New Testament Manuscripts

Papyri

Contents: P4 *P10 *P11 *P13 *P20 *P24 *P28 *P39 *P45 *P46 *P48 *P51 *P52 *P54 *P64+P67 *P74 *P75 *P78 *P79 *P90 *

Note: Many of the papyri, especially the Beatty and Bodmer papyri,have been subject to so much discussion that no attempt is made tocompile a full bibliography.


P4

Location/Catalog Number

Paris, National Library Suppl. Greek 1120.

Contents

Luke 1:58-59, 1;62-2:1, 2:6-7, 3:8-4:2, 4:29-32, 4:32-35, 5:3-8, 5:30-6:16

Date/Scribe

Dated paleographically to the third century.

Description and Text-type

Aland and Aland list P4 as Category Iwith a "Normal" text.Von Soden listed it as H (Alexandrian), which is quite clearly correct; has a numberof significant readings shared with ℵ and B.

It has been suggested that P4 is part of the same document asP64+P67. This is certainly possible on chronological grounds,and it is generally agreed that the handwriting used in the two is very similar.The latter pair, however, contain fragments of Matthew. IfP4 and P64+P67 are indeed one manuscript, theyrepresent quite possibly the earliest instance of a papyrus containing more thanone gospel. Peter Weigandt, Joseph van Haelst, C. H. Roberts, and T. C. Skeat have all argued for the identification, on the basis of script; Peter Head and Scott Charlesworthargued against it on other grounds. (For a more detailed summary, see TommyWasserman's article "A Comparative Textual Analysis of P4 and PP64+67." Thanks to James Dowden for pointing out this article.)

I personally do not think the evidence sufficient to draw firm conclusions; if Ihad to guess, I would guess the two fragments are from two different codices copiedby the same scribe. But I certainly would not be dogmatic. So far, the list ofmanuscripts in NA28 does not list them as being the same manuscript.

Other Symbols Used for this Manuscript

von Soden: ε34

Bibliography

Collations:

Sample Plates:

Editions which cite:
Cited in all editions since Von Soden.

Other Works:

The already-cited Wasserman article "A Comparative Textual Analysis of P4 and PP64+67" contains a great deal of detail about the readings of this manuscript. However, the self-referential nature of Aland-based textual criticism (if you already know the correct reading of the text, which theirmethod assumes, why do you have to study the manuscripts?) makes the conclusions quitedubious.


P10

Location/Catalog Number

Cambridge (Massachusetts), Harvard University, Semitic Museum, MS. Inv. 3736,Designated by its discoverers Oxyrhynchus papyrus 209.

Contents

Romans 1:1-7. The only variant for which it is cited in Bover, Merk, or NA27is Romans 1:1, where it agrees with B 81 m am cav dubl ful hub reg valin reading χριστου ιησουfor ιησου χριστου of P26 ℵ A G 1739 Byz.

Date/Scribe

Dated paleographically to the fourth century.

Description and Text-type

Aland and Aland list P10 as Category I.Von Soden did not list a type. Most authorities would probably list it asAlexandrian.

Deissman suggested that this was written for use in an amulet. He believesthe owner was probably named Aurelius Paulus. Grenfell and Hunt believed it tohave been a school child's practice copy. It almost certainly was not meant tobe part of a complete text of Paul.

Other Symbols Used for this Manuscript

von Soden: α1032

Bibliography

Collations:

Sample Plates:

Editions which cite:
Cited in all editions since Von Soden.

Other Works:


P11

Location/Catalog Number

Saint Petersburg, Russian National Library Gr. 258A

Contents

1 Corinthians 1:17-22, 2:9-12, 2:14,3:1-3, 3:5-6, 4:3-5:5,5:7-8, 6:5-9, 6:11-18, 7:3-6,7:10-14, with even the surviving verses oftendamaged (so much so that Tischendorf was unable to tell whether thefragments he had were of five or six leaves).

Date/Scribe

Dated paleographically to the seventh century. Some older manuals giveits date as the fifth century, but this was based on comparison withuncial manuscripts; a comparison with the style of papyri resulted inthe change.

Description and Text-type

Aland and Aland list P11 as Category II.Von Soden listed its text as "H or I."

In fact the text of P11 seems fairly ordinary (though its fragmentarynature makes a firm determination difficult; the Nestle text, for instance, citesit explicitly only about fifteen times, most often with the Alexandrian groupℵ A C 33, but also,with the Byzantine and "Western" texts; there appears to be some slightkinship with the later members of Family 1739,particularly 1881. Overall, the bestdescription of its text is probably "mixed," although most of the readingsare old. It does not appear to have any immediate relatives).

The most noteworthy thing about P11, therefore, is not its text butits history: It was the first biblical papyrus to be discovered (Tischendorf observedit in 1862), and the only oneto be cited in Tischendorf (as Q).

Other Symbols Used for this Manuscript

von Soden: α1020
Tischendorf: Qp

Bibliography

Collations:
Ellwood M. Schofield, The Papyrus Fragments of the Greek New Testament
See also K. Junack, Das Neue Testament auf Papyrus, Vol. 2: Die paulinischen Briefe

Sample Plates:

Editions which cite:
Cited in all editions since Tischendorf.

Other Works:
Kurt Aland, "Neutestamentliche Papyri," NTS 3


P13

Location/Catalog Number

London (British Museum, Papyrus 1532 verso) and elsewhere (Florence, Cairo).Designated by its discoverers P. Oxy. 657

Contents

P13 is an opisthograph, withthe epitome of Livy onthe reverse side. Presumably the manuscript originally containedall of Hebrews (it has been suspected that it contained othermaterial as well; a full-length scroll could contain rather more thantwice the material found in Hebrews); it now retains Hebrews2:14-5:5, 10:8-22,10:29-11:13, 11:28-12:17,with many minor lacunae. Despite the damage,P13 is the most extensive papyrus outside the Beatty and Bodmercollections.

Date/Scribe

Dated paleographically to the third or fourth century. It has beenspeculated that the scroll was carried to Egypt by a Roman official,then left behind and rewritten.

P13
Portions of two columns of P13, beginning with Hebrews4:2. Note the extensive damage (which is even worse in the lowerhalves of the columns). P13 is the only extensive NTopisthograph. Observe the surviving numbering at the top of the leftcolumn.

Description and Text-type

Aland and Aland list P13 as a free(?) text with "A number ofdistinctive readings, often with P46." Von Soden lists its text-typeas H.

The most substantial of the Oxyrhynchus papyri, P13 is also perhapsthe most important. As noted by Sanders and, later, the Alands, it frequentlyaligns with P46(and -- perhaps even more often -- with B for the portions of Hebrews where bothexist); Kenyon notes an 82% agreement rate between the two papyri, withsimilarities even in punctuation and pagination (even though the twocannot have had the same contents; a scroll simply could not contain tenPauline letters. It is possible that P13 contained Romans andHebrews, in that order, in which case it followed the same order as P46).P13 contains a number of singular and subsingular readings, but this seemsto be characteristic of the P46/B type. Since this type contains onlythree other witnesses (P46, B, and the Sahidic Coptic), P13is an extremely important witness which has not, so far, received sufficientattention (Zuntz, e.g., never even mentions it in his work on 1 Corinthians andHebrews).

Other Symbols Used for this Manuscript

von Soden: α1034
Designated P. Oxy. 657 in the Oxyrhynchus Papyri series.

Bibliography

Collations:
B. P. Grenfell and A. S. Hunt, Oxyrhynchus Papyri, Volume 4.
See also K. Junack, Das Neue Testament auf Papyrus, Vol. 2: Die paulinischen Briefe

Sample Plates:
Comfort, Manuscripts of the Greek Bible (1 page)
Comfort, The Quest for the Original Text of the New Testament (1 page; samephoto as above)

Editions which cite:
Cited in all editions since von Soden.

Other Works:
Comfort, Early Manuscripts & Modern Translations of the NewTestament, p. 37


P20

Location/Catalog Number

Princeton University Library, Am 4117 -- Oxyrhynchus Papyrus 1171

Contents

Portions of James 2:19-3:9

Date/Scribe

Dated paleographically to the third century.

Description and Text-type

P20 is a fragment of a single leaf, 11.5 cm. tall and somewhat less than4.5 cm. wide at the widest. It is the central portion of a leaf; both left and rightedges are damaged, as is the bottom. Portions of 20 lines survive on each side, withusually about twelve characters per line. The original seems to have had about 30-35characters per line, so the surviving portion is relatively slight. The hand is roughand hasty-looking; given the state of the manuscript, it is often difficult to distinguishthe letters.

The small amount of remaining text makes it difficult to classify the manuscript.The Alands list it as Category I, with a "normal"text. Von Soden lists it as H (Alexandrian). Schofeld reports that it only twicedeparts only twice from the "B-group," -- but of course this is a vaguegroup description. Still, the general feeling is that the manuscript is Alexandrian.

Other Symbols Used for this Manuscript

von Soden: α1019

Bibliography

Collations:
B. P. Grenfell and A. S. Hunt, Oxyrhynchus Papyri,volume 9.

Sample Plates:
W. H. P. Hatch, The Principal Uncial Manuscripts of the New Testament

Editions which cite:
Cited in Von Soden, Merk, Bover, NA26, NA27.

Other Works:
Comfort, Early Manuscripts & Modern Translations of the NewTestament, pp. 39-40
Ellwood M. Schofield, The Papyrus Fragments of the Greek New Testament


P24

Location/Catalog Number

Newton Centre: Andover Newton Theological School, Franklin Trask Library,O.P. 1230 (i.e. Oxyrhynchus Papyrus 1230)

Contents

Portions of Rev. 5:5-8, 6:5-8

Date/Scribe

Dated paleographically to the fourth century by the Alands, though some havepreferred the third century. The hand is unattractive and rather difficult; thecopyist was probably not a trained scribe.

Description and Text-type

P24 is a fragment of a single leaf, shaped rather like a very short,fat letter T turned upside down. The vertical stroke of the T contains two lines,with only about five or six surviving letters per line; the cross of the T containsportions of four lines, with about sixteen letters on the two central (and best-preserved)lines. The lines appear to have been fairly long -- about 30-32 letters per line -- soeven the best-preserved lines retain only about half the text of the relevant verses.

The fact that the manuscript has so many letters per line, and so many linesper page (there are over 1600 letters between Rev. 5:6 and Rev. 6:6, which at32 letters per line gives us some 50+ lines per page) implies a large papyrussize; Schofield thought it might have been a church Bible.

With only about 150 letters to examine, it is simply not possible to decideP24's text-type. The Alands list P24 asCategory I, but this is doubtless based primarilyon its date (early manuscripts of the Apocalypse being so rare); even they don'tventure a guess as to whether its text is free, normal, or strict. Comfort observesthat the manuscript has "only" three divergences from A, but in contextthis is quite a high number.

Other Symbols Used for this Manuscript

Bibliography

Collations:
B. P. Grenfell and A. S. Hunt, Oxyrhynchus Papyri,volume 10.

Sample Plates:
Comfort, Early Manuscripts & Modern Translations of the NewTestament, has plates of the entire manuscript.

Editions which cite:
Cited in Von Soden, Merk, Bover, NA26, NA27.

Other Works:
Comfort, Early Manuscripts & Modern Translations of the NewTestament, pp. 41-42
Ellwood M. Schofield, The Papyrus Fragments of the Greek New Testament


P28

Location/Catalog Number

Berkeley (Palestine Institute Museum), Pacific School of Religion Papyrus 2 --Oxyrhynchus Papyrus 1596

Contents

Portions of John 6:8-12, 17-22

Date/Scribe

Dated paleographically to the third century. The hand slants slightly andlooks hasty and unattractive. Numbers are spelled out (as, e.g., in P66)rather than written as numerals (as in P75). Its use of theNomina Sacra is incomplete; althoughwe findΙησουςabbreviated, in verse 9, we findανθρωπουςspelled out.

Description and Text-type

P28 is a fragment of a single leaf, ten cm. tall and five wide.The surviving portion is from the bottom of the leaf, and is broken on bothsides. Eleven lines survive on the recto, twelve on the verso (plus a few blotsfrom a thirteenth). About 13-15 letters survive on each line, out of an averageof perhaps 32 letters per line (the lines seem to have been somewhat irregular).

Textually, most scholars have regarded P28 as Alexandrian.The Alands list it as Category I, with a "normal"text. Grenfell and Hunt described it as eclectic, somewhat closer toℵ than B(though, given the list of variants below, I find it hard to see what ledthem to this conclusion).The small amount of surviving text makes any determination difficult, butthe description "eclectic" seems to fit; it has noteworthy differenceswith almost every important manuscripts. The following table shows thenotable readings of P28, with their supporters (the text is astranscribed by Finegan):

Readings of P28 and supportersOther readings
6:9 ταυτα τι εστιν P28 P66c P75 rell UBS ταυτα εστιν D*; τι εστιν ταυτα P66* e
6:11 ελαβεν ουν (P28 .λεβεν ο..) P66 A B D L W 892 al UBS ελαβεν δε ℵ* E F H 33 700 Byz; και λαβων G Θ f1 f13 565 (579 και ελαβεν)
6:11 ε...ριστησας εδ.... (i.e. ευχαριστησηας εδωκεν or similar) P28 P66 (P75 ..............εδωκεν, which could agree with P28 or with the later witnesses) N Γ 69 579 ευχαριστησηας διεδωκεν A B K L W f1 33 565 700 892 rell UBS; ευχαριστησηας και διεδωκεν ℵ D
6:11 τοις ανακειμενοις (P28 ...........ενοις but lacks space for a longer reading) P66 P75 ℵ* A B L N W f1 33 565 579 1241 al UBS τοις μαθταις οι δε μαθεται τοις ανακειμενοις D E F G H K Γ Δ Θ Ψ f13 892 Byz
6:17 και σκοτια ηδη εγεγονει (P28 ....σκοτια ηδ...) (P75 ....σκοτια ηδη εγεγονει) rell UBS καταλβεν δε αυτους η σκοτια ℵ D
6:17 ουπω προς αυτους εληλυθει ο Ιησους (P28 .....ηλυθει ο Ις) (P75 ηδ. .... προς αυτους εγεγον.. . Ις) B N Ψ ουπω εληλυθει προς αυτους ο Ιησους (L) W (f13 33 69 788 pc UBS; ουπω εληλυθει ο Ιησους προς αυτους D; ουπω εληλυθει Ιησους προς αυτους ℵ; ουκ εληλυθει προς αυτους ο Ιησους A E F G H (K) Δ Θ f1 565 579 700 892 Byz
6:19 σταδιους P28 P75-vid rell UBS σταδια ℵ* D
6:20 ο δε λεγει (P28 ο δε....) (P75 ...γει) rell UBS και λεγει ℵ
6:21 επι της γης P28 rell UBS επι την γην ℵ* f13 579 1424 pc
6:22 ειδεν οτι (P28 ...ιδεν οτι) ℵ D ειδον οτι (P75 ειδο....) A B L N W Θ 33 al UBS; ιδων οτι E F G H Δ Ψ 565 579 700 1241 Byz

(There are, of course, many other variants in this part of John, but P28is too fragmentary to testify to these, and the line lengths seemingly too irregularto testify to most of the add/omit variants.) NOTE: NA27 and related editionslist P28 as readingωσειπεντακισχιλιοιin verse 10. This is based solely on calculations of line lengths; the only survivingtext is -χιλιοι.This reading does appear likely -- the line is extremely short if the reading isως -- but is too uncertain for usto use it in determining textual groupings. A similar situation occurs in verse 19,θεωρουσιν τονΙησουν. P28breaks off in the previous line atεικουσι π....i.e. εικουσι πεντε,and all that survives of the textθεωρουσιν τονΙησουνis ν Ιν.The Aland Synopsis listsP28 as omitting τον,but this is based solely on line lengths and must be considered quite uncertain.

Other Symbols Used for this Manuscript

Bibliography

Collations:
B. P. Grenfell and A. S. Hunt, Oxyrhynchus Papyri,volume 13.

Sample Plates:
Finegan, Encountering New Testament Manuscripts

Editions which cite:
Cited in Merk, Bover, NA26, NA27.

Other Works:
Comfort, Early Manuscripts & Modern Translations of the NewTestament, p. 43
Ellwood M. Schofield, The Papyrus Fragments of the Greek New Testament


P39

Location/Catalog Number

Rochester (New York, USA). Ambrose Swabey Library, Inv. no. 8864 -- Oxyrhynchus Papyrus 1780

Contents

Portions of John 8:14-22

Date/Scribe

Dated paleographically to the third century. The hand is very clear and the survivingtext easily read; one suspects an expert scribe.

Description and Text-type

P39 is a fragment of a single leaf, preserving the entire height of themanuscript but only one edge. There are 25 lines per page, but only about six or sevensurviving letters per line (occasionally less, especially on the verso). There appearto have been about thirteen or fourteen letters per line (column?), meaning that abouthalf the text survives.

There is general agreement that the manuscript is Alexandrian. The Alandslist it as Category I, with a "strict"text. Grenfell and Hunt list it as aligning with B; Schofield goes further, claimingit never departs from B. When these authors wrote, of course, P75 wasnot known. In the area covered by P39, there are only a handful of differencesbetween P75 and B. P39 does not testify to verse 14,και/η.In verse 15, where P75 d f cop addδε,P39 is not extant, but line lengths make is more likely than not that itomits the word with B rell. The next variant in P75, the omissionof εγω in verse 22, occurs after theend of the manuscript (which actually breaks off at the end of verse 21; all that is visible of verse 22 is part of a stroke of the first letter).

Other Symbols Used for this Manuscript

von Soden: α1019

Bibliography

Collations:
B. P. Grenfell and A. S. Hunt, Oxyrhynchus Papyri,volume 15.

Sample Plates:

Editions which cite:
Cited in Merk, Bover, NA26, NA27.

Other Works:
Comfort, Early Manuscripts & Modern Translations of the NewTestament, p. 47
Ellwood M. Schofield, The Papyrus Fragments of the Greek New Testament


P45

Location/Catalog Number

Dublin, Chester Beatty Library, P. Chester Beatty I;Vienna, Austrian National Library, Pap. Vindob. G. 31974 (one leaf, containingMatt. 25:41-26:39)

Contents

P45 is surely in the worst condition of any of the substantialBiblical papyri. Even the surviving leaves (a small fraction of the originalcontents, estimated at 30 of 220 original leaves) are damaged; the most substantialpages are perhaps 80-90%complete, but many others are just small fragments. There are relativelyfew complete lines; many of the surviving leaves represent only about20% of the width of the original manuscript. Therefore any list of versesincluded in the manuscript will make it seem more substantial than it reallyis; very many of these verses survive only in part (often very small part).

With that said, the verses represented at least partly in P45 are:Matt. 20:24-32, 21:13-19, 25:41-26:39;Mark 4:36-40, 5:15-26, 5:38-6:3, 6:16-25, 36-50, 7:3-15, 7:25-8:1, 8:10-26, 8:34-9:8, 9:18-31, 11:27-12:1, 12:5-8, 13-19, 24-28;Luke 6:31-41, 6:45-7:7, 9:26-41, 9:45-10:1, 10:6-22, 10:26-11:1, 11:6-25, 28-46, 11:50-12:12, 12:18-37, 12:42-13:1, 13:6-24, 13:29-14:10, 14:17-33;John 4:51, 54, 5:21, 24, 10:7-25, 10:31-11:10, 11:18-36, 43-57;Acts 4:27-36, 5:10-20, 30-39, 6:7-7:2, 7:10-21, 32-41, 7:52-8:1, 8:14-25, 8:34-9:6, 9:16-27, 9:35-10:2, 10:10-23, 31-41, 11:2-14, 11:24-12:5, 12:13-22, 13:6-16, 25-36, 13:46-14:3, 14:15-23, 15:2-7, 19-26, 15:38-16:4, 16:15-21, 16:32-40, 17:9-17.

It is possible that the codex originally contained other books (e.g.the Catholic Epistles); unlike many of the major papyri, it is nota single-quire codex, but rather uses gatherings of two leaves, meaningthat it could have had many more leaves at the end.

All told, we have two leaves of Matthew, six of Mark, seven of Luke,two of John, and thirteen of Acts, with the leaves of Matthew being onlythe smallest fragments. The leaves of Mark and Acts are rather moresubstantial, but still badly damaged; those of Luke and John arerelatively complete. The leaves are broad enough, and the single columnof text wide enough, that these thirty leaves contain substantial amountsof text, but still only about 5% of the original contents.

Kenyon was of the opinion that the gospels were originally in the"Western" order Matthew, John, Luke, Mark, with Acts (andconceivably other material) following. Given the state of the manuscript,the fact that it used multiple quires,and the fact that it was brought to the west in pieces, this cannot beproved -- but Mark and Acts were discovered together, so it seems likely.

Date/Scribe

Dated paleographically to the third century.

Description and Text-type

It appears that P45 was originally the most extensive of allpapyrus manuscripts -- the only one to include more than one NT section.It has, however, been very badly damaged, meaning that relatively littletext survives. This makes an accurate assessment of the manuscript'stype rather difficult. Wisse, for instance, did not even attempt a profile.

When Kenyon first published the manuscript, however, he attempted toclassify it, stating that in Mark it seemed to be Cæsarean;in Luke and John, neither purely Alexandrian nor Western; in Acts, primarilyAlexandrian (although it has some of the smaller "Western" variants,it has few if any of the greater).

Kenyon, however, was probably led astray by Streeter's bad definition ofthe "Cæsarean" text and by all the bad work which followed fromthis. Two more recent works have re-examined the ground and produce a very differentconclusion.

The first and, in the long term, probably more important is E. C. Colwell,"Method in Evaluating Scribal Habits: A Study of P45, P66,P75" (1965; now available as pages 106-124 in Colwell'sStudies in Methodology in Textual Criticism of the New Testament).This showed that P45 is the result of a freely paraphrasedcopy; the scribe of P45 or one of its immediate ancestors feltfree to expand, paraphrase, and shorten the text. (Though Colwell noted thatdeletions were much more common than additions -- "The dispensable wordis dispensed with.")

The noteworthy point here is that this sort of editing is typical ofat least two other Gospel text-types, the "Western" and the"Cæsarean." (Though both of these add and harmonize morethan they delete.) Observe what this means: To a scholar who simplystudied the types of readings in P45 (as opposed to thepattern of readings, which is the true definition of a text-type), P45would appear to belong to one of the periphrastic text-types. Of the two,the "Cæsarean" is, of course, the more restrained, and alsohas more Alexandrian readings; P45, as an Egyptian manuscript,probably started with an Alexandrian text.

Thus, Colwell established that P45 needed to be examined moreclosely before it could be labelled "Cæsarean." Kenyon's"Cæsarean" classification was not rigorous, and was justwhat one would expect from a non-rigorous examination of a manuscript likeP45.

Colwell's implicit call for a more detailed study was supplied by LarryW. Hurtado in Text-Critical Methodology and the Pre-Caesarean Text: CodexW in the Gospel of Mark. This study suffers from major methodologicalflaws, but it pretty definitely establishes its main conclusion: ThatP45 and W do not belong with the so-called "Cæsarean"text. (Hurtado has also been interpreted to mean that the "Cæsarean"text does not exist. This conclusion, however, is premature, given hismethodology; see the discussion of the"Cæsarean" text in thearticle on text-types.)

So where does this leave P45? The truth is, very little controlledanalysis has been done of the manuscript. It was discovered too late for Von Soden.Wisse did not profile it. The Alands list it as Category Iwith a free text, but it seems likely that this assessment is based simply on whatthey think of the manuscript. The manuscript needs a re-evaluation before we canreally state firm conclusions. My own analysis indicates that the manuscript is in factcloser to B than to any other uncial. On the face of it, it would appear that P45comes from the Alexandrian tradition, but has been so heavily edited that it beginsto appear "Westernized."

Other Symbols Used for this Manuscript

Bibliography

Note: As with most major manuscripts, no attempt is made to compile a completebibliography.

Collations:
The basic publication remains Frederic G. Kenyon, Chester Beatty BiblicalPapyri (Part II, The Gospels and Acts, in two fascicles). Various authors(Gerstinger, Merk, Zuntz) have published supplements or additional analysis.

Sample Plates:
Aland & Aland, The Text of the New Testament (1 plate)
Sir Frederick Kenyon & A. W. Adams, Our Bible and the Ancient Manuscripts (1 plate)

Editions which cite:
Cited in NA16 and later, UBS, Merk, Bover

Other Works:
The two most important works are probably those already cited:E. C. Colwell,"Method in Evaluating Scribal Habits: A Study of P45, P66,P75" (1965; pp. 106-124 in Colwell'sStudies in Methodology in Textual Criticism of the New Testament).
Larry W. Hurtado in Text-Critical Methodology and the Pre-Caesarean Text: CodexW in the Gospel of Mark.


P46

Location/Catalog Number

Dublin, Chester Beatty Library, P. Chester Beatty II;Ann Arbor, University of Michigan Inv. 6238

Contents

86 leaves (out of an original total of 104), containing portions of Romans5:17-1 Thes. 5:28 (including Hebrews, following Romans.According to Sanders, only one other manuscript, 1919, has the Pauline Epistlesin this order).The surviving leaves (most of which are somewhat damaged) containRomans 5:17-6:3,6:5-14, 8:15-25,27-35, 8:37-9:32,10:1-11, 11, 24-33,11:35-15:9,15:11-end (with 16:25-27 followingchapter 15!);1 Cor. 1:1-9:2,9:4-14:14,14:16-15:15,15:17-16:22;2 Cor. 1:1-11:10, 12-21,11:23-13:13;Gal. 1:1-8,1:10-2:9, 2:12-21,3:2-29, 4:2-18,4:20-5:17,5:20-6:8, 6:10-18;Eph. 1:1-2:7,2:10-5:6, 5:8-6:6,6:8-18, 20-24;Phil. 1:1, 1:5-15,17-28, 1:30-2:12,2:14-27, 2:29-3:8,3:10-21, 4:2-12,14-23;Col. 1:1-2, 5-13,16-24, 1:27-2:19,2:23-3:11, 3:13-24,4:3-12, 16-18;1 Thes. 1:1, 1:9-2:3,5:5-9, 23-28;Heb. 1:1-9:16,9:18-10:20,10:22-30,10:32-13:25

The original contents of P46 are subject to debate. If themanuscript was indeed 104 pages long (and the quire numberings make it clearthat it was intended to be so), and the number of letters per page weremaintained, there is no possible way it could havecontained all the Pastoral Epistles; the remaining space would have allowedinclusion of 2 Thessalonians but not much more. But, of course, scribes hadto guess how many pages they would need in a single-quire codex. The Pastoralsrepresent only a little more than 10% of the Pauline corpus, and an scribe'serror of 10% in estimating the length of the codex is not impossible. This isespecially true in a case like P46 where the scribe did nothing toregulate his page size; there are no rules or prickings on the pages. Thus, while itseems fairly likely that P46 did not and was not intended to includethe Pastorals, the possibility cannot be denied that they were included onadditional leaves attached at the end. Sanders says that there were, onaverage, more lines per page in the second half of the codex than thefirst, implying that the scribe was trying to fit in more text,but if so, he didn't compress it nearly enough! But it might have given himenough space for some of the Pastorals. Sanders's suggestion (pp. 10-12)is that he intended to include 1 and 2 Timothy but not Titus. But this seemsto be based on Sanders's feeling that Titus is the least Pauline of the Pastorals.

Date/Scribe

Various dates have been proposed for P46, based entirely onpaleographic evidence. The earliest dates have been around the beginning ofthe second century (a date which has significant implications for the formationof the Pauline canon, but to which few experts subscribe);the latest have placed it in the third. Kenyon said early third century; Sanderssaid third without specifying whether early or late. The most widelyaccepted date today is probably that of the Alands, who place it circa 200C.E.

The scribe of P46 seems to have been a professional copyist,working in a scriptorium. The former conclusion is implied by the neat book hand. Thelatter is less certain, but Zuntz notes several places where the scribecame to a crux in copying and left a small gap in the manuscript. Zuntz theorizes,and this seems reasonable, that the scribe was unable to read or understandthe exemplar, and so left space to allow the corrector to settle the reading.

And Sanders notes that there are no rules or fixed borders on the pages,meaning that the margins vary widely. That the scribe was professional isimplied by the surprising straightness of the lines and the generally highlylegible writing, but he seems to have made few efforts to prepare a high-qualitybook.

It looks as if the scribe was writing for pay; Sanders observes that thereare stichoi counts on four of the books -- and all of them are somewhatlarger than expected; that for 2 Corinthians is extremely high. It is as ifthe scribe was cheating to increase the amount he was paid.

The Nomina Sacra are in evidence, and many ofthem have settled on their final usage rules(e.g. the singular θεως and its inflected forms are consistentlyabbreviated, but the plurals are not; similarly with κυριος) -- but some formsare not entirely fixed, since we see both υς and υις for υιος, and χς χρς for χριστος as well as ις ιης for ιησους.

There is very little punctuation by the original scribe, but Sandersreports that there are spaces between words at various points which tend tocorrespond with breaks in sense. Some of these spaces are much wider thanothers, but Sanders does not find a clear pattern to these. Some punctuationwas added by a later hand. Sanders finds only one accent, and a dozenbreathing marks. Initial iota is usually but not always marked with adiaeresis; upsilon is marked about two-thirds of the time. The mark almostnever occurs within a word, so the diaeresis might almost be treated asequivalent to a space, marking a new word.

Despite his apparent profession, the scribe left a great deal to be desiredin other areas as well;P46 contains a high number of peculiar errors. Zuntz thinks (andhere again I believe he is right) that the copyist did much of the copyingwhile tired or otherwise not at his best, as the errors seem to come in bunches,and are often quite absurd (e.g. writingΓΡΑ for ΓΑΡ).

The correctors weren't much better. The first corrector was the scribehimself, who occasionally spotted his own errors and attempted to repair them.The second corrector seems to have been contemporary, and employed as theδιορθωτης.But thisscribe wasn't all that much better; according to Zuntz, he missed the largemajority of the original scribe's peculiar errors. (This raises the possibilitythat the errors were in their common exemplar, but Zuntz does not believe this.)

A third corrector, working probably in the third century, made a handful of correctionsin a cursive script, as well as a line count. Zuntz thinks that this correctorwas a private owner of the manuscript,making corrections as he spotted them rather than systematically examiningthe manuscript.

Description and Text-type

The text of P46 has been the subject of a quiet but significantcontroversy, with too many scholars ignoring others' results. The first attemptsat studies were by the original editors, Kenyon and Sanders. Sanders, who hadaccess to the larger portion of the text, did the more detailed study -- althoughit was methodologically flawed, since he counted readings that agreed withvarious manuscripts, without so much as casting percentages. And, sadly, he didnot include 1739 in his table, although he did mention one unique reading itshared with P46. He found, by thismeans, that P46 had fully 199 significant variants that had no othersupport listed by Tischendorf. It also, for the small amont of text where theyoverlapped, had a very high rate of special agreements with P13; Sanderssays they show "very little difference" apart from their singularreadings. Aside from these, he showedthat P46 was closest to B, but its next closest ally, especiallyin Romans and Hebrews (the second book of the codex), was D. Hence, "Western."Sanders also showed pretty clearly, that P46did not align closely with the Alexandrian text as exemplified by ℵ A C 33.So the sense, when the manuscriptwas first found, was that it had mostly Alexandrian readings, but witha number of "Western" readings as well, especially in Romans. This wasdemonstrated, e.g., by the work of Sanders.

The only possible word for this description is "simplistic." Anumber of those so-called "Western" readings are not readingscharacteristic of D-F-G, but rather scribal blunders in P46.

There is something we must remember here. If two manuscripts display a mixtureof Alexandrian and "Western" readings, they may simply be mixed manuscripts.But if they display the same pattern of mixture, then they are geneticallyrelated.

It should also be noted that P46 and B have a number of singularagreements -- and that these agreements are by no means harmonistic adjustmentsor the like. Several of them (e.g. Col. 2:2,του θεουχριστου;Col. 3:6, omitεπι τουςυιους τηςαπειθειας)display strong signs of originality.

It was Zuntz who first tackled this issue head-on. In The Text of the Epistles:A Disquisition upon the Corpus Paulinum, he examined the text of Paul startingnot from the established Alexandrian/Byzantine/"Western" perspective butfrom the standpoint of P46. This proved an immensely (and probablyexcessively) laborious process; it took Zuntz a whole volume just to examine thedata for two books (1 Corinthians and Hebrews). Nonetheless, it produced a noteworthyresult: P46 and B form a group (along with a handful of other witnesses)which is clearly distinct from the main Alexandrian group found inℵ A C 33 81 1175 etc.

Zuntz proceeded to confuse the issue by calling this type "proto-Alexandrian,"Even though he found that, where the types differed, both the proto-Alexandrian andAlexandrian texts preserved original readings, he still gave the clear impression thatthe proto-Alexandrian text was a forerunner of the mainstream Alexandrian group. Ibelieve Zuntz knew better, but he did not really analyse the relations between his types,except on a reading-by-reading basis. This made his results hard to understand. Inaddition, Zuntz analysed the data only with respect to P46. This soundsreasonable, but in fact it has severe drawbacks. By his method, any manuscript whichhas a significant number of readings found only in P46+B, and not in theAlexandrian or Byzantine or "Western" texts, will appear to belong to theP46 type. So the Bohairic Coptic, which actually appears to be anAlexandrian text with some P46/B mixture, went into the P46/Btype, as did 1739 (which on detailed examination shows readings of all three othertext-types, plus some of its own, making it perhaps a text-type in its own right).

Unfortunately, Zuntz's research has not been pursued. Metzger's The Text ofthe New Testament, for instance, persists in describing it in terms ofAlexandrian and "Western" readings. And Zuntz's research needs to becontinued, as it focuses entirely on P46 and does not examine thetradition as a whole.

My own results imply that there are fully five text-types in Paul:The Alexandrian text of ℵA C 33 81 1175 1506 and the Bohairic Coptic; the P46/B type (consisting onlyof these two and the Sahidic Coptic; this type too seems associated with Egypt, andso needs a name); the Western text of D F G and the Latins, the Byzantine text, andthe Family 1739 text (in Paul, 1739 0121 0243 6 424** 630+2200 (Romans-Galatians) 1881;Origen's text is close to, but not identical with, that of this group). TheAlexandrian, P46/B, and 1739 texts are somewhat closer to each otherthan to the other two, but by no means a single text. But it should be noted thatthese results, like Zuntz's, have not been tested (though based on stronger statisticaltools than most scholars have used).

P46 should have been the most important papyrus ever discovered.P45 is too fragmentary and periphrastic to be important, P47too limited in extent, P66 too error-prone, and P72 andP75 too close to B to really contribute much. P46 should havechanged our view of the entire history of the text of Paul. Somehow, this seemsnot to have happened.

Still, some have argued strongly for the significance of particular readings.Sanders, p. 35, for instance, strongly suggests that P46's uniqueplacement of Romans 16:25-27 (which it locates after chapter 15) is original.

Other Symbols Used for this Manuscript

Bibliography

Note: As with most major manuscripts, no attempt is made to compile a completebibliography.

Collations:
Frederick G. Kenyon, The Chester Beatty Biblical Papyri. (P46is found in fascicle III, covering Paul)
Henry A. Sanders, A Third-Century Papyrus Codex of the Epistles of Paul(which includes both the material in Kenyon and the leaves not found in that book.Note: The 2015 reprint by Wipf & Stock is a fairly good but not perfect reprint;the text is mostly quite legible but the photos are not ideal. It also has the peculiarhabit of filling in the gaps on damaged pages with the text of the Textus Receptus.)
See also K. Junack, Das Neue Testament auf Papyrus, Vol. 2: Die paulinischen Briefe

Sample Plates:
Aland & Aland, The Text of the New Testament (1 plate)
Comfort, Manuscripts of the Greek Bible (1 plate, same page as the above)
Comfort, The Quest for the Original Text of the New Testament (1 plate; same page as above)
Metzger, Manuscripts of the Greek Bible (1 plate)
Metzger, The Text of the New Testament (1 plate -- again, the same leaf)
Sir Frederick Kenyon & A. W. Adams, Our Bible and the Ancient Manuscripts (1 plate)

Editions which cite:
Cited in NA16 and later, UBS, Merk, Bover

Other Works:
Perhaps most important of the many works on P46 is the one already mentioned,as it is the only one to treat P46 in light of its own text rather than bycomparison to the more recent uncials:
G Zuntz, The Text of the Epistles: A Disquisition Upon the Corpus Paulinum.


P48

Location/Catalog Number

Florence, Laurenxian Library, PSI 1165.

Contents

Portions of Acts 23:11-17, 25-29.

Date/Scribe

Dated paleographically to the third century. The script is considered quite similarto P13.

Description and Text-type

P48 is extremely defective even for the surviving portion of a leaf. Wehave portions of three margins, but the key word is "portions"; we have reallyonly about ten lines, from the middle of the page, and even those are damaged(e.g. one whole vertical strip of papyrus has been lost). The latter verses hardly existat all; the surviving material is just a few strings and strips extending down to thebottom margin of the page.

It has become traditional to regard P48 as "Western" -- theAlands, e.g., list is as having a Category IV text,free but related to D. It is worth noting, however, that P48 and D haveno common material at all.

Determining the actual text-type of P48 is extremely difficult simplybecause of its limited size. The Nestle-Aland text, for instance, reports ten readingsfrom the first section (Acts 23:11-17). Two of these readings are singular accordingto the apparatus, one is supported only by pc, and four are supported onlyby versions (usually Latin). One is supported by 614 h and the Harklean margin. Butseveral of these are really conjectural readings from the heavily damaged portionof the papyrus. At least one reading (23:16, insertεαν δεη καιαπεθανειν)is basedon only the barest handful of letters and is reconstructed on the basis of 614 hhark-marg. This can hardly be accepted as valid evidence of text-type.

Other Symbols Used for this Manuscript

Bibliography

Collations:
E. Lobel, C. H. Roberts, E. P. Wegener, Oxyrhynchus Papyri, Volume 18.

Sample Plates:

Editions which cite:
Cited in Merk, NA26, NA27, and the UBS editions.

Other Works:
Comfort, Early Manuscripts & Modern Translations of the NewTestament, p. 55


P51

Location/Catalog Number

Oxford (Ashmolean Museum, Papyrus Oxyrhynchus 2157).

Contents

Portions of Galatians 1:2-10, 13, 16-20. Every line of the survivingfragment is damaged (usually at both ends); every surviving verse ismissing at least a few letters.

Date/Scribe

Dated paleographically to the fourth or fifth century.

Description and Text-type

Aland and Aland list P51 as Category II.It is hard to see how they determined this, however, as the fragment is so small.Collating its text against P46ℵ A B D G K L81 30 365 1739 produced only eight variants where at least two of these manuscriptsagree against the others; in these eight readings, P51 showed the followingrates of agreement:

ManuscriptAgreement Rate
P463/7=43%
3/8=38%
A3/8=38%
B7/8=88%
D2/8=25%
G2/8=25%
K2/8=25%
L2/8=25%
813/8=38%
3304/8=50%
3652/8=25%
17395/8=63%

Thus P51 is quite close to B. This is confirmed by the original editors,who describe the text as "eclectic... its closest affinities seem to be with B,but an agreement with D F G against ℵ A B P46is worth noting." This reading is not, however, a true agreement with the"Western" witnesses; where D* F G readαποστολωνειδον ουδεναand the remaining witnesses haveαποστολωνουκ ειδον, P51 appears to conflate to read αποστολωνουκ ειδον ουδενα(It should be noted, however, that every letter of this reading is at least slightlydamaged; we should perhaps not place much importance on this variant.)It is curious to observe that P51 is not close to B's ally P46; asthe editors note, "None of the three peculiar readings of ...[P46]...find support here, nor does [P51] ever agree with P46 exceptwhen the latter is supporting B."The most interesting reading of P51 is, surely, in Gal. 1:5, where (alongwith H 0278 330) it readsω εστιν η δοξα.Thus, given the small amount of text we have to work with, we can hardly bedogmatic about P51's text.

Other Symbols Used for this Manuscript

Bibliography

Collations:
E. Lobel, C. H. Roberts, E. P. Wegener, Oxyrhynchus Papyri, Volume 18.
See also K. Junack, Das Neue Testament auf Papyrus, Vol. 2: Die paulinischen Briefe

Sample Plates:

Editions which cite:
Cited in NA26, NA27, and the UBS editions. (The editionof Merk also claims to cite it, but lists it as containing Matthew!)

Other Works:
Comfort, Early Manuscripts & Modern Translations of the NewTestament, p. 55


P52

Location/Catalog Number

Manchester, John Rylands Library, Gr. P. 457

Contents

Portions of John 18:31, 32, 33, 37, 38 (see transcriptionbelow)

Date/Scribe

Generally dated to the second century. C. H. Roberts, who first observedthe manuscript, dated it before 150 C.E. More recentobservers have tended to date it in the range of 110 to 125 C.E.

Description and Text-type

Aland and Aland list P52 as a normal text. However, it shouldbe noted that we really know nothing about the textual affiliationsof this manuscript, which contains roughly 118 legible letters. The mostnoteworthy feature of the manuscript is its age -- though even this should betaken with some caution. How certain can a paleographic determination be whenit is based on so small a sample?

I have also seen it stated that P52 comes from a single-quirecodex. However, we have only a fragment of a single leaf -- and no part of thebinding. Because it is written on both sides, it is safe to assume thatP52 is a codex. But we have no basis on which to claim that itis a single-quire codex.

The story of the manuscript is well-known. Acquired by Grenfell in Egyptin 1920, it went unnoticed among many other manuscript fragments until 1934,when C. H. Roberts recognized that it contained part of the Gospel of John.Impressed with the antiquity of the writing, he hastily published a booklet,An Unpublished Fragment of the Fourth Gospel in the John RylandsLibrary. Despite some caution among scholars about his early andprecise dating, almost all accept that it comes from the second century --simultaneously proving that the codex form and the Gospel of John werein use by that date.

The surviving fragment is only about 9 cm. tall by 6 cm. wide atits widest, counting lines makes it appear that the pages contained abouteighteen lines of about 32 letters per line. This implies a page sizeof about 22 cm. by 20 cm.

Textually P52 tells us little. The complete text istranscribed below:

recto
ΟΙΙΟΥΔΑΙ  ΗΜΕ
ΟΥΔΕΝΑΙΝΑΟΛ
PΕΝΣΗΜΑΙΝΩ
ΘΝΗΣΚΕΙΝΙΣ
ΡΙΟΝΟP
ΚΑΙΕΙP
  ΙΩ

verso
ΤΟΓ  ΝΝ  ΑΙ
ΣΜΟΝΙΝΑΜΑΡΤΥ
  ΤΗΣΑLΗΘΕ
  LΕΓΕΙΑΥΤΩ
    ΙΤΟΥΤ
    ΤΟΥΣΙ
      ΜΙ

As noted, it appears that P52 had about thirty characters per line.If so, then the likely reconstruction of the surviving lines isas follows (surviving characters shown in upper case, the rest in lower)

recto
ΟΙ ΙΟΥΔΑΙοι ΗΜΕιν ουκ εxεστιν αποκτειναι
ΟΥΔΕΝΑ ΙΝΑ Ο Λογος του ιυ πληρωθη ον ει-
ΠΕΝ ΣΗΜΑΙΝΩν ποιω θανατω ημελλεν απο-
ΘΝΗΣΚΕΙΝ ΙΣηλθεν ουκ παλιν εις το πραιτω-
ΡΙΟΝ Ο Πιλατος και εφωνησεν τoν ιν
ΚΑΙ ΕΙPεν αυτω συ ει ο βασιλευς των ιου-
δαΙΩν...

verso
(...lευς) ειμι εγω εις τουΤΟ ΓεγΝΝημΑΙ
και εληλυθα εις τoν κοΣΜΟΝ ΙΝΑ ΜΑΡΤΥ-
ρησω τη αληθεια πας ο ων ΤΗΣ ΑΛΗΘΕι-
ας ακουει μου της φωνης ΛΕΓΕΙ ΑΥΤΩ
ο πιλατος τι εστιν αληθεiα καΙ ΤΟΥΤο
ειπων παλιν εξηλθεν προς ΤΟΥΣ Ιου-
δαιους και λεγει αυτοις εγω ουδεΜΙαν

Observe the mis-spellings ofΗΜΕιν(line 1r),ΙΣηλθεν(line 4r).

Perhaps more interesting are the uses of the name of Jesus in lines 2r and 5r.Was the name abbreviated? This is an important and difficult question.Looking at the verso, we find the following line lengths: 28, 30 (38 ifεις τουτοis included),29, 28, 29, 28, 31. In the recto, if "Jesus" is abbreviated, we have 35,31, 31, 33, 28, 30; if it is expanded, 35, 34, 31, 33 (28 if we omitπαλιν),31, 30. This isproblematic, as the average line lengths on recto and verso are distinctlydifferent -- 29 for the verso, 31.33 or 32.33 for the recto. If we consideronly the recto, using the long forms produces less deviation for the linelengths (standard deviation of 1.97; it is 2.42 if we use the shortlengths). However, if we take all thirteen lines we can measure, usingthe abbreviations produces the lesser deviation (2.14, with a mean linelength of 30.1; without abbreviations the mean is 30.5 and the deviation2.30). On the whole, then, it is perhaps slightly more likely that themanuscript used the nomina sacra than not, butit is absolutely impossible to be dogmatic.

As far as interesting variants go, P52 tells us little.The following is a list of variants to which it attests (note that theseare all either idiosyncratic readings or of trivial importance, oftenboth):

By the nature of the case, P52 cannot help us with the variantadd/omit εγω (afterειμι in verse 37).

Other Symbols Used for this Manuscript

Bibliography

The bibliography for P52 is too extensive to be trackedhere. The basic article is the C. H. Roberts item (An UnpublishedFragment of the Fourth Gospel in the John Rylands Library)mentioned above. For more popular works on the subject see the lists below.

Collations:
Collations of P52 are common -- and often rather optimisticin their readings of almost obliterated letters. Many include reconstuctionsof the text as well. The following list includes some of the less scholarly,but more widely available, reconstructions:
Finegan, Encountering New Testament Manuscripts, pp. 85-100 (text, recontruction, and comparison with other manuscripts)
Metzger, Manuscripts of the Greek Bible, p. 62 (includes reconstructed text)
Salmon, The Fourth Gospel: A History of the Text, pp. 50-53

Sample Plates:
Almost every modern introduction to textual criticism includes photosof P52 (which is why no photo is included here). Examplesinclude:
Aland & Aland, The Text of the New Testament
Finegan, Encountering New Testament Manuscripts
Metzger, Manuscripts of the Greek Bible
Salmon, The Fourth Gospel: A History of the Text

Editions which cite:

Cited in all the recent Nestle-Aland editions and the like; it shouldbe noted, however, that P52 is so short that it plays no realrole in the critical apparatus.

Other Works:
Comfort, Early Manuscripts & Modern Translations of the NewTestament, pp. 55-56


P54

Location/Catalog Number

Princeton (University Library, P. Princeton 15).

Contents

Portions of James 2:16-18 (beginning withτου σωμτος),22, 24-25, 3:2-4. The manuscript is damaged onboth sides and at the bottom (though the defect at the bottom does notinvolve much text); in addition, the manuscript is broken in the middle(it in fact consists of two major pieces and some shreds), which explainshow a single leaf can contain four sections of text. All four sections aredamaged. The state of the fragment is so bad that it is hard to determineeven the line length, but it appears to have been about twenty characters;we have about ten characters in the surviving lines. A total of 29 linessurvive.

Date/Scribe

Dated paleographically to the fifth or sixth century. The hand is quitefirm and clear (or would be if the fragment were not so discoloured and faded).

Description and Text-type

Aland and Aland list P54 as Category IIIor possibly Category II. The Nestle text, however, cites it for only four readings(one of them, in 2:18, being subsingular); there just isn't enough text to make aclear determination of the manuscript's type.

Other Symbols Used for this Manuscript

Bibliography

Collations:
E. H. Kase, Papyri in the Princeton University Collections, Volume II
Ellwood M. Schofield, The Papyrus Fragments of the Greek New Testament

Sample Plates:

Editions which cite:
Cited in NA26, NA27, and the UBS editions.

Other Works:


P64+P67

Location/Catalog Number

P64: Oxford, Magdalen College Gr. 18.
P67: Barcelona, Fundació Sant Lucas Evangelista P. Barc. 1

Contents

(scraps of) Matthew 3:9, 3:15, 5:20-22, 5:25-28, 26:7-8, 26:10, 26:14-15, 26:22-23, 26:31-33. For instance, P64, the Oxford fragment (which contains the scrapsof Matthew 26) consists of three small scraps, the largest four lines tall with and thewidest being only about 12 letters across. The Barcelona fragment contains the portionsof Matthew 3 and 5, and is not in much better shape.

It appears that the page was originally about 200 by 130 mm., with two columns perpage and about 35-40 lines per page.

Date/Scribe

Dated paleographically to around the year 200.

Interestingly, the fragment appears to have Nomina Sacra -- but not in full form. The words κυριος and Ιησους are abbreviated, but the words do notappear to have lines above them.

Description and Text-type

Aland and Aland list P64+67 as Category Iwith a "Strict" text.

It has been suggested that P64+P67 are part of the same document as P4. This is certainly possible on chronological grounds.The latter manuscript, however, contain fragments of Luke. IfP4 and P64+P67 are indeed one manuscript, theyrepresent quite possibly the earliest instance of a papyrus containing more thanone gospel. I do not think the evidence sufficient to draw firm conclusions, however.For further discussion, see the entry on P4.

The text of P4 is clearly Alexandrian. The evidence forP64+P67 frankly does not strike me as sufficient to draw a firmconclusion.

It is interesting to note that the Alands rate P64+P67a "strict" text, P4 a "normal" text. However, neitheris extensive enough for such a judgment to be truly meaningful.

Other Symbols Used for this Manuscript

Bibliography

Collations:

Sample Plates:

Editions which cite:Cited in NA26, NA27, and the UBS editions.

Other Works:

Tommy Wasserman, "A Comparative Textual Analysis of P4 and PP64+67" discusses the text, but see the article on P4 for the frankly quite obvious limitations of the Alands' self-referential assessment of manuscripts.


P74

Location/Catalog Number

Cologne, Bodmer library. Bodmer Papyrus XVII

Contents

Contains most of Acts (1:2-5, 7-11, 13-15, 18-19, 22-25, 2:2-4,2:6-3:26, 4:2-6, 8-27,4:29-27:25, 27:27-28:31) and fragments of all sevenCatholic Epistles (portions of 75 verses of James, 16 verses of 1 Peter, 4of 2 Peter, 27 of 1 John, 4 of 2 John, 2 of 3 John, and 5 ofJude).

Date/Scribe

Dated paleographically to the seventh century.

Description and Text-type

Aland and Aland list P74 as Category I.Richards lists it as amember of his Group A3 (Family 1739), but even he admits "P74was classified even though there are only eight non-TR readings in 1-3 John by whichthe manuscript could be judged. We placed P74 in A3 becauseseven of its eight non-TR readings are group readings in A3, while only fiveof the eight are group readings in A2 [the main Alexandrian group], andjust three of the eight are A1 [Family 2138] group readings" (W. L.Richards, The Classification of the Greek Manuscripts of the JohannineEpistles, p. 139). However, Richards seems to have been betrayed by hisinaccurate groups and his small sample size. In the Catholic Epistles as a whole(meaning primarily James), P74 is not close to Family 1739.The following data examines all readings of P74 in the Catholics cited explicitly inNA27. There are exactly fifty such readings. Of these fifty, P74agrees with the Byzantine text in a mere six. Nine of its readings are singularor subsingular (i.e. not supported by any of the test witnessesℵ A B L P 33323 614 1241 1505 1739) It has six readings which have only one supporteramong the test witnesses. Its rate of agreements are as follows:

WitnessOverall
Agreements
Agreements supported only
by P74 and the listed witness
17 of 50 (34%)0
A30 of 49 (61%)4
B21 of 50 (42%)1
L11 of 50 (22%)1
P14 of 46 (30%)0
3321 of 44 (48%)0
32317 of 50 (34%)0
61414 of 50 (28%)0
124120 of 49 (41%)0
150514 of 50 (28%)0
173922 of 50 (44%)0

Thus P74's allegiance is clearly with A. If we omit P74'snine singular readings, they agree in 30 of 41 variants, or 73% of the time. Ais the only manuscript to agree with P74 over 70% of the time. In addition,A agrees with the larger part of P74's most unusual readings.

We also observe that P74's next closest relative is 33, which isfairly close to A.

Without adding statistics, we can observe that P74 seems to have asimilar text of Acts. Although it has been called Byzantine, in fact it is ahigh-quality Alexandrian text of that book, and deserves the Alands'Category I description.

Other Symbols Used for this Manuscript

Bibliography

Collations:
Rudolf Kasser, Papyrus Bodmer XVII: Actes des Apôtres, Epîtresde Jacques, Pierre, Jean et Jude
See also K. Junack, Das Neue Testament auf Papyrus (volumes 1 and 3, CatholicEpistles and Acts)

Sample Plates:
Aland & Aland (1 plate)

Editions which cite:
Cited in all UBS editions and in NA26 and NA27

Other Works:


P75

Location/Catalog Number

Cologny (Geneva), Switzerland, Bodmer library. Bodmer Papyrus XIV, XV

Contents

Contains major portions of Luke and John: Luke 3:18-22,3:33-4:2, 4:34-5:10,5:37-6:4, 6:10-7:32,7:35-39, 41-43, 7:46-9:2,9:4-17:15, 17:19-18:18,22:4-end, John 1:1-11:45, 11:48-57,12:3-13:10, 14:8-15:10.The volume, despite loss of leaves, is in surprisingly good condition, we even haveportions of the binding (which is thought to have been added later).We have all or part of 102 pages (51 leaves), out ofan original total of about 144 (72 leaves). Generally speaking, the earlier leavesare in better condition; many of the pages in the latter part of John have gone topieces and have to be reconstructed from fragments.

Date/Scribe

Dated paleographically to the third century (with most scholars tending towardthe earlier half of that century); Martin and Kasser, who edited the manuscript,would have allowed a date as early as 175. The scribe seems to have been generallycareful, writing a neat and clear hand (though letter sizes vary somewhat), and(with some minor exceptions) using a fairly consistent spelling. Colwell observedthat the natural writing tendencies of the scribe were strongly restrained by thetext before him, indicating a copy of very high fidelity. The editors ofthe codex argued that the copyist was a professional scribe. We do note, however,that lines are of very variable length (25 to 36 letters per line), as are thepages (38 to 45 lines per page). As P75 is a single-quire codex of(presumably) 36 folios, it has been argued that the scribe was trying to get moretext on a page to hold the codex to the available space.

Description and Text-type

The fact which has struck every examiner of P75 is its extremely closeresemblance to B. A number of statistical studies to this effect have been made;as far as I know, however, all have been done by textual critics with weakmathematical backgrounds and with inadequate controls. Thus, none of their figuresfor agreements between manuscripts can be regarded as meaning much. Still, theresult is unquestionable: P75 is closer to B than to any other manuscript,and vice versa. There are enough differences that P75 cannot be theparent of B, and is unlikely to be a direct ancestor, but P75 and Bcertainly had a common ancestor, and this ancestor must have been older thanP75. Moreover, both manuscripts have remained quite close to thisancestral text. The mere fact that the two agree does not tell us how goodthis ancestral text is (most scholars would regard it as very good, but thisis for other reasons than the closeness of the two manuscripts). The pointis that, good or bad, we are ableto reconstruct this text with great accuracy.

Interestingly, there has been no systematic study examining the text ofP75. The Alands, of course, list it asCategory I, with a strict text, but thisis based simply on the date and character of the manuscript; it is notreally an examination of the text. Wisse, for some reason, did not profileP75, even though it is the only papyrus of Luke substantialenough to allow such an evaluation (at least of Chapter 10).

The discovery of P75 has had a profound effect on New Testamentcriticism. The demonstration that the B text is older than B seems to haveencouraged a much stronger belief in its originality. The UBS committee, forinstance, placed the Western Non-Interpolationsback in their text based largely on the evidence of P75.

The irony, as E. C. Colwell pointed out in the essay "Hort Redivivus:A Plea and a Program" (p. 156 in the reprint in Studies in Methodologyin Textual Criticism of the New Testament), is that P75 shouldhave had no such effect. The existence of manuscripts such as P75had never been questioned. The major Bodmer papyri (P66, P72,P74, and P75) are important and influential witnesses,but they should have little effect on our textual theory. The truly significantwitnesses were the Beatty papyri -- P46, as Zuntz showed, shouldhave completely altered our view of the text of Paul (but somehow it didn't);P47 perhaps should have a similar if less spectacular effect on ourtext of the Apocalypse; and P45 (as Colwell showed) allows us tosee the sorts of liberties some copyists could take with the Biblical text.

This is not to deny the great value of P75. Since P66is a notably inaccurate copy, and P45 paraphrases (see Colwell,"Method in Evaluating Scribal Habits: A Study of P45, P66,P75," pp. 196-124 in Studies in Methodology), P75is the earliest substantial and careful manuscript of the Gospels. Most wouldalso regard it as having the best text. It does have a few limitations, however.It has been accused of omitting minor words such as personal pronouns (see page121 in the Colwell essay).

Other Symbols Used for this Manuscript

Bibliography

Note: As with most major manuscripts, no attempt is made to compile a completebibliography.

Collations:
Rudolf Kasser and Victor Martin, Papyrus Bodmer XIV-XV. Twovolumes; Volume I contains the Lukan material, Volume II the Johannine.
Supplementary portions of the text are found in Kurt Aland, "Neueneutestamentliche Papyri III," New Testament Studies #22.

Sample Plates:
Complete plates in Kasser & Martin. Sample plates in almost everyrecent book, including Aland & Aland, Metzger's Text of the NewTestament and Manuscripts of the New Testament, Finegan,Encountering New Testament Manuscripts, and anything ever publishedby Philip Wesley Comfort.

Editions which cite:
Cited in all editions published since its discovery -- includingNA25 and higher, all UBS editions, and even Hodges &Farstad.

Other Works:
Calvin Porter, "Papyrus Bodmer XV (P75) and the Text ofCodex Vaticanus," Journal of Biblical Literature 81.
E. C. Colwell,"Method in Evaluating Scribal Habits: A Study of P45, P66,P75," pp. 196-124 in Studies in Methodology


P78

Location/Catalog Number

Oxford, Ashmolean Museum. Oxyrhynchus Papyrus 2684

Contents

Portions of Jude 4-5, 7-8 (additional material illegible)

Date/Scribe

Dated paleographically to the third or fourth century.

Description and Text-type

P78 contains only a fragment of a single leaf, measuring a littleover 10 cm. across by 2.5 cm. tall. This suffices to hold three to four lines of text.There are two columns of about a dozen lines each. The surviving portion appearsto be the top of the page.

The verso portion is easily read, although written in a rather hurried,inelegant hand. The left-hand column begins with verse 4,αρνουμενοιand ends with verse 5,ειδοτας.Column 2 begins with verse 7,αιωνιουand ends with verse 8,ενυπνιαζομε[νοι].

The recto portion is in much worse shape, being practically illegible. The leftcolumn begins with verse 8,σαρκα μεν.The rest of this column is only marginally legible, and the second column cannotreally be deciphered (at least in visible light). The fragment thus contains atotal of only about 100 Greek characters.

Nonetheless its text is striking. The Alands classify it asCategory I (based on its date) with a "free"text. We observe several noteworthy readings:

Several of these may be the result of a hasty and careless scribe. Sadly, thefragment is so short that we cannot really draw further conclusions.

Other Symbols Used for this Manuscript

Bibliography

Collations:
L. Ingrams, P. Kingston, P. Parsons, J. Rea, Oxyrhynchus Papyri,volume 34.

Sample Plates:

Editions which cite:
Cited in UBS4, NA26, and NA27.

Other Works:
Comfort, Early Manuscripts & Modern Translations of the NewTestament, pp. 64-65


P79

Location/Catalog Number

Berlin, Staatliche Museen, P. 6774.

Contents

Portions of Hebrews 10:10-12, 28-30, the recto containing parts of 12 ασμε[νοι εσμεν]... 15 το διηνε[κες] and the verso including parts of 28 [αθετη]σας... 30 [οιδαμεν] γαρ τον.

Date/Scribe

Dated paleographically to the seventh century.

Description and Text-type

A single leaf, the surviving portion quite small (5.3 cm wide by 11.2 cm. high), from the inner side of the leaf. It is believed that it was in two columns (although all the surviving text is from a single column); 15 lines survive on the recto and 17 on the verso; it is estimated that it originally had 32 lines per column.

The text seems to average about 14 or 15 letters per column. Typically seven or eight letters per line survive, so we have about half the text.

The transcriptions in Horsley and in Das Neue Testament auf Papyrus do notentirely agree, with auf Papyrus marking many more letters as uncertain. One difference extends beyond marking a letter as uncertain; in 10:12, Horsley gives the first word of the verse as αυτος (read also by Dc K L 056 0142 0151; this is clearly the Byzantine reading) while auf Papyrus reads it as ουτος (found also in P13 P46 ℵ A C D* P Ψ 0278; this is clearly the reading of all the older text-types). There is one other apparently-unique reading in 10:11, where it omits πολλακις (apparently; the text is defective here, and there is no room for the word in the line, but the scribe could perhaps have squeezed it into the margin), but the only other variants are itacisms and spelling variants. There are several instances of the symbol ′′ to indicate a sense break.

The Alands describe P79 as Category II,but it's not obvious why, given that the only clearly Alexandrian reading (the αυτος/ουτος variant) is extremely uncertain.

Other Symbols Used for this Manuscript

Bibliography

Collations:
The compete text is printed in G. H. R.Horsley, New Documents Illustrating Early Christianity, 2, 1982, p. 139
See also K. Junack, Das Neue Testament auf Papyrus, Vol. 2: Die paulinischen Briefe (description and bibliography on pp. LXII-LXIII).

Sample Plates:

Editions which cite:
Cited in NA27 and NA28 and in UBS4 and UBS5.

Other Works:


P90

Location/Catalog Number

Oxford, Ashmolean Museum. Oxyrhynchus Papyrus 3523

Contents

Portions of John 18:36-19:7

Date/Scribe

Dated paleographically to the second century (making it, after P52,perhaps the oldest surviving New Testament papyrus). The script is considered similar tothe "unknown gospel," P. Egerton 2.

Description and Text-type

P90 contains only a part of a single leaf, about 15 cm. tall and nowheremore than six cm. wide. It appears that we have the entire height of the leaf, but onlya portion of its width, with thirteen or fewer characters surviving on each line (24 linesvisible on the recto, 23 on the verso). Even the surviving characters are often illegible.(So much so that, of the eleven readings noted in NA27, eight are markedvid.) The manuscript appears to have originally has about twenty charactersper line, meaning that even the best-preserved lines are missing a third of theirtext, and most are missing half or more. The hand is generally clear but notpolished.

Because the manuscript is so newly-discovered, it has not been classified accordingto any of the standard classification schemes. It does not appear to contain anynoteworthy variants. The following table shows its rate of agreement with somekey manuscripts in the variants cited in NA27:

MSAgreementsPercent Agreement
P665/1145%
7/1164%
A1/119%
B3/1127%
Dsup3/1127%
K2/1118%
L6/1155%
Q2/1118%
13/1127%

With such small samples, our percentages of agreement obviously don't mean much.But it will be clear that P90 is not Byzantine; it appears to be anAlexandrian witness of some kind. Comfort listed it as closest to P66(based probably on some relatively unusual readings they share),but his bias toward early papyri is well-known; in fact it looks closer toℵ. Its lackof kinship with B is noteworthy.

Other Symbols Used for this Manuscript

Bibliography

Collations:
L. Ingrams, P. Kingston, P. Parsons, J. Rea, Oxyrhynchus Papyri,volume 50.

Sample Plates:

Editions which cite:
Cited in NA27.

Other Works:
Comfort, Early Manuscripts & Modern Translations of the NewTestament, pp. 68-69