Latin borrowed a number of words, including names, from Greek. Are there any instances where the stress in Latin is not where expected but follows the Greek accent instead?

My impression was always that Latin completely ignores Greek accents and applies the highly regular stress system of Latin. This is all the more likely if the stress is realised differently in the two languages, Greek using pitch but Latin not — if I understood correctly. But I have ever seen this explicitly discussed. Does Latin always ignore Greek accent or are there examples of the contrary?

I think the most likely case of Greek effect on Latin stress is in names. I can well imagine Greek names being pronounced in Greek style within a Latin sentence, but I have never seen a mention of that. Perhaps someone else has?

This question arose from the question about stressing presbyter.

  • 2
    I don't see how we could know. The accent is not indicated in Latin writing.
    – fdb
    Commented Oct 11, 2020 at 18:12
  • 1
    @fdb Ancient grammarians could mention it. And perhaps stress influenced developments in Romance languages in a telling way.
    – Joonas Ilmavirta
    Commented Oct 11, 2020 at 18:53
  • 1
    Related: latin.stackexchange.com/q/9220/406
    – Draconis
    Commented Oct 11, 2020 at 21:02

2 Answers 2


Greek stress could be used in the medieval period.

Per An Introduction to the Study of Medieval Latin Versification, by Dag Norberg, translated by Grant C. Roti and Jacqueline de La Chappelle Skulby:

in the versification of the Middle Ages [...] the Greek accentuation since the end of antiquity had in a certain number of cases supplanted the Latin accentuation previously in use. In keeping with the Greek accentuation, St. Paulinus of Nola writes aby̆ssus; Fortunatus, emblĕma and problĕma; Aldhelm, machĕra and papĭrus. In the same way in both the metrical and rhythmic poetry we have the early appearance of éremus, ídolum, parádisus, spéleum, trópeum, báptismus, thésaurus, and sarcofágus. According to Lupus of Ferrières, blásphemus is a more correct pronunciation than blasphémus; Gottschalk of Orbais says in his grammatical work that ábyssus, báptisma, bútyrus, rómphea, among others, must be accented on the antepenult, even though the penult is long. In this group must be placed the Greek derivatives in -ία: sophia, philosophia, melodia, theoria, and so on, which in the Middle Ages ordinarily have a long i with an accent on the penult, even though examples of the classical accentuation and prosodic use of all of these words are by no means lacking.62

The prosodic treatment of proper nouns should also be the subject of special studies. [...] Greek proper nouns have often been accented following their Greek accent. Fortunatus scans, for example, Euphemīa (Εὐφημία), Paulinus of Aquileia accents Alexándria (Ἀλεξάνδεια)63, and others provide examples of accentuations of Árrius, Theódorus, Isídorus, Ágatha, Christophórus, Aégyptus, Antióchia.64

("Prosody and Accentuation" page 12)

[Note: I'm not sure why Ágatha and Christophórus are supposed to be telling of Greek influence: Wiktionary indicates final stress in Greek for Ἀγαθή and antepenult stress for Χριστόφορος. Also, I cannot find a Greek proper noun corresponding to Árrius.]

Italian seems to fairly consistently stress learned -ia words taken from Greek on the penult (e.g. teologia) and I'd be inclined to use that accentuation in Latin when using what is called the "Ecclesiastical" mode of pronunciation (that is, the pronunciation system that uses the Italian rules for pronouncing soft c, g as [tʃ], [dʒ] and has vowel length conditioned by stress rather than as an independently contrastive feature).


Weiss's Outline of the Historical and Comaprative Grammar of Latin, p. 112, gives some examples of exactly this phenomenon:

  • Philippus (Gk. Φίλιππος) scans in Old Latin verse as if its second syllable is light; this can be explained by "iambic shortening", in which an unstressed syllable following a light syllable can scan as light itself. This implies that the second syllable of this word was unstressed.
  • The Italian placename Taranto (Gk. acc. Τάραντα) has initial stress, though this may be due to direct continued influence from the Greek spoken in the region.
  • Italian senape "mustard", also with initial stress, from Lat. sināpis from Gk. σίνᾱπι.
  • A couple of Greek initial-accented loanwords show shortening of an originally long second vowel, which makes the regular Latin stress line up with the Greek accent: ἄγκῡρα > ancŏra "anchor", βούτῡρον > būty̆rum "butter".

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.