Peter M. Head

Originally Published in: The Book of Acts in its Ancient Literary Setting

(ed B.W.Winter & A.D. Clarke; The Book of Acts in its First Century Setting vol 1;

Grand Rapids: Eerdmans & Carlisle: Paternoster, 1993), 415–444.

[Here using SymbolGreek font]



This paper discusses the problems associated with the Western text of the book of Acts. A brief survey of the history of research on the subject suggests that the scholarly consensus until recent years is that the Western text should be regarded as secondary and non-Lukan. This position has been challenged by several scholars and the bulk of the paper is given over to an investigation of supposedly tendentious readings in the Western tradition (relating to Christology, the Holy Spirit, the status of the apostles, and the apostolic decree). The author concludes by suggesting that modern proponents of the originality of the Western text have not yet managed to account for the theological Tendenz of that text-type, which appears more like a secondary form of a more primitive original.

I. Introduction

The crucial textual problem facing both historians and exegetes in dealing with the book of Acts is the existence of two relatively distinct forms of the text. One such text-type is generally known as the Western text (common usage rather than accuracy demands that we continue to use this term).[1] The primary representative of this text is Codex Bezae, which contains over 800 more words than the commonly used NA26=UBS3 text for the extant sections.[2]

          The other form of the text might be described as the Alexandrian text, with Sinaiticus and Vaticanus as its primary representatives.[3] A commonly cited comparison suggests that the Western text is 8.5% longer than the Alexandrian text.[4]

          A further indication of the magnitude of the problems posed by the texts of Acts is supplied by the observation that Metzger’s Textual Commentary on the UBS3 edition devotes one-third of its pages to consideration of Acts, mainly because of the complexities of the Western text.[5]

          Textual critics have struggled to explain the origin and nature of the Western text of Acts since at least 1685 when Jean Leclerc suggested that Luke may have published two editions of Acts (a suggestion he subsequently rescinded).[6]

          In this paper we shall, in section II, briefly survey the discussions which contributed to the modern consensus that the Western text represents a later form of the text which arose from paraphrasing interpolations to the shorter (Alexandrian) original. This consensus has been challenged by various scholars, and in section III we shall investigate several of these alternative positions, including: i) the position of Blass that Luke was responsible for both versions of the text of Acts; ii) the view that the Alexandrian text is the result of a later redactor having abbreviated Luke’s original (‘Western’) text; and iii) the position of Strange which incorporates elements of all the other views.

          It will emerge from this historical survey that there are several crucial categories of variants which provide the basis for the consensus argument that the Western text is a secondary (non-Lukan) revision of a shorter Vorlage (best represented in the Alexandrian tradition). Section IV of our paper discusses several of these categories, focussing especially on variants which are said to reflect a distinctive (or at least secondary) theological Tendenz in the Western tradition, i.e. passages concerning Jesus, the Holy Spirit, the status of the apostles and the apostolic decree. The conclusion (section V) will briefly summarise our findings and suggest areas that need further exploration.

II. The History of Research: Towards a Consensus[7]

It was, in several ways, the pioneering work of Johann Jakob Griesbach which spelled the beginning of the end for the textus receptus.[8] Developing the earlier family-recension theories of Bengel and Semler, Griesbach assigned the textual material extant at the time to three different recensions: Alexandrian, Western, and Constantinopolitan.[9] This last group he believed to be a compilation from the two ancient recensions: the Alexandrian and the Western, of which he gave preference to the Alexandrian recension.[10] Subsequent investigations in the nineteenth century, informed by a flow of new material from Tischendorf and others, gave supremacy to the Alexandrian text-type in general (especially, for example, in Westcott & Hort, whose liking for Sinaiticus and Vaticanus is well known).[11] The support which many early papyri gave to an Alexandrian text-type means that the general situation, although markedly more complex, has not changed a great deal in the twentieth century. Most scholars accept (with variously weighty reservations) that the Alexandrian text-type is to be preferred to the Western text.[12]

          The Western text of the book of Acts, represented by variant readings found within codex Bezae (and others) has been regarded as essentially secondary to the shorter ‘Alexandrian’ text.[13] This was true in the era of Westcott & Hort,[14] and has been supported by, inter alia, J.H. Ropes,[15] F.G. Kenyon,[16] M. Dibelius,[17] E.J. Epp,[18] R.P.C. Hanson,[19] and B. Aland.[20] This consensus is clearly reflected in the texts of the favoured modern editions of the Greek New Testament (NA26, UBS3). Metzger reports that ‘more often than not the shorter, Alexandrian text was preferred [by the editorial committee]’.[21]

          The consensus position has also been upheld, during the course of the present century, by most commentators;[22] including C.S.C. Williams,[23] F.F. Bruce,[24] E. Haenchen,[25] R. Pesch,[26] H. Conzelmann,[27] and G. Schneider.[28] In this context it is also appropriate to mention the historical work of C.J. Hemer. Hemer handles the textual problem in an explicitly consensual manner: in general he prefers the ‘Alexandrian’ rather than the ‘Western’ text (clearly affirming his conviction that the Western text is not original), but he seeks to allow some flexibility, and regularly notes Western readings which impinge on historical questions. Despite his sympathy with the view that Luke may have produced two editions,[29] Hemer could find no Western reading which commended itself as original.[30]

III. Challenging the Consensus: the originality of the Western Text

It is F. Blass who is most often associated with the view that Luke was responsible for both forms of the text of Acts (although the position continues to have its modern advocates, as we shall see).[31] Blass argued that the two recensions have such a similar style that they must have come from the same author; that additional information such as that found in 1:5; 14:2; 21:16 could only have come from someone with knowledge of the events, and that the vocabulary of the Western variants is generally ‘Lukan’.[32] According to Blass, Luke wrote a rough draft of his history of the church while in Rome, and this was the full and longer version that is represented in the Western recension (b). Luke subsequently used this rough draft as the basis for his second edition, the version sent to Theophilus. This version was clearer, terser and more carefully composed (deleting superfluous phrases), and this is represented in the Alexandrian recension (a according to Blass).

          Blass won numerous and influential supporters, including T. Zahn,[33] E. Nestle,[34] F.C. Conybeare,[35] and J.M. Wilson.[36] While some critics attacked his reconstruction of Luke’s situation in Rome,[37] most focussed on what were felt to be the two crucial weaknesses of the theory. The first of these was the so-called pietistic Tendenz of many of the Western variants, especially in the inclusion of christological titles, a heightened emphasis on the Holy Spirit, and an elevated view of the apostles.[38] Such critics argued that if the Western text differs from the Alexandrian in a pietistic direction, then it can most readily be understood along the lines of other theologically motivated alterations which textual critics have generally accepted as secondary.[39] The second weakness which was perceived in the Blass hypothesis was the problem of inconsistencies between the Western and Alexandrian text-types. Thus, it was argued, that if in fact the same author had been responsible for both text-types, we would expect neither distinctive theological approaches, nor historical or geographical contradictions. Many critics have claimed, however, that just such inconsistencies do exist. A particular focus of interest in this area has been the divergent forms of the apostolic decree of Acts 15.[40] Both of these two areas of disagreement have continued to be of great importance in more recent discussions.

          Indeed, to some extent, the modern debate concerning the Western text revolves around the same issues which were current almost a century ago. Three crucial questions are: i) to what extent is the style and vocabulary of the Western text recognisably ‘Lukan’?; ii) can the Western text be explained as a secondary recension of a shorter original motivated by a theological Tendenz? and iii) do the differences between the Western text and the Alexandrian text constitute evidence against common authorship? However, recent discussions have added additional complexities in the area of identifying the western text and the related question of whether the evidence suggests an individual ‘redactor’ behind the original form of the Western text, or merely a loose collection of paraphrastic renderings.

          Of course, we should not give the impression that Blass’ position has been the only option for advocates of the Western text. Ramsay pointed out that although much of Blass’ discussion about the Western text had value, he had not established, even on his own terms, that the Eastern (Alexandrian) form was ‘equally original and good’.[41] A.C. Clark argued that Blass’ hypothesis was fatally flawed by the presence of contradictions between the two forms of the text.[42] He maintained the position that the Alexandrian form was a later revision of Luke’s original Western text.[43] Clark’s discussion was based on the palaeographical principle that lines of text (stivcoi) could be omitted by homoioarcton or homoioteleuton. The Bezan text is written in lines of sense (properly kw'la), as was the original (according to Clark), and Clark argued that the Alexandrian text was formed by the omission of these whole sense-lines from the original:

in Acts (not in the Gospels) G [i.e. the Alexandrian text-type] represents the work of an abbreviator who, having before him a MS. written in stivcoi similar to those found in D, frequently (not, of course, always) adopted the rough and ready method of striking out lines in his model, botching from time to time to produce a construction.[44]

Clark’s approach has not won scholarly support,[45] and was severely criticised by Kenyon who showed that there was no evidence for the use of such sense-lines in the early period.[46] Overall, the plausibility of this position will depend upon the plausibility of the deliberate redactional activity envisaged by Clark’s theory.

          Several recent works on the Western text have sought to provide support for the initial premise of Blass’ hypothesis: that in style and vocabulary the Western readings bear a remarkable similarity to undisputed Lukan style. R.S. MacKenzie noticed several ‘genuine Lucanisms’ in the Western variants to Acts 10:41; 13:31, 38, 41 and 17:26-28:[47]

i) the presence of sustrevfein in 10:41D (also in 11:28D; 16:39D; 17:5D corresponding to Lukan usage in 23:12 and 28:3);

ii) the use of a[cri nu`n in 13:31D (cf. 20:26D which also uses a[cri to designate a period of time; as Luke does in undisputed passages);

iii) the use of metavnoia in 13:38D (cf. Luke 3:16D; generally favoured by Luke);

iv) the use of kai; ejsighvsan in 13:41D (Luke regularly signals the silent reaction of listeners to speeches in Acts);

v) the use of ai{ma in 17:26D; mavlista in 17:27D and kaq∆ hJmevran in 17:28D (all, he argues, are regularly Lukan terms).

These proposals were investigated by R.F. Hull, and found wanting in detail and in method.[48] For example, he noted that Acts 10:41D ‘harmonizes the content of Peter’s sermon with the preamble of Acts (namely, 1:4) and the Emmaus story of Luke 24’.[49] In addition, although MacKenzie had argued that ‘the possibility of the Lucan character of some of the variant readings means, among other things, that they cannot be dismissed as simply the work of a particular reviser or of revisers’;[50] Hull pointed out that even if the ‘Lukan’ character of the Bezan vocabulary could be established MacKenzie’s conclusion would not follow. This has been reinforced by T.C. Geer in his critique of recent studies of the vocabulary of the Western text: ‘the only sure conclusion. . .is that if the compiler of this “Western” text was one of the New Testament authors, he would most likely have been Luke’.[51]

          The most wide ranging and thorough investigation of the vocabulary and style of the Western text is the combined work of M.-E. Boismard and A. Lamouille.[52] In an earlier study Boismard (aided by Lamouille) had argued that the Western variants of Acts 11:2 and 19:1 were Lukan both in vocabulary and style.[53] In their 1984 attempt to reconstruct and rehabilitate the Western text they took this further, suggesting that an original Western text (TO) and the Alexandrian text (TA) both reflect Lukan characteristics.[54] The original Western text-form gave rise to two derivative forms, one a pure form (TO1), and the other less pure (TO2).[55] Thus the original (Lukan) Western text needs to be reconstructed from the extant sources. Their overall theory then involves the priority of the Western text and a subsequent Lukan redaction:

Luke wrote a first redaction of Acts, of which we find an echo in the Western text. . .a certain number of years later he radically altered his initial work, not only from the stylistic point of view, as Blass had it, but also from the point of view of its content; that these two redactions were fused into one to give the present text of Acts, or more precisely, the Alexandrian text (in a purer form than that which we now have).[56]

In connection with the questions commonly posed to the Blass hypothesis, Boismard and Lamouille answer that there are clearly secondary elements in the Western text (they agree with Ropes that many elements are due to ‘the paraphrastic re-writing of a second-century Christian’),[57] but that these are not integral to the original Western text which they reconstruct from various sources. They postulate some years between Luke’s two editions in order to account for the change in content.[58]

          The material contained within Boismard and Lamouille’s two volumes is of importance for our study, as it is for the study of Luke-Acts in general, and we shall interact on some detailed points in what follows. Two issues of a general nature must be considered briefly here.[59] The first concerns the hypothetical nature of their reconstructed text, which has been criticised by both J.N. Birdsall and, to a lesser degree, J.K. Elliott.[60] In particular it is questionable whether their assumption of an earlier ‘pure’ form of the Western text is justified in view of the evidence of early ‘western’ papyri such as 29, 38, 48.[61] Further, it often appears that the main reason for the choice of an original occidental reading is often its proximity to Lukan style; this creates a degree of circularity in the structure of the whole argument, as Geer comments: ‘the Western text seems to be what Boismard makes it’.[62]

          The second problem is the issue we noted above in connection with MacKenzie: by limiting the comparison with the style and vocabulary of the New Testament, Boismard and Lamouille establish only that of all the New Testament writers the Western variants most closely resemble Luke.[63] This is clearly significant, but in view of the widespread scribal tendencies of harmonisation and assimilation, it is intrinsically probable that a secondary recension will stylistically resemble the Vorlage.[64] This is especially true if, as seems likely, the Western text conforms various passages to programmatic (Lukan) statements:[65] a later revisor will automatically look as if he is using Lukan style. Arguments from vocabulary and style cannot establish common authorship if the presence of inconsistencies and explicable tendencies in the variant readings demand a different explanation.[66] This is particularly true since the Western tradition (and Bezae in particular) in the Gospels is particularly marked by harmonisation, and harmonisation will invariably entail a degree of stylistic imitation.[67]

          The most recent attempt to address The Problem of the Text of Acts is the book of that name by W.A. Strange.[68] This book is now (in English) the most important challenge to the consensus position since Clark (1933), although, as he acknowledges, it is Boismard and Lamouille who have prepared the way for his argument. His general position can be summarised as follows: The Western version arose as a ‘commentary-like’ series of annotations (and is thus later and supplementary to the shorter text), but is ‘Lukan’ in style and vocabulary (following Boismard and Lamouille). The author of Acts left two separate rough drafts of Acts, one of which consisted of an annotated version of the other.[69] These two versions remained unused until they were eventually published separately some time between A.D. 150 and A.D. 175 in order to counter the Gnostic and Marcionite claims to Paul.[70]

          We shall deal below with Strange’s arguments concerning some of the key passages (particularly Acts 8:36-39 and the apostolic decree: 15:20, 29; 21:25). At this stage we should note that there are several weaknesses in Strange’s approach to the dating of these two editions. Firstly, it depends fundamentally on a massive argument from silence: he argues that since there is no conclusive evidence of its citation Acts did not exist until the middle of the second century.[71] Secondly, his position requires that two forms of Luke’s notes were kept, quite independently, across over eighty years, but never published or referred to (this is rather strange: who kept it? where? why? how?) In addition, this scenario fails to account for the prefaces (to Luke and Acts), which suggest a connected ‘published’ form. Nevertheless Strange offers important discussions of many Western variants, some of which I shall deal with below.

IV. The Theological Tendenz in Western Variants

We have mentioned several times that one of the key arguments used against the originality of the Western text is the presence within this text-type of variants which seem to reflect a later revision.[72] It is notable that Ropes, a key defender of the originality of the Alexandrian text, suggested that ‘of any special point of view, theological or other, on the part of the “Western” reviser, it is difficult to find any trace’.[73] Nevertheless, he did find some such traces, such as a hostile attitude to Judaism (14:5); a positive attitude to Gentile Christianity (24:5; cf. 2:17; 2:14); and a desire to emphasise the pre-conversion paganness of the devout (20:21; 26:15). Several other categories have also been suggested, and a large number of passages containing Western variants are involved.[74] In what follows we shall group these under the following headings: Christology (this includes material used in baptismal and confessional contexts), the Holy Spirit, and the Status of the Apostles; a final section will discuss the texts of the Apostolic Decree.[75]

1. Christology

A commonly cited tendency of some Western variants is the addition and/or expansion of honorific titles of Jesus. The most obvious examples include the addition of christological terminology to relatively simple titles of Jesus:

i) cristo" added to kurio" Ihsou" in:

1:21 (Dd 876 1108 1611 1765 1838 pc syh eth6mss mae). [TO2]

4:33 (Dd Ee 323 945 1739 2495 al r). [TO2][76]

8:16 (Dd vg3mss mae eth4mss). [TO2][77]

11:20 (D 33vid pc w mae). [TA=TO][78]

15:11 (C Dd Y 33 36 431 453 467 522 876 945 1175 1739 1765 1891 2298 it syrp bopt eth10mss geo Irenlat). [TO2][79]

16:31 (C Dd E Y 0120 ˜ syrp sa eth12mss). [TO2]

19:5 (Dd 257 383 614 1799 2412 2147 a(629) b gig syrp syrh* sa eth10mss geo). [TO2][80]

21:13 (C D gig syrp eth10mss CyrilAlex. Tert. Jer. Aug.). [TO2][81]

ii) kurio" added to Ihsou" cristo" in:

2:38 (Dd Ee 522 614 876 945 1611 1739 1891 2138 2412 pc b c p r syrph geo [sa mae: different order]; Cyp. Basil). [TO2][82]

5:42 (Dd 1898 gig h p syrp sa eth8mss).[83]

10:48 (Dd 383 81* 1311 pm p syrp geo). [TO2][84]

iii) kurio" added to Ihsou" in:

7:55 (Dd h p samss mae). [TO2]

18:5 (Dd 383) [not TO]

iv) kurio" and cristo" added to Ihsou" in 13:33 (D sa mae Amb. [kurio" without cristo": 614 & syrh*]). [TO2]

v) Miscellaneous similar additions:

9:20: o Cristo" before o uio" tou qeou (h l m sa mae Iren.). [TO2, TO=TA]

16:4: ton kurion Ihsoun Criston in wider alteration (D [syrh*]).

20:25: tou Ihsou to thn basileian (D gig [+ kuriou] sa Lucifer). [TO]

A very characteristic alteration occurs in several confessional and baptismal contexts:

vi) 6:8: The addition of dia tou onomato" kuriou Ihsou Cristou provides a source for Stephen’s signs and wonders (Dd Ee 33 257 431 614 876 913 1611 1765 2138 2412 p t sa mae Aug. [syrh: lacks Ihsou cristou]). [TO2, TO=TA][85]

vii) 9:17: The addition of en tw onomati Ihsou Cristou after epiqei" ep auton ta" ceira" in connection with Ananias restoring Paul’s sight (h mae t.3 vgms ps-Chrys.[86]). [TO2, TO=TA]

viii) 9:40: The addition of en tw onomati tou kuriou hmwn Ihsou Cristou occurs in words of Peter to Tabitha (it syrh* sa mae geo Cyp. Amb. Spec). [TO2, TO=TA][87]

ix) 14:10: The addition of soi legw en tw onomati tou kuriou Ihsou Cristou to Paul’s healing command (C Dd Ee 242 255 257 323 383 431 467 522 614 876 945 1175 1739 1765 1799 1891 2147 2298 2412 pm h syrphmg mae samss bohmss eth10mss geo). [TO2, TO=TA]

x) 18:4: An alternative rendering of the verse reads as follows: eisporeuomeno" de ei" thn sunagwghn kata pan sabbaton dielegeto, kai entiqei" to onoma tou kuriou Ihsou, kai epiqen de ou monon Ioudaiou" alla kai Ellhna" is slightly awkward, and introduces ‘the name of the Lord Jesus’ as the subject of Paul’s teaching (D h [a b c dem gig syrhmg][88]). [TO]

xi) 18:8: Two separate alterations have Western support: firstly the addition of dia tou onomato" tou kuriou Ihsou Cristou supplying the content of the Corinthian faith (257 383 614 1799 2147 2412 syrh*); and secondly (alternatively) the addition of pisteuonte" tw Qew dia tou onomato" tou kuriou hmwn Ihsou Cristou as a description of their baptismal faith (Dd h). [TO].

          These variants are widely known, and reflect a general trend which is present in many christologically motivated scribal alterations found in other parts of the New Testament textual tradition.[89] Most scholars have regarded this type of variant as both typical of the Western textual tradition and ‘a strong argument against the view that the Western text is prior to the non-Western’.[90] Boismard and Lamouille have argued that these christological alterations only secondarily infected the Western tradition (hence they are mostly regarded as TO2).

          According to Boismard and Lamouille the pure TO is the primitive form with a tendency to suppress christological titles,[91] evidence of the multiplication of christological formulae is consistently attributed to the degenerate form TO2.[92] Thus without disputing the principle (addition of christological titles is characteristic of secondary alterations) they use exactly this criteria to distinguish between TO and TO2. The success of their whole endeavour can therefore be measured by the plausibility with which they can reconstruct TO at these points. From this perspective, however, it must be noted that the evidence given in support of their reconstructed TO is very often limited to late versional and patristic sources (see above, especially re. 8:16; 15:11; 19:5; 2:38; 10:48). At other points there is only one Western variant, but this is attributed to TO2 rather than TO (see above, especially re. 9:20; 6:8; 9:17, 40; 14:10). As Strange suggests, the argument is circular: ‘copyists are more likely to add Christological titles, some Western witnesses do not have certain Christological titles, therefore these witnesses give the original Western readings, and therefore the Western text originally had fewer Christological titles’.[93] Although Boismard and Lamouille’s position is not impossible, it is far from self-evident, especially considering the widespread Western support for many of these readings, and the paucity of support for their TO readings.[94]

          In my opinion this type of pietistic variant provides the proper context for the investigation of the Western reading at 8:37. Strange suggests that this variant ‘is one of the most theologically significant in the entire work’,[95] and argues that the Ethiopian Eunuch’s confession of faith is original. The text is as follows:

eipen de autw [+ o Filippo" E] ei [ean E] pisteuei" ex olh" th" kardia" sou [- sou 323] exestin [swqhsei E] apokriqei" de eipen pisteuw ton uion tou Qeou einai ton Ihsoun Criston [ei" ton Criston ton uion tou Qeou E]

E 36 242 257 323 453 467 522 629 876 913 945 1522 1739 1765 1891 2298 it (i.e. a b c gig l m p r t w y dem) vgcl syrh* mae geo ethms Iren. Cyp. Aug.). [TO][96]

One of the main features which distinguishes this from the other variants listed above is its much more widespread support. Its absence from 45 74 S.01 A B C 33 81 614 cop etc. has stood against it in the eyes of most editors, primarily because ‘there is no reason why scribes should have omitted the material, if it had originally stood in the text’.[97] Strange finds a reason, arguing that the verse was omitted because of second century secrecy about Christian rites, as Celsus said: ‘Christians perform their rites and teach their doctrines in secret’.[98] The evidence adduced by Strange hardly provides the motive for the omission of a christological confession of such clarity, when such confessions were explicitly part of the public teaching of the church.[99] The passage is not about baptismal rites, but the content of baptismal faith, something that the Western text has taken pains to emphasise at other points (Strange makes his case only by isolating this passage from those listed above). This variant can be profitably compared with other Western expansions to baptismal scenes:

i) 2:41, which explains that those who believed Peter’s message were baptised (reading pisteusante" either in place of [Dd] or in addition to apodexamenoi [p r vgms syrphmg mae = TO]);

ii) 18:8, which explains the content of the faith of the Corinthians (discussed above);

iii) 19:5, which adds an explanation of the purpose of baptism (ei" afesin amartiwn: 38 Dd 257 383 614 1799 2147 2412 syrh* Chrys.) as well as an expanded christological appelation (mentioned above).

These examples support the argument of Zuntz that the narratives in Acts have been transformed in the western text into ‘scenes of a paradigmatic ecclesiastical character’.[100] The passages we have listed here suggest that in terms of transcriptional probabilities, the western variations could be regarded as in line with a common tendency to highlight christological material.[101]

2. The Holy Spirit

The Western text also includes numerous additional references to the Holy Spirit, or explanatory comments:

i) 6:10: the addition of tw agiw to pneumati (Dd Ee gig h p t g2 mae eth = TO2). This takes place within the context of a larger addition which emphasises the authority of Stephen (which we shall mention below).[102]

ii) 11:17: several western witnesses attest an addition to Peter’s question (which takes three forms):

tou mh dounai autoi" pneuma agion (467 p vgms mae Aug. BarS.) [TO]

t. m. d. a. p. a. pisteusasin ep autw (Dd)

t. m. d. a. p. a. p. epi ton kurion Ihsoun Criston (b vg3mss syrh* bhm ndl.2 prv tpl).[103]

iii) 15:7: the addition of en pneumati (Dd l Tert. = TO) or en pneumati agiw (257 614 1799 2412 syrhmg Cass.) to the description of Peter: ‘standing up in the [Holy] Spirit he said to them. . .’

iv) 15:29: the addition of feromenoi en tw agiw pneumati (Dd l IrenLat Tert. Ephr. = TO) to the closing statement of the letter from the Jerusalem council.

v) 15:32: the addition of plhrei" pneumato" agiou (Dd alone) to the description of Judas and Silas.

vi) 19:1: an alternative reading for the verse in which the Spirit tells Paul to return to Asia: eipen autw to pneuma upostrefein ei" thn Asian (38 Dd syrhmg Ephr. = TO).

vii) 20:3: an alternative reading in which the Spirit tells Paul to return through Macedonia: eipen de to pneuma autw (Dd gig syrhmg Ephr. = TO).

Although these references have been regarded by some scholars as strong evidence for the secondary nature of the Western text in emphasising the work of the Holy Spirit,[104] the alterations are basically either formal (adding to agion) or concerned with another subject altogether, that is, emphasising the Spirit’s endowment of the apostles and prophets of the Christian church.[105] When the following passages are taken into account this will emerge as a more clear cut tendency of the Western tradition.[106]

3. The Status of the Apostles

It has been a well recognised feature of the Western textual tradition that it tends ‘to emphasise the wisdom, authority, and power of the apostolic figures’.[107] This feature seems to have been significant in several passages given above (e.g. 6:10; 11:17; 15:7, 32) concerning references to the Holy Spirit. Additional passages which might be listed under this heading include the following:

i) 5:15: The Western tradition has two alternative additions, emphasising the effect of Peter’s healing ministry. Firstly: kai rusqwsin apo th" asqeneia" autwn ([Ee] a b w y vgcl ethms Lucif. = TO). Secondly: aphllassonto gar apo pash" asqeneia" w" eicen ekasto" autwn (Dd [e p mae: kai for gar] Chrom. = TO2); alternatively:[108]

ii) 6:10f.: An alternative rendering of this verse highlights the wisdom and boldness of Stephen: ouk iscuon antisthnai th oush en autw kai tw pneumati tw agiw w elalei, dia to elegcesqai autou" ep autou meta pash" parrhsia". mh dunamenoi oun antofqalmein th alhqeia (D [E] e h t w [syrhmg mae] = TO [except for tw agiw]).[109]

iii) 9:22: Western witnesses add (en) tw logw to the statement of that Paul grew strong (C Ee 467 h l p mae = TO; Bezae deficient).

iv) 13:8: an additional statement emphasises the positive response of Sergius Paulus to the preaching of Paul: epeidh hdista hkouen autwn (D* syrh* mae = TO2; cf. Ee: oti hdew" autwn hkouen = TO). This is matched by the addition, with a similar effect, in v. 12 of eqaumasen kai (D E syrp eth Ephr. = TO): ‘the proconsul, when he saw what was done marvelled, and believed. . .’

v) 13:43: a long additional statement (in two forms), emphasising the impact of Paul and Barnabas: egeneto de kata pasan polin fhmisqhnai ton logon (Ee w vgms mae = TO); alternatively: egeneto de kaq olh" th" polew" dielqein ton logon (tou qeou) (Dd syrhmg = TO2).

vi) 14:7: the addition of kai ekinhqh olon to plhqo" epi th didach autwn. . .(D [lacks autwn] h [b w] vgms mae = TO), emphasising the effect of the mission of Paul and Barnabas on the whole region (cf. olhn added to end of v. 6 in D E lat [mae]).

vii) 16:4: An alternative rendering of the verse: diercomenoi de ta" polei" ekhrusson kai paredidosan autoi" meta pash" parrhsia" ton kurion Ihsoun Criston, ama paradidonte" kai ta" entola" (Dd syrhmg Ephr. = TO). This emphasises the boldness of Paul and Timothy in much the same way as previously noted concerning Stephen.

viii) An additional feature is the emphasis on the apostolic obedience to the charge of preaching the gospel (1:2: add kai ekeleuse khrussein to euaggelion (D [gig t] syrhmg mae sa = TO2); 10:41: add kai sunestrafhmen [D it syrh mae]; cf. also 8:4; 11:1f.) even in the face of difficulties (note the occurence of qliyi" kai diwgmo" in 8:1; 13:50).

This tendency seems to be a well established feature of the Western tradition (as Boismard and Lamouille’s classifications indicate and as Strange also acknowledges).[110] In the instances surveyed this does not include the use of honorific titles for the apostles,[111] nor the hagiography of the later non-canonical Acts. In view of this Strange argues:[112]

there is a clear divide between the treatment of the apostles in the Western text of Acts and in the apocryphal Acts. In the latter, the Acts of Paul in particular, the apostles are brought firmly into the writers’ own age: Paul refutes Gnostic heresy,[113] encourages martyrdom,[114] and teaches clearly the custom of delaying baptism.[115] The Western text does not exhibit the second-century tendency to use the apostles to legitimate contemporary practice.

Nevertheless, in the passages listed above there are indications that the apostolic presentation has been contemporised in a way that is characteristic of these non-canonical second century writings, most particularly in the rendering of the setting and decree of the council of Jerusalem (also 8:37). This is not done by way of inventing new material (which is not really open to a reviser of manuscript material), nevertheless the Western text might stand part-way along the road to the non-canonical Acts.[116]

4. The Apostolic Decree

There are complex and difficult problems associated with the events of the council of Jerusalem and the terms of the apostolic decree recorded in Acts 15.[117] In particular the exact wording of the decree (Acts 15:20, 29; 21:25) has been much debated in view of the textual difficulties at this point.[118] The most recent discussion is in Strange (who has a useful presentation of the different text-forms of the decree on 88). According to Strange the consensus position is the one that has been defended by Kümmel and others, that is, the original text of 15:20 contained four prohibitions:[119]

ajlla; ejpistei`lai ajutoi`" tou` ajpevcesqai tw`n ajlisghmavtwn tw`n eijdwvlwn kai; th`" porneiva" kai; tou` pniktou` kai; tou` ai{mato"

74 S.01 A B C E ˜ latmss bo

From this original form (according to Kümmel et al) three alterations were made:

i) kai th" porneia" was omitted by 45 (also arm ethmss);[120]

ii) kai tou pniktou was omitted by Western witnesses (D d gig IrenLat Ambrosiaster Ephr. Aug. = TO);

iii) a form of the golden rule (kai osa an mh qelwsin eautoi" ginesqai eteroi" mh poein) was added, mainly in Western witnesses ([Dd] 242 323 522 536 945 1522 1739 1891 2298 it61 [eth] sa IrenLat).[121]

A somewhat similar situation arises in the two other places where the decree is cited (15:29 and 21:25, see Strange’s chart for details).

          Most scholars have argued that the longer reading reflects a ‘ritual’ understanding of the decree (eijdwlovquton refers to eating of meat sacrificed to idols; ai|ma refers to eating of blood; pniktovn refers to strangled things), formed under the influence of the Old Testament (whether Noachian commands or Lev. 17-18 need not detain us here). The Western readings have been regarded as an ‘ethical’ version of this original ‘ritual’ decision, designed to appropriate the decree in a later situation.[122] However, Strange argues (following Resch, Harnack, Ropes and others) that this distinction (between ‘ethical’ and ‘ritual’) cannot be maintained; that some elements of ‘ritual’ (particularly abstention from blood) were in any case widespread in the second century;[123] and that the form of the western text conforms with Luke’s desire to ‘make the Decree into a regulation having validity for Christians of all places and in all times’.[124]

          Strange does not, however, deal with related Western readings not only in the decree texts but in the surrounding contexts (he tends to isolate the Western readings in the decree from the Western readings throughout Acts, cf. above on 8:37ff.). The broader context into which the terms of the decree are located begins at 15:1 where the original issue posed ‘Unless you are circumcised according to the custom of Moses, you cannot be saved’ is considerably broadened in the western tradition by the addition of kai. . .peripathte (D syrhmg sa mae: TO): ‘Unless you are circumcised and live [lit. ‘walk’] according to the custom of Moses, you cannot be saved’.[125] Thus the issue is not limited to circumcision, but the broader question of Gentile Torah-observance.

          This broader issue is highlighted in the large western addition to 15:2, which identifies the content of the ‘discussion’ between Paul and Barnabas and these men: elegen gar o Paulo" menein outw" kaqw" episteusan diiscurizomeno" (Dd b gig w vgms syrhmg mae = TO). Thus in response to the claim that Gentiles should keep Torah, Paul asserts that ‘they should remain just as they were when they believed’.[126] This significantly shapes the context in which Peter’s statement ‘Now therefore why do you make trial of God by putting a yoke upon the neck of the disciples which neither our fathers nor we have been able to bear’ (15:20).[127] Note that the Western text signals its agreement with Peter by the addition of tw pneumati in 15:7 at the outset of his speech (cited above), and by recording a general agreement with Peter in v. 12: sugkatatiqemenwn de twn presbuterwn toi" upo tou Petrou eirhmenoi" esighsen (Dd [l] syrh* Ephr. = TO). These alterations result in a presentation of the council as a discussion of the principle of Torah-keeping among Gentile Christians. The different form of the decree itself (especially the presence of the golden rule) is in conformity with this.

          This agreement is clearly reinforced in the later passage about the decree (Acts 21:20-25). In this passage the decree (v. 25) is mentioned in the context of a report that many Jews had been concerned that Paul was teaching Jews to forsake Moses (see v. 20f.: ‘they have been told about you that you teach the Jews who are among the Gentiles to forsake Moses, telling them not to circumcise their children or observe the customs’). Thus Paul must give witness that he himself remains Torah-observant (v. 24), but what about the Gentiles? In the Western tradition we find the following (at 21:25):

But as for the Gentiles who have believed, they have nothing to say against you, for (ouden ecousin legein pro" se hmei" gar: D gig sa = TO) we have sent giving judgement that they should observe nothing of the sort except (krinante" mhden toiouton threin autou" ei mh: Dd ˜ gig syrh [C E Y 1739 pc]= TO) that they should abstain from what has been sacrificed to idols and from blood and from unchastity.

It is clear that ‘keeping nothing of the sort’ in the Western witnesses here refers to observance of the Torah as a whole; rather Gentile believers should should merely assent to the three items given. The terms of the (Western) decree do not in any way enforce Torah observance upon Gentile Christians. This understanding is also reflected in the actual text of the decree, which in the Western version of 15.29 reads:

that you abstain from what has been sacrificed to idols and from blood and from unchastity, and whatever they do not wish to be done to them, do not do to others (kai osa mh qelete eautoi" ginesqai eterw mh poein: Dd 242 323 522 614 945 1739 1799 1891 2298 2412 b l p syrh* sa (eth) IrenLat Cyprian = TO). If you keep yourselves from such things you do well, being sustained by the Holy Spirit (feromenoi en tw agiw pneumati: Dd l IrenLat Tert. Ephr. = TO). Farewell.

          The evidence adduced (classified as TO by Boismard and Lamouille) suggests that the Western text must be understood as a thoroughgoing attempt to address the question of Gentile Torah-observance in a more decisive manner than the Alexandrian text allows. The focus is not on table fellowship and food laws, but quite specifically upon the place of Torah in the life of believing Gentiles, and the answer given is negative.[128] The Western form of the decree emerges as a Christian, ethical document, plainly stating total freedom from Torah.

          It seems more likely to this writer that this Western form would have arisen in reaction to the Alexandrian tradition rather than vice versa. The alternative would require an Alexandrian redaction in a legalistic or historicising direction for which there is no evidence. Nor do the differences between these two approaches to the circumstances and terms of the decree suggest two versions from a single author. More plausible is the view of Zuntz, who described this case as ‘the extreme instance of. . .neglect of historical tradition in favour of moral ecclesiastical teaching’.[129]

V. Conclusions

The weaknesses perceived in Blass’ original hypothesis remain obstacles to the modern revival of the theory that Luke wrote two versions of Acts. We have seen that the Western text tends to include, specify, and emphasise matters in a way that is paralleled in secondary textual traditions in other parts of the New Testament (and in the re-use of New Testament material in subsequent non-canonical works). Recent attempts to either minimise the differences between the text-types (Strange) or distinguish between a pure and a later form of the Western text (Boismard and Lamouille) do not adequately account for the evidence we have surveyed. The consensus position in favour of originality of the ‘Alexandrian’ form of the text is preferable (as giving a plausible account of the theological tendencies of the ‘western’ tradition). In our opinion, the stylistic and vocabulary similarities between the two text-types (emphasised in particular by Boismard and Lamouille) do not represent an insuperable objection to this position (such stylistic similarities are a predictable outcome of such a revisionary practice).

          There remain significant problems which we have not addressed. Can the origin of the Western tradition be traced to one major revision with later accretions? Where might this activity have taken place? Was Acts affected to such an extent because it ‘achieved’ canonical status at a later date than, for example, the Gospels, and thus suffered more from ‘free treatment’? Notwithstanding the general conclusion that the Western text is secondary, might it occasionally record authentic historical information?[130] Although this possibility cannot be excluded, my own impression is that this is unlikely, as Kenyon concluded: ‘the attractive character of several of the d readings in Acts is to a considerable extent offset by the questionable company in which they are found’.[131]

          In many ways the Western form of Acts witnesses to a text of Acts put to use in the service of the church. Its expansions ‘provide the preacher and the missionary with suitable examples from the life of his authoritative predecessors and. . .give concrete directions for Christian life as it was meant to be lived in early Christian communities’.[132] As such these variations can be of use to the exegete in offering the earliest commentary on Acts. The argument of this essay, however, points to the likelihood that the Alexandrian text provides the closer approximation to the original text of the author, and should therefore be the primary object of the exegete’s attentions.[133]



[1]Griesbach suggested that ‘Western’ (occidentalis) was an appropriate term for this recension in view of its use ‘subsequent to the time of Tertullian by the Africans, Italians and Gauls and other Western nations’ (‘Prolegomena’ of Novum Testamentum Graece. Textum ad fidem Codicum versionum et patrum recensuit et lectionis varietatem adjecit [Second Edition; 2 vols; London & Halle, 1796 & 1806] vol. 1, LXXIV). This term has remained, despite the widespread recognition that numerous witnesses for this type of text are not from the West at all (something Griesbach was not unaware of).

[2]Bezae lacks the following passages: 8:29-10:14; 22:10-20; 22:29-end.

[3]For general information on the terminology and practice of NT textual criticism see B.M. Metzger, The Text of the New Testament: Its Transmission, Corruption, and Restoration (3rd ed; NY & Oxford: OUP, 1992) or B. & K. Aland, The Text of the New Testament. An Introduction to the Critical Editions and to the Theory and Practice of Modern Textual Criticism (ET E.F. Rhodes; Grand Rapids: Eerdmans & Leiden: Brill, 1987).

[4]Kenyon arrived at 8.5% by comparing the (Alexandrian) text of Westcott & Hort (18,401 words) with the (Western) text of A.C. Clark (19,983 words), see ‘The Western Text in the Gospels and Acts’, Proceedings of the British Academy 24 (1939) 287-315; p. 310.

[5]B.M. Metzger, A Textual Commentary on the Greek New Testament (London: UBS, 1975). The general discussion on pp. 259-272 is an excellent introduction to the problem.

[6]In a letter to R. Simon, see W.A. Strange, The Problem of the Text of Acts (SNTSMS 71; Cambridge: CUP, 1992) 3 & 205.

[7]For bibliographical information see A.F.J. Klijn, A Survey of the Researches into the Western Text of the Gospels and Acts (Utrecht: Kemink, 1949); A Survey of the Researches into the Western Text of the Gospels and Acts: Part Two 1949-1969 (NovTSS 21; Leiden: Brill, 1969).

[8]Although Bengel (1687-1752) and others had doubted many of the readings of the textus receptus, it was Griesbach who first abandoned the textus receptus in print (see Metzger, Text, 121; W.G. Kümmel, The New Testament: The History of the Investigation of its Problems [London: SCM, 1973] 74).

[9]‘Prolegomena’, NTG (1796) LXXIVf.; also Metzger, Text, 119f.

[10]According to Griesbach, Western readings occured in D, the Latin versions and older Latin Fathers as well as in the Syriac and Arabic translations (‘Prolegomena’, NTG (1796) LXXIV; cf. his earlier comments in ‘Praefatio’ § 2.5, Novum Testamentum Graece. Textum ad fidem Codicum, versionum et patrum emendavit et lectionis varietatem adjecit [Halae: vol 1, 1777; vol 2, 1775] vol 1, iii-xxxii: this volume deals with the Gospels and Acts).

[11]In fact, of course, B.F. Westcott & F.J.A. Hort believed in four groups: Syrian, Western, Alexandrian and Neutral (see ‘Introduction’, The New Testament in the Original Greek (2 vols; London: Macmillan & Co. Ltd, 1881) II, 178f. for summary; for discussion see Metzger, Text, 129-137). But as modern critics regard Sinaiticus and Vaticanus as generally ‘Alexandrian’ the language used above is not untenable.

[12]Some Western readings have been accepted as genuine and original, but this is generally despite their ‘Westernness’ (e.g. Mk. 2:22; 11:6; Lk. 24:52, 53) or because they in fact vary from Western tendencies (e.g. non-interpolations).

[13]The main witnesses for the Western text are Codex Bezae (D), three small papyri (29 38 48), a thirteenth century minuscule (614), a fifth century coptic manuscript in Middle Egyptian (mae), readings in the Harclean Syriac some of which are marked in that text by an asterisk (syrh*) others stand in the margin (syrhmg), an Old Latin manuscript h and citations in early Latin Fathers. See Aland, Text or Metzger, Text for descriptions of manuscripts and for fuller treatments of the main witnesses see M.-E. Boismard & A. Lamouille, Le Texte Occidental des Actes des Apôtres: Reconstitution et Réhabilitation (Synthèse 17; Paris: Editions Recherche sur les Civilisations, 1984; 2 vols) I, 11-95. Also available is B.M. Metzger, The Early Versions of the New Testament: their origin, transmission and limitations (Oxford: Clarendon, 1977).

[14]See their discussion in ‘Introduction’, 120-126, 172-175. A few years later E. Nestle was clearly aware that his position, favouring the Western text, was a minority position (Introduction to the Textual Criticism of the Greek New Testament [London: Williams & Norgate, 1901] 222, 226 = Einführung in das Griechische Neue Testament [ed 2; Göttingen: Vandenhoeck & Ruprecht, 1899] 185, 189f.).

[15]The Text of Acts (Vol III; The Beginnings of Christianity: Part I: The Acts of the Apostles [London: MacMillan & Co., Ltd., 1926]) esp. ccxv-ccxlvi.

[16]‘The Western Text in the Gospels and Acts’ (see n. 4).

[17]‘The Text of Acts: An Urgent Critical Task’, Journal of Religion 21 (1941) 421-431, cited here from H. Greeven (ed.), Studies in the Acts of the Apostles (London: SCM, 1956) 84-92.

[18]The Theological Tendency of Codex Bezae Cantabrigiensis in Acts (SNTSMS 3; Cambridge: CUP, 1966).

[19]‘The Provenance of the Interpolator in the ‘Western’ Text of Acts and of Acts Itself’, NTS 12 (1965) 211-230.

[20]‘Entstehung, Charakter, und Herkunft des sogenannt westlichen Texte: Untersucht an der Apostelgeschichte’, ETL 62 (1986) 5-65.

[21]Metzger, Textual Commentary, 272.

[22]W.W. Gasque suggests that Ropes’ discussion convinced ‘most scholars’ (A History of the Criticism of the Acts of the Apostles [Grand Rapids: Eerdmans, 1975] 98), unfortunately the textual problem receives little treatment in his work.

[23]A Commentary on the Acts of the Apostles (BNTC; London: Adam & Charles Black, 1957) 48-53; cf. also his Alterations to the Text of the Synoptic Gospels and Acts (Oxford: Blackwell, 1951) 54-82.

[24]The Acts of the Apostles: The Greek Text with Introduction and Commentary (Grand Rapids: Eerdmans, 1951, reprint of 1975) 43f.

[25]The Acts of the Apostles: A Commentary (Philadelphia: Westminster, 1971; ET from 14th German edition; Göttingen: Vandenhoeck & Ruprecht, 1965) 50-59; cf. also ‘Zum Text der Apostelgeschichte’, ZTK 54 (1957) 22-55; and (with P. Weigandt), ‘The Original Text of Acts?’, NTS 14 (1967) 469-481.

[26]Die Apostelgeschichte 1. Teilband (Apg 1-12) (EKK; Zürich: Benziger, 1986) 53-55.

[27]A Commentary on the Acts of the Apostles (Hermeneia; Philadelphia: Fortress, 1987) xxxv.

[28]Die Apostelgeschichte. Erster Teil: Einleitung. Kommentar zu Kap. 1,1 - 8,40 (HTKzNT; Freiburg/Basel/Wien: Herder, 1980) 167f.

[29]The Book of Acts in the Setting of Hellenistic History (ed C.H. Gempf; WUNT 49; Tübingen: J.C.B. Mohr, 1989) 55.

[30]On pp. 193-201 Hemer discusses a long list of passages where the Western and Alexandrian texts differ (1:23; 2:17-21; 3:11; 4:6; 5:18, 21; 7:43; 11:20; 11:28; 12:10; 12:25; 13:8; 13:14; 13:33; 14:13; 14:25; 15:20, 29; 15:34; 16:12; 18:7; 18:21; 19:9; 19:14, 16; 19:28; 20:4; 20:15; 21:1; 21:16; 27:5; 28:16), and where the Western variations ‘involve additional or discrepant historical details’, (p. 200). He did accept (like many scholars) that the western text might occasionally preserve a correct tradition or inference, and suggested that ‘the reviser had some knowledge of Asia Minor, as passages touching Lystra, Ephesus, Trogyllium and elsewhere can show’, (p. 200).

[31]F. Blass, ‘Die Textüberlieferung in der Apostelgeschichte’, TSK 67 (1894) 86-119 and Acta Apostolorum sive Lucae ad Theophilum liber alter. Editio Philologica (Göttingen: Vandenhoeck & Ruprecht, 1895). For an exposition in English see ‘The Question of the Double Text in St. Luke’s Gospel and in the Acts’ and ‘The Proofs for Two Distinct Text in the Acts’, in Philology of the Gospels (London: Macmillan, 1898) 96-112 & 113-137. Cf. also W.M. Ramsay, ‘Professor F. Blass on the Two Editions of Acts’, The Expositor, Series 5, Vol 1 (1895) 129-142, 212-225.

[32]Acta Apostolorum, 31f. (cf. Strange, Problem, 5).

[33]T. Zahn, Introduction to the New Testament (trans from 3rd German edition; M.W. Jacobus; Edinburgh: T. & T. Clark, 1909; 3 vols) vol. 3, 8-41.

[34]Introduction, 224. (See n. 14.)

[35]‘Two Notes on Acts’, ZNW 20 (1921) 41f.

[36]The Acts of the Apostles Translated from the Codex Bezae with an Introduction on its Lucan Origin and Importance (London: SPCK, 1923).

[37]Blass maintained that the reverse situation applied to the Gospel: the fuller first version was sent to Theophilus (and is Alexandrian) while the revised, shorter version was left in Rome (Western); this reversal attracted criticism from A. Jülicher (An Introduction to the New Testament [trans by J.P. Ward from 1900 2nd ed.; London: Smith, Elder & Co., 1904] 452-455); cf. also Kenyon, ‘Western Text’, 299-301.

[38]Especially T.E. Page, ‘Blass’ Edition of Acts’, Classical Review 11 (1897) 317-320 (Metzger, Textual Commentary, 262f. has substantial quotations); Ramsay, ‘Are There Two Lukan Texts of Acts?’, 466-469; also B. Weiss, Der Codex D in der Apostelgeschichte (TUGaL 2.1; Leipzig: J.C. Hinrichs, 1897).

[39]So, e.g. Williams, Alterations, 54-82; following M.-J. Lagrange, Introduction à l’étude du Nouveau Testament: II. Critique Textuelle; II: La critique rationnelle (EBib; Paris: J. Gabalda, 1935, 2nd ed.) 390-393. For a sympathetic assessment of the broad principle see P.M. Head, ‘Christology and Textual Transmission: Reverential Alterations in the Synoptic Gospels’, NovT 35 (1993) 105-129.

[40]See, e.g. Ramsay, ‘Are There Two Lukan Texts of Acts?’, 462-464. Ramsay was favourable to the historical information contained at certain points in the Western text; but he regarded it as generally secondary, suggesting ‘the first beginnings of Pauline legend’, St. Paul the Traveller and the Roman Citizen (London: Hodder & Stoughton, 1895; cited from 1903 seventh edition) 26. Also A. Harnack, Das Aposteldecret (Act. 15:29) und die Blass’sche Hypothese (Berlin: Reimer, 1899) = Studien zur Geschichte des Neuen Testaments und der alten Kirche. I: Zur neutestamentlichen Textkritik (AzK 19; Berlin/Leipzig: de Gruyter, 1931) 1-32. More recently Haenchen cites the divergent forms of the apostolic decree as refuting Blass (Acts, 51).

[41]‘Are There Two Lukan Texts of Acts?’, 461.

[42]The Acts of the Apostles. A Critical Edition with Introduction and Notes on Selected Passages (Oxford: Clarendon, 1933) xxif, xlviii-l (following P. Corssen, ‘Acta Apostolorum ed. F. Blass’, Göttingische gelehrte Anzeigen 158 [1896] 425-448). He draws attention to the following passages: 1:2; 4:6; 13:19f.; 13:33; 15:1f.; 15:20, 29; 18:19; 18:26f.; 19:14-16, cf. 21:16; 20:4. Note that this is one of the few matters upon which Clark and Ropes were in agreement (cf. Ropes, Text of Acts, ccxxix).

[43]See also his earlier work: The Primitive Text of the Gospels and Acts (Oxford: Clarendon, 1914).

[44]Clark, Acts, viii; cf. xlv-lii for the methods of the abbreviator.

[45]B.H. Streeter’s positive review is an exception, ‘The Primitive Text of the Acts’, JTS 34 (1933) 232-241.

[46]Kenyon, ‘Western Text’, 301-309.

[47]‘The Western Text of Acts: some Lucanisms in selected sermons’, JBL 104 (1985) 637-650. A ‘genuine Lucanism’ is defined as: ‘(1) a word or expression used only in Luke-Acts; (2) a word or expression used especially by the author of Luke-Acts when another word or expression might have been expected; or (3) a word or expression used in a manner we have come to regard as characteristically Lukan’, (p. 637).

[48]‘“Lucanisms” in the Western Text of Acts? A reappraisal’, JBL 107 (1988) 695-707.

[49]‘“Lucanisms” in the Western Text of Acts? A reappraisal’, 698.

[50]‘The Western Text of Acts: some Lucanisms in selected sermons’, 650.

[51]‘The Presence and Significance of Lucanisms in the ‘Western’ Text of Acts’, JSNT 39 (1990) 74; see pp. 70-72 on MacKenzie.

[52]M.-E. Boismard & A. Lamouille, Le Texte Occidental des Actes des Apôtres: Reconstitution et Réhabilitation (Synthèse 17; Paris: Editions Recherche sur les Civilisations, 1984; 2 vols). Henceforth cited as B & L.

[53]See M.-E. Boismard, ‘The Text of Acts: A Problem of Literary Criticism?’, New Testament Textual Criticism: Its Significance to Exegesis (FS B.M. Metzger; ed. E.J. Epp & G.D. Fee; Oxford: Clarendon, 1981) 147-157.

[54]They list over 950 examples of Lukan style in B & L, II, 195-335, sorted by alphabet, frequency and verse order, comparing frequency with Acts, Luke and the rest of the NT; and into categories: A (100% of occurences are in Luke-Acts); B (>80% are in Luke-Acts); C (>60% are in Luke-Acts).

[55]B & L, I, 97-118.

[56]‘Luc aurait écrit une première rédaction des Actes dont nous trouvons un écho dans le texte Occidental; il aurait, un certain nombre d’années plus tard, profondément remanié son oeuvre primitive, non seulement du point de vue stylistique, comme le voulait Blass, mais encore du point de vue du contenu. Ces deux rédactions auraient été ensuite fusionnées en une seule pour donner le texte actuel des Actes, plus exactement le texte Alexandrin (sous une forme plus pure que celle que nous avons maintenant)’. B & L, I, 9; ET from Strange, Problem, 31.

[57]B & L, I, 10 citing Ropes, Text of Acts, [no page given]; examples of such re-writing, attributed to TO2, are discussed in B & L, I, 111-118.

[58]Boismard & Lamouille have recently argued that the Alexandrian text is the third edition of Acts (Acts III); the reconstructed Western text (Acts II) helps in unravelling Luke’s sources (Acts I), M.-E. Boismard and A. Lamouille, Les Actes des deux Apôtres (3 vols; Études Bibliques NS 12-14; Paris: Gabalda, 1990). For a summary and sympathetic review see J. Taylor, ‘The Making of Acts: A New Account’, RevBib 97 (1990) 504-524.

[59]See further F. Neirynck & F. van Segbroeck, ‘Le texte des Actes des Apôtres et les caractéristiques stylistiques lucaniennes’, ETL 61 (1985) 304-339.

[60]See Birdsall’s review in JTS 39 (1988) 571-577; and Elliott’s in NovT 29 (1987) 285-289. Elliott (p. 286) draws attention to readings supported only by the Ethiopic (e.g. 27:38-40; 28:11-16) or a single Old Latin MS (27:2 based on h) or other single sources, as well as readings supported only by conjecture (13:27; 19:14).

[61]See Birdsall’s review (p. 576); and B. Aland, ‘Entstehung, Charakter, und Herkunft des sogenannt westlichen Texte’, who advocates a layered approach to the development of the Western text.

[62]‘The Presence and Significance of Lucanisms in the ‘Western’ Text of Acts’, 67.

[63]See especially Geer, ‘The Presence and Significance of Lucanisms in the ‘Western’ Text of Acts’, 70, 74; cf. also Birdsall’s review (p. 575).

[64]This is denied by J. Murphy-O’Connor in his review of B & L (RevBib 93 [1986] 599) but examples and parallels can be found in second century redactions of gospel material. Thus, even in the new material, the Secret Gospel of Mark bears significant stylistic parallels with Mark’s Gospel and Papyrus Egerton 2 contains significant parallels with John’s Gospel. Other examples of this type of situation would probably bear investigation (e.g. the various re-presentations of Matthew’s Gospel in the Ebionite and Nazarene Gospels).

[65]See below for the influence of passages such as 3:6 or 4:8; and cf. G. Zuntz, ‘On the Western Test of the Acts of the Apostles’, Opuscula Selecta: Classica, Hellenistica, Christiana (Manchester: Manchester University Press, 1972; essay originally dated 1939) 189-215.

[66]Cf. Geer, ‘The Presence and Significance of Lucanisms in the ‘Western’ Text of Acts’, 71f.

[67]Concerning Bezae see D.C. Parker, Codex Bezae: An early Christian manuscript and its text (Cambridge: CUP, 1992) 198-228 and his summary statments on pp. 247f and 279f. In general see T. Baarda, ‘DIAFWNIA - SUMFWNIA: Factors in the Harmonization of the Gospels, Especially in the Diatessaron of Tatian’, Gospel Traditions in the Second Century: Origins, Recensions, Text, and Transmission (ed. W.L. Peterson; Christianity & Judaism in Antiquity 3; London & New York: University of Notre Dame, 1989) 133-154.

[68]See n. 6.

[69]Thus ‘the copy of Acts from which the Western text was derived was an autograph annotated by the author’ (Strange, Problem, 175).

[70]For a general discussion see my review in EvQ (forthcoming).

[71]Evidence (admittedly inconclusive, although not mentioned by Strange) which suggests knowledge of Acts among the Apostolic Fathers and Justin Martyr is given by B. Weiss, A Manual of Introduction to the New Testament (ET A.J.K. Davidson; New York: Funk & Wagnalls, 1889; 2 vols) I, 39, 69; additional material is discussed in Haenchen, Acts, 3-14. According to patristic sources Acts was known (but rejected) by second century groups including: Ebionites (acc. Epiph., Haer. 30.16); Marcionites (acc. Tert., Adv. Marc. V.2), Severians (acc. Euseb, EH IV.29), Manichaeans (acc. August., De Util. Cred. II.7). In view of this, and the much later testimony of Chrysostom: ‘To many persons this Book is so little known, both it and its author, that they are not even aware that there is such a book in existence’ (Homilies on Acts, NPNF I.XI. [1889] 1) it is dangerous to assume the non-existence of the document on the basis of its non-appearance in the literature of the period.

[72]It should not be supposed that all passages containing Western variants exhibit distinctive theological tendencies. Many involve historical, geographical or grammatical differences from the Alexandrian text. Most scholars have focussed on theological differences as signalling a pattern of secondary development in the Western tradition more clearly than other sorts of differences, hence we have chosen to focus on these passages in what follows.

[73]Text of Acts, ccxxxiii.

[74]See especially P.H. Menoud, ‘The Western Text and the Theology of Acts’, SNTS Bull 2 (1951) 19-32 (cited here from his Jesus Christ and the Faith: A Collection of Studies [PTMS 18; Pittsburgh: Pickwick, 1978] 61-83); Williams, Alterations, 54-75; Epp, Tendency. There are at least two categories that I have not dealt with in this paper. The supposed anti-Jewish tendency (mentioned by Ropes, cited above; and discussed by Menoud, ‘Western Text’, 73f. and most emphatically by Epp, Tendency; and ‘The “Ignorance Motif” in Acts and Anti-Judaic Tendencies in Codex Bezae’, HTR 55 [1962] 51-62), the key passages include 3:17; 4:9; 13:45; 14:2-7; 18:12f. (reservations are voiced by C.K. Barrett, ‘Is There a Theological Tendency in Codex Bezae’, Text and Interpretation: Studies in the New Testament Presented to Matthew Black [ed. E. Best & R. McL. Wilson; Cambridge: CUP, 1979] 15-27). Secondly a possible anti-woman tendency (Ropes, Text of Acts, ccxxxiv; Menoud, ‘Western Text’, 76f.; B. Witherington, ‘The Anti-Feminist Tendencies of the “Western” Text in Acts’, JBL 103 (1984) 82-84), the key passages are 1:14; 17:4, 12, 34; 18:26.

[75]The lists of witnesses given in the following sections refers to divergences from NA26 and are derived from four sources: Clark, Acts; NA26; B & L, vol. two; Strange, Problem. In general I have followed B & L as the authoritative collection of manuscript and versional evidence (although I have used the sigla of NA26 except where citing directly from B & L). I have not cited all the patristic witnesses given in B & L and Strange. I have also noted B & L’s classification of the variants [TO, TO2 etc.]. Plans exist for a major critical edition of Acts (see C.D. Osborn, ‘The search for the original text of Acts; the international project on the text of Acts’, JSNT 44 (1991) 39-55.

[76]Ihsou Cristou tou kuriou is read by S.01 A 36 1175 pc vg. B & L’s TO lacks tou kuriou (with 808 1522 1831 ethms).

[77]B & L’s TO lacks kuriou: a(629) vg(D*). This is an example of the slim attestation available for B & L’s TO: the Latin column of a fourteenth century minuscule (Gregory-Aland: 629) and the original of a ninth century manuscript of the vulgate.

[78]B & L make no reference to this reading (II, 80).

[79]B & L’s TO lacks kuriou: ndl.1 (i.e. Netherlands version, from Latin: BM MS Ad. 26663) Tert. NB no Greek witnesses, very weak attestation for this reading.

[80]B & L’s TO lacks kuriou: Vg(S) Ps.-Vg(codd) Eth.7 Geo. N.b. ‘b’ =  University of Michigan MS 164 (Beuron 63).

[81]Cf. Clark, Acts, 138.

[82]B & L’s TO lacks cristou: IrenLat PhilM Boh(FS).

[83]B & L’s TO reads ton kurion Ihsoun (with C 33 257 pc mae).

[84]B & L’s TO lacks cristou: (383) 1175 Eth3.10.

[85]Cf. Acts 4:30.

[86]Cf. Clark, Acts, 55.

[87]Bezae deficient here. Note also the addition of paracrhma in Ee g m p r sa mae eth12mss.

[88]This verse is attested in several different forms, and I have oversimplified the situation somewhat here.

[89]See Head, ‘Christology and Textual Transmission’ for discussion and examples.

[90]Strange, Problem, 48.

[91]B & L, I, 107: ‘Le TO, au moins sous sa forme primitive (cf. infra), a tendance à supprimer les titres christologiques, alors que la tendance des copistes est manifestement à les multiplier’.

[92]B & L, I, 118: ‘C’est à TO2 qu’il faut attribuer la multiplication des formules «au nom de Jésus Christ», surtout à propos des miracles, comme aussi l’addition des titres «Christ» et «Signeur» dès que le nom de Jésus apparaît; à ce point de vue, le TO était beaucoup plus sobre que le TA’.

[93]Strange, Problem, 49 (see pp. 48-50 for further critical discussion).

[94]This is not to deny that some western variants may be secondary (or tertiary) developments; examples of this can be seen in various places, including readings unique to D (e.g. 3:13: cristo" added to Ihsoun; 16:40: kurio" in an added sentence); and readings attested among the more marginal western witnesses (e.g. at 11:17 a western addition has three forms, probably reflecting a continuous pattern of development, see below).

[95]Strange, Problem, 69.

[96]Strange gives a full list of versional and patristic support, Problem, 69; cf. p. 199f. (cf. B & L, II.61). Bezae deficient. N.b. ‘y’ indicates readings from Missale Mixtum (Migne, Pl. LXXXV),  cited from B & L.

[97]Metzger, Textual Commentary, 359.

[98]Origen, Contra Celsum, I.3; Strange cites other evidence from Pliny, Justin and Origen in support (Problem, 69f.).

[99]In addition, as Strange acknowledges, Irenaeus quotes the verse with enthusiastic approval (Against Heresies, III.12.8).

[100]Zuntz, ‘Western Text’, 193.

[101]The addition to the final verse in Acts provides a fitting summary to this category of Western variant: ‘dicens quia hic est Christus Jesus filius dei per quem incipiet totus mundus iudicari’ [m p dem vgmss syrh Ephraim Spec.] = oti outo" estin Ihsou" o uio" tou Qeou, di ou mellei olo" o kosmo" krinesqai. [TO2]

[102]Cf. a similar addition of to agion at 8:18 (45 74 A C D E Y ˜ latt syr bo = TO2) in a wide range of both Westen and non-Western sources.

[103]Cited from B & L, II, 79. bhm ndl.2 prv tpl are secondary versions (from the Latin). This provides clear evidence for a continued development of the western tradition.

[104]E.g. Lagrange, Critique textuelle, 389-394; Williams, Alterations, 56; Epp, Tendency, 116-118.

[105]So especially Zuntz, ‘Western Text’, 194f. [noting also 17:15D & Ephraem; 24:10; 26:1]. In a thorough study M. Black detects three categories where the Spirit: i) inspires utterance; ii) directs action; or iii) is the pre- (or post-) baptismal Spirit; ‘The Holy Spirit in the Western Text of Acts’, New Testament Textual Criticism: Its Significance to Exegesis (FS B.M. Metzger; ed. E.J. Epp & G.D. Fee; Oxford: Clarendon, 1981) 160-178.

[106]For similarly negative assessments of this argument see Menoud who said that ‘there is no theology of the Spirit peculiar to the Western writer’, ‘The Western Text and the Theology of Acts’, 75 [= p.30 SNTS Bull].

[107]Strange, Problem, 53; cf. earlier Lagrange, Critique textuelle, 391; supplemented by Williams, Alterations, 57f. (noting also 2:47; 5:38f.; 14:25; 16:35). Other passages such as 14:10 (the addition of euqew" [1838 mae eth10mss] or paracrhma [Ee] or euqew" paracrhma [D syrhmg]) should also be noted.

[108]We might also note, in connection with the status of Peter, the following: 2:14, where Peter is described as lifting up his voice first (prwto": D* vgmss mae; proteron: E); 5:29, where mention of the apostles is deleted so that Peter answers the Council charge alone (D [h syrp]); 8:24, where Simon Magus is reduced to tears (D* [syrhmg] mae); 9:34, where the long addition concerning Peter’s mission in 11:2 (D [p w mae], see Strange, Problem, 77f. for text and apparatus); 15:7 (see above).

[109]Strange, Problem, 53 has a full apparatus. Some Western witnesses also add estwto" en mesw autwn at the end of 6:15 (D h t mae), thus Corssen: ‘Stephanus steht vor seinem Richtern, er ist der Mittelpunkt der ganzen Scene’ (‘Acta Apostolorum ed. F. Blass’, 434; cited from Williams, Alterations, 57).

[110]Strange, Problem, 53-56. Also Zuntz: ‘the Apostles never act, decide, or speak without their dependence on their Lord and Master being stressed by themselves, or by the narrator’ (‘Western Text’, 194). This also helps explain the addition of ‘in the name of the Lord’ in 6:8; 9:14, 17; 14:10 etc.

[111]Strange notes the use of makavrio" and other titles in references to Paul in the Apostolic Fathers (e.g. 1 Clem 47:1; Ignatius, Eph. 12.2; Polycarp, Phil. 3.2), Problem, 214 (n. 53 to p. 54).

[112]Problem, 55.

[113]Acts of Paul, 7.

[114]Acts of Paul, passim.

[115]Acts of Paul, 25.

[116]See further D. MacDonald, ‘Apocryphal and Canonical Narratives about Paul’, in W.S. Babcock (ed), Paul and the Legacies of Paul (Dallas: SMU Press, 1990) 55-70 and C. M. Martini, ‘La tradition textuelle des Actes des Apôtres et les tendances de l’Eglise ancienne’, Les Actes des Apôtres. Traditions, rédaction, théologie (BETL xlviii; Gembloux: Duculot/Leuven: Leuven University Press, 1979) 21-35.

[117]See M. Simon, ‘The Apostolic Decree and its Setting in the Ancient Church’, BJRL 52 (1970) 437-460; D.R. Catchpole, ‘Paul, James and the Apostolic Decree’, NTS 23 (1977) 428-444; G. Zuntz, ‘An Analysis of the report about the “Apostolic Council”’, Opuscula Selecta: Classica, Hellenistica, Christiana (Manchester: Manchester University Press, 1972, essay from 1939) 216-251. Some of the historical and theological difficulties are addressed elsewhere in this collection (esp. D. Wenham). The language of Paul in Gal. 2:6 and his advice in 1 Cor. 8-10 do not seem (at least at first sight) congruent with his agreement to the decree in its Alexandrian form. This has led some scholars to opt for the Western version, since it provides a neater solution.

[118]See E. Bammel, ‘Der Text von Apostelgeschichte 15’, Les Actes des Apôtres. Traditions, rédaction, théologie (BETL xlviii; Gembloux: Duculot/Leuven: Leuven University Press, 1979) 439-446; C.K. Barrett, ‘The apostolic decree of Acts 15:29’, ABR 35 (1987) 50-59; T. Boman, ‘Das textkritische Problem des sogenannten Aposteldekrets’, NovT 7 (1964) 26-36; W.G. Kümmel, ‘Die älteste Form des Aposteldekrets’, Heilsgeschehen und Geschichte: Gesammelte Aufsátze 1933-1964. Marburger Theologische Studien 3 (1965) 278-288 (originally published in Spiritus et Veritas [FS K. Kundsin; ASTUL; Eutin: Andres Ozolin, 1953] 83-98).

[119]Kümmel, ‘Aposteldekrets’, (cf. Introduction, 181); UBS3 = NA26; Epp, Tendency, 109f.; Metzger, Commentary, 432.

[120]Menoud argued that this version represented the original (‘Western Text’, 65-71; following M.J. Lagrange, ‘Le papyrus Beatty des Actes des Apôtres’, RevBib 43 [1934] 168). This depended partly on the assumption that such a text would be enlarged rather than contracted, and partly on the assumption that the decree was only concerned with food laws. These assumptions cannot be maintained, so this position (dependent as it is upon very slim manuscript evidence) has been rejected by most scholars (including both Strange, Problem, 89 and Kümmel, ‘Aposteldekrets’, 282f.).

[121]Several different forms of the golden rule are found in these manuscripts, for example D has osa mh qelousin eautoi" ginesqai eteroi" mh poeite. N.b. ‘it61’ = Codex Armachanus (Trinity College, Dublin 52), cited from Strange.

[122]E.g. Metzger, Commentary, 432; Haenchen, Acts, 449f.

[123]Strange, Problem, 91-96.

[124]Strange, Problem, 104.

[125]Some other witnesses add a phrase identifying those who come from Judea as twn pepisteukotwn apo th" airesew" twn Farisaiwn (Y 614 pc syrhmg = TO2). This has a similar effect: by drawing v. 5 into closer relation with v. 1 the broader issue of Torah-observance is highlighted as the crucial matter for discussion.

[126]The Western texts also give the strong impression that these Pharisaic believers insisted on a judgement upon Paul’s position in Jerusalem. Note the addition of opw" kriqwsin ep autoi" after Ierousalhm in v. 2 (Dd 257 383 614 1799 2147 2412 syrh* = TO) and the description of these men in v. 5: oi de paraggeilante" autoi" anabainein pro" tou" presbuterou" (Dd syrhmg = TO2).

[127]There is no difference in the wording of the texts at this point. The difference arises from the changes previously made which focus attention on Torah-observance as a whole, thus inviting the understanding that the ‘yoke’ is the whole yoke of the Law (cf. Gal. 5:1-3: a passage which may in fact explain these alterations).

[128]The importance of harmonising Paul’s attitude to the law with that of Acts would be an important factor in this process.

[129]Zuntz, ‘Western Text’, 195; cf. also Bruce, Acts, 44.

[130]Kümmel: Western might be original at 12:10; 19:9; 20:4; 20:15; 27:5; Bruce: some Western additions may be original, at least a strong case could be made for their originality: 8:24 (Simon’s tears); 10:25 (Cornelius’ servant); 11:2, cf. 21:17; 12:10; 16:30; 18:4, 21; 19:9, 28; 28:16: ‘some of these added details give the impression of local knowledge, though others may have been deduced from the narrative, or even invented by the expander (though it is often difficult to see why)’ (p. 44).

[131]Kenyon, ‘Western Text’, 314; also Zuntz, ‘Western Text’.

[132]Zuntz, ‘Western Text’, 196.

[133]I would like to thank Dr S. Pickering (Sydney) for his detailed and helpful comments on an earlier draft and Dr P.E. Satterthwaite (Cambridge) for the translations from Griesbach (n. 1 and 10).