I've often heard it said that Aeolic Greek used -μι endings on contract verbs, like φιλημι in Sappho (for Attic φιλέω/φιλῶ).

However, I can't seem to find this supposed "φιλημι" anywhere. It doesn't appear in Loeb's collection of Sappho fragments, for example (or anywhere in Loeb's corpus at all), and Google hasn't turned up anything helpful.

Is this actually attested? If so, where?

  • (For example, Smyth §656 specifically mentions φιλημι, φιλημεν, and various other forms, though he doesn't specifically attribute any of them to Sappho. Colvin's dialectal reader mentions the general rule in Lesbian but doesn't cite any examples.)
    – Draconis
    Commented Sep 11, 2020 at 0:02
  • I think Thumb 1959 Handbuch der griechischen dialekte could be more useful; e.g. the so called verba vocalia in Aeolic ("der lesbische Dialekt") are discussed on pp. 102-103 (section 16).
    – Alex B.
    Commented Sep 11, 2020 at 19:10

1 Answer 1


A fragment quoted by Athenaeus preserves an elided form of φίλημι:

ἔγω δὲ φίλημ᾽ ἀβροσύναν, καὶ μοι τὸ λάμπρον
ἔρος ἀελίω καὶ τὸ κάλον λέλογχεν.

These lines also appear, in a slightly different version, in P.Oxy. 1787 fr. 1, a papyrus that contained, among other things, line endings from the "Tithonus poem". This is fragment 58 in the Loeb edition of Sappho and Alcaeus, translated there as:

"but I love delicacy ... love has obtained for me the brightness and beauty of the sun"

This poem was substantially filled out by a different papyrus published in 2004 (P.Köln inv. 21351 + 21376). However, the lines do not appear in the 2004 papyrus, so there is debate on whether or not they were part of the Tithonus poem.

  • Voigt 1971 also suggests fort. φί[λημμ(ι) for S. 65,5 and a question: why φίλημ' and not φίλημμ' (from φίλημμι)? I thought the first compensatory lengthening and consequent gemination happened in (North) Aeolic, so a geminate resonant is to be expected, cf. Miller 2014 writes "the outcome was -VRR- at least in Thessalian and Lesbian" (pp. 219-220, §19.1)
    – Alex B.
    Commented Sep 11, 2020 at 20:22
  • @AlexB. It looks like the papyrus does give φίλημμ’ (see the link).
    – TKR
    Commented Sep 11, 2020 at 20:36
  • @AlexB. But actually, why would this form be subject to the 1CL / Aeolic gemination? AFAIK that only involved s/y + resonant, while this would presumably be from *-eh₁-mi (that is unless the athematic treatment of contract verbs in Aeolic is a later analogical development, which seems likely to me offhand).
    – TKR
    Commented Sep 11, 2020 at 21:10
  • 2
    Also, Aeolic gemination happens instead of, not in addition to, the 1CL: e.g. ἔνεμμα, ἔστελλα for ἔνειμα, ἔστειλα. (Or if lengthening did occur, it was undone by Osthoff's Law.) I wonder what the evidence for the -μμ- forms is. I don't have access to Hamm's Grammatik zu Sappho und Alkaios, nor to Hock's 1971 dissertation The So-Called Aeolic Inflection of the Greek Contract Verbs, both of which might shed light on this.
    – TKR
    Commented Sep 11, 2020 at 22:32
  • 1
    "Aeolic gemination happens instead of, not in addition to, the 1CL" - ah, I see, thanks! I have Hamm 1971 and Thumb - I could add info later tonight.
    – Alex B.
    Commented Sep 11, 2020 at 22:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.