Chapter 8
The Age of Discoveries

The publication of the Revised New Testament by the two University Presses on 17th May, 1881, was the most sensational event in the annals of publishing. The public curiosity was intense, and the demand for early copies overwhelming. Bribes of as much as £5,000 had been offered for advance copies in vain. The Oxford Press alone sold a million copies on the first day. All day the streets about Paternoster Row were blocked by a stream of wagons carrying them to the railways for distribution. Five days later two Chicago newspapers printed the entire book as a supplement to their ordinary issue, half of the text having been received by telegraph before actual copies were available. A period of lively controversy followed, the new version being bitterly attacked by a few scholars (headed by Dean Burgon) who refused to abandon the 'received' text, maintaining that the authority of the Church outweighed the evidence of ancient manuscripts and the ordinary canons of textual scholarship. These critics had behind them the general unwillingness of the public to accept changes in words so well known and so much loved as those of the Bible, and for some time the Revision had a stormy passage. The verdict of scholars was overwhelmingly in favour of the revised Greek text as against the old 'received' text, and this issue may be taken as decided; but time has rather increased than diminished the weight of criticism of the literary shortcomings of the English. Increased knowledge (due in part to the discoveries of Greek papyri in Egypt) of the Greek in common use in the first century has shown that many of the verbal changes introduced by the Revisers were due to a pedantic application of the principles of classical Greek to a popular language which ignored them. And so opinion settled down to the general conclusions described in the last chapter.

It might have looked, and indeed did look, in 1881, as though the end of a period had been reached. Scholars had a new Greek text, based upon the most ancient authorities in accordance with the best principles of textual scholarship; and English readers had a revised English Bible based upon this Greek text. It looked as if nothing now remained to be done but to digest these results, and that further changes were not to be expected. Yet, as a matter of fact, a new period was just opening which may rightly be called the Age of Discoveries, since the half-century which has followed since 1881 has seen discovery after discovery widening our knowledge of the Bible text and its early history, and testing the results at which the scholars of 1881 had arrived by evidence with which they were totally unacquainted. It is the story of these discoveries which has now to be told -- discoveries which have gone on up to the time of writing, and to which more additions may be made before these pages are finished; and then an attempt will be made to sum up the results.

The discovery of the Codex Sinaiticus in the monastery of St. Catherine at Mount Sinai was the climax of the previous period, and it was from the same spot that the first discoveries of the new period came. First, in 1880, Dr. Rendel Harris found there a Syriac translation of a lost Christian work, the Apology of Aristides, a defence of Christianity addressed to the emperor Antoninus Pius by an Athenian philosopher about A.D. 140, very valuable for the early history and creed of the Christian community. A very curious result followed; for it appeared that the Apology had never really been lost at all, but that the original Greek had been embedded in a Christian romance composed about the seventh century, though hitherto there had been no means of identifying it, nor any reason to suppose that it was any earlier than the work in which it had been incorporated. Thus one more early Christian work was restored to our knowledge.

The next discovery was more distinctly Biblical. Encouraged by Dr. Rendel Harris's success, two Cambridge ladies, Mrs. Lewis and Mrs. Gibson, twin sisters and highly accomplished Orientalists, made another visit to Mount Sinai to search for treasures. Among other manuscripts which they examined, one was a palimpsest; that is, the original writing had been partly washed out, and another text written over it. The lower writing was seen to be a copy of the Gospels in Syriac, which might easily be important because of its evidently early date; so they took photographs of it and brought them home for detailed examination. Then it appeared that they had discovered a real prize; for it turned out to be, not the ordinary Syriac version of the Gospels (known as the Peshitta), but an earlier version, of which only one copy was hitherto known, and that very imperfect. This was a manuscript acquired by the British Museum in 1842, printed and privately circulated by Cureton in 1848, and finally published in 1858. It contained portions of the Gospels of Matthew, Luke and John, but of Mark only the last four verses. The new Sinaitic manuscript had portions of all Gospels, and evidently represented the same version as the Curetonian manuscript, but with considerable variations. Thus, while the Curetonian must originally have contained the disputed ending of Mark (16:9-20), the Sinaitic did not, ending the Gospel with verse 8, as do the Vaticanus and Sinaiticus and the earliest manuscript of the Old Latin version. On the whole, the Sinaitic Syriac appeared to represent the version in a rather earlier form than the Curetonian, besides supplementing many of its lacunae; so that we now had a substantial knowledge of this earliest Syriac translation.

Now this was a discovery of first-class importance; for this Syriac version is one of the earliest translations of the New Testament, having probably been made before the end of the second century, and we now had two witnesses for it, both written in the fifth century or earlier. The translation must have been made from Greek manuscripts in existence in the second century, which thus carries us back to a period long before our earliest Greek manuscripts. Further, it was apparent that this Old Syriac text differed in many details from the type represented by the Vaticanus and Sinaiticus, to which Westcott and Hort had given the name of 'Neutral', and which they maintained to be the best. It differed from it in very much the same way (though not always in the same passages) as did the Latin group which Westcott and Hort called 'Western'. While, therefore, in many passages it reinforced the Neutral text as against the 'received' or Byzantine text, it also gave strong support to those who were disposed to question Westcott and Hort's exclusive trust in the Neutral.

This was the more important, because by this time the centre of controversy with regard to the Bible text had shifted. At first, as has already been described, it was a contest between the adherents of the Byzantine text, which had so long dominated the Christian world, and the earlier types represented by a comparatively small number of authorities. That contest had been quickly decided, in the eyes of scholars, in favour of the earlier types, and the Byzantine text had been accepted as secondary and of relatively late origin. But the argument which ruled out the Byzantine text, viz. that readings characteristic of it were not found in Christian writers before the latter part of the fourth century, could not be used against the type of text which Westcott and Hort had labeled 'Western'. On the contrary, it was evident that all the early Christian writers, with the partial exception of Origen, had used texts which did not conform with the 'Neutral' type. If, therefore, all early non-Neutral readings could be grouped together in a single family, as the Neutral authorities were, there was good ground for claiming that this, rather than the Neutral family, had the support of the earliest patristic evidence. Many prominent scholars were inclined to take this view; and the issue now lay, not between 'Neutral' and 'Byzantine', but between 'Neutral' and 'Western'. Consequently the old Syriac version, which was certainly pre-Byzantine and non-Neutral, was hailed as an ally by the Westerners, and the position of Westcott and Hort, with their almost exclusive dependence on the Vaticanus, was to that extent shaken. As will be seen later, there were considerable qualifications to be made to this argument, but for the moment this was the general impression.

The question of the early Syriac text of the Gospels was complicated by another consideration to which attention must be drawn, both because of its own importance and because of a curious series of discoveries (one very recent) connected with it. It was known from early Christian writers that an Assyrian Christian named Tatian, who was born about A.D. 120, had produced about A.D. 170 a work called the Diatessaron ('Concordance of Four'), which was generally supposed to be a harmony of the four Gospels. The work, however, seemed to be completely lost, and the followers of the German school who about the middle of the nineteenth century maintained that none of our Gospels was written before about A.D. 140, and that therefore they are of little historical authority, denied that it was a concordance of our Gospels at all. So late as 1876 the anonymous author of an able work called Supernatural Religion strenuously maintained this point of view, even going so far as to deny that the work had ever existed; and Bishop Lightfoot, who used his immense learning to defend the authenticity and early date of the Gospels, could only bring arguments from probability against him. Yet all the time the decisive proof was lying, so to speak, under their noses. So far back as 1836 the Fathers of a convent in Venice had printed an Armenian version of a commentary by St. Ephrem (who lived in the fourth century) on this very work, which proved beyond doubt that it was in fact a concordance of the four canonical Gospels. Since, however, Armenian was a language almost unknown to Western scholars, no one took any notice of it; but in 1876, just before the appearance of Supernatural Religion, a Latin translation of this commentary was published by the same Fathers, which ought to have brought it to the notice of Biblical students. Strange to say, neither of the learned controversialists was aware of its existence, and it was not until 1880 that Dr. Ezra Abbot called attention to it. This at once led to further research, and first one and then another copy was discovered of an Arabic translation of the Diatessaron itself, which was published in 1888.

These discoveries finally disposed of any doubt as to what the Diatessaron was, and proved that by about A.D. 170 the four canonical Gospels held an undisputed pre-eminence over all other narratives of our Saviour's life; but they left several questions undetermined. Being in Arabic, it was not always certain what the exact text was from which it was translated; and scholars differed as to the original language of the Diatessaron. Was it originally composed in Greek or in Syriac? Since it unquestionably circulated mainly in the Syriac Church (though it was certainly also translated into Latin), and since the Arabic version was certainly made from the Syriac, many scholars were led to suppose that Syriac was its original language. Others, however, pointing to the fact that its title is Greek, maintained that it was first composed in Greek and then translated, in Tatian's own lifetime, into Syriac. In any case it is probably at least as old as the Old Syriac Gospels represented by the Sinaitic and Curetonian manuscripts, and it was in this form that the Gospels mainly circulated in the early Church in Syria. What its exact text was, and what influence it exerted on the text of the separate Gospels, remains uncertain, for the Arabic translation (which was probably made about A.D. 900) has unquestionably been partly assimilated, as almost invariably happens, to later texts; and it still remains one of the riddles which Bible students have to solve.

Here a most interesting, though rather tantalizing, discovery has to be recorded, which has only been made public within the last few years. Far away, on the banks of the Euphrates, the ruins of a Roman fort were discovered at a place called Dura by some English officers in 1920, just before the withdrawal of the British troops. What first attracted attention was some very remarkable wall-paintings of the first century and somewhat later. Subsequently (since the site fell into the French mandated area) it was systematically investigated by French and American excavators. It was discovered that in the last years of the Roman occupation of the fort (which was taken by the Persians in A.D. 256) the walls had been strengthened by the destruction of a quantity of houses, including a Christian church and a Jewish synagogue, and among the debris were found a number of papyri and vellum manuscripts which had been protected against damp in a manner to be paralleled only in Egypt. Among these, when they were examined at Yale in 1933, was found a vellum fragment of a copy of the Diatessaron in Greek. It is only a small scrap, consisting of some fourteen imperfect lines, containing the narrative of the petition of Joseph of Arimathaea for the body of our Lord, written in a hand of the first half of the third century (it must of course have been earlier than the destruction of A.D. 256). It is a mosaic made up of phrases from all four canonical Gospels, with some editorial adjustments. Small as it is, it throws some light on Tatian's method of compilation, showing that he dealt freely with his materials and did not give a general preference to any one evangelist. But its chief interest is that it shows that the Diatessaron circulated in Greek in a distant corner of Syria, about half a century after its composition. This has some bearing on the problem of its original language. If it had been composed in Syriac, it would naturally have circulated in Syria in that language, and the subsequent translations into Greek and Latin would have been reserved for countries in which those languages predominated; whereas if it were originally composed in Greek, it might have had some circulation in that language, even in Syria, before a Syriac translation was available. The proof is not decisive, since a military and trading station, such as Dura, would have had inhabitants who were not Syrians, and they might have brought a Greek copy of the work with them; but so far as it goes, it adds something to the case of those who advocate a Greek original, and in any case it proves that the Diatessaron existed in Greek before A.D. 250.

For the next discovery we must return from Mesopotamia to our familiar hunting-ground of Egypt. In the winter of 1906 an American gentleman, Mr. Charles L. Freer of Detroit, owner of a world-famous collection of Chinese and Japanese paintings which he has installed in a museum in Washington, was travelling in Egypt, and saw in the possession of a Cairo dealer a group of vellum manuscripts, or portions of manuscripts, of obviously early character and Biblical in their contents. They did not come within the scope of his normal interests as a collector, but he realized the importance of the opportunity and secured them. By so doing, he acquired for the United States one of the earliest copies of the Gospels in Greek that has come down to us.

The collection as a whole comprised four manuscripts, two of the Old Testament and two of the New. First there was a volume containing the books of Deuteronomy and Joshua, written in the sixth, or possibly the late fifth century. The numbering of the leaves shows that it originally contained the earlier books of the Pentateuch, from Genesis to Numbers, as well; and it may have had Judges and Ruth at the end, to complete the Octateuch. The second was a very fragmentary and much-damaged copy of the Psalter, which when found was just a solid lump of vellum that had suffered much from worms and damp. The most delicate skill and patience were necessary for its restoration, and even so, every leaf is imperfect, and of the earlier Psalms very little has survived. From the character of the handwriting, it appears to be of the sixth or seventh century, while the final quire, which no doubt replaces one that had been damaged at an earlier date, is probably of the ninth.

Of the New Testament manuscripts, one was of the Gospels, while the other originally contained the Acts, Catholic Epistles and Pauline Epistles, but the earlier portion, from Acts to Romans, has been wholly lost, and the rest is very imperfect. It is indeed only a collection of fragments, in as bad condition as the Psalter above described. In date it is not earlier than the sixth century, but it has a good type of text of the same character as the Sinaiticus, Alexandrinus and Vaticanus. The Gospel Codex is far more perfect and important. It contain all four Gospels, in an order which was not unusual in the West, viz. Matthew, John, Luke, Mark, which probably reflects the order of their popularity. Its writing is unlike those of the other early uncials, being a small, sloping hand which is rather difficult to date. but may be assigned to the fifth, or possibly the fourth century. The first quire of John is later, having been apparently inserted about the seventh century to replace one that was damaged or defective.

The text of this Washington Codex, as it is called, has some very curious features. It is by no means uniform in character, and must have been copied from several ancestors which did not belong to the same textual group. Thus (to use the classification of Westcott and Hort), Matthew is Syrian (i.e. Byzantine); Mark i-v.30 is Western; the rest of Mark does not conform to any of these groups, but to one which will have to be mentioned later; Luke i-viii.12 is Neutral; the rest of Luke is Syrian; John i-v.12 (the added quire) is Syrian and the rest of John is Neutral. One must therefore, conclude that this Codex was copied from a group of papyrus rolls which differed among themselves in textual character, and that the same exemplar was not even followed through a whole Gospel. This would only have been possible in a library where many copies of the Scriptures were available, and the scribe was not particular as to the text he was copying. This is quite natural when we remember that each book originally circulated in a separate roll, and there are other manuscripts which similarly have a different character in different parts (e.g. the Alexandrinus shows an early form of the Byzantine text in the Gospels, but elsewhere is Neutral); but one does not often see signs of so complicated a parentage as here.

One peculiar feature in the Washington Codex attracted immediate attention. It contains the disputed last twelve verses of Mark, but in the middle of them, after verse 14, is inserted the following additional passage:

And they answered and said, This generation of lawlessness and faithlessness is under Satan, who doth not allow the truth of God to prevail over the unclean things of the spirits. Therefore make manifest thy righteousness. So spake they now to Christ, and Christ said unto them, The term of the years of the dominion of Satan is fulfilled, but other terrible things draw near, and by reason of the sins of them I was delivered over unto death, that they may return to the truth and sin no more; that they may inherit the spiritual and incorruptible glory of righteousness which is in heaven.

The first two sentences were known from a reference in one of the writings of St. Jerome, who says that they were found in some copies of the Gospel, chiefly Greek ones; but the rest was entirely new. No one would suppose the passage to be authentic, but it shows how additions were liable to be made in copies of the Gospels and to obtain some circulation.

The next stage of our textual history affords an interesting example of the results of intensive study completed by happy discovery. While the search for early manuscripts of the Bible was proceeding, with the results already described, scholars had not neglected the study of the later manuscripts, on the chance of some of them having preserved traces of early types of text. So far back as 1877 two Irish scholars, W. H. Ferrar and T. K. Abbott, had published a study of four manuscripts which presented some peculiar features. Three of them were written in Southern Italy in the twelfth or thirteenth century; the fourth, now at Leicester, was written in England in the fifteenth century, but was evidently copied from a parent of the same family as the other three. The group is known as Family 13, from the number given in the lists to the first of them, or as the Ferrar Group, from the scholar who first called attention to them. They were plainly related, having peculiar readings which were not known elsewhere, or were only paralleled in very early manuscripts. A few other manuscripts were afterwards found to have traces of the same type; but it was not clear what significance or importance was to be attached to the peculiarities of a relatively late group, as to which all that could be said was that it seemed to have some affinity with the Old Syriac version. Its most outstanding variant was that it transferred the incident of the woman taken in adultery from St. John's Gospel (to which it certainly does not belong, being quite different in style and language) to a place in St. Luke, after xxi.38.

Another little group of four manuscripts was similarly isolated by Professor Kirsopp Lake, who published an account of them in 1902. It is headed by the manuscript which stands first in the catalogue of minuscules (and was slightly used by Erasmus), and is therefore known as Family 1. It preserves many readings which are found in early manuscripts such as the Vaticanus and Sinaiticus and Codex Bezae, or in the Old Syriac version, and Lake noticed that its peculiarities were most frequent in St. Mark, where it also seemed to show some affinity with Family 13. Once attention had been called to this point, it was observed that in other manuscripts also St. Mark seemed to have been less affected than the other Gospels by the process of revision which produced the Byzantine text. The reason no doubt is that St. Mark, bring the shortest Gospel, and containing less of the teaching of our Lord, was less often copied in early days and consequently less subject to alterations either by copyists or by editors. We have already seen that St. Mark displayed special characteristics in the Washington Codex, and we shall find the same in some other discoveries which have yet to be described.

The next step forward came from an unexpected quarter. In 1906, Professor von Soden, who was engaged in a very comprehensive edition of the Greek New Testament, called attention to a late and uncouth uncial manuscript. (now at Tiflis) which had belonged to a monastery in the Caucasus called Koridethi. It did not seem to be earlier than the ninth century and its scribe evidently knew very little Greek; but perhaps for this very reason he was not likely to make alterations (though he might and did make many mistakes) in the text he was copying, and he had certainly preserved a somewhat unusual text. Von Soden associated it with Codex Bezae, but in this he was certainly wrong; and when it was eventually published in full, in 1913, it was pointed out by Lake and others that at any rate in Mark it had strong affinities with Families 1 and 13. The next stage therefore was to combine this manuscript (to which the letter theta was assigned in the catalogue of uncials) with those two families, and to designate the whole as Family Theta.

The significance of the little Ferrar group was thus evidently growing in importance, but it took on altogether a new aspect when in 1924 Canon Streeter published (in his book The Four Gospels) the results of his researches into it. After emphasizing the connection of this Family Theta with the Old Syriac version (which was a proof of its antiquity, in spite of the relatively late date of the manuscripts preserving it), he established the remarkable fact that the great Christian scholar Origen (who died in A.D. 253) had in his later works, written after his removal from Egypt to Caesarea in A.D. 231, used a text of this type. Caesarea in Palestine was a noted centre of Biblical study, afterwards famous for a library of which the manuscripts used by Origen formed a conspicuous part, which was much used by St. Jerome, and in which there is good reason to believe the Codex Sinaiticus was at an early stage in its history. Streeter's conclusion therefore was that whereas Origen in his earlier works, written in Egypt, had used manuscripts containing the Alexandrian or Neutral type of text, he had found at Caesarea a different type which he had accepted as superior and had thenceforth used, and which for us was represented by Family Theta. He accordingly gave this family a new name, as the 'Caesarean' text, to be equated with Hort's 'Neutral' and 'Western' as a family of first-rate importance.

This seemed to be a neat and compact, as well as important, result whereby the originally insignificant Ferrar Group, of unknown origin, had been transformed into the Caesarean text, backed by the authority of the greatest scholar of the early Church, and comfortably housed in Palestine, in appropriate proximity to the Church of Syria. It was, however, almost immediately disturbed; for Professor Lake showed that, on a closer analysis of the quotations in Origen's writings, it appeared that in his first works produced after his migration to Caesarea he unquestionably used an Alexandrian text, and only later changed over to the Caesarean. There is also some indication (though weak, for lack of sufficient evidence) that in his last work at Alexandria he was using a Caesarean text. While therefore it may be legitimate to label this text as Caesarean, as having been used there by Origen and his disciple Eusebius, the truth may be that Origen brought it with him from Egypt, that he found manuscripts of the Alexandrian type at Caesarea and for a time used them, but then reverted to the Caesarean and thenceforth Adhered to it.

Another useful point made by Streeter was that the Washington Gospels might be added to the growing list of Caesarean authorities in respect of the greater part of St. Mark, the character of which, as indicated above, had hitherto been unidentified. The Georgian version has also been shown to be Caesarean in character; and if, as is probable, the Georgian version was derived from the Syriac, this is further evidence for the connection of this type with Syria. From all these discussions and discoveries one certain fact emerges, that the Caesarean text is now a definitely established entity, the character of which demands the further close study which it will undoubtedly receive from scholars.

NOTE: The "Caesarean text" is by no means a "definitely established entity." See the opinions of more recent scholars given here. --M.D.M.