I suppose that the infinitive of "possum" once was *potesse, but that the 'e' in the second syllable got lost, so it went from *potsse to "posse". But why did the 'e' in the second syllable get lost?

1 Answer 1


It's true that posse as a verb represents a kind of univerbation of potis + esse, but that doesn't mean every individual form does. The present infinitive, in particular, is just the stem pot- with the infinitive ending -se, which is the same as the infinitive ending -re you see in most verbs, tacked directly on.

-se is actually the older of the two forms; as most Latin verbs have stems ending in a vowel (amā-, monē-, tege-, audi-), the -s- regularly became voiced (Very Old Latin will have had amase [amaːze]), and rhotacism regularly turned z into r in Old Latin.
Posse is one of a handful of verbs whose stem doesn't end in a vowel, though, so the -s- was not subject to rhotacism; instead the final -t- of the stem assimilated to it, just as in the forms possum (pot-sum) and the like. *ts > ss is a regular sound change in Latin.

(Latin doesn't much like clusters of a consonant followed by s, so some kind of assimilation actually ended up taking place in most consonant-stem verbs: fer-se became ferre, vel-se became velle, ed-se became ēsse (with long ē because of Lachmann's law; later, though only from the 3rd century CE on, replaced by edere), and only es-se got away with being esse. If there had been any stems ending in -c-, they would have been preserved intact, though written with an x.)

Other forms in the paradigm of posse that aren't just pot- + a form of esse include all of the perfect, pluperfect, and future perfect, in both the indicative and the subjunctive: pot-fuī would yield poffuī, not potuī, and so on. Here, too, the verb wasn't perceived by speakers as literally pot- + esse, but as having a stem pot- that can take "normal" endings.

  • 2
    Also the participles: pote-ns not pot-ens
    – Draconis
    Commented Apr 17 at 21:20
  • @Draconis Very possibly, though I feel the model there could conceivably also be absens rebracketed as abs-ens rather than ab-sens instead.
    – Cairnarvon
    Commented Apr 17 at 21:50
  • The older ending of infinitive is reconstructed to be -(e)si (see this, this, this), so I would expect that forms for some verbs mentioned in your answer would be potsi or possi, amāsi, fersi, velsi, edsi, essi. Word-final short i was changed to e by vowel reduction (I think that it occurred after rhotacism).
    – Arfrever
    Commented Apr 18 at 5:20
  • @Arfrever It's hard to date these sound changes firmly, but Roman grammarians had an awareness of pre-rhotacised forms that they didn't seem to have of pre-reduced forms, so I'd place (the completion of) rhotacism after (most) vowel reduction and would reconstruct a stage *-se for the infinitive; to my knowledge we don't have any attested infinitives in Very Old Latin. People mostly do reconstruct an earlier stage *-si based on an s-stem locative, but a *-sa has also been suggested; *-e can reflect almost any short vowel and in verbal forms final *-i consistently disappears.
    – Cairnarvon
    Commented Apr 18 at 15:22
  • @Arfrever Though the infinitive obviously didn't appear in the sentence positions those other forms would have so it would probably have been exempt. I don't have an online source to link you, but I always recommend Weiss' Outline of the Historical and Comparative Grammar of Latin (and disrecommend Sihler's New Comparative Grammar of Greek and Latin).
    – Cairnarvon
    Commented Apr 18 at 15:24

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.