Matthew Davis (University of Houston, class of 2011)
The E3 (= West E, Escorialensis Y.I.1) has traditionally gotten less attention than its closest relative, the Venetus B (Marcianus Graecus Z. 453). As Dr. Dué mentioned in her preliminary notes, both texts were produced in Constantinople before making their way to Venice in the 15th C, where the E3 was eventually purchased for Phillip II in 1572. Allen dates both texts as 11th century, with the VB being no later than 1050, and sees them as derivatives of one branch of the manuscript family, and the more distantly related C (Laurentianus 32.3) of another. Still because of their nearly identical layout, the E3 has sometimes been considered to be a twin of the Venetus B, and in at least one case was even attributed to the same scribal hand. Fortunately, digital publication makes direct comparison of the manuscripts possible for the first time by anyone with internet access and an interest in doing so. Having been given early access to the images of the E3, undergraduates such as myself are doing just that. As it turns out, differences between the texts are numerous, and often too subtle to be fully expressed in an apparatus criticus.
A particularly strange example can be found at the end of folio 97 recto in the E3:
(Folio 97r of E3)
Line 413 above (ἄψορρον δ’ Ἰδαῖος ἔβη προτὶ Ἴλιον ἱρήν) ) is the last line of the corresponding folio in the Venetus B, 101r, and the reverse sides of both of these folios (E3-97v and VB-101v) begin with line 414. The last five lines shown here, those marked with an antisigma (used to mark passages where the order has been disturbed) are actually lines 7.430-7.434, which appear again in their proper place near the bottom of E3-97v. Typically, folios in these manuscripts have 24 lines each, and E3-97r is no exception; VB-101r is missing these duplicated lines and stops short at 19 lines, using the extra space instead for a vocabulary note. This note is in the later hand that characterizes a second set of scholia that is not present in E3. Curiously, the ink of this presumably later note resembles more closely the older set of scholia which both manuscripts share (shown below).
(Folio 101r of the Venetus B) This leaves us with two possible scenarios: they are either both ultimately derived from a text in which these lines were also duplicated (with the scribe of VB choosing to omit them) or the Venetus B was copied from the E3 or one of its derivatives. In either case, they are definitely not ʻtwins,ʼ and the E3 is definitely not a derivative of the VB.
Still, there is more evidence to consider. While in nearly every case the E3 follows the vertical differences of the VB when compared to the Venetus A (charted here), there is a plus verse in the E3, line 8.19 (E3-98r and VB-103r). If the scribe of E3 had been pulling from another family of manuscripts we should expect to find additional verses or scholia such as this, but it is far more likely for a single verse to be lost in transmission instead, especially considering these two manuscripts are overwhelmingly similar in their verse order. What makes this line even more interesting is that these two folios have the same number of lines, 21; it is only on the verso that the plus verse is accounted for in the layout. Line 8.22, the last verse on VB-103r, is the first on E3-98v, bringing the total number of lines on E3-98v to 25. For both of these changes to occur during the same copying, the scribe of the VB would have to have skipped line 8.19 accidentally, then also pulled the first line from the verso for the line count to match. (This presumes that the line was not struck intentionally. The line reads well with the one before it and is referenced as early as Plato, so I see no reason to suspect that its authenticity was in question.) This scenario seems even less likely than the previous one, where a scribe may have pulled the line from another text. The most reasonable explanation then, is that these changes occurred separately: in one branch (E3) the number of lines per page was altered when line 8.22 was moved onto the verso, and in another (VB), which must have still retained the original layout with 24 lines on the verso, a line was accidentally skipped on the recto.
These two observations support Allen then, though I suspect the split between the texts occurred a generation or more before he suggests. My own comparison was limited to books 7 and 8 (based on the chart linked above) but more systematic comparisons are surely on their way, and as more differences in layout and content come to light, further reconstruction of the manuscript family will surely follow.