Friday, December 24, 2010

Hernandez on Codex Sinaiticus: Revelation = Commentary-text!

It appears that Codex Sinaiticus is so unique and unusual in its text of Revelation, that it ought to be classed as a kind of "commentary" and/or collection of readings and interpretations collected over centuries.

Juan Hernandez Jr. of Bethel University, St. Paul, MN, presented a paper titled “Codex Sinaiticus: The Earliest Greek Christian Commentary on John’s Apocalypse?”  In an abstract he states:
The Apocalypse in codex Sinaiticus is a striking example of a fourth-century text that differs substantially from modern critical editions. It exhibits dozens of differences at key points, reflecting the concerns, interests, and idiosyncrasies of its earliest copyists and readers. Taken as a whole, Sinaiticus’s text of Revelation may constitute one of our earliest Christian commentaries on the book, disclosing its fourth-century milieu and anticipating the later concerns of Oecumenius and Andrew of Caesarea. This is no commentary in the contemporary sense, however. Sinaiticus’s readings range from the spectacular to the mundane and include the theological, the liturgical, the commonplace and even the infelicitous. It is a text ever in tension with itself, effective both in its capacity to obscure as well as in its regulation of meaning. Clarity and confusion co-reign and compete for our attention. Despite that, we can discern a concerted effort to elucidate the Apocalypse’s message by scores of changes throughout. Some of these are inherited. Others created. All affected the reading of the text.
Notice originally Posted on Fri, Jul 31, 2009, 
by Peter Nathan Sepher 

Sunday, December 19, 2010

Variations in Genealogical Stemma

In the previous discussion, we talked about sensibly and usefully displaying information in a convenient and intuitive way on a stemma, or genealogical chart.

Periodically, researchers try innovations in order to emphasize certain aspects of the textual data.  One such example is the stemma given by G. Mink in his article recently quoted by Nazaroo in discussing commentaries.  Our interest is however in Mink's innovative stemma, (diagram 17) shown below:

Click on this to Enlarge: Backbutton to return

From the chart, it will be seen that Mink offers two alternate 'routes' that readings could have arrived at each extant manuscript.  In other words, Mink has superimposed a 'main' ("most probable") stemma, and a secondary (2nd most probable stemma).

At once, it should be realized that the solid and dotted lines are not simply two alternate genealogical trees.  Each individual manuscript could have acquired readings from more than one source independently.    This means that the diagram encompasses ALL combinatorical possibilities, and this involves many hundreds of variations as to the actual transmission of readings.   Its not just two stemmas.   Mink has in a relatively simple and intuitive way shown all the most likely possibilities (1st and 2nd order) by which the extant manuscripts could have come to contain their texts.

He still uses vertical position to indicate manuscript date (and therefore has avoided the question of age of text vs. date of MS.).

Other useful innovations are possible.  In the following diagram, mixture (cross-copying of readings) is indicated in contrast to genealogical relationship (where the majority of the text was copied from).

Click to Enlarge: Backbutton to return.

Nowadays, it is acknowledged that manuscript relations are not as sharp as simple lines.  To show dependence and mixture of text-types as opposed to single manuscripts, something more realistic and representative of "flows and streams" of transmission is given below:

Click to Enlarge
Here one can imagine source-streams flowing out and expanding, and areas of influence overlapping.   This is much more representative of 'text-types', with manuscripts being represented in large quantities in the colored zones.

Much can be learned from various solutions to displaying human genealogies.  The important relations between individuals need to be retained in detailed family trees, and such methods give us ideas for similar problems with manuscripts.  In the following diagram, secondary texts, alternate readings, and the spread of influence can be placed inside boxes using symbols and color-codes if desired:

Where generational information is available, it may be more desirable to display that than manuscript ages:

Sometimes, genealogical dependence is unknown at the start of an investigation.  What is needed is to plot affinity or closeness of relation, without presuming dependence per se.   Such diagrams, while not 'true stemma', in that they don't indicate direction of dependence, can help to sort and solve relationships when coupled with other data.  Two such forms are as follows:

 In one diagram, size of node indicates importance, while in the other, the connection is stressed, and its strength is indicated by thickness or style of line.

When many manuscripts are to be displayed  and they are not closely related, and many others are missing from the stream, so that no real accurate stemmas are possible, a different method of display is preferable, as in the following diagram:

In other cases, genealogical connections are possible, but mixture may be as strong as genealogical dependence, and so it can be shown with similar lines:

In some situations, alternate routes of dependence are more important than spread of texts, and these details can be highlighted and elaborated with colors and symbols, as in the diagrams below;

In many cases, the earlier manuscripts are missing, and only the later copies survive.  Here genealogical relations must be deduced by working backward from extant MSS to the proposed source-structures:

 When groupings need to emphasized, another style of display can be used:

Color-coding can be effectively used to indicate groupings, and a more natural, organic display will illustrate the connections nicely:

Since the mass of manuscripts are quite late copies, a realistic understanding of the circumstances can help immensely in capturing the essence of things:

Above all, a look at the many options that are available can greatly stimulate both analysis and articulation/communication of data effectively.

Keep trying various methods, to see what is effective for your data, and someday your genealogical stemmas will be as nice as ours:


Tuesday, December 14, 2010

James Snapp - Tachygraphy in Luke

James recently posted on textual criticism (yahoo groups) the following tidbit:
This might be trivial; nevertheless: in Lk. 2:22, did Luke write the equivalent of "their," (AUTWN) "his," (AUTOU) or "her" (AUTHS)? Maybe the answer should be "No."

AUTWN has extremely strong support. D's AUTOU is probably a result of retro-translation. UBS-2 lists MS 76 as support for AUTHS, and this is repeated by the NET's online notes, but WW observes that Gregory and Hatch saw that this was not the case. Hatch went through the trouble of explaining why 76 was erroneously cited as a witness for AUTHS; he assigns the blame mainly to Scholz, in a footnote in an article, "The Text of Luke II, 22" in Harvard Theological Review (#14, from 1921). This issue of HTR is online at Google Books; I've added it to my virtual library-shelf of NTTC resources. It's the same annual in which Kirsopp Lake drew attention to Epistula Apostolorum. Hatch's article begins on p. 377; the digital page number in Adobe Reader is 396.

Decades after Hatch's article, UBS-2 still cited 76 as support for AUTHS, and so do the notes in the NET. (Sound effect: grunt-sigh of exasperation.) UBS-4 has no note here.

Hatch mentions that Origen discussed the problematic AUTWN in his fourteenth homily on Luke. Hatch also states that the Sahidic version supports AUTOU. Hatch proposes that Luke, misunderstanding a non-Greek source-document that was intended to refer to Mary's purification, wrote AUTOU, and that later scribes replaced AUTOU with AUTWN. He also mentions that although Erasmus and Stephanus printed AUTWN, the Complutensian Polyglott, and Beza, and the Elzevir text contained AUTHS, apparently as a Latin-based conjecture. (This is reflected in the KJV and NKJV.)

Metzger noted that "in cursive Greek script the pronoun was abbreviated AUT with the termination expressed by a "shorthand" stroke." Now, it occurs to me that an ambiguously abbreviated "AUT" could account for all the extant readings: scribes could independently unfold an ambiguously abbreviated "AUT" in different ways, as AUTWN or as AUTOU or as AUTHS.

2427 has some abbreviations of AUTOS, about which Colwell wrote something, but since it's a forgery let's take that off the table.

In an issue of The Journal of Hellenic Studies (XI, 1890), T. W. Allen wrote an article called "Fourteenth Century Tachygraphy." (It includes two Plates which are near the end of the volume.) On p. 291, he mentions "AUTOU is the XYZ of the Grotta Ferrata tachygraphs." ("XYZ" being a symbol in the printed text which cannot be duplicated without special fonts.) In a footnote he adds,
"As this mode of contracting AUTOS is rare, I may mention that it occurs (AUTOU and AUTON) frequently in the Paris MS. Coislin 387 (s. xi.)." (I think Paris MS. Coislin contains Porphyry's Isagoge.)

The Grotta Ferrata tachygraphs are, it seems today, not an easy subject to research in English, but in a short book published in 1889 book called Notes on Abbreviations in Greek Manuscripts, Professor Allen discusses them and their abbreviations. I'll skip most of the details; let's get to what he says about AUTOS on p. 9:
"AUTOS. A ligature for this pronoun worth recording occurs in some of the Grotta Ferrata mss. ; it consists of the A and U run together with the case-ending added : cf. AUTOS AUTOIS AUTHN hEAUTWN hWSAUTWS from Gr. Ferr. B. A. i. and Angel. B. 3. II. A similar combination of A and I occurs in AUTOU from Aed. Christ. 70 (a. 1104) and the ligature is probably common."

In the 1901 issue of the Journal of Hellenic Studies (accessible at Archive), F. W. G. Foat has an article, "On Old Greek Tachygraphy," beginning on p. 238. Foat helpfully profiles the MSS of all sorts, from all ages, that contain special tachygraphic symbols, monograms, etc, from the major MSS down to ostraca. Foat included a Plate (#XVIII) of Brit. Mus. Add. MS 33270, a third-century wax tablet (about the size of a Kindle) that seems to contain about 14 pages of tachygraphical writing. On p. 259 of his article, as Foat attempted to use a feature in the wax tablet to decipher part of another text (in Brit. Mus. Pap. cxxi. 14, col. 27) -- the details are rather dizzying so let's skip them for now -- he wrote:
"We have, as certainly belonging to the tachygraphy of this 3rd century, the small crossbar horizontally placed on an upright stem (in the familiar sense I think of AUT-), from forms common to the waxen-book, to this third-century papyrus, and to the scanty contents of the four Leipzig fragments, though not in the Rainer fragments. These [here a footnote shows the tachygraphical characters themselves] give the readings of AUTOUS, AUTON, AUTA or AUTO, and AUTOCHQON or some such word. But all these, like the last, are merely possible, for this may be an advanced stage of the system before us, and in dealing with such tachygraphy, a posteriori inferences, as we have seen, are almost worthless."
Well, you must see where I'm going with these tidbits: maybe Luke used ambiguous tachygraphy in 2:22, perhaps meaning AUTHS but allowing readers to interpret it as AUTWN, AUTOU, or AUTHS. This conjecture accounts for all its rival variants and does not seem to raise any new interpretive issues.

Yours in Christ,

James Snapp, Jr.

Greek Tachygraphy (shorthand)

click to enlarge

click to enlarge
  PAPYRUS, GREEK TACHYGRAPHY. Three fragments written on both sides, totalling 35 lines, [probably Egypt, 3rd to 4th century], significant examples of ancient Greek tachygraphy or shorthand, approx. 150 x 65mm, 150 x 35mm, 90 x 60mm, with six further very small fragments, two with writing, fragile, unexamined out of glass frame. Provenance: Ajasse sale, Lyon, 12 May 2005 -- the collection of Albin Schram.

Christie's  Lot 1 /Sale 7590,  Price Realised (pounds) 1,250 ($2,444)

Evidence of shorthand systems for writing Greek has survived from the 4th century BC but Greek tachygraphy is less well documented than the Tyronian system for Latin developed in the 1st century BC on the Greek model. Papyri discovered in Egypt, most famously from Oxyrhynchus and Antinoopolis, have proved an extremely valuable source of information since they include manuals of shorthand as well as tachygraphic texts (see H.J.M. Milne, Greek Shorthand Manuals, 1934). Yet their number is comparatively small and they have often survived in a more fragmentary state than the present items. As more is known about Greek shorthand, debate is growing on its possible role and influence in the transmission of texts, particularly of those now known as the New Testament.
(Anon, - Christie's Selling page)


Classical antiquity

The earliest known indication of shorthand systems is from Ancient Greece, namely the Parthenon in which a stone from mid-4th century BC was found. The marble slab shows a writing system primarily based on vowels, using certain modifications to indicate consonants. Hellenistic tachygraphy is reported from the 2nd century BC onwards, though there are indications that it might be older. The oldest datable reference is a contract from Middle Egypt, stating that Oxyrhynchos gives the "semeiographer" Apollonios for two years to be taught shorthand writing. Hellenistic tachygraphy consisted of word stem signs and word ending signs. Over time, many syllabic signs were developed.
- Wikipedia article Tachygraphy

Saturday, December 11, 2010

Introduction - Genealogy

Everyone understands the basic idea of a genealogy.  Manuscripts are copied from other manuscripts, and this forms a line of descent from the original to the latest copy.
Click to Enlarge
Not all copies are used again to make subsequent copies.  (Above, 1st generation copy #10 is used for a master, - then 2nd gen. copy #1, - 3rd gen. copy #253 etc.) Usually, either the best copies are used to make more copies, or else sometimes (under less favourable circumstances) the master-copy will just be whatever copy is available

By the time of Jesus, the craft of document copying was certainly very advanced.  The Jewish scribes had a long history and tradition of careful copying of their religious scrolls, and the Romans used documents extensively for administration of the Empire.   There were both organized religious groups and private-sector operations providing copying services.  Scribes were trained and employed in courts, temples, and libraries, and in places like Alexandria, where papyrus was manufactured, apparently copying services were organized businesses.

Because scribes were paid, manuscripts weren't simply copied.  The amount of papyrus (or parchment) was calculated, the number of lines written were counted, and the content was proof-read and corrected during manufacture.   Potential master-copies of existing works were examined for accuracy, and new copies were checked against the master-copy, and sometimes cross-checked with other good copies.   A reasonably effective system for correcting mistakes had evolved in major copying centers, complete with special symbols and conventions for dealing with variations between copies.

Because of this, many mistakes were caught before a manuscript went out the door of the copying-center.    Most people have heard of the "telephone game" in school, where children whisper a message in a long chain, and the end result is naturally a garbled message.   But the situation with manuscript copying is quite different.  Due to correction and cross-checking, the integrity of the copy is far better protected.   This can also be represented by a diagram:

Because of copy-correction and cross-checking, manuscripts are actually embedded in a protective network or web, which helps to stabilize the text of each copy.  The strength of these extra connections (and purity of text) is based on the quality of the proof-reading.

These extra connections and relationships between manuscripts created by cross-checking and proof-reading are not usually shown on genealogical trees (or stemmas), because the diagrams would become too complicated. 

Because reliable master-copies can be used over and over again to produce new copies, manuscripts of widely varying dates can be equally good, and equally close to the original in terms of copy-generations.   Realistic genealogical "trees" look like wide bushes rather than tall oaks.

The stemma below (from Edmer's 'Vita Wilfridi')  is 'realistic' only in that it shows surviving manuscripts, not in that it represents a true picture of the actual copying stream with all the manuscripts:

 Since errors only happen during copying or correction, and manuscripts can last an awfully long time without change, the actual date of manufacture or age of a manuscript is semi-independent of the "generation number" of a manuscript. What matters is not age at all, but rather three things control the quality of a text:

(1) the quality of the copying during each step,
(2) the quality of the correction during each step.
(3) The number of generations away from the original it is, because errors can be, but aren't necessarily, cumulative.

Nazaroo illustrated the independence of these two axis with his simple diagram and suggestion for a useful layout below:
click to enlarge

This idea was used successfully in the creation of a stemma for Lake's work:

click to enlarge

Wednesday, December 8, 2010

Codex Corsendoncensis GA-3 (12th cent.), Scribal Gloss

Codex Corsendoncensis is such a late manuscript, and valued so little, that it is not even found in the manuscript-lists or apparatus of modern critical editions like the UBS Greek NT.  Nor does Bob Waltz (Encyclopedia of NTTC) even give it an entry in his extensive description of cursives and miniscules.

Its sole claim to fame is that it was one of the few manuscripts used by Erasmus in constructing his Greek NT (1516), and that it contains a clumsy scribal gloss, (which in turn sparked Erasmus' famous annoyed and skeptical remark about Greek manuscripts).  John Marsh describes the gloss: 
"The common text at 2 Cor. 8:4-5. is ...δεξασθαι ημας (v5) και ου καθως ηλπισαμεν...   but the best, and most numerous authorities reject δεξασθαι ημας , which is probably an interpolation. The proprietor of an ancient MS. from which Corserdoncensis was copied, knowing that δεξασθαι ημας was contained in some MSS. but rejected by others, and wishing perhaps to rescue these words from the charge of spuriousness, wrote, with a reference to δεξασθαι ημας  the following note in the margin :  εν πολλοις των αντιγραφων ουτως ευρηται.  The industrious scribe, who wrote Corsendoncensis, taking these words for a part of Holy Writ, which had been omitted in the text and supplied in the margin, transferred them into the body of his own work, and wrote as follows;  ...δεξασθαι ημας (v5)  εν πολλοις των αντιγραφων ουτως ευρηται και ου καθως ηλπισαμεν..."   ( Marsh's Letters to Archdeacon Travis, P. 176  noted in The Critical review, or, Annals of literature, Volume 16  [Feb. 1796] edit.  T.G. Smollett, p. 450)

Jan Krans tells us:
"The reception history of the scribal blunder itself is interesting as well. It was mentioned for the first time by none other than Erasmus. In his 1519 edition, he added a long note to his annotation on these verses. The example must have been very welcome to him, for it is clear proof that even the sacred texts are not free from ridiculous scribal errors and that textual criticism is necessary, whatever theologians may say. He concluded the note by saying that 'we found innumerable places corrupted for this same cause'. Even before the 1519 edition, he had already mentioned the case in his apology against Faber Stapulensis. There, the point made was somewhat more specific. He warned his colleague to 'not naively trust (Greek) manuscripts. Keep thinking critically'. "

This remarkable statement by Erasmus is important, because he never produced any further examples from the "innumerable" instances he claimed to have found.    There plainly aren't that many in the few manuscripts he himself collated (certainly none like the blunder in GA-3).

From this seed of Erasmus, the rumour-weed about scribal glosses begins to grow:
"In subsequent centuries, the blunder was frequently mentioned in text-critical books. Bengel, in his 1734 Greek New Testament, refers to Erasmus.   Metzger also mentions the case in his Text of the New Testament (3-1992, p. 194; see also Metzger/Ehrman 4-2005, pp. 258-259), referring to Bengel."  (- Jan Krans, Weblog 2010)

Yet the fragility and weakness of this claim is apparent, even to Jan Krans:
"The example ... shows that marginal annotations which were mistakenly adopted into the text did occur. The question remains, however, how widespread the phenomenon actually is, and how it can be demonstrated in cases that are less clear than the one in miniscule GA-3." (- Jan Krans, ibid.)
 It is ironic that Erasmus, often hailed as the 'father' of the Textus Receptus turns out to be the ultimate source of the dubious generalization that "scribes frequently tended to mistakenly turn marginal notes into text", and that this maxim is actually based on a single late 12th century miniscule.

For Jan Krans' article, see here:
Jan Krans on GA-3


Tuesday, December 7, 2010

Bias against the PA

In his expose' of James White (a notable KJVonly Opponent), Jesse Boyd takes him to task for his shifty handling of the textual evidence regarding the PA (Jn 7:53-8:11):

  1. It is apparent that White takes his list of manuscript evidence straight out of the UBS4. However, he fails to include an asterix with mss. 1424. According to the UBS’ abbreviation system, the asterix refers to the original hand of the manuscript. Apparently, 1424 has the pericope in the margin. White does not want anyone to know that.
  2. White includes Codices L and Δ in his list of evidence, but he fails to mention that both mss. exhibit a blank space after John 7:52. It is clear that the scribes were aware of the fact that they were leaving something out.
  3. White argues that the pericope is absent from the majority of Lain versions. This is simply untrue. It is found in the Vulgate tradition and numerous mss of the Old Itala dating back to the second century. The UBS4 lists four italic mss hostile to the pericope, but seven plus the Vulgate that include it. White’s majority does not exist. Besides, there are thousands of Latin mss. that have not been examined by anyone, especially James White.
  4. White claims that the pericope is missing from the majority of Syriac mss. also. However, he fails to mention that it is found in the Palestinian Syriac and numerous later mss. of the Peshitta and the Harclean Syriac (even the UBS4 admits this). The pericope is also found in certain mss. of the Philoxenian Syriac. White’s assertion is misleading.
  5. White argues that the pericope is missing from the majority of lectionaries. Most assuredly, White has never seen an ancient lectionary in his life. John Burgon, on the other hand, personally handled over sixty of them. According to him, the passage was found rubricated in every one of them.. In other words, there were liturgical directions that instructed the reader to skip from 7:52 down to 8:12. This is hardly an omission of the pericope, for the pericope served no purpose on Pentecost Sunday when John 7:37-52; 8:12 was read. White also conveniently fails to mention that the Church selected nine out of the twelve disputed verses to be read on October 8. This reading, according to Burgon goes all the way back to the earliest of times in the Christian Church.
  6. White gives more credence to ten minuscules that place the pericope in different place in the Gospels than he does to the hundreds of them that retain the passage in the traditional place. His use of the exclamation point at the end of the sentence containing the aforementioned assumption indicates that he considers these witnesses to be great evidence against the passage. What White doesn’t promulgate is that the four mss. that place the pericope after Luke 21:38 come from a common archetype and are known for reworking the Scriptures (e.g. Luke 22:43 is removed and thrust into Matthew 26 between verses 39 and 40. White claims, “such moving about by a body of text is plain evidence of its later origin and the attempt on the part of the scribes to find a place where it ‘fits.’” This situation proves nothing of the sort save that a few irresponsible scribes wished to remove the passage from the Pentecostal Lesson because they believed it to treat the sin of adultery too leniently. Or, perhaps the mss. they were copying did not include the passage, but the scribe wished to include it where he had space (e.g. at the end of Luke or John).
  7. White says that some scribes included the pericope after John 7:44. It is apparent that White gets this information from Metzger who get it from Eberhard Nestle. In a footnote, Metzger promulgates that Nestle mentioned some Georgian mss. that placed the pericope after John 7:44. However, Nestle provided no specific mss. How do we know Nestle did not just make this up out of thin air.
  8. White says that John 7:52 and 8:12 go together. Are we to believe that “out of Galilee ariseth no prophet” is immediately followed by “Then spake Jesus again to them saying” without a word of explanation. As Edward Hills asserts, “Such impressionistic writing might possibly be looked for in some modern author. It is unthinkable in the simple narrative of John.” 68 Besides, what does White do with the fact that Jesus is nowhere in the vicinity in 7:45-52? Also, what is the purpose of the word “again” in 8:12? Without the pericope, there is no “again.”
  9. White says, “aside from issues of vocabulary and style.” What issues is he referring to? Of course, White gives absolutely no examples to back his claim. J.P. Lange, a scholar far more learned than White, says that the pericope exhibits the same “mystic twilight” that is common in John’s Gospel. 69
  10. White says, “The story of the woman taken in adultery interrupts the flow of the text.” How? The average reader does not see this.’
  11. White says that it is a “near certainty” that the pericope is not an authoritative part of the Bible. His “near certainty” is based upon one-sided evidence that is promoted with no documentation whatsoever.
  12. White claims that the story itself represents an oral tradition about Jesus that came to have its part in the Gospel of John over time. There is no concrete evidence for such a tradition whatsoever apart from the Gospel text itself. If all of this is true, why would a scribe insert the passage right in the middle of the passage preached from on the most important Sunday of the year--Pentecost?
  13. White does not provide a single sliver of documentation for the claims he makes on the pericope. Is the reader supposed to believe what he says simply because he says it. Apparently, White never consulted writers such as John Burgon, J.P. Lange, or Edward Hills.
  14. It is interesting to note the striking similarity between White’s words and those of Bruce Metzger in his Commentary on the Greek New Testament. White uses the term “earmark” as does Metzger. 70 This word is so uncommon that it is very doubtful it came from White’s vocabulary when Metzger was obviously sitting right in front of him.
  15. Concerning internal evidence surrounding the pericope, White writes, “aside from issues of vocabulary and style.” Metzger writes, “the style and vocabulary of the pericope differ noticeably . . .” 71 White writes, “interrupts the flow of the text.” Metzger writes, “it interrupts the sequence.” 72 Concerning the external evidence, White takes his list of witnesses straight out of Metzger. Also, he writes, “Both A and C most probably did not contain the passage, though both are defective in this section of John . . .” Metzger writes, “Codices A and C are defective in this part of John, but it is highly probable that neither contained the pericope.”73 What is going on here? White is taking information from Metzger, changing a few words, and offering no documentation. That is called “plagiarism” Mr. White, not scholarship.
 Conclusion: Fifteen counts of misrepresentation of evidence have been cited on one page of White’s book, The King James Only Controversy. This fact is scary. One can only wonder how much White bends the truth in the other 285 pages. This work is not scholarship. It seeks to trash the King James Bible through deception. Believe James White if you will, but I choose to believe God’s promise in Psalm 12:6-7. “The words of the LORD are pure words: as silver tried in a furnace of earth, purified seven times. Thou shalt keep them, O LORD, thou shalt preserve them from this generation for ever.”

68. Hills,89.
69. Lange,268.
70. see Metzger, 188.
71. Ibid.
72. Ibid.
73. Ibid.,187. 

Saturday, December 4, 2010

Estimates of Early MSS - Timothy J. Finney

Here is a quotation from

The Freer biblical manuscripts: fresh studies of an American treasure trove

 Editor Larry W. Hurtado (SBL, 2006),
Article: Manuscript Markup, Timothy J. Finney, p. 284:

"An important question in this respect is, What proportion of manuscripts do we have?  Taking Greek NT manuscripts of the Pauline Epistles as a case in point, about 15 remain from the time before Constantine.  Given that the population of the Roman Empire was about 50 million, it is possible to obtain a rough estimate of the number of copies that would have been produced during this era.  Assuming that Greek-speaking Christians constituted about 10 percent of the population by 300 C.E. and that there was one manuscript of the Pauline Epistles per 1000 Christians, there would have been about 5,000 of these manuscripts when Constantine became emperor.  We thus arrive at a survival rate in the region of 15 per 5,000, or 0.3 percent.  Even if this estimate is off by a factor of 10 in either direction, the fact remains that we have a very small remnant of the entire textual picture. (20)
20.  Those interested in a slightly more sophisticated treatment may wish to look at my manuscript-copying simulation program: Timothy J. Finney, "MSS: A Manuscript Copying Simulation" (2002).  Online: 

Real Relations between Text-types

click to enlarge, then backbutton...
Hort (1882) envisioned a line of descent and dependence as follows:

   (1)  Original Text (lost) = "Neutral text"

   (2)  binary split into two basic streams, Western & Alexandrian

   (3)  An early 'Lucian Recension' = a proto-Byzantine text

   (4)  further conflation of the two diverse streams into one mixed "Syrian text" (= Byzantine).

   (5)  further development of the circulating Byzantine text in language and style, and accruing of marginal insertions.

This linear model of descent has since been largely abandoned, because there is no real way to get to the Byzantine text by 'conflating' the Western and Alexandrian.

The real picture is far more complex, and more like the following:

The developing early Lectionary system, of dividing the NT into "pericopes" or Lections (Lessons) for public reading in church meetings, grew alongside the copying stream, and early copies of the text prepared for church use and public reading were modified on an ongoing basis.

Abberant texts like that of Vaticanus and Sinaiticus are just what they appear to be: Expensive and durable church copies made for public reading and worship services.   Copies like these served as a combination "Bible" and Lectionary for public reading, and have many idiosyncratic features caused by the back-influence of the Lectionary stream in development.

*thanks to Nazaroo for the graphics!


Sunday, November 28, 2010

P. Oxy. 1780: Jn 8:14-21

Recto: Jn 8:14-18a

Click to Enlarge
Click to Enlarge

Verso: Jn 8:18b-21

Click to Enlarge
Click to Enlarge


Preliminary Observations

(1) First the text is found to be John 8:14b-21. Since we have 50 consecutive half-lines spanning eight verses (and nothing else), the page is identified as belonging to a continuous-text papyrus with near-certainty.

(2) The columns can be seen to be left-justified, with very even placement of letters, and no spaces between words, except occasionally spaces are used to mark verses (16, 17 fall at a line-end, whereas 18, 19, falling mid-line, have a full space). The uneven right margin may be the result of copying in syllables; there doesn't seem to be a strong effort to avoid word-breaks, since there are six word-breaks in 25 lines (20% of the time).

(3) From knowledge of the text and layout we can determine average line-width (13-14 chars/line), and from extant line-ends, we can determine start of next line (and vice versa for visible line starts on recto side).

(4) A normal set of Noma Sacrae (Abbreviations of Sacred Names) seem to be used, with "Father", "Jesus", and even "Man" (ανων = ανθρωπων, line 22, side a): We know this because there isn't room for ανθρωπων in full here. Noun Cases are distinguished (πρα for Πατερα, line 09, side b, which helps reconstruct line 07).

(5) An initial difficulty in reconstructing the first 3 lines (v14) is resolved by recognizing a simple Word Order Reversal (W.O.R.) of two phrases, η μαρτυρια μου / αληθης εστιν. Since the change has no effect on the meaning of the verse, it is likely accidental. Internal evidence favours the reading here as opposed to the TR, the likely cause initially being a short homoioteleuton error, corrected by insertion afterward of the dropped phrase (since sense isn't affected, erasure wasn't needed):

Original:    περι εμαυτου η μαρτυρια μου αληθης εστιν (h.t.)

Dropped:   περι εμαυτου ......................... αληθης εστιν (next phrase)

Reinserted: περι εμαυτου αληθης εστιν η μαρτυρια μου (no erasing)

This process is a one-way process.  Had the TR order been original, the homoioteleuton error would have omitted both phrases, and no word-order reversal of these two could occur.

Longer homoioteleuton were minimized by switching to narrower columns (shorter spans). Shorter homoioteleuton are still possible regardless of line length, and were of concern. It is possible that the Word Order Reversal, once made, would be preferred since homoioteleuton is now less likely (with narrow columns). In this case, the editors of the Textus Receptus may have deliberately chosen the switched reading to avoid future errors, since the sense is not affected. From this example we can appreciate that even many W.O.R.s are really just partially corrected homoioteleuton.

(6) On line 10 of side a we see a short-form horizontal line in place of the (moveable)-N. This also seems to help fix the date to the 3rd century or later, as it was a scribal practice for ends of lines, often oblivious to any concept of adding N before vowels. This was a favourite technique of scribes who produced MSS like Codex B and Sinaiticus. So it may very well indicate a later document, i.e., 4th century.

(7) Turning to Side b we find three more cases of apparent homoioteleuton:

(a) Line 10, side b: There appears little room for the second "TA". This looks like a deliberate deletion caused by the scribe mistaking "TA TA" for a case of dittography. This may be a singular reading actually made by our copyist himself.

(b) Line 11, side b: The "O IHS" was clearly missing (no room in line). It is probable that the following "EN" was also clipped off, in a string of letters prone to just such a homoioteleuton blunder. Four whole letters are repeated: "HCENOIHCHCEN". This shows how easily the Noma Sacrae for "Jesus" could be knocked off the text. Since this is a well-known omission, it is not likely to have been created by our scribe, but was in his master-copy.

(c) Line 18, side b: Here it can be shown that our scribe is probably not responsible: The apparent dittography, OIC-OIC would more probably have arisen if the Noma Sacrae in use at the time was the Two Letter Form: IC for "Jesus". Our scribe seems to have simply copied the omission of Jesus' name from a previous MSS, since he uses a three letter form.

Although these cases appear in the reconstructed half of the column, they seem reasonably likely, because our copyist rarely extends a line more than one letter beyond lines nearby. Instead, he drops to the next line rather than cram, and gives each letter plenty of space for readability.

In summary, we find evidence of about 4 cases of homoioteleuton in 8 verses, with a rate of about 1 such error every other verse. About 3/4 of these however, seem to be errors copied from previous documents. This suggests that while copyists are more careful than they appear to be, errors of omission tended to accrue and build up at a rather alarming rate. This seems typical of 3rd century copying practice, and wasn't brought under control until new copying techniques and better proof-reading was enforced.


Sunday, November 21, 2010


Well, there is a difficulty here; photos are compressed so much and displayed so small, that they are hardly worth viewing, except as quick thumnails. This will make it difficult to discuss MS evidence and illustrate it with good pictures. We will have to limit discussions and/or provide external links to photos and charts etc.