I am learning classical Attic Greek at college, and we recently saw that, in a departure from the typical recessiveness of verbs, the accents for aorist infinitives always falls on the penult regardless of the ultima and is always a circumflex. Taking the verb παιδείομαι for example, the aorist infinitive is παιδεῦσαι, not παίδευσαι with the accent retreating as we might have expected (indeed this is actually the aorist imperative).

My question is, why is this the case for aorist infinitives? I hypothesized that perhaps, since infinitives are formally verbal nouns, the aorist infinitive keeps its accent on the penult and not the antepenult because noun accents are persistent. Obeying the accent rules of nouns would, at least, explain why it is a circumflex and not an acute. My professor said he wasn't sure of the reason, but added that aorist infinitives tend to have their accent on the same syllable as the masculine nominative singular of the verb's active participle. However, he did not know whether or not this is the reason why the accent of the aorist infinitive is where it is, whether it is just a coincidence, or whether there is some greater common reason for both.

If anyone had an explanation, or knows of any theories, I'd love to hear them. And if you do and added a link to a source I could read for more information, I'd greatly appreciate it. Indeed a related question could ask why second aorist infinitives also have peculiar infinitive accentuation, so an authoritative source on such matters would do brilliantly. Thanks!

  • 2
    Note that it's specifically finite verb forms that are recessive—participles, for example, aren't necessarily recessive either.
    – Draconis
    Mar 27, 2021 at 2:58
  • @Draconis Does this have to do with the fact that participles and infinitives are adjectives and nouns, respectively?
    – healynr
    Mar 27, 2021 at 3:21
  • 1
    Infinitives aren't true nouns, though they are verbal nominals. I think Draconis is onto something about that, though. Interesting question, I hadn't thought about the reason before.
    – cmw
    Mar 27, 2021 at 3:53
  • 1
    "Nominals" include pronouns, but it's essentially the same thing, just constructed on analogy with "verbal." I'm not sure where the dative idea originates, but a good reference on the topic would be Woodcock's A New Latin Syntax.
    – cmw
    Mar 27, 2021 at 4:22
  • 1
    @TKR By that I was merely referencing a rule we learned: "in nouns, whenever the accent falls on the penult, if the penult is long and the ultima is short, then the accent is circumflex". My only point was to give another example of how my "noun hypothesis" would fit, if that makes sense.
    – healynr
    Mar 27, 2021 at 4:59

1 Answer 1


The rule of thumb for the accentuation of infinitives is the following:

  • infinitives in -σθαι are accented recessively (e.g. λείπεσθαι), except in the thematic aorist (λιπέσθαι) and the perfect (λελύσθαι).
  • other infinitives are accented on the penult (λείπειν), except in the thematic aorist (λιπεῖν).

The accentuation of infinitives is relatively regular because infinitives are a late innovation: Proto-Indo-European didn't have them. Consequently, they haven't undergone many sound changes that make the accentuation appear more chaotic than it initially was. (There is one big one: -ειν goes back to -εσεν.)

The reason it differs from the accentuation of finite verb forms (i.e. all forms apart from infinitives and participles) is because that accent is a separate development altogether: it's believed that in (some stage of) Proto-Indo-European, those didn't carry an accent at all (the way many forms of εἰμί don't). The Greek recessive accent is a late innovation—late, therefore regular. (But there are a few exceptions, mainly in imperatives.)

It's important to see that it's the regularity of both the finite forms and the infinitives that's exceptional, not the fact that they're regular in different ways: Proto-Indo-European had a phonemic accent, meaning that (in nouns) its position couldn't be predicted at all.

As for why παιδεῦσαι has a circumflex rather than an acute: that's not because of anything specific to verbs, but because of what's often called the σωτῆρα rule:


  1. the accent is on the penult, and...
  2. the vowel of the penult is long, and...
  3. the vowel of the ultima is short, then...

... the accent is a circumflex.

The word σωτῆρα (acc. sg. of σωτήρ 'saviour') is an example.

-αι, being a diphthong, may not look particularly short, but for the purposes of accentuation (but not for poetic metrics) terminal -αι and -οι are usually counted as short. The exceptions are the various forms of the optative (because the ι was originally long there; παιδεύσαι), the old locative (because that's the result of contraction; οἴκοι 'homeward' vs. οἶκοι 'houses'), and certain interjections (αἰαῖ).

Accentuation in Greek is a fairly complex topic; don't let it discourage you, but don't underestimate it either. Most college-level textbooks will devote a significant section to the subject (there are a few lamentable exceptions, including the frequently recommended but execrable Reading Greek), and you'll undoubtedly be seeing more of it in the course of your higher education.

  • Thank you very much. My textbook is indeed the very one you decry; can you recommend me any better readings for accentuation?
    – healynr
    Mar 29, 2021 at 2:42
  • @healynr The Cambridge Grammar of Classical Greek, which is a very useful reference grammar to own anyway, has a pretty extensive and practical chapter on accentuation. It's a synchronic treatment of how it behaves, not particularly where it came from; if you wanted a comprehensive treatment of that AFAIK you'd have to gather all the pieces yourself from various works on PIE.
    – Cairnarvon
    Mar 29, 2021 at 11:33

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.