Friday, January 5, 2018

Lectionary 5 in Matthew 24:20-26

            Today, let’s look at the text on one page of a medieval lectionary and see how well it compares to the same passage in Codex Vaticanus (the flagship manuscript of the Alexandrian text of the Gospels) and Codex Bezae (the flagship manuscript of the Western text of the Gospels).  The passage is Matthew 24:20-26, and the lectionary is Lectionary 5, also known as Barocci MS 202, at the Bodleian Library. It was written in uncial lettering in the early 1000’s.   (I have not received a response from the Bodleian’s permissions-department, so no image of the manuscript is posted here – but you can see the zoomable, full-color page with Matthew 24:20-26 – page-view 301 out of 316, marked as fol. 147 at the top of the page – at the Digital Bodleian website.) 
            In the following comparison, the Tyndale House edition of the Greek New Testament was used as the standard of comparison.  Differences in the format of sacred names, contractions for και, and differing forms of letters are not counted as textual differences.  The total number of differences between the THEGNT-text and each witness will be given, as well as the number of differences without minor vowel-exchanges (itacisms) in the picture. 


20 – χειμονος instead of χειμωνος (+1, -1)
21 – omits τοτε (-4)
21 – ουδε instead of ουδ’ ου (+1, -2)
22 – η instead of ει (+1, -2)
22 – εκολοβοθησαν instead of εκολοβωθησαν (+1, -1)
22 – κολοβοθησονται instead of κολοβωθησονται (+1, -1)
23 – ηπη instead of ειπη (+1, -2)
24 – δοσουσιν instead of δωσουσιν (+1, -1)
24 – omits μεγαλα (-6)
25 – προηρηκα instead of προειρηκα (+1, -2)
26 – ειποσιν instead of ειπωσιν (+1, -1)

21 – θλειψις instead of θλιψις (+1)
23 – πιστευετε instead of πιστευσητε (a corrector has superlinearly written η (so as to read πιστευητε) (+1, -2)
24 – ψευδοχρειστοι instead of ψευδοχριστοι (+1)

20 – προσευχεσθαι instead of προσευχεσθε (+2, -1) 
21 – θλειψις instead of θλιψις (+1)
21 – ουκ εγενετο instead of ου γεγονεν (+5, -5)
21 – does not have του before νυν (-3)
23 – υμειν instead of υμιν (+1)
23 – εκει instead of ωδε (+3, -2)
23 – πιστευσηται instead of πιστευσητε (+2, -1)
24 – ψευδοχρειστοι instead of ψευδοχριστοι (+1)
24 – πλανηθηναι instead of πλανησαι (+3, -1)
25 – υμειν instead of υμιν (+1)
26 – υμειν instead of υμιν (+1)
26 – εξελθηται instead of εξελθητε (+2, -1)
26 – πιστευσηται instead of πιστευσητε (+2, -1)

RP2005:  better than
Codex Vaticanus.
            This yields the following results:  Codex Vaticanus has only has five letters’ worth of corruption in this passage, and is one letter longer than the text in THEGNT.  Lectionary 5’s text contains nine non-original letters and is missing 23 original letters.  With itacisms removed from consideration, Lectionary 5’s text remains ten letters shorter than the text in Vaticanus.
            Codex Bezae’s text is the least accurate of the three:  although it is about twice as old as Lectionary 5, Codex D has 24 non-original letters and is missing 15 original letters, for a total of 39 letters’ worth of corruption.  (Lectionary 5, with 9 non-original letters and with 23 original letters omitted, has 32 letters’ worth of corruption.  Without itacisms in the picture, Lectionary 5 has 13 letters’ worth of corruption, and D has 22 letters’ worth of corruption.)  
This data may raise some questions:
            ● If scribes tended to add to the text, how is it that a manuscript from the 400’s (or 500’s) has 24 non-original letters here, and a Byzantine manuscript from c. 1000, only has 9 non-original letters, if scribes tended to add to the text?  Apparently the scribes in the ancestral transmission-line of Lectionary 5 never got the memo that stated that they were supposed to gain accretions. 
            ● The RP2005 Byzantine Textform agrees more closely in this passage with the THEGNT and the UBS/NA compilations than Codex Vaticanus and Codex Bezae do.  Even the Textus Receptus – the base-text of the King James Version, compiled in the 1500’s – agrees with THEGNT and NA27 more closely in this passage than the early manuscripts Vaticanus and Bezae do.  How is it that compilations based on late manuscripts, whether many or few, have the best text in this passage?  
            ● Considering that the text of Matthew 24:20-26 in Codex B in the 300’s is longer than the text of Matthew 24:20-26 in Lectionary 5, why do some textual critics (looking at  you, Dan Wallace) continue to teach that copyists – particularly Byzantine copyists – gradually expanded the text?  How many times and in how many ways does the opposite need to be demonstrated before scholars and commentators will concede that no preference should be generally assumed in favor of the shorter reading?


maurice a. robinson said...

The claims in this essay regarding Lectionary 5 are flawed on several levels:

1. Comparison of readings from a lectionary against editions based on continuous-text MSS is questionable, given that lectionary textual transmission moved quite separately from that of continuous-text MSS.

2. The single-page passage from within a much longer lection appears more as selective cherry-picking rather than a more complete analysis.

3. Minor orthographic peculiarities are neither here nor there (i.e. they really do not count) when dealing with the matter of longer/shorter readings. Eliminate these entries, and the supposedly “substantial” differences are reduced to only three.

4. Of the three meaningful differences (all “shorter” in terms of letter count), these primarily reflect typical scribal accident (omission of short words in the cases of τοτε and ου [then necessarily writing ουδ as ουδε], followed by homoioteleuton in regard to omission of μεγαλα). None of these would seem to have any intentional alteration in view (where the longer text would reflect intentional expansion). In fact, many similar errors can be found in most MSS at various points with no suggestion that such “shorter” text implies intentional expansion in other MSS

5. In sum, the present essay really says nothing of significance at this juncture beyond pointing out that the scribe of Lectionary 5 has certain orthographic habits and some inclination toward minor accidental omission of sort words.

James Snapp said...

Maurice A. Robinson,

(1) Granting that lectionaries, by design, skip some passages and add incipit-phrases, and other things, those things can be taken into consideration. Lectionaries are part of the evidence, and evidence should not be ignored merely because it is different from other evidence.

(2) Granted, this sample is small. Others are welcome to do more detailed investigations. At least readers here are aware of Lect 5's existence and its online presence. (The link is not yet at CSNTM.)

(3) I don't see the validity of the complaint about minor orthographic variants being noted and considered, especially since I also provided a comparison /without/ itacisms in the equation. With or without them, Lect 5 has the shorter text of Mt. 24:20-26, right?

(4) "None of these would seem to have any intentional alteration in view" -- but the point here isn't a matter of intuiting the motive of the scribe or the error-inducing mechanism; it is to simply notice -- after seeing this evidence -- that in this particular case, to assume in favor of the shorter reading would guarantee erroneous results.

(5) "The present essay really says nothing of significance at this juncture beyond pointing out that the scribe of Lectionary 5 has certain orthographic habits and some inclination toward minor accidental omission of sort words." -- Of course Lectionary 5 is not going to become a first-order witness; nor is a six-verse sample going to become a full collation. But this is a witness that readers should be aware of (which has been done where else??), and while the data-set involved in this essay is small, in confirms that no assumption if favor of shorter readings should be made -- a conclusion with which you agree, do you not?

maurice a. robinson said...

As I said in my "Case" essay, neither the shorter nor the longer reading should be preferred as a matter of course.

As for lectionary versus continuous-text issues, the legitimate comparison base for Lect 5 would be other lectionaries within the same lection reading (also, such really should consider the entire lection and not only a snippet therefrom).

"Evidence should not be ignored merely because it is different from other evidence" -- of course not, but only within its proper placement and within boundaries related to that particular type of evidence.

"Lect 5 has the shorter text of Mt. 24:20-26, right?" -- Sure, but as noted, such "shortness" in this case readily falls under the category of minor scribal error, and offers nothing in particular regarding the usual eclectic interpretation regarding the "shorter reading" principle.

"This is a witness that readers should be aware of (which has been done where else??)" -- Obviously at least in Scrivener's collations from the 19th century. But then if readers should be aware of Lect 5, so also they should be aware of Lects 6, 7, 8, etc. But are any or all of the lectionaries really going to be probative for establishing the original text of the Gospels or Epistles that otherwise is established from the continuous-text MSS (traveling in quite different transmissional streams)? Obviously not -- and I say this with a full appreciation for the lectionaries and what they are doing, even if not on the same level with continuous-text MSS.

Daniel Buck said...

You left out a couple of variant readings in Codex Bezae, so you are going to have to redo all the math.
D reads προσευχεσθαι δε ινα μη γενηται η φυγη υμω_ χειμωνοσ μηδε σαββατου in v. 20. σαββάτου is a unique reading (which it shares with Origen's Homily on Matthew in P. Bon. 1).
D reads ουκ εγενετο απ αρχησ κοσμου εωσ νυν ουδε μη γενοιτο in v. 21, with οὐ missing.