New Sappho!

Sorry, I try not to overuse exclamation marks, but this is genuinely astonishing news: two poems, previously unknown, by Sappho have turned up, one of them complete with the five final stanzas complete [thanks, TR!]. Here is James Romm’s Daily Beast piece about it:

The two poems came to light when the owner of an ancient papyrus, dating to the 3rd century A.D., consulted an Oxford classicist, Dirk Obbink, about the Greek writing on the tattered scrap. Dr. Obbink, a MacArthur fellow and world-renowned papyrologist, quickly realized the importance of what the papyrus contained and asked its owner for permission to publish it. His article, which includes a transcription of the fragmentary poems, will appear in a scholarly journal this spring, but an on-line version has already been released.

And here (pdf) is Obbink’s draft paper, with the text (in both diplomatic and articulated versions) at the end. Now, if you’ll excuse me, I’m off to read the poems; I wanted to get the word out first! (Via Athanassiel’s MetaFilter post.)

Update. Here’s my quick attempt at a translation:

But you keep repeating “Kharaxos is coming
with a full boat”: that, I believe, is for Zeus
and all the gods to know; you should not be
thinking such things;

you should send me instead with strict instructions
to pray fervently to Lady Hera
that Kharaxos might arrive and bring his
ship safe and sound here,

finding us safe as well. As for the rest,
let’s entrust it all to the gods—fair weather,
after all, can come from a heavy gale
all of a sudden.

Those to whom the Lord of Olympus chooses
to send a god, one who will help them forthwith
out of their troubles, they are the blessed onesfortunate
and very wealthyblessed.

We ourselves, if Larikhos should raise
up his head and one day become a man,
we will be from the troubles that weigh us down
freed in an instant.

[Line 7 formerly read “that she might arrive here bringing Kharaxos”; thanks for the grammatical heads-up, TR!]


  1. Elision looks totally awesome there – the scribe consistently omitted final vowels of words before words starting with vowels. No spaces in the text, just a flow of syllables. Wow.

    For the Romans then, the vowel elision must have been an acquired taste, something perhaps dictated by their reverence to Greek poetry but not so easily lending itself to spoken / sung Latin … so they chose to keep the paired vowels in the text, but fused the pairs into single poetic syllables in recitation?

    (I never gave much thought to the origins of mandatory elisions in my beloved Spanish ballad meter, and now I’m amazed to realize how deep may be their roots!)

  2. Very exciting! I bet your translation is the first to go online.

  3. Astonishing news indeed! There are some lovely lines in these poems, too; I particularly like the description of fair weather coming suddenly out of great storms (εὐδίαι γὰρ ἐκ μεγάλαν ἀήταν αἶψα πέλονται). But the first poem can’t be complete; it surely wouldn’t begin with ἀλλά.

    (A couple of small corrections to your translation: “pray to Hera that Kharaxos will arrive here with his ship”; and πολύολβοι here is surely “very blessed” generally rather than just “very wealthy”.)

  4. Thanks, I’ve changed the “pray to Hera” line and the “complete” at the start.

    πολύολβοι here is surely “very blessed” generally rather than just “very wealthy”

    It may well be, but I can’t translate “they are the blessed ones/ and very blessed,” so one of us will have to come up with an alternative.

  5. Thanks! I was bummed to find that the paper had no translation, so I really appreciate you including one here. (I can’t help feeling, however, that not using Yorkshire dialect or possibly Scots to render Sappho’s Aeolic Greek represents an abdication of translatorial responsibility.)

  6. I’m actually a bit stumped by the penultimate stanza of that first poem. I’m not sure what to make of the genitive plural relative pronoun that begins the stanza — why is it genitive? The only way I can see to make the syntax work is something like “Those out of whose troubles Zeus orders a helper spirit to turn them, they become happy and very blessed”, but that seems rather convoluted. I wish we had a photo; I wonder if that των could be a misreading for τοις.

    And then in the first line of the last stanza, Obbink reads the verb as ἀέργη, which he seems to take from ἀεργός “idle” (“free to live his life as a member of the leading, leisured class”); but surely it should be read ἀέρρη “raises”, as you translate; otherwise there’s nothing to be done with τὰν κεφαλάν.

  7. They did set up a WordPress site for discussion.

  8. Yes, I’ve left a comment there, which is awaiting moderation.

  9. “I can’t help feeling, however, that not using Yorkshire dialect or possibly Scots to render Sappho’s Aeolic Greek represents an abdication of translatorial responsibility.” Aye, right enough.

  10. “they are the blessed ones
    and very wealthy.”

    they are the blessed ones
    and very fortunate.

  11. Bettany Hughes talks to Dr.Obbink in this Radio 4 programme, about Sapho and papirology –
    How do we put direct links in this new format?

  12. Same old way: <a href=”URL”>link text</a>.

  13. siggian: Thanks, “fortunate” was a good suggestion; I’ve emended the translation accordingly.

  14. It’s funny; I never thought about why elision is part of the text in Greek poetry but not Latin (I remember a professor blowing my mind with “dulc’et decor’est). It may be resistance by the Romans, or it may be the opposite, namely, that it was so obvious and accepted that it was transparent and there was no need to write it.

  15. I remember a professor blowing my mind with “dulc’et decor’est

    Funny, I say “decoru(m)’st,” with a nasal u rather than a full m. Do we have actual knowledge of how the elision worked?

  16. I’m with TR – I need to see the script.

  17. John Cowan says:

    It seems clear that it was the initial vowel of the following word that was always kept. Decorum est thus becomes decor’ est. My favorite: Monstrum horrendum, informe, ingens, cui lumen ademptum (Aeneid III:658), pronounced Mōnstr’ ’orrend’, īnform’, īngens etc. Elision at both ends of a word!

  18. It seems clear that it was the initial vowel of the following word that was always kept.

    Not that I’m doubting you, but have you got a cite for this?

  19. John Cowan says:

    Hat: If you look at the front (publicly accessible) page of this article by E.H. Sturtevant, you see a list of Latin words resulting from frozen elision, all of which involve omitting the first vowel (sodes < si audes reflects early /au/ > /o/; note that the comma after ecce belongs one word later). The thesis of the article is that elision was normal not only in verse but in prose and in conversation too.

    Sturtevant is also the author of the rule of thumb that sound-change operates regularly to produce irregularities, whereas analogy operates irregularly to produce regularities, of which I am very fond.

    David E.: Exactly so. I thought that vinometer was a perfectly cromulent English word, but as I had never heard it, I marked it “no”.

  20. Thanks, Sturtevant is quite convincing, and I will amend my reading practice accordingly. (One of my few rewards for being a survivor of Yale Graduate School is the ability to read JSTOR articles in full.)

  21. Heh. From Sturtevant:

    The dramatic poets very often present elision at a change of speakers, but we cannot suppose that in real life a Roman could foresee the reply to his question so as to treat his final vowel in one way if the reply was to begin with a vowel and in a very different way if the reply was to begin with a consonant.

    Note that you can read the whole thing here.

  22. Heh again, from p. 41:

    Quintilian’s remark can be squared with this only on the supposition that Quintilian fails to discriminate clearly between pronunciation and spelling — a failing of which many good scholars both ancient and modern have been guilty.

    And yet again, from p. 42:

    Probably Quintilian was the sort of person who would read English blank verse as if it were prose.

    I like this guy Sturtevant!

  23. Although you do get spellings like rixast, timendumst, etc., which suggest that at least in the case of est, the initial vowel could be elided.

  24. Aha, that’s probably why I had “decorum’st” in my head. Now I’m confused again.

  25. Inter-speaker elision happens in modern Romance drama, too. I wonder how it’s performed on stage. Come to think of it, what does one conventionally do with elided vowels when reading Spanish or Italian poetry aloud? Leave them out altogether? Pronounce them ultra-short? Nothing in particular?

  26. I had assumed that the repeat of “blessed” was an allusion to a weel-kent piece in the Bible. Or the Life of Brian.

  27. There is no repeat of “blessed”; the Greek has two different words.

  28. ‘There is no repeat of “blessed’: in your translation I meant.

  29. Elementary manuals for scanning Latin verse distinguish between ‘prodelision’ of ‘es’ (you are) and ‘est’ (he/she/it is) and ‘elision’ of every other word that begins with a vowel or H. They also tell students that in prodelision the following word (the ‘es’ or ‘est’) loses its vowel, while in elision the preceding word loses its vowel and the final M if it has one. So, yes, ‘horrendum est’ is explained as ‘horrendumst’ where ‘horrendum angit’ would be ‘horrendangit’ or ‘horrendmangit’. No one knows whether the elided vowel, or the final M, or both, were compressed or eliminated entirely, or conceivably fully pronounced with the elision merely nominal. Some think the M itself disappeared, but that whatever vowel sound was left behind was nasalized.

    Years ago, I compressed the basic rules of Latin meter for Ovid and Martial (4 basic meters) into two pages, with two more for Horace’s very miscellaneous meters. You can see them here in DOC and PDF format.

    FInally, yes it is accepted that prose was also elided except at the end of a so-called colon. (Is a colon a breathing-pause? Probably. It’s the prose equivalent of a line-end in verse.) Latin prose was highly rhythmical, with standardized ‘clausulae’ found at period end. (Classicists have only been studying this for a century or so.) In fact, it was so rhythmical that the semiliterate freedmen at Trimalchio’s feast use good clausulae: apparently Petronius couldn’t help writing rhythmically even when elegant rhythms were inappropriate.

    If you want to see some Latin prose divided into cola or clausulae, I’ve posted Seneca’s 80th letter here (each sentence has a hanging indent, and lines 7f-h and 8c are indented further since they are bits of verse quoted by Seneca). I didn’t clausulate it myself, but borrowed the analysis of B. L Hijmans, “Inlaboratus et Facilis: Aspects of Structure in Some Letters of Seneca” (Leiden, 1976). I’m working on some software to semi-automate the process of clausulating prose, because that’s the only way to display prose on the web with apparatus criticus on the side lined up with the text.

  30. So, yes, ‘horrendum est’ is explained as ‘horrendumst’ where ‘horrendum angit’ would be ‘horrendangit’ or ‘horrendmangit’.

    Thanks, now I’m unconfused again! And thanks for the links as well.

    in your translation I meant.

    There is no repeat of “blessed’ there either; that was precisely what I wanted to avoid.

  31. Thank you for getting my comment out of the spam filter in less than 15 minutes.

  32. Seneca Ep. Mor. 80:

    Hodierno die non tantum meo beneficio mihi vaco, sed spectaculi, quod omnes molestos ad sphaeromachian avocavit. Nemo inrumpet, nemo cogitationem meam impediet, quae hac ipsa fiducia procedit audacius. Non crepuit subinde ostium, non adlevabitur velum; licebit tuto vadere, quod magis necessarium est per se eunti et suam sequenti viam.

    He must be talking about Superbowl Sunday, which I look for precisely the reasons Seneca states: “omnes molestos ad sphaeromachian avocavit. Nemo inrumpet.”

  33. A possibly better reading for part of that troublesome penultimate stanza has now been suggested by Sasha Nikolaev at the WordPress site: reading ἐπ’ ἄρωγον for ἐπάρωγον, it could mean something like “Those whose daimon the Lord of Olympus chooses to turn around out of toils to succor / to be a helper, they become fortunate and very blessed”. I’m not totally sure this works, but it’s an improvement, anyway.

  34. The link for Obbink’s paper seems to be no longer working. Anyone know where to find it?

  35. Alan, I’ve just emailed you a copy.

  36. Thanks!

  37. John Cowan says:

    Hat, I know you don’t read newspaper comments, but this one is priceless:

    30 January 2014 3:53pm

    It’s Aeolian.. to me..

  38. TLS article on the new Sappho by Dirk Obbink:

  39. Thanks for that; I certainly agree with this: “So far from frigid juvenilia, the verses show an ear for balancing and texture, the pulse of rhythm, graceful shaping of words, and an atmospheric ending to a poem.” I don’t much care for the opening of the Pelling translation (“Oh, not again – ‘Charaxus has arrived!/ His ship was full!’”), but it gets much better after that. And now I really want a complete version of the second poem!

  40. No, it doesn’t seem like juvenilia to me either.

    I have been following the wordpress forum; there seem to be concerns in some quarters about the provenance of the papyrus, or rather the paucity of information about it. In fact that topic has pretty much edged out the linguistic and textual questions.

  41. John Cowan says:

    And now we can add the word barythymia ‘nervous depression, lit. heaviness of soul’ to the more active part of the English lexicon. Goodness knows, it’s needed nowadays.

  42. Alon Lischinsky says:

    @Michael Hendry:

    No one knows whether the elided vowel, or the final M, or both, were compressed or eliminated entirely, or conceivably fully pronounced with the elision merely nominal. Some think the M itself disappeared, but that whatever vowel sound was left behind was nasalized.

    I was under the impression that the orthographic -m in the Latin accusative singular desinence did not correspond to an actual /m~n~ŋ/, but to the nasalisation of the preceding vowel. Thus orthographic ‘horrendum’ would in practice be pronounced /hɔr.reːn.dʊ̃/.

  43. That was my impression too.

  44. An article on the discovery of the Sappho fragments in the Feb. 5, 2014 TLS by Dirk Obbink begins as follows:
    “An ‘Oxford secret’ is supposed to be a secret you tell one person at a time. Add social media and it’s across the world within hours, often in garbled form. In this case the ‘secret’ was the discovery of a fragment of papyrus….”
    The discovery is great, though the provenance is not clear. I wondered about the origin of the collocation “Oxford secret.”
    Provisionally, it *may* have been coined, or popularized, by Oliver Franks (1905-1992), Oxford philosopher, and Secretary of Supply during WWII, and Ambassador to the US, and diplomat involved in birth of the Marshall Plan, and also of NATO,** among other things. Remarkable man.
    (**My Dad, in the USN, was assigned as an assistant to Ambassador Averell Harriman at the time of NATO beginning.)

  45. David Marjanović says:

    Long, long ago…

    It may be resistance by the Romans, or it may be the opposite, namely, that it was so obvious and accepted that it was transparent and there was no need to write it.

    Also, there was only one way to do it (“elision” and “prodelision” elsewhere in this thread). The Greek response to vowel collisions is bewildering and involves not only elisions but fusions with full conservation of length (ο + α = ω for example).

    Thus orthographic ‘horrendum’ would in practice be pronounced /hɔr.reːn.dʊ̃/.

    I once read somewhere, and am convinced, that the supposed vowel length that precedes all non-final nasals is just syllable length, i.e. vowel + nasal sequences were at most pronounced as a long nasal vowel, not as a long vowel followed by yet another mora.

  46. Yes, that makes sense to me.


  1. […] Hat provides a first translation of the recently discovered poems by […]

  2. […] The online preprint (to ZPE 189 [2014]) is no longer available. However a translation by Tim Whitmarsh is available from The Guardian and another by Steve Dodson. […]

  3. […] the web preprint (to ZPE 189 [2014]) is now not available. however a translation by way of Tim Whitmarsh is on hand from The Guardian and some other with the aid of Steve Dodson. […]

Speak Your Mind