Chapter 2: Materials and Methods of Textual Criticism

A Historical Introduction to the New Testament
by Robert M. Grant

Chapter 2: Materials and Methods of Textual Criticism

In the previous chapter we considered the nature and contents of the New Testament as a whole, in order to see why it was that we were treating together these diverse documents produced within the Christian Church of the first century or even century and a half. Now we must go on to consider the documents not as a collection but as documents. Ordinarily most people read these documents in various English translations, and therefore they may sometimes be tempted to forget that they were originally composed not in English but in Greek. To be sure, some scholars have argued that parts, at least, of some of the books were written not in Greek but in Aramaic, and it may be worth while to state briefly why this view, while it may be interesting, can never be convincing. How does one prove that some text is not originally Greek but was translated from another language? (1) First, one must show that the Greek as it now stands is bad Greek. (If one is dealing with a really good translator, one cannot show that he has translated unless he has said so.) But most of the passages treated as bad Greek for this purpose can be shown to be at least acceptable in the Hellenistic Greek of the time. (2) Second, one must show that the Greek passage does not quite make sense. (3) Finally, one must show that the passage if retranslated into the other language does make sense, and that some very simple error could have resulted in the text we now have. This retranslation is harder to make than might be supposed, and where such efforts can be tested the proportion of successful retranslations is rather low; moreover, experts in Aramaic have a tendency to disagree as to what the original was. For these reasons, and because, after all, we do have the Greek text, it seems fitting to deal with it rather than with something else.

But do we have the Greek text? Or what kind of Greek text do we have?

Some early Christian writers were aware of the importance of old manuscripts for the study of the Bible, and copyists during the Middle Ages, both in the East and in the West, made efforts to find ancient models. It cannot be denied, however, that a much more vigorous concern for ancient writings and manuscripts arose at the time of the Renaissance. As far as Christian writings were involved, this concern was first expressed in regard to the works of the early Fathers. Such editions of the Greek New Testament as those of Erasmus (1516) and Robert Ètienne (Stephanus, 1551) were based on the available manuscripts which happened to come from the mediæval Greek Church, and contained a large number of accumulated errors.

In the sixteenth and seventeenth centuries much older manuscripts were discovered. The first of these to be found was the sixth-century Graeco-Latin Codex Bezae, named after the Reformed scholar Theodore Beza, who gave it to the University of Cambridge in 1581. Equally important was the Codex Alexandrinus, now in the British Museum, which in 1628 was sent to the King of England by Cyril Lucar, orthodox patriarch of Constantinople (formerly of Alexandria); he was grateful for English diplomatic assistance against Jesuit intrigues. This fifth-century manuscript contains the whole Bible in Greek, in addition to most of the two letters traditionally ascribed to Clement of Rome. Although the letters were published at Oxford in 1633, the New Testament did not appear until 1786, though scholars earlier made use of the manuscript.

Two more early manuscripts were discovered in the eighteenth and nineteenth centuries. The first of these contains Greek writings of the Syrian Father Ephraem; underneath them can be made out, with considerable difficulty, an incomplete Bible of the fifth century. This manuscript, the Codex Ephraemi Rescriptus (‘of Ephraem, written over’), is in the Bibliothèque Nationale at Paris. A later discovery was that of the highly important Codex Vaticanus (fourth century). In Napoleon’s time the French removed manuscripts from the Vatican Library and among them was this codex, which their scholars found to contain a text remarkably free from later additions. The manuscript was later returned to Rome, where it now is; it is still one of our most important witnesses to the early text.

Later in the nineteenth century, enthusiasm for antiquities led a German scholar named Constantine Tischendorf to search for manuscripts in the Convent of St. Catherine on Mount Sinai. There, in 1844, he found forty-three leaves of a fourth-century manuscript. Since the monks had been about to throw it away, they were willing to give it to him. Again in 1853, and once more in 1859, he returned to search for the rest of the manuscript, but without success. Just before he left after his last visit, the steward of the convent finally showed him the missing leaves; since they included the long-lost Greek text of the epistle of Barnabas, Tischendorf spent the night copying it. The manuscript finally reached the Imperial Library at St. Petersburg (now Leningrad), after a long series of legal disputes over its ownership. Called Codex Sinaiticus from its place of origin, it was sold in 1933 by the Russian Government to the British Museum for £ 100,000.

More manuscripts, of course, have been found along the way, but these five are probably the most important. They moved the clock back nearly a thousand years, and showed what the New Testament was like in the fourth and fifth centuries -- and even earlier, since they were obviously copied from still older originals.

What of the period before the fourth century? We should not expect to find manuscripts, or fragments, from the first century. First, there were probably very few of them. Second, the originals were probably worn out after repeated reading both private and public. Only in legend were the ‘authentic originals’ preserved The case is different for the second and third centuries, however. From the relatively dry rubbish heaps of ancient Egypt have come many fragments of New Testament books -- a few from the second century, and a considerable number from the third. Though interest in papyri arose as early as the eighteenth century, it was not until the end of the nineteenth that systematic investigations of sites began, especially at Oxyrhynchus, from which more than twenty volumes of papyri have been published.

The oldest papyrus fragment of any New Testament book is a scrap, about two inches square, which contains verses from the eighteenth chapter of John on both sides. The dates of non-dated papyri can be determined within a margin of about fifty years by comparative study of the styles of writing used in them. This papyrus scrap, now in the John Rylands Library at Manchester, has been assigned to the first half of the second century, perhaps earlier rather than later in the period. By filling in gaps at both ends of the lines, the length of the lines can be calculated; then by filling in gaps between the verses on the front and those on the back we can determine the number of lines to the page; and finally we can estimate the size of the little codex which contained the gospel. Probably the codex contained this book alone; a larger work would have been difficult to handle, given the size of the pages. Thus we see that early in the second century John was valued in Egypt, probably in upper Egypt; indeed, it may have been so highly valued that it was circulated apart from the other gospels.

From the third century come two highly important collections named after the modern Maecenases who purchased them from dealers. The first is the Chester Beatty papyri, containing nearly all of the New Testament except for some Pauline and deutero-Pauline letters and the Catholic epistles. The second is the Bodmer papyri, still in course of publication, including at least the Gospels of Luke and John, as well as I-II Peter and Jude and a number of apocryphal and patristic writings.

In addition to Greek manuscripts, there are other materials which can be used for the reconstruction of early New Testament texts. There are early versions of the New Testament books, especially in Latin, Coptic, Syriac and Armenian; manuscripts of some books in such versions come from as early as the third century. There are also quotations from the New Testament provided by the writers of the early Church, and though the manuscripts of the patristic writings are often late the quotations they give were not often altered by copyists. Sometimes, indeed, we can get back to a very early period in dealing with these quotations. Thus there is a papyrus scrap of the third book of Irenaeus, ‘Against Heresies’, which contains New Testament quotations. Irenaeus wrote about 180, and the scrap comes from the end of the second century or the beginning of the third. One could hardly get closer.

Finally, there are the lectionaries of various churches, containing excerpts from the New Testament arranged for liturgical reading. Though the lectionaries themselves are late, they Sometimes reflect early texts.

What we have endeavoured to show in dealing with the materials is that we do not lack an abundance of manuscripts and other relevant data by means of which we can get back to a period quite early in the history of New Testament transmission. Before the discovery of the Dead Sea Scrolls there was a considerable gap in the transmission of the Old Testament. Such a gap has not existed for some time in New Testament studies. Indeed, a basic difficulty in these studies is not that we have too little, but that we have so much that it is very difficult to control. There are about 4,700 New Testament manuscripts and at least 100,000 patristic quotations or allusions.

The methods employed in dealing with these materials are not free from difficulties. (1) It might be supposed that early manuscripts are naturally better than late manuscripts. Such is not necessarily the case, however, for a manuscript may be late in date but go back to an original text which was very early. Occasionally a group of late manuscripts can be traced back to a hypothetical ancestor, now lost, because of identical errors preserved in them. More often, however, various kinds of readings and ‘contaminations’ have come into the late manuscripts and ancestry is now impossible to trace. (2) It might be supposed that early versions would provide valuable evidence. Sometimes they do; but often they have been corrected from various kinds of Greek texts and we cannot definitely ascertain what the version’s original text was. (3) Patristic quotations are not always absolutely reliable: (a) the Church Father may have been quoting not from a text but from memory; (b) he may have used more than one manuscript; (c) his own works may not have been correctly transmitted; study of their manuscript tradition is required.

Beyond the methods employed in dealing with the manuscripts lie the methods used in relation to the errors they contain. Vaganay has analysed errors as unintentional and intentional. The first group includes (1) additions originating because the scribe wrote the same letter, syllable, word or clause twice (dittography); (2) omissions arising because of the presence of the same elements, sometimes because the scribe’s eye skipped a line which ended in the same way as one he had just copied (homoioteleuton); (3) confusions caused by the presence of different vowels or diphthongs which in Greek were pronounced almost identically (e, e, ei, ai, oi, u; also o and ô); (4) confusions of different letters which in ‘uncial’ writing (capital letters) looked much the same (E—C, O—0, r—T); and (5) confusions arising because the earliest manuscripts contained neither word-separation nor punctuation. The second group includes (1) corrections intended to improve spelling, grammar, or style; (2) ‘harmonizations’ between (a) parallel passages, (b) New Testament citations and Old Testament texts, and (c) New Testament texts and liturgical practice; and (3) exegetical-doctrinal interpolations, suppressions, and tendentious revisions. Exegetical-doctrinal modifications are not very common, but they do exist.

There are a few important passages in the New Testament in which it can be proved conclusively that textual alteration has taken place.

(1) The ending of the Gospel of Mark (16:9-20) is no part of what its author originally wrote: (a) Justin alluded to it and Irenaeus quoted from it; it is included in some important uncial manuscripts, mostly ‘Western’. (b) On the other hand, it is absent from the writings of Clement, Origen and Eusebius, and is omitted in Codex Vaticanus and Codex Sinaiticus, as well as in the older Latin and Syriac versions; the Freer manuscript contains a different ending entirely. (c) Therefore, though it was undoubtedly added at an early date, it is not authentic.

(2) The story about a woman ‘taken in adultery’ and forgiven by Jesus does not belong to the Gospel of John. (a) It occurs in the Byzantine text of the gospel, usually as John 7:53-8:1 but sometimes after John 7:36 or 21:24 (in a small group of manuscripts it is found after Luke 21:38). (b) No manuscript before the end of the fourth century contains it; no Church Father, in the same period, refers to it. (c) Therefore it is not authentic.

(3) A more difficult problem occurs in Luke 22:19b-20: (a) All but a few manuscripts include these verses, which are close to what Paul relates about the Last Supper in I Corinthians 11:24-5. (b) In Codex Bezae and the Old Latin version Jesus says simply, ‘This is my body’; there is no reference to what in the longer version is a second cup at the meal. (c) A Eucharist in which the wine preceded the bread seems to be found in the Didache; therefore some scholars have argued that the shorter version of Luke is the authentic one. (d) On the other hand, it may be that a scribe found the mention of two cups embarrassing and therefore deleted the second notice.

Other examples occur in the epistles.

(4) According to most of the uncials and the Fathers, the epistle to the Romans ends with a doxology (16:25-7). (a) But in the Byzantine text this doxology came at the end of the fourteenth chapter; in the third-century Beatty papyri it occurs after Romans 15:23; in some manuscripts (fifth century and later) it is to be found at the end of both the fourteenth and the sixteenth chapters. (b) Marcion omitted it entirely, as did the scribe of one tenth-century uncial -- though he left a space at the end of the fourteenth chapter. Marcion, who rejected the Old Testament, may well have deleted the doxology because it refers to ‘prophetic scripture’. (c) The passage may be an interpolation, though we cannot be absolutely sure.

(5) At the beginning of Ephesians the words ‘in Ephesus’ present a problem. (a) They are to be found in most manuscripts. (b) On the other hand, Marcion said that ‘Ephesians’ was really addressed to the Laodiceans; Origen omitted the words ‘in Ephesus’; they do not occur in the Beatty papyri, in Codex Vaticanus, or in Codex Sinaiticus (though in both these codices a corrector added them). (c) This evidence suggests that the words are not part of the original letter, though we must not suppose that the addresses of letters always became more specific with the passage of time; Origen omitted a mention of Rome in Romans 1:7 and one Greek manuscript leaves it out both there and in 1:15. Each case must be decided on its own merits.

(6) The text of I John has definitely been interpolated. (a) The later manuscripts of the Vulgate read as follows in I John 5:7-8:

There are three which bear witness on earth, the spirit and the water and the blood, and these three are one in Christ Jesus; and there are three who bear witness in heaven, the Father, the Word, and the Spirit, and these three are one.

In these words later Latin theologians found proof that the doctrine of the Trinity, only implicitly present in the New Testament, was actually stated in its text. (b) But all early Greek manuscripts, all early Church Fathers (including Jerome and Augustine), all early versions, and the older manuscripts of the Vulgate, read thus:

There are three which bear witness, the spirit and the water

and the blood, and the three are one.

(c) The ‘heavenly witnesses’ are no part of what John wrote. In cases like these, where the evidence of manuscripts, versions and early quotations is fairly straightforward, it is relatively easy to make decisions about the nature of the original, or more original, text. To be sure, it can still be argued that the additions are valuable for various reasons; but they should not be regarded as part of the earliest New Testament. The statement in I John about the three heavenly witnesses is valuable as an expression of the Church’s faith in the fourth century and later, but it does not come from the author of the epistle.

In most cases, however, the evidence is not so straightforward,, and it is usually necessary to apply some canons of criticism.

Some Principles of Textual Criticism

F. C. Grant has listed three basic principles of textual criticism which deserve further analysis. They are these:

1. No one type of text is infallible, or to be preferred by virtue of its generally superior authority.

2. Each reading must be examined on its merits, and preference must be given to those readings which are demonstrably in the style of the author under consideration.

3. Readings which explain other variants, but are not contrariwise to be explained by the others, merit our preference; but this is a very subtle process, involving intangible elements, and liable to subjective judgement on the part of the critic.

All three principles, indeed, contain a large measure of subjectivity. The first is more valuable negatively than positively; it means basically that all manuscripts and all types of manuscripts may contain errors. The second point introduces literary criticism (see the next chapter) into textual study, and makes us raise the question whether an author always writes in what we may call his style. If not, the principle is not altogether persuasive. The third brings us in the direction of historical criticism (see Chapter iv), and since it is admittedly subjective we need say no more than that the meaning of ‘explain’ is clearer than the means by which the principle is to be employed.

If we try to apply the three principles to a few examples we may be able to see more clearly how they work.

(1) In Mark 1:1 there is a significant variant. Is the ‘gospel’ that ‘of Jesus Christ’ or that ‘of Jesus Christ the Son of God’? The latter reading is found, sometimes with unimportant variations, in most of the early uncial manuscripts and in most of the quotations in the Fathers. The former reading occurs in the Sinaitic (first hand) and Koridethi uncials and in the writings of Origen. (a) The first principle indicates that we cannot immediately decide which reading is correct. (b) The second principle leads us to consider the fact that at high points in Mark’s gospel he speaks of Jesus as the Son of God (the Transfiguration, the trial before Caiaphas, the Crucifixion). Would Mark’s style, then, lead him to employ the expression at the beginning of his book? (c) It is hard to tell which of the two readings explains the other. In the earliest manuscripts words were not separated and ‘sacred names’ were abbreviated. Thus the words ‘of Jesus Christ the Son of God’ would read something like ITXTTTOT; confusion would be almost inevitable. But we cannot tell whether the longer or the shorter form is the original one.

(2) In John 1:18 either ‘the only-begotten God’ or ‘the only-begotten Son’ revealed God. (a) The witnesses to the text disagree, and we cannot give any text type the preference, though in general the earlier ones attest ‘God’. (b) The parallels in John’s own language are hard to assess properly. In John 1:14 we find ‘only-begotten’ (monogenes) without a noun, while in John 3:14 and 16 we hear of the ‘only-begotten Son’. By ‘Son’ John means ‘Son of God’, as the rest of his gospel makes clear. The Logos who became incarnate ‘was with God’ and ‘was God’ (1:1); but Jesus is addressed as God only after his ascension (20:28). The author’s style, on balance, does not seem to allow us to draw a conclusion. (c) We then ask which reading explains the other. Here we note that the earliest witnesses to ‘the only-begotten God’ were Gnostics of the second century. Were they responsible for this reading? Or did they make use of a text older than themselves, retaining ‘God’ though they would perhaps have preferred to read ‘only-begotten’ alone? Since this question cannot be answered, we finally ask whether ‘only-begotten’ by itself would not explain the existence of both ‘only-begotten God’ and ‘only-begotten Son’. Both ‘God’ and ‘Son’ may have been intended to give exegesis of one difficult term; but such a conclusion is purely conjectural.

(3) An example of a reading even more conjectural is provided in the Revised Standard Version at Jude 5. The manuscripts tell us that the people were saved out of Egypt by ‘the Lord’ (KC) or by ‘God’ (OC) or by ‘Jesus’ (IC) or by ‘God Christ’ (OC XC). To make a choice is exceedingly difficult. But by applying the third principle the revisers decided to read ‘he who saved the people’, supplying a Greek article (O) in place of any manuscript reading. Decisions will vary on this point; the author very much prefers to read ‘Jesus’.

In view of these examples -- to which many more could be added -- we may wonder whether or not the principles are fully adequate. At the same time, we must recognize that mere antiquity is no adequate indication of the goodness of a particular reading. Early manuscripts may contain multitudes of errors, conscious or unconscious; late manuscripts may preserve readings which seem to be correct.

For this reason, even the discovery of new papyri is not necessarily going to provide a more reliable New Testament text. Perhaps if papyri from the first century should turn up they could be given a considerable measure of confidence. None has turned up, however, and the most important and complete papyri we have come from the third century.

Should we, then, try to do nothing more than trace the history of the varieties of texts from the third century to the tenth or eleventh? This looks like a counsel of despair, and it is not greatly strengthened when it is suggested that the late history of the texts illustrates and illuminates the history of theology. The history of theology is known from the writings of theologians, and New Testament textual variants contribute practically nothing which was not, or could not have been, known independently.

The primary goal of New Testament textual study remains the recovery of what the New Testament writers wrote. We have already suggested that to achieve this goal is wellnigh impossible. Therefore we must be content with what Reinhold Niebuhr and others have called, in other contexts, an ‘impossible possibility’.( To call it an approximation would lessen the measure or paradox in the phrase.) Only a goal of this kind can justify the labours of textual critics and give credit to their achievements and to the distance between what they have achieved and what they have hoped to achieve.

If this, then, is the goal of the textual criticism of the New Testament, we are now able to state what attitude we should take towards the additions in the gospels and the epistles. They are not part of the original text, and they belong to the history of the Church rather than to the New Testament. They have as little, or as much, claim, to present the apostolic witness as does such a work as the Gospel of Thomas. The case is not very different when we consider the conjectural emendations intended to go behind disagreements in the manuscripts we possess. Such emendations obviously belong to the history of New Testament study, and emendations were being made as early as Origen’s time, not to mention that of Marcion.

On the other hand, if we virtuously claim that we are not making any emendations but are simply following what is written, the question of what is written will arise. Are we, so to speak, canonizing a particular manuscript or group of manuscripts? Is there some papyrus or other manuscript which deserves our total allegiance? It would appear that nothing of the sort exists, and that in making decisions about the text, just as in making decisions about the canon, it is still necessary for us to use our minds. Perhaps in consequence of the Fall, human reason has become totally corrupt, but since we are not dogs or cats we must still make use of it.