² Volume 1, which came out in 1963, has proven to be an invaluable source for textual critics for over three decades: Kurzegefasste Liste, updated in 1994. Volume 25 was published in 1995: Marcion und sein Apostolos, by Ulrich Schmid. The volumes we are specifically interested in are Text und Textwert der Griechischen Handschriften des Neuen Testaments. 1: Die Katholischen Briefe (3 vols.; ANTF 9–11Google Scholar; Berlin/New York: de Gruyter, 1987). (Not all of the 25 volumes are on the Greek text, but all of the studies are geared toward it.)

³ Aland said in the foreword to volume 2 of Text und Textwert that already it is possible to say with certainty which manuscripts belong to the ‘original’ text. ‘Wenden wir uns ihnen zu, die entweder für die Feststellung des urspriinglichen Textes oder/und der frühen Text-geschichte des Neuen Testaments von Bedeutung sind. Zum ersten Mal können wir mit Sicherheit sagen, welche Handschriften dazu gehören’ (2.ix, italics added).

⁴ Aland wrote that in the volumes now presented, all manuscripts, even those with the Byzantine text, are investigated in all details in the test passages (Text und Textwert l.v). See also Barbara Aland's statement that all of the 540 manuscripts of the Catholic Epistles have been classified, The Text of the New Testament (Grand Rapids: Eerdmans, 1989) 318.Google Scholar

⁵ Text und Textwert 1.viii.

⁶ Information on the sources is given below under Section 2.

⁷ Although the comparisons in this study rely upon collation data taken only from 1 John, when it comes to the assigned classifications, it should be kept in mind that the Institute's classifications are for all of the Catholic Epistles as a body, not for 1 John alone. More below.

⁸ This is also discussed, although with considerably more brevity, in a paper which I read to the 1995 Midwest SBL and which was later published in JBL (see footnote 23).

⁹ Text und Textwert l.viii and Text, 318.

¹⁰ Echoes of Westcott and Hort (their Syrian text). As early as 1965 Aland expressed this viewpoint in a paper he read for the 100th Annual Meeting of the SBL, ‘The Significance of the Papyri for New Testament Research’. This paper was published in The Bible in Modern Scholarship (ed. J. Philip Hyatt; New York: Abingdon, 1965) 325–40.Google Scholar More recently, the Aland view was expressed in several places: Text und Textwert 2.vii, viii (the Byzantine text stands outside the goal of this work), and xiii, and Text, 106, 142 and 321. This view was clarified somewhat in Barbara Aland's statement in Text (331) when she indicated that the Byzantine text ‘should not be totally eliminated from editorial considerations, but only to the extent that its manuscripts are merely reproductions of an identical text.’

¹¹ Text und Textwert 2.vii.

¹² Text, 317–18. We discuss below the procedures the Institute used in the actual application of the method.

¹³ Text und Textwert l.v, referring to Aland's, Barbara Das NT in syrischer Überlieferung 1: Die Groβen Katholischen Briefe (ANTF 7; Berlin/New York: de Gruyter, 1986).Google Scholar

¹⁴ Text, 318.

¹⁵ l.xii.‘… nur von einer begrenzten Kenntnis der Überlieferung ausging, die sich erst mit dem Fortschreiten der Kollationen vervollständigte.

¹⁶ Ibid.

¹⁷ Ibid. Originally the number of test passages for the Catholic Epistles was 103 (see ‘Significance’, 343). As early as 1972, most of the readings probably were fairly well established. While I was collating manuscripts at the Institute in 1972, Klaus Junack, knowing that I was working on the Johannine Epistles, told me that they had thirty-five test passages for those three epistles. That is precisely the number of readings for 1–3 John that were published fifteen years later in Text und Textwert (1.125–199). In response to my request to know what those thirty-five readings were, Junack stated that he did not feel comfortable sharing the information with me at that time.

¹⁸ ‘Farbe bekennen’ (Text und Textwert 2.vii)Google Scholar. A few pages later we read that from the Teststellen, the character of a manuscript is reliably composed. Aland added, however, that a final judgment rests with a full collation (2.xi).

¹⁹ Text, 318 and 321. These test passages (short units) have been ‘carefully selected and are spread over the complete range of a book (or corpus) of scripture like a net’ (Text, 318).Google Scholar ‘Can we actually evaluate the textual quality of manuscripts on the basis of these test passages? Are they actually significant and is their number sufficient? In brief, the answer is: yes. This answer is based on our practical experience with them in the Institute at Münster. These test passages have consistently produced remarkably accurate evaluations when used in a variety of different projects, and they amply suffice for the three main tasks outlined above’ (321–2).

²⁰ Text und Textwert l.xii.

²¹ Bib 70 (1989) 377–88.

²² Ibid. 383.

²³ ‘Test Passages or Profiles: A Comparison of Two Text-Critical Methods’, JBL 115 (1996) 253–71.Google Scholar It should be pointed out, however, that the Teststellen provide more information about a manuscript than deciding whether it is Byzantine or non-Byzantine. My study was done in view of one of the Institute's major goals: eliminating the Byzantine manuscripts with the use of the Teststellen. The study showed that it is possible by using the profile method to classify a previously unclassified manuscript as Byzantine or non-Byzantine in 1 John with just three readings– a reduction of 765 per cent. over the twenty-three Teststellen found in 1 John. In the course of the research, the first sign that there might be some disagreement in the classifications based on Teststellen and those based on full collations showed up in one of the manuscripts being used for my comparison. According to my study, MS 104 in 1 John, based on a full collation, is a solid representative of the Byzantine text, agreeing over 90 per cent. of the time with the Byzantine text. According to the classification in Text, MS 104 is listed as a Category III manuscript for the Catholic Epistles. We look at MS 104 below.

²⁴ The statistics we have used from the Institute in making the comparisons in this study are primarily those used for all ninety–eight test passages.

²⁵ ‘A Closer Look: Text und Textwert der griechischen Handschriften des Neuen Testaments: Die katholischen Briefe’, AUSS 34 (1996) 37–46.Google Scholar

²⁶ A clear and concise review was given of the three volumes of Text und Textwert by Elliott, J. K. in NovT 30 (1988) 187–9.Google Scholar

²⁷ In the introduction to volume 2 of Text und Textwert (x–xi and xiii–xxii), Aland gave detailed information on basic procedures they used for their Main List (Hauptliste) and their Supplementary List (Ergänzungsliste). provided, Barbara Aland this information in English in Text, 321–37.Google Scholar

²⁸ SBLDS 35; Missoula, MT: Scholars, 1977.

²⁹ See note 23.

³⁰ The complete list of test passages is given in Text und Textwert, 9.16–229; the twenty-three used for 1 John are on pp. 125–76. The same twenty-three test passages are also given in the IGNTP format in ‘Test Passages or Profiles’.

³¹ See note 25.

³² See note 4. We relied upon the data given in Text (129–42).

³³ According to our classifications, of these thirty manuscripts, twenty-four belong to the Byzantine text, five to the Mixed, and one to the Alexandrian. These manuscripts are identified in Table 1 with a ‘na’ in the first column under ‘Classifications’. The classifications given by the Alands are found in Text, 107–42Google Scholar

³⁴ We did make one complete cross-examination to see if any clues were evident to account for the differences. A comparison of the differing counts between Text and Text und Textwer for MS 01 is as follows: for each of the four categories of readings we give two numbers. The first number is the count given in Text, and the second number following the slash mark is the count found in Text und Textwert: number ‘1’ readings: 23/18; numbers ‘1/2’ readings: 6/5; number ‘2’ readings: 63/61; and special readings: 16/14, for totals of 108/98.

³⁵ In the Institute's approach, the number ‘1’ represents the Majority text, and the number ‘2’ represents the ‘original’ or ancient text.

³⁶ For this study, because of the nature of these readings, we did our own calculations for percentages of agreement by neutralizing their percentage value, and then placed these percentages alongside those given in Text und Textwert. The percentages in the second column are based on information from the table ‘Abweichungen vom Mehrheitstext’ (ANTF 9.394–8).Google Scholar

³⁷ The table of Handschriftenprofile is found in Text und Textwert 9.232–392. The percentages used to make the ‘probable’ classification were arrived at by calculating the percentages of agreement from the numbers given in Text, 107–42.Google Scholar These newly calculated percentages were then listed for each manuscript within a given category. In this way we were able to determine the lowest and highest levels of agreement a manuscript had within the range for those manuscripts already classified by the Alands. The range of percentages of agreement with the Majority text for the manuscripts within each category were: Category I ranged from 1 per cent. agreement to 29 per cent.; Category II ranged from 28 per cent, to 45 per cent, (overlapping percentages unaccounted for); Category III ranged from 46 to 80 per cent.; and the range for Category V was from 84 to 100. After we calculated the percentages for each of the twenty–seven (using the Handschriftenprofile), we simply compared the percentages with the ranges for the various categories, and plotted the unclassified manuscript accordingly.

³⁸ The technique used to assign a classification for these twenty-seven manuscripts resulted in having all twenty-seven projections confirmed in the Institute's favour! They are part of the ninety manuscripts that agree with our conclusions.

³⁹ The numbers are: 540 (Text, 318); 552 (Text und Textwert 2.ix, xi and xiii); and 662 [Apostolos] (Text, 79 and 83). A random count of six Teststellen in 1 John gave the following numbers of manuscripts: 560, 563, 559, 557, 555, and 573. See Text und Textwert 1.125–77.

⁴⁰ Of these ninety manuscripts, as noted above, twenty-seven manuscripts are in agreement as a result of analyzing data in Text und Textwert for manuscripts that were not classified by the Institute. Table 1 provides the specific information.

⁴¹ We established their percentages from the Handschriftenprofile (ANTF 9.232–392). See above.

⁴² These data are available in several tables in the three volumes.

⁴³ These definitions are all taken from Text, 335–7. We omit Category IV (Western text).

⁴⁴ In this analysis it is not possible to give a detailed breakdown for all nineteen manuscripts. We give, however, several specific examples of the differences and then some conclusions of a more general nature.

⁴⁵ Because of the size of these tables, they are not produced here.

⁴⁶ Solid Alexandrian according to my study, and manuscripts with a weak connection to the Alexandrian text according to Aland. The tabulations based on quantitative analysis for all of the following comparisons are available as well as the actual profiles of readings. It would be impractical to reproduce them in this paper.

⁴⁷ ANTF 11:47–8.

⁴⁸ Gregory 1175: Alexandrian or Byzantine in the Catholic Epistles', AUSS 21 (1983) 155–68.Google Scholar