THE TEXT OF THE GREEK BIBLE - A STUDENT'S HANDBOOK by Frederick G Kenyon, late Director & Principal Librarian of the British Museum. First published 1937 New edition 1949 Reprinted 1953. (Prepared for katapi by Paul Ingram 2015).
HOME | contents | << | introduction | textual groups: alpha - α ; beta - β ; gamma - γ ; delta - δ ; [tables:- δ text inclusions-omissions ; β-δ text compared : gospels, acts ] ; epsilon - ε ; the residue | bibliography | addenda |


The materials for textual criticism have now been described, and the course of textual theory and practice narrated, from Erasmus to the present day. It remains to consider what the present position of the subject is and how the theories of critics, from Mill and Bengel to von Soden and Streeter, have been affected by the most recent discoveries. The most convenient method would seem to be to deal in turn with each of the main textual groups, to set out the main authorities composing each of them, and to endeavour to form some estimate of their character, and to consider what progress has been made in elucidating the early history of the New Testament text.

For this purpose it will be best to discard, at any rate at first, the labels that have been attached by critics to the several families, since such labels tend to prejudge their character. After that character has been established it will be easier to see what title is most appropriate in each case. It is proposed therefore to deal with them under the first letters of the Greek alphabet: (α ) the Received Text, of which one of the earliest representatives (in the Gospels) is Codex A; (β ) the text which Hort calls Neutral, headed by Codex B; (γ ) the Caesarean text; (δ ) the Western text headed by Codex D; (ε ) the Syrian text; (ς ) such residue as may be found to be left over.

I.—The α Text.

This is the text found in the great majority of manuscripts, entrenched in print by Erasmus and Stephanus, and known as the Textus Receptus or Received Text, as opposed to the critical editions of modern times. It is Bengel's " Asiatic," Semler's " Oriental," Griesbach's " Constantinopolitan," Hort's " Syrian," von Soden's " K," and it is the text translated in our Authorized Version. Until 1881, in spite of the growing dissatisfaction of scholars, it held the field as the text in practically universal use, and when its position was then decisively challenged, a stiff fight was made in its defence by advocates such as Burgon. Only twenty years ago it was necessary to argue its claims in detail. Now it is possible to deal with them more summarily.

The essence of the case is that this is a text which has suffered progressive revision, not (or only to a very slight extent) on doctrinal grounds, but mainly in the interests of intelligibility and by means of verbal and stylistic alterations and by the assimilation (deliberate or unintentional) of parallel narratives. The proof of its secondary character rests partly on internal evidence, which shows by the application of the ordinary principles of textual criticism that its readings, when compared with those of the earlier authorities, can be explained as scribal or editorial modifications of them, while the opposite explanation is less possible or impossible; and partly on the evidence of the early Fathers, from which it appears that readings characteristic of this type first begin to be noticeable in the writings of Chrysostom. This argument is the foundation-stone of Hort's theory, and recent discoveries have done nothing to disturb it. It may now be taken as an ascertained fact that there is a type of text which begins to make its appearance about the end of the fourth century, that this type acquired predominance in the Church of Constantinople, and that it continued to be the text in general use throughout the Middle Ages, and finally was stereotyped in print.

It is not to be understood that this revision was made as a single deliberate act at a single time, or that it assumed its final form at once. It was rather the result of forces and tendencies which continued to operate over a long period. In part it was due to unconscious tendencies, which lead a scribe to substitute familiar phrases for those less familiar; in part to more deliberate but wholly innocent action in the supposed interests of the reader, taking the form of the insertion of names or pronouns to make a passage more clear, of modifying a passage which appeared open to misunderstanding, of removing apparent contradictions between two evangelists, of assimilating the narrative of one Synoptist to that of another, of the substitution of one word for another, or an alteration in the order of words. All of these are changes which might easily be made at a time when authoritative texts were not to be had, when accurate reproduction of a writer's words was of little account, and when what mattered was that the sacred Scriptures should be readily intelligible.

A few examples will perhaps make it easier to understand the nature of these changes. In Mt. vi.1 the α text has ἐλεημούνην instead of the δικαιοσύνην of the older MSS., substituting a more intelligible word for one which is rather unusual. Similarly in Mk. iii.29 κρίσεως instead of ἁμαρτήματος. In Mt. xv.8 the words ἐγγίζει μου ὁ λαὸς τῷ οτόματι αὐτῶν are added to complete the quotation from the LXX. In Mt. xix.17 the α text has been assimilated to Mk. and Lk., reading τί με λέγεις ἀγαθόν instead of τί με ἐρωτᾶς περὶ τοῦ ἀγαθοῦ of א B D, etc.; and in Mt. xx.22 a similar assimilation to Mk. is made by the introduction of the words καὶ τὸ βάπτισμα ὃ ἐγὼ βαπτίζομαι βαπτισθῆναι. In Lk. xi.2-4 the Lord's Prayer is amplified to correspond with the version in Mt., and the doxology in that version (Mt. vi, 13), which is not in א B D, etc., is probably a liturgical addition. Ecclesiastical usage probably also accounts for τὸ ποτήριον in Lk. xiv.23 instead of the simple ποτήριον. In Mk. i.2 the α text reads ἐν τοῖς προφήταις instead of ἐν τῷ Ἠσαίᾳ τῷ προφήτῃ, an alteration in the interests of accuracy, since the quotation includes words from Malachi as well as Isaiah. In Mk. xv.28 the words καὶ ἐπληρώθη ἡ γραφὴ ... ἐλογίσθη which are not in א A B C D, were probably added with a view to edification; and the omission of οὐδὲ ὁ υἱὸς in Mt. xxiv.36 may be due to a fear of doctrinal misunderstanding. In some cases two various readings in earlier authorities are combined or (in Hort's phrase) conflated. Thus in Mt. x.3 א B read Θαδδαῖος, D k read Λεββαῖος, and the α text has Λεββαῖος ὁ ἐποκληθεὶς Θαδδαῖος; and in Lk. xxiv.53, where א B C have εὐλογοῦντες and D has αἰνοῦντες, the α text has αὀνοῦντες καὶ εὐλογοῦντες.

Often, however, the alterations are slighter and more casual. These are to be found in great quantities throughout the Gospels, and may be illustrated by two passages taken quite at random. Thus in Mt. xxi.12 we find in the α text ὁ Ἰησοῦς for Ἰησοῦς ; ἱερὸν τοῦ Θεοῦ for ἱερὸν ; 13 ἐποιήσατε for ποιεῖτε ; 15 κράζοντας for τοῦς κράζοντας ; 18 πρωΐας δὲ ἐπανάγων for πρωῒ δὲ ἐπαναγαγών ; 23 ἐλθόντι αὐτῷ for ἐλθόντος αὐτοῦ ; 25 παρ' ἑαυτοῖς for ἐν ἑαυτοῖς ; 26 ἔχουσι τὸν Ἰωάννην ὡς προφήτην for ὡς προφήτην ἔχουσι τὸν Ἰωάννην ; 28 καὶ προσελθὼν for προσελθών ἀμπελῶνι ἀμπελῶνι μου for ἀμπελῶνι ; 38 λέγουσιν αὐτῷ for λέγουσιν ; 32 πρὸς ὑμᾶς Ἰωάννης for Ἰωάννης πρὸς ὑμᾶς ; 33 ἄνθρωπός τις for ἄνθρωπός ; 38 κατασχῶμεν for σχῶμεν ; 46 ὡς προφήτην for εἰς προφήτην. Similarly in Lk. xiv.26, οὐ δύναταί μου μαθητὴς εἶναι for οὐ δύναται εἶναι μου μαθητής ; 28 τὰ πρὸς ἀπαρτισμὸν for εἰς ἀπερτισμόν ; 29 ἐμπαίζειν αὐτῷ for αὐτῷ ἐμπαίζειν ; 31 συμβαλεῖν ἑτέρῳ βασιλεῖ for ἑτέρῳ βασιλεῖ συμβαλεῖν ; βουλεύεται for βουλεύσεται ; ἀπαντῆσαι for ὑπαντῆσαι ; 34 καλὸν τὸ ἅλας, ἐὰν δὲ τὸ ἅλας for καλὸν οὖν τὸ ἅλας, ἐὰν δὲ καὶ τὸ ἅλας. Each of these alterations is trifling in itself, but collectively they amount to an extensive modification in the text, and show how freely it was handled by scribes and editors throughout the period when what we now call the Received Text was being developed.

As has already been said, the α text is found in by far the greater number of our extant authorities, but some of the earlier ones show it less fully developed. Thus A and C in the Gospels contain a considerable number of readings of this type. So does W, except in Mk. So do the purple MSS. N O Σ Φ. It is fully established in the later uncials E, F, G, H, K, M, S, U, V, Y, Ω, which Legg classes together as representing this type of text. Of the minuscules all may be assumed to have this text until the contrary has been demonstrated; the most important of the exceptional minuscules will be specified later as belonging to other groups; and even these have not wholly escaped its influence. The only attempt to classify this great mass of materials is that made by von Soden, which has been described above; it will be remembered that his main divisions are K1, of which the earliest representative (Ω) is of the eighth century; Kx, which was dominant from the tenth to the twelfth century; and Kr, which prevailed from the thirteenth century onwards; each of these having many subdivisions.

The relatively late date and secondary character of the a text must now be taken as established. ← Among living scholars, that very strenuous worker, Mr. H. C. Hoskier, maintains the superiority of the α text. He was, however, more concerned to overthrow Hort's claim that the β text had escaped editorial revision; and this, as appears below, may be conceded. There are those, however, who are still uneasy on account of the immense numerical preponderance of the witnesses of this class, which they think must outweigh the small body of dissident testimony. In the past, this argument was accepted as decisive, and, as has been said above, all through the eighteenth and a great part of the nineteenth century the advocates of a revised text could hardly obtain a hearing. Yet in truth the New Testament only reproduces on a larger scale the phenomena which occur in the case of all ancient books. Whenever a classical text has been preserved in any considerable number of manuscripts, the large majority of these are practically ignored by modern editors. There are several hundreds of manuscripts of Virgil; but all editors depend almost wholly on three MSS. of the fourth or fifth century (two of which are seriously imperfect), with three others of equal antiquity which are only substantial fragments, and occasional use of perhaps a dozen others. Of Sophocles there are over a hundred MSS., but only three of these have independent value, and one of these stands out as the predominant authority for the text. For Plato two MSS. stand out so pre-eminently as to make all others unimportant; for Demosthenes, out of a total of some two hundred MSS., one is accepted as conspicuously the best, with occasional help from twelve or fifteen others, which seem to represent different families of less value. For Euripides (the most popular Greek author except Homer) there is no one outstanding authority; but four manuscripts of one family and two of another are accepted as the foundation of his text, and the mass of " Byzantine " manuscripts is ignored. When, therefore, modern editors of the New Testament depend on a relatively small number of (generally) early MSS., and ignore the vast majority of later date, they are doing nothing revolutionary, but are merely conforming to the established principles and practice of textual criticism. The only difference is that they have far more early authorities at their disposal than the editor of any other ancient book.

The a text therefore may be dismissed from further consideration. If we are. to arrive at the original text of the New Testament books, or at any close approximation to it, it must be by the use of the other families which can claim an earlier ancestry. As to its name, it will be best to discard Hort's title of " Syrian," both because of the danger of confusion with " Syriac," and because this type of text, though (since it appears first in the works of Chrysostom) the process of revision which produced it may have begun in Syria, did not have its main circulation in that part of the Christian world. " Arttiochian " is unsatisfactory for the same reason. The preferable title would appear to be " Byzantine," which makes no assertion as to its origin, but merely records the unquestioned fact that it is the text which dominated the whole Church of the Byzantine Empire.

II.—The β Text

Codex Sinaiticus

This is the text called " Alexandrian " by Semler and Griesbach, " Neutral " by Hort, and H (Hesychian) by von Soden. It is the most clearly marked of all, and has been the dominating influence in all modern critical editions. It is headed by the two great uncials, B and א, the former in particular being accepted as the type-specimen of the family. In fact, it may be said that, in the eyes of Hort (its great champion), the β text is the text of B, purged of its scribal errors and amended in a small number of cases (notably in the last chapters of Lk.) where some other authority appears, on grounds of textual criticism, to have preserved a superior reading. The uncials which, in addition to א, most often support it (though none of them has wholly escaped other influences) are L R T Z Ξ in the Gospels, Δ in Mk., and in second line P and Q, also Ψ (in Mk.), and sometimes A and C when these have escaped revision in the direction of the α text. Among the minuscules, the one which has preserved most β readings is 33, and after that 157. Most other cursives which retain a considerable number of early readings are considered by Hort to owe more to Western than to Alexandrian influences, notable among these being 565. With these Hort also reckons famm. 1 and 13, for which another designation has now been found. Of the versions the most prominent supporters of the β text are the Coptic, Boh. and Sah., the latter more fully so than was known to be the case in Hort's time (see p. 131); while the Aethiopic, though a mixed text, with influences of all kinds, retains many signs of its neighbourhood to Egypt. Jerome in the Vulgate relied largely on MSS. of this type; but the fundamental text on which his revision was based, and which he refrained from altering in minor details, was a late form of the Old Latin, i.e., a δ text already modified in the direction of the α type. Of the Fathers, Origen is the most important witness of the β text, though Clement of Alexandria often has readings of this type. In the Acts and Catholic Epistles A and C join this family, with 81, which Hort regards as the best of all the cursives; and the cursives generally retain more readings of this type than in the Gospels. For the Pauline Epistles the authorities are substantially the same, though it is to be noted that in not a few cases B joins the Western group. Thus in the parts of Romans preserved in P46, B 16 times joins the D F G group, whereas א, A, and C do so only once or twice each; and it is noteworthy that in 12 of these cases it is supported by the papyrus. In the other epistles this distinction is less observable; thus in 1 Corinthians B joins the D F G group again 16 times (10 times with P46), but א does so 10 times (4 times with P46), C 9 times (twice with P46), and A 7 times (once with P46); in 2 Corinthians the figures are B 10 (7), א 8 (6), C 8 (3), A 2 (0); in Ephesians B 5 (2), א 3 (0), C 0, A 9 (0); in Galatians B 7 (4), א 1 (0), C 1 (1), A 0; in Philippians B 4 (2), א 3 (0), C 1 (0), A 2 (1); in Colossians B 4 (2), א 0, C 1 (0), A 0. In general, the four uncials are nearer to the Western text in Romans than in the other epistles, though even there the differences (where the two texts are definitely distinguishable) exceed the agreements in the proportion of about 9 to 5. In the Apocalypse, where B is wanting, the B-text is represented by א A C P2, though the differences among them are greater than elsewhere; and with them P47 may now be associated.

For the place of origin of this family, its agreement with the Coptic versions points strongly to Egypt, and this is supported by its use by Origen (at any rate in some of his writings) and to some extent by Clement of Alexandria, and substantially by Athanasius, Didymus and Cyril of Alexandria. There is no direct evidence to show where B and א were written, and different views have been held. Hort was inclined to refer both to Rome, but their text is markedly different from that of the Old Latin version and the Latin Fathers, with the exception of Jerome; and his evidence points in the opposite direction, since he avowedly preferred Eastern authorities to Western. Caesarea has also been suggested, especially by Rendel Harris for א. א was certainly at Caesarea about the sixth or seventh century; but the fact that it was then corrected from MSS. at Caesarea is rather a proof that its own text originated elsewhere. There is some palaeographical evidence which points to Egypt, notably the use of the Coptic form of μ in the titles of some of the books in both MSS., and of a peculiar form of ω (with the middle upright excessively prolonged upwards), which is found in א, occasionally in A and B, and also in some Egyptian papyri. Palaeographical evidence is somewhat insecure, since we have no contemporary MSS. known to have been written elsewhere than in Egypt for comparison; but so far as it goes it supports the ascription to Egypt, which is now generally accepted. And if Egypt is the home of these MSS. and this text, one is bound to think especially of Alexandria: for that was the home of scholarship and the centre of the Christian Church, and hardly elsewhere could volumes of such magnificence have been produced. Further confirmation of the Egyptian origin of this family can be derived from the fact that of its principal members after B and א, T is a Graeco-Coptic bilingual, R was found in Egypt and Z has markedly Egyptian forms of letters.

With regard to its character, the general course of criticism and discovery has not maintained Hort's claim to its full extent. Hort held that the text, alone among the competing families, bears no mark of editorial revision, and therefore is entitled to the name of " Neutral." In this he was supported by Bernhard Weiss, who (while rejecting Hort's classification) came to the conclusion that β is the only New Testament MS. that has escaped deliberate revision, and that it alone has preserved the true reading in 280 passages in the Gospels, with a similar superiority in the other books. This conclusion was contested by Salmon and Hoskier in this country, ← G. Salmon, Some Thoughts on the Textual Criticism, of the New Testament (1897); H. C. Hoskier, Codex B and its Allies: a Study and an Indictment (1914). and by continental scholars in general, Hoskier describing it as an Egyptian text and arguing that it shows many signs of editorial revision, while Salmon concludes that it was the text which had the highest authority in Alexandria in the third century, and may have reached that city in the previous one. Continental scholars have tended to carry the matter farther, and to identify the β text not only with Alexandria but specifically with the name of Hesychius, who is known to have produced an edition of the Septuagint in that city.

The claim to uncontaminated descent and freedom from editorial handling cannot be considered quite apart from the question whether the β text may rightly be described as the text of Egypt. If it were the case that this text was universally current in Egypt, it would not be impossible to maintain that in that province, dominated by Alexandria with its scholarly traditions, a substantially true text had been preserved, without need of restoration by the hand of an editor. But the course of discoveries in Egypt does not confirm the theory of universal dominance. Here the evidence even of the smaller papyrus fragments is of value, and it is strongly reinforced by the more substantial Chester Beatty MSS. Although the papyri often offer readings of the β type, they are generally mixed with others that are not of this type, suggesting that the MSS. in general circulation were far from being universally of this class. One or two are quite definitely of the δ family. But if the β text is not the text of Egypt in general, its claim to uncontaminated descent becomes more difficult, since it would have had to descend by a singularly sheltered channel. This is not impossible, particularly in such a home of scholarship as Alexandria, but it is not very probable, especially when it is remembered that each book of the New Testament was originally a separate roll. The character of B is so homogeneous throughout the New Testament (though this cannot be said of the Old Testament) that it would be necessary to suppose that when its text first assumed codex form a complete set of uncontaminated rolls was available for the purpose. This in itself seems to imply the exercise of editorial selection, and the editor who selected the rolls may well have extended his supervision to the text.

The β text, therefore, is now generally regarded as a text produced in Egypt and probably at Alexandria under editorial care, which was so far accepted in the country that it is found in the Coptic versions and in many manuscripts of Egyptian origin, but which was not in universal use, at any rate in the third century. Whether it can be associated with Hesychius is another matter. This theory was propounded by W. Bousset, and has been accepted by von Soden, who on the strength of it uses the letter H to indicate this type of text. There is, however, singularly little evidence that Hesychius ever extended his labours to the New Testament. It rests on a single passage in Jerome's Epistula ad Damasum prefixed to his Vulgate: " Praetermitto eos codices quos a Luciano et Hesychio nuncupatos paucorum hominum adserit perversa contentio: quibus utique nec in veteri instrumento post septuaginta interpretes emendare quid licuit, nec in novo profuit emendasse, cum multarum gentium linguis scriptura ante translata doceat falsa esse quae addita sunt." It is clear therefore that Jerome regarded both editions as being marked by additions to the authentic text, and that he held them to be worthless. Of Lucian's edition, if that be identified, as it is by some, with the earliest form of the α text, the first of these criticisms might be true; but it is difficult to see how, on the strength of this quotation, the name of Hesychius is to be attached to a text which is conspicuously not characterized by additions, and which Jerome valued so highly that he used it as his main authority in revising the text of the Gospels. Further, since Hesychius was martyred in 311, he cannot have been the editor of a text which underlies the Sahidic version and was used by Origen. The β text, or something very like it, must have been in existence long before Hesychius was born; and even the nearest common ancestor of B and א could hardly be brought within the limits of his activity.

If, then, the extreme claims made by Westcott and Hort on behalf of the β text cannot be maintained, it does not follow that the text itself ceases to demand respect. Even if it is an edited text, it may be a well-edited text; and in the case of all ancient literature a well-edited text is the best that we can hope for. The claims of the β text in this respect can best be assessed when the other families, and especially its principal rival, the δ text, have been considered; for there too we may find some diminution of extreme claims.

What name should be assigned to it is not very material. Both Hort's " Neutral " and von Soden's " Hesychian " have been shown to be unsatisfactory, since both beg questions which are far from being decided. Nor will " Egyptian " do, since, as we have seen, it was by no means universally current in Egypt. On the other hand, some indication of its connection with Egypt is desirable; and since its two main representatives are most probably to be assigned to Alexandria, the name " Alexandrian," which Salmon gave to it, is perhaps the best available. As will be seen later, that title is no longer required for the group to which Hort assigned it.

III.—The γ Text

This letter was assigned by Hort to a type of text which he styled " Alexandrian," represented by no one MS. as the β text may be said to be represented by B, or the δ text by D, but comprising a number of readings which are found in MSS. usually associated with B, but are not present in that MS. and cannot be regarded as belonging to the β text. They are in fact non-Neutral Egyptian readings, and are found chiefly in such MSS. as א A C L T Ψ , where these diverge from B, and in the Coptic versions. Now it is certain, as was said above, that many readings were current in Egypt which did not find a place in the text; but the evidence of recent discoveries goes to show, first that they cannot be grouped together as a family, and secondly that they were current in provincial Egypt. There is far less reason to connect them with Alexandria than there is for א B, and they seem to fall into a wider category than the stylistic revision to which Hort assigns his γ group. It is proposed, therefore, to transfer the title " Alexandrian " to the β text, the contents of Hort's γ group to another category to be described below, and its indicating letter to the newly identified Caesarean text.


The genesis of the Caesarean text has been described above (pp. 173-9). It has hitherto been examined almost exclusively in Mk., in which Gospel its principal representatives are P45 W Θ famm. 1 and 13, and certain other minuscules, notably 28, 565, and 700. None of these manuscripts is to be regarded as a pure representative of the family, all except P45 having suffered more or less assimilation to the α type, while P45, which is anterior to the origin of the α text, has a pedigree which has not yet been established. But by taking the non-Byzantine readings in these MSS. it may be possible to go some way towards reconstructing the common text underlying them; and it is on this task that Prof. Lake and his colleagues are now engaged. His results are awaited with interest. Along with the Greek manuscripts, assistance is to be expected from the Syriac, Armenian, and Georgian versions, and of course from the quotations in Origen (in his later works) and Eusebius, from whose use of texts of this type the title of " Caesarean " has been taken for them.

As has been suggested above (p. 178), it now looks as if Caesarea should be regarded as the place of chief utilization rather than the place of origin. Prof. Lake has argued that the very slight evidence derivable from the early books of Origen's Commentary on Matthew, written before he left Egypt, rather favours his having used a Caesarean text then. It remains to be seen whether further research has enabled him to strengthen this proof; but meanwhile the appearance of P45, with a markedly Caesarean text, written probably before Origen's death and possibly even early in his Caesarean period, shows that texts of this type were circulating in Egypt and may quite well have originated there. Where it originated, however, is not very material. What is important is the claim that a distinct type of text has been identified, which may rank as a family alongside those which we know as the α , β , and γ texts.

This claim seems to be well established. The text which can be elicited from the authorities that have been enumerated is distinct from the ft text, and is certainly not to be confused (as it is by von Soden) with the text of which the leading representative is D. Streeter defined it as lying midway between Neutral and Western, and this is a description which would apply to the subsequently discovered P45. It has none of the extravagancies characteristic of D and the Old Latin, though it shares many minor variants with them. In the major variants it generally sides with the β text, but differs too much and too often to be regarded as a member of that family.

In the other Gospels the Caesarean text has not yet been established; and the task will be more difficult, because Mt. and Lk. always suffered more, thorough assimilation to the a text than always happened to Mk. Thus W and Θ fail to be available here, and P45 is practically non-existent in Mt. In Lk., of which a substantial portion of P45 has been preserved, its text is certainly about midway between β and δ texts, but its relation to them is not the same as in Mk., since it shows a higher proportion of agreement with both. It has, however, a large number of singular readings, which may perhaps prove to be Caesarean. Detailed examination of the quotations in Origen will be necessary before anything can be positively affirmed about it.

Whether there was ever a Caesarean text of the other books is quite unknown. W, Θ , fam. 1 and fam. 13 all fail us here. P45 in general agrees in Acts with the β text, but has a fair number of minor variants which are found in the representatives of the δ text (whether they are essentially δ text remains to be considered), and also a considerable number of singular readings. For the Epistles we have no authorities that have been identified as Caesarean.

The Caesarean text, which is the latest addition to our textual apparatus, requires much more study before its character can be regarded as determined. At present it appears to be a text with some claims to scholarship, akin in this respect to the β text, but essentially different from it.

IV.—The δ Text

Codex Bezae

This is Semler's " Occidental," Griesbach and Hort's " Western," and forms a part of von Soden's I; but the exact connotation of the term " Western," which is that commonly in use, varies, and much depends upon it. In general, it denotes a type of text, of which the leading representatives are Codex Bezae and the Old Latin, characterized by very marked and numerous divergences from both the α text and the β text. The term has, however, often been used to cover all readings which are neither α norβ , and to speak of them as if they formed a homogeneous family, the whole weight of which could be thrown into the scale together against the rival families. This was a natural consequence of Hort's division of all authorities into the three classes of Syrian, Neutral (with Alexandrian as a satellite) and Western; for any pre-Syrian reading that did not appear in the Neutral text fell inevitably into the category of Western. This hard and fast demarcation has been very much shaken by discovery and research since Hort's time, and the whole subject of the pre-Syrian non-Neutral texts requires reconsideration in the light of the latest knowledge.

The definition of the δ or Western text is indeed a prime difficulty. When faced by the question, one is tempted to use Jerome's phrase: " Respondeant quibus; tot sunt paene quot codices." In general, the extremer form of divergence is represented by D and the Old Latin; but the Old Latin authorities differ greatly among themselves, and are rarely unanimous in support of any particular reading. As a rule, the older or African form of the version, represented by the fragment k, e and Cyprian, comes nearest to D, and the Italian or European form (a, b, etc.) represents it with varying modifications. The consideration of the problem must, however, begin with the extremer form; for on the determination of its character the whole question depends. It will be convenient to deal with the Gospels first, then with Acts, and finally with the Epistles; for the conditions are different in each case.

In the Gospels there are two phenomena to consider: first a number of whole passages in which the δ text differs from the others, whether by addition or omission, and secondly a large number of verbal variations. Under the first heading, in general the δ text differs from the others on the side of inclusion. There are twenty-four passages in the Gospels where our principal authorities differ in respect of inclusion or omission.
They are as follows, with the principal witnesses on either side (ς standing for the Received Text):

    Inclusion Omission

1. Mt. xvi.2, 3. The sky at evening. CDW, 1, O.L., Pesh. ς אB, 13, O.S., Sah.
2. Mt. xvii.21. " But this kind never goes out except by prayer and fasting." CDW, O.L., Pesh. ς אB, 33, e, O.S., Sah.
3. Mt. xviii.11. " For the Son of man came to save the lost . " DW, O.L., O.S.c, Pesh. ς אB, 1, 13, 33, e, O.S.a, Sah.
4. Mt. xx.16. " For many are called but few are chosen " (from xxii.14). CD, 1, O.L., O.S., Pesh. ς אBL, Sah.
5. Mt. xx.28. Precedence at feast (see p. 90). D, O.L., O.S.c. אB, etc. Sah. Pesh. ς
6. Mt. xxi.44. The Stone that crushes. אBC.1, g1, O.S.c,Sah. Pesh. ς D, 33, O.L., O.S.a.
7. Mt. xxvii.49. The Piercing of the Side. אBC ADW, 1, O.L., O.S.a, Pesh. ς
8. Mk. xvi.3. The Resurrection. Addition in k ; variation in D All other authorities
9. Mk. xvi.9-20. The supplied ending. ACDW (with additional passage), O.L., O.S.c, Sah. (some MSS.), Pesh. ς אB, k, O.S.a, Sah. (some MSS.).
10. Lk. vi.4. Man working on Sabbath (see p. 91). D All other authorities
11. Lk. ix.55. " You do not know what spirit you are of," etc. D (part), O.L., O.S.c, Pesh. ς אABC etc., O.S.a, Sah.
12. Lk. xi.2-4. Lord's Prayer, additional clauses. ACDW, O.L., O.S.c (part), Sah. (part), Pesh. ς. אB, O.S.a
13. Lk. xxii.19, 20. The Last Supper. Nearly all authorities. D, a, e, etc.
14. Lk. xxii.43. The Agony in the Garden. א* D, 1, O.L., O.S.c, Pesh. ς אaABW, 13, i, O.S.a, Sah.
15. Lk. xxiii.34. " Father forgive them " etc. א* AC, 1, O.L.,O.S.c, Pesh. ς אaBDW, a, b, O.S.a, Sah.
16. Lk. xxiii.38. The Writing on the Cross.
א* AD, 1, O.L., Pesh. ς אcaBC, a, O.S., Sah.
17. Lk. xxiii.53. The Stone at the Sepulchre (see p. 92). D, c, Sah. All other authorities.
18. Lk. xxiv.6, 12, 36, 40. The Resurrection. Nearly all authorities. D, O.L. (some MSS.), O.S.
19. Lk. xxiv.51. The Ascension. Nearly all authorities. א*D, O.L. (some MSS.), O.S.a
20. Jn. v.3. The Stirring of the Water. D, 1, O.L., Pesh. ς אABC, O.S., Sah.
21. Jn. v.4. The Angel at the Pool. A, i, O.L., Pesh. ς

אBCDW, O.S., Sah.

22. Jn. vi.56. Addition : "As the Father is in me and I in the Father. Verily, verily I say unto you, if ye receive not the body of the Son of Man as the bread of life, ye have not life in him." D and in part a, ff2.

All other authorities.

23. Jn. vii.53, etc. The Woman in Adultery. D, O.L. (some MSS.) ς אABCW, O.L. (some MSS.) O.S., Sah., Pesh.
24. Jn. xii.8. " The Poor you always have with you." Nearly all authorities. D, O.S.a


It will be seen from this table that, if the passages at the end of Lk. (nos. 18, 19) be excluded, D is on the side of inclusion in 15 cases, and of omission in 6, while Β is in favour of inclusion in 4 and of omission in 18. In the passages at the end of Lk. the position is reversed, Β being for inclusion and D for omission. It is for this reason that Hort here deserts his almost invariable preference for B, labelling these passages " Western non-interpolations " (i.e. Neutral interpolations). Elsewhere the opposite character of the two texts is unmistakable, and the question for critics is whether these passages were in the original text and were struck out by the editor of the β text, or were originally absent and were inserted by the editor or by successive editors of the δ text. For it will be observed that not all these passages are found in all δ authorities. No. 10 is found only in D, no. 8 only in k, ← The addition in k after μνημείου in Mk. xvi.3 is : " Subito autem ad horam tertiam tenebrae diei [sc. diei tenebrae] factae sunt per totum orbem terrae, et descenderunt de caelis angeli, et surgent in claritate uini dï (so MS., qu. surgente ... filio dei) simul ascenderunt cum eo, et continuo lux facta est." Other singular or slenderly supported additions, which illustrate the character of the δ text, are : Mt. iii.16 after ὕδατος, " and a great light shone around " a, g1, Justin, Diat.
xvii.26, after υἱοί , " Simon said unto him, Yea. Jesus said to him, ' Give thou also to him as a stranger, ' " 713, Diat.
xx.33, after ὀφθαλμοί " Quibus dixit Jesus, Creditis posse me hoc facere? Qui responderunt ei, Ita, Domine," c ; + "and we may see thee," OSc.
xxvii.38, after δεξίων and εὐωνύμων the names of the thieves are added, " nomine Zoathan " (or Ioathas) and " nomine Cammatha " (or Maggatras), c, 1.
Mk. xii.22, after ἀπίθανεν, " sine filiis ; cui remanet mulier munda ? " k, c.
xiv.51, before καὶ κρατοῦσιν αὐτόν, " and there came many men," OSs.
Lk. xxiii.5, after ὧδε " et filios nostras et uxores avertit a nobis; non enim baptizantur sicut [et] nos, [nec se mundant], e, c.
xxiii.9, after αὐτῷ " quasi non audiens," c ; " as if he were not there," OSc.
Lk. xxiii.48, after ὑπέστραφον, " Woe to us, what hath befallen us," OS, Aphraates: " dicentes, vae nobis, quae facta sunt nobis propter peccata nostra; adpropinquavit enim desolatio Hierusalem," g1, Diat.
Jn. xi.39, after κύριε, " Why are they taking away the stone ? " OSa.

and W has a singular insertion in no. 9. In nos. 22, 23, D has the support oi only a portion of the O.L. authorities, א almost always sides with B, except in Lk. xxii, xxiii, where the original text includes, but the corrector marks the passages for omission. Sah. is almost always with א B. As between original inclusion or non-inclusion in general, it must be said that on ordinary critical grounds the presumption is in favour of the latter. It is very difficult to understand the mind of an editor who would strike out such passages as these, while it is easy to see why an editor who had these passages before him from some source should think that they deserved to be incorporated in the record. Also the character of the disputed passages must be considered. It will hardly be maintained that the singular additions in D, k, and W and those quoted in the footnote are authentic, and this casts doubt on the other passages where a few other authorities support D.

It is clear that the main core of the δ text lies in the West, in D (which is semi-Latin), the Old Latin version, and some of the Latin Fathers, such as Cyprian. This again tells against the authenticity of the passages in question ; for it is not reasonable to suppose that the original text was preserved in a pure form only in the West and was wholly obliterated in the East, from which it came. The occasional support which the δ text receives from the Old Syriac may be accounted for, as it is by Burkitt, by reference to the influence of Tatian's Diatessaron, if, as is probable, that was produced by him during his residence in Rome, and then taken by him to Syria, where, in its translated form, it first brought the Gospel story to t he Church of Edessa in its own tongue, and naturally liad some influence on the version of the separate Gospels which was subsequently produced there. Then when the revision which produced the α text took place, manyof these alterations were incorporated in it, which accounts for their appearing in the Peshitta and the Received Text.

But besides these major variations, it is necessary to consider the large number of minor divergences which characterize the δ text. These, just because they are individually small and scattered over the whole text, are difficult to present effectively;
but a few specimens may be given as samples of the rest.

  β text δ text

Mt.xviii.20. οὖ γάρ εἰσι δύο ἢ τρεῖς συνηγμένοι εἰς τὸ ἐμὸν ὄνομα ἐκεῖ εἰμὶ ἐν μέσῳ αὐτῶν. οὔκ eἰσιν γὰρ δύο ἢ τρεῖς συνηγμένοι εἰς τὸ ἐμὸν ὄνομα παρ' oἶς οὔκ εἰμι ἐν μέσῳ αὐτῶν (D g1. O.S.s Sah.).
Mt.xxii.13. δήσαντες αὐτοῦ πόδας καὶ χεῖρας ἐκβάλετε αὐτόν. ἄρατε αὐτὸν ποδῶν καὶ χειρῶν καὶ βάλετε αὐτόν (D, O.L., O.S.).
Mt.xxiii.27. οἴτινες ἔξωθεν μὲν φαίνονται ὡραῖοι, ἔσωθεν δὲ γέμουσιν ὀστέων κ.τ.λ. ἔξωθεν ὁ τάφος φαίνεται ὡραῖος, ἔσωθεν δὲ γέμει ὀστέων κ.τ.λ. (D, Diat.).
Mt.xxv.41. τὸ πῦρ τὸ αἰώνιον τὸ ἡτοιμασμένον τῷ διαβόλῳ. τὸ πῦρ τὸ αἰώνιον (τὸ σκότος τὸ ἐξώτερον Justin) ὃ ἡτοίμασεν ὁ πατήρ μου τῷ διαβὀλῳ (D, 1, O.L.)
Mk.viii.26. μηδὲ εἰς τὴν κώμην εἰσελθῃς. ὕπαγε εἰς τὸν οἷκόν σου, καὶ μηδένι εἴπῃς εἰς τὴν κώμην (D, q ; other O.L. MSS. otherwise).
Mk.x.27. παρὰ ἀνθρώπος ἀούνατον ἀλλ' οὐ παρὰ τῷ Θεῷ· πάντα γὰρ δυνατά ἐστιν παρὰ τω Θεῷ. παρὰ ἀνθρώποις ἀδύνατον, παρὰ δὲ τῷ Θεῷ δυνατόν (D, k, a, ff2).
Mk.xi.3. τί ποιεῖτε τοῦτο; τί λύετε τὸν πῶλον, (D, 13, O.L.)
Mk.xiii.2. οὐ μὴ ἀφέθῇ ὧδε λίθος ἐπὶ λίθον, ὃς οὐ μὴ καταλύθῃ. οὐ μὴ ... καταλύθη, καὶ διὰ τριῶν ἡμερῶν ἄλλος ἀναστήσεται ἄνευ χειρῶν (D, O.L.).
Mk.xiv.4. ἦσαν δέ τινες ἀγανακτοῦντες πρὸς ἑαυτούς. oἱ δὲ μαθηταὶ αὐτοῦ διεπονοῦντο καὶ ἔλεγον (D, a, ff2).
Lk.iii.22. ὁ υἱός μου ὁ ἀγαπητός, ἔν σοι ηὐδόκησα. ὁ υἱός μου ὁ ἀγαπητός, ἐγὼ σήμερον γεγέννηκά σε (D, a, b). From Ps. ii.7.
Lk.v.10. ὁμοίως δὲ καὶ Ίάκωβον καὶ ,Ἰωάννην, υἱοὺς Ζεβεδαίου, οἳ ἦσαν κοινωνοὶ τῷ Σίμωνι. καὶ εἶπε πρὸς τὸν Σίμωνα ὁ Ἰησοῦς, μὴ φοβοῦ· ἀπὸ τοῦ νῦν ἀνθρώπους ἔσῃ ζωγρῶν. καὶ καταγαγόντες τὰ πλοῖα ἐπὶ τὴν γῆν, ἀφέντες ἅπαντα ἠκολούθησαν αὐτῷ. ἦσαν δὲ κοινωνοὶ αὐτοῦ Ἰάκωβος καὶ Ἰωάννης, υἱοὶ Ζεβεδαίον· ὁ δὲ εἶπεν αὐτοῖς, δεῦτε καὶ μὴ γίνεσθε ἁλιεῖς ἰχθύων, ποιήσω γὰρ ὑμᾶς ἁλιεῖς ἀνθρώπων, oἱ δὲ ἀκούσαντες πάντα κατέλειψαν ἐπὶ τῆς γῆς, καὶ ἠκολούθησαν αὐτῷ (D, e).
Lk.v.14. eἰς μαρτύριον αὐτοῖς. D adds ὁ εἰς ἐξελθὼν ... Καφαρναούμ, from Mk i.45, ii.1.
Lk.xi.35, 36. σκόπει οὖν μῆ τὸ φῶς ... φωτίζῃ σε. D and O.L. substitute Mt. vi.23b; O.S.c combines Lk. xi.35 and Mt. vi.23b; O.S.s combines Lk. xi.35 and another phrase.
Lk.xi.53, 54. κἀκεῖθεν ἐξελθόντος αὐτοῦ ἤρξαντο oἱ γραμματεῖς καὶ οἱ Φαρισαῖοι δεινῶς ἐνέχειν καὶ ἀποστοματίζειν αὐτὸν περὶ πλειόνων, ἐνεδρεύοντες αὐτόν, θηρεῦσαί τι ἐκ τοῦ στόματος αὐτοῦ. λέγοντος δὲ ταῦτα πρὸς αὐτοὺς ἐνώπιον πάντος τοῦ λαοῦ, ἤρξαντο oἱ Φαρισαῖοι καὶ οἱ νομικοὶ δεινῶς ἔχειν καὶ συμβάλλειν αὐτῷ περὶ πλειόνων, ζητοῦντες ἀφορμήν τινα λαβεῖν αὐτοῦ, ἵνα εὕρωσιν κατηγορῆσαι αὐτοῦ (D,O.L.,O.S., Diat.).
Lk.xxiii.42, 43. καὶ ἔλεγεν, Ἰησοῦ, μνήσθητί μου ὅταν ἔλθῃς ἐν τῇ βασιλεία σου. καὶ εἶπεν αὐτῷ, ἀμὴν λέγω σοι, σήμερον μετ' ἐμοῦ ἔσῃ ἐν τῷ παραδείσῳ. καὶ στραφεὶς πρὸς τὸν Κύριον εἶπεν αὐτῷ, μνήσθητί μου ἐν τῇ ἡμερᾳ τῇς ἐλεύσεώς σου. ἀποκριθεὶς δὲ ὁ Ἰησοῦς εἶπεν αὐτῷ τῷ ἐπιπλήσσοντι, θάρσει, σήμερον ... παραδείσω (D).


These examples, which might be greatly multiplied, may serve to show the character of the discrepancies between the two texts. They are especially numerous in Lk., where, as will be seen from the specimens given, they sometimes amount to paraphrastic versions of considerable passages. This has given rise to the suggestion that Luke himself produced two editions of his Gospel ; but this theory will be better dealt with in connection with Acts, where the same suggestion has been made with greater force. It is clear that editorial revision has been at work on a large scale in one text or the other. The decision must be left to the application of the principles of textual criticism, which has to consider which text shows the stronger signs of authen-ticity ; and if in any considerable number of instances the verdict is in favour of one text, a presumption will be established in its favour in other cases where a decision is more difficult. In this connection it is relevant to notice that some of the variations in the δ text appear to be due to the importation of passages from the other Gospels or the Septuagint, and some to the influence of the Diatessaron ; it is possible also that some were derived from other narratives of our Lord's life, now lost. Also the support of the Old Latin MSS. is by no means uniform, and some additions or variations of the same nature appear in some of the Old Latin authorities but not in D. All this seems to suggest a lax handling of the text in the region in which the δ text had its origin, and seems, in the eyes of most scholars, to establish a presumption in favour of the β text, which, however, need not be decisive in every case.

The δ text in Acts requires separate consideration, and presents a very difficult problem, which is far from having received a final solution. Here the divergences are both more numerous and more substantial. The most notable of these have been quoted above in the description of D (pp. 92-4) ; but they are only a few among many.
Some other examples may be added here:

  β text δ text

i.2. ἀχρὶ ἧς ἡμέρας ἐντείλάμενος τοῖς ἀποστόλοις διὰ Πνεύματος Ἁγίου οὓς ἐξελέξατο ἀνελήμφθη. ἐν ᾗ ἡμέρᾳ τοὺς ἀποστόλους ἐξελέξατο διὰ Πνεύματος Ἁγίου, καὶ ἐκέλευσεν κηρύσσειν τὸ εὐαγγέλιον.
i.5. οὐ μετὰ πολλὰς ταύτας ἡμέρας. ὃ καὶ μέλλετε λαμβάνειν οὐ μετὰ πολλὰς ταύτας ἡμέρας ἕως τῆς πεντηκοστῆς.
ii.30. ἐκ καρπου τῆς ὀσφύος αὐτοῦ καθίσαι ἐπὶ τὸν θρόνον αὐτοῦ. ἐκ καρποῦ τῆς κοιλίας (καρδίας D) αὐτοῆ κατὰ σάρκα ἀναστῆσαι τὸν Χριστόν, καὶ καθίσαι ἐπὶ τὸν θρόνον αὐτοῦ.
iii.3. ὃς ἰδὼν Πέτρον καὶ Ἰωάννην. οὗτος ἀτενίσας τοῖς ὀφθαλμοῖς αὐτοῦ καὶ ἰδὼν Π. καὶ Ἰ.
iv.18. καὶ καλέσαντες αὐτούς. συγκατατιθεμένων δὲ αὐτῶν τῇ γνώμῃ φωνήσαντες αὐτούς.
v.18.   Adds καὶ ἐπορεύθη εἷς ἕκαστος εἰς τὰ ἴδια.
vii.4. Adds καὶ oἱ πατέρες ἡμῶν οἱ πρὸ ἡμῶν.
vii.24. Adds καὶ ἔκρυφεν αὐτὸν ἐν τῇ ἄμμῳ (from LXX).
viii.39. Πνεῦμα Κυρίου ἥρπασε τὸν Φίλιππον. Πνεῦμα Ἅγιον ἐπέπεσεν ἐπὶ τὸν εὐνοῦχον, ἄγγελος δὲ Κυρίου ἥρπασεν τὸν Φίλιππον ἀπ' αὐτοῦ.
ix.4, 5. τί με διώκεις ; εἶπε δέ, τίς εἶ, Κύριε; ὁ δὲ ἐγώ εἰμι Ἰησοῦς ὃν σὺ διώκεις· ἀλλὰ ἀνάστηθι. τί με διώκεις; σκληρόν σοι πρὸς κέντρα λακτίζειν (from xxvi.14). εἶπεν δὲ, τίς eἶ, Κύριε ; καὶ ὁ Κύριος πpὸs αὐτόν, ἐγώ εἰμι Ἰησοῦς ὁ Ναζωραῖος (from xxii.8) ὃν σὺ διώκεις. ὁ δὲ τρέμων καὶ θαμβῶν ἐπὶ τῷ γεγονότι αὐτῷ εἶπεν, Κύριε, τί με θέλεις ποιῆσαι; καὶ ὁ Κύριος πρὸς αὐτόν, ἀνάστηθι.
ix.7, 8.

μηδένα δὲ θεωροῦντες. ἠγέρθη δὲ ὁ Σαῦλος ἀπὸ τῆς γῆς.

μηδένα δὲ θεωροῦντες μεθ' οὗ ἐλάλει. ἐφη δὲ πρὸς αὐτοὺς, ἐγείρατέ με ἀπὸ τῆς γῆς. καὶ ἐγειράντων αὐτῶν.
x.33. ἐξαυτῆς οὖν ἔπεμψα πρός σε. ὃς παραγενόμενος λαλήσει σοι. ἐξαυτῆς οὖν ἔπεμφα πρός σε, παρακαλῶν ἐλθεῖν πρὸς ἡμᾶς.
xi.17. τίς ἤμην δυνατός κωλῦσαι τὸν Θεόν; τίς ... θεὸν τοῦ μὴ δοῦναι αὐτοῖς Πνεῦμα Ἅγιον πιστεύσασιν ἐπ' αὐτῷ;
xi.27, 28. εἰς Ἀντιόχειαν. ἀναστὰς δὲ εἷς ἐξ αὐτῶν, ὀνόματι Ἄγαβος, ἐσήμανε. εἰς Ἀντιόχειαν, ἦν δὲ πολλὴ ἀγαλλίασις. συνεστραμμένων δὲ ἡμῶν ἔφη εἷς ἐξ αὐτῶν, ὀνόματι Ἄγαβος, σημαίνων.
xii.3. ἰδὼν δὲ ὅτι ἀρεστόν ἐστι τοῖς Ἰουδαίοις. καὶ ἰδὼν ὅτι ἀρεστόν ἐστιν τοῖς Ἰουδαίοις ἡ ἐπιχείρησις αὐτοῦ ἐπὶ τοὺς πιστούς.
xiii.43. Adds ἐγένετο δὲ καθ' ὅλης τῆς πόλεως διελθεῖν τὸν λόγον τοῦ Θεοῦ.
xv.5. ἐζανέστησαν δέ τινες τῶν ἀπὸ τῆς αἱρέσεως τῶν Φαρισαίων πεπιστευκότες. οἱ δὲ παραγγείλαντες αὐτοῖς ἀναβαίνειν πρὸς τοὺς πρεσβυτέρους ἐξανέστησαν.
xv.12. ἐσίγησε δέ. συγκατατιθεμένων δὲ τῶν πρεσβυτέρων τοῖς ὑπὸ τοῦ Πέτρου εἰρημένοις, ἐσίγησεν.
xv.41. Adds παραδιδούς τε τὰς ἐντολάς τῶν πρεσβυτέρων.
xvi.4. ὡς δὲ διεπορεύοντο τὰς πόλεις, παρεδίδουν αὐτοῖς φυλάσσειν τὰ δόγματα τὰ κεκριμμένα ὑπὸ τῶν ἐποστόλων καὶ πρεσβυτέρων τῶν ἐν Ἱερουσαλήμ. διερχόμενοι δὲ τὰς πόλεις ἐκήρυσσον αὐτοῖς μετὰ πάσης παρρησίας τὸν Κύριον Ἰησοῦν Χριστόν, ἅμα παραδίδόντες καὶ τὰς ἐντολάς τῶν ἀποστόλων καὶ πρεσβυτέρων τῶν ἐν Ἱεροσολύμοις.
xviii.21 ἀλλὰ ἀποταξάμενος καὶ εἰπών, πάλιν ἀνακάμψω πρὸς ὑμᾶς τοῦ Θεοῦ θέλοντος, ἀνήχθη ἀπὸ τῆς Ἐφέσου. ἀλλὰ ἀπετάξατο αὐτοῖς εἰπών, δεῖ με πάντως τὴν ἑορτὴν ἐρχομένην ποιῆσαι εἰς Ἱεροσόλυμα· ἀνακάμψω πρὸς ὑμᾶς τοῦ Θεοῦ θέλοντας. Some add τὸν δὲ Ἀκύλαν εἴασεν ἐν Ἐφέσῳ· αὐτὸς δὲ ἀναχθεὶς ἧλθεν.
xxi.25. περὶ δὲ τῶν πεπιστευκότων ἐθνῶν ἡμεῖς ἐπεστείλαμεν, κρίναντες φυλάσσεσθαι.
περὶ δὲ τῶν πεπιστευκότων ἐθνῶν οὐδὲν ἐχουσιν λέγειν πρός σε, ἡμεῖς γὰρ ἀπεστείλαμεν κρίναντες μηδὲν τοιοῦτον τηρεῖν αὐτούς, εἰ μῂ φνλάσσεσθαι.

After xxii.28 D is deficient

xxiii.15. νῦν οὖν ὑμεῖς ἐμφανίσατε τῷ χιλιάρχῳ. νῦν οὖν ἐρωτῶμεν ὑμᾶς ἵνα τοῦτο ἡμῖν ποιήσητε· συναγαγόντες τὸ συνέδριον ἐμφανίσατε τῷ χιλιάρχῳ.
xxiii.23, 24. ἑτοιμάσατε . . ἀπὸ τρίτης ὥρας τῆς νυκτός, κτήνη τε παραστῆσαι ἵνα ἐττιβιβάσαντες τὸν Παῦλον διασώσωσι πρὸς Φήλικα τὸν ἡγεμόνα. ἑτοιμάσατε . . καὶ ἐπὸ τρίτης ὥρας τῆς νυκτὸς κελεύει ἑτοιμους εἶναι πορεύεσθαι. καὶ τοῖς ἑκατοντάρχοις παρήγγειλεν κτήνη παραστῆσαι, ἵνα ἐπιβιβάσαντες τὸν Π. διὰ νυκτὸς διασώσωσιν eἰs Καισάρειαν πρὸς Φήλικα τὸν ἡγεμόνα. ἐφοβήθη γὰρ μήποτε ἐξαρπάσαντες αὐτὸν οἱ ,Ἰουδαῖοι ἀποκτείνωσιν, καὶ αὐτὸς μεταξὺ ἔγκλημα ἔχη ὡς εἰληφὼς ἀργύρια.
xxiii.27. ἐξειλόμην, μαθὼν ὅτι Ῥωμαῖός ἐστιν. ἐρυσάμην κράζοντα καὶ λέγοντα εἶναι Ῥωμαῖον.
xxiv.6. Adds καὶ κατὰ τὸν ἡμέτερον νόμον ἠθελήσαμεν κρῖναι. παρελθὲν δὲ Λυσίας ὁ χιλίαρχος μετὰ πολλῆς βίας ἐκ τῶν χειρῶν ἡμῶν ἀττήγαγεν, κελεύσας τοὺς κατηγόρους αὐτοῦ ἔρχεσθαι πρός σε.
xxiv.10. λέγειν. ἀττολογίαν ἔχειν ὑπὲρ ἑαυτοῦ. ὁ δὲ σχῆμα ἔνθεον ἀναλαβὼν ἔφη.
xxiv.24. οὔσῃ Ἰουδαίᾳ μετεπέμψατο. οὔσῃ Ἰουδαίᾳ, ἥτις ἠρώτησεν ἰδεῖν τὸν Παῦλον καὶ ἐκοῦσαι τὸν λόγον. θέλων οὖν χαρίζεσθαι αὐτῇ μετεπἐμψατο.
xxiv.27. θέλων τε χάριτα καταθέσθαι τοῖς Ἰουδαίοις ὁ Φῆλιξ κατέλιπε τὸν Παῦλον δεδεμένον. τὸν δὲ Παῦλον εἴασεν ἐν τηρήσει διὰ Δρούσιλλαν.
xxv.24, 25. βοῶντες μῂ δεῖν ζῆν αὐτὸν μηκέτι. ἐγὼ δὲ κατελαβόμηνμηδὲν ἄξιον θανάτου αὐτὸν πεπραχέναι. αὐτοῦ δὲ τούτου ἐπικαλεσαμένου τὸν Σεβαστόν, ἔκρινα πέμπειν. ὅπως παραβῶ αὐτὸν εἰς βάσανον ἀναπολόγητον. οὐκ ἠδυνήθην δὲ παραδοῦναι αὐτὸν διὰ τὰς ἐντολὰς ἃς ἔχομεν τταρὰ τοῦ Σεβαστοῦ, ἐὰν δὲ τις αὐτοῦ κατηγορεῖν θέλῃ ἔλεγον ἀκολουθεῖν μοι εἰς Καισάρειαν οὗ ἐφυλάσσετο· οἴτινες ἐλθόντες ἐβόων ἵνα ἀρθῇ ἐκ τῆς ζωῆς. ἀκούσας δὲ ἀμφοτέρων κατελαβόμην ἐν μηδενὶ αὐτὸν ἐνοχον εἶναι θανάτου, εἰπόντος δέ μου, θέλεις κρίνεσθαι μετ' αὐτῶν ἐν Ἱεροσολύμοις, Καίσαρα ἐπεκαλέσατο.
xxvii.1. ὡς δὲ ἐκρίθη τοῦ ἀποπλεῖν ἡμᾶς εἰς τὴν Ἰταλίαν, παρεδίδουν τόν τε Παῦλον καὶ τινας ἑτέρους δεσμώτας ἑκατοντάρχῃ ὀνόματι, Ἰουλίω, σπείρης Σεβαστῆς, ἐττιβάντες δὲ πλοίω. οὔτως οὖν ἔκρινεν ὁ ἡγεμὼν ἀναπέμπεσθαι αὐτὸν Καίσαρι. καὶ τῇ ἐπαύριον προσκαλεσάμενος ἐκατοντάρχην τιὰά ὀνόματι Ἰούλιον, σπείρης Σεβαστῆς, παρεδίδου αὐτῷ τὸν Παῦλον σὺν καὶ ἑτέροις δεσμώταις. ἀρξάμενοι δὲ τοῦ ἀποπλεῖν εἰς τὴν Ἰταλίαν ἐπέβημεν πλοίῳ.
xxviii.16. ἐπετράπη τῷ Παύλῳ μένειν καθ' ἑαυτόν. ὁ ἑκατοντάρχης παρέδωκεν τοὺς δεσμίους τῷ στρατοπεδάρχῃ, τῷ δὲ Παύλῳ ἐπετράπη μένειν καθ' ἑαυτὸν ἔξω τῆς παρεμβολῆς.
xxviii.19. Adds ἀλλ' ἵνα λυτρώσωμαι τὴν ψυχήν μου ἐκ θανάτου.
xxviii.29. Verse omitted by β text. καὶ ταῦτα αὐτοῦ εἰπόντος ἀπῆλθον οἱ ,Ἰουδαῖοι πολλὴν ἔχοντες ἐν ἑαυτοῖς συζήτησιν.
xxviii.31.   Adds ὅτι οὗτός ἐστιν Ἰησοῦς ὁ υἱὸς τοῦ Θεοῦ, δι' οὗ μέλλει ὅλος ὁ κόσμος κρίνεσθαι.


A study of these variants (a few of which found their way into the Received Text) will show how marked in this book is the difference between the β and δ texts. It will be plain also that they are not the result of casual scribal errors, but must be due to deliberate alteration on one side or the other, the δ text being habitually the longer and fuller of the two. In his earlier work (The Primitive Text of the Gospels and Acts, Oxford, 1914) Prof. A. C. Clark laid stress mainly on his theory that the ancestors of our uncials were written in columns with very short lines, and that the shorter text was produced by frequent scribal omissions of one or more of such lines. Against this it could be argued (1) that since the number of letters in a line is never constant, the method of counting letters becomes quite unreliable except for short passages, since any moderately high number can be accounted for by some combination of multiples of a variable base, ← E.g. if the archetype be supposed to have lines of 10-13 letters, any number from 30 upwards can be accounted for as a combination of lines of such length, so that the argument loses all cogency. (2) that such narrow columns are extremely rare in the early papyri, (3) that accidentai scribal omissions would not account for the habituai correspondence of the omissions with breaks in the sense, (4) that most of the variants aie not due to simple omissions, but to differences in the wording. Even if the text were written in στίχοι,, or sense-lines, as it is in D, the last-named class of variants would remain to be explained ; and there is no evidence for the early use of stichometrical arrangement. None of the early papyri yet known shows it, ← The Chester Beatty papyrus of Ecclesiasticus (probably late4th cent.) is written stichometrically ; but this is poetry. and it is probable that it first came into use in bilingual manuscripts, where it was desired to show the correspondance between clauses in the two languages.

In his later work (The Acts of the Apostles, Oxford, 1933) Prof. Clark practically abandons this explanation in favour of the much more plausible one of deliberate editorial alteration. Such an explanation was first put forward at the end of the seventeenth Century by Jean Leclerc, who suggested that Luke had himself produced two editions of Acts. This hypothesis found no support at the time, and though it was mentioned with approval by no less a scholar than Bishop Lightfoot, it was not seriously considered until it was revived by the great German classical scholar, Friedrich Blass, in 1895. Blass' theory was that Luke originally wrote his Gospel in Palestine, and that when he came to Rome with St. Paul and was asked by the Christians there for a copy of his work, he wrote it out again with such alterations as an author naturally feels free to make in dealing with his own work. Similarly with regard to Acts, one copy was made for Theophilus, to whom it was addressed, and another for the Church at large. These two editions Blass would identify with the β and δ texts respectively ; and, on the rather doubtful assumption that revision implies abbreviation by the removal of the superfluous, he concludes that the δ text represents the later edition of the Gospel and the earlier of Acts, while the β text represents the earlier edition of the Gospel and the later of Acts.

There is an undeniable plausibility in this theory, which would account satisfactorily for many of the variations. An author is at liberty to alter his own work, and when every copy had to be made by hand, there was frequent opportunity for it. This would account not only for alterations in substance, but also for verbal variations which are in themselves indifferent. It is not surprising, therefore, that the theory was received with favour, even by scholars of the standing of Salmon and Nestle. There are, however, serious difficulties. D Substitutes the genealogy of Mt. for that which appears in the ordinary text of Lk. ; is it conceivable that Luke made this alteration in his second edition, or if, as Blass admits, this cannot be due to Luke himself, may not the editor who made this alteration be responsible also for others ? Further, it is hard to believe that Luke, in revising his Gospel, Struck out the Word from the Cross and the clauses relating to the Resurrection and Ascension; what could be his motive ? In Acts also, if the δ text is to be regarded as the earlier, it is difficult to understand an author omitting some of the details found in it. The gain in space is small, and the loss in descriptiveness is sometimes such as, in Clark's words, to reduce the narrative to " a colourless abstract," or even to contradict the original narrative. Having correctly described the break of the journey from Caesarea to Jerusalem at the house of Mnason (xxi.16), he would not be likely to alter it so as to make it incorrect. He would hardly have omitted a clause from the decisions of the Jerusalem Council (xv.20, 29), or have altered the language of the letter of Claudius Lysias (xxiii.26-30) or Paul's speech to Agrippa (xxv.24, 25). A further difficulty arises from the variations of the δ text in the other Gospels, which, though less numerous, are of the same character, such as the additions at Mt. xx.28 and Jn. vi.56. For these Blass finds other explanations which increase in complication and decline in probability.

For reasons such as these Prof. Clark does not accept the theory of a double Lucan authorship, but Substitutes that of an editor who by alterations and omissions reduced the δ text to the form of the β text. This is far more possible, and if Acts stood alone it would be undeniably attractive, though there would still be some difficulty in understanding the mentality of the editor who would cut out the often picturesque details contained in the δ text. Ultimately we come up against the question whether these passages are best explained as omissions by one editor or additions by another. Here different scholars have taken different views. Ropes, the other scholar who in recent times has made a detailed study of Acts, takes the opposite side to Clark, preferring the authenticity of the β text. For the δ text it is to be said that its characteristic passages seem to show personal knowledge of the incidents in question; e.g., the fuller account of Peter's mission to Cornelius, the seven steps in the story of Peter's deliverance from prison, the narrative of Paul and Silas in the prison at Philippi, the explanation of Apollos' journey from Ephesus to Corinth, the hours of Paul's teaching in the school of Tyrannus, the touching at Trogyllium on the voyage to Jerusalem, the residence of Mnason in a village between Cæsarea and Jerusalem, the mention of the στρατοπεδάρχης at Rome. ← A conspicuous instance where the δ text not only shows personal knowledge but is unquestionably right is in xx.4, where all MSS. except D and g read Γαῖος Δερβαῖος. D has Δουβ[.]ριος, d has douerius, and g doberius. Since Gaius is described in xix.29 as a Macedonian, he cannot have been a native of Derbe ; and Clark was the first to point ont that a town named Doberus in Paeonia is mentioned by Thucydides (ii, 98-100), and a tribe of Doberes near Mt. Pangaeus by Herodotus (vii, 113). The mutilated word in D should therefore be Δουβηριος for which Δερβαιος is a mistake by a scribe who knew nothing of Doberus but remembered Derbe. The error is scribal rather than editorial, since a deliberate editor would hardly have forgotten that Gaius was a Macedonian 16 verses before ; but it is a distinct point in favour of the δ text. It is also quite admissible that the original text of Acts should have been produced at Rome, and therefore should have been preserved in the Western family of authorities ; though it has to be remembered that the home of the Western text has to be looked for in Africa rather than in Italy.

On the other hand it is difficult to understand the motive for the abridgement of the text by the β editor. It is true that the β text in the Gospels also is marked by omissions rather than by additions; but there the contrast between the two texts is of a very different character, and the balance of probability seems on internal grounds to be decidedly against the insertions of the δ text; and in view of the general homogeneity of character of the β text in the New Testament it is hard to understand why he should in Acts have broken out into a passion for the pruning away of detail and even the alteration of statements of faet. But perhaps the greatest difficulty (which is not faced by Clark) is the existence of the δ text in the Gospels. There is no reason to ascribe a Western origin to the Gospels, and there are signs of a free handling which, as an evidence to character, teils against the authenticity of the δ text. It is a priori more probable that the details in the δ text of Acts were added by someone with personal knowledge of the events, than that they were originally present and were excised, without apparent reason, by a subsequent editor. What one would like to suppose (but for which there is no external evidence) is that one of St. Paul's companions transcribed Luke's book (perhaps after the author's death), and inserted details of which he had personal knowledge, and made other alterations in accordance with his own taste in a matter on which he was entitled to regard himself as having authority equal to that of Luke. Attractive as Prof. Clark's theory is at first sight, and admirable as is the learning and industry with which it is set out, it leaves many difficulties unremoved; and unless future discoveries should supply a solution, the problem must be solved according to the intrinsic probabilities of the methods of insertion or excision.

In the Pauline Epistles there is a quite distinguishable δ text, represented by the Graeco-Latin bilinguals D2 E3 F2 G3 with the Old Latin MSS. mentioned above (p. 140). It is however of much less importance than in the Gospels and Acts. There are no variations in substance; probably the writings of Paul had too much individuality to admit of the free handling which was admissible in the anonymous narratives of the lives of our Lord and the apostles. Most of the variants are purely verbal, or affect only the order of the words. Sometimes there are slight expansions, either because the expanded form was usual, or for the sake of slightly greater definition; e.g., Rom. x.15, the completion of the quotation by the addition of the words τῶν εὐαγγελιζομένων εἰρήνην, xii.17 the addition of οὐ μόνον ἐνώπιον τοῦ Θεοῦ, ἀλλὰ καὶ, 1 Cor. xvi, 22 the addition of Ἰησοῦν Χριστὸν to Κύριον, Gal. vi.17 the expansion of Ἰησοῦ to Κυρίου ἡμῶν Ἰησοῦ Χρίστοῦ, Heb. ix.18 the addition of διαθήκη after πρώτη. In such cases the presumption is always against the addition, as being more in accordance with the practice of scribes influenced by familiar phrases or anxious to make the text clear. In some cases, though the same motives are discernible, the addition is rather unintelligent ; e.g., Heb. x.28 the addition of καὶ δακρύων after οἰκτειρμῶν, 1 Cor. ix.7 the addition of καὶ πίνει after ἐσθίει, x.17 the addition of καὶ τοῦ ἱvὸς ποτηριού after ἄρτου (though the previous words are concerned with ἀρτος only), xvi.15 the addition of καὶ Φορτουνάτου or Φ. καὶ Ἀχαϊκοῦ because those names are mentioned below, Phil. iii.16 the insertion of τὸ αὐτὸ φρονεῖν, before τῷ αύτῷ στοιχεῖν.P46Some of the variants are, however, quite possible, and a minority of them are supported by our earliest authority, P46. Again it is noticeable also that (as mentioned above, p. 205) Β in several instances joins the δ group, and in most of these is supported by P46. On the whole, however, the δ variants in the Epistles belong to the category of minor verbal variants, and are not of the same class as those which command attention in the Gospels and Acts.

For the Apocalypse there is no Greek MS. representative of the δ text, which has to be sought in the few Old Latin MSS. that contain this book, g and h, and in the quotations in Primasius (covering practically the whole book), Cyprian and Tyconius.

The general tendency of modern discoveries and criticism has been to break up the unity of the δ text as at one time envisaged. It has been made clear that not all pre-Syrian and non-Neutral readings are to be classed as Western. The range of such readings, as attested by manuscripts, versions, and Fathers, is spread all over the Christian world, but they are far from forming a homogeneous text. The Old Latin MSS. differ widely among themselves, and the use by the early Fathers is sporadic and inconsistent. The difficulty of forming a " Western " text is illustrated by Blass' attempt to form such a text of Lk., which practically consists of an arbitrary selection of readings from authorities regarded as Western. Some of the authorities formerly so classed have now been more accurately placed in other groups. Famm. 1 and 13 have gone into the Cæsarean group, together with parts of W and Θ and some of the quotations in Origen and Eusebius, and the Armenian and Georgian versions. Others will be dealt with in the groups which have still to be considered. On the whole, it appears that the δ text should be confined to the truly Western authorities, the Graeco-Latin uncials and a few kindred minuscules, the Old Latin version (especially the African form of it), and the Latin Fathers. There are some who have questioned whether it ever truly existed in Greek, except as a Greek text revised from Latin sources; but against this must be set the recently discovered papyrus fragments, P38 and P48, which prove that a Greek text of this type was known in Egypt in the late third or fourth Century. But this does not prove that it originated in Egypt, where the majority of texts discovered are certainly not of this type; and when limited as above indicated the name of " Western " may properly be assigned to this family.

V.—The ε Text

It seems desirable to form a separate family under this title of the New Testament text as known in Syria. There was formerly a tendency to group this with the δ text, and to argue that the Western text was at one time dominant over the whole Christian world, and was only gradually reformed away ; though why there should be such an inveterate hostility to it did not appear. There is no doubt that the Old Syriac version in many cases agrees with the δ text; but further study, and especially the work of Burkitt, has given strong reason to believe that this Western colour is not original, but is due to the importation of a Western text in the form of Tatian's Diatessaron. In the 24 " major " variants quoted on pp. 216-8 the Old Syriac agrees with the δ text in 5 instances, but disagrees in 13 ; in 6 its evidence is divided, but in each case it is the later, or Curetonian, witness that has the 8 variant. In the 15 " minor " variants on pp. 221-3 the Old Syriac has the 8 text only twice, but differs from it 12 times (twice with readings peculiar to itself) ; once its testimony is divided, the Sinaitic this time supporting D. Other figures, more favourable to the β text, are given on p. 121. The Syriac New Testament therefore assumes the form of a local text, formed from the basis of a β text (or possibly a γ text), strongly coloured by Western influences, and pre-serving also a number of readings which are neither β nor δ, but belong to the unassorted readings which we believe to have been very numerous in the early centuries, before the families as we now know them were even approximately fixed. Subsequently this mixed text was further revised by Rabbula under the influences which produced the α text, and in this form, as the Peshitta, it became the accepted Bible of the Syrian Church. The true Philoxenian version appears to have been wholly of the a type, but never obtained much authority; and the Harkleian revision of it was due to the use of Greek MSS. of the δ type, and, though useful for the reconstruction of the δ text, tells us nothing of the text native to Syria. The Palestinian Syriac contains elements of ail kinds.

Together with the Syriac versions, the Armenian and Georgian give much help in the reconstruction of the ε text, the former having almost certainly been translated from Syriac, and the latter from the Armenian. From these it appears, as has been stated above, that the Georgian, and therefore presumably its Armenian and Syriac ancestors, had affinities with the Caesarean or γ text. Further study by the few scholars competent to deal with these languages may throw more light on this obscure subject; but for the present it seems best to treat the Syriac text as a group apart, having affinities with the β, γ, and δ groups (and in its later history with the a text also), and to be used with care for the constitution of these texts, and as evidence of the extension of their influence into other lands than those in which they originated.

VI.—The Residue

If the foregoing classification be accepted, there will remain a considerable residue of unclassifïed readings. They do not form a coherent text; their essence is that they are a congeries of unassorted early readings, left behind by the restriction of the β group to the text mainly represented by B, and of the δ group to the text mainly represented by D and the Old Latin. They include the readings of which Hort formed his γ group or " Alexandrian " text; they include also a number of readings found in the papyri ; they probably include also an unascertained number of readings which have found a home in other groups, especially the Western, but which are essentially of a different character. Although no coherent whole can be formed out of them, they are not therefore unimpor-tant. On the contrary, they throw a considerable amount of light on the early history of the New Testament text.

Hort's " Alexandrian " text was avowedly not to be found in any extant manuscript. It consisted of residual readings, found in manuscripts which on the whole are associated with the β text, though many of them also have a proportion of δ readings. They are not very closely defined, or perhaps definable, but are rather what is left after α, β, and δ readings have been identified and segregated. It is no long step to the suspicion that there never was an " Alexandrian " text in this sense; and this suspicion is strengthened by the papyri, which often, and indeed generally, offer texts which cannot be classified as wholly either β or δ, while they are certainly earlier than α. It is important to recognize the existence of this category, and not to sweep ail early non-β readings into the δ class. When, for example, the Chester Beatty fragments of Acts show 89 agreements with D, but not a single one of the more characteristic δ readings, it would be absurd to rank it among the δ family, or to regard it as giving even a qualified support to it as against its rivais. What it does is to add its confirmation to the belief that the textual history of the second and third centuries cannot be arranged in an orderly genealogical stemma, but presents rather a welter of unassorted variants, out of which the families that we find at a later date were eventually formed.

It is not to be understood from this that the New Testament text was at first a chaos of Democritean atoms, from which in course of time an ordered form emerged. At the first each book had its single original text, which it is now the object of criticism to recover; but in the first two centuries this original text disappeared under a mass of variants, created by errors, by conscious alterations, and by attempts to remedy the uncertainties thus created. Then, as further attempts to recover the lost truth were made, the families of text that we now know took shape. They were, however, nuclei rather than completed forms of text, and did ncft at once absorb all the atoms that the period of disorder had brought into existence. There was a residue of various readings, which attached themselves more or less fortuitously to the one centre or another, and which now confuse our conception of the several families. If it is true that none of the families which modern criticism has distinguished preserves uncontaminated the original text of the New Testament books, we must imagine for each of them an editor who at some point in the period of obscurity gave each of them its characteristic form. But he drew on a mass of material, much of which consisted of relatively minor details; and while we may, and indeed must, if we are to arrive at any conclusion, believe that one or other of them was the more right in his general principles, it does not follow that he was right in all details. We may believe Neutral or Western or Cæsarean to be on the whole preferable, with-out binding ourselves to accept it in all respects. There remains this floating residue, to be found in the representatives of each and all of them, which at once reminds us of the conditions out of which they came irito being, and presents the material out of which we must choose in completing the details of the best text at which it is possible, in the present state of our knowledge, to arrive.

If the present state of opinion with regard to the several textual groups be compared with that formu-lated by Westcott and Hort half a Century ago, it will be seen that there have been several modifications. It does not now seem probable that the β text (Hort's Neutral) can be regarded as having descended without editorial intervention and substantially unaltered from the original, but rather that it owes its present form to competent editorial treatment in accordance with the principles of trained textual scholarship. On the other hand, its principal rival, the δ text (Western), which at one time was regarded as a text pervading the whole Christian world and with earlier attestation than any other, has been broken up and reduced to a truly Western group, the extreme form of which offers wide variations from ail other groups, which (with some reservation in respect of Acts) do not seem on examination to gain in authority. Out of the group formerly classed as Western have been carved two other groups, the newly discovered Cassarean, with attestation from the time of Origen, and the text of the Church of Syria. These two last-named groups have affinities both with Egypt and the West, and ail take their rise from a period of much confusion of texts, which has left a residue of readings which may appear now in one and now in another of the main groups. In this residue is merged the group which Hort entitled Alexandrian. Finally the α text remains much as Hort left it, an essentially secondary text, based upon a process of revision in minor details which began about the end of the fourth century and continued for several centuries, affecting the descendants of ail the earlier groups to varying extents, and finally dominating the Byzantine Church until the invention of printing, when it became the Received Text of the whole Church until the rise of modern criticism under the influence of the discoveries and research of the last Century. Its progressive development has heen elucidated by the labours of von Soden.

In addition, recent discoveries have materially increased our knowledge of the condition of the text during the earliest Christian centuries. We now have texts, substantial in the case of the Gospels, Acts, and Apocalypse and almost complete in the case of the Pauline Epistles, which go back to the early part of the third Century. The interval between the composition of the books of the New Testament and the earliest extant manuscripts of them has been reduced by a hundred years, and we actually have evidence (small but decisive) of the circulation of the Fourth Gospel in the first half of the second Century—that is, within about a generation of the earliest date usually assigned to its composition. The general effect of this new evidence is to confirm the substantial integrity of our texts of the New Testament 'Scriptures, but also to show that in the third Century, and therefore presum-ably for some time back into the second, the text was in detail very far from being settled. Instead of a state of orderly descent, though with an ever-widening genealogical pedigree, from the original autographs to the extant copies of the fourth Century, we seem to see a period of increasing disorder, from which a state of comparative order was ultimately produced when the Church reached more settled conditions.

To understand the state of the New Testament text, it is necessary to form as clear a picture as the evidence permits of the circumstances of the early centuries. It is in this respect that the textual history of the New Testament differs materially from that of other ancient books. The works of classical literature were produced in peaceful conditions. They were copied by professional scribes. They were the work of recognized authors, whose individuality was respected, so that a scribe, though he might try to correct what seemed to him to be errors of transmission, did not feel at liberty to revise his author. They were not exposed to deliberate destruction, at any rate until, after many centuries, the Christian Church made war on pagan literature. The textual tradition which has come down to us is probably that of the great librar-ies, where good copies were preserved under the eyes of men of letters. In all respects they have had a respected and protected course until they suffered in the decline of civilization in the Middle Ages.

In all these respects the fortunes of the Christian Scriptures were different. In the earliest days the Christians were a poor community, who would seldom have been able to command the services of professional scribes. There were no recognized centres for the promulgation of authorized copies of the Scriptures. The New Testament books were not indeed Scriptures with a capital S. The sacred books of the earliest Christians were those of the Old Testament ; the writings of Mark and Luke and Paul were just memoirs and letters for present needs. Nor were there the usual literary standards to protect the integrity of their texts. The letters of Paul were, indeed, individual expressions of opinion which would not lightly be altered; but the anonymous Gospels and Acts were not regarded as the literary compositions of their authors, but as narratives of the life of our Lord and the work of His apostles, compiled with the purely practical object of disseminating the knowledge of their lives and teachings among the Christian community, and with no eye to a future which in any case would soon be curtailed by the Second Coming. There was no need to be meticulous in verbal accuracy. The substance was what mattered, and if additions, believed to be authentic, could be made to it, why should they not? Then there were little means, even if it had been thought needful, to secure uniformity of transmission. Each book circulated originally as a separate roll, and there was no fixed Canon of Christian Scriptures. Not every Christian community would possess a complete set of Gospels or of Paul's epistles, but each would supply itself as best it could from its neighbours. Many copies would be made by untrained provincial copyists, and there would be no opportunity of correcting them by comparison with other copies, except such as might be in the immediate neighbourhood. Such revision as there might be would be local and unmethodical. Then there was always the danger of destruction. Christianity was often a tolerated religion when books could be multiplied and distributed without hindrance; but there were also periods of persecution, sometimes local and occasionally general, when the Christian books were exposed to danger. There was much destruction of Christian property, and in the general persecutions the sacred books were special objects of search, when those which belonged to churches, which would be likely to be of superior quality, were the least likely to escape. So long as Christianity was at best tolerated and at worst persecuted, the transcription and circulation of the Scriptures were exposed to difficulties from which the pagan literature was free.

In circumstances such as these it was natural that varieties of readings should multiply. Apart from the growth of errors through untrained scribes, there was no restriction on casual alterations, made with the best intentions with a view to greater lucidity and the avoidance of possible misunderstandings. Such revision as there was would be local and casual, due to the initiative of individual bishops or scholars, and its influence would be confined to the immediate neighbourhood. This would tend to the creation of local types of text, extending at most to a province or to part of a province. At times copies would be brought from one part of the Empire to another, from Rome to Antioch or Alexandria or vice versa, and certain copies would be corrected from them, leading to the inter-mingling of local texts. Only when times were peaceful, and only decisively after the recognition of Christianity by Constantine, could revision be seri-ously taken in hand, and by that time the confusion had been created, the original autographs had long ago disappeared, and absolute verbal accuracy was no longer obtainable.

At what time and by whom the revisions were undertaken to which the families of text now known to us are due, we do not know. A few names are known to us. In the second Century Marcion produced editions of Luke and Paul for his own unorthodox purposes, and Tatian compiled his Diatessaron. Later on, as we have seen, Lucian and Hesychius, Origen and Eusebius revised the texts of the Septuagint, and the two former may have produced editions of the New Testament as well. At a later date we know of the labours of Jerome on the Latin Bible and Rabbula on the Syriac. All we can say is that by the fourth Century certain main varieties of tradition had been developed, and we can judge of the character of the editor or editors to whom they were originally due from the results. We can see also that they had to deal with a great welter of various readings and of mixed texts. When an editor set himself to deal with the varied material that he found before him, he could approach his task in several different ways. To some (and this was the commonest type) the governing idea was to make the text plain and easy. To this end the order of words was altered, names and pronouns were introduced, normal phrases were substituted for less usual ones, words which might be misunderstood or misinterpreted in an unorthodox manner were omitted or changed, narratives in one Gospel were assimilated to those in another, inexact quotations from the Old Testament were made exact, alternative readings in different manuscripts were combined, and so on. The result was an easy, intelligible text, at some sacrifice of character, and ultimately differing considerably in detail, though not in essential teach-ing, from the original. In this way such a text as the a text was produced.

Another editor, of more independent frame of mind, might treat the text more freely, varying phrases to suit his own taste, importing short passages from other sources, amplifying the narrative for the sake of effect, including rather than omitting, and attaching little importance to accuracy of transmission, though with no doctrinal motives. This would be the genesis of the δ text. At the opposite extreme to him would be the trained scholar, whose guiding principle would be accuracy, not edification, who would be thinking of the author rather than of the reader. He would be careful to consult the oldest manuscripts accessible to him, and would compare their variant readings in the light of critical science, considering which was most likely to give the author's original words. He would tend to omit superfluities or insufficiently attested words or passages, and to prefer the more difficult reading to the easier, as more likely to have been altered. These are the lines on which the closest approximation to the original would be arrived at. They are the established principles of textual science, which are applied to the editing of classical texts; and it is because the β text appears to have been formed on these lines that most modern scholars, even if they are not satisfied that it has preserved the authentic original substantially in uncontaminated integrity, neverthe-less hold that on the whole it offers us the purest text.

Of the other families, the Caesarean would appear to have been constructed on scholarly lines, though perhaps with less authority than the β text. It has, however, been too imperfectly established as yet for a final judgment to be formed. The Syrian text would appear to owe less to editorial revision on conscious lines, and to be a local text formed out of such mater-ials as were available, reflecting influences from all quarters. Such interchange of influence has indeed to be recognized as a constant factor in textual history. Hardly any manuscript is free from influences from authorities of a different complexion from itself. The readings characteristic of each family have for the most part to be extracted from manuscripts which in the main have been revised into conformity with the prevalent text,—generally and increasingly the Received Text, which ultimately submerged all others, leaving only here and there a few relics, like the wandering blocks of geology, of a more primeval form.

But to whichever family a critic may give his general preference, it by no means follows that he should regard it as always right. So long as the β text was regarded as practically free from editorial revision, it was natural to give it the preference except when it is obviously wrong—as every manuscript must sometimes be. But if once editorial treatment is admitted, then it stands to reason that the editor cannot always have been right. He had to make a number of choices where the balance of probability was quite undeter-mined, and, being human, he cannot always have chosen right. Therefore a rrtodern editor must be free to consider readings with an open mind, whether he finds them in a manuscript predominantly of β or γ or δ character. He may not be any more exempt from error than the ancient editor, but he has far more material to choose from, and centuries of experience to guide him, and he must make his decision to the best of his ability, though without dogmatism.

The natural conclusion is that while one family may be generally to be preferred, it is probable that readings found in other families will sometimes be right. This is the conclusion to which all the evidence derivable from the early papyri points, and notably that of the Chester Beatty papyri. For the present purpose it matters not whether the text of these papyri be regarded as good or bad. What is significant is that they prove that in Egypt in the early part of the third Century readings were in circulation which were derived from, or which eventually became attached to, all the principal families, together with a not inconsiderable number of which no other witness has survived. In the Gospels the Chester Beatty papyrus has readings which we associate with the β, γ, and δ texts, in varying proportions in the different books. In the Acts and Pauline Epistles we find a preponderance of agreement with the β text, but a quite considerable number of agreements with δ. And there are in addition many readings that belong to neither. We must therefore be prepared to find that the best manuscript or family is not always right.

This conclusion is strongly reinforced by what we have learnt from the papyri with regard to the texts of classical authors. Here again they have greatly enlarged our knowledge of the early history of these texts. There was a time when it was supposed that they were handed down in substantial purity through the Hellenic and Hellenistic periods, and that the corruptions which we find in the texts as they have reached us are due to unintelligent Byzantine scribes. We now have hundreds of papyri and fragments of papyri from periods earlier than the Byzantine (from the third Century B.C. to the third Century A.D.), and we know that this picture is a false one. Most of the corruptions found in our modem texts were already in existence in the earlier period. More than this, we find that this early evidence not infrequently supports readings hitherto known only from late and inferior manuscripts. It is true that, as a whole, the papyri confirm the general superiority of the manuscript or manuscripts which modem scholarship regards as the best for the several Greek authors, and to which it has often pinned its faith almost exclusively; but in a minority of cases it supports the " inferior " manuscripts against them. Indeed, it may be said as a general rule that the papyri support the better manuscript in two cases out of three, and the inferior manuscripts in the third. The evidence with regard to the classical texts therefore entirely supports the conclusion at which we have arrived in respect of the Biblical texts, that exclusive trust must not be placed in any one authority.

Although this conclusion suggests that absolute certainty in details is unobtainable, it is entirely justifiable to end on a note of hopefulness. In the first place, those who are attached to the Bible will be glad to find that ail the discoveries which have been so plentiful of late years have tended to confirm the authenticity and general integrity of our texts, and to establish them on a firmer basis than ever. The fears of those who, since the days of Mill, have been led by lack of faith to dread and doubt the free investigations of scholarship have been shown to be groundless. Truth has flourished in an atmosphère of free research. And secondly, if so many discoveries have been made in our own generation, there is every reason to hope that more discoveries may still be awaiting us in the sands of Egypt. A second-century Gospel is by no means an impossibility ; we already have a tiny fragment of one, and more extensive remains of two books of the Septuagint. But if such a treasure should appear, it will be well not to expect too much. It will probably show that errors already pervaded the sacred texts as they circulated in provincial Egypt; and in any case it would have to be remembered that it was originally only one copy among many, and that its evidence as to the best text may not be decisive. Another find even more to be desired is a substantial portion of the Diatessaron : for this might determine the vexed question of the disturbing influence of Tatian's work, both on the Syriac Gospels and the whole textual tradition, and might bring us appreciably-nearer to a comprehension of the history of the New Testament text, which is the final goal of textual criticism.

ἅπανθ' ὁ μακρὸς κἀναρίθμητος χρόνος
φύει τ' ἄδηκα·
κοὐκ ἔστ' ἄελπτον oὐδέν.



Westcott and Hort, von Soden, Streeter, Kenyon, opp. citt. : E. Nestle, Introduction to the Textual Criticism of the Greek New Testament (Eng. tr., 1901); K. Lake, The Text of the New Testament (1900 and later revisions) ; F. Blass, Acta Apostolorum secundum formam Romanam (1896), Evangelium Lucae secundum formam Romanam (1897), Philology of the Gospels (1898); A. C. Clark, The Acts of the Apostles (1933).


Page 46. l. 3. After xxxix, 29 add with gaps of 5 leaves.
l. 17. For not yet edited read Edited by F. G. Kenyon, Chester Beatty, Biblical Papyri, fasc. vii (London, 1937), the Princeton leaves by A. C. Johnson, H. S. Gehman, and E. H. Kase (Princeton, 1938).
At foot of page add Edited by G. Vitelli, Papiri Societa Italiana, vol. viii, fasc. 2, 1927.
Add also Pap. Fouad 1. Deut. xxxi.28-xxxii.7. Second Century B.C. Published in Biblical Archaeologist ix. 2, May 1946.
Page 64. 6 lines from foot. Progress with the Cambridge Septuagint has been delayed by the death of Dr. Brooke in 1939 and the ill-health and death (in 1947) of Dr. McLean. Part 1 of vol. iii, containing Esther, Judith and Tobit, appeared 1940.
Page 74. l. 3 ff. The photographie facsimile was published in 1937. The position of Hebrews is probably due only to its length, the epistles being arranged in descending order of length.
Page 80. In an elaborate study (Scribes and Correctors of the Codex Sinaiticus, 1938) Messrs. Milne and Skeat make it very probable that the number of original scribes should be reduced from four to three. Their reduction of the number of correctors from nine to two (who are identified with two of the original scribes) is less certain.
Page 103. l. 13 Codex Ξ. is assigned to the 6th cent. by W. H. P. Hatch (in Casey and Lake, Quantulacunque, 1937, but the argument, though plausible, is not conclusive. The MS. π should be included in this list, in view of the study by Silva Lake, Family π and the Codex Alexandrinus (Studies and Documents, part v, 1936).
Page 105. On Cod. 13 see K. and S. Lake, Family 13 (The Ferrar Group) Studies and Documents, part xi, 1941.
Page 180. After l. 15, add For a study of Ka see S. Lake, Family π and the Codex Alexandrinus, who decides that π is the archetype of this family, which is akin to A but not directly descended from it. In an appendix Mrs. Lake gives the readings of Kr in Mark.