Editor's Note: See also K. D. Clarke's rebuttal of paragraphs 8-9 and Petersen's response.

Non-Roman fonts used: SPIonic

This article is also available in transliteration and text-only formats.

D. G. K. Taylor, ed. Studies in the Early Text of the Gospels and Acts. Texts and Studies, third series, vol. 1. Birmingham, U.K.: University of Birmingham Press, 1999. Pp. xiii + 283. GBP £18.99/ US $55.00. ISBN: 1-902459-03-2. In the U.S.: Text-Critical Studies, vol. 1. Atlanta, Society of Biblical Literature: 1999. ISBN: 0-88414-007-5.

1. This pleasing paperback volume contains eleven studies (ten in English; one in Spanish with an English abstract), papers delivered at the Birmingham Colloquium on the Textual Criticism of the New Testament, held on 14-17 April 1997. The volume is broadly divided into two parts: one on "theoretical issues" (as David Parker puts it in the Preface) and one on detailed studies of a specific passage or witness. Well edited by David Taylor, the volume is a welcome and useful contribution to the discipline. This review summarizes each of the chapters, followed by a few concluding comments. We turn first to the five chapters that constitute the "theoretical issues" section.

2. The first chapter, "The Oxford Debate on the Textual Criticism of the New Testament, Held at New College on May 6, 1897: An End, Not a Beginning, for the Textual Receptus," by J. L. North (pp. 1-25), performs a great service by reminding us of the influence of one's own cultural context upon one's scholarship--here, the debate between Edward Miller (defending the positions of the late J. W. Burgon, who attacked the Westcott-Hort text and defended the "Traditional Text" [the textus receptus]) and William Sanday (defending Westcott-Hort). The straw that broke Burgon's back was the publication of the British "Revised Version" of the New Testament, which abandoned the textus receptus and was translated from a Greek text close to that of Westcott-Hort. As North observes, "Burgon's world was collapsing" (p. 7). The eternal verities upon which he had founded his life were being swept away: a Unitarian was a member of the revision committee (horror!), and this Unitarian had been invited to join with the Anglicans in the eucharistic service in 1870 that had inaugurated work on the "Revised Version" (double horror!). Beloved passages (e.g., the bloody sweat of Luke 22:43-44, and the "long ending" of Mark) were being struck from holy writ as if they were the product of an inebriated copyboy. Burgon's defense of the textus receptus was often vituperative, blind, and, when viewed with our knowledge of sources today, laughably nonsensical: North cites Burgon's opinion of Codex Vaticanus (B) as "one of the most vicious [MSS] extant" (p. 5), and he also notes his stubborn defense of the Peshitta as being older than the other Syriac versions (even after the discovery of the Curetonian MS, Burgon dated Syrp to the second century!).

3. By reminding us of the human dimension to these events--North cites an 1861 reviewer characterizing Burgon's writings as "screams of fear, rage and hate" (p. 24); North also observes that "Like orthodox theology, an infallible Bible was a matter of life and death and excused any amount of bad temper and violence of language" (p. 24)--and casting up before us once again Miller's defense of Burgon's position--which characterized the work of the likes of Burkitt, J. A. Robinson, Conybeare, Lake, and others as "subjective scholarship" (how contemporary that sounds!), "shallow and delusive sciolism," and which alleged that they obtained the acquiesce of their students through the exercise of "moral terrorism" (the equivalent of modern "political correctness") (p. 18)--North demonstrates how little has changed in the intervening century. In his last paragraph, North writes of Miller: "One cannot escape the feeling that here is an old man, angry, perhaps unbalanced, petulant, outraged that fellow-Christians could write about the New Testament in the way they did." This sad history, so well chronicled by North, reminds us that while scholarship is often contentious, it becomes especially so when mixed with theology; or, as Bertrand Russell put it, "Persecution is used in theology, not mathematics."

4. The second chapter is by Larry Hurtado: "Beyond the Interlude? Developments and Directions in New Testament Textual Criticism" (pp. 26-48). Hurtado uses Eldon Epp's 1973 lecture (published in 1974) on the "Twentieth Century Interlude in New Testament Textual Criticism" and Epp's more recent "New Testament Textual Criticism in America: Requiem for a Discipline" (published in 1979) as his points of departure. Hurtado cites various signs of vigor in the discipline (among which are the appearance of the IGNTP Luke volumes, the resurrection of the series Studies and Documents, the inauguration of new series such as The New Testament Text of the Greek Fathers, the birth of the electronic journal TC: A Journal of Biblical Textual Criticism, and the advent of a cadre of "younger" scholars in the field [many of whom are now middle aged]) and concludes that things are not as bleak as Epp opined. Hurtado points to five main areas where developments have already taken place that bode well for the discipline in the twenty-first century: (1) comprehensive examination of significant witnesses; that is, monographs devoted to a single father, document, or manuscript; (2) improved knowledge of scribal habits; that is, studies of the habits of scribes in general and also of particular scribes in particular manuscripts; (3) work on the "crucial" second century, meaning the erection of empirically founded theories about the establishment and dissemination of the NT documents; (4) a wider historical awareness, which includes awareness of the fluid theological and political milieux in which the NT documents circulated, and how these influenced their text; (5) computerization, which includes not only e-journals and e-lists, but also digitization and electronic collation (which, it must be noted, always requires manual entry of the texts to be collated).

5. The third chapter (pp. 49-85) focuses on a particular text and a particular scholar's study of that text. J. W. Childers critically examines M.-E. Boismard's use of the Syriac version of John Chrysostom's Homilies on John in Boismard's Un évangile pré-johannique (2 vols., 1993-1994). The importance of the Syriac version is simple: multiple Syriac manuscripts exist, dating from between the sixth/seventh and ninth centuries. The oldest Greek manuscripts are, by contrast, only from the ninth/tenth century and later.

6. As is typical of Boismard, his theory of the development of the text of the Homilies is quite complex, and Childers reprises Boismard's ideas fairly. However, he takes Boismard to task for errors in his presentation of the Syriac text. Since no edition of the Syriac exists, Boismard's evidence comes directly from his examination of the manuscripts--as does Childers's. Broadly speaking, the errors Childers detects fall into three categories (and the pitfalls detected here are cautionary for other workers in the area). (1) Childers finds errors in reading the manuscript, in transcribing the manuscript, and in proofreading. He cites several places where his examination of the manuscript shows that Boismard's reading is apparently simply in error. At other points, there appear to be typographical errors, which of course present an erroneous reproduction of what the manuscript supposedly reads. (2) A well-known problem in working with versional evidence is setting up equivalencies between two languages, especially when dealing with items as idiomatic as particles, conjunctions and prepositions, or technical terms and word order. Here Childers faults Boismard for "never clarif[ying] his method of interpreting and presenting the evidence of the Syriac. The reader is easily misled" (p. 5). Again he offers examples of these problems. (3) The selective presentation of readings can also skew the author's (= Boismard's) own view of the evidence and of course the reader's grasp of the evidence--for the reader is dependent upon not just the acute eye of the scholar but also his evenhandedness in presentation of the variant readings. Here again Childers offers examples of the omission from Boismard's monograph of "comparable distinctive readings [which] go unnoticed."

7. This is, at thirty-six pages, the longest study by a single author in the volume. This reviewer has not checked Childers's or Boismard's readings against the Syriac manuscripts of Chrysostom's Homilies on John (most of the MSS are in the British Library). However, the cautionary points are the most significant aspect of Childers's study for the readers of TC. Childers goes on to offer an alternative reconstruction of the textual history of the Homilies, based on what is presumably a more circumspect and accurate reading of the evidence.

8. "The Construction of Biblical Certainty: Textual Optimism and the United Bible Societies' Greek New Testament," by K. D. Clarke and K. Bales (pp. 86-93), is a brief examination of the number of variants classed as "A," "B," "C," or "D" across UBS1, 2, 3, 3corr., 4, with special attention paid to the differences between the 3rd corrected edition (1983) and the most recent, 4th edition (1993). Put in a nutshell, by the authors' count, the 3rd corrected edition had 126 "A" variants, 475 "B" variants, 699 "C" variants, and 144 "D" variants. The 4th edition, however, has (of the variants common to the two editions) increased the number of "A" variants to 346 (an increase of 220 variants), 479 "B" variants (+ 4), 312 "C" variants (- 387), and 9 "D" variants (- 135). Hence, according to the authors, there appears to be an increase (= the "optimism" of the title) in the confidence with which the editors regard their work.

9. There is, however, an obvious flaw that makes the (seemingly) impressive tables of numbers assembled by Clarke and Bales pointless. They failed to ascertain whether the definitions used in GNT3 to assign the A, B, C, and D rankings are identical with those in GNT4. If they were the same, then a comparison could, conceivably, be made. But if they were different--as they are--then any comparison is meaningless. It is astonishing that such a simple, obvious, elementary check was, apparently, not undertaken, especially when it is absolutely critical to the edifice Clarke and Bales wish to erect upon it. To wit: in the most recent "fourth revised edition" of the GNT (1993), the letters A, B, C, and D, signify, respectively, "the text is certain," "the text is almost certain," "the Committee had difficulty in deciding which variant to place in the text," and "the Committee had great difficulty in arriving at a decision" (GNT4rev, p. 3*). However, in the "third edition (corrected)" of the GNT (1983; the third edition appeared in 1975), the letters A, B, C, and D, signify, respectively, that the text is "virtually certain," "there is some degree of doubt," "there is a considerable degree of doubt," and "there is a very high degree of doubt" (GNT3corr, pp. xii-xiii). The differences are obvious. For example, in the current (fourth) edition the letter "C" means that "the Committee had difficulty in deciding which variant to place in the text," while a decade earlier, in the third edition, "C" meant "there is a considerable degree of doubt." The 1993 edition references the ease with which the Committee reached its decision; the 1983 edition references the degree of doubt the Committee had. These are two entirely different things, which cannot be compared--although that is precisely what Clarke and Bales do. This carelessness on the part of Clarke and Bales means that their results are useless.

10. G. P. Farthing's contribution, "Using Probability Theory as a Key to Unlock Textual History" (pp. 94-117) is one more attempt to bring mathematics to bear on textual variation in the NT. His approach is clear enough, focusing on what he defines as "non-unique reversible changes"; that is, changes which may happen at more than one point in a stemma and which may be "reversed" ("corrected") in the course of transmission. Accompanied by more than forty tables and diagrams, Farthing's approach is clearly described. Essentially, it seeks through probabilities to determine the most likely stemmatic relationship for a group of manuscripts. Collation of a significant number of variants is necessary, after which a computer is employed to test, one by one, each of the possible stemmatic configurations.

11. The application of mathematical, computer-driven models to stemmatics is an area of growing interest, especially since some of the principles are identical to work in genetics. However, problems remain. One step in Farthing's study that remains unclear is why, on the basis of the information in Table 5 (p. 110), he writes (p. 111): "The high number of agreements between [on the one hand] [MSS] 69 and 124 and [on the other hand] between 230 and TR suggest a common tradition with 69 and 124 at one end and 230 and TR at the other." But the data in Table 5, which lists the frequency of all possible MS groupings of four manuscripts (MSS 69, 124, 230, and the TR [for the purposes of his illustrations, Farthing regards the TR as a single MS]), lists 627 instances (in his sample passage) where MS 69 stands alone against the combination of MSS 124, 230, and TR; this is the highest number in the table. The combination of MSS 69 and 124 agreeing against the combination of MSS 230 and TR (the combination Farthing has singled out as the one worthy of further investigation) occurs 198 times; it is therefore not the highest number of agreements. MS 124 stands alone against the combination of MSS 69, 230, and TR 198 times, the identical number of times as the combination (69 + 124 vs. 230 + TR) Farthing judges worthy of further investigation. (Other combinations in Table 5 occur [in descending order] 149 times, 118 times, 42, and 30 times.)

12. This reviewer fails to see how Farthing, on the basis of the data in Table 5, settled on the MS combinations 69 + 124 vs. 230 + TR as the one he selects as "the" most instructive for further development of the stemma, when (a) this combination is not the most frequently occurring combination (which is instead MS 69 vs. 124 + 230 + TR), and, even more tellingly, (b) this combination is tied (at 198 agreements in combinations) with the grouping MS 124 vs. 69 + 230 + TR. If one is relying on numeric indicators, there must be some clear numeric indicator as to which of the seven combinations is to be pursued. None is discerned here. If the highest number of agreements were to be selected, the result would be 69 vs. 124 + 230 + TR. If one were to select the lowest number, the result would be 124 + 230 vs. 69 + TR. If one were to chose the second highest number of agreements (although why one would do this is unclear), then one is faced with a tie: both 69 + 124 vs. 230 + TR and 124 vs. 69 + 230 + TR have a frequency of 198. How does one choose between them? No explanation is given.

13. A second problem is one which is all too apparent to those who have worked extensively with early manuscripts: their double proclivity for (a) changing their textual character, sometimes in medias res, and (b) incorporating "odd" readings from a widely divergent and, generally speaking, "unrelated" source. An example of this is the insertion into Luke 23:48 by the Sinaitic Syriac and by Vetus Latina MS g1 (also known as Vulgate MS G [Codex Sangermanensis]) of a snippet from the Gospel of Peter 7. To the best of my knowledge, only these two manuscripts have this insertion. We know the source (presumably the Gospel of Peter [but did Peter get it from a very early copy of Luke?--one cannot exclude that possibility]), and we know that both place the insertion at exactly the same point in the text. How does one deal with such ad hoc changes which, despite their linguistic and geographic separation, are somehow clearly linked?

14. It is to be granted that stemmatics is a search for generalized relationships among manuscripts, and here Farthing's techniques may be of help. But it is also necessary to recognize that the datum put into the computer will often--especially in the pre-500 or pre-600 evidence--be, in the case of even a single manuscript, a pastiche from various traditions (e.g., the example from Luke 23:48, above). Hence, the new "regel de fer" imposed by the computer, which mechanically (instructed by its software) seeks to determine the most likely relationships without any sensitivity to the gravitas of a reading, may well produce results which are very misleading in the particular.

15. The remaining six chapters constitute the second part of the volume and are intended to focus on a particular passage.

16. The first chapter in this section is S. R. Pickering's "The Significance of Non-Continuous New Testament Textual Materials in Papyri" (pp. 121-141). Pickering makes the observation that NT citations are sometimes found embedded in non-NT papyri. This should be obvious to most; for example, the "Unknown Gospel" (as Bell and Skeat titled their edition of P. Egerton inv. no. 2) is well known to most adepts and is often cited. Pickering argues that these quotations, if that is what they can be called (in this context, where P. Eg. no. 2 may well be the second oldest piece of the NT text extant in manuscript form [only the comparatively few broken lines of P52 could rival its antiquity], the very concept of "quotation" is probably prejudicial and hence the word should be avoided), have been ignored and are deserving of attention. He would include among these "non-continuous" texts ostraca and amulets.

17. Pickering provides a convenient catalogue (pp. 133-136) of 21 "non-continuous" items (papyri, amulet texts, ostraca) containing portions of the Gospel of John and discusses as one example the fourth century Fayum P. Abinn. 19, which contains an allusion to Mark 9:41-42 or (as Pickering would prefer) Matt 10:42. An appendix to this chapter contains a transcription of nine lines from P. Vindob. G 2312, containing quotations/allusions/parallels to Ps 90:1-2, Rom 12:1-2, and John 2:1-2, supplemented with Pickering's notes. Those who are unaware of the existence of these "non-continuous" witnesses and their value in reconstructing the earliest layers of the NT text would do well to study Pickering's examples.

18. E. Güting's article, "The Relevance of Literary Criticism for the Text of the New Testament: A Study of Mark's Traditions on John the Baptist" (pp. 142-167), deals with Mark 1:2-3 and 9:12b. In both cases, Güting argues that glosses have been taken into the text of Mark, and transmitted as "gospel," when they are not. Because there is no manuscript evidence in the case of 9:12b and the evidence is divided in the case of 1:2-3, Güting holds that "our decisive witness" in these cases must be the literary style of "Mark." He embraces the literary-critical arguments of Carl Lachmann in the case of Mark 1:2-3 and extends them by examining the syntax and the "narrative design" of Mark. In the case of 9:12b, literary criticism also leads Güting (as it has others) to the conclusion that the question contained here is a gloss that eventually found its way into the transmitted text of Mark. Güting's conclusions are twofold. First, not all textual corruptions of NT documents are reflected in the manuscript tradition. Put differently, there may be places where, despite the agreement of all extant witnesses in a particular reading, we can nevertheless be quite certain that a corruption has taken place; it is simply that textual criticism, bound as it is to the use of manuscript evidence, cannot recognize such instances. Second, "textual criticism is methodologically dependent upon the results and perspectives of literary criticism" (pp. 166-167). In short, Güting is reminding textual critics that literary-critical observations are often the key that signals the existence of a disjunction in a narrative. It is this disjunction that prompts the search for textual variants--which may or may not be found. But the failure to find textual variants does not mean that the text has not been modified at this point, for the literary-critical observation (wrong pronouns or shifts in pronouns, tenses, disjunctions in the narrative, non-standard [for that author] style, etc.) still signals the existence of textual modification, even though it is unsupported in the manuscript tradition.

19. Whether Güting is correct in the particulars of the two cases he presents (this reviewer is inclined to agree with him) is not as significant as the larger point his examples illustrate: the fundamental importance for textual critics of careful literary-critical analysis, and the existence of textual corruptions in the NT which have left no trace in our present manuscript tradition.

20. Tjitze Baarda's study, "The Beatitudes of 'The Mourning' and 'The Weeping': Matthew 5:4 and Luke 6:21b" (pp. 168-191), is executed with all the exemplary care and circumspection we have come to expect of the dean of New Testament textual criticism. Beginning with the observation that the Matthean Beatitudes' blessing of those who "mourn" (Matt 5:4) is different from the Lucan Beatitude (6:21b, which blesses those who "weep"), Baarda sets out to discover what the Diatessaron read at this point. Did Tatian harmonize this disjunction or not? And if Tatian did harmonize these verses, then how did he do it?

21. Baarda begins, as all textual studies should, with a review of previous scholarship. He notes that the reconstructions of Ortiz de Urbina, Zahn, Hill, and Plooij are misleading, in part because of laziness (Ortiz de Urbina did not search widely enough, nor did he read his texts carefully enough), limited resources and lack of facility with the languages (Zahn had only one Diatessaronic witness at his disposal, Moesinger's Latin translation of the Armenian version of Ephrem's Commentary on the Diatessaron), and wrong translations (echoing Childers's criticism of Boismard in an earlier chapter in this volume): Plooij, like Zahn, was influenced in this passage by the infelicitous Latin translation by Moesinger of the Armenian of Ephrem's Commentary. On the basis of the Syriac Ephrem and Aphrahat, Baarda concludes that the order of the Eastern witnesses of the Diatessaron was Matt 5:3; 5:5; 5:4; 5:6; Luke 6:21b; Matt 5:7. The verse order 3-5-4 is also found in Syrsin D 33, and many Vetus Latina manuscripts. In the case of Luke 6:21b, Tatian seems to have changed Luke's second person plural verb (". . . ye shall laugh") to the third person plural (". . . they shall laugh"), which brings it into agreement with Matthew's third person plurals. The third person plural in Luke 6:21b is found in Ephrem, Syrsin, W, Vetus Latina MS e, and the Sahidic, Armenian, and Ethiopic translations, as well as MarcionTert. So much for the Eastern Diatessaronic tradition.

22. Turning to the Western Diatessaronic tradition, Baarda notes that its text seems to have been a bit different. Here the Latin gospel harmony family headed by Codex Fuldensis reads Matt 5:3; 5:5; 5:4; 5:6; 5:7; 5:8; 5:9; 5:10. The inversion of Matt 5:4 and 5:5 agrees with the Eastern tradition, but at least at first sight there seems to be no sign of the insertion of Luke 6:21b into this uniformly Matthean text. Upon closer examination of the Western vernacular harmonies, however, some ambivalence arises, for while the Latin harmony tradition of Matt 5:4 correctly reads lugere ("to mourn"), the Middle Dutch Liège Harmony reads wenen ("to weep," the text of Luke 6:21b). There is some support for this reading in the Old High German (OHG) of Codex Sangallensis (which reads vvuofan). Baarda, however, remains ambivalent about this OHG reading, for he knows all too well the difficulties in determining precise translation equivalencies in dead languages long ago (note that this is one of the problems signaled by Childers in his chapter on Boismard's work).

23. Baarda concludes that Tatian did not harmonize Matt 5:4 and Luke 6:21b into a single Beatitude, but kept them separate. Tatian did, however, combine the Lucan verse with the Matthean text: in the Diatessaron, the verses were presented in the order of Matt 5:5, 4, 6; Luke 6:21b; Matt 5:7. When he combined this Lucan verse fragment with the Matthean Beatitudes, Tatian brought the grammar of Luke 6:21b into line with that of the Matthean Beatitudes: he changed Luke's second person plural verb to agree with the third person plural verbs found in the Matthean Beatitudes. Finally, Baarda observes that Tatian appears to have given a "rather peculiar translation" of Matt 5:4, by rendering the Greek parakalew with the Syriac equivalent of "to pray, to beg." In the West, he notes that an editor of the early Latin Diatessaron may have harmonized Matt 5:4 with Luke 6:21b, giving rise to the ambiguous reading of the Old High German and to the apparently harmonized reading of the Middle Dutch Liège Harmony. This particular harmonization, however, seems to have been a distinctly European tradition, for it is absent from Eastern Diatessaronic witnesses.

24. Perhaps the most significant chapter in the volume is that by three of the conference's hosts, David Parker, David Taylor (who is also this volume's editor), and Mark Goodacre. Their study, titled "The Dura-Europos Gospel Harmony" (pp. 192-228) is, in the eyes of this reviewer, the last word on the famous "Dura Fragment" (Gregory/Aland 0212). It examines the frequently cited and generally accepted contention that this small, fourteen-line parchment was from a Greek copy of Tatian's Diatessaron. This identification was first claimed by the parchment's editor, C. Kraeling, and has been accepted rather uncritically by many later writers (F. C. Burkitt, M.-J. Lagrange, H. Lietzmann, K. Aland, B. Metzger, et al.). It has been taken as proof by many of these same scholars, as well as others, that the Diatessaron was originally composed in Greek and then translated into Syriac. Nevertheless, not all scholars agreed. From the beginning, the Dutch Diatessaron expert Daniël Plooij disputed both assertions. First, he cited textual evidence which suggested that the fragment had been translated into Greek from a Syriac Vorlage. Second, he suggested that the fragment was not part of the Tatianic tradition. Rather, Plooij pointed to the existence, in Syriac, of harmonies of the Passion Narrative that were unrelated to the Diatessaronic tradition. Since the few lines of the Dura Fragment center on Joseph of Arimathaea's securing Jesus' body, the subject matter of the Fragment was not a priori incompatible with what one would expect to find in a Greek fragment translated from one of these Syriac, non-Diatessaronic Passion harmonies. The evidence amassed by Plooij, along with the independent evidence of a Syriac Vorlage adduced by Anton Baumstark, convinced many experts in Diatessaronic studies that the Dura Fragment is most likely not from a Diatessaron. As a consequence, it has nothing to contribute to discussions about the original language of the Diatessaron.

25. Parker, Taylor, and Goodacre present a five-fold analysis of the Fragment. First, there is a careful paleographic analysis. On the basis of this and the physical context of the find, they suggest that the Fragment was copied "at some point between the second part of the second century and the building of the embankment [in which it was found, and which was constructed in 254], and we would prefer a late second century date" (pp. 198-199). (This is earlier than the date assigned by C. Kraeling, the Fragment's first editor, who dated the hand to the "first half of the third Christian century"; political events mandate that the Fragment was copied before the winter of 256/257 CE.) Second, there is a careful, annotated reconstruction of the text of the Fragment, including a treatment of the nomina sacra. Third, they examine the evidence produced by Plooij and Baumstark in favor of the argument that the Fragment is a translation into Greek from a Syriac Vorlage. Our authors, in a competent and evenhanded manner, find the evidence useless. Indeed, a close examination suggests composition in Greek, not Syriac. The fourth section ventures a reconstruction of the Diatessaron's text for this passage and compares this reconstruction with the Dura Fragment's text. Of the ten textual "items" they compare, only one is "Tatianic," and only two are "possibly Tatianic." By contrast, five are "non-Tatianic." Two more are "indeterminable." Our authors conclude: "The bulk of evidence is strongly against the fragment's being a part of Tatian's Diatessaron" (p. 225). The fifth and final section seeks to determine what kind of a Greek gospel text(s) underlies the Fragment. No decisive conclusion can be reached, for readings aligned with MS B et al. (the "Alexandrian" text) are found, but so are readings distinctive of MS D (the "Western" text), as well as readings characteristic of later Byzantine witnesses. The Fragment's text fails to align decisively with any known "group" or "family" of Greek MSS.

26. It has long been known that numerous gospel harmonies circulated in the early church: Justin Martyr apparently used one in Rome years before Tatian created the Diatessaron; the "Gospel of the Hebrews" seems to have been a gospel harmony; Clement of Alexandria seems to have sometimes cited the gospels in a harmonized form, leading Cerfaux to suggest that he might have been using a harmony; we have reports from the early church that harmonies were created by Ammonius of Alexandria and Theophilus of Antioch. Parker, Taylor, and Goodacre correctly point out the danger of assuming, as many have done with the Dura Fragment, that any harmonized text from the early church is automatically to be identified with Tatian's Diatessaron. Such an identification can only be done after a careful, analytical analysis of the text, for other harmonies were known and in use.

27. In "Where is Emmaus? Clues in the Text of Luke 24 in Codex Bezae" (pp. 229-244), J. Read-Heimerdinger addresses the issue of the variant reading Oulammaous in Codex Bezae (at Luke 24:13). Citing evidences of a greater Jewish influence on Luke-Acts than is generally acknowledged, Read-Heimerdinger seeks to unravel this curious reading in MS D. The key is found in the LXX, where MSS A and 370 at Gen. 28:19 read "Oulammaus," presumably a misreading (or mistranslation?) of the Hebrew ulam luz (which means "formerly [known as] Luz"). This is apparently the source of the variant with this spelling, which then migrates into Codex Bezae's text of Luke. (This reviewer notes that the rest of the LXX tradition apparently reads "Oulamlous.") Read-Heimerdinger supports this genesis of the Bezaean reading by a list (pp. 236-237) of episodic parallels between Genesis 28 and Luke 24 (e.g., Gen. 28:11, Jacob goes on a journey to flee from his brother = Luke 24:13, two disciples are on a journey which is a flight; Gen. 28:16, Jacob awakens = Luke 24:31, the disciples' eyes are opened). Read-Heimerdinger argues that "Oulammaous" is one more element of this Jewish subtext found in Luke 24 and that the decision to replace it with "Emmaus" is the result of later, notably Gentile, Christian insensitivity to the Jewish subtext of Luke 24. Whether it was "Luke" or the redaction of Luke found in Codex Bezae which is so suffused with Jewish thought-patterns that Luke 24 should be read in the light of Genesis 28 is something that will require further study and proof.

28. The final contribution, by J. Rius-Camps, is "La utilización del libro de Joel (Jl 2,28-32a LXX) en el discurso de Pedro (Hch 2,14-21): Estudio comparativo de dos tradiciones manuscritas" (pp. 245-270). Peter's speech, which is actually a sermon, in Acts 2:14-21 is significant (for the author of Luke-Acts, at least) as the first post-resurrection proclamation of the Christian message. Rius-Camps seeks to examine the use of the book of Joel in this speech, comparing the "Western Text" version of Acts with the "Alexandrian Text" version of Acts. A "synopsis" of the two traditions, and the Joel text that is cited by "Peter," is presented on pp. 261-267, accompanied by an apparatus. The conclusion is that the Alexandrian text's version of Peter's speech contains virtually literal quotations from Joel in the LXX. The Western text's version is much freer. Rius-Camps feels that the purpose of the Joel citations in the Alexandrian text is different from that in the Western text. In the Alexandrian text, their purpose is to show that "the outpouring of the Spirit [was predicted] in the Jewish Scriptures" (to quote the English abstract, p. 270), while their function in the Western text is to create a new "prophecy": using the Joel text, with its "prophecy addressed to the remnant of Israel as a guide," Peter in the Western text fashions a new "prophecy"; namely, that what has just taken place among the Jewish disciples is only the "first fruits" of what is "destined for all people." Rius-Camps favors the Western text as the more ancient, for he presumes that the "loose" citation of Joel in the Western text eventually awakened concern in a later age more concerned with literal, verbatim quotation. He also notes that "the revised text [= the Alexandrian text] loses the nuances of Peter's theological message by substituting a more straightforward historical account."

29. A few remarks in conclusion. The presentation of the volume with footnotes is superior, although the use of semi-glossy paper by the publisher makes it more tiring on the eyes than necessary. Two indices, one of biblical citations and one of authors (often omitted in "hurry-up" publications of proceedings), are welcome helps to the reader. Typographical errors are virtually non-existent, although the bibliographic form used in notes is sometimes defective by omission. For example, in Hurtado's essay the complete title and publication datum for a frequently cited volume (it is cited in notes 34, 37, and 40-45) are missing (one finds only the "short" title, Gospel Traditions, and nothing further). But this is a minor cavil, for the overall execution of the volume is excellent.

30. For the authors, the reviewer has two general remarks. First, and astonishingly for a collection of papers on text-critical studies, crucial readings are sometimes referred to, but never reproduced for the reader. This means that one sometimes needs three editions open on one's desk in order to see the texts upon which the author is basing his argument. Why not reproduce the text if you are discussing it and if it is so critical to your case? This is not only common sense, but also common courtesy to the reader. Second, in many of the essays one finds oneself taken into the author's private world. This may be an interesting and even cozy place, but the scholarly ideal of an agnostic view of the terrain is wanting. Rather, there is a tendency to set up and then investigate a single hypothesis. Only four of the essays investigate multiple hypotheses for the reader and then weigh the pros and cons of each in order to identify the "best" solution.

31. As is to be expected in a collection of conference papers, the quality of the contributions varies. Similarly, some hew more closely to the topic of textual criticism, while other wander further afield. In general, however, the papers are of high quality, and even where they stray from textual criticism per se, they will nevertheless be of interest to textual critics. Two papers deserve special mention, for this reviewer is of the opinion that they will still be cited a century from now: Baarda's study of the two Beatitudes in the Diatessaron and the Parker-Taylor-Goodacre study of the Dura Fragment. This reviewer singles them out not because they are close to the field in which he works nor because he agrees (or disagrees) with their conclusions, but rather because they are precisely what good papers should be: painstakingly thorough, completely honest analyses of a significant problem, to which they provide the definitive solution. More than that one cannot ask.

© TC: A Journal of Biblical Textual Criticism, 2002.

William L. Petersen
Religious Studies and Classics and Ancient Mediterranean Studies
Pennsylvania State University