenth century.


The elision and the lapse are often “corrected” from the Jerusalem MS, but the reading in the original text is noted in Harris and Gifford, 41; Franchi de' Cavalieri, 111; and Robinson, 63. The apparatus criticus shows that the MS actually reads Θουκριτάνων (i.e. Θουρβιτάνων). The “error” is important to a subsidiary argument that I wish to make.


PPL 6.3 (Rebillard, 310; Amat, 122): Et Hilarianus procurator, qui tunc loco proconsulis Minuci Timiniani defuncti ius gladii acceperat.


For Minicius Opimianus, whose full name was Titus Salvius Rufinus Minicius Opimianus, see Prosopographia imperii Romani2 M 622 and B. E. Thomasson, Fasti Africani: Senatorische und ritterliche Amtsträger in den römischen Provinzen Nordafrikas von Augustus bis Diokletian (Stockholm: Acta Instituti Romani Regni Sueciae, no. 53, 1996), no. 104, 79–80. This last is probably in need of revision, however: see T. D. Barnes, Tertullian, rev. ed. (Oxford: Oxford University Press, 1985), 334; cf. append. 18, “The Proconsul Rufinus,” 266–67. His homonymous grandfather was governor of Africa in 138–39. His name, a notice in Prosper Tiro, and a source on the martyrdom of Guddene at Carthage on 27 July 203, suggests the possibility that he was governor of Africa in 203–04 and that Perpetua's martyrdom was in March 204, not 203: see Barnes “Proconsuls of Africa in 202–204” in Early Christian Hagiography and Roman History (Tübingen: Mohr Siebeck, 2010), 304–07.


PPG 6.3 (Amat, 122); note that the Greek has a similar elision in the proper noun of Perpetua's home town: Θουρβιτάνων (but see n12 above for the restoration of the probable reading) for Θουβουρβιτάνων.


These points, and others, are noted by Harris and Gifford, 20.


See the important observations made by François Dolbeau concerning Amat's edition: Revue des études Augustiniennes 43 (1997): 350–53; for an overview see Mazzucco, “Il ruolo della versione greca nella ricostituzione del testo della Passio,” § 2 in “La versione greca e la versione latina,” 29–40, who confirms that the Greek translation does not depend on the Latin Passio that we possess.


In addition to Harris, I have consulted the modern editions in Franchi de' Cavalieri, Robinson, and van Beek. The Greek text was reprinted by Robinson, 60–95; more recently, it has been included as a parallel text in Amat.


Investigated in detail by Ake Fridh, Le problème de la passion des saintes Perpétue et Félicité (Göteborg: Acta Universitatis Gothoburgensis: Studia Graeca et Latina Gothoburgensia, no. 26, 1968) who was advancing, albeit along a different front, arguments first made four decades earlier by Walter Shewring.


For example, he translates the long complicated introduction written by the person whom I have called the “framer” of the whole account. There is no reason why the translator would have done this if the preface was not already an integral part of the Passio account.


See B. Rochette, “La traduction du latin en grecque à Byzance: un aperçu général,” in Latin in Byzantium, I: Late Antiquity and Beyond, ed. A. Garcea, M. Rosellini and L. Silvano (Turnhout: Brepols, 2019), 297–312, at 297–300. A similar method was followed in the Greek translations of classics like Vergil and Cicero found in the papyri; and also, as to be expected, in the translation of legal texts.


These have been so well analyzed by Mazzucco, “Il rapporto tra la versione greca e la versione Latina,” esp. § 3, “Particolarità e tendenze della versione greca,” 40–73, that there is no need to rehearse them in detail here.


On a larger scale, this inquiry intersects with the interests of J. P. Conant on the cultural dimensions of these transfers: “Europe and the African Cult of Saints, circa 350–900: An Essay in Mediterranean Communications,” Speculum 85 (2010): 1–46.


Most of the literary and epigraphical evidence available to the 1960s was covered by T. Kotula, “Utraque lingua eruditi. Une page relative à l'histoire de l'éducation dans l'Afrique romaine,” in Hommages à Marcel Renard, ed. J. Bibauw (Brussels: Collection Latomus 102, 1969), 2: 386–92. A thorough new study is needed.


See W. Ameling, “Femina Liberaliter Instituta—Some Thoughts on a Martyr's Liberal Education,” ch. 3 in Perpetua's Passions: Multidisciplinary Approaches to the Passio Perpetuae et Felicitatis, ed. J. N. Bremmer and M. Formisano (Oxford: Oxford University Press, 2011), 78–102, who has additional relevant remarks.


PPL, 13.4 (Rebillard, 318; Amat 150): Et coepit Perpetua graece cum illis loqui; PPG 13.4 (Amat, 150): Καὶ ἤρξατο ἡ Περπετούα ῾Ελληνιστὶ μετ᾿ αὐτῶν ὁμιλεῖν. Robert, “Une vision de Perpétue,” 254 and passim, seemed to accept on this basis that Greek was her primary spoken language, whereas I take the special notice of it by Saturus in his dream to suggest that speaking Greek was something that was sufficiently unusual as to be flagged by him. That is to say, she could speak Greek if called upon to do so (and in the ecclesiastical context of speaking with clerics she might be thought to be able to do in a dream). As much does not provide grounds for arguing that she wrote her prison “diary” in Greek, a piece of speculation shown to be very unlikely: A. A. R. Bastiaensen, “Heeft Perpetua haar Dagboek in het Latijn of in het Grieks Geschreven?” in De heiligenverering in de eerst eeuwen van het christendom, ed. A. Hilhorst (Nijmegen, Dekker & van de Vegt, 1988), 130–35.


Neither the forms Dinocrates nor Deinocrates are found in the cognomina listed in the indices of Corpus inscriptionum Latinarum 8; on the other hand, Saturus and variants are rather frequent: Corpus inscriptionum Latinarum 8, indices, 113.


Franchi de' Cavalieri, 97; for the editor's statement, see PPL 11.1: Sed et Saturus benedictus hanc visionem suam edidit, quam ipse conscripsit (Rebillard, 316; Amat, 142).


Fridh, Le problème de la passion, 30–45 (reiterated at 55).


See Reichmann “Die Passio Perpetuae,” 116; as Franchi de' Cavalieri (27) showed, these were technical Christian Graecisms that had already passed into Latin—much like the greeting used by the angels in Saturus' dream (12.2): Agios, agios, agios! See also, Fridh, Le problème de la passion, 81, referring to the fundamental work by Christine Mohrmann.


The primary source is Eusebius, HE, 5.23–27. On his African background, see L. Duchesne, ed., Liber Pontificalis: texte, introduction et commentaire (Paris: de Boccard, 1886; 2nd ed. 1955), 1: 137–38: he was son of a Felix; both the father and the son have cognomina commonly found in Africa.


G. A. Bisbee, Pre-Decian Acts of Martyrs and Commentarii (Philadelphia: Fortress Press, 1967), 118: they are ultimately derived from the commentarii of Justin's trial; see D. Minns and P. Parvis, ed. and comm., Justin, Philosopher and Martyr: Apologies (Oxford: Oxford University Press, 2009), 32–33; for the Greek A-C versions, see H. Musurillo, The Acts of the Christian Martyrs (Oxford: Clarendon Press, 1972), no. 4, 42–61.


C. Schick, “Per la questione del latino Africano. Il linguaggio dei più antichi atti dei martiri e di altri documenti volgarizzanti,” Rendiconti dell'Istituto Lombardo, classe di lettere, scienze morali e storiche 96 (1962): 191–234. The account of the martyrs of Lugdunum (Lyon) of 177 is a difficult case to unpack for our purposes. A letter from Christians at Vienne and Lugdunum to churches in Asia Minor and Phrygia, as reported in Eusebius, HE, 5.1, would most probably be written in Greek, so I think that we are not dealing with a translation in this case.


See D. B. Levenson and T. R. Martin, “The Ancient Latin Translations of Josephus,” ch. 21 in A Companion to Josephus, ed. H. H. Chapman and Z. Rodgers (Malden MA: Wiley, 2016), 322–44, at 323–24.


See the landmark study which emphasizes many of the unusual and sometimes unique aspects of the movement of translations from Greek to Latin: D. Feeney, Beyond Greek: The Beginnings of Latin Literature (Cambridge, MA: Harvard University Press, 2016), with implications for the reverse of this process.


B. Rochette, “Auteurs latins dans la littérature grecque,” ch. 4.2 in Le latin dans le monde grec: Recherches sur la diffusion de la langue et des lettres latines dans les provinces hellénophones de l'Empire romain (Brussels: Collection Latomus no. 233, 1997), 269–90, presents a truly pathetic scene of the number Greek translations of Latin literary texts that were done in antiquity.


The important studies are by X. Lequeux, “Latin Hagiographical Literature Translated into Greek,” ch. 14 in The Ashgate Research Companion to Byzantine Hagiography, ed. S. Efthymiadis (London: Routledge, 2011), 385–99, and S. Efthymiadis, “L'hagiographie grecque de l'Italie (VIIe – XIVe siècle),” in Corpus Christianorum: Hagiographie, 7, ed. M. Goullet (Turnhout: Brepols, 2017), 345–421, both of which were used in what follows.


As good a measure as any of this movement is the dated but still useful work by A. Siegmund, Die Überlieferung der griechischen Christlichen Literatur in der lateinischen Kirche bis zum Zwölften Jahrhundert (Munich-Pasing: Filser-Verlag, 1949); a recent up-to-date analysis focused on Late Antiquity reveals the same very exiguous number of Latin texts in Greek translation: see Rochette, “La traduction du latin en grecque,” 297–99.


H. Delehaye, “Cyprien d'Antioche et Cyprien de Carthage,” Analecta Bollandiana 39 (1921): 314–32, with comments by Conant, “Europe and the African Cult of Saints,” 25–26.


On what follows, see Reichmann, “Die Passio Perpetuae”—in addition to the martyr accounts of the Scillitani and Perpetua, Reichmann added the acts of Karpos and Euplos. In making this claim he was surely mistaken since no critical student of those texts today holds anything other than that the Latin version is a translation from the Greek. For the texts, see Rebillard, 38–45 (Greek text) and Musurillo, “The Martyrdom of Saints Carpus, Papylus and Agathonikê,” no. 1 in Acts of the Christian Martyrs, 28–37 (Latin text).


For specific consideration, with detailed bibliography, see C. G. Conticello, “Le projet d'un répertoire des traductions de textes chrétiens du latin au grec (IIIe-XVe S.). Quelques exemples,” in Eukarpa/Εὔκαρπα: Etudes sur la Bible et ses exégètes en hommage à Gilles Dorival, ed. M. Loubet and D. Pralon (Paris: Editions du Cerf, 2011), 225–33, specifically on the Acta Scillitanorum (227–33); and E. Dekkers, “Les traductions grecques des écrits patristiques latins,” Sacris Erudiri 5 (1953): 193–233, whose remarks are apposite.


A. Harnack, “Die griechische Übersetzung des Apologeticus Tertullians,” in Texte und Untersuchungen zur Geschichte der altchristlichen Literatur 8.4 (Leipzig: Teubner, 1892): 1–36; Reichmann, Römische Literatur, 14 and 101: a text mentioned (and quoted) several times by Eusebius: HE, 2.2.4–6 [= Apol. 5; 2.25.4 [= Apol. 5]; 3.33.3–4 [= Apol. 2]; and 5.5.5–7 [= Apol. 5].


Tertullian, de Corona, 6.3 (Corpus Christianorum, series Latina 2: 1047); see Reichmann, Römische Literatur, 14; other works of his, apparently, were in Greek and it is not an unreasonable assumption that he perhaps translated them as well: see Barnes, Tertullian, 68–69, 253, 277. In addition to the de Corona, the de Baptismo (15.2) refers to a Greek version. In his de virginibus velandis (1.1), Tertullian seems to refer to a Greek version obliquely by saying that he will make his argument “also in Latin,” Latine quoque.


Barnes, Tertullian, 253, in his discussion of the de Ecstasi, notes that there is not a whiff of evidence for an awareness of these Greek translations in eastern lands, so confirming the pattern we have already noted above for Greek translations of western secular texts; but in this case, it might well have been no more than a Latin text bearing a Greek title.


For the latter, there is a desperate need to provide an up-to-date version of W. Theiling, Der Hellenismus in Kleinafrika: der griechische Kultureinfluss in den römischen Provinzen Nordwestafrikas (Leipzig: Teubner, 1911 [reprint: Hildesheim: G. Olms, 1984]).


See the discussion by E. Rebillard, “The Beginnings of Christian Latin Literature,” in The Cambridge History of Later Latin Literature, ed. G. Kelly and A. Pelttari (Cambridge: Cambridge University Press, forthcoming) explicating the position accepted here. Tertullian began his literary career by writing his Christian works in Greek (including the ones noted above)—only later did he turn to write new works in Latin or revised versions of his Greek works in Latin. Most likely, these latter were not translations, strictly speaking, of already-existing works of his.


Discovered by Usener in 1891 (BHG 1645 = Bibl. Nat. graec. 1470, a ninth-century MS). See F. Ruggiero, “Atti dei martiri Scilitani,” Atti della Accademia nazioniale dei Lincei 388; Classe di scienze morali, storiche e filologiche; serie 8, vol 1, fasc. 2 (1991): 39–139, at 77–79 (Greek text); Robinson (113–17) also prints the Greek text. For context, see Rebillard, 351–59. Since Robinson and Ruggiero follow the manuscript readings more closely, their texts are to be preferred over the innovative speculations in Corsaro and Usener.


The Greek translation of the Scillitan martyrs was copied in the early ninth century in Rome as part of a series of hagiographical texts compiled by Methodios, later patriarch of Constantinople; and it was later recopied by Anastasios in 890. One suspects that this could indicate a possible track for our translation: see Lequeux, “Latin Hagiographical Literature,” 391.


The Scillitan martyr Speratus, along with Perpetua, did become known in the East. Conant (“Europe and the African Cult of Saints,” 12) thinks that his reputation traveled eastward from Sicily. The Scillitani are known to have had a cult at Naples, so a south Italian path of transmission to the East for them is also possible.


For a discussion of the Greek and Latin texts, see the edition of the Latin life by P. Laurence, ed., “Vita Sanctae Melaniae Senatricis Romae,” ch. 5 in La vie latine de sainte Mélanie: edition critique, traduction et commentaire (Jerusalem: Franciscan Printing Press, 2002 [Collectio minor: Studium Biblicum Franciscanum, no. 41]), 152–299; for the Greek, D. Gorce, ed., Vie de sainte Mélanie. Texte grec (Paris: Editions du Cerf, 1962) (SC 90), at 7–18 on the testimonia, 45–53 on the manuscripts. On the significance of the context, see P. Brown, Through the Eye of a Needle: Wealth, the Fall of Rome, and the Making of Christianity in the West, 350–550 AD (Princeton: Princeton University Press, 2011), 292. With Cardinal Rampolla's discovery of the first full Latin text in the Escorial, similar debates broke out over which version, the Greek or the Latin, was the original. As with the Greek text of Perpetua's martyrdom, the discoverer naturally thought that his was original, although in the reverse since Rampolla thought that the Latin, despite all of its manifest difficulties, was the original. The current consensus is that the Greek is the earlier and the Latin a later version, with both derived from a lost Greek original. In what follows, parallels with the Perpetua texts will become evident.


Ruggiero, Atti dei martiri Scilitani, 77; cf. Robinson, 112–13. For Sextus Quintilius Condianus, see Prosopographia imperii Romani2 Q 22: consul ordinarius with Caius Bruttius Praesens in 180 C.E.


PPG 21.2 (Amat, 176): τοῦ αἵματος τοῦ ἁγίου is simply perfusus est sanguine in the Latin; on the bodies, see PPG 21.7 (Amat, 178): ὅπως διὰ τῶν ἁγίων σωμάτων as opposed to simply in eorum corpore in the Latin. Similarly, the martyrs themselves and items connected with them are repeated additionally glossed as μακάριος/α—that is, as especially “blessed” (e.g. PPG 20.1, 21.7).


See Ruggiero, Atti dei martiri Scilitani, G, 59; 77, lines 1–2; 79, line 65 (late dating format only, omitting the traditional Roman form); cf. Robinson, 113.


See A. Ferrua, “Il giorno del mese,” Rivista di archeologia cristiana 61 (1985): 61–75.


See Conant, “Europe and the African Cult of Saints,” 28–30.


On the manner in which the acta have been edited, cutting out all the initial sections of the interrogation of the proconsular governor, see S. Fialon, Mens immobilis: Recherches sur le corpus latin des actes et des passions d'Afrique romaine (IIe-VIe siècles) (Paris: Institut d'Etudes Augustiniennes, 2018), 42; for the mention of the place: Ruggiero, Atti Scilitani, G, 78, line 66: 'Ισχλὴ τῆς Νουμηδίας.


In the MS most used for constructing modern editions of her martyrdom, the manuscript from Monte Cassino, her account is bound with a collection of Cyprian's letters, a collocation that is unlikely to be mere happenstance: Heffernan, Passion of Perpetua, 371, 375.


Amat, 278–79 and 292–93 = Rebillard, 332 and 342: APERP1, 1.1: Facta itaque persecutione sub Valeriano et Gallieno…; APERP2, 1.1: Valeriano et Gallieno consulibus, missa in Christianos persecutione violenta…


It is not, as is sometimes suggested, independently confirmed in the late inscription from the Mcidfa Basilica (most likely the Basilica Maiorum) at Carthage: Musée de Carthage Inv. 1715 = Corpus inscriptionum Latinarum 8.25038 = ICKarth 2.1 (Carthage). The reading of the date in this inscription—non(as) Mart(ias)—is restored. The small part of the vertical hasta could be a number of different things, so the “N” is restored on the basis of other evidence. See Y. Duval, Locus sanctorum Africae: le culte des martyrs en Afrique du IVe au VIIe siècle, 2 vols. (Rome: École française de Rome, 1982), 1: 7–17, esp. the contrast between fig. 11a (Delattre's restoration) and fig. 11b (the actual state of the fragments); and 2: 682–83. Amat, 25–27, for example, does not indicate that the restored words are not in the inscription as it survives, so it is important to note how heavily restored the inscription is. Duval and Ennabli have argued, based on its orthography, that the inscription is of either Vandal or Byzantine date. It is almost certainly the latter.


The argument has been made by J. Bremmer, “Felicitas: The Martyrdom of a Young African Woman,” ch. 1 in Bremmer and Formisano, Perpetua's Passions, 35–53, at 38n24, based on notes made by the editors of the editio princeps. It is based on two two-word selections possibly quoted by Augustine in Sermo 282.auct. = Erfurt 1 = I. Schiller, D. Weber, and C. Weidmann, Wiener Studien, 121 (2008): 260–64, at 264. The words in APERP2, 9.2 are repeated in Sermo, 282 auct. = Erfurt 1, 6.3: “in utere oneri” (uteris onus) and “virili virtuti” (virilis virtus). The argument is a bit tenuous, so some caution must be shown; as Fialon (Mens immobilis, 43n14) remarks, the arguments “me semblent peu probantes.” At least the following should be noted. First, the verbal echoes are weak, being not much more than a few words that are generally found elsewhere. Ambrose could use them with no sign that he was quoting anyone (De virginibus, 1.6). If not the precise words, very close equivalents are commonly found in secular literature, e.g. Columella, De Re Rustica, 6.24.4: “uteri gravetur onere;” Ovid, Heroides, 11.64: uteri crimen onusque mei; and Met. 10.481: “uteri portabat onus.” As Bremmer himself notes, virtute virili is used by Paulinus of Nola, Carmen, 26.159, and rather frequently thereafter by authors in a way that suggests that it was a commonly used phrase for this circumstance. Secondly, in every other instance it can be shown that Augustine is referring to the long Latin narrative, so why he should suddenly depart from his usual practice in this one case is unclear. The mentions of Perpetua in Augustine are in Sermo 159A (Mainz 42, Dolbeau 13); 280 (Migne, Patrologia Latina 38: 1280–85), 281 (Patrologia Latina 38: 1284–86); 282 (Patrologia Latina 38: 1285–86; a fuller text of which has now been found, see Erfurt 1 noted above); 335A (Patrologia Latina 46: 988–91 = Miscellanea Agostiniana 1: 219–21, one of three), and 394 (Patrologia Latina 39: 1715–16).


In addition to Augustine (see the references just cited above), see Quodvultdeus, Sermo de tempore barbarico, I: 5.2 and 5.6 (Corpus Christianorum, series Latina 60: 430–31) and Ps-Aug. Sermo, 396 (Patrologia Latina 39: 1715–16).


Harris and Gifford, 1 and 56: the MS also includes the martyrdom of Polycarp and the life of Porphyry of Gaza. Concerning other items, possibly indicating its liturgical aspect, see Mazzucco, “Versione greca e versione latina,” 48–49.


PPL and PPG 1.1 and 21.11 (Rebillard, 304 and 328; Amat, 98 and 182): Si vetera fidei exempla becomes Εἰ τὰ παλαιὰ τῆς πίστεως δόγματα; and et haec non minora veteribus exempla becomes Οὐχ ἧσσον τῶν παλαιῶν γραφῶν.


PPL and PPG 16.1 (Rebillard, 320; Amat, 158).


Typically used at the time: see, e.g., Passio Donatilla, Secunda et Maxima, 4 (Maier, Dossier du Donatisme, 1, 100): “quaedam puella, nomine Secunda, annorum circiter duodecim…” where it seems manifest that this is a fictitious creation of a “Secunda” who was about eleven years old and who was apprehensively considering the suitors who were planning to marry her when she reached the age of twelve years, the earliest legally permissible age to marry.


See Fialon, Mens immobilis, 45–46n24.


Some of this already noted by Franchi de' Cavalieri, 54n2.


The verb can have a spectrum of usage, but note its use three times elsewhere in the same account to indicate the inflicting of extreme physical pain in punishment: 18.9: Ad hoc populus exasperatus flagellis eos vexari per ordinem venatorum postulavit; 19.3: ipse (i.e. Saturninus) et Revocatus leopardum experti etiam super pulpitum ab urso vexati sunt; 20.9: non prius credidit nisi quasdam notas vexationis in corpore…; in other African martyr accounts, the word typically indicates harsh physical punishment, as in the African Acta Gallonii, 13–14, 48, 50 (P. Chiesa, “Un testo agiografico africano ad Aquileia: gli Acta di Gallonio e dei martiri di Timida Regia,” Analecta Bollandiana 114 [1996]: 265–67): Proconsul dixit: “vexate illum ungulis. Cumque vexaretur…”; Cumque impositus esset, proconsul dixit: “Ungulis vexetur”; Proconsul dixit: “Crudelius vexetur,” etc.


PPG, 3.3: … ἠθέλησεν τοὺς ὀφθαλμούς μου ἐξορύξαι· ἔπειτα μόνον κράξας ἐξῆλθεν… (Greek); APERPI, 3.3 … tunc pater eius…volens oculos eius eruere; et exclamans confusus egressus est.


Compare the Latin with the Greek at 21.7 (Amat, 178–80); already noted by Duchesne, “En quelle langue ont été écrits,” 46–47.


I strongly suspect that it is the manner in which the translator rewrote the original that led Louis Robert to think that the Greek text must be the original, especially because this scene, as Greek translator reshaped it, would make more coherent sense in the normal waking world as compared to what we read in the Latin.


See Harnack, “Die griechische Übersetzung,” 30–32, where he provides a resumé of his findings; E. A. Fisher, “Greek Translations of Latin Literature in the Fourth Century A.D.,” Yale Classical Studies 27 (1982): 173–215, at 203–07, adds little to what Harnack had already noted.


Harnack, “Die griechische Übersetzung,” 32–36 (translated quotation from 36).


As Dekkers (“Les traductions grecques,” 196) points out, the earlier versions of works by Tertullian in Greek were not translations but works originally written in Greek. If there was any translation it was in the other direction, but it seems doubtful that the later Latin versions were in any sense translations of the Greek. See Dekkers, ibid., 197–99, on what can be known of the Cyprian translation.


See Dekkers, “Les traductions grecques,” 206–12, for the details; the great African Father had to wait until the age of Maximus Planudes in the thirteenth century to find his Greek translator who chose to translate the de Trinitate.


See J. Farrell, “The Canonization of Perpetua,” ch. 16 in Bremmer and Formisano, Perpetua's Passions, 300–20, esp. pt. 4, “The Canon of the Saints,” 312–19, in the list of intercessory saints to whom prayers were directed nobis quoque peccatoribus. He expresses surprise that Felicitas ranks before Perpetua, but this was already the case in the Byzantine memorialization of the martyrs in the Basilica Maiorum at Carthage (where her precedence has also been the cause of some perplexity).


See M. Lapidge, “St. Felicitas and her Seven Sons,” entry 1 in The Roman Martyrs: Introduction, Translations, and Commentary (Oxford: Oxford University Press, 2018), 45–53.


Depositio martirum, § XII [in] Chronographus Anni CCCLIII = Chronica Minora, ed. Mommsen (Monumenta Germaniae Historica, auctores antiquissimi 9.1 1892: 13–148) at 71: Perpetua et Felicitas (non. Martias) and Cyprian (xviii kal. Octob.). See M. R. Salzman, On Roman Time: The Codex-Calendar of 354 and the Rhythms of Urban Life in Late Antiquity (Berkeley: University of California Press, 1990), 42–47. On the formation of the text, see R. W. Burgess, “The Chronograph of 354: Its Manuscripts, Contents, and History,” Journal of Late Antiquity 5 (2013): 345–96, at 382: the latest entry in the depositio is dated to 304, but the full list of depositions as we have it was probably completed in the mid-330s.


For the details, see Conant, “Europe and the African Cult of Saints,” 5–14.


See, e.g., Theod. Epp. P 22 (SC 40: 92–94); S 29 (SC 98: 86–90); 52–53 (SC 98: 128–30); 70 (SC 98: 152–54); on which see A. M. Schor, Theodoret's People: Social Networks and Religious Conflict in Late Roman Syria (Berkeley-Los Angeles: University of California Press, 2011), 157–68.


For a few of the specific ones noted here, see: H. Dolenz, Damous el-Karita: Die österreichisch-tunisischen Ausgrabungen der Jahre 1996 und 1997 in Saalbau und der Memoria des Pilgerheiligtums Damous el-Karita in Karthago (Vienna: Osterreichisches archäologisches Institut, 2001); and S. T. Stevens et al., Bir el Knissia at Carthage: A Rediscovered Cemetery Church (Ann Arbor: Kelsey Museum, 1993), 303–08: a simple basilica on the site was succeeded by a considerably larger construction in the early Byzantine period; and S. T. Stevens et al., Bir Ftouha: A Pilgrimage Church Complex at Carthage (Portsmouth RI: Journal of Roman Archaeology, Supplementary Series no. 59, 2005), 545: a huge complex constructed in the later 540s. For a contextual survey of these three basilicas, see Baratte et al., Basiliques chrétiennes, 2, no. 43, “Carthage” = Carthage 19, 20 and 23, at 132–42, 147–49.


See R. Miles and S. Greenslade, The Bir Messaouda Basilica: Pilgrimage and the Transformation of an Urban Landscape in Sixth Century AD Carthage (Oxford-Philadelphia: Oxbow Books, 2019): passim, but esp. 96–99, for a summary of the Byzantine phase of construction.


The bibliography is extensive. Most of it is provided with critical analysis by R. Bockmann, “Vandal Religious Policy and the Christian Topography of Carthage,” in Capital Continuous: A Study of Vandal Carthage and Central North Africa from an Archaeological Perspective (Wiesbaden: Reichert Verlag, 2013), 87–117.


On the Basilica Maiorum (Mcidfa), see Ennabli, ICKarth, 2: 3–12; and L. Ennabli, Carthage: une métropole chrétienne du IVe à la fin du VIIe siècle (Paris: Centre national de la recherche scientifique, 1997) no. 6, 19–20. For the basilica, see Baratte et al., Basiliques chrétiennes, 2, no. 43 “Carthage” = Carthage 21, at 142–45. For the inscription celebrating all the martyrs, see n67 above. For the other celebratory inscriptions, see ICKarth 2.3a = Corpus inscriptionum Latinarum 8.25038c; and 3b = Corpus inscriptionum Latinarum 8.25038b. One of them probably specifies the name of Secundulus: ICKarth 2 = Corpus inscriptionum Latinarum 8.25038a: […]n[o]n(as) Mar[tias…] / san(c) <imago crucis> [tus] / Sec[un] <imago crucis> [dulus]; ICKarth 2.2 and 3b preserve just enough of the dating of the natalitia of two of the other martyrs to make the reading non(as) Mart(ias) reasonably certain.


See S. Lancel, ed., Victor Vitensis, Historia persecutionis Africanae provinciae, 1.3.9 (Paris: Les Belles Lettres, 2002), 100–01: Et, ut de necessariis loquar, Basilica Maiorum, ubi corpora sanctarum martyrum Perpetuae et Felicitatis sepulta sunt, Celerinae vel Scillitanorum et alias quas non destruxerunt… It depends how one translates this sentence, but to my eyes the passage strongly suggests that the basilica of the Maiores was seized along with those of Celerina and the Scillitani. If so, the latter had basilicas of their own, although they seem to have been located in proximity to each other.


For the context, see J. Conant, Staying Roman: Conquest and Identity in Africa and the Mediterranean, 439–700 (Cambridge: Cambridge University Press, 2012), 163–65.


It seems that the name Maiulus is correctly read in the final line of the inscription, and martyr from Hadrumetum, Maiulus, is attested for 5 Ides May (11 May) in the Calendar of Carthage (Acta Sanctorum Nov. 2.2 [1931], lxx and 64). We might note, once again, the influence of Tertullian, since it is he who specifically mentions a Maiulus/Mavilus as one of the proto-martyrs in Africa: Tert. Ad Scap. 3 (Corpus Christianorum, series Latina 2: 1129). He was ordered to be thrown to the wild beasts at Hadrumetum by the governor Caecilius Capella: quod cum Adrumeticum Mavilum idem Caecilius ad bestias damnasset. Capella (Prosopographia imperii Romani2 C 735) was proconsular governor of Africa under Commodus in the years between 181–187 (Thomasson, Fasti Africani, no. 93: 72–73) which would place the death of Maiulus around the time of the Scillitan martyrs, the very first martyrs known for Africa. Capella must have ordered his execution at one of his conventus tribunals.


See N. Duval, “No. 23: Haïdra (Ammaedara): Basilique I dite de Melleus,” in Les églises africaines à deux absides: Recherches archéologiques sur la liturgie chrétienne en Afrique du Nord, 2: Inventaire des monuments-inteprétation (Paris: de Boccard, 1973), 191–99; N. Duval et al., Recherches archéologiques à Haïdra, II: La basilique I dite de Melleus ou de Saint-Cyprien (Rome: Ecole française de Rome, 1981), 114–19, 185–87; see Baratte et al., Basiliques chrétiennes, no. 121, “Haïdra” = Haïdra 1: 300–07: for the considerable additions and elaboration of the basilica that are Byzantine in date.


The inscription for the deposition of the remains of the bishop (antistites) and martyr Cyprian dates to 568–69: L. Poinssot, “Une inscription mentionnant les reliques de saint Cyprien,” Comptes rendus des séances de l'Académie des inscriptions et Belles-Lettres (1934): 249–54 = Duval-Prévot, Les inscriptions chrétiennes, no. 1: 20–22, figs. 4–5 : Hic abent[ur] / reliquie beat[i (or, beatissimi)]/ martiris et an/tistitis Cypria/ni d(e)p(ositae) a beat(o) Mel/leo ep(i)sc(op)o an(no) IIII / D(o)m(in)i Iustini Imp(eratori)s. The basilica itself was probably a Vandal period fifth-century construction (Duval, Recherches archéologiques, II, 204–07) and therefore had to be reclaimed by orthodox Catholics under the Byzantine occupation. On the later use of the cult of Cyprian, see Y. Duval, Loca sanctorum, 674–81.


For Victorinus, see N. Duval and F. Prévot, Recherches archéologiques à Haïdra, 1: Les inscriptions chrétiennes (Rome: Ecole française de Rome, 1975): no. 58 = Année épigraphique 1968: 638 = Année épigraphique 1971: 502; the words episcopus Vandalorum were added by a later hand, inserted into the inscription as a purposefully pejorative label. On the bishop Melleus, see Duval and Prévot, Les inscriptions chrétiennes, no. 3: 25–27, fig. 8 = Année épigraphique 1946: 22 (Haïdra, Sufetula): Melleus ep(is)c(opus) re/quiebit in p(a)c(e)… Under the word episcopus, a second hand added the letters VNTTS which, as Duval saw, must be a shortened form of unitatis: N. Duval, “Episcopus unitatis: à propos de trois inscriptions chrétiennes d'Afrique,” Bulletin de la société nationale des antiquaires de France (1959): 147–50. This designation also seems to have been added to the original inscription in an effort deliberately to contrast Melleus with Victorinus. For the cultural significance of these moves, see R. Whelan, Being Christian in Vandal Africa: The Politics of Orthodoxy in the Post-Imperial West (Berkeley: University of California Press, 2018), 29–32.


Corpus inscriptionum Latinarum 13.1770 (Lugdunum): praefectus classis Ravennatium, see Prosopographia imperii Romani2 A 409; coming from an important port city, celebrated in the mosaic panels in the Piazzale delle Corporazioni at Ostia; one suspects maritime expertise.


In taking this approach, I agree with Conant, “Europe and the African Cult of Saints,” 4: “the spread of collective and individual veneration of African holy men and women throughout Christendom was probably the product of the circulation of people, things, and ideas that characterized exchange in general in Late Antiquity.”


See Chiesa, “gli Acta di Gallonio,” 241–268; and “Pellegrino martire in urbe Bolitana e Pellegrino di Ancona: un' altra agiografia africana ad Aquileia?” Analecta Bollandiana 116 (1998): 25–56.


Hieron. De vir. ill. 97: “Fortunatianus, natione Afer, Aquiliensis episcopus, imperante Constantio, in Evangelia, titulis ordinatis, brevi et rustico sermone scripsit commentarios.” Fortunatianus was the author of the earliest western Latin commentary on the gospels, a new MS of which was discovered in 2012: see L. J. Dorfbauer, ed., Fortunatianus Aquileiensis: Commentarii in Evangelia = Corpus scriptorum ecclesiasticorum Latinorum 103 (Berlin-Boston: de Gruyter, 2017), who dates Fortunatianus to the 340s-360s and the text to some point in the reign of Constantius II (2–3).


See M. Humphries, Communities of the Blessed: Social Environment and Religious Change in Northern Italy, AD 200–400 (Oxford: Oxford University Press, 1999), 173–75: they were being founded in the latter half of the fourth century.


A context perhaps best caught by Brown, The Eye of a Needle, 275–77.


The literature is extensive. I shall refer here, selectively, a few items: A. Augenti and C. Bertelli, eds., Felix Ravenna: la croce, la spada, la vela: l'alto Adriatico fra V e VI secolo (Milan: Skira, 2007); and D. M. Deliyannis, Ravenna in Late Antiquity (Cambridge: Cambridge University Press, 2010).


The chapel was built around 500 and so is the earliest of the three cases considered here; see A. van den Hoek, “Female Splendor: The Role of Twelve Women Saints in the Apse of the Euphrasian Basilica (Parentium),” Eastern Theological Journal 1 (2015): 265–307, at 288–90, figs. 19–20.


The presence of Euphemia in such a prominent position should indicate that the program of mosaics dates to later than the 450s, after Council of Chalcedon (451) with which she was especially connected in the West: see A. Terry and H. Maguire, Dynamic Splendor: The Wall Mosaics in the Cathedral of Eufrasius at Porec, 2 vols. (University Park PA: The Pennsylvania State University Press, 2007), 2: 133, drawing attention to the work of the archbishop Helias of Aquileia in the restoration of the basilica of Euphemia at Grado and at the council at Grado in the late 570s that affirmed the decisions of Chalcedon. See also O. von Simpson, Sacred Fortress: Byzantine Art and Statecraft in Ravenna (Princeton: Princeton University Press, 1987), 86–87, on the significance of Euphemia at the head of the procession of female saints in Sant'Apollinare Nuovo.


Van den Hoek, “Female Splendor,” 291–93, figs. 21–23.


Perhaps additionally one Vincentia: see Deliyannis, Ravenna in Late Antiquity, 363–64n147: at least, she was thought to be African.


The Euphrasian basilica in the form that we have it with the mosaics concerned was constructed in the second quarter of the sixth century. For the mosaics see Terry and Maguire, Dynamic Splendor: there were extensive nineteenth-century restorations (ibid., chs. 1 and 2) but, as far as I can see, they do not affect the arguments advanced here.


See Van den Hoek, “Female Splendor,” 275, fig. 7: the apse mosaic features the Virgin Mary, with Euphrasius and Maurus to the viewer's left.


The regional mix of saints and martyrs at Sant'Apollinare are as follows (see chart: Deliyannis, Ravenna in Late Antiquity, 167): women: Italian: 13; African: 4; eastern: 4; other: 1; men: Italian: 21; African: 1; eastern: 2: other: 2


N. Duval, “L'église V (des saints-Gervais-Protase-et-Tryphon) à Sbeïtla (Sufetula) Tunisie,” Mélanges de l'archéologie et d'histoire de l'école française de Rome: Antiquité 111 (1999): 927–89, at 975–78; for the church, see Baratte et al., Basiliques chrétiennes, 2, no. 147, “Sbeïtla” = Sbeïtla 6, 398–402: the basilica is of Byzantine date. This is the sole appearance of the martyr Tryphon in Africa, surely suggestive of a deliberate decision to import. He only appears much later in Italy, at Capua and Naples, without doubt, as Duval states, having come there from Africa. Notably, his martyrdom is the very first one at the head of the MS that contains the Greek translation of Perpetua's martyrdom: Hefferan, Passion of Perpetua, 427.


See E. Cirelli, “Late Roman Lamps from the Port Area of Classe (Ravenna),” in Roman and Late Antique Lamps: Production and Distribution, Contacts on the Mediterranean (Zagreb: Institute of Archaeology, 2016), 93–106, at 94–96: mainly “Atlante X” and Hayes 2 types.


The sheer numbers are striking: over 1,300,000 pottery fragments, representing about 150,000 vessels, of which about 8,200 were transport amphorae: see Cirelli, “Late Roman Lamps,” 541–42, for the north African production; and 542–43, for the eastern production centers.


The following is offered as exemplary: E. Cirelli, “Typology and Diffusion of Amphorae in Ravenna and Classe between the 5th and the 8th centuries A.D.,” in Late Roman Coarse Wares, Cooking Wares and Amphorae in the Mediterranean, ed. N. Poulou-Papadimitriou, A. Nodarou and V. Kilikoglou (Oxford: Archaeopress, 2014 = BAR International Series 2616.1), 1: 541–52.


See D. Dobreva and A. Riccato, “Aquileia e il Vicino Oriente. Il commercio di prodotti Levanti in area Adriatica,” Aquileia Nostra 86 (2015): 111–39.


As is suggested by the sum of all the evidence relating to the problem: see Conant, “Europe and the African Cult of Saints,” 29: “The evidence is suggestive rather than conclusive, but with respect to the Byzantine world it hints that Italy may well have played a central role in the transmission of African cults to the eastern Mediterranean.”


Well noted by Duval, Loca Sanctorum, vol. 2, 682.


Both were probably celebrated on a mosaic of what has dubiously been called “the monastery of St. Stephen” at Carthage along with Speratus, Stefanus, Sirica, Saturus, and Saturninus: see ICKarth 3.297 = Corpus inscriptionum Latinarum 8.25036 and 25037 (Carthage); Duval, Loca Sanctorum, vol. 1, no. 3: 7–10; and Baratte et al., Basiliques chrétiennes, 2: no. 43: “Carthage” - Carthage 8, “Monastère de Saint-Etienne,” 121–22. The mosaic is of Byzantine date; the chapel was built first in the fifth century, but then redone in the Byzantine age. Another mosaic in the chapel contained the label Beatissimi martyres: ICKarth 3.298. Saturus was also celebrated in a baptistry at Carthage: Duval, ibid., no. 5: 12–13 (Byrsa).


On the Basilica Scillitanorum see Victor Vitensis n93 above.


See Van den Hoek, “Female Splendor,” 282–307, whose observation stands apart from her arguments on their relationship to liturgical practices.


H. Delehaye, ed., Synaxarium ecclesiae Constantinopolitanae e codice Sirmondiano nunc Berolinensi adiectis synaxariis selectis, Acta Sanctorum, Nov. Propylaeum (Brussels: Societas Bollandiana, 1902 [reprint 1985]), cols. 825–26, 440 and 536: Speratus: 17 July; Perpetua: 2 February, 4 March, and 14 March respectively (cf. Breviarium Syriacum, 7 March, p. liv). The synaxarion of Constantinople dates to the tenth century, but surely collates evidence available from earlier times: see Conant, “Europe and the African Cult of Saints,” 28–29.


See Delehaye, Synaxarium ecclesiae Constantinopolitanae, 16 April and 30 Aug (cols. 605–06 and 933), depending on a Latin text which, like that of the Acta of Perpetua and of the Scillitanorum, appears to have been “redone” in a rather fictive fashion in the fifth or sixth century. For the history of the Latin Passion (BHL 2895b) see Delehaye, Analecta Bollandiana 39 (1921): 247–76. By this route, Felix made his way into a kanon composed in the East in the ninth century. The Greek translation, alas, has been lost: see Lequeux, “Latin Hagiographical Literature,” 392–94.


A good case is made by Van den Hoek, “Female Splendor,” 298–301, plus her appendix, figs. 1–4, for a significant influence of the liturgy on the choice of the martyrs and saints. It is important to note the cautions of Deliyannis (Ravenna in Late Antiquity, 168) that there are other important saints and martyrs known for northern Italy who, for whatever reasons, were not selected for illustration.


Conant, “Europe and the African Cult of Saints,” 37: “if devotees of African saints traveled in the same ships that carried the region's exports abroad, it would help make sense of some of geographical and chronological patterns in the sources.”