Bruce Metzger on the “Western Text” of the Book of Acts

Bruce M. Metzger, A Textual Commentary on the Greek New Testament (Stuttgart: United Bible Societies, 1975), pp. 259-72.


The text of the book of the Acts of the Apostles circulated in the early church in two quite distinct forms, commonly called the Alexandrian and the Western. The former, which has been traditionally regarded as the authentic text of Acts, is represented by p45 p74 א A B C Ψ 33 81 104 326 and 1175. The other form is represented chiefly by D and the fragmentary papyri p29 , p38 , and p48, by the readings marked with an asterisk or standing in the margin of the Harclean Syriac version (syrh with *, syrh mg), by the African Old Latin ms. h (a fifth or sixth century fragmentary palimpsest that preserves about 203 of the 1007 verses of Acts), and by the citations of Acts made by Cyprian and Augustine. These, which are the primary witnesses to the Western text in Acts, are sometimes joined by others that present mixed texts with a relatively high proportion of Western elements. Among such are the Armenian version of the commentary on Acts by Ephraem Syrus, the Old Georgian version of Acts, several mixed Old Latin and Vulgate manuscripts, and a few Greek minuscule manuscripts that were included by von Soden in his I-group. More recent discoveries of witnesses with decided Western affiliations include a Palestinian Syriac fragment (syrms K) from the Kastellion Monastery at Khirbet Mird, dating from the sixth century, and a Coptic manuscript (copG67) written in the Middle Egyptian dialect and dated by its editor in the late fourth or early fifth century.

The two forms of text differ in character as well as length. The Western text is nearly one-tenth longer than the Alexandrian text, and is generally more picturesque and circumstantial, whereas the shorter text is generally more colorless and in places more obscure.

The relationship between the two forms of Acts has been the subject of much discussion; the chief theories that have been proposed are the following.

(1) Both forms of text proceed from the author, who produced two editions of his work. The first to make this suggestion appears to have been Jean Leclerc, who, however, later rejected his own hypothesis. In more modern times Bishop J. B. Lightfoot took a rather favorable view of this theory, and it was subsequently adopted and developed with much learning by the German professor of classics, Friedrich Blass. According to Blass, Luke, having made a rough draft of his history of the primitive church, perhaps on the back of some previous manuscript, desired to present a handsome copy of his work to his distinguished friend Theophilus. Not being rich enough to employ a professional scribe to make the copy, Luke had to make it himself; naturally, instead of slavishly following his first draft, he exercised the freedom that an author can lawfully take with a work of his own, in altering phraseology and deleting superfluities. From both forms of Acts, according to Blass, copies were made; the text current in most manuscripts represents the polished, second edition prepared for Theophilus, while copies were also made from the original (longer) draft, which Blass supposed was treasured and preserved in the Roman church.

Nothing in this theory is inherently unreasonable, and it attracted the support of a number of other scholars, including Theodor Zahn, Eberhard Nestle, J. M. Wilson, and A.J. Wensinck. Other scholars, however, found it difficult to understand the motives of the author in choosing to omit certain details found in the presumed earlier account; the gain in space is small and the loss in information and descriptiveness is sometimes great. Is it plausible that the author would have omitted a clause from the decrees of the Jerusalem council (15.20, 29), or have altered the language of the letter of Claudius Lysias (23.2630) or Festus’s speech to Agrippa concerning Paul’s culpability (25.2425)? Furthermore, sometimes the shorter form contradicts the longer form. For example, having described (in the first person plural) a break in the journey from Caesarea to Jerusalem at the house of Mnason (so the Western text of 21.16), the author would not be likely to alter it so as to suggest that Mnason lived in Jerusalem (as is implied in the shorter text).

It has also been pointed out that in many cases the text that Blass regarded as the earlier, unrevised form of Acts exhibits the clear characteristics of later additions. Thus, for example, in a devastating review of Blass’s edition, another classical scholar, T. E. Page, assembled numerous examples where the Western text heightens or exaggerates the emphasis of the passage, where it introduces religious formulae and substitutes for the simpler and natural names of Jesus fuller and more elaborate theological titles, and where it emphasizes words and actions as inspired by the Spirit.

For these and other reasons many scholars today are reluctant to adopt Blass’s theory of two editions of Acts.

(2) Soon after Blass popularized the theory of two editions of Acts, an Irish scholar, George Salmon, offered an alternative explanation to account for the textual phenomena of Acts. He suggested that "Luke may have continued to reside at Rome after the expiration of Paul’s two years [of Roman imprisonment], and may there have given readings of his work; and explanatory statements which he then made were preserved in the West." Although it is possible to point to examples of authors in antiquity who gave public readings of their literary works, it is difficult to imagine the historical circumstances that would account for the preservation in written form of the oral comments made by Luke.

(3) Other scholars explain the distinctive form of the Western text as due to interpolation. It is maintained that in the early ages of the church the text of the New Testament was not looked upon as sacred, and therefore scribes felt at liberty to modify the form as well as to incorporate from oral tradition all kinds of additional details. Thus the Western text, according to this explanation, represents a wild and uncontrolled growth of the text during the first and second centuries.

This view has been widely held by scholars of various backgrounds, such as Westcott and Hort, W. H. P. Hatch, F. G. Kenyon, and Martin Dibelius.

Still others have held that one of the rival texts is derived from the other, not merely by a haphazard accumulation of glosses added over the years by numerous scribes, but by a deliberate revision made early in the second century by someone who was not satisfied with the existing form of the book. The problem is to determine which form was primary and which was secondary. The following two theories give diametrically opposing answers to the problem.

(4) The view that in general the Alexandrian text preserves more accurately the work of the original author and that the Western text reflects the work of a reviser was set forth with great learning by James Hardy Ropes in his edition of the text of Acts, and has been championed more recently by R. P. C. Hanson, who, however, instead of referring to a Western reviser, prefers to speak of a Western interpolator.

An interesting hypothesis that Ropes threw out for further discussion is the suggestion that "the preparation of the ‘Western’ text, which took place early in the second century, perhaps at Antioch, was incidental to the work of forming a collection of Christian writings for general Church use which ultimately, somewhat enlarged, became the New Testament; in a word, the ‘Western’ text was the text of the primitive ‘canon’ (if the term may be pardoned in referring to so early a date), and was expressly created for that purpose."

(5) The opposite point of view, namely that the Western text of Acts is primary and the Alexandrian is a deliberate modification of it, was championed by Albert C. Clark, Corpus Professor of Latin in the University of Oxford. In his earlier publications Clark explained the shortened form as being the result of a scribe’s accidentally missing here and there one or more lines of his exemplar. Since, however, accidental omissions would not account for the regular correspondence of the omissions with breaks in the sense, nor does the theory explain the numerous differences in wording where no omission is involved, in a subsequent publication Clark practically abandoned the theory of accidental omission and revived the theory of a deliberate editorial shortening of the Western text. The Alexandrian abbreviator, he thinks, excised passages throughout the book for a variety of reasons; in some cases we can deduce that he eliminated what he considered to be otiose, but in other cases the excisions, Clark admits, show a singular want of taste.

Still other theories of a linguistic sort have been proposed over the years to account for the unusual phenomena of codex Bezae.

(6) J. Rendel Harris revived the theory of Mill, Wettstein, Middleton, and other eighteenth century scholars that "the whole of the Greek text of Codex Bezae from the beginning of Matthew to the end of Acts is a re-adjustment of an earlier text to the Latin version." The theory finds little or no support among present-day scholars.

(7) The view that codex Bezae embodies an appreciable amount of Semitic coloring has been examined and adopted in various forms by several scholars. Frederic Henry Chase sought to prove that the Bezan text of Acts is the result of assimilation of a Greek text to a Syriac text that antedated the Peshitta version. In the case of the Gospels, Julius Wellhausen frequently argued for the primitive nature of the readings in codex D. This point of view was discussed further by A. J. Wensinck in a study entitled, "The Semitisms of Codex Bezae and their Relation to the Non-Western Text of the Gospel of Saint Luke," and particularly by Matthew Black in his volume An Aramaic Approach to the Gospels and Acts, in which he gathers, classifies, and carefully evaluates a large amount of relevant material. According to Black, "The Bezan text in all the Synoptic Gospels, if less so in some respects in Mark, is more frequently stained with Aramaic constructions and idiom than the B א text." A somewhat similar conclusion concerning the Western text of Acts was also reached by Max Wilcox in his monograph (originally a doctoral dissertation written under the guidance of Black) entitled The Semitisms of Acts.

Another hypothesis that seeks to account for Semitisms in codex Bezae was proposed by a specialist in the Semitic languages, C. C. Torrey. After having published several monographs on details of Aramaic coloring in the Gospels and the first half of the book of Acts, Torrey advanced the theory that the Gospels and Acts were translated from Greek into an Aramaic "Targum" towards the end of the first century, and that this "Targum," being mistaken for the original Semitic text of these books, was very soon afterwards retranslated into Greek with constant reference to the existing Greek text. This retranslation, Torrey held, was the basis of the Western text in the Gospels and Acts.

Although F. F. Bruce described Torrey’s hypothesis as "very plausible [for] it seems to satisfy many of the linguistic phenomena better than any other," most other scholars have rejected it as too complicated to be probable. ...

After surveying the chief theories that have been offered to explain the origin of the Western text, one is impressed by the wide diversity of hypotheses and the lack of any generally accepted explanation. A failing common to many of the theories is the attempt to account for the Western text by concentrating upon only one aspect of the problem. The complex phenomena, however, that characterize the Western text in relation to the Alexandrian text include, as Haenchen points out in a brief but incisive discussion, at least three kinds or levels of variant readings. There are, first, not only for Acts but for the Gospels and the Pauline corpus as well, a great number of minor variants that seek to clarify and explain the text and make it smooth. Occasionally pious phrases are introduced. This form of text, widely current in the early church and used by Marcion, Tatian, Irenaeus, and others, cannot be regarded as a "recension," for it is not and never was a unity.

Secondly, there are variants of another kind, peculiar to the Western text of Acts. These include many additions, long and short, of a substantive nature that reveal the hand of a reviser. Working upon a copy of the "Western" text in the first sense, the reviser, who was obviously a meticulous and well-informed scholar, eliminated seams and gaps and added historical, biographical, and geographical details. Apparently the reviser did his work at an early date, before the text of Acts had come to be generally regarded as a sacred text that must be preserved inviolate.

Thirdly, there are still other variants which are not to be associated with the Western text as such, nor with its reviser, but which belong to a particular manuscript, namely codex Bezae. This witness, copied, according to Haenchen, about A.D. 500, exhibits a variety of scribal idiosyncrasies, some of which, though suggesting Aramaisms, are nothing more than errors of a scribe, or possibly two successive scribes. It follows, in the words of Haenchen’s conclusion, that "in none of the three cases does the ‘Western’ text of Acts preserve for us the ‘original’ text of that book; this is the lesson that we are gradually beginning to learn."

Since no hypothesis thus far proposed to explain the relation of the Western and the Alexandrian texts of Acts has gained anything like general assent, in its work on that book the Bible Societies’ Committee proceeded in an eclectic fashion, holding that neither the Alexandrian nor the Western group of witnesses always preserves the original text, but that in order to attain the earliest text one must compare the two divergent traditions point by point and in each case select the reading that commends itself in the light of transcriptional and intrinsic probabilities.

In reviewing the work of the Committee on the book of Acts as a whole, one observes that more often than not the shorter, Alexandrian text was preferred. At the same time the Committee judged that some of the information incorporated in certain Western expansions may well be factually accurate, though not deriving from the original author of Acts. ...