About Sappho 16, ll. 6-9

Question

There are apparently numerous versions of these lines. I only have access to two¹ however: Edmonds':

[…]· ἀ γὰρ πόλυ περσκόπεισα
κάλλος ἀνθρώπων Ἐλένα τὸν ἄνδρα
[κρίννε κάλ]ιστον

[ὸς τὸ πὰν] σέβας Τροΐας ὄλεσσε

and Campbell's:

[…]· ἀ γὰρ πόλυ περσκέ̣θ̣ο̣ı̣σα
κ̣άλ̣λο̣ς [ἀνθ]ρώπων Ἐλένα [τὸ]ν ἄνδρα
τ̣ὸν̣ [πανάρ]ι̣στον

κ̣αλλ[ίποι]σ̣' ἔβα 'σ Τροΐαν πλέοı̣[σα

Now that is a big difference, even outside lacunas. So I tried looking at Grenfell and Hunt's edition of P.Oxy. 1231, whence the poem mostly comes (see below for more), and found:

[…]ᾱγαρπόλυπερσκο̣π̣ε̣ı̣[.]α
[...]^λ̣ο̣σ̣[...]ωπωνελενα̣[..]νανδρα
[9 letters]ι̣στον

[7 letters]σ̣εβαστροϊα[..]λεσσ̣[..]

That seems to scream "Go Edmonds, screw Campbell". But let's look at the notes to the transcription. What I gather from them is:

• σκεδοισα and σκοπεισα are equally likely endings for l. 6, and since the notes only mention a «rather high stroke turning over» for the delta, I assume σκεθοισα would also be equally likely; so as far as P.Oxy. 1231 fr. 1 is concerned, Cambpell and Edmonds are equally likely right about that line;
• The other lines have no beginning in the papyrus at hand, so those must come from another papyrus, which we will discuss below;
• The notes don't even mention l. 9 (which is l. 21 in the papyrus), so this papyrus actually has a no-doubt λεσ in that line, whereas Campbell would have that as λεο, and read the doubtful sigma right afterwards as a doubtful iota;
• Neither the notes nor the transcriptions show any other P.Oxy. 1231 fragment that could contain beginnings for those lines.

So when Bonnaria states that P.Oxy. 1231 reads:

τ̣ὸν̣ [        ]ι̣στον

καλλ[        ]σ̣' ἔβα 'ς Τροΐαν πλέοı̣[

they are apparetly giving us a nice bit of bull… err, tauricopria. Or maybe they are mixing up papyri. There are two more papyri this poem comes from. One is PSI 123, 1-2, but that only gives the last line, τ' ἐξ ἀδοκή[τω], which gives no trouble. The second one is P.Oxy. 2166(a), published in volume XXI. Unfortunately, I can't find anything online about this: no images, no transcriptions. Campbell's critical notation suggests it must have read at least:

κ̣αλ̣λο̣σ[]
τ̣ὸν̣[]

κ̣αλλ[    ]σ̣εβαστροϊανπλεοι̣[

plus the first 5 letters of the following line and the first 2 of the one after that. This implies a papyrus conflict: l. 9 is different from one papyrus to the other, and doubtfulness cannot solve the problem. So we are faced with two options:

• Either we argue that the two are different poems, which seems viable if 2166(a) only has the little mentioned above (which could actually be even less, since we don't actually need it to read the σ̣εβαστροϊα part;
• Or we have to choose one or the other, and with what arguments do we point our finger at one and say "You're wrong!"?

So my question is: does anyone have access to anything that can shed light on what was actually on 2166(a), and/or to something pointing out extra margins of doubtfulness in 1231 to settle this dispute?

¹I also have a couple slightly different versions (which basically only differ in the lacuna at v. 8), but I kept the above two as prototypical, and also because all others have no critical notation. Researching again only yielded The Complete Poems of Sappho, giving Cox's version as reading κρίννεν ἄριστον. Maybe safopoemas.doc contained other versions. But that's beside the point anyways, so I won't look into it.

PS If you're thinking of suggesting I buy the Voigt and/or Lobel-Page edition of Sappho, don't. I'm already thinking of getting them for Xmas.
UPDATE: The dream is dead, but I may manage to borrow Voigt from a library if I find a way to get it from Brera library to one closer to the places I usually frequent.
UPDATE 2: The Voigt dream is dead, but I did buy Lobel-Page, and it's soon gonna be brought to me from England. Lobel is the editor of P.Oxy. XXI, so having an edition of Sappho coauthored by him seems like a good chance of finding answers. Also, Mr. Weimer in the comments offered to check if his library has that volume and look at the pages that may be relevant to this poem and the Gongyla one, that is:

• P. Oxy. XXI (1951) add. p. 125, 11 (and perhaps also P.Oxy. 2292-2293) for the Gongyla poem;
• P. Oxy. vol. 21, p. 122 [additional fragment of P.Oxy. 1231] for the poem at hand.

That would be great. And I found that on Amazon, but the cart had already been purchased and this book costs and needs shipping which would be free if it were in that cart, so that will have to wait at least till next Xmas.

Update

According to this, 2166(a) is not a source for this poem¹, whereas something in P.Oxy. vol. XXI is. I have also managed to find that volume on Amazon, so I may buy it.

¹Unless of course 2166(a) is in that volume, which should start at 2292, meaning the numbering doesn't line up with the volumes.

Update 2

Looking at Campbell again, I saw that P.Oxy. 2166(a) is said to be in Ox. Pap. XXI, so yeah, for some reason, the volume starts with 2292 and contains 2166(a). As if Oxyrhynchus papyri numbering weren't already weird enough…

Answer 1

see Obbink 2016 for further details.

Here’s Voigt 1971 conjecture:

Testimonia and critical apparatus in Voigt 1971:

Relevant pages from P.Oxy. XXI:

Answer 2

Not solving the problem entirely, but ahem-b-ok-ahem furnished me with an image of P.GC. inv 105, where on fr. 2 one can clearly see the endings:

ΟΗϹΑΙ
€ΘΟΙϹΑ
ΟΝΑΝ[]

ϹΑ
ΟΚΗωΝ
Α̣Γ̣ΑΥΤΑΝ

Ν̣ΟΗΜΜΑ
Ν̣Ο̣ΗϹΗ
ΝΕΜΝΑΙ

ΒΑΜΑ
ΝΠΡΟϹωΠω
ΟΙϹΙ

€Ν€̣ϹΘΑΙ
Χ̣ΗΝΔΑΡΑϹΘΑΙ
[....]ΑΙ

The final lacuna is actually not a hole, but no letter can be fully seen before the AI at the end. Vestiges are visible, though:

• The first vestige strongly suggests a M;
• Then we have what looks like an angle with vertex pointing upward (aka a caret or circumflex ^, but I felt like talking complicated :) ), perhaps the top of an A;
• Then there is some kind of "pit", perhaps the top of an Y; the bottom of the pit is exactly on the edge of the papyrus;
• Finally, there is a space which seems blank, though looking closely one can see what is either very faint ink, or a fold in the papyrus.

With that, the ending of that line might be ΜΑΥΤΑ, which suggests comparison with a previously known fragment, a quote by Ap.Dysc., reading ἔγω δ' ἐμαύτᾳ / τοῦτο σύνοιδα.

This certainly settles all our problems, modulo the line beginnings, which means the problem of l. 8 stays, but the rest is settled by these endings. However, this papyrus was published in 2014, well after Campbell published, so the problem of why Campbell had these readings remains, and new problems with older restorations are caused by the new endings in stanza 4 and stanza 6 (last one). But that is another story.

Update

As AlexB's answer made me notice, the ahem-b-ok-ahem finding mentioned above tells me that this mysterious extra source preserves only a grave accent in l. 12 (aka last line of stanza 3). Interesting info, but doesn't settle the problem of this question.