I'm wondering about the origins of the various non-finite verbal endings in Latin.

My understanding so far of their PIE origins:

  • Infinitives:
    • Present Active: -s-ey (dative of an s-stem verbal noun)
    • Perfect Active: no idea
    • Present Passive: thematic genitive either on bare stem or s-stem depending on conjugation?
  • Participles:
    • Present Active: -nt-s, other branches have o-grade in thematic verbs
    • Future Active: no idea
    • Perfect Passive: -t-os
    • Future Passive: related to the gerund?
  • Verbal Nouns:
    • Gerund: presumably related to the present active participle, but the voicing is odd
    • Supine: -tu-s

Is anyone aware of further information on any of these?

  • I'm aware this may also be suited to latin se, but as it is primarily concerned with historical linguistics I think it's better suited to here
    – Tristan
    Nov 4, 2022 at 17:32
  • 1
    Sihler (1995:610ff.) likens the present passive to Vedic infinitives based on datives of root nouns (yujé < *i̯ug-éi̯, rucé < *luk-éi̯) and says that the -r- was inserted later by analogy with the actives where applicable, e.g., act. *amāye-si vs pass. *amāye-ī*amāye-si vs *amāye-sī. In cases like *lege-si vs *leg-ī, apparently the parallel was too weak for the analogy to function. @Draconis But with something additional to account for the geminate s. Nov 4, 2022 at 18:50
  • 1
    The Latin perfect stem had four origins and four shapes -- reduplicative (dedi), sigmatic (visi), u (audivi), and long vowel (vīdi). Plus plenty of consonant gradation and vowel change, like vinco/vici or ago/egi.
    – jlawler
    Nov 4, 2022 at 19:11
  • 2
    @jlawler The stems, yes; but this is asking about the markers for the infinite forms, not the stems they’re attached to. Nov 4, 2022 at 20:11
  • 1
    @jlawler Janus is correct that I'm interested in the endings rather than stems. I've clarified the question
    – Tristan
    Nov 7, 2022 at 9:34

1 Answer 1


I've finally found sources for those options here that don't have completely transparent parallels and that don't really require explantion.

  • Infinitives:
    • Present Active: per Weiss 2021, this is either a locative or "directive" of a neuter s-stem, either: *-si or *-sa
    • Perfect Active: Weiss says this consists of the -is- formant also seen in the pluperfect subjunctive and the -se component seen in the active indicative, so also either a location or "directive" of a neuter s-stem: *-is-si or *-is-sa
    • Present Passive: Weiss argues (on the basis of some alternative forms in -Vrier cf laudārier "to be praised") this consists of an earlier -ii̯V, comparing to the Vedic instrumental gerund in -ya/-yā. That Sanskrit form would suggest an i-stem masculine or feminine (or I guess a neuter where the -n- failed to be inserted to avoid collision of the vowels), a PIE instrumental of an i-stem would be: *-s-i-h₁ (or *-s-y-eh₁ if the ending was stressed)
  • Participles:
    • Present Active: an nt-stem: *-nt-s
    • Future Active: per Fortson this is a ro-derivative of the instrumental of the supine in -tu-: *-tu-h₁-r-os
    • Perfect Passive: an o/eh₂-stem: *-t-os
    • Future Passive (aka Gerundive): per Jasanoff 2006, this goes back to an extension of the present active participle as an o/eh₂-stem: *-nt-in-os
  • Verbal Nouns:
    • Gerund: Jasanoff argues this is a secondary substantive use of the Future Passive (Gerundive) Participle, it's an o-stem: *-nt-in-om
    • Supine: a u-stem: *-tu-s

Weiss, Michael. 2021. Outline of the Historical and Comparative Grammar of Latin. Second edition. Ann Arbor: Beech Stave Press

Fortson, Benjamin W. IV. 2007. The origin of the Latin future active participle.

Jasanoff, J. H. 2006. The Origin of the Latin Gerund and Gerundive: A New Proposal

  • Weiss cites both the other sources used here when discussing the relevant formations and that was where I found the Fortson paper, but I'd actually found the Jasanoff one separately at an earlier date
    – Tristan
    Sep 22, 2023 at 8:11

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.