Optimal Substemmata Now Available for Acts

Just today I noticed that the CBGM for Acts has added substemmata to its repertoire. This is really exciting. When I produced the portion of the global stemma for the Catholic Letters in my thesis, Klaus Wachtel had to personally provide me substemmata using software on a very old Mac. And the data were not publicly available. The ability to do this online means that, using my faster approach explained here (pp. 167–168) and here (p. 102), one could probably put together a global stemma for Acts without too much trouble. I should say that I can’t find any documentation on the CBGM site for this new feature. If anyone has more details about it, please let me know.

A Previously Unidentified Folio of 093?

Earlier this week, I was working on Acts in the forthcoming textual commentary for the Tyndale House Greek New Testament, and I believe there is another page of 093 that the INTF does not currently recognise. Oddly enough, this is not the first time someone at Tyndale House working on Acts has stumbled across a previously-unknown page of a Greek New Testament manuscript.

According to the online Liste at the INTF, 093 is housed at the Cambridge University Library in the Cairo Genizah collection as “Taylor-Schechter Coll. 12.208”. That is slightly imprecise, but the print edition of the Liste (at least the first edition, which is the one I checked) correctly notes that 093 is two folios, each with its own shelfmark. The LDAB entry rightly gives them as Taylor-Schechter 12.208 and Taylor-Schechter 12.189. The manuscript itself is rather interesting. It is a palimpsest in which 6th-century Greek text (of Acts 24–25 and 1 Peter 2–3) was overwritten with Hebrew.

If you’ll hold that thought, we’ll hop over momentarily to the other side of the Atlantic.

As it turns out, the University of Pennsylvania also has a collection of manuscripts that came from the Cairo Genizah in their Center for Advanced Judaic Studies Library. Their online catalogue describes one manuscript, Halper 114, as an “Early midrash on Genesis 40:18-41:3; 46:28-47:1”. Note the following, however:

The hand may be identical to that of the ancient palimpsest copy of compendium of Midrash Rabbah Genesis, Cambridge T-S 12.208 and 189....

Another online catalogue at UPenn has a slightly longer set of notes (and I note that they do give a description for the undertext: “Palimpsest of Book of Acts(?) (Sixth century?)”) and includes the following:

Fragmented Greek characters, disposed vertically in 3 lines within center margins, along joint on hair side; no other Greek characters visible to naked eye, none visible on flesh side; some additional characters seem to appear with digital enhancement, but currently impossible to make out text; the identification with the Cambridge TS 12.208 and 189 bifolia is tentative, given ruling differences (Halper 30-32 lines; Cambridge 28-29 lines), and paucity of identifiable text, notwithstanding kinship between Hebrew texts.

I would suggest that it is not impossible to make out the Greek undertext, only very difficult. Still, I did some playing around, and I think I have found bits of Acts 21:13–14 in one of the more legible sections. As far as I know (and with my sincerest apologies to the INTF if this is not the case), the INTF does not currently recognise Halper 114 as part of 093 or as part of any other Greek New Testament manuscript.

[These images and the content of Center for Advanced Judaic Studies Library, Halper 114: Early midrash on Genesis 40:18-41:3; 46:28-47:1 are free of known copyright restrictions and in the public domain. See the Creative Commons Public Domain Mark page for usage details, http://creativecommons.org/publicdomain/mark/1.0/.]

Section of Halper 114, without my edits.

Section of Halper 114, with my very rough drawing on enough undertext for a preliminary identification

I can also see ε[ι]πο[ντες] after I stopped drawing in red there as well, but I admit I didn’t spend as much time on it as I could have.

My proposals:

1. It would be great if Halper 114 could be digitised with MSI so that the undertext can be read and edited (and I hear CSNTM has MSI capabilities now!).
2. It would be great if someone at the INTF could verify if Halper 114 is indeed a previously-unidentified (or at least previously-unrecognised in the Liste) folio of 093.

Finally, here we have another application of Head’s Rule, that “the best place to look for ancient manuscripts is in a library”.

The Text of Acts - Differences between Tyndale House Edition, ECM, and NA28

[warning: long post, you may want to skip to the conclusion at the end]

A little while ago I posted an analysis of all differencesbetween the NA28 and the Editio Critica Maior of Acts. I have now made a comparison between the Tyndale House Edition and the ECM, not knowing what I would find. Since the ECM and the Tyndale House Edition were prepared at the same time without knowledge of one another, it is also interesting to see how close they are to NA28.

I have kept things simple and will not comment on the 152 places where ECM-Acts gives a split guidance. Since the THGNT does not do split guidance (though it has diamonds in the apparatus), all these places can be counted as differences in a maximalist view. In each of the two and three-way splits the THGNT has one of the options of the ECM. This time I have also ignored brackets in the NA28 text, so as not to drown this overview with too many categories. I have noted brackets in NA28 and diamonds in the apparatus of the THGNT, though with no claim of being complete. I hope the tables are exhaustive, but improvements are always welcome.

1        THGNT and ECM Differ; NA28 Agrees with Neither

There is one place where each of the three versions has a different text:

Acts	THGNT	ECM	NA28
18:7	Omit  (with ♦ for Τίτου)	Τίτου	Τιτίου

This variant concerns the tricky business of the name of the man who owned the house next door to the synagogue. The ECM includes the singular omission by Alexandrinus of the preceding ονοματι in the variant unit, which, I think, distorts the way the external evidence is read. Together with D-05 this is the earliest evidence for just Justus, amply supported by the Byzantine tradition. The main transcriptional issue is whether the -τι  of ονοματι  was read as standing for the name ‘Titus’ or not (which in Latin would be abbreviated as ‘Ti.’).

2        THGNT and ECM Differ; NA28 Agrees with ECM

A much bigger group is formed by places where the THGNT (having started with Tregelles) disagrees with both ECM (which presumably started with NA28) and NA28. 

Some of the differences in the list are orthographic, yet they are ‘unusual’ orthographic variants—there are many other orthographic differences not listed here. I found the following 40 instances:

Acts	THGNT	ECM	NA28
1.11	βλέποντες	ἐμβλέποντες	[ἐμ]βλέποντες
2:31	ἅδου (with ♦)	ᾅδην	ᾅδην
4:9	σέσωσται	σέσωται	σέσωται
4:22	εγεγόνει	γεγόνει	γεγόνει
4:37	παρά	πρός	πρός
6:3	οὖν	δέ	δέ
7:11	εὕρισκον	ηὕρισκον	ηὕρισκον
7.13	Αὐτοῦ	Τοῦ Ἰωσήφ	[τοῦ] Ἰωσήφ
7:31	ἐθαύμασεν	ἐθαύμαζεν	ἐθαύμαζεν
7:43	Ῥεφάν	Ῥαιφάν	Ῥαιφάν
7:51	ταῖς καρδίαις ὑμῶν	καρδίαις	καρδίαις
8:18	τὸ ἅγιον (with ♦)	omit	omit
10:11	δεδεμένον καί	omit	omit
10:19b	ζητοῦσιν (with ♦)	ζητοῦντες	ζητοῦντες
11.23	τοῦ θεοῦ	τὴν τοῦ θεοῦ	[τὴν] τοῦ θεοῦ
12:6	προσαγαγεῖν	προαγαγεῖν	προαγαγεῖν
13:6	βαρϊησοῦς	βαριησοῦ	βαριησοῦ
13.14	ἐλθόντες	εἰσελθόντες	[εἰσ]ελθόντες
13.31	Εἰσιν (with ♦)	Εἰσιν νῦν	Εἰσιν [νῦν]
14:8	ἐν Λύστροις ἀδύνατος	ἀδύνατος ἐν Λύστροις	ἀδύνατος ἐν Λύστροις
15:4	Ἱεροσόλυμα	Ἰερουσαλήμ	Ἰερουσαλήμ
15:25	ἐκλεξαμένους (with ♦)	ἐκλεξαμένοις	ἐκλεξαμένοις
16.28	Παῦλος	ὁ Παῦλος	[ὁ] Παῦλος
16:40	ἐκ (with ♦)	ἀπό	ἀπό
17.22	Παῦλος (with ♦)	Παῦλος	[ὁ] Παῦλος
20:4	ἄχρι τῆς Ἀσίας (with ♦)	Omit	Omit
20:28	κυρίου (with ♦)	θεοῦ	θεοῦ
20:30	ἑαυτῶν	αὐτῶν	αὐτῶν
21:5	ἐξαρτίσαι ἡμᾶς	ἡμᾶς ἐξαρτίσαι	ἡμᾶς ἐξαρτίσαι
21:6	ἐνέβημεν	ἀνέβημεν	ἀνέβημεν
22:8	πρὸς ἐμέ	πρός με	πρός με
22:13	πρὸς ἐμέ	πρός με	πρός με
22:26	ἑκατόνταρχος	ἑκατοντάρχης	ἑκατοντάρχης
23:20	μέλλων	μέλλον	μέλλον
23:22	πρὸς ἐμέ	πρός με	πρός με
25:10	ἠδίκηκα (with ♦)	ἠδίκησα	ἠδίκησα
26:1	ὑπέρ	περί	περί
26:29	κἀγώ	καὶ ἐγώ	καὶ ἐγώ
27:8	Λασέα	Λασαία	Λασαία
27:16	Κλαῦδα	Καῦδα	Καῦδα

A few remarks. Contrary to ECM, I do not think 15:4 is an orthographic variant. There is quite some literature on the differences between Ιερουσαλημ and Ιεροσολυμα and variation between these two is surprisingly rare.

Perhaps one of the bigger differences is 20:28. Hurtado has a study (Texts and Artefacts, chapter 4) showing that this is one of a string of variants between κυριος and θεος in Acts. It is only here that the THGNT and the ECM have a difference of opinion.

The only variant where a whole phrase is at stake is at 20:4. Normally I would be inclined to go with the earliest evidence and omit ἄχρι τῆς Ἀσίας. However, there is a narrative difficulty that puts pressure on this phrase, and that is that in 20:16 we learn that Paul had no intention of spending time in Asia. At this point in 20:4, the journey has not reached Troas yet. That makes the insertion of this phrase difficult to explain. In addition, with the phrase there are two references to Asia if we count the noun Ἀσιανοί further down. It is a tight call though.

3        THGNT and ECM Differ; NA28 Agrees with THGNT

The group of readings where ECM has changed NA28 but has not been followed by THGNT consists of the following 25 readings.

Acts	THGNT	ECM	NA28
1:10	ἐσθήσεσι λευκαῖς	ἐσθήτι λευκῃ	ἐσθήσεσιν λευκαῖς
1:26	αὐτοῖς	αὐτῶν	αὐτοῖς
2:3	καὶ ἐκάθισεν (with ♦)	ἐκάθισέν τε	καὶ ἐκάθισεν
2:20a	-	ἥ before ἐλθεῖν	-
2:20b	ἡμέραν	τὴν ἡμέραν	ἡμέραν
4:4	ὡς	ὡσεί	[ὡς]
5:26	ἦγεν	ἤγαγεν	ἦγεν
5:33	ἐβούλοντο	ἐβουλεύοντο	ἐβούλοντο
7:25	ἀδελφοὺς αὐτοῦ (with ♦)	ἀδελφούς	ἀδελφοὺς [αὐτοῦ]
8:31	ὁδηγήσει	ὁδηγήση	ὁδηγήσει
9:8	οὐδέν	οὐδένα	οὐδέν
9:12a	ἐν ὁράματι	omit	[ἐν ὁράματι]
10:9	ἐκείνων	αὐτῶν	ἐκείνων
13:11b	ἔπεσεν	ἐπέπεσεν	ἔπεσεν
13:33	αὐτῶν	omit	[αὐτῶν]
14:10	Φωνῇ	τῇ φωνῇ	Φωνῇ
15:37	τόν Ἰωάννην	Ἰωάννην	τόν Ἰωάννην
16:13	ἐνομιζομεν προσευχήν	ἐνομιζετο προσευχή	ἐνομιζομεν προσευχήν
16:17	κατακολουθοῦσα	κατακολουθήσασα	κατακολουθοῦσα
19:14	τινος	τινες	τινος
20:21	Ἰησοῦν	Ἰησοῦν Χριστόν	Ἰησοῦν
23:5	ὅτι	omit	ὅτι
23:10	γινομένης	γενομένης	γινομένης
23.23	Τινὰς δύο	Δύο τινὰς	[τινὰς] δύο
28:5	ἀποτινάξας	ἀποτιναξάμενος	ἀποτινάξας

An interesting choice was made by ECM at 9:8 to prefer οὐδένα over οὐδέν. To me, οὐδένα before ἔβλεπεν shows clear influence of μηδένα θεωροῦντες in the previous verse, so that this is just a simple harmonisation to the immediate context.

Perhaps the most ‘controversial’ reading in this table is 13:33, where ECM has a ‘conjecture’ (according to the Commentary). Frankly, I am not sure if this is a true conjecture. It is only so if we accept the variation unit as given in ECM. However it may well be that the question of τέκνοις αὐτῶν and the presence/absence of ὑμῖν should be dealt with separately.

4        THGNT and NA28 Differ; THGNT and ECM Agree

There is one final list to produce so that we can make a three-way comparison between THGNT, ECM, and NA28, and that is the list where THGNT and ECM agree in their deviation from NA28. I have found 27 of these.

Acts	THGNT	ECM	NA28
1:15	ὡς	ὡς	ὡσεί
2:5	ἐν	ἐν	εἰς
2:33	-	-	[καί]
3:13a	-	-	[ὁ θεός]2
3:13b	-	-	[ὁ θεός]3
5:31	-	-	[τοῦ]
7:7	δουλεύσωσιν	δουλεύσωσιν	δουλεύσουσιν
7:22	-	-	[ἐν]
9:12b	-	-	[τάς]
9:21	ἐν	ἐν	εἰς
10:40	-	-	[ἐν]
11:22	-	-	[διελθεῖν]
12:11	-	-	[ὁ]
14:3	-	-	[ἐπί]
15:4	ὑπό	ὐπό	ἀπό
15:17	ὁ	ὁ	-
15:41	-	-	[τήν]
16:11	οὖν	οὖν	δέ
16:12[1]	πρώτη τῆς μερίδος	πρώτη τῆς μερίδος	πρώτη[ς] μερίδος τῆς
16:27	-	-	[τήν]
16:28	φωνῇ μεγάλῃ	φωνῇ μεγάλῇ	μεγάλῃ φωνῇ
19:15	-	-	[μέν]
20:5	προσελθόντες	προσελθόντες	προελθόντες
20:6	οὗ	οὗ	ὅπου
25:18	πονηράν	πονηράν	πονηρῶν
27:8	ἧν πόλις	ἧν πόλις	πόλις ἧν
27:23	-	-	[ἐγώ]

In the list of differences between ECM and NA28 the following is given as one variant, though I have split it up into a word order variant and the issue of the article before Paul, the latter is included above under section 2.

Acts	NA28	ECM	THGNT
16:28	μεγάλῃ φωνῇ [ὁ]	φωνῇ μεγάλῇ ὁ	φωνῇ μεγάλῃ

Interestingly, in over half of the differences between ECM and NA28 (27 out of 52) the THGNT agrees with the ECM.

5        Conclusion

[drum roll]

Time for one last table, the number of differences between THGNT, ECM, and NA28.

By ignoring all the brackets in NA28 and all the split readings in ECM, I have followed the same procedure as the list given in ECM-Acts I, 34*–35* where the differences between ECM-Acts and NA28 are found. That list gave 52 differences, besides the 152 split guidance readings. ECM-Acts and THGNT differ 76 times (section 1 + 2 + 3). And the total number of differences between THGNT and NA28 is 68 (section 1 + 2 + 4). There are orthographica included in the list, but the THGNT has many further differences as to iota subscript, -ει- for -ι-, and unassimilated prepositions that are not part of the comparison.

Number of differences	THGNT	ECM	NA28
THGNT		76	68
ECM	76		52
NA28	68	52

In terms of the percentages as used in the CBGM, the differences are fairly minimal. The text of Vaticanus, the witness most closely aligned with the ECM text, is recorded as agreeing in 6997 out of 7250 variant units, 96.51%. If we assume that 7250 is about the maximum number for any complete witness, then THGNT agrees with ECM in 7174 out of 7250, which is 98.95%.

Looking over the various lists quickly, it is surprising how few ‘big’ differences there are between the three editions. This is unlike some variants in the gospels. The tables above show that (unless one has a preference for the Textus Receptus, Codex Bezae, or the Byzantine text) there isn’t that much at stake in Acts.

---

[1] The reading of NA28 is given incorrectly in ECM, which omitted the final τῆς of NA28.