As I posted on the Wiktionary Tea Room:

Consulting Bergk's edition of Sappho, I have seem various instances of this genitive "Sapphonis" (e.g. «Sapphonis esse videtur») in the critical notes. This struck me as odd because I'd always known Sappho as a Greek name which as such was declined in Latin as in Greek. Tonight I opened the Sappho entry here, and I found confirmation of my knowledge, and no trace of the genitive Sapphonis. So I was wondering: is it a very late genitive of Sappho, or is there a whole other version of the name giving this genitive? And if the former, shouldn't we mention it in Sappho? And if the latter, do we have an entry covering that version?

Since I got no responses in over two weeks, I am cross-posting. Is it correct that, at least in Classical Latin, the Genitive form was the Greek form Sapphus? When did the genitive shift to Sapphonis, and why, if it's known? Is Sapphonis merely a form used in critical apparati?

Extra: Ngram

Here is the result of Ngram:

Ngram of Sapphus versus Sapphonis

It appears that Sapphus prevailed over Sapphonis almost always, except for that sudden incredible spike in Sapphonis around 1740 AD. Huh? What happened then?

  • Can you give us the actual numbers, how many occurrences is Ngram showing for 1730-1760 of Sapphus and Sapphonis in English texts?
    – Alex B.
    Commented May 24, 2020 at 12:52

2 Answers 2


Here's what I've been able to find so far (Loeb and Teubneriana).

Sapphonis (genitive) has two occurrences only (Charianus and Priscianus).

Accius 378:

Charisius, ap. G.L., II, 63, 19: Huius ‘Didūs Sapphūs Inūs.’ Sed melius esset secundum Latinam consuetudinem huius Sapphonis Didonis dicere . . . Pacuvius sic declinat . . . et Accius—

English translation (by Warmington):

"Charisius: Genitives ‘Didūs, Sapphüs, Inüs.’ But it would be better to say ‘Sapphonis, Didonis,’ according to a Latin usage. This is the inflection followed by Pacuvius . . . and Accius—"

So, the more frequent genitive form in Latin was Sapphus (since it is closer to Greek Σαπφοῦς, attested in e.g. Tz. π. Πινδ. μετρ. 20–22); thus, epistula Sapphus, amica Sapphus etc. This means that the word was never fully assimilated in Latin - hence its unusual (for Latin) declension paradigm.

Sapphus is the preferred genitive form in modern research too (e.g. Teubner).


The Greek female names ending in have two declension options, Greek style and Latin style. (The Greek style is not fully Greek; it has been Latinized but not fully.) For example, consider Dido:

  Greek Latin
NOM Dīdō Dīdō
ACC Dīdō Dīdōnem
GEN Dīdūs Dīdōnis
DAT Dīdō Dīdōnī
ABL Dīdō Dīdōne

As the L&S entry for Dido indicates, both versions are attested. L&S suggests that Iuno is exlusively Latin style and Sappho is exclusively Greek style. (As far as I know, Iuno is not of Greek origin; it is just an example of Latin style declensions of a female name ending in .) It is very easy to see why the endings might start mixing, especially since there were mixed cases like Dido.

A quick corpus search gives only one attestation for Sapphus and none for Sapphonis (or other Latin style cases). In my opinion this is too little evidence for definitive conclusions about what might be acceptable, and one has to look for other similar names.

The Ngram for Dido looks different, suggesting that what you see for Sappho is not universal to all names of this kind:

Ngram for Didus versus Didonis

  • 1
    How come the Greek style dative is not "Dīdŏĭ" (or "Dīdœ" to match the Greek diphthong) but "Dīdō"? Just wondering. So should we add a Latin-style inflection table to the Wiktionary and a usage note explaining that in Classical Latin there is only one attestation of Sapphus and zero of Sapphonis, whereas in more recent Latin one can find both, with the Greek form prevailing (at least according to Ngram)?
    – MickG
    Commented Aug 28, 2017 at 10:59
  • 2
    @MickG The Greek version is Latinized to some extent but not fully. I don't know why the final -i in dative was dropped. If the Wiktionary article is supposed to describe the name through the ages, then both Latin and Greek declension options should be included. I would also add a notice about other words with similar dual declension; Sappho is not unique.
    – Joonas Ilmavirta
    Commented Aug 28, 2017 at 11:10
  • 1
    The entry for Dido gives two declensions, with the Greek-style giving Accusatives Dīdō, Dīdūn and Dīdōn. How were the last two formed? I thought the Greek was Διδώ Διδοῦς Διδοῖ Διδώ Διδοῖ… was there an alternate Greek Διδών? EDIT Just found the answer: Ionic declension Δῑδοῦν (says Wiktionary).
    – MickG
    Commented Aug 28, 2017 at 11:27
  • 1
    @MickG I'm glad you answered your own question; I couldn't have. :) I have never seen accusatives like Didon and Didun in Latin. It would be interesting to know what era and author they are from. Perhaps someone could ask a separate question about them? It might be that Dido was borrowed as an undeclinable word and only the genitive was adopted separately. To me that looks like a plausible option to nominative, accusative, and dative (and ablative) having been separately Latinized to Dido.
    – Joonas Ilmavirta
    Commented Aug 28, 2017 at 11:35
  • 2
    I suspect, that the word "Didus" in the Google n-grams is the latin name of the extinct bird called dodo in English. Such are the joys of Google n-grams. Commented Aug 29, 2017 at 12:21

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.