I have noticed that active aorist infinitives ending in 'αι' may be accented in three different ways in Attic Greek, as follows (all from wiktionary.org):

  • As if the final 'αι' were short: κώλῡσαι, σιώπησαι.
  • As if the final 'αι' were long: ἁρπᾰ́ξαι, δῐδᾰ́ξαι, ὀρέξαι, ὀλέσαι.
  • A third way: ᾰ̓κοῦσαι, βουλεῦσαι, δεῖξαι, διῶξαι, ἐθελῆσαι, ἐρωτῆσαι, καθευδῆσαι, παιδεῦσαι.


  1. Is there some principle or pattern that will tell me which way a given verb will go? (I don't know that I have a large enough sample above from which generalizations can be hazarded.)

  2. When the word is accented as if the final 'αι' were long, am I safe to conclude that 'αι' is indeed long? Or can it be a short ultima with an accented penult?

  3. How can I account for the 'third way'? If I thought the ultima were short, I would expect the accent to fall on the antepenult. If I thought the ultima long, I would want an accute accent on the penult. The 'third way' does not fit either of these patterns.

  4. Is there any other kind of word that is accented as in the 'third way'?


I am aware that the final 'αι' can be consistently long or short in certain positions. For example:

  • short at the end of middle-passive indicative present third person singular (e.g. 'βουλεύεται') and
  • long at the end of active optative aorist third person singular (e.g. 'βουλεύσαι') .

1 Answer 1


Wiktionary is simply wrong if it says that κώλυσαι, σιώπησαι are infinitives: they can only be aorist middle imperatives. The aorist active infinitives are κωλῦσαι, σιωπῆσαι.

See this helpful handout by Helma Dik, which explains the principles:

  • The aorist middle imperative has the usual recessive accentuation.
  • The aorist active infinitive is always accented on the penult; this means that if the penult has a long vowel or diphthong, it gets a circumflex, otherwise an acute.
  • The aorist active optative has long -αι (as you mention) and the usual recessive accentuation.

The accentuation of the aorist infinitive may seem unusual, but it's really only finite verb forms that regularly have recessive accent, not infinitives and participles: many of the latter have their own peculiar accentuation, viz. the types λελυκέναι, λελύσθαι, λυθῆναι, λαβεῖν, λαβών, λυθείς, λελυκώς, λελυμένος.

  • Thank you. Should I think that 'αι' is short in both 'ἁρπᾰ́ξαι' and 'ᾰ̓κοῦσαι'? Or should I think it is long in 'ἁρπᾰ́ξαι' and short in 'ᾰ̓κοῦσαι'. (Actually, once the question of accenting is answered, I don't really have any other use for the length of the vowel for the present. But if facts about it exist, I would like to know them.)
    – Catomic
    Mar 14, 2017 at 7:57
  • 1
    @Catomic The -αι in ἀκοῦσαι must be short, because a properispomenon can only have a short final vowel (or, if you follow the moraic interpretation, a high pitch on the last-but-three mora is only possible if no consonant divides the last-but-one and the last-but-two). The -αι in ὀλέσαι could theoretically be long, but it is the same -αι as above, and it must be short as well.
    – Dario
    Mar 14, 2017 at 16:13

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.