Archives / 1963 / April


The Original Language of the New Testament

Sakae Kubo


The main part of the con­troversy over the original language of the New Testa­ment took place in the twen­ties and thirties of our cen­tury. The question was dis­cussed from many different angles, and a new discussion on the same basis would not prove profit­able. However, since the close of the major dialog between those who favored a Semi­tic (mainly Aramaic) origin of certain books of the New Testament and those who favored a Greek origin, new discov­eries and new studies have been made that throw new light upon the problem and lead to some modification of views in some aspects.

The area of controversy has been pri­marily over the four Gospels, and of these the most attention has been placed on the Gospel of John. However, Acts 1 to 15 and the book of Revelation have also been claimed as of Semitic origin. Even James, 1 Peter, and Hebrews have been in­cluded among these.' Though George Lamsa' would claim the whole New Testa­ment as originally Semitic, no serious scholar has paid much attention to his claim.

The Argumentation

The arguments put forth by the pro­ponents of Semitic origin' are based on Semitisms in the style and sentence struc­ture and in the usages in the several parts of speech and on mistranslations of Se­mitic originals.

Their equally able and well-equipped opponents answer by saying that many of the so-called Semitisms are not Semitisms at all and can be paralleled in non-Biblical Greek writing of the period, and that the others can be explained on the basis of the influence of the Septuagint and the Aramaic background of the writer or his material, and that the list of Semitisms by the different Semitists do not agree. In re­gard to the Gospel of John, Colwell con­cludes:

The fact that they disagree as to what should be included in the list of Johannine Aramaisms is very significant. The majority of the 54 "Semitisms" discussed in chapters ii and iii were taken from the work of Burney and Montgomery, only 7 coming from Torrey's article. The work of Montgomery and Burney was done independently; and as each feels that the "Semitisms" which he advocates could not escape the notice of an Aramaic scholar, a comparison of their findings should be interesting. There are 29 "Semitisms" from Burney's list, and 22 from Montgomery's. Only two of these are du­plicates! And the agreement is not even as high as that. In regard to one of these two, the use of the historical present, Burney is sure that its frequency is due to the Aramaic participle, but Montgomery is uncertain whether it is an Aramaism or good Greek usage. The extent of their agreement is that onoma auto is due to Semitic influence. It is ironic that their only agreement should be in error, for that onoma auto is a common construction in Hellenistic Greek is admitted by Burrows and has been demonstrated above.'

As for the mistranslations in regard to the same Gospel, Colwell says that of those suggested by Burney, Montgomery, and Torrey not one is common to all three.' Another approach to the problem of mis­translations was studied by the University of Chicago school—the study of the trans­lations of the Aramaic portions of Daniel by the translator of the Septuagint and Theodotion. Riddle, referring to an un­published doctoral dissertation of Merle Rife, points out that by comparing these two translations with the Masoretic text one cannot be sure how the original was written. His conclusion is that since the theory assumes an ability to predict an original text with a high degree of accuracy, and since this predictability is hardly en­couraged by the retranslation of the Sep­tuagint, there is an evident disparity be­tween the unquestioned and the theoreti­cal translation Greek." 6 Surprisingly little attention has been paid to this approach to the study of translation-Greek.

There were two other arguments espe­cially used by the Chicago school. The first was the lack of contemporary Aramaic lit­erature. "There are next to no remains of such a literature nor any reference to its existence in other literatures. All evidence points to the view that Aramaic was, like most languages, a non-literary speech, a vernacular and nothing more." The sec­ond argument was based on the socio-his­torical method for which the Chicago school was well known. This approach is very similar to form criticism. Riddle, on the basis of Romans 9 to 11, concludes that there were not many Jewish-Christians, and therefore no audience for Aramaic orig­inals.

Also, he feels that there are elements which clearly betray a Hellenistic atmos­phere rather than Palestinian. "Another emphasis is the distinction between mater­ials of gospel tradition which were pro­duced in Palestine and those which owed their rise to religious needs of Hellenistic communities."

The latter argument shows how this problem of the orginal language of the New Testament is tied together with prob­lems of Gospel origins as well as the Synoptic problem. Goodspeed complains that the proponents of Semitic origins (he means especially C. C. Torrey) completely disregard "the results of New Testament study in the fields of text canon, literature, history, introduction, and criticism, dis­missing them without examination as worthless.'

In text criticism the connection is seen by Goodspeed in that they do not "scruple to present rejected Greek readings where they serve its turn.' Textual criticism was further involved when A. J. Wensinck claimed that as a result of the "comparison of the Bezan text with non-Western texts of Luke, not only that there was much more evidence of Aramaic influence in Be­zan Luke but also that the isolation and establishment of Aramaisms in that text contributed substantially to the solution of the great textual problem. For if Ara­maic influence is more extensive in one text rather than another, the presumption is that the `Aramaized' text stands nearer to the kind of Greek which the Apostles wrote." "

An Analysis of the Arguments

In analyzing these arguments we find there are some areas of agreement. The major area of agreement is the fact of the decided Semitic background of some of these New Testament books. Proponents of both theories agree on this point. In this regard, therefore, careful distinction ought to be made between Greek composition in­fluenced by Greek translations of Semitic originals (i.e., the Septuagint) and the writer's own Aramaic background. De Zwaan clarifies this distinction thus:

A man may either have a) perfect or imperfect knowledge of Greek. We may call a) a "Greek" and b) for convenience a "Semite."

Now either of these two may attempt four things: on the one hand 1) translation from a Semitic dialect into idiomatic or 2) into Semiticis­ing Greek, or, on the other hand, 3) original com­position in idiomatic Greek, or 4) in Semiticising Greek.

A "Greek" trying 3) will produce no "Semitisms," a "Semite" cannot fail to do so. The same holds good of 1), but only approximately, the underlying Semitic may still shine through by means of what Psichari calls "negative Semitisms," that is, the use of locutions from a higher style, such as Attic, which would not naturally come in, but are pre­ferred because they square with a peculiarity of the translated document. We can, therefore, distinguish between "positive" and "negative" Semitisms, and, what is more important perhaps, between "pri­mary" and "secondary" ones.

Primary Semitisms are those which a "Semite" commits in 1) or 3). He is, however, always in danger of betraying himself by this cause even in cases 2) and 4), since the one source of these pri­mary or real Semitisms is his imperfect knowledge of natural Greek. The deviations however, which a man with perfect knowledge in this regard—for these ends, therfore, a "Greek"—may let pass in cases 1), 2), or 4) have a secondary cause, sec­ondary because they are due to an extraneous fac­tor: the exigencies of the reader whom he is ad­dressing, or of the documents he is translating. We have still left out of account the more or less per­fect knowledge which this "Greek" author may have of the Semitic idiom in which his source was written."

Another area of agreement (excepting Torrey among the later proponents for Semitic origin) is the generally accepted theory of Synoptic relationships. Even for Torrey the problem crops up in an Ara­maic form if not in Greek. De Zwaan, who accepts an Aramaic original for John, clearly comes out for the Synoptic theory.' Even Burrows, who hinted in an earlier ar­ticle " that he leans toward an Aramaic original of Mark, apparently argues on the basis of a Greek Mark in a later article."

Matthew Black, a distinguished Semitist and New Testament scholar, in his book An Aramaic Approach to the Gospels and Acts, in 1946 reviewed the problem of Ara­maic origins up to that time and gave a fresh study of the Aramaisms of the Gos­pels and Acts. His conclusion was that only one thing "can be regarded as in any de­gree established, that an Aramaic sayings-source or tradition lies behind the Synop­tic Gospels." " He also sees some Aramaic influence in the Marcan narrative or nondominical sayings, but says it may be attributable to "the kind of Greek which an Aramaic-speaking Jew would write.' The most crucial part of his conclusion comes when he answers the question, "What is the character of the Greek 'translation' in the Gospels where Aramaic sources can be shown to have been employed? In view of the results already obtained, we are bound to consider the Greek of the sayings of Jesus only; and in this connection, it cannot, I think, be sufficiently emphasized that in the majority of the longer con­nected parables, for example in Q the 'translation' is not literal but literary; in other words, it is doubtful if it can be justly described as translation at all in some cases, even where the evidence points to the existence and use of an Aramaic source. The Evangelists are for the most part writing Greek Gospels, even where they are dependent upon sources." "

This, then, is the generally accepted view of New Testament scholars today.

Judean Desert Discoveries

A new element in the picture that must be considered today, which was not pres­ent in the days of this controversy, is the discoveries in the Judean desert. These in­clude the discoveries at Qumran and Murabbaat. The discoveries at Qumran have brought to light Aramaic manuscripts in a form of Aramaic used in Palestine at the time of Christ. This is very significant for those who hold to the Aramaic origins of the Gospels. However, no use yet has been made of these documents in support of the Aramaic origins of the New Testament. Though many doctrinal relationships have been studied between the Qumran material and the New Testament, no study has yet been made of the possible linguistic con­nections between the Aramaic documents discovered thus far and the alleged Ara­maic originals of the New Testament. At least these discoveries have dissolved Good-speed's argument that there was no Ara­maic literature and that it was only a ver­nacular, nonliterary language. Another ele­ment in the discoveries at Qumran ought to be observed, and that is the proportion of texts in Hebrew rather than Aramaic. Milik in 1959 concludes from the discover­ies at Qumran and Murabbaat:

The copper rolls and the documents from the Second Revolt prove beyond reasonable doubt that Mishnaic was the normal language of the Judean population in the Roman period. . . . After A.D. 135 and the almost complete depopulation of Ju­dea, Hebrew ceased to be used as a colloquial language, although it was preserved in rabbinic circles.'9

While that conclusion may be too sweep­ing, at least it shows the change in attitude that has come about regarding the use of Hebrew at the time of Jesus as the result of these discoveries. Since the book was pub­lished, new discoveries have been an­nounced by Yigael Yadin" of more letters of Bar Koseba; and other documents written in Hebrew, Aramaic, Greek, and Nabataean have further confirmed the view that Hebrew was more of a living language than was formerly thought.

Using these discoveries, but approaching the problem from another angle, Jehoshua Grinte seems to show that the former con­tention, that when Hebrew is used to de­scribe a language in New Testament times it refers to Aramaic and not Hebrew, is wrong. While he may be right in this, his conclusion, that since Papias says Matthew made a record of the oracles of the Lord in the Hebrew tongue and therefore the Gos­pel of Matthew was written orginally in Hebrew, is far from proved. He strongly opposes an Aramaic original for Mark but admits a definite Aramaic background for it. But he contends that the tradition of a Hebrew original for Matthew is substan­tially correct and that the linguistic evi­dence in the book itself points in this di­rection. While admittedly there are He­braic expressions and the evidence may point more to a Hebrew than Aramaic background, the leap from the evidence he presents to the conclusion that Matthew was originally written in Hebrew is not justified at all. The evidence, like previous evidence for Aramaic originals, can be ex­plained without recourse to a Hebrew orig­inal. Besides, Grintz completely disregards Synoptic relationships as Torrey did. Some of the evidence he presents is found in Q sections (Matt. 8:10) or in sections where Matthew is following Mark (Matt. 27:41, 42; chap. 27:11, 57; chap. 15:22). Again, Grintz accepts the identity of the logia of Papias with the Gospel of Mat­thew apparently uncritically because it lends itself to his theory. At least he makes the reader think that he is not aware of the various interpretations placed upon the word logia, not to mention the distrust of some toward the reliability of Papias' wit­ness.

Another result of the Dead Sea discov­eries should lead to a more cautious atti­tude in regard to the comparison based on the Masoretic text of the translation-Greek phenomena in the Septuagint, es­pecially in Daniel. We now know that there were other texts besides the Masoretie type, and while the phenomena dis­covered above on the basis of the Maso­retie text will no doubt still be valid, at least some awareness that different originals may be the cause of the differences in translations will help to frame the conclu­sions more carefully.

Aramaisms in Codex Bezae

Another recent study bearing on the Aramaisms of Codex Bezae ought to be mentioned. Wensinck's studies led him to the conclusion that the Aramaisms in Co­dex Bezae are an evidence for their trustworthiness since they go back to a more primitive period. He accounted for the wide divergence between Codex Bezae and the B Aleph text by concluding that the latter was a later edition of an earlier edi­tion represented by Codex Bezae. Black, however, does not feel that such a theory is necessary. Instead, he explains the di­vergence simply on the basis of "two (or more) different redactions of what was substantially, if not verbally, the same orig­inal Gospel text. In what may be termed the `Bezan redaction' more of the primitive `Aramaized' Greek text has been left unre­vised than in the redaction—a word which we may now use in this connection—rep­resented by the Vatican and Sinaitic Un­cials." Torrey in his inimitable manner has fashioned a highly imaginative expla­nation for these divergences.' He considers the Bezan text as a tertiary translation into Greek of an Aramaic translation made from an earlier Greek version which was translated from original Aramaic docu­ments. The Aramaic originals had disap­peared by the end of the first century but there was still some need for Aramaic gos­pels, so the Greek translations of the orig­inal Aramaic were retranslated into Ara­maic for these Aramaic-speaking people. Early in the second century this unique Aramaic document attracted wide atten­tion because of some of its good readings and the awareness that Aramaic was the language of Jesus and His disciples; and the conjecture was made that in this codex had survived the text which the apostles themselves had written. Thus a careful, literal Greek translation was made, pre­serving all the Aramaisms. In this way Torrey explains the greater number of Aramaisms of the Bezan text and also ex­plains why he considers its unsupported readings as worthless. While Torrey's the­ory disallows the reliability of the Bezan text, the theories of Wensinck and Black lead to a greater trust in the reliability of Codex Bezae.

The above theories, however, based as they are on the supposed greater number of Semitisms in the Bezan text, have to be set aside or modified in view of the findings of James Yoder, who made a careful study of the distinctive readings of Codex Bezae. He arrived at "two significant conclusions: 1) when one takes into account not only the instances of Semitic phenomena in Codex Bezae, but also the Bezan variants which abandon Semitisms found in other MSS, the net increase of Semitisms is some­times inconsequential, while in other re­spects this MS actually reveals fewer Semi­tisms than found in the B Aleph text; and 2) ofttimes the data are concentrated in limited areas of the text, thus detracting from the supposed homogeneity of the Be­zan text."

Yoder's conclusions do not necessarily in­dicate that the Bezan text is not reliable, but they definitely show that the basing of its reliability on its more numerous Ara­maisms is no longer possible.

Thus we find new discoveries and new studies modifying previous points and en­lightening different aspects of the problem, but nothing has arisen to change the gener­ally held opinion that the New Testament books were all originally written in Greek, though admittedly influenced by trans­lated Semitic documents and the Semitic milieu of its contents and the Semitic back­ground of the writer in some cases.


comments powered by Disqus



1 Cf. J. B. Mayor, The Epistle of St. James (London: Macmillan and Co., 1897), p. ccxxxii.

2 George M. Lamsa, The Modern New Testament (Phila­delphia: A. J. Holman Co., 1940), pp. vii, xis.

3 Cf. Moulton and Howard, A Grammar of New Testament Greek (Edinburgh: T. & T. Clark, 1929), pp. 413-485. where Howard systematically discusses the various types of Semitisms and mistranslations.

4 Ernest C. Colwell, The Greek of the Fourth Gospel (Chicago: University of Chicago Press, 1931), p. 128.

5 Ibid., p. 129.

6 D. W. Riddle, "The Logic of the Theory of Translation Greek," Journal of Biblical Literature, 51 (1932), 23.

7 Edgar J. Goodspeed, Slew Chapters in New Testament Study (New York: Macmillan and Co., 1937), p. 154.

8 D. W. Riddle, "The Aramaic Gospels and the Synoptic Problem," Journal of Biblical Literature, 54 (1935), 37.

9 Goodspeed, op. cit., p. 56.

10 Ibid., p. 157.

11 Cited in Matthew Black, An Aramaic Approach to the

12 Gospels and Acts (Oxford: Clarendon Press, 1946), p. 6.

13J. de Zwaan, "The Greek of Acts," in Beginnings of Christianity (London: Macmillan and Co., 1922), II, p. 53. In Black, op. cit., p. 156.

14 Millar Burrows, "Mark's Transitions and the Transla­tion Hypothesis," Journal of Biblical Literature, 48 (1929), 117-123.

15 Burrows, "The Original Language of the Gospel of John," Journal of Biblical Literature, 49 (1930), 93-139.

16 Black, op. cit., p. 206.

17 Ibid.

18 Ibid., p. 209.

19 J. T. Milik, Ten rears of Discovery in the Wilderness of Judaea (London: SCM Press, 1959), pp. 130, 131.

20 In an illustrated lecture given at the University of Chi­cago in the fall of 1961.

21 Jehoshua M. Grintz, "Hebrew as the Spoken and Writ­ten Language in the Last Days of the Second Temple," Journal of Biblical Literature, 79 (1960), 32-47.

22 Black, op. cit., p. 214.

23 Torrey, C. C., Documents of the Primitive Church (New York: Harper and Brothers, 1941), pp. 118-139.

24 James Yoder, "Semitisms in Codex Bezae," Journal of Biblical Literature, 78 (1959), 317.

back to top