We use cookies to enhance your experience on our website. By continuing to use our website, you are agreeing to our use of cookies. You can change your cookie settings at any time. Find out more
Select Bible Use this Lookup to open a specific Bible and passage. Start here to select a Bible.
Make selected Bible the default for Lookup tool.
Book: Ch.V. Select book from A-Z list, enter chapter and verse number, and click "Go."
  • Previous Result
  • Results
  • Look It Up Highlight any word or phrase, then click the button to begin a new search.
  • Highlight On / Off
  • Next Result

The Oxford Handbook of Biblical Studies Provides a comprehensive survey of Biblical scholarship in a variety of disciplines.

Textual Transmission and Versions of the New Testament

J. N. Birdsall

No doubts may be entertained that the documents gathered into the collection which we call the scriptures of the New Testament are from the earliest years of the Christian movement, and most if not all from the first century of the Christian era. Traces of those documents are encountered in manuscript remains of the second century, in identifiable allusions and quotations by writers of the second century and increasingly of later centuries, and in early translations traceable to the second and third centuries (Latin, Syriac, and Coptic). If we take a late second-century writer, Irenaeus of Lyons, as a fixed point, we shall find in his work the main outlines of the later canon: that is, the accepted list of authoritative books. In verbatim quotation we shall find considerable evidence of the wording known uniformly at that point in subsequent years, but also of a number of variations of wording from standard editions which a proportion of later manuscripts present. His evidence is preceded historically by a more slender body of textual material and allusive reference in earlier second-century Christian writers. But from before about 150 CE we have little direct evidence of the spread or use of the New Testament books, and it is largely fragmentary or conveyed by implication.

To trace the earliest stages of the transmission of the New Testament documents, then, is a species of prehistory or archaeology. We need to use data largely from the documents themselves, and the specimens will be items to which attention has been drawn by those who seek to trace the growth and composition of the canonical books and to comment upon their meaning. These specimens will often not be patient of being dealt with by strictly text-critical methods, but will be simply interpreted as indications of what might be revealed if a significant sample of extensive text from this early period (say 50 to 150 CE) became available.

The scenarios differ according to the parts of the canon under investigation. It has frequently been attempted to link the composition of each of the gospels with a specific locality within which that gospel was current from its earliest period of existence. This hypothesis has carried the implication that for some unspecified period each gospel was the sole literary source of knowledge of Jesus in the church for which it was initially written or in the adjacent region. Recently voices have been heard urging that early circulation of each gospel within the whole Christian body should be envisaged, so that there would have been earlier interaction of written materials within the whole Christian body.

Either construct might provide an explanation of variants in the later text when the notion and entity of the ‘fourfold gospel’ made its appearance. Since the former model has been the background of discussion whenever this period and its data have been debated, we confine ourselves to the newly proposed possibility of an early period when gospels were in closer contact than we have hitherto assumed. For my own part, I should indicate that I consider the data to show that St John's Gospel maintained its isolation for a longer period than was the case with the three ‘synoptic gospels’. This would not have produced different effects in terms of textual criticism, but the time occupied by the creation of variants would have been longer in the case of John than of the others. This would be one factor in the contrasts between that gospel and the others.

Two or three phenomena immediately suggest themselves as more readily explained within a setting of fairly early parallel existence of three or four gospels. First, in several early quotations of or allusions to the teaching of Jesus in early Christian writings, such as the apostolic fathers and others, the text of sayings is found to present harmonization of parallel passages. Existence of separate documents side by side shortly after their composition would give ready occasion for this to come about by accident or design, memory playing its part. Second, in the same setting, we find centos made up of excerpts from different gospels as well as from sources not later included in the canonical four. Such could be created easily in a setting where, in addition to oral traditions, the written gospels had already spread widely. Third, it is sometimes the case that sayings or narrative statements in which literal rendering from a Semitic source has been perceived have a variant more in line with Greek idiom or style. These variants may exist side by side in documents which appear from later evidence to be contemporary in their earliest attestation. The background to this is more easily envisaged in a setting of multiplicity of gospel copies, where one strain would retain the Semitizing form (probably nearer to the original) while another would bear the mark of early correction.

Fourth, we can find in all gospels variations of wording and addition of words which seem to have had only a short early life span as part of a gospel text, although their echoes may ring through many centuries. We should be particularly mindful of these variant readings, since they could well be original to the autograph and later eased away, if there seemed to be some possible lack of accuracy or inclination to inaccurate teaching in them. Perhaps the piercing of the side of Jesus before his expiry known in St Matthew's Gospel in an ancient and influential text family but absent in the rest might be an instance of this. Another ancient variant is an addition regarding the appearance of fire upon the Jordan after the baptism of Jesus. This is known over a wide geographical range, and persists in East and West for many centuries, especially in harmonies of the gospels. The corrupting influence of harmonization could eventually happen whether at first only one gospel was known or when there was early contact between gospels, but its frequent and widespread appearance might more easily stem from very early comparison and exchange of different gospels. In Mark 10: 11, 12, we find in a few witnesses (Greek, Syriac, and Georgian) a form of prohibition of divorce in which the clause regarding the woman precedes that regarding the man. This appears to lie behind the allusion to this element of the Lord's teaching cited by Paul in 1 Cor. 7: 10, 11, but has been transformed in the other synoptic gospels to a ‘man first’ order.

We are not dealing with gospels alone. The letters of which most of the rest of the New Testament is composed would have circulated. In the case of Paul's letters, a recent hypothesis proposes that some of them were edited very early to form a corpus (Gamble 1977). We know in any event that letters were sent to churches other than their first, named recipients. Striking evidence here, attested in some manuscripts, is the absence of the phrases ‘in Rome’ and ‘in Ephesus’ in the opening verses of the respective epistles. The manuscripts which attest these are not the same in the two epistles. The many variations of text which often appear to spring from the complex style of the apostle may sometimes have arisen in the different churches which had become the recipients of a collection of his letters. The complexity of some sections is such that one speculates that distinct sets of attempted clarification have become tangled in transmission. We know far less about the letters attributed to other Christian leaders. The prophecy of John, the Apocalypse, has a structure which suggests that it was sent as a missive to a group of churches. The consequent dispersal of seven copies of the original to these recipients could have been the beginning of variation within the text, although there has been no testing yet of such a suggestion.

Acts presents us with very complex variation of its text in the earliest witnesses. Some scholars have stressed that this account of the expansion of the early church is a second volume of which the first is the Gospel according to St Luke. Perhaps the variation within the Gospel should be more closely associated with that of Acts and examined to discover if any tendencies may be observed differing in style by contrast with the other gospels. This was done in the past by Friedrich Blass (1895) and Theodor Zahn (1909). The more recent suggestion of W. E. Strange (1992) that the differences in Acts might be derived from intermediate forms of the work, each emanating from the same author, could be extended to see if it could be used to explain textual variation in Luke. Concentration on the Codex Bezae as representative of one strain of the text of Acts is misleading. That manuscript is of importance, but is idiosyncratic in a number of ways. Complex though the evidence is, we must, in discussing shared variants of the Codex Bezae, lay as much or more stress on Greek minuscule manuscripts, sometimes quite late in date, on the versions, especially Latin, Syriac, and Coptic, and on the quotations of early Latin Fathers and of Irenaeus. We may often find earlier strata there than in Bezae. This is so throughout the gospels too.

This discussion has taken us willy-nilly into the second century. In the concluding remarks of the preceding paragraph, it also shows an aspect of research on that century: namely, the constant encounter with fragmentary evidence and the necessity to extrapolate from evidence dated in later centuries back to probabilities of the earlier time. We have Greek manuscripts of the second century, but many are very fragmented, and none is full, though stretches of text can be studied. The Greek-writing fathers and other authors quote rarely. Their evidence has shown itself patient of interpretations both affirmatory and sceptical, largely dependent on presuppositions of the investigators and the fashions of their generations. Justin has gospel quotations, with the complicating factor that harmonization of the synoptics often shows itself. If the Gospel of John was known to him or to his predecessors, it was in some remote fashion which did not commend the text to quotation. Two early Christian thinkers and writers, soon rejected as unrepresentative of normative Christian belief, were very active in editions of other parts of the canon: Marcion and Tatian. Marcion produced editions of the Pauline letters and the Gospel of Luke, with severe editing based on his theology. Tatian produced a harmony of the four gospels, which is traceable in Syriac and other languages of Eastern Christianity and in Latin and the later languages of Western Europe. It is still debated whether he composed his work in a Greek form. The influence of Marcion and Tatian may be observed not only in the traces of the works they edited (i.e. Luke and the Pauline letters in Marcion's case, the four gospels in Tatian's) but in the whole body of biblical manuscripts: there is still ongoing detailed debate and a vast bibliography.

We find ourselves on more secure ground later in the century in the work of Irenaeus, bishop of Lyons. Originally from Smyrna, he was priest and later bishop to the considerable Greek immigrant population in Gaul. His main surviving work is directed against the Gnostic teachers who were active there. This five-book treatise Against Heresies (Adversus Haereses) survives as a whole in Latin, a translation probably made in the third century. Its trustworthiness has been shown from substantial Greek fragments and from a very literal translation of its two last books into Armenian. He quotes richly from the gospels, the Pauline letters, and the book of Revelation, while Acts and the Catholic epistles are far from unrepresented. From Revelation he quotes in accordance with the manuscripts which Joseph Schmid's (1955–6) work has shown to preserve the most reliable form of the text. The affiliation of the text quoted by him from all the rest of the New Testament is often in accord with the text traditionally termed ‘the Western Text’ (Klijn 1949). This term is a misnomer now, but was coined when the main witness to it, Greek witness apart, was the Old Latin version; over two centuries, considerable support in Syriac and Coptic has also been found. The Codex Bezae mentioned above remains the most striking Greek witness for the gospels and Acts. It contains Greek and Latin text on facing pages, and is of late fourth- or early fifth-century origin. It was probably produced in the East, a strong case having been made most recently for Beirut: some weighty opinion nevertheless still maintains Western origin in a Latin-speaking area. The unique readings of the Codex Bezae should not be taken as typical of the ‘Western Text’ without further support.

The first versions or translations were made in the second century. The Latin first appears in the province of Africa, where in the ninth decade of the second century the martyrs of Scillium carried gospels and Pauline letters. Tertullian, writing at the time, appears to know a Latin version, although he often translates for himself. The harmony of Tatian is the probable form in which the gospel reached the Syriac-speaking area (from the Mediterranean coast through Mesopotamia into what is now Iran and Central Asia). The fourfold ‘separated’ gospels in Syriac were probably created from the harmony text following a Greek model. By the time of Ephrem the Syrian (c.280–370) both are in use, but the harmony still dominates. Its traces are still discernible as late as the thirteenth century. In Egypt, Coptic versions seem to have been slower in appearing, perhaps because Christianity established itself first in the urban centres where Greek was spoken.

Moving from the late second century into the third, we enter a period of consolidation in all areas of the life of the church. In the field of text, there appears a text which in recent discussion has been called ‘Alexandrian’ since the main attestation of its use and existence is in the quotations made by Alexandrian and other Egyptian writers. The third-century father Origen is especially significant in this respect. Some extensive gospel texts on papyrus in the Bodmer collection and the great majuscule manuscripts (formerly termed ‘uncials’)—namely, Sinaiticus, Vaticanus, and Ephraimi Rescriptus—attest this text. It shows qualities which derive from the scholarship developed in Alexandria and other centres for the preservation of earlier Greek literature. It reveals a careful inclination to preserve the earliest attainable text even in the face of changes in language and in fashionable standards of style. At the same time, it is on its guard against expansion, innovation, and banality. Over 100 years ago, Westcott and Hort (1881) gave to it the misleading title of ‘Neutral’. This suggests that it has no faults, which is not so. But if they sought to say that as a rule it rejected various kinds of corruption and retained vividness of expression, they stated the truth.

Study of the quotations of Origen at first suggested that he used such a text in Alexandria but a different one after his move to Caesarea c.230 CE. There thus arose, other factors also playing their part, the theory of ‘local texts’. This postulated that the texts of the New Testament in a number of different locations diverged spontaneously. While this may be so where the texts in question were the earliest in the region, e.g. especially translations into another language, it does not apply to the phenomena which the study of Origen brought to the fore. Further study showed that both in Alexandria and in Caesarea two different gospel texts were available to him, and both these texts appear to be the product of deliberate editorial activity. He himself, however, is certainly not to be considered an editor who formed and promulgated such a text. From an assertion in his commentary on Matthew, we are made aware that he would not have dared to deal with the text of scripture in this way.

An important study of Origen's knowledge of and attitude to variations of text was made some years ago by B. M. Metzger (1968). From this, a number of interesting and significant points are to be derived: first, that he knew a number of variant readings in different parts of the New Testament; second, that he always discussed all the variants at a particular place, even if sometimes expressing preference for one or another. On occasion, but by no means regularly, he suggested the influence of scribal error. Third, it is clear from the details which he gives of the distribution of variants in his day, that readings which he knew may now be unknown to us from our data, or that those which he spoke of as the regular text may be exceptional to us or transmitted only in the ancient translations. His critical preference may be for readings in these categories. These things highlight that our data are preserved haphazardly, and our picture of the textual status quo of those early centuries may be thereby severely skewed in places, a fact we easily overlook.

The text identified in some of Origen's works written at Caesarea was initially called (inevitably) the ‘Caesarean text’. But many difficulties have revealed themselves over the years which have eroded the willingness of scholars to agree that here we have a unified text. Rather, we have a process in which the critical principle of choice and amalgamation is taking place over a long period of time. In the ‘Caesarean’ text, readings of the Alexandrian text have a significant place, but readings of the so-called Western text are equally abundant. There is also some independent revision of style. Thus we may say that it is a process, rather than a text, its differing attestations being moments or phases of the process. Some would see in it the signs of a process which eventually, a century or two later, issued in the earliest form of the Byzantine text.

During this period and into the years after Origen's death, stretching from the early third century into the early years of the fourth, we may perceive side by side throughout the Roman world, which was the central theatre of the development of the church, two significant experiences of the church. First, it is a church under intense persecution, the response of the authorities to its increasing growth. Origen died after torture. His successor at Caesarea, Pamphilus, died a martyr. We still possess evidence of his faithful work collating and correcting manuscripts while he was in prison. Colophons contemporary with the work reveal this. Sometimes the later church historian and commentator Eusebius visited Pamphilus and others in prison and shared in this last labour of love. For the second feature to be seen side by side with this is the church's increasing assimilation of learning and its robust encounter with the world, intellectual as well as moral and spiritual. This happened both in the Empire but also in the East beyond its bounds, under Persian or local rulers.

The spread of the faith and its intellectual defence, tested and tried in these two arenas, was the church's own primary task. The response of the wider church to this was that leaders of the churches outside the Empire and their younger scholars came to centres such as Caesarea and Antioch, studying there and assimilating the methods of precision in the biblical pursuits of the day. It is against this background that we may regard the new wave of translations. There was revision of the Latin in the West attested by writers of the third century in Europe, termed thus ‘the European Latin’. There are daughter versions of the earliest Syriac, pre-eminently the Armenian. The Armenian nation was brought into the church by the conversion of its king in the late third or early fourth century. It was given their own alphabet about a century later. This was the work of Mesrop-Mashtotz, and led to the translation of the Bible made from Syriac. In the fifth century, revision was made from Greek manuscripts. It is this later rendering alone that has survived in manuscripts, the earlier being accessible only by reconstruction from quotation and allusion.

Two other Christian cultures developed in the Caucasus, Armenia being a prime point for launching the evangelization. These are the cultures of Georgia and of the Caucasian Albania. Initial testimony was given in Georgia in the fourth century by Nino, an Armenian slave-woman. We have biblical manuscripts, albeit fragmentary, from the fifth century. These, and the witness of the earliest Georgian martyr, the queen Shushanik (martyred 17 October, 476), show the existence in Georgian translation of the gospels, the Pauline epistles, the Psalms, and some other parts of the Old Testament. Liturgy, hagiography, and homiletic materials are also known. This rich flowering of a whole sacred literature may be ascribed conjecturally not simply to translators in the homeland but to Georgian members of the monastic community in Palestine. Armenian sources lie behind much of this, but there was also awareness of the Greek. From this perhaps springs the fact that for the gospels we have two related yet distinct recensions. In our earliest fragments, however, we find the text of some manuscripts moving from blocks of one recension to blocks of the other. In quotations, vocabulary from either source may sometimes be mingled. But for Acts and the epistles, we have only one primary recension. The Revelation of John was not translated until the tenth century.

Until very recently, the language of the Christian Albanians and its alphabet were almost entirely unknown apart from some inscriptions and a late manuscript in Armenian depicting the alphabet. (Their culture was extinguished by waves of Muslim occupation.) But in recent new discoveries at St Catherine's Monastery on Mount Sinai, two palimpsests have come to light, the underwriting of which reveals a lectionary in the Albanian language. When it is more accessible, we shall learn much.

Another ancient language which was provided with an alphabet for biblical translation is Gothic, the oldest Germanic language known to us. It is extant now only in the manuscript remains of the biblical version and a fragmentary commentary on the Gospel of St John. The Goths had entered the Roman Empire by the third century, and Christianity had already been planted amongst them by that time. A Christian Goth named Wulfila was ordained bishop in the fourth century. As part of his ministry, he devised an alphabet and translated the New Testament and some few parts of the Old. The version migrated westwards with its readers, and was linked with the history of both Visigoths and Ostrogoths. Coming into contact with the Latin version, it interacted with it. The basic Greek from which the original translation was made was an early form of the Byzantine text, but its later history in contact with the Latin has made it a version sometimes difficult to interpret.

All this expansion of the church and development of texts and translations was in train during the period of persecutions before the accession of Constantine (308 CE). This led to the end of persecution and to the status of religio licita being granted to the church. The church could now devote yet more energy both to evangelization and to biblical scholarship. The latter also intensified as the quest for the definition of the faith grew, giving rise to dispute and schism. From within the literature generated in this context, we begin to gain a far denser knowledge of the textual form of scripture. We also perceive a new textual form emerging, new vehicles for the use of scripture in Christian worship, and in the manuscripts which survive from the succeeding centuries we find more complex presentations of the scriptural documents, for study or for worship.

The ‘new textual form’ here referred to has already made an appearance in our survey; it is ‘the Byzantine text’. This began to make its appearance in the fourth century. One of its earliest witnesses is the Freer Codex, housed in Washington DC, named from its donor, Charles Freer. It is a fifth-century text of the four gospels (Sanders 1912a, b). Its text of Matthew and Luke is referred to here; its other gospels present texts with other affiliations. Change of text is not uncommon. It shows that different manuscripts were present in the scriptorium from which the manuscript comes, used probably haphazardly as exemplars as the copying proceeded. A number of definite statements about the Byantine text's origin and attestation have been made in the past. But they need to be modified, or perhaps rejected. They include that Lucian, a scholar and martyr of Antioch, was its editor. But secure evidence is lacking, as careful studies by Metzger (1963) and Zuntz (1953) have revealed. Another is that the first Christian writer to make general use of that text was the great bishop and preacher of Antioch and of Constantinople, John Chrysostom. But when we seek to examine this, we find, first, that the text of his homilies is not yet certainly fixed, and so not a sound basis for conclusions; and second, that where soundings have been taken, it is only an approximation of the Byzantine text that is revealed. But of the existence of this text in many manuscripts dating from between the fourth and the fifteenth century, there is no doubt.

We see the Byzantine text appearing in strength in or about the fifth century and becoming dominant in the Greek Empire from about the tenth century. It has been most thoroughly studied by the early twentieth-century scholar Hermann von Soden (1902–13). There were a number of faults in his whole work, which was an edition of the New Testament. This led at first to neglect of his analyses, but subsequently several have proved to have much of value. His work on the Byzantine text is such a case. He found that over the thousand years of its growth and dominance three main varieties can be discerned within it, and some minor ones. A late form was used by the early editors of the printed Greek New Testament in the sixteenth century. It thus dominated devotion and learning for three centuries after that, in the guise of the Greek text called Textus Receptus by one of its publishers. This means ‘accepted text’, and suggested that it was definitive. It is generally rendered ‘Received Text’, not quite correctly. It is still in some quarters defended as original. Examination of the Byzantine text reveals it as an edited text which would have had its beginnings in the period just prior to the ‘Peace of the Church’. Its tendency, when contrasted with other texts of that period, is frequently to prefer polished variations differing in style from the wording of other groups but with little distinction in sense. One factor in this would have been to commend the Christian scriptures to the educated classes.

A useful aid to perceiving this, as I have found, is an edition of the Greek New Testament in that sixteenth-century form edited by the conservative scholar F. H. A. Scrivener in 1881. In this are highlighted points at which the ‘Received Text’ (i.e. the sixteenth-century imprint) was implicitly altered by the Committee which revised the English ‘Authorized Version’ in 1881. These alterations were occasioned by the revisers' acceptance of the text edited by Westcott and Hort, who preferred the ‘Alexandrian’ text. A study of these highlighted readings (i.e. the basis of alteration by the revisers) clearly reveals the main motivations of the creators of the Byzantine text.

Along with those created for such stylistic reasons are other instances: for example, frequent conflations which have the effect of preserving two prior readings, sometimes with tautology. Another feature is the retention or adoption of readings deriving from the Western text. Günther Zuntz has laid stress on the importance of these in his study The Text of the Epistles (1953) (a masterpiece which all should study). He emphasizes that readings shared by Western and Byzantine texts (even when the Western witness is not preserved in Greek, but only in a version) must be ancient, and hence, although often unacceptable, should necessarily be examined in any assessment of the total significance of the textual data. There are further readings preserved which are clearly erroneous. One which I have been able to examine appears to be the debris of a primitive mistake.

At Matt. 13:55, the name Ioannes appears in the list of the brothers of Jesus following Iakobos (that is, in place of forms of Ioseph). This seems to be an error arising from the familiar and frequent collocation elsewhere of ‘James and John’. It is found in various different manuscript texts. Amongst these are the Codex Sinaiticus (linked with Caesarea and the scholar and martyr Pamphilus), the Codex Bezae, full of ancient material but still as enigmatic as ever, about nine other manuscripts with the older form of writing, and a bevy of later ones with important traces of early text. It is not surprising to find attestation too in Origen—namely, in his Commentary on John—and it must antedate the earliest of these. (As a further sidelight on Origen, we should note that this is one quotation of three from this same verse in this commentary. The others read a form of Joseph!) The influence of this weighty tradition for Ioannes must have continued, as, according to von Soden's analyses, the earliest form of the Byzantine text has the reading, although it did not survive to be included in the printed text. With its wide attestation would also harmonize the further fact that the majority of lectionaries read Ioannes here. It is a hard decision whether it is an error of an early scribe or a slip of the author (as the conservative scholar Theodor Zahn (1909) thought), yet it almost ‘made it’ into the ‘Received Text’!

The notion of variation in the text of scripture has caused considerable unease over the past centuries, ever since more and more manuscripts became known, and variant readings of greater and greater moment were revealed. The changes are ascribed by those alarmed at their recognition to deliberate alteration with the intent to dilute the content of the statement of the Christian faith in its fulness. These fears still lie behind the conservative effort to publish a text based not on the earliest manuscripts but on the majority of manuscripts, most of which come from the latest centuries before the invention of printing (Robinson and Pierpont 1991). In the later twentieth century, we have seen another criticism levelled by scholars of radical inclination, who discern in the very texts which modern critical scholarship has established, changes at a very early stage, even before the fourth century CE. These are analysed as showing the traces of constant effort to ensure that no opinion, especially about the natures or person of Jesus Christ, judged erroneous and heretical in the controversies of those days, should have apparent foundation or purchase in the wording of scriptural passages (Ehrman 1993). Any passages falling under this judgement have been changed. While my judgement is that the classical statement of the case is an overstatement, this factor was certainly present, as has long been known to at least some textual critics, myself amongst them.

What, in conclusion, do we make of the situation from our survey of the data? Has scripture been handed down by ‘bogeymen’ intent on ‘corruption’ (a technical term much misunderstood and recently intentionally misused)? This is not my perception. The transmission of scripture has been the work of those who were primarily Christians, and must be seen in the context of their concern for the clear and true statement of the faith. They did their best to preserve the text, but saw the task in part as ensuring too that its message was comprehended. They could also make mistakes, and these were not always completely removed. It is the effects of the desire for absolute clarity, and especially for expressions which left no room for erroneous interpretation, considered useful in their day, which are today seen by a majority as hindrances to our perception of the pristine documents. However, many variant readings are early, some being errors, but others indicative of early tradition or exegesis. There is much valuable data latent in the critical apparatus as well as in the text of our modern critical editions.

  • Previous Result
  • Results
  • Look It Up Highlight any word or phrase, then click the button to begin a new search.
  • Highlight On / Off
  • Next Result
Oxford University Press

© 2022. All Rights Reserved. Cookie Policy | Privacy Policy | Legal Notice