¹ The results of their study may be found in their dissertations written in 1968 for the Claremont Graduate School; McReynolds: ‘The Claremont Profile Method and the Grouping of Byzantine New Testament Manuscripts’; Wisse: ‘The Claremont Profile Method for the Classification of the Byzantine New Testament Manuscripts: A Study in Method’.

² They used von Soden's classifications where no later group study was available (Wisse, p. 78); von Soden had formed 17 groups, 5 of which had been studied and confirmed by others (Wisse, p. 70).

³ ‘How could one be certain that an alleged group supported a variant reading before the existence of the group had been proven and all the members of the group had been taken into account? No alternative presented itself apart from using tentative group definitions borrowed from the students of manuscript groups in the past…Starting from scratch would pose much greater problems. It would mean selecting variants with enough manuscript support that they could possibly include a majority of a group. Then would follow a long and treacherous process of trial and error to find the manuscripts which read together often enough to suggest group relationships’ (Wisse, pp. 76–7).

⁴ ‘A Critique of a New Testament Text-Critical Methodology – The Claremont Profile Method’, J.B.L. 96 (1977), 555–66Google Scholar. This paper demonstrated and tests the method in I John.

⁵ McReynolds and Wisse used manuscripts in the International Greek New Testament Project (IGNTP) files at Claremont. My use of the 207 manuscripts in the IGNTP files was made possible by the gracious permission of Professor McReynolds. This study would not have been possible without access to the careful collations of this large number of manuscripts. Each of the 207 manuscripts was fully collated and all of the variants were considered in this study.

⁶ These procedures are covered in my study The Classification of the Greek Manuscripts of the Johannine Epistles, SBL Dissertation Series 35 (Missoula, Montana: Scholars Press, 1977).Google Scholar

⁷ It should be noted that the two CPM criteria are thoroughly sound once one has begun with the proper use of quantitative analysis.

⁸ The criteria are spelled out in McReynolds' dissertation, pp. 4–13.

⁹ The rationale for using ‘two-thirds’ is that there can be no question about there being a proper majority. See McReynolds, p. 5 and Wisse, p. 78.

¹⁰ Klaus Junack wrote in 1972 that ‘Der Haupteinwand gegen die Profile-method ist jedoch, dass jene Gruppen-profile a priori festgelegt werden müssen, die ganze Arbeit also allenfalls nur zu einer Bestätigung einer eingangs aufgestellten Prämisse dienen kann, obwohl eigentlich die Feststellung von Gruppen das Ziel und die Abgrenzung dieser Gruppen oder Untergruppen erst das Ergebnis einer eingehenden Durcharbeitung des Materials sein kann.’ ‘Zu den griechischen Lektionaren und ihrer Überlieferung der katholischen Briefe’, in Die alten Übersetzungen des Neuen Testaments. Die Kirchenväterzitate und Lektionare, ed. Kurt Aland, vol. 5 of Arbeiten zur Neutestamentlichen Textforschung (Berlin: Walter de Gruyter, 1972), p. 556Google Scholar. The criteria do reveal that some previously formed groups are illegitimate (see McReynolds, pp. 88–91), and also that some manuscripts do not belong to a group.

¹¹ In my studies, the group profiles all show an increase in the number of group readings. See footnote 22.

¹² I used the same 163 manuscripts plus an additional 44 that had been collated for the IGNTP files; approximately half of these additional manuscripts are uncials. After McReynolds had formed his profiles from studying the 163 manuscripts, he examined an additional 463 manuscripts and classified them according to their alignment with his group profiles. I refer to McReynolds in the comparisons because Wisse did not actually get into the classification process in his dissertation.

¹³ McReynolds actually formed 14 groups, but only one of the manuscripts (1424) was available to me from one of his groups (group 1424); he formed this group later while examining manuscripts in other collections. The 14 groups are as follows, with the number of manuscripts in the group placed in parentheses: 13 (8); 1 (6); Π^a (12); Π^b (11); Μ^a (9); Μ^b (5); 1424 (1): 1216 (6); Λ (8); Κ^r (17); Κ^x (34); Ω (14); and 7 (4). Twelve manuscripts were not classified in any of these 14 groups because they were considered ‘mixed’ or ‘wild’. Another 8 manuscripts were classified as ‘ϕ’ but a profile was not given for them because they do not actually have a profile; they do, however, share a mixture of the readings belonging to one of von Soden's three Ι^ϕ groups (McReynolds, p. 133). The total number of these manuscripts does not equal 163; it is not clear where the other eight manuscripts are in McReynolds’ lists. The answer may lie in the Κ^x manuscript count; see footnot 22, p. 56. McReynolds indicates what the von Soden designation was for the noes he has used on page 17: Ι^l = 13; Ι^na = 1; Ι^nb = 22; Ι^kac = п^a; Ι^kb = п^b; Ι^ϕa = 1424; Ι^ϕb = 7; Ι^ϕr = Μ^a; Κ^l = Ω Ι^r = Λ Ι^B = 1216.

¹⁴ The 64 units of variation are given on pages 104–6 of McReynolds' dissertation. Of the 64 readings, three were combined in my study with the preceding reading because they deal with the same unit of variation, that is, the second entry in these three cases is a second form of variation from the TR: units and, 11 and 12, and 60 and 61. CPM unit 19 was eliminated from my study since it is a reading from group 1424 (see footnote 13). One other CPM unit was also eliminated, in this case, for lack of support: unit 39 is a unique group reading for group 1216; but it is supported by only 6 of 11 (55 per cent) manuscripts profiled in this group by McReynolds. Furthermore, it is supported by only 3 of 6 manuscripts available to me.

¹⁵ originally had 60 extra readings, but 22 of these were discarded because they did not contribute to our picture of group relationships even after forming tentative groups by quantitative analysis.

¹⁶ The 97 readings are given in the Appendix; the 38 new group readings are marked with an asterisk.

¹⁷ The 97 readings are given in the Appendix; the 38 new group readings are marked with an asterisk.

¹⁸ See page 52, footnote I.

¹⁹ Wisse wrote, for example, that ‘the 163 minuscules included almost all those which had proven their value in the history of the study of the New Testament text (the “Neutral” and “Caesarean” minuscules, members of Family I and 13 etc.)’, p. 73.

²⁰ This term includes the Caesarean text; I have chosen to call it Mixed because it seems to describe the nature of the text so precisely. This text has both readings that are in the Alexandrian manuscripts but not in the Byzantine manuscripts, and readings that are in the Byzantine manuscripts but not in the Alexandrian manuscripts; furthermore, this text has readings that are uniquely its own, that is, in neither the Alexandrian nor the Byzantine manuscripts.

²¹ These three non-Byzantine groups had been confirmed by the Lakes.

²² It is significant that even with the CPM groups that were confirmed, the groups were still more sharply defined in this study by the increase in the number of group readings: the number of CPM group readings for a group is given first, and the number of group readings used in this study is given second: group i (23–27); group п² (9–18) – a doubling of the group readings; group I (26–32); group 22 (8–10); group ∧ (8–10); group 1216 (9–12); and group K^r (5–6). The CPM group reading count is based on the 163 manuscripts in the IGNTP files used by McReynolds to form his group profiles.

²³ All of these Byzantine groups were studied by von Soden. K^r was studied by David Voss, and M^b was a new group formed by McReynolds. McReynolds noted that group 7 has a distinctive profile in Luke only in chapter 2 (p. 12).

²⁴ In Appendix iv of his dissertation McReynolds listed 623 manuscripts that he classified. Most of the manuscripts are housed in various collections around the U.S., and McReynolds collated them in the units of variation which had become the CPM group readings. Of these 623 manuscripts, 139 are classified Kx in Luke To, and another 40 as K^x with extra readings (e.g. K^x + п or just K^x + ). It is not possible to tell just exactly how McReynolds handled these K^x manuscripts. He has given profiles foi K^x manuscripts in chapter io (p. 86), but only i of these are in the IGNTP files. On the other hand, there are 30 K^x or K^x + manuscripts classified by McReynolds in Appendix iv which are in the IGNTP files, but for which he gave no profiles in his K^x profile table. In his miscellaneous profiles (Appendix III, p. III), he profiles along with some manuscripts considered wild or mixed: one K^x manuscript (Gregory 28) and two K^x + п manuscripts (Gregory 157 and 565).

²⁵ See discussion below: Table I illustrates the points that are made in the discussion. McReynolds wrote that K^x represents the ‘uniform mass which lies between K¹ and K^r. The differences within the K^x manuscripts are not as great as the differences between K^x and K¹ or K^r’, and he further observes that von Soden allowed that there might be groups within the K^x group (p. 81).

²⁶ My own preference is to use Bz for the Byzantine text with raised numbers to indicate the groups within the text, e.g. Bz¹, Bz², etc., with the lowest number representing the group closest to the TR, and the highest number representing the group of manuscripts with the largest number of non-TR readings.

²⁷ In addition to the six divisious made up primarily of the CPM K^x group manuscripts, I have put together another six groups of Byzantine manuscripts; the additional six groups consist of manuscripts that come from several CPM groups: K^x, Ω, ϕ,п^b, 7, etc., including some of the manuscripts classified ‘mixed’ and K^x + and including some uncial manuscripts not examined by McReynolds.

²⁸ The other three K^x groups (K^x1, K^x2 and K^x4) are made up of either K^x or K^x + manuscripts in McReynolds' study except for one ϕ manuscript in K^x2 which was not profiled by McReynolds.

²⁹ The surplus readings which do not show up as group readings for any of the three group profiles are left Out here; a surplus reading is a non-TR reading which is supported by less than two- thirds of the members of the group. A surplus reading for McReynolds was a manuscript's reading that received less than fifty per cent of the support of the other member manuscripts. He nssde a distinction between primary and secondary readings; a primary reading was a reading supported by at least two-thirds of the members, and a secondary reading was a reading supported by less than two-thirds of the members, but by at least half of the members. Such a secondary reading could exist only if the riading were a primary reading in at least one other group.

³⁰ After some manuscripts were moved out of M^a that did not relate very well to the M^a group, unit 178 became a group reading for the new M^a group.

³¹ McReynolds classified this manuscript as Ω, but it was not included in his Ω profiles.

³² See footnote 29 for explanation of a surplus reading.