by FREDERIC G. KENYON
Late Director and Principal Librarian of the British Museum
Index 1. Pg.255.
Index 2. Pg.263.
The object of the present volume is to place at the disposal of students of the Greek Bible the results of the discoveries and researches which have been so plentiful in recent years. The standard works of Swete, Gregory, Scrivener, Hort, Nestle and others are a quarter of a century or more old, and contain no account of the labours of Burkitt, Lake, Streeter, Clark, Rahlfs, von Soden, Ropes and many others, or of discoveries such as the Freer MSS., tne Koridethi Gospels, and especially the Chester Beatty papyri. In the following pages the attempt is made to describe all of these, and to estimate their bearing on the history of the text of the two Testaments, and especially that of the New Testament. I have tried to state the evidence dispassionately, and I hope that the lists of characteristic readings of particular manuscripts and text-families will enable the student to form his own opinion ; but I have not tried to disguise my own judgment, for what it may be worth.
Naturally such a work is dependent throughout on the labours of those who have gone before. I have tried to indicate my principal obligations in the brief bibliographies appended to each chapter, and in references and footnotes throughout the text. If I have had to refer to some publications of my own, it is because I have had the good fortune to be intimately concerned with some of the recent discoveries, and no other scholars have yet dealt with them. I am quite aware that much more has yet to be said about them.
The textual criticism of the Greek Bible is a fascinating study ; much work is still needed on it, and much new material lies ready to the student's hand. I hope there may never be wanting a due supply of young scholars who will carry on the study, in which English scholars have taken so honourable a part in the past, of a Book which has such a special appeal to the English-speaking nations.
F.G.K. November, 1936.
In this second edition a few small errors in the text have been corrected, and some larger corrections and additions are appended on page 254. But there have been no large additions made to the textual material or criticism of the Greek Bible within the last twelve years, and in the main the book remains as first issued.
Textual criticism is a humble handmaid in the great task of Bible study, but its service is indispensable. Its business is to lay the foundations on which the structure of spiritual investigation must be built. For if we are to study any book to advantage, we must first be satisfied that it is authentic, and that we have it approximately in the form in which its author or authors wrote it. It is necessary, in the case of ancient books, to add the qualification " approximately," because of the conditions under which ancient books have come down to us, which make absolute certainty impossible. It is with these conditions, which apply to all ancient books, that textual criticism has to deal. There are special conditions applying to the books of the Bible, which will be explained in the following pages; but the fundamentals are the same for all works of Greek and Roman literature.
The basis of textual criticism, and the root of all the problems with which it has to deal, is the simple fact that before the invention of printing every copy of a book had to be written by hand; whence they are called manuscripts. Owing to the frailties of the human hand and eye and brain, it is impossible to copy large quantities of matter without making mistakes. These mistakes will be repeated by the next scribe who copies this manuscript, with additions of his own, so that as time goes on the text will tend to vary further and further from the true original.
Attempts may be made from time to time to correct them, either by comparison with other copies or by conjecture; but in either case there is no certainty that the corrections will be right or will always be the same. There thus grows up a number of various readings, as they are called, with competing claims which a modern editor has to consider. It is the function of textual criticism to supply the principles to guide him in his choice and to enable him to answer the questions with which he may be challenged, such as these: How early is the testimony at our disposal for the text of the book under consideration? What do we know of the manner in which the book has been handed down to us ? How can we decide between the various readings which the extant manuscripts present to us? What, on the whole, is the condition of the text as we have it ? Can we depend on it as substantially accurate, and go forward with confidence to consider its meaning and the lessons which it has to teach us?
In the following pages the attempt will be made to trace the textual history of the books of the Bible, and especially of the New Testament; to describe the conditions under which they were written and handed down in the earliest centuries; to enumerate the principal manuscripts in which they are contained, and to investigate their character; to set out the chief classes of variations, and to indicate on what lines we can estimate their merits. It is a task essential for the thorough study of the Bible, and for meeting the objections which critics may, and do, bring against it. The questions with which it deals are questions to which any student of the Bible should have an answer. It is the history of the Bible as a book; and if it is in
itself a non-spiritual science, it is the basis on which the spiritual study of the Bible stands. As was said by one of the greatest of textual scholars, Dr. Hort: " In respect of the Bible especially, it remains true that whatever helps our understanding helps also in the long run our praying and our working."
In tracing this history, much will have to be said about " families of manuscripts" and " types of text," and it will be as well at the outset to explain what is meant by such terms. In essence it is quite simple, though in working out it may become very complicated. If in a given manuscript of any work some words are wrongly transcribed, or a passage omitted, every manuscript copied from it, or from copies of it, will have the same mistake or the same omission; and if among the extant manuscripts we find that several have the same important mistake or omission, it is legitimate to argue that they are all descended from the manuscript in which that mistake or omission was first made. That manuscript itself may no longer exist, but its previous existence may be inferred, and its descendants form a group distinct from those which have not this particular defect. There are certain classical authors, the extant manuscripts of which are divided in this way, according as they have or do not have a particular passage, or a particular corruption. Or again (and this applies particularly to the Bible text) at a certain time some scholar or some person in authority, finding discrepancies among the copies accessible to him, may set himself to edit the text so as to put it in what he believes to be its correct form, or, it may be, to revise it so as to improve it or to make it more intelligible or more edifying; and every copy made from his revision, either directly or at some removes, will bear the traces of this revision, and will form what is called a family.
If the author's original manuscript had survived, it would of course be unnecessary to trouble about later and less accurate copies of it, or the work of revising editors; but since in the case of the Bible books, as also of all works of the classical authors and of nearly all mediaeval works, the original autographs and all early copies of them have disappeared, we have to do as best we can with such later copies as have survived. Where (as in the case of most classical authors) those copies are few in number and late in date, it is quite possible that in many passages the truth has survived in none of them, and can only be recovered, if at all, by conjecture; and such restorations can at best be regarded as probabilities, not as certainties. Where (as in the case of the Bible) the extant copies are very numerous, and some of them very early, it is permissible to hope that the true reading is to be found somewhere among them. To find it is the task of textual criticism. It is necessary to begin by examining the extant manuscripts, to see (with the aid of palaeography) which are the oldest and therefore probably (though not necessarily) nearer to the truth, and to ascertain which, by habitually sharing the same readings, show signs of being descended from a common ancestor earlier than themselves. Then it may be possible to determine which manuscript or group or family\of manuscripts is most often right, and then in cases of doubt it will be natural to give the preference to the manuscript or group or family which has shown itself generally superior; though it must be remembered that an authority which is generally right cannot really be expected to be always right. But to follow it may be the best we can do until better evidence comes to light.
We find, therefore, that in the case of all ancient literature the extant manuscripts of any work, if they are sufficiently numerous, do in fact tend to fall into groups, of which some are better and some are worse. Particularly is this the case with the books of the Bible, the extant manuscripts of which are far more numerous than those of any other book. The task of criticism, as will be seen when we come to narrate the history of the text, has been to determine which manuscripts show signs of being related to one another, what groups or families can thus be formed of them, how far back the ancestry of each group can be carried, and which, on the grounds of intrinsic merit, is generally to be regarded as superior. It is, however, a very complicated task, because the lines of descent from the common ancestors seldom or never run clear. Copies made from manuscripts belonging to one family were frequently corrected (or depraved) by comparison with manuscripts belonging to another family, and so the evidence becomes mixed and obscured. It is no light task, therefore, to trace the genealogy of our manuscripts, or to determine the true text of the earliest common ancestor of each group; and behind these group-ancestors we want, if we can, to get back to the common ancestor of them all, the author's original manuscript. The nature of this task will become clearer as the story proceeds.
The first essential for the comprehension of the textual history of the Greek Bible is to realize what Greek books were like at the time when the works comprising the two Testaments were written. This is a subject of which our knowledge has been greatly increased within quite recent times. It has always been known, from references and descriptions in classical literature, that books in Greek and Roman days were written on rolls of papyrus; but until the nineteenth century no actual specimens of such books were known to exist. The reason for the disappearance of old papyrus manuscripts lay simply in the perishable nature of the material. Papyrus becomes brittle with age if dry, and is rotted by damp; consequently in any normal climate papyrus manuscripts could not expect a long life. In one country alone were conditions more favourable to their preservation. In Egypt (south of the Delta) the climate is so dry that manuscripts buried in the soil beyond the limits of the inundation of the Nile may be preserved indefinitely, though becoming very brittle; and from Egypt, within the last few generations, great numbers of papyri have been disinterred, to which we owe most of our knowledge of the form and material character of ancient books.
One question which used to be put by sceptical critics of the Bible has therefore already been answered, namely, Why have we no copies of any of the books that come anywhere near the dates of their supposed composition? The answer used to be that all books before the adoption of vellum as the principal writing material early in the fourth century (as will be described later) had disappeared because the material on which they were written was so perishable. The Bible was only in the same case as all books of Greek and Roman literature, of which likewise no copies had
survived that were earlier than the fourth century. The only difference was in favour of the Bible; for there were more early vellum manuscripts of the Bible than of any other ancient book, and the interval between the date of composition and the earliest extant manuscript was less in the case of the books of the New Testament than in that of any work of classical literature. It is now possible to add that, thanks to the discoveries in Egypt, we now have many examples of books earlier than the fourth century; that among them are several substantial portions of books of the Greek Bible; and that the interval between the date of composition and the earliest extant manuscripts has been notably reduced, particularly by discoveries made within the last five years. These will be described presently.
Papyrus is a material made from the fibrous pith of a water-plant which formerly grew plentifully in the Nile, but is now extinct within the borders of Egypt itself. This pith, being cut in thin strips, was laid down in two layers, in one of which the fibres lay horizontally, and in the other vertically. These layers were fastened together by Nile water, glue and pressure, and were then polished to produce a smooth surface. The side primarily intended for writing was that on which the fibres lay horizontally, technically known as the recto; but it was quite possible to write on the other side, known as the verso. The papyrus was manufactured in sheets of various sizes, the better qualities being those in which the fibres were longest; and these sheets were fastened together side by side to make a roll. The roll might be of any length that was found convenient, and in ancient Egyptian times some rolls of sacred books, made for ritual purposes, extended to a length of 100 feet or more (the longest known is 133 feet long); but for practical purposes such a length was too cumbrous, and for works of Greek literature it would appear that a roll rarely exceeded 35 feet. The height also varied. The greatest height known is 19 inches, but for Greek works I2f inches is an outside measurement, and 9 or 10 inches may be taken as a normal height, while pocket volumes might be as little as 5 inches or even less.
The consideration of length is of some importance. A roll of 32-35 feet would contain, in writing of a medium size, books of the length of one of the longer Gospels, or a single book of Thucydides, but no more. Hence it follows that so long as the papyrus roll was the normal vehicle of literature, each of the Gospels and the Acts must have circulated separately. It was not possible to possess in a single volume all the four Gospels or all the Epistles of St. Paul, still less a complete New Testament. In the earliest days each book had its own separate history, and not every Christian community would have had a complete collection of all that we now know as the canonical books.
The writing on the roll thus formed was arranged in columns, which in the case of prose works were normally about 2½ or 3 inches wide. The margins between columns were normally small (½-¾ inch), allowing little space for annotation; but at the top and bottom they were wider, and here lines accidentally omitted in the text are not infrequently inserted. Wide margins are usually a sign of a sumptuously written MS. At the beginning of the roll a blank space was often left, both to protect the roll and to give the reader something to hold when reading the first column. Titles were usually given at the end of a work, not at the beginning.
The writing was normally on one side of the roll only, that on which the fibres lay horizontally; but occasionally, when matter was plentiful or papyrus scarce, the roll might be written both " within and without " (Ezek. ii, 10). Among the papyri found in Egypt there are also not a few in which the back (verso) has been utilized for a work different from that on the recto. Thus, to give a Biblical example, a copy of the Epistle to the Hebrews was about the beginning of the fourth century written on the back of an Epitome of Livy, probably of the third century. Such books are likely to have been copies for private use rather than for the market, but it is quite probable that in early days copies of the New Testament books were often produced in this way.
There were very few aids to the reader in ancient books. Words were not separated, punctuation was either wholly wanting or very incomplete (though not infrequently reading-marks were inserted by a later hand), paragraph divisions are rare, and in the earliest manuscripts capitals are not used. Occasionally a rough breathing is indicated, chiefly in the case of relative pronouns, where ambiguity might be caused (e.g., to distinguish οὐ from οὖ). Accents are almost unknown. It must consequently have been difficult to find a particular passage when required, and authors must have been tempted to quote from memory rather than take the trouble to hunt through a roll for the words required. Inexactness of quotation, therefore, should not surprise us.
Such was, in brief outline, the character of the papyrus roll, the form in which Greek and Latin books
were produced from the days of the great classical authors to the time at which the books of the New Testament were written. Until quite recently it has been supposed that it continued in full use up to the early years of the fourth century, when it was replaced by the vellum codex, or books in the modern form of leaves and pages. But discoveries in Egypt, especially some of quite recent date, have shown that not later than the early years of the second century the experiment was tried of utilizing the codex form for papyrus. It seems that this, if not actually the invention of the Christian community, was at any rate mainly employed by them; for whereas the roll continues in practically universal use for works of pagan literature all through the second and third centuries, the majority of Christian works are in codex form. The earliest examples known can be assigned with some confidence to the first half of the second century, and there are quite a number of the third, so that we are justified in concluding that it was a form in normal use.
A papyrus codex was formed by taking a number of sheets of papyrus, twice the size- of the page required, and folding them once in the middle, thus producing a quire or gathering, as it is called. The size of this quire would depend on the number of sheets thus treated at once. It was possible (and instances exist) thus to treat a single sheet, forming a quire of two leaves or four pages, like an ordinary double sheet of writing paper, and the codex would then be formed of a number of such two-leaf quires fastened together (as in a modern book) by threads through their inner margins. Or, to go to the opposite extreme, all the sheets estimated to be required for the work to be copied might be laid in a single pile and folded, thus producing a codex in a single enormous quire. Of this form (cumbrous as it seems) several examples are known, the number of sheets sometimes amounting to more than fifty and thus producing single-quire codices of over 100 leaves. . It is likely, however, that these were early experiments in the use of the codex form, and before long it was found more convenient to make up quires of 8 or 10 or 12 leaves, such as are found in mediaeval vellum manuscripts and in modern paper printed books.
The advantage of the codex form was that it could include much more matter than the roll, without becoming unduly cumbrous. The earliest papyrus codex known contained the books of Numbers and Deuteronomy, which would have required three rolls. Another, of the early third century, contained all four Gospels and the Acts, which would previously have occupied five separate rolls. Another, of the same date, contained all the Epistles of St. Paul, except (apparently) the Pastorals. It was therefore possible as early as the second century for the four canonical Gospels to be circulated as a single unit, and thus to be marked off from the other narratives of our Lord's life which we know to have been in existence; and we have an actual example of such a volume, as just stated, from the early part of the third century.
It was therefore in the form of the papyrus roll or the papyrus codex that the books of the Greek Bible were written and circulated from the third century B.C., when the Greek translation of the Old Testament was begun, to the end of the third century after Christ. But at the beginning of the fourth century a great
change occurred, which vitally affected the tradition of the sacred books. In the first place, Christianity was first tolerated and then accepted by the Emperor Constantine as the religion of the Roman Empire. Persecutions came to an end, and with them the destruction of Christian books, which had been a marked feature of the great persecutions of Decius (A.D. 249-251) and Diocletian (A.D. 303-305) and probably of the sporadic persecutions which occurred at other times; and a great demand arose for copies of the Scriptures in all parts of the Empire. It was just at this time that a revolution occurred in the world of book-production, by the substitution of vellum for papyrus as the material for the best books.
Vellum is a material produced from the skins of cattle, sheep and goats, and especially from the young of these animals, calves, lambs and kids. The hair is removed by scraping, and the skins washed, scraped with pumice, and dressed with chalk. The material thus produced is almost white in colour, very enduring in quality, easy to write on, and forming a good background both for black ink and for decoration in colour. The side from which the hairs have been removed is apt to be slightly the darker in colour, but it retains the ink better. In the Codex Alexandrinus, for instance, the writing has repeatedly faded on the flesh-side, while on the hair-side it remains clear and distinct. This is the material which from the fourth century to the fifteenth was the dominant material for book-production. Papyrus continued in use, certainly in Egypt and probably elsewhere, for non-literary purposes and for inferior copies of books, until the Arab conquest of the country in A.D. 640 closed its export trade and put an end to the production of Greek books,and at the other end of the Middle Ages paper began to come into use from about the twelfth century onwards; but for the majority of books, and for all the best books, vellum was regularly employed until the invention of printing, for which paper was both better adapted and cheaper.
Vellum was by no means a new material at the beginning of the fourth century. Its invention is ascribed by Pliny, on the authority of Varro, to Eumenes of Pergamum (probably Eumenes II, B.C. 197-159), who was ambitious of founding a library which would rival that of Alexandria. His rival, Ptolemy of Egypt (probably Ptolemy Epiphanes, B.C. 205-182), attempted to obstruct him by prohibiting the export of papyrus; whereupon Eumenes had recourse to the manufacture of vellum, which from its place of origin received the name of περγαμηνή, whence our word parchment, which in general is merely a synonym for vellum, though some writers have restricted it to the less good qualities, used for other purposes than for books. How long the embargo on papyrus lasted we do not know, and there is no evidence that vellum was used for books elsewhere than at Pergamum at this time. Certainly the supremacy of papyrus was not affected at Rome, where, so far as can be gathered from allusions in the classical writers, vellum was only used for note-books and inferior copies of literary works. Towards the end of the first century after Christ, Martial refers to manuscripts of Homer, Virgil, Cicero, Livy and Ovid in membranis, i.e., on vellum, and while their exact character is doubtful it is clear that they are not normal books, and that vellum has not yet superseded papyrus. From about the same period come the earliest extant fragments of vellum codices, a leaf of Demosthenes in the British Museum and one of Euripides at Berlin. But until the end of the third century the supremacy of papyrus remained virtually unchallenged.
In the first quarter of the fourth century, however, the superior advantages of vellum and the perishable nature of papyrus seem suddenly to have been realized. When Constantine ordered fifty copies of the Scriptures for the churches in his new capital, Constantinople, it is expressly stated by Eusebius (vit. Const, iv, 36) that they were on vellum (πεντήκοντα σωμάτα έν διφθέραις). A little later, but apparently before A.D. 350, Jerome records that the volumes in the celebrated library of Pamphilus at Caesarea, which had become damaged, were replaced by copies on vellum (Ep. 141: "quam [bibliothecam] ex parte corruptam Acacius dehinc et Euzoius, eiusdem ecclesiae sacerdotes, in membranis instaurare conati sunt "). This external evidence is confirmed by the extant manuscripts that remain from this period. The two earliest of the vellum codices of the Greek Bible, the Vaticanus and Sinaiticus, belong to just this time, the first half of the fourth century, while the Washington codex of the Gospels and the Sarravianus of the Pentateuch may belong to the second half. There are also manuscripts of Homer, Virgil, and Cicero which may probably be assigned to this century. All these are beautiful examples of the scribe's art, and as specimens of writing, apart from the adventitious aid of decorative illuminations, are as handsome as any
books in the world. On the other hand, the papyrus manuscripts that have survived from the fourth century onwards are inferior examples of book-production, being generally coarse in material and rough in writing. The supremacy of vellum is henceforth unquestionable, and from this time onwards we have an ever-increasing flood of extant copies of the Scriptures, which will be described in the chapters that follow.
E. M. Thompson, Introduction to Greek and Latin Paleography (Oxford, 1912).
V. Gardthausen, Das Buchwesen im Altertum (Leipzig, 1911).
W. Schubart, Das Buck bei den Griechen und Römern (Berlin, 1907).
W. Schubart, Griechische Palaeographie (Handbuch der Altertumswissenschaft, Vol. I, pt. 4, München, I925).
F. G. Kenyon, Books and Readers in Ancient Greece and Rome (Oxford, 1932).