Historia del Nuevo Testamento en Latín

Desde los orígenes hasta el final del siglo III

Este texto ofrece una amplia introducción a la historia y el desarrollo del Nuevo Testamento Latino. Se ofrece una nueva síntesis histórica, reuniendo las pruebas de los autores cristianos y los manuscritos bíblicos desde los primeros tiempos hasta la Baja Edad Media. Cada testimonio se considera en su contexto cronológico y geográfico, para construir el panorama general de la transmisión del texto.

La literatura presenta las características de los manuscritos bíblicos latinos y se examina cómo la tradición latina puede servir de testigo para el Nuevo Testamento griego. Además, se considera cada libro del Nuevo Testamento por separado, con detalles de los principales testigos y rasgos de especial interés textual. Se describen con detalle las tres principales ediciones eruditas del Nuevo Testamento en latín (la edición Vetus Latina, la Vulgata de Stuttgart y la Vulgata de Oxford). También se ofrece información sobre otras ediciones y recursos, lo que permite a los investigadores comprender la importancia de los distintos enfoques y conocer los últimos avances.

Keywords: Latin, New Testament, Vetus Latina, Old Latin, Tertullian, Cyprian, Itala, Afra, translation, textual criticism

The origins of the Latin New Testament are unknown. No-one is explicitly identified as a translator or reviser of the Bible before the end of the fourth century. Jerome and Augustine’s comments on the origins and previous history of the Latin translation have often been accepted without question, even though they are writing some two centuries later in justification of their own endeavours. A more reliable account has to be pieced together from surviving writings contemporary with the adoption of Latin in the early Church and the evidence of the biblical text itself. This results in a focus on Roman North Africa, where the shift from Greek to Latin appears to have preceded the same development in Italy and elsewhere in the Roman Empire. Nevertheless, the paucity of texts preserved from this time means that significant gaps remain and it can be difficult to contextualize the evidence which survives.

The Scillitan Martyrs
The earliest dated reference in Latin to the books of the New Testament is found in the proceedings of the trial of a group of Christians in Carthage, known as the Scillitan Martyrs, held on 17 July 180 (A-SS Scilitani):

Saturninus proconsul dixit: Quae sunt res in capsa uestra? Speratus dixit: Libri et epistulae Pauli uiri iusti.1

Saturninus the proconsul said: What are the objects in your carrying case?

Speratus said: Books and letters of Paul, a righteous man.

While the unpunctuated text of Speratus’ reply appears to attribute both ‘books’ and ‘letters’ to Paul, it has been suggested that a comma should be placed after libri and that the word should be interpreted as ‘gospel books’.2 Equally, it might be that Speratus originally qualified libri with a word which was not familiar to the court stenographer (e.g. libri euangeliorum, ‘books of the Gospels’), which was simply omitted from the record. Although the trial proceedings are in Latin, the administrative language of Roman North Africa, the language of the Christian texts themselves is unspecified: they may still have been Greek, although a quotation of 1 Timothy 6:16 by one of the martyrs resembles a form found in the fifth-century African writer Quodvultdeus (QU). These proceedings are the oldest Latin example of a series of court records involving Christians (‘Acts of the Martyrs’) which were collected and circulated; there are also similar Greek texts from elsewhere in the Mediterranean. In some churches, especially in Africa, there was a tradition of reading out the record during the annual commemoration of each martyr.

Tertullian
The first Christian author to write in Latin whose works survive is Tertullian (TE), active in North Africa at the end of the second century. Tertullian’s earliest writings date from around 196 or 197, and his output spans two decades. Later works show evidence of Tertullian’s adoption of the doctrines of Montanism; the Passion of Perpetua and Felicity (A-SS Per), an extended set of martyr acts written in Latin at the beginning of the third century, is believed to have been written by one of Tertullian’s circle. Tertullian wrote in Greek as well as Latin: although only his Latin works are extant, these bear witness to his knowledge of both languages (e.g. De baptismo 15 and De corona 6). The entire New Testament canon is represented in his quotations with the exception of 2 Peter, and 2 and 3 John.3 Although scriptural passages are found throughout his works, two writings are particularly significant for the history of the biblical text. In Adversus Praxean (TE Pra), Tertullian uses the Gospel according to John as the basis for a carefully-constructed refutation of Monarchianism. Adversus Marcionem (TE Marc) is an attack on Marcion, who several decades before had produced a form of the New Testament consisting of an expurgated text of the Gospel according to Luke and ten of the Pauline Epistles: the rest, Marcion alleged, had been corrupted by a group which he called Judaizers. In books four and five of Adversus Marcionem Tertullian examines Marcion’s treatment of Luke and Paul respectively, transmitting vital information about the nature and extent of this lost edition. Additionally, a set of prologues for the Pauline Epistles which appear to derive from (p.6) Marcion’s edition are transmitted only in Latin tradition: they are first attested from the middle of the fourth century.4

Tertullian’s biblical text poses numerous problems. He rarely, if ever, cites the same verse twice in exactly the same form, sometimes even within the same work. For example, the opening verse of John begins in principio at TE Pra 13.3 but a primordio at TE Pra 16.1. De anima and De baptismo have versions of John 3:5 both of which correspond to known Greek forms yet have little in common with each other:

nisi quis nascetur ex aqua et spiritu non inibit in regnum dei. (TE an 39) nisi quis renatus fuerit ex aqua et spiritu sancto non intrabit in regno caelorum. (TE ba 13.3)

Unless someone will be born of water and the spirit, they will not go in to the kingdom of God.

Unless someone shall have been born again of water and the holy spirit, they will not enter the kingdom of heaven.

Such examples can easily be multiplied. One suggestion is that Tertullian was using more than one version, reproducing Marcion’s text at one point and his own elsewhere, but the phenomenon is observable in the biblical books not included in Marcion’s New Testament and works not directed against Marcion. Furthermore, there is no external evidence that Marcion’s Gospel circulated in any other language than Greek.5 A more comprehensive explanation is that Tertullian was not working with a fixed form of the Latin Bible but produced his own translation as necessary, comparing Marcion with another Greek text of the biblical passage, and perhaps using different Greek manuscripts in other works. Support for this is found in the lack of overlap between Tertullian’s biblical text and the majority of surviving Old Latin forms. This distance from the rest of the Latin tradition is observed in the Vetus Latina editions, which use the siglum X to indicate text-types reconstructed from Tertullian’s quotations and other early authors who probably relied on a Greek original.6 Nonetheless, there remain occasional similarities between Tertullian’s quotations and Latin biblical manuscripts which suggest that he might have used a translation of the New Testament. Furthermore, two of his comments imply the existence of a Latin version of at least the Pauline Epistles. In De monogamia, he contrasts a Latin text of 1 Corinthians 7:39 omitting the second occurrence of uir eius (‘her husband’) with the reading in Graeco authentico, ‘in the authentic Greek’ (TE mon 11). Similarly, he says sicut inuenimur interpretatum (‘as we find it translated’) of a particular reading in Galatians 4:24 at TE Marc 5.4.7

(p.7) ‘Christian Latin’
Tertullian was a pioneer in the development of a Latin Christian vocabulary. Faced with the challenge of translating Greek terms which had a special Christian significance, Latin writers had three main options. Where a Greek word had been given an additional meaning (for example, μάρτυς‎ coming to mean ‘martyr’ as well as ‘witness’) this further sense could be attributed to an existing Latin word through the process of ‘semantic extension’. So, in some sources, the Latin for ‘witness’, testis, is also found in the sense of ‘martyr’. The second possibility was to create a calque, a new Latin word in which each morphological element corresponded to the Greek: the exact match between the words for ‘to enliven’, ζῳο-ποι-εῖν‎ and uiui-fic-are, is an example of this. Finally, the Greek term itself could be borrowed, usually becoming naturalized into Latin morphology: the word for ‘overseer’ or ‘bishop’, ἐπίσκοπος‎, was thus adopted as a noun, episcopus, for the technical Christian usage. In the earliest Christian writings, there is considerable fluidity in this technical lexicon. Numerous examples may be seen in Tertullian and other early biblical translations of initial attempts to create a Latin vocabulary through semantic extension which were later replaced with borrowings, for instance the use of tinguere to mean ‘to baptize’ as well as ‘to dye’ on the model of the Greek βαπτίζειν‎, or minister corresponding to διάκονος‎ (‘deacon’).8 Conversely, such texts may also have a liberal sprinkling of Greek borrowings for which a Latin term was normally preferred, such as horoma for uisio in the Passion of Perpetua and Felicity (A-SS Per 10.1). This bears witness to the bilingualism of early Latin Christian communities.

The origin of the use of Latin in a Christian context is usually associated with the liturgy. Just as, several centuries earlier, a scriptural reading in Hebrew was sometimes followed by an impromptu translation called a targum for the benefit of those attending Jewish worship not familiar with the language, the same is likely to have been the case in Christian gatherings. The only direct evidence we have of this practice is the account by the Spanish pilgrim Egeria, sometimes known as Aetheria, of her visit to the Holy Land between 381 and 384 (IT Ae). She explains how in Jerusalem the services were conducted by the bishop entirely in Greek, but there was a priest on hand who translated the sermon and biblical readings into Syriac and the same was done for those who understood only Latin. An origin in oral paraphrase is more plausible than the suggestion that Christian scriptures were translated into Latin primarily as a missionary strategy for reading by unbelievers, although examples of the (p.8) reading of the Bible by non-Christians mentioned by Tertullian and others are given below. Nevertheless, the early text of the Latin New Testament in surviving manuscripts stems from a written translation corresponding closely to Greek: it has even been proposed that it was originally an interlinear version written between the lines of a Greek text.9

The distinctiveness of the Latin language used by Christians with its Greek influence and unusual forms, some of which become standard in later Latin, led a group of twentieth-century scholars known collectively as the Nijmegen School to propose that ‘Christian Latin’ was a separate language (or Sondersprache).10 Not only did it feature numerous innovations for technical terms but it also appeared to have a different vocabulary for words which had no specialist Christian connotations. Examples of these include confortare (‘to comfort’), proximare (‘to approach’), or refrigerium (‘relief’). The theory also interpreted comments from early authors about ‘our writings’ as an indication of Christian linguistic peculiarity, as in the following line from Tertullian:

tanto abest ut nostris litteris annuant homines ad quas nemo uenit nisi iam Christianus. (TE an 1.4)

It is so remote that people agree to our writings, to which no-one comes unless they are already Christian.

In context, though, it was availability and use rather than language which posed a hindrance to potential users. From the beginning, the Sondersprache concept appeared to be an overstatement because the characteristics of Christian discourse were limited to the lexicon. Morphology and syntax were unaffected, apart from the influence of a Greek original on translations. Furthermore, vocabulary in Christian texts which is absent from classical Latin is sometimes attested in the early playwrights, especially Plautus, or other works written in a lower register.11 These terms therefore appear to be indicative of popular speech and form part of a continuum of colloquial Latin which eventually led to the Romance languages. The high volume of evidence preserved from ecclesiastical writers may have given the misleading impression that the phenomenon was distinctively Christian, rather than setting it within the broader context of non-literary and post-classical forms. Latin translations of the Bible and significant early writers such as Tertullian undoubtedly had an influence on Christian vocabulary and figures of speech (for example, some of the Semitic constructions transmitted through the Greek of the Septuagint), but there is no indication that non-Christians found it hard to read texts by their Christian contemporaries for linguistic (p.9) reasons. Indeed, there are numerous exhortations for non-believers to read the Bible for themselves, as in Tertullian’s Apology:

inspice dei uoces, litteras nostras, quas neque ipsi supprimimus et plerique casus ad extraneos transferunt. (TE ap 31.1)

Consult the words of God, our scriptures, which we do not ourselves hold back and which many situations bring to those outside the community.

Similar encouragements from Augustine are quoted below.12 While the ‘plain register’ (sermo humilis) of the biblical translations may have been an embarrassment for more literary converts, it was also treated as a virtue by apologists and contrasted with the esotericism of other religious discourse.13 The full Sondersprache hypothesis now finds few adherents, although careful linguistic analysis may still identify aspects of language use peculiar to Christian groups, as has sometimes been the case in subsequent generations.14

Cyprian and the First Latin Bibles
The biblical quotations of Cyprian, bishop of Carthage from 248/9 to 258, provide evidence for a Latin translation of the New Testament in third-century Africa. His numerous works, all in Latin, have a consistency in their scriptural text which indicates that they derive from a fixed version. A further indication of Cyprian’s reliance on a standard Latin form may be seen in a difference in vocabulary between his own writing and his biblical text, such as his use of caritas and gloria.15 There are even examples in Cyprian’s quotations of what seem to be copying errors within Latin biblical tradition, such as ut suscitentur for ut iudicentur at 1 Peter 4:6. If this is the case, it demonstrates that several generations of copies had preceded the text used by Cyprian. His two collections of testimonia, biblical extracts grouped under particular thematic headings, are of particular value: the three books of Ad Quirinum (CY te) date from 248 or 250 while Ad Fortunatum (CY Fo) is slightly later. The text-type reconstructed from Cyprian’s quotations in the Vetus Latina edition is given the siglum K.16 The oldest surviving Old Latin gospel (p.10) manuscript, VL 1 (Codex Bobiensis), has a similar form of text and is sometimes referred to as k.17 Although it was copied in Africa in the fourth century, its text appears to antedate Cyprian. This is most clearly shown by the ending of Mark, illustrated in Image 2. VL 1 is the only gospel manuscript in Greek or Latin which has the ‘shorter ending’ by itself, while Cyprian seems to be familiar with the ‘longer ending’ of Mark 16:9 onwards.18 Cyprian’s text of John in Ad Fortunatum is very similar to that of VL 2 (Codex Palatinus), while his citations of Acts are close to VL 55 (the Fleury Palimpsest).

There are a number of writings attributed to Cyprian which, although not authentic, may well be of a similar date and offer important evidence for early versions of the biblical text. De montibus Sina et Sion (PS-CY mont) has an intriguing form of John 2:19–21, reading fanum (a word often used in conjunction with pagan religious sites) rather than templum.19 There is some controversy over the date and location of De aleatoribus (PS-CY al): it has been dated as early as the end of the second century in Rome, but current preference is for a fourth-century African origin because its biblical text seems to be drawn from the testimonia in Cyprian’s Ad Fortunatum.20

Although there is surprisingly little overlap between Cyprian’s text and the biblical quotations of Tertullian, Cyprian features a number of the innovative early forms described above which were later replaced, such as baptiziator rather than baptista (e.g. Matthew 3:1, 11:12), similitudo for parabola (e.g. Matthew 13:35–6), praessura for tribulatio (e.g. Romans 5:3–4, 8:35), and even agape for caritas (e.g. Romans 8:35). There are preferences for certain renderings such as nequam for malum (‘evil’), quoadusque for donec (‘until’), ploratio for fletus (‘lament’), and even quoniam rather than quia or quod (‘because’) and fui rather than eram (‘I was’). Because many of these forms are peculiar to these African witnesses they are often described as ‘African’ readings or renderings, although they should not be considered as evidence for an African dialect of Latin (Africitas). The idea of Africitas was proposed by Sittl in 1882 but soon fell out of favour: Capelle 1913 and Löfstedt 1959 demonstrate that these terms were not confined to Africa. A full list of New Testament examples of such vocabulary is given in the work of Valgiglio of 1985.

A quick glance at the alternatives will show that most are forms common to Latin authors in general. In the present study, the designation ‘archaic’ is preferred to ‘African’ to represent these early terms which, although they may have been current in the community where the translation was first made, soon fell out of favour.

There is a degree of freedom in the earliest text which contrasts markedly with the traditional description of these as slavishly literal translations, full of (p.11) vulgarisms and infelicities, committing numerous grammatical and stylistic solecisms by reproducing Greek constructions.22 Words or phrases are sometimes omitted or paraphrased, especially if they are repetitive or explanatory in function.23 Burton has shown how the Old Latin Gospels often bear witness to considerable linguistic sensitivity on the part of the translators.24 Although there are frequently orthographical errors and nonsense readings in the oldest surviving manuscripts, the translation itself is not the work of an incompetent. Indeed, to have sufficient expertise both to read Greek and to write Latin is an indication of a relatively advanced degree of education. Augustine’s dismissive comment in De doctrina christiana has long been overapplied:

ut enim cuique primis fidei temporibus in manus uenit codex graecus et aliquantulum facultatis sibi utriusque linguae habere uidebatur, ausus est interpretari. (AU do 2.11.16)

For, in the first days of the faith, whenever a Greek manuscript came into the possession of someone who believed himself to have a modicum of ability in both languages, he hazarded his own translation.

In context, this refers to translations of the Old Testament, where Semitic idioms and points of obscurity may have resulted in greater confusion. The general direction is from a periphrastic early version (consistent with the Cyprianic text) towards an ever closer correspondence with a Greek text, culminating in the form adopted as the Vulgate.25 Jerome comments in the preface to his revision of the Gospels:

si enim latinis exemplaribus fides est adhibenda, respondeant quibus: tot sunt paene quot codices.26 (HI Ev)

If trust is to be placed in Latin originals, let them tell us which ones: there are almost as many as there are manuscripts.

The nature of the distinction between exemplaria (‘originals’) and codices (‘manuscripts’) is not immediately obvious, and may represent Jerome’s rhetorical attempt to establish the priority of his text founded on a new comparison with Greek: Latin copies were never ‘original’ and Jerome goes on to describe even his own version of the Gospels as a light revision of an existing text.27

(p.12) So was there one initial Latin translation of the New Testament, which then underwent numerous revisions, or were there multiple independent translations from which a handful—and eventually one in particular—became dominant? If we leave to one side the potentially unreliable comments of later authors and turn to the surviving textual evidence, the balance of probability favours the former. Editors of Old and New Testament books in the Vetus Latina series have reached the conclusion that in each case a single Latin translation underlies all the surviving evidence for the Old Latin tradition.28 This does not remove the possibility that other translations were made at an early stage, but little if anything of these remains. As noted above in the case of Tertullian, variation in Latin biblical quotations in the initial centuries is often likely to indicate direct use of a Greek source rather than an alternative Latin version in circulation. Even the Gospels, for which the surviving manuscript evidence goes back furthest, display shared features for which the simplest explanation is a single common original. These range from the sequence Matthew–John–Luke–Mark in the majority of Old Latin codices to common omissions, patterns of rendering and even particular words.29 Individual forms may have been substituted here and there, absent text supplied and paraphrases brought into closer conformity with a Greek source, but the overall shape remains remarkably consistent. Furthermore, as early as the middle of the fourth century, there is evidence for a conservatism in Latin Christian culture pertaining specifically to the biblical text.30

Occasions when the Latin tradition agrees on a reading not or poorly attested in Greek provide evidence in favour of a single original translation. One of the best known examples is Mark 9:15, where all pre-Vulgate Latin manuscripts have gaudentes, ‘rejoicing’ rather than ‘running’, apparently due to the misreading of προστρέχοντες‎ as προσχαίροντες‎ (as found in Codex Bezae).31 At Luke 1:9, early Latin tradition agrees on introitus eius, ‘his entrance’, even though Greek witnesses only have τῷ λόγῳ‎, ‘word’. Most Old Latin manuscripts reverse the sequence of phrases in Luke 9:62, with ‘looking backwards’ before ‘putting his hand to the plough’. A second indication of the general uniformity of the Latin tradition is agreement on a particular reading in one location when multiple alternative renderings are attested elsewhere. Although the choice of word may in some cases be prompted by sensitivity to context, this is not always the case. For example, (p.13) although there are five different translations preserved of the relatively rare Greek word for ‘inn’, τὸ κατάλυμα‎, at Luke 22:11 and Mark 14:14, at Luke 2:7 all but one manuscript has diuersorium. Again, every instance of ὁ λίθος‎ (‘stone’) in Matthew is translated by lapis apart from Matthew 27:60, where all Old Latin manuscripts (apart from one known to be influenced by the Vulgate) have saxum, despite reading lapis for the same stone seven verses later. The verb ‘to eat’, ἐσθίω‎, is normally translated by manducare, but almost all manuscripts switch to edere at Matthew 15:27 and cenare at Matthew 26:26. The treatment of words which occur only once in the Greek New Testament is also instructive. These include ἀμαράντινον‎ at 1 Peter 5:4 and ἐπιούσιον‎ in the Lord’s Prayer at Matthew 6:11 and Luke 11:3. If the surviving manuscripts derived from independent translations, one would expect variation in these unusual words for which there was no obvious Latin equivalent, rather than universal agreement on immarcescibilem and cottidianum respectively. Even the more common word τέκτων‎ (‘builder’) in Matthew 13:55 and Mark 6:3 is always rendered by faber, while at John 21:5 almost all witnesses have pulmentarium for the unique προσφάγιον‎. It has been suggested that the early Latin translators may have had some connection with Jewish communities because of their treatment of technical terms: the use of cena pura rather than praeparatio for παρασκευή‎ (‘the day of preparation’) appears to reproduce a Jewish practice.32

There are also practical reasons which may explain how a single translation could become widespread. Books circulated relatively quickly and easily around the Mediterranean. If a Latin translation were known to exist already, users might have preferred to make a copy of that (with their own adjustments) rather than start from scratch. While the need for Latin copies of the New Testament probably arose at around the same time in different communities, meaning that early translators may have worked in parallel, it is not impossible that a single original could have exerted a wide influence through multiple subsequent copies. The likelihood of this would be increased if, like Jerome’s Gospels later on, it had some form of prestige through association with an authoritative writer or ecclesiastical figure. It is worth observing that, despite the probable origin of Latin translations in an oral context, all the surviving evidence is literary.

Nonetheless, the theory of a single translation of all biblical books in the early third century is not without its problems, given both the ongoing debate at this time about the scriptural canon and the nature of biblical codices. Pandects, that is manuscripts containing the Old and New Testament in a single volume, are unknown until the appearance of the Greek Codex Vaticanus (GA 03) and Codex Sinaiticus (GA 01) around the end of the fourth century. Even after that (p.14) it was still the norm for books to be circulated in smaller collections, such as the Gospels or the Pauline Epistles. The earliest Latin pandects appear to have been assembled in the fifth century from multiple manuscripts.33 It therefore remains possible that the single versions claimed to underlie the surviving Latin tradition had various origins: while Africa provides the earliest evidence for the Gospels, the Pauline Epistles may have originated elsewhere.34 A detailed survey of the translation technique in different books is needed to determine whether or not this is the case. At the same time, the loss of most of the early Latin New Testament manuscripts makes it difficult to quantify the amount of revision and extent of variation across Latin tradition. Almost all pre-Vulgate witnesses have a greater or lesser number of readings which are now unique to them but which may have had wider currency in manuscripts which no longer survive.

An early revision of the Latin text of the Gospels around the time of Cyprian is attested in a set of capitula (chapter titles), part of the prefatory material commonly found in later gospel manuscripts. By a remarkable accident of preservation, this series (KA Cy) is only present in a handful of much later manuscripts, which have a Vulgate form of the biblical text.35 Nevertheless, the affiliation of the passages quoted in these lengthy summaries corresponds very closely to the text of Cyprian and VL 1. The inclusion of the story of the Woman taken in Adultery (John 7:53–8:11) suggests that these capitula post-date Cyprian, who does not quote the passage. On the other hand, some of the renderings are more ancient than Cyprian, indicating different layers in the biblical text even at this early stage. Another set of capitula for Matthew (KA D) also has similarities with VL 1. The provision of chapter titles and other paratextual information goes hand in hand with a revision of the biblical text as part of the creation of a new edition of the Bible.

Christian Authors in Europe
Greek continued to be the first language of the early Church at the turn of the third century in Europe, as shown by Irenaeus of Lyons and Hippolytus of Rome. Even so, there is also evidence for the use of Latin at this time. In the Greek text of the Shepherd of Hermas, the Latin word statio is borrowed as a way to speak of ‘fasting’.36 Jerome identifies Victor, bishop of Rome in the (p.15) 190s, and his contemporary Apollonius as ‘Latins’ (latinorum; HI ill 53), implying that their theological treatises and writings against the Montanists may have been in Latin. However, there was ongoing traffic between the metropolis and the colonies. Many prominent Roman Christians were of North African origin, including Pope Victor and possibly also Minucius Felix, whose apologetic treatise Octavius (MIN) is set in Rome. Cyprian corresponded regularly with Roman clergy, including the presbyter Novatian (NO). Few of Novatian’s writings have survived, because of his excommunication as a heretic. The most substantial is his treatise De trinitate (NO tri). His biblical text has sometimes been claimed to be a separate Roman tradition (Vetus Romana) or even a witness for the gospel harmony known as the Diatessaron, but is textually similar to later Italian tradition, especially VL 3 (Codex Vercellensis) in the Gospels.37

The tradition of referring to the Old Latin versions as Itala derives from a comment by Augustine on the Old Testament:

in ipsis autem interpretationibus, Itala ceteris praeferatur; nam est uerborum tenacior cum perspicuitate sententiae. (AU do 2.15.22)

As for the translations themselves, the Itala is preferable to the rest; for it keeps more closely to the words and gives the sense with clarity.

The word Itala has been much debated, with some scholars suggesting that it is corrupt and others that it referred to Jerome’s Vulgate.38 However, the adjective Italus is used elsewhere by Augustine (AU ord 2.5.15 and 2.17.45) and, given his designation of other biblical manuscripts as African (e.g. AU re 1.21.3), the best interpretation is that this is a geographical term indicating pre-Vulgate translations of Italian origin, perhaps those which he encountered during his time in Milan.39 Another observation later in the same paragraph indicates that certain places were renowned for the quality of their biblical texts:

libros autem Noui Testamenti, si quid in latinis uarietatibus titubat, graecis cedere oportere non dubium est, et maxime qui apud ecclesias doctiores et diligentiores repperiuntur. (AU do 2.15.22)

As for the books of the New Testament, if the variety of Latin manuscripts leads to any uncertainty, there is no doubt that they should give way to Greek ones, especially those which are found in more learned and responsible churches.

(p.16) The use of Itala for the entire tradition, or even just Old Latin texts which do not preserve the most archaic features, is therefore unduly restrictive and has generally been abandoned.

Some third-century Latin writers continued to use Greek biblical texts. This is the case for Victorinus of Poetovio (also known as Ptuj or Pettau), who wrote a commentary on the Apocalypse (VICn Apc) before his martyrdom in the Diocletianic persecution around the year 304. Like Tertullian, Victorinus is referred to by the siglum X (or Y) in the Vetus Latina edition, to indicate his dependence on Greek: the text of Revelation in the biblical lemmata in the commentary appears to be Victorinus’ own translation. The original version of this work is preserved in a single manuscript: most later users encountered it in the form of an edition made by Jerome (HI Apc), although the biblical text in this version was reworked by a series of later revisers.40 Victorinus also wrote a commentary on Matthew, which is now lost. His gospel quotations exhibit frequent harmonization.41

Early Translations of Other Works
The date of the translation of Irenaeus’ Adversus Haereses (IR), written in Greek around 180, is unknown: it may be third-century, or from around the end of the fourth.42 The biblical quotations in this writing are often of textual interest. It is not clear whether they were translated directly or whether reference was made to an existing Latin version. The date of translation is also unknown for the Latin versions of Clement’s First Epistle to the Corinthians (CLE-R) and the ‘Vulgate’ version of the Shepherd of Hermas (HER V). It has even been suggested that these predate Tertullian.43 Along with the translation of the Epistle of Barnabas (BAR) produced in Africa before Cyprian, which appears to rely on an existing translation of the Latin Bible, they offer interesting comparisons for the translation technique and vocabulary of the Latin Bible.44

The Muratorian Fragment (or Muratorian Canon; AN Mur) has been the focus of a considerable amount of scholarly attention, as one of the earliest surviving lists of canonical books. Preserved in a manuscript of the seventh or eighth century (Milan, Biblioteca Ambrosiana, Cod. J 101 sup.), it was rediscovered by Muratori in 1740. The fragment mentions four Gospels, the Acts of (p.17) the Apostles (attributed to Luke), thirteen letters of Paul, one of Jude and two of John, and the Apocalypses of John (i.e. Revelation) and Peter. The status of the latter is said to be dubious; there is no reference to the epistle to the Hebrews, and letters of Paul to the Laodiceans and Alexandrians are dismissed as forgeries. The order of the Pauline correspondence is unusual (Corinthians, Ephesians, Philippians, Colossians, Galatians, Thessalonians, Romans) and it is not clear whether the Catholic Epistles precede Acts or the Apocalypse. The dating of the text hinges on the reference to the Shepherd of Hermas at the end of the list:

Pastorem uero nuperrime temporibus nostris in urbe Roma Hermas conscripsit sedente cathedra urbis Romae ecclesiae Pio episcopo fratre eius et ideo legi eum quidem oportet se publicare uero in ecclesia populo neque inter prophetas completo numero neque inter apostolos in finem temporum potest.45

Most recently, in our own times in the city of Rome, Hermas wrote The Shepherd while his brother bishop Pius was occupying the episcopal seat of the church of the city of Rome. Therefore he should indeed be read, but he can neither be read aloud to the people in church among the prophets, whose number is complete, nor among the apostles, as he is after their times.

The death of Pius I in 157 means that the earliest possible date for the list is around the year 170, although most scholars hesitate to take this at face value.46 Armstrong suggested that the Muratorian Canon could have come from the prologue to Victorinus of Poetovio’s lost Commentary on Matthew, but Guignard’s recent study offers compelling reasons to identify it as a fourth-century Latin translation of a Greek original.47 The order of the Gospels, with Luke and John at the end, is another argument in favour of its origin in Greek tradition.

In conclusion, the adoption of Latin in the early Church was a gradual development, lasting at least a century. The origins of the translation of the New Testament are obscured by the continuing use of Greek texts by authors familiar with both languages who made ad hoc translations of their biblical quotations into Latin. Even so, a Latin translation of most if not all books of the Bible, probably made in the first half of the third century, was used by Cyprian in North Africa. What is more, the surviving evidence for each book points to a single original translation which was subsequently revised in different ways on numerous occasions. This accounts both for peculiarities shared across the whole of Latin tradition and the diversity of texts arising from internal revision or comparison with Greek. Examples of such revision (p.18) are already attested by the time of Cyprian, along with paratextual material to aid readers. Claims of a distinctive form of language in the early Church, often described as ‘Christian Latin’, are difficult to sustain. Nevertheless, the early biblical translations, including features from an initial period of experimentation, exert a strong and lasting influence on most Christian writing in Latin throughout its history.