Non-Roman fonts used: SPIonic, SPAchmim
This article is also available in transliteration and text-only formats.
1. The need for a comprehensive edition of the New Testament does not need to be argued at length; it is obvious. The present edition aims to make available the full range of resources necessary for scholarly research in establishing the text of the New Testament and reconstructing its history. It includes all the variants found in the Greek manuscripts selected and in the quotations by the Greek Fathers. It further includes all the non-intraversional variants of the most important early versions (Latin, Coptic, Syriac), as well as the evidence of the other versions to the extent these are available in scholarly editions and research publications.
2. On the basis of this documentation, fully representing the textual evidence surviving from the first thousand years, the edition makes possible comprehensive studies in textual criticism, as well as in the history of textual transmission. It makes for a better understanding of the external criteria of textual criticism by shedding light on the textual character and value of the individual documents in the context of the whole textual tradition, and it provides the infrastructure necessary for achieving the ultimate goal of textual criticism, which is the establishment and confirmation of the text.
3. You may well ask how such goals can ever be achieved, or how they (hopefully) have been achieved. This involves the questions of:
4. The selection of the material to be presented raises a very ticklish problem, as everyone with any experience knows. On the one hand, the edition must be comprehensive, yet, on the other hand, it must not be a diffuse mass cluttered with details; therefore, it must be selective. In order to satisfy both of these requirements we have examined the textual character of all the surviving manuscripts of the Catholic Letters in a group of 98 test passages. The results (which proved to be reliable by the full collations) have shown that 372 of the 522 complete manuscripts and larger fragments of the Catholic Letters agree with the Majority Text in at least 90% of the test passages. The greater number of these almost identical manuscripts cannot be cited individually in the present edition but must be represented by a relatively small selection. All the remaining manuscripts, namely, those which differ from the majority in more than 10% of the test passages, are fully represented in the edition! In addition, 20 Apostolos lectionaries have been chosen from among 400 lectionaries on the basis of the test passages and of further full collations; these demonstrate once again the dependence of the lectionary text on the Byzantine text. Apart from these 20 Apostolos lectionaries, a total of 182 continuous-text manuscripts were selected in this way for inclusion in the editon, including several manuscripts attesting a pure or an almost pure Byzantine text, so that the substance of this text type is very well represented. (It should be noted that the list of manuscripts of the Catholic Letters comprises more entries than the list for James given in fig. 2.)
5. This process of selection ensures that the critical apparatus contains all the known readings which have appeared in the history of the text from its earliest beginnings through the formation and final establishment of the Byzantine text in the 11th/12th century (the Koine edition Kr is incorporated). The quality of what is not given in the apparatus is demonstrated by a couple of readings found in manuscripts not selected for the edition (see Supplement, p. B32). Those readings are, as everyone can see, secondary readings derived from Byzantine ones at most and having no value for establishing the original text or its history. Relatively many of the manuscripts included are closely related to the Byzantine text. These codices Byzantini are represented by the group symbol Byz. There are 97 such manuscripts summarized under this symbol. When members of this group differ from the group reading, they are always cited explicitly with the variant reading they attest, so that the manuscripts represented by the symbol Byz can readily be identified in every passage.
6. In addition to these primary witnesses, the edition includes all the Greek patristic quotations to the time of John of Damascus (7th/8th century) plus some important later authors. The difficult task of distinguishing between quotations and allusions is somewhat alleviated by the fact that the edition contains all the textual variants found in the manuscript tradition of the first millennium. The text of the Letter of James preserved in the writings of the Fathers corresponds in most instances to variants known in the manuscript tradition; in other New Testament writings the situation may differ. Readings with exclusively patristic support are cited only rarely, and usually then only if they are attributed to manuscripts which no longer survive. (Allusions have been considered only if they clearly reflect a known reading.)
7. Among the early versions, the Latin, Coptic, and Syriac are fully cited in the edition as witnesses to the Greek text. The Greek base of each has been reconstructed to the extent possible, and a descriptive index system (described below, par. 15) has enabled us to offer comments on the major versions in each passage. Thus the versions are treated as full witnesses. The lesser versions are cited selectively, only where the relation to their translation base is unequivocal. They are cited from published editions where these are available.
8. The layout of the edition is characterized by precision, simplicity, and usefulness for a wide variety of scholarly interests, and the Introduction and all the explanatory material are given in both German and English. The page layout comprises three principal elements on each page: a primary text line offering the basic text of the edition, an overview of readings exhibiting the range of variants extant for a textual unit, and a critical apparatus in which the variants are cited with their supporting witnesses. With this three-part arrangement the scholar can tell at a glance precisely what is relevant to his or her particular interests (see fig. 1).
9. The primary line presents the established text of the edition. In each verse the words of the primary line are identified by even numbers, with odd numbers reserved for the spaces between them. Accordingly every word, every sequence of words, and every space between words is identified explicitly, avoiding the necessity for a repetition of the full lemma at any point. The text of the edition has been established afresh on the basis of all the evidence assembled. In the Letter of James it differs in two instances from the text common to NA27 and UBS4 (1:22 a)kroatai\ mo/non; 2:3 h2 ka&qou e0kei=). This is not due to any unwillingness on the part of the editors to make changes but rather to the relative integrity of the tradition and also to the quality of scholarship in the Nestle-Aland and UBS editions. Of course the number of changes may be greater in other books.
10. There has been an important innovation, however, where there are variant readings which the editors consider as of almost equal value to the text. In such instances the beginning and end of the problematic reading are marked with a dot in the primary text line, and a dot is also placed before the respective variant in the overview. The wisdom both of textual critics and of exegetical scholars is especially solicited in these instances for their contribution toward solving a problem that still remains open.
11. For example, in James 2:3, the reading of ECM is su\ sth=qi h2 ka&qou e0kei=. The b-reading here, su\ sth=qi e0kei= h2 ka&qou, is read by NA27/UBS4, and it is almost of equal value with the a-reading. But we believe that the more difficult but well attested a-reading su\ sth=qi with the adverb e0kei= placed after h2 ka&qou deserves preference for the text. The reader may have been irritated by the position of the adverb, which may have been expected immediately after sth=qi (cf. variants b-e) in parallel with su\ ka&qou w{de (words 24-30). Therefore, it seems to have been transposed in variants b and c. In addition to putting e0kei= after su\ sth=qi, the Byzantine reading d augments h2 ka&qou with the adverb w{de. But every reader should notice: here the text is not absolutely safe.
12. The overview of variant readings exhibits without supporting evidence all the readings which differ from the basic text (the primary text line) immediately below the relevant words. This permits a quick survey of the existing variants, and it is particularly useful to exegetes for an impression of the quality of the tradition and the possible causes for alterations of the text. Individual variants are identified by their numerical addresses and by lower case letters. The text of the primary line (the a-reading) is not repeated. This index system provides a neat transition to the third element of the page layout: the critical apparatus.
13. Here the identifying number and letter addresses are followed by the readings and their supporting witnesses in the order of Greek manuscripts, Church Fathers, and versions. All the witnesses supporting a text unequivocally are cited in full. In order to do this a special system of markings is necessary, the meanings of which will be readily grasped by anyone using the edition. The following are the most important points.
14. The full citation of all the witnesses supporting a reading (an explicit or positive apparatus) is given only when 15 or more Greek manuscripts differ from the primary line text. The majority of the (more than 800) instances of textual variation involve only a few manuscripts at most. It would be pointless and unnecessarily repetitious to list the whole of the Greek manuscript evidence for the a-reading in each of these instances. Consequently, in these instances only the witnesses which differ from the a-reading are noted (a partially implicit or negative apparatus). Three dots (...) then stand for the Greek witnesses supporting the a-reading, followed by the Greek manuscripts requiring some qualification (because of defects or corrections), by the Church Fathers and versions. It is important to remember that all manuscripts not explicitly mentioned may be assumed to support the a-reading! In every passage all manuscripts with lacunae or other defects are listed under the sign "-", so that a simple subtraction of these is all that is required to identify all the manuscripts attesting the a-reading.
15. Another new marking used in ECM is a double arrow 
beside two or more letter addresses, indicating that it is impossible to tell which of these readings is supported by the witnesses cited.1 It also serves to indicate which readings the witnesses definitely do not support. This is one of the markings which makes it possible to treat the versions as full witnesses. If the reading of a versional witness cannot definitively be associated with one or more readings in the apparatus, it is marked with a question mark.2
16. An edition such as this should not try to anticipate the results of any research it is designed to facilitate. So the present edition attempts only to provide open access to actual evidence, without suggesting any inferences about manuscript groupings of families or about any microstructures within the manuscript tradition. Even the symbol Byz is used only as a pragmatic term to designate a group of manuscripts sharing agreement in 80% or more of the variant readings characteristic of the Byzantine text (that means in all the readings where the undivided Byzantine text differs from the primary text line). The symbol Byz is not to be understood as representing a defined entity. Thus, the present edition does not pretend to be adequate for specialised studies in the Byzantine text. It does include a great number of purely Byzantine manuscripts and also manuscripts closely related to the Byzantine text, so that it offers a reliable survey of this form of the text (see figs. 2 and 3). Yet not all the defective and singular readings, nor the late developments of individual witnesses to the Byzantine text, have been included. Manuscripts representing exclusively the text of this late period had to be set aside as contributing neither to the recovery of the original text nor to the history of the text.
17. One useful aspect of the edition is the supplements which accompany each installment and provide additional information. These supplements support and expand the critical apparatus. They contain not only the usual lists of abbreviations, manuscripts, and Church Fathers, but also detailed information on the lacunae, defects, and scribal blunders of all the manuscripts used in the edition.
18. The supplements also contain a list of the Byzantine variants in the Letter of James (i.e., all the readings where the undivided Byzantine text differs from the primary line text--cf. fig. 3) and an index of the patristic quotations and their references. This documentation makes it possible to trace back a notation in the critical apparatus to the father's edition used. Expanded information about versional evidence where the version's textual base is not clear is also given in the supplements. These passages are indicated in the critical apparatus by a question-mark or the reference sign ">". This makes it possible to give a more complete account of the witness of the major versions. Even the reader unskilled in these languages is able to use them as witnesses to the Greek text. In principle, if a Latin, Coptic, or Syriac witness is not mentioned, it has a lacuna.
19. Finally, for every volume of the ECM there will be a companion volume of supplementary studies explaining how the text has been established. For the Catholic Letters this will appear as the fourth installment, containing descriptions of the New Testament manuscripts, with definitions of their textual character in the light of the total evidence and an arrangement of the manuscripts (and manuscript groups) by their role in the development of the text. The preservation of the source material in databases makes this study possible, and it also allows for greater precision and a reevaluation of external criteria for judging variant readings (i.e., a better insight into the textual character of manuscripts in the light of their genealogical contexts). Publication of this material will not only promote a more comprehensive survey of the textual scene but also stimulate further studies in textual criticism and in the transmission and exegesis of the text. A textual commentary will discuss all passages where difficulties are found in the textual tradition. Thus on the one hand scholars will have open access to the material, yet on the other hand we will also publish the results of our own investigation of the text.
20. The Editio Critica Maior aims to provide the basis for establishing, confirming, and exploring the text of the New Testament. It is intended for all who are engaged in these endeavors. It is by no means simply for textual critics but also for exegetes and for historians of the New Testament text and of its reception through the centuries. It is our hope that this edition will promote cooperation among these disciplines.
21. Some may wonder why the edition has begun with the Catholic Letters, and accordingly with the Letter of James. No theological reasons of any kind lie behind this decision, but rather a variety of carefully weighed text-critical considerations. The Catholic Letters are preserved in a relatively small number of manuscripts compared with the whole of the New Testament textual tradition, although compared with other textual traditions the number (about 520 manuscripts) is still very large and appears almost overwhelming. Since the Editio Critica Maior claims to be based on the whole of the manuscript tradition while exhibiting only the evidence of a manageable selection that nevertheless accurately represents the full range of its development, we decided to begin with these letters. We proceeded to develop the methodology for distinguishing the Koine text from the earlier text forms and published it in the first volumes of the series Text und Textwert (Kurt Aland, ed. 1987).
22. There are several advantages to this approach. Because of the position of the Catholic Letters in the early history of the canon, when they were neither universally accepted nor uniformly valued, the manuscript tradition for these letters experienced a greater degree of freedom than did other parts of the New Testament. The text was less rigorously controlled, so that manuscripts where variant readings would be corrected or suppressed in other parts of the New Testament escaped comparable attention here, with the result that a complex variety of interrelated manuscripts and groups survived. If there is any part of the New Testament characterized by diversity in its transmission, it is here. We proceeded therefore to develop the computer-aided methodology for editing the NT text here, for it proved to be a good model of textual transmission. Here we can trace the Byzantine text through its stages of development. We have a 
72 demonstrating all the freedom of an uncorrected private copy. But now we also have a 100, just identified and kindly made available to us before its publication by the editors of the Oxyrhynchus Papyri for inclusion in this edition. 100, from the 3rd/4th century and containing the text of James 3:13-4:4 and 4:9-5:1, exhibits a remarkably strict text which differs only rarely from the primary line text. The tradition of the Catholic Letters lies between the extremes represented by 72 and 100, Vaticanus and the Koine text. There is no better area to gain experience in textual criticism, at least among the New Testament letters.
23. The Catholic Letters will now appear in regular sequence. The second installment will include the Letters of Peter, the third the Letters of John and Jude. Then the fourth installment will conclude with supplementary studies of the Catholic Letters. After that we will proceed to the Acts of the Apostles. But it is impossible to give an overall schedule here and now. It will depend too on our success in enlisting colleagues to share in the endeavor of establishing, confirming, and exploring this most important text of the Christian culture.
This paper is a revision of a presentation given to the New Testament Textual Criticism Section of the Society of Biblical Literature at the 1997 annual meeting in San Francisco, Michael W. Holmes, presiding. Presentations by Peter H. Davids, Bart D. Ehrman, D. C. Parker, William L. Petersen, and Klaus Wachtel (co-editor, Editio Critica Maior) also appear in this issue of TC. See also the critique of the volume by J. K. Elliott.
© TC: A Journal of Biblical Textual Criticism, 1998.
1At 2:3/44-48, for example, 74 attests to variants d or e, but not to the others (cf. fig. 1). In the supplements, you find the reason under "Further information on Greek witnesses (marked )" (p. B33): all that can be read of 2:3/44-48 in 74 is kaqo]u w?[de, an u followed by a letter which probably is an w and certainly is neither a second u (of upo) nor an e (of ekei). Obvious scribal blunders are treated similarly. For example, ms. 400 is recorded for reading a at 2:3/50-56 with the qualification "f4" (cf. fig. 1). This means that the scribe of 400 obviously meant to write upo to upopodion mou but made a mistake. In the "List of errors in the Greek manuscripts" (p. B13), the actual reading is recorded (upo to popodion mou). 
2At 2:3/26-30, for example, a Sahidic manuscript (K:Sms) is noted with "?". Under "Further information on versional witnesses" in the supplements (p. B36), the reading is cited in the original Coptic (6moot kalws nteikaqedra) and translated into German and English ("sit down well in this chair"). A versional witness is defined as a text-type (Latin) or a distinct translation (e.g. the Syriac Harklensis and Peshitta).
