2d pers. pres./fut. indicative medio-pass.: ει or ηι?

Greek Learning Greek
1

-ΗΙ is the older spelling in the Ionic alphabet. (However, -EI was the spelling in the old Attic alphabet, before the adoption of the Ionic alphabet. That occurred officially in Athens in 403 but I believe came into general use in Athens somewhat earlier.) -ΕΙ becomes prevalent in Attic in the fourth century BCE, but was not carried over into the koine, by which time ει, η and ηι were in the process of merging or had merged in pronunciation. See Smyth 628.

In his introduction to the new OCT of Plato’s Republic, Slings complains that, with a few exceptions (βουλει οψει), he was “forced” (coactus) to follow the practice of vol. 1 of the new OCT edition of Plato’s complete works (on which, see the BMCR review previously linked to by mwh). He notes that if you follow this practice, you should also write ποληι and βασιληι, and not πολει and βασιλει, which no one does.

He would also prefer to print iota adscripts, but again, he was forced to print iota subscripts.

The new OCT Plato:

http://bmcr.brynmawr.edu/1997/97.01.07.html

2

Allen has the following chart on the development of Attic ηι to ει from epigraphic evidence. He suggests that the pronunciation merger happened first, and was complete early in the 4th century.

However, what is really useful is his footnote where he mentions A. S. Henry’s Epigraphica article in CQ:

https://www.jstor.org/stable/637728?seq=1#page_scan_tab_contents

From Henry, orthography obviously took some time to catch up, with some forms keeping ηι longer than others. At the end of the 4th century, at least, when Henry’s data starts, ει seems to have a slight preponderance over ηι in inscriptions, which moves fairly linearly to a major preponderance by the beginning of the 2nd century. Extrapolating the trend backwards to Plato’s time, it seems that ει and ηι would have been a tossup in written communication.

You could also object to τῇ. The restoration of ῃ in most of these forms was a late 2nd-century thing.

Plato may likely have pronounced the diphthong as ει, but still wrote it as ῃ much of the time.

3

According to Slings, the consensus of the mss. of the Republic make us certain that Plato wrote -ει in the Republic.

Henry’s article starts in the late 4th century – after Plato – and doesn’t address the 2nd sing. indicative pres./fut. medio-passive. This is probably because second-person singular forms rarely if ever appear in the inscriptions on which his data are based.

4

In re orthographica volumen novae editionis Platonicae primum sequi coactus, iota post vocales ᾱ η ω subscripsi in textu, in apparatu autem notavi iota vel adscriptum vel omissum prout aut unus e tribus libris ADF adscripsit aut omnes omiserunt. eandem rationem secutus semper formas secundae personae singularis indicativi medii thematicorum, ut dicunt, temporum -ῃ non -ει praebui (paucis formis sicut βούλει ὄψει exceptis), quamquam consensus librorum ADF nos certos facit Platonem in Republica -ει scripsisse. hoc quidem non possum non monere, quicumque Platoni formas quales sunt λύῃ pro λύει adscribat, ei et πόλῃ βασιλῇ non πόλει βασιλεῖ scribendum esse, id quod nemo scribere ausus sit.

I thought that ADF all derive from Thrasyllus’ 1st century A.D. edition, which split things up into tetralogies.

EDIT: cross-posted

Henry’s article starts in the late 4th century – after Plato – and doesn’t address the 2nd sing. indicative pres./fut. medio-passive. This is probably because second-person singular forms rarely if ever appear in the inscriptions on which his data are based.

Yes. Only 3rd-person subjunctive, and yes because second-person singular forms are too rare. But what seems to have happened is a language-wide change in the pronunciation of ηι, not anything limited to one form.

5

ADF are said to derive from a common archetype, but pinning the archetype and the division of Plato’s corpus into tetralogies on Thrasyllus alone is highly speculative. Plato’s text was worked over by a number of scholars in the Hellenistic and Roman periods.

https://books.google.com/books?id=Eewd2gEKo8kC&printsec=frontcover&source=gbs_ge_summary_r&cad=0#v=onepage&q&f=false

mwh probably has something interesting to say about this.

6

On AD and F there are some remarks in
https://www.jstor.org/stable/4432062
commenting on the book you link to.
(The pun on Slings’ name in the latter part was unintentionally cruel, for his name apparently has some unseemly connotation in Dutch, and the pun gave malicious glee to his enemies. He was a fine scholar in the grand Dutch tradition, and I had great respect for him.)

7

https://forvo.com/word/slinks/ Unfortunate, but it does make the pun wickedly funny.

From the article:

And his postulated “bifurcation of the D and F traditions” already in the second century (p. 78) and other such inferences, even the conventional assumption that ADF have a common ancestor (p. 65 and passim), are founded on a more rigidly stemmatic (or “vertical”) model of the ancient tradition than the ancient texts exemplify. The process of transmission evidenced by the papyri gives grounds for querying the presupposition that common errors all derive from one and the same source.

It seems to indicate that the theory of Thrasyllus’ version as common ancestor is less than solid. If there were ancient collations of readings, original ηι might have survived, and that it didn’t perhaps tells us something.

I wouldn’t mind some printed versions of different texts that got me used to common manuscript variations that get edited away nowadays. I feel like it would make me a better reader. There are manuscript facsimiles, and I look at them, but handwriting is still much slower for me to read than print.

8

The Dutch word slinks means ‘deceitful’, ‘furtive’, ‘sly’.

9

mwh: Thank you for the links you provided to thought-provoking articles.

10

Thrasyl(l)us does seem to have had something to do with the transmission, but it’s far from clear what. Our main source both for Thrasyl(l)us and for the organization of the Platonic corpus is Diogenes Laertius (3.56ff.), where Thrasylus is said to have said that Plato published his dialogues in tetralogies, on the analogy of tragedians’ tetralogies. That’s obviously absurd. Originally the dialogues were put out individually, and that’s how they circulated for centuries, in roll form. It’s quite unclear how the tetralogic arrangement got established.

Thrasyllus is credited, rightly or wrongly, with the double titling of the dialogues—“Phaedo or On the Soul” for example. Socratic dialogue being effectively a new genre, dialogues could be labeled as if they were plays (Phaedo, Crito, etc, cf. Theseus, Medea, etc) or as if they were treatises (e.g. περι δικαιου, the alternative title of the Republic). I don’t know how Plato titled them, if at all.

They were variously classified (by neo-platonists?) as “ethical,” “logical,” “protreptic,” etc, the Republic being “political.” The organization of the Demosthenic corpus is comparable, and that controlled the transmission (but not the organization within any one group, curiously). But the grouping of the Platonic dialogues into sequenced quartets, which eventually prevailed, follows different principles (some more discernible than others), and its origin is obscure. It’s quite artificial (and requires the number of genuine dialogues to be a multiple of four). It can’t be very early.

Just as the individual dialogues vary enormously in size (from just a few columns to thousands, as if to defy book-making conventions), so do the tetralogies (the eighth comprises Clitopho, Republic, Timaeus, and Critias!), so it was not at all a practical arrangement, or not until the codex became sufficiently capacious in later antiquity, and hardly even then. It’s more of a theoretical organization.

There were other orderings too. Again Diog.Laert. is our source, and there’s no agreement on what his sources were. Apparently there was one arrangement in which the dialogues were sorted into trilogies (headed by Rep. Tim. Crit.). This one is associated (like so much else, including colometrization of lyrics) with Aristophanes of Byzantium, but Alexandrian organization of the corpus is unlikely to have been like that. DL records various alternative sequences. Philosopher-teachers’ opinions on the order in which the dialogues should be read were very varied, and this may have affected their sequence in some codex texts. There’s no evidence.

Thrasyllus does seem to have had a significant role in the transmission (reportedly he organized the Letters too), but his work will have been one of systemization. He won’t have exercised any influence on the text. There’s been much written about it but it’s all up in the air really.

As to –ηι vs. –ει, I neither know nor care. Ancient manuscripts would be expected to modernize orthography to some extent unless there was a strong tradition against it, but within any given tradition it’s rare to find consistency. I certainly wouldn’t use them to establish Plato’s own practice. In non-Attic authors the given orthography is often an imposition by ancient scholars and cannot be original, but that would hardly apply here, unless there was puristic insistence on the older spelling.

11

Thesleff says that the Corpus Platonicum was established probably during the tenure of Xenocrates, i.e. late 4th century. The Corpus would have been kept in the Academic library, and was used in producing copies for the whole of the Greek-speaking world. The Academic authority was immense, and constant comparison to this standard edition was made. This is apparently seen in how uniform the tradition (to a great extent) is. A sign of this levelling, though, is in the tradition which says that the Theaetetus had a variant preface (and in the pseudepigraphy like the Halcyo).

The order of the texts in the manuscripts may originate in the Xenocratic edition. Thesleff thinks that the trilogical grouping which mwh mentions is from the early Hellenistic period, possibly even the Old Academy, and the tetralogic grouping is based on that. What exactly Thrasyllus did in his Tiberian edition I don’t know, but mwh sketches it out and speculates on it quite well in his post.

12

Here there is need to keep track of the μεταγραμματισμός ; that was when they started to use eta to write long e, and omega to write long o, as far as I remember.