The First Written Gospel.

Results of Some of the Recent Investigations.

By Lester Bradner, Jr., Ph.D., Berlin.

Taken from THE BIBLICAL WORLD, Volume 1, Issue 6.


Students of the synoptic problem are familiar with the fact that the theory proposed by Weisse in 1838, that an extra-canonical writing, the “Logia” of the apostle Matthew mentioned by Papias, and the Gospel of Mark are the sources of the other two gospels, has been gradually gaining ground until it is now acknowledged with slight variations by nearly all the foremost New Testament scholars. The attainment of unity of opinion, however, on this one point leaves several important questions still to be discussed.1

As respects the relation of these two sources the view of Bernhard Weiss that Mark’s Gospel is a secondary written source based upon the Logia as primary, seems to be gaining ground. As concerns date, Weiss puts this, " The oldest gospel,” between 64 and 66 A. D.2 If it should be established by later research that the writing was know to Paul, and that references to it are found in his epistles, as Marshall and Resch promise to show, then of course the terminal date of the original must be set perhaps a decade earlier. But that it should have been written much earlier than this latter date is quite unlikely. With reference to its extent there is not as yet perfect unanimity. Weiss maintains that the discourses of the Logia were not thrown together without definite order, or in such extensive combinations as we find in Matthew, but were fitted into a brief historical framework, not strictly chronological and yet, sufficient to indicate to the reader or hearer under what circumstances the words were spoken. And farther, the arrangement and sequence of material seem to have been controlled by a desire to follow somewhat the historical course of Jesus’ ministry, beginning with an introduction concerning the preaching of John the Baptist. Then, touching on Jesus’ baptism and temptation, it gave at some length the Sermon on the Mount. Then through a series of accounts of both miracles and discourses, it proceeds as far as the sayings concerning the second coming, and closes, according to Weiss, with the account of the anointing in Bethany, the prophecy, on this occasion, of Jesus’ death supplying in a meagre way the place of the story of the Passion. Nevertheless, the possibility that the close of Christ’s career was narrated is not excluded.3 Wendt, in his attempt at a reconstruction of the Logia, in the first volume of his Lehre Jesu, dissents from quite such an extensive view of the contents of the Logia, and confines himself more to passages which have no verbal parallels in Mark.4 On the other hand he considers the whole of Luke’s long insertion 9: 51-18:14 as drawn from the Logia, whereas Weiss would reserve portions of it not paralleled by Matthew for the third Luke source. Marshall would include in the Logia several paragraphs additional to those assigned to it by Weiss. Resch supports these additions, and would include still other material from extra-canonical sources.

It being possible, therefore, by a close critical method to determine within certain limits the probable extent of this earliest document, the question what language the writer used becomes of interest. And on this point have centered the investigations of two of the most recent writers, Resch5 and Marshall.6 In times past this question has not seemed of very great difficulty or even importance. Those who believed in such a source at all were mostly content to accept the testimony of Papias that it was written Ἐβραΐδι διαλέκτῳ “in the Hebrew dialect,” and many thought that in the translations of that original made by each evangelist himself, lay the key to the variations of the synoptics in common passages. According to the latest results these men seem in the main to have followed the right track, only failing to note one or two critical steps which are essential to an exact conclusion. In the first place it should be clear from the correspondences in the Greek text of our synoptics that a first-hand translation by the evangelists themselves is out of the question unless in Mark’s case alone. The common document which lay in their hands for incorporation into a new gospel was not Hebrew but Greek. Otherwise we could not find passages agreeing so exactly in words and the order of words, even down to minute particles. Secondly, just what is meant by Papias’s expression, “in the Hebrew dialect.” For besides the true Hebrew, the Aramaic, which was closely allied and the general speech of the people in the time of Christ, also frequently goes under the name of Hebrew, by virtue of an inexact use of the term. To be sure, the knowledge of the original language is not so important a question as the extent of the document when we consider that our evangelists worked from translations; nevertheless it has relative significance from the fact that a determination of the original, and therewith the possible variations of translation, would confine within narrower and clearer limits the departures of the several evangelists from the original text; thus enabling us to classify more closely their modifications. At the same time, our interest in a more exact knowledge of the teaching and theology of the earliest church requires us to recover, if possible, the original language in which the fountain head of the triple synoptic stream first flowed from the apostle’s pen.

The above mentioned scholars, Marshall in England, and Resch in Germany, pursued their labors for the settlement of this question separately and independently until they had reached the final conclusion: Marshall, that Aramaic was the original language; Resch, that it was Hebrew proper. That opposite results have been obtained does not invalidate the process so much as it might seem, owing to the affinities of the two languages, and the interest lies in the similar methods used by the two men, and the possibility of their producing reliable results. Of Marshall’s work I shall not attempt to speak in detail as it is accessible to English readers in the volumes of The Expositor (1890-1892), and on this account I should like to give a somewhat fuller survey of Resch’s researches, and the more since his must be regarded as more comprehensive, and in this way more valuable than Marshall’s.

The work in which Resch first called attention to his results was issued in 1889 under the title "Agrapha, Extra-canonical Gospel-Fragments.”7 It is a very complete collection, comparison, and discussion of all those quotations of the sayings of Christ (Herrenworte) which have no place in our canonical gospels, but are found either in the other New Testament writings, or in apocryphal books, ancient liturgies or the patristic works. This work is intended to serve as an introductory textual study for the support of the hypothesis first stated there, but carried out more fully in his more recent book, that the original gospel was written in Hebrew; that the different translations of this original furnish a key to many of the variations of our synoptical gospels; that in extent it included all of these extra-canonical gospel fragments which may be proved genuine, and that in character it was a gospel of the type almost completely deleted by the subsequent process of canonization, and whose sole remaining representative (and this perhaps only partial) is the peculiar Codex Cantabrigiensis or Bezae (Codex D). This first publication has already been noticed in Sanday’s articles in The Expositor already referred to, so that I shall confine myself to the second, a smaller volume of 16o pages, bearing the general title “Extra-canonical Parallel-Texts to the Gospels.” But this title belongs to the whole of a larger work of which the present part is only the “Critical Basis as to Texts and Sources,” and to be followed by treatments of the Gospels of the Infancy, the Gospel of John, the three synoptics and Acts. The whole is then to be succeeded by a work on the “Parallels of the canonical gospels.” This program of an investigation so comprehensive and yet so compact, involving many years of critical work, ought to be welcomed by every lover of New Testament scholarship, and the results, whatever position may be taken as to their main drift, cannot fail to be an important addition to the history and equipment of gospel criticism.

The particular part of this plan now under discussion treats in its first few chapters of the Canon. Its formation is shown to be the work and the necessary and rightful work of the church. Three epochs of Canon criticism are distinguished: a) the time of its formation; b) the age of the Reformation; c) the present historical effort to determine the exact process of canonization, and to reach behind it to pre-canonical writings. As to the Canon of the gospels, three epochs are again noted: a) collection of the traditional gospels into a quadruple gospel canon; b) this gospel canon attains sole authority in the church; c) the fixing and final purification of the text of the canonical gospels. The date, 140 A. D., is regarded as the latest possible for the termination of the first step. The second cannot have taken place before Irenaeus’s time, nor after Origen’s. In the third epoch three steps can be traced: a) the recension which produces the archetype of the Codex Bezae, the Syrian text of Cureton, and the old Latin versions, as well as a part of the oriental (1st half of the 2d Cent.); b) the work of Origen; and c) the final fixing of the text at the closing of the canon. Chapter or paragraph 3 discusses canonical and extra-canonical texts, showing the need of investigation of the latter as the occasion for the present and preceding volumes, and the following chapter groups together the sources whence a collection of these texts which have escaped the conforming zeal of the recensors may be gathered. They are in brief: (1) The Greek Codex Bezx; (2) The old Latin versions; (3) The old oriental, especially Syriac versions; (4) Tatian’s Diatessaron; (5) The patristic quotations from the gospels; (6) The New Testament Apocrypha and Pseudepigrapha; (7) The Liturgies of the primitive church.8 The question of what is to be regarded as the final source of a collection of extra-canonical texts so rich and extensive yet retaining withal so close a connection with the text of the synoptics, forms a transition to the consideration, in chaps. 5-8, of the pre-canonical gospel. As to critical principles with regard to the synoptic problem, the author upholds the position of Professor Weiss, whom he considers the surest guide in this territory despite the more recent work of Wendt, Ewald, Feine and Mandel. Bousset9 he thinks, has made some advances upon Weiss by using the extra-canonical texts as well in his investigations. Resch’s statement of his principles is then as follows: (1) The priority of Mark. (2) The existence of a pre-canonical source. (3) The “two-source” theory (for Matthew and Luke ).10 (4) The secondary character of the first canonical gospel, (5) and of the third, (6) as well as the second. (7) Various translations of the pre-canonical gospel. (8) The very early use of this gospel (reaching back to Paul). (9) Its later influence (upon patristic quotations and readings in the gospels such as the Codex Bezx contains). A number of analogies for redactional processes are then pointed out, such as the Apostolical Constitutions, the Clementine Homilies and Recognitions, and the Sybilline Oracles, as well as the Acta Pilati and examples in the Old Testament. Thus fortified as to the justification of his method, the writer enumerates the sources which are to furnish the contents of this pre-canonical gospel. They are: (1) All parallels between Luke and Matthew where Mark is lacking. (2) The discourse material (including parables) which Matthew alone furnishes. (3) That of the same category peculiar to Luke. (4) Many passages of narrative and discourse common to all three (especially those designated by Weiss). (5) A number of genuine “Agrapha.” (6) Parallels in other New Testament books to the synoptic texts. As subsidiary indications come in: (7) All texts pointing by their Hebraisms to a Hebrew original. (8) Such varying expressions of the synoptics as can be shown to be different translations of a common Hebrew original. (9) Variant readings of like character in the Codex Bezae. The only departure shown in this list from the position taken by Weiss is the addition of material from extra-canonical sources and from other parts of the New Testament. The latter is a point of much importance since it involves the acquaintance of Paul with this first written gospel and his use of quotations from it, as well as quotations in the other epistles and the Apocalypse. Resch supports the hypothesis by exhibiting parallel quotations between the epistles and synoptic gospels as follows: Romans and I Corinthians, ten each; 2 Corinthians, two; Galatians, four; Ephesians and Philippians, five each; Colossians, one; 1 Thessalonians, five; 1 Timothy, three; 2 Timothy, two; 1 Peter and James, three each; and Apocalypse, nine; making in all sixty- two cases of parallel texts.11

Chapter 7 takes up the discussion as to the language of this original gospel. Having already (Agrapha § 6.) stated and supported the belief that this was the Biblical Hebrew, the author advances a few additional points in favor of that position. He maintains that a literary use of the Hebrew was still kept up in apostolic times in Palestine, alongside the Aramaic and Greek spoken by the common people, and therefore it was possible that the author of the earliest gospel should have chosen any one of the three. The actual choice would depend somewhat on his purpose. A point of considerable weight is the following: “If
he chose the learned’ Hebraic idiom, his work obtained in advance an esoteric character which would best explain its disappearance from view in the old Catholic Church.” It is in fact remarkable that we have not more external evidence of the existence and use of such an important document, and it seems as if an Aramaic original would have had a wider circulation and more mention. On this ground therefore the Aramaic hypothesis seems less fitted to explain the facts. In this field Resch must of course meet the arguments of Marshall who favors the Aramaic. This he does partly by showing that an Aramaic original contradicts the direct testimony of Jerome and Epiphanius that the language was Hebrew and not Syro-Chaldaic ( = Aramaic), and partly by arguing that the few Aramaic elements at present found in the synoptic text, proper names, nouns and sentences, do not necessarily require the hypothesis of an Aramaic original. The proper names and nouns are not sufficient evidence, and as for the sentences, two of them (ἐφφαθά  Mark 7:34, ταλιθάκούμ Mark 5:41) probably do not belong to the pre-canonical gospel while the words from the cross (ἐλωΐ ἐλωΐ, λεμὰ σαβαχθανεί Matt. 27:46; Mark 15: 34) show evidence in the Western text of having been handed down in Hebrew rather than the Aramaic. At the same time Resch admits that the question must rest maimly on the evidence as to which dialect best explains the divergencies of the synoptic parallels. He criticizes Marshall’s work, however, as confined too closely to the canonical text, without a comparison of patristic and extra-canonical parallels, and as being based not so much upon an analysis of the discourses, where the evidence of a source is clearest, as upon the more treacherous ground of the narrative passages, where the redaction is always a greater disturbing factor. As to Marshall’s effort to explain a large number of divergencies by the assumption of such accidental changes and corruption in the Aramaic text as the confusion of similar consonants, or the omission or transposition of two adjacent letters, or the different vocalization of the same consonants, all cases which have their exact parallels in the relation of the Septuagint to the Hebrew text of the Old Testament, Resch admits that such mistakes may have been made, even adds a few examples of his own, but considers that Marshall has carried this method entirely too far. Besides, he rightly argues that there is not the same latitude for texual corruption in MSS. only a few decades old as in the case of the Old Testament books, where the transmission is reckoned by centuries. So the author decides to remain by the Hebrew hypothesis, in accordance with the literal testimony of tradition, and believes that in all necessary and likely cases given by Marshall, a Hebrew root will explain the phenomena equally as well as an Aramaic.12

In chapter 8, the author takes a farther step and proceeds to deal with the Greek translations of this original Hebrew gospel. Recurring to the various versions of the Old Testament, he notes some general characteristics which mark different styles of translation: e.g. the careful literal tendency, the freer and yet faithful reproduction of an original, and the loose paraphrasing style. He then asks whether any such varying types can be discovered in the translations used by our evangelists, and what evidence remains of various translations. The answer is found in the comparison of our synoptical gospels with each other and with extra-canonical parallels. Mark, as the first redactor of our original gospel, may be regarded as pursuing a paraphrastic and eclectic method. He is, in truth, the interpreter (ἐρμηνεύτης), and his gospel is a sort of Targum on the primitive work. Consequently his type of translation is the least reliable source for a reconstruction of the Logia text. It is even possible that we have traces of his use of two different translations, which Marshall thinks may explain his pleonasms, but no stress is laid on this point. Quite a different character belongs to the version used by our first evangelist,’ in that it is so strongly marked by Hebraisms. That he was well versed in Hebrew, there is no doubt, and we might imagine that he had made his own translation except that the Hebraisms seem to point to an older version than the type used by Paul and Luke. These three types are then displayed by several tables showing the various renderings of the same Hebrew word, while another table of some length collates the resemblances between the Lucanian type and the parallels in the Pauline epistles. Luke, then, owing to his similarities in some cases and his divergencies in others with respect to Matthew’s reproduction of the Logia, is credited with the use of two different types of translations, one of which is the Pauline-Lucanian and the other the Matthaean. It is not, however, strictly necessary that these variant types should each have covered the whole extent of the Logia.

The author now enters the field of extra-canonical parallels, and shows by a number of tables that the same variety of translation of a common Hebrew word exists among them as in the gospels. In the course of this investigation, Resch discovers a surprising resemblance between the changes made in the evident recension of the Gospel of Luke in the Codex Bezx, and the Matthaean type of Logia-translation which leads him to regard it as trustworthy testimony to a type of translation independent of and yet closely related to the Matthaean. A second point of importance is the establishment of still another type called the Alexandrian. This is maintained on the ground of the uniform predominance of certain extra-canonical readings in Clement of Alexandria, and his followers, Origen and Macarius, as well as in Alexandrian gospel-fragments, the Egyptian gospel and the Faijum fragment. Especially to be noted here is the fact that among these variants are to found such as could only have arisen from a conscious translation of the Hebrew text, as for example, Matt. 18:3, where for στρέφεσθαι καὶ γίνεσθαι a harsh literal translation of the Hebrew idiom shudh wehaya [h], Clement reads αὖθις γίνεσθαι. The survey closes with the hope that more work will be spent in noting Hebraisms in the various Greek versions, and in the endeavor to reconstruct the pre-canonical gospel by translations of its probable contents back into Hebrew.

On the whole, among the various topics treated, four appear to be of leading importance, namely: the value of the family of “Western Texts,” the use made of canonical and extra-canonical parallel texts, the language of the original gospel, and its translations with their types. A judgment in the matter of New Testament texts will not be ventured upon by the present reviewer. The opinion even of such eminent critics as Westcott and Hort seems somewhat divided. Some of the Western readings they hesitate to reject, a few are considered superior to non-Western readings, but the general criticism is made that the Western text is on the whole less pure and trustworthy. Its characteristics are said to be a love of paraphrase, assimilation of clauses for the purpose of harmonizing, and its readiness to adopt alterations or additions from extraneous sources (Cf. Westcott and Hort, Greek Test.) If Resch can point out a trustworthy origin for these variations by means of the different translations of a common Hebrew text, it would certainly relieve this family of texts from much unjust criticism; and it would perhaps be an advantage to think that the less pleasing elements of an old text had been subsequently pruned out, than that they had been introduced arbitrarily into a text originally pure. At least it may be said that Resch is not entirely alone in his position as regards the the value of these texts, while Westcott and Hort probably hold to the more conservative side.

To fail to make use of the extra-canonical as well as canonical parallels to synoptic passages would certainly be a mistake, and even if all of Resch’s conclusions on this matter cannot be made good, he has rendered a great service in the collection of the material, and in showing at least the possible limits of its application. And yet there is always the danger of becoming mechanical in the effort to prove too much, and of losing sight, as Sanday notes in his criticism of Resch’s previous work, of the latitude which must be granted to each evangelist for his individuality of style, and for the use of synonyms. And this must be especially true of an age when the letter of the sacred text was of much less importance than its meaning. Not that this excludes the possibility or probability of such a theory as Resch advances, but that it must render its application at times precarious. Sanday remarks (in the article in The Expositor previously referred to) that there are always two unprovable steps with respect to parallel texts in the epistles: a) that they come from a written gospel, and b) that this gospel was the Logia. The first objection may perhaps be set aside in case the parallel in the language is sufficiently close, and the second can be controlled somewhat through the analysis itself of the synoptics which constructs the extent Of the source. But it is certainly a question deserving consideration how far we may rightly extend the compass of this original gospel. Resch’s tendency would be to make it include the greatest possible amount, especially when it comes to the identification with it of all the genuine Agrapha. On the other hand, the additional material secured by each of the two later synoptics and the claim of the third evangelist that he was superseding both in exactness and extent, a number of existing works on the life of Christ should make us wary of pressing too much material upon this earliest source, and failing to allow that oral tradition may have carried down many sayings of Jesus which the evangelists have not included in their accounts.

The original language of the Logia is a point to be settled by experts in Semitic dialects, and by a patient testing of the relative applicability of each language to the various points of the problem. We are, I think, to assume that neither Resch nor Marshall has offered the whole of his evidence. And while one may be inclined to prefer the completeness and apparent harmony of Resch’s results to the occasional “tours de force” and necessary supposition of textual corruption which belong to Marshall’s hypothesis, a premature judgment would be unfair. The matter must be appealed, as Marshall himself says, to the general consensus of scholarly opinion. Finally, as to various types of translation, if the theory can be sustained without infringing too largely upon the individualities of the various writers, it would have the advantage of relieving us from many arbitrary and complicated theories of redaction, while the attainment of an approximate original would prove a great advance in securing an accurate historical and literary knowledge of the earliest written record of Christianity. No one who is conversant with the matter will hesitate to admit the presence of the problem and its importance, and it is to be hoped that American scholars will not leave the list of contestants for the prize of its solution to be so completely filled out by German and English names as it has been in the past.


1) The reader who desires to pursue the subject further may be referred to the New Testament Introductions of Weiss and Holtzmann, the article “Gospels” by Abbott in the Encyclopedia Britannica, the interesting discussion on The Present State of the Synoptic Problem by Sanday in The Expositor for 1891, the nearly contemporary articles by Marshall in the same magazine, and a German work just published by Alfred Resch, spoken of later in this article. Mention should also be made of Bruce’s Kingdom of God, Wendt’s Lehre Jesu, and Johann Weiss’s Das Reich Gottes.

2) Sanday agrees substantially with Weiss, and thinks that the Logia could hardly have been written before 63-68 A. D.

3) Weiss remarks as to the probable extension of the source beyond the limits that can be distinctly proved: “A source which contained the words of the Baptist, the baptism and temptation of Jesus, must necessarily have had a sort of introduction; and the last piece that can be proved to belong to it, the account of the anointing, points by its very prophecy of the death of Jesus, immediately to follow, to a conclusion of his history.” 1 should be inclined to suggest the extent of the Gospel of Mark as a good criterion for the probable historical compass of the Logia account.

4) The Logia included, according to Wendt, the original of Luke 3:7-9, 16 f; 4:16- 30; 5:39; 6:20-49; 7:2-10, 18-50; 8:1-3; 9:51—18:14; 19:1-27. 37-44; 20:18; 21:34-36; 22:14-17, 26-32, 35-38, with their parallel passages in Matt.: and also the following, peculiar to Matt.: 5:14; 7:6; 9:27-30; 12:5-7; 13:24-30, 47-50, 52; 16:17 f; 17:24-27; 18:19 f; 19:10-12; 20:1-16; 23:1-12; 25:36-41.

5) Aussercanonische Paralleltexte zu den Evangelien; textkritische und quellen-kritische Grundlegungen, Leipzig, 1893.

6) Articles in The Expositor, 1890-1892.

7) Numbered v. 4. in the series of Texte und Untersuchungen, edited by Gebhardt and Hamack, in which Resch’s new work appears as X. i.

8) The first four of these sources are shown to have a peculiar relationship between themselves as all pointing back to an archetype dating not later than 140, and exhibiting a variety of the pre-canonical text whose best extant representative is the Codex Bezae. The views of J. Rendel Harris as to the “Western text” are discussed but not upheld, Resch strongly preferring Credner’s positions.

9) Die Evangeliencitate Justin des Märtyrers in ihrem Werthe für die Evangelienkritik.

10) It seems as if Professor Stanton (cf. Expositor, March 1893,) were justified in his objection to the use of this name as characterizing the Weiss theory. The solution as applied to the whole synoptic question would more correctly be termed a “one-source,” or an “Urevangelium” (primitive gospel) theory. Strictly taken, two sources are to be assumed only for Matthew, since Mark has but one, the Logia, and Luke is acknowledged by Weiss himself to have three.

11) I select a few of the more striking instances:

Rom. 2:19 cf. Matt. 15:14
Rom. 8:15 cf. Mark 14:36
Rom. 8:26 cf. Matt. 20:22
Rom. 8:26 cf. Mark 10:38
Rom. 16:19 cf. Matt. 10:16
1 Cor. 10:27 cf. Luke 10:8
1 Cor. 12:28 cf. Luke 11:49
1 Cor. 12:28 cf. Matt. 23:34
Gal. 4:14 cf. Matt. 10:40
Phil. 2:8f cf. Luke 14:11
Phil. 2:15 cf. Matt. 5:14
1 Thess. 5:2f cf. Luke 21:34
1 Thess. 5:13 cf. Mark 9:50
1 Tim. 2:5f cf. Matt. 20:28
1 Tim. 5:18 cf. Matt. 10:10
1 Tim. 5:18 cf. Luke 10:7
2 Tim. 2:12 cf. Matt. 10:33
1 Pet. 3:14 cf. Matt. 5:10
James 5:12 cf. Matt. 5:34-37
Rev. 3:3 cf. Matt. 24:43f
Rev. 3:5 cf. Matt. 10:32
Rev. 3:5 cf. Luke 12:8
Rev. 3:20a cf. Matt. 24:33
Rev. 3:20b-21a cf. Luke 22:30
Rev. 11:2 cf. Luke 21:24
Rev. 13:10 cf. Matt. 26:52
Rev. 14:4 cf. Luke 9:57

12) Marshall, in his criticism of the present work (Critical Review, Jan., 1893, p. 73) says in reply, that he is not prepared to accept the high authority that Resch claims for the Codex Bezse and the Western Text, and as for the variant readings in the fathers, they do not furnish sufficient proof of a Hebrew original. Then, in regard to text corruption, it is quite as possible for an Aramaic text under the conditions of the time as for the ancient Hebrew MSS. Moreover, Resch has not proved the superiority of the translation back into Hebrew, as against the Aramaic. And finally, Marshall prefers an independent position on the synoptic problem and thinks he cannot be condemned for not adhering in every case to the results of Weiss.