Chapter III: Theory and Praxis of the Textual Criticism of the New Testament.129 (3)

Zahn also calls attention frequently to the different manuscripts which still exhibit a text agreeing with that credited to Marcion, and which are precisely the Western witnesses, the Old Latin manuscripts, and D of the Greek.[214] Compare, _e.g._, on Luke v. 14, 34, 39; vi. 25 f., 31, 37; viii. 45; ix. 6, 16, 22; x. 22, 25; xi. 20, 41; xii. 14, 31, 58 f.; xviii. 35; xx. 36; xxi. 27, 30; xxiv. 6, 26, 37. But there are also passages where Marcion parts company with D and its associates—_e.g._, vi. 22, 26, 29; xi. 4. In Paul, too, the number of passages displaying agreement between Marcion and D_{2} G_{2} preponderates: Gal. ii. 5; iii. 14_b_; v. 1, 14, 24; 1 Cor. i. 18; 2 Cor. v. 4; 1 Thess. iv. 16; Eph. i. 9, 13; iii. 10; iv. 6; v. 28 ff. The agreement between Marcion’s text and that of the minuscule 157 was previously emphasised by Zahn—_e.g._, in Luke xvi. 12, where the reading τὸ ἐμόν instead of τὸ ὑμέτερον (ἡμέτερον B L) is supported as yet by this Greek manuscript alone and three old Latin (e i l), and in xxi. 30, where only one other of the minuscules collated by Scrivener supports D 157 in reading προβάλωσιν τὸν καρπὸν αὐτῶν.[215] In Luke xxiv. 26, D and Marcion are our only witnesses for the reading ὅτι instead of οὐχί. How is this to be explained? Zahn, _e.g._, holds that it is a mere coincidence that Marcion’s reading, “prophetas suos,”[216] in 1 Thess. ii. 15 agrees with τοὺς ἰδίους προφήτας read by D_{2} E_{2} K_{2} L_{2}, _i.e._, the representatives of the Antiochean recension, with which Marcion elsewhere very seldom agrees, seeing he founds throughout upon a Western text. In the great majority of cases the explanation seems to be simple enough. Marcion began his career at Rome, so that we may naturally expect him to give us a Western text. So far, therefore, one might be tempted simply to ignore that text as hitherto, although a text attested by Marcion and the Church in common is surely entitled, even in respect of its antiquity, to much more consideration than has been paid to it heretofore. The importance, and at the same time the difficulty, of the problem is increased by the fact that we find the same text as his, or at all events one of a similar sort, represented in a totally different quarter—viz. in Tatian.[217]

[Sidenote: Tatian.]

(_h_) =Tatian=[218] has already been referred to in a general way above (p. 97 ff.): we shall now give the testimony of the early church regarding him verbatim. If we leave out of account the somewhat doubtful reference in Hegesippus (p. 96), and an equally uncertain allusion to the title of his Harmony in Origen,[219] the testimony from purely Greek sources is confined to a few sentences in Eusebius,[220] a notice in Epiphanius,[221] and a scholion in a manuscript of the Gospels.[222]

For more exact information we are indebted solely to the Syrian church. The Greek writer Theodoret gives us most details.[223] The notices contained in Syriac and Arabic sources are more numerous than the Greek, but they are shorter and must be omitted here.[224] It becomes necessary, therefore, to consider very carefully whether any vestiges of Tatian’s work are preserved in our witnesses for the text, and how these may, and indeed must, be used in its criticism. I assume as having been demonstrated by Zahn, that Tatian’s Diatessaron was a Syriac work, and I take it as very probable that the Curetonian Syriac and the Lewis Syriac present us with two works based on, or at least influenced by, that of Tatian. To what extent the same is true of the Peshitto as well need not be considered here, the main problem being to elucidate the connection between Tatian and the _Western_ witnesses. And here we are at once confronted with a matter of great uncertainty—viz., whether there might not also have been a _Greek_ Harmony of the Gospels either antecedent to the Diatessaron or contemporary with it, which Tatian himself made or employed. Zahn thinks not, mainly because from the side of the Greek Church we have almost no notice whatever of the existence of anything of this sort, nor of Tatian’s own work either. Harnack seems not to be convinced of the correctness of Zahn’s position.[225] He even declares that Harris’s _Preliminary Study_[226] has only confirmed his “conviction that Tatian composed a Greek Harmony of the Gospels.”[227] That treatise is accompanied by a facsimile of the fragment of Mark in Codex W^d, “the contents of which display an affinity with the text of the Diatessaron (with the original text?).”[228] At all events Harnack is of opinion that Harris’s conclusions with regard to a Pre-Tatian and a very early Harmony of the narrative of the Passion are very premature, and in his judgment should either not have been put forward at all in a _Preliminary Study_ or suggested with more deliberation. G. Krüger also puts “this Combined Gospel written in Syriac (Greek?)” in his _History of Early Christian Literature_, § 37. On the other hand, Hogg in § 12 of his Introduction,[229] _Non-Syriac Texts of the Diatessaron_, says nothing of a Greek text, and in § 19, where he raises the question, “In what language was it written?”, he speaks only of the “view favoured by an increasing majority of scholars, that it was written in Syriac,” and then asks, on this view, “was it a translation or simply a compilation?” and lastly, which is the main question, “what precisely is its relation to ... the Western text generally?”

In his first work, written prior to the publication of the Arabic text, Zahn very frequently pointed to the fact that the so-called =Western= witnesses—_i.e._, Codex D and the Old Latin manuscripts, agree so often with Tatian.[230] His explanation of this phenomenon is very simple—viz., that Tatian returned from Rome to his old home in Syria about the year 172, and took with him from the West his text, which was just the Western text. [Sidenote: Tatian and the Western Text.] This view would present no difficulty if it were only the case that the Diatessaron shared the peculiarities of the Western text, but is the fact not rather the converse of this—viz., that D, the leading representative of that text, shares the peculiarities of a Harmony of the Gospels, might we say, in short, of the Diatessaron? Not only are certain readings the same in both texts, but the Western text seems actually to exhibit features which can scarcely be regarded otherwise than as the outcome of a Harmony. I have given expression to this opinion ere now; it struck me forcibly when I was collating the Codex Bezae for my _Supplementum Novi Testamenti Graeci_. In order to afford a more convenient survey of the vast number of variants, I followed the paragraphing of Westcott and Hort’s edition. Now look at the variants there. Whereas the majority consist of quite separate and disconnected readings, I was obliged at the beginning of the pericopæ regularly to copy half a line or even a whole line from D, its text differed so much from that of our present editions at the beginnings of the pericopæ, and there only to the same extent. See, _e.g._, Luke v. 17, 27; vii. 1, 18; ix. 37; x. 1, 25; xi. 14; xii. 1 to the end; xxiv. 13. It is true this phenomenon is most frequently observed in Luke, where I had previously explained its appearance in another way by supposing like Blass that it was due to the author having issued two editions of that Gospel. But neither is it altogether absent from the other Gospels. It occurs most seldom, as might be expected, in Matthew, but examples may be seen in xvii. 22, 24; xx. 29. In Mark see iii. 19; iv. 1; vi. 7. There are other features besides this which are difficult of explanation on any other grounds. For these I may briefly refer to the second of the works relating to this part of the subject, _The Syro-Latin Text of the Gospels_, by F. H. Chase,[231] in which a special chapter is devoted to the question of “Harmonistic Influence” (pp. 76-100). The writer calls attention there to three points, viz.:—

1. “The text of Codex Bezae shows constant indications of harmonistic influence.” This, however, is nothing new. Jerome, _e.g._, complains of amalgamations of this sort. But then,

2. “In such harmonized passages readings occur which we are justified by other evidence in considering as Tatianic readings.”

3. “There are other signs of the influence of Syriac phraseology in, or in the neighbourhood of, such readings due to harmonistic influence.”

I waive consideration of this last point, but as regards the second it is noteworthy, and bears out what I have said above, that Chase in this connection goes almost entirely by passages from Luke with the exception of Matt. xxi. 18; xxiv. 31 f.; xxvi. 59 ff.; and Mark viii. 10; xiii. 2; x. 25 ff. From Luke he instances iii. 23-38; iv. 31; v. 10 f., 14 f.; vi. 42; viii. 35; xi. 2; xx. 20; xxi. 7; xxiii. 45 ff.; xxiv. 1.

I should like, however, to call attention here to one passage to which Chase refers in another connection—viz., the extensive interpolation after Matt. xx. 28 (Chase, pp. 9-14). It is true, as Zahn expressly points out,[232] that neither Ephraem nor Aphraates, who were our only sources for the Diatessaron prior to 1881, “shows any traces of this long and in part apocryphal interpolation, nor yet of Luke xiv. 7-10, from which the most of it is taken.” But in the Arabic Tatian,[233] Luke xiv. 1-6 and xiv. 7-11, 12-15 are found after Matt. xx. 1-16 at the end of § 29 and the beginning of § 30 respectively. The verse Matt. xx. 28, regarding which Zahn was uncertain whether it was in Tatian or not, seeing that neither Ephraem nor Aphraates mentions it, is found in § 31, 5 between Mark x. 44 and Luke xiii. 22, while Matt. xx, 29_a_ (+ Mark x. 46_a_) follows a little further down in § 31, 25. So far, indeed, this result is not favourable to our theory. But I ask myself in vain how else this interpolation is to be explained except as an attempt at harmonising. Now, seeing that its text is found in one Syriac, two Greek, and half a dozen Latin witnesses (the particulars are given in the critical note, p. 255), the further question arises, Whence comes it? The most ready answer will be, “it comes from the Greek, whence it passed to the Latin on the one side and to the Syriac on the other.” As for the Latin, it is certain that the majority, perhaps even all, of the Latin forms are derived from the Greek. But are the Syriac as well? Or is not rather the converse true, however strange it may seem at the first glance, that the Greek is a translation of the Syriac? There is the word δειπνοκλήτωρ, _e.g._, which strikes me as it did Chase, as being particularly strange. I admit that I should not care to build a hypothesis of this magnitude on this one word and this one passage alone. I would merely submit it generally as a question to be kept in view in further investigations. And I would supplement it by another question whether, in the case of the first being negatived, it may not be true after all, _pace_ Zahn, that there was a Greek Harmony alongside the Syriac and probably going back to the same author. May not the close resemblances traceable between Tatian and the Western text be also accounted for on the supposition that instead of Tatian being influenced by the latter, it really goes back to Tatian?

I would ask this question specially in regard to the Western text of the _Pauline Epistles_. What is meant by the statement of Eusebius cited above as to Tatian’s treatment of these Epistles? Μεταφράσαι may certainly mean to translate, but then one translates an entire text and not φωνάς τινας merely, and, moreover, one does not translate ὡς ἐπιδιορθούμενος αὐτῶν τὴν τῆς φράσεως σύνταξιν “with a view to improving the phraseology and syntax.” Do not our Western witnesses present us with a work of this description? I am well aware that such hypotheses are like that regarding the author of the _Nibelungenlied_ where there was a great poem without a name and one or two great names without poems, and so various combinations were made, for each of which something could be said, while none of them could be said to be proved. That may be the case here too. But at present I feel disposed to attribute a considerable share in this peculiar “Western” text to Tatian. [Sidenote: Syro-Latin.] And as this name “Western,” the inappropriateness of which has long been recognised, becomes on this supposition more inappropriate still, I am inclined to recommend the freer adoption of the nomenclature familiarised by the work of Chase, I mean that of “Syro-Latin.” In his preface Chase puts in a plea for its use, citing a sentence from the _Dublin Review_ of July 1894, p. 52, in which H. Lucas says: “The time is, we hope, not far distant, when the term Western will give place to the term Syro-Latin, the only one which truly represents, in our opinion, the facts of the case.” Just as when we wish to indicate those languages and tribes that extend from the Indian to the German and Keltic we say Indo-Germanic, or Indo-Keltic, if we wish to be more exact and avoid wounding the sensibilities of the French, so the term Syro-Latin would be the best designation for a form of text whose characteristics are as distinctly traceable among the Syrians in the East as among the Greeks in the centre and the Latins in the West. But be that as it may, one thing is clear, that many problems here await solution. But they will not for ever defy methodical investigation.

The foregoing was all written before I saw the analysis given by
Zahn in his _Geschichte des Kanons_, i. 383 ff. Reading it, I am
surprised that his conclusions have not been followed up by a
thorough investigation of the subject long ere now. Personally, I am
precluded at this moment from even making an attempt in this
direction. Zahn says: “The quotations of Aphraates frequently
presuppose a different Greek text (of the Pauline Epistles) than
that lying at the foundation of the Peshitto. The repeated
resemblances to Western texts, Claromontanus, Boernerianus (D G),
Tertullian, and other Latin witnesses are particularly striking. In
the earliest Syriac Gospel the same phenomenon appears even more
conspicuously. How is it to be explained? Shall we suppose that this
type of text was dispersed equally throughout _all_ parts of the
Church during the second century? In that case we should have to
regard it as the earliest form at which we can arrive, on the
principle laid down by Tertullian: quod apud multos unum invenitur,
non est erratum sed traditum. But,” says Zahn, “even those who
venerate the Western tradition of the text—_i.e._ those who, like
myself, are of opinion that it does not get nearly its due share of
attention from present-day critics—will decline to assent to this
proposition. Because the result of this view would be to establish
the rule that the so-called Western tradition _invariably_ deserves
the preference over those others, even over our oldest Greek
manuscripts themselves. Even if we limited it to those elements of
the text in which the furthest East agrees with the furthest West,
the result would still be a text to which no cautious critic would
pin his faith. A more natural explanation of this striking condition
of things is required.” Zahn finds this in the supposition that
there was formerly a close intimacy between the Syrian Church and
Rome. “Just as the Princes of Edessa had much direct intercourse
with Rome, so to all appearance had the Church there.” In proof of
this, he points to the early intrusion of Marcion’s doctrines and
Bible into Mesopotamia, to the participation of the Church of Edessa
in the Easter controversy and its agreement in that matter with
Victor of Rome, and to the Abgar Legend which connects Edessa with
Zephyrinus of Rome (199-216) by way of Antioch, and represents Peter
as sending the Epistles of Paul from Rome to Edessa. “Considering
the anachronisms that legends usually exhibit, may we not take this
to be the expression of an historical fact, viz. that a text written
in the West formed the basis of the earliest Syriac version of the
Pauline Epistles? This supposition is confirmed by the earliest
history of the Gospel among the Syrians—viz. by the
Diatessaron.”[234] After a most thorough discussion of all the
questions relating to that book (pp. 387-422), Zahn discovers in
this part also (the Gospels) an intimate connection between the text
on which it is based, and the form assumed by the text of the
Gospels in the West during the second century. And he believes that
it will be difficult to find a more feasible explanation of the
remarkable agreement evidenced by these two texts in the very matter
of their textual corruption and licence than this, that this text
came from Rome to Syria. And so the final question arises, “whether
a connection does not exist between the first Gospel and the first
text of Paul and the Acts in the Syriac, and whether the _entire_ N.
T., as the _Doctrine of Addai_ says, was not a present which Tatian
brought with him from Rome to his countrymen, and adapted for their
use by means of a free translation and revision?” Zahn thinks that a
positive answer cannot be given, but he refers pointedly to what
Eusebius says regarding Tatian’s treatment of the Pauline Epistles,
and is led to suppose that those changes were introduced on the
occasion and in the form of a translation from the Greek into
Syriac, and that the reason why Eusebius had such hazy notions
regarding it as well as the Diatessaron, is most likely that both
the books were in Syriac, and used only in the Syrian Church. A
closer investigation of the Pauline Epistles in the Syriac is needed
to decide these questions.

To these propositions of Zahn I have but the one objection stated
above, that the expressions used by Eusebius point far too plainly
to a revision of the phraseology of the Pauline Epistles, which
could have been done only on the original Greek.[235] Zahn himself
points out that the words of Eusebius remind us of what is elsewhere
said of the Theodotians (_Eccl. Hist._, v. 28, 15. 18; see above, p.
200).[236]

[Sidenote: Western Text.]

(_i_) =The Western Text.=—We thus find ourselves face to face with what has been called the only burning question of the textual criticism of the New Testament—the question, namely, of the place to be assigned to the so-called Western text. Our treatment of the external testimony has led us back through Lucian, Pamphilus, Hesychius, and Origen, to Marcion and Tatian, that is, into the middle of the second century. But the question is now whether we must stop here, or whether it is not possible to ascend with certainty even somewhat higher by means of an investigation of the material afforded by the manuscripts themselves. The “Higher Criticism,” _e.g._, seeks to get behind the Synoptic Gospels to the documents which the authors or editors used in their composition; is it not possible for the “Lower Criticism” to recover with certainty at least the primitive text of the New Testament books? And is that not most readily found in the so-called Western text? We have been obliged to make frequent reference to it already; the question for us now is, “What is the value of Codex Bezae and its associates?”

It was observed by Theodore Beza himself, the scholar whose name the Codex justly bears, that the text of this manuscript differed in so many respects from that of others, especially in Luke and Acts, that he could give no explanation of it satisfactory to himself. He was not led to suppose that the alterations were due to heretics; nevertheless, like a cautious man, he thought it more advisable to preserve the Codex than to publish it. Eight hundred years before, Bede was similarly impressed by the Codex which we now know by the name of Laudianus, E_{2}. He indicated “quaedam quae in Graeco sive aliter seu plus aut minus posita vidimus.” He was uncertain “utrum negligentia interpretis omissa vel aliter dicta, an incuria librariorum sint depravata sive relicta ... namque graecum exemplar fuisse falsatum suspicari non audeo.” When the manuscripts began to be more systematically collated, Bengel declared that the criticism of the text would be much simplified if one were not bound to trouble himself with these codices, which, as being written in Greek and Latin, he called _vere bilingues_. Old students of the Maulbronn College have told me that Ephorus Bäumlein the most distinguished philologist of our Institute in this century, and editor of _Disquisitions on the Greek Particles_ and similar works, was always referring to the Codex Cantabrigiensis, though they themselves never rightly understood about this Codex, or indeed about such things at all. I do not know who it was from whom I myself first heard of it; certainly there was no particular importance attributed to it in my student days or at the college where I was. On the other hand, Tischendorf admitted its claims in opposition to all the other Greek manuscripts in several passages, such as Mark ii. 22; xi. 6; Luke xxiv. 52, 53, etc. In other places he did so at first, but changed his opinion afterwards—_e.g._, in Acts xi. 12, while in others again, like Acts xiii. 45, he was inclined to accept its testimony, asserting expressly: Ceterum D quantopere passim inter omnes testes excellat constat. One of the things for which Westcott and Hort deserve credit is the attention they have directed to Codex Bezae and its associates. Some of their remarks upon it will be found in the note below.[237]

[Sidenote: Lagarde.]

That peerless scholar, P. de Lagarde, has even greater claims to honourable mention in this connection, though but little regard was paid to his representations during his lifetime. As early as 1857, he said of Codex Cantabrigiensis: facile patet, quum similibus libris careamus et ultra Evangelia et Actus nondum cogitem, totius editionis meae quasi fundamentum futurum esse hunc codicem Cantabrigiensem, sed eum eis librarii vitiis purgatum quae vitia esse agnita fuerint (_Gesam. Abh._, p. 98). His chief merit, however, lies not in his having estimated Cantabrigiensis so highly, but in having assigned a lower value to the other manuscripts. By comparing D with the earlier versions, and particularly by relying on the testimony of Epiphanius, he recognised in it a representative of an “editio emendatorum orthodoxorum temeritate corrupta” (_ibid._, p. 96). Compare also his _Übersicht über die Bildung der Nomina_, p. 213, where he instances ἔταξαν ἀναβαίνειν of the “emendati” for παρήγγειλαν ἀναβαίνειν of D.[238]

[Sidenote: Blass.]

General attention was first directed to the question of the Western text, when Blass came forward with his view that it was quite wrong to present the problem in the shape of an alternative between D and A B, because both groups were right, D and its associates representing a first edition of the Acts and a second of Luke, and the other group conversely. I hailed this solution of the difficulty at once as a veritable Columbus Egg, and to this day I am firmly persuaded that Blass’s theory is nearer the mark than the previous estimate of the Western text. [Sidenote: Zahn.] Readers may, perhaps, be struck by the fact, which Zahn has since made public (_Einl._, ii. 348), that in his practical class at Erlangen, in the winter of 1885-6, he set as the subject of the prize essay an “Investigation of the materially important peculiarities of Codex D in the Acts,” and made a note at the time of the result which he hoped to see the investigation arrive at—viz., “either the author’s first draft before publication or his hand copy with his own marginal notes inserted afterwards.” Zahn himself got no further, but he was not surprised when Blass came forward with his clearly-defined and thoroughly-elaborated hypothesis. In one point, certainly, Zahn does not agree with Blass, and that is in the application of the hypothesis to Luke. He holds that the text which Blass restored as the Roman form or second edition of Luke is essentially nothing but a bold attempt to restore what is called the Western text; that the question to which such different answers have been made as to the value of this type of text—for it is not to be spoken of as a recension in the proper sense of that term—is by no means confined to the Third Gospel, but touches the others as well and the Pauline Epistles also; that the reason why the divergence of the Western text from that exhibited in the oldest manuscripts and the great majority of Greek witnesses is more conspicuous in Luke, is simply that we have the additional testimony of Marcion for that Gospel, but the question is essentially the same in all the cases; that whereas in Acts we have two parallel texts both possessing equal authority, in Luke the case is different, where in determining what the evangelist actually wrote, we have to choose one or the other of two mutually exclusive propositions; that this verdict on the text of Luke, however, in no way invalidates the conclusion come to as regards the text of the Acts. But further, Zahn, who even before this had avowed himself an “admirer of the Western text,” stands up determinedly for the view that this same Western text, which I shall, like Zahn and Blass, indicate henceforward as β, contains much that is original. He says that just as we must beware of a superstitious idolatry of what are styled the best manuscripts,[239] which goes hand in hand with a disparagement of much older tradition (Marcion, Irenæus), so we have equally to be on our guard against a morbid preference for every interesting and fanciful excrescence of the riotous tradition of the second and third centuries. Such a preference would logically imply that the scholars who took in hand to revise the text about the beginning of the fourth century simply corrupted it, somewhat after the fashion of those who set themselves to “improve” our Church hymns in the age of Rationalism. More than twenty years ago, when I was a Tutor at Tübingen, I had the impression to which I frequently enough gave utterance in debate with my colleagues, that modern textual criticism is going altogether on the wrong tack. The textual study of the New Testament was out of my province at that time, and is really so still, were it not that, as Augustine says, it is necessary for everyone who devotes himself to the holy Scriptures to take up such studies. Nor am I inclined thus far to fall foul of the system to which Westcott and Hort devoted the labours of a lifetime, and in the building up of which they had at their command such an apparatus as is far beyond the reach of a German, especially of one who is not attached to any University. And as for the results of Zahn’s researches, I prefer to look upon myself here as a mere learner and admirer. In the presence of such doughty warriors I feel like a spectator upon the battlefield of New Testament textual criticism, and I would beg that what follows, as well as what has been already said, be regarded as but suggestions, the acceptance or rejection of which by others may perchance serve to bring a younger generation nearer to the goal. In this spirit I have in my _Philologica Sacra_ (1-15th March 1896) taken as a starting-point the reading in Luke xxii. 52, λαοῦ = ναοῦ = ἱεροῦ, which is not mentioned at all by Tischendorf, and have sought by means of one or two analogous cases to show “how frequently D preserves the correct reading.” I have instanced ἑπταπλασίονα, Luke xviii. 30; φάντασμα, xxiv. 37; δέρριν καμήλου, Mark i. 6 ; ἠνοιγμένους, i. 10, which might, however, be inserted from Matt. iii. 16, Luke iii. 21 ; ὀργισθείς, Mark i. 40; ὁμοιάζει, Matt. xxvi. 73.

In the first sketch of this Introduction, written in the year 1895, I referred to the addition found in Matt. xxvii. 49, which is manifestly taken from John xix. 34, and is read by many authorities, among these being א B C.[240] I said then: “Only two possibilities are conceivable. Either the passage stood here originally, and was removed at an early date on account of its variance with John xix. 34, or it is an interpolation. In the latter case, it must have been inserted at a very early date, and all the witnesses containing it, which elsewhere are so frequently and so widely divergent, must then go back to one and the same exemplar. Because the third possibility—viz. that the same sentence was inserted in different copies in the same place quite independently of each other, no one will consider to be at all likely. But if the second supposition is to be held as correct, then we see just what amount of importance is to be attached to the concurrence of our oldest witnesses, particularly our chief manuscripts א B C L. They are not streams flowing independently from the same fountain of Paradise: they flowed together for a good part of their course, and were considerably polluted before they parted company.”

[Sidenote: Salmon.]

Two years later, when the first edition was issued, I added: “This too must now be asserted with far greater emphasis, that the concurrence of B א, on which so much stress has been laid hitherto by almost all textual critics, proves nothing at all. In Sirach the common archetype of B א was younger than the origin of the Latin version, manifestly a good deal younger, because it already contained errors that had not yet made their appearance in our other manuscripts (or in their sources). Salmon (p. 52) is of opinion that the text which Westcott and Hort have restored is one that was most in favour in Alexandria in the third century, and that came there, perhaps, in the century previous. This is not far from Bousset’s view that B perhaps contains the recension of Hesychius. Salmon calls the results of Westcott and Hort ‘an elaborate locking of the stable door after the horse has been stolen.’ Burgon’s paradox, that the reason why B and א have survived is that they were the worst, seemed to Salmon at first to be a joke, but he now thinks it not improbable that they were set aside on account of their divergence from the form of text that acquired ascendancy _at a later time_. If that be so, then they met the same fate that they themselves prepared for the primitive form they supplanted; and just as they, with the help of Tischendorf and Westcott and Hort, dislodged the _Textus Receptus_ of the sixteenth and seventeenth centuries from the hands of theologians, and made themselves the _Textus Receptus_ of the end of the nineteenth century, so perchance will Codex D, which the builders despised, become the foundation-stone of a new structure. In _Urtext_, p. 54, Oscar v. Gebhardt alludes to the objections raised partly against the entire method of Westcott and Hort, partly against their particular estimate of Codex Vaticanus, and partly also against the position they have assigned to what they call the Western text, and he too says: ‘If these objections are valid, then the sure foundation which they seemed at last to have secured for the text of the New Testament begins once more to totter.’”

Since this was written, my impressions have been greatly confirmed, particularly by Zahn’s _Einleitung_; only I must admit that I am now less in a position than ever to make any definite proposals as to the way in which the goal of the textual criticism of the New Testament is to be reached. To follow one witness or one group of witnesses through thick and thin, which would really be the only consistent course, will seemingly not do.[241] And the “eclectic method” to which Bousset was led in his work on the Apocalypse as the only possible one, is surely the opposite of the genealogical, which we must acknowledge to be in theory the only correct method. But first of all, a fresh application of it would require to be made. And as the task is too great for any single worker, might it not be well if, in the exegetical classes of our Theological Faculties, the separate witnesses were either examined anew, or, conversely, selected passages of the text, quite small passages—a single chapter, or a single epistle like 2 or 3 John or Philemon—were given out to different students to examine thoroughly all the witnesses for each passage, and the results then compared with one another? Furthermore, the critical apparatus would require at once to be lightened of all those manuscripts which are unmistakably recognised to be the representatives of a definite recension, and the Lucianic recension printed separately with or without an apparatus, just as was done by Lagarde himself for half of the Old Testament. Finally, the Western text would require to be much more thoroughly examined than has hitherto been the case. It is true that Weiss has given a special part of _Texte und Untersuchungen_ to an examination of Codex D in Acts, but without prejudice one may be quite sure that a solution of the problem is not to be found in the way in which Weiss seeks it. No doubt he establishes among other things the fact, that in the Speeches of Peter β displays almost no variation, but then he makes no attempt to explain this fact or make any use of it. It is an indication of considerable progress to find Zahn going so carefully into matters of the text in an Introduction to the New Testament, and his appreciation of the Western text is most gratifying. [Sidenote: Luke and Acts in D.] At the same time the reader will naturally ask whether Zahn’s verdict on the β text in Luke is not fatal to his own conclusions with regard to Acts. Is it not true in this connection that “he who says A must also say B”? If you admit that there were two editions of Acts, you must make the same admission in the case of Luke. And conversely, if there was no second edition of the Gospel, must you not then look for some other explanation of the variations in Acts? For it seems quite certain that the variants in Luke xxiv. are most closely related to the text of Acts i. Or how else are the readings in Luke xxiv. 51-53 to be explained? Westcott and Hort have one way of explaining them. They say that καὶ ἀνεφέρετο εἰς τὸν οὐρανόν “was evidently inserted from an assumption that a separation from the disciples at the close of a Gospel must be the Ascension. The Ascension apparently did not lie within the proper scope of the Gospels, as seen in their genuine texts; its true place was at the head of the Acts of the Apostles as the preparation for the Day of Pentecost, and thus the beginning of the history of the Church.” That is all very well, and it may also be the case that προσκυνήσαντες αὐτόν in v. 52 is the natural result of the insertion of καὶ ἀνεφέρετο εἰς τὸν οὐρανόν. But how then are we to account for the interchange of εὐλογοῦντες and αἰνοῦντες in the next verse, which is found in precisely the same groups of witnesses?[242]

If this explanation then is insufficient on account of verse 53, it may be confidently asserted that the omission of the Ascension and the Worship of the Exalted Lord by any later scribe is all but inconceivable from the moment that Luke was separated from Acts and placed among the Gospels. If such a thing were possible at all, it would be in the case of προσκυνήσαντες αὐτόν, as there is no express mention in Acts i. of the disciples worshipping. On the other hand, the omission becomes quite conceivable as soon as the author added a δεύτερος λόγος to the πρῶτος. So far these variants appear to me to fit in very well with Blass’s theory and with no other. Zahn, as far as I can see, has nowhere expressed any opinion regarding them, certainly he says nothing of the variation between αἰνοῦντες and εὐλογοῦντες, which is the one of most importance critically, though it is of least consequence materially.

Graefe, following on the lines of Birt and Rüegg, supposes that the shorter form was due to want of space, that Luke was glad to get the shorter form all into his roll at the foot the first time he wrote it out, and sent off the book to Theophilus in that form, hoping to deal with the Ascension in the second of his books. In the second edition he had sufficient space to admit of the insertion of καὶ ἀνεφέρετο εἰς τὸν οὐρανόν, then of προσκυνήσαντες αὐτόν, and finally of εὐλογοῦντες.[243] These additions he made, feeling, rightly enough, that there could be no more fitting conclusion to his Life of Jesus than a brief allusion to the Ascension, which he had already described more particularly in the Acts. At the same time he substituted ἕως εἰς for ἔξω πρὸς. Graefe thinks that all these changes are connected with the alterations made also in the introduction to the Acts, though he omits to say what the connection is.

Weiss, father and son, omit the words καὶ ἀνεφέρετο εἰς τὸν οὐρανόν as “a gloss derived from Acts i.,” and “likewise” the words προσκυνήσαντες αὐτόν in verse 52 (is this also a gloss from Acts i.?). Which text they hold to be correct in verse 53 they do not say.[244]

Αἰνεῖν[245] is the specifically Lucan word for “to praise,” while εὐλογεῖν in this sense does not occur in Acts at all, and only in the first two chapters of Luke. Further, as any concordance will show, αἰνεῖν is the regular equivalent of הלל and εὐλογεῖν of ברך, while αἰνεῖν is rarely used for ברך or εὐλογεῖν for הלל. This confirms the supposition that αἰνοῦντες, which is preferred by Tischendorf but rejected by Westcott and Hort, is the original reading.

In order to show the full extent of the difficulty of the problem, we shall take along with this passage from the end of the Gospel a single instance from the Acts. How does the Apostolic Decree read in ch. xv.? “To judge any matter before knowing the facts of the case is inadmissible.” So Hilgenfeld says in his magazine, adding that the matter of the Apostolic Council, as it is called, and the Decree have been so judged. He himself restores the whole text in this passage to the form that Blass has adopted as the _Forma Romana_—_i.e._, to confine ourselves to this point of main importance, he omits “things strangled.”[246] On the other hand, Harnack, in the article to which reference will be made below, comes to the conclusion that the Eastern, _i.e._ the common, text is the original, and the Western a later correction made subsequent to the _Didache_, and not earlier than the first decade of the second century.[247] The same conclusion is reached by Zahn in his extremely careful discussion of the question (_Einl._, ii. 344 ff.): “The two texts are here mutually exclusive, and therefore cannot both be derived from the same author (xv. 20, 29, xxi. 25).” But he immediately adds: “The fact that Blass, in this important point, as in many another of less consequence, declares a certain thing to be an original element of the text which turns out to be simply an early corruption in no wise detracts from the correctness of his hypothesis.” That is quite true and must be borne in mind in connection with the objection raised by Wendt, that “manifest clerical errors are found in the actual β text.” The passages are also used by Corssen as an argument against Blass. The remarks of the latter in reply to the strictures of Corssen (_Evang. sec. Lucam_, p. xxvi) seem to me to be not without reason, but in any case it is strange that alterations should have been made in an official document like the Decree in Acts xv., no matter whether these were due to the writer himself or a later intermediary. That there was some method in the alteration is shown by its recurrence in three places. But again I must emphasise the superiority of Codex D. Whereas in ch. xv. 20, 29 the shorter text is represented by other witnesses as well, in ch. xxi. 25 it is supported by D with the sole addition of Gigas Holmiensis.[248] I have not to decide the question here; I simply commend it to a searching investigation, in which attention must be paid to the apparently meaningless differences in the use of particles and synonyms, of simple and compound words, and such-like seeming minutiæ. I can only repeat how frequently the thought occurred to me when I was comparing Scrivener’s edition with that of v. Gebhardt for my _Supplement_, that here was no alteration of a later scribe, and what then? The simplest explanation was that both readings were due to the author himself, who on the one occasion purposely set down the one reading and on the other the other.

There is another question in connection with the Western text which has been even more neglected than the former—viz. the amount of importance to be attached to it in the case of the Pauline Epistles. What about Eusebius’s reference to Tatian’s work on these Epistles? I frankly confess that not till the printing of this work was begun did I become aware, mainly from Zahn’s _Einleitung_, how many problems are here waiting to be solved, and for this reason as well as others I must for the present forbear making any attempt in this direction.

Here I can only indicate a few of the most general rules of textual criticism, and thereafter adduce a number of New Testament passages which are of interest from a critical point of view.

[Sidenote: Rules of criticism.]

(_k_) =General Rules of Textual Criticism.=—In its essence the task of the textual critic resembles that of the physician, who must first of all make a correct diagnosis of the disease before attempting its cure. Manifestly the first thing to do is to observe the injuries and the dangers to which a text transmitted by handwriting is liable to be exposed. A correct treatment must be preceded by a correct diagnosis.

The injuries which a text receives will vary according as it is multiplied by =Dictation= or by =Copying=. The fifty Bibles which Eusebius prepared at once for Constantine would be written to dictation. In the early times of the Church, copying, as has been already mentioned, would doubtless be the more usual method of multiplication. Here, however, we must make a single exception in the case of Paul, who for the most part did not write his Epistles with his own hand. He evidently dictated them. He certainly did not have them simply written out from his own rough draft.

[Sidenote: Illegibility.]

(1) In the case of copying, errors originate first of all, though not most frequently, in a word or group of letters being =illegible=, or in their being read for some psychological reason otherwise than as they were intended. However attentive the copyist may be, he may still be in doubt as to the way in which a word or passage should be read, and may decide wrongly. Proper names, _e.g._, are often very doubtful.[249] More frequently however the mistake will be due to inattention. The context may lead the copyist to expect a certain word; he sees one like it, and inserts the former in its place.

[Sidenote: Homoioteleuton.]

(2) It frequently happens, particularly in copying the old _scriptio continua_, that the eye of the scribe jumps from one word or group of letters to another the same or similar to it, either before or after it. In the former case the intervening words will be repeated, in the latter they will be omitted. Scholars designate these errors as =dittography= and =elision= respectively; printers know them under the name of a marriage and a funeral. The former mistake is not so serious, because it is at once detected on reading over the copy. A peep into any manuscript will show how frequently this error occurs, the repeated words being enclosed in brackets or surmounted with dots.[250] In Codex B such passages give us an opportunity of observing the beauty of the original writing, because the painstaking man who retraced the old writing with fresh ink in the eighth, or tenth, or eleventh century, or whenever it was, adding at the same time accents and punctuation marks, left these untouched. This kind of mistake very often happens in passages where a group of characters catches the eye for any reason, such as the occurrence of the abbreviation mark, Θ̅Σ̅, Ι̅Λ̅Η̅Μ̅, Α̅Ν̅Ο̅Σ̅, etc., and at the transition to a new page or leaf. The omission of a piece of the text of various length by homoioteleuton is as common, and is more serious.[251] Any critical apparatus will show the frequency of its occurrence. We often find there the note “a voce alterutra ... ad alterutram desunt” or “a voce 1^o ad vocem 2^o (3^o) transilit,” or “vox ... alterutra et intermedia desunt.” Compare, _e.g._, Codex D, Matt. xviii. 18 from γῆς to γῆς; x. 23; xxiii. 14-16. The result is worst when the mistake is not discovered till afterwards and the two fragments are patched together in some way with more or less success. Lacunæ that have not been doctored are very helpful in determining the relationship between different texts.[252]

[Sidenote: Confusion.]

(3) As errors of the tongue and the memory[253] rather than of the eye may be reckoned the =Transposition= and =Confusion= of particular combinations of letters or entire words. The former occurs so frequently in connection with a liquid, that in some cases it ceases to be a mistake. Thus we have on the one hand the confusion of κορκοδειλος with κροκοδειλος, Καρχηδων with Carthago, and on the other, εβαλον with ελαβον, βηθαραβα with βηθαβαρα, John i. 28; κιρνατω with κρινετω, ποντον with τοπον, בכרין, _talent_, Matt. xxv. 14-30, with כרכין, _cities_, Luke xix. 17, 19. Akin to this is the confusion of vowels with a similar sound, to which are to be ascribed all cases of =itacism=, as it is called—ἔγειρε and ἔγειραι, —εσθε and —εσθαι, ἑταίροις and ἑτέροις, χρηστὸς ὁ κύριος and χριστὸς ὁ κύριος, 2 Cor. xii. 1; φορέσωμεν and φορέσομεν, ἔχωμεν and ἔχομεν, Rom. v. 1; μετὰ διωγμῶν and μετὰ διωγμόν. Manifestly mistakes of this sort would occur more readily in dictation than in copying.

A third class of errors of a more conscious or semi-conscious description is due to the =substitution= of words or forms of similar meaning. Thus, for λέγει we may have εἶπε or ἔφη or ἀπεκρίνατο, or the simple form may be replaced by the compound or _vice versa_, or one preposition may be substituted for another.[254] Separate words are very frequently transposed without seriously affecting the sense. Thus, in Acts iv. 12, we find nearly all the possible permutations of the three words ονομα εστιν ετερον actually represented—viz., in addition to this (2) ονομα ετερον εστιν; (3) ετερον ονομα εστιν; (4) εστιν ετερον ονομα; (5) εστιν ονομα ετερον.[255] On Luke xvii. 10, Merx says (_Die vier kanonischen Evangelien_, p. 246): “Let it be observed that the position of ἀχρεῖοι fluctuates between (1) δοῦλοι ἀχρεῖοί ἐσμεν; (2) δοῦλοί ἐσμεν ἀχρεῖοι D, and (3) ἀχρεῖοι δοῦλοί ἐσμεν. Such fluctuations are due to the different arrangement of a word that did not originally belong to the text, but was appended as a note and afterwards incorporated with the text. Such fluctuations point to the interpolation of the fluctuating word.” This judgment has to be accepted with caution. For one thing, it is not at all clear which word it is that fluctuates. In this particular case, one might say that δοῦλοι fluctuates as much as ἀχρεῖοι, and the copula still more. Moreover such an interpolation becomes at once an integral part of the text, and its insertion is no longer visible. Only if several copies were made of that exemplar in which the interpolation was first introduced could fluctuation of this sort originate. Such transpositions are much more frequently of a harmless order, as each one may perceive for himself. The writer’s thoughts fly faster than his pen and anticipate a word that should not come in till later. One of the most frequent cases of transposition is that of Ἰησοῦς Χριστός and Χριστὸς Ἰησοῦς in the Pauline Epistles.

[Sidenote: Additions.]

(4) Akin to this last is a class of mistakes originating in the border region between the unconscious and the conscious or intentional—viz. that of =Additions=. One can readily understand how easy it was to insert a κύριος or ὁ κύριος ἡμῶν, a μου after πατήρ on the lips of Jesus, the subject at the beginning of a sentence, especially of the first sentence of a pericope, or the object in the form of a pronoun. Bengel proposed to omit the name of Jesus in some twenty-five places, for which he was ridiculed by Wettstein, as may be learned from my work on Bengel, p. 74. Now, everyone admits that Bengel was right. Under the head of “Interpolationes breviores,” Wordsworth and White first give examples “de nomine Jesus,” then of “Christus, Dominus, Deus,” and then of “Pronomina.” It is evident that in this way the wrong word may be supplied now and again. Perhaps one of the most interesting cases is Luke i. 46. All our present Greek witnesses make Mary the composer of the Magnificat, but Elisabeth’s name is attached to it in three Old Latin manuscripts, in the Latin version of Irenæus, according to the best manuscripts, and in some manuscripts known to Origen (or to his translator, Jerome: the passage, unfortunately, is found as yet only in the Latin).[256]

[Sidenote: Corrections.]

(5) To the category of conscious alterations belong first of all =grammatical corrections=, then =assimilations= to parallel passages, =liturgical changes= introduced from the Evangeliaria, as, _e.g._, the addition at the close of a pericope of the words ὁ ἔχων ὦτα ἀκουέτω which occurs in all sorts of manuscripts in the most diverse passages, or indications of time, such as ἐν τῷ καιρῷ ἐκείνῳ at the beginning of a pericope,[257] and lastly, alterations made for =dogmatic= reasons, if any such can be established. It is impossible to deny that dogmatic conceptions had some influence on the propagation of certain readings if not on their origin—as, _e.g._, on the form assumed by the words in Matt. xix. 17, τί με λέγεις ἀγαθόν, or on the omission of the words οὐδὲ ὁ υἱός in Mark xiii. 22; compare also above, p. 106. On the whole, however, there is no real ground for the scepticism that was for a time entertained with respect to our texts in this connection. A sober criticism will be able in most cases to restore the correct form. Its conditions will be apparent from what has been said in the foregoing.

[Sidenote: Canons of Criticism.]

Gerhard von Maestricht laid down forty-three Critical Canons, and Wettstein set forth in his New Testament his _Animadversiones et Cautiones ad examen variarum lectionum Novi Testamenti necessariae_ (vol. ii. 851-874). In 1755 J. D. Michaelis added to his _Curae in versionem Syriacam Act. Apost._ his _Consectaria critica de ... usu versionis Syriacae tabularum Novi Foederis_.[258] Bengel reduced all the rules to a single one. Quite recently Wordsworth and White comprehended the rules they followed in the preparation of the text of their Latin New Testament in four sentences. Of these the first two apply to a version only, and therefore do not concern us here;[259] while the fourth (_brevior lectio probabilior_) is but another form of Bengel’s canon. The third alone may be regarded as new and deserving of attention—viz., _vera lectio ad finem victoriam reportat_. That is to say, if a phrase is repeated in several passages in the same or similar terms, and displays variants in the earlier passages, the reading of the later passage will, as a rule, be the correct one, the reason being that copyists are apt to consider a certain reading to be an error the first time it occurs, and therefore to alter it, but come in the end to admit it as correct.

I would once more briefly emphasize the following propositions:—

(1) The text of our manuscripts must not be regarded as homogeneous,
but must be examined separately for each part of the New Testament.
A manuscript that exhibits a very good text in one book does not
necessarily do the same in the others. The same thing holds good of
versions and quotations.